As follow-up to a blog of April 2 on The Jobless Trajectory: State-by-State, E. D. Hovee has done further analysis to compare recent job layoffs with coronavirus related mortality — on a state-level basis. The following graphic is a plot comparing cumulative jobless claims filed during the weeks ending March 14, 21 and 28 with COVID-19 mortality as of March 29.
In the chart, cumulative unemployment claims filed over this three week period are calculated as a % of total covered employment, by state. Mortality is calculated in terms of deaths per million residents, by state.
At first glance, there would appear to be little evidence of any clear statistical relationship between coronavirus deaths and job layoff experience, by state. Even if there were to be a clear statistical relationship, no direct causality is asserted at this point. In other words, no assertion that high mortality necessarily causes a state to exact massive job shutdowns, and no assertion that job shutdowns affect mortality one way or the other.
However, there may yet be a story to tell. As suggested by the graph, four potential clusters of activity are noted:
First is the experience of the 6 high mortality states — New York, Louisiana, Washington, Vermont, New Jersey and Michigan — with mortality to date ranging between 10-50 deaths per million residents. While each has a slightly different story to tell, all but Vermont are states with relatively high population densities. Vermont is a low population state but in the outer orbit of the Boston and New York spheres of influence — as is obviously the case for New Jersey. Washington state got it early on (clustered in one facility) and Louisiana suffered in the aftermath of Mardi Gras. Michigan is in the early stage of a significant increase in virus cases in the Detroit area. And New York is both highly dense and internationally focused, both of which put the city and metro area at greater risk.
Second is the experience to date of three states experiencing high job loss of more than 12 to nearly 14% — Pennsylvania, Rhode Island, Nevada — albeit with minimal coronavirus mortality experience to date. One could argue that these states are bearing an unnecessarily high economic burden relative to virus-posed risk, so far.
Third is South Dakota clustered with 10 other states that have experienced less than 4% job loss and less than 10 deaths per million population. Several of these states, in addition to South Dakota — Arkansas, Colorado, Mississippi, Utah, West Virginia and Wyoming — have below average population densities. Four states — Connecticut, Florida, Georgia and Texas — are more densely populated but with the southeastern states politically resistant to early job shutdown.
Finally, there are another 30 states that have experienced job losses ranging from 4% to less than 10% but also with relatively low mortality rates. These are critical swing states that may be experiencing somewhat disproportionate job loss relative to reduced mortality — but perhaps investing now in shutdown to better stave off potentially harder pandemic-related times to come.
Currently, there is considerable debate as to whether more stringent national guidelines should be extended (or mandated) across all 50 states uniformly. The alternative is to continue providing state-by-state discretion to adapt to changing conditions when and if warranted.
This analysis, while preliminary, suggests that when it comes down to determining the right tradeoff between human and economic survival, the right choice is not necessarily “one size fits all.” Neither the circumstances nor the policy prescriptions appropriate for New York are necessarily well suited for South Dakota or Wyoming.
In the weeks (and perhaps months) ahead, the death toll will inevitably increase. So will the loss of gainful employment and business opportunity — some quickly recoverable and some not. The pivotal and ever more painful decision will be how to make the trade-off between potential lives lost versus impoverishment for many of the more numerous survivors. .
If there is to be a national mandate, let it be a prescription that sets certain threshold targets, which when violated, will become mandatory at the state level. Otherwise, let state-by-state discretion prevail. And let both the national and state-local level decisions be guided by both real-time heath and economic data — each in tension and balance against the other.
