Minimum wages and jobs: The overseas story

Does increasing the minimum wage lead to job losses? Will an increase in pay for those workers covered by the current National Wage Case see their colleagues losing jobs at some stage down the track?

The various employer lobby groups certainly think so. And for a long time conservative politicians have been fond of the mantra: “one man’s [sic] pay rise is another man’s job”. Neo-classical economists have also promoted this logic as a kind of conventional wisdom, arguing that setting the minimum wage at a level above the “market-clearing wage” will see a drop in labour demand and a subsequent fall in employment.

Of course, there are economic models to “prove” this, but the process is somewhat circular because the models are based on the immutable laws of supply and demand and the logic is ineluctable: if the price rises, demand falls. If we turn to the real world, we can ask: is there empirical evident for this conventional wisdom that job losses follow from pay rises for the low wage workforce? Unfortunately, there is little useful research in Australia which directly addresses this question, but there are large numbers of studies done overseas which shed light on the issue.

Before we look at some of this evidence, a couple of qualifications are needed. First, the impact of an increase in the minimum wage will have only a small impact on the average wage, since most workers already receive wages higher than the minimum wage. Secondly, the labour market for minimum wage workers is essentially one among many labour markets. In the minimum wage labour market, which is basically a market for low skilled workers, the substitution of labour is common. This means that the impact of relative wages matters. In other words, if you increase the minimum wage for one group of low skilled workers, employers might turn to another group of low skilled workers and employ them instead. If this happens, the net employment effect might be negligible, but the adverse effect on a particular sub-group of workers might be more serious.

We’ve have now had several decades of empirical research looking at the wages and employment connection. Some of the earlier economic studies were a bit dubious. They often repeated the same economic modelling exercise against similar data. For example, an overview on the impact of minimum wages in the early 1980s concluded that a 10 per cent increase in the minimum wage would have an adverse impact on employment of between 1 and 3 per cent. The authors of this study admitted, however, that their conclusion was not based on a large body of research because “one could argue that there really are not 25 independent studies”. Their use of the same labour force data and the similarity of their methodologies meant that most of the studies simply replicated each other.

More sophisticated studies arrived during the late 1980s, as researchers began to make use of longitudinal data, that is, studies of the same subject over time. There were detailed studies of firms, which allowed researchers to examine the impact of increases in minimum wages on the employment patterns within those firms.

One of the most celebrated, or notorious (depending on your point of view) of these was a study by two American economists, David Card and Alan Krueger, into the fast-food industry. They compared the impact of a rise in the minimum wage on employment in fast-food restaurants in New Jersey with similar restaurants in Pennsylvania where no such wage increase had occurred. The firms affected by the wage increase were regarded as a “treatment group” and the firms where no wage increases occurred were regarded as a “control group”. A comparison between the employment outcomes for the two groups resembled an experiment and its results have had a major impact in the field for the last decade and a half.

What made Card and Krueger’s research so important was its controversial findings alongside its methodological rigour. It reversed the conventional wisdom by showing that increasing statutory minimum wages had no negative effects on employment, and it did so in a way which survived critical scrutiny. Subsequent criticism was answered by means of a reanalysis of the New Jersey wage increase using payroll data which basically confirmed Card and Krueger’s earlier survey-based findings.

It may be the case, of course, that increasing the minimum wage has less impact on jobs in the United States because the minimum wage is so low in that country. It might be argued that in countries where the ratio of minimum wages to average wages is much higher - such as in Europe and Australia - the negative effect on jobs of increasing minimum wages will be much more obvious. So what does the research from Europe show?

Not surprisingly, recent European research has produced contentious results. Several studies of the impact of the minimum wage in France, the Netherlands, Spain and the United Kingdom failed to find strong evidence showing that increasing minimum wages invariably led to employment losses. This led one researcher to comment:

Our study of the experience of minimum wages in four European countries finds very little evidence of important employment effects associated with minimum wages. Effects are typically small and in some cases go the opposite way in terms of predictions of the orthodox model of the labour market. On the whole our results seem in line with the recent US work that fails to find any evidence of job loss associated with minimum wages. The emphasis seems to have shifted from “how negative are the employment effects?” to “is there an employment effect?” and, in some circles, “what potential is there for a positive effect?”

However, at the same time, other research on French minimum wages did find a stronger relationship between increases in minimum wages and employment losses, particularly for young workers. The OECD attempted to summarise this literature by looking at over 20 major studies carried out during the 1990s which examined the impact of the minimum wage on employment. The OECD found that the results of these studies varied considerably, and there were no clear, unambiguous findings. This led them to conclude tentatively:

Young workers may be most vulnerable to job losses at a high level of the minimum wage. There is less evidence available on the employment effects, if any, for other groups such as women and part-time workers.

Turning to our own neighbourhood, one of the most interesting studies in this field has been done by New Zealand researchers who examined very large increases in youth wages in the period following 2001 and found evidence of positive employment responses. They studied changes to the minimum wage system which had led to wage increases for 18 and 19 year-olds in the order of 69 per cent, and increases for 16 and 17 year-olds of about 41 per cent. It is worth noting how these increases compare with the wage increases studied by Card and Krueger: they were only in the range of 19 to 27 per cent. Clearly, if there was to be an adverse impact on employment, then wage increases of this size should have made it evident. Instead, the New Zealand researchers found:

Positive employment responses to the changes for both groups of teenagers, and that 16-17 year-olds increased their hours worked by 10-15 per cent following the minimum wage changes.

In summary, we could say that the research on the issue of wages and jobs remains strongly divided, but the conventional wisdom is no longer conventional. It is now very much an open question whether pay increases for low wage workers really do lead to job losses down the track. There are some sub-groups, such as young people, for whom the evidence raises concerns. But when it comes to the low wage workforce as a whole, there is no convincing body of contemporary evidence that the conventional wisdom applies.

One of the problems for researchers who are wedded to outmoded economic theories is their pursuit of a generally acceptable and universally applicable answer to this question of wages and jobs. Such a goal is elusive, if not utopian. There are so many unique factors in a local, or even national economies, that universally applicable findings are almost impossible to obtain. The gradualist logic of the Australian Industrial Relations Commission seems to be the most sensible approach. Instead of relying on the casuistry of economic theory, these industrial judges have examined the actual consequences of each National Wage case in the context of the Australian economy, and then acted accordingly.