Abstract
The conceptual framework of this paper assumes that macroeconomic performance depends on the interplay between the economic environment and policies. Declining labor shares, wage moderation, and employment performance in Germany and the Netherlands have been presented. A number of policy changes are under way, but additional reforms may be needed to fully reap the benefits of the new economy. The tax reform package marks a radical and constructive shift in German tax policy, and the pension system requires a sea of change in public policy reforms.
II. Declining Labor Shares, Wage Moderation, and Employment Performance in Germany and the Netherlands15
A. Introduction and Overview
26. Since the early 1980s the share of labor in national income has trended significantly downward in both Germany and the Netherlands, owing to prolonged periods of moderate wage increases. In Germany, the downward trend in the labor share was accompanied by an upward ratcheting of the unemployment rate, while in the Netherlands the unemployment rate declined (Figure II-1).
Labor’s Share, Unemployment Rate, Real Wages and Average Labor Productivity in Germany and the Netherlands, 1970–99 1/
Citation: IMF Staff Country Reports 2000, 142; 10.5089/9781451810424.002.A002
Sources: OECD analytical database; and staff calculations.1/ Prior to 1991, west Germany only. The discontinuity in both the real wage and the average labor productivity is due to unification.Labor’s Share, Unemployment Rate, Real Wages and Average Labor Productivity in Germany and the Netherlands, 1970–99 1/
Citation: IMF Staff Country Reports 2000, 142; 10.5089/9781451810424.002.A002
Sources: OECD analytical database; and staff calculations.1/ Prior to 1991, west Germany only. The discontinuity in both the real wage and the average labor productivity is due to unification.Labor’s Share, Unemployment Rate, Real Wages and Average Labor Productivity in Germany and the Netherlands, 1970–99 1/
Citation: IMF Staff Country Reports 2000, 142; 10.5089/9781451810424.002.A002
Sources: OECD analytical database; and staff calculations.1/ Prior to 1991, west Germany only. The discontinuity in both the real wage and the average labor productivity is due to unification.27. This chapter analyzes what has been driving the decline in labor share in both countries and offers an explanation for the rather different unemployment performance. Labor’s share is defined as the ratio of aggregate labor income to national income, which can also be expressed as the ratio of the average real wage per worker to average labor productivity. Because a reduction in the labor share implies that real wages grow by less than average labor productivity, it is often concluded that a decline in the labor share must go hand-in-hand with falling unemployment. However, models described in Section B and the Appendix show that a decline in the labor share is often the result of adverse labor market shocks that, when they interact with labor market institutions, may well be associated with rising unemployment. Section C thus takes a detailed look at the type of labor market shocks that have affected Germany and the Netherlands and examines how policies and labor market institutions have dealt with them. In doing so, it sheds light on why across-the-board wage moderation has apparently been less successful in lowering unemployment in Germany than in the Netherlands. Conclusions are presented in Section D.
B. Unemployment and Labor Share
28. Theoretical models point to the importance of distinguishing between shocks to labor supply and demand in explaining the relationship between labor share and unemployment. The adjustment of wages and employment to such shocks depends on the institutional structure of the labor market and may take many years to complete.
29. The basic insight of such models is relatively straightforward.16 Wage moderation might be expected to follow an adverse shock to supply or demand, and thereby reduce labor share. But, at least in the adjustment period, substitution of capital for labor may raise unemployment. In the case of an adverse shock to labor demand—for example, because of a labor-saving change in production technology—the initial impact will be a decline in the labor share as both the real product wage and employment fall. Over time, firms respond to lower wages and higher profits by increasing capital and rehiring workers. Eventually, the economy will return to the employment level and the real product wage that prevailed before the shock but with more capital and a higher level of economic activity. The fact that activity is higher implies a lower labor share.
30. By contrast, a labor supply shock—for example, a failure of wages to adjust to a slowdown in the rate of growth total factor productivity, or an increase in unemployment benefits—will lead initially to an increase in labor share. In response to the resulting lower profits, firms will scale down labor, capital, and output leading to a gradual reversal of the increase in labor share. The new long-run equilibrium will be characterized by a return to the original capital-labor ratio and labor share, but at a lower level of employment (and hence higher level of unemployment). In both the cases of adverse supply or demand shocks, adjustment to the new equilibrium will be slow if the elasticity of substitution between labor and capital is low or there is considerable rigidity in wage setting.
31. The basic model can be extended to incorporate a pertinent feature of the German labor market—namely that labor is not a homogeneous factor of production. In the simple case where the labor market is split into high- and low-skilled sectors, shocks need not hit each sector symmetrically. As a result, there are additional channels of adjustment through the relative wages between skilled and unskilled workers (the skill premium). Attempts to resist a change in the skill premium because, for example, wage increases are coordinated between sectors or because minimum wage levels cannot be lowered, will perpetuate the impact of shocks and, at best, make higher unemployment more persistent.
C. Labor Market Performance in Germany and the Netherlands
32. Using business sector data, this section attempts to identify the labor market shocks in Germany and the Netherlands over the last three decades, and to shed light on why the countries’ employment performance has differed despite similarities in wage setting institutions. The sustainability of across-the-board wage moderation as a strategy for lowering unemployment is also analyzed.
Labor market shocks
33. Using the model detailed in the Appendix, shifts in labor supply and demand can be derived. The results suggest that, with respect to labor supply, Germany has been slower than the Netherlands in reversing early adverse shocks to wage setting, particularly at the beginning of the 1990s when its strategy of moderate wage growth was derailed by unification. With respect to labor demand, both countries suffered sizable adverse shocks, although the initial shock was larger in the Netherlands.
34. For the 1970s, labor market developments in Germany and the Netherlands are well captured by adverse labor supply shocks, triggered by a failure of wages to adjust to a slowdown in productivity growth, and to the impact of the oil price hikes (Figure II-2, panel I).17 The first panel of Figure II-2 shows that the real product wage in both countries moved away from a level that would allow the economies to return to their 1970 unemployment rates without exerting inflationary pressure. By 1980 the real product wage in both Germany and the Netherlands was too high by about the same amount (assuming, that is, that the real product wage in both countries is equally sensitive to changes in unemployment). The increase in the real product wage also pushed up the labor shares but, as the models would have predicted, a prolonged decline in both variables subsequently set in.
Labor Market Shocks in Germany and the Netherlands, 1970–99
Citation: IMF Staff Country Reports 2000, 142; 10.5089/9781451810424.002.A002
Sources: OECD analytical database; and staff calculations.Labor Market Shocks in Germany and the Netherlands, 1970–99
Citation: IMF Staff Country Reports 2000, 142; 10.5089/9781451810424.002.A002
Sources: OECD analytical database; and staff calculations.Labor Market Shocks in Germany and the Netherlands, 1970–99
Citation: IMF Staff Country Reports 2000, 142; 10.5089/9781451810424.002.A002
Sources: OECD analytical database; and staff calculations.35. In the 1980s the lagged labor demand adjustments to the labor supply shocks of the 1970s coincided with new shocks to labor demand (Figure II-2, panel 2). The demand shocks were similar in profile in both countries, although the initial drop in labor demand was somewhat larger in the Netherlands. This was probably triggered by a slightly deeper recession than in Germany.18 As a result of the labor demand shocks, there was a sharp decline in the labor share in the first half of the 1980s and a large increase in the unemployment rate in both countries.19 Although it is difficult to determine exactly what caused the shifts away from labor, the framework discussed in Section B suggests the following explanations:
adverse labor demand shifts in response to the earlier shifts in the labor supply curve;
a sharp rise in the real interest rate in the beginning of the 1980s that lowered the long-run labor demand curve, leading to a lower real product wage and higher unemployment; and
technological bias away from unskilled labor, which—under the assumption that unskilled labor and capital are better substitutes than skilled and unskilled labor—would lead to a genuine shift away from labor.20
36. The steady reversal of the labor supply shock since the mid-1980s (i.e., the downward movements in Figure II-2, panel 1) can be viewed as the supply-side reaction to the adverse labor demand shocks of the early 1980s, as both countries tried to halt the surge in unemployment through wage moderation. In the Netherlands, the Wassenaar Agreement of 1982 marked the introduction of a widely-supported strategy of wage moderation that was implemented to turn around the dismal performance of the Dutch economy and to stop excessive wage growth. The government played a prominent role in this process by strongly encouraging wage moderation and by threatening to interfere if wage growth were to get out of hand. In Germany, wage moderation followed a more haphazard pattern but invariably occurred after large labor shakeouts. Contrary to the Netherlands, the German government had no role in the bargaining process.21 In both countries the wage bargaining mechanism was wed to the principle of wage solidarity so that wage moderation applied across the board.
37. In the early 1990s, the shock that featured most prominently was the adverse shock to wage-setting that arose in Germany when, after unification, wages in east Germany embarked on a process of catch-up to western levels. The catch-up was not justified by productivity levels, which were much lower in the east due to an outdated and non-market oriented capital stock. Moreover, to the extent that the supply of unskilled labor increased most, wage rigidities exacerbated the shock’s negative impact on total employment.
Moderation across the board; why has the Dutch approach been sustainable?
38. With the exception of German unification, the reasons behind the decline in the labor shares in Germany and the Netherlands appear to be very similar. The same is true for the approach that both countries have taken to wage moderation. Why then did unemployment steadily decline in the Netherlands but not in Germany?
39. Wage moderation in the Netherlands was perhaps successful in lowering unemployment because it was more aggressive, particularly at the outset. Not only did wage growth slow relative to average labor productivity growth, but real wage increases were often below what was warranted by the rise in labor-augmenting technology leading to a sharp decline in the real product wage (Figure II-3). German wage moderation was less aggressive, so that the negative differences between actual and warranted wage growth were usually smaller than in the Netherlands (Figure II-4). While the German process of moderate wage growth was derailed by unification, the Dutch strategy was helped by a reduction in the income tax wedge (which softened the impact on net wages of the limited increases in gross wages, Figure II-5),22 and by an increase in the participation rate, exercising downward pressure on wages.
The Real Product Wage in and the Netherlands,
Citation: IMF Staff Country Reports 2000, 142; 10.5089/9781451810424.002.A002
Sources: OECD analytical database; and staff calculations.1/ The growth rate of the warranted real wage is determined by the growth rate of the laboraugmenting technology parameter, a.The Real Product Wage in and the Netherlands,
Citation: IMF Staff Country Reports 2000, 142; 10.5089/9781451810424.002.A002
Sources: OECD analytical database; and staff calculations.1/ The growth rate of the warranted real wage is determined by the growth rate of the laboraugmenting technology parameter, a.The Real Product Wage in and the Netherlands,
Citation: IMF Staff Country Reports 2000, 142; 10.5089/9781451810424.002.A002
Sources: OECD analytical database; and staff calculations.1/ The growth rate of the warranted real wage is determined by the growth rate of the laboraugmenting technology parameter, a.
Growth in the Actual Growth in the Warranted Germany and the Netherlands,
(In percent)
Citation: IMF Staff Country Reports 2000, 142; 10.5089/9781451810424.002.A002
Growth in the Actual Growth in the Warranted Germany and the Netherlands,
(In percent)
Citation: IMF Staff Country Reports 2000, 142; 10.5089/9781451810424.002.A002
Growth in the Actual Growth in the Warranted Germany and the Netherlands,
(In percent)
Citation: IMF Staff Country Reports 2000, 142; 10.5089/9781451810424.002.A002
Average Labor Tax Wedge in Germany and the Netherlands, 1982–97
(In percent) 1/
Citation: IMF Staff Country Reports 2000, 142; 10.5089/9781451810424.002.A002
Source: OECD.1/ Average tax wedge for a married couple, single earner (average income) with two children taking account of social insurance contributions, personal income taxes, and family allowances; as a percentage of gross labor costs.Average Labor Tax Wedge in Germany and the Netherlands, 1982–97
(In percent) 1/
Citation: IMF Staff Country Reports 2000, 142; 10.5089/9781451810424.002.A002
Source: OECD.1/ Average tax wedge for a married couple, single earner (average income) with two children taking account of social insurance contributions, personal income taxes, and family allowances; as a percentage of gross labor costs.Average Labor Tax Wedge in Germany and the Netherlands, 1982–97
(In percent) 1/
Citation: IMF Staff Country Reports 2000, 142; 10.5089/9781451810424.002.A002
Source: OECD.1/ Average tax wedge for a married couple, single earner (average income) with two children taking account of social insurance contributions, personal income taxes, and family allowances; as a percentage of gross labor costs.40. The large degree of wage moderation in the Netherlands, plus the fact that it was achieved mostly through across-the-board limits on wage growth, suggests that the Netherlands did not suffer much from the disequilibrium that is typically created by across-the-board wage moderation in the face of non-neutral shocks. How can this be explained?
41. First, the Netherlands did experience some increase in the wage differential between high- and low-paid workers, while in Germany the differential narrowed (Figure II-6). The increase was due primarily to a nominal freeze of the Dutch minimum wage during most of the 1980s and 1990s, an increase in the differentiation of the minimum wage according to age, and the establishment of wage scales below the sectoral (bargained) minimum pay.
Developments in Earnings Inequality in Germany and the Netherlands, 1982–94 1/
Citation: IMF Staff Country Reports 2000, 142; 10.5089/9781451810424.002.A002
Source: OECD Employment Outlook, 1996.1/ Defined as the ninth over the first decile in gross earnings.Developments in Earnings Inequality in Germany and the Netherlands, 1982–94 1/
Citation: IMF Staff Country Reports 2000, 142; 10.5089/9781451810424.002.A002
Source: OECD Employment Outlook, 1996.1/ Defined as the ninth over the first decile in gross earnings.Developments in Earnings Inequality in Germany and the Netherlands, 1982–94 1/
Citation: IMF Staff Country Reports 2000, 142; 10.5089/9781451810424.002.A002
Source: OECD Employment Outlook, 1996.1/ Defined as the ninth over the first decile in gross earnings.42. Putting aside the slight increase in the differentiation in gross earnings, there are a few reasons why wage differentiation may have been less urgent in the Netherlands:
The policy measures of the early 1980s induced a rebound of the Dutch economy and a recovery of firms’ profits. By the end of the 1980s, per capita GDP growth in the Netherlands had surpassed that in many other European countries and during most of the 1990s also exceeded that in Germany. Higher growth reduced the bias against the unskilled that is typical of large cyclical downturns.
The lower capital-intensity in the Netherlands relative to Germany also reduced its demand bias away from unskilled labor. Economic growth after the mid-1980s was very labor intensive. This shift to a more labor-intensive growth pattern was largely endogenous to the policy of wage moderation, which (as employers became convinced that wage moderation was a long-term strategy) made labor more attractive. It was also aided, however, by a large and expanding service sector which, in the early 1980s, already occupied two-thirds of all workers.23 Indeed, both the capital-output and the capital-labor ratio show that capital-for-labor substitution was less pronounced in the Netherlands (Figure II-7).
From the early 1980s onward, a sharp increase in the labor force participation rate took place in the Netherlands (Figure II-8). As in many other countries, the composition of the labor force also moved to a greater share of skilled labor, thus putting downward pressure on the wages of the skilled. The fact that much of the increase came through part-time workers and women who had been out of work for a while perhaps also implied that these entrants had a lower reservation wage than existing labor market participants. This development contrasts sharply with the labor supply shock in Germany: unification raised the relative supply of low-skilled labor24 and was accompanied by a policy of wage catch-up for the east.
The Netherlands achieved an increase in the cost differential of skilled and unskilled labor through a differentiation in non-wage labor costs that largely left intact the relative gross wage. Starting in 1983, taxes and social contributions of employees were cut substantially.25 In the beginning of the 1990s several measures were taken to cut the employers’ costs of hiring unskilled workers, mainly through cuts in the employers’ social contributions (Figure II-9). The Dutch social insurance system differs from the German system in that it does not strictly adhere to the contribution-benefit parity and thus enjoys a larger degree of freedom in restructuring contributions. An example of these measures is the social contribution concession (the SPAK) for employers hiring workers at an hourly wage of up to 115 percent of the minimum wage. The scheme operates together with large subsidies for hiring long-term unemployed; if workers meet both criteria, the employers’ contributions fall to zero.26
Finally, the Netherlands addressed the high reservation wages at the low end of the market by a cut in the unemployment and disability replacement rate from 80 percent to 70 percent of the last earned wage.
Capital-Labor and Capital-Output Ratios in Germany and the Netherlands, 1970–99
Citation: IMF Staff Country Reports 2000, 142; 10.5089/9781451810424.002.A002
Sources: OECD analytical database; and staff calculations.1/ The discontinuity is the direct result of unification.Capital-Labor and Capital-Output Ratios in Germany and the Netherlands, 1970–99
Citation: IMF Staff Country Reports 2000, 142; 10.5089/9781451810424.002.A002
Sources: OECD analytical database; and staff calculations.1/ The discontinuity is the direct result of unification.Capital-Labor and Capital-Output Ratios in Germany and the Netherlands, 1970–99
Citation: IMF Staff Country Reports 2000, 142; 10.5089/9781451810424.002.A002
Sources: OECD analytical database; and staff calculations.1/ The discontinuity is the direct result of unification.
Labor Force Composition in Germany and the Netherlands, 1973–97
Citation: IMF Staff Country Reports 2000, 142; 10.5089/9781451810424.002.A002
Sources: OECD Employment Outlook 1996 and 1999; Sociale Nota 1999; and staff calculations.1/ Defined as a share of total working age population.Labor Force Composition in Germany and the Netherlands, 1973–97
Citation: IMF Staff Country Reports 2000, 142; 10.5089/9781451810424.002.A002
Sources: OECD Employment Outlook 1996 and 1999; Sociale Nota 1999; and staff calculations.1/ Defined as a share of total working age population.Labor Force Composition in Germany and the Netherlands, 1973–97
Citation: IMF Staff Country Reports 2000, 142; 10.5089/9781451810424.002.A002
Sources: OECD Employment Outlook 1996 and 1999; Sociale Nota 1999; and staff calculations.1/ Defined as a share of total working age population.
Employers’ Social Contributions in Germany and the Netherlands, 1979–99
(In percent) 1/
Citation: IMF Staff Country Reports 2000, 142; 10.5089/9781451810424.002.A002
Source: OECD, The Tax/Benefit Position of Employees, several isssues.1/ Employers’ social insurance contributions for a married couple, single earner (average income) with two children; as a percentage of gross labor costs.Employers’ Social Contributions in Germany and the Netherlands, 1979–99
(In percent) 1/
Citation: IMF Staff Country Reports 2000, 142; 10.5089/9781451810424.002.A002
Source: OECD, The Tax/Benefit Position of Employees, several isssues.1/ Employers’ social insurance contributions for a married couple, single earner (average income) with two children; as a percentage of gross labor costs.Employers’ Social Contributions in Germany and the Netherlands, 1979–99
(In percent) 1/
Citation: IMF Staff Country Reports 2000, 142; 10.5089/9781451810424.002.A002
Source: OECD, The Tax/Benefit Position of Employees, several isssues.1/ Employers’ social insurance contributions for a married couple, single earner (average income) with two children; as a percentage of gross labor costs.43. To sum up: The effectiveness and sustainability of across-the-board wage moderation in the Netherlands depended greatly on timing and the right combination of additional policy measures. Wage moderation was effective because it was introduced as the economy was in a severe recession, which led to broad agreement on the necessity of wage moderation and allowed for an aggressive approach. The across-the-board aspect of the approach was feasible because additional measures (including tax cuts, and reforms at the low end of the market) either put downward pressure on wages at the high end of the labor market, or ensured differentiation through non-wage labor costs. The result of the Dutch strategy has been a large decline in total unemployment over the last two decades, a development that has also put the unskilled in a somewhat better position. Whether the Dutch strategy has promise for the future is less clear; the current unemployment rate is low and shortages of skilled labor have arisen. This could put upward pressure on the wages of the skilled and, if wage solidarity is maintained, on those of the unskilled.
D. Conclusions
44. The evolution of the labor share and a country’s employment performance depend on the type of shocks that affect the labor market, as well as on the institutional structure of this market. Some form of wage moderation is essential to prevent rising unemployment in the face of an adverse shock. In a market with heterogeneous labor, moreover, moderation would need to be aggressive, given that it must aim to bring wage increases more in line with labor productivity growth, particularly at the low end of the labor market. Data suggest that in the Netherlands wage growth has, in the past, substantially undercut warranted wage growth, while in Germany wage moderation has been more modest and, in recent years, hampered by the lingering wage effects of unification.
45. If labor markets are heterogeneous, across-the-board wage moderation can, in effect, establish equilibrium in only one market. In the long run, therefore, the across-the-board aspect of wage moderation will become problematic as labor shortages at the high end of the market will put upward pressure on wages. Ideally, the skill premium should be allowed to change with relative changes in the demand for different types of labor. In the Netherlands, the problem of little differentiation in gross wages has in the past been alleviated by a differentiation in overall labor costs (through employer subsidies), while a somewhat smaller bias away from unskilled labor, a growing workforce, and a changing skill-composition of this workforce may have rendered wage differentiation less urgent than in Germany.
APPENDIX A Formal Model of a Labor Market with Homogeneous Labor
46. This Appendix summarizes the formal model underlying the homogeneous labor market framework of Section B and the method used for deriving the labor market shocks in Germany and the Netherlands that are analyzed in Section C.
A simple model of the labor market
The model comprises the following equations:
47. The wage-setting equation relates the effective or real product wage (w/a) to the unemployment rate and a shift-parameter, z, that captures other relevant labor market conditions:
Ceteris paribus, real wages will grow at the rate of a (the level of labor-augmenting technology), so that w/a remains constant.
48. The short-run labor demand relation is determined by the first derivative of the production function with respect to labor. Using a CES production function,
—in which A is a multiplicative constant, K is capital, aL is labor in effective terms, b is the share parameter on labor, and ρ is a function of the elasticity of substitution between capital and labor (ρ=(σ-1)/σ)—the first-order condition (including the mark-up, μ, over the wage) is:
49. The long-run labor demand relation is determined by the zero-net-profit condition that a firm’s revenues from output just cover the cost of production:
with the real interest rate (r) and the rate of depreciation (δ) reflecting the user cost of capital. After substituting (2) and (3) into (4), the condition can be rearranged to yield:
For a given user cost of capital, this condition pins down the capital-labor ratio and, by implication, the real product wage.
50. The labor share (SL) that is compatible with this framework is defined as:
In equilibrium, labor is paid its marginal product and employment equals equilibrium employment, so that the labor share becomes:
which, after substituting (2) for
51. The equilibrium labor share is a positive function of the share parameter, b; a negative function of the mark-up, μ; and a positive or negative function of the capital-labor ratio, depending on the sign of ρ. In the short run (when the elasticity of substitution is probably low) ρ will be negative. In the medium and long run, ρ is probably positive.
52. In the steady state, the economy satisfies the labor supply relation, as well as both labor demand relations. Capital, labor, and output grow at the same pace, as do real wages and technology (thus maintaining a constant real product wage). The labor share that is compatible with the steady state is one in which labor earns its marginal product (adjusted for the mark-up) and employment equals equilibrium employment. Given that in the steady state the ratio of capital to effective labor is constant, the labor share will be constant as well.
53. Out of steady state, the labor share can change:
The immediate effect of an adverse labor supply shock will be an increase in the labor share but, as quantities adjust, the labor share will move back to its original level (Figure A-1, panels 1 and 2). Suppose that the wage-setting curve shifts inward because wages fail to adjust to a slowdown in technological progress (a).27 The adverse shock will initially raise both the real product wage and the labor share, but in response to high wages and low profits, firms will scale down labor, capital, and economic activity. This causes a leftward shift of the short-run labor demand curve and a gradual reversal of the initial increase in the wage and the labor share. The new long-run equilibrium will again be on the long-run labor demand curve and, as the parameters of this curve were not affected, both the capital-labor ratio and the labor share are unchanged from their original values.28 In the long run, an unfavorable shift to the wage-setting curve thus comes entirely at the cost of higher unemployment.
The immediate effect of an adverse shock to short-run labor demand will be a decline in the labor share that persists even after the economy has reached a new steady state (Figure A-1, panels 3 and 4). Suppose that the short-run labor demand curve shifts because of an increase in the mark-up.29 The shock will initially reduce both the real product wage and employment. Over time, firms respond to lower wages and higher profits by increasing capital and rehiring workers. Eventually, the economy will return to the employment level and the real product wage that prevailed before the shock but with more capital and a higher level of economic activity. Although the equilibrium is on the same long-run labor demand curve as before, the fact that activity is higher implies a lower labor share.
Shocks to Labor Supply and Labor Demand
Citation: IMF Staff Country Reports 2000, 142; 10.5089/9781451810424.002.A002
Shocks to Labor Supply and Labor Demand
Citation: IMF Staff Country Reports 2000, 142; 10.5089/9781451810424.002.A002
Shocks to Labor Supply and Labor Demand
Citation: IMF Staff Country Reports 2000, 142; 10.5089/9781451810424.002.A002
54. The effect of an adverse shock to the long-run labor demand condition will be a permanent decline in the real product wage and the level of employment, while the labor share can rise or fall depending on the elasticity of substitution between labor and capital (Figure A-1, panels 5 and 6). An increase in the user cost of capital will lower the long-run labor demand curve and trigger substitution toward a smaller capital-labor ratio and a lower real product wage. Despite the labor-for-capital substitution, it is the reduction in capital that dominates the impact on labor demand and reduces employment. If the elasticity of substitution between capital and labor is less than unity, the declining capital-labor ratio will compress the labor share, while with an elasticity of substitution greater than unity, the labor share will rise.
Calculating Labor Market Shocks30
55. Labor supply is determined by the wage-setting relation, which can be estimated using
log (w/a)=-βu+z,
where β measures the sensitivity of the real product wage (w/a) to changes in the unemployment rate (u). Changes in z reflect labor supply shocks, so that the evolution of the wage-setting curve can be represented by the evolution of z (given by log (w/a) + βu). The technology parameter, a, is estimated by the Solow residual scaled by the labor share, while the real wage is the nominal wage of the business sector deflated by the GDP-deflator of the business sector. The sensitivity parameter β is taken to be 1 for both countries (a value consistent with findings by Blanchflower and Oswald, 1995). At any point in time, figure II-2 (panel 1) shows the difference between the actual real product wage and the wage that would allow the economy to return to its 1970 unemployment rate without any inflationary pressure. The size of the gap increases with β because, if wages are more sensitive to changes in the unemployment rate, a reduction in this rate will put more upward pressure on wages.
56. Labor demand is determined by the first-order condition that labor is paid its marginal product, which can be rearranged and estimated as:
log (w/a)=log (b)-log (1+μ)-(1-ρ) log (aL/Y).
Shifts in labor demand come from changes in the share parameter or the mark-up, so that the evolution of labor demand can be represented by the evolution of b or μ (changes in b or μ are observationally equivalent and given by log (w/a) + (l−ρ) log (aL/Y)). The elasticity of substitution is taken to be 1, but other values generate similar results. Incidentally, with an elasticity of 1 (ρ=0), labor demand shocks mirror changes in the labor share because b and μ are the only source of change in the labor share in this case.
References
Blanchardz, O., 1997, “The Medium Run,” Brookings Papers on Economic Activity (2), pp. 89–158.
Blanchardz, O., 1998, “Revisiting European Unemployment: Unemployment, Capital Accumulation, and Factor Prices,” written for a conference in honor of Michael Bruno, Jerusalem.
Blanchardz, O., and J. Wolfers, 1999, “The Role of Shocks and Institutions in the Rise of European Unemployment: the Aggregate Evidence,” NBER Working Paper No. 7282.
Blanchflower, D., and A. Oswald, 1995, “International Wage Curves,” in Freeman and Katz (eds.), Differences and Changes in Wage Structures, University of Chicago Press.
Caballero and Hammour, 1997, “Jobless Growth: Appropriability, Factor Substitution, and Unemployment,” NBER Working Paper, 6221.
Central Planning Bureau of the Netherlands, CPB Reports, several issues, The Hague (Netherlands).
Dutch Economic Institute, Economisch Statistische Berichten, several issues, Amsterdam (Netherlands).
Krusell, P., L. E. Ohanian, J-V. Rios-Rull and G. L. Violante, 1997, “Capital-Skill Complementarity: A Macroeconomic Analysis,” Federal Reserve Bank of Minneapolis, Research Department, Staff Report 239.
Ministry of Social Affairs in the Netherlands, 1999, Sociale Nota. Sdu Publishers, The Hague (Netherlands).
Nickell, S., 1997, “Unemployment and Labor Market Rigidities: Europe versus North America,” Journal of Economic Perspectives 11 (3), pp. 55–74.
Nickell, S., and B. Bell, 1997, “The Collapse in Demand for the Unskilled and Unemployment across the OECD,” Oxford Review of Economic Policy 11 (1), pp. 40–62.
OECD, Employment Outlook (various issues).
OECD, Economic Surveys: the Netherlands, (various issues).
OECD, Economic Surveys: Germany (various issues).
Van der Willigen, Tessa, 1995, “Unemployment, Wages, and the Wage Structure,” in: United Germany: The First Five Years, Eds.: Corker, Robert J., et al., IMF Occasional Paper No. 125 (Washington: International Monetary Fund), pp. 21–50.
Watson, Maxwell C., Bas B. Bakker, Jan Kees Martijn, and Ionannis Halikias, 1999, The Netherlands: Transforming a Market Economy, IMF Occasional Paper No. 181 (Washington: International Monetary Fund).
Prepared by Caroline Kollau.
The models are based on Blanchard (1997, 1998) and are set out in detail in the Appendix.
In the Netherlands, the positive impact of the rise in energy prices on the revenues from natural gas accrued to the government. Similar to other oil-importing countries, the private sector suffered a severe decline in profits.
See Watson et al. (1999) for details on the Dutch crisis of 1981–82. From the mid-1970s, per capita GDP growth in the Netherlands stayed well below that of its neighboring countries and underperformance was particularly pronounced from 1979 onward.
The unemployment rate in the German business sector doubled from below 4 percent in the 1970s to 8 percent in 1983. The rate in the Netherlands rose from 4 percent to 11 percent.
This would also arise if the substitution of skilled for unskilled labor were limited by the amount of skilled labor, or if the share of unskilled labor were larger than the share of skilled labor. Apart from technological change, the bias against unskilled labor could also come from increased international trade and specialization toward non-labor-intensive products.
More recently, however, the government became more involved with the initiation (in 1998) of the Alliance for Jobs. The Alliance is a social partner forum (of employers, trade unions and government) established to coordinate a comprehensive approach to improving labor market conditions.
Relevant for the impact on wages is the decline in the tax burden rather than the fact that income taxes were initially higher than in Germany.
In Germany, the service sector also expanded. But, while in the Netherlands this development led to an increase in total employment, the German service sector had to absorb excess employment from the manufacturing sector—a sector that had proven extremely sensitive to the shock of unification.
Although the average level of schooling in eastern Germany was at least as high as in western Germany, most east German workers had to be retrained after unification as they were ill-adapted to a western-style economy.
To the extent that income tax cuts gravitated toward the skilled (e.g., because of cuts in the top rates), this relaxed the gross wage demands from the skilled, which—with fixed differentials—could trigger lower wage demands from the unskilled.
See Watson at al. (1999) and CPB report 98/2 for further details on the cuts in employers’ social contributions.
Similar shifts would result from an increase in the level or duration of unemployment benefits; an increase in the tax wedge between gross and net pay (assuming that some of the tax burden is shifted to the employer); a higher incidence of long-term unemployment; or a failure of wages to adjust to an increase in labor supply.
Adjustment to the new equilibrium may take years and is affected by the elasticity of substitution between labor and capital—a low elasticity will slow the adjustment process.
Similar shifts would come from a decline in the share parameter on labor (i.e., a change in production technology biased against labor).
