Designing labor market regulations in developing countries Updated

Measuring the effects of labor market regulations

Quantitative analysis typically relies on econometric techniques to isolate the effects of regulations such as employment protection legislation and the minimum wage from the effects of other variables. The older, more traditional, approach is to explain the cross-country variation in outcomes of interest (such as the national unemployment rate) by differences in country labor market rules, controlling for other factors that might affect these outcomes. This approach is still used, but it now shares the stage with methodologies that analyze what happens to these outcomes of interest in a single country where there is either within-country variation in the regulations (e.g. by state) or a change in the regulation, so that before-and-after effects can be observed. Another issue is measuring labor market regulation. This is particularly challenging in the case of job security rules, which are inherently qualitative and multi-dimensional. Researchers have relied on established indices such as the OECD Indicators of Employment Protection and the World Bank's Doing Business Employing Workers data. However, in recent years, a number of new indices have been developed, such as the ILO's EPLex and the Labor Regulation Index produced by the University of Cambridge Centre for Business Research.

Minimum wages

Much of the evidence in developing countries has examined whether higher minimum wages—by making labor more costly—reduce employment and increase unemployment. Clearly, country context matters a lot and the impact depends on labor market conditions, the minimum wage level, and compliance. Several studies have identified some adverse employment effects [2]. For example, increases in the minimum wage have been found to reduce employment in Brazil, China, Colombia, Costa Rica, Hungary, Indonesia, Nicaragua, and Thailand. However, the effect is generally modest, and some studies find no overall employment impact, for example in Mexico and South Africa.

Not surprisingly, where there are employment effects, they are concentrated in the lower part of the wage distribution, where the minimum wage actually “bites.” For example, a study for Costa Rica found that a 10% increase in the minimum wage was six times more likely to affect formal sector employment for workers earning within 20% of the minimum wage than for formal sector workers overall.

So the effects of the minimum wage—positive and negative—are generally concentrated among young people and unskilled workers (and, in some cases, women), because these groups tend to have lower wages. Positive effects can stem from the higher wages that these workers receive when the minimum wage is raised. However, studies in a number of countries in Asia, Eastern Europe, Latin America, and sub-Saharan Africa have found that raising the minimum wage reduced employment for these low-wage groups.

An important question for developing countries concerns the implications, if any, of minimum wages for the informal sector. Although the minimum wage is not enforced in the informal sector, it can still affect the sector. Economic theory predicts that increasing the minimum wage, by raising relative labor costs in the formal sector, would shift employment to the informal sector. A meta-analysis in middle-income countries finds that the probability of being employed in the informal sector increases as the minimum wage rises [3]. However, while some studies have found that higher minimum wages do shift employment from the formal to the informal sector, other studies have found no increase in informality. How is that possible? One surprising observation from several studies is that increasing the minimum wage often increases wages in the informal sector as well as in the formal sector—the so-called “lighthouse effect.” The minimum wage is seen as a benchmark wage for unskilled labor throughout the economy, even in the informal sector where it is not binding.

The empirical research, largely from Latin America, shows that higher minimum wages generally reduce earnings inequality by raising wages in the lower part of the income distribution. However, the equalizing effects of raising the minimum wage disappear when it is set fairly high, since this benefits primarily higher wage workers [4].

Advocates often present a higher minimum wage as an antipoverty policy. However, raising the minimum wage might not reduce poverty if a higher minimum wage reduces employment or if workers benefitting from the increases are not in poor households. So poverty reduction strategies require other more targeted instruments besides the minimum wage. In most countries, policymakers can expect that raising the minimum wage will shift more of the national income to labor.

Job security rules

Employment effects are also an open question in the case of job security legislation. A collection of studies analyzing the employment effects of changes in employment protection in several Latin American and Caribbean countries finds inconclusive results, both for cross-country regressions and within-country analyses [5]. For some countries, the relationship between the strictness of employment protection laws and employment is significant and negative (Argentina, Colombia, and Peru), but for others no significant relationship is found (Brazil and three Caribbean countries).

In general, negative employment effects have been most prevalent in industries with high turnover. Much of the analysis of the employment impact of job security legislation in developing countries has concerned Latin American countries. This issue has also received substantial attention in India where it has been the subject of considerable controversy. Many researchers have found that stringent layoff rules have constrained formal sector employment growth but others have seriously questioned the data and methods used to arrive at these findings.

For developing countries, an important consideration is whether employment protection legislation has different effects on formal and informal employment. Conventional dual-sector theories predict that more costly job security rules would shift production from the formal to the informal sector. While not all studies have come to this conclusion, most have found that more protective regulations are associated with modestly lower formal sector employment and some increase in informal employment [6].

Job security rules do not have the same effects on all types of workers. Not surprisingly, effects are most favorable for workers covered by these protections—typically, prime-working-age males and better-educated workers are overrepresented in this group. Workers in uncovered jobs gain no benefit from the protections and can even be hurt if some employers are reluctant to hire formal workers when job security rules are strong. Youth, women, and the less-skilled are overrepresented in this group.

Rules discouraging temporary contracts and making dismissals costly lengthen both job tenure and unemployment duration. These two effects, together, explain the common finding of modest or insignificant overall employment effects of stricter employment protection laws. At the same time, the pace of labor reallocation slows when job security protections are strong, resulting in a less dynamic labor market, with smaller flows between jobs and between employment and non-employment.

Does the strictness of employment protection legislation have implications for productivity? The results are mixed for developed countries, which have been studied most often. The results are also inconclusive for some limited analyses for developing countries. One study finds that increased employment protection has a significant negative impact on total factor productivity growth (of around –1% annually) where the rule of law is strong, but no effect where the rule of law is weak (as it is in many developing countries) [7]. Another study concludes that stricter employment protection does not robustly affect labor productivity, although it does affect output, primarily through a decline in new firms [8].

These results reflect the diverse, and often opposing, ways in which employment protection legislation may affect productivity. By slowing labor reallocation, strict job security rules can limit the potential efficiency gains from the movement of workers from low- to high-productivity sectors and firms. However, higher productivity can result where firms benefit from the experience of longer-tenure workers or when they adjust to lower flexibility by investing more in capital or in the training of the existing workforce.

Setting regulations on the “plateau” and avoiding the “cliffs”

Overall, the empirical evidence on the effects of labor regulations is mixed and, in some ways, inconclusive. To summarize what is known: minimum wages and job security rules have clear distributional effects, most obviously by narrowing earnings inequality, at least in the case of minimum wages. Increases in the minimum wage and the degree of protection provided by job security rules often lead to some shifts in jobs from the formal to the informal sector. They can also shift the composition of formal employment away from groups such as youth, the less-skilled, and in some cases, women. Effects on efficiency are less evident. Negative employment effects of higher minimum wages and stricter employment protection have been found in some countries but not all and when they are observed, the magnitude of the effect appears to be modest. Studies on output and productivity effects do not reach strong conclusions either way, though more research is needed in developing countries.

The results regarding efficiency effects (on employment and productivity) may come as a surprise to those who base their intuitions on textbook models. How can policies that raise the price of labor not always lead to fewer jobs? And how can policies that slow labor reallocation not inevitably have negative consequences for productivity? Three factors come into play in accounting for these results:

First, firms and workers can make different types of adjustments when faced with changes that raise the price of labor or alter the balance between flexibility and security. For example, employers can respond to higher minimum wages by cutting back on employment or moving some jobs “off the books,” but they can also take another route, such as substituting more skilled workers for less skilled ones, or adjusting other labor cost components. When reforms to job security rules reduce flexibility, employers can invest more in training or other productivity enhancing measures. They may reduce these investments when rule changes increase their options for a more flexible workforce. When minimum wages rise, low-wage jobs may decline in many situations, but in others the change can induce more workers into the labor market and, under certain conditions, lead to higher employment.

Second, the modest efficiency effects observed in developing countries must take into account the large informal sectors, which limit the coverage of regulations, and weak compliance where the rules do apply. In the case of minimum wages, most developing countries have non-compliance rates (i.e. the share of wage workers earning less than the statutory minimum) in the 20–50% range [9]. Certainly, weak enforcement of bad rules should not be recommended for regulating the labor market. However, while improvements in coverage and enforcement would alter the effects of labor regulations in developing countries, the findings are not that different in developed countries, where compliance is much higher [10].

Third, most countries appear to understand the risks of labor regulations at the strong and weak extremes, so they usually set rules in a range where many of the potential negative effects are avoided. Within this range most of the effects of labor market regulations are redistributive, while the effects enhancing efficiency roughly cancel out those undermining it. When countries operate in this range—called a “plateau” by the World Bank in its 2013 World Development Report on jobs [1] (Illustration)—labor market regulations can (partially) meet their intended goal of addressing labor market imperfections without serious consequences for efficiency.

While the empirical literature suggests that most countries operate on this plateau, this does not mean that policymakers never set regulations that lead to undesirable outcomes. That occurs when regulations are at the edges of the plateau—on the “cliffs,” in the World Bank's metaphor (Illustration). Most familiar to economists is the cliff where overly protective job security rules or too high minimum wages become an obstacle to employment or productivity growth. For example, several sub-Saharan countries have minimum wages that are higher than the average wage or the average value added per worker [11].

The other cliff may be less familiar but should be of equal concern for policymakers seeking good regulation. That cliff is characterized by minimum wages or job security rules that do not exist, are too lax, or are too weakly enforced. For example, 8% of the ILO's 186 member countries, mostly lower-income countries, have no minimum wage at all. Countries on this cliff do not address the labor market imperfections that motivate rule-setting in the first place, such as imperfect information or uneven bargaining power between employers and employees. The challenge, then, is to set regulations that mitigate these imperfections without falling off one cliff or the other.

Locating the “plateau” and the “cliffs”

What kinds of information and analysis are needed for policymakers to determine whether their regulations are on the plateau or on a cliff? There is no clear-cut answer but evidence can provide important inputs into decisions about labor market rule-setting that will avoid the under- or overregulation traps. For example, while most countries involve social partners in the setting of minimum wages, quantitative analysis that takes into account variables such as trends in GDP, productivity, employment, and informality can help ground decisions about minimum wage adjustments in the country's economic reality, or even be used as a formula for making adjustments.

There are informational and analytical challenges in determining the contours of the plateau. Other factors may be having an effect as well as the regulation under review. For example, if the introduction of more protective job security rules is followed by higher unemployment, that might not mean that the reform has had a negative impact if the economy has been slowing at the same time.

Another challenge is to assess the effects of a regulation over a long enough time. Studies too often focus on short-term effects of half a year to a year, when the real effects may take longer to emerge. Some research has shown that the longer-term effects of the minimum wage, both positive and negative, can be stronger than the contemporaneous ones [12]. Another challenge is to consider all of the important effects that rule-setters should be taking into account. Employment, wages, and productivity get most of the attention from analysts, while more difficult-to-measure outcomes such as economic insecurity and fairness are not assessed, even though they are clearly important.

These challenges can be addressed to some extent. Gathering information from a variety of sources can help in understanding the effects of labor market regulations. Household, firm, and administrative data can be used to track how outcomes of interest differ before and after regulatory reforms are introduced. In countries where regulations are determined by states or provinces, variations by jurisdiction can be exploited to identify effects. Different econometric techniques can be used to isolate the effects of regulatory changes by controlling for other factors that might influence the outcomes of interest.

Qualitative information can also provide useful insights. Although the contours of plateaus and cliffs are context-specific, benchmarking with similar countries can provide guidance on whether minimum wages or job security rules are within a reasonable range. Insights can also be gained from subjective perceptions of employers and workers. For example, enterprise surveys, such as the World Bank's Investment Climate Assessments, ask employers to assess how labor regulations (and other factors) affect their operations.

While these surveys are limited by their subjectivity, they do capture perceptions based on the actual effects of regulations. Comparative perceptions can be found through cross-country studies, such as the World Economic Forum's Global Competitiveness Report, which includes manager opinion surveys. The importance workers attach to job security and decent pay can be identified through national opinion surveys, and internationally comparable estimates are available through the World Values Survey and similar data-collection efforts.

The ultimate test of whether regulations are on the plateau is whether the costs of the rules are reasonable and whether reforms would have a substantial positive effect on the outcomes that regulators care about. Two questions can help evaluate this:

First, is there evidence that employers or workers are trying to overcome or bypass the regulations? One obvious indicator is the level of informal employment in the labor market. However, care must be taken in interpreting this finding. Higher than expected informality could be due to employers trying to avoid the costs of compliance, which would signify that the country is on the overregulation cliff. However, if the informality is the result of workers having little incentive to work in the formal sector because there is no advantage in wages or job security, then the country might be on the underregulation cliff.

Second, is there evidence that firms that are inherently more likely to be heavily affected by the regulations are experiencing worse outcomes than firms for which the rules are less relevant? As a case in point, when job security legislation is very protective, industries that are naturally more volatile, with high turnover, are more likely to grow slowly, while more stable industries might exhibit no significant ill effects.

Of course, valid empirical assessments of a country's labor market regulations depend on the availability of reliable and relevant data. Ideally, countries would possess regularly generated and representative data on workers and firms in both the formal and informal sectors. Panel data, which follow workers or firms over time, are especially useful in assessing the effects of regulatory changes. Having good data is only the first step. The data must also be accessible to analysts, and there must be channels for conveying the results of analysis to the policymaking process.