1. TRANSPORT ECONOMICS 1.1. Organization Course }Transportation Economics explores the efficient use of society’s scarce resources for the movement of people and goods. }This course carefully examines transportation markets and standard economic tools, how these resources are used, and how the allocation of society resources affects transportation activities. Method }This course is unique in that it uses a detailed analysis of econometric results from transportation literature to provide an integrated collection of theory and applications. }Its numerous case studies illustrate the economic principles, discuss testable hypothesis, analyze econometric results, and examine each study’s implications for public policy. }These features make this a well-developed introduction to the foundations of transportation economics. } Lectures and seminars (2/0 + 0/2) I.Transport economics II.Transportation demand – the divisible goods case III.Transportation demand – the discrete goods case IV.Firm production and cost in transport – the long run V.Firm production and cost in transport – the short run VI.Competition, concentration and market power in transport VII. 1. Empirical Project }The major task will be to write an empirical project in transport economics. }Deadlines: } 30th October …. Proposal } 30th December ….. Project } } Structure of proposal (recommended) 1.A concise statement of the problem 2.Comments on the information that is available with one or two key references 3.A description of the research design that includes } a) an economic model } b) the econometric estimation and inference methods } c) data sources } d) estimation, hypothesis testing and prediction 4.The potential contribution of the research } } Structure of project (recommended) 1.Statement of the problem 2.Review of literature 3.The economic model 4.The econometric model 5.The data 6.The estimation and inference procedures 7.The empirical results and conclusions 8.Possible extensions and limitations of the study 9.Acknowledgements 10.References 11. Exam }Evaluation: }20% - seminar’s activity }40% - defense of empirical project }40 % - exam from transport economics Contact }Zdeněk Tomeš }tomes@econ.muni.cz 1.2. Transport and the Economy What is transport economics? }Transport economics is an applied area of economics that is concerned with the efficient use of society’s scarce resources for the movement of people and goods from an origin to a destination. Transport and GDP •Decoupling of economic and transport growth •Freight and passenger traffic growth •The impact of world recession • Transport and Prices •Do transport prices grow more quickly than other prices in the economy? •Why? Transport Performance World Comparison Transport Infrastructure •The state must provide the needed physical capital in the way of roads, rail trackage, airports, ports and other infrastrucutre in order to facilitate the movement of people and goods across space. • Quality of transport infrastructure •How extensive are these networks and are they keeping up with the transport demands of the public? •How well is the nation’s transportation network meeting the performance needs of shippers and travelers? •Are users receiving safe and reliable service or is the system providing deteriorating services? • Transport and Energy •In addition to its impact upon the economy and employment, transportation is a major user of energy • •Although transportation is a huge consumer of fuels, there have been dramatical technological imrovements in energy efficiency Transport Challenges •Safety •Mobility •Economic growth and trade •The human and natural enviroment 1.3. The statistical analysis of economic relations Economic model }Economic model – simplified description of behavior which purports to identify primary cause and effect relationships between economic variables }Economic models are: ¨Explanatory ¨Representative ¨Qualitative }Models are useful in formulation of policy because they identify the important causal linkages between economic variables. Econometric model }An econometric model is a statistical specification of an economic model that enables the analyst to quantify the economic relationships identified in the theoretical framework. An Econometric Model of Airline Demand }To illustrate the relationship between an economic and econometric model, consider the following economic model of the demand for airline passenger miles: } }Miles = f(Price per Mile, Per Capita Income) } = β1 + β2 Price Per Mile + β3 Per Capita Income + ε } Least Squares Regression }The method of obtaining parameter estimates by minimizing the sum of squared residuals is referred to as the method of least squares. }The least squares estimator is the formula that is used to produce the least squares estimate. }OLS estimator is unbiased, consistent and efficient Estimating the Demand for Airline Travel }Hypotheses: 1.All else constant, an increase in Price per Mile is expected to reduce the quantity of passenger miles demanded 2.All else constant, an increase in Per Capita Income is expected to increase the demand for airline travel } }OLS estimation: } }Miles = 15.7 - 8.7 Price Per Mile + 0.031 Per Capita Income } } Confidence Intervals }Each estimated coefficient is a random variable that has an associated probability distribution }A standard method that is used for determing how close each coefficient estimate is to the true parameter values is to construct a confidence interval around our estimate t statistics }It can be shown that statistics (βi – βi)/sβi has a t- distribution with degrees of freedom equal to the sample size minus the number of parameters estimated. }sβi is the square root of the estimated sample variance and is referred to as the standard error of the estimate. }If we want to construct a 95% confidence interval, than we want to identify lower and upper bound numbers for the true parameter value Confidence Intervals for Airline Travel } }Miles = 15.7 - 8.7 Price Per Mile + 0.031 Per Capita Income } (1.0) (0.0013) } } }Confidence intervals: }Price per Mile (-10.80; -6.60) }Per Capita Income (0.0284; 0.0337) Hypothesis tests }A major objective of many empirical studies is to test hypothesis regarding the relationship between the dependent and independent variables. }The methodology for hypothesis testing first requires that we identify two hypotheses, a null hypothesis (Ho) and an alternative hypothesis (Ha) } }Ho: β = 0 }Ha: β ≠ 0 } Issues in hypotheses testing }One tail and two tail tests }Type I and Type II error }Significance level (1%; 5%; 10%) }Reject x Not reject null hypothesis Goodness of fit and R2 }In addition to examining whether our estimated model is consistent with the implications of economic theory, it is of interest to know how well model fits the data. }R2 = goodness of fit = the statistics that gives the proportion of variation in the dependent variable that is due to all of the explanatory variables in the model Cross section and time series data }Cross section data reflect observation on individuals, households, firms, countries, or some other unit of analysis at a given point of time. Usual problem: heteroscedasticity → solution: robust standard errors. }Time series data are different because each observation corresponds to a different time period. Problem (1): autocorrelation → solution: robust standard errors. Problem (2): unit roots → risk: spurious regression → solutions: trend inclusion; differencing, cointegration. Model specification }A time trend is a variable that is included in time series models to capture the influence of factors that are correlated with time but are excluded from the model because they are either unquantifiable or the data are unavailable. A time trend is often use to reflect technological change. }A dummy variable is a variable that either has a value of 0 or a value of 1. It is often use for analysis of structural changes. Functional Form Marginal effects Elasticity measures The Airline Demand Model Revisited The Airline Demand Model Revisited Elasticity measures 1.4. Summary Summary (1) }An economic model identifies the most important factors that explain the behavior of some underlying economic variable of interest. Economic models are explanatory, generally characterize average behavior, and identify the qualitative relationships between variables. }An econometric model is a statistical model that enables one to quantify the casual relationships identified in an economic model. Econometric models are used to make statistical inferences on the importance of “explanatory variables” in explaining the variation in the “dependent” variable. Summary (2) }A common method for estimating parameters in an econometric model is the method of least squares. The “least-squares estimates” are estimates of an empirical model’s parameters obtained by minimizing the sum of squared errors from a sample of observations. Least-squares estimators are unbiased, consistent, and efficient. Summary (3) }Hypothesis tests are used to statistically determine whether variation in an explanatory variable explains at least part of the variation in a dependent variable. All hypothesis tests involve a null hypothesis, which identifies an hypothesized true value of the parameter, and an alternative hypothesis. If the alternative hypothesis specifies a two-sided effect (positive or negative), the hypothesis test is two-tailed; if the alternative hypothesis specifies a one-sided effect, the hypothesis test is one-tailed. Summary (4) }In regression models, t-statistics are used to test hypotheses on the statistical significance of the explanatory variables. In large samples (more than 120 degrees of freedom), if the calculated t-statistic under the null hypothesis was greater than 1.960 (2.576) in absolute value, then we would reject the null hypothesis at a 0.05 (0.01) level of significance. On a one-tail test, we would reject the null hypothesis at a 0.05 (0.01) level of significance if the calculated t-statistic was greater than 1.645 (2.326) in absolute value. } Summary (5) }A regression model’s goodness of fit gives the proportion of variation in the dependent variable which is “explained” by explanatory variables included in the empirical model. }Cross-section and time series data are common forms of sample information used to estimate empirical models. Cross-section data provide information on observations units at a point in time. Time series data provide information on an observation unit over a period of time. }Common statistical specifications for estimating empirical models are testing hypotheses on economic relationships include the linear model, the log–linear model, the linear–log model, and the double–log model.