Wage Distribution and Labor Market Discrimination

Labor income, also known as earnings or wages, is obtained by providing one’s labor services to the labor market. In contrast, non-labor income, or unearned income, refers to income received outside of the labor market and includes mostly returns from assets (wealth) such as rents, interest, and dividends, as well as transfer payments such as pensions, unemployment benefits, and social assistance.

The causes of disparities in income can be quite complex and multifaceted.Fundamentally, these disparities are the result of differences in individual characteristics, such as preferences, abilities, and human capital, as well as institutions that constrain and change the way individual skills are rewarded and act as regulators, such as in the case of minimum wages. Remember that individual endowments can only generate returns within a particular institutional structure. The labor market is regarded as the primary institution that rewards these abilities and individual differences.

We observe individual variances in skill formation due to three factors. One reason for variations in skill formation among individuals is that innate abilities may not be equally distributed among all people. This can be thought of as human capital capacity. While we have limited understanding of how this capacity is determined, we can assume that skill development continues throughout an individual’s lifetime. It is commonly believed that individuals with stronger abilities tend to invest more in acquiring skills, leading to an increase in their lifetime earnings over time. A second reason has to do with personal preferences and choices. People may choose to utilize their abilities in different ways, which can result in differences in skills among individuals. Individuals may choose to specialize in a high- earning profession or start working at an earlier age, which can affect their life-time earnings. A third reason is related to external factors that are beyond an individual’s control. These external factors can include things like supply and demand conditions in the market or economic shocks that can affect the development of skills and productivity. The conditions in the labor market, such as unemployment and the matching of relevant skills to jobs, also have an impact on how skills are developed and rewarded.

Workers with higher education tend to earn more and are found at the right end of the wage distribution, while those with lower education and young workers tend to earn less and are found on the left end. Individuals with intermediate education (secondary education) and in professions that require certain skills are concentrated around the median of the distribution. 

The distribution has several noteworthy characteristics, one of which is that there are more workers earning lower wages than those earning higher wages. This outcome leads to the median wage being lower than the average wage. This is a common observation in other countries as well: the distribution of wages and income is positively skewed. As the median wage moves away from the average wage, the distribution becomes skewed. There could be multiple reasons for this skewness in the wage distribution.

In the case of Türkiye, as seen in Figure 7.1, there are two concentrations in the wage distribution. The first concentration is on the left, mostly for those with low wages. This group is likely to include workers with lower levels of education and/or jobs that do not require qualifications. These wages are likely to be set with reference to the minimum, and the wage rate is not likely to be negotiated individually. The dashed line in the graph represents the minimum wage level, and it can be observed that the greatest concentration of wages is around this level. The smaller hump-shaped curve to the right of the distribution represents the distribution of public employees, whose wages are generally higher and concentrated at about 2 times the minimum wage. Like the minimum wages, public sector wages are determined through collective bargaining rather than market conditions or individual productivity. Therefore, the two regions where the wage distribution is most concentrated are the result of an institutional setting where collective bargaining predominates over market dynamics.

The differences in skills and productivity are a major factor in the accumulation of human capital over a lifetime.

Wages tend to level off or even stagnate slightly as individuals get older. This can be explained by the gradual decrease in the marginal return to human capital over time. As people age, their productivity may decline, and their skills and human capital may depreciate or even become obsolete. Workers reduce their human capital investments as they age since they may not have enough time to recoup their investment before retirement.

Manual jobs tend to have a less steep increase in earnings with age compared to cognitive jobs. This could be due to the fact that manual jobs offer fewer opportunities for skill development over time, and given that a significant portion of the labor force in developing countries is employed in manual jobs, this points to the importance of human capital in determining earnings.

We measure firm-specific experience as the number of years spent (or tenure) on the current job. As education increases, the amount of time spent within a firm also increases. One of the factors that is likely to cause a slowdown in human capital accumulation is job turnover, as discussed above. The transfer of skills and return on experience are more likely to occur when workers stay in the same profession and sector.

The scale used by the OECD is the square root of household size. We divide household income by the OECD scale to obtain what is referred to as the equivalised household income and use it in the calculation of inequality indices. If the monthly income of a four-person household composed of 2 children and 2 adults is 25,000 TL, the equivalised household income is 12,5000 TL and not 6,250 TL.

The Lorenz curve is a graphical representation of the distance between the actual distribution of income and absolute equality for various segments of the population.

The Gini coefficient is a statistical measure that quantifies the inequality of income distribution among households. A Gini coefficient of 1 represents absolute inequality, while a coefficient of 0 represents absolute equality.

In Türkiye, the top 10% earns 31.6% of the total income, while the bottom 10% receives only 2%. The performance of Türkiye is closer to that of Latin American countries than the more equal distributions seen in most European countries. For instance, the richest 10% in Brazil receive more than 40% of the total income, while the bottom 10% only receive about 1%. Many Latin American countries exhibit a similarly unequal distribution, with the bottom 10% getting a very small portion of total income.

There are two main ways to reduce income inequality through policy: taxation and redistribution. Taxation can be used to decrease income disparities by applying different tax rates to different income groups. A progressive tax system, where tax rates increase with income, can slow the accumulation of wealth in high-income groups and provide more resources for redistribution. Redistribution policies, like transfers, can also be used to decrease income differences. Distributing transfers inversely proportional to income can help shrink the income gap. Countries that effectively implement these policies tend to have a more equal distribution of income.

In many countries, wage negotiations are based on hourly rates, and hourly wages are the most widely used definition for measuring inequalities. In countries where wages are negotiated on a monthly basis, there can be difficulty determining which definition of working hours to use, particularly when working hours are also a part of the negotiation. In Türkiye, the significant variation in working hours makes it challenging to study monthly wages as a labor market outcome. For example, two workers earning similar monthly wages but with different working hours will be in the same rank (percentile) based on monthly wages, but one will have to work harder than the other to get the same wage. Since working hours are a characteristic of jobs in general, it can be said that hourly wages better reflect existing labor market inequalities.

ILO defines the minimum wage as follows: “Minimum wages have been defined as the minimum amount of remuneration that an employer is required to pay wage earners for the work performed during a given period, which cannot be reduced by a collective agreement or an individual contract.”

In Türkiye, the implementation of a minimum wage began in 1951, and it is defined as the lowest wage level that allows workers to meet their basic needs, which is similar to the general definition of a living wage. The current process of determining the minimum wage is carried out by the Minimum Wage Determination Commission in accordance with ILO conventions. The national minimum wage is set by the government following tripartite consultations, which include participation from the government, trade unions, and employer associations.

One limitation is that the Z variable, which represents group membership, may also indicate a productivity factor. Additionally, the coefficient β may not represent the same return for different groups, even if they have similar characteristics. Another limitation is that the formation of human capital may vary between groups due to discrimination. For example, in a less developed region, the school enrollment rate of women may be lower due to limited employment opportunities and low human capital returns, which can result in women having less human capital and lower expectations in the labor market. In addition to the limitations noted above, if discrimination not only occurs in the labor market but also affects workers’ access to human capital, this creates a correlation between X and Z in the estimation equation, leading to a biased estimate of α. This highlights the importance of considering the impact of discrimination not only on wages but also on the formation of human capital, and the need for more complex and nuanced approaches to understand and address wage disparities between groups. This may include additional control variables or different estimation techniques to control for the correlation between X and Z and produce unbiased estimates.

Using group averages to determine discrimination can be misleading in various ways. Factors that influence wages, such as human capital, experience, and region of residence, may not be evenly distributed among different groups. For instance, the low wages of migrant workers may not solely be a result of discrimination. There may be other factors at play such as low
human capital or language barriers that prevent them from accessing higher- skilled jobs despite having the necessary skills. Another example can be given in the context of gender. Since women’s labor force participation increases
with education, the average human capital of working women will be higher than men’s, which will increase women’s average wages. Another factor to consider is segregation in the labor market. Even if the average wages appear to be equal, individuals may face barriers to accessing certain job opportunities. Discrimination may manifest in recruitment through matching with similar group members and in-group bias. It’s important to consider segregation in recruitment to better understand discrimination. Finally, expectations and beliefs in the labor market can also play a role in creating wage differences. Groups may limit themselves to certain occupations or skills based on the feedback they get from the labor market, which can be hard to observe in a model. Even though the difference in average wages may be attributed to human capital or other observable factors, underlying expectations and beliefs may also be contributing to this difference.