Minimum wages are a bit like minimum speed limits. For a while, they can seem not to matter too much. Then all of a sudden they start to bite.
State Highway 73 runs from Christchurch over to the West Coast. The first part of it is flat and straight. Suppose we’re all driving along to the West Coast. But we’re a bit annoyed at a few slow drivers. So we put up a minimum speed limit – say at Kirwee where the road is still nice and flat. 50 kph minimum speed for cars with 2 litre engines, no minimum speed for cars with smaller engines. No problem. The slow cars just speed up a bit. Then we increase the minimum limit, say around Darfield: 70 kph for the powerful cars, 50 kph for the ones with smaller engines. Still nobody notices. So we up the minimum speed to 75 kph for all cars regardless of engine size at Springfield. Things are fine for a bit – a couple of scooters had to pull over and the smallest cars have turned off their air conditioners. Then we hit the Southern Alps and even reasonable cars have to work hard to get up the hills. The smaller cars can’t keep up with that minimum speed limit and have to pull over.
That’s what happened mid-2008 after Labour eliminated the differential youth minimum wage. The slow increases in the youth minimum wage in the period up to 2008 hadn’t seemed to have had much effect. Dean Hyslop and Steven Stillman’s rather nice 2004 paper for the New Zealand Treasury showed no effect of those earlier increases, but during a period in which employers were starving for workers of any sort. And the mountains still loomed.
Starting mid-2008, youth unemployment rates began skyrocketing relative to the adult unemployment rate. We were in the midst of a pretty serious global recession. But relative to the adult unemployment rate – a decent measure of the overall state of the job market – the youth unemployment rate jumped far more than it had in prior recessions. Using a very simple regression in which the youth unemployment rate is taken as a function of the adult rate, the regression residual – the portion of youth unemployment that is not explained by movements in the adult rate – went through the roof soon after the elimination of the youth minimum wage.
The graph below shows that relationship. If the adult unemployment rate perfectly predicted the youth unemployment rate, the blue line would be flat at zero. When it’s below zero, the youth unemployment rate is less than we might have thought given the adult rate; when it’s above, the youth unemployment rate is worse than expected. By December quarter 2009, the youth unemployment rate was more than seven percentage points higher than would have been expected given the previous relationship between adult and youth unemployment rates. That’s around 12,000 kids out of work who wouldn’t have been had the youth unemployment rate followed the same trend relative to the adult rate that it had in prior recessions. As of September quarter 2010, the excess residual of just under six percentage points suggests about 8,500 youths who would otherwise have had jobs.
New Zealand’s peculiarly high youth unemployment rate was confirmed in data released by the OECD in December. Our overall unemployment rate is low, but our youth unemployment rate is a higher multiple of the adult rate than that of any OECD country other than Sweden and Luxembourg. While Labour was quick to blame changes in youth training under National, the far more plausible culprit is Labour’s elimination of the lower youth minimum wage and National’s refusal to reverse the changes.
Like the minimum speed limit, the minimum wage really only starts having noticeably pernicious effects when the road gets a bit rough. Otherwise, it’s not a binding constraint. But even when it does start to bind, it can be reasonable to argue that the unemployment costs are low relative to the income benefits for those who do keep their jobs. If relatively few people lose their jobs, and those who do are mostly kids rather than household breadwinners, the policy might still be a winner. Unfortunately, that take seems not to mesh well with the empirics. David Neumark and Olena Nizalova found that the longer an inner city American black youth was exposed to a higher minimum wage as a teenager, as compared to a counterpart in another state with a lower minimum wage, the lower would be that youth’s earnings even into their late twenties. These persistent effects suggest that kids who fail to get jobs because of high minimum wages always stay a few steps behind their counterparts who gained early experience. And it’s not hard to guess whether it’s disadvantaged or comfortably middle class kids who bear that burden. In Canadian data, Aninda Sen found that higher minimum wages increased the proportion of families in poverty. Many families just over the poverty line, with second or third earners on the minimum wage, were pushed into poverty when those additional earners lost their jobs. The gains to those who kept their jobs were small relative to the losses for families losing critical supplemental incomes.
The analysis of the New Zealand data presented above remains somewhat back-of-the-envelope. But what other candidate explanations have we for not just an increase in the youth unemployment rate, but an increase so disproportionate relative to the adult unemployment rate and relative to the relationship between the youth and adult unemployment rate in prior recessions?
It was disappointing but not surprising that National killed Sir Roger Douglas’s bill that would have allowed the youth minimum wage to diverge from the adult rate. Few of the kids currently unemployed would have credited a National change in policy with their employment, but both they and their parents might have blamed the policy for lower wages. It might have been the right political decision. But it sure wasn’t the right decision for those kids likely to suffer a lifetime of lower earnings as consequence.
1. Hyslop and Stillman:
2. OECD and youth unemployment: Neumark and Nizalova linked here: http://offsettingbehaviour.blogspot.com/2010/01/minimum-wage-empirics.html
3. Sen: http://offsettingbehaviour.blogspot.com/2011/01/minimum-wages-and-poverty.html
4. The method by which I came up with the NZ numbers is described here: http://offsettingbehaviour.blogspot.com/2010/11/youth-unemployment.html