AP Images

About the author: Abigail Wozniak is an associate professor of economics at the University of Notre Dame.

Many Americans take pride in the idea that our country offers frequent opportunities for determined individuals to improve their economic lot in life. We imagine earlier generations moving West when agricultural conditions deteriorated during the 1930s; or moving North to fill in-demand blue collar jobs in our manufacturing centers in the 1940s and 1950s; or quickly moving through a series of entry level jobs before settling into the right job match. Yet economists who study these issues have reached a high level of consensus that these types of transitions have declined over the past three to four decades.

Regardless of how labor transitions are measured—as a change of employer, change of industry or occupation, change of location, or movements into or out of work—all are at substantially lower levels today than they were at the close of the 1970s. Rates of employer-to-employer job change have declined 25 percent, and inter-state moves are down 15 percent or more. According to one summary measure, overall fluidity has declined by 10 percent to 15 percent since the late 1970s.

The numbers are clear and there is widespread agreement—something about labor mobility in the United States has certainly changed. But it is less clear, and there is less agreement, about whether this change is good or bad. One leading economist remarked at a recent conference that if we replaced the word “mobility” with “turnover” then we would be celebrating its decline.

This remark highlights that transitions can happen for good and bad reasons. Transitions can indicate that workers are taking advantage of improved opportunities and are reaching better work and location arrangements. Or they can indicate that opportunities are scarce and require extensive searching. Conversely, the decline in transitions may indicate that workers are better matched to or better compensated by their current firms, requiring fewer changes. Or the decline may mean that opportunities for advancement have become fewer and farther between, and possibly, that the risk of an unsuccessful change has become greater.

What do we know about why
labor mobility has declined?

The broad consensus around declining U.S. labor mobility is a recent development. Although some key research on the question began in the late 1990s, interest in labor market adjustments among scholars and the public took off as the country began what would be a long recovery from the Great Recession of 2007-2009. It was clear to all that many changes would have to be made to return U.S. workers to a situation like that of the mid-2000s. Families would have to move to booming cities from elsewhere. College graduates would have to switch into jobs that more intensively used their skills. Workers who had become discouraged and left the labor force would have to search for and return to employment. Yet as scholars began to look more systematically at measures of these transitions at the close of the last recession, it became apparent that not only had such transitions declined during the recession but in fact had been in decline for decades.

Given that the consensus of broad-based decline in labor mobility is a recent development, it is not surprising that scholars have yet to settle on a single explanation. There is a long list of potential explanations, however, that have been considered, and there has been some success in determining which of these are most plausible.1

Here are four plausible explanations that deserve more investigation:

  • Rising compensation flexibility may mean workers are paid what they earn more consistently, reducing the need to change employment to adjust wages. The data to fully investigate this are limited and restricted to a small set of researchers. There is some evidence that earnings volatility has risen, which might reflect more frequent compensation adjustments, but there is little evidence of an ongoing trend toward greater volatility.
  • Declining firm dynamism (fewer start-up firms) may have reduced opportunities for workers to jump ship from stagnating firms to high-potential new firms, lowering overall labor movement. There is ample evidence of a decline in the rate of new firm formation. But the links between this and overall labor market fluidity are the subject of ongoing research.
  • Employers may be sharing fewer productivity gains with workers, limiting their incentive to change their employment situations. This is equivalent to an explanation that says bargaining power of workers has declined. Stagnant wage growth may be one symptom of this, but as with compensation flexibility, the data to fully investigate this are lacking, and it seems at odds with an increasingly competitive product market.
  • Lower social capital and trust may make both employers and workers reluctant to change their situation, slowing overall fluidity in the labor market. There is strong evidence of declining trust and social capital (connectedness) in the United States dating back to the 1970s, but as with declining firm dynamism, the connection between the two declines has not yet been fully tested by researchers.

Here are seven explanations that have been investigated and found to play little role:

  • Aging population and other demographic changes lead to fewer transitions in the labor market as workers age and become more settled (for example, through home purchases), but these changes are too modest to explain the overall fluidity decline.
  • Occupational licensing now affects one-quarter of the U.S. workforce, and licensing requirements may slow movement into and out of licensed jobs, or across states. But evidence of the rise in licensing at the state level does not appear related to state declines in fluidity, casting doubt on in as an explanation for the overall decline.
  • More sophisticated job search and recruiting may lead to better matching between workers and firms, reducing the need for employment changes, but since this is not reflected in worker wage growth, the decline in fluidity seems unlikely to be the result of better worker-firm pairings.
  • As jobs become more technology-intensive, firms may increase employer-provided training to workers, raising their incentive to retain workers and making worker knowledge more specialized. This could reduce employment transitions, but there is little direct evidence to support an increase in employer-provided training.
  • Health insurance-related job lock is an unlikely explanation, since fluidity has fallen for workers both with and without employer-provided health insurance.
  • Dual-career spouses can face challenges when co-locating in a city, making them reluctant to move once a workable arrangement is reached. However, they should make other types of employment changes at similar rates, and their rise in the population is too modest to explain the fluidity decline.
  • Changes in homeownership, land regulation, industrial regulation, and formalization of hiring have been tested and do not correspond to declining fluidity at the state-level.

Why do people feel like their economic
situation is unstable if fluidity is declining?

Before transitioning to an examination of policies that could boost labor mobility, it is worth pausing to acknowledge what appears to be a puzzle: If employment transitions are declining then why do workers feel so insecure in their jobs? Can it really be the case that workers are staying with their jobs longer when there is broad consensus that the era of “career jobs” is over? The answer to the second question is yes. Job duration has increased at the same time that labor market fluidity has declined. This is in part because of a decline in short-duration jobs—those lasting less than a year or less than a quarter. Yet this brings us back to the first question: Why, then, do workers feel less secure? Potentially it is because the incidence of very long duration jobs has also fallen. So the era of career jobs is ending, but the era of staying with an early employer for months or years longer than one’s parents did is here.

Workers may also feel a heightened sense of insecurity because finding a new job after losing one has become harder. The data show that movement into employment out of non-employment or unemployment has also declined. Separating from an employer without a new job in hand therefore means a longer period of unemployment and job searching than in the past. Although the length of unemployment spells received increased attention during the Great Recession, these are longer-run trends that date back several decades and reflecting a changed landscape of employment in the United States.

How the next Administration
should act on labor mobility

The forces behind declining labor mobility likely have deep roots. The fact that they have been in operation for at least three decades suggests they are unlikely to be affected by short-run policy. The fact that the forces themselves remain to be fully determined mean that it is not yet clear what the appropriate long-run policy responses would be. Still, there are at least five concrete policy steps that are appropriate now:

  • Step 1: Reform the Unemployment Insurance system to reflect the fact that unemployed workers face longer spells of unemployment, and are more skilled and older than in the past
  • Step 2: Develop a pilot program of relocation vouchers for young workers, and use gold-standard methods to evaluate its success
  • Step 3: Assist community colleges and four-year colleges in counseling students who will face longer tenures with any given employer and heightened difficulty changing employment
  • Step 4: Pivot the policy focus on occupational licensing to emphasize job access and rationalizing the burden on practitioners across fields and states
  • Step 5: Develop and improve access to data on the specifics of how firms hire and compensate workers

These steps will inform the ultimate long-run policy response and greatly help workers adapt to the situations they face today. Let’s examine each of them in turn.

Reform the Unemployment Insurance system to reflect the fact that unemployed workers face longer spells of unemployment, and are more skilled and older than in the past

Given longer tenures that workers have with a given employer, the typical worker may experience unemployment spells that are more distant than in the past. With the decline of “career jobs,” more experienced and higher-skill workers will enter the Unemployment Insurance system. Both forces suggest that UI should be reformed to better serve clients who are at more advanced stages of a developed career. Older UI requirements may be a hindrance to such workers. In many states, for example, job search is monitored by requiring UI recipients to apply regularly for available jobs in their fields. This may not be appropriate for more experienced workers, who may rightly pass on applying for a job in their field that entails significantly less responsibility than they had before. Such workers also may be actively searching without applying for jobs, for example, by attending networking events or arranging informational interviews.

Yet it is important to consider the rising share of unemployed workers who face longer spells of unemployment than in the past. These workers will need longer access to assistance in order to keep them connected to the labor market and prevent large negative consequences to their household budgets. There are many ways to support such workers. The UI system could use data it has at its disposal to try to identify workers most at risk of a long unemployment spell, and direct them to enhanced resources sooner.

Benefits also could be offered at a tapered rate to such workers. The idea here is to lower the monthly benefit amount for at-risk workers in order to extend the total months of benefits. Such workers also might qualify for enhanced benefit access, for example, while they are enrolled in a training program. Increased targeting of benefits streams to those at risk for long-term unemployment seems likely to lead to greater welfare gains than other types of expanded access to UI, such as expansions to allow receipt while holding part-time employment.

Develop a pilot program of relocation vouchers for young workers, and use gold-standard methods to evaluate its success

There is limited but compelling evidence that relocation can benefit individuals who are strongly encouraged to do it, leading to higher earnings, better health outcomes, and better schooling outcomes for children.2 The evidence is limited, however, and there is also evidence that some families and individuals fare poorly after relocation. It is therefore appropriate to proceed with a limited program using cutting-edge design and careful monitoring to evaluate its impacts. Evidence and theory suggests that the groups most likely to benefit from such a program are young workers and families with young children. Relocation entails fewer benefits for older children and can potentially be detrimental.

The limited evidence also suggests that any voucher amount would need to be fairly large to encourage take up, perhaps equal to 20 percent to 50 percent of the price of a modest home. Such a program might be financed by allowing individuals to borrow (in whole or in part) against future Unemployment Insurance or Earned Income Tax Credit entitlements. The program could, but need not be, targeted by place of origin. Such programs have been tried on a small scale in the United States in the past. Anecdotally these seem to have had low take-up, and there is little evidence of success.3 Improved targeting and high-subsidy amounts could lead to substantially greater successes for such a program.4

Assist community colleges and four-year colleges in counseling students who will face longer tenures with any given employer and heightened difficulty changing employment

Students need to understand that their first job will have a greater impact on their lifetime career path than it did for their parents. They should be encouraged to search more intensively during their first major period of job search. This is a key time in which marginally greater investment in job search could have payoffs decades into the future. Students should be encouraged to be more ambitious than “just finding a job” by sorting out what they want to do later. That first job is increasingly important for their career path, and they should plan accordingly.

Pivot the policy focus on occupational licensing to emphasize job access and rationalizing the burden on practitioners across fields and states

Many licensing practices can be reformed to rationalize the system and promote access for new entrants, but the evidence is scant that this will substantially jump start labor mobility. Policy on licensing should therefore primarily focus on access and appropriate burden (requirements) and secondarily on labor fluidity.

Develop and improve access to data on the specifics of how firms hire and compensate workers

All available evidence points to key changes in the way that firms hire and compensate workers, but researchers have limited access to the best available data for investigating these issues and no access to other key information because it simply is not collected. While we collect large amounts of information on individuals and families, the information we collect on what firms are doing to attract and retain workers is extremely limited. The last time the U.S. Bureau of Labor Statistics was able to field a survey to ask firms about their employer provided training was in 1995.5

  1. Key surveys of this literature include: Steven Davis and John Haltiwanger, “Labor Market Fluidity and Economic Performance,” NBER Working Paper #20479 (Cambridge, MA: National Bureau of Economic Research, 2014); Raven Molloy, Christopher L. Smith, Riccardo Trezzi, and Abigail Wozniak, “Understanding Declining Fluidity in the US Labor Market,” Brookings Papers on Economic Activity (2016, forthcoming).
  2. Emi Nakamura, Josef Sigurdsson and Jon Steinsson,“The Gift of Moving: Intergenerational Consequences of a Mobility Shock,” NBER Working Paper #22392 (Cambridge, MA: National Bureau of Economic Research, 2016); Bruce Sacerdote, “When the Saints Go Marching Out: Long-Term Outcomes for Student Evacuees from Hurricanes Katrina and Rita,” American Economic Journal: Applied Economics 4, no. 1 (2012): 109–135, Tatyana Deryugina, Laura Kawano, and Steven Levitt, “The Economic Impact of Hurricane Katrina on its Victims: Evidence from Individual Tax Returns,” NBERWorking Paper 20713, Cambridge, MA: National Bureau of Economic Research, 2014).
  3. Brianna Briggs and Peter Kuhn, “Paying for the Relocation of Welfare Recipients: Evidence from the Kentucky Relocation Assistance Program,” UKCPR Discussion Paper DP2008-01 (2008).
  4. A related proposal was made in Jens Ludwig and Stephen Raphael, “The Mobility Bank: Increasing Residential Mobility to Boost Economic Mobility,” (Washington, DC: The Hamilton Project, 2010), http://www.hamiltonproject.org/assets/legacy/files/downloads_and_links/FinalLudwig_DiscussPaper_Oct2010.pdf.
  5. “President Obama’s Upskill Initiative” (Washington, DC: The White House, 2015) and available at https://www.whitehouse.gov/sites/default/files/docs/150423_upskill_report_final_3.pdf.