The high cost of minimum wage

The debates about the merits of increasing the minimum wage and of liberalizing trade have a key feature in common: there are two main dimensions for evaluating outcomes. Thus, arguments based on a policy measure’s effect on total incomes are met with counter-arguments about their effect on the distribution of income, and vice-versa.

The free trade debate has, I would say, reached some sort of consensus: expanded trade increases total incomes, but the resulting redistribution of income generates winners and losers. In rich countries, such as Canada, the gains are mainly concentrated among those at the top end of the income distribution. Thus, free trade accentuates income inequality.

Proponents of freer trade have come to accept that the increase in total incomes is not powerful enough of an argument to overwhelm concerns about inequality. The recent enthusiasm for “inclusive growth” is due to both an acknowledgment that increasing inequality is bad in itself, as well as a recognition that those who lose out from free trade are often in a position to block liberalization. If there are to be gains from trade, they will have to be broadly distributed.

The minimum wage debate is also conducted along these dimensions, but there is not yet a consensus. The argument for a higher minimum wage is the mirror image of the argument for freer trade: total incomes will fall, but this will be more than compensated by a decrease in income inequality.

The case for increasing the minimum wage has problems in both dimensions: the losses in total income are typically underestimated (when they are not being dismissed out of hand) and the putative reductions in income inequality are almost certainly being overstated.

Let’s examine total incomes first. Labour demand curves slope down: everything else being equal, higher wages reduce the quantity of labour employers demand. And fewer people with jobs means less total income. If the theoretical point is clear — and I’m not aware of a compelling theoretical argument suggesting that employers will react to higher minimum wage by hiring more workers — the empirical evidence is not. Estimates for the disemployment effects of minimum wage increases vary. Pierre Fortin of the Université du Québec à Montréal has, I think, a compelling explanation as to why.

Studies that have found weak disemployment effects — most famously the one published by David Card and Alan B. Krueger in 1994 — were generally performed using data in which the existing minimum wage was well below the average wage. (The U.S. federal minimum wage had been frozen during the Reagan-Bush years, and inflation had eroded its real purchasing power by the early 1990s.) On the other hand, stronger disemployment effects were obtained using data where the minimum wage was already 45 per cent to 50 per cent of the average wage or higher. Fortin’s conjecture is that when the minimum wage is already very low compared to the average wage, increases have little effect on employment. But if the minimum wage is already relatively high, then increasing it will produce stronger disemployment effects.

In the case of Alberta, it’s reasonable to expect that the disemployment effects of a minimum wage increase would be less than they might be elsewhere: Alberta still has the highest average wages in Canada. Even so, it’s still a mistake to leverage the more or less-uncontested point that small changes in the minimum wage will produce small losses in employment into the much stronger claim that large changes in the minimum wage will have no effect on employment.

But even if you take into account the disemployment effects, an increase in the minimum wage will redistribute income toward minimum wage workers. And a case can be made for exchanging a small reduction in total incomes for lower poverty rates and reduced income inequality. But here is the real problem: there is little evidence to support the claim that increasing the earnings of minimum wage workers will result in a progressive redistribution of income.

There have been several recent Canadian studies examining the relationship between the minimum wage and poverty. None find that an increase in the minimum wage significantly reduces poverty rates, and several find that an increase in the minimum wage actually increases poverty rates. The basic issue is that the poverty rates among minimum wage workers correspond almost exactly with the poverty rates in the general population: working at minimum wage is uncorrelated with being in poverty.

Another question that requires unpacking is: who is paying for the increased incomes of minimum wage workers? In the short term, the increased labour costs will likely be absorbed by the employers, but they will eventually be passed along in the form of reduced employment and higher prices.

There is as yet no Canadian evidence on who consumes goods and services produced by minimum wage workers, but a recent U.S. study by Stanford University’s Thomas MaCurdy finds the burden of the minimum wage increase is mainly borne by those with low incomes, and the benefits are redistributed more or less evenly across the income distribution. “These incometransfer outcomes sharply contradict portraying an increase in the minimum wage as an antipoverty initiative.”

Proponents of minimum wage increases have so far focused on downplaying their costs, but they have yet to demonstrate that the benefits outweigh those costs, or indeed that there are any benefits. If it can’t be shown that higher minimum wages reduce income inequality and poverty rates, then what is the point of increasing them?