Following the initial claims numbers

admin | April 17, 2020

This document is intended for institutional investors and is not subject to all of the independence and disclosure standards applicable to debt research reports prepared for retail investors.

The initial unemployment claims figures over the past several weeks have shown a stunning flood of workers seeking jobless benefits.  There have been just over 20 million initial claims over the past four weeks, or 22 million seasonally adjusted. That works out to more than 13% of total payrolls as of February.  The state-by-state figures offer some insight into how the coronavirus economic slump is hitting various states and may provide a blueprint on where claims could go from here.

State-by-State Data

The weekly initial and continuing claims data are reported by state as well as on a national basis.  States have taken a variety of approaches to social distancing, as some were quick to implement sweeping lockdowns while others have permitted more economic activity to continue.  In addition, state labor offices have shown a varying ability to handle the unprecedented flood of applications.  There have been frequent media reports of people who, despite repeated attempts, could not get through on the phone or online to file an application.

Exhibit 1: Virus cases versus severity of layoffs

Source: Labor Department, Worldometers

A scatter plot showing cumulative initial claims over the past four weeks as a percentage of total payrolls as of February vs. the incidence of the virus offers a rough look at how the varying severity of the virus across the country is related to the magnitude of layoffs (Exhibit 1).  The outliers are informative.  For reference, the U.S. average is right in the middle of the chart, at roughly 2,000 cases and claims of 13.3% of payrolls.

Outliers offer interesting insights.  The two dots within the oval at the upper left reflect states that have seen fewer claims than would be expected for the incidence of the virus.  The dots represent Connecticut and Washington, D.C.  In the case of D.C., a high percentage of workers are federal government employees and will not be laid off, so it makes sense for D.C. to be an outlier to the upper left.  Connecticut appears to be among the worst-performing states in terms of handling the inflow of claims, a hypothesis that is supported by the fact that it was one of a few states that saw the number of new filers rise on a week-to-week basis in the latest release.  Similarly, New York, the dot at the top of the chart, appears to lagging somewhat in terms of claims as well. The struggles of the New York state unemployment office have been well documented, and New York posted a substantial increase in the latest week, suggesting that it is still working through a backlog of filers.  Still, overall, this paints an encouraging picture.  New York probably has several hundred thousand filers that should have already been processed, but Connecticut is too small to have much impact on a national basis, and there are not many other states that look to be radical outliers.

In contrast, the three dots in the rectangle at the bottom right represent states with heavy claims relative to the incidence of the virus. Two of the three represent Hawaii and Puerto Rico, economies where the absence of travel and tourism have undoubtedly had a disproportionate economic impact.  The third dot represents Kentucky, which appears to be an outlier, in that its layoffs far exceed what the incidence of the virus would suggest.  Kentucky implemented one of the earliest aggressive lockdown regimes, which of course exacerbated layoffs but appears to have been relatively effective at limiting the spread of the virus. Incidentally, the dot directly above the rectangle represents Nevada, another state heavily dependent on travel and tourism.  The outliers in the bottom right mainly represent special cases.  If there were, say, six states to 10 states in the bottom right corner of the graph, well below the line, it might suggest that a small number of states had efficiently processed claims but that everyone else was still well behind.  There is not much evidence of that, which is a good sign.

With the likely exception of New York, most states are not far behind in terms of processing.  Indeed, Pennsylvania offers a reassuring observation.  State officials have estimated claims from the beginning of the flood of new filers, in essence sidestepping any processing glitches in reporting its numbers.  It also was another state, like Kentucky, that had an early and aggressive lockdown.  Even so, the Pennsylvania observation is more or less right on the trend line drawn in.

Looking ahead

So, the good news appears to be that there is not a massive backlog of past filers yet to be counted.  The bad news is that layoffs will most likely continue. In the week ended April 11, the latest observation, most states recorded a week-to-week decline in initial claims.  I expect that trend to continue going forward, though the number of new filers is likely to remain exceedingly high for at least the next few weeks. Large states that posted substantial weekly declines in the last period, most notably California and Michigan, may continue to see moderation, while New York and a few others work to catch up.  My estimate for the coming week is around 4,000,000, and, if that expectation proves reasonably accurate, I would not be surprised to see ongoing significant drops in late April.