Labour Market Still Labouring: Will It Get The Support It Needs?

Last week’s employment report offered several positive indicators for a nascent recovery. Payroll employment jumped by 4.8 million or 3.6%. The official unemployment rate dropped sharply from 13.3% to 11.1%. And temporary unemployment fell by 4.8 million to 10.5 million, showing that businesses around the country are indeed trying to get back into the swing of things.

Even though the absolute levels of employment are still far below February’s levels, these statistics point to a rebound in GDP growth and are constructive for markets. At this point in the recovery cycle what matters is the sharp change for the better (or impulse) rather than the level.

That said, levels matter too at some point. There are good reasons to believe the underlying labour market is in worse shape than official statistics imply. To the extent policymakers focus on the official statistics, they may fail to take more aggressive steps to support the economy and labour market. This in turn could slow or stunt the recovery after the initial surge, with adverse consequences for both GDP growth and markets for years to come.

In this note, we take a closer look at the unemployment rate and a growing disconnect between the employment report and the weekly overall jobless claims data.

Structural Unemployment is Still on Hold

Two enormous shifts in historical norms have distorted the recent monthly labour market report.

First, in April the labour force collapsed by 4.8 million workers or 3.9%. This is extraordinary. During recessions, over the entire post-war period (including the GFC), the labour force has never fallen by more than 0.3% of the pre-recession level, and it recovered quickly. Virtually all of those people were reclassified as not in the labour force and ‘want a job now’.

Critically, they were unemployed but not counted as such in the official labour market report. Presumably they failed to meet the BLS criteria for being unemployed and in the labour force because they were unable to look for work during the coronavirus lockdowns.^[1] If we include these people in the labour force and count them as unemployed, our pro forma unemployment rate jumps to 13.6% (Chart 1). This is a significant improvement from 19% in April and, as noted above, that improvement is the focus of markets now.^[2]

Second, the vast majority of unemployed people have been furloughed and are counted as temporarily unemployed. Historically, temporary unemployment has been about 10-15% of total unemployment and modestly higher during recessions. In April, temporary unemployed workers accounted for 78% of total employed. That dropped to 58% in the June employment report.

The good news here is that that the unemployment rate could continue to fall sharply as temporarily unemployed workers are recalled.

If we subtract temporary unemployment from total unemployment, we get structural unemployment, or people who have been laid off and are seeking new jobs or who are entering the job market. This has improved modestly from 8% in April to 7.2% in June but is hardly showing the robust recovery of overall unemployment. And structural unemployment is still well below the period after the GFC recession, when it was in the 8-8.5% range in 2009 and 2010.

The risk, of course, is that furloughed workers may morph into the permanently unemployed and need many months to find new jobs, which would push structural unemployment higher. To put a broad range on it, if all temporary unemployed became permanent, structural unemployment would rise to 13.6%. We certainly don’t expect that to happen! But we also note that after the GFC structural employment gradually fell to 7.2% (today’s level) in December 2011. It took another eight (!) years to fall to 3% in December 2019. The point being, structural unemployment tends to recover very slowly. If it increases in coming months, that could have serious consequences for economic growth in coming years.

Will Total Continuing Jobless Claims Soon Follow Payroll Employment?

As unemployment soared, so did initial applications for unemployment insurance – spiking to 47 million so far. Continuing claims, on the other hand, capped out at about 24 million because many of those initial applicants were recalled due to the Payroll Protection Program or because they failed to qualify for benefits.

Historically, continuing claims have run roughly one-half of total unemployment either because some workers fail to qualify for benefits, or their benefits have expired. But thanks to the CARES Act,^[3] unemployment benefits were expanded and hence official continuing claims data has tracked changes in payroll employment fairly closely (Chart 2), albeit with a lag in April probably because of delays in processing applications.

However, the CARES Act also provides the Pandemic Unemployment Assistance program (PUA) for workers who would normally not qualify for unemployment benefits, such as freelancers and gig workers (e.g., Uber drivers). Continuing PUA claims have been running over 10 million a week and have not declined in line with normal continuing claims. Consequently, combined continuing and PUA claims have been flat at about 30 million since early May (Chart 2).

If you take the PUA data at face value, the combined continuing and PUA claims data implies an unemployment rate of about 19.1%. There are caveats: the PUA continuing claims data lags by two weeks and could soon start declining. And some double counting is possible, making the true total somewhat lower.

But perhaps the more important point is that people getting PUA coverage seem to be largely off the radar of the official employment and jobless claims data. This is further confirmation that the labour market is in worse shape than the already dismal unofficial data, despite sharp improvements in May and June.

What We’re Watching For Now…

For now, what matters for the economy and markets is the trend. The labour market is likely to keep improving as temporarily furloughed workers are recalled, although the spike in COVID-19 cases in many states may slow this process somewhat.

The key things to watch for are whether or to what extent furloughs seem to be morphing into permanent layoffs, and any changes in the PUA trends. Those factors could well determine when the current surge in the economy runs its course and whether markets remain mostly upbeat.

Medium term, the big question will be whether policymakers focus on the official job market data or the (in our view) broader and generally weaker labour market. The risk is that emerging fiscal policy and other social support programs will be geared more to the former. In this scenario, inadequate support could stunt economic growth for years, and ultimately have knock-on effects for markets.

Likewise, more aggressive policy now probably means a more robust recovery sooner. Equity markets so far are still pricing in some variation of that scenario. We’ll see how that goes.

1. To qualify to be in the labour force (and to be counted as unemployed) an unemployed person must have been actively looking for work for the past four weeks. ↑

2. During the coronavirus period we assume the labour force stays at the February level and subtract monthly employment to derive our pro-forma unemployment. We assume most of the newly unemployed will soon return to the labour force. ↑

3. The Coronavirus Aid Relief and Economic Security Act was signed into law on 27 March 2020. ↑