Economist Brian Easton has a crunch of minimum wages facts and thinking

This is a re-post of an article originally published on pundit.co.nz. It is here with permission.

An introductory economics course student is likely to meet the below graph. It shows that if the minimum wage (in purple) is above the point where the labour supply curve (in red) and labour demand curve (in blue) cross there will be less employment. That is what students are taught. But research shows the world does not work that way.

Early in an introductory economics course, a student will meet an example of what happens when there is an intervention in a market. For instance, what will happen if the government introduces a minimum wage and what happens if it is increased? The short answer from the basic model is that there will be unemployment or that unemployment will increase.

A good teacher – many are not, I’m afraid – will warn that there are assumptions on which the model rests, although even he or she will not have time to go through all of them.

Even so, there is a strong lobby for increasing the minimum wage. As a result, economists go about estimating how much additional unemployment an increase will generate. After all, if a 10 percent increase in the minimum wage were to increase unemployment by 100 workers, one might favour a higher wage; on the other hand, if the 10 percent increase doubled unemployment, then one might be reluctant to support an increase.

The size of the estimate is a scientific question, but unlike much of science, economics cannot usually do experiments. The econometrics will be dependent upon those confounded assumptions for which it cannot control.

In 1992, the state of New Jersey increased its minimum wage from $US4.25 to $5.05 per hour, but there was no change in the adjacent state of Pennsylvania. Comparing the two was a ‘natural’ experiment which would enable econometricians to calculate how much extra unemployment was generated by a higher minimum wage.

When David Card and Alan Krueger analysed the data they found that there was ‘no evidence that the rise in New Jersey's minimum wage reduced employment at fast-food restaurants in the state. ... we find that the increase in the minimum wage increased employment’.

They concluded that ‘taken as a whole, these findings are difficult to explain with the standard competitive model’. Bang goes the story in the introductory course.

Yet it continues to be taught, long after the evidence pointed in the opposite way. In that time, Card went on to analyse many more natural and quasi-natural experiments, producing many insights into the economics of immigration, education, job training and inequality. His achievements have been recognised by his sharing the 2021 Nobel Memorial Prize in Economic Sciences ‘for his empirical contributions to labour economics’.

(The other half went to Joshua Angrist and Guido Imbens ‘for their methodological contributions to the analysis of causal relationships’ – best known in the area of the economics of education. Perhaps not incidentally, Angrist’s doctoral supervisor was Card and he supervised Esther Duflo who was a 2019 Nobel Laureate. Krueger was not eligible because he died in 2019.)

Natural and quasi-natural experiments do not often occur, but Card et al. have raised the standard of innovation, sophistication and ingenuity in empirical economics.

Among Card’s other doctoral students was New Zealand economist Dean Hyslop who with his Motu colleague David Maré has just published a paper reviewing what we know about the minimum wage in New Zealand. (Hyslop has also published jointly with Card, including an award-winning paper.) Their findings are considerably more sophisticated than those of introductory economics.

Their main conclusion is that minimum wages over the past two decades have become a de facto teenage wage-setting policy. Some 57 percent of 16-to-17-year-old employees, and 43 percent of 18-to-19-year-old employees are paid at or below the minimum wage. Yes, there are minimum wage earners among older adults but they make up only a small proportion of them – 5 percent of 25-to-64-year-olds. (This suggests that while most of the young start off on, or near, a minimum wage their relative pay rates go up as they get older.)

In the last 20 years, the adult minimum wage has increased 75% in CPI-adjusted real terms and, following the abolition of the youth minimum wage, the real minimum wage for 16 to 19-year-old workers has increased 125%. (Median wages increased 35 percent in the period.) Yet the researchers found ‘no clear evidence that increases in the minimum wage have led to adverse employment effects’, although they wonder whether there have been ‘downside risks for youth and low skilled workers’ employment’.

Where do we go from here? The paper opens with a dedication to Jas McKenzie, who was one of the researchers’ mentors. It goes on, ‘the last time we saw Jas, he expressed the view minimum wages have become New Zealand’s de facto wages policy’. Unfortunately Jas died last year so we will have to work out the implications without his help.

A couple of things struck me about Maré and Hyslop’s findings. While the higher minimum wage pushed up the wages immediately above it, it did not push up the whole wage structure. This is evident from the median wage not rising as fast as the minimum wage. The implicit assumption in the first-year-course illustration is that a rise in the minimum wage is passed on throughout the wage structure. It isn’t.

A second finding was that minimum wage workers are broadly spread across the entire household income distribution which suggests that rises in minimum wages will not be very effective as a redistributive income support policy. That is consistent with my findings when I was exploring these issues many decades ago, although there were some important differences in the labour force. (Mothers were then just sneaking into the paid workforce.)

The second finding will not stop many advocating higher minimum wages because they think it will reduce inequality. Scattered throughout the political spectrum, there are the ideologues impervious to research findings. Some will seize those Card (and Krueger) findings which suit them and ignore others which do not. Drunks use lampposts in a similar manner – to provide support rather than shed light.

To do so, is to lose the richness and nuances in economics that are evolving from new data bases and the technical innovations like those Card have developed. It is to treat a first-year course in economics in a way one would never do if the first-year course was in science, say. It is the beginning of the journey, not the end.

Brian Easton, an independent scholar, is an economist, social statistician, public policy analyst and historian. He was the Listener economic columnist from 1978 to 2014. This is a re-post of an article originally published on pundit.co.nz. It is here with permission.