The new federal government, when it was an opposition party, set great store by the phrase “evidence-based policy,” implying that the previous federal government just made things up, which in my view it generally didn’t. As a university professor, I’m not really going to argue against evidence-based policy (though I have: at times “values-based policy” can be better). But sometimes—maybe even oftentimes—evidence seems to count very little in deciding whether a policy is smart or not. Take the minimum wage (please!).

A new Fraser Institute study on the minimum wage by Robert P. Murphy, Charles Lammam and Hugh MacIntyre provides an evidentiary base for great skepticism about whether the minimum wage achieves its presumed anti-poverty goals.

Much of the economic debate around the minimum wage over the last quarter century has concerned its effects on unemployment. Several famous case studies in the United States in the 1990s seemed to show there weren’t big reductions in unemployment or even any reductions in unemployment—maybe even increases in employment—following hikes in minimum wages in New Jersey, Texas and California. The new study’s authors spend a chapter going over the evidence, concluding that many other studies, including several recent ones, find effects more consistent with the “law of demand,” namely, that if you raise something’s price (in this case, unskilled labour) the demand for it will fall.

But whether a policy has harmful side effects or not isn’t actually the first piece of evidence you want about it. The first piece of evidence you want is that it stands a chance of doing what it’s supposed to, which in the case of the minimum wage presumably is to help out poor people. In this connection, my favourite part of the study is “Table A1: Demographics of Minimum Wage Earners in Canada, 2014,” found on page 40. This table, based on special compilations from Statistics Canada (see the study, page 39), tells us who Canada’s minimum wage-earners are and, more importantly, aren’t.

Go to the very bottom line. Reading from left to right, with the relevant column entries in bold, what the table says is that 700,000 Canadian workers were members of a low-income household in 2012, using Statistics Canada’s Low Income Cut-off as a bar. They represented 4.9 per cent of Canada’s 15.1 million employees that year. Only 116,000 of them earned minimum wage, however. Those 116,000 were just 12.5 per cent of the 1.1 million Canadians who earned the minimum wage that year. And they were just 16.6 per cent of the 700,000 low-income workers. It seems the minimum wage has very little to do with low income.

How is that? Rows higher up in the table tell us 36.4 per cent of those earning the minimum wage were 15 to 19 years old, while another 22 per cent were 20 to 24. Moreover, 58.1 per cent were part-time rather than full-time workers. Essentially the same per cent (56.8) were a son, daughter or other relative living in a family.

In sum, not all minimum wage workers are poor (far from it), while not all poor people are working for minimum wage (far from that, too, in fact). If your goal in raising the minimum wage is to reduce poverty, well, the minimum wage just doesn’t seem to have a lot to do with poverty.

Which maybe gets us back to the notion of values-based policy.

More and more you hear people saying we need to raise the minimum wage because it’s just not right in a country as rich as Canada that people should work for less than $X per hour, filling in for X whatever their idea of a dignified wage is. “How can someone raise a family on $X an hour?” they invariably say. Well, as the table shows, hardly anybody is trying to do that. Only 5.7 per cent of those earning minimum wage are the head of a household with no spouse while, coincidentally, only another 5.7 per cent are single and living alone.

Re-phrase the question: What’s a dignified wage for teenagers and those only slightly older to get the basic job experience that will allow them to move up the wage ladder? Seems to me almost anything above zero is more than dignified. In fact, given the very valuable lessons imparted to first-time workers, maybe even the zero lower bound needn’t stop us.