Chapter 12. Youth Unemployment in Europe: Okun’s Law and Beyond

Petya Koeva Brooks, and Mahmood Pradhan
Published Date:
October 2015
  • ShareShare
Show Summary Details
Angana Banerji, Huidan Lin, Sergejs Saksonovs and Rodolphe Blavy 

Youth unemployment increased sharply in Europe in the aftermath of the global crisis in 2008–09 and remains at historic highs in the current weak recovery. A large proportion of these increases can be explained by output dynamics and the greater sensitivity of youth unemployment to economic activity compared with adult unemployment. Labor market institutions also play an important role, especially the tax wedge, minimum wages, and spending on active labor market policies (ALMPs). Youth unemployment has moved up the policy agenda in Europe, and policies to deal with this issue have been formulated at both the European Union (EU) (European Commission 2012, 2013b) and national levels.

This chapter assesses the youth unemployment problem in advanced European countries, especially the euro area. It documents the main trends in youth and adult unemployment before and after the crisis, identifies the cyclical and structural drivers of youth and adult unemployment, and outlines elements of a comprehensive strategy to address the problem.

Stylized Facts

Youth unemployment rates were at unprecedented levels in the euro area in 2014.1 The global crisis has reversed a decade-long trend of modest declines in youth unemployment; the youth unemployment rate in the euro area in mid-2014 (some 23 percent) was well above the precrisis rate (15 percent in 2007).

Adult unemployment also ticked up after the crisis, but less so than youth unemployment. Although unemployment rates typically tend to be higher for youth than for adults,2 these differences increased sharply after the global crisis (Figure 12.1).

Figure 12.1Euro Area Unemployment Rate


Source: Eurostat.

Youth unemployment rates vary widely across the euro area in magnitude as well as in trajectory. These cross-country differences were exacerbated during the crisis. The hardest hit euro area countries had unprecedented increases in youth unemployment rates, ranging from 25 percent in Ireland to 43 percent in Spain, on average, during 2007–13 (Figure 12.2). In countries that fared better, youth unemployment rates increased only marginally (Austria, the Netherlands) or even fell (Germany). Precrisis youth unemployment rates, however, have had little bearing on youth unemployment dynamics since the crisis.

Figure 12.2Changes in the Youth Unemployment Rate since 2007

Sources: Eurostat; and IMF staff estimates.

Large and persistent youth and adult unemployment rates lower potential output as the result of hysteresis effects (skill attrition and depreciated human capital) or the outward migration of skilled labor. Youth unemployment erodes social cohesion and institutions. For individuals it may lead to scarring—a lower probability of future employment and lower wages.


This chapter analyzes the relative significance of two main drivers of youth unemployment—business cycle fluctuations and the institutional setup and features of the labor market. It contrasts the impact of these factors on the youth versus the adult labor market.

The analysis covers 22 advanced European countries—18 in the euro area,3 plus Denmark, Sweden, Norway, and the United Kingdom. It is based on annual data from 1980 through 2012, although the actual size of the sample varies depending on data availability, especially for labor market characteristics.

Labor market features are grouped into several interrelated categories: the opportunity cost of working, hiring costs, the role of collective bargaining, measures of labor market duality, education and training, and spending on ALMPs. A number of different measures are used in each category. Labor market features vary widely across countries but change slowly over time. (See Annex 12.1 for definitions.)

Data on labor market characteristics are usually not available for the full sample period, and are especially limited for new entrants to the euro area. Data gaps make it infeasible to produce country-by-country estimations to determine the country-specific effects of institutional variables and the business cycle. As a workaround for the data limitations, a two-pronged approach was adopted. One prong was to assume that the impact of labor market features was the same across countries while allowing the impact of the business cycle to vary. This approach was implemented using one labor market factor at a time as well as a combination of factors. The second prong also assumed common effects of labor market factors across countries; however, the model allowed the impact of labor market institutions to vary across countries via the interaction of labor market variables with the business cycle. To ensure robustness, only one institutional variable was considered at a time. (See Annex 12.2 for additional details on this methodology.)

The Business Cycle

Okun’s Law, proposed by Arthur Okun in 1962, is the empirical regularity that changes in unemployment rates and output growth are negatively related.4 Many studies confirm this relationship for overall unemployment, but research on youth is less common. Some authors highlight the sensitivity of youth unemployment to the business cycle (OECD 2006; Scarpetta 1996; Scarpetta, Sonnet, and Manfredi 2010; European Commission 2013a).

Empirical analysis unambiguously confirms that Okun’s Law holds (Figure 12.3), regardless of how the business cycle is measured—by real GDP growth or the output gap (that is, the difference between actual output and its potential). Cyclical factors explain about 50 percent of the changes in youth unemployment rates (and 70 percent of the increase in unemployment rates in stressed euro area countries) and about 60 percent of changes in adult unemployment rates across all advanced European countries.

Figure 12.3Euro Area: Okun’s Coefficients

Sources: Eurostat; and IMF staff estimates.

Note: The gray shaded bars and dots indicate statistically insignificant results. Data labels in the figure use International Organization for Standardization (ISO) country codes.

The sensitivity of unemployment rates to the business cycle (Okun’s coefficient) varies across countries (Figure 12.3). Estimates range from not significantly different from zero (Austria5) to – 1.9 in Spain, that is, a 1 percent decline in growth increases the youth unemployment rate by almost 2 percentage points. Countries with the biggest increases in youth unemployment rates since the crisis tend to be those that are most affected by the business cycle (they have the largest Okun’s coefficients).

In every country the estimated Okun’s coefficient is higher for youth than for adults, almost three times as large on average. This may be due to both the nature of the youth labor force (described previously) and special features of youth employment, specifically concentration in cyclically sensitive industries and in small and medium-sized enterprises (SMEs).

Youth employment is concentrated in sectors that tend to be more sensitive to the business cycle: manufacturing, construction and real estate, wholesale and retail trade, and hotels and restaurants (Figure 12.4). Together these sectors comprise between 65 percent and 75 percent of youth employment in countries where youth unemployment increased the most after the global crisis.

Figure 12.4Youth Employment by Sector, 2000–07


Sources: Eurostat; and IMF staff estimates.

Note: Data labels in the figure use International Organization for Standardization (ISO) country codes.

Box 12.1.Small and Medium-Sized Enterprises (SMEs): The Role of Financing Constraints and Youth Unemployment

Methodology. The Okun’s Law framework is augmented with the interaction of GDP growth and the average share of SMEs (Figure 12.1.1), or with financial constraints (Figure 12.1.1). Financing constraints are measured as the percentage of firms in the industry and services sectors that, in surveys, report financial constraints as a factor limiting production. The surveys do not separate SMEs as a separate category, but it is reasonable to assume that they are more affected by financial constraints than larger firms.1

Figure 12.1.1Small and Medium-Sized Enterprises’ Share in Employment

(Percent; average 2008–13)

Source: European Commission.

Note: Data labels in the figure use International Organization for Standardization (ISO) country codes.


  • Greater financial constraints are associated with higher youth unemployment. Controlling for country-specific fixed effects and output gaps, an additional percentage point of firms reporting financial constraints raises youth unemployment rates by 0.3 percentage point (industry) to 0.4 percentage point (services). The effect on adult unemployment rates is smaller (0.2 percentage point).
  • A percentage point increase in the average employment share in SMEs (or the SME share of value added) lowers Okun’s coefficient by 0.01, making unemployment more procyclical.
1 There are data limitations. The average SME share is used because the data are available only from 2008 and stay broadly constant. The percentage of firms reporting financial constraints is small (often zero) and relatively unchanged; for example, before the financial crisis only 2 percent of industrial firms reported financial constraints, on average, compared with 2012-13, when the average was 9.3 percent.

SMEs6 employ the majority of the labor force, with the average employment share especially high in some southern European countries—more than 75 percent for Italy, Portugal, and Spain, for example. SMEs face unique financial constraints in the current environment of financial fragmentation and private sector deleveraging. These constraints appear to increase the extent to which youth unemployment rates respond to growth (Box 12.1).

Labor Market Features

Output changes, on average, explain much of the increase in youth unemployment, but not in every country in Europe (Figure 12.5). Excluding countries worst affected by the crisis—Greece, Ireland, Latvia, Portugal, and Spain—cyclical factors explain, on average, about 35 percent of the changes in youth unemployment rates across advanced European countries (for example, 33 percent and 27 percent, respectively, in France and Italy, both of which have high youth unemployment rates). What are the other explanatory factors? In particular, to what extent do labor market institutions and skills play a role in explaining labor market outcomes across advanced European economies as well as within individual countries?

Figure 12.5Proportion of Change in Youth Unemployment Explained by Growth

Sources: Eurostat; IMF staff estimates.

A large body of literature discusses the role of labor market characteristics in determining aggregate unemployment in Europe. Some of the labor market characteristics considered in the literature to have an impact on youth unemployment include unionization (Bertola, Blau, and Kahn 2007); hiring and firing regulations; minimum wages and hiring costs (Bernal-Verdugo, Furceri, and Guillaume 2012); and labor market flexibility (OECD 2006; Choudhry, Marelli, and Signorelli 2012).

Labor market features have significant effects on the levels of youth and adult employment and unemployment, but not on changes.7 Put simply, the rapid divergence of youth unemployment rates in the aftermath of the crisis has not been accompanied by dramatic changes in labor market features.

A number of labor market features have an impact on youth unemployment rates; in particular, lower hiring costs (tax wedge, minimum wages relative to the median wage) and higher spending on ALMPs per unemployed person, especially on training, reduce unemployment. The opportunity costs of working (for example, benefit replacement rates) and low skill levels tend to raise unemployment, whereas collective bargaining (for example, union density) and labor market duality (for example, protection of workers) have mixed effects. These effects remain significant after controlling for several labor market features at a time (Figure 12.6, panel 1).

Figure 12.6Effects of Labor Market Features on Youth Employment

(Percentage points)

Sources: Eurostat; Organisation for Economic Co-operation and Development (OECD); IMF World Economic Outlook; and IMF staff estimates.

Note: GRR = gross replacement rate; NRR = net replacement rate; ITRAP = inactivity trap; MIN2MED = ratio of minimum to median wage; TWEDGE = tax wedge; UDENS = union density; ABP = adjusted bargaining power; EPT = OECD temporary employment protection index; TSHARE = share of temporary employees for a given age group; TPROB = share of temporary employees on probation in total temporary employees for a given age group; ALMPTOT = total spending on ALMP (thousand euro per unemployed); ALMPTRAIN = spending on ALMP training policies. See Annex 12.2 for model details. Only significant results are shown for the univariate model (panel 2).

Allowing for interactions between labor market features and the business cycle reveals significant cross-country differences in the impact of labor market features on labor market outcomes (Figure 12.6, panel 2). For instance, a 1 percentage point increase in the tax wedge increases youth unemployment rates by between 0.6 and 1.4 percentage points in univariate models with the interaction term.8

Higher Hiring Costs, Higher Unemployment

Taxes on employers and employees, in combination with statutory minimum wage rates, affect both the supply of and demand for labor. High tax rates on labor income depress the supply of labor and drive a wedge between marginal productivity and the reward for work. If higher taxes translate into higher wages, then the increase in labor costs can reduce labor demand and increase unemployment. High payroll taxes and employers’ social security contributions are even more likely to raise labor costs in the presence of wage floors generated by statutory minimum wages. If employers succeed in shifting the tax burden to employees in the form of lower wages, the labor supply could shrink, especially for low-wage earners (who would conceivably include youth). OECD (2012) shows that since 2007 young people have, on average, been at a big disadvantage in countries in which the minimum wage is relatively high as a percentage of median pay.

Hiring costs remain above average for several countries. The tax wedge declined in the Scandinavian countries and Germany, while either remaining unchanged or even increasing in other European countries (in particular, Greece and Spain) (Figure 12.7). The ratio of the minimum wage to the median wage has remained unchanged for the vast majority of advanced European countries while marginally increasing in France, Greece, Portugal, and Spain (Figure 12.8).9

Figure 12.7Tax Wedge


Source: European Commission.

Note: Latest data for Cyprus are for 2007. Data labels in the figure use International Organization for Standardization (ISO) country codes.

Figure 12.8Ratio of Minimum to Median Wage

Source: European Commission.

Note: Data labels in the figure use International Organization for Standardization (ISO) country codes.

This chapter’s empirical analysis indicates that greater hiring costs—larger tax wedge, higher minimum wages relative to the median wage, or both—are associated with higher youth and adult unemployment rates and lower employment rates for both youth and adults. A 1 percentage point increase in the tax wedge raises youth unemployment rates 0.3–1.3 percentage points.

The effect on adult unemployment is smaller (about 0.4–0.5 percentage point). Higher minimum wages (relative to median wages) raise youth unemployment by 0.4–1.2 percentage points. Because many young people are hired at minimum wage jobs, they may be particularly vulnerable to increases in the cost of hiring.

Higher Opportunity Cost of Working, Higher Unemployment

High unemployment benefits raise unemployment by reducing the willingness to search intensively for jobs or to accept job offers, and by increasing the reservation wage, that is, the salary at which the unemployed would be willing to work instead of receiving unemployment benefits (both of which lower labor supply). Moreover, tax and benefit systems can interact to create an unemployment or inactivity trap that arises when individuals who qualify for social protection benefits have little financial incentive to work because the combined effects of increased tax payments and the withdrawal of income-tested benefits offset the potential gain in disposable income from increased earnings.

Overall, the picture remains largely unchanged during 2001–12, regardless of which indicators are used to capture the incentives of the unemployed to seek work (the net benefit replacement rate or the inactivity trap; Figures 12.9 and 12.10). Germany and some Scandinavian countries, which have reduced benefit replacement rates and indicators of the inactivity trap, provide the few exceptions. In contrast, these indicators have remained unchanged for the vast majority of advanced European countries while marginally increasing in the euro area countries under stress.

Figure 12.9Net Replacement Rate


Source: European Commission.

Note: Data labels in the figure use International Organization for Standardization (ISO) country codes.

Figure 12.10Inactivity Trap


Source: European Commission.

Note: Data labels in the figure use International Organization for Standardization (ISO) country codes.

Higher opportunity costs of working are associated with higher youth and adult unemployment rates. In univariate models with the interaction term, a 1 percentage point increase in net replacement rates raises youth unemployment rate 0.1–0.5 percentage point depending on the country, more than it does for adults (0.1–0.2 percentage point). Country-specific estimates find significant and positive effects of higher marginal tax rates on income (inactivity trap) on youth unemployment for most countries, with a stronger effect on youth unemployment than on adults. Youth may be more sensitive to net replacement rates because unemployment benefits allow them time to find a more desirable job. However, an aggregate indicator may mask country-specific differences in eligibility for unemployment benefits; for example, in some countries, people who have never had a job may not be eligible for unemployment benefits.

Dual Labor Markets, Shifting Composition of Unemployment

Dual labor markets feature a high share of temporary employment contracts with lower employment protection (Figure 12.11). Studies show that the impact of employment protection legislation (EPL)—legislation governing the hiring and firing of employees—on labor market outcomes is small and ambiguous.10 It can lower job separation rates by increasing the cost of firing, but also lower the job finding rate by increasing employers’ reluctance to hire workers in the first place. Labor market duality has been associated with a lower youth employment rate in a sample of 17 OECD countries during 1960–96 (Bertola, Blau, and Kahn 2007).

Figure 12.11Protection of Temporary Workers


Source: Organisation for Economic Co-operation and Development.

Note: Data labels in the figure use International Organization for Standardization (ISO) country codes.

Young workers tend to be employed on temporary contracts more so than adult workers. The disparity between adults and youth in this regard is particularly large in Italy, Portugal, and Spain, which have had some of the largest increases in youth unemployment. In Spain, labor market adjustments have focused on shedding workers on temporary contracts—mainly youth.

Higher protection for temporary contracts lowers unemployment rates for youth and adults, but the effects for youth are stronger because a higher share is employed on temporary contracts (Figure 12.12). A unit increase in the EPL rating11 lowers youth unemployment rates by 2.5–5.0 percentage points and adult unemployment by 1.5–2.0 percentage points. A higher share of youth on temporary contracts increases youth unemployment and lowers employment by 0.3–0.4 percentage point, but it has no significant effects on adult employment or unemployment rates. Hence stronger labor market duality can shift the composition of employment toward adults.

Figure 12.12Share of Temporary Workers

(Percent, 15–24 years)

Source: Eurostat.

Note: Data labels in the figure use International Organization for Standardization (ISO) country codes.

Stronger Collective Bargaining, Limited Effect on Youth Unemployment

A higher incidence of collective bargaining has the potential to lower employment, but the impact of collective bargaining depends on the level at which the bargaining occurs. Firm-level bargaining tends to limit wage increases beyond productivity levels, thereby having less of an impact on employment and unemployment rates. Very centralized or coordinated bargaining systems may also be less detrimental to employment because the overall impact on the labor force can be explicitly incorporated into the bargaining process, thereby minimizing the effect on unemployment. Thus, the relationship between the strength of collective bargaining and unemployment tends to be hump shaped, having the worst effects on unemployment when collective bargaining systems are neither fully centralized nor fully decentralized.

Figure 12.13Union Density


Source: Organisation for Economic Co-operation and Development.

Note: Data labels in the figure use International Organization for Standardization (ISO) country codes.

Overall, higher union density has a limited (not significant) effect on youth unemployment. Some specifications indicate that a 1 percentage point increase in union density could lower youth unemployment rates 0.2–0.6 percentage point.12 However, this finding is not robust to alternative specifications, including variations in control variables or allowing country-specific interactions. The results from some specifications suggest that higher union density may be associated with an altered employment composition as well, perhaps because employers prefer to hire nonunionized youth instead of unionized adults to preserve the flexibility to adjust the workforce as needed, leading to higher employment for youth and lower employment for adults.

More Vocational Training, Less Unemployment

Educational attainment may have a large impact on employability (OECD 2013). Vocational training and expanded access to training could play a significant role in reducing temporary work and contribute to making temporary jobs a stepping stone toward open-ended contracts (OECD 2004).

The share of workers in the population with low education has been declining steadily across all countries, but the level of formal education may not provide a complete picture of the skills of the young unemployed. Vocational training and apprenticeships are also important forms of attaining skills, but are difficult to measure. Survey data on the reasons for temporary contracts suggests that temporary contracts for youth are associated with education, training, or probation in countries with low youth unemployment rates such as Austria, Germany, and the Netherlands (Figure 12.14). The share of those in education, training, and probation is relatively smaller in Cyprus, Greece, Portugal, and Spain.

Figure 12.14Reason for Temporary Work Contract among Youth

(Percent of unemployed, 2012)

Source: Eurostat.

Note: See Abbreviations and Acronyms section for composition of EA17. Data labels in the figure use International Organization for Standardization (ISO) country codes.

Access to vocational training—measured by the share of temporary workers in probation or vocational training—significantly reduces youth unemployment, by about 0.3 percentage point, but has no significant effect for adults. A higher share of individuals with low education has generally no significant effects on youth unemployment or employment rates, but has a strong negative effect on adult unemployment and employment rates. Low education may be less of an obstacle for youth employment, perhaps because young workers can be more easily trained than adults.

More Spending on Active Labor Market Policies, Lower Unemployment

According to OECD (2006), most macroeconometric studies have found significant positive effects of spending on ALMP, especially spending on training, on aggregate unemployment. However, microeconomic evaluation studies of ALMPs find that the effectiveness of programs vary, and that apparently similar programs can yield very different outcomes (Card, Kluve, and Weber 2010; Kluve 2010). Studies also show that ALMPs that specifically target young people are not very effective regardless of the type of the program (that is, they have a lower probability of yielding positive results).

Spending on ALMPs varies widely across countries, and several countries have increased spending in this area since the crisis (Figure 12.15). Given dramatic increases in unemployment during the crisis, ALMP funds have had to be distributed across greater numbers of the unemployed.

Figure 12.15Spending on Active Labor Market Policies (ALMPs)

(Thousand euros per unemployed)

Higher spending on ALMPs, especially training, is associated with significant reductions in youth unemployment rates. Additional ALMP spending of 1,000 euro per unemployed reduces youth unemployment by about 0.3 percentage point and adult unemployment by about 0.1 percentage point. It also raises employment rates by 0.2 percentage point for youth and 0.1 percentage point for adult.


The youth unemployment problem in the euro area is multifaceted and varies across countries. Substantial cross-country differences in the composition and dynamics of youth unemployment suggest that no single policy at the EU or national level is likely to solve the problem. The solution would need to target the country-specific factors affecting youth unemployment. Past experience indicates that there is no one-size-fits-all approach to tackling youth unemployment (ILO 2013).

A comprehensive strategy for tackling youth unemployment in the euro area should focus on creating conditions for sustainable growth, given the higher sensitivity of youth unemployment to the business cycle. In the short term, policies to restore the housing sector and improve competitiveness in trade and tourism activities would be particularly beneficial given that these are also sectors in which youth employment is concentrated. Historically, euro area countries have reduced youth unemployment rates by growing much more strongly than they are currently expected to (Box 12.2).

Labor market reforms will help. The empirical results show that growth explains about half the increase in youth unemployment overall, and about a third in some high youth unemployment countries (such as France and Italy). Therefore, growth alone cannot solve the youth unemployment problem. Empirical analysis also shows that labor market reforms would pay dividends. As the economic recovery solidifies and unemployment rates return closer to their historical averages, labor market institutions may play an increasingly large role in labor market dynamics. Reforms could include lowering hiring costs by reducing the tax wedge and reconsidering minimum wage policies (which largely affect youth) to increase labor demand; changing unemployment benefits to provide better incentives for moving from inactivity to work; improving skill levels and work-related training; and implementing ALMPs.

Reforms may complement business cycle effects. Labor market institutions affect the sensitivity of youth unemployment rates to the business cycle in Greece, Ireland, Italy, and Spain (Figure 12.16). For example, an increase in the tax wedge increases the cyclical responsiveness of youth unemployment in both Greece and Spain. Similarly, ALMP spending seems to significantly reduce the effect of cyclical changes on youth unemployment in Ireland, Italy, and Spain. Thus, a reduction in ALMP spending per unemployed in Ireland and Spain in the aftermath of the crisis may have aggravated the effect of the sharp drop in growth on youth unemployment rates.

Figure 12.16Coefficients for Interaction between Labor Market Features and Output Gap

Source: IMF staff estimates.

Note: GRR = gross replacement rate; NRR = net replacement rate; ITRAP = inactivity trap; MIN2MED = ratio of minimum to median wage; TWEDGE = tax wedge; UDENS = union density; EPRT = OECD temporary employment protection index; TSHARE = share of temporary employees for a given age group; TPROB = share of temporary employees on probation in total temporary employees for a given age group; LOWEDUC = share of population with lower secondary education; ALMPTOT = total spending on ALMP (thousand euro per unemployed); ALMPTRAIN = spending on ALMP training policies.

Box 12.2.A Historical Perspective on Growth and Youth Unemployment

Youth unemployment rates have been very high in the euro area after the global economic and financial crisis, but such unemployment levels are not unprecedented (Figure 12.2.1). In Spain, for example, youth unemployment rates are close to (about 90 percent of) the previous peaks (1986 and 1994). Youth unemployment rates remain below their historical peaks for France and Italy.

Figure 12.2.1Euro Area: Youth Unemployment Rate

(Peak = 100)

Source: Eurostat.

Note: See Abbreviations and Acronyms section for composition of EA18. Data labels in the figure use International Organization for Standardization (ISO) country codes.

Euro area countries succeeded in reducing high youth unemployment rates in the past, but they have done so in the context of stronger growth than currently envisaged (Figure 12.2.2). In some cases, growth rates would have to be double or even triple the current forecasts.

Figure 12.2.2Average Growth during Unemployment Reduction and Current Projections


Sources: Eurostat; IMF World Economic Outlook (WEO); and IMF staff estimates.

Given the estimated effects on youth unemployment, the amounts of ALMP spending required to make a sizable dent in historically high youth unemployment rates would be too large to be feasible. Thus, ALMP spending will need to be complemented by growth and other labor market reforms to yield the maximum effect. Empirical analysis shows that ALMPs are likely to be more effective if they are part of a broader, comprehensive strategy to address structural impediments to greater youth employment; for example, higher tax wedges reduce the effectiveness of ALMP spending in Austria and Germany.

However, ALMP is not a panacea. ALMP programs need to be designed and monitored properly—meta analysis of such programs show that the impact and cost-effectiveness of ALMPs vary significantly based on their design.

Annex 12.1. Data Definitions
Output gap(Real GDP – Real potential GDP) as a percentage of real potential GDP.WEO
GDP growthYear-over-year growth of GDP, constant price.WEO
Unemployment rateUnemployed population as a percentage of labor force in corresponding age cohort.Eurostat
Net replacement rateNet benefits replacement rate is defined as the ratio of net income while out of work (mainly unemployment benefits if unemployed, or means-tested benefits, if on social assistance) divided by net income while in work. A lower net replacement rate is associated with a greater incentive to search for and take up a job when unemployed.European Commission Tax and Benefits Indicators Database
Gross replacement rateAverage of the gross unemployment benefit replacement rates for two earnings levels, three family situations, and three durations of unemployment.OECD
Inactivity trapThe implicit tax on returning to work for inactive persons; measures the part of additional gross wage that is taxed away when an inactive person (not entitled to receive unemployment benefits but eligible for income-tested social assistance) takes up a job. This indicator measures the financial incentive to move from inactivity and social assistance to employment.European Commission Tax and Benefits Indicators Database
Minimum wage/Median wageMinimum wage relative to median wage for full-time workers. This ratio is set to zero for countries without a national minimum wage.OECD
Protection of temporary workersStrictness of employment protection for temporary contracts.OECD
Share of temporary workersTemporary employees as a percentage of the total number of employees.Eurostat
Tax wedgeThe tax wedge is defined as the proportional difference between the costs of a worker to their employer (wage and social security contributions, that is, the total labor cost) and the amount of net earnings that the worker receives (wages minus personal income tax and social security contributions, plus any available family benefits). The tax wedge measures both incentives to work (labor supply side) and to hire persons (labor demand side).European Commission Tax and Benefits Indicators Database
Union densityTrade union density corresponds to the ratio of wage and salary earners that are trade union members, divided by the total number of wage and salary earners (OECD Labour Force Statistics). Density is calculated using survey data, wherever possible, and administrative data adjusted for nonactive and self-employed union members otherwise.OECD
Adjusted bargaining powerEmployees covered by wage bargaining agreements as a percentage of all wage and salary earners in employment with the right to bargaining, adjusted for the possibility that some sectors or occupations are excluded from the right to bargain (removing such groups from the employment count before dividing the number of covered employees over the total number of dependent workers in employment).The Quality of Government Institute, University of Gothenburg
Share of temporary workers on probationProportion of total temporary workers on probation (other reasons for being on temporary contracts include “could not find a permanent job,”“did not find a permanent job,”“in education or training”).Eurostat
Share of low-educated workersPersons with lower secondary education attainment.Eurostat
ALMP total spending per unemployedTotal ALMP spending per unemployed.Eurostat
ALMP spending per unemployed on trainingALMP spending on training per unemployed.Eurostat
Note: ALMP = active labor market policy; OECD = Organisation for Economic Co-operation and Development; WEO = IMF World Economic Outlook.
Note: ALMP = active labor market policy; OECD = Organisation for Economic Co-operation and Development; WEO = IMF World Economic Outlook.
Annex 12.2. Methodology

Estimating Okun’s Coefficient

The estimation of Okun’s coefficient for individual countries was conducted using the following specifications:

in which

Urateit: Unemployment rate for the youth and adult age groups in country i, year t

Urate*it: Structural (equilibrium) unemployment rate in a certain age group in country i, year t (estimated by using Hodrick-Prescott filter, with λ = 100)

Growthit: GDP growth rate in country i, year t

Output gapit: Output gap in country i, year t

Countryi: Country fixed effect

The estimated bi is the Okun’s coefficient.

Impact of Labor Market Institutions

Economists have advanced a number of models for the unemployment rate that are consistent with using the unemployment rate in levels as a dependent variable in reduced-form equations. For example, Nickell and Layard (1999) develop a wage-bargaining model with numerous identical firms, showing that the equilibrium level of the unemployment rate will be decreasing in any exogenous factor that increases the job separation rate (represented in this case by the output gap), increases the search effectiveness of the unemployed (represented by ALMPs), lowers the benefit replacement ratio, lowers the strength of workers in wage bargaining (union density), or raises the elasticity of product demand facing the firm. The latter argument even suggests scope for including variables associated with product market regulation into unemployment equations. Other examples of similar models include Scarpetta (1996) and Bassanini and Duval (2006), who estimate a specification very similar to the one in this chapter.13

The univariate model with interaction terms assumes that (1) the effects of the business cycle may depend on labor market features, (2) this dependence may be different across countries, and (3) the effect of the structural variable itself does not depend on the country, except indirectly via the business cycle. These assumptions, together with data limitations, mean that structural variables can only be considered one at a time; otherwise the high number of parameters to estimate relative to the size of the sample prevents efficient estimation of the coefficients.

We consider the following specification:

in which ui,t is one out of six dependent variables (youth and adult unemployment and long-term unemployment rates, as well as employment rates), ci is a dummy variable equal to 1 if the dependent variable is from country I; yi,t* is the and xi,t is a given labor market feature. Finally εi,t is the error term with standard assumptions.

The marginal impact of the change in labor market feature xi,t on the level of unemployment or employment is given by the following partial derivative:

That is, the impact of a change in a labor market feature differs depending on the country considered and its output gap. Crucially, therefore, the marginal effects of the change in the structural variable will depend on the value of the output gap yi,t* at which they are evaluated. In this chapter, this point is the country-specific average output gap (average growth rate for the specification in differences). The standard errors for the marginal effects are computed using the delta method.

The second specification considers several labor market features at a time and assumes that the impact of labor market features, if any, is common across all countries. It allows the impact of the business cycle (output gap) to vary across countries.

In which

Uratei,t: Unemployment rate for the youth and adult age groups in country i, year t

Xijt: Labor market institution j, in country i, year t

A variety of robustness checks are performed, for example, including time effects, using different measures of the output gap and youth unemployment, and others. These results are available upon request.


    BallL. M.D.Leigh and P.Loungani.2013. “Okun’s Law: Fit at Fifty?Working Paper 18668National Bureau of Economic ResearchCambridge, Massachusetts.

    • Search Google Scholar
    • Export Citation

    BassaniniA. and R.Duval.2006. “The Determinants of Unemployment across OECD Countries: Reassessing the Role of Policies and Institutions.Economic Studies 42Organisation for Economic Co-operation and DevelopmentParis.

    • Search Google Scholar
    • Export Citation

    Bernal-VerdugoL. E.D.Furceri and D.Guillaume.2012. “Labor Market Flexibility and Unemployment: New Empirical Evidence of Static and Dynamic Effects.Comparative Economic Studies54 (2): 25173.

    • Search Google Scholar
    • Export Citation

    BertolaG.F. D.Blau and L. M.Kahn.2007. “Labor Market Institutions and Demographic Employment Patterns.Journal of Population Economics20 (4): 83367.

    • Search Google Scholar
    • Export Citation

    CardD.J.Kluve and A.Weber.2010. “Active Labor Market Policy Evaluations: A Meta-Analysis.Economic Journal120 (548): F45277.

    • Search Google Scholar
    • Export Citation

    ChoudhryM.E.Marelli and M.Signorelli.2012. “Youth and Total Unemployment Rate: The Impact of Policies and Institutions.” Unpublished.

    • Search Google Scholar
    • Export Citation

    European Commission. 2012. “Moving Youth into Employment.Communication from the Commission to the European Parliament the Council the European Economic and Social Committee and the Committee of the Regions727Brussels.

    • Search Google Scholar
    • Export Citation

    European Commission. 2013a. “Labor Market Developments in Europe 2013.European Economy 6/2013Brussels.

    European Commission. 2013b. “Working Together for Europe’s Young People: A Call to Action on Youth Unemployment.Communication from the Commission to the European Parliament the Council the European Economic and Social Committee and the Committee of the Regions447Brussels.

    • Search Google Scholar
    • Export Citation

    International Labor Organization (ILO). 2013. Global Employment Trends for Youth 2013: A Generation at Risk. Geneva: International Labour Office.—dgreports/—dcomm/documents/publication/wcms_212423.pdf.

    • Search Google Scholar
    • Export Citation

    KluveJ.2010. “The Effectiveness of European Active Labor Market Programs.Labour Economics17 (6): 90418.

    NickellS. and R.Layard.1999. “Labor Market Institutions and Economic Performance.” In Handbook of Labor Economics Vol. 3C edited by O.Ashenfelter and D.Card.Amsterdam: North-Holland.

    • Search Google Scholar
    • Export Citation

    Organisation for Economic Co-operation and Development. 2004. Employment Outlook. Paris: OECD Publishing.

    Organisation for Economic Co-operation and Development. 2006. OECD Employment Outlook: Boosting Jobs and Growth. Paris: OECD Publishing.

    • Search Google Scholar
    • Export Citation

    Organisation for Economic Co-operation and Development. 2012. “Under Shock: How to Spread Macroeconomic Risks More Fairly.” In Economic Policy Reforms 2012: Going for Growth. Paris: OECD Publishing.

    • Search Google Scholar
    • Export Citation

    Organisation for Economic Co-operation and Development. 2013. Education at a Glance 2013: OECD Indicators. Paris: OECD Publishing.

    • Search Google Scholar
    • Export Citation

    ScarpettaS.1996. “Assessing the Role of Labour Market Policies and Institutional Settings on Unemployment: A Cross-Country Study.OECD Economic Studies26: 4398.

    • Search Google Scholar
    • Export Citation

    ScarpettaS.A.Sonnet and T.Manfredi.2010. “Rising Youth Unemployment during the Crisis: How to Prevent Negative Long-Term Consequences on a Generation?OECD Social Employment and Migration Working Paper No. 106OECD PublishingParis. doi:10.1787/5kmh79zb2mmv-en.

    • Search Google Scholar
    • Export Citation

This chapter is based on Euro Area Policies: Article IV 2014 Consultation—Selected Issues, IMF Country Report 14/199,2014.

The authors thank Jesse Siminitz and Xiaobo Shao for excellent research assistance and Katherine Cincotta for general assistance. They also thank Ana Lamo of the European Central Bank and Alessandro Turrini of the European Commission, the participants of seminars at the European Central Bank and the European Commission, as well as IMF European Department country teams for helpful feedback and comments.


Henceforth, youth refers to individuals ages 15–24 years old, and adults refers to individuals ages 25–64 years old. Unemployment refers to the unemployment rate.


The youth labor force tends to be smaller than the labor force for other age cohorts because young individuals may choose to pursue full-time education, although participation in education does not necessarily exclude participation in the labor force (for example, through part-time work or apprenticeships). The youth labor market is also characterized by frequent search and matching as individuals look for better jobs, using intermediate stages for accumulating experience (and perhaps occasionally dropping out of the labor force).


Euro area countries include Austria, Belgium, Cyprus, Estonia, Finland, France, Germany, Greece, Ireland, Italy, Latvia, Luxembourg, Malta, Netherlands, Portugal, Slovak Republic, Slovenia, and Spain.


See Ball, Leigh, and Loungani (2013) for an extensive discussion on this topic.


Even in those cases, some studies have found that Okun’s Law holds for measures of hours worked.


SMEs are defined as firms with fewer than 250 employees, turnover of less than €50 million, or balance sheets of less than €43 million.


These results are robust to changes in specification. Results for changes in unemployment rates are available upon request. Results discussed in the text refer to the multivariate model unless otherwise specified.


Several studies find that higher labor tax wedges raise unemployment, and the impact of the tax wedge is strengthened when combined with the impact of minimum wages and the strength of collective bargaining.


The minimum wage was frozen in Portugal in the second half of 2011 and cut in Greece in the second half of 2012 under the financial assistance programs.


Labor market duality is measured by the Organisation for Economic Co-operation and Development’s (OECD’s) employment protection indicator for temporary and permanent workers, and the share of temporary workers as a percentage of total employees. In-sample variation in the data is larger for temporary employment protection indicators than for permanent employment protection indicators.


The rating is on a scale from 1 to 6.


This result is based on the OECD’s indicator on union density, which measures the incidence of unionization among the employed but does not measure the degree of centralization (Figure 12.13).


Bassanini and Duval (1996) estimate a reduced-form equation (ui,t = Σj βjXj + χGi,t + αi + λt + εi,t) consistent with a variety of theoretical models of the labor market equilibrium (job search, wage setting), in which unemployment is regressed on a series of structural variables (in vector χ), an output gap measure (G), as well as country and time fixed effects. The analysis in this chapter departs from this specification by including interaction terms and excluding time fixed effects.

    Other Resources Citing This Publication