Job quality for Americans with disabilities

1 Introduction

In the United States (U.S.), employment outcomes for individuals with disabilities are generally measured as labor force participation, employment, wages earned, or hours worked (Sevak, Houtenville, Brucker, & O’Neill, 2015). These measures are routinely collected and reported by federal agencies on a monthly basis, providing one way of monitoring the labor market activity of persons with disabilities (Kessler Foundation & University of New Hampshire, 2018). Such measures are blunt tools, however, that do not fully capture the nuances of work experiences. In addition to the traditional measures used above, the concept of job quality has recently gained prominence on the international stage as an equally important way of conceptualizing employment participation across nations and across certain sub-populations (Findlay, Warhurst, Keep, & Lloyd, 2017). The purpose of this paper is to use publicly available data to develop and implement a measure of job quality that can be used to enhance the monthly tracking of employment participation for persons with disabilities in the U.S.

2 Literature review

Measures of job quality vary in focus and scope. Whereas economists might identify wages as the primary driver of job quality, sociologists might favor autonomy in the workplace, and occupational medicine professionals might identify workplace safety as the most important indicators of job quality (Olsthoorn, 2013). On the international stage, multidimensional measures of job quality are frequently used. In 2011, Munoz de Bustillo, Fernandez-Macias, Anton, and Esteve created a Job Quality Index, a measure designed for use in international and national work and based on data from the European Working Conditions Survey. The Job Quality Index includes five dimensions: pay, intrinsic quality of work, employment quality, health and safety, and work-life balance. In 2015, an international forum of governments and central banks of advanced economies (i.e., the G20) signed the Ankara Declaration, formally recognizing the importance of job quality on economic development and adopting a measure of job quality that includes not only economic dimensions such as pay, but also a host of other dimensions such as labor market security and quality of the working environment (OECD, 2017). The United Nations Economic Commission for Europe (UNECE) published a handbook on measuring quality employment, outlining seven different dimensions of job quality: 1) safety and ethics of employment; 2) income and benefits from employment; 3) working time and work-life balance; 4) security of employment and social protection; 5) social dialogue; 6) skills development and training; 7) employment-related relationships and work motivation (UNECE, 2015).

Outside of the international realm, individual researchers have developed different measures of job quality as well. Mustosmaki, Oinas, and Anttila (2016) developed a multidimensional measure of job quality, including the possibility to develop and use skills at work, the level of task discretion, perceived work pressures, and job insecurity. Similarly, Horowitz (2016) named multiple dimensions of job quality: individual task discretion, monetary compensation, job security, low work intensity, and safe work conditions. Some researchers have adopted a more parsimonious approach, using only a few indicators to measure job quality. Using data from the U.S. Current Population Survey (CPS) and examining trends in good jobs from 1979 to 2011, Jones and Schmitt (2016) defined “good jobs” as those that paid at least $19 per hour in constant 2011 dollars, had employer sponsored health insurance, and had some form of retirement plan. They found that the share of African American workers in good jobs declined over time (from 1979 to 2011), even while human capital (educational attainment) increased for this population. Overall, other research has suggested a steady decline in the proportion of U.S. jobs that would meet this definition of “good jobs.” Kalleberg (2011) attributes this decline to several institutional changes, including within the realms of the economy (shift to service sector work and increasing inequality between workers and employers), government (reduced regulations protecting workers), and general social and political ideologies (shift toward individualism and associated decrease in unionization).

As a result of the decline in access to good jobs, there is an associated increase in what could be termed “bad jobs.” Goh, Pfeffer, and Zenios (2015) framed poor quality jobs as those that negatively affected health. Using this lens, Goh et al. (2015) identified low social support at work, high job demands, low job control, job insecurity, shift work, lack of health insurance, and high levels of unemployment and layoffs as poor job quality attributes. Jones and Schmitt (2016) described bad jobs as those that had low wages, no health insurance benefit, and no retirement or pension accounts. These studies and others (Kelleberg, 2011) demonstrate that certain socio-demographic groups are more vulnerable to the restructuring that resulted in workers being funneled into jobs with low wages or poor working conditions.

Job quality varies by characteristics one might expect such as years of experience, skill level (Feldstead, Gallie, Green, & Zhou, 2007; Gallie, 2007; Green et al., 2013; Kalleberg, 2011), and educational attainment (Clark, 2005). Researchers have found differences between high and low skilled workers in terms of not only wages (Fernandez-Macias, Hurley, & Storrie, 2012; Goos & Manning, 2007; Tahlin, 2007) but also in terms of job quality (Feldstead et al., 2007; Gallie, 2007; Green et al., 2013; Kalleberg, 2011). However, higher job quality is not equitably distributed by education or skill level, with women (Schmitt & Jones, 2012; McCall, 2001) and minorities (Jones & Schmitt, 2016) facing additional barriers to securing jobs of good quality. Older, white males are more likely to hold higher quality jobs (Burchell, Fagan, O’Brien, & Smith, 2007; Ficapal-Cusi, Diaz-Chao, Sainz-Ibanez, & Torrent-Sellens, 2018).

To this point, the emphasis on job quality may be particularly relevant when monitoring the employment participation of persons with disabilities. In February 2018, the employment-to-population ratio was 30.5% for people with disabilities and 73.5% for people without disabilities (Kessler Foundation & the University of New Hampshire, 2018). While understanding the representation of people with disabilities within the labor force is essential, little is known about the quality of jobs held by workers with disabilities. With respect to the perceived importance of job quality, Ali, Schur, and Blanck (2011) found that non-employed people with disabilities held similar views as non-employed people without disabilities about the importance of flexible work hours, income, and job security, which could each be taken as a measure of job quality. Workers with disabilities are more likely to hold part-time or temporary jobs (Schur, 2003) but are not more likely than others to have jobs that offer flexible hours (Presser & Altman, 2002). For persons with intellectual disabilities, competitive employment is defined as working for at least minimum wage among coworkers without intellectual disabilities. Heyman, Stokes, and Siperstein (2016) defined high quality jobs for persons with intellectual disabilities as those with higher salaries, availability of employer-sponsored health insurance, and a higher number of working hours per week. They found that most workers with intellectual disabilities did not have high quality jobs using these criteria. The literature to date does not extend study of job quality to people with disabilities more generally.

3 Methods

4 Results

Table 1 shows the distribution of workers in each demographic group overall and also separately by disability status and full-time work status. Workers with disabilities are significantly more likely to work part-time (18.3% of all workers with a disability compared to 11.6% of all workers without a disability work part-time, p < 0.001). As people age, they are more likely to have a disability. This is reflected in the profile of workers: those with disabilities are generally older (36.4% of workers with a disability are aged 55 to 64 compared to just 17.4% of workers without a disability, p < 0.001) and less educated (the modal category for workers with disabilities is some college, compared to college degree and higher for workers without disabilities, p < 0.001). Those with disabilities are also less likely to report being in excellent or very good health (43.6% compared to 70.8% of those without disabilities) and less likely to be living in urban communities (84.6% compared to 87.4% of those without disabilities).

Part-time workers are also less likely to be in excellent or very good health (66.9% compared to 70.5% of full-time workers), have lower levels of education, and are generally younger (over one quarter of them are aged 18–24, compared to just 7.2% of full-time employees). Women are also substantially over-represented among part-time workers, with 69.6% of part-time employees being female.

Table 2 shows percent of workers holding good, fair, or bad jobs. Most notably, there is a significant difference in job quality by both disability status and particularly full-time status. While 26.2% of workers without a disability meet our good job criteria, just 21.6% of workers with a disability meet it. Even more pronounced is the difference by work status: 28.7% of full-time workers have a good job, but just 6.1% of part-time workers do.

There are significant differences in the components of job quality by disability status as well as full-time status. Workers with disabilities are significantly under-represented among those with good (above average) wages (p < 0.001): just 37.7% of workers with disabilities earn above average wages compared to 45.3% of workers with no disability. Likewise, part-time workers also tend to earn lower wages (p < 0.001), even after adjusting their earnings for number of hours worked. Just 23.1% of part-time workers have above average earnings, while 48.1% of full-time workers do. Access to employer-sponsored health insurance and pension plans are also more restricted to full-time workers and those with no disability. The majority of full-time workers (82.9%) are offered health insurance as a benefit of employment and 48.9% are offered some type of retirement plan. For part-time workers, those numbers are just 41.0% and 17.6%, respectively (each difference is statistically significant at p < 0.001). The differences between workers with a disability and those without are not as pronounced, but show a similar trend: 75.5% of workers with a disability have access to employer-sponsored health insurance and 43.4% have retirement benefits. For workers with no disability, 78.1% have health insurance and 45.3% have retirement benefits.

The next two tables examine the impact of socio-demographic characteristics on the odds of a worker holding a good job. Table 3 is restricted just to those who are full-time employed (those who reported usually working at least 35 hours per week). Model 1 examines the role of having a disability, independent of any other characteristic. This shows that those with a disability have 15% decreased odds of having a good job compared to full-time workers with no disability (p < 0.01). The impact on the odds of a full-time worker with a disability having a good job is unchanged (see Model 2, Table 3) after controlling for sex, age, education, race, and urbanicity (OR = 0.85, p < 0.05). However, after also controlling for self-reported health status of workers, disability status has no significant predictive power on the odds of holding a good full-time job. In this third model, we also note that the odds of holding a good job increases substantially with age: those aged 25–34 have almost three and a half times increased odds of holding a good job, those aged 34–54 have 7.4 times increased odds and those 55–64 are almost 9 times more likely to have a good job than those aged 18–24 (p < 0.001). Similarly, each increased level of education is associated with an increased odds of having a good job: high school graduates have 3.2 times the odds of holding a good job compared to those with less than a high school diploma, those with some college have 6 times the odds and those with at least a college degree have 13.2 times the odds of holding a good job compared to someone who did not complete high school.

The results for part-time workers (those working less than 35 hours per week) are shown in Table 4. While the role of race, age, and health status appear to be similar in predicting the likelihood of holding a good job for part-time workers, these models differ from those presented for full-time workers in Table 3 in a few respects. First, on the topic of our primary analytic interest, disability status is not an important predictor of holding a good part-time job whether we control for other sociodemographic characteristics or not. Second, the direction of the role of gender differs for part-time and full-time workers. While men had 53% increased odds of holding a good full-time job after controlling for all other factors (Model 3, Table 2), men have a 33% lowered odds of holding a good part-time job compared to women (Model 3, Table 3). One area in which the associations seen among full-time workers is similar though much more pronounced is with respect to educational attainment. Each additional increase in education results in substantially higher odds of holding a good part-time job. For instance, part-time workers with at least a college degree are over 24 times more likely to have a good job than those with less than a high school diploma (p < 0.001).

Table 5 takes the estimates that resulted from the final models (Models 3 in Tables 3 and 4) to calculate conditional margins. These margins are the average predicted probability of having a good job if everyone in the sample had the characteristics of a typical worker with a disability. We vary only the disability indicator to examine its impact. The modal values of a worker with a disability include being aged 34 to 54, male, non-Hispanic white, in excellent or very good health, living in an urban area, and having some college education but no four year degree. This method shows an average predicted probability of having a good job as 0.39 for workers with characteristics of someone without a disability (p < 0.001) and a predicted probability of 0.36 for workers with the characteristics of a worker with a disability (p < 0.001).

The results shown identify some important associations with respect to disability and job quality as well as more general findings regarding job quality. First, disability was significantly associated with decreased odds of having a good quality job, until we controlled for self-reported health status. In that respect, our first research hypothesis is not confirmed. The public often conflates disability with health status although they are distinct concepts. A person may have a disability and still have very good health; in fact, the modal category in our sample was excellent or very good health (43.6% among those with a disability). While workers with a disability may be more likely to experience poor health (2.6% compared to just 0.5% in poor health among workers with no disability), it appears that health status, rather than disability, may be affecting the ability to obtain and retain a good quality job. Previous research addressing employee accommodations for health problems has found that the physical strenuousness of the job and the severity of the health problem are two factors that impact an employer’s ability to retain employees with health problems (Gould-Werth, Morrison, & Ben-Shalom, 2018). Innovative disability management and other workplace health-related programs that can help employers and employees navigate the challenges faced by that who are experiencing health problems can help to improve access and retention of “good jobs” by this population.

Second, persons in part-time jobs rarely hold positions that would qualify as “good jobs,” as we hypothesized in our second research hypothesis. Only 6.1% of part-time workers have a job that meets these criteria. The part-time workers who do have a good job are generally highly educated women who are not just starting out in their careers. This suggests that these positions might be good not just in the tangible economic metrics we examine, but perhaps also in that they require fewer hours of work – a benefit which is conducive to managing family demands. Other part-time workers may choose to sacrifice either pay or health/retirement benefits so that they can have flexible work or a job that is “good” in some non-instrumental manner. However, it is likely that some portion of part-time workers accept part-time jobs as a last resort when full-time work is not available. In addition, the fact that women are more likely to have a good part-time job than men may be due to men being less willing to accept part-time work. The data we use here is unable to assess the role of these non-economic factors.

Third, persons with higher levels of education are significantly more likely to hold high quality jobs. As persons with disabilities generally have lower levels of educational attainment than persons without disabilities (Lauer & Houtenville, 2018), this finding highlights the importance of efforts to ensure that persons with disabilities have equal access to primary, secondary and post-secondary educational opportunities. Maintaining funding for in-school supports for persons with disabilities is particularly important as access to such supports has been found to be associated with higher levels of educational attainment (U.S. Department of Education, 2016). In addition, the most common level of educational attainment found among workers with disabilities was some college, suggesting that education beyond high school leads to more engagement in employment. In contrast, 56 percent of persons with disabilities who were age 25 or older had only a high school education or less in 2016 (Lauer & Houtenville, 2018).

Fourth, for full-time workers, our results replicate those of others (Burchell et al., 2007; Ficapal-Cusi et al., 2018; McCall, 2001; Schmitt & Jones, 2012) who have found that women and persons who are not White, non-Hispanic are less likely to hold good jobs, while controlling for other characteristics. Further research can examine whether these findings are due to worker choice, discrimination, or other factors.

Fifth, while improving the quality of jobs has recently received international attention (Findlay et al., 2017), systemic changes in rates of pay, availability of health insurance, and provision of retirement savings programs by employers would be required to improve the proportion of workers holding “good jobs” per our quality measure. As union protection has lessened over the years (Kalleberg, 2011; Rosenfeld & Kleykamp, 2012), worker protections have eroded along with access to benefits such as health insurance and retirement plans. Wages for the average worker have also been fairly stagnant, particularly since the Great Recession (Autor, 2011). As Schmitt and Jones (2012) noted, “bad jobs” have been on the rise in recent decades. As workers face restricted choice in the quality of jobs they obtain, more and more workers will be looking to government programs for assistance in alleviating wage compression, finding affordable health care coverage, and saving for retirement. The ability of public programs to fully address these needs is of concern, particularly as policymakers seek to reduce funding for income and in-kind assistance programs, restructure public health insurance programs, and continue to grapple with diminished resources with which to fund retirement programs. Employers who are interested in attracting and retaining the best employees should therefore prioritize offering “good jobs”: paying more than median wages and offering employer-sponsored health insurance and a retirement plan.

6 Conclusion

Job quality is a complex concept. Using the objective economic criteria of above average pay and access to health insurance and retirement benefits, we found that full-time workers with a disability have a lower rate of job quality than those with no disability. However, after accounting for employee health status, disability was not a statistically significant predictor of holding a good quality full-time job. Other characteristics such as age, gender, and race, and human capital measures such as educational attainment, were also significantly associated with having a good job. A lower proportion of part-time workers, particularly young and less educated workers, held good quality jobs.

This work contributes to a growing body of research on job quality by considering the ways in which workers with a disability may be at risk for holding lower quality jobs. While the focus on economic measures is an important step, future research should consider data collection efforts that would allow for an examination of the role of disability on subjective and non-economic aspects of job quality. Social inclusion is one example of the former; teleworking and flexible work schedules are examples of the latter. Existing survey data with these measures generally do not have an indicator of disability status.

Conflict of interest

The authors have no conflict of interest to report.