Wage subsidy has no empirical basis for SA

In its document “Confronting Youth Unemployment: policy options for South Africa”, the National Treasury proposes the youth employment subsidy as a means to address youth unemployment.

The document argues that “the magnitude of the employment challenge in South Africa, particularly among its youth, suggests that an employment subsidy is an appropriate policy intervention to boost labour demand”.

DA supporters take part in a march on the headquarters of Cosatu on March 15, which descended into chaos. The writer argues that it is young people who should get the subsidies, not employers. Picture: Reuters. Credit: Reuters

One point that the document emphasises throughout, as a motivation for the subsidy, is that “the gap between productivity and real wages for young workers is an important constraint to job creation… a gap between real wages and productivity undermines competitiveness, discourages businesses from hiring workers and pushes unemployment higher”.

The Treasury document therefore motivates for the youth wage subsidy on the ground that it “aims to narrow the gap between entry-level real wages and productivity for young people”.

Interestingly, nowhere does the Treasury document compute this alleged gap between real wages and productivity for young people. There is no empirical evidence to show that this gap exists. In addition, the value of the subsidy that is proposed is not linked to this alleged gap.

Logic would dictate that since the Treasury maintains that entry-level wages are higher than productivity, the subsidy amount to firms should be calibrated so as to close this gap.

However, instead of following this scientific route, the document adopts an example which “assumes a maximum subsidy value of R12 000. This is approximately half of the average income of a formal-sector worker aged 18 to 29.” Nowhere does the document justify this assumption, which implies that young workers who earn less than R24 000 are generally 50 percent more expensive for employers, or their productivity is half of what is required to justify their current real wage.

In this critique, we address this fundamental point.

We show that the alleged gap between the real wage earned by young workers and their productivity is the opposite.

Young people are actually paid less than their productivity.

From a neo-classical perspective young people in SA are exploited. This exploitation is greater for the younger segment of the youth. This result implies that government should grant a subsidy to young workers to augment their current wage, rather than grant it to employers. The government must tax employers and transfer the money to young workers in order to eliminate the gap between the productivity-based real wage and the actual wage that is earned by young people.

The basic concept that neo-classical economics, an analytical framework that is dominant but of which I am not an advocate, analyses productivity by using the so-called production function.

This function relates inputs, such as labour and capital, to output. Under perfectly competitive conditions, neo-classical theory says the real wage paid to the worker must equal the marginal productivity of labour. If the real wage exceeds marginal productivity, there will be unemployment.

This is what the Treasury means by “a large gap between real wages and productivity for young people”? Fortunately, the simplicity of neo-classical theory allows us to compute this gap easily. Neo-classical theory tells us that the marginal productivity of young labour is equal to the labour share of youth labour multiplied by the ratio of output to youth labour.

In the SA context workers aged between 15 and 34 are roughly 45 percent of the employed. In addition, despite its downward trend, the average labour share over the past 16 years is roughly 52 percent.

This labour share includes older workers and should therefore be modified to reflect that labour share of young workers.

Neo-classical theory tells us that we can perform this adjustment by multiplying the average labour share with the elasticity of output with respect to youth employment. According to the Treasury, this elasticity of output is estimated to be 0.83 but this is clearly wrong, because it overestimates the share of income that goes to young workers relative to older workers.

Instead, based on wages of the young relative to old workers and the proportion of young workers in total employment, we estimate the elasticity of output with respect to employment to be 0.37.

This translates into a youth labour share of 20 percent.

The SA Reserve Bank reports that the ratio of real output to labour was R88 127 in 2010. Applying the neo-classical formula we thus get the marginal product of youth labour to be R39 168 a year, which translates to R3 264 a month.

Notice that had we used the wrong Treasury estimate of the elasticity of output with respect to youth employment, the marginal productivity of young people would be R89 400 a year, which would translate to R7 450 a month.

Now that we have established the marginal productivity of a worker aged between 15 and 34, we turn to the computation of the real wage that is actually earned by such workers. We are aware that the Treasury document admits that young people who earn below the tax bracket earn on average a nominal R2 000 a month. But we do not know the split between those who are below the tax bracket and those who are not. Therefore we turn to the Monthly Earnings of South Africans Report (2010), since it is the only officially published source of data on youth employment earnings. Based on this report, I estimate that the average nominal monthly wage for a person aged between 15 and 34 is R3 680. Translating this nominal wage into a real wage, using the gross domestic product deflator for 2010, which was 1.45, we arrive at the real wage of R2 538.

What this analysis has shown is that the so-called “large gap between the real wage and productivity for young people” is actually the opposite.

Neo-classical economics tells us that, based on their productivity, young people are supposed to be paid on average R3 264 a month (but this would be R7 450 using the Treasury document assumptions!).

Consequently young workers are paid 78 percent of what they should be paid, and based on Treasury’s assumptions this would be 34 percent. This means that the state must grant a subsidy to young workers, rather to employers.

The situation is worse for young workers aged between 15 and 24, who are 10 percent of the total employed. By implementing the neo-classical calculation I find that these young workers are highly productive. Their marginal productivity is R2 787 per month. On the other hand their real wage averages R2 251 per month. This segment of young workers is therefore paid 81 percent of what they should be paid, or they are robbed on average of R536 monthly by their employers, neo-classically speaking.

I am aware that the subsidy targets young people between the ages of 18 and 29, whereas here I used the data for people aged 15 to 34.

The reason is that officially published data from Statistics SA on earnings does not break down the workforce in a way that is consistent with the Treasury target group. However, the fact that younger workers are slightly less exploited than older workers counterbalances the likely disparity of including workers from age 30 to 34. The point remains that young people are robbed of R726 of their earnings by employers monthly, neo-classically speaking.

Based on the logic of neo-classical economics, the youth wage subsidy that is given to employers has no empirical basis in SA.

Youth labour is extremely cheap and employers exercise buying power by paying young people well below their productivity. Young workers must therefore be given a subsidy, raised from taxing employers, so that they augment their current wage. Furthermore, since youth wages are below productivity, the argument that the high rate of youth unemployment can be explained by “real wages that are above productivity” falls flat.

n Malikane is an associate professor of economics at Wits University