- In searching for explanations for the steep decline in the U.S. labor force participation rate analysts have rightly stressed the importance of retiring baby boomers.
- Increase in disability share accounts for 20-25% of the drop in the labor force participation rate since 2007 (vs ~45% for retirements).
- We expect this shift to be essentially permanent, but growth in the disability share will also probably slow down.
Zach Pandl, Portfolio Manager and Strategist | April 7, 2014
In searching for explanations for the steep decline in the labor force participation rate analysts have rightly stressed the importance of retiring baby boomers. This trend appears to be the single most important factor depressing participation over the last six years. However, we have been surprised in recent conversations with investors about the limited agreement on another important driver: the steady increase in the share of workers leaving the labor force due to injuries or other serious disabilities. The rise in the so-called disability share—the percent of the adult population out of the labor force due to disability—has also contributed to the structural weakness in the U.S. participation rate, and this colors our interpretation of current labor market statistics.
Workers leave the labor force for a variety of reasons, including disability, retirement, education and discouragement about job prospects. For the labor force as a whole, disability is the third most common cause of nonparticipation, accounting for 16% of the working age population outside the labor force. However, disability accounts for a much larger share of nonparticipation for certain age groups. For workers in their 40s and 50s, almost half left the labor force because of injury or other disabilities (helpful charts here). In addition, disabled workers have accounted for a rising share of persons outside of the labor force in recent years.
Disabled workers are eligible for federal benefits through two sources: social security disability insurance (DI) and the Supplemental Security Income (SSI) program. As of 2012, 7.9 million persons received DI benefits only, 1.4m received SSI only, and 3.5m received benefits through both programs (including dependents of disabled beneficiaries). After reforms to federal disability benefits in the mid-1980s the US disability share steadily increased (Figure 1).
Figure 1: Recipients of disability insurance benefits
When workers leave the job market because of injury or disability they reduce the labor force participation rate. Thus, along with the surge in retirements, growth in disability could help explain the falling participation rate and historically low employment-to-population ratio. There are two main sources of disability statistics: (1) the Current Population Survey (CPS), which is the same underlying data used to calculate the unemployment rate, and (2) disability benefit recipient data from the Social Security Administration. The two measures are not exactly the same, and the CPS does not use DI or SSI benefits to determine disability status. However, we believe both sources capture the same basic economic phenomenon (nonparticipation due to disability), and in any case the numbers are in the same ballpark.
The table below summarizes the key data. From the end of 2007 through the end of 2013, the number of workers outside of the labor force due to disability increased by around 2 million (CPS data are from Fujita (2014)). As a share of the working age population, the increase ranged from 0.6%-points to 0.9%-points. Over this period, the labor force participation rate fell by 3.2%-points. Thus, these figures suggest that growth in disability accounted for 19-28% of the total decline in participation. By comparison, retirements account for about 45% of the decline in the participation rate since 2007.
Academic research points to three broad explanations for the growth in the disability share:
1. Demographics. An aging population puts upward pressure on the disability share because disability is more prevalent among older workers. Thus, there has been a natural tendency for disability claims to increase as baby boomers age. In addition, more women are now eligible for DI benefits due to rising labor force participation prior to the last 15 years. Daly (2013) estimates that these types of structural factors accounted for 35% of the increase in the disability share through 2011.
2. Benefit and program changes. Most research finds that the attractiveness of benefits has played a role in program growth. For instance, Autor and Duggan (2003) explain that the DI benefit formula, combined with a widening income distribution, caused an increase in the earnings replacement rate for the DI recipients. The increase in the real value of Medicaid benefits may also have contributed to program growth. Separately, reforms passed in 1984 (the Disability Benefits Reform Act) relaxed eligibility rules and allowed more disabled workers to qualify. According to CBO (2012), these changes “led to a substantial expansion in the share of DI beneficiaries with mental or musculoskeletal disorders, many of whom enter the program at younger ages than do people with other types of disabilities and many of whose applications are largely judged by using subjective criteria”.
3. Weak economy. DI and SSI make up a part of the nation’s social safety net (see Burkhauser et al (2013)), and just like other social programs, demand tends to rise during economic downturns. The correlation between DI applications and the unemployment rate is evident in the data (Figure 2), but there is also a large body of research more carefully demonstrating the cause/effect relationship—such as Black et al (2002), who study DI and SSI growth in regions that experienced booms and busts in the coal industry.
Figure 2: Unemployment rate and DI applications
The sizable increase in disabled workers as a share of the population should affect how analysts interpret current labor market statistics, in our view. By definition, DI and SSI recipients are not able to work and therefore rarely return to the labor market (according to Stapleton et al (2010), only 4% of a sample of DI recipients in the 1990s had returned to work after ten years). Thus, the trend increase in the disability share has likely permanently lowered the labor force participation rate and employment-to-population ratio. Indeed, when we adjust the employment ratio for the effect of disability and retirements it looks broadly similar to other labor market statistics—a point made in a speech earlier this year by San Francisco Fed President Williams (Figure 3).
Figure 3: Adjusted employment-to-population ratio
Looking ahead, we expect that growth in the disability share will slow. First, DI applications have already declined sharply, possibly due to improving economic conditions. Second, as disabled baby boomers move into their mid- to late-60s, benefits will shift from the DI program to standard social security (self-reported status in the CPS data tends to shift from disabled to retirement categories at this time as well). Lastly, after the latest increase, the U.S. disability share is more in line with OECD norms.