M A S A R Y K U N I V E R S I T Y
Faculty of Economics and Administration
Field of study: Public Economics
EFFICIENCY EVALUATION IN THE
MUNICIPAL SOLID WASTE MANAGEMENT
(HODNOCENÍ EFEKTIVNOSTI
V ODPADOVÉM HOSPODÁŘSTVÍ OBCÍ)
Dissertation
Dissertation supervisor: Author:
doc. Mgr. Ing. Martin DLOUHÝ, Dr., MSc. Ing. Michal S T R U K
Brno, 2018
Author's name and surname: Ing. Michal Struk
Dissertation title: Efficiency evaluation in the municipal solid waste
management
Department:
Dissertation supervisor:
Public Economics
doc. Mgr. Ing. Martin Dlouhý, Dr., MSc.
Year of defense: 2018
Abstract
The aim of the submitted thesis "Efficiency evaluation in the municipal solid waste
management" is to suggest ways for improving municipal solid waste management in South
Moravian Region based on a detailed analysis of selected relevant efficiency factors. Chosen
efficiency factors were identified both using available literature and quantitative analysis of
collected data. Data were collected from more than 500 municipalities in South Moravian
Region in the Czech Republic and subsequently used in a two-stage Data Envelopment
Analysis in order to identify specific efficiency factors. Selected efficiency factors
(convenience of waste separation, presence of incentives, effects of age structure and impact
of competition among waste management providers) were examined in more detail and
suggestions how these findings can be utilized in municipal waste management practice are
provided.
Předložená práce "Hodnocení efektivnosti v odpadovém hospodářství obcí" se zaměřuje na
identifikaci a přezkoumání faktorů efektivnosti ovlivňujících odpadové hospodářství obcí
v Jihomoravském kraji za účelem poskytnutí návrhu k jeho vylepšení. Zkoumané faktory
efektivnosti byly vybrány na základě studia příslušné domácí a zahraniční literatury
a kvantitativní analýzy primárních a sekundárních dat sesbíraných pro více než 500 obcí
v Jihomoravském kraji v České republice s využitím dvoustupňové analýzy obalu dat.
Vybrané faktory efektivnosti (vnímané pohodlí při třídění odpadu, existence pobídek, efekt
věkové struktury obyvatel a dopad konkurence mezi odpadářskými společnostmi) byly
podrobně prozkoumány. Výsledkem je soubor zjištění a doporučení pro možné využití
získaných poznatků v praxi za účelem dalšího zvyšování efektivnosti v odpadovém
hospodářství obcí.
Keywords
Municipal waste management, efficiency, DEA, waste separation, age structure, competition
Abstrakt
Klíčová slova
Odpadové hospodářství obcí, efektivnost, DEA, třídění odpadu, věková struktura, konkurence
Declaration
I hereby declare that the dissertation "Efficiency evaluation in the municipal solid waste
management" is my own work carried out under the supervision of doc. Mgr. Ing. Martin
Dlouhý, Dr., M S c , and I have duly acknowledged all sources in accordance with the law, the
internal regulations of Masaryk University, and the binding internal documents of Masaryk
University and of the Faculty of Economics and Administration.
In Brno, May 2018
A u t h o r ' s signature
Acknowledgements
I would like to express my gratitude to Martin Dlouhý, Jana Soukopová, Jiří Hřebíček and
Simon Struk for lot of various help during writing this dissertation, numerous municipal
representatives for providing me with the data, and my family for the patience and tolerance
(mostly), especially my wife Janka.
CONTENT
I N T R O D U C T I O N 9
1 E F F I C I E N C Y A N D M U N I C I P A L W A S T E M A N A G E M E N T 13
]. 1 DEFINITION OF EFFICIENCY 14
1.2 EFFICIENCY A N D THE PUBLIC SECTOR 18
1.3 METHODS OF EFFICIENCY MEASUREMENT 23
1.4 W A S T E M A N A G E M E N T 27
1.4.1 W A S T E M A N A G E M E N T IN T H E ENVIRONMENTAL PROTECTION 28
1.4.2 W A S T E M A N A G E M E N T IN T H E C Z E C H REPUBLIC 31
1.4.3 W A S T E M A N A G E M E N T IN T H E C Z E C H MUNICIPALITIES 33
1.5 EFFICIENCY IN THE MUNICIPAL WASTE M A N A G E M E N T - LITERATURE OVERVIEW 43
1.5.1 IDENTIFIED FACTORS OF EFFICIENCY 52
2 M U N I C I P A L S O L I D W A S T E M A N A G E M E N T I N T H E S O U T H M O R A V I A N R E G I O N 55
2.1 SOUTH M O R A V I A N REGION IN THE C Z E C H REPUBLIC 55
2.2 B A C K G R O U N D OF THE D A T A COLLECTION 56
2.3 D A T A COLLECTION PROCESS 62
2.4 D A T A DESCRIPTION 76
2.5 QUALITATIVE FINDINGS 80
3 E F F I C I E N C Y E V A L U A T I O N I N T H E M U N I C I P A L S O L I D W A S T E M A N A G E M E N T 87
3.1 S A M P L E DESCRIPTION 87
3.2 D A T A E N V E L O P M E N T A N A L Y S I S 89
3.3 D E A EVALUATION OF EFFICIENCY 98
3.3.1 IDENTIFIED FACTORS OF EFFICIENCY 103
3.4 DISCUSSION OF THE FINDINGS A N D LIMITATIONS 103
4 E F F I C I E N C Y F A C T O R S I N T H E S O U T H M O R A V I A N M U N I C I P A L S O L I D W A S T E
M A N A G E M E N T Ill
4.1 INCENTIVES, DISTANCE A N D KERBSIDE COLLECTION I l l
4.2 A G E STRUCTURE 128
4.3 COMPETITION A N D PUBLIC VS. PRIVATE NATURE OF THE WASTE M A N A G E M E N T PROVIDER 148
4.4 DISCUSSION OF THE FINDINGS A N D SUGGESTIONS FOR PRACTICAL APPLICATIONS 158
C O N C L U S I O N S 163
L I T E R A T U R E 167
S U M M A R Y , L I S T O F F I G U R E S , T A B L E S A N D A B B R E V I A T I O N S 189
7
8
INTRODUCTION
Waste management is today commonly perceived as a service provided by the public sector.
More specifically it is considered as one of the standard services provided by the
municipalities. Arguments for providing municipal waste management as a public service
derive from the nature of the service itself and the related private sector failures. This is
understandable, as waste generation creates externalities that impose costs also on the subjects
different from those that generate the waste. Public provision of services dealing with issues
when the producer of the externalities does not bear the costs is then a natural consequence.
But while on the one hand public provision of some goods or services can, to a certain point,
correct the market failures it was intended for, on the other hand it can, and often does create
other problems known as public sector failures. One of them is the question of the efficiency
of public expenditure. Unlike in private sector, where companies primarily aim for profit
generation, thus the question of pursuing efficiency is naturally incorporated in their
decisions, in public sector the main goal is the benefit of the society, with efficiency being
less of an important factor in decision making. The conflict then becomes clear with
efficiency on one side and benefits to the society on the other.
In this thesis I focus specifically on municipal waste management, which is in developed
countries commonly accepted as a public service and provided accordingly by the public
entities, whether directly by them or by contracting external providers. I do not question
whether the provision of waste management as a public service is meaningful, that is to be
decided elsewhere. My goal in this thesis is to examine the ways how municipal waste
management can be improved (both from the supply and demand side) and I do it primarily
from the economic perspective, however, with some overlaps to the environmental
perspective as well. My objectives in this thesis are therefore to
1) describe situation in municipal solid waste management in a selected Czech region;
2) identify relevant efficiency factors in this region using both collected data and
available literature, and based on further analysis provide policy suggestions for
improving municipal waste management in the region.
9
In the first chapter I provide foundations for the term efficiency, concentrate on the efficiency
in the public sector, and describe common ways how to measure efficiency. Then, after the
definition of the key concepts that this thesis is based on, I provide an introduction to the
municipal waste management in the Czech Republic, as it represents the focus are of the
thesis and it is necessary in order to orientate better in the examined area. After that I provide
an overview of the studies relevant to the issue of efficiency of waste management, from
which I identify efficiency factors that literature usually lists as important with respect to
municipal waste management. Deduction method was then used for creating the assumptions
about the role of the selected factor, of which some are then more thoroughly examined in the
fourth chapter in order verify the assumptions.
In the second chapter I describe the data that I used in this thesis and the data collection
process together with some initial findings that emerged already during the data collection
part of my research. The reason why this amount of space is devoted to this part is because in
order to examine the issue of efficiency of municipal waste management the data had to be
collected directly at the municipal level for this specific purpose. The data collection thus
represents a significant part of the research itself. Method of analysis of the original collected
data was used in order to identify important factors that influence M S W M from the
perspective of municipal representatives.
In the third chapter I employ the collected data in an empirical analysis using Data
Envelopment Analysis together with the secondary data from public databases and based on
induction identify the potential factors that could play an important role in the municipal
waste management. As the primary data was collected only from selected region (even though
vast majority of municipalities participated), the validity of the findings has its limitation
regarding the whole Czech Republic. Nevertheless, the sheer size of the sample and the
relative homogeneity of Czech municipalities make it possible to generalize the findings to a
certain extent.
In the final fourth chapter I examine more thoroughly selected factors that emerged in the
previous efficiency evaluation. In separate subchapters I examine effects of incentives,
10
distance, kerbside type of waste collection, age structure of the municipal population,
competition presence and ownership nature of the waste management service provider.
Method of verification of the assumption regarding the role of selected factors was used by
employing relevant statistical methods.
Finally, based on the results I conclude the findings of my thesis and discuss the potential of
incorporating these findings in municipal waste management design and suggest ways how to
improve the current practices in the area of municipal waste management.
11
12
1 EFFICIENCY AND MUNICIPAL WASTE MANAGEMENT
"Efficient - productive of desired effects; productive without waste"
"Efficiency - the quality or degree of being efficient"
(Merriam-Webster Online Dictionary)
Pursuing efficiency is a permanent process in both public and private sector. Considering the
homo economicus concept, efficiency is something that is naturally connected with rational
decision making. Thus the efficiency is something we all are doing simply by nature without
necessarily thinking about it. It is generally accepted that, when given several options, one
chooses the option that, in his own perspective, provides him with most of an output while
consuming the least of an input. Aim of this chapter is to further explore the concept of
efficiency and link it to the waste management that is examined further in this thesis.
When considering individual decision making, the problem of efficiency might be seen as a
much more simple - if assuming that we make rational decisions, it implies that these
decisions are also efficient. Of course we run to the problem of what rationality is and the
question whether one makes rational decisions. But this raises another issue - who decides
what is rational and what is not? The issue with the rationality is in the fact that when
considering individual rationality, it is relatively easy to define it. People make choices
willingly, and if they choose an option, it is the option that, based on their current knowledge
and ability to evaluate given options that are able to fulfill their desires in an accepted way,
was the best out of the given range. The other thing is how this choice might turn out later,
but this does not contradict with the fact that initially people made a subjectively rational and
efficient decision.
In the case of collective decision making the problem of rationality becomes trickier. The
question here is who should decide what is rational and what is not. Although there is a
philosophical concept of objective rationality that can be hypothetically worked with,
defending such concept without setting many additional conditions and further restrictions
becomes next to impossible. Nevertheless, in reality there exists something like this "quasi-
13
objective' rationality in decision making, represented, for instance, by transferring the
decision making to the deputy elected in the voting process. The outcomes of such decision
making, however, depend in the end on the personal qualities of the deputy himself.
Although a simple word, there still is no consensus in what the term efficiency exactly means.
The issue is that it has different meanings depending on the context, although every time the
definitions follow the same idea. In the first subchapter general aspects of the efficiency are
provided in order to define how it should be understood in this thesis. In the second
subchapter I sketch out the relationship between efficiency and the public sector, together
with the specifics related to the public sector. In the third subchapter I provide the basic
overview of the methods commonly used for efficiency evaluation (including the role of
econometrics as an important tool that can significantly improve the primary results of
efficiency analyses). In the next part I provide the overview of the waste management and its
relation to environmental protection and then focus more specifically on the basics of waste
management in the Czech Republic and its municipalities. Finally, in the last subchapter I
provide an overview of the important efficiency studies performed in the area of M S W M and
some concluding findings of these studies.
1.1 Definition of efficiency
At this point I provide several definitions of the term efficiency, explain how it is understood
in the rest of the thesis and point out the specific aspects of efficiency when concerning
quantitative analysis. Very simple definition of efficiency can be "the absence of wastage"
(Vlček, 1992:27). However, this definition is, in my opinion, more philosophical than
practical (although it is correct) and makes it difficult to define a way of evaluating whether
some unit is efficient or not, or alternatively how efficient the unit is when compared with
other evaluated units.1
Much more quantifiable definition of efficiency offers for example
Ochrana (2005:164), who defines efficiency as the maximization of ratio between outputs and
inputs (presumably costs). This definition basically says that a unit is efficient if it is able to
1
Here I would like to stress the part "compared to other evaluated units". In my opinion one cannot state that
something is efficient solely by itself. Efficiency evaluation has to include a set of at least two units in which
these units can be compared together with the result of one unit being at least as efficient that the other. Without
other units for evaluation, or at least some benchmark, it would be like competition with just one contestant this
contestant would always finish both first and last.
14
produce more from consuming less than other units it is compared with. The concept behind
this goes back to the Pareto efficiency, which is in the words of Musgrave (1984:295)
described as "economists consider an arrangement efficient if resources are used in a way
which does not leave a possibility of alternative arrangements under which somebody could
be better off without anyone being worse off."
However, in practice it is often impossible to know whether there is no other possible
arrangement of resources that could provide a better result. Therefore it is for our purposes
probably the most practical to define efficiency as a ratio between a set of outputs of a
transformation process and a set of inputs used in this transformation process (Jablonský,
Dlouhý, 2004:12). Such definition of efficiency automatically adds the assumption that there
is some level of wastage by the decision making unit, which in my opinion seems as an
appropriate real world assumption. Efficiency of a unit then can be understood as a rate of this
unit's performance with respect to an ideal 100% efficient unit. And finally one more
philosophical definition, according to which is the efficiency of spent resources considered as
the pursuing of the given goals set as the basic criterion of rationality, and it is in the very
center of evaluating the organizational process of the pursued policy. Acquiring the state of
economic efficiency then means the situation where the choices between the goals and the
alternative ways of achieving them are made in the way that the maximal total benefit from
the available resources is achieved (Malý In Strecková, 2005:7).
After defining efficiency as a certain measurement of unit's performance, we logically
encounter the issue of how to measure this efficiency. This is still an unsolved problem not
just in the public sector. There are numerous methods of how to measure efficiency. Medveď,
Němec, Orviská and Zimková (2005), or Ochrana (2001, 2011) provide an overview of this
issue with respect to the 3E (economy, efficiency, effectiveness) concept in the public sector.
Commonly used evaluation methods are for example single criterion methods (also known as
input-output methods) consisting of CMA (Cost-Minimization Analysis), CBA (Cost-Benefit
Analysis), C E A (Cost-Effectiveness Analysis) and C U A (Cost-Utility Analysis), various
2
This concept of considering efficiency is usually called "allocative efficiency''. However, it assumes that one is
able to compare all available arrangements of resources. In my opinion this concept works well in theory, but the
complex situations in real life might make it impossible to even consider all possible arrangements, not speaking
about their further evaluation and comparison.
15
multicriterial methods or financial methods like FV (Future Value), NPV (Net Present Value),
ROI (Return of Investment), PM (Payback Method) and IRR (Internal Rate of Return). Listed
methods provide relatively easy way of comparing evaluated units, but on the other hand are
quite limited when it comes to more complex decision making problems that include multiple
inputs and outputs. More advanced methods for efficiency evaluation can be found in the field
of operation research, for example in Jablonský (2002). The fact is, however, that each of
these methods has its pros and cons, and none of them is perfect. Thus, each time some of
these methods are used, the related limitations of the method should be kept in mind.
Mintzberg (1996) calls it "the myth of measurement", noting that "many activities are in the
public sector precisely because of measurement problems. If everything was so crystal clear
and every benefit so easily attributable, those activities would have been in the private sector
long ago". And besides that, the most efficient unit does not have to be the "besf unit, unless
we define best strictly as the most efficient. Also there is a difference between outputs and the
outcome. A unit can perform very well in certain output indicators (and thus be considered
most efficient) but still be outperformed, for example, by overall or combined result of other
units. All this depends on how is efficiency defined, as well as the selection of method for the
efficiency evaluation.
But let us move further with a more specific type of efficiency that is the main concern of this
thesis - economic efficiency, and to be even more specific technical efficiency. The
relationship between the two is simple, as economic efficiency consists of technical and
allocative components. The technical component refers to the ability of avoiding waste, either
by producing as much output as technology and input usage allow, or by using as little inputs
as required by technology and output production. Thus, the analysis of technical efficiency
can have an output-augmenting orientation or an input-conserving orientation. Allocative
component refers to the ability of combining inputs and/or outputs in an optimal proportion in
light of prevailing prices. Optimal proportions then satisfy the first-order conditions for
optimization problem assigned to the production unit (Fried et al., 2008:20).
3
For example Lovell (2002:23) provides a brief discussion of the use of these "techniques" in the public sector
with listing their pros and cons.
16
For the clarification I stress here that in the further text I use the terms efficiency generally to
refer to the technical efficiency and I leave out the discussion about allocative efficiency,
which requires the presence of price information. Just one more note on allocative efficiency
- it corresponds to the ratio of cost efficiency and technical efficiency (Porcelli, 2009).
Nevertheless, the costs aspects of MSWM will be included in the practical part of the
analysis.
Probably one of the first definitions of technical efficiency was provided by Koopmans
(1951), who explained it in a bit of Pareto style as "a producer is technically efficient if an
increase in output requires a reduction in at least one other output or an increase in at least
one input, and if a reduction in any input requires an increase in at least one other input or a
reduction in at least one output". Some refer to this as "Pareto-Koopmans" condition for
technical efficiency (Ray, 2004:2). Koopmans had here already recognized the potential of
linear programming for economic analysis and optimization, which was developed later.
Around the same time Debreu (1951) started to define what has later, after merging with
Farrell (1957) became known as Debreu-Farrell measure of technical efficiency. With an
input-conserving orientation it is defined as: "(one minus) the maximum equiproportionate
reduction that is feasible with given technology and outputs" (Fried et al., 2008:20). Or in
other words as a "... (radial) reduction in all inputs that still allows the production of given
outputs, a value of one indicates technical efficiency and a score less than unity indicates the
severity of technical inefficiency" (Porcelli, 2009). Or alternatively, any deviation of this
measure from one was interpreted as "a deadweight loss suffered by society due to inefficient
utilization of resources" (Ray, 2004:3). In case of an output-augmenting orientation, the
Debreu-Farrell measure is defined as the maximum radial expansion in all outputs that is
feasible with given technology and inputs. In both orientations, a value of unity (1) indicates
technical efficiency because no radial adjustment is feasible, and a value different from unity
(less than 1) indicates the severity of technical inefficiency (Fried et al., 2008:20).
Although Debreu-Farrell measure might seem at first a bit difficult to understand, it creates
the foundations of how the efficiency is being understood when acquiring results of more
17
complex efficiency measuring techniques such as later presented DEA, where efficient units
acquire value of one (unity), while inefficient units acquire lower values, depending on the
level of their inefficiency.
Already mentioned Farrell (1957) made his breakthrough contribution in the topic of
efficiency by constructing a linear programming model using the actual input-output data of a
sample of firms with the result of a numerical measure of technical efficiency of individual
firms in the sample. Farrell recognized that a function fitted by the standard OLS regression
could not serve as a production frontier, as the analyzed points would lie on both sides of the
fitted line. Such line obviously would not represent an efficient frontier, but rather an average
one. Therefore Farrell approximated the underlying production possibility set by the convex
hull of a cone containing examined input-output bundles (Ray, 2004:3). Nevertheless, it took
many years before further progress in this area was made, and, as some say, Farrell's
contribution to the area of efficiency measuring was ignored for more than two decades
(Russell, 1998).
1.2 Efficiency and the public sector
When speaking about efficiency in the public sector, a good starting point is to explain the
differences between the efficiency in public and private sector. Private sector, when it comes
to efficiency, offers one very elegant shortcut through the institute of prices (carrying
exhausting amount of information, however trivial and clear the price solely by itself might
appear). In private sector pretty much everything has its price, everything is monetized. Thus
one has the option of reducing efficiency measuring to the examination of the firms' profit,
even though it is in fact a very strong simplification. On the other hand, in public sector the
decision maker often faces outputs that are not expressed in financial means (the reasons for
that being for example political, moral or due to the absence of a market equivalent). This
means that in case of efficiency evaluation in the public sector we do not face just one
problem of specifying what efficiency is, but also the problem of assessing the effects or the
outcomes that are produced in the public sector. Even such simple comparison then reveals
that the two sectors are very often not interchangeable.
18
The objectives pursued by the public and private organizations are different, as the private
sector aims for profit, while the public sector seeks not only to obtain economic benefits, but
also to obtain social benefits, with the stated primary objective to ensure the public welfare
(however, profit maximization might still be the secondary goal). The goals of the public
sector are generally driven by the political decisions and the political decisions are driven by
the political representatives elected by the people. And the people usually elect those
representatives who seem to be the most capable or suitable to fulfill their needs and desires.
The trick is, however, in the fact that these needs and desires have often intangible form, and
as such they cannot be easily expressed directly in the form of money or not even in any
tangible form at all. Among these desires are, for instance justice, equality, safe environment,
freedom, productive life, etc. And how does one express such things in the monetary form?
This is also the reason why private sector in general is not interested here, as, although it
might be simple to express the costs, it is next impossible to evaluate the outcomes in terms of
money. Or one can try, but any results would be immediately subdued to a wide criticism, as
these thing have very broad range of values in the perspective of individual people.
Also there is a problem that in the public sector in many cases the direct and immediate
economic benefit is missing. For example, if a school is built in a village, the effort involved
in this investment can be easily identified as all the costs incurred for the construction, the
material basis, the wages, etc. But how can we measure the benefits in this case? We can
identify the social benefits, such as increasing literacy, ensuring better labor market, higher
living conditions, etc. But how should we express these benefits in the form of money? Or
another example - when building a highway by the public sector we should not look only at
the direct benefits, but also consider indirect ones besides collection of highway taxes (if
applied at all), like for instance reduction of the road accidents, reduction of the traveling
time, improved conditions for the economic development of the region, etc. If taking into
account only the directly expressed monetary benefits, the calculated efficiency would be
much lower than the real one, possibly resulting in no investment and development at all.
The private sector thus rather seeks primarily economic benefits, while showing a reduced
concern for the social and environmental issues, even though nowadays many companies are
starting to improve the mentality and trying to place the social responsibility vision with the
19
one of obtaining profit. This aspect of public sector (taking into account more than just direct
financial aspects) should be kept in mind when planning any efficiency evaluation. For a
better understanding of the key differences between public and private organizations I provide
the following Table 1.
Table 1: Main differences between public and private organizations
Public organizations Private organizations
Are usually monopolies Operate on competitive markets
Serve the citizens Maximize the investment's profit
Are driven directly or indirectly by the politicians, Leaders of companies are responsible to the
who should reflect the interests of the citizens shareholders, to the boards; they seek profit
maximization
State organizations are more rigid due to the process Are more flexible, easier to manage because the
of decision making and implementation decision is taken by a single leader
Distribute, redistribute and regulate resources Produce and distribute resources
Are sometimes poorly funded Are financed under based on their productivity or if
the investment decision is feasible
Citizens are often poorly informed and suspicious of Investors and shareholders are well informed and the
the government ongoing activities of the company and the market
evolve
Source: Mihaiu et al. (2010)
Next important thing to keep in mind when dealing with the efficiency in the public sector is
the 3E concept, where Es stand for economy, efficiency4
and effectiveness, or alternatively
the overall performance (Nemec et al., 2010:52). This concept is sometimes referred to as
Value for Money (VFM) and is among the most important concepts in the field of public
finance. To put it in very simple terms it stands for economy, efficiency and effectiveness
means to achieve the desired goals. A common mistake in practice is to confuse effectiveness
with efficiency which are different concepts but work complementary in the 3E's approach.
Table 2 provides basic overview of 3E criteria. Let us look at each of the Es in more detail.
Table 2: Overview of 3E criteria
Criterion Means of monitoring Best quality
Economy Cost minimization Ideally the lowest price
Efficiency Costs per unit of outcome Lowest cost per outcome unit ratio
Effectiveness Rate of fulfillment Highest rate of objective fulfillment considering costs
Source: Ochrana (2006:52)
4
Important note here is that the term efficiency in the 3E concept and the efficiency as used in the rest of the
thesis are not identical, although they share some common aspects. Outside of 3E concept, efficiency in this
thesis refers to the technical efficiency, as described in the subchapter 2.1.
20
The first E, Economy, stands for minimizing the cost of resources ("doing things at a low
price"). It covers the financial side on which an activity is achieved. The desired objectives
and goals should be achieved with as little as possible use of the resources (Nemec et al.,
2010:48). Because we live in a world in which economics permeates every aspect of our lives,
it is fundamental to balance the use of resources in order to achieve the right goals with the
reasonable cost. There is a difference between Efficiency and Economy, as the former states
the volume of resources and the way of using them, while the latter looks primarily at the
costs.
The second E, Efficiency, stands for performing tasks with reasonable effort ("doing things
the right way"). Efficiency means the right use of resources in order to accomplish a task. It is
understood as a "wise" consumption of inputs used in the transformation process towards the
desired goals. It is important not just to attain the goals, but it is also fundamental to see how
these goals can be reached with the least effort and/or resources. Therefore when choosing
between comparable option, the most efficient is one that has the lowest costs per unit of
output (Nemec et al., 2010:49). Improving efficiency thus means that the company is in a
trend of reducing wastage and consuming less to deliver the same amount of value.
The third E, Effectiveness, stands for the extent to which objectives are met ("doing the right
things"). Effectiveness should therefore be understood as the level at which the goals are
being achieved, for instance on a scale of 0-100%. A successful performance in terms of
effectiveness is then to fully meet the beforehand set quantitative indicators. Or alternatively,
such usage of resources that the set goals and objectives are optimally met (Nemec et al.,
2010:49).
Next important term to define, when dealing with the efficiency in the public sector, is public
expenditure. Very simple definition might be, let us say "the money that is somehow spent by
the government". In literature we can find for example a broad definition from Sivak et al.
(2007:202), according to which "the existence of public expenditure is related with the
character and functions of the public sector and its relation with the private and non-profit
sector. The goal of the public sector is to ensure the maximization of the welfare to the most
21
people possible." And the most obvious "too/" for ensuring the maximization of the welfare
to the people is "to spend financial means in order to reach the welfare." In other words this
says that public expenditures are public sector's means of ensuring the welfare to the people.
Sivak et al. (2007:205) offers yet another definition in which "public expenditure is financial
expression of the government's interference", and Hillman (2009:3) would probably add "... in
order to achieve its social objectives" (or in other words "that, what is in principle expected
to be sought by consensus"). However, here we encounter the philosophical question of what
is good or what are the legitimate objectives of the government, as well as the definition of
public interest and whether we truly are able to compare between the various alternatives. But
trying to solve such questions is far beyond the goal of this thesis. Let us just assume that we
are able to compare various goals and find consensus on which of them is "best" and that we
are also able to agree whether something is or is not a public interest.
When considering the determination of the specific efficiency indicator that we are planning
to use for the efficiency evaluation in the public sector, we can follow the scheme in Figure 1.
I assume here that the efficiency is provided by the relationship between the effects, or the
outputs, and the efforts or the inputs. The relationship is apparently simple, but practice often
proves the contrary, because identifying and measuring inputs and outputs in the public sector
can often become a difficult operation. Following scheme shows the main differences when
making efficiency evaluation in the public and private sector.
Figure 1: Scheme of determining the specific efficiency indicator
Efficiency
Input Output
- easy to quantify in the
private sector
- might be difficult to
quantify in the public
sector
Source: Mihaiu et al. (2010)
Private sector
1 1
*
Public sector
- easy to identify and
quantify in the
monetary form
- depending on the
output difficult to
quantify and compare
22
Based on the provided scheme, the direct factors of influence on the efficiency are:
1) The inputs. In the public sector the resources are sometimes harder to quantify than in
the private sector, because public services often overlap and resources from several
sources are used. But, in general, the inputs are given by the costs incurred for the
project/service in matter.
2) The outputs. In the public sector these are usually much more difficult to quantify than
in the private sector, because they often have both economic and social dimension. In
the private sector the outputs have typically market value and thus are from the private
sector decision maker perspective easily evaluated. On the other hand, in the public
sector is this process often much more cumbersome. To evaluate the outputs from the
public sector we must carefully define the indicators that will be evaluated, and
through which a level of efficiency will be determined. Such process might often get
complicated and also kind of vague in some areas.
1.3 Methods of efficiency measurement
There are multiple methods that can be used for measuring efficiency. One can choose from
the very simple ones to the relatively complex ones. The choice should depend both on the
importance of the result and the ability of the evaluator to fully utilize the specific method,
including the ability to provide sufficient and relevant data for the evaluation. Following
subchapter describes three basic categories of such methods. But one should also remember
that further analysis is often necessary, in order to find also reasons that make analyzed unit
inefficient and how this can be improved (Jablonský, Dlouhý, 2015:7).
Single criterion methods
Very common are single criterion efficiency measurement methods that use a calculated ratio
between the inputs and the outputs. Probably the most known method in this category is Costeffectiveness
Analysis, or CEA. This method is used for calculating ratios of a certain output
(typically a physical unit) and the related costs used in the transformation process when
comparing several alternatives. Necessary with this method is to have comparable outputs
from the evaluated alternatives. Advantage of this method is in the fact that outputs can be
23
used in the evaluation process in their natural form, without the need for any additional
adjustments. The result is then presented typically in the form of costs per unit produced, or
alternatively units produced per certain amount of costs. On the other hand, CEA does not tell
anything about whether it is worth to spend the resources in order to obtain the outputs. But
this should not be seen as a flaw of the method, as the goal of the CEA is to compare the
alternatives and not to assess the reasonability of producing certain output. This should be
decided either politically or by using other relevant methods. Another evaluation method,
Cost-utility Analysis, or CUA, can be used in cases with somewhat more complex outputs for
instance in healthcare, when using quality adjusted life years provided by some kind of
treatment instead of just additional life years that would be used in the CEA. Both of these
methods can be well used in less complex decision making problems where we are able to
abstract from additional conditions and characteristics of evaluated units, as these methods are
able to cover only very few factors that might have the influence on the overall efficiency of
the production unit (Jablonský, 2004).
Multicriterial methods
In cases of more complex measurement of efficiency one can use multicriterial evaluation
methods based on the aggregation of inputs and outputs. These methods share the ability to
incorporate multiple inputs and outputs and then calculate a ratio of this sum of inputs and the
sum of outputs. Important is, however, to assign the appropriate weights in order to obtain a
relevant numerator (weighted sum of outputs) and denominator (weighted sum of inputs) in
the calculation. When compared to the CEA, the obvious advantage is the ability to include
larger amount of evaluated unit's characteristics in the evaluation. However, there still remain
issues with assigning the proper weights to the individual characteristics and the issue of
comparing units of a notably different size.
Frontier analysis methods
For even more complex efficiency measurement one can use methods based on the
construction of the production frontier from the set of evaluated units. Most important
methods representing this field are Stochastic Frontier Analysis, or SFA and Data
Envelopment Analysis, or DEA (explained in detail in subchapter 3.2). The origins of both
24
methods come from Farrell's (1957) study, with the difference being in certain aspects like
the use of parametric approach instead of nonparametric. One can generally say that after
Farrell's seminal study there have been two main paths of further research - one became later
known as DEA while the second as SFA.
SFA represents an efficiency evaluation method based on econometric approach. It is a
parametric method that considers the presence of a statistical noise, which is one of its main
differences when comparing with DEA. As put by Greene (2008:93), it is "essentially a
regression model that is fit with the recognition of the theoretical constraint that all
observations lie within the theoretical extreme". Measurement of efficiency is then the
empirical estimation of the extent to which observed units achieve estimated frontier. The
origins of SFA come from Farrell (1957), who suggested that one could usefully analyze
technical efficiency in terms of realized deviations from an idealized frontier isoquant, which
falls naturally into the econometric approach. The inefficiency is thus identified as the
disturbances in the regression model (Greene, 2008:93). One of the discussants of Farrell's
study, Sturrock (1957), then conceptualized the proposed production frontier as stochastic. He
opposed that the output producible from an input bundle would be subject to chance
variations beyond the control of the firm and argued against using 100% efficiency that is
required in DEA (Ray, 2004:5). The pioneering works in this area came from Aigner and Chu
(1968), who considered production frontier to have a Cobb-Douglas (log-linear) form with
assigned random non-negative variable, let us say u, associated with technical inefficiency.
There are various approaches for estimating u, for instance Aigner and Chu (1968) used linear
programming, Afriat (1972) assumed u to have gamma random distribution and for estimation
used method of maximum likelihood, while Richmond (1974) used a specific modified OLS
technique. However the most notable work in this area came several years later by Aigner,
Lovell and Schmidt (1977) and Meeusen and van den Broeck (1977) who both, just one
month apart from each other, proposed the stochastic frontier model that is being applied by
the researchers up till now. Third, although less known "founding" paper about SFA came
from Battese and Corra (1977).
25
The initial model of SFA proposed by Aigner, Lovell and Schmidt (1977) and Meeusen and
van den Broeck (1977) has the form of
In qt = x[p + Vi-Ui
where x[fl determines the shape of the production function, v, stands for symmetrical random
error to account for statistical noise (either positive or negative), and w, is a random nonnegative
variable associated with the technical inefficiency. Together vt — ut = £t or simply
the error term. Such model is called stochastic frontier production function because the output
values are bounded from above by the stochastic variable exp(x-/?). The goal then is to
minimize ut subject to ut > 0 (Coelli et al., 2005:242). The economic logic behind this
specification is that the production process is subject to two economically distinguishable
random disturbances - statistical noise represented by vt and technical inefficiency
represented by ut. Errors vt are assumed to have symmetrical distribution, component ut is
assumed to be distributed independently of vt while being non-negative. These assumptions
are rather strong ones (Mastromarco, 2008). Stochastic frontier models are estimated mainly
by Modified Ordinary Least Squares (MOLS), or by maximizing the likelihood function
directly. More about these methods and about the theory behind SFA can be found in Fried et
al., (2008) or in more advanced Kumbhakar and Lovell (2000).
The main advantage of SFA is the presence of a disturbance term representing noise,
measurement error, and exogenous shocks beyond the control of the evaluated unit. This is in
clear contrast with DEA, where efficient units have to lie directly on the efficiency frontier
created by the peers, and any deviation from this was considered as inefficiency. This
allowance for statistical noise, together with nonparametric nature of SFA, is a very strong
argument for SFA, if choosing between the two. Another feature of SFA is its attempt to
determine the absolute economic efficiency against some imposed benchmark unlike DEA,
which evaluates efficiency between the analyzed samples. And, as SFA is of statistical nature,
it allows for statistical testing of hypotheses (Worthington, 2014). Further disadvantages
related to the econometric/statistical nature of SFA are explained in Greene (2008).
26
Econometric evaluation of additional efficiency factors
Finally, econometric tools can be very well used in the efficiency measurement for analyzing
the relevance of additional factors that can influence efficiency of the production units. In
many cases we might have additional data about the production that we cannot simply utilize
in the previously listed methods. These can include, for instance, presence of various
qualitative characteristics. We can include such characteristics (after, for instance
transformation into the dummy variables) in the subsequent regression analysis with the
efficiency obtained from the previous analysis as the dependent variable, together with
various other characteristics as the independent variables. Results of the regression can point
us in the further direction of where to focus our research. Two-Stage DEA that is used later in
this thesis represents one such method, where initial efficiency score is calculated using
standard DEA and the obtained score is then used in the subsequent regression with the
additional available data in order to identify factors that can have significant effect on the
efficiency score of the evaluated units.
1.4 Waste management
In this thesis I focus specifically on the efficiency in the field of waste management, which is
today commonly secured by the public sector as a public service. In this subchapter I provide
an overview the waste management and specifically municipal waste in the Czech Republic
primarily from the economic perspective. This provides reader with an idea of the importance
of this field. The examination of efficiency and factors that influence municipal waste
management which represent the primary focus of this thesis then naturally follows.
This subchapter firstly introduces the reader to the environmental protection in the Czech
Republic from primarily expenditure perspective and then focus on the area of waste
management, as it is the main field of research of this thesis. Then, in order to show how
waste management is connected with the public sector and specifically municipalities I
describe in detail what are municipal obligations in the waste management area, what actually
this service means, and how it works in practice, including financing, expenditures, practical
aspects, etc. Provided perspective will be primarily economic and not environmental.
27
Inclusion of this introduction is necessary in order to understand the importance of further
search for ways how to improve municipal waste management.
1.4.1 Waste management in the environmental protection
Waste management is part of a broader category of environmental protection. Environmental
protection in general represents activities or practices of protecting natural environment on all
kinds of levels from individual to global, and for the benefit of both individual people and
environment as a whole.
Various aspects of environmental protection are studied extensively and, let us say that, there
is a global consensus on the importance of it. The importance of environmental protection and
related environmental policies started to emerge in 1970s as a result of the book Limits to
Growth (Meadows et al., 1972). This book was the first widely noticed report that warned
about the consequences of economic and population growth with a finite supply of resources.
While there is a substantial body of both criticism and defense of the main message of the
book, it definitely brought a worldwide attention to the issue of environmental protection.
One of the consequences of the report is continually increasing role of environmental policies
in all aspects and increasing pressure to keep in mind environmental aspects in all other kind
of public policies, and primarily expressed in the focus of environmental policy itself.
One might ask why is environmental protection such an issue? The reason for that is the fact
that the environment is basically the most important source of production factors for the
economy, it defines the basic conditions for the production itself, and therefore it is in the
very interest of whole society to avoid or at least limit the negative effects of production
process and human behavior on it. The main problem here is the fact that individually rational
human economic decisions have in the end often negative impact on the whole society. These
issues have been known to the economic theory for a long time and include concepts of the
tragedy of commons, prisoner's dilemma and negative externalities. However, it is not among
the goal of this thesis to address these issues, but merely focus on a selected part of the
environmental protection area and help to improve it.
28
One way how to examine the focus of environmental policy is through the focus of the
environmental protection expenditures, or EPE. The areas of protection where expenditures
flow are thus the demonstration of the policy priorities - important issues are those that get
financed. Let us look at the development of EPE in the Czech Republic in last years that show
specific areas of environmental policy focus.
Figure 2: Capital (investment) and current (non-investment) EPE in the Czech Republic,
2003-2015
Investment expenditure [CZK bil., c.p.] Non-investment costs [CZK bil., c.p.]
o o o o o o o * - < - * - > - < - * - O O O O O O O * - ! - * - * - * - * o
o o o o o o o o o o o o o o o o o o o o o o o o o
< N C M C N r « J C s t < N C s l C N C M C S I ( N C v | O J C N C M C S ( N C N C N < N C N < N C N C N C N C N
• Waste water management
Air and climate protection
• Waste management
• Landscape and biodiversity protection
Reduction of noise and vibration;
protection against radiation
• Protection and remediation of soil,
groundwater and surface water
Research and development; other
activities in the environmental
protection
Source: MZP (2017a:5)
Figure 2 shows the development of EPE since 2003. There are two parts of the figure - capital
(investment) and current (non-investment) expenditure. The difference between them can be
demonstrated with the differentiation in the type of activity they are used for. Capital
expenditure is used usually for purchasing fixed assets or an infrastructure that is projected to
last for a longer period (typically multiple years), and usually without knowing exactly when
the assets will need to be renewed. This is also the reason why such expenditure does not
occur regularly, but more likely in times when there is either a need for such an investment or
there are available funds for financing it. On the other hand, current expenditure happen
regularly and much more often, as they are used typically for covering repetitive activities and
are more likely to cover services instead of purchasing fixed assets with long term lifespan.
29
The difference between these two categories also explains the different structure of the
expenditure. For example let us look at the largest expenditure categories from both parts of
the Figure 2, wastewater management in capital expenditure and waste management in current
expenditure. In case of wastewater management the majority of expenditure is used for
infrastructure like sewage system and water treatment plants and slowly covering more and
more municipalities in the Czech Republic. Once this infrastructure is operating, regular
expenditure is usually relatively low, as this is a kind of infrastructure is designed to last for
decades. However, initial investments are usually very high and in many municipalities can
represent amount comparable to several years' total municipal budget. Nevertheless, there are
usually ways how to co-finance such projects from national or international sources and thus
somehow reducing the pressure on municipal budgets. On the other hand, waste management
represents a different kind of activities with different structure of expenditures. Majority of
waste treatment facilities are privately owned and private companies offer waste management
services to the public entities. As waste is practically a product of almost any kind of
consumptions and people need to consume in order to survive, the result is a continuous
stream of waste that needs to be taken care of, requiring regular waste-related services.
In Figure 2 we can see that in the category of capital expenditure major areas are wastewater
management and air and climate protection. In practice, wastewater areas usually cover
infrastructure for water management and treatment, like improving existing or in some cases
even constructing whole sewage system together with corresponding water treatment
facilities. In case of air and climate protection, majority of finance flow to the measures of
reducing emissions of hazardous material into the air, and can cover mostly improvements of
certain transport corridors and reduction of impacts of stationary pollution sources (like
heating plants or incinerators).
In the category of current expenditure, the major area is waste management. As was
mentioned above, this is a consequence of the nature of the waste generation producing a
continuous stream of waste that has to be taken care of. While there might be some measures
that could in certain aspects reduce the amount of generated waste or somehow improve the
existing waste streams, the basic need of waste collection, transport and treatment remains
largely the same, and requires repetitive care and thus repetitive expenditures as well.
30
The advantage of this repetitive occurrence of waste management expenditure is that is
produces a substantial amount of data that can be quantitatively analyzed. When comparing
with the capital expenditure, we often encounter only isolated data points that are often spread
wide both in time and space.
From the perspective of historical development, the sum of EPE has practically doubled in the
shown time period. Moreover, in relative terms it has grown from somewhere around 1.5% of
GDP up to 2.0% of GDP in the recent years (according to the data from the Czech
Environmental Information Agency - CENIA, 2017), further stressing the importance of
environmental protection from the perspective of the public sector. And out of this sum
including both capital and current EPE, current expenditure on waste management represents
roughly 40%. Financially-wise, this makes the area of waste management practically the most
important part of environmental protection in the Czech Republic.
1.4.2 Waste management in the Czech Republic
In this subchapter I provide overview of the waste management in the Czech Republic. The
development of waste generation in the Czech Republic is shown in the Figure 3.
Figure 3: Total waste generation in the Czech Republic, million tons, 2009-2016
I Industrial waste I Municipal waste Waste generation per capita
2009 2010 2011 2012 2013 2014 2015 2016
Q.
u
i—
ai
a.
t/i
E
i—
aao
Source: ISOH (2018)
31
Total generation of waste was in recent years above 30 million tons. However, it is important
to differentiate between two main types of waste, which is commercial (industrial) waste and
municipal waste. Commercial waste represents practically any kind of waste that is a byproduct
of commercial activities and can cover anything from office waste to heavy industry
waste. Municipal waste, on the other hand, represents, simply put, waste being produced in
the municipalities, primarily by the residents, but not from the commercial activities.
Figure 3 shows a large disproportion between these two types of waste. If calculated in
relative terms, per capita commercial waste is around 2500 kg, while it is only around 500 kg
in case of municipal waste. The reason is that commercial waste covers also extremely heavy
waste fractions compared to the standard municipal waste, probably most notably demolition,
or in general building and construction waste. While taking relatively little volume, weightwise
can this waste fraction be very significant. For instance, partial reconstruction of an
apartment can generate demolition waste of several tons, while the whole family living there
might not generate even one ton of municipal waste during the whole year. The total amount
of generated waste thus heavily depends on the type of prevailing industry in the country, and
can be significantly different even between the various regions. Relative comparison of waste
generation by categories is shown in the Figure 4, where municipal waste accounts for around
16% of the total waste, while building and construction waste takes up more than 60%.
Figure 4: Waste generation in the Czech Republic in 2016 by categories
• Building and construction waste
• Municipal waste
• Waste from waste treatment facilities
• Waste from thermal processes
• Packaging waste
Other waste
Source: MŽP (2017d:253)
32
Another distinction between commercial and municipal waste is in the responsibility for the
waste and subsequently the costs of the waste treatment. According to the Czech Waste Act,
commercial waste is responsibility of the commercial body that generated it and this body is
also required to cover the related costs. This practically means that it is in the best interest of
the commercial bodies to efficiently deal with their waste and seek for any possibilities of
either reducing the total amount of waste generated by them or somehow utilizing the waste,
once it had been generated. These commercial bodies also directly contract the waste
management (WM) companies and pay for the collection and treatment of their waste.
On the other hand, municipal waste is, according to the Czech Waste Act, the responsibility of
the municipality. While it is clear that it is actually citizens who generate the waste, it is the
municipality that has to secure waste collection and the subsequent treatment. In terms of
costs that are naturally part of this process, it is again the municipality that has to pay for it.
However, the municipality can raise revenues for W M from legal municipal fees it is allowed
to charge to the citizens. This W M fee is defined in the relevant legislation and it is then up to
the municipality to collect it and cover the W M expenditure with it. But municipality cannot
set the fee at whatever level it decides to. First, Waste Act defines how municipality is
supposed to calculate this fee, including certain limits. And second, setting up basically any
municipal fee is also a political decision and higher fees are naturally connected with the
negative reactions. Municipalities therefore try to balance the fee level between sufficient to
cover the M W costs and being politically passable. Nevertheless, the sole fact that it is the
public finance that is used to cover the expenses of municipal waste management brings up
the question of efficiency and the factors that influence it. In the rest of the thesis I will focus
on the municipal waste managements and examine what affects the efficiency in this area.
1.4.3 Waste management in the Czech municipalities
Czech Waste Act (Act 185/2001. Coll.) defines in Part 1, §4(l)x that the municipality is the
originator and the owner of the municipal waste since the very moment this waste has been
discarded in the designed place (excluding commercial waste, as was described above). This
practically means that, independently from whoever produced the waste in the municipality, it
is the municipality that is responsible for the whole process around the waste management. It
33
is then up to the municipality how its waste management will be set up and it is the
municipality that is fully responsible for it. Waste Act further allows the municipality to
define its waste management system in the form of a specific local legislative act valid within
the limits of that municipality. Specific responsibilities of the municipality include provision
of designated places where waste can be stored for later collection. This can differ between
the municipalities and also within the municipality based on the housing structure. In the
Czech Republic the two most common systems for waste collection are door-to-door
collection of individual waste bins in case of family houses and large shared waste bins
typical in densely populated areas with apartment buildings. Another responsibility of the
municipality is to provide a possibility for separate waste collection for at least hazardous
waste, paper and cardboard, plastics, glass, metals and biowaste. Separate collection of
additional waste fractions is then up to the individual municipality.
If we then consider only municipal waste, generated amount drops significantly from the
figures of the total generated waste. Following Figure 5 shows the development of municipal
waste generation in the Czech Republic in recent years.
Figure 5: Municipal waste generation in the Czech Republic, million tons, 2009-2016
i Municipal solid waste Other municipal waste Municipal solid waste per capita
5
M 4
c
o
•*-»
o °
1
2
1
0
r 450
400
- 350
•*-»
- 300
cap
- 250
i—
a;
Q .
- 200
•ams
- 150 Cuo
O
- 100 2
- 50
- 0
2009 2010 2011 2012 2013 2014 2015 2016
Source: ISOH (2018)
34
Total amount of generated municipal waste in recent years is stable at around 5 million tons
annually. It can be split into two categories of municipal solid waste and other municipal
waste that differ in the actual origin of the waste. Other municipal waste represents waste that
is produced in the municipality, but does not end up in waste bins for mixed waste assigned to
the households. Practically it covers waste generated during street cleanings, waste from
public bins and other kinds of waste that is generated in the municipality but not the one in the
bins assigned to the dwellings. It also covers separated waste than is collected in the specially
designated bins for recyclable waste, including for instance paper and cardboard, plastics,
glass or biowaste. Figure 5 shows that this kind of waste is slowly growing in the recent
years, which is likely the result of the improving waste separation in the Czech Republic.
Municipal solid waste, on the other hand, represents the waste directly produced by the people
in the municipality and ending up in the mixed waste bins located usually next to the
individual dwellings. Figure 5 shows that the amount of this waste is slowly decreasing. As
was mentioned, this is likely at least partially caused by the improving waste separation.
There is a clear connection between the two - the more people separate waste, the less mixed
residual waste they generate. Other reasons for decreasing mixed waste generation might be
changes in the behavior of the people, like preferring consumption of products causing less
waste (reusable products, less packaging waste, etc.).
Difference between these two categories of waste becomes obvious when it comes to the costs
they generate. Other municipal waste is usually much better separated into the relatively
homogenous waste streams. For instance, if we have collected recyclables, these could be
often directly used for extracting secondary raw materials or otherwise valorized. Collected
biowaste can be often used straight away for composting or biogas generation, paper, glass
and plastics can be after little treatment used for recycling and so on. This means that these
kinds of waste are very likely to become a resource for further production and often can
induce more revenues than costs. In the Czech Republic most of the municipalities have a
contract with authorized packaging company E K O K O M and get paid some money based on
the amount of waste they separate while paying some fees for the separate waste collection.
Depending on the collected amount, waste separation can be for the municipality in the end
profitable. Alternatively, municipality can take collected recyclables directly to the material
35
recovering facility and get paid for it. In case of biowaste, municipality can set up its own
composting plant or biogas plant and either get revenues from selling the product or utilize
the product for its own purposes. In the end, if the process for further treatment of separated
waste is set correctly, instead of only costs it can generate also revenues for the municipality.
On the other hand, mixed residual waste does not generally provide any ways of generating
revenues from this stream. After the collection is such waste transported and undergoes
further treatment, usually in the landfill or incinerator. Compared to collected recyclables,
unit costs are usually several times higher, making mixed waste practically the most
expensive waste stream to treat (excluding some special types of hazardous waste, that public
usually does not deal with). From the perspective of municipality it is then preferable to have
more separated waste instead of mixed waste. Besides environmental benefits it can also lead
to notably lower municipal solid waste expenditures (typically in the sense of lower payments
to the W M company), for instance as a result lower frequency of mixed waste collection, as
less mixed per capita is generated. Factors that can lead to increased separated to mixed waste
ratio are examined in more detail in later parts of this thesis.
Figure 6: Municipal waste treatment in the Czech Republic, million tons, 2010-2016
• Energy utilization of municipal waste • Material utilization of municipal waste
Landfilled municipal waste • Other municipal waste treatment
6 i—
2010 2011 2012 2013 2014 2015 2016
Source: MZP (2017c:71)
36
In terms of waste treatment, two most common ways of dealing with municipal waste are
landfilling of predominantly mixed waste and material utilization of typically separated waste.
Third, slowly growing category is energy utilization, usually meaning incineration of the
waste in order to generate energy. Other ways of municipal waste treatment (like incineration
without energy utilization) represent marginal ways of waste treatment. Division of municipal
waste according to the type of waste treatment is shown in the Figure 6. Compared to the
other European countries, large percentage of waste landfilling represents a problem with
respect to achieving specific environmental targets set by the European Union, even though
this indicator dropped notably in the recent years. However, current levels of fees for different
types of waste treatment in the Czech Republic make it difficult to further decrease
landfilling, as it is the cheapest option for dealing with the mixed municipal waste and W M
companies naturally respond to that. Compared to some other countries in EU, landfilling fees
there are sometimes double or even quadruple. However, an alternative in the form of energy
utilization requires significant investments into the new infrastructure, while currently
existing capacities already operate at the designed level. Another issue is also the distance
factor, as the frequency of existing landfills is also much higher in comparison with only few
thermal treatment plants for municipal waste, which makes energy utilization of the municipal
waste even more expensive due to the transportation costs. Nevertheless, current national
waste management plan counts with gradual increase of the landfilling fees which should
divert current waste stream more towards other treatment options than landfilling.
In terms of securing municipal solid waste management (MSWM) there are few usual options
for the municipality. Most of the municipalities choose to contract an external provider of
M S W M and pay for the service (either fixed yearly per capita fee or fee based on some
quantitative indicator, like generated amount of waste). From the perspective of the
municipality this is probably the most comfortable way, as the municipality just has to pay the
fee and external company takes care of the rest. Some municipalities contract only certain
parts of the M S W M services, for instance service related to mixed municipal waste, and take
care of waste separation on their own, or contract a different company for this, but that is up
to the municipality. The other, much less common option is to have a municipal company that
takes care of MSWM. Such company usually takes care also of other technical services in the
municipality. These are likely to be found in larger municipalities, where economies of scale
37
make it economically viable to have such municipal company. Alternatively, some of these
municipal companies historically evolved and besides home municipality provide service also
to several surrounding municipalities at fully commercial basis in order to better utilize their
capacity and again take advantage of economies of scale. In case of more common external
providers, the companies can be divided based on their ownership. There are privately owned,
publicly owned (usually by a group of municipalities), with mixed ownership (part is private,
part is owned by the municipality or a group of municipalities), or already mentioned
municipal companies.
Municipality is responsible for financing MSWM. In order to do that, it has the option of
charging its citizens with related waste fee (technically municipality can either charge a local
fee for municipal waste system or a so called waste management fee - I will not go into
details of legal differences between the two, as it is not important for our purposes and a large
majority of municipalities use the first type - the local fee, while the other is relatively rare).
While majority of the municipalities choose to charge such fee, it is not obligatory and some
municipalities finance their W M from different sources, depending on the local specifics.
Examples include raising other, for instance property taxes (well applicable if there is a
significant amount of commercial activity in the municipality). And some municipalities do
not need to collect waste fees or local taxes at all, as they get sufficient resources from the
central budget based on the current system of tax assignment to the local governments
(typically if some local corporations or residents pay substantial income taxes and the
municipality that gets part of it is relatively small).
Nevertheless, large majority of the municipalities charge waste fees. These are typically
calculated as annual fee per capita, sometimes with discounts for some specified groups of
people (elderly, children, etc.). This type of fee represents a flat fee that does not reflect
individual waste behavior. On the other hand, it is possible use "performance-based" fees that
reflect, for instance, how much waste one produces or alternatively how much waste one
separates. This issue is examined in more detail later in the thesis. In case of most common
flat fee per capita, the municipality is limited by the legislation when setting the level. Until
2012 this limit was 250 CZK base fee plus up to another 250 CZK depending on the actual
municipal waste expenditure. Since 2013 was this limit raised to 250 CZK base fee plus up to
38
750 CZK variable part. Thus maximum level of waste fee charged by the municipality can go
as high as 1000 CZK per capita per year. However, majority of municipalities charge lower
fees, even though it does not sufficiently cover their W M expenditures. There are several
reasons for that. First one is that variable part of the fee can be calculated from the municipal
expenditure on solid waste, which represents only part of the municipal waste expenditure,
even though it is usually the largest part. The rest of the municipal W M expenditure has to be
then covered by the fixed 250 C Z K per capita part. If these other costs are higher,
municipality has to find other sources for financing them, but cannot raise variable part of the
fee higher that the level calculated from the solid waste expenditure. In a practical example,
let us say that total per capita municipal waste expenditure is 800 CZK, but out of this amount
only 500 is represented by the solid waste expenditure. In such case municipality can set the
local fee to be 250 CZK fixed part plus 500 CZK variable part, resulting in 750 CZK per
capita total fee. The remaining gap of 50 CZK has then to be covered from other municipal
sources, even though the fee level is not at its top limit. And the second reason for charging
lower waste fees is political. People generally dislike any tax or fees increase and waste fees
are no exceptions. This political aspect was strongly present in 2013, once municipalities
could charge much higher waste fees, but most of them did not raised the fee level as high as
they could according to the variable part calculation (or did not raised it at all), as they did not
want to upset their citizens.
Figure 7: Histogram of Czech municipal solid waste expenditure per capita, 2016
• Municipalities in expenditure level cohort
1000
300 400 500 600 700 800 900 1000 1100 1200 1300 1400 1500 1600 1700 More
Municipal solid waste expenditures per capita, 2016 (CZK)
Source: own calculations based on MONITOR and CZSO
39
Figure 7 shows the calculated frequency of per capita municipal expenditure on waste
management in 2016 from all 6254 Czech municipalities that reported their waste
management expenditure. The average per capita expenditure was 927 CZK and median value
was 849 CZK. More than half of the municipalities had per capita expenditure in 700-1100
CZK range. However, over 43% of municipalities had average per capita expenditure higher
than 1000 CZK. Considering the maximum fee level set by the legislation, it means that more
than 40% of municipalities had to partially finance their W M even if they set the fee level at
the allowed maximum. However, majority of the municipalities have their waste management
fee set at much lower level, usually around 600-700 CZK (plus not including discounts for
some specific groups and the fact that the municipal fee collection rate is typically lower than
100%). This results in the observation that majority of the municipalities need to co-finance
their W M from some other sources, that could not therefore be used for other purposes, as
municipality is by law responsible for its W M independently from how it raises revenues for
it. Overall number suggest that municipalities have to co-finance somewhat around 30% of
W M expenditure, but there are naturally outliers on both sides with municipalities that are
fully able to finance their W M from collected fees, as well as those that finance large majority
of W M expenditure from other sources than collected fees.
Figure 8: Development of municipal environmental protection expenditure by categories,
2009-16
I Water protection
Biodiversity and landscape protection
I Waste management
I Other EPE areas
50 000
45 000
40 000
35 000
S 30 000
§ 25 000
I 20 000
15 000
10 000
5 000
0
2009 2010 2011 2012 2013 2014 2015 2016
Source: M Z P (2017c:414)
40
I provide Figure 8 in order to illustrate the overall importance of W M among other categories
of municipal EPE (values are in the current prices). W M expenditure represents more than a
quarter of total EPE. However, as was mentioned earlier, waste expenditures are mostly
current, while other EPE are mostly capital. This can be seen especially in case of water
protection (including wastewater) that fluctuates notably in various years while the W M
expenditure remains at around 9-10 billion CZK each year.
If we look even further into the structure of municipal waste expenditure in Figure 9, we see
that usually more than 70% of expenditure goes to the collection and transport that is directly
related to the waste generation by the citizens. But here I need to note that based on the
experience from collecting the data from the municipalities, this division of expenditure might
not be always accurate, as some municipalities simply do not bother with more precise
reporting of their W M expenditure into individual categories and report everything under
collection and transport category. In the end these expenditures are still waste expenditure,
but the assignation to the specific categories might not be that precise.
Figure 9: Development of municipal waste management expenditure by categories, 2009-16
l Collection and transport of MSW
Other municipal waste expenditure
l Treatment of MSW
Per capita M W expenditure
12 000
10 000
8 000 — 18.2%
ISI
U 11.9%
lion
6 000
2 4 000 1
19.7%
11.3% 10.5%
16.4%
11.1%
15.7%
11.0%
19.0%
12.2%
64.3%
13.0%
2 000
0
1600
1400
1200 3
a.
1000 5
Q.
800 §
I I I I I I I I I
2009 2010 2011 2012 2013 2014 2015 2016
600
400
Source: M Z P (2017c:414)
In practice, municipal W M is set up mostly in the following way. Practically each house in
the Czech Republic has its own mixed residual waste bin (door-to-door collection). In case of
single family houses these are bins with volume of mostly 110-240 liters, while in case of
multifamily apartment buildings there are usually one or more larger bins with volume of
41
1100 liters. Sometimes there are multiple mixed waste bins available, especially in case of
larger apartment buildings. Contracted W M company then comes in scheduled times to
collect this waste and transport it for the further treatment, usually landfill or incinerator, with
the possibility of further intermediate mechanical-biological treatment for additional
separation of selected waste fractions. Frequency depends on the specific contract between the
municipality and the W M company. In case of smaller municipalities consisting primarily of
single family houses it is usually every week or every fortnight, occasionally less frequent,
and sometimes alternating with biowaste collection, if such bins for houses are used in the
municipality. Also, in parts of larger municipalities with comparable housing structure is this
collection frequency usually every week. In case of apartment buildings that are usually found
in larger municipalities, the provided large bins are emptied more frequently, typically twice a
week, as there is generally much less available bin capacity per person generating waste due
to the limited space around these buildings. Increased frequency of waste collection naturally
results in higher municipal W M costs.
In addition to the mixed residual waste, municipalities provide also options for separate waste
collection. In practically every municipality in the Czech Republic it is possible to separate
paper and cardboard, plastics and glass, and since 2015 also biowaste (green municipal
waste). These are usually collected at designed drop-off sites at one or multiple locations in
the municipality (depending on the size of the municipality) and consist of several larger
colored containers for each waste fraction that are usually also collected by contracted W M
company (but this can be contracted separately). There is not a single rule of where these
drop-off sites are located or how many bins they should contain - it depends solely on the
decision of the municipality as it needs to respect the space limitation and the usual amount of
the incoming waste. In some cases, typically in small municipalities, there is only one bin for
each waste fraction emptied every month. On the other hand, some municipalities might have
a large drop-off site with 10+ containers that are emptied practically every week and can
include bins for other separately collected waste fractions as well. In addition to that, larger
municipalities (usually 1000+ citizens) have often at least one household waste recycling
center (HWRC) with professional staff that provides additional options for waste separation,
including bulky waste, WEEE (waste electrical and electronic equipment), metals and so on.
In comparison with drop-off sites HWRCs offer better options for waste separation, but are
42
also limited by the opening hours that can vary from few hours per weeks in case of smallest
to those open practically every day from morning to evening in case of the largest ones. In
some municipalities there is an alternative to the drop-off sites in the form of kerbside
collection of selected recyclables (usually paper and cardboard and plastics). And finally, the
municipality is required to collect hazardous waste at least twice a year or provide a safe place
where people can dispose of it. Besides that it is up to the municipality to provide options for
collecting other waste fractions, either by using stationary collection points or regularly
scheduled collection.
In addition to that, there is also mentioned other municipal waste that people do not directly
dispose of but is still generated in the municipality. This includes waste like street sweepings,
waste from public bins or other waste from public spaces (including green waste). Collection,
transport and treatment of all this mixed municipal waste, separated waste and other
municipal waste sum up to the total waste expenditure that municipality has to cover, with the
primary role played by the mixed municipal waste. My goal in this thesis is to explore what
are the factors that affect the efficiency of the municipalities in this area. I will focus
specifically on the expenditure side of the municipal W M and examine what are the reasons
and factors that cause notably different efficiency (expressed typically by the per capita
municipal waste expenditure) in various municipalities. While doing so I will try to identify
both those factors that cause higher expenditures (and thus decreasing the efficiency) and
those that cause lower expenditure (and thus increasing the efficiency).
1.5 Efficiency in the municipal waste management - literature
overview
In this subchapter I provide an overview of the relevant studies that examined efficiency in
the field of waste management. Most of the included studies use frontier analysis methods
(namely DEA and SFA) that are suitable for examining a complex issue of efficiency in cases
where multiple inputs and outputs come into play. Both Czech and foreign studies are
included. First group of the studies focuses primarily on the organizational form of W M
service provision, second group employs primarily statistical methods for identifying relevant
factors and last group of studies is included in order to show the variability of the potential
43
efficiency factors that might have important influence on W M . The objective of this
subchapter is both to provide overview of the available research in this area and to identify
potential efficiency factors that would be later examined using available data collected for this
study.
Bel and Warner (2008) published a review in which they examined studies dealing with the
effect of privatization with respect to the cost reduction in WM. Generally they have found
little support for a link between privatization and cost savings, resp. the observed savings are
not systematical. The issue identified in the review is that in the research is rather too much
focus on the competition, which actually might not work that well in quasi markets, such as
municipal waste management provision. An important comment in the study is that with
absent competition, ownership makes little difference on costs borne by the municipalities.
Study then concludes that there is no systematical optimal choice between public and private
delivery of the service, and responsible managers should thus approach the issue in a
pragmatic way. Cost savings are simply not systematically found if looking at the issue from
company organizational type. Specifically, waste collection is characterized by weak
competition or collusion due to the trend to concentration in the market. According to the
authors, public vs. private debate places too much emphasis on the ownership when primary
attention should be given to the market structure, regulations, incentives and the level of
contract completeness. In waste collection weak competition between companies erodes any
potential cost savings. Authors suggest that the debate on privatization needs to move beyond
a debate on competition and ownership and instead look more closely at the costs of
contracting and the organization of service itself. These are considered the primary features
that will determine cost savings, whether under public or private production. The reasons why
private sector fails to deliver consistent cost savings is a result of inadequacy of the
theoretical framework based primarily on assumption about competition and ownership.
In the Czech conditions efficiency of waste management as a public service was examined in
more detailed by Ochrana et al. (2007), who focused on the role of the W M company
organizational form factor in the efficiency of the service. Study analyzed which form of
service provision is preferred, which criteria are important when selecting W M provider and
what reasons can lead to the change of the W M provider. Authors collected more than 900
44
survey replies from Czech municipalities and analyzed the answers together with the available
data about related municipal expenditure. The outcome of the study is that that the own
production of services is the most efficient, but this is due to the fact that certain related costs
of the service production are not directly assigned to the production of these services by the
municipalities, and therefore the reported expenditures are undervalued. Using only
expenditures directly reported by the municipalities on these services is thus incorrect when
comparing with the external provision of these services. The least efficient, on the other hand,
are municipal companies that were arbitrary selected without any competition. But as long as
the competition exists, the form of ownership actually might not matter at all, as all
companies compete among them.
There are more available studies from the Czech Republic that focused specifically on
contracting out selected public services (including waste management), with the conclusion
being basically in accordance with studies already mentioned. Merickova Mikusova et al.
(2017) examined various aspect of contracting-out public services in Slovakia and Czech
Republic. Authors here compare own and external production of selected public services
(including waste management) in the sense of question whether to make or to buy the service?
Study then verifies the hypothesis whether the contracting actually improves the effectiveness
of selected service delivery. Based on quantitative data analysis, this hypothesis was not
confirmed, both in Slovakia and the Czech Republic. Nemec et al. (2007) examined
implementation of contracting-out on the local level in Slovakia and the Czech Republic for
several public services, including WM. Study concludes that level of contracting-out is in
both countries relatively high, but the positive impacts from this type of public service
provision were not proved in either country. The main issues seen by the authors are in the
contractor selection process, less focus on the quality and potentially corruption. Suggested
solutions include pursuance of increased transparency, regular testing of service arrangements
and necessity to improved public procurement standards. Finally, Merickova et al. (2014)
based on the data from the Czech Republic and Slovakia again do not confirm that either
internal or external provision is of waste management is better and the focus should move
beyond the public vs. private provision debate, as already suggested by previously mentioned
studies.
45
Worthington and Dollery (2001) examined efficiency of M S W M using data of 103 local
governments in Australia. Study uses DEA with inputs including number of properties with
W M service, population density, cost index, occupancy rate (people per household),
population distribution and expenditure, while outputs included total waste, recyclable waste
and recycling rate. Provided results suggest that the W M inputs could be reduced to 65%
based on the observed best-practice while the productivity losses due to the scale effect were
estimated to account for 15% of total inputs. Study also suggests that strong inefficiency can
be in urban municipalities caused by the congestion and other issues related with high
population density, while in case of rural municipalities is the main problem the inability to
attain an optimal scale of operations. Comparable to this study is Domberger et al. (1986)
from U K who used collection frequency, population density and distance to disposal sites in
their costs function. As an output they used units serviced (pickup points) instead of
population, as the number of pickup points seemed to the authors as more important than
number of serviced people. Then density was again identified as having negative impact on
total costs although the lower proximity would suggest lower costs.
De Jaeger et al. (2011) published a study in which they examined impacts of some local
policies aimed at MSW cost reduction (basically on efficiency). DEA was used in order to
identify important efficiency factors. Used sample included 299 municipalities in Belgium.
Better results in terms of acquired efficiency were exhibited by municipalities that are
members of joint ventures focused on waste collection (WM outsourcing), or members of
voluntary agreements to reduce MSW. Different frequency or using of weight-based pricing
system appears to have no impact on the efficiency, but in this case the authors were
comparing weight-based pricing and pricing per bag and not system with and without unit
pricing, in which cases adoption of unit-based pricing tend to have a very significant effect on
the waste behavior of the people. Overall, results show that aiming at the MSW reduction
does not lead to lower efficiency of public service provision, on contrary. Used DEA model
included inputs of costs and outputs of residual waste and 4 other variables standing for
different types of recyclable waste fractions. After calculation of efficiency scores authors
analyzed effects of additional variables like population characteristics, density, political
orientation and income, what represents a common multiple-stage DEA approach. According
to the results, young and old people have positive effects on the efficiency, population density
46
has negative effect, what is comparable with observation by Worthington, Dollery (2001) that
densely populated areas suffer from congestion which in practice offsets any efficiency gains
from relative shorter distances to waste pick-up points. Finally political orientation towards
left spectrum was also identified as having strong positive effect on the efficiency.
Rogge and De Jaeger (2012) used a modified DEA for evaluating efficiency of municipal
W M using data from 239 municipalities in Belgium. Modified DEA was used as authors
criticize the limited explanation power of the basic DEA models, especially considering
increasing complexity of MSWM. Inputs in the constructed models were municipal
expenditures, while outputs were various types of collected waste fraction - residual waste,
green waste, bulky waste, packaging waste and other waste categories. In addition, 6
background characteristics of the municipalities (population, density, income, collection
frequency, pricing scheme and number of collected waste fraction) were analyzed against
efficiency scores of selected municipalities, as it is common in multiple-stage DEA approach.
However, this study was more of a demonstration of the modified DEA model that a study
focused on identification of important factors affecting efficiency and little results have been
drawn. Following this study, authors used practically the same model in Rogge and De Jaeger
(2013) where the second stage analysis of additional municipal characteristics is more
thorough and uses background characteristics such as percentage of old people and young
people, median income, political orientation, collection frequency and waste payment scheme.
Based on this municipalities were then divided into categories like residential, rural major
centers, etc. Significant and negative factors were identified to be the percentage of young
people in the population, political orientation towards green and social political parties, while
results for the income were mixed. Second presented model focused more on billing system
and collection frequency and was not corrected for demographic characteristics. Significant
and positive variables were then identified to be income level, and political orientation
towards green and social parties. Again, study does not provide any clear policy
recommendations or identify efficiency factors suitable for adoption, as this was not the
primary objective of the authors.
Simoes et al. (2010) computes the influence of selected variables on the efficiency of the
Portuguese urban solid waste services using DEA. Evaluated units were waste utilities - these
47
could be understood (after simplification) as companies (in the form of concessions, private,
or controlled directly by the state) providing waste management services. Variables used in
the model were operational and capital costs as inputs, and treated solid waste and recycled
waste as outputs. As explanatory variables were used GDP per capita, distance to the
treatment facilities, population density, regulation, type of management, composting and
incineration services. The identified inefficiencies were relatively significant and notable
reduction in costs could be potentially achieved. Theoretical optimal size of the system of
W M was at around 300,000 inhabitants. Strong regulation seems to be important as well, as it
sets certain levels for the services and contributes to the improvement and quality of the
service provided. Negative effects on performance were identified in case of rural areas or
high density population. Private utilities seem to be performing better public ones.
Composting and incineration also resulted in positive effects, even though in case of
incineration it was predictable.
Bosch et al. (2000) analyzed efficiency W M of 75 Spanish municipalities in Catalonia region
using DEA, SFA and deterministic frontier method. Inputs used were containers, vehicles and
covered kilometers. Outputs included biowaste, which was perceived as the most important
waste fraction in Spanish case. Focus of the study was primarily on the method comparison.
Based on that, DEA was considered to be the best method, as it is flexible enough to compute
specific features of the production process, unlike more rigid formulated models. In terms of
important factors influencing efficiency, authors stress that presence of competition seems to
be more important than private-public management dichotomy.
Garcia-Sanchez (2008) examined performance of W M and street cleaning in 38 Spanish
municipalities with population over 50,000. Study uses common two-stage DEA approach
with inputs staff, vehicles and containers and outputs being waste tonnage, collection points,
collection points density and km of washed surface area. After initial calculation of the
efficiency scores, significance of selected factors was examined in the calculation of an
efficiency index. These factors included population, tourist index, mean town temperature,
surface area of the municipality, population density, per capita income, economic activity
index and prevailing political ideology of the municipal population. Results show that there is
a potential for improvement estimated up to 8% reduction in the inputs. Considering the
48
analyzed factors in the subsequent analysis, only tourist index proved to be significant.
Additionally study asks whether public vs. private waste management of the W M company
makes a difference with the results showing that the distribution of acquired efficiency among
them is identical and does not show that either of them is optimal.
Benito-López et al. (2011) used D E A for identifying determinants of efficiency in the
provision of municipal waste management and street cleaning using data of over 1000
Spanish municipalities. Input in DEA model was per capita costs and outputs were residual
waste and a specific W M quality indicator (responsivity of the provided service). Additional
sociodemographic factors were used in the second stage of DEA in order to identify factors
with potential effect on the efficiency. Significant and positive variables were identified to be
per capita income, index of tourism importance, general economic activity in the municipality
and the prevailing political orientation of the population towards progressive parties. Negative
effect was again identified in case of population density.
Passarini et al. (2011) examined indicators of W M efficiency related to different territorial
conditions of municipalities using data from a selected Italian region. Performance indicator
percentage of waste separation was used for evaluating performance of W M systems. This
was tested against income variables in a correlation analysis, but no significant relationship
was identified. With respect to territorial variables, municipalities located in plains showed
both higher waste generation and separation, while those on the hills produced less waste but
had also lower separation percentage - effective collection service is according to the study a
problem in hilly areas (the same holds for rural areas). In terms of density, the most efficient
seem to be low density municipalities, while most densely populated areas show the worst
results.
Guerrini et al. (2017) applied a specific robust non-parametric method introduced by Daraio
and Simar (2005) to identify performance drivers of M S W M in 40 Italian municipalities.
Input of the model was expenditures and outputs were amount of residual and separated
waste. Several exogenous variables were then used in order to examine their relevance
considering efficiency of WM. Significant positive factors were identified to be the number of
49
inhabitants per house and kerbside collection (even though it is more expensive than street bin
system. Significant negative factors were the percentage of non-residents, size of the
municipality (smaller municipalities are more efficient), density, and tourist flow.
Moore et al. (2005) used DEA to identify efficiency factors in 46 selected US cities over 6
years period in 11 types of municipal services, including waste management, together with
multiple potential factors that could influence the efficiency. DEA model for waste
management used inputs staff and expenditure, while the outputs were number of citizen
served. However, study then does not go any further due to the lack of data for such
additional analysis and basically suggests that further studies are to be conducted.
Vishwakarma et al. (2012) used SFA in order to evaluate efficiencies of W M services in
selected 22 urban cities in India. Provided model assumes that municipalities want to
maximize collected amount of waste using inputs such as labor, costs, density, number of
available vehicles and vehicle technology level. Various functional forms of the classical
Cobb-Douglas function were used in the model. But study does not go beyond the calculation
of the efficiencies and only comments that larger cities tend to have higher efficiency scores
and actually that the costs do not seem to affect collection capacities in any municipality. On
the other hand, amount of labor and available vehicles has positive effect on the output.
Negative effect of the density was identified. The suggestions are then standard comments
that the less efficient municipalities should adopt appropriate practices and management
models from those that are more efficient, as according to the model they could improve their
capacity even without additional inputs. But beyond such general recommendation the study
does not provide any specific answer what factors are actually to be adopted.
Huang et al. (2011) analyzed performance of municipal solid waste collection using DEA
with the data from 307 local governments in Taiwan. Authors used five "key performance
indicators" for the evaluation including cost per unit volume of MSW collected, quantity of
MSW unit collected per unit collection time, quantity of MSW collected per collection
vehicle, population served per collector, and quantity of MSW collected per vehicle-mile. The
50
result of the study is then information about which local governments perform best among the
analyzed sample.
In the Czech conditions Fiala (2007) used DEA to examine W M of 68 municipalities. For
each municipality he identified 19 parameters including, for instance, population, amount of
solid waste, usage of PAYT charging scheme, rates of landfilling, incinerating, or recycling,
expenditure or revenues. Results are presented for selected model, where the inputted
parameters were amount of municipal solid waste and expenditures on municipal solid waste
treatment while output was amount of separated waste. An observation was made, that main
factor of efficiency seemed to be the presence of a PAYT charging scheme.
Finally I include two more studies from developing countries in order to show both the
differences and similarities related to the W M . Wilson et al. (2006) examined role of the
informal sector in W M in developing countries in order to pride also a bit different view at the
W M issue. Waste management in developing countries often heavily relies on the informal
sector. The issue in such countries is therefore not how to set up a new system for W M but
how to successfully integrate informal sector into the waste management planning. The fact is
that, despite the health and social problems associated with the informal recycling, possibility
to participate in W M provides significant economic benefits to the people in this informal
sector. Experience shows that it can be actually highly counterproductive to establish new
formal waste system without taking into account informal system that already exists. The goal
then becomes to build on the best practices and experience of the informal sector, while
working on improving efficiency and the living and working conditions of those involved. On
one hand it is true that such conclusions have a limited validity for purposes of this thesis, as
the role of the informal sector in waste management is in the Czech Republic much weaker
compared to the developing countries, but on the other hand this still, it provides an
interesting insight to the important aspects of waste management that are behind the
performance of waste management in a significantly culturally and economically different
environment. One thing we can learn from such study is that the informal sector can represent
a powerful force in the waste management. It does not have be in the forms of waste pickers,
but for instance in the voluntary green watches or the overall movement towards higher waste
separation, preference of waste minimization, etc. that can further improve existing waste
51
management also in developed countries where the formal sector is often expected to be the
one and only responsible body for the state of the waste management.
Second study is a review by Guerrero et al. (2013), and provides an interesting insight to the
factors that were identified as important in relation to waste management in the studies
focused on W M in developing countries. These include factors like family size, education,
income, household attitudes towards waste separation, support by the real estate company,
community commitment, public participation, fee structure, location of the household,
transport practices, route planning, improper bin collection, proper construction of the bins
themselves, lack of information about the waste pickup schedule, insufficient infrastructure,
poor roads, number of vehicles for waste collection, presence of market for recycled
materials, unreliable data about waste management, organization of the informal sector and
promotion of micro-enterprises for extending available waste management services and
others. As one can see, multiple of them are relevant for developed countries as well.
1.5.1 Identified factors of efficiency
Based on the provided literature overview I have identified potential efficiency factors that
seem to appear most frequently. I will divide the factors into two groups according to their
potential for further research from my perspective.
First group of efficiency factors includes those that can be either somehow directly influenced
by the municipality or can be relatively easily predicted, thus municipality can use this
knowledge in order to adjust the existing W M accordingly. These efficiency factors include
presence of competition, type of W M company ownership, waste collection frequency,
availability of kerbside collection and incentive programs, population density, rurality of the
municipality, territorial conditions and its age and household structure. In case of some listed
factors the effects are relatively easy to predict, for instance higher collection frequency
directly translates to higher expenditure, while in case of other like presence of incentive
programs, are the resulting effects on W M less clear.
52
The second group of factors includes those that are likely beyond the direct influence of the
municipality or are primarily affected by some external factors making them difficult to
predict. These second group of factors include political orientation, income of the people,
amount of tourists, general economic activity in the municipality, household occupancy or the
amount of non-residents in the municipality. This second group of factor includes basically
those are somehow defined, but generally municipality cannot do that much about them, so
they offer mostly ex-post explanation of the acquired efficiency results and their utilization
for planning W M in a municipality is more limited.
Important note here is that provided literature overview has its limits. As one can see, many of
the studies are focused on the same countries and thus the identified efficiency factors can be
biased. But this is also the problem of the overall availability of such studies from various
countries, especially if research in this area is not being done in other countries. If studies for
other countries were available, factors could have been different, as suggested by including
last two studies of W M in developing countries. However, in my opinion, provided studies
could be relatively well applied to the Czech condition, thus the identified factors should be
mostly relevant as well.
53
54
2 MUNICIPAL SOLID WASTE MANAGEMENT IN THE
SOUTH MORAVIAN REGION
The purpose of this chapter is to generally describe the waste management in the South
Moravian Region in the Czech Republic. I have selected this region for collecting municipal
waste management data from individual municipalities. First parts of the chapter contains
detailed description of the data collection process from available databases and individual
municipalities, which was necessary, as significant amount of the utilized data is available
only directly from the municipalities and is not present in any centralized database. The final
parts of the chapter include description of the collected data and additional extensive list of
qualitative findings obtained during the data collection.
In this thesis I primarily perform analysis of the quantitative data, or the data that could have
been transformed from the qualitative to the quantitative (usually in the form of a dummy
variable). However, as one can see from the list of additional qualitative findings from the
interviews with the municipal representatives including notes, opinions, comments, etc., there
is a substantial amount of factors that can have a notable influence in municipal waste
management, but are either difficult to be included in the quantitative analysis or are too much
municipality-specific. Nevertheless, I include these findings, as in my opinion they are of
value and importance when trying to understand better MSWM, and in order to show the
variety of factors that can influence waste management in the individual municipalities.
2.1 South Moravian Region in the Czech Republic
South Moravian Region is located in the southeast part of the Czech Republic and borders
with Austria and Slovakia. Regional capital Brno is the second largest city in the Czech
Republic. It is home to several universities and research centers and is well known for trade
fair center with a long history. Region has very good transport connection with neighboring
countries and capitals (Prague, Vienna, Bratislava) both by car and train. Region has
increasing potential in the areas of computer technology, telecommunications, software
development, biotechnology and other high-tech fields, thanks to the strong impact of the
universities in the region. Besides that more than 60% of the area is agriculture land, of which
83% is arable. The region is also well known for winegrowing (90% of vineyards in the
55
Czech Republic are located in this region, including numerous small wine producers). Region
offers also large potential for tourism, especially in summer thanks to the relatively warm
climate (warmest in the Czech Republic) and many historical and cultural sites, together with
several protected natural areas (Jihomoravsky kraj, 2018). Basic data are provided in Table 3.
Table 3: Basic data about South Moravian Region, 2017
Size 7187.8 km2
Population 1 178 815
Population density 164 inhabitants/km2
Number of municipalities 673
Number of towns and cities 49
Districts 7
Regional capital Brno (population 377 973)
Highest and lowest points 836m and 150m
Source: Jihomoravský kraj (2018)
Concerning the focus of my research, this region offers large amount of municipalities that
produce a large dataset, the region includes municipalities from very small with few dozen
citizens to towns with over 10 000 citizens, municipalities are located both in rural areas and
on the suburbs of large towns with most of them being at relatively close physical distance
from the center of the region. Such variety of municipalities offers also large variety of ways
how to deal with the municipal waste management with many of them being unique and
inspiring. The rest of the chapter provides a detailed description of the whole data collecting
process together with the basic description of the collected data.
2.2 Background of the data collection
My first question when considering the analysis of M S W M was what data to analyze. The
basic data to collect when examining M S W M are obvious - the amount of generated waste
and the expenditures, as I am looking at M S W M from the economic perspective.
The amount of waste in municipalities is typically measured in tons per year. As there are
multiple waste streams that are generated in the municipality, I had to differentiate between
them. Description of what kinds of wastes are collected in the municipalities is provided in
56
the chapter 1. The basic distinction between the fractions of municipal waste can be defined
for our purposes as the waste that is being separately collected with possible later recycling or
energy utilization, and residual mixed waste that ends up in the landfill or in the incinerator.
With residual mixed waste is the situation from the perspective of data collection simple, as it
consists of just one category (20 03 01 - mixed municipal waste, according to the waste
catalogue5
). With recyclables comes the problem that there is no one unified system how
different fractions are collected. Especially large municipalities collect and report a wide
variety of waste fractions, while some small municipalities collect and report only few
fractions besides residual waste, and some municipalities not even those. Thus in case of
recyclables we are dealing with a notably heterogeneous data. While the information is
available for multiple waste fractions, in order to have comparable data I have focused only
on selected waste fractions that are most commonly collected by the municipalities. Besides
obvious residual mixed waste, the most common fractions are plastics, paper and cardboard,
glass, and recently also biowaste (primarily green waste from the gardens).
While the Waste Act requires the municipalities to provide collection of also other waste
fractions, such as bulky waste, WEEE, or hazardous waste, the comparison of collections of
these fractions on municipal level has its issues. Unless the municipality has an HWRC, the
collection of these "additional" waste fractions is done usually through occasional mobile
collections that naturally cost money. Such mobile collection can be notably costly for small
municipalities that often cannot even fully utilize the minimum provided capacity of a mobile
collection truck. An economic thing to do is then to decrease the frequency of collection of
these waste fractions - resulting sometimes in less than once per year optimal collection
frequency (which is practically against the legal obligations). Alternatively the municipality
can make an agreement with a neighboring municipality that has an HWRC and ask its
inhabitants to discard their infrequently collected waste fractions there, either individually or
after some "informal" agreement among group of them in order to optimize the transport
capacities and costs. In both cases is such system incomparable with larger municipalities
where the collection of these waste fractions works on a regular basis.
5
Vyhláška č. 93/2016 Sb., o katalogu odpadu, interactive catalogue available at https://www.katalogodpadu.cz/.
57
Moreover, certain waste fractions, like WEEE, cloths or some other hazardous waste fractions
are sometimes collected by the private subjects (for instance in designated boxes in shops,
schools, appropriate public places, etc. or using take-back/discount/refund programs). Such
cases create again a distortion among municipalities and direct comparison at the municipal
level is then full of various issues and caveats. Due to that I omit quantitative examination of
certain waste fractions from the latter analysis and focus only on the main ones.
Expenditures related to M S W M are generally less tricky, although some issues exist here as
well. First, we need to distinguish between current and capital expenditures that have been
already mentioned in the chapter 1. Current expenditures are those spent on, simply put, dayto-day
or other regular activities related to MSWM, like waste collection, waste treatment,
and separation - or in other words activities with the waste itself. Capital expenditures are
related usually to irregular investments in the field of MSWM, such as the purchase of a new
collection truck once the old one broke down, one time purchase of home composters that
municipality provides to the households at reduced costs, setting up an HWRC, etc., or let us
say an "infrastructure" used by the municipality for generally dealing with the municipal
waste. Larger municipalities are more likely to have capital expenditures more often, as they
administer larger body of infrastructure, while in case of smaller municipalities capital
expenditures might occur only once in couple of years.
The next issue is related to the extent at which municipalities provide MSWM. If all
municipalities provided M S W M themselves, this would be irrelevant. But in this case we
should take into account the concepts of comparative advantages and economies of scale. For
a municipality it might be more efficient to contract a professional W M company to secure its
MSWM. While one municipality might rely fully on this external company, the second can
still take care of some of the W M related activities, while the third can fully secure M S W M
on its own. In case of largest municipalities it actually might be, due to the legal issues, the
only possible way how to handle their MSWM. It is logical that the first municipality will
most likely have little to no capital expenditures, while the second one some, and the third one
the highest, even though these municipalities might be providing practically the same level of
MSW services to their population. While this might not seem as an issue at first, as in all three
cases certain investments have to be made - in case of the third municipality such expenses
58
come up as capital expenditures, while in case of the first municipality it is up to the external
company to deal with this - and the costs from the perspective of the municipality show up as
the current expenditure that are reflected in the regular payments for the W M service to the
external company. But in practice is this typically not that much of an issue, as large majority
of the municipalities use external companies from securing their MSWM, and therefore face
more or less the same situation in this area.
In this thesis I focus only on the current MSWE. These expenditures are more likely to be
comparable among various municipalities due to the activities they primarily cover - which is
waste collection, transport and treatment. While in case of municipalities contracting external
company are these expenditures used to pay flat fee per person, per ton or some other agreed
sum, in case of municipalities providing M S W M themselves it is used for covering wages,
landfill fees, transport, etc. directly. But as the external company has practically the same type
of expenditures, current MSWE represent in my opinion a sufficiently comparable measure.
Besides generated waste amount and related expenditures I have collected also several other
quantitative municipal data that are in my opinion relevant for MSWM. These include:
Population. Amount of the people living in the municipality is clearly connected with the
amount of generated waste and therefore costs as well. The very definition of waste is that it
represents the goods that people dispose of - goods that people throw away. The direct link
between the amount of the people and the amount of generated waste is then there to be
expected. Moreover, it also enables to calculate relative per capita measures, as these are
necessary if comparing municipalities of different sizes.
Age structure. In case of age structure I expect that people at different life stages exhibit
different waste related behavior. For instance young children are largely affected by their
parents, adults that go to work and earn money tend to generate more waste as they consume
more, and retired people with limited income behave differently as well.
59
Other socioeconomic data like unemployment rate, amount of students, or education of the
people, that, according to some studies, might have some effect on the waste generation and
related costs.
Another often mentioned factor related to the municipal waste management is people's
income that is directly connected to the waste-related behavior of the people. The relation is
simple - with higher income people are able to buy more, consume more, and as a result
produce more waste. Many foreign studies analyze different waste-related behavior of
households depending on different income levels. However, in case of the Czech Republic,
information about income levels is not available at the municipal level. Although this should
be possible from the tax returns or from the data from social security office even at the
municipal level, publicly it is not available. I have contacted directly Czech Statistical Office
(CZSO) with a question regarding this issue and how they calculate the value of average and
median wage. Based on their response, this is calculated using the information from a
representative sample and not from actual data from the individual people. With this data
CZSO is able to estimate income levels for the whole country, regions, and in some cases the
largest municipalities, but it is not possible for individual municipalities.
One way to overcome this issue, which is very common in foreign studies, is to focus only on
a small sample, let us say few households or few streets. In this case the researcher can ask
the household directly about their income and relate it to their waste generation. However, the
size of the sample is in such cases very limited, can practically cover only small areas and the
results can only be extrapolated to the whole with a high possibility of bias. Therefore in this
thesis I do not include income levels among the factors I analyze, simply because in the
Czech Republic this information is not available at the municipal level.
Next group of the data that I have collected are data directly from the municipal
representatives of the individual municipalities, as in practice these data are not available
anywhere centrally. These data include:
60
Specific information how MSWM in individual municipality works, how is M S W M
provided and what kind of M S W M infrastructure is available in each municipality, both for
residual mixed waste and the recyclables. Some elements of local W M are relevant for
multiple waste fractions, while some can be specific for one or the other. In case of residual
mixed waste, collected information include data on waste collection frequency, whether
households have individual or shared waste bins and if there are any incentives related to the
household waste generation. In case of recyclables the collected data reflects the available
options for waste separation. This includes the prevailing system of waste separation (if any,
but in almost all municipalities people have the options for waste separation) - this can be
done most commonly by a bring-in scheme or a kerbside collection, frequency of recyclables
collection, variety of waste fractions that are collected, presence of any kind of incentives
related to the recyclables, etc.
Type of WM company - information whether the company securing M S W M is private,
mixed, municipal or publicly owned. Multiple foreign studies examine differences among
these companies and come with various results. The question these studies ask is which of
these companies provide M S W M the cheapest, or better, or the most efficient - whether the
private companies are the most expensive because they focus on profit, or if the private
companies are cheapest due to their effort to be efficient or due to the economies of scale. Or
if the best company is mixed, because it connects positives of private ownership with the
positives of public control, or the opposite, that it brings together the tendency towards
inefficiency from public sector and higher costs due to the profit pursuit from the private
sector.
Available competition - I have asked municipal representatives about the competitive
environment in M S W M in their municipality from their perspective, whether they are able to
choose the provider of MSWM, or they practically have to accept the offer from the only one
available company. It is then natural that if municipality can choose from multiple offers or
has multiple available companies that could provide MSWM, that it is more likely to get a
better deal. Another perspective is the bargaining power of the municipality depending on its
size. While in case of small municipalities it might not be very profitable for a W M company
from greater distance to provide W M services for a single municipality, it is significantly
61
different in case of the largest municipalities. In those the amount of required W M service
creates a greater motivation for the companies even from a far distance to offer M S W M
service provision as the possibilities for economies of scale emerge.
Political situation. Finally, I was initially thinking that the political orientation of municipal
council and mayor might have some impact on MSWM, but I later have abandoned this idea.
In the Czech Republic majority of municipalities are relatively small, with only several
hundred people, and do not have a fully professional administrative staff - the mayor is often
employed only for a part time while having other regular day job, and this holds usually even
more for other municipal representatives. In case of small municipalities the largest problem
seems then to be to secure all kinds of obligations that a municipality needs to provide and the
differences between the political orientations of individuals often become secondary. The
situation in largest municipalities is notably different, with larger council, professional
administrative staff containing many people and highest officials focusing primarily on the
strategical decisions rather than operative decisions and actions. However, in the ends there
are actually only few larger municipalities that could provide some data regarding this aspect,
which prohibit from a more reasonable quantitative analysis.
2.3 Data collection process
Data collection process for this thesis can be divided into two, resp. three parts. First group of
the data was collected from the databases and involved practically only computer with
internet connection. Second group of data regarding M S W M was collected directly from the
municipalities by personal visits. Third group of data was collected directly from the
municipalities by telephone interviews (and in few occasions also by e-mail correspondence).
I provide a detailed description of the data collection in the municipalities, as it might be of
help to others planning to conduct this kind of data collection (not necessarily in the area of
MSWM) and they might also benefit from my experience in terms of what to expect, what to
be prepared for, and how to plan the data collection process - while it might not seem as that
important, when performing a larger scale data collection, mistakes and insufficient planning
62
can cause a lot of delays or time wastage, which would be, especially with respect to the focus
of this thesis, inefficient.
Public databases
The first group of collected data contains statistical information about municipalities and their
MSWM. These data include various municipal socioeconomic data, waste generation amounts
and related municipal expenditures.
Waste generation related data are provided by the Ministry of the Environment of the Czech
Republic (MZP), specifically at https://isoh.mzp.cz (in Czech only), which stands for Waste
management information system (Informační systém odpadového hospodářství). This
information system includes modules with aggregated data on waste generation for various
waste fractions, registry of authorized subjects in the W M area including addresses of waste
related facilities, and more. Most important source of information for this thesis is the module
VISOH. This module includes, among others, information about waste generation reported by
the Czech municipalities. One issue I have encountered is that the data is publicly provided
only at aggregated levels, with the smallest unit of aggregation being area falling under the
municipality with extended power (ORP, Obec s rozšířenou působností). There are 216 such
municipalities in the Czech Republic, while there are more than 6200 individual
municipalities. The organization responsible for this information system, CENIA was kind
enough to provide data about waste generation for selected years at the municipal level. With
such data, it was possible to conduct further analysis with respect to the individual MSWM.
Municipal solid waste expenditures for individual municipalities are in the Czech Republic
publicly available dating back to 2001, respectively 2000. Throughout the years the
information systems providing this data changed, but the public availability remained. In all
cases the information system was run by the Ministry of Finance of the Czech Republic (MF
CR). Data from 2001-2009 are available in ARIS at http://wwwinfo.mfcr.cz/aris/. Data from
2010-2012 are available in UFIS at http://wwwinfo.mfcr.cz/ufis/. From the user's perspective
these two information systems are practically same with same available information and same
functionality. They provide individual municipal information about expenditures and
revenues according to the relevant legislative classification, income statements, balance
63
sheets, budgets and some other financial information. Some of this information is available
also for nonprofit subjects under the municipalities. User can also generate aggregated data
sheets for larger administrative units, even for the whole republic, depending on the specific
financial report. Generated data can be exported and further processed by a computer.
Same data for 2000-2013 are available also at the site http://www.rozpocetobce.cz, run by a
voluntary association, and it provides better options for time comparisons, provides simpler
navigation for the common user, is graphically more appealing, and offers possibilities for
generating some simple analyses for the selected municipality. Compared to the two previous
systems it is much more user friendly, although exporting options are less complex.
Municipal financial data since 2010 are available in information system MONITOR available
at http://monitor.statnipokladna.cz/, run by the MF CR. Most recent available data are usually
no older than two calendar quarters and they are reported quarterly for the most recent year
and annually for the previous years. Compared to the other mentioned information systems,
availability of current quarterly data is a notable benefit. MONITOR offers simple navigation
for the user, is quite user friendly and offers possibilities for some time comparisons and
analyses of the municipal finance. It is also possible to generate aggregated reports for some
larger administrative units than municipalities. From the perspective of this thesis, the most
important functionality of this information system is that the visitor can download complete
source data used by this information system that contains expenditures and revenues for all
individual municipalities and other municipal organization in the Czech Republic in a
common data format. This makes is relatively simple, although not trivial, to make further
analyses of the municipal financial data. Public availability of such data at the municipal level
has a great potential not just for the academic purposes, as it provides practically anyone with
the option to analyze selected municipal financial data. While in case of the user interface it
might be at first difficult to find some specific data, as the navigation was designed for a
common user and often aggregates data from certain fields, in case of the source data it is
rather simple for the informed users to quickly find what they seek.
64
One important note to make here is that one cannot fully rely on the provided data, whether it
is the waste generation or the municipal finances. The issue stems from the fact that this data
is in most cases reported manually by the responsible person at the municipality. With larger
quantity of the data is might happen that the reported value is incorrect, misreported in the
wrong/inappropriate/inaccurate category, or missing at all.6
Reasons for that are insufficient
attention by the reporting person, insufficient knowledge of where to report what value (or
whether to report it all), or simply insufficient time for data reporting, even though
municipalities are obliged to reported this data by the law. Reason behind all this is likely in
the size of the municipalities (over 6200 municipalities with median population slightly above
400), as smaller municipalities usually do not have any full-time professional staff, and if,
such person has to cover wide variety of tasks in order to provide and secure everything a
municipality is supposed to do. In such cases it is then obvious that many activities are
prioritized over reporting some statistical data to the central bodies. On the other hand, such
errors occur, according to my long-term experience, relatively sporadically and usually can be
dealt with by either omitting the municipality from the analysis if the data is missing, or by
trimming the dataset from the evident outliers. The fact that the Czech Republic has so many
municipalities requires a lot of work with the data, but it also creates a large dataset, where
such reduction has only a marginal effect on the final sample.
Socioeconomic data about individual municipalities at annual basis are available from CZSO
public database site https://vdb.czso.cz, together with some more detailed and less frequently
collected data (namely about housing structure) available in the Census section at
https://www.czso.cz/csu/sldb. Census in the Czech Republic occurs every decade, with the
last taking place in 2011. CZSO also allows downloading of the majority of the source data in
a common electronic format for possible further analyses. Some additional municipal
information, or let us say information available in a more user friendly way, can be also found
at Regional Information Service site at http://www.risy.cz/cs. However, to my knowledge it
does not offer the possibility do download the source data, only viewing them for individual
municipalities. On the other hand, it provides some useful thematic maps.
6
If any doubts about the data arise, one should double-check, for example with the relevant municipality. More
on these limitations can be found for instance in Ochrana et al. (2007:30), or Bakos et al. (2009).
65
Pilot study
Second group of the data used in this thesis contains information about M S W M in a selected
sample of municipalities. The primary research objective of this thesis is to identify and
examine the efficiency factors in MSWM. Although there are numerous studies that examine
factors that influence MSWM, in my case I did not want just to adopt these factors. Instead of
that I have selected specific municipalities from the region, set up meetings with the
responsible person for the waste management in a given municipality and visited them in
person in order to better understand how M S W M works in the municipalities. I had prepared
a list of general questions regarding M S W M that I have asked these people during the
interview, together with some questions regarding available waste generation data and related
MSWE for the double-check purposes. Nevertheless, majority of the interviews consisted of
the explanation of M S W M in the given municipality, how it was run, what kind of system is
used, what is working, where are the problems, how are people accepting the system, where
the municipality sees further opportunities and so on. The thing is that each municipality is in
its own way a unique entity and M S W M in no two municipalities is absolutely similar. In
many cases were the provided information about the specifics of M S W M in a given
municipalities very interesting and practically impossible to predict unless the responsible
person brought them up. Unfortunate fact is that such information in not otherwise available,
there is no database of how municipalities specifically provide their M S W M and in each
municipality the municipal representatives have to adjust the system to the local unique
conditions. The result of these visits were contacts with the responsible municipal
representatives, adjustments to the list of factors that are possibly relevant in M S W M and a
much better idea about the practical aspects of M S W M in Czech municipalities.
One thing that might distort one's perception of M S W M is that a person usually lives only at
several places during the lifetime for a sufficient time in order to get at least some idea how
M S W M in that municipality works, if at all. Based on this experience, we are likely to expect
that M S W M in other municipalities is more or less the same as in our municipality based on
our own limited experience. But this expectation does not hold anymore once you start to
examine M S W M in dozens and even hundreds of municipalities with each of them having
some unique features and providing M S W M slightly or even notably different. Face to face
conversations with the responsible municipal representatives took usually around an hour,
66
depending on the time limitations from their side and their willingness to share the
information on the issue. However, a greater problem was to find a suitable time for the
conversations for both sides, as these municipal representatives have to fulfill their regular
duties. And finally, as the conversations took place at the given municipality, it took some
time to actually get there. All this required careful time planning and necessity to follow often
very tight time schedule. Some of the days on which these interviews occurred practically
consisted of alternation between half an hour (even up to one hour in case of more remote
municipalities) driving to the municipality, some time for preparation for the interview (as the
municipal representatives react much better if the interviewer comes prepared and has at least
some awareness about the visited municipality) and around one hour of interview with the
municipal representatives. This could begin very early in the morning, as from the experience
municipal representatives strongly prefer to work from early morning and are often not
available anymore later in the afternoon. Moreover, in many cases the municipal
representatives are not available every day during the whole working week, as they can have
other jobs and other duties. Also, in order to efficiently use available time, I tried to chain
multiple municipalities in the same travelling direction - it is not very wise to plan interviews
in the municipalities that are hundred or so kilometers apart from each other, one will then
spend most of the time by just transporting. Due to the time demand of the interviews, limited
availability of the municipal representatives and the necessary time for transport between the
municipalities, I was usually not able have more than 3-4 interviews per day - which was,
based on the practical experience, rather a solid result, due to the above mentioned issues.
One can ask why I did not choose to perform these interviews over the phone. While
theoretically this might seem as a much simpler and more efficient option, in practice it often
comes with several issues. From my personal experience municipal representatives are much
more willing to share their information once they get proper explanation and reasons for the
study plus they have contact with the interviewer. Also many people are not willing to spend
hour or so on the phone speaking with a complete stranger that they have never seen and
cannot associate with in any way. I have found face to face contact in such kind of situation to
work much better, although the drawback in the form of the time demand resulting from
transportation is very high. At this stage quality mattered more than quantity.
67
An important thing related with the interviews was to talk about M S W M with the appropriate
person. In my case I firstly always visited the website of the municipality and looked for the
information related to the W M , or environment in general. Nowadays majority of
municipalities in the Czech Republic have some website, therefore in most cases I was able to
acquire at least some basic information about the municipality in the examined area.
Municipal website usually provides information about the current municipal representatives,
often with direct telephone and e-mail contacts, and if not, at least a general contact to the
municipal administration office. In addition to that, municipal office hours are among other
things that are mostly available on the website, so I could already suggest appropriate times
for the visit with respect to the office hours. Most of these scheduled visits then occurred in
the municipal representative's office, but I have also visited a person running municipal waste
company at their headquarters, or a person responsible for both M S W M and a landfill within
the cadaster in the office directly at the landfill. In all cases I was initially looking for the
person officially responsible for M S W M in the given municipality according to the provided
information on the website.
These persons were at first contacted with an e-mail describing the purpose of the research
and how the research would benefit from the information they could provide. The e-mail also
included a part in which I was asking the e-mail recipient to forward the e-mail if it was not
addressed to the correct person, which proved useful multiple times. Sometimes the
information on the website were not current anymore, sometimes I was not able to clearly
identify the correct person (or such information was not provided at all), sometimes the head
of the administrative section forwarded it to some other person that was more specialized in
the agenda relevant for my research, and sometimes the information on the website were
simply too brief and only few main contacts were provided - therefore I had to send the email
to the available address and hope, someone will look at it, consider it worth a while and
not delete it right away. After some time ranging from hours to few weeks (as some of the
recipients were on holidays) I got most of the replies often with direct and working contacts to
the correct persons and could start making the arrangements for the personal visits.
From my experience municipal representatives were in most cases surprised that anyone is
interested specifically in their work and often initially suspected some commercial offers or
68
something of that kind. In some cases they also seemed to be rather careful and not willing to
tell anything that might be perceived as a negative aspect, perhaps worrying that it might be
used against them and possibly even have a later negative effect with respect to their career as
if they already had some experience with getting victimized if they expressed negative
opinions about certain things.
This part of the research is often called a pilot study - an initial part of the research where
researcher tests data collecting on a smaller scale in order to get some feedback and possibly
better understand the examined issue. In order to get more relevant information about MSWM
I visited mostly larger municipalities with over 1000 population (27 visits in total), even
though majority of the municipalities in the region are less than half this size. But this was
later more than balanced during the telephone interviews, where I have contacted by far
mostly small municipalities and only very few remaining larger ones. I should note here that
the main difference between M S W M in these municipalities is that in a small municipality
usually only few (2-3) measures of dealing with the waste are used, while in larger
municipality M S W M is much more complex and covers wide array of waste related activities,
which actually makes M S W M in larger municipalities much more alike. On contrary, the
differences between individual small municipalities in terms of M S W M are in my opinion
much greater than in case of the larger ones, even though description of M S W M in a single
small municipality is much briefer compared to a larger municipality, as there simply is not
that much variety of activities done.
During each of these visits I have taken notes extensively, as these interviews were often full
of new information regarding individual municipality's waste management, and if there were
multiple visits chained one after another for several days, one cannot simply both remember
and distinguish easily between the municipalities. In some cases the interviewed municipal
representative could not provide certain additional, mostly statistical information during the
time of the visit, and either sent it to me later or I called back to get it at a scheduled time.
This was usually a case of the number of drop-off sites for separated waste in case of larger
municipalities, development of the waste fee in previous years, amount of collected waste for
individual waste fractions or the amount of the expenditure spend on M S W M in the previous
years. In some cases the data I had obtained from central databases did not corresponded with
69
the data provided by the municipal representative, as was already mentioned. However, the
deviations were usually only marginal. In such cases I have replaced the data from databases
with the data provided by the municipality.
During these visit I also asked the municipal representative about any notable changes in the
development of the examined M S W M characteristics, like the amount of the development of
the generated waste or MSWE. This part was from my perspective very important, as there is
usually no other way how to identify the reasons for the changes than to ask the people right
at the source. Some of these observations are discussed later in the qualitative findings
subchapter. In one such case I identified a sudden drop in the amount of MSWE without any
obvious reason like the decrease in the amount of MSW. After directly asking municipal
representative about this observation, the accountant remembered that the W M company
forgot to send the invoice for the second half of the year on time, and the municipality paid it
in the next year. Without such information directly from the municipality, the municipality
would suddenly become among the most efficient, as the amount of waste remained stable
with the costs being cut in half. But after taking into account the additional information about
the late invoice and subsequent payment, the municipality showed no more dramatic changes
between the years.
Telephone interviews
After finishing these visits and collecting some of the missing data that were not provided
directly during the personal visits, I have created a set of questions that I distributed in the
next stage of the data collection process to the municipal representatives of other
municipalities within the South Moravian Region. At first in the form of e-mail and after a
rather modest response rate I contacted individual municipalities by the phone, which
produced much better results in terms of amount of the collected data. Also by doing this I
was able to get immediate response and feedback, I was able to make my points more clear, if
necessary, and also got some additional information while talking generally about the current
state of W in the examined municipality. Depending on the conditions and willingness of the
other side to communicate, such telephone calls lasted mostly between 5-20 minutes, on
average about 10 minutes. During this time I introduced myself, explained the purpose and
70
the objectives of my research and mentioned already sent e-mail (what many of them recalled
and sometimes apologized for not replying due to the lack of time or so). In some cases the
municipal representative requested to resend the questions instead of answering them directly,
as they preferred doing it at their own pace, while many other representatives requested
scheduling another telephone sessions due to their other ongoing activities. In majority of
cases this rescheduling worked out, and at the next time they had both time and sometimes
already some relevant material prepared in advance.
Sometimes, during the initial phone call I was given contact to a different person that was
more qualified to answer my questions and later tried to contact him. Also some contacted
representatives tried to excuse themselves from the interviews due to the insufficient
knowledge about their MSWM, but in many cases I was able to persuade them that majority
of questions are easy to answer for almost anyone from the office at a small municipality,
which almost always turned out to be true and I was able to get sufficiently qualified answers.
The advantage of municipal staff in the small municipalities is that basically everyone knows
at least something about everything, so even the accountant or a clerk have sufficient
knowledge about MSWM in their municipality from the perspective of my research. And if
they did not know the answer to something, they often simply asked their colleague at the
next desk for the immediate answer. However, speaking directly with the mayor or vice
mayor often had the advantage of providing a broader perspective of the examined issue and
connected it with the other aspects of running the municipality. Only in one or two cases
during the entire data collection I was directly turned down, as the municipal representative
did not was to participate in the research at all without any reasoning whatsoever.
Interesting observation during this data collection was often negative attitude of many
municipal representatives towards providing any kind of quantitative data about their
MSWM. First of all, such data are in my opinion from their nature public, whether it is the
amount of generated waste or the related expenditure. Even according to the Czech
information law such information cannot be kept secret. Also many people seemed surprised,
once I told them that I am in fact just double checking the data that are already public.
Another thing is when looking at this from the perspective of a municipal representative - as
being a public subject dealing with the public money, why should it not provide such
71
information if someone asks for them, especially considering that municipality represents in
many aspects a monopoly for providing certain services? Sure, it is understandable that
providing such data might require some additional labor, depending on the extent of the
request, but still.
In some cases I have identified a notable change in interannual data about the generated waste
amount or MSWE. In such cases I have confronted the municipal representatives with this
observation and asked them, if they recall anything that might have caused such change.
While in many cases they did not provide any answer, as they either did not remember or they
did not have a clue, those who recalled something often provided important and interesting
answers that one practically cannot acquire by any other means apart from just asking
directly. Some of the mentioned reasons can be straightly translated into the factors that
notably influence the efficiency of MSWM. For instance a sudden change in MSWE might be
caused by the change of the W M company and/or from the new contract. In case of waste
generation the reason might be the introduction of green waste collection, new HWRC, other
something else. Again, all such changes have one thing in common- unless you have direct
contact with the responsible people in the municipality, there is practically no way how to get
such knowledge from the centralized databases. Interesting and important findings from these
parts of the interviews are included in the subchapter about qualitative findings.
At the end of the interviews I always asked the municipal representatives an open questions
about their own observations in MSWM, what they perceive to be working, what should be
further improved, what kind of issues they face, what kind of improvement they have
implemented or plan to implement and so on. This part of the interview goes beyond the
efficiency evaluation, as the answers provided here are very difficult, if at all possible, to
transform into the quantitative form for the later analysis. Nevertheless, the very large variety
of the answers, full of interesting observations and ideas, provides a valuable source for a
more complex understanding of the problems and challenges in the field of MSWM. Many of
the mentioned things are the result of very creative thinking of how to deal with the things
while taking into the account unique local conditions - but on the other hand, many of them
can be generalized and serve as a good (or in some cases bad) example to the other
municipalities, that might benefit from the knowledge that something like that can be done,
72
and that there is already someone that has tried it and some experience with it. It occasionally
happened, that while one municipality had for instance creatively dealt with the separated
plastics waste fraction by selling it for profit to the nearby factory, another municipality
practically on the other side of the hill few kilometers away lamented that they are paying too
much to the W M company for dealing with separated plastics. Findings from these parts of
the interviews are again included in the qualitative findings subchapter.
The actual collection of the data by telephone interviews took place during most of the
summer and part of the fall 2014. It practically consisted of sitting in front of the computer
since morning to the late evening, speaking with the municipal representatives and taking
notes from their answers. While this might not seem as a difficult task, there were couple
issues related with the practical part.
First necessary thing is to make the contact at all. While majority of municipalities have at
least some website, information there is not always current. Although it sometimes took some
effort to find a correct and working phone number, in almost all cases I was eventually able to
make the contact at least with some clerk from in municipality.
The next thing was to contact the appropriate person. If the person I have contacted at first did
not have sufficient knowledge about M S W M and could not be persuaded to continue with the
interview, like I mentioned before, I was usually given the contact to the responsible person
or at least a time window when such person should be available at the office. While this
worked in majority of the cases, I have encountered municipalities in which even after
multiple calls and rescheduling I was still not able to interview the responsible person,
whether it was because the person was always unavailable or did not have time for me.
The thing that turned out to be crucial was to contact the municipality (or at least try to)
during the office hours. As majority of the municipalities provide only their landline number
on their website, it is in the most cases pointless to try to reach them outside these office
hours, as there is simply no one there to answer the phone. This is usually not a problem in
73
case of larger municipalities with professional full-time staff, but as I have said, majority of
the municipalities in the region are small with limited office hours. It was relatively common
that some municipalities have office hours only 2-3 times per week for just few hours. And
also many municipalities tend to have their office hours at the almost same exact time, thus
one cannot schedule the calls that easily, as there are multiple time constraints. In extreme
cases the office hours of municipality were just one hour per week or even biweekly. Based
on this I eventually created a time schedule of office hours for the remaining municipalities
from the sample. Then, at the start of each day during the interviewing phase I have filtered
out those municipalities that should have office hours and created a day call plan. Here I
stress the word should, as in same cases the office hours of the municipality did not seem to
apply according to the provided information, even after repeating the attempts over the course
of several weeks.
Due to the observation that many municipalities tend to have limited office hours at the exact
same time, after several weeks of interviews I eventually ended up with a bunch of
municipalities with office hours condensed into only few time windows during the week.
From my experience I can say that the most common municipal office hours occur on
Wednesday afternoon and Monday morning, while Tuesdays, Thursdays and Fridays are the
least popular. Although it might be surprising, there were actually few municipalities that
have their office hours during the weekend. However, the reason is pragmatic - such
municipality was most likely including a large recreational area with recreational housings
and their count sometimes greatly exceeding the number of permanent housings - thus there
was much more people around during the weekends, and it was simply more convenient for
these people to meet with the mayor and administrative staff and solve their problems during
the weekend and not during the regular working days. This is just one of the examples of how
variable the municipalities can be.
Moreover, one has to deal with the things like holiday, especially during the summer, when
the responsible municipal representative could not be available for a week or two. In other
cases the person with sufficient knowledge about M S W M might be in the office only for few
unscheduled hours per week - and during this time this person has to deal with other more
important issues than to be interviewed about M S W M for some research purposes. If such
74
person does not provide personal contact, there simply is no other way how to get the desired
information. But even in such cases one has to be prepared that many of these municipal
representatives work for municipality only part time, and it might be inappropriate to disturb
them outside their office working time.
Next issue to consider is that collecting these data over the phone can be quite physically
demanding, at least in the initial phase when there are many municipalities with available
office hours. Each of the calls contained up to 20 minutes of talk, followed by approximately
the same time for completing the notes, sending further questions that were not answered
(usually regarding waste and expenditure statistics) together with my e-mail and phone
contacts, or eventually a supplementary information about the research. However, besides the
call itself, especially in later stages it proved challenging to find the next available
municipality, and thus another 10, 20 or even more minutes could be spend without any
successful contact. Altogether, I considered contacting and collecting the data from around 10
municipalities per day to be above average success rate, especially at the later stages of the
data collection. As I mentioned, in the final stages of the data collection process I ended up
with only few time windows per day with only few municipalities to contact. And if I was not
successful, I tried calling another day, or eventually another week. So in the end majority of
this process consisted of a lot of time spent on the phone talking to strangers about things you
are only partially familiar with. Although the very basics of M S W M are often same in every
municipality, each municipality usually has some unique features that should be noted for the
possible later utilization. If you contact 10 or even more municipalities per day, it adds up to
2-3 hours of time spent just talking. According to my calculations (and the bill that the
university got later, as it was the place where I have made majority of the calls), altogether I
have spent more than 100 hours (6000 minutes) just on the phone talking with the municipal
representatives, not including the time spent with completing notes, planning calls, doing
research about individual municipalities and making further contacts with the municipalities.
Time period of the collected data
Last thing to address here is the period for which I have collected the data. As mentioned
before, municipal financial data are available practically since 2001, and since 2010 in a
75
relatively user friendly form that allows making large set analyses. However, in case of waste
generation data, the situation is worse, as I was able to acquire data at the municipal level
only for years 2009-2012. In case of socioeconomic data the situation depends on the type of
the data. For example, for variables like population is the data easily available for practically
every municipality for every year, and there is a high change that such data is correct. But
other data like housing structure is not collected on the annual basis and in such cases I had to
rely on the most current available data that are from the national census in 2011. With such
limited availability of the data I used the 2011 values of certain data as a constant in some of
the further analyses, as there simply are no other more current data available. However, in my
opinion this in most cases should matter that much, as with data like this I do not expect a
large interannual differences in the relatively short period that I was examining.
And finally, in case of the data collected from the municipalities, the latest data are usually
from 2013 or 2014. In case of the historical municipal data collection, I have decided not to
go further than 2009. The reasons are both that I do not have any older municipal data on
waste generation and that from the perspective of the municipal representatives it is already
quite far in the past. In many cases the municipal representative simply did not remember
details about M S W M from more than 5 years ago, for instance because the personnel has
changes after the election, or the person was not responsible for M S W M at that time, or there
was a new person employed on the position and thus might not have any knowledge about
their M S W M prior to his appointment.
2.4 Data description
In this subchapter I provide description of the data I have collected from the sample of
interviewed municipalities together with the explanation of the individual variables. Variables
were selected based on initial examination of the overall process of W M in the municipalities
during pilot study and based on the typical data used in relevant literature dealing with
M S W M efficiency. In the end I was able to make contact with 535 municipalities in the South
Moravian Region. Each municipality was identified by its name and a territorial unit number
(ZUJ, základní územní jednotka). List of the collected municipal data includes:
76
Identification of WM company providing MSWM services in the municipality, including
the information about the ownership type of the company (private, public, mixed or
municipal), length of the current contract, previous W M company and amount of other
potential W M companies that could provide M S W M in order to obtain information about the
level of competition in this area. Based on the interview, approximately 25 different W M
companies operate in the examined sample of municipalities and include all types of company
ownership. Length of the contract varies from more than 20 years to freshly contracted W M
companies, depending on the perceived satisfaction of the municipality with the provided
W M services and the price for it. In terms of perceived competition, approximately 20% of
municipalities suggested only 1 relevant W M company, 20% suggested 2 relevant W M
companies, 30% suggested 3 W M companies, 15% suggested 4 W M companies and
remaining municipalities suggested 5-9 possible competitors available.
Waste collection frequency, any seasonal changes in the frequency, development of the
collection frequency in the previous years, and any differences between different parts of the
municipalities. Collection frequency can have significant impact on the overall costs of
MSWM. Usually municipalities with less frequent collection are able to lower their MSWE.
On the other hand, cooperation of citizens is also important and decreased frequency needs to
be accompanied with decreased generation of residual mixed waste in order not to overstock
the available capacities of the bins. Depending on the specific municipality, there might be
seasonal differences due to the reasons like generation of green waste or a higher turnout of
people in case of municipalities with significant amount of seasonal housings. Frequency of
collection can also be different for parts of the municipality with single family houses and
apartment buildings. Almost 90% of the examined municipalities have on average biweekly
collection of MSW. Only few municipalities have less frequent MSW collection. These are
typically very small municipalities where it is not economical for the W M company to
provide more frequents collection due to the capacity reasons. On the other hand, around 10%
of municipalities have average collection frequency under two weeks, usually due to the
apartment buildings that require more frequent collection of 1-2 times per week, as the
capacity of the provided shared bins is limited.
77
Waste fee payment development. Most common way to finance M S W M is to collect annual
fees from the municipal population. Fee level may vary up to 1000 CZK per capita, as was
mentioned in the chapter 2. Majority of the municipalities keep their fee level constant for
multiple years, only few make regular adjustments based on the MSWE development or other
factors. Most of the municipalities also did not take immediate advantage of increased
maximum fee level in 2011 and made mostly only a marginal adjustment to the previous fee
level. Besides 5, all examined municipalities use annual fee per capita. Remaining few
municipalities charge individual households or houses, or individual waste bins.
Approximately 60% of the municipalities charged 450-500 C Z K annually per capita,
respectively 90% charged 400-600 CZK. There were also few municipalities that did not
charge any W M fees and cover MSWE from other sources, but in the Czech conditions this is
quite rare. From the other perspective, around 25% of municipalities were able to practically
fully cover MSWE from collected waste fees, 20% had to co-finance 5-20% of MSWE, and
half of the municipalities had to co-finance 30-50% of MSWE from other sources than fees.
Collection options for recyclable waste fractions. This includes availability of HWRC,
drop-off sites, biowaste collection and kerbside collection. Collected information included
presence and amount of such options for waste separation and their development. Higher
availability of waste separation option should usually lead to increased waste separation and
thus reduction in the residual mixed waste generation. This should eventually lead to lower
MSWE. 30% of municipalities in the examined sample had some form of HWRC (opening
hours varied between every day to once in two weeks, larger municipalities often have
multiple HWRCs). Drop-off sites were present in practically every municipalities except 9,
were HWRC or kerbside collection was used instead. Biowaste collection availability in
municipalities increased from 7% in 2009 to more than 60% in 2014, before it became
mandatory for all municipalities in the following year. Around 15% of the municipalities used
kerbside collection for waste separation, which is also often mentioned as a very important
factor increasing overall waste separation.
Presence of some incentive program. This includes presence of any system that directly
influences people's waste related behavior. It can include waste fee discounts based on the
amount of separated waste, provision of additional waste bin only for an extra fee (or just
78
simply making evidence of the amount of existing waste bins in order to check the costs with
the W M company), limited amount of the waste bins calculated based on the amount of
people living in the house and so on. This was relevant only for approximately 3% of the
municipalities in the sample at the time of the data collection, but another 2% were already
planning to introduce some kind of motivation. In the recent years is this topic becoming
more and more popular among the municipalities, as many early adopters show quite
promising results.
Perceived positive and negative aspects of the waste separation, biowaste collection or any
other factors relevant for M S W M from the perspective of the municipal representatives,
including any additional notes and comments from their side. This final group of collected
data represents a very wide group of replies and is described in more detail in the following
subchapter.
Table 4 presents basic statistical information about the interviewed sample of municipalities. I
provide here data only from 2011, but in the selected subchapters later I use data from other
examined years as well. Used databases occasionally did not contain MSW or MSWE values
for several municipalities, and thus these municipalities were not included in the calculation
of given aggregated values in the table. Therefore only in case of population values the full
sample of 535 municipalities was used. I also omitted the largest municipalities in the region
from the further analysis. Based on my experience, M S W M in such municipalities is
significantly different from M S W M in the smaller municipalities that constitute majority of
the sample, and including very large municipalities could create some bias in the results.
Table 4: Basic descriptive statistics about interviewed sample of municipalities, 2011
535 municipalities, 2011 minimum median average maximum
Population 33 557 1239 34 073
MSW generation per capita (kg) 0.1 210.2 218.5 1 373.4
Total waste generation per capita (kg) 0.4 257.1 247.7 1 672.6
MSWE per capita (CZK) 7.68 538.39 587.47 5 299.69
Source: Own construction using data from CZSO, MONITOR and CENIA
79
In the later chapters I typically work with smaller subsamples due to the missing, inconsistent,
or likely incorrect data in some categories. This was done primarily in order to have a more
consistent dataset for the individual analyses that would contain as little as possible incorrect
values. In each subsequent case I provide a detailed description statistics about the subsample
chosen for the specific analysis.
The issue of incorrect data is a part of the wider group of methodological issues connected
with the evaluation of the public finance in general, discussed for instance in Sumpikova et al.
(2005:205). Besides incomplete or incorrect data we can sometimes encounter the issue of
time inconsistency of the data (typically when the methodology for reporting some indicator
changed), changes in the quality aspects that are difficult to capture in the simple numerical
form, time lag in the adoption of certain measures, identification of the effects of selected
analyzed measures and the definition of baseline for comparison (what would happen if the
measure would not be used?), and also the potential alternative use of the finances spent on
adoption of the new measure. Such questions are, however, beyond the scope of this thesis.
2.5 Qualitative findings
This subchapter contains additional data and information collected during the interviews with
the municipality representatives. However, the nature and often uniqueness of the following
data greatly limits them to be used in a quantitative analysis. Nevertheless, I include some of
the most interesting ones and discuss them briefly, as some of them might be inspirational for
some other municipalities.
Presence of a landfill within the municipal cadaster can have notable impact on MSWM.
Such municipalities often greatly benefit from the landfill presence, as they can get access to
the landfill services at a great discount or the W M company could make a deal of securing
M S W M completely free of charge to the municipality in exchange for the possibility to run
the landfill. On top of that, municipality might be able get additional revenues from such
company as well. Logically, such municipalities should not be then simply compared with the
other municipalities, as their position in terms of M S W M is very much different. However,
80
with the ongoing trend in EU towards reducing landfills as much as possible, there should be
eventually less and less such places.
Distance to the nearest WM company facilities can play a role in a way that W M company
in such cases can provide variety of services to the municipality at often very favorable
conditions. This is simply because the W M company can "squeeze" servicing this
municipality between servicing other further municipalities and by doing that better utilize its
available capacities. For instance when the garbage truck goes back to the facility and is not
full yet, it is almost without any additional costs to pick up some more municipal waste from
the close by municipality along the way. Also any occasional demands from the close by
municipality can be dealt with practically immediately due to the minimum distance and
again with the little costs.
Presence of recreational areas in or close to the municipality can affect the costs of M S W M
significantly in case of small municipalities. In larger municipalities this tends to dissolve
among the larger population. But in some cases the proportion of the tourists to the permanent
population might be very high, and thus additional costs arise. While waste produced by the
tourists should actually be treated as a commercial waste by the tourist facilities, it is not
always the case, for instance if the tourists use municipal waste bins, that are then reported as
MSW, and thus affect MSWE.
Large amount of recreational housings compared to the permanently inhabited ones can
also cause issues in MSWM, especially in very small municipalities. For example one
municipality in the sample contains 50 houses with 45 of them being recreational. The
municipality has only one large shared waste bin which would be on average sufficient, but
due to the nature of the dominant recreational housings, waste generation is very irregular,
resulting sometimes with overflow of the waste almost immediately after the scheduled
pickup, while in other cases is the bin empty even until the next scheduled pickup. Waste
generation in such municipalities can depend on factors like holiday, current season or even
weekend weather - basically any factor that attracts people to visit their recreational housings.
In this specific municipality is, based on the observations of municipal representatives, only
81
20% of total annual waste generated in the first half of the year, and from the whole waste
amount only 40% is generated by the permanent residents and 60% by recreational residents.
Time lags after adopting some waste related measures can also have a significant impact
on MSWM. In some municipalities it might take a very long for the new measures to become
fully utilized, while in the other people get accustomed almost immediately. An example
could be biowaste collection. Some municipalities reported significant change in MSW
generation already in the following year after the biowaste collection introduction, while in
others it took 3-4 years to have some notable impact. This of course translates also into
MSWE. Appropriate communication strategy of the new measures and related information
campaigns are often crucial in such cases and can make the difference between successful and
fast adoption and an unnecessary wastage of resources (both in terms of time and money) on
the other hand.
Littering and illegal dumping is often an issue in many small transiting municipalities,
including also those on the state borders. If the neighboring country imposes high fees on the
generated MSW, those foreign people might take the risk and dump their waste illegally
abroad. While not something done in a larger scale, for those few small municipalities it can
cause a notable pressure on the budget, as dealing with illegal dumping is usually quite costly.
A similar problem with illegal waste dumping in the form of littering can happen in
municipalities with some important transport corridor in their cadaster. Trespassers can
simply stop by along the road and produce some waste, leaving it on the spot. And if others
see it, it often does not take a long time before small illegal dumps emerge. Another example
of the impact of trespassing people was from municipality that had installed 2 large shared
containers in addition to the regular individual waste bins. These containers were installed
next to the main road crossing in the municipality. It always took relatively short time before
these containers were filled up, primarily by the trespassers. Municipality then had to
frequently order pickups from the W M company for the additional costs. But once the
municipality moved these 2 containers to a less visible place for the trespassers (but still
easily accessible for the locals), total waste generation of the municipality decreased by
notable 35%.
82
Municipalities with new housings can face the problem that incomers already require W M
services but do not pay the full fees as they do not have permanent residency there.
Analogous problem occurs in the largest municipalities where people come to work or study,
but again do not have a permanent residency there. MSWE in such municipalities are then
higher, while the collected fees do not increase accordingly.
Geographical conditions can have an impact as well. A remote municipality has often very
few choices of W M companies and usually has to accept otherwise less favorable offers in
order to have at least some W M services provided. Rough terrain can also become an issue
for more efficient waste collection using larger trucks. In winter it might even prohibit waste
collection from the remote parts or municipalities at all and several such municipalities thus
provide only centralized bins for their inhabitants.
Gasification of the municipality can cause a great decrease of the generated MSW, once
people no longer use fossil fuels for heating and discard subsequent burned waste together
with mixed residual waste. For this reason, in the municipalities without gasification one can
often identify notably increased waste generation during the heating season in winter
(especially if the winter is colder), while in majority of the municipalities is the waste
generation during winter typically lower, with the reason being little to no green waste
generation during this period.
The way how the WM company counts the collected waste can have impact on the final
MSWE. This is usually not an issue in larger municipalities that can fill up whole or multiple
garbage trucks, but it might become important for small municipalities from which the truck
collects waste in a sequence. If there is no scale installed on the truck, it might be unfavorable
for some while favorable for other municipalities, as the amount of waste assigned to
individual municipalities depends on the judgement of the W M employer or the truck
operator. Especially in very small municipalities this can cause a notable distortion of the
data.
83
Anonymity in larger municipalities, especially among people living in apartment buildings,
can lead to the worse waste separation and more mess around waste bins, which requires
additional costs from the municipality. In smaller municipalities such inferior behavior is
much less frequent.
A municipality can try to sell separated waste on its own at the market and earn additional
revenue this way. In one case a municipality rented a machine for construction waste
processing and then sold the product on the market. Costs per processing one ton of such
waste were 100, while selling price was 500, showing that proper M S W M can provide the
municipality with the opportunities for additional revenues and not include just the costs.
A municipality can provide some parts of MSWM using its own capacities, for instance
using municipal employees or unemployed citizens.
In some municipalities some people might still prefer burning their waste or dumping it
illegally in the woods instead of discarding it to the designated waste bins, even if they have
already paid their waste fee. But according to several municipal representatives, sometimes
people just do not want to give up their old habits. One municipality dealt with such illegal
dumping by installing warning signs in the usual waste dumping areas saying that these areas
are under surveillance. Even if there was no camera installed, problems with illegal dumping
soon practically disappeared.
A municipality can have an agreement with the neighboring municipality that its people
can use drop-off sites and HWRC in that neighboring municipality free of charge. This might
result in distorted observations, as one municipality shows less separated waste while the
other shows more. An even larger distortion can happen in municipalities with few permanent
residents and a lot of recreational housings. Per capita residual and separated generation of
waste calculated using the number of permanent residents produces logically very high values
in such cases.
84
Personal involvement of the municipal representative (usually mayor or his deputy) can
make a big difference in MSWM, especially if this person is willing to try new progressive
things in W M and is willing to devote sufficient time into communicating the benefits of the
new measures to the citizens. Professional education of the municipal representative in the
W M field can improve this even more.
Biowaste collection has great potential in some municipalities. There were reports that after
introducing biowaste collection, MSW generation decreased by up to 50% together with the
MSWE decrease by 40%. Presence of local farmers willing to utilize collected biowaste can
result in additional revenues for the municipality, while lowering the costs for the biowaste
treatment and collection, as the farmers can take care of this waste by themselves.
The quality of M S W M services provided by the W M company can vary notably and simple
fact that one company is more expensive might not mean that it is actually worse for the
municipality. Services from a more expensive W M company can, for example, include higher
frequency of waste collection, more flexible reactions should some waste overstocks happen,
better response time if some other related issues occur, provision of additional extra services,
etc.
And finally, probably the most important thing affecting W M in individual municipalities is
the behavior of the people. Persuading people to cooperate correctly with the current waste
system in the municipality and to follow the given rules represents the greatest opportunities
in improving the performance of MSWM.
85
86
3 EFFICIENCY EVALUATION IN THE MUNICIPAL SOLID
WASTE MANAGEMENT7
This chapter contains efficiency analysis in order to identify the potential factors of efficiency
that are examined more thoroughly in the following chapter 4.
3.1 Sample description
I conduct this efficiency analysis on a sample of municipalities from South Moravian Region
in the Czech Republic. Municipalities in the Czech Republic are rather scattered with a lot of
small independent municipalities. This has an advantage of a possibility to acquire a large set
of data, but it has also a drawback that the data gathering becomes often difficult and time
consuming, unless the data is centrally provided. South Moravian Region consists of almost
700 municipalities with total population of almost 1.2 million with regional capital Brno
(population 380 000). Besides Brno there are no other larger municipalities with second
largest Břeclav having population of 60 000. Nevertheless, due to the significant difference in
the size compared to the rest of the sample, I exclude these largest municipalities. I have
collected general data regarding population and housing structure from CZSO, data regarding
municipal expenditure were collected from MF CR, and data regarding the generation of
municipal waste were acquired from CENIA. These characteristics are often used in the
researches regarding M S W M efficiency. For additional data about MSWM-related
infrastructure (in many studies identified as a very important factors affecting M S W M
efficiency) I have contacted majority of the municipalities in the region by either e-mail or
telephone call. However, some municipalities provided insufficient data, and thus were
removed from the sample. Finally I have excluded several municipalities for which I have
acquired obviously incorrect data from either MF CR or CENIA (typically extremely low
amounts of expenditure or generated waste caused most likely by incorrect data reporting).
After these adjustments I ended up with sufficient data for a sample of 463 municipalities.
Table 5 shows that majority of municipalities in the sample are relatively small, which is
common in the Czech Republic. Almost three quarters of municipalities in the sample have
population of less than 1000 with slightly more than 300 separate dwellings. I include share of
7
This chapter includes parts from Struk (2015a)
87
the condominium dwellings on the total amount of dwellings, as this has been suggested by
the municipal representatives as one of the factors having an impact on the overall state of
MSWM. In addition I include the ratio of dwellings used for recreational purposes to the
regular dwellings, as this has been also suggested as another important factor in MSWM,
especially if this ratio is high (I do not provide a share from the total dwellings, as recreational
dwellings generally do not count towards regular dwellings, and therefore it is possible for a
municipality to have more recreational dwellings than total regular dwellings). Further
descriptive statistics about the selected data of the sample follow. Individual columns show
data for sorted municipalities at each quarter of the sample plus the average value.
Table 5: Descriptive statistics of the analyzed sample of 463 municipalities by quartiles
Q0 Q l Q2 Q3 Q4 Average
Population 36 305 589 1036 34 073 1083
Dwellings 14 109 198.5 339 13 813 381
Condominium share 0.0% 0.0% 4.9% 10.2% 89.7% 8.2%
Recreational ratio 0.0% 1.6% 5.3% 13.2% 215% 10.2%
Built-up area (ha) 1.5 7.2 11.7 18.3 281.8 16.93
Density (per ha) 8.9 40.0 51.4 64.5 202.8 53.6
Drop-off sites 0 2 3 5 145 4.9
Population/drop-off site 0 130 195 287 1412 234
MSWE (000 CZK) 21 165 319 587 34 799 660
Separated waste (t) 0 19 46 110 5179 135
Separated waste share 0.0% 20.4% 29.4% 38.7% 70.1% 29.8%
Source: CZSO, MONITOR, ISOH, own research
Considering the data about the generated municipal waste, I note that as separated waste I
take into account the standard types of wastes that are usually collected at drop-off sites and
HWRCs: paper and cardboard, plastics, glass, WEEE, metals, biodegradable waste (typically
garden waste), and bulky waste8
. Together with MSW I consider all this waste as the total
municipal waste. Commercial waste was not taken into account here, as this research is aimed
solely on municipal waste. Also, as was mentioned in the beginning, according to the Czech
law, commercial subjects are responsible for dealing with their own waste, and thus
municipality is not involved in this process and should not have any related expenditures.
Besides characteristics included in Table 5 I have collected data about the MSW collection
I exclude hazardous and remaining minor types of wastes generated in the municipalities, as these are not
regularly collected by municipalities or in all HWRCs. However, this should not matter much as these kinds of
wastes attribute for a very small fraction of total municipal waste, thus creating rather insignificant distortion in
the data.
88
frequency. In few municipalities there are instead of regular bins few large containers in the
center of the municipality (these are usually very small municipalities with geographical
situation that does not allow for a regular door-to-door collection). In such cases is the
container emptied when it is full, what happens on average once every 2-3 months. On the
opposite side are usually the largest municipalities that have weekly collection of MSW, and
in cases of areas consisting mainly of apartment buildings with larger bins for multiple
dwellings is the collection frequency even twice or more per week. On average, the collection
frequency of MSW in the sample is 2.2 times per month.
Next I have collected information about HWRC opening days per week. Slightly over one
quarter of municipalities in the sample have an HWRC (including also those unofficial ones,
but in fact offering roughly standard HWRC services). Opening days per week span from 0.25
(open once per month) in the very small municipalities up to 6 times per week in the largest
municipalities (where there are usually multiple HWRCs as well). The average amount of
opening days of HWRC (if at least some variation of HWRC is present) is 2.2 times per week.
Additionally I have collected information about several characteristics that generally have
impact on the efficiency of M S W M and can be represented in the binary form. These include
information whether people in the municipality have an option to dispose of their waste in a
HWRC free of charge (either HWRC within their municipality or some neighbor municipality
that agrees with such disposal, usually as a part of an intermunicipal cooperation),
information whether municipality makes some kind of waste bin evidence (usually by issuing
yearly stickers, stamps or barcode stickers), information about the presence of some incentive
program (typically providing discounts based on the amount of separated waste by individual
dwellings participating in the program), information whether municipality provides some kind
of kerbside collection of recyclables, and information whether some kind of collection of
green garden waste was available to the people in the municipality.
3.2 Data Envelopment Analysis
Data Envelopment Analysis (DEA) is, according to Ramanathan (2003:25), a methodology
based upon an interesting application of linear programming. It was originally developed for
89
the performance measurement and has been successfully employed for assessing the relative
performance of a set of firms (or alternative units) that use a variety of identical inputs to
produce a variety of identical outputs.
DEA utilizes techniques such as mathematical programming which can handle large numbers
of variables and relations (constrains) and this relaxes the requirements that are often
encountered when one is limited to choosing only few inputs and outputs because the
techniques employed will otherwise encounter difficulties. Relaxing conditions on the number
of candidates to be used in calculating the desired evaluation measures makes it easier to deal
with complex problems and to deal with other considerations that are likely to be confronted
in many managerial and social policy contexts (Cooper et al., 2006:2).
And finally the last two definitions: DEA is a nonparametric, mathematical programmingbased
technique for efficiency measurement (Thanassoulis et al., 2008:251), and DEA is a
nonparametric method of measuring the efficiency of a decision-making unit (DMU) such as
firm or a public sector agency (Ray, 2004:1). With this basic idea of what is DEA about we
can move to its historical background and principles.
History
Building upon the foundations set by Koopmans (1951), Debreu (1951) and Farrell (1957), a
technique that later became known as Data Envelopment Analysis was introduced into the
Operations Research field by Charnes, Cooper and Rhodes in 1978. Their original CCR
model (named after their surnames) was applicable to technologies with constant returns to
scale (CRS), which was somewhat limiting. However, after few years a major breakthrough
occurred, as Banker, Charnes and Cooper (1984) extended their original model into BCC
model which was able to accommodate technologies with variable returns to scale (VRS)
(Ray, 2004:1).
In the following years DEA usage experienced a boom as it became a valid alternative to
regression analysis for the efficiency measures. Relevant literature on either theoretical
90
aspects of DEA, as well as various practical applications, began to grow exponentially in
1990s and 2000s. Popularity of the DEA technique grew along with the development of
software tools for its solving. As DEA consists of solving linear programming problems, with
increasing numbers of DMUs and considered inputs and outputs, the actual computing
became very time-consuming while requiring a lot of precise procedures. But with the
appropriate software, solving DEA problems became much less demanding in terms of
computation, and thus making this method very popular. A list of the programs solving DEA
problems is provided later.
The popularity of DEA can be illustrated by the amount of the published literature dealing
with the topic. For instance Ray (2004:1) states than in 1992 there was almost 500
publications dealing somehow with DEA, while in 2002 there was already more than 3 000
DEA related publications. According to the search in Web of Science, in 2012 there was
almost 6 000 references, while Google Scholar returned more than 40 000 items when
searching for Data Envelopment Analysis. Recently in 2018 Google Scholar returned over
260 000 results, illustrating the popularity of this method, even though notable portion of the
results might be just mentioning DEA in the text and not be directly dealing with DEA. Table
6 shows development of DEA papers in WOS and Scopus databases in 2004-17 period.
Table 6: Published DEA-related papers indexed in Web of Science and Scopus, 2004-17
o
LO
o
ID
O
1--.
o
oo
o
cn
o o <H
<H
r\l
<H
no
<H
LO
<H
ID
<H <H
o o o o o o o O O O o o o Or\l r\l r\l r\l r\l r\l r\l r\l r\l r\l
WOS 240 272 313 372 480 690 691 787 780 838 962 1165 1202 1271
Scopus 298 353 429 476 634 952 985 1045 1066 1165 1214 1093 1204 1357
Source: own construction using WOS and Scopus
Very good explanation of the DEA history prior to its " reanimatiorT by Charnes, Cooper and
Rhodes (1978) is available from Forsund and Sarafoglou (2002) in their study "On the
Origins of Data Envelopment Analysis". This study covers thirty years of the related
development in the field of economics since Farrell's paper (and accompanying discussion) in
1957, and elaborates the connection between the two.
91
Next thirty years of DEA development after the CCR publication in 1978 is described by
Cook and Seiford (2009). Their paper focuses on the methodological developments while
omitting the practical applications that appeared during the period.
Principles
The basic idea behind DEA is the creation of an " efficiency frontier" by the units with the best
output-input ratios9
. According to Ramanathan (2003:25), "DEA is a linear programmingbased
technique for measuring the performance efficiency of organizational units which are
termed Decision-Making Units (DMUs). This technique aims to measure how efficiently a
DMU uses the resources available to generate a set of outputs". The basic efficiency measure
of DEA is the ratio of total outputs to total inputs. Important note is the fact that efficiency is
evaluated only within the units from analyzed set and thus results cannot be generalized for
the units outside the set. Figure 10 illustrates a simple DEA example with both constant
returns to scale and variable returns to scale.
Figure 10: Example of CRS and VRS efficiency frontier, case with 7 DMUs, 1 output, 1 input
25
CRS frontier /
20
___©VRS
frontier
J 15
•M
Q . 10
• x
0'
/ e o
y f o
o ° 8
d
c
3
o
b
/
5 /
/
/
/
/
0 r0
0 50 100 150 200
Inputs - Money
Source: own construction
Figure 10 provides a situation with 7 DMUs (marked "a-g"), with each of them consuming
certain amount of inputs - money, and producing certain amount of outputs - goods. D M U
9
Such methods as DEA or SFA are sometimes called .frontier" or "distance" methods.
92
"e" is clearly the most efficient yielding highest output to input ratio. Together with point
"0,0" it forms constant returns to scale (CRS) efficiency frontier while points "0,0", "e", "/'
and "a" form variable returns to scale (VRS) efficiency frontier.10
In this example (one input,
one output) the visualization in simple, but with more inputs and outputs are the results of the
models usually presented only in numerical form.
DEA method, through linear programming, provides a way of how multiple inputs and
outputs can be "stacked" together in order to create one total output and one total input
(through assorting appropriate weights for each partial input and output), so that the relevant
efficiency ratios can be computed for each DMU. After that, the "most" efficient units
(usually called peers) are identified and get efficiency score of 1 (unity). Other DMUs that do
not lie on the efficiency frontier get efficiency score other than 1 depending on their distance
from the closest efficiency frontier constituted by one or two peers.11
These different scores
tell how the inputs or alternatively outputs (depending on the DEA model specifications) need
to change in order to make inefficient D M U efficient (so that its position moves to the
efficiency frontier) and are called "slacks".
When calculating DMUs efficiency scores, ^-values are used. Together with the efficiency
score they provide information about the actual position of examined D M U in respect to its
peers. Based on the linear programming, ^-values tell to what extent was the peer used in
order to construct a vector on which examined D M U lies. And the efficiency score simply
tells where on this vector the D M U is situated with 1 meaning it is right on the efficiency
frontier formed by the peers and values less than 1 meaning the distance from this efficiency
frontier.
Here can be seen the analogy with the concept of production possibility frontier from microeconomic theory area
under the efficiency frontier is the production possibility set, however, only results on the frontier are
considered efficient. D E A can be, in certain aspect, seen as a "reverse process" of production function analysis.
1 1
An inefficient D M U can get efficiency score either above or below one, depending on what exact type of
model is computed. However, scores are reciprocal and thus tell the same information about the distance from
the frontier.
93
Formally, an input-oriented measure of technical efficiency of the D M U t in the sample of N
DMUs with input vectors Xj and output vectors y;-, (J = 1, ...,N) is then obtained from the
following linear program:
min 6 ,s.t.
N
Xj Xj < 8xt,
J'=I
N
:y, > yt,
; ' = i
N
7=1
Ay > 0,7 = 1 JV
Then we can define x*t = Y/j=1Xj XJ = 6*xt for the optimal value of objective function 6* and
optimal multipliers XJ, j = 1, ...,N. Then (xl.y^) is the efficient input-oriented projection of
(xtyt) and its technical efficiency is defined by eff(xtyt) = 6* (Struk and Matulová, 2016).
Forsund (2018) provides an accessible economic interpretation of DEA. Further explanation
of mathematical programming aspects of DEA together with the theory behind DEA can be
found in D E A textbooks, for example Ramanathan (2003) or in Czech in Jablonský and
Dlouhý (2004, 2015), and also in more comprehensive books like Cooper et al., (2006), Fried
et al., (2008), or Ray (2004). This thesis does not intend to describe the computational aspects
of DEA and for such the reader should refer to the provided literature.
Intuitively, as was already mentioned, with more inputs and outputs, as well as amount of
DMUs, the computation of DEA results becomes increasingly time-demanding. However, the
results of DEA are in practice usually calculated using some available DEA-solving programs
or spreadsheet extensions and thus the researcher performing efficiency evaluation is not
limited by these computational aspects. A good list of available software for solving DEA
94
problems can be found at webpages www.gapem.org/default.php ?ref=papers and
www.opendea.de/verfuegbaredea.php.12
Selected DEA solving programs are listed in Table 4.
Although presented list in the Table 7 contains many options, it is necessary to note that some
of the programs are commercial and require a purchase in order to be utilized. On the other
hand, there are still free of charge opportunities for DEA problems solving, especially if
problems do not contain large datasets (as some of the commercial programs offer free demo
versions with limited number of DMUs, inputs and outputs).
Table 7: List of DEA solving programs
Program Author
DEA-Excel Jablonský, 2008
Deafrontier Zhu, 2003
D.E.A.O.S. Int. Institute of Operations
DEAP Coelli, 1996
DEAQual Kamakura
DEA Solver Online Uni. Hohenheim
DEA Solver Pro Saitech
EMS Uni. Dortmund
FEAR Wilson, 2008
Frontier Analyst Banxia
GAMS/DEA Gams
iDEAs Keck Virtual Factory Lab
ISYDS (SIAD) Angula-Meza et al., 2005
KonSi DEA Analysis Konsi
LIMDEP Limdep
MDeap Deslierres
myDEA Wilken, Harren
OnFront Emq
PIM-DEAsoft Deasoftware.co.uk
Qhull Uni. Minnesota
xlDEA Prodtools
Warwick DEA software Uni. Warwick
Win4DEAP Deslierres
Source: gapem.org, opendea.de, nb.vse.cz/~jablon/
When making actual calculation, the only necessary thing with available programs is to have
the data. Some of the programs require specific format of the data, while other allow direct
Both sites are unavailable at the time of the writing of this thesis. Fortunately, they can be still accessed using
Wayback Machine at web.archive.org/web/20101219151559/http://www. gapem.org/default.php?ref=papers, and
web.archive.org/web/20160403085152/http://opendea.de/index.php/verfuegbaredea
95
input of the data. After that there are usually some options regarding the formulation of the
desired DEA model, nevertheless, even without the deeper knowledge of the DEA theory one
can choose basic ones and acquire results that should be relatively easy to interpret. Some of
the programs would even produce a graphic output and some other additional information
concerning examined DMUs and the dataset.
Pros and Cons
DEA has several advantages as well as some disadvantages. The most notable advantage is
probably the ability to handle multiple inputs and outputs that can be expressed in various
means. This becomes especially useful when dealing with decisions in a complex
environment. Among other advantages is, for example, no need for a specified production
function, as DEA "constructs" the production frontier from the available data, DEA's ability
to uncover relationships that might not be clear to see, or the fact that one can analyze and
quantify the sources of inefficiency for every unit from the set. Another advantage implied by
the previous one is the unnecessary presence of money, which can be very useful when
conducting analyses in the public sector where unmonetized inputs and outputs are often
likely to occur.
However, DEA has limitations too. The main of them is the choice of the inputs and outputs
that enter the evaluation (as some important ones can be omitted and thus the analysis would
yield incorrect results). Also the choice of the evaluated units is an important factor, as DEA
evaluates only the units within the analyzed set and the results cannot be simply generalized
for the whole population. A single one new unit can notably change the efficiency of the rest
of the sample. On the other hand, this is more likely to be a defining feature that emerges
from the principles of this method rather than a disadvantage in its exact nature. Also units
lying outside the cone formed by efficiency units can be sometimes still considered as
efficient. More about this can be found in already mentioned textbooks on the DEA topic like
Ramanathan (2003), or Cooper et al., (2006).
Moreover, as DEA is a nonparametric method, there is no production, cost or profit function
estimated from the data. This leaves a rather very limited possibility for evaluating marginal
96
products, partial elasticities, marginal costs, or elasticities of substitution from a fitted model.
This prevents the usual conclusions about the technology, which is possible from parametric
functional form (Ray, 2004:2).
And finally, as DEA is of nonstatistical nature, the solutions to the problems by linear
programming produce no standard errors and leave no room for hypothesis testing. With DEA
any deviation from the calculated efficiency frontier is treated as inefficiency, rejecting any
random shocks (Ray, 2004:2). Lack of standard errors in DEA efficiency measures stems
from the fact that stochastic properties of inequality-constrained estimators are not well
established in the econometric literature. Even with simple DEA problems this becomes a
very complex issue, and, according to Ray (2004:5), the solution is unlikely to arise in near
future, although there are several lines of research addressing this problem.
Fields of Research
When it comes to the bibliography, there are several comprehensive studies that list the
development of publications concerning DEA right from the very beginnings.
Emrouznejad's web page dedicated to DEA, http://deazone.com, can serve as a good source
for papers containing practical DEA results and discussing its various aspects. Besides
provided references to several thousands of DEA studies, Emrouznejad et al. (2008) had
published selected statistics about the DEA literature covering roughly 30 years after the
initial CCR study, together with the recently updated study of first 40 years of scholarly
literature in DEA (Emrouznejad and Yang, 2018).
Another very comprehensive study comes from Gattoufi et al. (2004b), who published a 70pages
long list of DEA-related studies, including 13 studies prior to the CCR publication in
1978 that are, according to Forsund and Sarafoglou (2002), related to the DEA concept. This
list contains more than 1 800 studies, dissertations, proceedings or book chapters up to 2001.
Along with that, Gattoufi et al. (2004a) have published another study about the "taxonomyfor
data envelopment analysis", where they present a scheme for classifying DEA literature
97
according to several criteria. Just to illustrate the popularity of DEA, Seiford in his "CyberBibliography"
from 2005 lists 2800 DEA-related studies, which means about 1000
publications increase in four years when compared with Gattoufi et al. (2004b).
When it comes to the real-world areas of DEA focus, the ideal applications include areas
where there are many DMUs that use comparable inputs and produce comparable outputs. For
instance, some notable publications in the field of environment are Coli et al., (2010) who
have evaluated environmental efficiency of Italian provinces; Honma (2014) who evaluated
eco-efficiency of 31 Asia-Pacific countries and regions; Korhonen and Luptacik (2004) who
evaluated eco-efficiency of power plants; or Kuosmanen and Kortelainen (2005) who
evaluated eco-efficiency of road transportation in selected Finnish towns. In the other areas
related to this thesis, public services, selected notable publications are Afonso and Fernandes
(2008) who created comprehensive comparison of local government efficiency using the data
for Portuguese municipalities; Kirkpatrick et al., (2006) who used DEA to compare efficiency
of state and private sector provision of water services in Africa; or Woodbury and Dollery
(2004) who examined municipal water services in Australia. Here, one can easily see that
DEA is especially popular for analyzing network industries. This list is by no means
exhaustive and rather reflects the research focus of the thesis. Numerous other applications of
DEA can be found in banking, seaports or agriculture sectors.
3.3 DEA evaluation of efficiency
I use Two-stage DEA (2SDEA) for the efficiency analysis, first introduced by Ray (1988).
2SDEA represents an extension to the standard DEA by adding a regression analysis of the
computed efficiency scores of units on other relevant variables. Adding a regression in the
second stage also overcomes one of the pitfalls in the DEA methodology discussed by Dyson
et al. (2001) - an assumption that all units within the sample operate in the same environment.
However, some limitations of this methodology still exist, as noted by Simar and Wilson
(2011). Nevertheless, the primary advantage of employing 2SDEA is that it can both point out
which units perform the best and suggest what might cause it. Depending on the nature of
such identified factors, other units might be able to adopt them in order to increase their own
efficiency. Hoff (2007) provides further discussion of approaches used in the second stage of
98
DEA. Liu et al. (2013) conclude that two-step DEA became a prevailing approach in recent
years, what is in my opinion natural, as knowing what makes units efficient is in my
perspective even more important than just knowing which of them are efficient. 2SDEA has
already been used for measuring efficiency in the sector of municipal waste management, for
instance by Garcia-Sanchez (2008) in Spain, Marques and Simoes (2009) in Portugal, or De
Jaeger et al. (2011) in Belgium.
First stage of this method consists of calculating the efficiency of decision making units
(municipalities) using DEA. By doing this each municipality gets a score between 0 and 1
depending on the levels of inputs and outputs. In some studies this is the final result as well,
as the efficient units are identified together with the estimation of how much can they
improve in order to become efficient and therefore acquiring the efficiency score of 1. With
2SDEA I extend this and use the calculated efficiency scores in a subsequent regression
analysis with various characteristics of M S W M in order to explore the possible effects of
these additional characteristics on M S W M efficiency.
I provide two models regarding the DEA calculation. Here I simplify the whole process of
M S W M into a black box where on one side we insert the amount of MSWE (inputs) and on
the other side we get certain amount of served units of population or dwelling (outputs). It is
possible to consider this process the other way round, but the results would be the same - we
either consider that efficient municipalities are those with more outputs for the expenditure, or
those with less expenditure in order to serve the initially inputted amounts of population,
dwelling, etc. I assume the one and only input to be the amount of MSWE, as I perceive it as
the resource that is inputted into MSWM. While some studies like Garcia-Sanchez (2008) use
multiple inputs like staff, vehicles and containers, in my opinion it can be simplified to the
one input in the form of expenditure. Mentioned inputs can be then purchased and it is up to
the D M U (or in this case waste management service provider) to create a best mix depending
on the perceived marginal productivity of each of such production factors. Moreover, due to
the nature of our sample, I do not expect significant differences between the costs of the
production factors between waste management providers, therefore a common input in the
form of expenditures seems as appropriate. On the other side of the process, or as outputs, I
put the amount of municipal population, amount of dwelling within the municipality and the
99
size of built-up area served by MSWM. I perceive these three parameters as the output of the
whole process of MSWM. Such input-output specification of the DEA model is comparable
with for instance Rogge and De Jaeger (2013).
DEA offers a possibility to choose from various models depending on the exact thing one is
exploring. In this case I use one of the most common models developed by BCC, which
extends the basic CCR model by allowing variable returns to scale. I use this model instead of
the basic CCR one due to the large variance in the size of the municipalities I analyze. As the
size of the municipalities in terms of population spans from several dozen up to several
thousands, I assume that there is not some constant effect with increasing size, but rather
some variability in the scale effect throughout the sample. Furthermore I set the model to be
output oriented, meaning that given the "fixed" amount of expenditure I consider the
municipality to be more efficient if it is able to produce more outputs, in this case to provide
M S W M to more people, dwellings, larger area (efficiency model 1), and additionally produce
more separated waste and have higher share of separated waste to MSW (efficiency model 2).
An alternative interpretation would be that the more efficient units (municipalities) are those
with less expenditure given the "fixed' amounts of outputs.
Therefore the first efficiency model has 1 input (MSWE) and 3 outputs (population, dwellings
and built-up area). Second efficiency model has additional 2 outputs (amount of generated
separated waste and share of separated waste on the total municipal waste), resulting in more
stress towards the factors directly related to the waste separation, when compared to the first
model. In these models I specifically do not include total amount of collected residual waste,
unlike in many studies, as in my opinion in such cases the efficiency could possibly go the
opposite than desired direction - a more efficient municipality would be the one that is
generating more waste for the fixed amount of costs, which I do not perceive as a desired
result, at least from the environmental perspective, even though strictly economically
speaking it makes sense.
Both models have been calculated with OSDEA software. In the first model 8 efficient units
(also called peers, or the units with 100% efficiency) were identified, while in the second
100
model 20 units were identified as efficient (all efficient units from first model were included).
Increased number of efficient units is the usual consequence of adding more inputs/outputs, as
we are dealing with a more dimensional optimization problem. Following Figure 11 shows
the distribution of acquired efficiency scores among the units in the sample.
Figure 11: Frequency distribution of calculated efficiency scores of municipalities
Efficiency score in model 1 • Efficiency score in model 2
70
2- 5 0
'u
§ 40
E
- 30
§ 20
0
1 10
0
V *V s?ef *V *v *V <?
Calculated efficiency score cohorts
Source: own calculations
Regarding the efficient municipalities, they are relatively diverse with no significant features
that only they would have, compared to the rest of the sample. There are both larger and small
municipalities, they have various waste collection frequency, only half of them have HWRC,
one third has adopted kerbside collection, their expenditure per capita are about the same as
the rest of the sample, so there is no clear direct connection between the presence of some
factor and a high level of acquired efficiency score. The only thing they have in common is a
significantly higher than the average generation of separated waste per capita, together with
higher than average share of separated waste on the total generated municipal waste.
Factors influencing efficiency
In the second stage of DEA I used calculated efficiency scores as the dependent variables in
the subsequent regression analysis using simple OLS technique for estimating effects of the
independent variables. As the potential independent variables I chose various characteristics
101
of MSWM, with each of them practically creating own hypothesis regarding the effect of this
factor on the M S W M efficiency. Used characteristics were chosen based on the availability of
the data collected from the municipalities and those available from public databases. The
limitation here is that some important influencing variable might have been omitted, and the
results could thus be biased, but this is a feature of such model and it is up to researcher to try
to specify the model to the best of his knowledge and the availability of the data. I need to
note here that the exact values of the coefficients are not that important as they are related to
the constructed efficiency score. More important is the trend suggested by the calculations
and the significance of the effect compared to the other variables. Results of the calculations
are provided in the Table 8.
Table 8: Effects of selected M S W M characteristics on M S W M efficiency
Effects of selected M S W M characteristics Efficiency model 1 Efficiency model 2
on the efficiency (in %)
MSW collection frequency/month -3 2 4 * * * -3.75**
(1.11) (1.50)
Waste bin evidence 5 95*** 4.21**
(1.41) (1.90)
HWRC open/week 1.22* 1.89**
(0.64) (0.86)
Availability of free of charge HWRC -0.67 -0.81
(1.51) (2.04)
Amount of separation drop-off sites 0.30*** 0.30***
(0.07) (0.10)
Population/Separation drop-off sites 0.03*** 0.01***
(0.00) (0.00)
Kerbside collection of recyclables 3.84** 6.20***
(1.70) (2.29)
Separation incentive program 5.32 10.17*
(4.02) (5.42)
Green waste collection -0.02 -0.73
(1.55) (2.09)
Share of condominium dwellings 6.62 2.22
(6.71) (9.04)
Share of recreational dwellings -9.64** -10.20*
(4.16) (5.61)
Population/hectare of built-up area 0.13**
(0.04) (0.05)
Constant 46.44*** 52.02***
(3.29) (4.44)
No. of observations 463 463
Adj. R2
0.38 0.21
*, **, *** stand for 10%, 5%, 1% statistical significance; standard errors in parentheses
Source: own calculations
102
3.3.1 Identified factors of efficiency
Conducted DEA showed several statistically significant municipal characteristics with respect
to the W M management in the analyzed sample of municipalities. From my perspective they
can be divided into several categories based on their nature:
1) Convenience factors: drop-off sites density and amount, kerbside collection
2) Incentive factors: incentive presence, waste bin evidence
3) Direct costs affecting factors: collection frequency
4) Indirect factors: recreational dwellings share, population density, HWRC opening
hours
From the perspective of this thesis I find the most interesting first two categories, as in my
opinion they offer the largest potential for further analysis and also can be directly influenced
by the municipality. The direct influence can be also seen in case of collection frequency in
the third category, but this relation is in my opinion quite straightforward (higher collection
frequency translates into higher costs), and thus does not require further studying. Fourth
category then contains factor that cannot be easily influenced by municipality and/or are
related also to some other factors. HWRC opening hours for instance tend to correlate with
the size of the municipality - larger one has more frequent opening hours, plus the overall
separation options in a larger municipality a greater, thus the actual identification of relation
between efficiency and HWRC opening hours might be a bit tricky. Moreover in this case we
are dealing with a large heterogeneity in the data, as some HWRC are opened once per month
while other every day, not speaking about the other issues like what actually municipalities
consider as an HWRC. Final two characteristics, share of recreational dwellings and
population density are indirect factors that are typically set and municipality has little
measures how to practically influence them, but they could provide and explanation of certain
relationships and municipality can adjust its W M accordingly.
3.4 Discussion of the findings and limitations
In both models MSW collection frequency per month has negative effect on the efficiency.
Most of the municipalities in our sample have a biweekly collection of MSW, but with higher
share of multifamily dwellings (typically areas with apartment buildings in the larger
103
municipalities), the trend leans toward weekly collection, which decreases M S W M efficiency.
On the other hand, some municipalities have only monthly collection, and in several cases
even less frequent collection periods (typically if a municipality has large central containers
instead of smaller bins for each family house), which also increases their M S W M efficiency.
Positive effect on efficiency with lower frequency makes sense, as less frequent collection
directly transfers into lower MSWE, although the effect, as observed in per capita MSWE of
municipalities with different frequency, is far from linear. Contrary to our results, De Jaeger
et al. (2011) did not find any significant effect of different collection frequency on M S W M
efficiency, although this might be caused by the differences among the sociodemographic
characteristics of Flemish and Czech municipalities (triple the population density, much
flatter terrain, average municipality more than tenfold larger, etc.).
Evidence of the amount of waste bins used in the municipality increases the observed
efficiency of municipalities. The most common form of bin evidence is issuing yearly stamps
or stickers that are physically attached to the bin, and declaring that the corresponding
dwelling had paid its yearly MSW collection fees. An alternative to the stamps is for instance
a sticker with a barcode, or in some very small municipalities the sole visual evidence by the
municipal representative. Despite issuing stamps, in majority of municipalities there is no
limit on the amount of bins, as long as the fees have been paid, so having an evidence system
should not directly lead to the increased efficiency, even though the results suggest that with
any kind of bin evidence present in the municipality. The reason behind this, as suggested by
some municipal representatives, might be the observations that with evidence system people
in the municipality voluntarily start to act more economically and often use rather one fully
stuffed bin instead of two or three partially filled. Additionally, with bin evidence the
municipality is able to control the amount of costs billed by the collection company much
better - some municipalities reported a sudden drop in the amount of billable bins after the
introduction of bin evidence, even if there was no visual difference in the amount of bins after
the introduction of the bin evidence system.
More frequent HWRC opening days have positive effect on the efficiency. This is caused by
the better availability of waste separation options for the people, who are therefore more
likely to separate. If available, HWRCs in municipalities are on average open twice per week,
with opening frequencies spanning between once a month up to six times per week.
104
Availability of a HWRC, which is typically free of charge to the people with permanent stay
in that municipality (with occasional visits from other people being often tolerated), usually
results in both less generated MSW together with higher ratio of separated waste to total
generated municipal waste. Additionally I have examined the effect of the sole availability of
a free of charge HWRC to the municipal population, whether directly in the municipality, or
in some neighbor municipality as a result of intermunicipal cooperation. According to Bel and
Mur (2009), Bel et al. (2014) or Poldnurk (2015), intermunicipal cooperation leads to the
reduction in overall MSWE in municipalities with smaller population. In this case I was
looking only at the availability of an HWRC. According to the results, there were no
significant observed effects on the efficiency, suggesting that the availability of a free of
charge HWRC outside of a municipality does not seem to matter that much - people likely do
not consider this as an option, as it includes travelling to the other municipality.
The most common way of waste separation in the examined sample is separation through
drop-off sites within the municipality, usually called separation "nests". These sites are used
primarily for separation of paper, plastics and glass and, except very few, all municipalities in
our sample have at least one such site. Therefore I do not examine whether the simple
presence has some effect on the efficiency. With increasing average amount of people served
by one drop-off site the efficiency slightly increases, meaning that municipalities with less
frequent drop-off sites are more efficient, what might seem a bit counter-intuitive. Higher
drop-off sites frequency means that they are closer to the people and the proximity, as
concluded by Gonzalez-Torre and Adenso-Diaz (2005) and others, acts as one of the
important determinants in peoples' decision whether to separate their waste or not. The
explanation behind this might be that municipalities from my experience often report
expenditures related to the collection of separated waste from drop-off sites together with
MSWE, and therefore with more drop-off sites the total amount of MSWE increases. Other
explanation might be that in rural and less populated areas (which applies to many
municipalities in our sample) people tend to separate less, as was observed by Poldnurk
(2015). I have also examined the effect of increased density of drop-off sites with respect to
the built-up area, but the results were not significant.
105
Some municipalities have adopted a kerbside collection of recyclables, mostly paper and
plastics, occasionally also beverage containers. Such collection system of recyclables
represents a very convenient way of waste separation for the people, as they can separate
practically right at their household. Then, during scheduled dates, they simply leave the
separated waste in front of their property. The most common type of kerbside collection in the
Czech Republic is through plastic sacks; alternatively some municipalities have plastic
containers. Containers are on one hand more durable, but on the other hand have significantly
higher initial purchasing costs, are less flexible and the manipulation requires more effort - it
has to be emptied right on the street and then taken back to the house, compared to throwing
plastic sack into the truck without requiring any further attention from the people at the
household. Presence of a kerbside collection system increases the efficiency quite notably,
especially considering efficiency model 2, which includes the level of waste separation as one
of the outputs. Such results are consistent with observations of Dahlen and Lagerkvist (2010),
who concluded that municipalities with property-close collection of recyclables (in our case
kerbside collection using plastics sacks) show twice as high separation rates. Folz (1991)
came to the basically same conclusion that compared to drop-off sites, kerbside collection
results in almost twice as high participation of the public in recycling. Kerbside collection
represents a very convenient type of waste separation system for the people, and in several
studies has been identified as the most important determinant of peoples' decision whether to
separate their waste or not, what subsequently affects the overall M S W M efficiency.
In addition to the kerbside collection, some municipalities have adopted an incentive program
for the waste separation, which allow people to get discounts from the waste collection fee
based on the amount of waste they separate or some other benefits. This usually works with
providing a unique identifier to each dwelling as a part of the kerbside collection and then by
weighting the amount of separated waste from this dwelling. With such program adopted,
people should be motivated to separate more, thus reducing the amount of total MSW, and
resulting in the increased M S W M efficiency. However, I observed statistically significant
positive effects of motivation system only in the case of efficiency model taking into account
the waste separation levels, although in this case the effect was quite strong. Significance of
such incentive program is supported by Hornik et al. (1995), who conclude that monetary
106
incentives work well in encouraging waste separation behavior. Additionally, Schultz et al.
(1995) observed that using a lottery instead of fixed discounts results in even more separation.
Separate collection of biologically degradable waste should naturally decrease the amount of
MSW and increase efficiency in the first model. Additionally, presence of such separation
system should have even higher effect in the efficiency model 2, where the amount of
separated waste is also considered. However, in the regression analysis this assumption
proved insignificant. The reason behind this, suggested by several municipal representatives,
is that majority of the municipalities in our sample consist primarily of single family houses
where people are historically used to utilize their own green waste, resulting in the underuse
of available options for green waste separation. On the other hand, other municipal
representatives reported that, after some time, people often get accustomed to the new
possibility of green waste separation, and previously underused capacities suddenly cannot
keep up with the increasing volume of the green waste. With respect to that, if I take into
account that in 2012 (the year from which I have collected the data for this analysis) the
separation of green waste was something rather new for the majority of the municipalities (if
they have adopted something like that at all), we might simply face the fact that it was too
early to examine the effect of such factor on the M S W M efficiency.
According to the municipal representatives, M S W M in areas where majority of dwellings are
in multifamily apartment buildings is usually considered as more problematic. People in such
areas are more anonymous, usually share one large waste bin for multiple dwellings, and
generally are less motivated to care about their waste-related behavior. Lebersorger and Beigl
(2011) have observed such negative effect of increased percentage of condominium-type of
dwellings on MSW generation, supporting the suggestions of the municipal representatives.
Negative effect of a higher share of apartment buildings applies to the observed lower rates of
waste separation as well. Ando and Gosselin (2005) confirmed this in a survey, according to
which people living in condominiums reported lower waste separation rates than people living
in single family houses. In contrast to the suggestion of municipal representatives, the main
reason identified behind this behavior was not that people from condominiums have less of a
sense of owning their waste stream, and thus caring less about their waste-related behavior,
107
but rather less convenience for such people with respect to the waste separation. Nevertheless,
in our analysis this effect has not been confirmed, and results were insignificant.
The amount of dwellings used for recreation has an effect on M S W M efficiency as well. In
municipalities with large recreational areas municipal representatives often said that such
areas are problematic due to the seasonality in their waste generation (people use such
dwelling mainly during summer and during the weekends), and that again such people care
rather little about the state of their surroundings, if they visit these places only occasionally.
Our results show that higher share of recreational dwellings in the municipality has a negative
effect on M S W M efficiency. Observed effect may seem marginal, but in case of several
municipalities with more recreational than regular dwellings this can have a relatively large
impact on the overall efficiency. Increasing effect of higher share of recreational dwellings
(called sometimes secondary residences) on MSW generation, leading subsequently to the
decreased efficiency, has been also observed by Lebersorger and Beigl (2011).
The last factor that was examined is the density of population with respect to the built-up area
of the municipality (the area where houses and related infrastructure are built, excluding
lakes, fields, forests, etc. that are included in the total municipal area). Here I have observed
positive effect of increased density on the efficiency. Although the results may again appear
small and marginal, we need to consider that the population density is usually in the range of
several hundred people per hectare, resulting in relatively strong overall effect. Positive effect
of increasing density has been also observed by Benito-Lopez et al. (2011) or Dubin and
Navarro (1998), but only in municipalities with population under 20,000. In this sample there
are only few municipalities larger than that, what might also contribute to the significance of
the density factor. Such results therefore suggest presence of some economies of scale in
more densely populated municipalities, but as was mentioned, this might hold true only until
certain size of the population. To support that, Worthington and Dollery (2001) conclude that
the efficiency of M S W M provision notably decreases with density, which confirms previous
results, as in their case they were examining M S W M of rather large metropolitan areas.
108
In this chapter I had explored the potential factors influencing M S W M efficiency. As an
appropriate and common method for efficiency evaluation I used two-stage DEA. In the first
stage I have calculated the efficiency in models using MSWE as the input, while using the
municipal population, built-up area and the amount of dwellings as the outputs, resp. adding
the amount of separated waste and the ratio of separated waste to total municipal waste in the
second efficiency model, further stressing the environmental aspect of analyzed efficiency.
The sample used in the research consists of 463 municipalities in the South Moravia Region
in the Czech Republic and the data were collected with respect to year 2012.
After calculating the M S W M efficiency scores I have subsequently applied OLS regression
analysis using various M S W M characteristics collected from the interviews with municipal
representatives. I have identified positive effects on the efficiency in case of lower MSW
collection frequency, presence of waste bin evidence system, more frequent opening days of
an HWRC, availability of a kerbside collection of separated waste, an incentive program for
the waste separation, and higher population density. On the other hand I have identified a
negative effect of increased density of drop-off sites, increased share of apartment-type
dwellings on total dwellings, and increased ratio of recreational dwellings to regular
dwellings. A l l of these observations, except negative effect of drop-off site density, are
consistent with previous researches in this area. Contrary to the assumption, the availability of
green waste separation did not prove to have a significant effect on M S W M efficiency, which
might be caused by it simply being a relatively new separation option to the majority of
municipalities at the time of the data collection. Compared to the previous researches in this
area, I have focused on a broader range of factors usually connected with M S W M efficiency
and provided results regarding their effects on M S W M efficiency.
The contribution of the conducted analysis is two-fold. First, by using proposed two-stage
DEA one is able to identify efficient units in the first stage. Subsequently, in the second stage
of analysis one can identify which specific factors related to M S W M have a statistically
significant contribution to the efficiency and whether this contribution is positive or negative.
In case of factors that can be influenced or adopted by the municipalities, such knowledge can
be utilized by less efficient municipalities for improving their efficiency in this area.
109
Conducted analysis has some drawbacks too. Data from multiple years could provide a better
understanding of the issue, but some of them simply are not available. Central institutions that
collect them do not have an obligation or the will to share them, what limits further analysis
of newer data. Collecting data directly from the municipalities can be demanding as well.
Increased range of the municipal characteristics could have been used in the analysis, but this
is again related to the availability of the data and substantial time requirements in order to get
them. In some cases the data might be incorrect as well, whether due to the mistake while
reporting them by the municipal representatives or while processed at the centralized
database. In such cases municipality might become a peer even if the reality is different and
subsequently influence efficiency scores of other DMUs in the sample.
Used efficiency evaluation method - DEA, has its limitations. First of all, it unsurprisingly
takes into account only municipalities included in the sample. Therefore the results cannot be
simply generalized and are valid only for the municipalities included in the analysis, as there
is always a possibility that there is also some other municipality with some other mix of
characteristics that, if included in the analysis, would become a peer and thus would affect the
results. But with sufficient sample that covers majority of the municipalities in the examined
area, the likelihood of existence of additional municipalities that could notably change the
calculated results becomes slim.
Inclusion of qualitative findings might add further explanation to the factors influencing
efficiency of MSWM. But such large variability of these factors among the examined sample
of municipalities makes it difficult to incorporate them directly into the analysis.
Nevertheless, I at least included these qualitative findings in the previous chapter.
110
4 EFFICIENCY FACTORS IN THE SOUTH MORAVIAN
MUNICIPAL SOLID WASTE MANAGEMENT
The following chapter examines in more detail selected factors of efficiency identified either
in the relevant literature or during the quantitative analysis of the collected data. Based on the
conducted DEA evaluation I chose to examine in more detail factors related to the
convenience in WM, specifically presence of kerbside collection and the density of drop-off
sites, and the presence on an incentive program related to municipal waste separation. Based
on the literature review I have chosen to examine more in detail age structure of the
municipality, presence of the competition and the ownership nature of the W M company. I
have selected these factors primarily due to the research potential they, in my opinion offer,
and also due to the availability of data, which is a necessary condition for any further analysis.
Besides results of the analyses themselves (including the specification of the used datasets
and methods), in each following subchapter I include a brief overview of the literature
relevant to the examined issue.
4.1 Incentives, distance and kerbside collection13
In this subchapter I consider a better solution for M S W M to be the one that increases waste
separation. In the previous chapter I explained that higher waste separation has usually direct
impact on the lower mixed residual waste generation. Municipality then does not have to pay
costly mixed waste treatment and instead pays much less for separated waste or even profits
from selling such waste. This clarification is necessary in order to avoid using simple
input/output measure like costs per ton of waste, where better solution would be both one with
lower costs or alternatively one with higher waste generation, which is not something desired.
On the other hand, higher waste separation rate represents a desired output - logically, if all
waste is separated, there is no mixed residual waste, which I perceive as the desired state.
When considering waste separation, Folz (1999) stresses the availability of infrastructure
plays a key role in the peoples' willingness to separate. But the sole availability of facilities is
1 3
This subchapter is a shorter version of Struk (2017a).
I l l
not enough, and Williams and Taylor (2004) note that the public needs to be educated in order
to think of waste as a resource rather than as materials that just need to be thrown away.
Further important factors relevant for waste separation, namely the convenience of the overall
process, are provided in Kirakozian (2015), Derksen and Gartrell (1993), Chen and Tung
(2010), Oskamp et al. (1991), and McCarty and Shrum (1994). Importance of the convenience
is further stressed by Williams and Taylor (2004), Gonzalez-Torre and Adenso-Diaz (2005),
Domina and Koch (2002), Perrin and Barton (2001), Hage et al. (2009), and Ando and
Gosselin (2005). Instead of convenience, multiple studies, such as Gonzalez-Torre and
Adenso-Diaz (2005), Hage et al. (2009), or Belton et al. (1994) mention distance (proximity)
to the waste separation facility, what can be also interpreted as the level of convenience or the
effort required. Ayalon et al. (2013) suggests also the importance of proper targeting of the
waste policies to the specific household groups.
The available waste separation infrastructure can be divided into groups based on the level of
separation convenience they provide. Gonzalez-Torre and Adenso-Diaz (2005) have
categorized these from to the least convenient, HWRC (according to Schultz et al. (1995)
most comprehensive but also most expensive) to the most convenient in the form of a
kerbside collection p people separate their waste correctly. In the middle stands a drop-off cheaper
than HWRC, but less convenient than kerbside collection (for further insight on
practical applications of kerbside collection throughout the world reader is referred to
Gillespie and Bennett (2011), Derksen and Gartrell (1993), Zheng et al. (2017), Wilson and
Williams (2007), or Saphores and Nixon (2014).
In case of the Czech Republic, kerbside collection is mostly realized in the form of a "sack
collection". Compared to the kerbside collection using bins, plastic sacks are for one-time use
only, do not require as high initial investment, and can be easily manipulated with. On the
other hand, given the nature of the sack, limited types of recyclables can be collected with this
method, with paper (including cardboard) and plastics being the most common. Such system
is then easily connected with rewards, forming effectively and waste related incentive
program. Strong effects of introducing practically any kind of rewards is mentioned by
Schultz et al. (1995) or Williams and Taylor (2004). More on the issue of incentives related to
kerbside collection in the Czech Republic is available in Struk (2017a), where the most
112
common form of a "reversed" PAYT (a system where people get discounts from usual waste
fee based on how much waste they separate contrary to traditional Pay-As-You-Throw, where
people pay for each additional waste they generate. For a further discussion of the various
aspects of incentive programs in W M is the reader referred to Sauer et al. (2008), Slavik and
Pavel (2013), Bilitewski (2008), Dijkgraaf and Gradus (2004), Fullerton and Kinnaman
(1996), Bucciol et al. (2015), or Shaw and Maynard (2008).
The research goal of this subchapter is to evaluate whether kerbside collection, higher density
of drop-off sites and having an incentive program result in higher separation rates for paper
and plastics in a sample of municipalities from the South Moravian Region in the Czech
Republic.
Material and Methods
Municipal solid waste management 2012 data of 455 municipalities in the South Moravian
Region in the Czech Republic were used in this study. The municipalities were selected based
on the data that were accessible via email survey or phone call. Sizes of the municipalities
ranged from population of less than 200 to over 3000. Figure 12 shows frequency distribution
of these municipalities based on their population size. Compared to the whole Czech
Republic, municipalities sampled in this study are a bit larger, but otherwise follow the same
distribution. Few municipalities with significantly larger populations were not included in this
study, as they usually have much more complex M S W M practices, and cannot be easily
compared with small municipalities.
Waste quantity data were acquired from ISOH (Waste Management Information System), a
Czech reporting system for waste producers administered by CENIA. Further quantitative and
qualitative data, like amount of drop-off sites in the municipality, presence of kerbside
collection, or usage of an incentive program were collected directly from the municipalities.
The qualitative data were transformed into the dummy variables that were used in the
subsequent analysis.
113
Figure 12: Frequency distribution of municipality population size of the sample, 2012 state
100
90
80
70
60
50
40
30
20
10
0
1
1 1 1 1 1
1 1 1 1 1
1 1 1 1 1
1 1 1 1 1
& ^ c# <# 0c£
Municipality population size
Source: CZSO
Qualitative data were collected during 2014 via telephone calls directly to the people
responsible for solid waste management in the municipalities. The query consisted of a
standard set of questions about MSWM, including questions regarding the presence of an
HWRC, drop-off sites (sites count and collected waste fractions), availability of kerbside
collection (including pickup frequency and collected waste fractions), provision of any
incentives related to the separated waste collection and details on how incentives are applied.
Incentive programs in the sample took mostly form of mentioned "reversed" PAYT. Other
minor incentive programs included a standard PAYT or a waste fee differentiation depending
on the chosen waste collection frequency (practically realized, for instance, by different color
of the sticker on the waste bin, so the waste collector can clearly identify the bins to collect).
As shown in Table 9, data were collected from 455 municipalities (673 total in the region),
with average population of 923 (median 580) and built-up area of 0.158 km2
(16 hectares).
Data on built-up area were used instead of the total area (9.50 km2
on average per
municipality) as it represents the actual area where people live and generate waste, excluding
the forests, lakes, meadows, etc. Each municipality includes on average 319 dwellings and
4.12 drop-off sites. This equates to 4.47 sites per 1000 population, 12.94 sites per 1000
dwellings, or 0.29 sites per hectare.
114
Table 9: Descriptive statistics of the municipalities in the sample
455 municipalities (2012 values)
Bottom
value
1st
quartile
Median
3rd
quartile
Top value Average
Population1
37 301 580 1027 21681 923
Dwellings1
14 107 196 340 8456 319
Area (ha)1
42 503 783 1215 5484 950
Built-up area (ha)1
1.5 7.2 11.7 18.2 212.3 15.8
Drop-off sites3
0 2 3 5 113 4.1
Drop-off sites per built-up ha1 , 3
0 0.17 0.26 0.37 1.74 0.29
Residual MSW (t)2
9 59 117 223 4123 196
Residual MSW per capita (kg)1,2
64 170 208 242 1505 219
Paper and plastics (t)2
0 4 9 21 642 20
Paper and plastics per capita (kg)1,2
0 11 16 23 84 18
Source: 'CZSO, 2
ISOH and3
own data
From the Table 9 I point out the top residual MSW per capita value of 1505 kg, which is very
high compared to the rest of the municipalities in the sample. The reason behind it is that the
given municipality has permanent population of approximately 50, but also additional more
than 200 recreational dwellings within the cadaster. Therefore the actual amount of people
generating waste is much higher than the official count, leading logically to the very high per
capita values (using the permanent population count). The second highest value, around 1000
kg per capita, was reported in the municipality with landfill within its cadaster, and the W M
company in exchange does not require any payments for M S W M in this municipality. As a
result, people there generate as much waste as they like, because they are not charged for it in
any way, and even people from other municipalities bring in their waste, avoiding charges in
their municipality. Besides these two, the highest per capita values were up to 700 kg.
To reach stated research goal of this subchapter, I used simple statistical methods for the
comparison of municipalities' performance, and then answered the following three questions:
1) whether the presence of a kerbside collection system has a positive impact on the
separation rate;
2) whether the presence of an incentive program has a positive impact on the
separation rate, and
3) whether a higher density of drop-off sites has a positive impact on the separation.
115
For testing the hypotheses I used OLS (ordinary least squares) regression technique, which
estimates the coefficients of selected variables in a specified linear function with the goal of
minimizing the sum of the squares of the differences between observed data points and points
predicted by the specified function. As a dependent variable I used the separated waste rate,
representing the percentage of separated waste (in this case paper and plastics) from the total
sum of separated plus residual waste, like in for example Bucciol et al. (2015).
After the necessary transformation of the qualitative data into dummy variables, I used the
following variables in the regression analysis:
Separation rate (dependent variable) - the percentage of separately collected paper and
plastics waste fractions in the municipality from the total generated municipal solid waste
(including residual waste, recyclables waste fractions, etc.)
Kerbside collection (independent variable) - a dummy variable, 1 if municipality was
using kerbside collection for the separate collection of paper and plastics waste fractions,
0 if not
Incentive program (independent variable) - a dummy variable, 1 if municipality was using
some kind of incentive program in order to increase the separate collection of recyclable
waste fractions, 0 if not
Drop-off site density (independent variable) - the amount of drop-off sites in the
municipality per hectare of built-up area, amount of drop-off sites based on the
information provided by the municipal representatives
I continued with the regression analysis and tested the three hypotheses:
Municipalities with kerbside collection system have higher separation rate (Hi)
Municipalities with incentive program have higher separation rate (H2)
Municipalities with higher density of drop-off sites have higher separation rate (H3)
116
I tested each hypothesis separately and then tested the joint hypothesis that can be formally
written as
separation rate = /?0 + ^kerbside + f>2incentive + /33drop- off density + e
where /?0 stands for constant,
/?n stands for coefficient of independent variable, and
e stands for random error component
Results
First I show the results of a statistical analysis of the sample of municipalities divided into the
three categories based on their recyclables collection system (Figure 13). The first category
contains municipalities that use only drop-off sites and no kerbside collection or incentive
program. A total of 379 municipalities fall under this category, with average paper and
plastics separation rate of 7.7% (median 7.5%).
The second category contains municipalities that use only kerbside collection system for
paper and plastics but no incentive program or drop-off sites. There were 7 municipalities in
this category, with an average paper and plastics separation rate of 9.7% (median 9.1%).
The third category contains municipalities that implemented some kind of incentive program
with either drop-off sites or kerbside collection, or both. There were 10 municipalities in this
category, with an average separated paper and plastics rate of 17.2% (median 14.5%). If we
consider subgroup of municipalities with incentive program and both drop-off sites and
kerbside collection (6 units), their average paper and plastics separation rate is 21.0%.
117
Figure 13: Average and median paper and plastics separation rates
20%
18%
16%
14%
12%
10%
8%
6%
4%
2%
0%
] Average paper and plastics separation rate
I Median paper and plastics separation rate
7.7% 9.7%
Drop-off Kerbside
17.2% 14.5%
Drop-off and/or Kerbside
+ Incentive
*standard errors are 0.2% for drop-off, 1.9% for kerbside, and 2.5% for drop-off/kerbside with incentive
Source: own construction
In addition, 55 municipalities form the sample used some combination of kerbside collection
and drop-off sites (usually at some remote part of the municipality) without any incentives
with average separated paper and plastics rate of 9.6% (median 9.0%) and 4 municipalities
that used only occasional mobile collection of recyclables with average separated paper and
plastics rate of 4.4% (median 3.1%).
Table 10 contains results of the OLS regressions. The first model tests Hi. Results indicate
that kerbside collection increases separation rate by 2.9%. Considering the base level of
separation represented by the constant (7.7%), this means that a kerbside collection system
increases the paper and plastics separation rate by almost 40%. Acquired result that this factor
can increase municipal separation level is statistically significant, thereby confirming Hi.
The second model tests H2 . Results indicate that incentive program increases separation rate
by 9.3%, which more than doubles the total separation rate compared to the base level
represented by the constant (7.9%). Acquired result that this factor can increase separation
level is statistically significant and thus confirms H2.
118
Table 10: Estimated effects of collection systems and incentives on separation rates (in %)
Hi H2 H3
Kerbside Incentive Drop-off sites Joint model
collect, model program model density model
Kerbside collection
2 gg***
(0.65)
2.63***
(0.65)
Incentive program
9 26***
(1.56)
8.04***
(1.55)
Drop-off sites density ^ 21*** 3 95***
(per hectare) (1.16) (1.12)
Constant
7 70*** 7 93*** 7 20*** g41***
Constant
(0.25) (0.23) (0.41) (0.42)
No. of observations 455 455 455 455
Adj. R2
0.039 0.070 0.014 0.115
P-value (F) <0.001 <0.001 0.006 <0.001
*, **, *** statistical significance of the coefficient at 10%, 5%, 1%; standard errors in parentheses
Source: own calculations
The third model tests H 3 . Results indicate that additional drop-off site per built-up hectare of
municipal area increases separation rate by 3.2%. A density of one drop-off site per built-up
hectare is in practice quite high. In our sample, the average density of 0.29 sites per hectare of
built-up area corresponds to one drop-off site per area of a circle with 105 m radius. Again,
acquired result that higher density of drop-off sites can increase separation level is statistically
significant and proves our assumption in H 3 .
The final joint model includes all three examined factors together. It shows again that all of
the individually examined factors are statistically significant. Based on this model it is
possible to estimate the impacts of individual examined factors on paper and plastics
separation rates for a municipality with information about the presence of the factors.
Here I note that in the sample only 11 out of 455 municipalities did not have drop-off sites therefore
the constant in the regression model can be used as a proxy for the municipality with
only drop-off sites for waste separation plus eventually some other factors not captured in the
data (the constant in the joint model excluding these 11 municipalities was 6.42, which is
practically identical to the joint model of the full sample). None of these 11 municipalities
have an incentive program, but 7 have kerbside collection. Average separation rate of this
subgroup was 7.8%, which is comparable with municipalities with only drop-off sites (Figure
119
13). Remaining 4 municipalities that do not have drop-off sites or kerbside collection have
average separation rate of 4.4%. Their waste separation is conducted through occasional
mobile collection of recyclables with possible use of some municipal property as a temporary
storage facility. Nevertheless, this subgroup of municipalities is very marginal and represents
less than 1% of the sample population and thus has practically no impact on the results.
Table 11 shows differences in average total MSW generation and amount of separated paper
and plastics between municipalities without incentive program (455 municipalities) and those
with some kind of incentives (10 municipalities). With incentive program present is the
amount of MSW 20% lower (28% lower for just residual waste), while the amount of
separated paper and plastics is 71% higher (regression model suggests even higher increase).
Table 11: Average amount of generated waste per capita with and without incentive program
Municipal average value
Total MSW Residual waste Paper and plastics
Municipal average value
per capita (kg) per capita (kg) per capita (kg)
No incentive program (445 municipalities) 238 220 18
Some incentive program (10 municipalities) 190 158 31
Change if incentive program is present -20% -28% +71%
Source: own calculations
Finally, I have examined whether the amounts of generated MSW and collected recyclables
change with the size of the municipalities. Municipalities were sorted according to their size
and equally divided into five groups (quintiles) from the smallest to the largest (Figure 14,
values represent averages from individual municipal values, information about sizes of
municipalities are under quintiles). In case of the total considered generated MSW (in our
case the sum of residual MSW, paper and plastics) we observe a slightly higher (20%)
generated amount among the smallest municipalities. In the rest of the sample average values
are stable and vary only by few per cent. In case of separated paper and plastics per capita we
can see an increasing trend with the municipality size, with the increase between the first and
the fifth quintile being approximately 36% on average.
120
Figure 14: Average amount of generated waste per capita according to the municipality size
• Total MSW • Separated paper and plastics
1st quintile 2nd quintile 3rd quintile 4th quintile 5th quintile
pop. <250 pop. 255-446 pop. 447-735 pop. 748-1212 pop. >1213
Source: own construction
Discussion of the results
In this study I did consider only MSW (waste produced by households in the municipalities)
that represented 17% of the total waste generation in the Czech Republic in 2012. I did not
consider industrial and commercial waste, which in many developed countries accounts for
majority of the total generated waste. Compared to MSW, industrial and commercial waste
streams seem to be notably better developed in the Czech Republic, as the utilization of these
wastes was in recent years close to 80%, compared to less than 50% of MSW. Also the
landfilling rate of industrial and commercial waste was slightly over 10%, compared to more
roughly 50% in case of MSW (MZP, 2017b). The reasons are both in more stringent
industrial and commercial waste legislation, and that it is much easier to clearly identify the
producer of this waste, unlike with the municipalities, where the producers are the
households. Also in case of industrial and commercial waste the payments for waste treatment
are notably higher, creating much stronger incentives for the companies to take better care of
their waste.
I also did not cover other important types of recyclables, such as glass and green waste, as
these types of waste were not typically collected through kerbside collection in 2012. If I
included them, the total separation rates would be significantly higher, as glass and green
121
waste are much heavier types of waste compared to considered plastics or paper. According to
MZP (2017b), total MSW separation rate in the Czech Republic is over 30% in the recent
years, compared to 10% if considering just paper and plastics examined in this study.
As was mentioned, convenience of waste separation, usually interpreted as the perceived
availability of waste separation sites (Derksen and Gartrell, 1993; Chen and Tung, 2010), or
alternatively the distance to the nearest waste separation site (Gonzalez-Torre and AdensoDiaz,
2005), is a key factor of participation in separation. The willingness of people to
separate can thus be defined as a function of the perceived availability/distance to the nearest
separation site (in our case represented by variables of kerbside collection and drop-off sites
density). Much empirical research was conducted on this issue with the only difference
between the results in the actual magnitude of the effect of distance, as it is affected by many
variables that might be related to the specific place where the research took place. These
include, for instance, the usual habits of the households, housing structure, type of waste
collection systems, education of the people, or how well the people are informed about the
available waste separation options and benefits of it (Martin et al., 2006).
I did not examine whether individuals decide to separate waste, like Knussen et al. (2004),
Knussen and Yule (2008), or Miliute-Plepiene et al. (2016), but how much they separate.
However, the psychology behind this is basically the same. With less effort needed/more
convenient separation system/less distance to the nearest separation site (Perrin and Barton,
2001) and, eventually, a proper incentive, people will participate in separation more often,
resulting in a higher recyclables separation rate. Results showed highly significant effects in
waste separation under these different scenarios.
Also, since it is common to have both an HWRC and a drop-off site right next to each other,
practically forming just one waste collection facility, I did not test the presence of an HWRC
as a factor affecting paper and plastics separation rate.
122
Concerning the quality of the collected recyclables, it is ensured by the collecting company
that checks it on every pick-up, whether it is the drop-off site or kerbside collection. If the
recyclables are contaminated by other inappropriate waste fractions, it is further considered as
regular residual mixed residual waste and charged accordingly. Based on my experience, this
happens very rarely, and municipalities are very active in explaining this to the people, who in
the end bear the additional costs. Very little contamination of recyclables in case of kerbside
collection was observed also by Wilson and Williams (2007).
Results from Figure 13 show the potential of implementing incentive program to the
municipal waste separation. While the difference between the pure drop-off sites collection
and pure kerbside collection is more than a quarter in favor of kerbside collection,
combination of these two systems did not show any notable difference compared to the pure
kerbside collection, both in terms of average separation rates or statistical significance of their
interaction, unlike in Sidique et al. (2010). Implementing incentive program, however,
doubled the separation rates, and this strong impact was verified also in the regression model.
While both groups of municipalities (kerbside collection without incentives; any incentives
implemented) are relatively small, standard errors of averages show that achieved values are
statistically significant, suggesting that the impact is consistent. However, one drawback with
this collection system is that the kerbside was generally reported by municipal representatives
as more expensive than drop-off sites, had municipality previously used them and therefore
could compare. Thus, it is in the end more of a political decision whether municipality is
willing to accept higher costs for higher separation rates and convenience. This issue becomes
even more important in case of implementing incentives. If not properly planned, increased
separation rates might be outweighed by the additional costs. Moreover, as Thogersen (2003)
or Halvorsen (2012) note, shifting an intrinsic motive to separate waste based on one's belief
to purely economically motivated behavior could become an issue, once the incentives are no
longer perceived as paying off.
Positive impact of kerbside collection on waste separation levels (tested in Hi) was
confirmed. Multiple studies, for instance Ashenmiller (2011), Dahlen et al. (2007), Wilson
123
and Williams (2007), or Dahlen and Lagerkvist (2010), confirm significant impact of door-todoor
separation systems on waste separation and estimate the increase in separation rates up
to 100% (in our case 40% increase). Our relatively lower results might be caused by, for
instance, already high separation rate without kerbside collection. Also, in some
municipalities the kerbside collection was implemented only recently, and it might take some
time until the households understand and adjust to this collection system, as appropriate
communication strategy of the municipality plays a crucial role (Mee et al. 2004).
The generally most convenient separation system, kerbside collection, showed the highest
paper and plastics separation rates in our sample. Such results confirm previous studies. For
instance, Barr et al. (2003) concluded that access to the kerbside waste separation (in our case
in the form of the sack collection) greatly enhances waste separation behavior, Folz (1991)
concluded that the participation of public in the waste separation is almost twice as high with
kerbside separation system compared to the drop-off sites system, or Abbott et al. (2011)
identified expansion of a kerbside waste separation as a key factor in UK's improved
recycling performance.
An interesting related observation reported by the local representatives is that the sole
visibility of the kerbside waste separation (in our case, a sack filled with recyclables at the
edge of the property) may motivate others (i.e. neighbors) to take part in the waste separation
process. The "created" social pressure on the non-participants can be perceived as another
benefit of the kerbside separation system that tends to work in smaller municipalities with
limited anonymity of the people. Positive effects of such social interactions between
neighbors on recycling behavior have been examined in more detail by Shaw (2008).
Positive impact of incentives on waste separation levels (tested in H2) was confirmed. Similar
results were acquired by, for instance, Allers and Hoeben (2010) or Bucciol et al. (2015).
Incentive programs are generally accompanied by a kerbside collection system, or some other
system that allows matching of the waste producer with the acquired reward. Bucciol et al.
(2015) calculated combined effect of kerbside collection and incentive on the waste
separation rate of municipalities from selected Italian district to be almost 17%, while in our
124
case it was almost 11%. However, mentioned study considered also additional recyclables,
therefore the results for paper and plastics could be in fact similar to ours. When considering
evidence from using standard PAYT, Sidique et al. (2010) estimated that incentives in the
form of variable pricing could increase the rate of recycling by up to one third, and were the
most important factor in the total separation rate. These results suggest that if an incentive
program for waste separation is present in the municipality, the willingness of the people to
separate waste tends to increase dramatically. But unlike in the case of standard PAYT,
"reversed' PAYT used in the Czech municipalities motivates municipalities only to increase
waste separation, and not to dispose of the waste in questionable ways.
Finally, testing H 3 confirmed positive impact of higher drop-off site density on waste
separation. However, the real impact should not be overestimated. Average density in our
sample is 0.29 sites per hectare of built-up area. According to the joint model, to match the
impact of a kerbside collection system it would require increasing the density from 0.29 to
0.96, which means more than tripling the amount of drop-off sites. Especially in larger
municipalities with many drop-off sites this is unlikely to be economical. NIMBYism (Not In
My Back Yard) might become an insuperable issue here as well (Byrne and O'Regan, 2014).
And at last, our model assumes constant returns to scale of additional drop-off sites, while
municipal representatives mention rather decreasing returns to scale, thus further worsening
the potential of this measure.
The calculated adjusted R 2
values in estimated models were relatively low, but the statistical
significance of the examined factors was very strong in all cases. Inclusion of control
variables such as municipality size produced slightly higher adjusted R 2
values, but the
coefficients of examined independent variables remained practically similar, thus I do not
report these models. Low R 2
indicates weak ability of the model to predict the values of
dependent variable, suggesting that additional variables should be used to explain more of the
variance. However, in our case the goal is not to predict the separation rates, but to estimate
the impacts of selected independent variables. In cases like this we are therefore interested
primarily in statistical significance of the coefficients of independent variables rather than R 2
values of the models. Low R 2
values are reported also in comparable models from other
studies, for instance Ashenmiller (2011) reports pseudo R 2
of 0.125, or Sidique et al. (2010)
125
report pseudo R2
of 0.08, and 0.11 when using log of separation rate as a dependent variable.
In other studies, like Dahlen et al. (2007), R values are not reported at all, as the focus is on
the coefficients of independent variables. I have also found evidence of models with R higher
than 0.6 in Bucciol et al. (2015), but in this case panel data from 10 year period were used in
models with 15-20 independent variables, unlike in our case. Our examined variables thus
might not be the major contributors to the separation rate in individual municipalities, but
they consistently explain certain trend, which is the focus of the study.
Results in Table 10 stress the environmental benefits of an incentive program. Even though
there are only 10 municipalities with incentives, the decrease in residual waste and increase in
separated waste is notable. Linderhof et al. (2001) and Dijkgraaf and Gradus (2004) analyzed
waste generation of municipalities for several years after implementing incentive program and
report even larger decrease in the residual mixed waste in general, although recyclables did
not increase as much as in our case. But this might be caused simply by higher initial level of
the waste separation in their samples. While some could argue that the decrease in residual
mixed waste generation is a result of, for instance, illegal dumping, with the employed
"reversed" PAYT system I do not consider this to be likely cause.
Size of the municipality (Figure 14) could have some impact on the amount of generated and
separated waste. Higher average values of generated waste per capita in the smallest
municipalities in our sample are, in my opinion, caused partially by the fact that these
smallest municipalities are primarily located in the countryside and often include second
homes or cottages, where people generate waste, but are transient residents, thus not
influencing amount of permanent population.
On the other hand, higher amount of collected recyclables per capita in the largest
municipalities in our sample is most likely caused by generally better separation options in
larger municipalities with more drop-off sites, and at least one HWRC. Also, due to the higher
population density, distances to the closest separation options are shorter, which equals to the
higher convenience, thus making people more likely to separate.
126
This research could be further expanded by examining impacts of analyzed factors on other
types of recyclables, although not many of them are consistently collected using drop-off sites
or kerbside collection. Adding multiple years could also provide additional information on
waste generation and separation change throughout time. Using more variables, for instance
socioeconomic characteristics of municipalities, or M S W M related characteristics like waste
collection frequency, could also shed some more light on the issue, but such data are often not
centrally available, making it difficult to create an appropriate dataset for the analysis.
Finally, in this study I have not examined the added benefits of examined measures in
economic terms. Increased convenience of the kerbside collection and higher amount of
separated waste could be possibly evaluated by cost-benefit analysis, or appropriate nonmarket
valuation method. Performing a cost-benefit analysis by the interested municipality
with its own data and taking into account its unique characteristics is generally recommended
before considering adoption of any of the examined measures. Especially appropriate
valuation of social impacts in the economic part of the cost-benefit analysis can prove to be
difficult, while possibly having the crucial impact on the analysis outcome.
Partial conclusions
I examined the waste separation systems of drop-off sites, kerbside collection and the impact
of incentives on waste paper and plastics separation rates. My findings confirmed that the less
effort needed for waste separation, and the lesser distance to the nearest waste separation site,
the more waste people separate. The presence of drop-off sites and kerbside collection system
resulted in approximately 7.7% and 9.7% paper and plastics separation rates. However,
incentivizing the local population can increase separation rates significantly to 17.2%.
Impacts of analyzed factors on the waste separation rates were examined also in the
regression analysis, and were found to be statistically significant. Compared to the most
common collection through the drop-off sites, kerbside collection increases separation rates
by 2.6%, incentive program by 8.0% and 1 additional drop-off site per hectare (sample
average is 0.29 per hectare) by 3.9%. Adding an incentive program can thus more than double
the usual separation rates.
127
Results from this research can be utilized by municipalities planning to significantly improve
their waste separation rates. Introducing an incentive program can have a great potential to
increase the waste separation, without necessarily high investments from the municipality, if
appropriately designed. However, it is recommended that a cost-benefit analysis is to be done
before adopting any of the analyzed measures.
The limitation of these results are in the fact the examined efficiency factors were identified in
the DEA models using sample of municipalities only from South Moravian Region in the
Czech Republic. Therefore the validity for the rest of the Czech Republic might be
questioned. In my opinion limited initial sample does not present that much of a problem. The
reasons for that is the fact that the Czech Republic is not that large country and the differences
between individual regions are not that large. If excluding Prague (which is practically one
large city, and in such M S W M tend to work differently than in an average municipality),
average monthly wage in individual Czech regions was in 2012 between 760 EUR and 875
EUR (based on data from CZSO), which translates to 15% difference. Furthermore, used
dataset includes more than 500 municipalities, therefore any outlier should have only
marginal effect on the final results, if they are present. Additional remarks could be towards
the prevailing lifestyle of the people in various parts of the country, but South Moravian
Region includes both larger and smaller municipalities, municipalities in dominantly
agricultural areas as well as in more remote woods and in general should not be significantly
different from common municipalities in other regions of the Czech Republic.
4.2 Age structure14
In this subchapter I focus on identification whether age structure of the municipality has some
effect on it W M , and if what is this effect for various age cohorts. Previous initial studies of
this issue in the Czech Republic were made by Soukopová et al. (2013), Soukopová and
Kalina (2012), or Hřebíček et al. (2013). Foreign studies of the effects of sociodemographic
factor include, for instance, Hockett et al. (1995), Daskalopoulos et al. (1998), Dennison et al.
(1996), Sterner and Bartelings (1999), Dyson and Chang (2005), Martin et al., (2006),
1 4
This subchapter is a shorter version of Soukopová, Struk and Hřebíček (2017), to which MS contributed
significantly with the parts dealing with methods, material, discussion and conclusions; and Struk and
Soukopová (2016), to which MS contributed significantly with all the parts.
128
Bandara et al. (2007), Beigl et al. (2004, 2008), Hage and Soderholm (2008), Johnstone and
Labonne (2004), Martin et al. (2006), Sterner and Bartelings (1999), Mazzanti and Zoboli
(2008), Benitez et al. (2008), Pickering and Shaw (2015), Gellynck et al. (2011), Lebersorger
and Beigl (2011), Brown (2015), or Talalaj and Walery (2015). However, none of these
studies includes a more detailed analysis of the age structure that simply dividing the
population to those roughly under 14, those between 15-64, and those over 65, what is our
opinion greatly insufficient, and a more detailed analysis of this issue seems appropriate.
The main aim of this subchapter is then to examine how the age structure influences
municipal waste generation, waste separation and related waste expenditure using statistical
and econometric analysis. We have used available data regarding age structure, municipal
waste generation and waste expenditure from more than 5000 Czech municipalities. The
results show notable differences between people of various age, and how the average waste
generation of people changes throughout their life. Our findings could help municipalities to
better focus their measures on specific age groups in order to achieve better M S W M results.
The area of study in both cases of waste generation and waste expenditure comprises the
whole Czech Republic. The Czech Republic has one of the lowest levels of MSW generation
in the EU, with slightly over 300 kg/year/capita (Eurostat, 2015). The reasons for lower MSW
generation are closely related to the population's purchasing power, consumer behavior and
the frequency with which consumer goods are replaced, which is lower in Central and Eastern
European countries than in the countries of Western Europe. A door-to-door collection system
of MSW, that has been dominantly used in the Czech municipalities since 1990s contributes
to this low generation of MSW as well.
Waste expenditure models method and material
We previously analyzed the cost of MSW with respect to the multiple socio-demographic
parameters available in data obtained from municipalities during the national census that took
place in 2011 (Struk, 2015b). Several models from this study produced statistically significant
effects of parameters such as age and housing structure, and the amounts of people that are
129
economically active, unemployed or retired on final M S W M expenditure levels. In this part of
the study we focus on the effects of various age categories on M S W M expenditure levels.
The analysis, which involved both explorative data analysis and multiple regression analysis,
was conducted at the municipality level and used data from the years 2011 and 2014 due to its
availability and quality. The program packages Microsoft Excel 2011 and gretl (version
2015d) were used for the analyses.
The initial sample consisted of all Czech municipalities (6,251 in 2011 and 6,253 in 2014),
but this was reduced to approximately 6,150 municipalities. First, the sample was reduced by
removing all municipalities reporting the absence (or a zero value) of MSW expenditure
(resulting in new totals of 6,212 in 2011 and 6,222 in 2014). The sample was then sorted
based on the MSWE per capita level, and the extreme values were removed from the top and
bottom of the list (trimming the top and bottom 0.5% (31+31) municipalities from the list),
resulting in a sample of 6,150 municipalities. Besides these changes, no other adjustments to
the sample were made.
One thing to note here is that these data are based on expenditure reported by the
municipalities themselves. It is therefore possible that some municipalities have reported
incorrect data, but it is practically impossible to double-check this for all 6,000+
municipalities which provided the data. Still, we only expect such misreporting to happen in
case of very few and rather small municipalities that often lack qualified personnel, or due to
individual mistakes. The overall distortion in the results due to this should be therefore
practically marginal, if any.
In our MSWE models, the basic cost function for MSWE takes the following form:
MSWE = f(Pop )
MSWE represents the total expenditure of municipalities in the Czech Republic on the
management of municipal solid (residual) household waste and is explained as a function of
130
population. MSWE includes the collection, transportation, and disposal of MSW (such
services typically take the form of regular collection/pick-up, transportation to a waste
treatment facility and the MSW treatment itself). It does not include expenditure on recycling
or irregular capital/investment expenditure such as the purchase of new collection vehicles,
new treatment plants, etc. We used the population of the municipality as the independent
variable as it is available practically instantly from the statistical office and in general are
unlikely to contain incorrect entries. Moreover, in several calculations population variable
showed very high correlation with MSW generation, therefore it can be seen also as a certain
proxy for waste generation - more on this later in the subchapter.
In order to develop the models we have used the standard ordinary least squares regression
(OLS) method and have calculated multilinear regression from the available data, as some
other authors (Hockett et al., 1995; Lebersorger and Beigl, 2011). We have transformed the
values obtained for absolute expenditure and population into relative values (expenditure per
capita; the proportion of the population found in each age category) and therefore have not
incurred the problem of heteroskedasticity that would otherwise usually occur with such data
(i.e. when estimating a model with some municipalities that are far larger than the others).
The normality of the distribution of the residuals has also been confirmed. Because of this, the
use of OLS is justified in our case. Due to the significance of the acquired results it was not
necessary to apply more advanced econometric models.
Our dependent variable in the expenditure models is then MSWE per capita (MSWMEpc),
while the independent variables Aget_j are the proportions of people living in a municipality
in certain age categories from year i to year j (i < j are integers) with regard to the total
amount of people living (with permanent residency) in that municipality. The age categories
are 0-14; 15-19; 20-29; 30-39; 40-49; 50-59; 60-64; 65-69; 70-79; and over 801 5
.
1 5
The categories 0-14; 15-19; 60-64 and 65-69 exist due to the need to ensure compatibility with CZSO methods
- In the past, CZSO used to divide the population into only three age categories: 0-14; 15-64 and 65+.
131
A formal description of Model 1, based on the cost function (above), is then:
-39
+ /?5^e4 0 _49 + /36AgeS0_S9 + P7Age60_64 + p8Age6S_69
+ jM#e7 n_7 9 + PioAgeso+ + a
In Model 2 we used only the independent variables Age2o-29, Agew-w, Age^.\% Ageso-59, Age^o-
64, Agee5-69, Ageing and Age%o+ because we focused only on variables which we expected to
have a greater impact on the independent variable. In Model 3 we then simply tested just
economically active inhabitants, using the variables Age^.^, Age^.^, Agem.^, Age^o-64 and
^£^65-69. hi this last model we did not include Age2o-29 variable as, especially in last years,
people often tend to study until their late twenties and delay moving out from their parents.
The population of the Czech Republic at the end of 2014 was 10,538,275. In terms of
population density, the Czech Republic is above the European average, with 131 inhabitants
per km2
. Its age structure shows a clear increase in the over 65+ age category, which refers to
retired people by default (amount of people retiring in recent years is strongly affected by
people born after World War II, who began retiring after 2010). Overall, the average age
increased in 2011-2014 from 41.1 to 41.7 years (CZSO, 2018). When looking at the basic
categories used to describe the age structure in the Czech Republic (the three main categories
0-14; 15-64; and 65+) in Table 12, we can see a 2% increase in the amount of people in the
oldest age category, while the proportion of the people in the "productive" age category
decreased by almost 3%. The proportion of the population in the oldest age category is most
likely to keep on increasing thanks to the improvements being made to healthcare provision,
which are causing people to live longer overall.
Table 12: People in the main age categories in the Czech Republic, 2011 and 2014
Age Amount of Share of total Amount of Share of total Change in share
people in 2011 (%) people in 2014 (%) (%)
0-14 1,488,928 14.32 1,601,045 15.19 0.88
15-64 7,267,169 69.87 7,056,824 66.96 -2.91
65+ 1,644,836 15.81 1,880,406 17.84 2.03
Source: The authors, based on data from CZSO
132
Age category data were acquired from the CZSO1 0
. For 2011 we used data from the national
census, while for 2014 we used statistical data from that year. We have chosen 2014 because
the CZSO publishes more detailed data about age categories in municipalities only since
2014. The normality for the share of each age category in the total municipal population of
each of the municipalities from the sample has been visually checked via a frequency
histogram, not showing any notable deviations from the normal distribution. MSWE data
were acquired from the MONITOR web portal run by the MF CR. It provides complete linked
open data concerning revenue and expenditure for every Czech municipality in the given
years. We have focused specifically on the category of current municipal expenditure related
to MSW collection, transportation and subsequent treatment.
Waste generation models method and material
Data on waste quantity were acquired from CENIA. At the time of our research, the most
recent data were provided for 2013. These data contain amounts of annual production of
generated waste fraction for the Czech municipalities. Although it is natural that every single
municipality produces at least some waste, these data are not available for all municipalities.
This might be caused either by some technical errors, or simply because responsible
municipal representative did not provide the data. Especially in case of very small
municipalities with no professional staff the municipal representative might simply prefer to
do more important things than reporting such data.
For clarification we note that as MSW we consider mixed residual waste - unsorted waste
that ends up in a landfill or in an incinerator, without any further utilization. As separated
waste we consider a sum of any municipal waste that is separated and then somehow utilized
- it consists mostly of separated paper, plastics, glass, biodegradable municipal waste, metal,
or other relevant waste fractions (more details on the issue of municipal waste content are in
1 6
Municipal socio-demographic data are published by CZSO as linked open data. They are publicly
downloadable, along with a detailed description of the method by which they are collected and what these data
actually represent. We have used data from this source regarding municipal age structure, such as the amount of
people in various age categories in each Czech municipality. To be more specific, the CZSO has published data
for each 5-year age cohort since 2014. However, the data from 2011 is taken from the national census, in which
data for each 10-year age cohort were collected. After merging these two sources, we were able to acquire
comparable data regarding age structure for 10-year age cohorts for both 2011 and 2014. By doing this we got 10
age groups that we use in the analysis, compared to only 3 age groups available in the previous years and in
many foreign studies.
133
Chapter 1). Depending on the available separation options in individual municipalities, some
of these fractions might not be separately collected, and therefore ended up as mixed residual
waste in examined years - but we did not examine this in more detail in this study.
Age structure data for individual municipalities in 2011 and 2014 were acquired from CZSO,
as was described in the previous section. As the most recent waste data were available for
2013, we decided to use the age structure data from 2014 together with the waste generation
data from 2013. We assume that the changes in the age structure between the two consecutive
years are marginal, and therefore would not create any significant distortion in the results.
OLS method was again used for the analysis. As we have two sets of cross-sectional data that
are not from consecutive years, we have analyzed both datasets separately. Problem of
heteroscedasticity was avoided again by using per capita values for waste generation and
percentage of people in the individual age groups on total municipal population. Considering
the normality of the data distribution, in case of the age groups we did not observe any
notable irregularities. In case of waste generation per capita, the distribution was a bit
between normal and a Poisson distribution, with slightly longer tail towards higher per capita
values and slightly higher concentration of data points at the lowest per capita values. The
longer right-tail is simply explained by the fact that the data are naturally limited on the left
side by the zero value, but are not limited in their maximum value - in this kind of leftbounded
data this is normal. Higher concentration of the very low values is caused by the fact
that many municipalities reported very low amount of total generated waste, which we
assume was not always correct, but we do not have any measures how to safely identify
municipalities with certainly incorrect data. In order to eliminate at least the extreme outliers,
we have trimmed our sample at several levels, and provide the models for each of these
trimmed subsamples in order to check, whether the calculated results are consistent.
Considering the sample, there was 6251 total municipalities in the Czech Republic in 2011
and 6253 in 2013. In the analysis we left out municipalities that did not report any values for
generated MSW or separated waste - both cases are most likely due to some errors, as
providing M S W M is compulsory for the Czech municipalities according to the law, together
with providing at least some basic options for waste separation. We used several subsamples
134
for the subsequent regression analysis, varying in the trim level from the full sample.
Description of these subsamples is provided in the Table 13, together with the mean and
median values per capita for both years and for MSW and separated waste.
From the data description we can see that MSW generation in the Czech Republic is
approximately 300 kg of total municipal waste per capita per year (residual plus separated
waste fractions). Compared to the EU-27 average of 500 kg per capita per year, this is a
notably lower value. According to Závodská et al. (2014), the reasons for that are partly
historical with the traditional thriftiness and lower standard of living in the Czech Republic
compared to the western countries, both leading to a lower wastage rate, and partly due to the
observed fact that in case of the Czech municipalities, reported data cover only waste
produced by the households, while in the case of some other countries these data cover also
waste from municipal sweepings and public parks. This is a result of slightly different
definitions of municipal waste and its fractions in various countries, as stressed in Hřebíček
and Soukopová (2017). Therefore the actual difference in MSW per capita generation in the
Czech Republic is not that much lower compared to other E U countries, as it might seem
from the data. In our case this methodological issue does not matter much, as we are dealing
with the data only from the Czech Republic, which are therefore consistent.
Table 13: Reduction of the amount of municipalities in the subsamples
20ÍI 20Í3
Trim level Reduced Final amount in Reduced Final amount in
(bottom+top) municipalities the subsample municipalities the subsample
Residual 0.33% 17+17 5082 18+18 5268
municipal 0.50% 26+26 5064 27+27 5250
solid waste 1.00% 51+51 5014 53+53 5198
<20% & >500% 255+6 4855 170+6 5128
Separated 0.33% 19+19 5752 20+20 5990
solid waste 0.50% 29+29 5732 30+30 5970
1.00% 58+58 5674 60+60 5910
<20% & >500% 592+66 5132 488+42 5500
Source: own calculations
135
Results and Discussion - Municipal solid waste expenditures
The results of OLS estimates of MSWEpc based on data concerning the age structure in
municipalities from 2011 and 2014 are presented in Table 14.
Table 14: OLS estimation, Czech municipalities, 0.5% trimmed sample
Dependent variable: MSWE per capita (CZK/year)
Model 1 Model 2 Model 3
(2011) (2014) (2011) (2014) (2011) (2014)
const -221.1 -820.9** -16.12 94.62 -47.03 -68.88
(191.6) (373.9) (98.85) (197.5) (60.42) (117.3)
Age0-14 404.5*
(229.0)
1270***
(431.1)
-411.5
(283.4)
-313.8
(594.4)
-^fif e
20-29 -285.3 189.0 -494.4*** -741.9**
(234.3) (454.2) (170.3) (339.3)
Age30-39 593.3*** 859.2* 477.9** 166.7 444.6*** 393
(226.8) (455.4) (197.4) (400.4) (138.2) (276.1)
Age40_49 1609*** 2494*** 1268*** 1400*** 1210*** 1566***
(230.8) (435.5) (193.7) (381.8) (169.8) (318.3)
Ageso-59 1955*** 3026*** 1757*** 2074*** 1716*** 2022***
(237.1) (486.1) (152.5) (306.7) (144) (296.2)
Age60-64 2643*** 3030*** 2430*** 2107*** 2523*** 2294***
(261.6) (515.5) (187.6) (370.4) (173.3) (344.8)
^ge
6S-69 1207*** 3085*** 1028*** 2H4*** 1316*** 2490***
(286.4) (524.7) (221) (379.9 (209) (351.9)
Age70-7<i 854.6*** 1676*** 666.5*** 764.2**
(285.7) (542.1) (185.7) (342.2)
Ageeo+ -370.1*
(217.8)
-962.0**
(419.1)
n 6150 6160 6150 6160 6150 6160
Adj. R2
0.0668 0.0244 0.0672 0.0232 0.0644 0.0211
Standard errors in parentheses
statistically significant at 10%; ** statistically significant at 5%, *** statistically significant at 1%
Source: own calculations
Model 1 clearly shows the trend of gradually rising MSWEpc with increasing age up to the
60-64 age group, where MSWEpc reaches what is by far its maximum value. After that peak
there is a sudden drop in MSWEpc, which continues further with increasing age. We found the
very high MSWEpc peak for the 60-64 age category (compared to other age categories) very
interesting. We have also estimated econometric models with absolute MSWE values and
136
absolute quantities of people in municipalities (although there are naturally problems with
heteroskedasticity even if we use robust standard errors or heteroskedasticity-corrected
models), but the results suggest the same trend with MSWEpc peak at the 60-64 age category.
This result has very strong statistical significance (p-value<0.01) in basically every model we
have estimated (see Struk, 2015b for additional models from 2011). This result is interesting
because it does not correspond to the outcomes of studies from outside the Czech Republic
(Sterner and Bartelings, 1999; Beigl et al. 2008).
The aspect of ageing was also examined in Sterner and Bartelings (1999) and Beigl et al.
(2008), but no conclusions were reached regarding its importance, unlike the results shown
above. In our opinion this might be caused by the insufficient definition of what constitutes
the senior population in these studies.
As ageing is part of a natural and continuous process of change in individuals, it is important
to clearly define the moment when middle-aged people become seniors. Multiple approaches
and methods exist for defining seniority objectively (Stuart-Hamilton, 1999). Most
commonly, the start of seniority is based on the chronological (calendar) age, but this measure
itself does not provide any deeper information, as it is usually just an administratively set age.
In modern western societies the beginning of seniority is usually considered to be the moment
when people leave work in order to retire. This usually occurs around the age of 60-65. The
majority of gerontologists agree with this age. Vohralikova and Rabusic (2004) note in this
context that "today it is the norm to work until the age of 60 or 65", but do not provide any
further explanation. In the Czech Republic the retirement age has increased in recent years
due to the parametric adjustments occurring as part of social reforms (the calculated
retirement age in 2011 was approximately 62 for males and 59 for females, while in 2014 it
was already 63 and 60, respectively). However, based on the current formula for calculating
the size of pensions, if a person is willing and health-wise fit enough to do so, it is beneficial
for him or her to stay in work if such an option is available (and many people take this
opportunity) in order to secure higher retirement compensation later. Additionally, many
studies further stress the key impact of retirement on consumption behavior, which
subsequently affects related waste generation.
Once we had obtained such a result, we began to think about what the reason might be for
such a spike in MSWEpc in the 60-64 age category in 2011 and in the 65-69 age category in
137
2014. The first idea was that people at such age are usually the most experienced individuals
in the job market, and therefore should earn the most money and be able to consume the most
goods, resulting in the highest amount of generated waste. Well, this is partially true - such
people are on the one hand the most experienced, but on the other they also often have
problems adjusting to new technologies and processes, they are starting to experience health
problems and are generally starting to prefer having more leisure time instead of staying at
work. It may therefore hold for some at this age that they are at the peak of their earning
potential, but many others are long past that point. For this reason we discarded the incomeconsumption
theory behind the MSWEpc peak. Then, instead of giving further considerations
to what might have caused this spike, we simply asked the people around us who are in this
age category about whether and how they have changed their waste-related behavior. One of
the authors of this study is actually in this age group, and comments from his perspective
regarding waste generation correspond to the statements made by other people of a similar
age who were asked.
From the obtained answers we have deduced that once people reach 60 years of age, they
begin to prepare themselves for their actual retirement, which in the Czech Republic usually
occurs a few years after one reaches 60 (as we have mentioned earlier in the text). One
important part of their preparation is often the physical adjustment of their household (or
moving to a different place if the individual can afford it), consisting of (among other things)
reconstruction work, the replacement of ageing household goods, and going through all the
things one has accumulated during one's life, deciding what to keep and what to discard. This
is all usually done with the goal of getting one's household into a state where it will require
minimal further adjustments over what may possibly be the next few decades. What all these
"preparatory" activities have in common is the generation of waste, and if one considers that
this is a kind of "once-in-a-lifetime general clean-up" of things one has accumulated during
the past forty years or so, there is very likely to be a lot of waste generated (at this point we
would add that, unlike in for example the USA, in the Czech Republic it is rather uncommon
to hold something like a yard sale of the old unnecessary things, resulting in probably much
less waste generated during this period). This idea also supports the observation that despite
the general decline in income that people experience after reaching their 60s (which means
that they do likely consume less than before), they are still able to generate more waste by
138
discarding previously accumulated goods. Moreover, with the over 70 age category we have
estimated a notable decrease in the coefficient's effect on their MSWEpc. This is in
accordance with the observed behavior of elderly people, who tend to generate a rather low
amount of waste, as noted by Pickerin and Shaw (2015). Based on all this we have come to
the conclusion that this MSWEpc spike in the 60-64 age category is not actually caused by the
age itself, but rather by the retirement process, which occurred around the age of 60-64.
With the continually increasing age at which people in the Czech Republic retire, it occurred
to us to verify this conclusion by looking at data from later years. This became possible with
the data from 2014, which cover the same variables as in 2011. We theorized that if the
retirement age itself increased in combination with the tendency of many people to further
extend their working career past their retirement age, there should be some kind of observable
shift in the 60-64 MSWMEpc spike from 2011 to the next age category (65-69) in 2014. The
model estimated from 2014 data did indeed show exactly this kind of shift. This is depicted in
Figure 15, which displays the levels of estimated coefficients for the age categories obtained
from Model 3 for 2011 and 2014.
Figure 15: Municipal solid waste expenditure per capita for 2011 and 2014
12011 «2014
3000
2500
•f. 2000
£ 1500
Q .
F5 IOOO
u
500
30-39 40-49 60-64 65-69
Source: own construction, based on data from MONITOR and CZSO
The 2011 spike for the 60-64 category is clearly visible, while in 2014 (apart from the overall
higher MSWEpc levels) there is no longer a single spike for the 60-64 age category, but rather
a shift in the high MSWEpc levels to the next 65-69 category. This observation further
139
supports our theory concerning the linkage between the retirement process and the high levels
of MSWEpc at around the age when retirement occurs.
Another explanation may be the fact that the average age of retirement has been increasing
over the given 3 years thanks to parametric retirement reform (the regular retirement age was
61 in 2011 and 64 in 2014). Moreover, the key moment from the perspective of consumption
is the start of retirement, as emphasized in many studies.
In Model 2 we left out the 0-14 and 15-19 age groups. One reason for doing so is that keeping
all age categories represented by their proportion of the total population would result in
obvious collinearity. Another reason is that, unlike the rest of the age structure data provided
by the CZSO which is divided into age decades, the 2011 data for children is divided into the
0-14 and 15-19 categories (instead of 0-9 and 10-19). Finally, children in the Czech Republic
generally stay at home with their parents until their late teens, what affects their waste-related
behavior to a large extent. However, from Model 1 it is clear that children tend to have rather
low MSWEpc, and are thus comparable with people in the over 70 age category. In particular,
children aged 15-19 tend to cause very little MSWEpc, while the MSWEpc level of the
youngest children is affected by their diapering period in their first years. The results of
Model 2 are similar to the results of Model 1, with the difference that the variable Age2o-29
became statistically significant. These results are very interesting because they show that
young people tend to contribute notably less to the waste expenditures than people in other
age groups.
In Model 3 we focused on variables which are related to economically active inhabitants
(omitting 20-29 age group, as explained earlier). The results of Model 3 further confirm the
results obtained in Models 1 and 2.
Results and Discussion - Municipal solid waste generation
Results for per capita generation of MSW are in Table 15, and per capita generation of
separated waste in Table 16. Both tables include several models differing in the trim level of
140
the full sample resulting in slightly different subsamples, but otherwise contain same
variables. In order to avoid collinearity, we had to leave out one of the age groups - in our
case age group 0-14. This age group was left out specifically because it contains the largest
age span, and the members of this group can have very different waste-generating behavior.
For instance little infants can produce notably more residual waste per day than average
people in the form of diapers, while teenagers are more likely to produce very little waste due
to their budget constraints - they simply cannot generate that much waste if they do not
consume that much. Such tendencies to behave very differently within one age group are
much less likely to be present in other age categories.
Table 15: Residual MSW generation per capita for the individual age groups
0.33% bottom+top 0.50% bottom+top 1.00% bottom+top <20% & >500% mean
trimmed sample trimmed sample trimmed sample trimmed sample
2011 2013 2011 2013 2011 2013 2011 2013
const 144.4** 106.5** 143.6** 85.34* 140.3*** 91.24* 81.82 100.4*
(57.92) (51.26) (56.42) (50.10) (53.67) (47.91) (58.18) (52.20)
Age 15-19 -188.5 -13.52 -184.0 34.90 -162.3 -3.379 -203.6* -175.2
(117.5) (115.6) (114.4) (112.8) (109.1) (108.0) (116.1) (117.7)
Age 20-29 -201.9** -229 8*** -174.6** -208.5*** -133.3* -165.2** -184.5** -239 4***
(87.29) (79.15) (85.08) (77.24) (80.94) (73.78) (88.10) (80.64)
Age 30-39 -13.69 37.20 -23.46 98.46 -9.791 109.3 123.3 45.39
(106.8) (100.1) (104.1) (97.86) (99.02) (93.72) (107.9) (102.1)
Age 40-49 80.07 169.1* 67.70 206.1** 75.67 205.4** 174.9* 189.5**
(92.76) (86.69) (90.41) (84.60) (86.02) (80.86) (92.60) (88.28)
Age 50-59 300.9*** 296.4*** 294 7*** 293 3*** 270.3*** 248.8*** 431.2*** 359 1***
(80.06) (72.83) (78.01) (71.03) (74.34) (67.83) (78.71) (73.66)
Age 60-64 479.8*** 309 1*** 478.4*** 313 2*** 412.8*** 324.1*** 552 5*** 314.3***
(94.99) (86.74) (92.66) (84.54) (88.42) (80.76) (94.02) (88.38)
Age 65-69 421.1*** 469.3*** 415.7*** 446.8*** 468.6*** 399 8*** 618.3*** 567.0***
(112.2) (88.93) (109.5) (86.88) (104.3) (83.32) (111.5) (90.57)
Age 70-79 231.7** 266.9*** 242 9*** 301.0*** 175.3** 243.5*** 228.0** 297 9***
(93.22) (81.89) (90.83) (80.09) (87.17) (76.64) (92.26) (83.91)
Age 80+ -99.67 7.751 -87.97 7.837 -13.41 17.92 45.80 55.78
(112.3) (99.38) (109.9) (96.93) (104.6) (92.78) (110.9) (99.78)
#of obs. 5082 5268 5064 5250 5014 5198 4855 5128
Adj. r2
0.0283 0.0232 0.0292 0.0218 0.0266 0.0183 0.0409 0.0326
*, **, *** stand for statistical significance at 10%, 5%, 1%, standard errors in parentheses
Source: own calculations
We also stress out that the provided values in the models should not be taken literally - for
instance understanding the negative values as that the presence of people of this age group in
the municipality actually decreases the generated MSW. Instead the calculated coefficients
should be perceived more likely as the suggested trend. The correct interpretation is that the
age group with lowest (even negative) coefficient values tends to contribute the lowest to
MSW generation, while the group with the largest coefficient values does the opposite.
Provided models show notable differences of the MSW generation per capita between age
groups. Lowest values of coefficient were acquired by the youngest age groups 15-19 and 20-
29, with values for the latter being consistently statistically significant regardless of year or
the subsample. Such results suggest, that the people aged 20-29 contribute the lowest. From
our perspective are these results in accordance with the fact that people in this age group are
usually still students, or are at the beginning of their working careers, in both cases with
limited budget, subsequently resulting in low consumption and low waste generation.
The following age group 30-39 does not have any statistically significant results, but from the
values we can expect that their waste generation per capita is increasing, which would be in
accordance with the common fact that these people are usually working, and thus able to
consume more, resulting in more generated waste.
Age group 40-49 has significant results in year 2013, suggesting that their contribution to the
MSW generation is becoming higher compared to the younger groups. In our opinion this is
caused by their further extended budget due to their longer job experience.
Age groups spanning from 50 up to 79 have all high and significant coefficients regardless of
year and subsample. Such results suggest that people in these age groups contribute the most
to the overall MSW generation. The highest (peak) values can be observed among the 65-69
and in some cases also the 60-64 age groups. This suggests that throughout the lifetime,
people are likely to generate the highest amount of MSW during this age period and after that,
the amount decreases. Our interpretation of these observations, in our opinion, cannot rely
solely on increasing consumption budget of these people, as people in the Czech Republic
usually commence retirement at around the age 62 (plus minus few years, depending on the
sex and both the availability and willingness to continue working), and also people at this age
often prefer not to work that hard at their jobs as when they were younger. We have observed
a similar thing when examining municipal expenditure related to waste management, and our
idea here is again the same as with the MSWE model in previous section - that people at
142
around this age prepare their households for the retirement, what often includes some
reconstruction, refitting the household with different equipment, replacing it, and getting rid
of things that will likely not be used anymore.
Finally, the oldest age group 80+ does not produce any significant results, but from the values
we can expect that these oldest people tend to produce very little amounts of waste. This is
most likely the result of decreasing consumption budget, as they do not work anymore, and
that they tend not to buy a lot of new things, as they likely already have what they needed.
If we look at the differences between the years 2011 and 2013, in general they are not that
high (a constant has to be included into the consideration, as it is statistically significant in
most cases). We can observe that in case of the age group 20-29 the effects on decreasing
MSW generation are bit stronger, suggesting that people in this age group are continuing to
generate less MSW per capita. Same positive trends can be seen in the age groups 50-69,
where in all cases the contribution to MSW generation seems to be lower than in 2011. On the
contrary, in case of the age group 40-49 the effects seem to be stronger, suggesting that per
capita generation of MSW by this age group increases - but in this case only 2013 results are
statistically significant (2011 coefficient is significant only in the last subsample). In case of
eldest people in the age group 70-79, the effects are approximately the same in both years.
Based on the provided results for MSW per capita generation, we suggest further researches
to focus more on the age groups 40-79, and specifically on the age group 60-69. Results
suggest that these groups produce largest amounts of MSW per capita, therefore they presents
the greatest possibilities for campaigns and measures oriented towards waste reduction,
especially in case of people around the age when commencing retirement. On the other hand,
it seems that the people in the age group 20-29 (and partially 15-19) seem to be doing already
well in these aspects, without need to extend the activities focused on them much further.
In Table 16 we provide models with similar independent variables as previously, but this time
the dependent variable is the amount of separated municipal waste per capita.
143
Table 16: Separated municipal waste generation per capita for the individual age groups
0.33% bottom+top 0.50% bottom+top 1.00% bottom+top <20% & >500% mean
trimmed sample trimmed sample trimmed sample trimmed sample
2011 2013 2011 2013 2011 2013 2011 2013
const 45.72 8.935 51.62* 15.24 54.26** 23.06 43.02 23.69
(29.79) (23.63) (27.95) (22.50) (25.28) (20.79) (26.56) (22.78)
Age 15-19 75.24 170.1*** 89.48 166.5*** 84.30 145.7*** 104.9* 112.8**
(62.16) (53.76) (58.34) (51.15) (52.39) (47.32) (54.34) (51.45)
Age 20-29 43.52 113 5*** 52.64 100.5*** 57.09 60.89* 85.33** 68.09*
(45.54) (36.69) (42.72) (34.92) (38.63) (32.28) (40.60) (35.03)
Age 30-39 -76.50 22.19 -83.16 15.14 -80.68* 7.200 -61.50 11.26
(54.86) (45.68) (51.47) (43.51) (46.59) (40.21) (49.05) (44.09)
Age 40-49 -3.570 -19.63 -29.84 -38.09 -13.26 -28.83 3.284 -29.71
(47.83) (39.47) (44.93) (37.58) (40.67) (34.70) (42.67) (38.26)
Age 50-59 31.29 69.75** 16.04 81.50** -4.874 88.39*** 11.45 83.44***
(40.88) (33.54) (38.38) (31.90) (34.71) (29.49) (36.12) (32.28)
Age 60-64 141.2*** 111 9*** 133 1*** 97 31*** 86.84** 97 27*** 93.51** 97.23**
(47.84) (39.79) (44.92) (37.86) (40.69) (34.90) (42.51) (38.21)
Age 65-69 -14.97 37.30 -59.11 22.07 -59.58 0.3651 -60.11 18.81
(57.13) (40.85) (53.62) (38.98) (48.25) (36.00) (49.97) (39.48)
Age 70-79 19.03 99 75*** 29.83 88.65** 33.41 79.36** 53.10 89.34**
(47.88) (37.46) (44.93) (35.65) (40.34) (32.94) (41.70) (36.04)
Age 80+ 99.90* 109.9** 106.7** 108.8*** 81.67* 60.23 143.3*** 147.1***
(56.09) (44.20) (52.63) (42.04) (47.34) (38.97) (50.11) (43.15)
#of obs. 5752 5990 5732 5970 5674 5910 5132 5500
Adj. r2
0.0050 0.0070 0.0064 0.0083 0.0055 0.0068 0.0080 0.0097
*, **, *** stand for statistical significance at 10%, 5%, 1%, standard errors in parentheses
Source: own calculations
Compared to the results of the MSW per capita generation models, we get less statistically
significant results. Interesting observation is that in case of the youngest age group the values
are among the highest (however, statistically significant mostly only in 2013). This suggests
that the youngest people seem to separate the most, while also generating the least MSW. In
our opinion this is consistent with previous note that these young people have already been
affected by the waste-related campaigns at schools, and they try to behave proenvironmentally.
If they have been raised with relevant information for which they, as
students, had enough time to absorb, it often might seem natural to them to behave in this
way. When we were contacting municipal representatives for gathering data about MSWM,
many of them said that the school in their municipality plays an important role in the waste
management in the way of educating young people and creating pro-environmental habits
through various related activities. Once teenagers at schools were properly explained how to
separate waste and what are the benefits of this activity, they often became actively leading
the rest of the family towards higher separation (or to the waste separation at all). Therefore it
seems logical that in municipalities with higher percentage of people in this age group the
separation amounts of waste would be higher. Some of this effects can be probably
observable also among the next age group, 20-29, but this probably requires more time, as the
people in this age group might not have been exposed to such campaigns towards waste
separation and recycling when they have been teenagers.
The next two age groups, 30-49, seem to be separating waste the least, but the results are not
statistically significant. Based on what the values suggest, it might be that part of these people
tries so separate, while the other part does not, and therefore the results are inconsistent. It
might also be that part of these people is keeping the habits of waste separation, while for the
other part it might be less convenient to do so, for example due to less available time for it.
Age groups 50-64 seems to be contributing to the higher separation rates, and compared to the
previous age group we perceive it as a positive shift. On the other hand, these positive effect
might be simply a side product of the fact that these people tend to produce the highest
amounts of MSW, and therefore if separating even just part of it, it will positively influence
the total separation levels.
In case of the age group 65-69 are the effects on waste separation inconsistent, with the
coefficients very close to zero. The interpretation might be the same as in case of the age
groups 30-49, that part of the people in this age group is actively separating, while the other is
not. If we accept the idea of people refitting their households at around this age (due to
commencing retirement), another explanation might be that people are simply getting rid of
all kinds of things that do not fit into any available separated waste fractions, and thus
contributing mostly to MSW generation.
Finally, in case of the eldest people, we observe positive and mostly statistically significant
results, suggesting that eldest people contribute to the higher separation. Especially in case of
the age group 80+ the coefficients are among the highest, which is an interesting observation
when considering, that this age group produces usually the lowest amounts of waste. In our
opinion this might be caused simply by the fact that these people have usually enough time
for waste separation and while doing it, they relate themselves with this pro-environmental
145
behavior, share their experience with other peers and generally perceive waste separation as
something worth devoting their time to.
As we do not have that many statistically significant results in case of models of per capita
separation levels, we do not draw any conclusions regarding the shift of the values between
years 2011 and 2013 like in case of MSW generation models.
Partial conclusions
In this subchapter we analyzed how age structure can influenced MSWE and MSW
generation, including waste separation. The added value of our study lies in the novel and
innovative approach, giving consideration to new factors affecting M S W M . Society is
affected by demographic change and as a result, demographic characteristics represent
significant determinants that can influence MSWM. Understanding the interdependencies
between age variables and the decisions made by householders regarding the production and
management of waste can help municipal representatives to affect and manage M S W M and
influence certain groups of people regarding their waste generation and waste separation.
We have estimated several models for both MSWE and MSW generation per capita using age
structure data with people divided into 10 age groups. More than 6 000 municipalities in the
Czech Republic have been used as a sample in our estimations using data from 2011 and
2013/2014. During our estimations we made an interesting discovery that the highest levels of
MSWE per capita occur for the 60-64 age category. We have come to the conclusion that this
MSWE per capita spike is not actually caused by the age itself, but instead by people
coincidentally retiring at around that age. Based on our observation of the behavior of people
around the time of their retirement, we have identified that this process is connected with
many activities generating rather large amounts of waste (reconstruction of the home, the
replacement of household goods, sorting and discarding one's belongings accumulated during
previous decades, etc.). In our opinion it is therefore the actual process of retirement occurring
around this age that causes such a spike in MSWE per capita. We also estimated the same
models for 2014, where, based on the fact that the retirement age in the Czech Republic is
slowly increasing and at the same time more and more people are staying in their jobs past
146
their retirement age, we expected a shift in the high MSWE per capita levels to the next 65-69
age category. This expectation proved correct in 2014 model, further supporting our theory of
the connection between high MSWE per capita and the start of the retirement period.
The results of our research show that senior citizens of a certain age (after the start of
retirement) have a surprisingly important influence on the generation of MSW and on
expenditure devoted to dealing with it. These conclusions drawn on the basis of our models
can also be tested in other countries where comparable data is available.
In case of age structure and the MSW generation and separation, our models including Czech
municipalities from years 2011 and 2013 suggest that waste-related behavior of the various
age groups is notably different. We have identified that young people under 30 tend to
produce lowest amounts of residual waste per capita, while also being the best in the waste
separation. On the contrary, people of age 50-79 seem to be generating the most mixed
residual waste, while not separating correspondingly that much. Interesting finding was also
that the eldest people, while generally producing little waste overall, seem to be among the
best at waste separation.
Such findings can be again beneficial for the municipality when planning to improve the
results of its municipal W M and planning to target information campaigns to the specific age
groups (suggested also by Ayalon et al., 2013), instead of treating everyone as the same.
Results in Table 16 suggest that there is no need to focus more attention to the youngest and
the oldest people, as their waste-related behavior seems to be already relatively good (low
MSW generation and high waste separation), and the improvement would most likely be
marginal. Instead of that, the municipalities should focus more on the middle age groups 30-
49 and 65-69, possibly including also the age groups between them. People in these age
groups seem to produce inconsistent results regarding their waste generation and waste
separation level, thus representing the largest possibilities for improvement, both in the area
of MSW generation and the separation levels, which subsequently affect also MSWE.
147
4.3 Competition and public vs. private nature of the waste
management provider17
In this subchapter I focus on the differences in the W M company ownership with respect to
the municipal expenditure together with the possible importance of the competition in the
relevant W M providers market. This study builds upon the available body of literature
(Domberger et al., 1986; Kinnaman and Fullerton, 1999; Reeves and Barrow, 2000; Dijkgraaf
and Gradus, 2003; Ohlsson, 2003; Bel and Warner, 2008; Bel and Mur, 2009; Bae, 2010;
Jacobsen et al., 2013; Bel et al., 2010; Simoes et al., 2012; Szymanski, 1996; Gomez-Lobo
and Szymanski, 2001). In the Czech Republic was this issue with respect to W M services also
examined by Nemec et al. (2005), Ochrana et al. (2007), Nemec et al. (2007), Merickova et al.
(2014), or Nemec et al. (2017).
The aim of this subchapter is therefore to examine how changing a M S W M provider affects
MSWE in given municipality. Unlike in other studies, which usually compare the differences
between public and private providers, I use the empirical evidence to examine the difference
in costs once the change occurred. In addition to that I also look at the difference between
changing from private to public company and vice versa as well.
Methods and Material
There are two primary sources of our data. Financial data were acquired from the web portal
run by MF CR, MONITOR. This portal allows anyone to freely download complete detailed
data for individual fiscal years for all municipalities in the Czech Republic. I used the current
expenditure of municipalities on M S W M in order to compare day-to-day expenditure of
municipalities on provision of M S W M and did not use capital expenditures, as was explained
in the chapter 1. Current expenditure calculated per capita (using municipality population data
available by CZSO) generally provide a good basis for expenditure comparison among the
municipalities, as MSWE covers approximately same things both in smaller and larger
municipalities.
1 7
This subchapter is a shorter version of Struk (2017b)
148
I have collected municipal financial data for several consecutive years, making it possible to
see their development over the time and also allowing to check whether there are any issues
with the data, suggested for instance by a very high difference between individual years. This
is usually a result of municipality reporting their financial data incorrectly. A typical example
is reporting both current and capital expenditures as current, or when municipality does not
differentiate between expenditures from a certain subgroups and report only aggregated
expenditure under the most common category from such group. Reasons behind such
mistakes were already explained earlier, and again I add that in my experience they do not
seem to occur very often, thus the acquired results should be biased only marginally, if at all.
The second source of the data used in this subchapter was interviews with the municipal
representatives from the sample of municipalities that was described in detail in chapter 3. As
I examined the change of the W M company, I focused only on those municipalities that made
such change. Unlike in the case of financial data, there is no centralized source where
municipalities report how they secure their W M . Out of 500+ contacted municipalities 70
experienced a change of W M provider in the last several years. However, due to the very
large interannual differences caused by reporting expenditures related to building an HWRC
site together with the current expenditure I have dropped 4 municipalities, resulting in a final
sample of 66 municipalities used in this study.
When it comes to the actual W M provider in these municipalities, most of them use an
external W M company, as having an own company would not make economic sense. Due to
that they contract external private, public, or a mixed W M company. I now define each of
these ownership types in a way they are used throughout this subchapter, except for private
company, which is self-explanatory.
A municipal W M company is usually historically created by a larger municipality for which it
provides MSWM, and sometimes in addition provides this service also for a few surrounding
municipalities. According to the interviews, such company is usually less focused on profit
and is often part of a larger municipal company generally dealing with various technical
municipal services. Providing M S W M for additional municipalities then serves as a way to
149
better utilize available infrastructure with more focus on the profit (thus practically acting like
a private W M company, but in some cases it can still offer very reasonable conditions).
A public W M company then stands for a company owned by an association of municipalities
where individual municipalities act as the shareholders based on their size and respective
financial investments into the W M company. Each municipality still keeps paying this
company provision for the M S W M services, but it also participates on the profits of the
company. However, during the interviews I have noted occasional disillusions from municipal
representatives with being part of such company, as small municipalities seem to have very
little vote in it compared to few largest municipalities that in practice make all the decisions.
A mixed W M company is usually a result of previous decision of a larger municipality to
partially outsource W M provision but still have some participation in the decision making and
profit, in exchange for having an economically strong partner from the private sector. In such
cases the mixed company often takes form of a newly created company owned partially by an
already established player in the W M market together with the municipality. If relevant, such
company also provides M S W M for surrounding smaller municipalities, just as in case of the
municipal W M company.
Based on the telephone interviews with the responsible municipal representatives, or
alternatively with the municipal representatives with the sufficient knowledge of the topic, I
matched each municipality with the W M company, time horizon when this company provided
M S W M in the given municipality and the type of the W M company ownership.
After that I have created a dataset for several consecutive years with information about
municipalities and their W M company, its ownership type, related municipal expenditures
and any change in W M company, had it occurred during the examined time period.
Tables 17-19 include some basic characteristics of the sample used in this study.
150
Table 17: Description of the sample of municipalities (with respect to the year of the change)
66 municipalities Bottom value Median Top value Average
Population 76
MSWE per capita 291 CZK
930 9 555 1 391
518 CZK 926 CZK 539 CZK
Source: CZSO, MONITOR
Table 18: Year of the waste management company change, 66 municipalities
Year 2008 2009 2010 2011 2012 2013 2014
No. of changes 2 8
% of the sample 3% 12%
5 6 14 19 12
8% 9% 21% 29% 18%
Source: own data
Table 19: Waste management company ownership before/after the change, 66 municipalities
Ownership Private Public Municipal Mixed
Before the change 50
% of the sample 76%
8 6 2
12% 9% 3%
After the change 31
% of the sample 47%
29 4 2
44% 6% 3%
Source: own data
In order to be able to analyze the effect of company change I had to adjust the data, as these
were collected for a broader time horizon and thus difficult to compare. I have adjusted the
dataset in order to have data in the format of MSWE in the year before the change (year -1),
in the year when the change of the W M company occurred (year 0), in the subsequent year
(year +1), etc. After that I have aligned the individual municipal data in order to have
matching periods. By doing this I did not need to consider in which absolute year the change
occurred, and constructed the relative timeline, which is more important for our purpose. So
instead of having 66 changes occurred in 7 years horizon I acquired dataset with company
change occurring in the same relative period.
With the municipal data about the W M company change aligned to the same year 0, I
calculated the relative differences in MSWE per capita between year 0 and year -1 and +1.
Calculating these differences allows us to directly see the change of MSWE once the W M
company change occurred.
151
Our sample of municipalities was also divided into the groups based on the type of change in
terms of W M company ownership. As I mentioned, I differentiated between 4 types of
ownership. After the data collection I have concluded that there are 5 most common situations
with the W M company change in terms of ownership that are discussed later. After such
distinction, each of the situations can be analyzed separately and compared to the other.
Results and Discussion
Figure 16 shows the difference between aggregated data from municipalities in year -1 and
year 0 (after the municipality changed W M company). Provided data are MSWE per capita.
The results show that once the municipality changed its W M company, average per capita
expenditure decreased in the following year on average by 6% (median value was 4%). While
this might not seem that significant, in terms of municipal finance where budgets are often
very limited, even such relatively small change can make a difference. Moreover, in
municipalities that perpetually extend their contract with the W M company, it is common that
MSWE increase each year by few percent. Reduction in MSWE instead of a common increase
thus represents even greater savings for the municipality.
Figure 16: Interannual change in MSWE after the change of the W M company
(66 municipalities), standard errors for averages included
• Average MSWE per capita • Median MSWE per capita
700 r MSWE
before the company change MSWE after the company change
Source: own construction
152
Figure 17 shows the relative interannual changes in MSWE of individual municipalities in the
examined sample. In this figure we see that the change of the W M company does not always
lead to the decrease in MSWE. There might be several reasons for that:
1) new W M company might provide a broader range of waste services, what logically
results in higher costs. For instance the collection frequency might be increased,
additional waste fractions might be separately collected, etc.
2) although being more expensive than before, the new company might still be actually
cheaper than if the previous W M company was still in place. Jump increases in costs
required by the original W M company have been mentioned by several municipal
representatives as the important decision factor when changing the M S W M provider.
3) there might be some additional costs included in the reported MSWE by the
municipality that coincidentally occurred in the same year as the W M company
change. For instance, during this period many municipalities begun with separate
collection of biowaste, which required purchasing composters or additional bins for
the biowaste. Even though such purchases occur irregularly, technically they can
count as current expenditure, leading to the increased reported MSWE in the given
year, and thus possibly resulting in an overall increase of MSWE by several per cent.
Figure 17: Relative changes in MSWE of individual municipalities
after the change of W M company (66 municipalities)
40%
30%
20%
10%
0%
-10%
-20%
-30%
-40%
-50%
-60%
ill
QQDDDDD'
. • • • • 0, 1
Relative change in MSWE after the change of the waste management company
Source: own construction
153
Nevertheless, in the Figure 17 we see that majority of municipalities experienced a decrease
in MSWE after they changed their W M company. Almost 30% of them report decrease in
MSWE by up to 10%, while additional almost 30% report even higher MSWE reduction, with
few municipalities saving even more than 40%. On the other hand, there are few
municipalities that report increase in MSWE by over 30%, but based on my experience, such
increase is very likely due to the one of the above mentioned reasons. Overall, half of the
municipalities from the sample did not experience more than ±10% MSWE difference after
the change.
Szymanski and Wilkins (1993) and Gomez-Lobo and Szymanski (2001) mention that while
there might be great cost savings in the initial year, this advantage tends to diminish quickly
in the following years, resulting again in higher costs and likely next public tender. In case of
part of our sample I was able to obtain MSWE data also from the second year after the W M
company change, and I present the results in Figure 18.
Figure 18: Comparison of MSWE before and two years after the change of the waste
management company (52 municipalities), standard errors for averages included
• Average MSWE per capita • Median MSWE per capita
700 .—
600
MSWE before the company change MSWE two years after the company change
Source: own construction
Figure 18 shows that two years after the company change, the savings still exist (on average
3.5% lower MSWE compared to the period before the change), but these are beginning to
diminish, which is in accordance with the mentioned observation in other studies. With
154
respect to that, competitive tendering every few years might have the potential to keep
MSWE down. On the other hand, any individual municipality needs to decide how often it
should opt for the new tendering, as such action brings additional costs to the municipality.
Jacobsen et al. (2013) mention biannual tendering system in order to get the company with the
best offer. One of the municipalities in our sample began to utilize such biannual tendering
through electronic bidding application and was able to get much better contracts than before,
although this is to a large extent caused by the rather poor starting condition of the W M
provision in this particular municipality.
On the other hand, especially in case of smaller municipalities, costs associated with the
tendering might be higher than the potential savings, and thus the change of the W M company
will likely occur less frequently. The general suggestion here would then be to actively pursue
public tendering as soon as the costs with the current W M company start to seem higher than
they should be and the costs of the tender could be offset by the potential savings.
Finally in Figure 19 I show the changes in MSWE differentiated by the type of the company
ownership before and after the W M company change, if such kind of change occurred at least
4 times. We can see that savings are present in case of any kind of change.
Slight savings are reported whether municipality changes from private W M company to
public (in our conditions owned by an association of municipalities) W M company or vice
versa. Larger savings seems to be possible if switching between private and municipal
companies, but again, data suggest that this goes both ways. Based on that I cannot make a
clear conclusion whether in terms of savings it is better to go with the private or public W M
company. The conclusion that there is rather little difference in waste related costs between
public and private companies is accordance with many authors (Domberger et al., 1986;
Szymanski, 1996; Bel and Fageda, 2010), as well as the observation that much more
important that ownership type of the company is the level of competition present (Gomez-
155
Lobo and Szymanski, 2001; Dijkgraaf and Gradus, 2007; Bel and Warner, 2008, Ochrana et
al, 2007:99).
Figure 19: MSWE change according to the company ownership change (62 municipalities)
o%
LU
I -5%
<u
BO
c -10%to
>
-15%
-20%
I Relative annual change in MSWE according to the ownership
-1%
J
-2%
-8%
-18%
1•-4%
Public -> Private Private -> Public Municipal ->
(6) (29) Private (6)
Private -> Private -> Private
Municipal (4) (17)
*I omitted 2 changes from public to mixed and 2 changes from mixed to private W M companies in this
figure, as such changes are too few to show any relevant results
Source: own construction
However, if we look at the individual municipalities, in case of changing from private to
public W M company there were highest amount of cases where MSWE increased. Thus even
though there has been highest amount of changes from private to public companies (same
trend was observed also by Gradus et al. (2014) in Netherlands), average savings in such
cases were in our sample practically marginal. This is somehow in contradiction with Bel et
al. (2014), who observed significant savings after municipalities switched from private W M
companies to companies owned by a group of municipalities, or Bel and Costas (2006), who
suggested that intermunicipal cooperation (which is in our case represented by a W M
company owned by an association of municipalities) might be a good alternative for small
municipalities with limited amount of potential external W M providers.
Observed increase of MSWE in multiple cases when switching to the public provider might
partially explain mentioned disillusion of some municipal representatives expressed in the
interviews after becoming a member of the association of the municipalities in order to utilize
156
M S W M service of the relevant public W M company. Based on that it is advisable for the
municipality considering this kind of change to carefully examine whether the potential
savings are truly there, compared to the other available options.
According to Massarutto (2007), even better results can be achieved when competitive
tendering is used for dividing more specific activities along the W M value chain. However,
based on my experience with the municipal representatives, such separate competing for
specific activities in M S W M are very scarce. In my opinion the problem might also be the
small average size of the municipality in the Czech Republic, where it does not make much
economic sense for the W M companies to compete for only specific activities and thus
dividing M S W M into distinct activities might be relevant only in the largest municipalities.
But the results in those few municipalities where such separate tenders happen seem
promising so far. Together with the stronger preference of the short-term contracts, suggested
by Simoes et al. (2012) this might become a good strategy for the municipalities in order to
cut down MSWE and keep them there.
Partial conclusions
This study focused on the changes of costs in time, once the municipality switched to a
different company providing W M services. Results show that a municipality is likely to
benefit from such change and that majority of municipalities in our sample of 66
municipalities experienced decrease in their MSWE (on average by 6% in the first year after
the change including the whole sample), in some cases even more than 20-30%. Compared to
the previous situation, lower MSWE levels seem to hold also for the following year, although
savings tend to diminish later.
On the other hand, in some municipalities the costs increased, but based on my experience
this was likely caused by other factors, namely the extension of activities included in the
provided W M service. In other cases, where municipalities experienced MSWE increase after
the W M company change, was this increase often notably less steep than with the previous
provider, what can be perceived as an actual saving compared to the situation without change.
157
Finally I examined the differences in MSWE when switching from public to private MSWM
providers and vice versa, but did not find any significant patterns. It seems that, in accordance
with other authors, it does not matter much whether the W M provider is public or private, but
whether the municipality is willing to regularly engage in competitive tendering. By doing
this the municipality seems to be most likely to get best available services at reasonable costs.
The validity of the results might be questioned, as the data used in the analysis are solely from
one region in the Czech Republic. This is true, it is a limitation of this study. However, data
for municipalities from other regions are not available and it would require additional
collection that, especially considering the fact how fragmented Czech municipalities are,
requires significant amount of time. But in my opinion even results from a territorially limited
sample represent useful findings about the important efficiency factors in the municipal waste
management. It is possible that in some regions there are some additional factors that might
cause a bit different results, but the factors that I have included in this study (ownership
nature of the W M company and the presence of competition) in my opinion apply generally in
the whole country, plus the absolute size of the sample (over 60) municipalities suggest that
the results could be generalized to a certain extent also to the whole Czech Republic.
4.4 Discussion of the findings and suggestions for practical
applications
Throughout this chapter I have examined several factors influencing M S W M efficiency that
have been identified earlier using either 2SDEA or in the relevant literature, and seem to play
an important role in the overall performance of MSWM. These findings include positive
effect of improved convenience of the waste separation on the municipal waste generation and
the amount of separated waste, strong positive effect of implementing incentives in MSWM,
several interesting observations regarding relationship of age structure and MSW generation
and expenditures, and finally the importance of available competition in the field of MSWM,
regardless of the ownership nature of the W M company. Incentives and ownership
arrangements have also been identified as important factors in M S W M area by Simoes and
Marques (2012), together with the effects of economies of scale and benchmarking. However,
in related research by Soukopova et al. (2014) we did not find economies of scale to be
158
present in M S W M area in the Czech Republic. Benchmarking in M S W M was, according to
the interviews with the municipal representatives in our sample, also very rarely used, if at all.
Convenience of the waste separation was identified as a key factor when trying to improve
MSWM. This can be translated into the either availability of the separation options, such as
presence of drop-off sites or HWRC, or the distance to the nearest waste separation site. Both
of these work with the necessary required time in order to separate waste, with logically less
effort needed leading to higher waste separation rates. I specifically examined how well
municipalities separate paper and plastics in scenarios when they have basic drop-off sites,
kerbside collection of recyclable waste fractions, or their combination. According to the
findings, drop-off sites produce on average approximately 7.7% waste separation while
kerbside collection produces 9.7%. While this might not seem to be much, I stress that only
recyclable waste fractions paper and cardboard and plastics were considered, thus total
separation level would be notably higher. Density of drop-off sites was also examined with
the finding that one additional drop-off site per hectare increases waste separation by 3.9%.
But with the average density of just 0.29 drop-off sites per hectare in the sample this might
not be that much important in the practice. However, kerbside collection seems to have a
significant effect on the waste separation. All in all it should not be surprising, as it represents
likely the most convenient option for waste separation. Unlike with the drop-off sites, people
do not need to carry their separated waste anywhere, they just put it in front of their property
at scheduled times for the pickup. Compared to the (in the Czech Republic) standard waste
separation through drop-off sites it increases the waste separation level by 20%. Moreover, if
implemented appropriately, kerbside collection can significantly reduce the problems
associated with the drop-off sites. These include for instance the requirements for a well
accessible public space for the containers placement together with the frequent NIMBY effect
from the locals. On the other hand, kerbside collection might be difficult to implement if the
prevailing housing structure consists of apartment buildings, or the requirement to store the
separated waste fractions inside one's household or at one's property in between the pickups,
what might not be acceptable for some people. Still, when trying to increase waste separation
level in the municipality, kerbside collection seems like an appropriate solution, where the
benefits from higher waste separation (and subsequently lower mixed residual waste
generation) should outweigh the extra costs associated with it. In cases when kerbside
159
collection almost fully replaced drop-off sites it even resulted in comparable or lower
expenditures for the municipality, further stressing the viability of such solution in MSWM.
Incentive programs related to the waste separation produced very strong results with respect
to the achieved waste separation rate. In the model with drop-off sites and kerbside collection
adding an incentive program practically doubled the amount of separated waste. A typical
practical application of an incentive program is in the form of a PAYT scheme with variable
charges based on individual household's waste generation. In Czech Republic mostly in
adjusted (reversed) form in order to reward those who separate more waste. In my opinion
such system has multiple advantages, as people are less motivated to exhibit certain kinds of
unwanted behavior usually related with standard PAYT systems. People in general seem to
prefer systems where they get rewarded instead of punished, even if the final result in terms
of paid fees is practically the same. On the other hand, providing rewards based on the waste
separation requires a more complex planning and possibly other necessary measures. In case
of the Czech Republic, incentives in W M are usually connected with kerbside collection,
where adding a sticker with the barcode allows W M company to clearly connect individual
household with the amount of waste it had separated. However, in the end it is up to the
decision makers (municipal representatives) what they prefer, and most importantly how well
are they able to communicate the benefits of the new measures and ways how they work.
Again as was mentioned above, depending on the specifications of the incentive system, it
might lead both to the increased waste separation and the lower overall MSWE.
Age structure of the municipality has been in this thesis identified as an important proxy for
predicting both MSW generation and MSWE. Importance of such finding lies in the
availability of the data regarding municipal population. Presented models showed differences
between waste and expenditures connected with the waste behavior of 10 various age cohorts
in the sample of municipalities. An interesting and novel finding was that people in the 60-69
age group tend to generate notable amounts of MSW compared to other age groups,
subsequently resulting in much higher MSWE attributed to this age group. The proposed
theory behind this is that people in this age group are usually around the time of commencing
retirement, what is connected with significant adjustments of their households in order to
prepare for their retirement. Such process naturally includes generation of large amount of
160
waste by discard many items from their households with no expected use in future. In case of
estimated models explaining waste separation, notable difference was identified between
various age groups. Young people together with the elderly were identified as those
separating the most, while people of middle age were identified as those separating the least.
Comparable results were obtained in the models explaining MSW generation. Such findings
suggest that middle aged people represents the largest potential for improving M S W M results
in terms of waste separation and minimization, and the campaigns focused towards these
issues should be aimed primarily at them, as the other age categories seem to be performing
well already. Alternatively, municipal representatives could try to adjust the current charging
systems for M S W M services in order to take into account the age factor and vary the height
of local taxes (waste charges) according to the citizen age. This might be perceived as being
more just than the existing waste disposal charging system, and could distantly evoke a kind
of a PAYT charging scheme. However, practical enforcement of such change instead of the
widely used flat fee per capita seems to me at the moment as politically very challenging.
Competition presence in M S W M represents, according to the obtained results, another
important factor in economic performance of MSWM. The worst case seems to be the one
where municipality stays with one W M provider, does not look for any other and simply
perpetually extends the current contract with continually increasing costs due to both inflation
and higher service charge by the W M company. I understand that for some remote
municipalities this might be in fact the only available option, but in case of majority of the
municipalities a regular engaging in W M services tendering should produce benefits in form
of lower MSWE. In the analyzed sample majority of the municipalities that changed W M
company report saving in the following year, on average by 6% if including the whole
sample. MSWE savings seem to hold even in the second year after the company change. And
even in some cases where MSWE did not decrease, new W M provider often brought
additional services or improved the quality of the current ones, which might be worth the
extra costs. As usual with the competition, people as the final consumers are often those who
benefit most from such situations. Besides impact of the company change to MSWE, I also
examined the types of ownership of the W M companies participating in the change. The
literature dealing with the question of whether the public services should be provided by
public companies or contracted out to the private ones mostly does not provide a consensus
161
on the matter. This holds also for our case, when it seems that it does not matter whether the
W M company has public or private ownership, but whether there is sufficient competition
present at the market. The results show that it actually does not matter whether the
municipality changes W M provider from private to public or vice versa - what matters is the
fact that the change itself occurred. Subsequent savings in MSWE were reported in either
case. The municipal representatives should therefore be encouraged to actively pursue regular
public tendering of W M services. Practical experience and foreign literature suggest public
tenders every 2-3 years. Also, even if the W M company does not actually change, the new
contract should at least lead to a more reasonable MSWE levels and reflect currently available
options at the M S W M market.
There are generally two main limitations that I can see in provided studies of selected factor
influencing efficiency. First one is the everpresent possibility that there are some additional
significant factors that influenced the observed results but were not included in the analysis.
This is generally true for practically any models that do not have coefficient of determination
close to unity. But in my case my goal was not the complete explanation of the municipal
performance, but rather the examination of selected efficiency factors that might be
potentially important. And as long as they were identified as statistically significant, the
objective was met and the conclusion about these specific factors could have been drawn.
This leads us to the second limitation, which lies in the availability of data. In an ideal
situation, all thinkable data that could be related to the municipal waste management would
be available for a long time period and for all municipalities. Unfortunately, this is far from
true, and I had to rely on selected variables that are available in the central public databases
and the data I was able to collect. Therefore, strictly speaking, my results should be relevant
only for the limited sample and factors that were included in the examination. However, due
to the used sizes of the samples and sometimes also multiple years, the results should also
have general validity to a certain extent. But I am aware that especially in uncontrolled real
world there are usually many factors that contribute to the final performance of (not just)
municipal waste management, including many municipality-unique and qualitative ones that
were not considered in the analysis. The more I dive into the municipal waste management,
the less I seem to know about this complex issue.
162
CONCLUSIONS
This thesis was focused on the municipal solid waste management and the related factors that
influence its efficiency. The area of waste management was chosen specifically, as it
represents an interference of environmental protection and public service. As was shown in
the first chapter, waste management represents the most important area of environmental
protection expenditure in the Czech Republic, and so is the case in many individual
municipalities. Efficiency improvements in this area thus can have significant and positive
effect on the municipal budgets and on the environment as well.
The objectives of this thesis were to 1) describe situation in municipal solid waste
management in a selected Czech region; and 2) identify relevant efficiency factors in this
region using both collected data and available literature, and based on further analysis provide
policy suggestions for improving municipal waste management in the region.
First, I have collected relevant literature in order to identify potential factors influencing
efficiency of municipal waste management. Next I have collected data from more than 500
municipalities in the South Moravian Region and spoke directly with the municipal
representatives about the municipal waste management in their municipalities and about what
they consider important in this area. Most interesting were probably the numerous qualitative
findings from individual municipalities that show how much variety can exist among the
municipalities in the area of waste management and that there is no one-size-fits-all solution.
This data collection provided me with ideas which factor could be potentially important when
dealing with municipal solid waste efficiency and which of them to focus on in further
analyses. Based on the analysis of the collected data using Data Envelopment Analysis I have
chosen to further examine convenience and incentives factor, and based on the literature study
I have chosen to examine factors of age structure of the municipality and the type of waste
management company ownership including switching between them by the municipalities.
Convenience of the waste separation from the perspective of municipal citizens proved to be a
very important factor in their decision to separate waste. I note here that higher waste
163
separation is typically connected with lower mixed residual waste generation. This is an
important note, as separated waste acts more like a resource and often can be traded in the
relevant market for a profit, while residual waste does not have that much further use and
requires significant costs in order to be treated appropriately. Availability of the infrastructure
such as drop-off points and waste recycling centers is a necessary but not sufficient condition
for improved waste separation. Results show that even more important might be the perceived
convenience of the whole process, usually (and also in our case) represented by the distance
to the closest relevant waste separation site. Less effort needed in order to separate translates
into higher separation rates of municipal waste. The most convenient solution from this
perspective is a kerbside collection, even though in might not be possible to always
implement this system of waste separation. These limitations exist especially in cases where
there is no simple way how to match generated waste and the household that generated it typically
in areas with multi family dwelling such as apartment buildings. In these cases is the
waste separation through standard densely located drop-off sites accompanied by an
occasional waste recycling center probably the best available option.
Incentives proved to have probably the strongest impact on the waste separation levels in
municipalities, and the inclusion of such measures often doubled the previous waste
separation values. Incentive systems in the municipal waste management can be very well
paired with kerbside collection, and every municipality in the sample that implemented both
of these measures reported very high waste separation levels compared to the other
municipalities. However, in the end it is up to the individual municipality and its
representatives to devote necessary time and effort to developing an appropriate system that
takes into account unique characteristics of the given municipality. Here it definitely holds
that one size does not fit all, and what works very well in one municipality might not work at
all in the other. Very important thing here is to appropriately communicate the benefits of the
new system and how it works. If done correctly, people will do the rest.
Age structure of the municipality can act as a proxy for both waste generation and waste
expenditure in the municipality, as both these values are naturally related. Our models showed
notable differences between waste behaviors of different age groups. The most important
findings of this part were probably very high levels of waste generation of the 60-69 age
164
group, which is in our opinion caused by one's preparation for the retirement. On the other
hand, young people tend to produce among the least waste, together with highest waste
separation tendencies. Such observations might be utilized in a better design of the waste
related campaigns that should focus more on the middle aged people, where the potential
benefits are highest, instead of young people, where relatively good results have already been
achieved.
Finally, change of company providing municipal waste management seems to effects related
municipal expenditure generally in a positive way. Results show that majority of
municipalities did benefit from such change and it is suggested to make such public tender
every 2-3 years in order get best available prices and levels of service. Even though this might
not be easily applicable in all municipalities, those that identify at least some competition
around should not be reluctant to change their waste provider, as it can often lead to the
improved waste management in the municipality. And even if they keep the current provider,
public tendering might at least result in better prices and higher quality of the provided
service, if the current provider wants to stay competitive with the others entering the tender.
Besides that I have also examined whether there is any difference in the type of the waste
company ownership and based on the results I conclude that there does not seem to be
anything like that. Improvements after the change of the company happened regardless of
whether the municipality switched from private to public company or vice versa. Therefore I
conclude that it is not the ownership of the company that matters, but the presence of
competition among these companies. With sufficiently competitive environment and the
willingness of the municipalities to change their waste provider, should some better offer
occur, every waste company regardless of the ownership has to stay competitive in order to
survive. A by-product of such situation should be then a provision of an efficient municipal
solid waste management.
Due to the employed methods the validity of the results is limited by the used data and sample
of municipalities, and there is always a possibility that some important factor was omitted
from the analysis. Nevertheless, the used samples were in majority cases relatively large and
the objective was to examine the effects of the specified potential efficiency factors, of which
165
several proved to be statistically significant. Therefore the results should to a certain extent
valid for the whole Czech Republic as well.
Nevertheless, in my opinion some policy recommendations could be drawn from the provided
results. First of all, convenience seems to play a key role for the people when deciding
whether to separate. Thus the proper infrastructure needs to be established. However,
infrastructure sometimes is not sufficient, and other measures, like for instance incentive
programs focused on waste separation could be employed. Results show that with proper
setting such systems are able to both perform better economically and environmentally.
Besides that municipalities should pay closer attention to the inner factors like, for instance,
examined age structure. Results show that there are notable differences between waste related
behavior of various age groups and the municipal approach should reflect that, in order not to
waste money, and instead use them where they can make the most difference. Least but not
last, role of the competition show both in this study and multiple other to be the key factor
when trying to secure best available conditions for municipal waste management provision.
Municipalities therefore should take active part and regularly pursue competitive tendering
and not become accustomed with the current solution for too long. Results also show that it
actually should not matter whether the waste management is provided by the public or private
company, as long as the market is competitive enough.
166
LITERATURE
[1] Abbott, A., Nandeibam, S. & O'Shea, L. (2011). Explaining the variation in household
recycling rates across the UK. Ecological Economics. 70(11), 2214-2223.
https://dx.doi.Org/10.1016/i.ecolecon.2011.06.028.
[2] Afriat, S. N. (1972). Efficiency Estimation of Production Functions. International
Economic Review, 13(3), 568-598. https://doi.org/10.2307/2525845.
[3] Afonso, A. & Fernandes, S. (2008). Assessing and explaining the relative efficiency of
local government. The Journal of Socio-Economics, 37, 1946-1979.
https ://doi.org/10.1016/i .socec.2007.03.007.
[4] Aigner, D. J. & Chu, S. F. (1968). On Estimating the Industry Production Function. The
American Economic Review, 58, 826-839. www.jstor.org/stable/l 815535.
[5] Aigner, D. J., Lovell, C. A. K. & Schmidt, P. (1977). Formulation and Estimation of
Stochastic Production Function Models. Journal of Econometrics, 6, 21-37.
https ://doi .org/10.1016/0304-4076(77)90052-5.
[6] Allers, M . A. & Hoeben, C. (2010). Effects of unit-based garbage pricing: a differencesin-differences
approach. Environmental and Resource Economics. 45(3), 405-428.
https ://dx.doi.org/10.1007/s 10640-009-9320-6.
[7] Ando, A. W. & Gosselin, A. Y. (2005). Recycling in multifamily dwellings: does
convenience matter? Economic Inquiry. 43(2), 426-438.
https://dx.doi.org/10.1093/ei/cbi029.
[8] Ashenmiller, B. (2011). The effect of bottle laws on income: new empirical results. The
American Economic Review, 101(3), 60-64. https://doi.org/10.1257/aer. 101.3.60.
[9] Ayalon, O., Brody, S. & Schechter, M . (2013). Household waste generation, recycling
and prevention. OECD Studies on Environmental Policy and Household Behaviour
Greening Household Behaviour: Overview from the 2011 Survey, 219-245.
https://dx.doi.org/10.1787/9789264214651-ll-en.
[10] Bae, S. (2010). Public versus private delivery of municipal solid waste services: The
case of North Carolina. Contemporary Economic Policy, 28(3), 414-428.
https://doi.Org/10.l 111/i. 1465-7287.2009.00180.x.
167
[11] Bakoš, E., Soukopová, J. & Kaplanova, B. (2009). Porovnání vykazovaných
environmentálních dat se zaměřením na úroveň municipalit. In Účetnictví a reporting
udržitelného rozvoje na mikroekonomické a makroekonomické úrovni, Praha: Lindě, 79-
85, ISBN 978-80-86131-82-5.
[12] Bandara, N. J., Hettiaratchi, J. P. A., Wirasinghe, S. C. & Pilapiiya, S. (2007). Relation
of waste generation and composition to socio-economic factors: a case study.
Environmental Monitoring and Assessment, 135(1-3), 31-39.
https ://doi.org/l 0.1007/s 10661 -007-9705-3.
[13] Banker, R. D., Charnes, A. & Cooper, W. W. (1984). Some models for estimating
technical and scale inefficiencies in data envelopment analysis. Management Science,
30(9), 1078-1092. https://doi.Org/10.1287/mnsc.30.9.1078.
[14] Barr, S., Ford, N. J. & Gilg, A. W. (2003). Attitudes towards recycling household waste
in Exeter, Devon: quantitative and qualitative approaches. Local Environment: The
International Journal of Justice and Sustainability, 8(4), 407-421.
https://dx.doi.org/10.1080/13549830306667.
[15] Battese, G. E. & Corra, G. S. (1977). Estimation of a Production Frontier Model: With
Applications to the Pastoral Zone of Eastern Australia. Australian Journal of
Agricultural Economics, 21, 169-179. https://doi.Org/10.l 11 l/j.1467-
8489.1977.tb00204.x.
[16] Beigl, P., Lebersorger, S. & Salhofer, S. (2008). Modelling municipal solid waste
generation: A review. Waste Management, 28 (1), 200-214.
https://dx.doi.org/10.1016/i.wasman.2006.12.011.
[17] Beigl, P., Wassermann, G., Schneider, F. & Salhofer, S. (2004). Forecasting municipal
solid waste generation in major European cities. In iEMSs 2004 International Congress:
Complexity and Integrated Resources Management, Osnabruck, Germany.
https://scholarsarchive.byu.edu/iemssconference/2004/all/83/.
[18] Bel, G. & Costas, A. (2006). Do public sector reforms get rusty? Local privatization in
Spain. The Journal of Policy Reform, 9(1), 1-24.
https ://doi.org/10.1080/13841280500513084.
168
[19] Bel, G. & Fageda, X. (2010). Empirical analysis of solid management waste costs:
Some evidence from Galicia, Spain. Resources, Conservation and Recycling, 54(3),
187-193. https://doi.Org/10.1016/i.resconrec.2009.07.015.
[20] Bel, G., Fageda, X. & Mur, M . (2014). Does Cooperation Reduce Service Delivery
Costs? Evidence from Residential Solid Waste Services. Journal of Public
Administration Research and Theory, 24(1), 85-107.
https ://doi .org/10.1093/iopart/mus059.
[21] Bel, G., Fageda, X. & Warner, M . E. (2010). Is private production of public services
cheaper than public production? A meta-regression analysis of solid waste and water
services. Journal of Policy Analysis and Management, 29(3), 553-577.
https ://doi .org/10.1002/pam.20509.
[22] Bel, G., & Mur, M . (2009). Intermunicipal cooperation, privatization and waste
management costs: Evidence from rural municipalities. Waste Management, 29(10),
2772-2778. https://doi.Org/10.1016/i.wasman.2009.06.002.
[23] Bel, G. & Warner, M . (2008). Does privatization of solid waste and water services
reduce costs? A review of empirical studies. Resources, Conservation and Recycling,
52(12), 1337-1348. https://doi.Org/10.1016/i.resconrec.2008.07.014.
[24] Belton, V., Crowe, D. V., Matthews, R. & Scott, S. (1994). A survey of public attitudes
to recycling in Glasgow (UK). Waste Management & Research, 12(4), 351-367.
https ://dx.doi.org/10.1006/wmre. 1994.1024.
[25] Benitez, S. O., Lozano-Olvera, G., Morelos, R. A. & Vega, C. A. D. (2008).
Mathematical modeling to predict residential solid waste generation. Waste
Management, 28(S1), S7-S13. https://doi.Org/10.1016/i.wasman.2008.03.020.
[26] Benito-Lopez, B., del Rocio Moreno-Enguix, M . & Solana-Ibanez, J. (2011).
Determinants of efficiency in the provision of municipal street-cleaning and refuse
collection services. Waste Management, 31(6), 1099-1108.
https ://doi.org/10.1016/i. wasman.2011.01.019.
[27] Bilitewski, B. (2008). From traditional to modern fee systems. Waste Management,
28(12), 2760-2766. https://dx.doi.Org/10.1016/i.wasman.2008.03.032.
169
[28] Bosch, N., Pedraja, F. & Suárez-Pandiello, J. (2000). Measuring the Efficiency of
Spanish Municipal Refuse Collection Services. Local Government Studies, 26(3), 71-
90. https://doi.org/10.1080/03003930008434000.
[29] Brown, D. P. (2015). Garbage: How population, landmass, and development interact
with culture in the production of waste. Resources, Conservation and Recycling. 98, 41-
54. https://doi.org/10.1016/i.resconrec.2015.02.012.
[30] Bucciol, A., Montinari, N. & Piovesan, M . (2015). Do not trash the incentive! Monetary
incentives and waste sorting. The Scandinavian Journal of Economics, 117(4), 1204-
1229. https://dx.doi.org/10.1111/sjoe. 12122.
[31] Byrne, S. & O'Regan, B. (2014). Attitudes and actions towards recycling behaviours in
the Limerick, Ireland region. Resources, Conservation and Recycling, 87, 89-96.
https://dx.doi.Org/10.1016/i.resconrec.2014.03.001.
[32] CENIA (2017). Celkové výdaje na ochranu životního prostředí.
https://issar.cenia.cz/prehled-klicovych-indikatoru-podle-hlavnich-temat/financovani-
ochrany-zivotniho-prostredi/celkove-vydaje-na-ochranu-zivotniho-prostredi/.
[33] Charnes, A., Cooper, W. W. & Rhodes, E. (1978). Measuring the efficiency of decision
making units. European Journal of Operational Research, 2(6), 429-444.
https://doi.org/10.1016/0377-2217(78)90138-8.
[34] Chen, M . F. & Tung, P. J. (2010). The moderating effect of perceived lack of facilities
on consumers' recycling intentions. Environment and Behavior, 42(6), 824-844.
https://dx.doi.org/10.1177/0013916509352833.
[35] Coelli, T. J., Rao, D. S. P., O'Donnell, C. J. & Battese, G. E. (2005). An Introduction to
Efficiency and Productivity Analysis. New York: Springer, 349 p. ISBN 978-0-387-
25895-9. https://dx.doi.org/10.1007/bl36381.
[36] Coli, M., Nissi, E. & Rapposelli, A. (2011). Monitoring environmental efficiency: An
application to Italian provinces. Environmental Modelling & Software, 26(1), 38-43.
https ://doi.org/10.1016/i .envsoft.2010.07.004.
[37] Cook, W. D. & Seiford, L. M . (2009). Data Envelopment Analysis (DEA) - Thirty
Years On. European Journal of Operational Research, 192(1), 1-17.
https://doi.Org/10.1016/i.eior.2008.01.032.
[38] Cooper, W. W., Seiford, L. M . & Tone, K. (2006). Introduction to Data Envelopment
Analysis and Its Uses. Springer-Verlag, 354 p. ISBN 978-0-387-28580-1.
https ://doi .org/10.1007/0-387-29122-9.
[39] CZSO (2018). Czech Statistical Office - various data, https://www.czso.cz.
[40] Dahlen, L. & Lagerkvist, A. (2010). Evaluation of recycling programmes in household
waste collection systems. Waste Management & Research 28(7), 577-586.
https://dx.doi.org/10.1177/0734242X09341193.
[41] Dahlen, L., Vukicevic, S., Meijer, J. E. & Lagerkvist, A. (2007). Comparison of
different collection systems for sorted household waste in Sweden. Waste Management.
27(10), 1298-1305. https://dx.doi.Org/10.1016/i.wasman.2006.06.016.
[42] Daraio, C. & Simar, L. (2005). Introducing environmental variables in nonparametric
frontier models: a probabilistic approach. Journal ofproductivity analysis, 24(1), 93-
121. https://doi.org/10.1007/sl 1123-005-3042-8.
[43] Daskalopoulos, E., Badr, O. & Probert, S. D. (1998). Municipal solid waste: a
prediction methodology for the generation rate and composition in the European Union
countries and the United States of America. Resources, Conservation and Recycling,
24(2), 155-166. https://doi.org/10.1016/S0921-3449(98)00032-9.
[44] De Jaeger, S., Eyckmans, J., Rogge, N. & Van Puyenbroeck, T. (2011). Wasteful wastereducing
policies? The impact of waste reduction policy instruments on collection and
processing costs of municipal solid waste. Waste Management, 31(7), 1429-1440.
https ://doi.org/10.1016/i.wasman.2011.02.021.
[45] Debreu, G. (1951). The Coefficient of Resource Utilizations. Econometrica, 19(3), 273-
292. https://dx.doi.org/10.2307/1906814.
[46] Dennison, G. J., Dodd, V. A. & Whelan, B. (1996). A socio-economic based survey of
household waste characteristics in the city of Dublin, Ireland. II. Waste quantities.
Resources, Conservation and Recycling, 17(3), 245-257. https://doi.org/10.1016/0921 -
3449(96)01155-X.
[47] Derksen, L. & Gartrell, J. (1993). The social context of recycling. American
Sociological Review, 58(3), 434-442. https://dx.doi.org/10.2307/2095910.
171
[48] Dijkgraaf, E. & Gradus, R. (2007). Collusion in the Dutch waste collection market.
Local Government Studies, 33(4), 573-588.
https ://doi.org/10.1080/03003930701417601.
[49] Dijkgraaf, E. & Gradus, R. (2004). Cost savings in unit-based pricing of household
waste: The case of The Netherlands. Resource and Energy Economics, 26(4), 353-371.
https://dx.doi.Org/10.1016/i.reseneeco.2004.01.001.
[50] Dijkgraaf, E. & Gradus, R. (2003). Cost savings of contracting out refuse collection.
Empirica, 30(2), 149-161. https://doi.org/10.1023/A: 1024175730230.
[51] Domberger, S., Meadowcroft, S. A. & Thompson, D. J. (1986). Competitive tendering
and efficiency: the case of refuse collection. Fiscal Studies, 7(4), 69-87.
https://doi.Org/10.l 111/j. 1475-5890.1986.tb00530.x.
[52] Domina, T. & Koch, K. (2002). Convenience and Frequency of Recycling Implications
for Including Textiles in Curbside Recycling Programs. Environment and Behavior,
34(2), 216-238. https://doi.org/l0.1177/0013916502342004.
[53] Dubin, J. A. & Navarro, P. (1988). How markets for impure public goods organize: the
case of household refuse collection. Journal of Law, Economics & Organization, 4(2),
217-241. http://www.istor.org/stable/764922.
[54] Dyson, R. G., Allen, R., Camanho, A. S., Podinovski, V. V., Sarrico, C. S. & Shale, E.
A. (2001). Pitfalls and protocols in DEA. European Journal of Operational Research,
132(2), 245-259. https://doi.org/10.1016/S0377-2217(00)00149-1.
[55] Dyson, B. & Chang, N. B. (2005). Forecasting municipal solid waste generation in a
fast-growing urban region with system dynamics modelling. Waste Management, 25(7),
669-679. https://doi.org/ 10.1016/i.wasman.2004.10.005.
[56] Emrouznejad, A., Parker, B. R. & Tavares, G. (2008). Evaluation of research in
efficiency and productivity: A survey and analysis of the first 30 years of scholarly
literature in DEA. Journal of Socio-Economic Planning Science, 42(3), 151-157.
https://doi.Org/10.1016/i.seps.2007.07.002.
[57] Emrouznejad, A. & Yang, G. L. (2018). A survey and analysis of the first 40 years of
scholarly literature in DEA: 1978-2016. Socio-Economic Planning Sciences, 61, 4-8.
https://doi.Org/10.1016/i.seps.2017.01.008.
[58] Eurostat (2015). Municipal waste generation and treatment, by type of treatment
method. ec.europa.eu/eurostat/web/products-datasets/-/tsdpc240.
[59] Farrell, M . J. (1957). The Measurement of Technical Efficiency. Journal of Royal
Statistical Society, Series A, General, 120, Part 3, 253-281.
https ://dx.doi.org/10.2307/2343100.
[60] Fiala, P. (2007). Analýza efektivnosti odpadového hospodářství českých obcí metodou
DEA. In Sauer, P. (ed.) Příspěvky k metodologii ex-post analýz v politice životního
prostředíll. Praha: SEPTIMtisk, 87-105, ISBN 978-80-904032-0-8.
[61] Folz, D. H. (1999). Municipal recycling performance: A public sector environmental
success story. Public Administration Review, 59(4), 336-345.
https://dx.doi.org/10.2307/3110116.
[62] Folz, D. H. (1991). Recycling program design, management, and participation: a
national survey of municipal experience. Public Administration Review, 51(3), 222-231.
https://dx.doi.org/10.2307/976946.
[63] Forsund, F. R. (2018). Economic interpretations of DEA. Socio-Economic Planning
Sciences, 61, 9-15. https://doi.Org/10.1016/i.seps.2017.03.004.
[64] Forsund, F. R. & Sarafoglou, N. (2002). On the Origins of Data Envelopment Analysis.
Journal of Productivity Analysis, 17, 23-40. https://doi.org/10.1023/A: 1013519902012.
[65] Fried, H. O., Lovell, C. A. K. & Schmidt, S. S. (eds.) (2008). The Measurement of
Productive Efficiency and Productivity Growth. Oxford University Press, 638 p., ISBN
978-0-19-518352-8. http://dx.doi.Org/10.1093/acprof:oso/9780195183528.001.0001.
[66] Fullerton, D. & Kinnaman, T. C. (1996). Household responses to pricing garbage by the
bag. The American Economic Review, 86(4), 971-984.
http://www.jstor.org/stable/2118314.
[67] Garcia-Sanchez, I. M . (2008). The performance of Spanish solid waste collection.
Waste Management & Research, 26(4), 327-336.
https://doi.org/10.1177/0734242X07081486.
[68] Gattoufi, S., Oral, M . & Reisman, A. (2004a). A taxonomy for data envelopment
analysis, Socio-Economic Planning Sciences, 38(2-3), 141-158.
https ://doi.org/10.1016/S0038-0121 (03)00022-3.
[69] Gattoufi, S., Oral, M . & Reisman, A. (2004b). Data Envelopment Analysis Literature: A
Bibliography Update (1951-2001). Socio-Economic Planning Sciences, 38(2-3), 159-
229. http://dx.doi.org/10.1016/S0038-0121(03)00023-5.
[70] Gellynck, X., Jacobsen, R. & Verhelst, P. (2011). Identifying the key factors in
increasing recycling and reducing residual household waste: A case study of the
Flemish region of Belgium. Journal of Environmental Management, 92(10) , 2683-
2690. https://doi.org/10.1016/i.ienvman.2011.06.006.
[71] Gillespie, R. & Bennett, J. (2011). Willingness to pay for kerbside recycling the
Brisbane Region, EERH Research Report No. 97, Available at
crawford.anu.edu.au/research units/eerh/pdf/EERH RR97.pdf.
[72] Gomez-Lobo, A. & Szymanski, S. (2001). A law of large numbers: bidding and
compulsory competitive tendering for refuse collection contracts. Review of Industrial
Organization, 18(1), 105-113. https://doi.Org/10.1023/A:1026545306923.
[73] Gonzales-Torre, P. L. & Adenso-Diaz, B. (2005). Influence of distance on the
motivation and frequency of household recycling. Waste Management, 25(1), 15-23.
https ://dx.doi.org/10.1016/i.wasman.2004.08.007.
[74] Gradus, R., Dijkgraaf, E. & Wassenaar, M . (2014). Understanding Mixed Forms of
Refuse Collection, Privatization, and Its Reverse in the Netherlands. International
Public Management Journal, 17(3), 328-343.
https ://doi.org/10.1080/10967494.2014.935237.
[75] Greene, W. H. (2008). The Econometric Approach to Efficiency Analysis. In Fried, H.
O. et al. (eds.) The Measurement of Productive Efficiency and Productivity Growth.
Oxford University Press, 92-250, ISBN 978-0-19-518352-8.
http://dx.doi.Org/10.1093/acprof:oso/9780195183528.001.0001.
[76] Guerrero, L. A., Maas, G.,& Hogland, W. (2013). Solid waste management challenges
for cities in developing countries. Waste management, 33(1), 220-232.
https://doi.Org/10.1016/i.wasman.2012.09.008.
[77] Guerrini, A., Carvalho, P., Romano, G , Marques, R. C. & Leardini, C. (2017).
Assessing efficiency drivers in municipal solid waste collection services through a non-
174
parametric method. Journal of Cleaner Production, 147, 431-441.
https ://doi.org/10.1016/i .iclepro.2017.01.079.
[78] Hage, O. & Sóderholm, P. (2008). An econometric analysis of regional differences in
household waste collection: the case of plastic packaging waste in Sweden. Waste
Management, 28(10), 1720-1731. https://doi.Org/10.1016/i.wasman.2007.08.022.
[79] Hage, O., Sóderholm, P. & Berglund, C. (2009). Norms and economic motivation in
household recycling: empirical evidence from Sweden. Resources, Conservation and
Recycling, 53(3), 155-165. https://dx.doi.Org/10.1016/i.resconrec.2008.ll.003.
[80] Halvorsen, B. (2012). Effects of norms and policy incentives on household recycling:
An international comparison. Resources, Conservation and Recycling, 67, 18-26.
https://dx.doi.Org/10.1016/i.resconrec.2012.06.008.
[81] Hillman, A. (2009). Public Finance and Public Policy: Responsibilities and Limitations
of Government. Cambridge University Press, 859 p. ISBN 978-0-521-49426-7.
https://doi.org/10.1017/CBO9780511813788.
[82] Hockett, D., Lober, D. J. & Pilgrim, K. (1995). Determinants of per capita municipal
solid waste generation in the Southeastern United States. Journal of Environmental
Management, 45(3), 205-217. https://doi.org/10.1006/iema.1995.0069.
[83] Hoff, A. (2007). Second stage DEA: Comparison of approaches for modelling the DEA
score. European Journal of Operational Research, 181(1), 425-435.
https ://doi.org/10.1016/i .eior.2006.05.019.
[84] Honma, S. (2014). Environmental and Economic Efficiencies in the Asia-Pacific
Region, Journal of Asia-Pacific Business, 15(2), 122-135.
https://doi.org/10.1080/10599231.2014.904185.
[85] Hornik, J., Cherian, J., Madansky, M . & Narayana, C. (1995). Determinants of
recycling behavior: A synthesis of research results. The Journal of Socio-Economics,
24(1), 105-127. https://doi.org/10.1016/1053-5357(95)90032-2.
[86] Hřebíček, J., Kalina, J. & Soukopová, J. (2013). Integrated economic model of waste
management: Case study for South Moravia region. Acta Universitatis Agriculturae et
Silviculturae Mendelianae Brunensis, 61(4), 917-922.
https://doi.Org/10.l 1118/actaun201361040917.
[87] Hřebíček, J. & Soukopová, J. (2017). Municipal Waste Strategy for Approaching
European Union Recycling Targets. In Cossu, R., Stegmann, R.. Proceedings Sardinia
2017, 16th International Waste Management and Landfill Symposium. Sardinia, Italy:
CISA Publisher, 1-24. ISBN 978-88-6265-010-6.
[88] Huang, Y., Pan, T. & Kao, J. (2011). Performance assessment for municipal solid waste
collection in Taiwan. Journal of Environmental Management, 92(4), 1277-1283.
https ://doi.org/10.1016/i .ienvman.2010.12.002.
[89] ISOH (2018). Informační Systém Odpadového Hospodářství, https://isoh.mzp.cz.
[90] Jablonský, J. (2004). Modely hodnocení efektivnosti produkčních jednotek. Politická
ekonomie, 52(2), 206-219. https://doi.Org/10.18267/j.polek.458.
[91] Jablonský, J. (2002). Operační výzkum: kvantitativní modely pro ekonomické
rozhodování. Praha: Professional Publishing, 322 p., ISBN 80-86419-42-8.
[92] Jablonský, J. & Dlouhý, M . (2015). Modely hodnocení efektivnosti a alokace zdrojů.
Praha: Professional Publishing, 199 p. ISBN 978-80-7431-155-0.
[93] Jablonský, J. & Dlouhý, M . (2004). Modely hodnocení efektivnosti produkčních
jednotek. Praha: Professional Publishing, 183 p. ISBN 978-80-86419-49-5.
[94] Jacobsen, R., Buysse, J. & Gellynck, X. (2013). Cost comparison between private and
public collection of residual household waste: multiple case studies in the Flemish
region of Belgium. Waste Management, 33(1), 3-11.
https ://doi.org/10.1016/i.wasman.2012.08.015.
[95] Jihomoravský kraj (2018). Základní údaje o Jihomoravském kraji, https://www.kr-
jihomoravsky.cz/Default.aspx?ID=27204&TypeID=2.
[96] Johnstone, N . & Labonne, J. (2004). Generation of household solid waste in OECD
countries: an empirical analysis using macroeconomic data. Land Economics, 80(4),
529-538. https://doi.org/10.2307/3655808.
[97] Kinnaman, T. C. & Fullerton, D. (1999). The economics of residential solid waste
management. NBER Working Paper No. 7326. National bureau of economic research.
https://doi.org/10.3386/w7326.
[98] Kirakozian, A. (2015). Household Waste Recycling: Economics and Policy. GREDEG
Working Paper Series, No. 2015-09. https://ideas.repec.Org/p/gre/wpaper/2015-09.html.
[99] Kirkpatrick, C , Parker, D. & Zhang, Y. (2006). State versus private sector provision of
water services in Africa: An empirical Analysis, The World bank economic review,
20(1), 143-163. https://doi.org/10.1093/wber/lhiOOl.
[100] Knussen, C , Yule, F., MacKenzie, J. & Wells, M . (2004). An analysis of intentions to
recycle household waste: The roles of past behaviour, perceived habit, and perceived
lack of facilities. Journal of Environmental Psychology, 24(2), 237-246.
https://dx.doi.Org/10.1016/i.ienvp.2003.12.001.
[101] Knussen, C. & Yule, F. (2008). "I'm Not in the Habit of Recycling" The Role of
Habitual Behavior in the Disposal of Household Waste. Environment and Behavior,
40(5), 683-702. https://dx.doi.org/10.1177/0013916507307527.
[102] Koopmans, T. C. (1951). An Analysis of Production as an Efficient Combination of
Activities. In Koopmans, T. C. (ed.) Activity Analysis of Production and Allocation,
Cowles Commission for Research in Economics. Monograph No. 13, New York: Wiley,
33-97.
[103] Korhonen, P. & Luptacik, M . (2004). Eco-efficiency analysis of power plants: An
extension of data envelopment analysis, European Journal of Operational Research,
154(2), 437-446. https://doi.Org/l0.1016/S0377-2217(03)00180-2.
[104] Kumbhakar, S. C. & Lovell, C. A. K. (2000). Stochastic Frontier Analysis. Cambridge:
Cambridge University Press, 333 p. ISBN 0-521-66663-5.
https ://doi .org/10.1017/cbo9781139174411.
[105] Kuosmanen, T. & Kortelainen, M . (2005). Measuring Eco-efficiency of Production with
Data Envelopment Analysis. Journal of Industrial Ecology, 9(4), 59-72.
https://doi.org/10.1162/108819805775247846.
[106] Lebersorger, S. & Beigl, P. (2011). Municipal solid waste generation in municipalities:
Quantifying impacts of household structure, commercial waste and domestic fuel.
Waste Management, 31(9), 1907-1915.
https ://dx.doi.org/10.1016/i.wasman.2011.05.016.
[107] Linderhof, V., Kooreman, P., Alters, M . & Wiersma, D. (2001). Weight-based pricing
in the collection of household waste: the Oostzaan case. Resource and Energy
Economics, 23(4), 359-371. https://dx.doi.org/10.1016/S0928-7655(01)00044-6.
177
[108] Liu, J. S., Lu, L. Y., Lu, W. M . & Lin, B. J. (2013). Data envelopment analysis 1978-
2010: A citation-based literature survey. Omega, 41(1), 3-15.
https ://doi.org/10.1016/i .omega.2010.12.006.
[109] Lovell, C. A. K. (2002). Performance Assessment in the Public Sector. In Fox, K. J.
(ed.) Efficiency in the Public Sector. Norwell: Kluwer Academic Publishers, 11-35,
ISBN 0-7923-7563-7. https://dx.doi.org/10.1007/978-l-4757-3592-5.
[110] Marques, R. C. & Simoes, P. (2009). Incentive regulation and performance
measurement of the Portuguese solid waste management services. Waste Management
& Research, 27(2), 188-196. https://doi.org/10.1177/0734242X08095025.
[Ill] Martin, M., Williams, I. D. & Clark, M . (2006). Social, cultural and structural
influences on household waste recycling: A case study. Resources, Conservation and
Recycling, 48(4), 357-395. https://dx.doi.Org/10.1016/i.resconrec.2005.09.005.
[112] Massarutto, A. (2007). Municipal waste management as a local utility: Options for
competition in an environmentally-regulated industry. Utilities Policy, 15(1), 9-19.
https://doi.Org/10.1016/i.iup.2006.09.003.
[113] Mastromarco, C. (2008). Stochastic Frontier Models. CUE 2008, Salento: University of
Salento, Available at: http://www.camillamastromarco.it/CIDE/STFR.pdf.
[114] Mazzanti, M . & Zoboli, R. (2008). Waste generation, waste disposal and policy
effectiveness: Evidence on decoupling from the European Union. Resources,
Conservation and Recycling, 52(10), 1221-1234.
https ://dx.doi.org/l 0.1016/i .resconrec.2008.07.003.
[115] McCarty, J. A. & Shrum, L. J. (1994). The recycling of solid wastes: Personal values,
value orientations, and attitudes about recycling as antecedents of recycling behavior.
Journal of Business Research, 30(1), 53-62. https://dx.doi.org/10.1016/0148-
2963(94)90068-X.
[116] Meadows, D. H., Meadows, D. L , Randers, J. & Behrens, W. W. (1972). The limits to
growth. New York: Universe books, 205 p. ISBN 0-87663-165-0.
[117] Medved', J., Nemec, J., Orviska, M . & Zimkova, E. (2005). Verejne financie.
Bratislava: Sprint, 464 p. ISBN 80-89085-32-6.
178
[118] Mee, N., Clewes, D., Phillips, P. S. & Read, A. D. (2004). Effective implementation of
a marketing communications strategy for kerbside recycling: a case study from
Rushcliffe, UK. Resources, Conservation and Recycling, 42(1), 1-26.
https ://dx.doi.org/10.1016/i .resconrec.2003.12.003.
[119] Meeusen, W. & Van den Broeck, J. (1977). Efficiency Estimation from Cobb-Douglas
Production Functions with Composed Error. International Economic Review, 18(2),
435-444. https://dx.doi.org/10.2307/2525757.
[120] Merickova, B. M., Nemec, J. & Soukopova, J. (2014). The Economics of Waste
Management: Evidence from the Czech Republic and Slovakia. Lex Localis, 12(3), 431.
https://doi.Org/10.4335/12.3.431-449(2014).
[121] Mihaiu, D. M., Opreana, A. & Cristescu, M . P. (2010). Efficiency, effectiveness and
performance of the public sector. Romanian Journal of Economic Forecasting, 13(4),
132-147. https://ideas.repec.Org/a/rjr/romjef/vy2010i4p 132-147.html.
[122] Mikusovä Merickovä, B., Nemec, J. & Vozärovä, Z. (2017). The Efficiency of
Contracting Out Local Public Services in the Czech Republic and Slovakia. In Douglas
J. Besharov, Karen J. Beahler, Jacob Alex Klerman. Improving Public Services. New
York: Oxford University Press. 265-285, ISBN 978-0-19-064605-9.
https ://doi.org/10.1093/acprof :oso/9780190646059.003.0012.
[123] Miliute-Plepiene, J., Hage, O., Plepys, A. & Reipas, A. (2016). What motivates
households recycling behaviour in recycling schemes of different maturity? Lessons
from Lithuania and Sweden. Resources, Conservation and Recycling, 113, 40-52.
https://dx.doi.Org/10.1016/i.resconrec.2016.05.008.
[124] Mintzberg, H. (1996). Managing Government, Government Managed. Harvard
Business Review, 74(3), 75-83.
[125] Moore, A., Nolan, J. & Segal, G. F. (2005). Putting out the trash: Measuring municipal
service efficiency in US cities. Urban Affairs Review, 41(2), 237-259.
https://doi.org/10.1177/1078087405279466.
[126] MONITOR (2018). Information portal of the Ministry of Finance of the Czech
Republic, http://monitor.statnipokladna.cz/.
179
[127] Musgrave, R. A. & Musgrave, P. B. (1984). Public finance in theory and practice. New
York: McGraw-Hill, 824 p., ISBN 0-07-044126-X.
[128] MZP (2017a). Environmental Protection in the Czech Republic: Experience and KnowHow,
htt^si/ywvvwur^
4b90fd93a463b9f7cl2581bd002d701d.
[129] MŽP (2017b). Souhrnná data o odpadovém hospodářství ČR v letech 2009 - 2016.
https://www.mzp.cz/C1257458002F0DC7/cz/odpadove hospodářství data 2015/SFIL
E/QQDP-Souhrnna data 2009 2016-20171006.pdf.
[130] MZP (2017c). Statistická ročenka životního prostředí České republiky 2016.
https://www.mzp.cz/cz/statisticka ročenka životního prostredí publikace.
[131] MZP (2017d). Zpráva o životním prostředí České republiky v roce 2016.
https://www.mzp.cz/cz/zpravy o stavu životního prostredí publikace.
[132] Nemec, J., Meričková, B. & Šumpíková, M . (2007). Contracting out at local
government level in Slovakia and the Czech Republic. E+ M Economics and
Management, 10(3), 5-13.
[133] Nemec, J., Merickova, B. & Vítek, L. (2005). Contracting-out at local government
level: theory and selected evidence from the Czech and Slovak Republics. Public
Management Review, 7(4), 637-647. https://doi.org/10.1080/14719030500362900.
[134] Nemec, J., Ochrana, F., Pavel, J. & Sagát, V. (2010). Kontrola ve veřejné správě. Praha:
Wolters Kluwer ČR, 160 p., ISBN 978-80-7357-558-8.
[135] Ochrana, F. (2011). Veřejné výdajové programy, veřejné projekty a zakázky. Praha:
Wolters Kluwer ČR, 220 p., ISBN 978-80-7357-644-8.
[136] Ochrana, F. (2006). Programové financování a hodnocení veřejných výdajů : teorie a
metodika hodnocení veřejných výdajů a veřejných služeb v systému programové alokace
zdrojů. Praha: Ekopress, 189 p., ISBN 80-86929-13-2.
[137] Ochrana, F. (2005). Nákladově užitkové metody ve veřejném sektoru. Praha: Ekopress,
175 p., ISBN 80-86119-96-3.
[138] Ochrana, F. (2001). Veřejný sektor a efektivní rozhodování. Praha: Management Press,
246 p. ISBN 80-7261-018-X.
180
[139] Ochrana, F., Fantová Šumpíková, M., Pavel, J., Němec, J. et al. (2007). Efektivnost
zabezpečování vybraných veřejných služeb na úrovni obcí. Praha: Nakladatelství
Oeconomica, 152 p., ISBN 978-80-245-1259-4.
[140] Ohlsson, H. (2003). Ownership and production costs: choosing between public and
private production and contracting-out in the case of Swedish refuse collection. Fiscal
Studies, 24(4), 451-476. https://doi.Org/10.l 11 l/i.l475-5890.2003.tb00091.x.
[141] Oskamp, S., Harrington, M . J., Edwards, T. C , Sherwood, D. L., Okuda, S. M . &
Swanson, D. C. (1991). Factors influencing household recycling behavior. Environment
and Behavior, 23(4), 494-519. https://dx.doi.org/10.1177/0013916591234005.
[142] Passarini, F., Vassura, L, Monti, F., Morselli, L. & Villani, B. (2011). Indicators of
waste management efficiency related to different territorial conditions. Waste
management, 31(4), 785-792. https://doi.org/10.1016/i.wasman.2010.11.021.
[143] Perrin, D. & Barton, J. (2001). Issues associated with transforming household attitudes
and opinions into materials recovery: a review of two kerbside recycling schemes.
Resources, Conservation and Recycling, 33(1), 61-74.
https ://dx.doi.org/10.1016/S0921 -3449(01 )00075-1.
[144] Pickerin, P. L. & Shaw, P. J. (2015). Age and ageing as factors influencing waste
management behaviour. In R. Cossu, P. He, P. Kjeldsen, Y. Matsufuji, D. Reinhart, R.
Stegman (eds.) SARDINIA 2015 - 15th International Waste Management and Landfill
Symposium Proceedings, Padova, CISA. ISBN 9788862650212.
[145] Porcelli, F. (2009). Measurement of Technical Efficiency - A brief survey on parametric
and non-parametric techniques, University of Warwick, www2.warwick.ac.uk/
fac/soc/economics/staff/phd students/porcelli/porcelli dea sfm.pdf.
[146] Poldnurk, J. (2015). Optimisation of the economic, environmental and administrative
efficiency of the municipal waste management model in rural areas. Resources,
Conservation and Recycling, 97, 55-65.
https ://doi.org/10.1016/i .resconrec.2015.02.003.
[147] Ramanathan, R. (2003). An Introduction to Data Envelopment Analysis: A toolfor
performance measurement. New Delhi: Sage Publications India, 201 p. ISBN 0-7619-
9760-1.
181
[148] Ray, S. C. (2004). Data envelopment analysis: theory and techniques for economics and
operations research, Cambridge University Press, 353 p. ISBN 0-521-80256-3.
[149] Ray, S. C. (1988). Data envelopment analysis, nondiscretionary inputs and efficiency:
an alternative interpretation. Socio-Economic Planning Sciences, 22(4), 167-176.
https ://doi.org/10.1016/0038-0121 (88)90003-1.
[150] Reeves, E. & Barrow, M . (2000). The impact of contracting-out on the costs of refuse
collection services. The case of Ireland. The Economic and Social Review, 31(2), 129—
150. https://ideas.repec.org/a/eso/journl/v31 v2000i2p 129-150.html.
[151] Richmond, J. (1974). Estimating the Efficiency of Production. International Economic
Review, 15(2), 515-521. https://doi.org/10.2307/2525875.
[152] Rogge, N. & De Jaeger, S. (2012). Evaluating the efficiency of municipalities in
collecting and processing municipal solid waste: A shared input DEA-model. Waste
Management, 32(10), 1968-1978. https://doi.Org/10.1016/i.wasman.2012.05.021.
[153] Rogge, N. & De Jaeger, S. (2013). Measuring and explaining the cost efficiency of
municipal solid waste collection and processing services. Omega, 41(4), 653-664.
https://doi.Org/10.1016/i.omega.2012.09.006.
[154] Russell, R. R. (1998). Distance Functions in Consumer and Producer Theory. In Fare,
R. et al. (eds.) Index Numbers: Essays In Honour ofSten Malmquist. Springer
Netherlands, 7-90. ISBN 978-94-010-6035-6. https://doi.org/10.1007/978-94-011-4858-
0 2.
[155] Saphores, J. D. M . & Nixon, H. (2014). How effective are current household recycling
policies? Results from a national survey of US households. Resources, Conservation
and Recycling, 92, 1-10. https://dx.doi.Org/10.1016/i.resconrec.2014.08.010.
[156] Schultz, P. W., Oskamp, S. & Mainieri, T. (1995). Who recycles and when? A review
of personal and situational factors. Journal of Environmental Psychology, 15(2), 105-
121. https://dx.doi.org/10.1016/0272-4944(95)90019-5.
[157] Seiford, L. M . (2005). A Cyber-Bibliography for Data Envelopment Analysis (1978-
2005). Available at http://en.convdocs.org/docs/index-8834.html.
182
[158] Shaw, P. J. (2008). Nearest neighbour effects in kerbside household waste
recycling. Resources, Conservation and Recycling, 52(5), 775-784.
https://dx.doi.Org/10.1016/i.resconrec.2007.ll.004.
[159] Shaw, P. J. & Maynard, S. J. (2008). The potential of financial incentives to enhance
householders' kerbside recycling behaviour. Waste Management, 28(10), 1732-1741.
https ://dx.doi.org/10.1016/i.wasman.2007.08.008.
[160] Sidique, S. F., Joshi, S. V. & Lupi, F. (2010). Factors influencing the rate of recycling:
An analysis of Minnesota counties. Resources, Conservation and Recycling, 54(4), 242-
249. https://dx.doi.Org/10.1016/i.resconrec.2009.08.006.
[161] Simar, L. & Wilson, P. W. (2011). Two-stage DEA: caveat emptor. Journal of
Productivity Analysis, 36(2), 205-218. https://doi.org/10.1007/sl 1123-011-0230-6.
[162] Simoes, P., Cruz, N. F. & Marques, R. C. (2012). The performance of private partners
in the waste sector. Journal of Cleaner Production, 29, 214-221.
https ://doi.org/10.1016/i .iclepro.2012.01.027.
[163] Simoes, P., De Witte, K., & Marques, R. C. (2010). Regulatory structures and
operational environment in the Portuguese waste sector. Waste management, 30(6),
1130-1137. https://doi.Org/10.1016/i.wasman.2009.12.015.
[164] Simoes, P. & Marques, R. C. (2012). On the economic performance of the waste sector.
A literature review. Journal of Environmental Management, 106, 40-47.
https ://doi.org/10.1016/i .ienvman.2012.04.005.
[165] Sivák, R. et al. (2007). Veřejné financie. Bratislava: Iura Edition, 301 p. ISBN 978-80-
8078-094-4.
[166] Slavík, J. & Pavel, J. (2013). Do the variable charges really increase the effectiveness
and economy of waste management? A case study of the Czech Republic. Resources,
Conservation and Recycling, 70, 68-77.
https://dx.doi.Org/10.1016/i.resconrec.2012.09.013.
[167] Soukopová, J., Hřebíček, J. & Valta, J. (2015). National Environmental Data Facilities
and Services of the Czech Republic and Their Use in Environmental Economics. In
Denzer R, Argent R.M., Schimak G., Hřebíček J. (eds.) Environmental Software
Systems. Infrastructures, Services and Applications. ISESS 2015. IFIP Advances in
Information and Communication Technology, vol. 448, 361-370.
https://doi.org/l0.1007/978-3-319-15994-2 36.
[168] Soukopová, J. & Kalina, J. (2012). Mathematical model of economics of municipal
waste management. In Jaroslav Ramik and Daniel Stavařek. Proceedings of the 30th
International Conference Mathematical Methods in Economics 2012, Karviná: Silesian
University in Opava, School of Business Administration in Karviná, 823-829.
ISBN 978-80-7248-779-0.
[169] Soukopová, J., Malý, I, Hřebíček, J. & Struk, M . (2013). Decision Support of Waste
Management Expenditures Efficiency Assessment. In Hřebíček J., Schimak G, Kubásek
M., Rizzoli A.E. (eds.) Environmental Software Systems. Fostering Information Sharing.
ISESS 2013. IFIP Advances in Information and Communication Technology, vol. 413,
651-660. https://doi.org/10.1007/978-3-642-41151-9 61.
[170] Soukopová, J., Nemec, J., Matějova, L. & Struk, M . (2014). Municipality Size and
Local Public Services: Do Economies of Scale Exist? NISPAcee Journal of Public
Administration and Policy, 7(2), 151-171. https://doi.org/10.2478/nispa-2014-0007.
[171] Soukopová, J., Struk, M. & Hřebíček, J. (2017). Population Age Structure and the Cost
of Municipal Waste Collection. A case study from the Czech Republic. Journal of
Environmental Management, 203(P2), 655-663.
https://dx.doi.Org/10.1016/i.ienvman.2016.03.030.
[172] Sterner, T. & Batterings, H. (1999). Household waste management in a Swedish
municipality: determinants of waste disposal, recycling and composting. Environmental
and Resource Economics, 13(4), 473-491. https://doi.Org/10.1023/A:1008214417099.
[173] Strecková, Yvonne a kol. (2005). Faktory efektivnostifungování veřejného sektoru a
obecné poznatky o vlivu řízení veřejného sektoru na rozvoj region. Brno: Masarykova
univerzita, 243 p. ISBN 80-210-3623-0.
[174] Struk, M . (2017a). Distance and incentives matter: The separation of recyclable
municipal waste. Resources, Conservation and Recycling, 122, 155-162.
https://dx.doi.Org/10.1016/i.resconrec.2017.01.023.
[175] Struk, M . (2017b). The Effect of Waste Management Company Change on Municipal
Solid Waste Expenditures. In Cossu, R., Stegmann, R. (eds.) Proceedings Sardinia
184
2017, 16th International Waste Management and Landfill Symposium. Sardinia, Italy:
CISA Publisher, 1-10. ISBN 978-88-6265-010-6.
[176] Struk, M . (2015a). Efficiency factors in municipal solid waste management. In Raffaello
Cossu, Pinjing He, Peter Kjeldsen, Yasushi Matsufuji, Debra Reinhart, Rainer Stegman
(eds.) SARDINIA 2015 - 15th International Waste Management and Landfill Symposium
- Symposium Proceedings. Padova, Italy: CISA, 2015, 1-13. ISBN 978-88-6265-021-2.
[177] Struk, M . (2015b). Estimating municipal solid waste expenditures - the case of South
Moravian Region. In Dagmar Spolková, Lenka Matějova (eds.) Proceedings of the 19th
International Conference Current Trends in Public Sector Research. Brno: Masarykova
univerzita, 324-331. ISBN 978-80-210-7532-0.
[178] Struk, M . & Matulová, M . (2016). The Application Of Two-Stage Data Envelopment
Analysis On Municipal Solid Waste Management In The Czech Republic. In Marián
Reiff, Pavel Gežík. Quantitative Methods in Economics; Multiple Criteria Decision
Making XVIII. Bratislava: University of Economics, 349-355. ISBN 978-80-972328-0-
1.
[179] Struk, M . & Soukopová, J. (2016). Age Structure and Municipal Waste Generation and
Recycling - New Challenge for the Circular Economy. In 4th International Conference
on Sustainable Solid Waste Management, 23rd - 25th June 2016 Limassol, Cyprus.
[180] Stuart-Hamilton, I. (1999). Key ideas in psychology. Jessica Kingsley Publishers, 336 p.
ISBN: 978-1-84642-702-2
[181] Sturrock (1957). Discussion on Mr. Farrell's Paper. Journal of Royal Statistical Society,
Series A, General, 120, Part 3, 285-285. https://dx.doi.org/10.2307/2343100.
[182] Szymanski S. (1996). The impact of compulsory competitive tendering on refuse
collection services. Fiscal studies, 17(3), 1-19. https://doi.Org/10.l 11 l/j.1475-
5890.1996.tb00491.x.
[183] Szymanski, S. & Wilkins, S. (1993). Cheap rubbish? Competitive tendering and
contracting out in refuse collection-1981-88. Fiscal studies, 14(3), 109-130.
https://doi.Org/10.l 111/i. 1475-5890.1993.tb00489.x.
185
[184] Šauer, P., Pařízková, L. & Hadrabová, A. (2008). Charging systems for municipal solid
waste: Experience from the Czech Republic. Waste Management, 28(12), 2772-2777'.
https://dx.doi.Org/10.1016/i.wasman.2008.03.030.
[185] Sumpíková, M., Ochrana, F., Pavel, J. a kol. (2005). Veřejné výdajové program a jejich
efektivnost. Praha: Eurolex Bohemia, 230 p., ISBN 80-86861-77-5.
[186] Talalaj, I. A. & Walery, M . (2015). The effect of gender and age structure on municipal
waste generation in Poland. Waste Management, 40, 3-8.
https://dx.doi.Org/10.1016/i.wasman.2015.03.020.
[187] Thanassoulis, E., Portela, M . C. S. & Despič, O. (2008). Data Envelopment Analysis:
The Mathemetical Programming Approach to Efficiency Analysis. In Fried H. O. et al.
(eds.) The Measurement of Productive Efficiency and Productivity Growth. Oxford
University Press, 251-420, ISBN 978-0-19-518352-8.
http://dx.doi.Org/10.1093/acprof:oso/9780195183528.001.0001.
[188] Thogersen, J. (2003). Monetary incentives and recycling: Behavioural and
psychological reactions to a performance-dependent garbage fee. Journal of Consumer
Policy. 26(2), 197-228. https://dx.doi.Org/10.1023/A:1023633320485.
[189] Vishwakarma, A., Kulshrestha, M . & Kulshreshtha, M . (2012). Efficiency evaluation of
municipal solid waste management utilities in the urban cities of the state of Madhya
Pradesh, India, using stochastic frontier analysis. Benchmarking: An International
Journal, 19(3), 340-357. https://doi.org/10.1108/14635771211242996.
[190] Vlček, J. (1992). Výkladový lexikon pojmů tržní ekonomiky. Praha: Victoria Publishing,
215 p., ISBN 80-85605-26-0.
[191] Vohralíková, L. & Rabušic, L. (2004). Čeští senioři včera, dnes a zítra. VUPSV,
Výzkumné centrum Brno. 90 p. http://praha.vupsv.cz/Fulltext/vz 149.pdf.
[192] Williams, I. D. & Taylor, C. (2004). Maximising household waste recycling at civic
amenity sites in Lancashire, England. Waste Management, 24(9), 861-874.
https://dx.doi.Org/10.1016/i.wasman.2004.02.002.
[193] Wilson, D. C , Velis, C. & Cheeseman, C. (2006). Role of informal sector recycling in
waste management in developing countries. Habitat international, 30(4), 797-808.
https ://doi.org/10.1016/i .habitatint.2005.09.005.
[194] Wilson, C. D. H. & Williams, I. D. (2007). Kerbside collection: A case study from the
north-west of England. Resources, Conservation and Recycling, 52(2), 381-394.
https://dx.doi.Org/10.1016/i.resconrec.2007.02.006.
[195] Woodbury, K. & Dollery, B. (2004). Efficiency Measurement in Australian Local
Government: The Case of New South Wales Municipal Water Services, Review of
Policy Research, 21(5), 615-636. https://doi.Org/10.llll/i.1541-1338.2004.00098.x.
[196] Worthington, A. C. (2014). A review of frontier approaches to efficiency and
productivity measurement in urban water utilities. Urban Water Journal, 11(1), 55-73.
https://doi.org/10.1080/1573062X.2013.765488.
[197] Worthington, A. C. & Dollery, B. E. (2001). Measuring efficiency in local government:
An analysis of New South Wales municipalities' domestic waste management function.
Policy Studies Journal, 29(2), 232-249. https://doi.org/10.11111\. 1541 -
0072.2001.tb02088.x.
[198] Závodská, A., Benešová, L., Smyth, B. & Morrissey, A. J. (2014). A comparison of
biodegradable municipal waste (BMW) management strategies in Ireland and the Czech
Republic and the lessons learned. Resources, Conservation and Recycling, 92, 136-144.
https://dx.doi.Org/10.1016/i.resconrec.2014.09.007.
[199] Zheng, P., Zhang, K., Zhang, S., Wang, R. & Wang, H. (2017). The door-to-door
recycling scheme of household solid wastes in urban areas: a case study from Nagoya,
Japan. Journal of Cleaner Production, 163, S366-S373.
https://dx.doi.Org/10.1016/i.iclepro.2016.03.106.
187
188
SUMMARY, LIST OF FIGURES, TABLES AND
ABBREVIATIONS
Summary
Provided thesis examines selected efficiency factors related to municipal waste management,
Objectives in this thesis are to 1) describe situation in municipal solid waste management in a
selected Czech region; and 2) identify relevant efficiency factors in this region using both
collected data and available literature, and based on further analysis provide policy
suggestions for improving municipal waste management in the region. After the initial
definition of the efficiency in the public sector, the target area of municipal waste
management in the Czech Republic was introduced. Relevant methods for measuring
efficiency were described and an overview of literature dealing with the issue was presented.
This overview resulted in an identification of the most common efficiency factors related to
municipal solid waste management. In addition to that, own collection of primary data from a
sample of more than 500 municipalities in South Moravian Region was performed, in order to
identify other potential efficiency factors relevant to municipal waste management. These
primary data were accompanied by available secondary data from public databases. Collected
data was then used in a quantitative analysis employing Two-stage Data Envelopment
Analysis, what provided additional efficiency factors that were considered as relevant with
respect to the municipal solid waste management. Based on literature overview and
quantitative analysis of the primary and secondary data, specific efficiency factors were
selected for the further detailed analysis. These included separation convenience factors
(availability of kerbside collection, density of drop-off sites), presence of incentives related to
the waste management, ages structure of the municipality, level of competition present in the
waste management provision and the nature of the waste management company ownership.
Analyses of these factors showed that both convenience and incentives play an important role
in the waste management and can lead to substantial improvement in both economic and
environmental terms. Age structure of a municipality is also an important factor to consider,
as various age cohorts can have notably different effect on waste generation and expenditure.
Proper utilization of these finding has a lot of potential in improving the overall performance
of municipal waste management. Finally, results also showed that not the ownership nature of
the waste company is important, but instead the sufficient level of competition among them.
189
Figures
Figure 1: Scheme of determining the specific efficiency indicator 22
Figure 2: Capital (investment) and current (non-investment) EPE in the Czech Republic, 2003-2015 29
Figure 3: Total waste generation in the Czech Republic, million tons, 2009-2016 31
Figure 4: Waste generation in the Czech Republic in 2016 by categories 32
Figure 5: Municipal waste generation in the Czech Republic, million tons, 2009-2016 34
Figure 6: Municipal waste treatment in the Czech Republic, million tons, 2010-2016 36
Figure 7: Histogram of Czech municipal solid waste expenditure per capita, 2016 39
Figure 8: Development of municipal environmental protection expenditure by categories, 2009-16 40
Figure 9: Development of municipal waste management expenditure by categories, 2009-16 41
Figure 10: Example of CRS and VRS efficiency frontier, case with 7 DMUs, 1 output, 1 input 92
Figure 11: Frequency distribution of calculated efficiency scores of municipalities 101
Figure 12: Frequency distribution of municipality population size of the sample, 2012 state 114
Figure 13: Average and median paper and plastics separation rates 118
Figure 14: Average amount of generated waste per capita according to the municipality size 121
Figure 15: Municipal solid waste expenditure per capita for 2011 and 2014 139
Figure 16: Interannual change in MSWE after the change of the W M company (66 municipalities), standard
errors for averages included 152
Figure 17: Relative changes in MSWE of individual municipalities after the change of W M company (66
municipalities) 153
Figure 18: Comparison of MSWE before and two years after the change of the waste management company (52
municipalities), standard errors for averages included 154
Figure 19: MSWE change according to the company ownership change (62 municipalities) 156
Tables
Table 1: Main differences between public and private organizations 20
Table 2: Overview of 3E criteria 20
Table 3: Basic data about South Moravian Region, 2017 56
Table 4: Basic descriptive statistics about interviewed sample of municipalities, 2011 79
Table 5: Descriptive statistics of the analyzed sample of 463 municipalities by quartiles 88
Table 6: Published DEA-related papers indexed in Web of Science and Scopus, 2004-17 91
Table 7: List of DEA solving programs 95
Table 8: Effects of selected M S W M characteristics on M S W M efficiency 102
Table 9: Descriptive statistics of the municipalities in the sample 115
Table 10: Estimated effects of collection systems and incentives on separation rates (in %) 119
Table 11: Average amount of generated waste per capita with and without incentive program 120
Table 12: People in the main age categories in the Czech Republic, 2011 and 2014 132
Table 13: Reduction of the amount of municipalities in the subsamples 135
Table 14: OLS estimation, Czech municipalities, 0.5% trimmed sample 136
190
Table 15: Residual MSW generation per capita for the individual age groups 141
Table 16: Separated municipal waste generation per capita for the individual age groups 144
Table 17: Description of the sample of municipalities (with respect to the year of the change) 151
Table 18: Year of the waste management company change, 66 municipalities 151
Table 19: Waste management company ownership before/after the change, 66 municipalities 151
Abbreviations
2SDEA - Two-Stage Data Envelopment Analysis
3E - economical concept including Economy, Efficiency and Effectiveness
CEA - Cost-effectiveness Analysis
CENIA - Czech Environmental Information Agency
CUA - Cost-utility Analysis
CZK - Czech koruna (currency), also known as Kc
CZSO - Czech Statistical Office
DEA - Data Envelopment Analysis
D M U - decision making unit
EPE - environmental protection expenditure
GDP - gross domestic product
HWRC - household waste recycling center, also known as civic amenity (CA) site
MF CR - Ministry of Finance of the Czech Republic
MSW - municipal solid waste
MSWE - municipal solid waste expenditure
MSWM - municipal solid waste management
MZP - Ministry of the Environment of the Czech Republic
OLS - Ordinary least squares
PAYT - Pay-as-you-throw
SFA - Stochastic Frontier Analysis
WEEE - waste electrical and electronic equipment
W M - waste management
191