Disability prevalence is rising alongside employment

The increase is driven by cognitive difficulties among younger people and needs to be monitored because of its potentially serious long-term policy implications

Photo by charlesdeluvio on Unsplash

After the onset of the pandemic, the employment-population ratio for people with disabilities has risen to its highest level on record (23.0 percent in August). Perhaps more importantly, however, the share of the population reporting a disability has also increased over that period and is now about 0.8 percentage points higher than it was in February 2020, representing nearly 3.1 million additional people reporting disabilities. Among types of disabilities, difficulty concentrating, remembering, or making decisions shows the largest increase, and these increases appear more often among younger people. Qualitatively, disability increases of this nature seem consistent with consequences of “long Covid,” but this has not been definitively established. Increased cognitive difficulties among young people could reduce the productive capacity of the economy and/or increase health care and disability insurance costs, though increased employment within this group cautions that these potential consequences are far from certain.

Cognitive difficulties have driven increased disability reporting

Over the years leading up to the pandemic, self-reported disability in the Current Population Survey (CPS) had declined slightly, a trend that continued through 2020.¹ Since roughly the beginning of 2021, however, the share of people at least 16 years of age reporting that they have some kind of disability has increased fairly steadily, as shown in Figure 1, rising about 0.9 percentage points to 12.7 percent on average over the three months ending August 2023, a share that represents a population that is about 10 percent larger than it was in February 2020.

The increase in disability reporting, earlier stages of which were documented by Ne’eman and Maestas (2022), has been especially pronounced among people aged 25-54 (the so-called prime-age population). This group has also seen its disability rate rise by about 0.9 percentage points, but from a lower base. This increase brings the prime-age disability rate to 6.9 percent over the last three months, corresponds to nearly 1.3 million additional prime-age people with disabilities, and represents a 16 percent increase in that population. In contrast, the population not reporting any disabilities has grown by about two percent since February 2020, with slower growth among prime-age people.

Figure 1: Disability prevalence by age

Looking at changes by type of disability (Figure 2) reveals an especially pronounced increase in the rate at which CPS respondents report having serious trouble with concentrating, remembering, or making decisions because of a physical, mental, or emotional condition. Over the last three months, 4.4 percent of people reported such difficulty, an increase of 0.6 percentage points since the three months ending February 2020. Though people can and do report disabilities of multiple types, this suggests that the increased prevalence of cognitive difficulties has been at least involved with the majority of the overall increase in disability prevalence. The share reporting only trouble with concentrating/remembering/making decisions and no other disabilities was 1.4 percent, about 0.3 percentage points above its pre-pandemic level. Difficulties with only hearing or seeing saw the next largest increases of this kind, each rising by less than half as much.

Figure 2: Change in disability prevalence by type

Even though cognitive difficulties are more common among older people, the increases in cognitive difficulties seen since the onset of the pandemic have been driven by younger people, as shown in Figure 3. Since February 2020, rates of cognitive difficulties have increased by about 1.1 percentage points among people 16-24 and 25-34 and by 0.7 percentage points among people 35-44, in each case representing at least a 25 percent increase in prevalence.

Figure 3: Change in prevalence of cognitive difficulties by age

Qualitatively, the increased prevalence of cognitive difficulties among younger people arising during the Covid-19 pandemic sounds like a potential consequence of “long Covid.” One meta-analysis highlighted cognitive impairment as one of the most common persistent symptoms of Covid-19, with about 0.2 percent of those infected with Covid experiencing it 12 weeks or more after diagnosis. Through 2022, the Centers for Disease Control and Prevention estimated that about 77.5 percent of people ages 16 and older had been infected with Covid-19 at least once. Based on these rates, one might expect to see about 400,000 additional people reporting cognitive difficulties when responding to the CPS. Given that other types of disability have seen smaller increases in prevalence in the CPS than cognitive difficulties since the beginning of the pandemic, this would suggest that most of the overall increase of about 3.1 million people is not explained by this channel.

Alternatively, using changes in the rate at which people are absent from work due to their own illness as a proxy for long Covid, similar to Goda and Soltas (2022), suggests that a potentially large share of the increase in disability reporting could be related to long Covid. However, the relationship between long Covid prevalence and changes in disability reporting is sensitive to specification, and some specifications suggest little role for long Covid in the recent increase in disability.

Employment among people without disabilities has lagged during the recovery

In addition to shrinking as a share of the population, people without disabilities have seen their employment recover more slowly than people with disabilities. Figure 4 shows that the employment-population ratio (EPOP) for this group remains about 0.6 percentage points below its February 2020 level; for the prime-age non-disabled population, EPOP is roughly in line with its pre-pandemic level. Similarly, the labor force participation rate (LFPR) among people without disabilities has yet to consistently exceed levels before the pandemic. In contrast, EPOP and LFPR among people with disabilities surpassed their pre-pandemic levels in mid-2021 and now exceed those levels by nearly four percentage points, and by nearly eight percentage points among prime-age people with disabilities. Among prime-age people, the combined increase in the population of people with disabilities and employment within that population is large enough to account for more than the entire increase in overall EPOP.

Figure 4: Employment and participation by disability status and age

The superior performance of employment among people with disabilities is at most partially explained by two highly salient features of the pandemic-era economy: tight labor markets and work from home (WFH). Estimates suggest that employment among people with disabilities has been more responsive to labor market tightness as measured by the state-level quit rate since the pandemic. The magnitude of this difference, however, is relatively small, explaining only about one-sixth of the recent increase in EPOP among people with disabilities.²

While people with and without disabilities have seen increases in employment in occupations that are amenable to WFH since the onset of the pandemic, they have seen very different patterns of employment growth in occupations that are not.³ People with disabilities have seen employment in non-WFH occupations grow roughly proportionally to its share of employment within this group prior to the pandemic. People without disabilities have seen employment in non-WFH occupations remain below its pre-pandemic level, more than offsetting growth in WFH occupations. As the labor force in this group also remains below its pre-pandemic size, employment has only limited scope to grow in one sector before it begins to crowd out employment in the other. Increased salience of and access to WFH options during the pandemic may have contributed to the increase in LFPR among people with disabilities and allowed employment to grow as well.

Increased disability could have serious implications and needs to be monitored

Increasing disability among young people is potentially concerning for a variety of reasons. If disability limits the amount or kind of work they can do, it could reduce the productive capacity of the economy. Higher rates of work-limiting disability could also increase costs associated with the Social Security Disability Insurance program. If disability requires medical treatment, it could increase healthcare spending, including in public programs like Medicaid, which serves millions of people with disabilities. Increases in cost for public programs could be particularly large if disability persists through the remainder of young people’s lives.

These consequences, however, are far from certain. So far, the increased disability has been accompanied by increased employment among those with disabilities, including specifically among those reporting cognitive difficulties, suggesting that any reduction in productive capacity or increase in work-limiting disability may be limited. Similarly, through 2022, neither SSDI applications nor awards had risen, according to Social Security Administration data. It is also far from clear whether or how long this increase in disability will persist. To the extent that the disability increase arises from Covid-19, it could slow if infections become less common due to increased immunity or additional treatments are developed.

The recent shift in disability reporting is still a developing trend. Continued monitoring of data from surveys like the CPS and programs like SSDI and Medicaid will help inform our understanding of what has changed and how. Careful attention should be paid to how a given source defines and measures disability, since these details vary and are likely to be important to understanding what already seems to be a nuanced situation.

1. In the CPS, disability status is measured based on a series of questions that ask if respondents have serious trouble with seeing, hearing, walking or climbing stairs, dressing or bathing, doing errands alone, or concentrating, remembering, or making decisions because of a physical, mental, or emotional condition. In this exercise, any respondent answering yes to any of these questions is considered to have a disability.

The onset of the pandemic disrupted ordinary CPS interview procedures, leading initial interviews that are normally conducted in person to be conducted by phone. This reduced response rates among newly sampled households and led to shifts in the composition of the sample toward those that had already completed several interviews before the pandemic. This may have affected how disability was measured (Ne’eman and Maestas, 2022). Much like certain labor market indicators (see Krueger, Mas, and Niu, 2017, for example), the rate at which people report having a disability differs systematically across months in sample in the CPS, with people who are in their first four months in sample being more likely to report having a disability. The CPS sample shifted away from newly sampled households most acutely during 2020 before gradually reverting, with each of the eight months in sample returning to approximately its normal share of the overall sample by early 2022, suggesting limited scope for this issue to affect disability measurement in the most recent months. Setting aside the pandemic, response rates have been in long-term decline across a variety of government surveys, with unclear implications for measurement quality across a wide range of topics.

2. To understand the relationship between employment and labor market tightness, I estimate

\(log(EPOP_{st}) = \alpha + \beta \ log(Quit Rate_{st}) + \sigma_{s} + \tau_{t} + \varepsilon_{st}\)

in moving 12-month windows by disability status, where s indexes states and t indexes months. As shown below, point estimates suggest that employment among people with disabilities has been more responsive to labor market tightness, as measured by the quit rate at state level, since the end of the initial, acute pandemic period (as it was late in the pre-pandemic expansion).

3. Work-from-home status is assigned at the occupation level based on Dingel and Neiman (2020). Thanks to Sabrina Pabilonia for sharing implementation code.