ECONOMISTS ADDING COVID VALUE.

I wrote last week how people might be asking themselves “What bloody use are economists?” after the Bank of Canada declined to provide numerical forecasts for the next few months but contented itself with qualitative statements about the current crisis. It’s not unreasonable justification was that unique historical events are very hard to forecast — a proposition I demonstrated by implying the new Bank of Canada governor would be a woman. He isn’t. He is Tiff Macklem, dean of business at the University of Toronto and a Bank of Canada and Department of Finance veteran. Good luck to him as he steps aboard the raging bear that is our current economy. I’m a big believer in meritocracy but it does seem strange that in the 85 years now of the bank’s existence no francophone or female has headed it — even stranger, as a simple matter of prediction, that the current government didn’t take a “Because it’s 2020!” approach.

Though we economists may not be great at forecasting, the latest weekly email from the National Bureau of Economic Research suggests we may yet have our uses. Everybody is working on COVID-19 these days, economists included. And big data is proving its worth.

Three U.S.-based economists (at Texas, California Berkeley and Chicago: thank goodness for the internet) use the “Nielsen Homescan” regular survey of Americans to double-check the official employment data. They estimate 20 million jobs lost by April 6, “far more than jobs lost over the entire Great Recession.” On the other hand, the rise in the unemployment rate is only about two percentage points. Why? People have quit the labour force. The decline in the labour force participation rate is seven percentage points, “an unparalleled fall that dwarfs the three percentage point cumulative decline that occurred from 2008 to 2016,” and which has been a really big deal in the labour market analysis of the last decade. Fingers crossed it’s temporary.

Columbia University’s Daniel Hamermesh, one of the best-known U.S. labour economists, uses data from the 2012-13 American Time Use Survey to forecast how people

THE DECLINE IN THE LABOUR FORCE PARTICIPATION RATE IS SEVEN PERCENTAGE POINTS, ‘AN UNPARALLELED FALL ...’

are doing in lockdown. In that survey, married couples reported their satisfaction rose most as a result of time spent with their better half — though presumably there can be too much of a good thing. Singles, on the other hand, felt worse the more they were on their own. Hamermesh then runs simulations suggesting that, though singles are unambiguously worse off, couples may actually be better off as a result of the lockdown. Everyone’s well-being is hit, however, if job and income losses are big enough.

Three other Columbia economists look at the distribution of COVID-19 testing in New York City. Using zip code data on both tests and incomes, they find testing is almost perfectly egalitarian. “The 10 per cent of the city’s population living in the richest zip codes received 11 per cent of the COVID-19 tests and 29 per cent of the city’s income.” The bottom 10 per cent by income got 10 per cent of the tests. On the other hand, the chance of testing negative is much higher in the richest zip codes: 65 per cent versus 38 per cent in the poorest.

MIT’s Jeffrey Harris argues that “New York City’s multi-tentacled subway system was a major disseminator — if not the principal transmission vehicle — of coronavirus infection during the initial takeoff” in March. (This is research published in April, mind you.) “Maps of subway station turnstile entries, superimposed upon zip code-level maps of reported coronavirus incidence, are strongly consistent with subway-facilitated disease propagation,” with local lines more to blame than express lines. Canadian cities with subways will want to take note.

Two econometricians from Northwestern and Cornell take on the problem of inferring the true COVID-19 infection rate, given that testing has been neither comprehensive nor random and tests are not completely accurate. Using data through April 6, they conclude infection rates are higher than reported and the “infection fatality rate” — deaths among people who become infected — is substantially lower than the “case mortality rate,” deaths among those testing positive. But “given that the tested fraction of the population has been very low, one can barely draw any conclusion about the population infection rate without making assumptions that bound the rate of infection in the untested subpopulation. … As has been widely recognized, random testing of populations would contribute enormously.” Everyone agrees. Money moved from income support to testing, as well as expanding bed capacity, would surely make sense.

Finally, taking to heart that economists’ real job is to investigate trade-offs, four economists from California San Diego, San Diego State, Colorado and Bentley University in Massachusetts calculate, using many assumptions, that California’s lockdown reduced deaths by as much as 1,661 in its first month and that each life saved cost at least 400 jobs.

Does that mean it was worth it? For that you need a priest, rabbi, philosopher or your own internal moral guidance system. We economists have our own views on such things but they are just that, our own views.