One of the most hotly debated labor market topics today is the apparent shortage of workers in restaurants, hotels and other leisure and hospitality businesses. Media reports have documented staffing issues in fast food and lodging, hiring bonuses and rising wages in the sector, and the occasional restaurant closure prompted by employees quitting en masse. Of course, the question hanging over all these stories is how much expanded unemployment benefits are to blame (on this, see the latest estimates from UMass Amherst Professor Arindrajit Dube).
Yet there are also other factors affecting employment in low-wage sectors. Childcare needs and safety concerns have played some role in Leisure and Hospitality (L&H) staffing issues, though these factors likely faded in recent months. As numerous news reports have highlighted, workers in L&H jobs may simply be switching to other industries due to low pay, poor and worsened working conditions, and the career introspection shock brought about by the pandemic.
While there has been plenty of media coverage of the industry-switching phenomenon, it hasn’t been well quantified, as far as I can tell. This post attempts to put some numbers to the question of switching, particularly for restaurants and hotels. My best estimate—and it’s a very rough one—is that an additional 1 million workers switched from L&H to other industries during the pandemic (this is above and beyond the typical flow out of L&H). If this number is right, it represents about 6% of the pre-pandemic employment level of the sector (16.9 million), or 46% of the current shortfall in L&H employment relative to February 2020 (2.2 million).
My methodology is straightforward. The Current Population Survey asks around 60,000 people a month about their working situations, including their industry. The survey follows families over consecutive months so we can see what people do over time. For each of the 13 sectors defined by the Bureau of Labor Statistics, I calculate the net switching into that industry. If a worker is in the target industry in month 0 but in another industry in month 1, I define it as a switch out. Vice versa for switches in. Net switching is defined as (switches in) – (switches out), so that positive numbers reflect net switching into that industry. Switches are defined only on those who are in the labor force in both months.
The figure below shows the overall flow of switchers for all sectors, excluding agriculture and mining, with L&H highlighted for clarity. Because the series are noisy, I use 3-month moving averages. The vertical dotted line is March 2020. The pandemic prompted an increase in switching out of the L&H sector, from a monthly average of around -50,000 to roughly -200,000 for most of 2020. The increased outflows that hit L&H make the sector stand out among all the others.
Since L&H is such a large industry, it may appear to be a pandemic outlier simply due to its size. To rule this out, we can look at the same data as a rate: What is the proportion of each industry’s labor force that switches to another industry each month? The figure below provides an answer. Here again L&H again stands out among industries, moving from an average switching rate of -0.3% pre-pandemic to around -1.2% in 2020.
The charts show that, while many workers switched out of L&H over the course of the pandemic, the trend has reversed in recent months. This reversal is even more evident when we look at the raw data in addition to the moving average. These are plotted for L&H below. June 2021 saw the highest one-month positive reading since 2007. This is not too surprising, given the further reopening of the economy, vaccination progress, and rising wages and hiring bonuses in the sector, which should lure some workers back from other low-paying industries.
A somewhat puzzling finding is that the switching out of L&H seems to begin in February 2020, a month before the pandemic really hit. It’s a noisy series, but the February move is 3.7 standard deviations. Did workers foresee the effects of the pandemic this much? It warrants further study.
The charts above make it clear that a large number of workers left restaurants and hotels and entered other industries during the pandemic. The question is whether this trend was economically meaningful. This requires first an estimate of the cumulative departures from L&H—above and beyond the typical rate—since 2020. Even in normal times, around 50,000 more workers switch out of L&H than switch in. Many workers’ first job is at a coffee shop or a McDonalds, after which they find better jobs. Since I’m not measuring flows into industries from out-of-the-labor-force (including first jobs), the switch rate is usually negative for L&H. Some other sectors have positive switch rates in normal times.
The simplest way to calculate the number of workers who switched out of L&H due to the pandemic is just to add up how much the pandemic outflows exceeded some pre-pandemic average. If we use the 2015-2019 average net monthly switch (-48,000), the resulting figure is 1.2 million cumulative additional switchers out of L&H since January 2020. A more involved procedure is to predict what the switch rate “would have been” using a pre-pandemic regression on the switch rate with a linear time trend and month dummies. This yields a very similar estimate of 1.1 million cumulative switches out of L&H due to the pandemic. Given the noisiness of the series and the novelty of this analysis, these numbers should be treated with some caution.
If this rough estimate of 1 million industry switches out of L&H is accurate, is it economically meaningful? As noted above, this 1 million compares to a pre-pandemic L&H employment level of 16.9 million and a current employment shortfall of 2.2 million. A million is significant relative to these totals. Another point of comparison is the level of posted job openings in L&H, which was 1.6 million as of May, or about 500,000 more than in May 2019. However we look at it, 1 million is an economically meaningful number of people to switch out of L&H over the last year and a half.
Needless to say, switching isn’t the whole story. There are plenty of workers who left restaurant and hotel jobs during the pandemic and plan to remain in their original industries, but have yet to return for one reason or another. Yet the data here suggest that departures from L&H for greener pastures have played a big role in the ongoing staffing challenges the sector faces. It’s not just a bunch of anecdotes.
Note: Data from IPUMS CPS. Month-to-month observations are weighted using the IPUMS panlwt variable. This is preliminary work in a rough draft form, so I hope that this post invites comments. I haven’t seen any similar analyses but if they are out there, I would love to know about them. I’m happy to share Stata code and data.