Has Staffing Decoupled from Labor Markets?

By Aaron Haskins & Hugo Malan, Kelly Science, Engineering, Technology & Telecom

Introduction

Conventional wisdom in the staffing sector says that high GDP growth periods combined with tight labor markets create the perfect conditions to drive demand for the industry’s services. This describes recent conditions perfectly: between the first quarter of 2022 and the first quarter of 2024, real GDP grew by 4.7%, the private sector added almost six million jobs, and the unemployment rate maintained its lowest average for any two-year period since the 1960s. One would reasonably expect staffing employment and sales to have correspondingly surged. 

Instead, temp employment fell by over 12%—almost 400,000 workers—and the industry’s total sales plunged by a massive 17.2% over the same two-year period.ⁱ Most major staffing companies reported year-on-year declines in organic real revenue in 2023 as the industry tipped into what one could reasonably call a sectoral recession.ⁱⁱ 

This surprising outcome presents three clear questions for leaders in the staffing industry. First, why have staffing industry trends diverged so wildly from aggregate employment and economic growth? Second, has the relationship between the industry and the broader economy fundamentally changed, or is this just a transitory situation that will resolve itself? And third, what does this mean for the industry moving forward? 

Simple Explanations Fall Short. 

The American Staffing Association’s 2023 Industry Playbook noted that “many industry CEOs believe that this year’s declines may be...reflective of moderation or rebalancing in hiring coming off of some of the over-hiring as the economy emerged from the pandemic-induced recession....”ⁱⁱⁱ 

Others have proposed that the slowdown was the effect of a recent surge in labor productivity, as new technology like AI tools and a recent emphasis on efficiency has allowed employers to grow output with limited hiring. 

It’s true that 2021 featured the fastest employment growth in any two-year span since the mid-1980s. It is also true that the labor productivity growth rate has sharply accelerated from 1.1% CAGR in the 2010s to 1.6% since 2019. Either one of these factors could well have contributed to employment growth in 2023 slowing to 2.3% from the 4.3% mark set the year before. 

The problem with both of these simple explanations, however, is that total employment still continued to rise through 2022 and 2023, not fall. In fact, apart from 2021 and 2022, 2023’s 2.3% employment growth rate was the highest since 1999. The important question is why temp employment, specifically, declined while total employment continued to rise strongly, and neither over-hiring nor productivity acceleration provide a satisfactory answer for this. And the observation that temp employment growth rates tend to be more volatile than total employment growth doesn’t actually answer anything; it’s a fact, but not a causal explanation. 

No, explaining why temp employment and staffing demand have steadily fallen since their peak in early 2022 requires a more nuanced story. 

Labor Demand, Staffing Demand, and Relative Preferences. 

The Bureau of Labor Statistics groups the staffing industry’s various contract labor solutions, from contingent staffing and short-term contractors to more comprehensive SOW-based outsourcing, as “Temporary Help” in its detailed employment numbers. These don’t encompass the entirety of the industry’s sales, as revenue from other products such as direct hire recruiting and workforce program management products like Recruitment Process Outsourcing and Managed Service Providers makes up a significant portion of industry sales and doesn’t show up in temp employment figures. But, in general, BLS-reported temporary help employment levels are highly correlated with inflation-adjusted staffing industry sales performance, which makes it a useful proxy for overall sector performance: 

Per basic economic theory, the demand for temps is a product of two factors: the total demand for labor, and the relative preference for these forms of employment versus other labor options. This latter factor is key: how closely changes in temp employment track total employment depends largely on how willing employers are to engage temps to fill their labor needs and how willing workers are to take temp and contract positions. 

When the preference for temps is reasonably stable from year to year, temp demand is closely tied to total labor demand. As that preference increases, temp employment can grow significantly faster than total employment. And, of course, the reverse is also true: declining willingness to use staffing services can result in shrinking staffing employment even when overall employment is growing. 

Economists can’t directly observe the average preference for temps, but it can be measured after the fact using the temp penetration rate, the percentage of total employment classified as Temporary Help. And while temp employment levels have historically been viewed as closely correlated with the broader labor market (though generally leading overall employment), the data shows that changes in the penetration rate are far more useful to understanding temp employment trends than broader employment growth.  

Between 1983 and 2019, there were seven years in which temp employment declined. Total employment only fell in four of those years, but temp penetration declined in all of them. The reverse story is similar: of the five years in which total employment shrank, temp employment fell in only three of them, and the years in which temp employment grew also saw rising temp penetration, sufficient to overcome declining total labor demand. There’s only one year in the period where temp penetration fell yet temp employment increased, as total labor demand grew by enough to slightly offset the small decline in staffing’s share. 

Temp penetration growth is also a far stronger predictor of temp employment growth than GDP: temp employment managed to fall in seven years that saw positive GDP growth (ranging from 2.0-4.8%), but only once did the penetration rate fall without temp employment following suit, and temp employment has never grown without a simultaneous increase in the penetration rate.  

Since Q1 2022, total employment went up, but temp employment declined, which inevitably means a decline in the temp penetration rate greater than the increase in total employment: indeed, the penetration rate in Q1 2024 was down a whopping 15.8% versus Q1 2022, while total employment was up by 4.7%. 

So, if the decline in staffing sales is at least in large part due to this apparent shift in labor market preferences, what explains the sudden decline in the use of temp labor to meet hiring demand? 

Hiring Preferences: Anxiety, Leverage, and Cost. 

It’s critical to consider the preferences of both employers and workers. As the demand side, employers won’t use staffing solutions to fulfill their labor needs unless there’s some benefit gained over other options in the market, like hiring in-house employees, using self-employed contractors, or bringing in bench-based consultant services. And as the supply side, workers’ willingness to take temp and contract roles is obviously critical to staffing employment success. 

In both cases, like in any market with competing options, the choices individuals make are based on the relative costs and benefits of the various choices. At the individual level this is of course highly variable, but at the level of the economy, there are clear forces affecting how appealing staffing is versus other labor arrangements for both employers and workers. Specifically, three major drivers together appear to provide a reasonable causal story to explain what happened in 2023. 

Anxiety. 

The first is short-term anxiety, or how concerned decision-makers are about the potential for a downturn in the near future. The Federal Reserve Bank of Philadelphia calculates the “Anxious Index” as the consensus probability of negative GDP growth in the coming quarter. This measure has an interesting nonlinear relationship with temp penetration growth: the two are clearly inversely correlated, but that correlation turns out to be much stronger for higher Anxious Index readings than for lower. Extreme anxiety seems to affect preference for staffing quite strongly, while low anxiety has relatively little impact. 

Now, normally the Anxious Index closely mirrors GDP growth, rising as growth slows and falls, then falling as GDP rises. But on occasion the two have diverged, with anxiety rising but the economy managing to avoid falling into negative growth. From 1983 to 2019, the full span of available data before the post-pandemic period of interest, there were five years in which annual GDP growth was at least 2%, but economic anxiety averaged 20% or greater. Temp penetration rates fell in four of those five (1991, 2001, 2002, and 2008).  

At the quarterly level, 71% of quarters with anxiety over 25% saw a declining penetration rate in the following quarter, while only 7% of quarters with anxiety below 10 saw the same.¹ This directionality, in which anxiety levels in a given quarter are most correlated with changes in the following quarter’s penetration rate, suggests a causal relationship: hiring managers appear to adjust their preference for temps to their anxiety levels, rather than anxiety increasing in response to changes in temp staffing. 

At first glance, this response may seem counterintuitive. Received wisdom in the staffing industry is that periods of uncertainty increase demand for temporary labor, as companies hesitate to increase fixed labor costs and instead opt for the flexibility of staffing solutions. But the time horizon matters. Economic conditions further in the future are usually more uncertain and thus riskier than what will happen in the next quarter or two, but occasionally perceptions of immediate risk can spike while future risk remains stable. A well-known effect of such spikes is an inverted yield curve, a rare situation when markets price near-term risk higher than long-term risk. And, indeed, a review of the data shows that spikes in Anxious Index levels are correlated with drops and inversions in yield curves. 

When this happens in labor markets, the elevated near-term anxiety may strongly depress demand for temporary labor, even as the more sanguine medium-term outlook can sustain internal hiring and overall job growth. 

And much like an inverted yield curve, this dynamic can potentially be self-reinforcing: as fears shift preferences away from temps, falling temp employment affects macroeconomic forecasts and increases perceptions of an imminent recession in a positive feedback loop.  

Labor Leverage. 

For their part, workers generally seem to prefer the stability and benefits associated with in-house employment over temp and contract positions, and are less likely to accept such jobs when they have the power to look for a more preferred alternative. The labor leverage ratio (LLR), originally conceived by Aaron Sojourner at the W.E. Upjohn Institute for Employment Research, is the ratio of how many workers are voluntarily leaving their jobs to those who are dismissed or laid off by their employers. As Sojourner put it, a termination usually occurs when one party judges it has better prospects outside of the current employment relationship. Quits typically mean workers are confident in their ability to find a better job, while layoffs and dismissals occur when employers decide the hassle of replacing or downsizing a role are worthwhile, so a higher LLR value implies a stronger bargaining position for workers, while lower values suggest employer strength.^iv The LLR is also not particularly strongly correlated with the Anxious Index, possibly because it’s a more complex dynamic involving both workers’ views of the labor market and employers’ labor capacity decisions, instead of a single reading of short-term expectations. But whatever the reason, this low correlation means they aren’t merely two sides of the same coin, and labor leverage can be considered as an additional factor that can help explain temp employment dynamics. 

Labor leverage is, however, correlated with the temp penetration rate. The relationship isn’t as strong as that between anxiety and penetration rates, as workers are always somewhat beholden to what jobs are available on the market, while employers have more control over whether the positions they offer are in-house employment or staffing roles. But LLR doesn’t have to explain temp penetration on its own to be useful; it instead helps fill in the gaps, explaining those periods where anxiety doesn’t tell the whole story. In the 7% of quarters where temp penetration declined despite low anxiety, all of them had strong labor leverage. More tellingly, those quarters with anxiety between 10% and 25% were far more likely to see decreasing penetration rates when LLR values were elevated, 56% to 20%. When all else is equal, the more bargaining power workers have, the less likely they are to opt for temp work. 

Relative Costs. 

And finally, private-sector employers are in the business of making a profit, so the last part of the story is cost: the higher the cost of staffing labor relative to in-house employment (both the direct cost of the associated labor and the cost of finding and hiring), the less likely employers are to use staffing services. This one is harder to see in the historical record, as there’s limited data on average staffing bill rates to compare to average wages and benefits costs, but it’s a simple and uncontroversial economic reality. If the cost of staffing relative to hiring in-house rises enough, it will begin to outweigh any relative benefits, and preferences will begin to shift.  

This cost factor also might help explain recent declines in staffing revenue associated with direct hire recruiting services. Recent analysis^v has suggested that the 2022-23 surge in immigration is at least somewhat responsible for the growth in the US labor force (up 3.7% from 2021, versus 0.9% population growth), which helps explain why employment has managed to rise without a corresponding decline in unemployment. This, in turn, means that despite strong labor demand, it’s been getting easier to find candidates; SIA’s Staffing Industry Pulse Survey shows the average reported recruiting difficulty has been declining steadily since mid-2022.^vi Thus there’s less apparent need for external recruiting assistance, and with the cost to acquire in-house employees remaining fairly low, the higher comparative cost of fee-based recruiting could plausibly have resulted in employers choosing to do their own candidate sourcing instead of engaging staffing services. 

To summarize this section, three major forces seem to affect the relative preference for staffing. On the demand side, employers are less likely to choose staffing solutions to meet their labor needs when they’re anxious about an impending economic downturn or when the comparative cost of staffing services shifts the relative value versus alternative options. And on the supply side, most workers prefer not to take temp or contract roles whenever possible, so the more bargaining power they have, the less likely they are to go into staffing positions. 

The Story of Staffing in 2022-23. 

As it turns out, these forces neatly explain the situation in mid-2022 through 2023: 

Despite strong GDP growth, economic anxiety has been elevated at near-record levels. The beginning of the temp employment downturn in Q2 2022 coincides with the Fed raising interest rates to tackle inflation and sparking fears of an impending recession, with the average Anxious Index level jumping up to over 40% by the end of that year and remaining exceptionally high through 2023. There has never been a year with positive temp penetration growth and an average anxiety level over 30%, so it’s unsurprising that temp penetration fell by 9% in 2023, far more than enough to result in declining temp employment even with total employment increasing. As noted earlier, this rare confluence of elevated anxiety about the immediate future despite a stable medium-term outlook is similar to the perceptions driving an inverted yield curve, and in fact the yield curve inverted in late 2022 and remained so throughout 2023. 

But if that weren’t enough, the 2021-23 period has also set historic records for labor leverage, with almost 3.5 quits for every dismissal in early 2022. This came down some in 2023 as the Great Resignation ended and the rate of layoffs began to increase as companies focused on expense controls in an era of rising labor and capital costs, but the ratio was still higher than ever before, as far back as the data goes. Even if employers were open to staffing solutions, workers in 2023 still had sufficient bargaining power to hold out for their preferred employment arrangements—for more evidence of this, just consider the struggles executives have faced in their efforts to get employees back to the office. 

And finally, there’s the cost factor: median staffing bill rates increased by over 17% from 2020 – 2023,^vii versus a 13% increase in the total cost of internal employment in the same period.^viii Employers considering staffing solutions versus alternative options to fill their labor needs in 2022-23 certainly had to consider this shift in the relative cost against the benefits, and given the context of other rising costs affecting their bottom lines, this likely affected their preference away from staffing. 

The past two years, then, could be considered a perfect storm, with conditions ripe to drive both employer and worker preferences away from staffing solutions. As anxiety rose to extreme levels, bill rates continued to creep up faster than average employment costs, and labor leverage remained at historic highs, it should have been unsurprising to see these forces push down temp penetration growth into negative rates. Combine that market shift away from temps with the decreasing difficulty of finding candidates reported by SIA reducing the apparent need for external recruiting assistance, and it’s unsurprising that staffing sales have steadily declined from the post-pandemic highs of 2021. 

Looking Forward. 

This more nuanced story explains why staffing industry trends over the past two years have diverged so wildly from aggregate employment and economic growth. It also helps to answer the second key question: there is little sign of any fundamental changes in the relationship between the staffing sector and the broader economy. Any apparent decoupling is merely the product of powerful forces that, in combination, have been more than enough to overcome the growth in total employment, shifting employer and worker preferences away from staffing enough to see negative temp employment growth even in a strong economy and tight labor market. 

As for what it means for the industry moving forward, that largely depends on broader macroeconomic trends. The most recent Anxious Index reading in Q2 of 2024 was 18.7%, far below the 2023 average, and the Labor Leverage Ratio in May 2024 was down to 2.09, the same level as March 2019. This suggests that the decline in the temp penetration rate should begin to slow and reverse in the coming months—the slowing has already begun, with the average monthly rate of decline from January through May 2024 at only -0.50%, versus an average of -0.85% from April 2022 through December 2023.  

The one force that may be more enduring, however, is the shifting relative cost of staffing: if average bill rates continue increasing faster than the cost of internal employment, more employers will start to see the economic benefits of doing their own recruiting and hiring workers internally, especially if the difficulty of sourcing candidates (and thus the relative up-front cost of acquisition) continues to fall. This could potentially result in the temp penetration rate stabilizing at a significantly lower level than it was in the pre-pandemic period. 

But even at a lower level, once the penetration rate stabilizes, temp employment will likely resume its long-term correlation with total employment levels. As long as the economy continues to grow and businesses continue to expand output, even in the face of accelerated productivity growth, total employment should also continue increasing, and staffing sales along with it. 

For more information on how Kelly Science, Engineering, Technology & Telecom can support your need for STEM talent, visit our web site today. 

Aaron Haskins & Hugo Malan are executives at Kelly Services, a leader in workforce solutions.

_____________

ⁱ American Staffing Association

ⁱⁱ Data from public filings. Robert Half: -11.7%. Randstad: -6%. ManpowerGroup: -4.0%. Kelly Services: -3.2%.

ⁱⁱⁱ American Staffing Association

^iv Sojourner & DiVito, “The Labor Leverage Ratio: A New Measure That Signals A Worker-Driven Recovery,” 2022, https://rooseveltinstitute.org/2022/02/04/the-labor-leverage-ratio-a-new-measure-that-signals-a-worker-driven-recovery/

^v Edelbert & Watson, “New immigration estimates help make sense of the pace of employment,” 2024, https://www.hamiltonproject.org/wp-content/uploads/2024/03/20240307_ImmigrationEmployment_Paper.pdf

^vi Staffing Industry Analysts, US Staffing Industry Pulse Survey Report, March 21, 2024

^vii Staffing Industry Analysts, Staffing Industry Benchmarking Consortium, Overview Reports for 2021, 2022, & 2023

^viii US Bureau of Labor Statistics, Employment Cost Index: Total Compensation, All Civilians

¹ The Anxious Index typically reports anxiety levels for the quarter associated with the survey responses rather than when the responses were collected. To assess causality, this analysis instead matches index readings to the quarter of the survey responses: a given quarterly index reading in this analysis is the level of anxiety felt in that quarter about the potential for a downturn in the following quarter.