The Hidden Costs of Disability

January 10, 2025

The Hidden Costs of Disability

Zofsha Merchant, Erin Troland, and Douglas Webber¹

1. Introduction

Disability can be devastating financially for households, as it can severely limit earnings potential (Meyer & Mok, 2019; Benito, Glassman, & Hiedemann, 2016; Jolly, 2013). The earnings penalty for households with disabilities is estimated to range from 15 to 70 percent of earnings, depending on the nature of the disability (Meyer & Mok, 2019). Households receiving disability insurance (either employer-based or via Social Security) receive payments based on their prior earnings to help account for this earnings penalty. However, less well-studied is how disability may affect households' financial well-being through more hidden channels: those other than the direct earnings penalty. We analyze the effect of disability on overall financial well-being, controlling for the direct earnings penalty and a wide array of other financial metrics. This method helps us estimate the financial impact of disability separate from the more commonly studied earnings penalty because we can control for differences in earnings across households with and without disabilities. Subjective financial well-being includes households' assessment of other components of their finances, such as expenses and ability to build wealth.

We present and empirically analyze a conceptual framework with several mechanisms by which disability negatively impacts household finances through channels other than the earnings penalty for the disabled individual. For example, one channel is unpaid household labor. An adult with a disability may be limited in their ability to contribute to household production. As a result, the household might spend more on prepared food or household cleaning services. All else equal (crucially, including paid labor income), households without such limitations have higher capacity for unpaid labor and may spend less on such services. Given these disparities, as well as gender norms around caregiving activities, we also investigate whether the disability earnings penalty or one of the "hidden costs" varies by the gender of the household head in coupled households, which we define as the wage earner.²

To analyze the full set of financial costs of disability, we use Survey of Household Economics and Decisionmaking (SHED) data from 2019-2023. Our main sample includes cross-sectional data for around 2,400 households. We examine households with the following characteristics: at least two adults (either spouses or partners), single-income, and with the respondent of prime working age (25-54). We also limit our sample to heterosexual couples for measurement purposes for our analysis of female sole-earner households versus male sole-earner households. We focus on households with disabilities in which one adult is not in the labor force due to disability or health limitation (a "work limitation" definition of disability). In around one-fourth of the households in our sample, the adult not in the labor force reports not working because of a disability or health limitation.³ We also examine a sample of almost 11,000 individuals in one-adult households. We measure financial well-being as the share of respondents who report doing at least okay financially, representing the top two of four categories: "doing okay" or "living comfortably." We run regressions of self-reported financial well-being on an indicator for disability. We compare the "financial well-being penalty" across specifications of no controls and controls for demographics. To isolate the financial costs of disability on top of the direct earnings penalty, we then control directly for household income.

We find a sizeable role for the hidden financial costs of disability. Controlling for demographics and income alone does not eliminate the well-being gap between households with and without disabilities, suggesting a sizeable role for more hidden costs of disability that is over 40 percent of the overall gap. To give a sense of the scale of this penalty, for our main sample of single-earner, two-adult households with a spouse/partner, it is roughly equivalent to a drop in annual household income of $25,000. For one adult households, which are less able to spread the burden of disability across multiple people, this penalty is even larger.

We also examine how the gender of the sole earner affects financial well-being among heterosexual couples. First, we find that female sole earners are more likely to have a partner with a disability than male sole earners. Then, we explore the determinants of financial well-being gaps between these households, including income (gender wage gap) and disability. Without any controls, female-headed single earner households are around 9 percentage points less likely to be doing at least okay financially than male-headed single earner households. Unlike the other results from our analysis, this gap in financial well-being appears to be driven entirely by differences in income, i.e. the gender wage gap.

We contribute to our understanding of the impact of disability in two ways: (i) we quantify understudied financial costs of disability, and (ii) we explore a gendered component of these costs. First, we bring attention to various costs that often go unquantified and undiscussed in economic evaluations of disability, such as additional expenses and unpaid care. A large body of existing work examines a single cost of disability: the paid labor (earnings) penalty. Specifically, disabilities that affect an individuals' ability to work permanently reduce total labor earnings and hours worked (Stephens, 2001; Charles, 2003; Meyer and Mok, 2019; Jolly & Wagner, 2023). Households with severe or chronic disabilities face large earnings and consumption penalties (Meyer & Mok, 2019). Labor earnings volatility is higher and wealth accumulation is also lower for households with disabilities (Jolly & Wagner, 2023; Meyer & Mok, 2019).

However, fewer analyses consider other costs that disability imposes. Meyer & Mok (2019) find via time use surveys that disabled men spend relatively more time using medical services, sleeping, and watching TV rather than food preparation or home production. Spouses/partners can also be affected. Wives' reported time caring for their spouses rises by approximately 2-3 hours per week, which could come at the cost of fewer hours working (Lee, 2020). We build on prior work by analyzing additional, hidden financial costs of disability using subjective financial well-being such as reduced capacity for unpaid household labor and increased expenses (discussed further in the conceptual framework section). While we are unable measure these costs separately, it is useful to have a conceptual model of how they differ from one another, because efficient policies to address each would likely be very different. Further, future work that can separately measure these costs can use our methodology which can uncover estimates of the value of unpaid labor using subjective financial well-being. Thus, we contribute to the unpaid labor literature by introducing a new methodology for valuing unpaid labor.

Second, we explore a gendered component of how disability affects single-earner, two adult households. Much of the research on disability and work focuses on men and male household heads. We find that, among heterosexual couples, the negative financial consequences of disability are disproportionately concentrated households in which the woman works outside the home. Previous work examines other demographic characteristics of the impacts of disability. Earnings losses are greater, and recovery of wages is weaker for individuals with less education, who are nonwhite, or who are older (Charles, 2003; Prinz et al., 2018).

2. Measurement of Disability

Disability can be defined and measured in many ways. The Committee on the Rights of Persons with Disabilities defines individuals with disabilities as "those who have long-term physical, mental, intellectual or sensory impairment which in interaction with various barriers may hinder their full and effective participation in society on an equal basis with others" (United Nations, 2006). To measure or identify individuals with disabilities, surveys may try to measure the categorization or severity of the disability. For example, the American Community Survey categorizes disability into six broad areas: hearing, visual, cognitive, ambulatory, self-care, and independent living disabilities (Erickson et al., 2022). Based on their data, in 2022, 13.4 percent of all individuals in the US reported any type of disability. On the other hand, the 'severity' of a disability may depend on how chronic it is. Disability also may be measured by what activities an individual is able to do by themselves, such as working or caring for oneself. Ultimately, the measurement and identification of individuals with a disability largely depends on the purpose for using that data (United Nations Economic and Social Commission for Asia and the Pacific, n.d.).

For our analysis, we focus on disability which prevents individuals from working at a point in time, whether that be temporarily or permanently.⁴ For this type of disability, the earnings penalty is large, as the disabled individual does not earn any income at all. Moreover, a person who cannot work outside the home is likely limited in their ability to do unpaid household labor inside the home. For people with disabilities who are in the labor force, the full financial costs of disability may be lower than those in our sample.

3. Conceptual Framework and Methodology

In this paper, we use a new methodology to isolate the financial costs of disability aside from the direct earnings penalty. In particular, we examine single-earner, two-adult households where the non-earner reports a disability or health limitation as a reason for not working outside the home. A disability that prevents someone from working outside the home is also likely to limit their ability to work inside the home.

(i.) Conceptual Framework

Two-adult households with a disability generally face different financial challenges compared to other households in several key ways, all else equal. We classify the financial costs of disability into four categories. For households with disabilities:

1. Paid labor income is lower (earnings penalty)
2. Expenses to care for the disabled adult are higher (e.g. health care costs)
3. Care work for the non-disabled partner may be higher, reducing the time they can spend on other household production⁵
4. The capacity of the disabled partner to contribute to the household production function is reduced (unpaid household labor penalty)

In our conceptual framework, we aim to net out the effect of the disability earnings penalty on financial well-being (1) to measure the more hidden financial costs of disability coming from (2)-(4).

Theoretically, costs (1) and (2) have straightforward effects on financial well-being. Income is lower and expenses are higher, which reduces financial well-being. For costs (3) and (4), the effects involve how the household responds to having less capacity for household production, which can also lead to higher expenses. Both adults may have reduced capacity for household labor. The disabled adult's disability can limit their household labor capacity (cost 4). The partner may need to spend more time doing care work, reducing the time they can spend doing routine household labor (cost 3). As a result, the household may increase spending on substitutes for household labor, such as eating more meals out, child care, or a cleaning service. By using a subjective measure of financial wellbeing, we can capture the effect of reduced household labor capacity on a household's budget. We also note that these effects apply to single-adult households with disabilities, who would experience costs (1), (2), and (4), but not (3).

It is also possible that a household cannot afford to increase spending on substitutes for household labor, and instead, the quality of life goes down as such work is simply unable to be done. Individuals could certainly view this as a decline in financial well-being (since they are unable to purchase these basic needs/wants), but they could also conceptualize this as a decline in overall life satisfaction. For this reason, we also estimate the relationship between disability and life satisfaction.

(ii.) Methodology for Valuing Hidden Costs of Disability

We value these hidden financial costs with a methodology using subjective financial well-being and controlling directly for the earnings penalty. Existing literature on valuing disability's effects use subjective measures such as health satisfaction and overall life satisfaction. These studies find sizable effects on both with some evidence of adaptation over time (Braakmann 2014; Oswald and Powdthavee, 2008; Lucas, 2007; Ferre-i-Carbonelle and van Praag, 2001). We use this methodology because the hidden financial costs of disability are difficult to measure. Thus, instead of directly measuring these costs, we use a household's subjective financial well-being to capture all of these additional financial costs associated with disability.

In particular, the financial impact of disability on unpaid household labor is difficult to measure. Currently, unpaid household labor has been measured largely through time use surveys documenting the number of hours spend on household production, then placing a monetary value on those hours. There are three predominant valuation methods. The first is the global substitute method, which is the cost of hiring one individual (a general housekeeper) to do all of the unpaid work. The second is the specialist substitute method, which is the cost of hiring various individuals to perform tasks based on their specialization at their market wage rate. The third method is the potential earnings method where unpaid work is calculated based on the loss of earnings for the individual doing the housework. Here the amount of hours the individual is spending on unpaid work is multiplied by their market wage rate (Reddy, 2020). We introduce a new methodology for valuing unpaid labor-one using subjective financial well-being. Our subjective financial well-being methodology could be used when households differ in their capacity for household production. Differences between these households' subjective financial well-being can be, in part, attributed to this capacity difference, which can then be used to measure the value of household labor.

4. Data and Empirical Strategy

We use data from the Survey of Household Economics and Decisionmaking (SHED) from 2019-2023. This is a large, nationally representative survey of adults produced by annually by the Federal Reserve Board of Governors. Approximately 11,000 to 12,000 individuals take the survey each year.

The SHED is well-suited for this analysis because it not only asks respondents about employment and disability, but also general questions about their financial well-being and overall life satisfaction. Specifically, every respondent is asked the following subjective question on their financial well-being: "Overall, which one of the following best described how well you are managing financially these days?", with respondents choosing from finding it difficult to get by, just getting by, doing okay, or living comfortably. If respondents report any of the latter 2 options, they are coded as 'doing at least okay financially'. This question is critical for our analysis because it allows us to estimate the hidden cost of disability. Additionally, the life satisfaction scale measured in the SHED is a Likert scale of 1-10. To make this comparable to our financial well-being variable, the life satisfaction dependent variable is coded as a 1 for values 7-10, and 0 otherwise (this leads to nearly identical population means for both life satisfaction and financial well-being).

We examine households with disability as those single-earner, two adult households where one adult is not working due to a disability. Our sample consists of around 2,400 households. We focus our analysis on households with a prime-age (25-54) respondent and those where the two adults are either married or partners (as opposed to adult roommates or multi-generational households). Single-earner, two-adult households are defined as those households in which only one adult reports doing any work for pay or profit in the last month. To identify these households, we code households as single-earner households only if the respondent reports doing work and their partner is not doing any work for pay or profit in the last 12 months, or vice versa.⁶ To categorize male versus female-headed single earner households, we condition on the respondent's gender and their sexual orientation being heterosexual (the survey does not ask about the gender of household members besides the respondent).

In the first four years of our data (2019-2022), we know the employment status of both adults in the household, but only know the disability status of the respondent. If the respondent is employed and their spouse/partner is not, we do not have the reason why the spouse/partner is not employed for these years. As a result, in these years, only the person with a disability reports financial well-being. If respondents reported not working for pay or profit in the past month, they were asked what contributed to them not working. The answer choices always include health limitation/disability.⁷ Respondents could select multiple responses for not working. In our sample, 24 percent of individuals report that they are not working due to a disability. This is higher than other reports of overall rates of disability because our sample excludes dual and no earner households. Among those who report disability as a reason for not working, 40 percent (10 percent of the entire sample) report disability being the only reason for not working, citing no other reason. Through these questions, we were able to identify single-earner households where one household member was not working due to a disability or health limitation and their spouse or partner is working. The respondent with a disability would then answer the subjective financial well-being question.

To determine whether it mattered whether the person answering the financial well-being question was the person with a disability, we incorporated new data from the 2023 SHED. In these data, if respondents reported that their partner was not working, they were asked why their partner was not working (with the similar range of answer choices). Thus, we were also able to identify single earner households where the respondent or their partner was not working due to a disability and have subjective financial well-being data from the non-disabled partner. Through a series of robustness checks, we found that results are similar for respondents only with a disability (2019-2022 data) versus those for both respondents and their spouses/partners with a disability (2023 data).

Table 1 shows that households in our sample are relatively representative of U.S. prime-age adults generally. Race, education, geography, income, and financial well-being are similar in our two-adult household sample as the entire nationally representative SHED sample. As would be expected, a sample of all prime-age two-adult households is slightly older and more likely to have kids than the average household.⁸

Table 1: Descriptive Statistics of Full Sample versus Analysis Sample (Single-Earner Two Adult Households with a Spouse/Partner)

We use standard OLS regressions with year fixed effects to estimate the hidden financial costs of disability separate from the earnings penalty. We estimate the raw difference in well-being between households with and without disabilities, then add controls for demographics and income to isolate the more hidden costs.

5. Results

We find that the hidden costs of disability, aside from the earnings penalty, account for over 40 percent of the difference in financial well-being between households with and without disabilities that prevent being in the labor force. Our sample of single-earner, two-adult households with disabilities still have lower financial well-being even when controlling for differences in demographic factors and income (the earnings penalty). These households have consistently lower financial well-being across all specifications (Table 2). Unconditionally, households with a disability are 21 percentage points less likely to report doing at least okay financially (column 1). However, these differences could be driven by differences in demographic factors, rather than disability itself, such as age, education and race.⁹ Controlling for demographic factors, the well-being gap declines to 17 percentage points (column 2).

Table 2: Financial Well-Being Differences for Single-Earner Two Adult (Spouse/Partner) Households by Disability

To measure the more hidden costs of disability, we control directly for income (Table 2 column 3). The coefficient on disability in this regression represents the remainder of the financial costs of disability on top of the direct earnings penalty, assuming we can control for all factors that would generate differences in financial well-being between these groups besides disability. Controlling for income reduces the well-being gap by around half, however there is still a well-being gap of 9 percentage points.¹⁰ Finally, in column 4, we control for a measure of health care costs to account for increased health care costs, described as disability cost (2) in our conceptual framework. Ideally, we would have a direct measure of health care costs, but we proxy for this measure using an indicator for whether the respondent reported an unexpected, major medical expense in the past 12 months. Even when controlling for this measure of health care costs, households with disabilities still have lower financial well-being (a difference of 7 percentage points).

We also explored our disability cost (3) from the conceptual framework using a question that measures increased care work by the non-disabled partner. This question is available for 2023 only and asks respondents whether they regularly provide unpaid help or take care of an adult relative or friend who needs assistance due to aging, disability, or illness. While not shown in this table, adding this variable to the regression in Table 2 (demographic and income controls) makes the well-being gap decline to around 5 percentage points. However the sample size is cut to around a third of our main sample, so the results are underpowered.

Next, we examine the role of gender for the single-earner among heterosexual couples. Among heterosexual couples where one partner is not in the labor force, men are more likely to report not working due to disability/health limitation. Women are more likely to report not working because of care responsibilities. Therefore, working women are more likely to be the sole earners when one member of the couple is not working due to disability. Among female-headed heterosexual couples, 31 percent have a male partner with a work-limiting disability compared to 22 percent for their male-headed counterparts. Further, the gender of the household head matters for financial well-being. Among all households where one member of the couple is not working, regardless of the reason, there is a nearly 9 percentage point gender gap in financial well-being (Table 3, column 1). Given these disparities, as well as gender norms around caregiving activities, a natural question is whether disability adds any additional financial well-being penalty on top of the gender wage gap faced by female headed households. As Table 3 illustrates, this gender gap in financial well-being is explained almost entirely by household income disparities, or the gender wage gap (column 2).

Table 3: Financial Well-Being Differences for Single Earner Two-Adult Households (Spouse/Partner), by Gender of Wage Earner

Table 4 presents the results of analogous models to Table 2, except for households with only one individual. The negative impact of disability on financial well-being among these households is considerably larger across all specifications and illustrates the importance of separately estimating these models by household type. Despite the large differences in coefficients, it is notable that as a share of the gap (e.g. column 4 divided by column 1), the unexplained cost proportion is similar across household types. The total cost of disability is large and multifaceted, but to some extent it can be attenuated when shared across multiple people.

Table 4: Financial Well-Being Differences for One Adult Households by Disability

Finally, Table 5 presents the results for models in which the dependent variable is overall life satisfaction rather than financial well-being. Following our binary coding of life satisfaction as described above, we find that disability leads to a greater decline in overall life satisfaction across every specification. The remaining disability penalty after controls (column 4) for overall life satisfaction is larger in magnitude than the remaining penalty for financial well-being, suggesting that other components of disability affect life satisfaction. While not surprising, this is further evidence that the negative consequences of disability go far beyond direct financial impacts.

Table 5: Life Satisfaction Differences for Single-Earner Two Adult Households by Disability

Taken together, the above results emphasize the important role that the incidence of disability plays in household financial well-being over and above the impact that disability has on contemporaneous income and the other financial variables we are able to control for.

6. Conclusion

The impact of disability on households' financial situations can be devastating. A large literature is devoted to estimating the direct earnings losses caused by disability, but relatively little attention is paid to the other costs that households face. We produce an overall estimate of the non-direct, "hidden" costs that disability imposes on households, which includes the value of unpaid household labor and the cost of normal household labor that is outsourced due to the disability.

While we are unable to disentangle the various drivers of the financial well-being penalty that we estimate in this study, which surely vary significantly across households, the magnitude of the overall "hidden costs" penalty is large. Just as disability can impact many facets of an individual's life, the financial impact is similarly broad-based and goes beyond the direct loss of income experienced by the disabled adult. Understanding exactly what drives this financial well-being penalty will be an important contribution of future work, as the design of effective policies would look very different depending on the source of these hidden costs. For instance, a policy aimed at addressing costs of care for the disabled could have relatively little impact if the primary cause of financial instability is instead the high cost of contracting out household production activities (e.g. cooking, errands, etc). Regardless of the root causes, the shifting age distribution of the U.S. population will likely lead to an aggregate increase in the type of costs discussed in this note, and thus there will be an increased need to understand them.

References

Benito, S. G., Glassman, T. S., & Hiedemann, B. G. (2016). Disability and labor market earnings: Hearing earnings gaps in the United States. Journal of Disability Policy Studies, 27(3), 178-188. https://doi.org/10.1177/1044207316658752

Braakmann, N. (2014). The consequences of own and spousal disability on labor market outcomes and subjective well-being: evidence from Germany. Review of Economics of the Household, 12, 717-736. https://doi.org/10.1007/s11150-012-9164-7

Burkhauser, R. V., Weathers II, R. R., & Schroeder, M. (2006). A guide to disability statistics from the Panel Study of Income Dynamics. Cornell University Technical Series paper (06-02).

Charles, K. K. (2003). The longitudinal structure of earnings losses among work-limited disabled workers. Journal of human Resources, 38(3), 618-646. DOI:

Erickson, S. R., Houseworth, J., & Esler, A. (2022). Factors associated with use of medication for behavioral challenges in adults with intellectual and developmental disability. Research in Developmental Disabilities, 123, 104182. https://doi.org/10.1016/j.ridd.2022.104182

Erickson, W., Lee, C., & von Scrader, S. (2022). 2022 Disability Status Report. Cornell University Yang Tan Institute on Employment and Disability. https://production-disabilitystatistics-org.s3.amazonaws.com/reports/2022-PDF/2022-StatusReport_US.pdf

Ferrant, G., Pesando, L. M., & Nowacka, K. (2014). Unpaid Care Work: The Missing Link. OECD Development Center.

Ferrer-i-Carbonell, A & van Praag, B. (2001). The Subjective Costs of Health Losses due to Chronic Diseases: An Alternative Model Appraisal. IZA Discussion Papers, No. 313,

Goyat, R., Vyas, A., & Sambamoorthi, U. (2016). Racial/ethnic disparities in disability prevalence. Journal of Racial and Ethnic Health Disparities, 3, 635-645. https://doi.org/10.1007/s40615-015-0182-z

Jolly, N. A. (2013). The impact of work-limiting disabilities on earnings and income mobility. Applied Economics, 45(36), 5104-5118. https://doi.org/10.1080/00036846.2013.818212

Jolly, N. A., & Theodoropoulos, N. (2023). Health shocks and spousal labor supply: an international perspective. Journal of Population Economics, 36(2), 973-1004.

Jolly, N. A., & Wagner, K. L. (2023). Work-limiting disabilities and earnings volatility. Labour Economics, 81, 102333. https://doi.org/10.1016/j.labeco.2023.102333

Lee, S. (2020). Household responses to disability shocks: Spousal labor supply, caregiving, and disability insurance. Canadian Labor Economics Forum Working Paper Series, (No. 21).

Lucas, R. E. (2007). Long-term disability is associated with lasting changes in subjective well-being: Evidence from two nationally representative longitudinal studies. Journal of Personality and Social Psychology, 92(4), 717-730.

Meyer, D., & Mok, W. (2019). Disability, earnings, income and consumption. Journal of Public Economics, 171, 51-69. https://doi.org/10.1016/j.jpubeco.2018.06.011

Oswald, A. J., & Powdthavee, N. (2008). Does happiness adapt? A longitudinal study of disability with implications for economists and judges. Journal of public economics, 92(5-6), 1061-1077. https://doi.org/10.1016/j.jpubeco.2008.01.002

Prinz, D., Chernew, M., Cutler, D., & Frakt, A. (2018). Health and economic activity over the lifecycle: Literature review. NBER Working Paper Series (24865). DOI: 10.3386/w24865

Reddy, S. (2020). Measurement Brief: Unpaid Work (PDF). University of Maryland WEDGE Program.

Stephens, M. (2001). The long-run consumption effects of earnings shocks. Review of Economics and Statistics, 83(1), 28-36.

United Nations. (2006). Convention on the Rights of Persons with Disabilities. United Nations Human Rights Office of the High Commissioner.

United Nations Economic and Social Commission for Asia and the Pacific. (n.d.). Part One: Issues on Disability Measurement (PDF).

Public link

Please use the above public link if you want to share this noodl on another website.