Current and Expected Economic Hardship and Satisfaction With Family Life in Europe

Abstract

This study investigated how people’s satisfaction with their family life is influenced by economic circumstances. Expectations were formulated that people who experienced or expected economic hardship would be less satisfied with their family life. Additionally, it was hypothesized that current and expected economic hardship would amplify each other’s consequences on satisfaction, and that current and expected economic hardship was more harmful for people with children and when the rise of unemployment in a country was larger. Multilevel analyses were conducted using a sample from the European Quality of Life Survey 2012 (N = 13,013 in 30 countries). Results indeed indicated that people who experienced or expected economic hardship were less satisfied with their family life. Expecting a financial decline was (slightly) more harmful for people in larger families. Generally, current and expected economic problems were not more harmful for parents or when a country’s rise of unemployment was larger.

Introduction

Many families have experienced economic hardship during the recent economic crises in Europe, such as the loss of employment or income (Eurofound, 2015; Standing, 2011). Although direct consequences of economic hardship are largely employment-related, they likely spill over to people’s family life because economic difficulties may foster conflicts between partners due to stress, frustration, or depression and because it complicates making future plans for the family. While previous research often studied individual consequences for a person who loses a job or income, all family members may suffer from precarious circumstances, implying that economic crises hit more people than the ones directly involved. This study thus aims to enlarge our knowledge about people’s satisfaction with family life and the consequences of economic hardship for family life satisfaction by simultaneously considering two key dimensions of hardship; namely, current and expected economic hardship; and by examining under which circumstances economic hardship reduces family life satisfaction more strongly.

While economic hardship encompasses both current economic hardship and people’s uncertainty for their economic future (Standing, 2011), most research on economic circumstances and family well-being focusses on one of the two. This article will therefore improve on this research by simultaneously examining current economic hardship and expectations about future economic hardship. More precisely, this study focusses on (current and expected) financial difficulties and unemployment as two primary indicators of economic hardship. In prior studies, a focus on the consequences of current economic hardship dominated research on family well-being. These studies generally concluded that people who perceived or experienced economic hardship reported a lower quality of their partner relationship, reported more conflict between family members, and were less satisfied with their family life (see for instance, Conger, Conger, & Martin, 2010; Currie, Duque, & Garfinkel, 2015; Dew & Yorgason, 2009; Shim, Lee, & Kim, 2017; Vinokur, Price, & Caplan, 1996; Williams, Cheadle, & Goosby, 2015). The current study explicitly focuses on family life satisfaction, which can be defined as “a conscious cognitive judgment of one’s family life in which the criteria for the judgment are up to the individual” (Zabriskie & Ward, 2013, p. 449). This judgment includes an evaluation of the relationship with their spouse, children, and parents. Because satisfaction with family life is strongly related to family functioning, cohesion, and communication (Zabriskie & Ward, 2013), it is a valuable overarching indicator to study family well-being.

This article will extend previous research findings also by investigating whether expected economic hardship in the near future reduces family life satisfaction. People’s perceptions and expectations on future problems likely influence families regardless of actual current hardship (Rosino, 2016). Families make plans for the future, and these plans are partly based on their perceived economic prospects. Expecting economic hardship may make planning for the future troublesome and more difficult (Hofmann & Hohmeyer, 2013), and the uncertainty about a family’s economic future deteriorates its members’ well-being (Modrek & Cullen, 2013; Standing, 2011). Prior research on the relationship between people’s expectations about their future economic hardship and family well-being is very scarce, but available research on relationship quality has shown that couples who are uncertain about the future report lower quality relationships, less affection, and poorer family communication (Kinnunen & Pulkkinen, 1998; Larson, Wilson, & Beley, 1994; Mauno, Cheng, & Lim, 2017). Other studies suggested that people are more prone to postpone long-term family investments and commitments when they experience economic uncertainty (De Lange, Wolbers, Gesthuizen, & Ultee, 2014; Hofmann & Hohmeyer, 2013). By simultaneously investigating current and expected economic hardship, it is possible to disentangle whether expected economic hardship harms family life satisfaction over and above current hardship, and whether the expectation of future economic difficulties is even more harmful for those who already experience economic hardship.

Another important contribution of this study is its assessment whether the impact of economic hardship on family life satisfaction depends on different contextual circumstances. Whether economic circumstances translate to perceptions of hardship and stress may depend partly on the family and societal context (Boss, Bryant, & Mancini, 2017). Previous studies primarily investigated direct associations between current economic hardship and family well-being as well as the mechanism explaining them (e.g., Williams et al., 2015). What remains unclear is whether economic hardship reduces family life satisfaction differently for people in different situations. This article aims to bridge this gap by studying two possibly moderating factors: the presence of children in the household and the macro-economic situation in a country. Previous studies indicated that parents are more rejecting of financial risks than childless people (Chaulk, Johnson, & Bulcroft, 2003), and being able to provide for their children showed to be especially important for mothers’ well-being (Mistry, Lowe, Benner, & Chien, 2008). These findings highlight the importance parents place on taking financial responsibility for their children. Because parents have more financial responsibilities than childless couples do, economic problems may especially be harmful for their personal and family well-being. Research also indicated that children have more psychological and behavioral problems when they experience economic hardship (Conger et al., 2002; Conger et al., 2010; Neppl, Senia, & Donnellan, 2016). Their problems could potentially transfer to parents’ satisfaction with family life, again leading to the expectation that economic problems hit couples with children harder than couples without children.

Previous research has mainly studied the influence of financial difficulties on families during macro-economic recession (see for instance, Aytaç & Rankin, 2009; Conger et al., 1990; Kwon, Rueter, Lee, Koh, & Ok, 2003). It remains unclear if this macro-economic context affects the degree to which hardship influences a person’s family life (Boss et al., 2017). Related research in the field of personal well-being suggests that economic difficulties influence a person’s well-being differently in diverse macro-economic contexts. Empirical support for this expectation is mixed. The well-being of lower educated women is more strongly affected by changes in macro-level unemployment than the well-being of other women (Currie et al., 2015), whereas macro-level unemployment does not alter the influence of personal unemployment on well-being (Eichhorn, 2012; Oesch & Lipps, 2012). In the current study, it is acknowledged that not all countries experienced the same degree of economic hardship during the previous crisis, and this variation is used as a natural experiment. By studying changes in unemployment in various European countries, this article investigates whether the severity of an economic crisis influences how people’s perceptions of their own labor circumstances affect their satisfaction with family life.

In sum, this study builds on previous research in several ways. First, it studies whether the expectation of economic hardship influences satisfaction with family life alongside with or in combination with current hardship. Second, it investigates whether the associations between current and expected economic hardship and family life satisfaction differ for people with or without children and for people in countries with a lower or higher rise in unemployment. In this article, we also take differences between men and women into account by studying whether economic hardship influences men’s and women’s family life satisfaction differently. Men’s employment is often regarded to be more important for families due to a larger economic contribution to the household and due to internalized and societal traditional gender roles (Shim et al., 2017; Williams et al., 2015). Hence, men and women may react differently to economic hardship, which could affect their family life satisfaction in another way (Shim et al., 2017; Williams et al., 2015).

The research question reads as follows:

• Research Question: To what extent do current and expected economic hardship affect satisfaction with family life, and how do these associations differ between families with and without children and in times of greater unemployment?

This research question will be answered using data from the European Quality of Life Study (EQLS) of 2012, which provided information on 13,013 partnered individuals of working age across 30 European countries. The EQLS data provide a unique opportunity to investigate simultaneously about the relationship between economic hardship and satisfaction with family life at the height of the European economic crisis across countries in various degrees of economic downturn. This cross-national variation in (changes in) unemployment enables us to study consequences of a country’s unemployment situation for the impact of economic hardship on a person’s satisfaction with their family life.

Theoretical Background

The Family Stress Model: Main Effects

Previous research focusing on current economic hardship and conflict within a partner relationship often departs from the family stress model (Conger et al., 1990; Conger et al., 2010). This model was designed to study the consequences of the Midwest farm crisis in the 1980s for families and has found confirmation during other crises across the globe (e.g., Aytaç & Rankin, 2009; Falconier & Epstein, 2010; Kwon et al., 2003). A basic proposition of family stress theory states that negative economic events increase the economic pressure people experience (Conger et al., 1990; Conger et al., 2010). Economic pressure includes unmet material needs, inability to pay bills, and financial cutbacks. How people react toward negative economic conditions depends on their perception of these conditions (Boss et al., 2017; Hill, 1958). Previous studies emphasized the importance to differentiate between objective and subjective experiences of stressful circumstances. Objective negative economic events such as job loss lead to subjective experiences of these events, resulting in stress (Hill, 1958; Rosino, 2016). The subjective experiences or perceptions do not have to originate from actual events and can induce stress similarly to objective experiences (Hill, 1958). When people perceive an event as stressful or frustrating, it is presumed they become more emotionally and behaviorally distressed, as expected from, for instance, frustration–aggression arguments and the ABC-X model (Berkowitz, 1989; Hill, 1958). Emotional consequences of economic pressure may include depression and anger. Both influence satisfaction with family life negatively because these increase “aggressive or angry responses, such as criticism, defensiveness, and insensitivity, as well as withdrawal of supportive behaviors” (Conger et al., 2002, p. 181). Anger also causes hostility between partners and therefore augments possible conflict, while depression induces emotional withdrawal and distancing, reducing satisfaction with family life (see also Conger et al., 2010).

Aside from the indirect consequences of economic pressure via stress, anger, and depression, economic pressure also has a more direct effect on people’s satisfaction with family life. Economic hardship causes partners to have arguments over financial issues, and these quarrels proved more problematic and recurrent than arguments about other issues (Papp, Cummings, & Goeke-Morey, 2009). Next to inducing financial pressure, unemployment reduces people’s structured time, social contact, sense of purpose, status, and activity, which are important psychological needs (Paul & Moser, 2009). The lack of fulfilment of these needs results in more distress, which in turn might result in people being less satisfied with their family life. Therefore, first hypothesis reads as follows:

• Hypothesis 1: When people experience current economic hardship, they are less satisfied with family life.

Similar to present-day economic hardship, the expectation of economic hardship likely influences family life. Perceptions of the future can shape current behavior and thus influence families (Mantler, Matejicek, Matheson, & Anisman, 2005; Rosino, 2016). People who are uncertain about their future economic situation likely anticipate an economic pressure in the near future and therefore may become more distressed, anxious, and both emotionally and physically exhausted (Mantler et al., 2005; Standing, 2011). According to the family stress model, such emotional consequences induce more aggressiveness and angry responses toward a partner, and less supportive behavior, such as showing empathy and interest, and listening (Conger et al., 2002). Economic insecurity has been found to reduce marital adjustment, family communication, affective involvement, and the quality of the relationship (Kinnunen & Pulkkinen, 1998; Larson et al., 1994; Mauno et al., 2017). Accordingly, expecting future economic hardship may have similar consequences for a person’s family life satisfaction as current economic hardship. Moreover, expecting economic hardship may result in postponement of long-term family commitments. People may be less inclined to marry or to have children when they are uncertain about future employment, earnings, or their households’ economic situation (De Lange et al., 2014; Hofmann & Hohmeyer, 2013). Delay of long-term commitments has been found to be negatively related to perceived quality of partner relationships (Wiik, Keizer, & Lappegård, 2012). Therefore, the second hypothesis reads as follows:

• Hypothesis 2: When people expect economic hardship, they are less satisfied with family life.

Moreover, people who currently experience economic hardship may be less able to handle future economic hardship. When people experience a decline in income, they need to adjust their spending to their new level of income. However, when people are already barely able to make ends meet, an additional loss of income would require extra financial cutbacks even though families may be unable to do so without cutting back on important needs. Therefore, expecting economic hardship may be especially stressful for families that currently are already experiencing economic hardship. Hence, the third hypothesis reads as follows:

• Hypothesis 3: Expected economic hardship is more detrimental for people’s satisfaction with family life when they currently experience more economic hardship.

Conditional Influences of Having Children and Macro-Level Unemployment

It is likely that not all families react to economic hardship to the same degree. The impact of current and expected economic hardship on family life satisfaction is argued to depend on the presence of children in the household and the rise of unemployment in a country. First, the extended family stress model—which incorporated parenting and children’s well-being into the original family stress model (Conger et al., 2002; Conger et al., 2010)—argues that the experience of economic pressure not only results in lower quality interactions with the partner but may also lead to more aggressive or angry responses in the interaction with children (Conger et al., 2002; Conger et al., 2010; Neppl et al., 2016). It is argued that emotional and relational consequences of economic pressure make parenting more harsh, inconsistent, and uninvolved (see Conger et al., 2010, for a review). The extended family stress model hypothesizes that children in families with economic difficulties experience more psychological (e.g., anxiety and depression) and behavioral problems (e.g., aggressive and antisocial), and children are less attached to their parents (Neppl et al., 2016). We expect that such behavioral problems would also relate negatively to parental family life satisfaction. In addition, because parents likely have more financial responsibilities than childless people (Chaulk et al., 2003), current and expected hardship may be more stressful for parents than nonparents. Negative economic events, may be perceived as a more substantial problem and therefore will be experienced as more stressful by parents than by people without children (Boss et al., 2017), because of their need and strong desire to financially take care of their children (Mistry et al., 2008). Taking all arguments together, it is hypothesized as follows:

• Hypothesis 4: (a) Current and (b) expected economic hardship are more detrimental for people’s satisfaction with family life if they have children.

Second, macro-economic circumstances may influence whether personal hardship is perceived as stressful (Boss et al., 2017). Specifically, the unemployment situation in a country may influence the impact of economic hardship on family life satisfaction. High and rising unemployment rates signal few job vacancies. Moreover, rising unemployment levels in a country amplify stress among people who experience or expect economic hardship since they observe limited opportunities to improve their income or employment situation (Oesch & Lipps, 2012). Congruent with these arguments, Currie et al. (2015) found that a rise in unemployment was especially harmful for the well-being of lower educated mothers, who could be considered to be in more precarious positions than higher educated mothers. Additionally, people were also found to experience more work stress when there had been more layoffs in their company (Modrek & Cullen, 2013), suggesting that being at risk of becoming unemployed is more stressful in situations where many people lose their jobs. Our final hypothesis is as follows:

• Hypothesis 5: (a) Current and (b) expected economic hardship are more detrimental for people’s satisfaction with family life in countries with a large rise in unemployment.

Data and Method

EQLS of 2012 was used to test our hypotheses. EQLS data were collected in EU27 countries in late 2011, and the beginning of 2012 using face-to-face interviews. In the summer of 2012, interviews were held in seven additional countries, namely Croatia, Iceland, Kosovo, the Former Yugoslav Republic (FYR) of Macedonia, Montenegro, Serbia, and Turkey. The response rate was 41.3% in the EU27 countries and 44.7% in the other countries (see www.eurofound.europa.eu). Because of missing information on country characteristics, Kosovo, FYR Macedonia, Montenegro, and Serbia were omitted from the analyses. Only people who cohabited with a different-sex partner and lived independently of their own or partner’s parents were selected for this study. The sample was further restricted to people of working age (18 to 65 years old) and whose main daily activity was either being employed, unemployed, or homemaker, therefore excluding people who were retired, unable to work due to long-term illness or disability, or in education. Next, we selected people whose partner was also of working age (18 to 65 years old) and were also either employed, unemployed, or homemaker. These criteria ensured that this study solely included people whose financial situation reflected their personal or their partner’s circumstances rather than their parents’ resources, and referred to people’s satisfaction with their own family life instead of satisfaction with their family of origin. Following these inclusion criteria, the sample consisted of 13,013 individuals across 30 countries. Missing values on independent variables were multiply imputed per country using all other independent variables and main daily activity of both partners, and work hours as predictor variables. The question on expected job loss was only applicable to people in employment. Consequently, analyses regarding expected job loss were based on a restricted sample of 10,177 employed persons.

Measurements

The dependent variable satisfaction with family life was measured by asking people how satisfied they were with their family life. The scale ranged from 1 (very dissatisfied) to 10 (very satisfied). This question was generally interpreted as referring to the nuclear family (Chapman & Guven, 2016; Greenstein, 2009; Shim et al., 2017). Similar questions have been included in, for instance, the Satisfaction With Family Life Scale (Zabriskie & Ward, 2013). The distribution of this dependent variable was highly skewed (M = 8.47), which was taken into account in the robustness analyses by using a negative binomial multilevel modelling. These analyses led to similar conclusions as using the original scale.

As aforementioned, current economic hardship was operationalized as experiencing financial hardship and unemployment. Financial hardship was measured using the concept of making ends meet. Respondents were asked the following question: “Thinking of your household’s total monthly income: Is your household able to make ends meet?” Answer categories ranged from 1 (with great difficulty) to 6 (very easily). The scale was reversed so that higher scores implied more financial hardship. Unemployment was measured with a binary variable for the respondent and the partner separately: not unemployed (0) versus unemployed (1). Partner’s unemployment was included for its importance for couple’s well-being (Blom, Kraaykamp, & Verbakel, 2017).

Expected economic hardship was indicated by the expectation that the future household situation was likely to be worse and by expecting job loss. Expected financial situation was measured with the question: “When it comes to the financial situation of your household, what are your expectations for the 12 months to come, will the next 12 months be (a) better, (b) worse, or (c) the same?” The same was used as the reference category. For employed people expecting job loss was measured with the question: “How likely or unlikely do you think it is that you might lose your job in the next 6 months” The answer categories ranged from very unlikely to very likely. Because of nonlinearity this indicator was included categorically, and for reasons of parsimony the categories were reduced to three groups: “job loss is unlikely,” “job loss is neither likely nor unlikely,” and “job loss is likely.”

The presence of children was measured as the number of children younger than 25 years living in the household. The age of 25 years was chosen as upper limit since parental financial responsibilities often continue beyond children’s age of 18 years. The variable consisted of three categories; “no children younger than 25 years” (0), “one or two children younger than 25 years” (1), and “three or more children younger than 25 years” (2). The macro-economic circumstances in a country were measured as the change in unemployment as percentage of the economically active population (aged 20 to 64 years) between 2008 and 2011, derived from Eurostat. Individual-level control variables included were gender, age, educational attainment (The International Standard Classification of Education classification, in three categories), and support. Support was measured as the mean of whether people could depend on help from (0) nobody versus (1) institutions, family members, relatives, friends, neighbors, or someone else in the following situations: help around the house when ill, advice about a serious personal or family matter, help with looking for a job, wanting someone to talk to, and urgently raise money to face an emergency. All these control variables could potentially influence the level of economic hardship people experience and their level of family life satisfaction. The country-level controls that were included were GDP per capita in current market prices in Euros divided by 1000, and social protection expenditure as percentage of the GDP¹, which were derived from Eurostat. These factors could influence family life satisfaction and economic hardship, and are related to the level of contextual unemployment (e.g., GDP affects family life satisfaction, Greenstein, 2009).

Table 1 shows descriptive statistics for all variables for the full sample and for the subsample of employed-only. The full sample is used when analyzing the influence of current financial hardship, current unemployment, and expected financial situation. The sample of only employed respondents is used when analyzing the influence of expected job loss, since one’s expectations about potential job loss are inherently only asked among those currently having a job. All continuous variables were grand-mean centered. Table A1 in the appendix shows the mean satisfaction with family life, the macro-level indicators (before grand-mean centering), and the number of respondents per country.

Table 1.

Descriptive Statistics for the Full and the Employed Sample, Before Grand-Mean Centering.

	Minimum	Maximum	Amount imputed cases	Full sample (N = 13,033)		Only employed sample (n = 10,177)		Countries (N = 30)
				M or %	SD	M or %	SD	M or %	SD
Satisfaction with family life	1	10	0	8.47	1.66	8.54	1.59
Current economic hardship
Financial hardship	1	6	116	3.35	1.25	3.18	1.18
Unemployed (not = ref.)	0	1	0	8.50
Partner unemployed (not = ref.)	0	1	0	6.38		4.70
Expected economic hardship
Future financial situation
Same	0	1	938	50.96		53.06
Better	0	1	938	19.48		18.58
Worse	0	1	938	29.55		28.36
Expected job loss
Unlikely	0	1	494			69.29
Neither likely nor unlikely	0	1	494			16.11
Likely	0	1	494			14.60
Male (female = ref.)	0	1	0	43.16		50.02
Age	18	65	0	42.26	10.03	42.58	9.85
Education
Less than upper secondary	0	1	97	24.81		18.48
Upper secondary	0	1	97	45.21		46.70
Tertiary	0	1	97	29.97		34.82
Children younger than 25 years
None	0	1	0	12.31		13.13
One or two	0	1	0	53.34		53.27
Three or more	0	1	0	11.29		9.72
Support	0	1	118	0.95	0.13	0.96	0.12
Country characteristics
Change in unemployment %	−1.7	10.3	0					3.64	3.64
GDP/1000	7.75	113.24	0					34.02	22.39
Social protection expenditure	12.9	31.2	0					22.73	5.38

Source. EQLS (2012), Eurostat.

Analytical Strategy

The associations between the four indicators of economic hardship and family life satisfaction were considered bivariately first (see Figure 1). Next, multivariate analyses were conducted using multilevel analyses to take into account that individuals were nested in countries (intraclass correlation is .033). The explained variance is calculated based on the Snijders and Boskes’ method (R² S&B; LaHuis, Hartman, Hakoyama, & Clark, 2014). The main effects of economic hardship and the interaction between current and future economic hardship were shown in Table 2. Model 1 showed the main associations between satisfaction with family life and financial hardship, unemployment, and expectations about the financial situation and includes all control variables. The associations between expected job loss and satisfaction with family life are presented in Model 2 as these were based on the subsample of employed people, since inherently only employed people may be expected to lose their jobs. Models 3 and 4 showed the interaction between current and expected economic hardship. Table 3 shows the interactions between economic hardship, the presence of children and change in macro-level unemployment. These gender interactions were included one-by-one in Models 5 to 14, and were controlled for all other variables. Differences between men and women are indicated by bold coefficients in Tables 2 and 3. Difference were tested by simultaneously interacting all indicators with gender in additional analyses. All additional analyses are available on request from the corresponding author.

Figure 1.

The association between economic hardship and satisfaction with family life, bivariate results.

Table 2.

The Influence of Economic Hardship on Satisfaction With Family Life.

	Model 1		Model 2		Model 3		Model 4
	Full sample		Only employed		Full sample		Only employed
	B	SE	B	SE	B	SE	B	SE
Intercept	8.421***	0.067	8.554***	0.070	8.422***	0.067	8.547***	0.070
Current economic hardship
Financial hardship	−0.175***	0.014	−0.162***	0.015	−0.172***	0.020	−0.181***	0.018
Unemployed (not = ref.)	−0.168**	0.053			−0.165**	0.053
Partner unemployed (not = ref.)	−0.149 *	0.060	−0.115	0.074	−0.146 *	0.060	−0.119	0.074
Expected economic hardship
Expected financial situation (same = ref)
Better	0.147***	0.041	0.133**	0.044	−0.202***	0.038	0.132**	0.044
Worse	−0.223***	0.038	−0.154***	0.039	0.149***	0.042	−0.156***	0.039
Expected job loss (unlikely = ref)
Neither likely nor unlikely			−0.264***	0.045			−0.286***	0.046
Likely			−0.145**	0.051			−0.138**	0.052
Male (female = ref)	0.111***	0.029	0.098**	0.031	0.114***	0.029	0.099**	0.031
Age	−0.002	0.002	−0.002	0.002	−0.002	0.002	−0.002	0.002
Education (less than upper secondary = ref.)
Upper secondary	0.039	0.039	−0.010	0.045	0.037	0.039	−0.010	0.045
Tertiary	0.086*	0.043	0.047	0.048	0.084*	0.043	0.047	0.048
Children younger than 25 years (none = ref.)
One or two	0.084**	0.031	0.050	0.034	0.084**	0.031	0.052	0.034
Three or more	0.086	0.050	0.111	0.057	0.086	0.050	0.112*	0.057
Support	0.730 * **	0.115	0.457***	0.133	0.719 * **	0.115	0.467***	0.133
Macro-level indicators
Change in Unemployment %	0.030	0.018	0.020	0.017	0.030	0.018	0.020	0.017
GDP	0.002	0.003	0.001	0.003	0.002	0.003	0.001	0.003
Social protection expenditure	−0.002	0.013	−0.007	0.012	−0.002	0.013	−0.007	0.012
Interaction coefficients
Financial hardship × Better financial future					−0.046	0.032
Financial hardship × Worse financial future					0.052	0.033
Financial hardship × Expected job loss, neither							0.113**	0.042
Financial hardship × Expected job loss, likely							0.018	0.042
Variance
Residual	2.579	0.032	2.366	0.033	2.577	0.032	2.364	0.033
Country	0.079	0.022	0.068	0.020	0.079	0.022	0.068	0.020
R2 (S&B)	0.039		0.032		0.039		0.033
Log likelihood	49881.336		37713.823		49399.881		37705.138

Note. SE = standard error. N = 13,013 for the full sample, n = 10,177 for the only employed sample, across 30 countries. R² (S&B) is the explained variance as calculated by Snijder’s and Boskes’ method (see LaHuis et al., 2014). Values given in bold are the significant (p < .05) differences between men and women. Results based on multilevel models.

Source. EQLS (2012).

^*p < .05. **p < .01. ***p < .001.

Table 3.

Current and Expected Economic Hardship and Satisfaction With Family Life, and Their Dependency on the Presence of Children and Change in Macro-Level Unemployment, Only Relevant Coefficients Are Shown.

Financial hardship (full sample)	Model 5		Model 10
	B	SE	B	SE
Financial hardship	−0.179***	0.021	−0.178***	0.022
Children younger than 25 years (none = ref.)
One or two children	0.084**	0.032
Three or more children	0.097	0.050
Financial hardship × One or two children	0.019	0.025
Financial hardship × Three or more children	−0.045	0.037
% Change unemployment			0.028	0.017
Financial hardship × % Change unemployment			0.002	0.007
Respondent’s unemployment (full sample)	Model 6		Model 11
	B	SE	B	SE
Unemployed (not = ref.)	−0.254**	0.083	−0.185**	0.060
Children younger than 25 years (none = ref.)
One or two children	0.074*	0.033
Three or more children	0.060	0.052
Unemployment × One or two children	0.104	0.108
Unemployment × Three or more children	0.295	0.170
% Change unemployment			0.027	0.018
Unemployment × % Change unemployment			0.018	0.017
Partner unemployment (full sample)	Model 7		Model 12
	B	SE	B	SE
Partner unemployed (not = ref.)	−0.124	0.094	−0.162*	0.084
Children younger than 25 years (none = ref.)
One or two children	0.081*	0.032
Three or more children	0.113*	0.051
Partner unemployment × One or two children	0.046	0.123
Partner unemployment × Three or more children	−0.424 *	0.195
% Change unemployment			0.026	0.018
Partner unemployment × % Change unemployment			0.018	0.024
Future financial situation (full sample)	Model 8		Model 13
	B	SE	B	SE
Expected financial situation (same = ref)
Better financial future	0.131	0.068	0.157**	0.051
Worse financial future	−0.209***	0.058	−0.201***	0.049
Children younger than 25 years (none = ref.)
One or two children	0.064	0.044
Three or more children	0.202**	0.073
Better financial future × One or two children	0.058	0.086
Better financial future × Three or more children	−0.141	0.132
Worse financial future × One or two children	0.030	0.072
Worse financial future × Three or more children	−0.278 *	0.117
% Change unemployment			0.022	0.019
Better financial future × % Change unemployment			0.002	0.015
Worse financial future × % Change unemployment			0.019	0.016
Expected job loss (only employed sample)	Model 9		Model 14
	B	SE	B	SE
Expected job loss (unlikely = ref)
Job loss, neither likely nor unlikely	−0.451***	0.074	−0.258***	0.045
Job loss, likely	−0.146	0.078	−0.144**	0.052
Children younger than 25 years (none = ref.)
One or two children	−0.007	0.040
Three or more children	0.099	0.068
Job loss, neither × One or two children	0.325***	0.094
Job loss, neither × Three or more children	0.154	0.168
Job loss, likely × One or two children	0.021	0.096
Job loss, likely × Three or more children	−0.092	0.157
% Change unemployment			0.013	0.017
Job loss, neither × % Change unemployment			0.019	0.014
Job loss, likely × % Change unemployment			0.018	0.015

Note. N = 13,013 for the full sample, n = 10,177 for the only employed sample, across 30 countries. The values given in bold are the significant (p < .05) differences between men and women. Results based on multilevel models; controlled for all other variables included in Table 2.

Source. EQLS (2012).

^*p < .05. **p < .01. ***p < .001.

Results

Bivariate Analyses

A first step was to test whether people who experienced or expected economic hardship were less satisfied with their family life. Figure 1 shows the association between current financial hardship and satisfaction with family life and indicated that people indeed were less satisfied with their family life when they experienced more financial hardship. In addition, people who themselves were unemployed or whose partner was unemployed indicated to be less satisfied with their family life. Figure 1 also showed that people, who expected their future financial situation to worsen, reported a lower satisfaction than people who expected financial stability. People who expected to be better off financially displayed most satisfaction with their family life. Last, people who found it likely that they would lose their job showed less satisfaction with their family life than people who expected job stability, but people who found it likely to lose their job were more satisfied than people who found it neither likely nor unlikely. So, uncertainty seems most harmful for a person’s family life satisfaction.

Exploration of the associations between the measurements of economic hardship (not shown) supported, first, the general idea that employment and financial situations go hand in hand: current unemployment (of either respondent or partner) was associated with more current financial hardship and expected job loss was associated with worse expected financial situation. Second, it revealed that those currently in disadvantaged positions (either in unemployment or financial hardship) expected either better or worse financial situations in future compared with those currently in advantaged positions, who in turn were more likely to expect no change of their situation.

Satisfaction With Family Life and Economic Hardship

Table 2 reports the associations between economic hardship and family life satisfaction from the multilevel analyses holding all other aspects constant. Model 1² shows that people who experienced more financial hardship were less satisfied with their family life (b = −0.175). Also, people expressed lower satisfaction when they or their partner were unemployed (b = −0.168 and b = −0.149, respectively). The negative association between partner’s unemployment and satisfaction with family life was only relevant for women, indicating women’s lower satisfaction with their family life when their partner was unemployed. All in all, findings supported hypothesis 1 on the detrimental effect of current economic hardship for family life satisfaction.

Model 1 also showed that people who expected their financial situation to worsen were less satisfied with their family life (b = −0.223), while people who expected to be better off in the near future showed more satisfaction (b = 0.147). Model 2³ (based on employed people only) adds to this that family life satisfaction was higher among people who found it unlikely to lose their job as compared with people who found it likely or who found it neither likely nor unlikely (b = −0.145 and b = −0.264, respectively). These findings supported Hypothesis 2. Analyses gave no indication of gender differences in these effects. All other economic hardship indicators in Model 2 showed similar results for people who were employed compared with the full sample (in Model 1) with the exception that partner unemployment was no longer negatively related to satisfaction with family life. Additional analyses showed that of all indicators of economic hardship, only the influence of current financial hardship varied over countries.

Regarding the control variables in the models, men showed to be more satisfied with their family life than women. Age did not affect family life satisfaction. The higher educated reported more satisfaction with their family life than the lower educated. People with one or two children younger than 25 years in the household were more satisfied with their family life than people without children in the full sample, but not in the sample of only employed respondents. People who expected to receive more support from the social network were more satisfied with their family life (women more than men). The country-level indicators did not influence satisfaction with family life, with the notable exception of social protection expenditure which was negatively related to satisfaction with family life for men.

Models 3 and 4 in Table 2 showed whether expecting financial hardship (Model 3) or expecting job loss (Model 4) was especially harmful for people’s satisfaction with family life when they currently experienced financial hardship, as expected in Hypothesis 3. Model 3 showed that expecting to be financially worse off in the next year did not amplify the negative influence of current financial hardship. The interaction results of Model 4 were presented in Figure 2 to facilitate interpretation. These results indicated that among those with most current financial hardship, family life satisfaction was not associated with expected job loss, whereas we expected that in such situation, differences in family life satisfaction would be especially large between those who find it likely versus unlikely to lose their job in the next year. In sum, we reject Hypothesis 3; expecting economic hardship was not more detrimental for people’s satisfaction with family life when they currently experienced more economic hardship.

Figure 2.

Moderation between financial deprivation, expected job loss, and family life satisfaction.

Conditional Influences

Table 3 shows the models with interactions between the presence of children, change in country’s level of unemployment, and economic hardship; the variances for these models are shown in Table A2 in the appendix. In line with Hypothesis 4, this study found some support that economic hardship reduced family life satisfaction especially if three or more children lived in the household. More precisely, having an unemployed partner reduced satisfaction with family life for women with three of more children, but not for childless women (Model 7), and expecting a financial decline was more detrimental for family life satisfaction for women with three or more children compared with childless women (Model 8). However, contrary to the hypothesis, current financial hardship (Model 5), unemployment (Model 6), and expected job loss (Model 9) were not more detrimental for people with children than for childless people. All in all, Hypothesis 4 cannot be confirmed unequivocally.

Last, results concerning the conditional influence of change in the percentage of unemployed in a country (Hypothesis 5) were reported in Models 10 to 14 in Table 3.⁴ No indication was found that economic hardship was more detrimental for people’s satisfaction with family life in countries that had experienced a larger rise in unemployment, for all interaction coefficients were nonsignificant.

Robustness Analyses

Several checks for robustness were conducted. First and foremost, multilevel negative binomial analyses designed for count-variables were conducted because this type of analysis is designed to deal with highly skewed dependent variables (the scale of family life satisfaction was reversed for these analyses). Differences were limited. The analyses only indicated that the main effect of partner unemployment and its interaction with the presence of children were not related to satisfaction with family life. Other results were robust.

Second, we restricted the age selection from 25 to 55 years to deal with possible selectivity in the youngest and oldest age groups; generally lower educated are employed at a younger age, and are less likely to retire early. These analyses did not lead to different conclusions, but in these models the interaction between financial decline and presence of children was no longer significant.

Third, in our current measurement of individual and partner unemployment, homemakers and employed people were combined in the reference category for reasons of parsimony. Differentiating employed people from homemakers provided the same conclusions.

Fourth, the reference category for the presence of children included both people without children and people with children who had left the household. Possibly, the absence of moderation by presence of children was caused by this ambiguity. Therefore, we distinguished two groups, people without children in the household (a) younger than age 40 and (b) older than age 40. This led to similar conclusions, but expected financial decline and presence of children was no longer significant.

Fifth, it was checked whether the measurement of the macro-economic circumstances influenced the associations between economic difficulties and satisfaction with family life. Alternatives that were tested included the moderating influence of (a) relative change in unemployment levels between 2008 and 2011, (b) the unemployment percentage in a country in 2011, (c) GDP per capita, (d) the average GDP growth between 2008 and 2011, and (e) the GDP growth in 2011. All models showed that the macro-economic circumstances did not (consistently) influence the effect between economic hardship and family life satisfaction.

Overall, our robustness checks suggested that the main findings were generally robust. The direct associations between financial and employment difficulties and satisfaction with family life were robust for all measurements and selections that were made, with two exceptions. First, unemployment of the partner was not negatively related to satisfaction in our negatively binomial analyses. Second, interactions between unemployment of the partner and future financial situation with the presence of children were not very robust. So, these results should be interpreted with care.

Discussion and Conclusion

This study investigates how satisfaction with family life is influenced by various forms of economic hardship. More specifically, it studied whether people were less satisfied with their family life if they experience more economic hardship or expect more economic hardship in the near future. Additionally, we investigated if expecting economic hardship was especially harmful for people’s satisfaction when they currently experience more economic hardship.

Our main finding is that people who currently experience and perceive economic hardship were less satisfied with their family life. This conclusion is in line with previous studies (e.g., Shim et al., 2017; Williams et al., 2015). While previous studies mostly focused on single countries, our research indicated that this negative association is generally applicable to a wide range of European countries; overall people who experienced financial hardship or were unemployed showed to be less satisfied with their family life. Our study also included individual and partner’s unemployment, since the employment of both partners is of importance for couples’ well-being (Blom et al., 2017). The unemployment of individuals and their partners demonstrated to have an additional negative effect on family satisfaction even when financial hardship was taken into account. This may be explained by a potential reduction of people’s social contacts, sense of purpose, and status when becoming unemployed (Paul & Moser, 2009).

This study is among the first to study both current and expected economic hardship and their consequences for people’s satisfaction with their family life. We showed that—aside from current economic hardship—people who expect a financial decline or job loss were less satisfied with their family life. Therefore, this research provides clues that people’s expectations about their future financial and employment situations are meaningful in understanding evaluations of family life. Insecurities about future finances or potential job loss, however, did not seem to amplify the negative consequences for families in current economic hardship. In sum, we conclude that people’s expectations of their financial and employment future shape people’s family life satisfaction over and above their current experiences.

It was further investigated whether economic difficulties were more harmful for family life when children live in the household or if national unemployment rates had risen sharply during the recent European economic crises. Remarkably, this study concluded that having children or living in a country with high-unemployment risks did not condition the relationship between economic hardship and family life satisfaction. It must be noted that for having children some results supported the idea that hardship had more detrimental consequences for satisfaction in large families. Specifically, the results indicated that having more children amplified the negative influence of expecting a financial decline for one’s family life satisfaction. Similarly, the partner’s unemployment was more detrimental for women’s family life satisfaction when they had more children. These results, however, were not very consistent or robust. Possibly, we did not find economic difficulties affecting parents’ and nonparents’ satisfaction differently because conflicting mechanisms counterbalanced each other. On the one hand, parents have more financial responsibilities, which may strengthen the importance of their economic situation when assessing their family life satisfaction. On the other hand, a household’s economic situation may be less important for parents’ satisfaction with family life, because their satisfaction with family life is primarily constructed around their parental status, in contrast to people without children. This may serve as a protective factor against the influence of economic problems on their family life. In sum, our study indicated that economic difficulties had a relatively universal effect on people’s satisfaction with family life, for associations were largely independent of family composition, macro-economic circumstances, and gender.

This study is, to our knowledge, the first study to investigate the consequences of economic hardship on people’s satisfaction with family life across a wide range of countries. Additionally, our findings indicated that financial hardship affected people’s family life differently across European countries; in some countries financial problems affected the family more than in others. A country’s economic situation, however, did not appear to explain these cross-national differences. We encourage other researchers to investigate how other contextual circumstances affect how a person’s economic situation influence people’s family life, such as a family culture and the welfare state (Boss et al., 2017). Although this study benefitted from using cross-national data, using such data comes with some drawbacks. A limitation was the measurement of satisfaction with family life which proved highly skewed and was measured with a single item. Future research therefore could benefit from using more comprehensive scales, such as the Satisfaction With Family Life scale (Zabriskie & Ward, 2013). Furthermore, macro-economic conditions were measured on the country-level, which can be regarded as a limitation. Possibly regional economic conditions, job sector-specific conditions, or the economic conditions of the extended social network would more strongly reflect people’s perception of their economic circumstances, which could influence how people perceive their personal economic hardship and its consequences.

Aside from the factors studied in this article, other precarious labor market conditions may be important for the quality of family relationships. Future researcher could investigate the influence of other precarious labor market conditions such as underemployment and flexible contracts. In addition, we encourage future studies to investigate whether resources, for instance (social) support, may diminish the negative consequences of economic difficulties for families. Negative economic events potentially do not strike every family to the same degree. Some families are better able to handle negative economic events than others due to their resources, such as their social support system. The resources of the family may mitigate potential negative effects of economic problems for family life, as for instance theorized in the ABC-X model (Boss et al., 2017).

Finally, we encourage future research to investigate into specific mechanisms for the relationship between expected economic circumstances and satisfaction with family life, as theorized by the family stress model (Conger et al., 2010). Other studies have investigated the proposed mechanisms with respect to current economiccircumstances and found confirmation for them (e.g., Conger et al., 2002; Mistry et al., 2008; Williams et al., 2015). However, the specific mechanisms how expected economic problems affect family relationships are relatively unknown and we were unable to test this specific mechanism which we encourage future research to do. Relatedly, in this study, we were unable to explore reverse causation since we used cross-sectional data. Although previous longitudinal studies had similar conclusions regarding current economic hardship and family functioning (e.g., Neppl et al., 2016; Shim et al., 2017), further longitudinal examination is needed on expected economic circumstances and family functioning.

This study started by noting that people in Europe experienced more economic hardship as a result of recent economic crises in Europe (Eurofound, 2015). Our research showed that negative consequences of economic difficulties spill over to people’s family life. In addition, not only people who actually experience a decline of income or job loss seem affected by the economic crisis; also people who expect to face income or job loss in the near future feel less satisfied with their family life. Therefore, it can be concluded that the negative consequences of economic crises for families are far more widespread than merely the individuals who lose income or employment.

Appendix

Table A1.

Descriptive Statistics of Satisfaction With Family Life and Country Characteristics Per Country.

	Mean satisfaction with family life	Change unemployment percentage of 2008-2011	Social protection expenditure	GDP per capita	Total sample, N	Employed sample, n
Austria	8.57	0.5	28.2	51.12	384	350
Belgium	8.23	0.1	28.4	47.70	313	267
Bulgaria	8.05	5.6	16.1	7.75	278	232
Croatia	8.75	5.1	20.0	14.54	302	220
Cyprus	9.18	4.1	20.5	31.84	432	270
Czech Republic	8.08	2.2	19.5	21.66	408	367
Denmark	9.00	4.0	31.2	61.30	361	333
Estonia	8.35	7.0	15.5	17.45	237	201
Finland	8.80	1.5	28.0	50.79	325	309
France	8.49	1.8	30.7	43.81	792	686
Germany	8.47	−1.7	27.3	45.94	890	704
Greece	8.33	10.1	29.1	25.87	320	219
Hungary	8.50	3.3	21.5	14.03	274	238
Iceland	9.02	4.2	23.7	45.97	481	450
Ireland	8.73	8.3	21.9	52.83	398	271
Italy	8.12	1.7	27.3	38.36	841	613
Latvia	8.26	8.5	14.9	13.78	261	203
Lithuania	8.61	9.7	16.2	14.37	292	236
Luxembourg	8.68	−0.1	21.9	113.24	405	337
Malta	8.64	0.7	18.0	22.35	373	256
Netherlands	8.10	2.0	28.2	53.54	377	326
Poland	7.98	2.5	18.1	13.89	686	532
Portugal	8.42	5.1	24.2	23.19	298	226
Romania	8.95	1.6	16.2	9.20	379	290
Slovakia	8.31	4.0	17.4	18.14	324	295
Slovenia	8.54	3.9	24.1	24.98	289	250
Spain	8.49	10.3	24.9	31.83	538	337
Sweden	8.85	1.8	27.7	59.59	308	291
Turkey	8.13	−0.8	12.9	10.58	781	293
The United Kingdom	8.46	2.2	28.3	41.02	686	575

Note. Mean satisfaction with family is mean satisfaction in the total sample.

Source. EQLS (2012), Eurostat.

Table A2.

Variances for the Interaction Models Between Economic Hardship and the Presence of Children in the Household and the Change in Unemployment in Table 3.

	Model 5		Model 6		Model 7		Model 8		Model 9
	Full sample		Full sample		Full sample		Full sample		Employed sample
	B	SE	B	SE	B	SE	B	SE	B	SE
Variance
Residual	2.578	0.032	2.578	0.032	2.577	0.032	2.577	0.032	2.362	0.033
Country	0.079	0.022	0.079	0.022	0.079	0.022	0.079	0.022	0.069	0.020
R2 (S&B)	.039		.039		.039		.039		.033
Log likelihood	49405.634		49405.809		49402.696		49399.873		37699.741
	Model 10		Model 11		Model 12		Model 13		Model 14
	Full sample		Full sample		Full sample		Full sample		Employed sample
	B	SE	B	SE	B	SE	B	SE	B	SE
Variance
Residual	2.567	0.032	2.577	0.032	2.574	0.032	2.570	0.032	2.365	0.033
Financial hardship	0.008	0.003
Unemployment			0.016	0.021
Partner unemployment					0.083	0.056
Better financial future							0.019	0.017
Worse financial future							0.025	0.019
Country	0.072	0.021	0.077	0.022	0.077	0.022	0.077	0.025	0.069	0.020
R2 (S&B)	.045		.040		.041		.042		.032
Log likelihood	49377.128		49405.139		49400.518		49393.427		37710.911

Note. N = 13,013 for the full sample, n = 10,177 for the only employed sample, across 30 countries. R² (S&B) is the explained variance as calculated by Snijder’s and Boskes’ method (see LaHuis et al., 2014). Results based on multilevel models.

Source. EQLS (2012).