Interpreting job market statistics demands a lot of care right now. The pandemic has muddied the statistical waters and created the illusion that unemployment rates are significantly higher in Canada than in other countries.
Statistics in the service of partisan politics are often, to put it gently, rather elastic in their meaning, so it is natural to wonder: do we really lead the pack in the dubious distinction of having the highest rate of unemployment?
In a medical sense, COVID-19, as highly contagious as it is, can be thought of as the great leveller. No one has immunity, and we face the health risk of this virus with a sense of our common humanity.
But in a socio-economic sense, it is not as contagious. The jobs some of us hold give us an economic immunity, and we face the economic risk of this virus with a very different sense of our interconnectedness.
Normally, I don’t venture into to predicting month-to-month changes in the unemployment rate, but this month is an exception for two reasons. The changes are certainly going to go well beyond the statistical noise inherent in the Statistics Canada survey, so there is no chance that the picture will be clouded. And history really isn’t a guide to what is coming next (in the very short term), so sophisticated models based on past data don’t have a particular advantage. My bets are on an almost doubling of the Canadian unemployment rate between February and March, with even this being an understatement because the official survey preceded some of the more dramatic shutdowns that happened later in the month. I’m suggesting that we are even probably close to 15% right now.
On Thursday, April 9th Statistics Canada will release the results of the Labour Force Survey for the month of March 2020. COVID19 makes this one of the most scrutinized releases in the 75 year history of the survey, reporting as it will on jobs and unemployment during the week of March 15th to March 21st. Here’s what you need to know, and what to look for.
Angus Deaton, the Princeton University economist, wrote in the opening paragraph of his acceptance speech for the 2015 Nobel Prize in economics that:
Measurement, even without understanding of mechanisms, can be of great importance in and of itself—policy change is frequently based on it—and is necessary if not sufficient for any reasoned assessment of policies, including the many that are advocated for the reduction of national or global poverty. We are wise to remember the importance of good data, and not to neglect the challenges that measurement continuously poses (Deaton 2016, page 1221).
This nicely sums up the tone of a previous post, that a conversation about Canadian public policy directed to poverty has not been well served by the confusing and conflicting information provided by official statistics.
Just how should we measure poverty in a way that is most helpful for public policy?
The two most commonly used poverty rates produced by Statistics Canada tell very different stories. The patterns are curious, and confusing. The two statistics—the poverty rate according to the Low Income Cut-off and that according to the Low Income Measure—track each other rather closely up to the early 1990s, then diverge quite markedly as the Low Income Cut-off falls steadily to an unprecedented low, while the Low Income Measure drifts upward. Which statistic should we believe?
The cyclical patterns also differ, with the Low Income Measure registering higher poverty during recessions only before the 1990s, and in a way that is more muted and lagging the movement in the Low Income Cut-off. It also signals a rise in poverty only well after the onset of the 1990/92 recession, and both measures show no upturn in poverty during the Great Recession, which began in 2008 and led to a significant fall in employment.
For something that is central to so many policy debates, the Canadian “poverty” rate is notoriously confusing, and it is easy to imagine that public policy may be misled. The first step in devising a poverty reduction strategy is understanding what these numbers mean, and whether they are useful. Is poverty at unprecedented lows, or has it been stuck at high levels for decades? Both views can’t be right, but they can both be wrong.