acroeconomics A European Text Second Edition Michael Burda and Charles Wyplosz OXFORD UNIVERSITY PRESS 6 6.2. Demand and Supply in the labour Market 6.2.1. Labour Supply and the Labour-Leisure Trade-off 6.2.2. Labour Demand, Productivity, and Real Wages 6.2.3. Labour Market Equilibrium 6.2.4. The Interpretation of Unemployment ,6.3. A Static Interpretation of - , Unemployment 6.3.1. Involuntary Unemployment and Real Wage Adjustment 6.3.2. Collective Bargaining and Real Wage Rigidity 6.3.3. Social Minima and Real Wage Rigidity 6.3.4. Efficiency Wages and Real Wage Rigidity Labour Markets and Equilibrium Unemployment 6.4. A Dynamic Interpretation of Unemployment 6.4.1. Labour Market States and Transitions 6.4.2. Stocks, Flows and Frictional Unemployment 6.4.3. job Finding and the Duration of [Unemployment '6.5. The Equilibrium Rate of Unemployment 6.5.1. The Concept 6.5.2. The European Experience 6.5.3. Actual and Equilibrium Unemployment 6.5.4. Equilibrium Employment and Equilibrium Output 6.6. Summary Appendix Labour is the source of all value RarlM.ir\ In our present day complicated economic life we are likely to be confused by the ma 11 j i n dus-trial operations and money transactions. But net income remains exactly what it wa,- Lo pi uni-tive Robinson Crusoe on his island—the enjoyment from eating the berries we pick, so lo speak, less the discomfort or the labor of picking them. Irviii" lisher 1 - In the last chapter, available output was determined by the endowment of capital and technical sophistication. The supply of labour was simply taken as given, regardless of the wage or other variables. At any moment of time, therefore, the economy relied on what history had left, i.e. capital, labour, and knowledge. In practice, however, income is not derived just from fixed endowments. Even in storybooks, life is not so simple: Robinson Crusoe had to expend time and ef-[fort to gather and transport the fruits that he would ^eventually consume. Indeed, most people are able to ! choose whether or not they will work, and sometimes ! how much they will work or at least how much effort ! they will put in their work. Households work so that they can consume, but they also desire to spend some i time networking, which is called leisure or free time. : The supply of labour is presented as a trade-off be-i tween consumption and leisure. Labour must also be in demand. For that, it must have value to firms. How the markets value labour and how demand and supply interact is the subject matter of this chapter. This chapter deals with an important market. Marx had a point when he viewed labour as the most important factor of production. Everything we use stems from labour. Raw materials are drawn from the earth by human hands; equipment used in this process is produced from labour and previously manufactured equipment, itself the output of labourers and capital in a more distant past. Even the knowledge embodied in people—sometimes called 'human capital'—comes from our own efforts at mastering skills and techniques, as well as the time our teachers spend i n irMny to educate us. As before, we begin the analysis by studying the be-" haviour of a representative household lli.it supplie^ labour. Next, we look at demand by a representative;' firm. This naturally leads into the st.nu.liii J conlron tation of demand and supply. Yet, tht l.ibom maiket is not a standard 'market'. Workers are not identical^ and the quality of labour services is difficult lo istej-tain and harder to monitor. Unlike jn.Kh'ni> <>i raw materials, workers can decide whether they w< mid like to work for a particular employer aad luuL-i which conditions to render labour services, hi Met, the employment relationship involves expliv.il .uid implied contractual arrangements with durable fealui es luch as personal bonds or firm-specific i oinpeienct- and knowledge. The labour market is also charatiautd by unique institutions, such as labour law 01 collective bargaining, and is the object of o nuplcs. It-pal anu; customary rules. Finally, the labour iii.ukei * Jv"-' namic market, with suppliers of labour entei itig exiting unemployment at a remarkabk ralc.'NW'*"0 . how these interactions help undersl uid ihc ton«p of equiUbrium unemployment, wh ich may dil 111« 01 , actual unemployment. LABOUR MARKETS AND EQUILIBRIUM UNEMPLOYMENT 137 6.2. Demand arid Supply in the Labour Market 6.2.1« Labour Supply and the Labour-Leisure Trade-off ]n oi der to consume, most households need income, jn modern societies, earning income means working, oi supplying labour to firms in return for a wage or viljrv- Supplying labour is something that millions of f-imilies do every day. Labour has a cost too: every luuii worked is an hour less of free time. Because households value both consumption and leisure, they balance the two, just as they balance consumption of ^oods today against saving and goods tomorrow. As ■vm. incus on the consumption—'eisu-e trade-off we ignore ibe intertemporal aspects which were consid-eredin( hapten ■ !,aiid ">. In iIk iiou-1 uiiili.>i parable, Robiiiii'ii < m-.<>i in I hi-. Ji.ipiu is as-,Linkdto Lnnsuiiie liJI dial hi. i-n n-. din in" i.iJi pi mm 6.2.1.'. Fefa.s-:=i Crusoe's preferences with regai"dto consumption and leisure are shown in Figure 6.1 using indifference 0 ; Leisuie »K Figure 6.1. Household Pi eforences An indifference curve shows the Mte Ht vhirh a repreri sentative household substitutes rnnsumpcion for leisuie, holding utility ronstant Higher uirve-i correspond to higher levels of utilitv. The maximum •urinuiiL of time available is hours. curves.1 Each indifference curve shows how readily Crusoe substitutes consumption for leisure, holding constant his level of utility or satisfaction. The shape of the indifference curves reflects a decreasing marginal rate of substitution: the greater a household's consumption relative to its leisure, the more consumption it is willing to give up for an additional unit of leisure, or the higher is the marginal rate of substitution of consumption for leisure. As always, higher indifference curves correspond to higher levels of utility. 6.2.1.2. The budget constraint Crusoe's 'budget constraint' is fixed by the total amount of time, denoted €, available over any given period (a day, a month, a year, or more). Time is a scarce resource and it has a price. The price of an hour of leisure is its opportunity cost: how much can be earned from working instead. Equivalently, the price of leisure is measured in terms of the consumption goods that cannot be consumed for lack of earned income: the price of leisure is then called the real (consumption) wage. In practice, it is measured as the ratio of nominal wages (W) to the consumer price index With £ hours at his disposal and facing an hourly real wage w- MP, the value of Crusoe's total time endowment in terms of consumption is tw. The budget constraint states that this endowment can be allocated between consumption, with value C, and £ hours of leisure, with value tw? (6.1) Iw=£w+C. Alternatively, the budget constraint can be expressed in terms of'cash flow' (more accurately, coconut flow). When Crusoe spends -Hiours of leisure, he works £ -€ hours and earns w(£ - €) coconuts. Since Crusoe does not save, this income w(-£ - €) is spent on consumption; so (6.2) w(e--E) = C, which is the same as (6.1). The budget constraint is shown in Figure 6.2 as AB. Its negative slope (-w) 1 This is the same idea as in Ch. 4, except that here we look at two 'goods' in a given period, rather than atdifferent points in time. 2 The nominal budget constraint is iW= tW+ PC. To write it in terms of consumption goods as in (6.1), we simply divide by P, the price of consumption goods. 138 THE REAL MACROECONOMY ,, Leisure t Figure 6.2. The Household Budget Constraint The household has ( hours at its disposal (measured by Lhe dictiinru OA) toi either leisure ot i oik For ev<_ry un t M leisme that it qi"es up it can earn a qjcintiL/1" ot consumption goods The leal waqe iv detti mines the slope of the budget line AB measures the trade-orT of consumption for leisure of-i fered by the market: how much consumption must be given up to get an additional unit of leisure. The distance OA is Crusoe's endowment of time, or the fixed number of hours he has at his disposal. The distance OB measures the value of that endowment in terms of consumption goods. It is the total amount of consumption attainable when leisure is zero. If Crusoe had inherited some initial wealth, the budget constraint would be shifted vertically by that amount, which he could consume without having to work at all. If the real wage changes, the budget line rotatesaround point A, which measures his time endowments. With a wage increase, for example, more consumption can be afforded (OB increases) and the budget line rotates clockwise. If instead the real wage falls, OB declines, and the line turns in a counter-clockwise direction. 6.2.1.3. Optimal choice: the individual labour supply schedule Crusoe maximizes his utility by choosing the highest possible indifference curve without violating his budget constraint. This is achieved at point R in Figure 6.3, where the indifference curve is tangent to the budget line. At this point, given the going market wage w, he cannot make himself better off by further trading leisure against consumption: the marginal rate of o _ Consumption Leisure 1 Worl Leisure' ', Figure 6.3. Optimal Choice Given the hudqot (.oivtMinr, the hijhpst possible util iLy is adiieved at point R, nhoiL aw irdirferLnrp curve is tangent tothf budget line. substitution of consumption for leisun. 11 u 11 lio vi ay., are equal. In the first panel of Figure 6.4, an increase in the real wage changes Crusoe's consumption-leisure choice from R to R'. It is useful to disiinguish two effects. First, Crusoe faces a higher opi'mu-nin i."M of leisure in terms of consumption goods. Work has become relatively more attractive and a rational Crusoe responds by choosing less leisiiK nul mon. consumption. This is the substitutio- JillI I hi. ■■cl ond effect works in the opposite diret ium \n nurtan" in the wage allows Crusoe to enjoy both more con-;, sumption and more leisure: this is tlv rin.«into (.-fleet For this reason, moving from R to R', Crusoe's con-^ sumption increases, but the effect on leisure (and-, work) is ambiguous. The right-hand side panels of Figure 6.4 show ho^ the relative strengths of the income and substil • effects translate into different individual '.lWom *>UP" ply curves. The labour supply cm i hou s It ^ m"111 labour an individual is willing to supply at e; of the real wage. In panel (a) the income and institution effects exactly cancel: leisure—jiid lab* ply—remain unchanged. This is the' benchra where labour supply is inelastic, m iiniv."-P',i|S,xi; 0 the real wage. If the substitution effect predonin,jL'"" Crusoe responds to higher real wages by redw'S ■ leisure time: labour supply is elastic: (panel-panel (c) the income effect dominates: both consun t tion and leisure increase. In this case labour sup J1 % " LABOUR MARKETS AND EQUILIBRIUM UNEMPLOYMENT 139 Figuie 6.4. Labour Supply When the teal wage inu cases, the Dudget line rotates aiound point A (the endowment of Ume lemains unchanged) and becomes steepei, becduse a unit of leisui° is exchanged loi more units of consumption This alio vs both consumption and leisure to increase al the same time (income effect) Because leisuic is more expensive, however, some is given up (substitution effect) In rase (rj) income and substitution effects exacth cancel in the lert-hcind pniipl jnd labour supply, sho in in the light hjnd panel, is vertical In rase (b) the substitution pffc-cl dominate1-, leisuir5 is reduced, and the laboui supply schedule is upivaid sloping In rase (c) lh& income effect dominntos, leisuip inaeases, and labour supply is back-,wai.dT.but.idinq. 140 THE REAL MACROECONOMY Table 6.1. Annual Total Hours Worked and Average Wages, 1870-1987 18/0 1913 19J8 1987 Annual hours worked per person France 2945 2588 1848 1543 Germany ZW 2584 2316 1620 UK 2084 2624 2267 1.557 USA 2964 2605 2062 Í608 Sweden 2945 2588 2204 1466 Real wage (index 1870 = 100) France 100 ?05 335 1358 Germany 100 1S5 285 1227 UK 100 157 256 589 USA 100 189 325 '643 Sweden 100 270 521 1439 Sources: Hdurs worked ai;e from Maddison'(199'1);s wages are from Mitchell (1978, 1983) and OECD, Main Econo'ipklndi-pators; German wage data from 1913^38'are approximated using average labour productivity growth backward-bending as increases in the real wage actually reduce the supply of labour. In practice, the response of labour supply to a change in wages depends on the time horizon under consideration. In the short run, individuals do not seem to react much to changes in the real wage (the inelastic case). In the next section, it is shown that aggregate labour supply is generally more elastic than individual supply in the short run. In the long run, labour supply is backward-bending. This is what Table 6.1 shows. Over the last 100 years, real wages have increased by five- to fifteen-fold, while working hours have declined by one-half. The evidence from Table 6.1 has to be interpreted carefully, though, as labour supply varies according lo sex. For men, the average work-week, the retirement age, and the rate of labour force participation (the proportion of working-age people working or registered as unemployed) have fallen secularly since 1900. J vork >s added nutpirl 'nupases but at a declining rate. This additional output ii tne marginal pioriucti 'ity of laboui (MPL) Flic- rav i the ..ost of using l houis of work'ahen +ne hourly real, wage is w The distance bet' 'een the production curve " and the ensi line itpiosents the firm s piofit. It is ata ^ maMinuin at point A, v* heie the e ui ve ii parallel to OR, i.e. where MPL = w. hours) used. The link between outputFamI ..mplov-ment I is captured by the production mnc i ion «.bowii in Figure 6.7. The slope of the production lunaion -measures me marginal productivity of labour' Wl\ the quantity of additional output obtained '\ hen oni. _ more unit of input (an hour) is used. The shape or the curve reflects the principle of decreasing ma i g'lial P'1'" ductivity; the MPL is declining as the annum "f labour employed increases. In deciding how much labour to emplov, the Pun looks for the highest possible profit given liV liourlv ^ real wage w. The line OR represents the ccl of labour^ to the firm: its slope is w since I hours ot <*oiA tost wL. For each level of employment, profits mca'-uiM as the vertical distance between the curvi uVpicting. the production function and the labour cn^t Imc c »■ It is at a maximum at point A, where: the cui \ >-ls Va*_~ allel to OR and the MPL is equal to the rea I w a?e. It th* j j MPL exceeds the real wage, hiriivj one mui c. houi <>. work raises profits by MPL and costs by >nh W' u ■*-'-- fit plying an increase in profits. The firm won IJ th"^',V** fore hire the extra hour and would continue ^'''"^j^v* until the MPL has declined to the point *l,L'rt',l&V-* equal to the real wage. In the opposite case, in w the real wage exceeds the MPL, the firm can increas? its profit by reducing its demand for labpi.r l,L'CJV5rt% LABOUR MARKETS AND EQUILIBRIUM UNEMPLOYMENT 143 <• Figure 6.8. Labour Demand Curve j;The firm maximizes profits when MPL = w, so its de-i mand tor labour n given by the declining marginal i' product o1 labour (MPL) curve il is optimal to set labour such that MPL = w, the MPL schedule in Figure 6.8 is also the labour demand curve. 6.2 2 2. Wage end profit shares "The portion of GDP that goes to workers is called the " wage shareand is given by wLI Y. The rest, the profit, or 'capital share (1 - wL/Y), goes to firm owners.4 Obviously, these two shares move in opposite directions. But what is the effect of an increase in wages on the Ijhuurshare? At the given employment level L, labour s'eosts (wL) rise because an hour of work (w) costs more. With higher wages, however, total employment (L) is Reduced as firms move down their labour demand .schedules. The effect on the numerator of the labour share is therefore ambiguous. At the same time, a de-vcline in employment generally means a decline in output, and therefore a decline in the denominator. If the demand for labour is not very elastic, neither labour demand I nor output Y change much in response to an increase in Lhe real wage. Then both total labour costs and the labour share increase, while the profit share declines. Changes in the labour share are important for two ^reasons. First, the distribution of income between capi-tal and labour has political elements that can affect the behaviour of governments. Second investment responds to profitability, which is related to the profit -j A This is not quite accurate: in fact, rental income, patent and royalty, payments, and other minor factor incomes have been ignored. share. Investment, in turn, affects growth.5 When labour demand is rather inelastic, therefore, exogenous wage increases may reduce profitability, Tobin's q, investment, and the standard of living. 6.2.2.3. Shifts in the demand for labour Now consider the effects of an increase in the capital stock K, which was assumed constant so far. Figure 6.9(a) shows that this normally raises MPL—the production function becomes steeper at every level of production. The labour demand curve shifts out in panel (b). A technological improvement that shifts out the production function produces a similar effect.6 This helps account for the fact that wages have grown secularly over time. 6.2.3. Labour Market Equilibrium We now have the building blocks for understanding the labour market: a supply curve derived from household behaviour, and a demand curve derived from firm behaviour. The interaction of supply and demand for labour is depicted in Figure 6.10. Equilibrium occurs at the intersection of the two curves (point A). At wage w* the market clears (there is no excess demand or supply): L* is the number of hours firms want to hire and households want to work. Both the real wage rate and employment are endogenously determined in the labour market. This simple characterization of the labour market will serve as the benchmark for the rest of the chapter. Figure 6.11(a) provides an example of its usefulness. It depicts an increase in labour productivity, which has occurred over centuries as the result of capital accumulation and technological advances. The labour demand curve shifts outward; the supply curve is unaffected. The effect is an increase in living standards as real wages increase. If the labour supply curve is perfectly vertical, employment remains unchanged and labour income wL rises proportionally to the real wage. If the supply curve is backward-bending, employment 5 The inverse relationship between real wages and the profit share is illustrated in Ch. 1, Fig. 1.4. The link between investment and profits is studied in Ch. 4. The growth effects were analysed in Ch. 5. The Appendix presents a formalization of this section. 6 Ch. 5 discussed technical change. An exception would be a fobour-sflvmgtechnical change, such as robots, which would reduce the demand for labour at any wage and shift bade the demand for labour. 144 THE REAL MACROECONOMY MPLr' MPL Labour dfft& Figure 6.9. An Increase in Labour Productivity Labour bt Lomrs mnrf productive oither b-Ccjusc moro i špital is put in plarr or bi'i luse of ferhnoloqunl progress In panel (ci), at- ony lu\el of innnur input, morr ojtpul i produced and the production function is everywhere steeper. The MPL increases and the demand foi Idboui 5cht>duli shifts up in panel (Ď) illllllii Supply Demand ' Labour hours Figure 6.10: Equilibrium in the Labour Market Labour market equilibrium occurs at "point A where (demand and supply are equal. The, real wage "w* clears the market at employment levelJL*. If total labour endowment is L, the,distance;(L-i*) is'voluntary unemployment. ' (man-hours) declines while real wages increase. Table 6.1 seems to indicate that this has been the case over the past century. The second panel of Figure 6.11 shows that an exogenous increase in the supply of la- bour leads to an increase m employment m the short run, but also to a reduction in real wages.7 6.2.4. The interpretation of ■ Unemployment While the supply-and-demand apparatus allows usto-,. evaluate the effect of various changes on equilibrium. £ employment and real wages, it is disappointing m one j crucial respect. At point A in Figure 6.10, labour sup-, ^ plied is equal to labour demanded. Any uneinpli resources correspond to voluntary decisions of hou-*^ ^ i holds and firms. If total available labour supnh i*- ^ -■ unemployment (measured as hours of work m >r 11 u oT "r* is measured as L -1*. Since point A is on the labour supply curve, it corresponds to the optimal behaviour. < 'm and I" - L' is labour supplied but nui ik-m.imJLd bv the market. If the real wage were t xLJiik l<> i» • J^i, mand would increase, supply would decrease, Mid employment be restored at I*. illmg apO-f■ able to work at the wage W but cannot find a lob-matter how hard he or she tries. This is a U\ ics"g the existence of involuntary unemployment niustoe . explained by real wage rigidity, which we e\amu£ next. ' 6.3.2. Coiiective Bargaining and Real Wage Rigidity For sustained real wage rigid11> In m-u i a« LABOUR MARKETS AND EQUILIBRIUM UNEMPLOYMENT 147 0 W L- L" Involuntary unemploymen Labour Figure 6.12 Involuntary Unemployment At the real wage late iv vorl '.'is supply L ofclabour, "but fimv demand, and lint un\j L The qiunui' L , /' uhiihr .uppliod b, hour<" holds hut iv>t demanded by firms, represents involuntary unemployment II.the* real * ^aqe i 'ere to adju .1 t'i Ihu levi 1 \\i , ihe market v ould lIi-vji at pnmt C. tunable to supply their labour sendees at wages below y/*, or firms must be unwilling to take up such offers, ' pr be unable to make their own. What important institutional feature bas been overlooked so far? Labour unions are one of the most fundamental and universal institutions that operate in modern economies. 'They are matched by equally powerful employers5 associations, such as the CBI in the UK, the CNPF in France, the BE)A in Germany, or the SAF in Sweden. The perfect-cqmpetition description of labour markets of Section 6.2 contrasts sharply with the bilateral bargaining commonly observed in Europe between employers and unions. We first characterize the economic role of trade unions to discover howunemploy-ment can be voluntary from the perspective of trade unions and nevertheless be involuntary from the viewpoint of the individual household.8 6.3.2.1. Labour unions and their rationale The employer-employee relationship has inherently confiictual aspects. One reason for this is the distribu- 8 It should be stressed that we limit ourselves strictly to the economic significance of trade unions. As the history of the labour movement amply demonstrates, unions have had an enormous influence on modern society which goes beyond economics. tion of income; economic principles assert that income should be split according to marginal productivity, but marginal productivity is not easily measured, and economic principles are not always adhered to. Another more subtle reason is that firms need to monitor effort at work, a key element of productivity, which is under the control of each individual employee. Individual workers facing a large employer are in a poor bargaining position. They have little influence over their own wage rate and may not even feel safe discussing working conditions, fearing reprisals in the form of a salary cut or dismissal. They may also feel pressure from powerful employees to accept conditions that would not be acceptable under competitive conditions. To resist such pressures, but especially to achieve higher pay levels and a voice in the day-to-day operation of the workplace, workers have organized themselves into trade or labour unions. As Table 6.3 shows, union organizations vary considerably from country to country. Scandinavian countries have a tradition of centralized unionization; Britain is organized according to craft; France, Italy, and Spain have unions with ties to political parties. These differences reflect social history as well as the costs and benefits associated with union membership. The costs are dues that members must pay. The benefits vary, ranging from higher wages and protection from arbitrary employer decisions to more specific advantages, including priority for certain jobs and income supplements when unions are on strike. In some countries, many advantages accrue to all workers, so there is little point in paying the union dues. This is the case in France, for example.9 In other countries, such as Belgium and Scandinavia, unions manage funds that hand out some social benefits. In the USA, some unions even issue credit cards and provide other services to their members. 6.3.2.2. The economics of labour unions Unions have two major economic objectives: higher real wages, and more jobs.10 It is useful to think of their preferences in terms of indifference curves as shown in Figure 6.13. The slope of the indifference curve represents the willingness of the union leadership to trade off employment for wages. Panel (a) describes the average union which accepts a trade-off between employment andhigher wages. A 'hard-line' union which 9 This is an example of the so-called free-rider problem. If no workers pay dues, the union disappears and no one is protected. So some workers must pay the dues for all to have a union. 10 This is a simplification, of course. Unions care for other things too, such as safety at work, working time, workers' say in worldng conditions and organization. 148 THE REAL MACRO ECONOMY Table 6.3. European Trade Unions: Structure and Membership, 1970 and 1988 Country Structure and principal unions Union density8 1970 1988 Austria Umbrella/industrial (OGB) 59.8 45.7 Belgium Party, religious (FGTB, CSC, CGSLB, CNC) 46 0 53.0 Denmark Umbrella (LO) 60.0 73.2 Finland Umbrella (SAK) 51.3 73.3" France Party, religious (CGT, CFDT, CFTC, GGC, FO) 22.3 12.0 West Germany Umbrella/industrial (DGB, DAG) 33.0 33.8 Ireland Mostly crafts in 1CTU, fragmented 53.1 52.4C Italy Party, religious (CGIL, CISL, UIL) 36.9 39.6 Neth. Party, religious (FNV, CNV, RMHP, AVC) 37.0 25.0 Norway Umbrella (LO, AF, YS) 50.6 57. V Sweden Umbrella (ILO, TCO, SACO/SR), 67.7 85.3 SwitZj Mostly plant-level (SGB) 30.7 26.0 UK Mostly crafts (96 in TUC, fragmented) 44.8 41.5 memo: USA Mostly local plant-level (AFL-CIO) 22.8 16.4" Union density isthe percentage of employed workers who are union members. (Unemployed work ers are not counted.) * b 1989 : 1987 Sources: OECD, Employment Outlook, July 1991; Barnouin (1986); Roberts (1985) 5 , Employment Employment Employment (a) Average ' - ib) Hard-line (c) Jobs first ■ Figure 6.13. Trade Unions' Indifference Curves When a trade union values both higher wages and more employment, its preferences are described by indifference curves. A 'hard-line' union is not willing to give up much in lo>, 'er wages to raise employment. A union mainly preoccupied with employment is represented by steep indifference curves. LABOUR MARKETS AND EQUILIBRIUM UNEMPLOYMENT 149 ooesfgr high wages is represented in panel (b): here the curves are flatter as the union is not willing to give lip much in terms of real wages for additional em-]0y,nent. A union that cares more about jobs than waacs is characterized in panel (c) by steep indiffer-(.mi curves. In contrast to Section 6.2, union indifference curves jgp$K»e-the^representative individual's indifference ,curves, symbolizing the fact that the active agent in tin. labour market is not the individual, but his or her 'trade union—or, more generally, the collective bargaining process. The 'budget line' faced by the union is i lu demand for labour. From Section 6.2.2 we know fhat labour demand is given by the MPL (either a firm's MPL, or that of an industry or the entire economy: if firms are all alike, their individual demand does not differ from the collective one defended by an employers' association). The optimal choice for the union is the tangency point between the highest indifference curve and labour demand. When the demand for labour shifts out—as a result of capital accumulation or technological progress, for example—the succesive tangency points map out a collective labour supply curve as represented in Figure 6.14. The curve describes the most desired joint evolution of real wages and employment from the union's viewpoint. The shape of the collective labour supply curve Employment . ' Employment ', l' >'* ' Employment ■ Employment ; '(c) , I'd) . . Figure 6.14. The Collective Labour Supply Curve The collective labour supply curve is obtained by connecting the points of tangency between the indifference curves and a shifting labour demand schedule. Panels (c) and (d) respectively describe unions with a target wage level wTand target employment level LT. Panels (a) and (d) give rise to steep schedules, whereas (b) and (c) are associated with real wage rigidity and a more elastic supply of labour. 150 THE REAL MACRO ECONOMY reflects the preferences of the union, as shown by the various panels of Figure 6.14.11 Panel (a) is an example of a 'hard-line' union with a steep trade-off of wages for employment; panel (b) is the counter-example of a 'jobs-first' union requiring only moderate wage increases to supply more labour. Panel (c) shows a union with a well defined real wage target ready to accept a wide latitude of employment outcomes, yielding a flat collective labour supply curve at the target level. The other extreme is panel [d), which aims at a given level of employment, and is willing to set wages at whatever level is necessary to keep this group employed.12 The appendix provides a formalization of these ideas. 6.3.2.3. Employment effects of collective bargaining The collective labour supply curve resembles the individual supply curve of Figure 6.6, but has different origins. Collectively, through their unions, workers feel that they have more strength and accordingly aim at better outcomes. In particular, for a given amount of labour supplied, they ask for higher real wages: the union-driven collective labour supply curve lies above the individual labour supply curves. In Figure 6.15, without the union equilibrium would occur at point B: individuals would be willing to provide employment L, at wage w2. They cannot, however, because the wage wt is set through negotiations between the firms and the trade union, and individuals cannot simply underbid their employed colleagues. Unemployment {L\ - LJ is involuntary for affected individuals, but voluntary from the union's point of view. Why do unions enforce wage rigidity apparently against the will of unemployed workers? One reason is that the leadership is elected by the employed. Unemployed workers are always a small minority of the membership, even at record high unemployment rates of 10% or even 20%. Furthermore, unemployed workers often give up their membership or lose interest in union affairs. Unions end up representing those who work, not those who are unemployed. Employed workers look for high real wages (for themselves) at the cost of some unemployment (for others). Box 6.3 illustrates how the relentless rise of unemployment in Europe after the two oil shocks can be explained by this effect. 11 A technical note: this is true only for a given shift of the labour demand curve. Here we consider only parallel shifts; one could imagine however, that other types of shifts would change both the position and the slope of the labour demand curve. 12 This group may be the current union members, or the subset of the membership that is employed, or some core of so-called 'insiders', those members with sufficient seniority. Individual supply curve B ÜHüü Employment > Figure S.I 5. Labour Market with a Trade When a l.ibour union repi esents woi kers at vj aqe nego-^1 tiations, labour market equilibrium ocLiirs appoint/4 •* If the union collective labour supplj luivpis abovethe^ .ndi"dual labour supply tuive, the real wage w,MsJ highei and employment/., (hours, or nunibefcot work.-? eis) lov/or than at point B, which would be. the ou^>£ comd if individuals w^rc negotiating individually The y result is the cxisionce ot union-vclunUii} individual- 'r involuntary unemployment (i , L) itislhtdiffeience^ between actual employment L and tho amount cjf work L'i that workers are individually willing to supply,] at the real wage rate wv V The split between unions and unemployed w mUlim cannot go too far, though. After unemploynieiil increased to high levels in Europe in the 1970s and I Ws. unions have become more employmentrconscjoui and real wage growth has moderated. One town is that members become worried that they too might become unemployed. Another is that the loss 1 n mHA"^ "^jj bargaining agreements reached with unions. fcl',^i^j end, employers' associations do not control "ie-jipyj* mand for labour: this is the prerogative of the ,ni^sf* vidua! companies. While it is in firms' interest to, v i-gj 1 -I J wages low, it is also in their intei.-si U >U-i-P ^ of their competition high, or at La -I m present Ifl^ competitors from hiring cheap labour. Thus, 1 ■> ers' associations also contribute to real wage njs LABOUR MARKETS AND EQUILIBRIUM UNEMPLOYMENT 151 BOX 6 3 "I IH HYSTERESIS EFFECT lillllltfllll^^ Sisrhetwo oil shocks of the mid-1970s and early 1980s correspond to a significant inward shift of the aggregate labour demand-curve in Figure 6.15, hence to less employment 11 trade unions respond to a narrower membership by calling for higher wages, the collective labour ' supply curve shifts upward. Owing to-the behaviour of the 'insiders' who have jobs, employment prospects for outsiders' are reduced. After the oil shocks are absorbed. the employment level is permanently reduced. That such an effect—dubbed the 'hysteresis effect'—has been observed in several European countries is suggested by the step-wise increase in the unemployment rate following each oil shock (Figure 6.16). In contrast, in the USA and Japan, where unions may have been less 'hard-line', unemployment rates increased at the time of each oil shock but then reverted towards earlier levels. 1365 1970 1975 , 1980 1985 1990 figure 6.16. Hysteresis in EC/EU Unemployment, 1965-1994 The unemployment rate increased in both the EC/EU and the USA after the two oil shocks. In Lhe USA the rate has since come down, to about where it was in the early 1970s. In the EC/EU, Lhe unemployment rate seems to suffer from hysteresis, stabilizing wherever unexpected shocks move it. Source: IMF, World Economic Outlook, October 1995 6.3.3. Social Minima and Real Wage Rigidity Involuntary Unemployment occurs when real wages are downwardly rigid. Beyond monopolistic behaviour by trade unions and employers' associations, several other institutional and economic factors contribute to wage rigidity. Among institutional factors frequently suspected are Social minima, or minimum standards for income and earnings mandated by the government for reasons of social equity or protection. Minimum wages prevent wages from declining below some level. Many countries legislated minimum wages long ago for a variety of reasons. One was to prevent employers with too much market power from depressing wages artificially. Another reason was to protect young people from exploitation. With schooling rudimentary and poverty endemic, for many youngsters on-the-job training was the only way to get started; unscrupulous employers would offer very low wages, sometimes below minimal survival needs. Social protection was and still is justified; but, paradoxically, 152' THE REAL MACROECONOMY Individual supply curve Employment Figure 6.17. Minimum Wages Minimum wages reduce the demand for labour below the level that would result with either union-negotiated wages or individual-supplied labour. youth unemployment may be exacerbated, by mini-jmum wages. Their effect is to deter firms from hiring j workers whose MPL is below the minimum wage rate. I Those most likely to be hurt are poorly educated young i people with no job experience and older workers with obsolete skills. The effect is quite widespread, because, once a floor is set, it pushes up the lower echelons of the wage pyramid, possibly affecting better qualified workers as well. The range of qualifications for which the MPL is below the real wage is wider than just the very lowest echelons. Figure 6.17 illustrates the effect of minimum wages. To serve any purpose at all, the minimum wage H'.,.:,. must be higher than the wage w0 that would obtain"' otherwise, and which is itself higher than what individuals would accept (wj) if only to cover their part of the cost of investing in human capital. The result is ■ unemployment (I, - Imin) even higher than the level' (Ii - L0) implied by the union-set wage. Some evidence on the effect of minimum wages is presented in Box 6.4. 6.3.4. Efficiency Wages and Real Wage Rigidity Another reason why real wages may not dechne in the S presence of involuntary unemployment is that iirnis do not wish to reduce them. The phenomenon is often called efficiency wages, and it is related to the dif- • / ; ficulty for firms to observe work effort.: By paying a," worker a wage in excess of his marginal productivity,' firms may attempt to elicit more work. A worker who is dismissed for lack of effort is unlikely to obtai n si1 J1 a good deal elsewhere, especially if dismissals an. in-terpreted as a sign of poor work effort. Firms nui ilso_ _ pay efficiency wages to obtain a better selection ofap-j|pJ plicants and to keep workers from quitting too n( In capital-intensive industries, where shirking«. i > aid seriously disrupt the production process and v li ete a high-quality work-force is of primary impoitan^v, firms have a strong incentive to offer efficiency wages'. ' ^ In this case, the function of real wages is not just lo^ equate demand and supply in labour: markets \s a ■ result, wages will not generally be able to satjsU both functions. Real wages will tend to be rigid and ma\ be set above the market-clearing level, as in Figure 6. 6.4. A Dynamic Interpretation of Unemployment 6.4.1. Labour Market States and Transitions Any person can find herself in one of three situations: employed, unemployed, or out of the labour force. Figure 6.18 displays these three states and the various flows that describe how workers shift from one situation to another. A striking aspect of labour markets in developed economies is the size of these flows. Table 6.5 shows that the flow of individuals moving into and out of unemployment per year is almost twice the stock of unemployment at any given time. In contrast to the static picture painted in Section 6.3, labour markets, are remarkably dynamic, even when unemployed is high and stable. -' 1L There are three ways of becoming unemplagg First, new entrants to the labour market join thej a\ bour force before they have found a job but a16;-^ successful, at least initially. Second arc separation from jobs. Voluntary separations from the viewpoint (quits) account for roughly 50% to 6«^ all separations from employment m the UK, and'u 70% in the USA. Yet quits rarely lead to ^eJ| ment: most workers who quit immediately. LABOUR MARKETS AND EQUILIBRIUM UNEMPLOYMENT 153 6.4. MINIMUM WAGES AND YOUTH UNEMPLOYMENT BJ3s stnkinq that teenager", in the USA often work during She'summer when their Euiopean f oiiriterparts go on Jjiation. One rca.on iniqht bo that < -ages thai must he Spafd-for young, unskilled Inboui are too high in Europe fhe US nnnimuni uuge jniounts 10 ahout i third of the sljjyetrage manufacturing wage while many European rrountnes it exceeds SO1-" This is one reason thy tillinq Ration attendants and groLery shop assistants have all 'i6ui disappeared in most European countiiGs Table 6 4 5§hows non-omploymcnt and unemployment rates 1or fefeung people in a number of countries as uell as Hie 1 average minimum wage as a fraction of the a"eiage Qivagp. In interpreting the table, it is important tn note jjW Denmark exempts those below 13 from collective J&fjifrgaining agreements The UK does not have a mini •Trtium wage and teenagers are generalh not uniornred It is especially rioteworLh\ that in Franco, Belgium, the Netherlands, and Portugal, the minimum v aqe as a fraction of the median wage is high mejiung that a giLat many jobs are paid the minimum wage In France the minimum wage i ailed thij SMIC (Sdl.iin1 Minimum Inter per onel de Croissance), ir an important element of the collective bargaining system It is set by a council on which both the government and unions are represented Many government employees receive the SMIC In 1987, roughly 8°o of vaqt> earners in industry, commerce and ser»n.es ore covet ed by the SMIC a much hiqher proportion than in the USA (less than 4°o) Wnile the minimum viage in gpneral lias a npgative effect on youth emoloyment, it may Ipad to a suhstitu tion of adults for youths, at the same time ine leasing the employment of the former ,,Table 6.4. Youth Unemployment ana Minimum Wages, 1990 Miniriium vuge as % of: Nonumploynient ratio * Unemployment rate Ai or aqe uaqc Median wage Aqe 14-19 ^ge 20-24 Aqt 20-24 '""Belgium 66 94 1 46.6 13 0 "'Canada 45 2 21.4 'Denmark NonL' 4/9 29 2 14.4 * France 61 91 4 46.0 18.5 -Germany 70 2 29 0 38 -Italy Noiip 86 3 5Ufa 26.6 'Netherlands 72 68.7 31 6 8.9 Portugal 74 65.7 32.1 8.6 Spain 34d 70.0 34.1 30.6 Not il 49.3 29 3 13 1 JJSA Iflll^llillllll 5; :í?S IffiltfKllfff i 48.2 24.2 8.8 •J iThenonemployment ratio is defined as the fraction of the population in the specified age" bracket that is currently not employed. 'iZ'y Youths under 18 are exempted fřonj union wage agreements. ^ ' iA 1 ' v v ,, * ''5 Union agreements restrict employment of'youths in industry and many service sectors. Sources. OECD Jobs Study (1994), Part II; S. Bazen and G. Benhayoun, 'Low Pay and Wage Regulation in the European Community', hfttsh Journal of Industrial Relations, 30 (1992): 623-38, Eurostat, OECD Employment Outlook, 1995, Labour Force Statistics', 1971-91 another job (transition from employment to employment) and most of those who do not get a new job leave the labour force, often for reasons of pregnancy, return to school, retirement, etc. Finally, job losers— those who are zWoluntarily separated—tend to flow into unemployment. Job loss may occur when short-,term contracts expire (common in France and Spain), factories close or relocate, or because of layoffs (more common in the USA and Denmark). 6.4.2. Stocks, Flows, and Frictional Unemployment The fact that no two positions and no two persons are the same means that finding a job is not always easy and may take time. It requires pairing a worker and an unfilled job opening or vacanq'. The matching of skills, occupation, industry, and geographical location requires a large amount of information. The more 154 THE REAL MACROECONOMY Job finds Unemployed Unsuccessful new entries Discouraged Figure 6.18. A Map of Labour Markets Every individual is in one of three states: employed, unemployed] or out of the labour force. At any time, large numbers of workers are flowing from one state to another. ŠBĚm mm mm Mi Table 6.5. Unemployment: Flows and Stocks .in 1987 (millions) Unemployment flows llllflÉSI Into Out of uncmnl unenipl Unemployment stocks Hance Goimany UK 4.12 A.'n . 3.73 3.64 3.03 3.48 2.73 2.50 2.70 Source::Burda and Wyplosz (1994) , efficient the labour markets are, the faster the match is achieved. In the mean time, frictional unemployment occurs. This is an unavoidable result of the dynamics of labour force movements, the normal process of job loss and creation. In addition to the efficiency of the job matching process, frictional unemployment depends on the number of job separations and the number of vacancies. If we ignore the flows from and to 'Not in the labour force' in Figure 6.18, the number of workers who become vmemployed (per month or per year) represents a fraction s, called the separation rate, of existing jobs (I). While unemployment rises by sL workers, it declines as unemployed workers find jobs. If we use /to denote the job finding rate, i.e. the fraction of the unemployed (U) who go into employment, the change in unemployment in a given period is given by 118 (6.4) AU=sL-fU. Frictional unemployment can be understoo stock ^unemployment that is expected to oiun, on average, when unemployment remaiiii -.1 ^.id* Loiul ity of flows into and out of unemploymeni o^.ui> when AU= 0 in (6.4), or when (6-5) Uf = JL. It is convenient to express unemployment a-, i proportion of the labour force N l ( . llvu shows that the frictional rate of unemploymt uf=Uf/N = - (6.6) „ .... s + f Both separation and finding rates can he infe'lj'rctdJ^ as probabilities: the average probability s of lo^ng ^J." job when currently employed, and the avcraj't P'™£li ■' ability/of finding a job whenunemployed#fll|tliy™y**-'" unemployment is larger, the less frequent arej i ih finds. ^ and the more frequent are job separations.:;! The separation rate s has two components strW- ^ -rural and cyclical. The structural aspect is link Ji0 ^ \ ease with which firms dismiss workers. It is . countries where legal and social restrictions c\t^ I ' tL in most European countries) than in eountrif " 4 * redundancies are more acceptable (e.g. tJie K an,. 13 Since?/= L + c/(see(6.3)), 1 = (LIN) + (UlN) piv"'"*^"** by N, and substituting LIN= l — u gives (6.6),, ; v, By demographic group: UK, 1984 ] Aged 16-19 3.33 22.1" Aged 20-24 1.33 16.9 Aged 25-54 0.74 8.8 Aged 55-64 0.47 8.3 White 0 92 10.4 Non-white 1.43 , 20.1 Male 0.78 11.2 Female 1.17 10.2 By skill: UK, 1984 Professional/managerial 0.50 5.3 Clerical 0.88 8.0 Other non-manual 1.14 12.2 Skilled manual 1.02 12.6 Other manual 1.32 15.5 Source. Layard era/ (1991) theTJSA). The cyclical aspect simply refers to the fact that during recessions the probability of losing a job rises and so, therefore, does fractional unemployment. Table 6.6 shows that for a given country (the UK), job separation rates vary considerably across, age, region, and skill groups. Those specific labour force groups that exhibit higher separation rates of inflow into unemployment have higher frictional, and probably therefore overall, unemployment rates. 6.4.3. Job' Finding and the Duration of Unemployment The finding fate /depends on the effectiveness of the matching process. It depends on how hard unem- ployed workers look for jobs, how many job openings are available, and how easy it is to spot an opportunity. It may also depend on incentives to remain unemployed, and unemployment insurance may therefore exert a perverse effect. Because unemployment often represents personal trauma, unemployment benefits or assistance respond to a widely perceived need for solidarity and social conscience. Table 6.7 shows that imemployment benefit systems vary considerably from country to country, with respect to eligibility criteria, income replacement, and the period over which they are paid. At the same time, unemployment benefits have adverse side-effects. They may encourage unemployed workers in declining industries to wait for an unlikely recovery rather than to retrain and change sectors. They also act as a disincentive for looking for a job, or 156 THE REAL MACROECONOIvn Table 6.7. Unemployment: Conditions for Eligibility dnd Benefits —------ —--_-- ----- - —---— — —-----—-- Eligibility Maximum Replacement conditions duration" EiTiployment Period Benefit Assistance Single Married Austria 15C -ks 5 yrs 30 wks Jtlllljj Belgium 90 wks 27 mos 60 De mndil' (C) 3>rs 30 mos 64 " ~ Finland 4 yis 2 yrs 59 . France 12 mos. 24 mo1- 30 mos 59 Gorman^ 3 yrs 4 /is '2 mos 58 • Gteecc 7 mos I2 mos 5 inn, Irclcind 48 i Nop ay 2 yrs 80 . .\ks 11 wkse 1 yrs 52 .\ks USA 20'.\ks 1 yrs 26 .vks 50 V Water Situation on 1 Jan 1989 for a worker40 years old Married' refers to dependent spouse ■ Required period of employment during a reference period, for maximum duration of benefits (e g. to quafifyan Austrract must have been working for at least 156 wks over 5 yrs before becoming unemployed). h Period of eligibility formaximum unemployment benefits and foliow-up assistance > - c Ratio of maximum benefits to previous earning (both before tax) d Denmark, Finland, and JMorWay use a rolling referehce period system " The criterion is. a Tevel of insurance contribution amounting to about 11 wks in most cases. * ', * So ut.* OECD Eiry]ovi'i",tOuifook Ju','1591 111 Sil:' 8W» as an incentive for being 'choosier'.'5 If the benefits are generous, and particularly if they are long-lasting, some unemployed workers may require longer to find 15 Strictly speaking, this applies only to those who already qualify for benefits. In many European countries and the USA, prior work experience is required before one can draw unemployment insurance benefits. In this case, individuals will be more willing to accept the first job. This is often called the 'entitlement effect'. a job that they are willing to accept. 'I In*, is cspn.ulh' true for low wage-earners (see Box 6.2). As Ciii be seen in Figure 6.19, there is a tendencv loi penple to remain unemployed longer in countrn. i «■ hen. unemployment benefits are more generous, paying iru*1' income over longer periods. As the finding ) it-1 &e~. clines, factional unemployment rises. This.eonlirni* _ an uncomfortable trade-off between social centi'ip and economic efficiency. LABOUR MARKETS AND EQUILIBRIUM UNEMPLOYMENT 157 The Equilibrium Rate of Unemploymem .5.1. The Concept f.all unemployment were voluntary, it would hardly ttract any attention. [High and mostly involuntary [^employment means that labour markets do not friction like other markets. A large number of im-ierfeclions, arising from both economic and institu-ional factors, forces us to qualify the pure-competition laradigmof Section 6.2 and to consider an alternate definition of equilibrium to the equality of de-nand for labour by firms and the supply of labour by ouseholds. 'Labour market equilibrium occurs when the un-jmployment rate stabilizes. Because of imperfections, abour markets may be in equilibrium and yet unemployment may not be limited to voluntary unemploy- '.• per i<-> receive. Generosity depends on tin lo»el of tliL' ben^filb and on the period uf time ovei which thuy arc seized. The diagram indicates thai * here the s^sti-rn r more generous:the: proportion ol long-term uiiL-mplo\cd in the total population ot unemplo\ed li ndsto i * Source- Burda (1988) J ment. Equilibrium unemployment can be viewed as the sum of factional and structural unemployment: (6.7) Equilibrium unemployment = factional unemployment + Structural unemployment. Frictional unemployment occurs because it takes time for a match to occur between a worker seeking a job and a vacancy needing to be filled. It depends on the efficiency of the labour market, including the eagerness of both parties to find a match quickly. The frictional unemployment rate may well vary over time, not just because the market's efficiency changes but because economic conditions make it more or less likely for people to find jobs or to become unemployed. Structural unemployment has many causes. The common thread is that the supply of labour is mediated by a number of institutions and regulations. Collective labour supply, which is brought into balance with labour demand in equilibrium, does not quite match individual supply behaviour. Some workers are involuntarily unemployed even when real wages equate the collective supply of labour with the demand of firms. Estimates of equilibrium rates of unemployment are provided in Table 6.8. The contrast between Europe and North America is striking. The equilibrium unemployment rate was generally very low in Europe in the 1960s. Since then it has risen considerably while remaining stable in the USA. A comparison of Table 6.8 and Figure 6.16 shows that actual unemployment Table 6.8. Estimates of Equilibrium Rates of Unemployment 1966-75 1976-85 1986-95 Germany 1.2 4.9 7.1 Italy 5.9 8.3 11.3 Japan 1.6a 2:1 2.5 Spain 2.9" 11.6 20.1 UK 2.6 7.1 8.4 USA 5.2 7.5 6.3 " 1968-75;b 1967-75. Source: OECD 158 THE REAL MACRO ECO NO MY BOX 6.5. THE EUROPEAN WAGE SHOCK AND EUROSCLEROSIS The 1960s witnessed an increase in trade union militancy in several European countries. The May 1968 general strike in France was soon followed by labour unrest in Italy, Belgium, and Germany. In the UK, these were the years when Marxist-led unions exerted influence on the ruling Labour Party. 'Hardline' European unions succeeded in pushing real wages higher and increasing their downward rigidity. Figure 6.20 illustrates the effects of the wage shock by providing a comparison with the USA. The persistence of high unemployment despite improved economic conditions is a consequence of the limited job creation reported in the figure. This dismal performance led to the ilcpiessing bout of 'Eurosrlerosis' ir the early 1980s ' F.n iny jn apparently unstoppable nsinq tide of unemployment, many leaders simply accepted it as stmc tural, and thtrelore beyond policy influence Yet unions cannot overlook completely the plight of the unemployed Over the latter part nf the 1980s unions gtadu ally boi.ed to market pressure and in some countries such Real wage ill' i i i i iii i i i ■85. 1970 72 74 76 78 80 82: 8» 86 88 90 [b) EC/EU Figure 6.20. Employment and Real Wages: Europe and the USA, 1970-1992 Over the period 1970-92, real wages increased by more than 50% in Europe while employment stagnated. At the same time, as real wages stagnated in the USA, employment rose by 50%. (Real wages = total compensation per employee deflated by the GDP deflalor) -9 t"it -3f II has followed the same pattern. To begin to understand this dramatic evolution, we return to the two components of the equilibrium rate of unemployment. 16 The term 'Eurosclerosis' was coined by German economist Herbert Giersch of the Kiel Institute of World Economics. 6.5.2. The European Experience.' I The evidence suggests that J uvoixan "'T-"inrlL,u?Jt rose when large numbers of woi kv. i -. lo>i lh<-*,r 10 ^ the time of the oil shocks. The e\p>-Lied -ul"",L'1"j,'j£ return to pre-oil-shock levels has been thwa'K m LABOUR MARKETS AND EQUILIBRIUM UNEMPLOYMENT 159 i .is- jnarjv KC countries by a fall in the finding rate, so exit from unemployment has become increasingly harder. ic- development of the social safety net to be blamed ,. having pro vi ded workers with the incentive to wait iul iheir unemployment? Circumstantial evidence— joj example Figure 6.20—points in that direction tthtai based on a comparison between Europe, where 1 the- safety net has become extensive, and the USA. Yet there- is some disturbing counter-evidence. The social -,dt<-tynet is even more developed in Sweden and Nor-where long-term unemployment has remained er. This implies that what really matters is not the ty net itself, but the disincentives that it may gen-iiau. Unemployment benefits, for example, provide an llternative to finding a job, and help transform temporary unemployment into permanent—structural I —unemployment. Long-term unemployment has become increasingly widespread, and as workers ■TaduaDy lose their human capital and contact with "the active labour force, they become unsuitable for any \acdncy. ,J ' The strildngly different evolution of the equilibrium iin<. tnployinent rate across countries also points to the importance of institutions in shaping the levels of w jc.es. This concerns the process of wage bargaining. I be comparison between the EC and the USA in Box 6.5 shows thai high unemployment in the EC is re-laud to steep real wage increases, amounting to what has been dubbed the European 'wage shock'. Labour tusts1 consist not only of wages: labour taxes (social '"security and retirement contributions) have also been allowed to rise steeply. Of importance too is the regulation of the use of labour (length of the work-week, dismissal procedures, part-time work, etc.).17 6.5.3. mum ■Unemployment1 It can take a long time, often years, before real wages actually adjust to their long-run values shown in Figures 6.10 and 6.15. In the meantime, actual unemployment can deviate from equihbrium unemployment. 7 Ch. 17 provides a more detailed exposition. 1 8 This section briefly presents an issue to be explored at length in Ch. 12. Demand Figure 6.21. Actual and Equilibrium Employment When unions negotiate on behalf of workers, market equilibrium occurs at point A and equilibrium unemployment is L - L Actual employment and unemployment may differ if the real wage is slow to move to its equilibrium level w. If it is above the market equilibrium level (w, > w), firms reduce employmenttoZ., and actual unemployment exceeds equilibrium unemployment. Conversely, below-equilibrium real wages (w2 < w) allow firms temporarily to find a way of con-necting-with structurally unemployed workers willing to work at lower wages than the union-set level. The resulting unemployment rate is lower than the equilibrium level. Actual employment is below, and actual unemployment above, equihbrium when the real wage is above the equilibrium level, as at point A1 in Figure 6.21. When the real wage is low, firms may be able temporarily to move away from the union-set collective labour supply curve towards the individual labour supply curve (point A2), for example by using agencies specializing in temporary jobs or overtime work. Workers may have overestimated the real wage by underestimating the rise in the price level. Firms may be willing to hire more workers at the going wage. In such situations employment is above, and unemployment below, the equihbrium level. 6.6. Summary 1. Households trade off leisure against consumption (more generally, labour income An increase in wages can induce more labour supply if the substitution effect domm ates (elastic labour supply), less labour supply if the income effect dominates; (back ward-bending), or no change at all (inelastic supply). 2. Individual labour supply seems to be inelastic in the short run and backward-bending in the long run as real wage gains are taken partly in the form of additionalieisuct. Aggregate labour supply is more responsive in the short run as real wage increases draw new individuals into the labour force. 3. The demand for labour by firms depends on its (marginal) productivity which is determined by the available technology and the capital stock. Firms hire labour to the point at which the marginal productivity of labour is equal to the real wage. Thelabour. demand schedule is shifted outwards by an improvement in technology or an i tk i 1uv in the capital stock. 4. Equilibrium employment and the wage level are given by the intersection ol lab»ur demand and labour supply. Improvements in technology or increases in capital- wifej be reflected in higher wages if labour supply is inelastic, and in higher employm labour supply is elastic. 5. Involuntary unemployment arises when real wages do not clear the market so tE&p not all labour supplied by households is actually hired. 6. Labour unions care about real wages and employment. In determining their target wage, given the demand for labour firms, they ask for higher real wages than iftiffin labour market were perfectly competitive. While the resulting unemployment uie^'-(optimal and) voluntary for unions, it may be involuntary for individuals., 7. Very centralized or decentralized wage negotiations deliver lower real-\\a*''- less unemployment than negotiations taking place at intermediate levels of centra^l^^ tion (industry by industry, or by craft). . -j,, e ' \ 8. Because firms cannot easily monitor work effort or wish to elicit lower tu' novel v» y improve worker quality, they may offer efficiency wages. This is yet anofhei ieas0,^ why real wages may be set above market-clearing levels. ..v::.: ' i',^ "* n's *'* 9. Labour markets are also characterized by widespread government inteneiitio ^ ^ Minimum wages, designed to protect workers, can actually cause unemplo1 men , ^ 10. The labour market is characterized by a considerable amount of flow b ^ different states (employment, unemployment, not in the labour force). Se.'i j-important aspect and results in frictional unemployment. Alongside structui al u ^ ployment, it is a source of equilibrium unemployment. ■ - % 11. Search maybe more or less efficient. It is affected by go\ernment k'b'u1' policies. Unemployment benefits, designed to make unemployment n>or^ provide disincentives to qm^ldyfindinganewjob,therebyincreasingfri>-til,ni "ijt 1 LABOUR MARKETS AND EQUILIBRIUM UNEMPLOYMENT 161 ployment. Other programmes, such as training and relocation subsidies, reduce fric-tional unemployment. 12. Because of distortions and regulations, equilibrium unemployment is never zero or entirely voluntary. Individuals may be willing to work at lower wages than those prevailing in equilibrium, but they cannot underbid. This is the sense in which real wages are downwardly rigid. 13. Real wages are slow to adjust to disequilibria, if only because they fulfil many other roles. As a result, actual and equilibrium unemployment may differ for long periods of time. © leisure • consumption-leisure trade-off • real (consumption) wage « income effect, substitution effect • labour supply: individual and aggregate • labour force «> participation rate c man-hours • marginal productivity of labour (MPL) • labour demand » wage and profit shares • voluntary and involuntary unemployment • real wage rigidity • trade unions, labour unions » collectively volimtary/mdividually involuntary unemployment • collective labour supply curve • hysteresis • mhrimum wages • efficiency wages • unemployment benefits • unemployment stocks and flows • separation and separation rate • fmdingrate • frictional unemployment « equilibrium unemployment • structural unemployment Exercises: Theory 1. Suppose that the household in Figure 6.3 receives an inheritance. What is the effect on its decision to work and to consume? According to this result, do rich people work more or less than poor people? 162 THE REAL MACROECONOMY 2. Suppose Robinson Crusoe is paid a higher wage ('overtime') if he works more than 8 hours a day, but only has 16 hours at his disposal. (a) Draw his budget constraint in this case. (b) Does the existence of overtime necessarily make him better off? (c) Show Crusoe's optimal behaviour for normal' indifference curves. Under w 11 n p conditions will he choose to work overtime? Under which conditions will he 1> % refuse? - c. 3. Normally Crusoe sleeps eight hours a day. Suppose it were possible for him to stay awake an extra hour per day without any impairment of his abilities. What would' _ iJ> the effect be (a) on his consumption? (b) on the amount of leisure taken? How , JJ' much would he be willing to pay for this innovation? *Jf 4. How should the equilibrium rate of unemployment respond (a) to an exogenous » -r ^ decrease in investment over several years? (b) to an exogenous increase in produc- V ">| tivity? How does your answer depend on the institutions of wage determination? ■ ^ y 5. What is the effect on the labour market of a minimum wage that is actuallylowei than the equilibrium wage? Show in your graphical answer the new eqiulibrium .'V wage and the level of employment. -?llt 6. A new labour tax is imposed which is proportional to wages. Individuals care about' - , after-tax wages. (a) Draw the old and new individual supply curves. (b) What is the effect on the equilibrium wage and employment levels? (c) What is changed if wages are set through negotiations with a trade union whiclj also cares about after-tax wages? 7. Many countries have housing rent control; i.e. rent increases are tightly controlled ^ below levels that would prevail in a free market. How might this affect equilibnuifr|p unemployment? (Hint: think about labour mobility and mismatch.) 8. What is the effect on wages, employment, and unemployment of a wave oi unmi gration: j (a) in the absence of minimum wage legislation? (b) in the presence of minimum wage legislation? ■<■::•■ fli^l? 9. How should Figure 6.13 be redrawn if unemployment benefits are generall1 avail able and provide a floor under which wages cannot fall? (Hint: the.'trade union * would never accept wages lower than the unemployment benefits.); Whal i-> implication for the collective labour supply curve? LABOUR MARKETS AND EQUILIBRIUM UNEMPLOYMENT 163 Exercises: Applications 1. One of the hnmediate consequences of opening the borders between East and West Germany was the potential for migration between the two regions. Because of a more productive capital stock and more know-how, wages in the West were about three times as much as those in the East. Consequently many East Germans moved to the West. What are the consequences for this migration for real wages (a) in West Germany? (b) in East Germany? (c) for employment in the two regions? Can you explain why West German trade unions were eager to organize their comrades in the East? 2. It is often said that unemployment insurance in Europe was always generous and thus cannot be responsible for the present high unemployment on the Continent, which has emerged only since the mid-1970s. Examples are Germany, France, and Belgium. Evaluate this statement. 3. In Japan the bonus system is widespread. Workers often receive 30% of their pay in the form of a profit-contingent payment, which can go up or down depending on the fortunes of the enterprise in which they work. What are the implications of such a system for real wage rigidity and equilibrium employment? 4. It is a fact that in fishing communities and tourist areas the average rate of unemployment over long periods of time is higher than in highly industrialized regions. How could our concept of frictional unemployment explain this fact? 5. Many newly industrializing countries, such as Korea, have underdeveloped or repressed trade union movements. These countries also have a very low wage share, and thus a high rate of return on physical capital. How might this be explained? 6. Youth employment is very low in Germany, at least in comparison with other European countries. It is often asserted that this is because training for youth is subsidized by the state. Is this a good use of public money? 7. It is often the case that unemployment benefits are paid out of a fund financed by taxes levied on the firms proportionately to their wage bill. How might this affect equilibrium unemployment? 8. It is sometimes suggested that the massive influx of women into the labour force is a cause of unemployment. (a) Draw the effect on the labour market as described by Figure 6.10. What does it mean for real wages and employment? (b) Using Chapter 4, note that more available labour raises the MPK. Assuming that, as a result of more investment, the capital-stock increases: how does your answer to (a) change? Suggested Further Reading Two textbooks provide further analyses: Ehrenberg, Ronald, and Smith, Robert (1988), Modern Labour Economics, 3rd edn., Scott Foresman, Glenview, 111. and London. Layard, Richard, Nickell, Stephen, and Jackman, Richard (1991), Unemployment, Oxford University Press. High European unemployment has been studied in detail. A survey is: Bean, Charles (1994), 'European Unemployment: a Survey, Journal of Economic Lit-: erature, 32: 573-619. See also the special issue of the Swedish Economic Policy Review, 1 (1-2), Autumn 1994 ; (articles by Lars Calmfors, Jorgen Elmeskov, Charles Wyplosz, Patrick Minford, Rudiger Dornbusch, Jacques Dreze and Henri Sneessens, and Richard Jackui m . Some interesting contributions are Bentolila, Samuel, and Dolado, Juan (1994), 'Labour Flexibility and Wages: Lessons' from Spain', Economic Policy, 18: 53-100. Blanchard, Olivier J., and Summers, Lawrence H. (1986), 'Hysteresis and tlu' Wun- pean Unemployment Problem', NBER Macroeconomics Annual, 1: 15-77. Blanchflower, David, and Oswald, Andrew (1994), The Wage Curve, MIT Prei.s, ( am bridge, Mass. ' Burda, Michael (1988), 'Wait Unemployment in Europe', Economic Policy, 7: 39H«";f 426. Calmfors, Lars, and Driffill, John (1988), 'Bargaining Structure Corporatism andl» Macroeconomic Performance', Economic Policy, 6:13-62. Lindbeck, Assar (1993), Unemployment and Macroeconomics, MIT Press,; CambridgS^ Mass. , .J Saint-Paul, Gilles (1993), 'On the Political Economy of Labour Market Fie NBER Macroeconomic Annual, 8: 151-86. A controversy on the effects on minimum wages has arisen: see Card, David, and Krueger, Alan B. (1995), Myth and Measurement: the New E^'iotmc? * of the Minimum Wage, Princeton University Press. J; Machin, Steven, and Manning, Alan (1994), 'Minimum Wages, Wajv I 'ispt i'-ion anf, £ Employment: Evidence from the UK Wage Councils', Industrialand labour Hons Review, 47 (2): 319-29. "''^ Kennan, John (1995), 'The Elusive Effects of Minimum Wages', Journal ofE.onofflt<-< \ Literature, 33:1949-65. LABOUR MARKETS AND EQUILIBRIUM UNEMPLOYMENT 165 Appendix: Labour Supply and Demand, and the Collective Labour Supply Curve Household Labour Supply ibinson Crusoe maximizes his utility function defined over ttsumption Cand leisure €: 5.1) U=U(CJ), Dject to the budget constraint 6.2) w£ = w( + C, lere w is the real (consumption) wage and ( is Crusoe's le-endowment. The first-order condition is found by sub-Tiling C from (A6.2) into (A6.1) and maximizing with pect to £: «) V,IUC Theiratio'offhe marginal utilities vis-á-visleisure (t/f) and t consumption {(./.; is the marginal rate of substitution and ' the; (absolute value of the) slope of the indifference curve. This measures the ratio of the changes in leisure and consumptionthat keep the utility level unchanged. At the opti-muru,5.the amounts substituted are such that this ratio is equal to the real wage rate. Hhe?iadividual labour supply curve is the (imphcit) func-■tioii household labour supply L as to real wage w defined by \(5.3): (A6.4) : I'M, ^iisre the wealth argument has been suppressed. Aggregation over many households results in (Afi.5) U=If(w,L), where L is the number of individuals of working age. The Demand for Labour by Firms Since capital is fixed, we can write the production function simply as Y = F(K, L) = F(L). Express the firm's profit in , terms of units of output, or as ,(A6.6) XT = P(L) - wL. 'Choosing employment to maximize (A6.6) gives (A6.7) MPL = F(L) = w. The inverse of relationship (A6.7) determines the demand for labour; (A6.8) L = L(w) = F'-\w), with L'(w) < 0. Wages and the Labour Share The wage share is s = wL(w)/F(L). What is the effect of an increase in the real wage on s? Differentiating this expression yields (A6.9) dw~~ Y Define two elasticities as 1 + - wL' wF'L' Y r\Lw = elasticity of labour demand to the real wage _ dL w _ wL' Aw L L QYL = elasticity of output with respect to labour - iY 1 - F'L ~ dL Y ~ Y (A6.9) can be rewritten as (A6.10) |i = l[i_tl fi_eK)]. aw Y Since F' = w, Qyl is ^so labour's share in value added and (1 - Oyx) is the profit share. Thus, the condition for the wage share to be increasing when the real wage rises is that X[L„ < 1/(1 - Qyi), or that the elasticity of labour demand be less than the inverse of the profit share. This is the more likely to happen the smaller is the profit share and the less elastic is the demand for labour, i.e. if the demand for labour does not decline 'too much' in response to an increase in real wages. Collective Labour Supply Consider a labour union that has utility given by U(w - w, L- L). In this formulation (sometimes known as a Stone-Geary specification), utility is dependent on the excess of wages and employment above exogenous reference levels, here w and I. We assume that Uw > 0, UL > 0 while LTm„ < 0, Uu < 0, and UwL > 0. We consider only the case of a monopoly labour union, which sets the wage unilaterally.19 The union maximizes utility subject to the labour demand curve (A6.8), by choosing the real wage w that maximizes U(w— w,L(w)- I). The first-order condition is (A6.ll) = -L'(w) 19 For more sophisticated treatments of the collective bargaining process, which allow for bargaining between unions and employer associations, see an advanced textbook, e.g. Booth (1995). 166 THE REAL MACROECONOMY where the derivatives are evaluated at the optimum value. Equation (A6.ll) has an interpretation similar to that of (A6.3): the union picks a wage such that the marginal rate of substitution of employment for wages is equal to the slope of the labour demand curve. This corresponds to points of tangency in the various panels traced out in Figure 6.14. We can solve this condition by considering a specific form of the utility function:20 (A6.12) U(w- w,L - I) = [cc(w_- w)1/p + (l-a)(I- L)5p]"p, with 1 > p > -m. First consider the case y = 5 = 1. It can be shown that the elasticity of substitution between adjusted employment and wages is given by (1 - p)"1. Asp approaches 0, the function assumes the form (A6.13) U(w- w,I - I) = (w- w)a(L - I)""*', whereas the case p —> -» gives 2D This form exhibits constant elasticity of substitution between isoeiastic functions of 'normalized' employment and wages. (A6.13') U(w~ w,L- L)= min[(w- w), (I - L)], and p = 1 yields the linear form (A6.13") U(w- w,L~ L) = a(w- w) + (l-a)(I- I). An alternative class of union preferences arises when p = 1" with y or 5 less than unity: (A6.14) U(w- w,L- I) = a(w- w)1 + (l-a)(L- I)6. Panels_(n) and {b) of Figure 6.14 correspond to the case w = 1=0. The union in panel (a) places relatively high, weight on wages, whereas the union in panel (b) places more; weight on employment. In panels (c) and (d), the reference;' level for wages and employment are nonzero, and p = 1 is assumed. In (c) y = 1, 5 < 1, w = 0, but I > 0 would lead to vertical collective labour supply curves. L > 0 cuuld be understood as the 'membership' or reference gro up. la panel (d),y< 1,8 = 1, L = O.but w > 0; mis implies a horizontal*' collective labour supply curve, where w can be thought of as the 'target wage' for the union. ^^^^ amSi MR