Mr. Market Gets The US Labour Data Right

So, let’s get this straight. The Bureau of Labor Statistics somehow misclassified a bunch of people as employed when they should have been reported as temporarily unemployed. Twice. In April and May.

As a result, the May unemployment rate should have been 16.4% versus the official 13.3%. And April should have been 19.5% versus the reported 14.7%.

What? Me, Worry?

Does anyone seriously think Mr. Market cares? All that matters is the direction. Everyone thought it was going to be worse; consensus estimates were for the unemployment rate to be 19.5%. Instead, employment was up. Unemployment was down, however you measure it. What was there to dislike?

For What It’s Worth…

There were other anomalies.

• In April, people not in the labour force who do not want a job jumped by 1.7 million people (to 92.1 million people). Then in May that category reversed, dropping by 1.7 million people (Table 1). Either those people (who didn’t want a job) suddenly got a job and joined the labour force, or, then again, maybe they were employed and in the labour force all along. Take your pick.
• The BLS reported that, due to the coronavirus lockdowns, the response to the establishment and household surveys were about 10 and 15 percentage points lower than normal, respectively. The BLS said their sampling was still robust, but at the very least the margin of error is larger.

The monthly labour market report is what is it is, but there is no denying that it is a noisy mess these days.

*Adds recent labour force leavers who want to work and misclassified workers

**Excludes temporary laid off; adds recent labour force leavers who want work

Rest Assured, it Wasn’t All Roses

Keep three points in mind.

First, apart from the statistical anomalies, the improvement in the labour market was entirely due to temporary workers being recalled. Given that the establishment and household surveys are designed to cover the week including the 12^th of the month, very little of this improvement came from states reopening their economies – it was almost entirely due to the Payroll Protection Program, which provides loans so businesses can call back laid-off workers.[1] Absent that, few temporary unemployed people would have been recalled.

Second, as we have pointed out in previous Macro Hive articles (here and here), we expect the temporary unemployment rate to fall rapidly as states reopen their economies. There really is no new information here.

Third, other key components of the labour market report were little changed. Permanent layoffs increased modestly from 2.6 to 2.9 million, and people who recently fell out of the labour force but who

want a job dropped from 5.4 to 5.1 million. These are presumably people who lost their jobs due to coronavirus cutbacks but couldn’t search for a new one due to the lockdowns. If we are correct in assuming that many or most of these people are permanent rather than temporary, they will rejoin the labour force and ranks of the unemployed in coming months. If history is any guide, finding new employment will be tough.

Structural Unemployment Creeps Higher

Our measure of the structural unemployment rate rose slightly from 6.4% to 6.6%.[2] The good news is that it isn’t getting worse. During the Great Recession, this measure rose to 9%.

The open question is, how many temporarily unemployed people will morph into permanently unemployed? It is too soon for most companies to know how the coronavirus lockdowns will affect their businesses, so the path of least resistance is to keep people on temporary layoff or rely on the PPP for now. If structural employment remains near present levels, we can look forward to the economy steadily returning to normal over the coming year and some form of V-shaped recovery.

If structural unemployment worsens, the economic outlook becomes less sanguine.

How Did Forecasters Miss So Big? Or Did They?

Forecasters were presumably working off of April’s labour report, where unemployment rate was 14.7% and 23.1 million people were unemployed. The forecast unemployment rate of 19.5% is consistent with about 30.6 million unemployed people. It is easy enough to work with the weekly initial and continuing jobless claims data to arrive at an estimate close to this.[3] Forecasters probably missed partly because they underestimated the uptake on the PPP and partly because some temporarily unemployed people who were recalled never filed for or received unemployment benefits.[4]

Alternatively, it is possible that forecasters classified the available data a bit differently. Maybe they did not classify temporarily unemployed people as employed (or assumed the BLS would fix this anomaly, as it was flagged in the April labour report after all); and did not exclude some 5 million temporarily unemployed people from the labour force. When we make these adjustments, we arrive at a pro forma unemployment rate of 19.6% – in line with consensus forecast (Table 1).

Mr. Market Has a View

It is easy enough to work the numbers to suggest the labour market report was nowhere near as robust as the headline implied. But the one thing the market will be focusing on is whether it is improving. In fairness, our pro forma unemployment rate for May is a meaningful improvement over April’s pro forma 22.2% unemployment rate.

The one certainty is that we can expect continued volatility in the monthly labour report. The weekly jobless claims and continuing claims data is hardly perfect, but it’s the closest thing to real time data on whether and how quickly the labour market is coming back to life.

If all this noise and uncertainty pushes some uncomfortable buttons, then listen to Mr. Market. He’s got it all figured out.

[1] The Payroll Protection Plan provides low interest loans to businesses with 500 or fewer employees to cover payroll and other expenses. If loans are used to recall workers at their previous salaries for at least two months (and other requirements are met), the loan will be forgiven and become a grant.

[2] Our measure of structural unemployment excludes temporary unemployment and adds recent labour force dropouts who want a job. We assume they were unable to seek work because of coronavirus lockdown.

[3] I was able to do so by calculating a four-week total initial claims less four-week total ending jobless benefits, lagged one week because continuing claims are reported with a one-week lag.

[4] The drop in the BLS’s pro forma unemployment rate from 19.5% to 16.4% would have been consistent with 5 million additional unemployed people who weren’t receiving jobless benefits being recalled to work.

Over a 30-year career as a sell side analyst, John covered the structured finance and credit markets before serving as a corporate market strategist. In recent years, he has moved into a global strategist role.

(The commentary contained in the above article does not constitute an offer or a solicitation, or a recommendation to implement or liquidate an investment or to carry out any other transaction. It should not be used as a basis for any investment decision or other decision. Any investment decision should be based on appropriate professional advice specific to your needs.)