The Recent Rebound In Monthly Employment: Contradictory Micro And Macro Indications

Summary

• A variety of sources of information suggested that the deceleration in jobs growth through early last year should have continued. Instead, there has been a rebound.
• A closer look indicates that the rebound was primarily in leisure and hospitality jobs.
• Industry conditions reports from that sector do not confirm any upturn in business over that time.
• But a top-down approach looking at real retail sales suggests there should have been an improvement.
• Since real retail sales have stalled since last August, the January report this Friday is of extra importance in terms of the near-term outlook in the jobs market.

Introduction

On Friday, I wrote that I had some distrust of the strong jobs numbers that have been posted since May of last year. Basically, it boiled down to the long-term forecast indicating dramatic slowing into Q3 and Q4 of 2019, buttressed by a short-term softening in jobs in Q1 and Q2 last year. Further, the leading sectors of employment - temporary jobs, manufacturing, and construction - in particular showed deceleration or declines in that time frame (and two of those have continued up until now). And then there was the downward -510,000 revision from March 2018 through March 2019.

Since employment is a coincident indicator, we should have seen slowing of the overall numbers late last year. Instead we saw a rebound. Indeed, the BLS revised the number since last May "upward" by roughly 100,000. Why? I've taken a further look, and the answers look different in the micro and macro views.

The Micro view

First of all, here's a look at the q/q gains in the goods producing vs. service providing jobs sectors over the last several years:

Goods producing jobs (outside of construction) have turned flat since the end of 2018, while service providing jobs growth increased significantly in the latter part of last year.

With an assist from Jeff Miller, here's a breakout of the sources of job gains last month that help point us at the sectors where those recent gains have come from:

The big gains were in two sectors: education and health, and leisure and hospitality. A further breakdown between the two sectors shows that most of the outsize gains were in the leisure and hospitality sector:

Health care jobs are in line with their longer-term trend, growing more than jobs as a whole, as the Boomer generation isn't getting any younger.

Looking strictly at leisure and hospitality, we see that only 394,000 jobs were added in the two years between July 2017 and July 2019. In the seven months since, another 300,000 jobs have been added! That's more than a triply of monthly growth:

When it comes to leisure and hospitality, the two big components I think of are hotels and restaurants. As it happens, both of those industries publish monthly reports on business conditions. Here's what they show.

Via Bill McBride, here is a graph of the hotel occupancy rate by week for the last two years (plus the median and lowest years):

At no point did 2019 significantly outperform 2018, and for several short periods, it slightly underperformed. No reason to think hotels were increasing their hiring.

Here are the current and expectations Restaurant Conditions Indexes:

Both actually decelerated considerably for almost all of 2019. No reason to think there was an explosion in hiring by restaurants either.

In short, when we do the ground-up analysis, the surge in hiring in the leisure and hospitality sector over the past eight months in the jobs report looks ripe for a downward revision.

The Macro view

But now let's take a top-down look. Leisure and hospitality is an important component of retail sales. Real retail sales tend to lead jobs by roughly three to six months. So the next two graphs compare real retail sales (blue) with jobs numbers (red).

The first graph looks at the absolute values in each, normed to 100 as of December 2018:

We see the big hit to real retail sales in the first four months of the year, followed by a strong rebound in late spring and summer, a rebound that at the time I wrote was likely to lead to better jobs reports in the ensuing months. Well, with the exception of actual job losses in February ex-Census hiring, payrolls continue their upward climb. (As an aside, the February decline in payrolls is further vindication of my call at the time that the government shutdown in January of last year had brought on a brief "mini-recession").

Here is the same data, presented as the q/q% change. Note that I've amplified the jobs metric better to show the changes:

And there it is. Real retail sales gains decelerated from Q3 2017 through Q1 2019, and rebounded strongly in the two quarters thereafter. Jobs gains followed suit in Q3 and Q4.

Conclusion

I'll be candid here and say I find it difficult to reconcile the poor micro data from the industries with the in-line macro data based on real retail sales. Perhaps ultimately revisions will "split the difference" in slight downward changes.

Going forward, note that real retail sales actually declined in Q4 2019, having peaked in August, for only the 4th time since 2010. Not only does this suggest more lackluster jobs reports in the next few months, but it also heightens the importance of January's retail sales report, which will be released on Friday.

Disclosure: I/we have no positions in any stocks mentioned, and no plans to initiate any positions within the next 72 hours. I wrote this article myself, and it expresses my own opinions. I am not receiving compensation for it (other than from Seeking Alpha). I have no business relationship with any company whose stock is mentioned in this article.