Skip to main content
NCBI home page
NIHPA Author Manuscripts logoLink to NIHPA Author Manuscripts
. Author manuscript; available in PMC: 2024 Oct 1.
Published in final edited form as: J Marriage Fam. 2023 Oct;85(5):1110–1124. doi: 10.1111/jomf.12925

Unplanned Costs and Benefits: Gender and Spousal Spillover Effects of Retirement on Health

Lauren Newmyer 1,2,*, Kendal L Lowrey 1, Yuliana Levchenko 1
PMCID: PMC10798816  NIHMSID: NIHMS1906371  PMID: 38250186

Abstract

Objective:

Our study assesses how women and men’s health indicators are shaped by their spouse’s retirement.

Background:

The retirement process can reshape the health of a retiree, but these effects can also extend onto the health of spouses. Although past research has largely focused on how men’s retirement might negatively shape their wife’s health outcomes, it is possible that wives’ retirement has detrimental effects on their husband’s health as well.

Method:

Using data from the Health and Retirement Study (HRS), we employ a fuzzy regression discontinuity design to identify the causal effects of spousal retirement on indicators of physical and mental health in married older adults.

Results:

Our results suggested that men, not women, experience the most negative spousal spillover effects of retirement on their health outcomes. We found the most support for spillover effects on spouses’ physical health outcomes. Additionally, men who are not working when their spouse retires experienced the most negative health effects.

Conclusion:

Women and men’s health is differentially affected by spousal retirement, where men might be the most negatively affected by their spouses’ transition in the U.S. context. These results contradict conventional wisdom that undergirds numerous untested assumptions underlying prior research on this significant life transition.

Keywords: Marriage, Retirement, Aging, Older Adults, Health, Gender

Introduction

Retirement is an important life transition for many older adults. Research on this topic has steadily expanded as many populations shift to contain a growing and longer-lived older adult population (Bongaarts, 2004; Lee & Mason, 2010). As retirement frequently occurs at older ages (often corresponding with poor or declining health), previous studies have focused on the varied relationships between retirement and individuals’ health outcomes (Behncke, 2012; Coe & Zamarro, 2011; de Breij et al., 2020; Kim & Moen, 2001), particularly with respect to physical and mental health and the associated relationships with stress and health behaviors (Behncke, 2012; Belloni et al., 2016; Eibich, 2015; Müller & Shaikh, 2018; Tran & Zikos, 2019).

Given that the decision to retire is not always made alone and spouses tend to plan this transition together (Hospido & Zamarro, 2014; Szinovacz & Davey, 2005), it is important to observe how one spouse’s transition is linked to the other’s outcomes. Glen Elder’s (1998) perspective of “linked lives” contends that social connections play a significant role in decision-making and shaping life outcomes. Thus, the decision to retire not only affects the individual but also shapes the life of his/her spouse. Consequently, retirement is a joint transition that affects both spouses and must be examined as such to understand the outcomes of both parties involved (Kim & Moen, 2002). Therefore, “spousal spillover” effects are when the choices, behaviors, and experiences of one spouse affect the other. Although spousal spillover effects have been documented in multiple areas of health (Falba & Sindelar, 2008), less work has examined how a spouse’s retirement is linked to an individual’s health.

Spousal spillover effects of retirement on health have largely been overlooked in the U.S. The majority of past research has examined international contexts (Atalay & Zhu, 2018; Bertoni & Brunello, 2017; Müller & Shaikh, 2018; Zang, 2020) and has predominantly focused on women as they are assumed to experience the most health effects (Atalay & Zhu, 2018; Bertoni & Brunello, 2017; Johnson, 1984; Zang, 2020). This past research has discovered inconsistent results on whether a partner’s retirement is harmful to the health of his/her spouse, and has largely focused on health behaviors (Müller & Shaikh, 2018; Zang, 2020) and mental health (Bertoni & Brunello, 2017; Zang, 2020). The United States differs widely from these contexts in access to and utilization of healthcare and healthcare expenditures (Kumar & Ozdamar, 2004), profile of older adult health (Verdery et al., 2021), and in the retirement process itself (OECD, 2007). For that reason, studies that use international samples of adults may not be generalizable to older, married Americans, thus pushing the need for more research in various contexts on older adults (Veron & Council, 2001). Given that both men and women have high rates of formal workforce participation within the U.S. context and decisions to retire are often made jointly for married couples, it is important to consider spousal spillover effects on both men and women from retirement as a determinant of older adult health.

The current study investigates the spousal spillover effects of retirement on health in the context of the United States using the Health and Retirement Study (HRS), leveraging a fuzzy regression discontinuity design. In this study, we focus on indicators of physical and mental health to investigate how the retirement of a spouse might affect the other spouse’s general health. We also investigate how the retirement status of the other spouse buffers or encourages the spillover effects from his/her spouse’s retirement. This study highlights the implications a spouse’s retirement might have on the other’s health and how these effects are uniquely gendered. Implications from this study highlight the need to consider spousal health, in addition to individual health, during and after retirement to fully capture the health needs of older adults.

Background

Marriage and Health

There is a large body of research that has found that married individuals experience better health outcomes than those who are not (Koball et al., 2010; Lillard & Waite, 1995; Margolis et al., 2022; Rendall et al., 2011; Uecker, 2012; Waite, 1995). Research has linked marriage to positive benefits for life expectancy (Lillard & Panis, 1996; Lillard & Waite, 1995; Rendall et al., 2011), mental health (Barrett, 2000; Frech & Williams, 2007; Margolis et al., 2022), and chronic conditions (Hughes & Waite, 2009; Zhang & Hayward, 2006). However, marriage might not always hold just benefits for health, and instead it might contain a combination of both protective and negative advantages (Mikucka et al., 2021). For example, after marriage people sometimes reduce physical activity (Nomaguchi & Bianchi, 2004) or gain weight (Teachman, 2016) which could lead to negative health outcomes.

Marriage and health are associated due to multiple factors such as spousal influence (Cobb et al., 2016; Reczek et al., 2020; Umberson, 1992; Wilson, 2002), good health behaviors (Averett et al., 2013; Koball et al., 2010; Uecker, 2012), and high economic security (Killewald, 2013). Research has also found that marital health advantages can sometimes be explained by selection (Kane, 2016; Lipowicz, 2014; Waldron et al., 1996) or that accounting for selection reduces the size of health benefits from marriage (Buckles & Price, 2013). Furthermore, positive health benefits might be more modest in recent cohorts than in the past (Tumin, 2018). Health benefits of marriage might also be most likely for couples who experience continuous marriage (Tumin & Zheng, 2018).

Men often experience more health advantages of marriage than women (Rendall et al., 2011; Waite, 1995). Due to gendered inequalities surrounding caretaking, women often serve as the caretakers of their families which can include monitoring diets as well as encouraging healthy behaviors (Brown et al., 2016; Courtenay, 2000; Rook et al., 2011). Additionally, women often serve as confidants for their spouse and are also the most affected by their partner’s stress (Behler et al., 2019). The ways men and women deal with their emotional stress is highly gendered with the latter more likely to internalize their stress, and the former more likely to act on these emotions through behaviors such as drinking (Rosenfield et al., 2005). These factors contribute to the environment of marriage being particularly beneficial for men in comparison to women as their health might improve through the benefits of marriage.

Marriage and Retirement

The retirement process might reshape marriage, as retirement does not just impact one individual, but the marital union as a whole (Kim & Moen, 2002). Retirement is a major transition for individuals, as well as for their partners, because it involves the exit from one role to another which sometimes causes individuals great distress (Ebaugh, 1988). Role theory suggests that the way individuals live their daily lives is shaped by the roles they fill (Linton, 1936; Mead, 1934). Because a career often plays a significant role in shaping a person’s life, the transition to retirement can be complicated by a feeling of a lack of identity when an individual is no longer in the workforce (Silver & Williams, 2018). The transition to retirement along with the stress that may occur can extend onto the marital relationship itself, which might change marital stability (Davey & Szinovacz, 2004; Stancanelli, 2014).

When spouses retires, couples might be pushed to restructure their lives to contend with their spouse’s transition. It is possible that gendered caretaking roles might continue after a spouse’s retirement, which could improve the health of men. However, it is possible that spousal retirement could cause economic issues for the couple which might result in poor health outcomes for one or both spouses. We will explore these two divergent pathways further below.

The Retired Husband Syndrome

In the 1980s, a doctor documented a phenomenon he coined the term “Retired Husband Syndrome” (Johnson, 1984). He argued that his married female patients complained about their health declining post their husband’s retirement. They claimed that not only had their gendered caretaking and homemaking continued, but that it also increased. In contrast, their husband’s health was unaffected or improved as they experienced more benefits from their wife’s caretaking post-retirement. This proposition has been tested in contexts such as Japan (Bertoni & Brunello, 2017), which found support for negative mental health effects on women post their husband’s retirement. However, in China, these deleterious effects were largely not observed in women post their husband’s retirement (Zang, 2020). These mixed results suggest that the “Retired Husband Syndrome” might be context dependent and driven by differences in gender roles and family structures.

Under this framework, women’s health might suffer when their husband retires as they unequally bear the increased burden of the caretaking of their home, their spouse, and themselves. Additionally, they may find their mental and physical health declining due to increased stress from their retired husband, as they likely serve as his main confidant (Cutrona, 1996) and are likely to experience their spouse’s stress through spousal stress transmission, or shared stress (Behler et al., 2019; Siegel et al., 2004; Tower & Kasl, 1996). In sum, this pathway suggests that a wife’s health will decrease when her husband retires, and her husband’s health will either be unchanged or improved.

The Family Stress Model

However, it is also possible that retirement shapes spousal health through shared economic stress. The family stress model proposes that economic issues can cause strain on the marital relationship (Conger et al., 1994; Conger et al., 1999; Masarik et al., 2016). Although the original theoretical framework discussed macro-level economic factors, research has consistently shown that economic strain within families can shape negative health and life outcomes (Aytaç & Rankin, 2009; Kwon et al., 2003; Neppl et al., 2016). This economic strain can increase emotional distress (Kinnunen & Pulkkinen, 1998; Neppl et al., 2016). Additionally, economic problems can increase marital problems for couples (Aytaç & Rankin, 2009; Neppl et al., 2016), which might also shape negative health outcomes.

Men may be the most likely to experience negative health outcomes due to spousal retirement as economic hardship might manifest as more damaging for men in contrast to women. This pressure might manifest as negative mental and physical health symptoms. In contrast, women might feel less responsible for economic strain when her husband retires, as men have been traditionally viewed as the breadwinner and, thus, the wife perceives her income as more supplementary than essential. These factors might lead to spousal retirement being the most damaging to husbands’ health.

Using data from the Health and Retirement Study, we adopt a fuzzy regression discontinuity design and consider how spousal spillover effects from retirement shape indicators of physical and mental health. Unlike past studies that evaluated spousal spillover effects, we assess these outcomes for both men and women, instead of limiting our analysis to just one spouse (Atalay & Zhu, 2018; Bertoni & Brunello, 2017; Zang, 2020). By examining both genders, we seek to understand whether spousal spillover effects are more prevalent for men or women and how these results fit into these existing theoretical frameworks.

Methods

Data

Our study uses data from the Health and Retirement Study (HRS) (Health & Retirement Study, 2020). The HRS provides a representative sample of about 20,000 non-institutionalized older adults in the United States. Data collection began in 1992 and is ongoing. These data are uniquely equipped to study spousal effects on health because respondents and spouses are interviewed in each wave when the spouse is available. The addition of these spousal interviews increases the sample to be over 20,000 when spouses are included in analyses. In our analysis, we use data from 1998–2016 that are from the 2016 RAND Longitudinal file.

To assess the effects of retirement on one’s spouse, we implement strict sample exclusion criteria. As we are concerned with spousal spillover effects among older adults, our sample is limited only to respondents who are 50 years or older in the HRS with interviews that include spousal information. Due to the small sample of same-sex couples in the HRS, we remove these couples from our analytical sample. This yields an analytical sample of 24,776 married respondents from Wave 4 (1998) to Wave 13 (2016). This sample includes both dual and single-earning households. Appendix A contains more details on the creation of the analytical sample.

Measures

Dependent Variables

Physical health.

We measure physical health through measures of functional limitations and being obese. Functional limitations are assessed in two ways: Activities of Daily Living (ADLs) and Instrumental Activities of Daily Living (IADLs). The presence of ADLs is constructed as an index measure of whether a respondent has difficulty with bathing, getting dressed, eating, getting in and out of bed, or walking across a room. The presence of IADLs is similarly created as an index measure of activities that people have difficulty completing, including using the phone, managing money, taking medications, shopping for groceries, and preparing hot food. These measures are coded on a scale from zero to five based on respondents’ number of identified limitations. RAND calculates participants BMIs by dividing respondent’s weight (kilograms) by the square of their height (meters). We exclude biologically implausible BMI values <12 or >70 following past guidelines (Cheng et al., 2016). Respondent’s scores are then converted to either being obese (30.0 and above) or not (CDC, 2021).

Mental health.

For mental health, we assess the Center for Epidemiologic Studies Depression (CES-D) scale (Radloff, 1977). The CES-D score is calculated by the sum of six “negative” questions and the negation of two “positive” questions. The six questions that are negative ask respondents whether they experienced the following feelings most of the time: depression, everything is an effort, restless sleep, loneliness, sadness, and could not find the effort to get going. This score is then subtracted by the responses to two positive questions that assess a respondent’s feelings of happiness and whether he/she generally enjoyed his/her life. We use this score as a scale from zero (no symptoms) to eight (most depressive symptoms).

Independent Variables

Retirement status.

Respondents are asked to respond whether they consider themselves to be completely retired, partially retired, not retired at all, or if the question is irrelevant in each wave of the HRS. The HRS did not give a definition of what they consider each complete versus partial retirement. Although there is slight variation across the years in how HRS coded the irrelevant category which might influence the responses, it is consistently defined as “doesn’t work for pay or is a homemaker.” This measure allows us to identify when a respondent or his/her spouse, completely retires from the workforce. We use this variable to construct our treatment status variable (a spouse’s complete retirement). We use complete retirement as we hypothesize that completely retired spouses would spend more time at home and in contact with his wife or her husband. Additionally, we create a variable for a spouse’s own retirement status as either working or not in relation to when his/her spouse retires.

Control Variables

We control for several sociodemographic factors in our analysis. These variables include the respondent’s age, age squared, level of educational attainment, racial identity, and the year of observation. Educational attainment is measured as less than high school, GED or high school graduate, some college, or college and above. See Appendix B for sample descriptive statistics.

Analytic Strategy

In our study, we leverage an experimental design using the fuzzy regression discontinuity method. Regression discontinuity methods allow for the ability to compare similar individuals before and after a cutoff point, which in our study is a spouse’s retirement. The model accomplishes this by acting as if some respondents received a “treatment” while others did not. The model compares these individuals to determine any differences from the treatment, while also accounting for individual-level characteristics. This unique modeling approach allows us to identify the causal effects on health from spousal retirement while also controlling for endogenous individual-level characteristics; traditional regression model approaches are unsuitable for this research question as they may overestimate spillover effects as they are unable to fully account for individual-level characteristics (Carpenter & Dobkin, 2009; Imbens & Lemieux, 2008). Additionally, the method allows us to specifically examine the period before and right after a spouse’s retirement, as examining a longer period afterward might be biased as health might also be shaped by other factors such as declining health, Medicare adoption, or a spouse’s mortality. Past research that has assessed spousal spillover effects of retirement on health outcomes has adopted fuzzy regression discontinuity designs as well (Müller & Shaikh, 2018; Zang, 2020), of which we model our analysis similarly.

Regression discontinuity designs require the research to identify both a “treatment” variable and a “running variable.” The treatment (retired) and control (non-retired) groups are partially determined by the running variable – which in our study is the spouse’s age. Although sixty-five is the age most often associated with full retirement in the United States because it is the time when Social Security and private pension benefits are generally distributed, as well as when Medicare health coverage starts, many older adults choose to retire early. Although individuals might retire at any age, individuals often retire between the ages of 62 (the earliest age that Social Security benefits can be distributed) to 66 (full retirement age for individuals born between 1943 and 1954) due to the distribution of social security and other pension benefits (Social Security, 2021). In our study, we choose to focus on early retirement (which we classify as 62) due to the increase in the probability of retirement due to the distribution of social security benefits. Our data justify our decision, as our HRS sample has a natural increase in individuals who chose to retire early at 62. Additionally, we can disentangle the effects of Medicare adoption (often at age 65), by examining early retirement age (in this case age 62) as our treatment variable. In our research design, individuals right before and right after the cutoff (age 62) are assumed to be very similar, with retirement status as the sole distinguishing factor. However, some people over age 62 may remain in the workforce (noncompliers), which requires the use of a fuzzy regression discontinuity design. Although we understand that individuals may retire for a variety of reasons before the age of 62, these individuals are likely a select group and we, therefore, do not focus on them in our analysis. Descriptive statistics for compliers versus noncompliers are shown in Appendix C. For further technical details on the fuzzy regression discontinuity model see Appendix D.

A regression discontinuity design requires several theoretical assumptions to be met. We estimated the McCrary sorting test to check that age (our running variable) was not manipulated (McCrary, 2008), wherein we found no evidence of assumption violation. Next, we assessed whether there was a discontinuity in the probability of exposure at the cutoff point for both husbands and wives. Figures 1 and 2 show that we met this assumption for both wives and husbands, as there is a clear discontinuity in retirement rates at our cutoff point of age 62. Lastly, we found that although related to retirement status, the health outcomes were normally distributed across our sample.

Figure 1:

Figure 1:

Open in a new tab

Discontinuity Plot of Retirement Rates (Wives)

Note: Retirement age is normalized at the cutoff (62).

Figure 2:

Figure 2:

Open in a new tab

Discontinuity Plot of Retirement Rates (Husbands)

Note: Retirement age is normalized at the cutoff (62).

In Table 1 we present the results of our models which assess the effects of spouses’ retirement on their partner’s health outcomes by gender. We present the results from our sex-stratified fuzzy regression discontinuity models including conventional and bias-corrected models which correct for this overestimation. Each health outcome is estimated separately using a fuzzy regression discontinuity model. Although we present all estimates in our results section, we base significant findings on bias-corrected estimates, which account for possible undersmoothing (incorrect estimates of the regression line at the cutoff point) by adjusting standard errors. To diagnose the role of a spouse’s own retirement status in these spousal spillover effects, we stratify our models by both sex and retirement status (see Table 2 for wives’ outcomes and Table 3 for husbands’ outcomes). The retirement status of each respondent is recorded in the same wave his/her spouse retires and is used to distinguish spousal subgroups within each gender. These subgroups include non-working spouses, whether retired, not in the labor force, or jointly retired and working spouses, those who are still in the workforce. This analysis aims to capture how a spouse’s own retirement might also influence spousal spillover effects on his/her health outcomes. We conduct a sensitivity analysis for joint retirement, which shows its inclusion does not bias our results.

To assess the robustness of our regression discontinuity design, we also conduct a placebo test by assigning the treatment cutoff to an age when many people do not retire (at age 50 instead of age 62). As expected, the age 50 cutoff yields no significant results (results available upon request). As a robustness check we provide parametric results in Appendix E.

Results

Regression Analysis of Health Outcomes

We first present our sex-stratified analysis for our measures of physical and mental health, shown in Table 1. Overall, our results suggest that a spouse’s gender is an important factor for whether he/she experience spillover effects from his/her spouse’s retirement. Interestingly, we found that these spillover effects were most likely to be negative and prevalent among men in contrast to women.

Table 1:

Spousal Spillover Effects of Retirement by Sex on Health

Wife’s Outcomes Husband’s Outcomes
Conventional Bias-Corrected Number of Observations Conventional Bias-Corrected Number of Observations
Physical Health
ADLs −0.03 (.30) −0.07 (.30) 45,601 0.64* (.27) 0.85** (.27) 38,222
IADLs −0.24 (.26) −0.33 (.26) 45,596 0.51+ (.27) 0.67* (.27) 38,216
Obese 0.17 (.21) 0.16 (.21) 44,529 0.21 (.18) 0.31+ (.18) 38,127
Mental Health
CESD Score 1.10 (1.12) 1.74 (1.12) 43,697 0.68 (.64) 0.93 (.75) 33,753
Covariates Yes Yes Yes Yes
Open in a new tab
+

p<.10;

*

p<.05;

**

p<.01

Standard errors clustered by the individual in parentheses. Covariates include respondent’s education, race, age, age squared, and year.

For physical health measures, our results suggested no significant spillover effects for women. However, results showed multiple negative spillover effects for men on their physical health, particularly their functional limitations. We observed a significant increase in both the ADLs (0.85, p<.01) and IADLs (0.67, p<.01) measures of functional limitations. This is a substantial increase as the variable scale is from zero to five, which means this is almost a one point increase in functional limitations for men following their spouse’s retirement. However, our results did not suggest that one’s odds of becoming obese change for either men or women following their spouse’s retirement. Interestingly, our results implied no significant association between spousal retirement and changes in the CES-D scale for wives or husbands.

Analysis by Retirement Status

In Tables 2 and 3, we stratified our results by spouse’s own retirement status to better understand how this might influence spousal spillover effects on health from retirement. Table 2 shows the results for wives, with non-working wives’ results on the left and working wives’ results on the right. For non-working wives, our analysis shows no evidence of significant spousal spillover effects from retirement on our physical or mental health outcomes.

Table 2:

Spillover Effects by a Wife’s Retirement Status

Non-Working Wife Working Wife
Conventional Bias-Corrected Number of Observations Conventional Bias-Corrected Number of Observations
Physical Health
ADLs 0.32 (.409) 0.30 (.41) 28,696 −0.44 (.39) −0.58 (.39) 16,655
IADLs −0.03 (.335) −0.13 (.34) 28,619 −0.45 (.30) −0.62* (.30) 16,655
Obese 0.17 (.276) 0.04 (.28) 27,975 0.18 (.26) 0.35 (.26) 16,308
Mental Health
CES-D Score 0.88 (1.378) 1.22 (1.38) 27,334 1.13 (1.15) 1.90+ (1.15) 16,130
Covariates Yes Yes Yes Yes
Open in a new tab
+

p<.10;

*

p<.05;

**

p<.01

Standard errors clustered by the individual in parentheses. Covariates include respondent’s education, race, age, age squared, and year. Non-working wives include wives who are not in the formal workforce, retired wives, and those who jointly retired with their spouse.

Table 3:

Spillover Effects by Husbands’ Retirement Status

Non-Working Husband Working Husband
Conventional Bias-Corrected Number of Observations Conventional Bias-Corrected Number of Observations
Physical Health
ADLs 0.65* (.308) 0.78* (.31) 27,317 0.24 (.52) 0.42 (.52) 10,891
IADLs 0.49+ (.279) 0.60* (.28) 27,314 0.66 (.72) 1.13 (.72) 10,888
Obese 0.13 (.207) 0.24 (.21) 27,245 0.81 (1.00) 1.33 (1.00) 10,865
Mental Health
CES-D Score 0.35 (.628) 0.36 (.69) 23,818 0.36 (2.11) 1.47 (2.11) 9,957
Covariates Yes Yes Yes Yes
Open in a new tab
+

p<.10;

*

p<.05;

**

p<.01

Standard errors clustered by the individual in parentheses. Covariates include respondent’s education, race, age, age squared, and year. Non-working husbands include husbands who are not in the formal labor force, retired husbands, and those who jointly retired with their spouse.

However, our results showed that wives might be most susceptible to experiencing spousal spillover effects from retirement if they were still in the workforce. Across our measures of physical health, our results indicated a negative and significant effect for spousal retirement on a wife’s IADLs score. Results indicated that spousal retirement could decrease a working wives’ IADLs score by 0.62 points (p<.05) on a zero to five scale. This means that spousal retirement decreased the IADLs score of a working wife by about half a point which was sizeable given the overall average score for IADLs was 0.24 (Standard Deviation: 0.80) for wives. Again, results showed no significant effects for obesity or the CES-D score of working wives.

Table 3 is set up similarly as Table 2 for our results for husbands by their own retirement status. These results again highlighted that a spouse’s own retirement status might shape the effects of spousal spillover from retirement on health. Our results indicated the most spillover effects occur for a husband’s physical health when he was not working. A wife’s retirement was significantly associated with increases in ADLs (0.78, p<.05) and IADLs (0.60, p<.05) measures of functional limitations for her non-working husband. Again, we saw no significance of spousal retirement and non-working husbands’ odds of being obese or their CES-D score. Our results suggested less negative health effects from spousal retirement for working husbands than non-working husbands. We saw no association between spousal retirement and the physical or mental health of a working husband.

We conducted an additional sensitivity analysis to assess whether the results for non-working spouse were biased due to the inclusion of joint retirement. These tests showed similar significant and negative health effects for non-working husbands as those shown in Table 3. We found that a wife’s retirement was significantly associated with increases in functional limitations for her non-working husband. Results for wives were substantially and significantly like those presented in Table 2. These results highlighted that spousal spillover effects occured to some extent regardless of sex and retirement status. However, non-working husbands might have experienced the highest concentration of these negative effects.

Discussion

Previous research on retirement and health has predominantly focused on spousal spillover effects of retirement on women, or individual effects of retirement on health. This narrowed focus has limited the understanding of how spousal retirement might shape health across both genders. To further understand the spousal spillover effects of retirement on health, we observe the retirement process among older adults in the Health and Retirement Study. Overall, we found that the retirement status of an individual’s spouse may have significant effects on the other spouse’s health. However, these effects might vary based on the own spouse’s retirement status. Moreover, men’s and women’s health is differentially affected by spousal retirement, where men might be the most negatively affected by their spouses’ transition in the U.S. context. These results contradicted conventional wisdom that undergirds numerous untested assumptions underlying prior research on this significant life transition. This finding highlights the need to consider the husband’s experience of his spouse’s retirement, as well as the wife’s. Additionally, as U.S. women continue to have high levels of engagement with the workforce, research must adopt new understandings of how gendered effects on health from spousal retirement might be changing from previous patterns.

Our results for men indicated that a husband’s response to their wife’s retirement might have been underestimated in the past, as previous studies have largely omitted observing them. We found more significant and negative effects for husbands than wives. Across measures of physical health, we found evidence of an increase in functional limitations across both measures of ADLs and IADLs. However, for wives we found no significant spousal spillover effects on either physical or mental health indicators.

In further analyses, our results indicated that non-working spouses were likely to experience more negative health spillovers than spouses who were working when their spouse retired. We found that the increases in functional limitations for men were concentrated in our subgroup of men who were non-working when their spouse retired. However, our results showed no significant effects for non-working wives. Additionally, our results suggested that working wives might experience less functional limitations when their spouse retires.

Our results highlight the need to consider the negative effects men might experience due to their spouse’s life transitions, such as retirement. Past research that has investigated spousal spillovers on health from retirement has looked predominately at samples of women (Bertoni & Brunello, 2017; Zang, 2020). This focus has stemmed largely from the literature that uses the Retired Husband Syndrome as a framework (Johnson, 1984). However, this framework neglects to understand how men experience their wife’s retirement and how they might restructure their duties and activities post this retirement. For example, men might decrease in their physical activity (Müller & Shaikh, 2018), which might lead to increases in functional limitations. Alternatively, spouses commonly take on more household duties post spousal retirement which might shed insight on why a husband’s, particularly those who are out of the workforce, number of functional limitations might increase when his wife retires (Leopold & Skopek, 2015). However, these trends might differ from those seen in East Asian countries such as China and Japan, where other research has investigated the Retired Husband Syndrome (Bertoni & Brunello, 2017; Zang, 2020) due to stricter gender roles than in the United States, which includes higher housework burdens. In contexts such as Europe, gender roles can be less rigid than in the United States, which makes spousal spillover effects on retirement related less to gender (Müller & Shaikh, 2018).

A family stress model might be a more appropriate framework to adopt when assessing both men and women’s health following spousal retirement, in contrast to the Retired Husband Syndrome. In our study, we see the most negative health effects present in husbands, particularly those who are non-working when their wife retires. It is possible that some of these negative physical health effects could be due to economic strain, particularly due to historical gender roles which place men as the main breadwinner of the family. The stress of this role might be present regardless of a man’s own employment status, as we see the most negative health effects for non-working husbands. Future research should leverage more detailed quantitative or qualitative data to better understand breadwinning pressures to assess if they drive negative health effects post spousal retirement. Although we find no significant results for effects on mental health, this result might occur due to our use of the CES-D scale. As the CES-D is a measure of depressive symptoms it may miss other important mental health outcomes such as stress and anxiety which can be shaped by one’s spouse (Behler et al., 2019). More research is needed on the effects of spousal retirement on mental health measures that are beyond the CES-D scale.

Although our results offer insight into how married individuals in in their 50s through 70s respond to their spouse’s retirement, these results may not be generalizable to other age groups. Our findings may not apply to couples where one spouse retires earlier in the life course. Additionally, our analysis only extends to opposite-sex couples because they are reported to differ from same-sex couples in stress management (Behler et al., 2019). Furthermore, our research design leverages the discontinuity in health outcomes before and after a spouse’s retirement. More research is needed to show the long-term effects of retirement spillover, as some health effects might not occur until later in the life course.

Future research should also consider how health insurance gaps might shape spousal health post spousal retirement. It is possible that gaps in health insurance could lead to the negative health effects we see for men as they no longer may be insured under a wife’s work insurance policy. Spouses might experience gaps in healthcare coverage if their spouse retires and they are ineligible for Medicare. Our analysis captures the effects of retirement on health for couples who were married during the period of HRS data collection (1992–2016). Later cohorts might differ in health as well as partnership variation and should be examined. Future research should also consider how marital quality might influence spillover effects.

To fully meet the health needs of older adults, policymakers must take a broader look at the marital unit in the retirement process. Past research has largely focused on the effects of retirement on the retiree (Behncke, 2012; Coe & Zamarro, 2011; de Breij et al., 2020), which overlooks the impacts of one spouse’s retirement on the other. As we show here, older adult health is linked to spousal spillover effects which highlights an important need for the consideration of the family unit rather than just the retiree in shaping policy decisions related to older and retiring adults.

Supplementary Material

Appendix

Acknowledgements:

This research was supported by the Population Research Institute at Penn State University, which is supported by an infrastructure grant by the Eunice Kennedy Shriver National Institute of Child Health and Human Development (P2CHD041025). The Health and Retirement Study is sponsored by the National Institute on Aging (NIA U01AG009740) and is conducted by the University of Michigan. The authors thank Ashton Verdery, Steven Haas, Brandon Wagner, Shawn Bauldry, and Andrew Fenelon for helpful feedback on previous drafts.

References

  1. Atalay K, & Zhu R (2018). The Effect of a Wife’s Retirement on Her Husband’s Mental Health. Applied Economics, 50(43), 4606–4616. 10.1080/00036846.2018.1458198 [DOI] [Google Scholar]
  2. Averett SL, Argys LM, & Sorkin J (2013). In sickness and in health: An examination of relationship status and health using data from the Canadian National Public Health Survey. Review of Economics of the Household, 11, 599–633. 10.1007/s11150-012-9143-z [DOI] [Google Scholar]
  3. Aytaç IA, & Rankin BH (2009). Economic crisis and marital problems in Turkey: Testing the family stress model. Journal of Marriage and Family, 71(3), 756–767. 10.1111/j.1741-3737.2009.00631.x [DOI] [Google Scholar]
  4. Barrett AE (2000). Marital Trajectories and Mental Health. Journal of Health and Social Behavior, 41(4), 451–464. [PubMed] [Google Scholar]
  5. Behler R, Donnelly R, & Umberson D (2019). Psychological Distress Transmission in Same-sex and Different-sex Marriages. Journal of Health and Social Behavior, 60(1), 18–35. 10.1177/0022146518813097 [DOI] [PMC free article] [PubMed] [Google Scholar]
  6. Behncke S (2012). Does Retirement Trigger Ill Health? Health Economics, 21, 282–300. 10.1002/hec [DOI] [PubMed] [Google Scholar]
  7. Belloni M, Meschi E, & Pasini G (2016). The Effect on Mental Health of Retiring During the Economic Crisis. Health Economics, 25(Suppl. 2), 126–140. 10.1002/hec [DOI] [PubMed] [Google Scholar]
  8. Bertoni M, & Brunello G (2017). Pappa Ante Portas: The Effect of the Husband’s Retirement on the Wife’s Mental Health in Japan. Social Science & Medicine, 175, 135–142. 10.1016/j.socscimed.2017.01.012 [DOI] [PubMed] [Google Scholar]
  9. Bongaarts J (2004). Population Aging and the Rising Cost of Public Pensions. Population and Development Review, 30(1), 1–23. [Google Scholar]
  10. Brown TH, Richardson LJ, Hargrove TW, & Thomas CS (2016). Using Multiple-Hierarchy Stratification and Life Course Approaches to Understand Health Inequalities: The Intersecting Consequences of Race, Gender, SES, and Age. Journal of Health and Social Behavior, 57(2), 200–222. 10.1177/0022146516645165 [DOI] [PMC free article] [PubMed] [Google Scholar]
  11. Buckles KS, & Price J (2013). Selection and the marriage premium for infant health. Demography, 50, 1315–1339. 10.1007/s13524-013-0211-7 [DOI] [PubMed] [Google Scholar]
  12. Carpenter C, & Dobkin C (2009). The effect of alcohol consumption on mortality: Regression discontinuity evidence from the minimum drinking age. American Economic Journal: Applies Economics, 1(1), 164–182. [DOI] [PMC free article] [PubMed] [Google Scholar]
  13. Centers for Disease Control and Prevention. (2021). Adult obesity facts. Department of Health and Human Services. [Google Scholar]
  14. Cheng FW, Gao X, Mitchell DC, Wood C, Still CD, Rolston D, & Jensen GL (2016). Body mass index and all-cause mortality among older adults. Obesity, 24(10), 2232–2239. 10.1002/oby.21612 [DOI] [PubMed] [Google Scholar]
  15. Cobb LK, Godino JG, Selvin E, Kucharska-Newton A, Coresh J, & Koton S (2016). Spousal influence on physical activity in middle-aged and older adults: The ARIC Study. American Journal of Epidemiology, 183(5), 444–451. 10.1093/aje/kwv104 [DOI] [PMC free article] [PubMed] [Google Scholar]
  16. Coe NB, & Zamarro G (2011). Retirement Effects on Health in Europe. Journal of Health Economics, 30, 77–86. 10.1016/j.jhealeco.2010.11.002 [DOI] [PMC free article] [PubMed] [Google Scholar]
  17. Conger RD, Ge X, Elder GH, Lorenz FO, & Simons RL (1994). Economic stress, coercive family process, and developmental problems of adolescents. Child Development, 65(2), 541–561. 10.1111/j.1467-8624.1994.tb00768.x [DOI] [PubMed] [Google Scholar]
  18. Conger RD, Rueter MA, & Elder GH (1999). Couple resilience to economic pressure. Journal of Personality and Social Psychology, 76(1), 54–71. 10.1037/0022-3514.76.1.54 [DOI] [PubMed] [Google Scholar]
  19. Courtenay WH (2000). Constructions of Masculinity and Their Influence on Men’s Well-Being: A Theory of Gender and Health. Social Science & Medicine, 50, 1385–1401. [DOI] [PubMed] [Google Scholar]
  20. Cutrona CE (1996). Social support in couples. Sage Publications. [Google Scholar]
  21. Davey A, & Szinovacz ME (2004). Dimensions of Marital Quality and Retirement. Journal of Family Issues, 25(4), 431–464. 10.1177/0192513X03257698 [DOI] [Google Scholar]
  22. de Breij S, Huisman M, & Deeg DJH (2020). Macro-Level Determinants of Post-Retirement Health and Health Inequalities: A Multilevel Analysis of 18 European Countries. Social Science & Medicine, 245. 10.1016/j.socscimed.2019.112669 [DOI] [PubMed] [Google Scholar]
  23. Ebaugh HRF (1988). Becoming an ex: The process of role exit. University of Chicago Press. [Google Scholar]
  24. Eibich P (2015). Understanding the Effect of Retirement on Health: Mechanisms and Heterogeneity. Journal of Health Economics, 43, 1–12. 10.1016/j.jhealeco.2015.05.001 [DOI] [PubMed] [Google Scholar]
  25. Elder GH (1998). The Life Course as Developmental Theory. Child Development, 69(1), 1–12. [PubMed] [Google Scholar]
  26. Falba TA, & Sindelar JL (2008). Spousal concordance in health behavior change. Health Services Research, 43(1), 96–116. 10.1111/j.1475-6773.2007.00754.x [DOI] [PMC free article] [PubMed] [Google Scholar]
  27. Frech A, & Williams K (2007). Depression and the psychological benefits of entering marriage. Journal of Health and Social Behavior, 48(2), 149–163. 10.1177/002214650704800204 [DOI] [PubMed] [Google Scholar]
  28. Hospido L, & Zamarro G (2014). Retirement Patterns of Couples in Europe. IZA Journal of European Labor Studies, 3(12). [Google Scholar]
  29. Hughes ME, & Waite LJ (2009). Marital Biography and Health at Mid-Life. Journal of Health and Social Behavior, 50(3), 344–358. [DOI] [PMC free article] [PubMed] [Google Scholar]
  30. Imbens GW, & Lemieux T (2008). Regression Discontinuity Designs: A Guide to Practice. Journal of Econometrics, 142, 615–635. 10.1016/j.jeconom.2007.05.001 [DOI] [Google Scholar]
  31. Johnson CC (1984). The Retired Husband Syndrome. Western Journal of Medicine, 141, 542–545. 10.1097/00006534-198605000-00093 [DOI] [PMC free article] [PubMed] [Google Scholar]
  32. Kane JB (2016). Marriage Advantages in Perinatal Health: Evidence of Marriage Selection or Marriage Protection? Journal of Marriage and Family, 78(1), 212–229. 10.1111/jomf.12257 [DOI] [PMC free article] [PubMed] [Google Scholar]
  33. Killewald A (2013). A reconsideration of the fatherhood premium: Marriage, coresidence, biology, and fathers’ wages. American Sociological Review, 78(1), 96–116. 10.1177/0003122412469204 [DOI] [Google Scholar]
  34. Kim JE, & Moen P (2001). Is Retirement Good or Bad for Subjective Well-Being? Current Directions in Psychological Science, 10(3), 83–86. 10.1111/1467-8721.00121 [DOI] [Google Scholar]
  35. Kim JE, & Moen P (2002). Retirement transitions, gender, and psychological well-being: A life-course, ecological model. Journals of Gerontology - Series B Psychological Sciences and Social Sciences, 57B(3), 212–222. 10.1093/geronb/57.3.P212 [DOI] [PubMed] [Google Scholar]
  36. Kinnunen U, & Pulkkinen L (1998). Linking economic stress to marital quality among Finish marital couples: Mediator effects. Journal of Family Issues, 19(6), 705–724. [Google Scholar]
  37. Koball HL, Moiduddin E, Henderson J, Goesling B, & Besculides M (2010). What do we know about the link between marriage and health? Journal of Family Issues, 31(8), 1019–1040. 10.1177/0192513X10365834 [DOI] [Google Scholar]
  38. Kumar A, & Ozdamar L (2004). International comparison of health care systems. International Journal of the Comupter, the Internet and Management, 12(3), 81–95. [Google Scholar]
  39. Kwon HK, Rueter MA, Lee MS, Koh S, & Ok SW (2003). Marital relationships following the Korean economic crisis: Applying the family stress model. Journal of Marriage and Family, 65(2), 316–325. 10.1111/j.1741-3737.2003.00316.x [DOI] [Google Scholar]
  40. Lee R, & Mason A (2010). Some macroeconomic aspects of global population aging. Demography, 47, S151–S172. [DOI] [PMC free article] [PubMed] [Google Scholar]
  41. Leopold T, & Skopek J (2015). Convergence or Continuity? The Gender Gap in Household Labor After Retirement. Journal of Marriage and Family, 77, 819–832. 10.1111/jomf.12199 [DOI] [Google Scholar]
  42. Lillard LA, & Panis CWA (1996). Marital Status and Mortality: The Role of Health. Demography, 33(3), 313–327. [PubMed] [Google Scholar]
  43. Lillard LA, & Waite LJ (1995). ‘Til Death Do Us Part: Marital Disruption and Mortality. American Journal of Sociology, 100(5), 1131–1156. [Google Scholar]
  44. Linton R (1936). The study of man: An introduction. D. Appleton & Company. [Google Scholar]
  45. Lipowicz A (2014). Some evidence for health-related marriage selection. American Journal of Human Biology, 26, 747–752. 10.1002/ajhb.22588 [DOI] [PubMed] [Google Scholar]
  46. Margolis R, Chai X, Verdery AM, & Newmyer L (2022). The Physical, Mental, and Social Health of Middle-Aged and Older Adults without Close Kin in Canada. The Journals of Gerontology, Series B: Psychological Sciences and Social Sciences, 77(7), 1350–1360. 10.1093/geronb/gbab222 [DOI] [PMC free article] [PubMed] [Google Scholar]
  47. Masarik AS, Martin MJ, Ferrer E, Lorenz FO, Conger KJ, & Conger RD (2016). Couple Resilience to Economic Pressure Over Time and Across Generations. Journal of Marriage and Family, 78(2), 326–345. 10.1111/jomf.12284 [DOI] [PMC free article] [PubMed] [Google Scholar]
  48. McCrary J (2008). Manipulation of the running variable in the regression discontinuity design: A density test. Journal of Econometrics, 142(2), 698–714. [Google Scholar]
  49. Mead GH (1934). Mind, self and society. University of Chicago Press. [Google Scholar]
  50. Mikucka M, Arránz Becker O, & Wolf C (2021). Revisiting marital health protection: Intraindividual health dynamics around transition to legal marriage. Journal of Marriage and Family, 83(5), 1439–1459. 10.1111/jomf.12784 [DOI] [Google Scholar]
  51. Müller T, & Shaikh M (2018). Your retirement and my health behavior: Evidence on retirement externalities from a fuzzy regression discontinuity design. Journal of Health Economics, 57, 45–59. 10.1016/j.jhealeco.2017.10.005 [DOI] [PubMed] [Google Scholar]
  52. Neppl TK, Senia JM, & Donnellan MB (2016). Effects of economic hardship: Testing the Family Stress Model over time. Journal of Family Psychology, 30(1), 12–21. 10.1037/fam0000168 [DOI] [PMC free article] [PubMed] [Google Scholar]
  53. Nomaguchi KM, & Bianchi SM (2004). Exercise time: Gender differences in the effects of marriage, parenthood, and employment. Journal of Marriage and Family, 66, 413–430. 10.1111/j.1741-3737.2004.00029.x [DOI] [Google Scholar]
  54. OECD. (2007). Comparing pension policies of OECD countries. 10.1787/pension_glance-2007-3-en [DOI]
  55. Radloff LS (1977). The CES-D Scale: A Self-Report Depression Scale for Research in the General Population. Applied Psychological Measurement, 1(3), 385–401. 10.1177/014662167700100306 [DOI] [Google Scholar]
  56. Reczek C, Thomeer MB, Gebhardt-Kram L, & Umberson D (2020). “Go See Somebody”: How Spouses Promote Mental Health Care. Society and Mental Health, 10(1), 80–96. 10.1177/2156869319834335 [DOI] [PMC free article] [PubMed] [Google Scholar]
  57. Rendall MS, Weden MM, Favreault MM, & Waldron H (2011). The Protective Effect of Marriage for Survival: A Review and Update. Demography, 48, 481–506. 10.1007/s13524-011-0032-5 [DOI] [PubMed] [Google Scholar]
  58. Rook KS, August KJ, Stephens MAP, & Franks MM (2011). When does Spousal Social Control Provoke Negative Reactions in the Context of Chronic Illness? The Pivotal Role of Patients’ Expectations. Journal of Social and Personal Relationships, 28(6), 772–789. 10.1177/0265407510391335 [DOI] [PMC free article] [PubMed] [Google Scholar]
  59. Rosenfield S, Lennon MC, & White HR (2005). The Self and Mental Health: Self-Salience and the Emergence of Internalizing and Externalizing Problems. Journal of Health and Social Behavior, 46(4), 323–340. 10.1177/002214650504600402 [DOI] [PubMed] [Google Scholar]
  60. Siegel MJ, Bradley EH, Gallo WT, & Kasl SV (2004). The Effect of Spousal Mental and Physical Health on Husbands’ and Wives’ Depressive Symptoms, among Older Adults: Longitudinal Evidence from the Health and Retirement Survey. Journal of Aging and Health, 16(3), 398–425. 10.1177/0898264304264208 [DOI] [PubMed] [Google Scholar]
  61. Silver MP, & Williams SA (2018). Reluctance to Retire: A Qualitative Study on Work Identity, Intergenerational Conflict, and Retirement in Academic Medicine. Gerontologist, 58(2), 320–330. 10.1093/geront/gnw142 [DOI] [PubMed] [Google Scholar]
  62. Social Security. (2021). “Normal retirement age.” https://www.ssa.gov/oact/ProgData/nra.html. [Google Scholar]
  63. Stancanelli E (2014). Divorcing Upon Retirement: A Regression Discontinuity Study. IZA Discussion Paper, (8117). [Google Scholar]
  64. Szinovacz ME, & Davey A (2005). Retirement and Marital Decision Making: Effects on Retirement Satisfaction. Journal of Marriage and Family, 67, 387–398. [Google Scholar]
  65. Teachman J (2016). Body weight, marital status, and changes in marital status. Journal of Family Issues, 37(1), 74–96. 10.1177/0192513X13508404 [DOI] [PMC free article] [PubMed] [Google Scholar]
  66. Tower RB, & Kasl SV (1996). Gender, Marital Closeness, and Depressive Symptoms in Elderly Couples. The Journals of Gerontology Series B: Psychological Sciences and Social Sciences, 51B(3), P115–P129. 10.1093/geronb/51b.3.p115 [DOI] [PubMed] [Google Scholar]
  67. Tran DB, & Zikos V (2019). The Causal Effect of Retirement on Health: Understanding the Mechanisms. The Australian Economic Review, 52(4), 427–446. 10.1111/1467-8462.12340 [DOI] [Google Scholar]
  68. Tumin D (2018). Does Marriage Protect Health? A Birth Cohort Comparison. Social Science Quarterly, 99(2), 626–643. 10.1111/ssqu.12425 [DOI] [Google Scholar]
  69. Tumin D, & Zheng H (2018). Do the Health Benefits of Marriage Depend on the Likelihood of Marriage? Journal of Marriage and Family, 80(3), 622–636. 10.1111/jomf.12471 [DOI] [Google Scholar]
  70. Uecker JE (2012). Marriage and mental health among young adults. Journal of Health and Social Behavior, 53(1), 67–83. 10.1177/0022146511419206 [DOI] [PMC free article] [PubMed] [Google Scholar]
  71. Umberson D (1992). Gender, Marital Status and the Social Control of Health Behavior. Social Science & Medicine, 34(8), 907–917. [DOI] [PubMed] [Google Scholar]
  72. Verdery AM, Newmyer L, Wagner B, & Margolis R (2021). National Profiles of Coronavirus Disease 2019 Mortality Risks by Age Structure and Preexisting Health Conditions. The Gerontologist, 61(1), 71–77. 10.1093/geront/gnaa152 [DOI] [PMC free article] [PubMed] [Google Scholar]
  73. Veron J, & Council NR (2001). Preparing for an Aging World: The Case for Cross-National Research. In Population (French Edition) (Vol. 56). 10.2307/1534892 [DOI] [Google Scholar]
  74. Waite LJ (1995). Does Marriage Matter? Demography, 32(4), 483–507. [PubMed] [Google Scholar]
  75. Waldron I, Hughes ME, & Brooks TL (1996). Marriage Protection and Marriage Selection-Prospective Evidence for Reciprocal Effects of Marital Status and Health. Social Science & Medicine, 43(1), 113–123. 10.1016/0277-9536(95)00347-9 [DOI] [PubMed] [Google Scholar]
  76. Wilson SE (2002). The Health Capital of Families: An Investigation of the Inter-Spousal Correlation in Health Status. Social Science & Medicine, 55, 1157–1172. [DOI] [PubMed] [Google Scholar]
  77. Zang E (2020). Spillover effects of a husband’s retirement on a woman’s health: Evidence from urban China. Social Science & Medicine, 245. 10.1016/j.socscimed.2019.112684 [DOI] [PubMed] [Google Scholar]
  78. Zhang Z, & Hayward MD (2006). Gender, the Marital Life Course, and Cardiovascular Disease in Late Midlife. Journal of Marriage and Family, 68(August), 639–657. 10.1111/j.1741-3737.2006.00280.x [DOI] [Google Scholar]

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Supplementary Materials

Appendix
NIHMS1906371-supplement-Appendix.docx (38.6KB, docx)

RESOURCES