Did Minimum Wage Policy Changes Impact Home Health Workforce?

Abstract

The shortage of home health aides has been exacerbated in recent years partially because of low wages. Minimum wage (MW) policy changes may alleviate this workforce shortage. This study examined the effects of MW policies on wages and employment of home health aides. We performed a county-level longitudinal analysis using 2012 to 2018 national data. The study cohort included 2,496 counties and focused on all workers in the home health industry. Outcome variables included wages and the employment of home health aides. Key variables of interest included the consumer price index adjusted state MW and a set of variables that captured the effect of the Fair Labor Standards Act (FLSA) extension. This study found that home health aides’ hourly wages were $1.00 higher (p = .011) in states that increased their MWs from below $8 to above $10. The FLSA extension was associated with $1.15 higher wages in states with higher MWs (i.e., state MW above $10 in 2014). The FLSA extension was associated with higher employment of home health aides in less-competitive markets, rather than high- or average-competitive markets. This study suggests that state MW increases combined with the FLSA extension may help maintain the current home health workforce and improve their wages.

Introduction

The shortage of home health aides has been a long-lasting issue. There are nearly 3.5 million job in the home health industry, with more than 15% unfilled in 2020.¹ This is a critical problem for the older population who require longterm care services and supports (LTSS), particularly those with physical and cognitive impairments. This population continues to grow rapidly, expecting to increase by over 40% between now and 2035.^2–4 Home health industry provides services to over 80% of the older population needing LTSS.⁵ Workers in the home health industry, commonly known as home health aides, provide basic care and essential support on daily activities to enable their customers with functional or activity limitations to live independently at home.⁶ The shortage of home health aides has been further exacerbated during the COVID-19 pandemic,^7,8 that may force out those who wish to get care at home to nursing homes and hospitals.^9,10 Low wages are one of the potential reasons for the shortage of home health aides.^11,12

The wages and employment of home health aides may be influenced by minimum wage (MW) policies. The federal MW law has extended its coverage to home health aides in the 2015 amendment to the Fair Labor Standards Act (FLSA), which established a federal MW rate and the scope of coverage.¹³ Before the 2015 FLSA amendment, home health aides were excluded from the MW and overtime pay protections at the federal level.¹⁴ This FLSA extension, also known as “Home Care Extension,” extends MW and overtime protections to all workers employed by home care agencies and other third parties.¹⁵ Although 15 states and Washington, DC had extended these protections prior to 2015, home health aides in the remaining 35 states gained the MW and overtime protections for the first time due to the 2015 FLSA extension.¹⁶ Additionally, many states have increased their MW rates in the last decade. The current federal MW of $7.25 per hour has been in effect since 2009.¹⁷ In the absence of federal-level action, 31 states have increased their MWs above the federal MW by 2020.¹⁸

Although some studies have examined the impact of MW policies on employment,¹⁹ there is very limited evidence on how the MW policies affect home health workforce in the US. A recent study found that state MW increases were associated with higher nursing assistants’ wages and employment in the US,²⁰ but it only examined the effects of state MW increases but did not consider the FLSA extension effects nor include home health aides in the study sample. Therefore, the objective of this study was to examine the effects of MW policies—state MW increases and the FLSA extension—on workers of the home health industry.

Methods

Data Sources

We obtained workforce data from the Quarterly Workforce Indicators (QWI), which track firm and worker characteristics over time, thus enabling longitudinal analyses.²¹ The QWI provided county-level payroll data on employment and earnings within the home health industry. The Department of Labor (DOL) documented the annual state-level MW and region-level consumer price index (CPI).²² The Agency for Healthcare Research and Quality (AHRQ) Social Determinants of Health (SDOH) Database creates linkable SDOH-focused data at the county-level, including hospital beds, population, and the median house price in a county.²³ The Home Health Compare (HHC) data contains characteristics of home health agencies. We linked these datasets mentioned above at the county-year level between 2012 and 2018.

Study Population

The study population included 2496 counties and all workers in the home health industry (NAICS 6216).²⁴ These workers, also known as home health aides, were non-professional home care workers who provided basic care and essential support on daily activities.⁶ The 2496 counties were from 50 states and Washington, DC. We excluded 642 counties because these counties had missing wages or employment data in QWI.

Measures

Outcome variables included employment and wages of home health aides. Employment was defined as the employment of home health aides per 1000 population in a county. Wages were defined as the average hourly wage of newly hired home health aides.²¹ We converted the quarterly measures in QWI into annual measures.

Key variables of interest included state-level MW rates between 2012 and 2018 and a set of variables capturing the influence of the FLSA extension in a state. We adjusted the state MWs using the 2018 regional (i.e., Northeast, Midwest, West, and South) CPI. To account for the threshold effect of MW policies, we categorized state-years into 3 groups according to the state MW rate in a particular year: (a) MW <$8 per hour; (b) MW ≥$8 and <$10 per hour; and (c) MW ≥ $10 per hour. We used $8 as the cutoff point because the nominal federal MW was about $8 in the baseline year (i.e., 2012) after adjusted CPI. We used $10 as another cutoff point because the average wage of the newly hired home health aides was about $10 in the study sample.

To explore the effect of the FLSA extension, we categorized states into 3 groups. The reference group included 14 states and Washington, DC (referred as FLSA-control states) that extended MW and overtime protections to home health aides before the baseline year.²⁵ We then divided the rest of the states (referred as FLSA-affected states), into 2 groups. These states extended MW and overtime protections to home health aides after 2015 (except that Hawaii extended MW protections to home health aides in 2013²⁶). One group included 35 states with CPI-adjusted state MWs less than $10 per hour in 2014 (referred as “FLSA low-wage” states). Another group included Oregon, with adjusted state MWs of more than $10 per hour in 2014 (referred as a “FLSA high-wage” state). We used $10 as the cutoff point because the average wages of the newly hired home health aides were about $10 in the study sample.

We also obtained a set of county-level time-varying covariates, including hospital beds per 1000 adults aged 65+, percent of adults aged 65+, and the median house price, and state-level covariates, including the Medicaid expansion policy and the average wages of registered nurses (RNs). These county- and state-level covariates may influence the needs for home health services or the wages of home health aides.

Analysis

The primary analyses were conducted at the county-year level. We first described the distributions of workforce measures, state MW, and other covariates at the baseline year. Then we fit several linear regressions with county and year fixed-effects to examine the impact of state MW increases and the FLSA extension on workforce outcomes, accounting for time-varying covariates. We clustered the error term at the state level because these MW policies were state-specific policy interventions.

We used a difference-in-differences approach with interactions between the FLSA extension-based state groups (i.e., FLSA-control states, FLSA low-wage states, and the FLSA high-wage state) and an indicator representing years after the FLSA extension. This interaction term captured changes in outcome variables from before to after the FLSA extension among the states affected versus states not affected by the FLSA extension. State fixed effects and post-year dummy were not included because county and year fixed effects had absorbed their effects. To check the parallel trend assumption for the difference-in-differences model, we tested whether the time trend differed between FLSA-control and FLSA-affected states prior to the FLSA extension by adding interactions between the FLSA extension-based state groups and each year dummies. In addition, to check the robustness of the findings, changed the cutoff points of the MW from $10 to $9 to test if the threshold effect of MW policies still existed. In addition, we excluded county-fixed effects but used county-random effects instead in regression models. We also added the FLSA extension-based state groups as another regressor to compare the differences between FLSA-affected and FLSA-control states on wages and employment across the study period.²⁷

Lastly, we stratified analyses by the market competition level (i.e., competitive, average, and less competitive) to examine whether and how the effect of minimum policies on employment varied among these markets. Market competition was calculated based on the Herfindahl–Hirschman Index (HHI). The HHI is defined according to the market share of home health agencies in a county in 2014.²³ We used 0.15 and 0.25 as cutoff points as suggested by the literature.²⁸ The market competition in this study reflected the competition between employers in hiring workers. We did not account for the competition between workers in searching for jobs since the shortage of home health aides would make this competition much smaller than the competition between employers.

All analyses were performed using SAS 9.4 (SAS Institute Inc.) and STATA 17 (StataCorp LLC. College Station, TX, USA) This study has been reviewed and approved by the University of Rochester Research Subjects Review Board. All authors have no conflicts of interest.

Results

Descriptive Analyses

At the base-line year (i.e., 2012), the mean hourly wage of newly hired home health aides was $10.0, the employment of home health aides was 3.7 per 1,000 population, and the mean state MW was $8.7 per hour (shown in Table 1). Figure 1 shows changes in employment and wages of home health aides during the study period. While there was an increasing trend in hourly wages for newly hired home health aides over the years, the employment was relatively stable.

Table 1.

Distribution of Outcome Variables, Key Variables, and Covariates by State Groups, CY 2012.

	State group
	FLSA-control sates	FLSA-affected states
	Mean (SD)	Mean (SD)	p-Value	All
Number of counties (%)	607 (24.32)	1889 (75.68)		2496 (100.00)
Outcome variables
Newly hired home health aides’ average hourly wage ($)	10.02	10.03	.982	10.03
SD	(7.18)	(5.35)		(5.75)
Home health aides employment per 1000 population	2.90	3.88	.016	3.70
SD	(6.31)	(6.33)		(6.34)
Key variable
State minimum wages adjusted for CPI (2018 $)	8.85	8.62	<.001	8.66
SD	(0.71)	(0.45)		(0.52)
Covariates
Hospital beds Per 1000 65+ population	19.44	20.94	.242	20.65
SD	(20.02)	(25.68)		(24.67)
Percent of 65+ population	15.74	15.42	.111	15.48
SD	(3.85)	(3.96)		(3.94)
Median house price ($10K)	17.57	12.99	<.001	13.89
SD	(9.23)	(7.97)		(8.43)
Registered nurses’ wages ($)	32.62	29.45	<.001	30.08
SD	(2.52)	(3.70)		(3.72)
Market competition in 2014	%	%		%
Competitive	14.47	9.58		10.51
Average	9.50	9.11	.008	9.16
Monopolistic	76.03	81.31		80.33

Note. FLSA represents the 2015 Fair Labor Standards Act amendment which extends minimum wage and overtime protections to all home health aides employed by home care agencies and other third parties at the federal level. FLSA-control states included 14 states and Washington DC that have extended state minimum wage and overtime protections to DCWs before 2012. FLSA-affected states included the rest of the states, which extended minimum wage and overtime protections to DCWs after the 2015 FLSA (except that Hawaii extended minimum wage protection to home health workers in 2013).

Figure 1.

Home health industry workforce factors changes among state groups, from 2012 to 2018.

Note. FLSA represents the 2015 Fair Labor Standards Act amendment which extends minimum wage and overtime protections to all home health aides employed by home care agencies and other third parties at the federal level. FLSA-control states included 14 states and Washington DC that have extended state minimum wage and overtime protections to DCWs before 2012. FLSA-affected states included the rest of the states, which extended minimum wage and overtime protections to DCWs after the 2015 FLSA (except that Hawaii extended minimum wage protection to home health workers in 2013).

On average, home health aides in FLSA-affected states had similar wages ($10.03 vs $10.02, p = .98) compared to those in FLSA-control states in 2012, but the employment of home health aides was higher in FLSA-affected states than that in FLSA-control states in 2012 (3.88 vs 2.90 per 1,000 population, p = .016). FLSA-affected states had lower state MWs than FLSA-control states ($8.62 vs $8.85, p < .001) in 2012. Additionally, FLSA-affected states had a lower median house price (129.9k vs 175.7k, p < .001) and lower wages of registered nurses ($29.45 vs $32.62, p < .001) than FLSA-control states.

Regression Analyses

Table 2 presents the results of linear regressions. After adjusting for the county- and state-level covariates, both state MW increases and the FLSA extension were associated with higher wages among the newly hired home health aides. Specifically, states that increased MWs from below $8 to above $10 per hour were associated with $1.00 higher wages (p = .011). In the FLSA-high wage sate, the FLSA extension was associated with $1.15 higher wages (p < .001) compared to FLSA-control states. However, neither state MW increases nor the FLSA extension were associated with the employment of home health aides.

Table 2.

Minimum Wage (MW) Policy Effects on Home Health Workforce, CY 2012 to 2018.

	Newly hired home health aides’ hourly wages	Home health aides employment per 1k
	Coefficient (SE)	Coefficient (SE)
State MW (Ref: State MW <$8)
State MW at ($8 and $10)	0.613 (0.416)	0.0124 (0.105)
State MW ≥ $10	1.004* (0.394)	−0.00373 (0.135)
FLSA states × years after extension (Ref: FLSA control states)
FLSA states 2014 MW < $10 × post	−0.138 (0.372)	−0.0787 (0.206)
FLSA states 2014 MW ≥ $10 × post	1.154** (0.353)	0.233 (0.203)
County-level covariates	Yes	Yes
State-level covariates	Yes	Yes
County fixed effects	Yes	Yes
Year fixed effects	Yes	Yes
Number of counties	2496	2496

Note. County-level covariates included hospital beds, percentage of the older population, and house price. State-level covariates included Medicaid expansion policy and average wages of registered nurses. Full models were shown in Supplemental Table A1.

^*p < .05.

^**p < .001.

The findings from the sensitivity analysis on changing the cutoff point from $10 to $ 9 per hour are presented in Supplemental Table A4. The effects of the 2 MW policies on home health workforce outcomes were insignificant when we changed the cut off. There was no statistically significant difference in the time trend between FLSA-control states and FLSA-affected states prior to the FLSA extension (shown in Supplemental Figure A1). The findings from the sensitivity analysis on adding the direct effects of FLSA-affected state indicators were consistent with the main results (shown in Supplemental Table A5).

Stratified Analyses

We first compared the wages for newly hired home health aides and employment of home health aides by market competition level in the baseline year (i.e., 2012). We found that wages for newly hired home health aides were lower in less-competitive markets than in average and more competitive markets ($9.8, $10.5, and $11.0, respectively, p < .001). The employment of home health aides was lower among less-competitive markets than in average and competitive markets (3.2, 3.6, and 5.8 per 1000 population, respectively, p < .001). Stratified regression analyses suggested that the effects of the 2 MW policies varied across different markets. Specifically, state MW increases and the FLSA extension were associated with higher wages only among less-competitive markets but not among competitive or average markets (shown in Supplemental Table A2). The FLSA extension was associated with higher employment of home health aides only among less-competitive markets but not among competitive or average markets (shown in Supplemental Table A3).

Discussion

This study examined effects of state MW increases and the FLSA extension on home health aides’ wages and employment. We found that a larger increase in the state MW (from less than $8 to above $10) and the FLSA extension in the high-MW state was associated with higher wages, and such relationship was mainly driven by less competitive markets. In addition, FLSA extension appeared to have been related to higher employment of home health aides in less competitive market but not in other markets.

Consistent with studies in other long-term care industries,^19,20 we found that state MW increases and the FLSA extension were associated with higher wages for home health aides. However, this relationship became significant only when there was a relatively large wage increase (i.e., increase to above $10 from below $ 8 per hour). As suggested by a prior study,²⁹ MW policy changes may affect workforce outcomes of low-income workers only if wages rise to a high enough level. The average wages of home health aides in this study were about $ 10 per hour in 2012. Thus, MW policies may affect the entire industry only when the MW rates are close to workers’ average wages.

Moreover, stratified analyses suggested that the relationship between MW policies and home health aides’ wages were driven by the findings from less-competitive markets. One potential reason is that home health aides in less-competitive markets generally have lower wages. Thus, these workers are more likely to gain wage increases after MW policy changes than workers in more competitive markets where wages were already higher. Indeed, in our study sample, home health aides in less-competitive markets had lower wages than those in more competitive markets.

The increase in MW did not appear to be associated with employment of home health aides. As suggested by economic models, MW policy may not influence employment when there is a labor supply shortage.³⁰ A recent study found that state MW increases were not associated with higher utilization of paid home care services among older residents with dementia.³¹ MW policies would not increase the use of home care if the policy did not increase employment. Given the market shortage of home health aides, home health employers may have to recruit workers from other industries. In this case, employers need to pay a higher wage than the worker’s current wage to motivate them to change jobs because there are costs associated with changing jobs, such as spending time searching for jobs, training to gain new skills, etc. As a results, state MW increases may not increase the labor supply if the increased wages cannot cover the costs associated with changing jobs. Additionally, modest wage increases after MW policy changes may not provide strong enough incentives to workers in other industries markets. In less-competitive markets, especially the market switching to the home health industry.

However, we found that the FLSA extension was associated with higher employment of home health aides in less-competitive markets, rather than high- or average-competitive markets. In less-competitive markets, especially the market with very few home health employers, a home health aide who wants to a higher wage job may have to switch to other industries due to the lack of availability of home health jobs. With the FLSA extension, home health aides were covered by the MW protection, and the wage gaps between the home health industry and other competing industries diminished. Thus, although the FLSA extension may not increase the recruitment of new workers, the policy may help retain current workers since workers earn higher wages after the FLSA extension.

This study had several limitations. First, although we have controlled for a set of time-varying county-level characteristics in the analyses and county fixed-effects in the model, there may still be time-varying unobserved factors (e.g., labor supply and demand for home health aides) that may influence the home health industry workforce. Second, our sample included only 1 high-wage state—Oregon. However, we could still detect a significant effect of the FLSA extension by including the state in our difference-in-differences analysis. Third, in some states, within-state variations in MWs may exist, and we did not account for these variations. However, the state-level minimum wage policy would still apply in all areas of the state. Lastly, QWI data did not provide information about working hours. The MW policy could influence working hours without significantly affecting employment.

Despite these limitations, this is the first study that used national data to examine effects of state MW increases and the FLSA extension on home health aides. In addition, we examined the threshold effect of MW policies and how the impact of MW policies varied across different markets, which could inform decision-makers based on their local conditions.

Implications

Given the high labor demand and workforce shortage, it is important to retain current home health aides, especially after the pandemic of COVID-19. On the one hand, home health aides played a central role in providing care and have reported increased volume and intensity of tasks during the pandemic.^32,33 On the other hand, the pandemic may have exacerbated the shortage of home health aides since many workers have concerns about being infected. Moreover, there are concerns that hospitals and other employers hired away home health aides with better pay and benefits during the pandemic,^34,35 which has persisted leaving many positions unfilled. Findings of this study suggest that increasing the wages of home health aides and reducing the wage gap between the home health industry and other competing industries could help to retain home health aides. MW policy alone may not significantly improve the home health workforce. Endeavors that provide skills training, value-based payment, and career advancement opportunities are needed to strengthen this workforce.³⁶ Medicare and Medicaid home and community-based services plans may consider increasing the reimbursement rate for home health services to increase the wages of home health aides and retain these workers. Home health agencies also need to take positive adaptations and responses to MW policies, for example, improving the working environment, developing curriculums on career development, adjusting work-flow, and strengthening coordination.³⁷