This Selected Issues paper analyzes the challenge of population aging for Belgium. It argues that the aging strategy should be broadened to include more explicitly the objective of raising employment rates to foster potential growth. The paper discusses assumptions underlying the official aging projections, and presents an alternative baseline scenario on the basis of unchanged policies. It discusses the feasibility of strategies that rely exclusively on either fiscal or labor market adjustment, and illustrates the benefits of a two-pronged strategy. The paper also examines employment effects of reductions in labor taxes in a wage-bargaining model.

Abstract

This Selected Issues paper analyzes the challenge of population aging for Belgium. It argues that the aging strategy should be broadened to include more explicitly the objective of raising employment rates to foster potential growth. The paper discusses assumptions underlying the official aging projections, and presents an alternative baseline scenario on the basis of unchanged policies. It discusses the feasibility of strategies that rely exclusively on either fiscal or labor market adjustment, and illustrates the benefits of a two-pronged strategy. The paper also examines employment effects of reductions in labor taxes in a wage-bargaining model.

II. Employment Effects of Reductions in Labor Taxes in a Wage-Bargaining Model10

A. Introduction

26. Employment rates in Belgium are among the lowest in the euro area, as are average hours worked in the market sector (Figure 1). The causes of the lower labor utilization are well understood. Low employment rates appear largely to be the result of disincentives to work and high labor costs. Indeed, Belgium has an unusually wide safety net, with the number of beneficiaries exceeding the number of employed. Generous early retirement schemes discourage labor force participation of the old, while high replacement rates for the low paid and generous unemployment benefits keep the low-skilled and young on benefits. High labor costs also hurt labor demand. Moreover, limited geographical labor mobility has helped generate regional employment disparity, as Wallonia has exceptionally low employment rates.

Figure 1.
Figure 1.

Belgium: Labor Market Indicators, 2003

Citation: IMF Staff Country Reports 2005, 076; 10.5089/9781451803211.002.A002

Sources: OECD; and Fund staff estimates.1/ In percent of population aged 15-64 years.2/ In percent of population aged 55-64 years.3/ Income tax plus employee and employer contributions as percent of labor costs (single persons without children).4/ In dollars, PPP adjusted.5/ Percent of GDP (1999-2002 average).

27. The tax wedge between labor costs and workers’ take-home pay is the widest in the euro area, with likely adverse effects on employment. It rose from 38 percent to 51 percent during 1960-2000 ( Nickell, 2003). Based on the coefficients estimated by various studies, the 13 percentage point increase in the tax wedge could have raised real labor costs by about 20 percent and reduced the employment rate by as much as 4 percentage points. Belgium’s average effective income tax rate at the average production worker (APW) salary is the highest in the OECD countries. The marginal effective tax rate on labor income is substantially higher than in its three neighboring countries (France, Germany, and the Netherlands) across all income ranges (Table 1).

Table 1.

Belgium: Tax Wedge on Labor Income in 20031/

(In percent of labor costs)

article image
Source: OECD.

Total tax wedge is defined as the combined central and sub-central government income tax plus employee and and employer social security contribution tax, as a percentage of labor costs defined as gross wage earnings plus employer social security contributions. The tax wedge includes cash transfers.

28. Raising employment has been a priority of the government, and policy efforts have largely focused on labor market programs and tax cuts to promote employment (Box 1). Including training programs, subsidized private employment, unemployment compensation and early retirement (which is funded through the unemployment system), public spending on labor market programs in Belgium is one of the highest within the euro area (Figure 1). Reducing direct taxes on labor has since the early 1980s been a centerpiece of the government strategy, which aims to lower replacement rates for social benefits over time to encourage labor supply and decrease labor costs to stimulate labor demand (Carey, 2003).11 Changes in employers’ social security contributions (SSC) affect the wedge between labor costs and labor income, while changes in employees’ SSC and personal income tax modify the wedge between gross and net income. These tax cuts have often been targeted at the low-income earners,12 as these types of measures are regarded as very effective in raising employment, because the tax wedge tends to have an especially negative effect on employment at the lower end of the labor market.

29. These efforts have yet to raise labor utilization significantly. The employment-population ratio did rise appreciably during 1988-2000. This increase largely reflected an increase in female participation in the labor force, whereas the employment-population ratio for men (aged 15-64) remained virtually unchanged (Figure 2). However, based on OECD data, employment measured by the average annual hours worked per person aged 15-64 in the market sector declined by 10 percent during 1983-2003 and by 4 percent during 1993-2003, despite a brief increase in 1995 and 1997.13 Admittedly, this declining trend in working hours since the early 1980s is not a uniquely Belgian phenomenon, as it is also observed in many other EU countries, in particular, France, Germany, and the Netherlands. It is due to, among others, increasing female participation and the associated rise in part-time employment.

Figure 2.
Figure 2.

Belgium: Labor Supply, 1983-2003

(In percent of population; unless otherwise indicated)

Citation: IMF Staff Country Reports 2005, 076; 10.5089/9781451803211.002.A002

OECD and the Belgian Central Economic Council (CCE).1/ Working hours per worker per year. Data for 1983-98 are based on OECD, while data for 1999-2003 are based on the CCE.

Belgium: Reforms of Taxes on Labor Income Since 1983

1983-84: The “Planplus un” and the “Plan d’insertionprofessionnelle des jeunes” reduced employers’ SSC for the hiring of the unemployed and the young workers.

1985: Personal income tax reform (the “Grootjans” law) led to a 0.6 percentage point reduction in the marginal labor income tax rate. Reforms included (1) indexation of tax brackets and taxable threshold; (2) a slight decrease in tax rates; and (3) an increase in the taxable threshold.

1988: Personal income tax reform (the ”Maystadt” law) led to a 2.2 percentage point reduction in the marginal labor income tax rate. Reforms included (1) indexation of tax brackets and taxable threshold; (2) reshuffling of the tax schedule; (3) separate taxation of labor income for married couples with two earners; (4) income-splitting for married couples with one earner; (5) tax rebates for replacement income; and (6) a new system of child burden rebates.

1993-2000: Employers’ SSC for targeted groups were lowered several times. Since 2000, employees’ SSC rates have been reduced slightly.

2001-05: Income tax reform introduced in 2001 and to be implemented fully by 2004 is expected to lower the marginal labor income rate gradually by 3.2 percentage points between 2000 and 2005. In the context of an “employment conference” in 2003, new initiatives were developed, including reducing the tax wedge for high-skilled workers.

The 2001 tax reform consists of four pillars:

  • first pillar—easing the tax burden on labor by (1) introducing an income tax credit (€500 per year) for low-income earners; (2) raising the flat-rate deduction for working expenses; (3) shifting tax brackets upward; and (4) abolishing the top two marginal tax rates;

  • second pillar—introducing neutrality between married and unmarried couples, particularly with regard to (1) the income floor for tax exemption; (2) the tax rebate for replacement incomes; and (3) separate taxation on nonlabor income;

  • third pillar—improving tax treatment for those with dependent children by introducing (1) a refund for their income tax reductions; and (2) a supplementary tax reduction and a higher income ceiling for single parents, and by excluding part of the child support from the other parent for tax-related income assessment; and

  • fourth pillar—introducing more environmentally sound taxation.

30. Empirical evidence suggests that there is an overall adverse tax wedge effect on employment, hence reducing the tax wedge can be expected to raise employment. There is an extensive research based on panel data for OECD countries, with many studies finding significant tax wedge effects on labor costs and employment. For instance, Prescott (2004) finds that virtually all the large differences between the U.S. labor supply and those of France and Germany are due to the differences in tax rates. Similarly, Daveri and Tabellini (2000) conclude that the increase in the tax wedge fully explains the increase in unemployment in Europe in recent decades and that a 10 percentage point increase in the tax wedge raises real labor costs by 5 percent in the long run for a group of countries including Belgium. Other studies find a somewhat smaller negative tax wedge effect on employment ( Nickell and Layard, 1999; Elmeskov and others, 1998). Nickell (2003) concludes that a 10 percentage point increase in the tax wedge reduces employment by between 1 and 3 percent of the working-age population. For Belgium, Elmeskov and others (1998) finds that of the 1.7 percentage point increase in structural unemployment during 1990-95, about a third was due to the widening of the tax wedge during that period.

31. The employment effect of tax cuts with different wage formation mechanisms has been examined in two recent studies on Belgium, both using general equilibrium macroeconomic models. Burggraeve and Du Caju (2003) investigate the labor market and fiscal impact of a budget-neutral reduction in employers’ social security contributions, taking into account the wage formation process in Belgium. Using the National Bank of Belgium’s (NBB) quarterly macroeconomic model, they conclude that the impact of the tax reductions on employment depends on the outcome of the wage formation process. More recently, Stockman (2004) assesses the impact of the 2001 personal income tax reform on wages and employment, using the Federal Planning Bureau’s LABMOD model. When fully implemented, the 2001 reform is expected to lower the personal income tax rate (as a percent of gross wages) by about 3 percentage points in 2005. Stockman finds that the reform could raise employment by 0.6 percent (24,000 workers) in 2004 and by 59,000 workers in the long run under the right-to-manage wage bargaining model; however, the impact would be larger under the search model, with employment increased by about 86,000 workers in the long run.

32. This paper analyzes the effect of tax changes on employment under different wage formation mechanisms. Unlike the previous studies on Belgium, which focus on the changes in the average effective tax rates, this paper investigates the impact of changes in the marginal effective income tax rate on the Belgian employment. The theoretical part of the study derives the equilibrium employment under two wage-setting mechanisms: a competitive market model based on Prescott (2004) and a wage-bargaining model based on Nickell (2003). The effects of the tax wedge on employment under these models are assessed. In the empirical part of this chapter, both models are calibrated to simulate the effects of fiscal reforms during 1980-2001 on employment under different wage formation mechanisms.

33. The rest of the paper is organized as follows: Section B reviews the current wage formation process in Belgium. Section C outlines the theoretical models used for analysis. Model simulations and their implications are discussed in Section D. Section E concludes.

B. Wage Formation in Belgium

34. The wage-bargaining system in Belgium can be characterized as one with intermediate wage coordination, as opposed to centralized and decentralized wage formation systems. Although firm-level agreements have gained considerable importance over the last few years, wages are predominantly determined at the sector level. The wage negotiation process has three key components: the Interprofessional Agreement (IPA) followed by sectoral negotiations; administrative extension of labor contracts; and indexation of wages to prices during the contract. The IPA is a product of negotiations within the joint committees (comités paritaires), which have equal employer and employee representation and are organized by sectors of economic activities.14 The IPA determines a national wage norm and a national minimum wage, which serve as the upper and lower limits for the sectoral- and enterprise-level wage negotiations.

35. The wage norm sets a ceiling for the growth of nominal hourly labor costs for Belgian enterprises (the Law of July 1996), with a view to preserving external competitiveness vis-à-vis France, Germany, and the Netherlands. It is set every two years (usually in October) in the IPA for a two-year period by the social partners, based on estimates provided by the Central Economic Council (Conseil central de I’economie, or CCE). In particular, the CCE estimates the nominal wage norm as the weighted average of the expected increase in nominal labor costs in Germany, France, and the Netherlands, according to projections published by the OECD’s Economic Outlook and corrected for average working hours. In theory, the estimated margin can be lowered to reflect any excessive increases during the previous two-year period in Belgium vis-à-vis its three neighbors. The nominal wage norm is adjusted by projected inflation to determine the room for real wage increases.

36. Legally-guaranteed minimum wages set the floor for wages. Since the mid-1980s, national minimum wages in Belgium have declined steadily in relation to APW wages, broadly in line with rates in the Netherlands but lower than the rates in France (Figure 3). Nonetheless, the overall level of national (and sectoral) minimum wages still appears high, pricing the young out of jobs. OECD estimated that about 12 percent of full-time employees earn up to 115 percent of the statutory minimum wage (€1,131 per month). Moreover, the age-related sectoral and regional hourly minimum wages were well in excess of the national minimum wage, thus diminishing the effectiveness of fiscal and other measures to address inactivity traps.

Figure 3.
Figure 3.

Minimum Wages as Percent of APW, 1975-2000

Citation: IMF Staff Country Reports 2005, 076; 10.5089/9781451803211.002.A002

Source: OECD.

37. The wage indexation rule implies that nominal gross wages are automatically adjusted with the health index of consumer prices (HICP).15 The implementation of the indexation, however, is subject to sectoral negotiations. In some cases, wages are adjusted at certain intervals (e.g., every two, three, or four months); while in other cases, they are adjusted when the health index of consumer prices exceeds a certain threshold.

38. In theory, intermediate wage formation tends to generate higher wage increases than centralized wage formation or decentralized wage bargaining (Nickell and Layard, 1999; and Elmeskov and others, 1998). It tends to limit sectoral wage differentiation and result in insufficient centralization to account fully for macroeconomic spillovers of wage settlements, but enough to generate market power.16 Both labor unions and business representatives have been dissatisfied with the process, including the estimate of the norm by the CCE: the labor representatives view indexation as nonnegotiable, and business representatives prefer a centralized process to resist wage demands. Moreover, if the wage norm is binding, tax cuts might not have the desired effect on gross wages. Hence, tax cuts affect employment only through higher demand due to higher disposable income. Similarly, if unions have strong bargaining power, they may more easily resist employers’ attempts to reflect the higher payroll taxes in lower wages. Therefore, the effect of reducing the tax wedge may not be as significant as expected.

39. In practice, the wage norm has not always been adhered to (Table 2). Two factors seem to explain this outcome: first, the wage norm appears not to be binding in sectors with labor shortages; second, the norm may have served as a floor rather than a ceiling, thus pushing up the general wage level. Overruns of the wage norm, whether due to higher-than-anticipated inflation or due to wage drift, were not corrected in subsequent agreements despite a legal provision permitting such action. In addition, cuts in taxes and employer and employee social security contributions appear to have translated into higher take-home pay rather than lower labor costs.

Table 2.

Belgium: Developments Under Interprofessional Wage Agreements, 1999-2004

(Hourly labor costs, percent change)

article image
Sources: Belgian authorities; and Fund staff estimates.

France, Germany, and the Netherlands.

C. Theoretical Framework

40. The basic competitive market model is taken from the standard theory used in Nickell (2003) and Prescott (2004). Using a representative agent model, the household faces a labor-leisure decision and a consumption-savings decision. The household’s preferences are given by

maxE{Σt=0ρt[1nCt+α1n(1Nt)]}

s.tCt+It=wt(1τt)Nt+(rtδ)Kt+Tt

where C is consumption, I investment, w real wage, r the real return on capital, and τt is the tax wedge between the real labor cost per employee facing the representative firm and the real after-tax net wage.17 The preference parameter a measures the value of leisure relative to consumption. With the population of working age normalized to one, 1-N is leisure. Even though leisure includes time allocated to working in the nonmarket and underground market sectors, the important point is that these sectors are not taxed. All taxes are used for government consumption G and transfer payments T to the household.

41. Suppose W is the nominal labor cost per worker and P is the price of the firm’s product. Assuming τ123, and τ4 are employers’ SSC, employees’ SSC, income tax, and the consumption tax, respectively, the real after-tax consumption wage is given by

waftertax=wt(1τt)=Wt(1τt1)(1τt2)(1τt3)P(1+τt4)

and

τt=(1τt1)(1τt2)(1τt3)(1+τt4).

42. The representative firm determines labor demand by maximizing its profits for given factor prices:

maxΠ=Σt=0{ρt[PtYtWtNt]}

s.tYt=Ct+It+Gt=λtKtβNt(1β)

Assuming a depreciation rate of δ, the capital stock is

Kt+1=(1δ)Kt+It.

43. In equilibrium, the marginal rate of substitution between leisure and consumption is equal to their price ratio, and the marginal product of labor is equal to the real cost of labor

α/(1N)1/C=w(1τ)

(1β)YN=w.

These equations solve for the labor supply under a competitive market:

Nt=(1β)(1β)+CtYtα1τ(1)
dNt/dτ=α(1β)[(1β)+CtYtα1τ]2(1τ)2<0.

In this model, the intratemporal factor affecting labor supply is captured by 1-τ, which distorts the relative prices of consumption and leisure at a point in time. The size of its impact on labor supply depends crucially on the preference parameter a. The term C/Y captures intertemporal factors. Equilibrium C/Y is a function of future tax rates, productivity, and the current capital stock.

44. Now we consider a situation where wages are determined in a wage-bargaining model, which seems more suitable for Belgium (Estevão, 2001). In this model, unions and firms bargain over the wage level taking into account the labor demand curve. In particular, for each firm i, wages are determined by a Nash bargain that maximizes

[Ni(wi)2(wi(1τ)+YnA)]θΠi

subject to the labor demand derived from the firm’s profit maximization:

(1β)YiNi(wi)=wi,(2)

where Yn is real, after-tax, per capita nonlabor income, A is the expected alternative income if not employed in firm i, and Π is the firm’s profit. The parameter y measures the extent to which the worker takes into account the employment effects of the wage bargain. Collective bargaining is therefore associated with a higher y. The parameter 6 captures the relative strength of the worker in the bargain.

45. Expected alternative income A for a representative worker consists of income generated by employment in another firm (with probability N) and nonemployment benefits (with probability 1 − N). Specifically, A is given by

A=N(1τ)wi+(1N)b(1τb),(3)

where b represents the after-tax nonemployment benefits, which are assumed to subject to a much lower income tax rate τb (in our example, we assume τb = 0 for simplicity). Hence, b/(l − τ) w is the net replacement rate.

46. The first-order condition implies that real after-tax wages in the bargaining model are determined as a markup over the worker’s alternative income. Specifically,

w(1τ)=(1+1B1)A,whereB=θγ+βθ(1β).(4)

This markup is larger when the bargaining power of the worker is stronger (higher θ) or when the bargaining process is more decentralized (lower γ).

47. Assuming identical firms with wi = wb, equations (2) and (3) solve for employment and the real wage under the equilibrium (N(wb), wb). It can be shown that ∂N/∂τ<0 so long as the after-tax benefits are less than the after-tax wage (otherwise no one would work). Moreover, the size of ∂N/∂τ depends on the features of labor market institutions, specifically the parameters γ and θ.

D. Simulations

48. We first simulate the impact of changes in the tax wedge on employment under the competitive market model, using equation (1). Specifically, the capital share parameter β is set to 0.3224 and the preference parameter α set to 1.54, based on Prescott (2004). This is consistent with the conjecture that across countries, idiosyncratic preference differences average out and households tend to have almost identical preferences. Employment is measured by hours worked per year per person aged 15-64.

49. To run the simulation, the marginal labor income tax rate is estimated based on the following equation:

τmarginal=1.68τincome+1.13τemployerSSC+τemployeeSSC,

where factors 1.68 and 1.13 are selected based on the average and marginal rates for four different income groups published by the OECD (Table 1) and reflect the fact that marginal income tax rates are higher than the average tax rates.

50. The results indicate that the impact of changes in the tax wedge on employment is much larger under a competitive market than within the current institutional setting (Table 2). We look at the following two periods when fiscal reforms led to substantial reductions in tax wedge:

  • 1985-91—the Grootjans law in 1985 and the Maystadt law in 1988, which reduced the marginal tax income rate by 3.6 percentage points; and

  • 2000-03—the fiscal reform of 2001, which has lowered the marginal rate by 1.4 percentage points.

51. Simulation results indicate that in a competitive market, the income tax reforms of 1985 and 1988 would have increased employment by more than 12 percent during the period 1985-91 (Table 3). This seems to be consistent with Estevao’s (2001) finding that policy changes since the early 1980s have apparently not boosted employment. This result also holds when using the OECD data on tax wedges for simulation. Nonetheless, these estimates should be interpreted with caution. For example, the lack of an impact on employment during 2000-03 may simply reflect the slowdown in economic activity during this period.

Table 3.

Belgium: Impact of Changes in Tax Wedge on Employment, 1985-2003

article image
Source: Fund staff estimates.

In percentage points.

Percent change during the period.

52. The difference between Belgian labor supply and that of France and Germany appears due to differences in their tax wedges. For example, for the period 1993-96, with the same marginal tax wedge (about 60 percent), our predicted labor supply for Belgium is very similar to that in Germany and France as estimated by Prescott (2004).

53. Simulated effects of changes in the tax wedge on employment under a wage-bargaining model are based on a reduced-form employment equation derived from the first order conditions captured by equations (2), (3), and (4). This equation is then calibrated to reproduce the employment effect of tax reforms during the period 1985-91. Specifically, for the parameter β, a value of 0.5 is selected, and a value of 0.007 is chosen for the parameter y, based on existing macroeconomic models for Belgium ( Jeanfils, 2000).18 Once the model has been calibrated, it is used to estimate the impact of the 2001 tax reform.

54. Under the wage bargaining model, the 2001 tax reform would have a positive effect on employment; moreover, this effect is larger when the wage bargaining power is weaker (Table 4). The 2001 tax reform is expected to lower the average and the marginal effective tax rates on labor by 2.0 and 3.2 percentage points, respectively, during the period 2000-05.19 This would raise employment by 1.6 percent during this period and by 0.5 percent in 2003 alone—the latter is very close to the 0.6 percent estimated by Stockman (2004). The estimated effect under the wage bargaining model is much smaller than the effect estimated under the competitive market model discussed above (4.3 percent).

Table 4.

Belgium: Employment Effects of Labor Tax Cuts in a Wage-Bargaining Model

article image
Source: Fund staff estimates.

In percentage points.

Percent change during the period.

E. Concluding Remarks

55. Since the early 1980s, reducing the tax wedge on labor has been an important component of the authorities’ strategy to reverse the decline in employment rates. Social security contributions were lowered several times, as were income taxes, most recently in the 2001 tax reform (the full effect of which will be felt in 2007). Budget constraints limited progress, however, even inducing reversals such as in 1993 when taxes were raised. Tax reductions were often targeted at low-income earners. These cuts, to the extent that they reduce labor costs, stimulate labor demand. They promote labor supply directly and probably also indirectly as net take-home pay rises compared to social benefits. Given empirical evidence that the rise in the labor tax wedge from 38 percent in 1960 to 51 percent in 2000 may have lowered the employment rate by up to 4 percentage points, such a strategy makes sense.

56. Current labor market institutions impede the effectiveness of the tax cut strategy to raise employment. The strong bargaining position of labor unions implies that reductions in social security contributions and income taxes end up mostly in take-home pay. Our estimates suggest that the 2001 tax reform (at a budgetary cost of 2 percentage points of GDP) will raise the employment rate by 1.6 percentage points by 2005. By comparison, if union bargaining power were reduced by half, the employment effect would be 2.2 percentage points, while a fully competitive market would deliver 4.3 percentage points.

References

  • Burggraeve K., and P. Du Caju, 2003, “The Labour Market and Fiscal Impact of Labour Tax Reductions,National Bank of Belgium, Working Paper No. 36, March (Brussels: National Bank of Belgium).

    • Search Google Scholar
    • Export Citation
  • Carey, D., 2003, “Tax Reform in Belgium,OECD Economics Department, Working Paper No. 354 (Paris: OECD).

  • Daveri, F. and G. Tabellini, 2000, “Unemployment and Taxes: Do Taxes Affect the Rate of Unemployment?Economic Policy, April (Great Britain).

    • Search Google Scholar
    • Export Citation
  • Elmeskov, J. J. P. Martin, and S. Scarpetta, 1998, “Key Lessons for Labour Market reforms: Evidence from OECD Countries’ Experiences,Swedish Economic Policy Review, Vol. 5, pp. 205– 52

    • Search Google Scholar
    • Export Citation
  • Estevão, M., 2001, “The Effects of Tax Changes on Belgian Employment,in Belgium: Selected Issues. IMF Country Report No. 01/45 (Washington: International Monetary Fund).

    • Search Google Scholar
    • Export Citation
  • Hendrickx, K., and others 2003, “Un Nouveau Modele Macro-econometrique du Marche du Travail,Belgian Federal Planning Bureau, Working Paper No. 08-03.

    • Search Google Scholar
    • Export Citation
  • Jeanfils, P., 2000, “A Model with Explicit Expectations for Belgium,National Bank of Belgium, Working Paper No. 4, March (Brussels: National Bank of Belgium).

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Meyermans, E., 2004, “The Macroeconomic Effects of Labour Market Reforms in the European Union,Belgian Federal Planning Bureau, Working Paper No. 12-04.

    • Search Google Scholar
    • Export Citation
  • Nickell, S., and 2003, “Employment and Taxes,CESIFO Working Paper, No. 1109 (Munich: Center for Economic Studies and the IFO Institute).

    • Search Google Scholar
    • Export Citation
  • Nickell, S., and R. Layard, 1999, “Labour Market Institutions and Economic Performance,” In O. Ashenfelter and D. Card (eds.), Handbook of Labor Economics, Vol. 3, (Amsterdam).

    • Crossref
    • Search Google Scholar
    • Export Citation
  • OECD, 2003, OECD Economic Surveys: Belgium, Volume 2003/1 (Paris: OECD).

  • Prescott, E., 2004, “Why Do Americans Work So Much More Than Europeans?NBER, Working Paper No. 10316 Cambridge, Massachusetts: National Bureau of Economic Forecasts

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Stockman, P., 2004, “Personal Income Tax Reform in Belgium: The Short-, Medium- and Long-run Impact on Wages, Employment and Value Added Re-examined by LABMOD,Belgian Federal Planning Bureau, Working Paper No. 11-04.

    • Search Google Scholar
    • Export Citation
  • Thomas, A., 2002, “The Costs and Benefits of Various Wage Bargaining Structures: An Empirical Explanation,IMF Working Paper No. 02/71.

10

Prepared by Jianping Zhou (jzhou1@imf.org).

11

Policies to lower social security contributions were envisaged as early as in 1981. A detailed description is presented in Estevao (2001).

12

For example, the substantial reductions in employers’ social security contributions in 1999.

13

Paid vacations, sick leave, and holidays are not accounted as working hours; neither is the time of someone working in the underground economy or in the home sector. There is a structural break in the OECD’s annual working time series for Belgium, because in 1999, the NIS changed the ranking of the questions in the Labor Force Survey, which might explain the large drop in working time in 1999. The OECD has recently revised the data, which are yet to be published.

14

The number of joint committees now exceeds 100, because economic sectors are narrowly defined.

15

Defined as the national index of consumer prices, excluding the prices of alcoholic beverages, tobacco, and motor fuels.

16

On the drawbacks of such an intermediate position, see the OECD Employment Outlook, July 1997, and Thomas (2002).

17

For simplicity, we assume no investment or capital income tax. Although introducing these parameters would affect the price of investment relative to the consumption good, it would not alter the price of consumption relative to leisure, which is the focus of the analysis.

18

The existing estimates of the long-run elasticity of labor demand in Belgium vary from −0.2 to −2.0.

19

These are our estimates based on Stockman (2004) and OECD (2003).

Belgium: Selected Issues
Author: International Monetary Fund