A. Introduction

1. Labor market conditions in Colombia present a challenge. At end–2004, the unemployment rate amounted to 12 percent, and about one–third of the labor force was considered underemployed. The steady economic growth in 2003–04 helped lower the unemployment rate from 20 percent in 2000. However, while cyclical factors play an important role, it seems clear that structural rigidities also affect labor markets. Anecdotal evidence suggests that rigidities introduced by regulations may drive employment into the informal sector as employers seek to avoid these requirements. Jobs in the informal sector are often less secure, raising concerns about the stability of future incomes. In short, a lack of jobs in the formal sector is often considered to reflect structural problems, and could contribute to high poverty rates in Colombia.

2. Concerns about structural rigidities prompted a major labor market reform in December 2002 aimed at increasing flexibility. Unemployment and informality have declined since the reforms, with the economy generating hundreds of thousands of new jobs. At the same time, the economic recovery solidified and gained momentum, thus making it difficult to distinguish between the impact of the labor market reforms and the employment gains that typically accompany stronger economic growth.

3. This paper will examine conditions in Colombian labor markets. The following section reviews labor market developments leading up to as well as since the reforms. Subsequent sections examine structural issues in the labor markets, then review the labor market reforms, and analyze the impact of reforms versus stronger growth. The final section explores remaining challenges.

B. Recent Labor Market Developments

The economic crisis in 1999 interrupted a sustained improvement in labor markets

4. Two decades of solid macroeconomic performance contributed to a firming of labor market conditions through the mid-1990s. Strong economic growth and moderating inflation provided a sound framework for job growth, as real GDP growth was robust through 1998. Macroeconomic stability improved as well, with inflation decelerating from over 30 percent in the early 1990s to below 20 percent in 1997, with further declines to single digits since then. These economic conditions fostered a steady improvement in labor markets. Unemployment rates declined from 14 percent in the mid-1980s to below 8 percent in 1994. Strong demand for labor lifted wage growth contributing to a decline in poverty rates.

Real Wage Growth

(In percent per year)

Citation: IMF Staff Country Reports 2005, 162; 10.5089/9781451808872.002.A001

• Download Figure
• Download figure as PowerPoint slide

View Full Size

Real Wage Growth

(In percent per year)

Citation: IMF Staff Country Reports 2005, 162; 10.5089/9781451808872.002.A001

• Download Figure
• Download figure as PowerPoint slide

Real Wage Growth

(In percent per year)

Citation: IMF Staff Country Reports 2005, 162; 10.5089/9781451808872.002.A001

• Download Figure
• Download figure as PowerPoint slide

5. Employment conditions worsened considerably, though, during the economic crisis of the late 1990s. Real GDP stagnated in 1998 and contracted by more than 4 percent in 1999 as the regional crisis spilled over to Colombia. The unemployment rate rose from 7½ percent in 1994 to over 20 percent by the end of the decade. The poor job market cut across all major demographic groups, with significant increases in unemployment among both men and women, adult workers, youth, and older workers. Longer-term unemployment worsened, with those looking for work for more than 25 weeks rising from less than 40 percent of total unemployed prior to the crisis to 65 percent in 2000.

Unemployment Rate

(In percent)

Citation: IMF Staff Country Reports 2005, 162; 10.5089/9781451808872.002.A001

• Download Figure
• Download figure as PowerPoint slide

View Full Size

Unemployment Rate

(In percent)

Citation: IMF Staff Country Reports 2005, 162; 10.5089/9781451808872.002.A001

• Download Figure
• Download figure as PowerPoint slide

Unemployment Rate

(In percent)

Citation: IMF Staff Country Reports 2005, 162; 10.5089/9781451808872.002.A001

• Download Figure
• Download figure as PowerPoint slide

6. These unemployment trends understate the deterioration in labor markets, as the quality of employment worsened as well. A modest decline in overall employment masked a shift from those with jobs in the formal sector, to underemployment, informal and self-employment. Informal sector employment rose from an already-high 52 percent to 60 percent of total employment. Self employed workers, who often are unable to find higher-paying jobs in the formal sector, rose more than 11 percent from 1997 to 1999, from 1.65 million to 1.84 million. This shift toward lower-quality jobs hurt earnings, with real wages declining 20 percent from mid-1998 to 2000. At the same time, the share of workers earning the minimum wage nearly doubled, from 20 percent of all workers to 37 percent.

7. Labor force participation rose to bolster family incomes.Women and younger family members entered the labor force to supplement or replace income from the primary wage earner. The female participation rate rose from 47 percent to 57 percent between 1996 and 2000, while the participation rate among adult males, in contrast, was stable at 75 percent.

C. Sources of Labor Market Rigidity

10. Despite these gains, unemployment rates remain elevated and the informal sector still accounts for an unusually large share of total employment. This section addresses several sources of structural rigidities in Colombian labor markets.

• Nonwage labor costs. Nonwage costs are high in Colombia and increased further as a result of labor market reforms in the early 1990s, to over 50 percent of total labor costs. Such costs provide disincentives to hiring and can reduce allocative efficiency in labor markets. Chief among these costs are severance payments, pension costs and other non–wage levies. Funding for special fiscal funds imposes a significant nonwage cost to employers through a 9 percent surcharge on wages, including 3 percent for the Cajas de Compensacion, which encompass a number of programs in the workplace. The absence of an unemployment insurance scheme has reinforced the reliance on severance payments as a source of financing for unemployment spells, as well as making the funding of employment–related programs highly procyclical.

  Nonwage labor costs undercut labor market performance in Colombia, according to several studies. Bernal and Cárdenas (2003) estimate a relatively high cost elasticity of labor demand, and suggest that nonwage costs were responsible for much of the rise in unemployment in the 1 990s. Kugler (1999) finds that lower severance costs following the 1990 reforms improved labor market flexibility by increasing both separations and hiring, with a net positive impact on employment, including increased hiring from the informal sector into the formal sector. Greater employment flexibility likely improved allocative efficiency in the labor markets.

• Minimum wage. Colombia has a relatively high minimum wage that applied to over one–third of all workers in 2002 and appears to act as a binding constraint on hiring. Moreover, researchers have found the minimum to have broader effects in wage-setting in the formal sector not just at wages near the minimum but also at levels further up the pay scale, and in the informal sector as well (Maloney and Nuñez, 2000). This broader effect on wage setting behavior may exacerbate the severity of the impact of minimum wages on labor market flexibility. Moreover, minimum wages in Colombia appear to have worsened income inequality by depressing employment of less-skilled workers, especially women and younger workers (Arango and Pachon (2004), Angel-Urdinola (2004)).

  Economic distortions due to the minimum wage have likely intensified in recent years, as the real minimum wage rose significantly in the late 1990s due in part to the backward-looking process of adjusting nominal minimum wages for inflation. The nominal minimum wage is adjusted annually based roughly on the previous year’s inflation rates. During a period of declining inflation rates, this process resulted in nominal increases faster than current inflation. As a result, the real minimum wage rose nearly 20 percent over the past decade, inhibiting the labor market adjustment to the economic crisis.

• Other labor market regulations. Colombia has historically had strict regulations regarding the workweek, overtime wages, severance payments and job security, and other aspects of the labor market. Such labor market regulations may have detrimental effects on employment. Botero et al. (2003) find labor market regulation across 85 industrialized and developing economies to be associated with a larger informal economy, lower labor force participation, and higher unemployment, especially among younger workers. According to an index of labor market regulation the authors construct, Colombia’s labor markets were more heavily regulated than three–quarters of the 41 middle–income countries in their sample. Interestingly, they also find that richer countries regulate labor markets less than do poorer countries, but offer a more generous social safety net (Table).

• Low worker skills. Commentators have noted an apparent mismatch between skill levels in the supply and demand for labor, with the surplus of low–skilled workers accounting for a high share of unemployment and underemployment. Furthermore, the rapid pace of technological change may have worsened the skills gap in recent years. To address the level of workers’ skills the Ministry of Social Protection has included education and training in its strategy to improve labor market performance.

• Security. A long guerrilla civil conflict and drug–related violence have disrupted economic activity, rendered certain areas of the country unsafe for commerce, investment, and mineral and oil exploration and development. Crime and drug activity may have also diverted labor and capital to protection and security as well as socially unproductive activities. Cárdenas (2002) compares the level of conflict in Colombia to a broad range of countries and argues that crime and security concerns were significant factors in the drop in productivity in recent years following several decades of robust gains.

CPI Inflation and Change in Nominal Minimum Wage

(In percent per year)

Citation: IMF Staff Country Reports 2005, 162; 10.5089/9781451808872.002.A001

• Download Figure
• Download figure as PowerPoint slide

View Full Size

CPI Inflation and Change in Nominal Minimum Wage

(In percent per year)

Citation: IMF Staff Country Reports 2005, 162; 10.5089/9781451808872.002.A001

• Download Figure
• Download figure as PowerPoint slide

CPI Inflation and Change in Nominal Minimum Wage

(In percent per year)

Citation: IMF Staff Country Reports 2005, 162; 10.5089/9781451808872.002.A001

• Download Figure
• Download figure as PowerPoint slide

Employment Laws Index for Selected Countries, by GNP Per Capita

Source: Botero et al (2003)

¹Measures restrictions against alternative labor contracts, such as part-time work.

²easures the flexibility of employment contracts with regard to working time requirements (working hours, mandated rest, overtime), mandatory payments for nonworking days (paid annual leave, holidays, maternity leave), and minimum wage legislation.

³easures legal protections against dismissal and costs to employers to terminate employment contracts.

⁴The overall index is aggregated from subcomponents that measure the degree of labor market protection or regulation regarding alternative employment contracts, conditions of employment, and job security. Higher values indicate greater labor market rigidities.

Employment Laws Index for Selected Countries, by GNP Per Capita

	Subindicies:
Selected Countries	Alternative Employment Contracts 2	Conditions of Employment 3	Job Security 4	Overall Index 1
Bottom 25 percentile of GNP per capita (median)	0.56	0.73	0.38	1.72
Uganda	0.44	0.77	0.50	1.71
Zambia	0.56	0.59	0.00	1.15
Bolivia	0.39	0.87	0.57	1.82
Ukraine	0.72	0.84	0.68	2.24
Indonesia	0.83	0.50	0.43	1.75
Middle 50 percentile of GNP per capita (median)	0.56	0.75	0.41	1.75
Philippines	0.39	0.65	0.57	1.61
Ecuador	0.57	0.62	0.67	1.86
Dominican Republic	0.56	0.77	0.33	1.65
Jamaica	0.56	0.48	0.13	1.16
Peru	0.23	0.74	0.70	1.67
Colombia	0.56	0.82	0.62	1.99
Russian Federation	0.78	0.75	0.68	2.21
Panama	0.87	0.84	0.67	2.38
Turkey	0.72	0.81	0.20	1.74
Venezuela	0.85	0.84	0.64	2.32
Mexico	0.53	0.77	0.71	2.01
Brazil	0.85	0.86	0.69	2.40
Chile	0.70	0.55	0.31	1.56
Uruguay	0.72	0.52	0.03	1.27
Argentina	0.39	0.72	0.44	1.55
Top 25 percentile of GNP per capita (median)	0.56	0.39	0.21	1.25
Canada	0.56	0.49	0.17	1.22
United Kingdom	0.56	0.26	0.20	1.02
Hong Kong	0.56	0.19	0.01	0.76
Singapore	0.56	0.19	0.11	0.85
United States	0.56	0.29	0.08	0.92
Japan	0.59	0.64	0.19	1.42
Median, all countries	0.56	0.65	0.34	1.66

Source: Botero et al (2003)

¹Measures restrictions against alternative labor contracts, such as part-time work.

²easures the flexibility of employment contracts with regard to working time requirements (working hours, mandated rest, overtime), mandatory payments for nonworking days (paid annual leave, holidays, maternity leave), and minimum wage legislation.

³easures legal protections against dismissal and costs to employers to terminate employment contracts.

⁴The overall index is aggregated from subcomponents that measure the degree of labor market protection or regulation regarding alternative employment contracts, conditions of employment, and job security. Higher values indicate greater labor market rigidities.

View Table

Employment Laws Index for Selected Countries, by GNP Per Capita

	Subindicies:
Selected Countries	Alternative Employment Contracts 2	Conditions of Employment 3	Job Security 4	Overall Index 1
Bottom 25 percentile of GNP per capita (median)	0.56	0.73	0.38	1.72
Uganda	0.44	0.77	0.50	1.71
Zambia	0.56	0.59	0.00	1.15
Bolivia	0.39	0.87	0.57	1.82
Ukraine	0.72	0.84	0.68	2.24
Indonesia	0.83	0.50	0.43	1.75
Middle 50 percentile of GNP per capita (median)	0.56	0.75	0.41	1.75
Philippines	0.39	0.65	0.57	1.61
Ecuador	0.57	0.62	0.67	1.86
Dominican Republic	0.56	0.77	0.33	1.65
Jamaica	0.56	0.48	0.13	1.16
Peru	0.23	0.74	0.70	1.67
Colombia	0.56	0.82	0.62	1.99
Russian Federation	0.78	0.75	0.68	2.21
Panama	0.87	0.84	0.67	2.38
Turkey	0.72	0.81	0.20	1.74
Venezuela	0.85	0.84	0.64	2.32
Mexico	0.53	0.77	0.71	2.01
Brazil	0.85	0.86	0.69	2.40
Chile	0.70	0.55	0.31	1.56
Uruguay	0.72	0.52	0.03	1.27
Argentina	0.39	0.72	0.44	1.55
Top 25 percentile of GNP per capita (median)	0.56	0.39	0.21	1.25
Canada	0.56	0.49	0.17	1.22
United Kingdom	0.56	0.26	0.20	1.02
Hong Kong	0.56	0.19	0.01	0.76
Singapore	0.56	0.19	0.11	0.85
United States	0.56	0.29	0.08	0.92
Japan	0.59	0.64	0.19	1.42
Median, all countries	0.56	0.65	0.34	1.66

Source: Botero et al (2003)

¹Measures restrictions against alternative labor contracts, such as part-time work.

²easures the flexibility of employment contracts with regard to working time requirements (working hours, mandated rest, overtime), mandatory payments for nonworking days (paid annual leave, holidays, maternity leave), and minimum wage legislation.

³easures legal protections against dismissal and costs to employers to terminate employment contracts.

⁴The overall index is aggregated from subcomponents that measure the degree of labor market protection or regulation regarding alternative employment contracts, conditions of employment, and job security. Higher values indicate greater labor market rigidities.

D. The Labor Law Reform of 2002

11. The Labor Market Reform Law 789 of December 2002 aimed to increase the flexibility of the labor market while protecting vulnerable segments of the population. The reform sought to stimulate demand for labor by increasing the flexibility of labor contracts, encouraging training through apprenticeships, and expanding protection for the unemployed through training programs and a temporary unemployment subsidy. Restrictions on the length of the workday and workweek were relaxed or eliminated, making it easier for employers to hire for peak periods and weekends, or to use multiple shifts in factory work. Mandatory overtime charges for nighttime, holidays, and Sunday work were also reduced. Severance payments for unilateral dismissal—which may make employers reluctant to hire new workers in the face of uncertainty about future business demand—were lowered. The contract for apprentice programs, meanwhile, was changed to stimulate demand for apprentices, reducing wages during a training stage to 50 percent of the minimum wage, and to 75 percent during an on–the–job employment phase of apprenticeship.

E. The Impact of the Labor Market Reforms

12. It is difficult to separate the effects of labor market reforms in generating employment gains from the general effects of economic recovery. Gaviria (2004) suggests that economic growth has been the main driver of labor market improvements, and that the overall effects of the reforms have been slight. Relying on indirect evidence from a survey of businesses, as well as an empirical examination of employment in the commercial, service, and manufacturing sectors, Gaviria finds no evidence that the reforms stimulated employment. He does find, however, that efforts to increase apprenticeship have been largely successful. Also, underemployment due to insufficient hours fell significantly, presumably as a result of the reform to increase the flexibility of the workday and workweek, suggesting some further beneficial impact of the reforms.

13. These findings have been somewhat controversial. One concern centers on the validity of the control group for the impact of the reforms. The study assumes that the reforms had little or no effect on the manufacturing sector, which can thus serve as a control group. Accordingly, the methodology compares employment trends in commercial and service sectors relative to manufacturing. Finding little difference in employment growth across sectors, these results are interpreted as indicating little impact of the reforms. This lack of divergence is also consistent, however, with the reforms having a positive effect, but one that is similar across all three sectors, including manufacturing. Furthermore, some analysts have observed that the sample of businesses in the paper’s survey is not very representative.

14. Other research, in contrast, provides more direct support for the effectiveness of the reforms. One method of testing the impact of the reforms is to examine whether the relationship between real GDP growth and employment has changed since the reforms were enacted. López et. al. (2004) find that employment has, in fact, risen more rapidly for a given increase in real GDP since the reform than in prior periods. ² Specifically, according to these estimates, employment was 3.4 percent higher in 2004 than it would have been in the absence of reforms, representing a net increase of 260,000 jobs. Moreover, this partial–equilibrium analysis may understate the overall effect if the reforms, by stimulating labor demand or allocative efficiency of labor markets, lifted real GDP growth as well.

15. The quality of jobs improved, with a marked shift from informal to formal employment. The sectoral detail of employment suggests that gross flows may have been more important than net flows, with significant net hiring out of the informal sector into the formal. Formal employment rose 325,000 between mid-2002 and mid-2004, with most of the gains occurring at larger firms. Informal employment, in contrast, rose modestly in 2003 but declined in 2004, decreasing slightly in net terms (López et. al., p.28). While more rapid real GDP growth would be expected to boost labor demand in both the formal and informal sectors, the reform eased restraints on formal employment. Thus, the greater strength of employment in the formal sector is consistent with these reforms having had some of the desired effects. Other signs of improved job quality include a decrease in underemployment, as well as a decline in labor force participation rates by secondary household workers, presumably because the incomes of primary workers have stabilized.

16. Spells of unemployment have shortened in duration as greater labor market flexibility encouraged flows from unemployment to employment. Nuñez (2004) estimates the impact of the reforms by comparing movements between unemployment and employment in the formal sector, where the flexibility of labor contracts improved under the new legislation, with the flows in and out of the informal sector, where the labor laws do not formally apply. Increased flows out of unemployment in the formal sector reduced the duration of unemployment spells since the reforms were enacted. Furthermore, the improvements were greatest among younger and less–qualified workers, suggesting the reform has had positive consequences on the distribution of income as well.

17. Workers receiving benefits under social security rose as well. Two factors combined to increase the number of workers covered by government benefits. First, coverage ratios rose in both formal and informal sectors. Coverage within the formal sector increased 1¼ percentage point to nearly 90 percent. Coverage was up a similar amount in the informal sector (but from a much lower base), rising to 45 percent. Second, the employment shift from the informal to formal sectors helped boost overall coverage, which increased 2½ percentage points (López et. al., p. 29). While the timing of this shift toward the formal sector and increased coverage is suggestive of an impact of the reforms, though, one cannot rule out that it is a result of stronger economic growth, as real GDP continued to grow in 2003 and into 2004.

18. Greater flexibility of the workday appears to have bolstered employment generation since the reforms. Gross employment data do not allow one to distinguish which aspects of the reform had the greatest impact. To overcome this, the Ministry of Social Protection conducted a survey of some 200 firms. Responses suggest that more flexible regulations of work schedules played a major role in more rapid employment generation. Over three–quarters of respondents indicated the changes to the workday and reduced overtime charges for holiday and Sunday work were the most aspect of the reforms. Reduced charges for dismissals were also considered important (López et. al. pp. 29–30).

19. The reforms appear to have stimulated a sharp increase in apprenticeships. The total number of apprentices rose 89 percent in the first year following the reforms (Gaviria, p. 20). Most of this increase has been in the on–the–job phase of apprenticeship, while the number in training has not risen substantially (López et. al., p. 33).

20. Programs to support the unemployed, however, have fallen short of the reform’s targets. Administrative difficulties in these programs appear to have slowed their execution. As of mid-2004, for example, only 69 percent of the resources available for unemployment subsidies and been utilized, while spending on training programs only achieved 36 percent of the target (Gaviria p. 18).

21. Results to date may be understated if the reforms have fuller effect only over a longer period. The difficulties in evaluating the effectiveness of the reforms are compounded by the relatively short time period that has elapsed since their enactment. As hiring decisions and job search processes tend to take a considerable amount of time to complete, the full effect of the reforms to increase the flexibility of labor markets is unlikely to have been realized in just two years. Other aspects of the government’s strategy to reduce structural unemployment, including apprenticeship and training programs, may take even longer to bear fruit.

F. Remaining Challenges

22. Reforms to date are perhaps best characterized as an important first step whose ultimate results are not yet fully apparent, while not being a complete solution to structural labor market issues. Colombian labor markets still have several sources of structural weakness. Some—like the high real minimum wage, and heavy nonwage labor costs—are specific to labor market regulations and could be addressed through policy changes. More general influences may stifle labor demand as well, including the skill mismatch in the workplace, the pace of overall real GDP growth, and the security situation.

23. Real minimum wages remain high. There is strong political opposition to reducing the minimum wage, and the Constitutional Court has ruled that the minimum wage must continue to be adjusted for past inflation. Thus, while a gradual reduction in the real minimum wage would likely have a favorable impact on employment generation, particularly in the formal sector (with the attendant increase in wages, benefits, and coverage under social security), as well as the distribution of income, progress over the near term appears unlikely.

24. Nonwage labor costs—if not lowered—may continue to weigh on formal sector employment. One hurdle in reducing the nonwage costs is the absence of an alternative revenue source for the public sector for employment–related programs and other social benefits. A key part of removing the disincentives to employment, therefore, will be to strengthen the overall revenue framework to reduce the reliance on nonwage levies.

25. Further efforts to build human capital, through education, training, and apprenticeships, would likely bolster employment. Several studies have suggested that a large supply of unskilled or low–skilled labor relative to labor demand contributes to higher unemployment, especially in the face of technological change in recent years (for example, see Arango, Posada, and Uribe). Education and training may facilitate the movement of these workers into higher skilled positions. Furthermore, as the minimum wage and high nonwage costs may have a greater effect on lower–skilled workers, such training efforts that successfully boost skill levels may diminish the negative effects of these other structural problems as well.

26. More general macroeconomic and societal improvements will also ameliorate labor market issues. Two factors in particular will be critical in reducing unemployment over the medium term. First, overall GDP growth will need to be sustained in line with potential growth rates in order to generate sufficient demand for labor. Second, further stabilization of the security situation would continue to reduce the disruptions that crime and violence have caused to economic activity and investment spending, and increased security costs. Improvements in the security situation may have some offset on unemployment rates, though, if a large number of former combatants need to be reabsorbed into the domestic labor force.