Partners’ adjustment to older workers’ retirement: testing the role of preretirement expectations in a 10-year panel study

Abstract

Objectives: Retirement is not only an important later-life transition for the retiring individual, but also for his or her life partner. This study aims to improve our understanding of the partner’s adjustment to the retirement of the older worker, by paying attention to the multidimensional nature of adjustment, and by examining to what extent preretirement expectations are predictive of postretirement experiences. Well-established adjustment predictors – i.e. preretirement resources and characteristics of the work and retirement context – are also taken into account.

Method: Analyses are based on Dutch three-wave multi-actor panel data, collected between 2001 and 2011 among 724 partners of older workers who transitioned into retirement during the course of the study.

Results: Only a minority of the partners reported adjustment difficulties to the retirement of the employee. About 20 percent reported at least some financial problems, 8 percent reported relationship problems, and 10 percent reported problems with shared leisure time. Expected problems in all three domains were predictive of experienced problems in the same domain. For expected financial problems, a cross-over effect was observed: expected financial problems were also predictive of experienced adjustment difficulties with regards to shared leisure activities.

Conclusion: Not only the older worker, but also the partner develops expectations on different dimensions about the shared postretirement future, and these expectations are related to postretirement experiences. Retirement counseling may therefore not only be relevant for older workers, but also for their partners, and needs to take the multidimensional character of retirement processes into account.

Keywords:

• Adjustment
• couples
• multi-actor
• retirement process

Retirement is a meaningful transition in older adulthood, with multiple implications for daily living. The reduced commitment to the workplace often implies an increase in the time spent with one’s spouse (especially if retired); financial changes in retirement have implications for the household financial management; and the changes in daily routine, when employment does no longer fill the better part of the day, affect everyday behavior, most notably, everyday leisure consumption. Previous research on retirement adjustment has generally focused on the factors that affect retirees’ retirement adjustment quality (i.e. their level of “psychological comfort regarding the retirement life”; Wang, Henkens, & van Solinge, 2011, p. 204). Research on the partner’s psychological comfort with regard to the retirement of their spouse is still scarce though. The aim of this study is to improve our understanding of partners’ retirement adjustment quality, by conceptualizing retirement as a multidimensional and longitudinal process embedded in the couple’s life course context. Our research questions are: How much difficulties – in terms of finances, relationship, and leisure – does the partner experience after retirement of the older worker, and to what extent are the partner’s preretirement expectations predictive of these difficulties?

In the retirement literature the interlinkage between the behaviors and experiences of spouses is well-recognized, particularly in studies based on the ‘linked lives’ principle from the life course perspective (Elder & Rockwell, 1979). This earlier literature is generally focused on the question how partners (i.e. their characteristics, pressures by the partner, relationship transitions) affect workers’ retirement processes (e.g., Henkens & Van Solinge, 2002; Moffatt & Heaven, 2017; Topa, Depolo, & Alcover, 2017), or on antecedents of joint retirement (e.g., Eismann, Henkens, & Kalmijn, 2017; Ho & Raymo, 2009; Hospido & Zamarro, 2014). Much less is known about how retirement of the older worker affects the psychological comfort of the partner. The first contribution of this study is therefore that we examine the implications of retirement within the couple’s life course context, by focusing on the partner’s adjustment quality after the retirement transition of the older worker.

The second contribution of this study is that attention will be paid to the multidimensional nature of the partner's adjustment process. To measure adjustment, earlier studies have either used general measures of psychological comfort (such as indicators of happiness, psychological well-being, and poor mental health, see Calvo, Haverstick, & Sass, 2009; Segel-Karpas, Ayalon, & Lachman, 2018; Szinovacz & Davey, 2004; Yeung, 2017; Yeung & Zhou, 2017), or have asked retirees directly about how they experienced their transition to retirement (Van Solinge & Henkens, 2008). Only recently, scholars have moved beyond these unidimensional approaches, and have emphasized the importance of studying retirement adjustment in a multidimensional way: among retirees the incidence of retirement adjustment difficulties appears to vary between financial and non-financial adjustment dimensions, and predictors differ as well (Damman, Henkens, & Kalmijn, 2015). In this study, we also take such a multidimensional approach, but focus on the partner. More specifically, we directly ask partners about the extent to which they experience problems (and therefore have limited psychological comfort) with regard to three dimensions of the older worker’s retirement transition: finances, the relationship, and shared leisure. Although some of these dimensions have been studied separately – such as the relationship dimension in studies on marital quality (Bushfield, Fitzpatrick, & Vinick, 2008; Davey & Szinovacz, 2004; Kulik, 2001; Moen, Kim, & Hofmeister, 2001) – our contribution is that among partners three different adjustment dimensions are examined, making it possible to see whether the incidence and predictors of adjustment difficulties vary between dimensions.

Third, this study contributes to the literature by paying explicit attention to the role that preretirement expectations play in explaining postretirement experiences, by using multi-actor panel data covering a 10-year period. Thus, we do not only examine how partners experience the retirement transition of the older worker, but also examine how they – in preretirement years – expected their shared postretirement life to be. We measure preretirement expectations on the same three dimensions as our outcomes (i.e. money, relationship, and leisure time). This approach allows us to study discrepancies between partners' expectations and experiences, testing whether, on average, partners experience more, or rather less problems than initially expected. Moreover, by taking this approach we are able to examine whether expectations are predictive of outcomes on the same dimension (this congruence may only be modest, as for instance has been shown for changes in activities upon retirement, e.g., see Fitzpatrick, Bushfield, & Vinick, 2005), and whether there are cross-over effects (i.e. whether expectations on one dimension are related to the experiences on another dimension).

Background

In the scientific literature, retirement is mostly conceptualized as a decision making process, as a career stage, or as an adjustment process (Wang & Wanberg, 2017). Employing an adjustment perspective on the transition to retirement means that the process of retirement should be viewed longitudinally, starting with retirement planning, continuing with the decision to retire, and culminating with adjustment (Shultz & Wang, 2011). Thus, the initial stage of expectations and planning should be taken into account when studying adjustment. According to expectancy theory, expectations potentially carry great consequences. They can be thought of as implicit theories (a set of organized beliefs), and as such, as mechanisms that guide the perception of stimuli and the interpretation given to new information. Stimuli that match the implicit theory one holds are more readily perceived. Similarly, ambiguous situations will be interpreted to match the existing knowledge (Dweck, 1996; Dweck, Chiu, & Hong, 1995; Kirsch, 1985). Focusing on retirement expectations, expecting difficulties in the relationship with the spouse (for example) could direct attention to disagreements, and heighten reactivity to the situation that meets the expectations (that is – the disagreement), thus generating the expected response (experience). It might also encourage one to interpret the behaviors of the partner in a way that matches the expectations. That is, if one expects her husband to be more judgmental when he is home, she might more readily perceive any comment as “being judgmental”. In other words, expecting difficulties may increase the probability of experiencing difficulties, thus making the retirement transition more difficult and harming adjustment.

Studies conducted among retirees have shown that expectations are significant predictors of adjustment to retirement (Gall & Evans, 2000; Taylor, Goldberg, Shore, & Lipka, 2008; Taylor, Shultz, Spiegel, Morrison, & Greene, 2007; Van Solinge & Henkens, 2005, Van Solinge & Henkens, 2008; Yeung, 2013). Despite the importance of retirement expectations for adjustment among retirees, little is known about the expectations among partners and the predictive value of these expectations for postretirement experiences. However, it is safe to assume that as one member of a couple approaches retirement, his or her spouse also develops certain expectations regarding the shared future, i.e. couples' joint lives after the older worker makes the transition. Retirement is an institutionalized phenomenon, and not only do “older workers carry an exquisite consciousness that retirement awaits them within the life-course program (Ekerdt, Kosloski, & DeViney, 2000, p.4),” but also so do their partners. From the life course notions of linked lives and lifelong development (e.g., Settersten, 2003) it can therefore be expected that partners formulate expectations as well. From expectancy theory, it can be anticipated that these expectations of the partner, will impact the actual adjustment experiences of the partner after retirement of the employee.

This study focuses on the partner's expected and experienced difficulties in three areas: financial resources, quality of the relationship with the spouse, and shared leisure time. These domains were chosen as they all represent areas in which retirement of one spouse is expected to have significant effects: As retirement is first and foremost a process that defines one's standing in the labor force, it is a step with meaningful financial implications. Second, the changes in work-home activities balance often implies that couples have more time together, thus, in certain relationships could cause strain. Third, the reduced commitment to the workforce allows greater engagement in leisure activities. Research indicates that these three areas are often mentioned by late-life couples as sources for difficulties and disagreements (Henry, Miller, & Giarrusso, 2005).

In this research we take a detailed perspective, evaluating similarities and differences between domains, assuming intra and inter-domain effects. We hypothesize that the more problems partners expect in preretirement years with regards to (H1) finances, (H2) relationship, and (H3) leisure after retirement of the employee, the more postretirement adjustment problems they will experience in the respective domains. We also hypothesize a cross-over effect for financial expectations. Given that money enables consumption of leisure activities, and given that the adequacy of financial resources is a source of satisfaction in the household (Smith & Moen, 2004), we hypothesize that the more financial problems partners expect in preretirement years, the more postretirement problems they experience on the (H4) relationship, and (H5) leisure dimensions.

When examining the impact of preretirement expectations on postretirement adjustment experiences of partners, well-established predictors of retirement adjustment derived from studies conducted among retirees or partners will be taken into account (see reviews by Barbosa, Monteiro, & Murta, 2016; Wang et al., 2011; Wang & Wanberg, 2017). Earlier studies have shown that resources such as income and health (Hill & Dorfman, 1982; Topa, Moriano, Depolo, Alcover, & Moreno, 2011; Zaniboni, 2015), and characteristics of the couple's work and retirement context such as involuntary retirement of the older worker (Fisher, Chaffee, & Sonnega, 2016; Moffatt & Heaven, 2017; Van Solinge & Henkens, 2005), type of decision-making (Dorfman & Hill, 1986; Smith & Moen, 2004), and work situation of the partner (Curl & Townsend, 2014; Moen et al., 2001) affect the postretirement adjustment process. By taking resources and characteristics of the couple's work and retirement context into account, we will examine whether psychological factors – i.e. preretirement expectations about postretirement life – play an additional explanatory role above and beyond these well-established structural factors among spouses.

Method

Sample

To test the hypotheses, data of the NIDI Work and Retirement Panel were analyzed. These three-wave panel data were collected by the Netherlands Interdisciplinary Demographic Institute among older workers and their partners between 2001 and 2011. During the course of the study, the retirement landscape in the Netherlands was characterized by an early exit culture (De Vroom, 2004). The mean retirement age of Dutch employees has been around 61 between 2001 and 2007, and has increased to 63 in 2011 (Statistics Netherlands, 2013).

The first wave of data collection took place in 2001. Data were collected from all workers aged 50-64 in three large Dutch multinational private-sector organizations, and a random sample of civil servants aged 50-64 years. A hard-copy questionnaire was sent to 3,899 older workers; 2,403 questionnaires were returned (response rate 62%). Wave 2 data collection was performed in 2006-2007 among surviving and traceable participants of the first wave. A total of 2,239 questionnaires were sent out; 1,678 were completed (response rate 75%). In 2011 the third wave of data collection took place. A questionnaire was sent to all 1,638 surviving and traceable respondents of the second wave, resulting in 1,276 returned questionnaires (response rate 78%). In all three waves of data collection, the employee’s partner (if applicable) was asked to participate in the study and to fill out a separate questionnaire. Among participating employees who indicated having a partner, the partner response rates were 92% at Wave 1, 91% at Wave 2, and 89% at Wave 3.

Given that this paper focuses on adjustment of the partner to the retirement of the employee, we analyze those partners whose spouse (i.e. the employee) had transitioned from employment to full retirement during the study (that is, fully retired between the first and second or third wave of data collection). In total 1,080 employees transitioned into full retirement during the study period. For 782 of these employees, both a preretirement (i.e. Wave 1) and a postretirement (i.e. Wave 2/3) partner questionnaire were available. Partners who did not answer all three dependent variables of the study (N = 58) were excluded from the analyses, resulting in an analytic sample of 724 partners to recently retired employees.

Measures

Dependent variables. Partners were asked to rate the extent to which they had experienced adjustment problems resulting from the retirement of their spouse. We used the answer provided at the wave most proximate to the employee’s transition to retirement. Partners were asked to rate on a scale ranging from 1 “very much” to 5 “not at all” the extent to which they had experienced problems with regards to: a. money, b. relationship with their spouse, and c. the use of common leisure time, following their spouse’s transition into retirement. Responses were reverse coded, such that higher scores on these variables indicate more adjustment difficulties.

Independent variables. Preretirement expectations of partner – Partners were asked in the preretirement Wave 1 questionnaire to rate on a scale ranging from 1 “very much” to 5 “not at all” the extent to which they expected problems with regards to a. money, b. relationship with their spouse, and c. the use of common leisure time, following their spouse’s retirement. The answers were reverse coded, such that higher values indicate more expected adjustment problems.

Preretirement resources – The financial wealth of the couple was assessed using a single item obtained at Wave 1. The employee was asked to estimate his/her total wealth (including house, savings, stocks, etc., minus debts/mortgage). Responses ranged from 1 “less than 10,000 guilders to 7 “more than 1 million guilders”. In the analyses we used the natural logarithm of the class averages (transformed to euros). Subjective health was measured for both employee and partner, using a single item: “How would you characterize your health in general?” (1 = very good to 5 = very poor). We reverse coded the variables, such that higher values reflect better health.

Work and retirement context – Information on preretirement work hours of the employee was provided by the participating organizations (range 0.10–1.00, where 1 represents a full-time work week). We multiplied these values by 40 to obtain the formal number of work hours per week. The employee’s age at retirement reflects the age at which the employee made use of an (early) retirement arrangement. Whether or not the transition to retirement was voluntary was measured by a single question “Was your decision to retire entirely voluntary or not?” The responses were coded into a dummy variable, where the value 1 indicates that the employee retired voluntarily. To measure the partner’s work and retirement situation information about the partner’s work status before and after the retirement of the employee was combined into the following categories: (1) partner is not working (neither before nor after retirement of the employee), (2) partner retired during study period, (3) partner is not yet retired.

Control variables – In the analyses we controlled for the gender of the employee (0 = male, 1 = female), and the age difference between partners, which was determined by subtracting the partner’s age at Wave 1 from the age of the employee. Also the time that has elapsed since retirement of the employee (i.e. the number of years between measurement of the dependent variables and the age of making use of an (early) retirement arrangement), and the organization the employee used to work for are taken into account.

Analyses

Linear regression analyses were conducted to test the hypotheses. The number of missing values on the independent and control variables was low (the highest percentage is 4% missing values on the wealth measure). These missing values were imputed by applying a multiple imputation procedure in Stata 14 (mi impute chained). The variables having missing values were imputed 25 times by using the information of the dependent, independent, and control variables. For all these 25 datasets the models were then estimated, and the results were combined (Stata 14: mi estimate).

Results

Descriptive results

The descriptive statistics of the dependent, independent, and control variables are presented in Table 1. Regarding the dependent variable – the partner’s experienced adjustment problems to the retirement of the employee – the results generally show that only a minority of the partners had experienced adjustment problems. The share of experienced problems is the highest on the financial adjustment dimension (M = 1.77, SD =0.88), with 19 percent of partners reporting quite some or (very) much money-related problems since the retirement of the employee. With regards to the relationship with the spouse (M = 1.54, SD =0.74) and the use of common leisure time (M = 1.65, SD =0.72) the share of partners reporting problems is lower. About 8 percent of the partners had experienced quite-some or (very) much relationship problems, and about 10 percent reported problems with shared leisure time.

Table 1. Descriptive statistics (before multiple imputation).

	Mean (or %)	SD
Dependent variables:
Experienced problems - money	1.77	0.88
Experienced problems - relationship	1.54	0.74
Experienced problems - common leisure time	1.65	0.72
Control variables:
Employee gender - female	21%
Age difference between partners	0.72	4.09
Time elapsed since retirement employee	2.55	1.88
Preretirement resources:
Wealth (log)	11.55	1.39
Subjective health employee	4.08	0.82
Subjective health partner	4.17	0.77
Work and retirement context:
Work hours employee	36.89	7.08
Age at retirement employee	59.35	2.75
Voluntary retirement employee	76%
Partner’s situation (ref = not working)
Retired during study period	31%
Not yet retired	25%
Preretirement expectations of partner:
Expected problems - money	2.04	0.96
Expected problems - relationship	1.54	0.71
Expected problems - common leisure time	1.69	0.75

When looking at preretirement expectations of the partner, the descriptive statistics suggest that partners on average expected to have more financial adjustment problems than they experienced, t(712) = 7.06, p < .01. About one out of four partners expected quite some or (very) much money-related problems when their husband/wife stops working (see Figure 1). For the other dimensions, the mean differences between expectations and experiences are not statistically significant and percentages of partners expecting problems are lower: 9 percent of the partners expected quite some or (very) much relationship problems in preretirement years, and 12 percent expected problems on the leisure dimension.

Figure 1. Expected and experienced financial, relationship and leisure problems of partners.

Results of multivariate analyses

The results of the multivariate linear regression analyses are presented in Table 2. The models are estimated in two steps. In the first step, the partner’s adjustment problems to the retirement of the employee are examined in relation to resources and characteristics of the work and retirement context (Models 1a, 2a, and 3a). In the second step, measures of preretirement expectations of the partner are added to the model, to see whether these add to our understanding of the partner’s adjustment difficulties beyond the structural factors (Models 1b, 2b, and 3b).

Table 2. Linear regression models of the partner’s adjustment problems to the retirement of the employee, coefficients and standard errors, N = 724.

	Model 1a		Model 1b		Model 2a		Model 2b		Model 3a		Model 3b
	Money		Money		Relationship		Relationship		Leisure		Leisure
Explanatory variables	Coef.	SE	Coef.	SE	Coef.	SE	Coef.	SE	Coef.	SE	Coef.	SE
Control variables:
Employee gender - female^a	−0.05	0.12	−0.04	0.12	−0.23*	0.11	−0.22*	0.11	0.00	0.11	0.03	0.10
Age difference between partners	0.02*	0.01	0.02*	0.01	−0.01	0.01	−0.00	0.01	−0.00	0.01	−0.00	0.01
Time elapsed since retirement	−0.04*	0.02	−0.03#	0.02	−0.01	0.02	−0.01	0.02	−0.02	0.02	−0.03#	0.02
Preretirement resources:
Wealth (log)	−0.09**	0.02	−0.06*	0.02	0.01	0.02	0.01	0.02	0.01	0.02	0.02	0.02
Subjective health employee	−0.11**	0.04	−0.09*	0.04	−0.02	0.03	−0.02	0.03	0.01	0.03	0.01	0.03
Subjective health partner	−0.14**	0.04	−0.10*	0.04	−0.13**	0.04	−0.09*	0.04	−0.15**	0.04	−0.12**	0.04
Work and retirement context:
Work hours employee	−0.01	0.01	−0.01	0.01	−0.01	0.01	−0.01	0.01	−0.00	0.01	−0.00	0.01
Age at retirement employee	−0.04**	0.01	−0.05**	0.01	0.01	0.01	−0.00	0.01	0.01	0.01	0.00	0.01
Voluntary retirement employee^b	−0.25**	0.07	−0.22**	0.07	−0.10	0.07	−0.09	0.06	−0.08	0.06	−0.06	0.06
Partner’s situation (ref = not working)
Retired during study period	−0.01	0.08	−0.02	0.07	0.04	0.07	0.03	0.07	0.05	0.07	0.05	0.07
Not yet retired	0.07	0.09	0.05	0.08	0.14#	0.08	0.11	0.07	0.10	0.07	0.09	0.07
Preretirement expectations of partner:
Expected problems - money			0.29**	0.03			0.05	0.03			0.09**	0.03
Expected problems - relationship			−0.03	0.05			0.20**	0.05			0.05	0.05
Expected problems - common leisure time			0.06	0.05			0.07	0.05			0.14**	0.05
Constant	6.70**	0.91	6.03**	0.87	1.92*	0.79	1.89*	0.77	1.47#	0.78	1.37#	0.76
F	6.12**		10.49**		2.03*		4.59**		2.09*		3.93**

# p < .10,

* p < .05,

** p < .01.

Note: In all models, organization is controlled for by including organizational dummy indicators.

^a Coding: 0 = male, 1 = female.

^b Coding: 0 = retired involuntarily, 1 = retired voluntarily.

According to our first hypothesis, there is a positive relationship between expected and experienced financial problems. In support of this hypothesis, expected money problems are significantly and positively related to experienced problems (see Model 1b), above and beyond the effects of resources and contextual factors (b=.29, p<.01). The effects of preretirement resources (i.e. wealth and health) remained significant. Similarly, retiring at a younger age and involuntary retirement of the employee were also found to increase the partner’s experienced problems with regards to money.

In Model 2a and 2b, the extent to which partners experienced relationship-related problems after their spouses’ retirement is examined. The findings show that partners who expected more relationship problems before retirement of the employee, also experienced more of these problems (b=.20, p<.01), which supports hypothesis 2. The fourth hypothesis was not supported: the effect of expected financial problems on experienced relationship problems was not statistically significant. Furthermore, the findings show that hardly any of the studied resources, and transition characteristics has a statistically significant association with relationship problems. Only the subjective health of the partner plays an explanatory role: the better the health of the partner is, the less likely he or she is to experience relationship-related problems after their spouse retired (b = −.09, p < .05).

In Model 3a and 3b, we examined experienced problems regarding the use of common leisure time. The findings regarding retirement expectations (see Model 3b) suggest that both leisure expectations (b = .14, p < .01), and money expectations (b = .09, p < .01) are significantly related to experienced leisure problems, as was expected in hypotheses 3 and 5. The only resource associated with experienced leisure-related problems is the partner’s subjective health, suggesting that poorer health limits the couple’s ability to enjoy common leisure time, and increases leisure-related problems.

Discussion

This study contributes to the relatively limited body of research examining the implications of older workers’ retirement within the couple’s life course context. Whereas in the literature the importance of the family context in retirement processes is broadly acknowledged (see review by Matthews & Fisher, 2013), still little is known about the psychological comfort of the partner with regard to the retirement transition of the older worker. Our study contributes to our understanding of partners' adjustment processes, by examining both the partner’s preretirement expectations and the partner’s postretirement experiences with regard to the retirement of the older worker, using multi-actor data from a 10-year longitudinal study, and by paying explicit attention to three central dimensions of potential experienced adjustment difficulties: finances, the relationship, and shared leisure.

In descriptive terms, our findings show that relatively few partners were having severe negative postretirement experiences after retirement of the employee. Most negative experiences have to do with financial resources after retirement. About 25 percent of the studied partners expected at least some financial adjustment problems, and about 20 percent reported that they had experienced some financial problems. Few spouses expected and experienced problems in their relationship and with filling shared leisure time (i.e. 8 percent and 10 percent experienced problems with the relationship and leisure respectively). In explanatory terms, our findings showed a clear relationship between preretirement expectations and postretirement experiences (on the same dimension). Moreover, the study results highlighted the importance of examining the multiple dimensions of the partner’s adjustment experience separately.

First, paying attention to the multidimensional nature of the adjustment process advanced our understanding of partners' retirement adjustment, by enabling us to examine cross-over effects: retirement adjustment, by enabling us to examine cross-over effects: whether expectations on one dimension are related to experiences on another dimension. Our findings suggest that expected financial problems are also predictive of experienced adjustment problems with regards to shared leisure activities. This underscores the notion that financial difficulties are linked to stress and adjustment problems (Segel-Karpas, Bamberger, & Bacharach, 2013; Van Solinge & Henkens, 2008). Expected difficulties with regards to shared leisure activities and relationship with the spouse, did not demonstrate the same crossover effect, and were not predictive of experienced problems on the other dimensions. This might suggest that partners succeed in managing their time separately. That is, it could be that those who anticipate leisure problems (or, indeed, experience these), manage to navigate their time, such that they do not spend their leisure together, but rather willingly develop their own leisure routine, thus not harming the relationship with the spouse. Similarly, those who anticipate (and experience) problems with the spouse, do not see solitary (or even joint) leisure activities as problematic. The question of how couples negotiate leisure and relationship in retirement is a possible direction for future research, perhaps taking a “micro” perspective on everyday behavior using daily diaries.

Second, examining the multidimensional nature of partner’s adjustment appeared to be important, because it showed that both the incidence and predictors differ considerably between financial and non-financial retirement adjustment difficulties. We would not have been able to detect these nuances, if we would have asked the partners about their general psychological comfort about the retirement of the older worker. Whereas well-established correlates of adjustment – such as resources (wealth and health) and characteristics of the retirement of the retiree (e.g., age of retirement; voluntary retirement) – were significantly associated with experienced financial adjustment difficulties, no significant association with the non-financial dimensions was observed. In couples that are less wealthy and have a relatively poor health situation, partners, on average, report more financial adjustment problems, as compared to more wealthy and healthy couples. Spouses also report more financial adjustment difficulties when the employee retired at a younger age, and when the employee retired involuntarily. Only the subjective health status of the partner was associated with all three adjustment dimensions of this study. Partners who reported better self-perceived health were less likely to experience adjustment problems with regard to money, the relationship, and shared leisure. This underlines the importance of good health as a central resource for developing a satisfactory postretirement lifestyle for couples.

When interpreting the study findings some limitations should be kept in mind. First, based on our study it is not possible to disentangle the mechanism between expectations and experiences. The question remains whether the expectations are influential because they guide the interpretation of future situations (as predicted by expectancy theory), or because they are reflections of the partner’s preretirement status (e.g. negative expectations regarding the relationship with the spouse after retirement might imply that the relationship is strained). Second, the sample is based on the partners of former employees from four large Dutch organizations and is therefore not nationally-representative. Third, the study is conducted in the Netherlands, a country that is well-known internationally for its relatively generous pension system. Therefore, the generalizability of the study findings – and the descriptive findings in particular – to other countries is unknown. An important direction for future research is therefore to examine partners' multidimensional adjustment processes to retirement of the older worker also in less generous welfare regimes.

Despite the limitations noted above, this study contributes to the retirement literature by taking the often neglected perspective of the spouse. Practitioners might benefit from tailoring retirement workshops for spouses or couples, helping them to reduce the anxious anticipation, and finding ways to better cope with the retirement transition. On the one hand, preretirement counseling both for the older worker and for the spouse may be crucial to help partners being well-prepared for the upcoming retirement transition. Given that finances seem to be the most meaningful source of worry, financial education of the partner may be especially valuable. On the other hand, while in the retirement literature counseling often refers to preretirement initiatives, postretirement counseling may also be worthwhile for some couples, particularly when partners experience adjustment problems that they had not foreseen. In sum, retirement counseling may not only be relevant for older workers, but also for their partners. It may also benefit from taking the multidimensional character of retirement processes into account.

Disclosure statement

No potential conflict of interest was reported by the authors.

Additional information

Funding

This work was supported by the Netherlands Organization for Scientific Research NWO [VICI Grant 453-14-001 to K.H.; VENI Grant 451-17-005 to M.D.] and Netspar.