Labor markets in poor countries differ fundamentally from those in richer ones. A central distinguishing feature consists in their very low levels of wage employment, and high self-employment. For instance, in countries with GDP per capita below $5,000 (in 2017 int. $), the average urban self-employment rate is 45%, which exceeds the average wage employment rate of only 43%. The discrepancy is even larger in rural areas. In the United States, in contrast, wage and salary workers account for about 85% of the labor force, and own-account workers make up only about 5% of employment.
Because of these stark differences, the creation of wage jobs has been identified as a key development challenge, and the wage employment rate is part of the United Nations Millennium Development Goals. But why is wage employment so low, and self-employment so high, in developing countries?
Most existing work on the topic has focused on factors related to self-employment, like the ease of entry or lack of regulation. In a new IZA discussion paper, Markus Poschke proposes a different channel: the study links the high self-employment rate in poor countries to the functioning of labor markets. This connection is motivated by two new facts shown in the study: (1) Although the unemployment rate is not higher in poor countries, this is entirely due to their high levels of self-employment. Relative to wage employment, unemployment is significantly higher in poor countries, implying that wage jobs are either harder to find or likely to last less long. (2) On top of this, self-employment is generally higher in countries with high unemployment relative to wage employment.
Self-employment as the lesser evil?
These patterns suggest a simple mechanism: when wage employment jobs are less attractive or harder to find, the unemployed may opt for self-employment instead – even if self-employment per se is not very attractive either.
Because a direct empirical investigation of this mechanism is elusive, the study evaluates its importance by developing a new model and combining it with new data on labor market flows across countries. This new data set, which has just recently been assembled by Kevin Donovan, Will Lu and Todd Schoellman, comprehensively reports how many individuals move between wage employment, unemployment and self-employment every quarter in 37 countries spanning a wide range of income per capita. The countries covered include Brazil, Mexico, France, South Africa and the United States.
The new model builds on canonical models of “search” in the labor market, which analyze individual choices about labor market states, like job search, job to job moves etc., in response to fundamental characteristics of the economy, like the volatility of jobs and the difficulty of finding them. It extends existing models by also considering self-employment. As a result, the new model can account for all movements among wage employment, unemployment and self-employment in the 37 countries in the data set. Integrating self-employment into the model is essential for understanding labor markets in poor countries, where self-employment is much more prevalent than unemployment.
Frequent labor market shocks
The unemployed in the model face a choice: should they opt for job search or become self-employed? How many of them enter self-employment naturally depends on the characteristics of each economy. Once the model has been estimated for 37 countries, it turns out that a particularly important such characteristic are shocks to firms. Where shocks are more frequent, there is a greater risk that a job will end. This is why such shocks are an important determinant of the unemployment rate, as is well known from the existing literature.
The new study finds an important additional effect: on average, greater shocks raise the self-employment rate just as much as the unemployment rate, since they make job search less attractive compared to self-employment. This effect is particularly strong in countries where self-employment is already high for other reasons, like lower costs of entry, or lower regulation or taxation of the self-employed compared to employers.
In a comparative analysis, the study also sheds light on factors contributing to differences in unemployment and self-employment rates across countries. Countries with high self-employment and high unemployment relative to wage employment (like many poor countries) mostly exhibit cheap self-employment entry, frequent shocks to jobs, and greater difficulty in determining whether a worker and a job are a good fit (screening).
Negative effect on output
This analysis thus shows that the forces governing job creation and destruction, like the difficulty of hiring and shocks to jobs, also affect self-employment – and may even affect it more strongly than the unemployment rate. In addition, they affect overall income in an economy: where jobs are less attractive, individuals are willing to pursue self-employment opportunities that would not be worthwhile elsewhere. As a consequence, output declines by much more than indicated by the change in unemployment alone.
The study highlights the interdependence of different labor market states for understanding any of them. In particular, it is impossible to understand high rates of self-employment in poor countries in isolation, since they are an outcome of the broader situation in the labor market.