Abstract
Objectives. We investigated to what extent current financial distress explains the relationship between life-course socioeconomic position and well-being in Southern, Scandinavian, Postcommunist, and Bismarckian welfare regimes.
Methods. We analyzed individuals (n = 18 324) aged 50 to 75 years in the Survey of Health, Ageing, and Retirement in Europe, 2006–2009. Well-being was measured with CASP-12 (which stands for control, autonomy, self-realization, and pleasure) and life satisfaction. We generated a life-course socioeconomic index from 8 variables and calculated multilevel regression models (containing individuals nested within 13 countries), as well as stratified single-level models by welfare regime.
Results. Life-course socioeconomic advantage was related to higher well-being; the difference in life satisfaction between the most and least advantaged was 2.09 (95% confidence interval = 1.87, 2.31) among women and 1.65 (95% confidence interval = 1.43, 1.87) among men. The weakest associations were found among Scandinavian countries. Financial distress was associated with lower well-being and attenuated the relationship between life-course socioeconomic position and well-being in all regimes (ranging from 34.26% in Postcommunist to 72.22% in Scandinavian countries).
Conclusions. We found narrower inequalities in well-being in the Scandinavian regime. Reducing financial distress may help improve well-being and reduce inequalities.
Well-being among older people is shaped by socioeconomic circumstances across the life course.1–5 The degree to which an individual’s childhood socioeconomic circumstances or current wealth affects his or her later well-being is likely to be a function of the combination of policies individuals are exposed to throughout their life course, such as whether the country’s education system is equitable and whether an individual receives an adequate pension in later life. Welfare regimes embody the history, values, and generosity of welfare states and are therefore an important tool for examining inequalities in health and well-being among older people from a life-course and comparative perspective.
The welfare regime approach to studying the effects of welfare policy on health and inequalities is based on the notion that welfare states can be grouped into ideal types (regimes) based on their shared policies, political traditions, and ideologies, which persist over time.6,7 For example, the Nordic (or Scandinavian) countries are considered to cluster together into a distinct welfare regime, which is typically characterized by universal welfare provision, relatively generous benefits, a commitment to full employment, and income protection.6,8 For this reason, it has often been hypothesized that Scandinavian countries would exhibit narrower socioeconomic inequalities in health compared with the less generous welfare states found in Southern Europe, where more family-based arrangements persist.9
However, there is evidence that this is not always the case.10 Results from studies comparing socioeconomic inequalities in health between countries tend to differ depending on the measures of socioeconomic inequality and outcomes used. For example, some studies have found socioeconomic inequalities in mortality to be narrowest in Southern countries such as Italy and Spain,11 but others have found Nordic countries to exhibit the narrowest health inequalities, at least among men.12 Those investigating self-reported outcomes have also produced inconsistent findings, some again finding that social democratic countries did not have the narrowest inequalities,13,14 whereas others investigating mental health and well-being measures found narrowest inequalities in Scandinavian countries.15,16
The World Health Organization states that health is the complete state of mental well-being—not simply the absence of disease or infirmity.17 Yet, few studies have examined health inequalities with measures of well-being,18 and even fewer have examined factors that may act to mediate or moderate the relationship between social conditions and later well-being,16 in contrast to studies on mortality and physical health.19–21 Well-being is increasingly recognized as an important epidemiological outcome among both older people who may have health conditions that do not necessarily have an impact on their day-to-day lives and among younger populations who may be free from physical and mental health conditions.
Governments are now beginning to gather population well-being data to address the shortcomings of economic measures such as gross domestic product, which fail to capture what is thought to make life worthwhile.22,23 It is therefore important to quantify inequalities in well-being and investigate potential mediating factors that could be subject to intervention. The type of welfare regime individuals are exposed to throughout their life course could be a key modifier of the relationship between life-course socioeconomic position (SEP) and well-being among older people. For example, welfare states that have more generous basic pensions may be able to counteract the potentially negative effect on well-being in older age associated with experiencing a lower social class throughout working life, by helping to alleviate financial distress.
This study brings together life course and political economy approaches to studying inequalities in well-being. In particular, it asks 3 pertinent research questions. (1) Is increased socioeconomic advantage over the life course associated with higher well-being in early old age? (2) Can the relationships be explained by current financial distress? (3) Are there differences in the associations between welfare regimes? The study extends previous research on inequalities in well-being4,5,16 by examining the influence of a complex measure of life-course SEP and explores financial distress as a potential mediating factor.
METHODS
The study population comprised individuals (n = 18 324) who participated in both the second and third waves of the Survey of Health, Ageing, and Retirement in Europe (SHARE) and were aged 50 to 75 years at wave 2. SHARE is a longitudinal panel survey collected via face-to-face computer-assisted personal interview.24 Wave 2 was collected during 2006–2007 and included representative data from 13 countries. The target population of wave 2 comprised all noninstitutionalized individuals born in 1956 or earlier (and their partners). The third wave collected retrospective life histories of respondents during 2008–2009, by using the lifegrid method.25 Response and attrition rates are detailed elsewhere.26 In this study, we focused on individuals who were born in their current country of residence.
Outcome Variables
Well-being was measured during the second wave via CASP-12, which stands for the 4 domains of control, autonomy, self-realization, and pleasure.27 CASP-12 is a shortened version of CASP-19, a validated needs satisfaction-based measure of positive quality of life in early old age.28–30 CASP-12 contains 12 questions relating to feelings about experiences in life. Respondents were asked to rate (on a 4-item Likert scale ranging from often to never) how often, in the past 4 weeks, they experienced particular thoughts and feelings, such as whether life has meaning and the future is full of opportunities. Positively worded items were reverse-coded so that the sum of the 12 items ranged from 12 to 48 and higher scores reflected higher well-being. We excluded individuals (n = 748) missing information for this variable from the analysis.
The second well-being outcome examined was life satisfaction. Life satisfaction reflects the cognitive evaluation of one’s life and was measured by asking participants “On a scale from 0 to 10 where 0 means completely dissatisfied and 10 means completely satisfied, how satisfied are you with your life?” Previous research has treated life satisfaction as a continuous variable,31–33 and studies have demonstrated that results operationalizing life satisfaction as continuous and conducting linear regression are not substantially different from those using ordered logistic regression.32,33 For this reason, as well as to help the interpretation and comparison of results, we treated life satisfaction as continuous. We excluded respondents (n = 189) missing data for this variable.
Exposure Variables
We captured life-course SEP by using a composite index generated from 4 childhood and 4 adulthood socioeconomic variables. We selected the variables to represent different aspects of the socioeconomic experience individuals are exposed to throughout their life course. The childhood variables (number of books owned, number of rooms per capita, father’s occupational skill level, and the number of amenities in the household including a fixed bath, hot and cold running water supplies, inside toilet, and central heating) were collected via retrospective recall during the third wave of SHARE and correspond to when the participants were aged 10 years. The adulthood measures of SEP included the respondent’s highest education level, main occupational skill level, current household income, and current household wealth, collected mainly during the second wave. Further details on the theoretical meaning and the derivation of each variable are found in a supplement to the online version of this article at http://www.ajph.org.
We derived the life-course socioeconomic index from the sum of the 8 socioeconomic variables’ standardized relative ranks in the cumulative population distribution.34 To obtain the standardized ranks, we first sorted the sample from the least advantaged to the most advantaged according to the socioeconomic variable of interest (e.g., education level, a 3-category variable). We then counted the number of people in the least-advantaged category and allocated a rank to them, which was equal to the midpoint of the number of people in this category. We repeated this for each further category and we then added their ranks to the cumulative number of people already allocated a rank. We then divided the rank scores by the total number of individuals in the sample, which generated a standardized socioeconomic rank ranging from zero (the least advantaged) to 1 (the most advantaged). We then summed each of the 8 standardized socioeconomic ranks to generate the life-course socioeconomic index, which was divided by 8 to range from zero to 1, zero representing the least-advantaged score and 1 the most-advantaged.
We chose to give equal weight to each socioeconomic indicator making up the measure as it is not clear which indicators should be given greater or lesser weight and any decision taken would be rather arbitrary. This method to generate the life-course socioeconomic index is derived from the calculation used to generate the slope index of inequality. As we ranked each variable by country, gender, and cohort (born pre-1946 or post-1945) it is a relative measure that takes into account the differing socioeconomic distributions for these groups.4,16 For example, the number of people categorized as having a low education level in the cohort born post-1945 was 2863 (36.2%) compared with 4330 (53.0%) in the cohort born pre-1946.
Current financial distress was measured during the second wave by the household’s ability to make ends meet; these terms are used interchangeably throughout. Participants were asked to think about their household’s total monthly income and rate the degree to which they felt able to make ends meet: with great difficulty, with some difficulty, fairly easily, or easily. This was treated as a continuous variable as both measures of well-being increased in a stepwise and linear fashion as the ability to make ends meet increased. We also examined a number of other variables that could be considered to potentially mediate or confound the relationships. These included the individual’s age (treated as a continuous variable and including a squared term to account for the nonlinear association between age and well-being), employment status (retired, employed, homemaker, or other), and cohabitation status (living as a couple or living as single), all recorded during the second wave. We considered employment and cohabitation status to be potential mediators of the relationship between life-course SEP and financial distress, which then influences well-being. Full details of each questionnaire item used and the operationalization of the variables are contained in the data available as a supplement to the online version of this article at http://www.ajph.org.
We grouped the 13 countries into 4 welfare regime types: Southern (Greece, Italy, and Spain), Scandinavian (Denmark and Sweden), Postcommunist (Czech Republic and Poland), and Bismarckian (Austria, Belgium, France, Germany, the Netherlands, and Switzerland).4,16 A description of the key features of each welfare regime is found elsewhere.35
Statistical Analyses
We first examined descriptive statistics for each variable by welfare regime and gender. We calculated linear random-intercept multilevel models containing individuals nested within countries for the 2 measures of well-being and then we adjusted for age and age-squared, followed by the welfare regime type to estimate the percentage of the variance at the country level. We tested statistical interactions between welfare regime dummy variables and the life-course SEP index, as well as financial distress, to examine whether there was evidence to suggest effect modification.
Next, we calculated single-level linear regression models separately for each welfare regime, adjusting for country fixed effects. The models first explored the age-adjusted associations between the life-course SEP variable and well-being and then examined the associations when we included employment and cohabitation status, followed by financial distress.
We performed all analyses with Stata/MP version 12.1 (StataCorp LP, College Station, TX) and we calculated them separately by gender. As SHARE does not provide weights at the country level, we chose to present unweighted results. The SHARE team imputed missing data for income, wealth, and education level; further details of the multiple imputation procedure are provided elsewhere.24 We excluded individuals with missing exposure data that were not multiply imputed (n = 1479) from the analysis.
RESULTS
The sample analyzed consisted of 16 071 individuals: 8675 (54.0%) women and 7396 (46.0%) men. (Data broken down by country and gender are available as a supplement to the online version of this article at http://www.ajph.org.) Mean well-being scores were highest in Scandinavian, followed by Bismarckian welfare regimes, and lowest in Southern and Postcommunist countries. (Table 1 contains descriptive statistics for the sample.) Few individuals experienced financial distress in the Scandinavian and Bismarckian welfare regimes, but in the Postcommunist and Southern regimes around a fifth of the sample reported that they had great difficulty making ends meet. The intraclass correlation for the null CASP-12 multilevel models was 0.19 among women and 0.15 among men (data available as a supplement to the online version of this article at http://www.ajph.org). Less variation was apparent at the country level for life satisfaction—0.13 among women and 0.10 among men.
TABLE 1—
Descriptive Statistics for the Sample by Gender and Welfare Regime Among Adults Aged 50–75 years: Survey of Health, Aging, and Retirement in Europe, 2006–2007
| Women, No. (%) or Mean ±SD |
Men, No. (%) or Mean ±SD |
|||||||||
| Variable | SO | SC | PC | BM | Overall | SO | SC | PC | BM | Overall |
| Participants | 2338 (26.95) | 1351 (15.57) | 1488 (17.15) | 3498 (40.32) | 8675 (100) | 2131 (28.81) | 1212 (16.39) | 1128 (15.25) | 2925 (39.55) | 7396 (100) |
| Life-course SEI | 0.50 ±0.15 | 0.50 ±0.16 | 0.50 ±0.14 | 0.50 ±0.16 | 0.50 ±0.16 | 0.50 ±0.15 | 0.50 ±0.16 | 0.50 ±0.14 | 0.50 ±0.16 | 0.50 ±0.15 |
| Age, y | 61.35 ±7.14 | 61.16 ±6.87 | 60.48 ±6.89 | 61.16 ±6.95 | 61.10 ±6.99 | 62.76 ±7.09 | 61.90 ±6.93 | 61.07 ±7.05 | 61.89 ±6.88 | 62.02 ±7.00 |
| CASP-12 | 34.38 ±5.98 | 40.55 ±4.51 | 34.73 ±6.09 | 38.88 ±5.65 | 37.22 ±6.17 | 35.68 ±5.61 | 40.65 ±4.40 | 35.95 ±5.82 | 39.55 ±5.32 | 38.07 ±5.73 |
| Life satisfaction | 7.15 ±1.74 | 8.57 ±1.30 | 6.77 ±2.02 | 7.78 ±1.54 | 7.56 ±1.75 | 7.47 ±1.53 | 8.54 ±1.30 | 7.10 ±1.87 | 7.92 ±1.38 | 7.77 ±1.56 |
| Ability to make ends meet | ||||||||||
| With great difficulty | 500 (21.39) | 30 (2.22) | 320 (21.51) | 221 (6.32) | 1071 (12.35) | 410 (19.24) | 17 (1.40) | 215 (19.06) | 134 (4.58) | 776 (10.49) |
| With some difficulty | 981 (41.96) | 136 (10.07) | 698 (46.91) | 684 (19.55) | 2499 (28.81) | 867 (40.69) | 104 (8.58) | 495 (43.88) | 471 (16.10) | 1937 (26.19) |
| Fairly easily | 655 (28.02) | 448 (33.16) | 386 (25.94) | 1330 (38.02) | 2819 (32.50) | 638 (29.94) | 413 (34.08) | 333 (29.52) | 1121 (38.32) | 2505 (33.87) |
| Easily | 202 (8.64) | 737 (54.55) | 84 (5.65) | 1263 (36.11) | 2286 (26.35) | 216 (10.14) | 678 (55.94) | 85 (7.54) | 1199 (40.99) | 2178 (29.45) |
| Employment status | ||||||||||
| Retired | 842 (36.01) | 615 (45.52) | 1005 (67.54) | 1529 (43.71) | 3991 (46.01) | 1305 (61.24) | 571 (47.11) | 641 (56.83) | 1670 (57.09) | 4187 (56.61) |
| Employed | 450 (19.25) | 617 (45.67) | 304 (20.43) | 1059 (30.27) | 2430 (28.01) | 732 (34.35) | 584 (48.18) | 355 (31.47) | 1038 (35.49) | 2709 (36.63) |
| Other | 91 (3.89) | 98 (7.25) | 122 (8.20) | 263 (7.52) | 574 (6.62) | 94 (4.41) | 57 (4.70) | 132 (11.70) | 217 (7.42) | 500 (6.76) |
| Homemaker | 955 (40.85) | 21 (1.55) | 57 (3.83) | 647 (18.50) | 1680 (19.37) | . . . | . . . | . . . | . . . | . . . |
| Cohabitation status | ||||||||||
| Living as a couple | 1894 (81.01) | 1029 (76.17) | 1058 (71.10) | 2578 (73.70) | 6559 (75.61) | 1893 (88.83) | 1025 (84.57) | 966 (85.64) | 2543 (86.94) | 6427 (86.90) |
| Living as single | 444 (18.99) | 322 (23.83) | 430 (28.90) | 920 (26.30) | 2116 (24.39) | 238 (11.17) | 187 (15.43) | 162 (14.36) | 382 (13.06) | 969 (13.10) |
Note. BM = Bismarckian regime; CASP = control, autonomy, self-realization, and pleasure measure; PC = Postcommunist regime; SC = Scandinavian regime; SEI = socioeconomic index; SO = Southern regime.
In age-adjusted multilevel models (Table 2), the difference in CASP-12 scores between the most and least advantaged according to the life-course socioeconomic index was 7.86 (95% confidence interval [CI] = 7.12, 8.59) among women; the equivalent result among men was 6.80 (95% CI = 6.04, 7.57). The index was also associated with life satisfaction; among women the difference between the least and most advantaged was 2.09 (95% CI = 1.87, 2.31) and among men it was 1.65 (95% CI = 1.43, 1.87). Substantive differences in the association between the life-course socioeconomic index and well-being were apparent between welfare regimes, as demonstrated by statistically significant interaction terms (Table 2, model 2). Compared with the Scandinavian regime, the association between the life-course socioeconomic index and well-being was consistently stronger in the Southern and Postcommunist welfare regimes. In Bismarckian countries, the association was always larger than in the Scandinavian regime, particularly for life satisfaction among women. However, the differences were not as apparent as between the other welfare regimes.
TABLE 2—
Age-Adjusted Linear Multilevel Models Containing Interaction Terms Between the Life-Course Socioeconomic Index, Ability to Make Ends Meet, and Welfare Regime for Well-Being by Gender, Among Adults Aged 50–75 Years: Survey of Health, Aging, and Retirement in Europe, 2006–2007
| Women, b (95% CI) |
Men, b (95% CI) |
|||||||
| Variable | Model 1: Life-Course SEI | Model 2: Life-Course SEI | Model 3: Ability to Make Ends Meet | Model 4: Ability to Make Ends Meet | Model 1: Life-Course SEI | Model 2: Life-Course SEI | Model 3: Ability to Make Ends Meet | Model 4: Ability to Make Ends Meet |
| CASP-12a | ||||||||
| Main effect | 7.86*** (7.12, 8.59) | 4.68*** (2.91, 6.44) | 2.40*** (2.27, 2.52) | 1.81*** (1.45, 2.17) | 6.80*** (6.04, 7.57) | 3.93*** (2.13, 5.73) | 2.21*** (2.08, 2.34) | 2.11*** (1.72, 2.51) |
| Welfare regime | ||||||||
| Southern | . . . | −9.03*** (–11.70, –6.35) | . . . | −5.25*** (–7.60, –2.90) | . . . | −7.22*** (–9.43, –5.00) | . . . | −2.41* (–4.61, –0.21) |
| Scandinavian (Ref) | . . . | . . . | . . . | . . . | . . . | . . . | . . . | . . . |
| Postcommunist | . . . | −8.40*** (–11.36, –5.44) | . . . | −5.78*** (–8.32, –3.24) | . . . | −7.35*** (–9.85, –4.85) | . . . | −2.64* (–5.04, –0.25) |
| Bismarckian | . . . | −2.39 (–4.81, 0.02) | . . . | −2.55* (–4.74, –0.35) | . . . | −1.84 (–3.85, 0.16) | . . . | −0.52 (–2.63, 1.59) |
| Interactions | ||||||||
| Southern | . . . | 6.19*** (3.93, 8.45) | . . . | 0.61** (0.18, 1.05) | . . . | 5.14*** (2.83, 7.45) | . . . | 0.06 (–0.40, 0.52) |
| Scandinavian (Ref) | . . . | . . . | . . . | . . . | . . . | . . . | . . . | . . . |
| Postcommunist | . . . | 5.21*** (2.63, 7.80) | . . . | 1.04*** (0.55, 1.53) | . . . | 5.45*** (2.67, 8.23) | . . . | 0.24 (–0.29, 0.76) |
| Bismarckian | . . . | 1.80 (–0.30, 3.90) | . . . | 0.56** (0.15, 0.97) | . . . | 1.84 (–0.30, 3.99) | . . . | 0.06 (–0.39, 0.51) |
| Life satisfactionb | ||||||||
| Main effect | 2.09*** (1.87, 2.31) | 0.81** (0.28, 1.33) | 0.54*** (0.50, 0.58) | 0.33*** (0.22, 0.44) | 1.65*** (1.43, 1.87) | 1.01*** (0.50, 1.52) | 0.47*** (0.43, 0.51) | 0.48*** (0.36, 0.59) |
| Welfare regime | ||||||||
| Southern | . . . | −2.36*** (–3.01, –1.71) | . . . | −1.51*** (–2.08, –0.94) | . . . | −1.44*** (–1.99, –0.90) | . . . | −0.43 (–0.96, 0.09) |
| Scandinavian (Ref) | . . . | . . . | . . . | . . . | . . . | . . . | . . . | . . . |
| Postcommunist | . . . | −3.42*** (–4.14, –2.69) | . . . | −2.34*** (–2.95, –1.73) | . . . | −2.50*** (–3.12, –1.88) | . . . | −1.25*** (–1.82, –0.69) |
| Bismarckian | . . . | −1.11*** (–1.70, –0.52) | . . . | −1.17*** (–1.72, –0.63) | . . . | −0.77** (–1.26, –0.27) | . . . | −0.26 (–0.77, 0.26) |
| Interactions | ||||||||
| Southern | . . . | 1.90*** (1.23, 2.57) | . . . | 0.22** (0.08, 0.35) | . . . | 0.81* (0.15, 1.46) | . . . | −0.03 (–0.16, 0.10) |
| Scandinavian (Ref) | . . . | . . . | . . . | . . . | . . . | . . . | . . . | . . . |
| Postcommunist | . . . | 3.21*** (2.44, 3.98) | . . . | 0.44*** (0.29, 0.59) | . . . | 2.11*** (1.32, 2.90) | . . . | 0.17* (0.01, 0.32) |
| Bismarckian | . . . | 0.71* (0.09, 1.34) | . . . | 0.18** (0.05, 0.30) | . . . | 0.37 (–0.24, 0.98) | . . . | −0.06 (–0.19, 0.07) |
Note. CI = confidence interval; SEI = socioeconomic index. Model 1 contains age, age-squared, and the life-course SEI. Model 2 contains age, age-squared, and the life-course SEI interacted with welfare regime variables. Model 3 contains age, age-squared, and ability to make ends meet. Model 4 contains age, age-squared, and ability to make ends meet interacted with welfare regime variables.
CASP-12 = control, autonomy, self-realization, and pleasure measure; a shortened version of CASP-19, a validated needs satisfaction-based measure of positive quality of life in early old age.
Life satisfaction = the cognitive evaluation of one’s life, measured by asking participants “On a scale from 0 to 10 where 0 means completely dissatisfied and 10 means completely satisfied, how satisfied are you with your life?”
*P < .05; **P < .01; ***P < .001.
Financial distress was also strongly associated with well-being (Table 2, model 3). Among women, a 1-unit increase in the ability to make ends meet was related to 2.40 (95% CI = 2.27, 2.52) higher CASP-12 scores; the same result among men was 2.21 (95% CI = 2.08, 2.34). There were gender differences in the interactions between the financial distress and welfare regime variables (Figure 1). Compared with the Scandinavian regime, the ability to make ends meet was more strongly related to women’s well-being in the other regimes, particularly in Postcommunist countries, whereas, among men, the ability to make ends meet had a generally similar association with well-being across welfare regimes. The only exception was for life satisfaction, in which, compared with the Scandinavian regime, the ability to make ends meet was more strongly associated in Postcommunist countries.
FIGURE 1—
Age-adjusted predicted mean life satisfaction by financial distress for each welfare regime among (a) women aged 50–75 years and (b) men aged 50–75 years: Survey of Health, Aging, and Retirement in Europe, 2006–2007.
We then stratified models by welfare regime to examine the influence of cohabitation, employment status, and financial distress on the association between the life-course socioeconomic index and well-being in the different welfare regimes. (Table 3 displays results for life satisfaction. Results for CASP-12 are available in the supplement to the online version of this article at http://www.ajph.org.)
TABLE 3—
Age-Adjusted Single-Level Linear Regression Models for the Association Between the Life-Course Socioeconomic Index and Life Satisfaction by Gender and Welfare Regime Among Adults Aged 50–75 Years: Survey of Health, Aging, and Retirement in Europe, 2006–2007
| Women, b (95% CI) |
Men, b (95% CI) |
|||||||
| Variable | Model 1 | Model 2 | Model 3 | Model 4 | Model 1 | Model 2 | Model 3 | Model 4 |
| Southern | ||||||||
| Life-course SEI | 2.65*** (2.21, 3.09) | 2.55*** (2.11, 2.99) | 2.39*** (1.93, 2.84) | 1.57*** (1.10, 2.05) | 1.75*** (1.32, 2.17) | 1.69*** (1.26, 2.12) | 1.54*** (0.11, 1.97) | 0.60** (0.15, 1.06) |
| Cohabitation | ||||||||
| Living as couple (Ref) | . . . | . . . | . . . | . . . | . . . | . . . | . . . | . . . |
| Living as single | . . . | −0.57*** (–0.75, –0.39) | −0.55*** (–0.73, –0.37) | −0.49*** (–0.67, –0.32) | . . . | −0.42*** (–0.63, –0.22) | −0.40*** (–0.60, –0.20) | −0.48*** (–0.68, –0.29) |
| Employment status | ||||||||
| Retired (Ref) | . . . | . . . | . . . | . . . | . . . | . . . | . . . | . . . |
| Employed | . . . | . . . | −0.06 (–0.29, 0.17) | −0.06 (–0.29, 0.16) | . . . | . . . | 0.12 (–0.08, 0.31) | 0.10 (–0.09, 0.29) |
| Other | . . . | . . . | −1.28*** (–1.65, –0.91) | −1.15*** (–1.51, –0.79) | . . . | . . . | −0.71*** (–1.04, –0.37) | −0.57*** (−0.90, –0.25) |
| Homemaker | . . . | . . . | −0.04 (–0.21, 0.13) | 0.02 (–0.15, 0.18) | . . . | . . . | . . . | . . . |
| Ability to make ends meet | . . . | . . . | . . . | 0.40*** (0.32, 0.48) | . . . | . . . | . . . | 0.39*** (0.31, 0.47) |
| % attenuation | . . . | 3.77 | 9.81 | 40.75 | . . . | 3.43 | 12.00 | 65.71 |
| Scandinavian | ||||||||
| Life-course SEI | 0.90*** (0.47, 1.33) | 0.66** (0.23, 1.09) | 0.47* (0.04, 0.89) | 0.25 (–0.19, 0.68) | 1.08*** (0.63, 1.54) | 0.74** (0.28, 1.19) | 0.62** (0.16, 1.08) | 0.34 (–0.12, 0.79) |
| Cohabitation | ||||||||
| Living as couple (Ref) | . . . | . . . | . . . | . . . | . . . | . . . | . . . | . . . |
| Living as single | . . . | −0.59*** (–0.75, –0.43) | −0.57*** (–0.73, –0.41) | −0.48*** (–0.64, –0.31) | . . . | −0.72*** (–0.92, –0.51) | −0.69*** (–0.89, –0.49) | −0.58*** (–0.78, –0.38) |
| Employment status | ||||||||
| Retired (Ref) | . . . | . . . | . . . | . . . | . . . | . . . | . . . | . . . |
| Employed | . . . | . . . | 0.22* (0.01, 0.43) | 0.14 (–0.07, 0.35) | . . . | . . . | 0.20 (–0.02, 0.43) | 0.16 (–0.06, 0.38) |
| Other | . . . | . . . | −0.62*** (–0.94, –0.31) | −0.66*** (–0.97, –0.36) | . . . | . . . | −0.32 (–0.70, 0.05) | −0.21 (–0.58, 0.16) |
| Homemaker | . . . | . . . | 0.04 (–0.53, 0.60) | 0.06 (–0.50, 0.62) | . . . | . . . | . . . | . . . |
| Ability to make ends meet | . . . | . . . | . . . | 0.24*** (0.15, 0.34) | . . . | . . . | . . . | 0.40*** (0.30, 0.50) |
| % attenuation | . . . | 26.67 | 47.78 | 72.22 | . . . | 31.48 | 42.59 | 68.52 |
| Postcommunist | ||||||||
| Life-course SEI | 3.94*** (3.26, 4.61) | 3.62*** (2.95, 4.30) | 3.44*** (2.75, 4.13) | 2.59*** (1.88, 3.29) | 3.10*** (2.35, 3.85) | 2.94*** (2.20, 3.69) | 2.46*** (1.72, 3.21) | 1.67*** (0.91, 2.43) |
| Cohabitation | ||||||||
| Living as couple (Ref) | . . . | . . . | . . . | . . . | . . . | . . . | . . . | . . . |
| Living as single | . . . | −0.61*** (–0.83, –0.39) | −0.61*** (–0.83, –0.39) | −0.45*** (–0.67, –0.23) | . . . | −0.62*** (–0.92, –0.32) | −0.48** (–0.78, –0.18) | −0.51*** (–0.80, –0.21) |
| Employment status | ||||||||
| Retired (Ref) | . . . | . . . | . . . | . . . | . . . | . . . | . . . | . . . |
| Employed | . . . | . . . | 0.37* (0.03, 0.71) | 0.31 (–0.02, 0.65) | . . . | . . . | 0.36* (0.02, 0.70) | 0.34* (0.01, 0.66) |
| Other | . . . | . . . | −0.23 (–0.64, 0.17) | −0.04 (–0.44, 0.36) | . . . | . . . | −0.99*** (–1.38, –0.59) | −0.81*** (–1.20, –0.43) |
| Homemaker | . . . | . . . | 0.10 (–0.43, 0.63) | 0.13 (−0.39, 0.65) | . . . | . . . | . . . | . . . |
| Ability to make ends meet | . . . | . . . | . . . | 0.54*** (0.42, 0.67) | . . . | . . . | . . . | 0.47*** (0.34, 0.60) |
| % attenuation | . . . | 8.12 | 12.69 | 34.26 | . . . | 5.16 | 20.65 | 46.13 |
| Bismarckian | ||||||||
| Life-course SEI | 1.57*** (1.25, 1.88) | 1.33*** (1.02, 1.64) | 1.23*** (0.91, 1.54) | 0.60*** (0.28, 0.92) | 1.41*** (1.10, 1.71) | 1.30*** (0.99, 1.60) | 1.09*** (0.78, 1.40) | 0.61*** (0.30, 0.93) |
| Cohabitation | ||||||||
| Living as couple (Ref) | . . . | . . . | . . . | . . . | . . . | . . . | . . . | . . . |
| Living as single | . . . | −0.67*** (–0.78, –0.56) | −0.65*** (–0.77, –0.54) | −0.50*** (–0.61, –0.38) | . . . | −0.46*** (–0.60, –0.31) | −0.43*** (–0.57, –0.28) | −0.42*** (–0.57, –0.28) |
| Employment status | ||||||||
| Retired (Ref) | . . . | . . . | . . . | . . . | . . . | . . . | . . . | . . . |
| Employed | . . . | . . . | 0.17* (0.00, 0.33) | 0.13 (–0.03, 0.29) | . . . | . . . | 0.24** (0.08, 0.41) | 0.22** (0.06, 0.38) |
| Other | . . . | . . . | −0.39*** (−0.61, –0.18) | −0.25* (–0.46, –0.04) | . . . | . . . | −0.49*** (–0.71, –0.27) | −0.29** (–0.51, –0.07) |
| Homemaker | . . . | . . . | 0.07 (–0.08, 0.21) | 0.09 (–0.06, 0.23) | . . . | . . . | . . . | . . . |
| Ability to make ends meet | . . . | . . . | . . . | 0.40*** (0.34, 0.46) | . . . | . . . | . . . | 0.33*** (0.27, 0.39) |
| % attenuation | . . . | 15.29 | 21.66 | 61.78 | . . . | 7.80 | 22.70 | 56.74 |
Note. CI = confidence interval; SEI = socioeconomic index. Model 1 contains age, age-squared, the life-course SEI, and country dummy variables. Model 2 contains age, age-squared, the life-course SEI, country dummy variables, and cohabitation status. Model 3 contains age, age-squared, the life-course SEI, country dummy variables, cohabitation status, and employment status. Model 4 contains age, age-squared, the life-course SEI, country dummy variables, cohabitation status, employment status, and ability to make ends meet.
*P < .05; **P < .01; ***P < .001.
When we included the ability to make ends meet variable in the models, the effect size for the life-course socioeconomic index was greatly attenuated and the attenuation was larger than that observed for cohabitation and employment status. This was a consistent finding among both genders, outcomes, and all welfare regimes. However, the degree of attenuation by financial distress varied among welfare regimes; for life satisfaction it ranged from 34.26% among women in the Postcommunist regime to 72.22% among women in the Scandinavian regime. In Scandinavian countries, the association between the life-course socioeconomic index and life satisfaction was no longer statistically significant in the fully adjusted model; in this group the index had a relatively weaker association compared with the other regimes. We observed the strongest relationship between the life-course socioeconomic index and well-being, which persisted even when we adjusted for financial distress, among women in the Southern and Postcommunist regimes.
DISCUSSION
Increased socioeconomic advantage over the life course was associated with higher well-being in early old age. This relationship was present in each welfare regime and for both well-being outcomes, but the magnitude of the associations varied by welfare state. The associations were strongest among the Southern and Postcommunist regimes and weakest in the Bismarckian and Scandinavian countries. Life-course SEP also exhibited a stronger association with well-being among Bismarckian countries compared with Scandinavian, but not to the same extent as between the other regimes. Financial distress had a substantial independent impact on well-being and also attenuated the relationship between life-course SEP and well-being to a large extent. The extent of attenuation varied between welfare regimes. In particular, the life-course socioeconomic index exhibited little or no association with life satisfaction in the Scandinavian regime after we included financial distress.
These findings suggest that feeling able to make ends meet is a key determinant of well-being in early old age. This is likely because of resultant stress and anxiety associated with financial distress, which may lead to feelings of being unable to control one’s life. It also hinders participation in many costly activities that may bring self-realization and pleasure to those who can afford it, but lead to added worry and feelings of being left out among those who cannot.36 Other studies have also demonstrated that financial issues are likely to be a key mediator of socioeconomic inequalities in health.37,38 However, this study is to our knowledge the first to investigate financial distress as a mediator of life-course inequalities in well-being in early old age, which may explain some of the contrasting findings of a recent study that found that Nordic countries experienced relatively large social class inequalities in self-rated health that persisted with the inclusion of financial strain.38
Although in this sample Scandinavian and Bismarckian welfare states contained fewer individuals reporting that they experienced great difficulty making ends meet, this group still experienced poorer well-being compared with those who felt able to get by with ease. As the relationship between the life-course socioeconomic index and well-being was greatly reduced when we included the ability to make ends meet variable, it suggests that feelings of financial distress are important mediators of the relationship between life-course SEP and well-being. In some groups, particularly among those in the Scandinavian regime, where life-course SEP had a weaker association with well-being, feeling able to make ends meet resulted in the complete explanation of the association with life satisfaction. This perhaps suggests that Scandinavian welfare states were better able to prevent socioeconomic disadvantage over the life course translating into greater financial distress in early old age and in turn lowering well-being. Financial distress may also be a better measure of socioeconomic circumstances, which is able to capture the multidimensional nature of poverty and resulting social exclusion.39
Financial strain in early old age may, in part, also reflect socioeconomic circumstances over the life course. Resources available at this life stage are likely to reflect the accumulation (or not) of wealth as well as the possibility of inheritance from family. However, not all individuals on low income reported feeling financial distress, and feeling able to make ends meet is likely to be influenced by a range of factors, including social comparisons with others.36
The results of our study should be considered in light of its strengths and limitations. Key strengths include the use of cross-nationally comparable data and the inclusion of 2 well-being outcome measures, which enabled us to examine the consistency of the results. The use of an innovative relative measure of life-course SEP that considered different dimensions of SEP from childhood to early old age and took into account potential differences in the distributions of the socioeconomic variables by country, gender, and cohort was also an important strength. However, the analyses were unweighted and attrition and survival bias may have affected our results, though we expect that this may lead to an underestimation of some results,40 for example, among Postcommunist countries with lower life expectancies. We also cannot rule out the possibility that cultural factors influenced how people responded to the questionnaire items. Furthermore, we have not explored the separate associations between childhood and adulthood SEP and well-being or examined the potential contribution that moving between socioeconomic advantage and disadvantage across the life course may make to later life well-being.
Our findings have important implications for policy. Alleviating financial distress in early old age could help to improve the quality of life of this population and reduce the inequality in well-being that is associated with the accumulation of socioeconomic experiences across the life course. However, policies that tackle socioeconomic inequalities as they arise across the whole life course are required to ensure that the difference in well-being between the most and least advantaged does not widen among future generations of older people. Further research is required to uncover which particular policies matter for reducing socioeconomic inequalities in well-being among older age groups. However, our results demonstrate that welfare regimes that have tended to be more generous in terms of social protection appear to be more effective at reducing life-course inequalities in well-being among both genders, as well as financial distress among women.
Acknowledgments
No specific funding was received for this study. This article uses data from Survey of Health, Ageing and Retirement in Europe (SHARE) wave 2 release 2.5.0, as of May 24, 2011, and SHARELIFE release 1, as of November 24, 2010. The SHARE data collection has been primarily funded by the European Commission through the 5th Framework Programme (project QLK6-CT-2001-00360 in the thematic programme Quality of Life), through the 6th Framework Programme (projects SHARE-I3, RII-CT-2006-062193, COMPARE, CIT5- CT-2005-028857, and SHARELIFE, CIT4-CT-2006-028812), and through the 7th Framework Programme (SHARE-PREP, No. 211909, SHARE-LEAP, No. 227822, and SHARE M4, No. 261982). Additional funding from the US National Institute on Aging (U01AG09740-13S2, P01 AG005842, P01 AG08291, P30 AG12815, R21 AG025169, Y1- AG-4553-01, IAG BSR06-11, and OGHA 04-064) and the German Ministry of Education and Research, as well as from various national sources, is gratefully acknowledged. (See http://www.share-project.org for a full list of funding institutions.)
The authors would like to thank Vittal Katikireddi for providing comments on a draft of the article, the anonymous reviewers for greatly improving the article, and the participants of SHARE for making the study possible.
Human Participant Protection
This study is an analysis of previously collected data and, therefore, ethics approval was not required. Ethics approval for SHARE was obtained by the SHARE team (http://www.share-project.org).
References
- 1.Niedzwiedz CL, Katikireddi SV, Pell JP, Mitchell R. Life course socio-economic position and quality of life in adulthood: a systematic review of life course models. BMC Public Health. 2012;12(1):628. doi: 10.1186/1471-2458-12-628. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 2.Bowen ME, Gonzalez HM. Childhood socioeconomic position and disability in later life: results of the Health and Retirement Study. Am J Public Health. 2010;100(suppl 1):S197–S203. doi: 10.2105/AJPH.2009.160986. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 3.Blane D, Webb E, Wahrendorf M, Netuveli G. Life course influences on quality of life at age 50 years: evidence from the National Child Development Study (1958 British birth cohort study) Longitudinal Life Course Stud. 2012;3(3):346–358. [Google Scholar]
- 4.Niedzwiedz CL, Katikireddi SV, Pell JP, Mitchell R. The association between life course socioeconomic position and life satisfaction in different welfare states: European comparative study of individuals in early old age. Age Ageing. 2014;43(3):431–436. doi: 10.1093/ageing/afu004. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 5.Wahrendorf M, Blane D. Does labour market disadvantage help to explain why childhood circumstances are related to quality of life at older ages? Results from SHARE. Aging Ment Health. 2015;19(7):584–594. doi: 10.1080/13607863.2014.938604. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 6.Esping-Andersen G. The Three Worlds of Welfare Capitalism. Cambridge, UK: Polity Press; 1990. [Google Scholar]
- 7.Bergqvist K, Aberg Yngwe M, Lundberg O. Understanding the role of welfare state characteristics for health and inequalities—an analytical review. BMC Public Health. 2013;13(1):1234. doi: 10.1186/1471-2458-13-1234. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 8.Bambra C. Going beyond the three worlds of welfare capitalism: regime theory and public health research. J Epidemiol Community Health. 2007;61(12):1098–1102. doi: 10.1136/jech.2007.064295. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 9.Mackenbach JP. The persistence of health inequalities in modern welfare states: the explanation of a paradox. Soc Sci Med. 2012;75(4):761–769. doi: 10.1016/j.socscimed.2012.02.031. [DOI] [PubMed] [Google Scholar]
- 10.Brennenstuhl S, Quesnel-Vallée A, McDonough P. Welfare regimes, population health and health inequalities: a research synthesis. J Epidemiol Community Health. 2012;66(5):397–409. doi: 10.1136/jech-2011-200277. [DOI] [PubMed] [Google Scholar]
- 11.Mackenbach JP, Stirbu I, Roskam A-JR et al. Socioeconomic inequalities in health in 22 European countries. N Engl J Med. 2008;358(23):2468–2481. doi: 10.1056/NEJMsa0707519. [DOI] [PubMed] [Google Scholar]
- 12.Popham F, Dibben C, Bambra C. Are health inequalities really not the smallest in the Nordic welfare states? A comparison of mortality inequality in 37 countries. J Epidemiol Community Health. 2013;67(5):412–418. doi: 10.1136/jech-2012-201525. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 13.Eikemo TA, Huisman M, Bambra C, Kunst AE. Health inequalities according to educational level in different welfare regimes: a comparison of 23 European countries. Sociol Health Illn. 2008;30(4):565–582. doi: 10.1111/j.1467-9566.2007.01073.x. [DOI] [PubMed] [Google Scholar]
- 14.Espelt A, Borrell C, Rodriguez-Sanz M et al. Inequalities in health by social class dimensions in European countries of different political traditions. Int J Epidemiol. 2008;37(5):1095–1105. doi: 10.1093/ije/dyn051. [DOI] [PubMed] [Google Scholar]
- 15.Kok R, Avendano M, Bago d’Uva T, Mackenbach J. Can reporting heterogeneity explain differences in depressive symptoms across Europe? Soc Indic Res. 2012;105(2):191–210. doi: 10.1007/s11205-011-9877-7. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 16.Niedzwiedz CL, Katikireddi SV, Pell JP, Mitchell R. Socioeconomic inequalities in the quality of life of older Europeans in different welfare regimes. Eur J Public Health. 2014;24(3):364–370. doi: 10.1093/eurpub/cku017. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 17. World Health Organization. Preamble to the Constitution of the World Health Organization as adopted by the International Health Conference, New York, 19–22 June, 1946; signed on 22 July 1946 by the representatives of 61 States (Official Records of the World Health Organization, no. 2, p. 100) and entered into force on 7 April 1948. 1948.
- 18.Motel-Klingebiel A, Romeu Gordo L, Betzin J. Welfare states and quality of later life: distributions and predictions in a comparative perspective. Eur J Ageing. 2009;6(2):67–78. doi: 10.1007/s10433-009-0112-1. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 19.Eikemo TA, Hoffmann R, Kulik MC et al. How can inequalities in mortality be reduced? A quantitative analysis of 6 risk factors in 21 European populations. PLoS ONE. 2014;9(11):e110952. doi: 10.1371/journal.pone.0110952. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 20.Laaksonen M, Talala K, Martelin T et al. Health behaviours as explanations for educational level differences in cardiovascular and all-cause mortality: a follow-up of 60 000 men and women over 23 years. Eur J Public Health. 2008;18(1):38–43. doi: 10.1093/eurpub/ckm051. [DOI] [PubMed] [Google Scholar]
- 21.Stringhini S, Dugravot A, Shipley M et al. Health behaviours, socioeconomic status, and mortality: further analyses of the British Whitehall II and the French GAZEL prospective cohorts. PLoS Med. 2011;8(2):e1000419. doi: 10.1371/journal.pmed.1000419. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 22.Stiglitz JE, Sen AK, Fitoussi JP. Commission on the Measurement of Economic Performance and Social Progress. Mismeasuring Our Lives: Why GDP Doesn’t Add Up. New York, NY: The New Press; 2010. [Google Scholar]
- 23.Costanza R, Kubiszewski I, Giovannini E et al. Development: time to leave GDP behind. Nature. 2014;505(7483):283–285. doi: 10.1038/505283a. [DOI] [PubMed] [Google Scholar]
- 24. Mannheim Research Institute for the Economics of Aging. SHARE release guide 2.5.0 waves 1 and 2. Mannheim, Germany: Mannheim Research Institute for the Economics of Aging; 2011.
- 25.Berney LR, Blane DB. Collecting retrospective data: accuracy of recall after 50 years judged against historical records. Soc Sci Med. 1997;45(10):1519–1525. doi: 10.1016/s0277-9536(97)00088-9. [DOI] [PubMed] [Google Scholar]
- 26.Börsch-Supan A, Brandt M, Hunkler C et al. Data resource profile: the Survey of Health, Ageing and Retirement in Europe (SHARE) Int J Epidemiol. 2013;42(4):992–1001. doi: 10.1093/ije/dyt088. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 27.Knesebeck O, Hyde M, Higgs P, Kupfer A, Siegrist J. Quality of life and well-being. In: Börsch-Supan A, Brugiavini A, Jürges H, Mackenbach J, Siegrist J, Weber G, editors. Health, Ageing and Retirement in Europe. First Results From the Survey of Health, Ageing and Retirement in Europe. Mannheim, Germany: Mannheim Research Institute for the Economics of Aging; 2005. [Google Scholar]
- 28.Howel D. Interpreting and evaluating the CASP-19 quality of life measure in older people. Age Ageing. 2012;41(5):612–617. doi: 10.1093/ageing/afs023. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 29.Hyde M, Wiggins RD, Higgs P, Blane DB. A measure of quality of life in early old age: the theory, development and properties of a needs satisfaction model (CASP-19) Aging Ment Health. 2003;7(3):186–194. doi: 10.1080/1360786031000101157. [DOI] [PubMed] [Google Scholar]
- 30.Wiggins R, Netuveli G, Hyde M, Higgs P, Blane D. The evaluation of a self-enumerated scale of quality of life (CASP-19) in the context of research on ageing: a combination of exploratory and confirmatory approaches. Soc Indic Res. 2008;89(1):61–77. [Google Scholar]
- 31.Deindl C. The influence of living conditions in early life on life satisfaction in old age. Adv Life Course Res. 2013;18(1):107–114. doi: 10.1016/j.alcr.2012.10.008. [DOI] [PubMed] [Google Scholar]
- 32.Eichhorn J. Happiness for believers? Contextualizing the effects of religiosity on life-satisfaction. Eur Sociol Rev. 2012;28(5):583–593. [Google Scholar]
- 33.Pittau MG, Zelli R, Gelman A. Economic disparities and life satisfaction in European regions. Soc Indic Res. 2010;96(2):339–361. [Google Scholar]
- 34. Niedzwiedz C. An Examination of the Relationship Between Life Course Socioeconomic Position and Quality of Life Among Europeans in Early Old Age and the Influence of the Welfare Regime [PhD thesis]. Glasgow, Scotland: University of Glasgow; 2014.
- 35.Bambra C. Health inequalities and welfare state regimes: theoretical insights on a public health “puzzle.”. J Epidemiol Community Health. 2011;65(9):740–745. doi: 10.1136/jech.2011.136333. [DOI] [PubMed] [Google Scholar]
- 36.Arber S, Fenn K, Meadows R. Subjective financial well-being, income and health inequalities in mid and later life in Britain. Soc Sci Med. 2014;100:12–20. doi: 10.1016/j.socscimed.2013.10.016. [DOI] [PubMed] [Google Scholar]
- 37.Aldabe B, Anderson R, Lyly-Yrjänäinen M et al. Contribution of material, occupational, and psychosocial factors in the explanation of social inequalities in health in 28 countries in Europe. J Epidemiol Community Health. 2011;65(12):1123–1131. doi: 10.1136/jech.2009.102517. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 38.Shaw RJ, Benzeval M, Popham F. To what extent do financial strain and labour force status explain social class inequalities in self-rated health? Analysis of 20 countries in the European Social Survey. PLoS ONE. 2014;9(10):e110362. doi: 10.1371/journal.pone.0110362. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 39.Nolan B, Whelan CT. Using non-monetary deprivation indicators to analyze poverty and social exclusion: lessons from Europe? J Policy Anal Manage. 2010;29(2):305–325. [Google Scholar]
- 40.Mein G, Johal S, Grant RL, Seale C, Ashcroft R, Tinker A. Predictors of two forms of attrition in a longitudinal health study involving ageing participants: an analysis based on the Whitehall II study. BMC Med Res Methodol. 2012;12(1):164. doi: 10.1186/1471-2288-12-164. [DOI] [PMC free article] [PubMed] [Google Scholar]