Individual, family and neighbourhood factors related to life satisfaction and perceived discrimination among low-income, non-immigrant mothers in seven European countries

ABSTRACT

We examined associations between dimensions of social exclusion and maternal life satisfaction and maternal perceived discrimination for low-income, non-immigrant mothers of children at preschool and primary school age in seven European countries (Czech Republic, England, Germany, Greece, Italy, Poland, Portugal – N = 1227), using standardized survey data from a large-scale interview study. Life satisfaction and perceived discrimination were found to be associated, indicating that both dimensions are important to consider in relation to subjective well-being. The results of linear mixed effect regression models demonstrated that lack of resources in multiple dimensions of social exclusion were linked to our well-being measures, including objective life condition variables (material deprivation, poor neighbourhood quality; both outcomes) as well as more relational aspects (unemployment, less social support; life satisfaction) and adult literacy related difficulties (perceived discrimination). These findings reaffirm the importance of combatting social exclusion. There are implications for public policy, emphasizing the importance of joined-up policies that tackle different forms of exclusion.

KEYWORDS:

• social exclusion
• life satisfaction
• perceived discrimination
• family disadvantage
• international study

The primary interest of the present study was to explore associations between indicators of social exclusion and life satisfaction and perceived discrimination in disadvantaged families in Europe. Throughout Europe the fight against poverty and social exclusion has repeatedly been high on the political agenda (European Commission, 2022; European Parliament, 2022). There are estimated to be 95 million people in the EU at risk of poverty and social exclusion, including one-fifth of the EU population in 2021 living in households with dependent children (Eurostat, 2022). Social exclusion and poverty can have significant negative impacts on family life. Economic hardship can create stress for parents, families and communities, and contribute to difficult family relationships, (e.g. Barnett, 2008; Gershoff et al., 2007; Mayer, 1997) and lower parental well-being (e.g. Gutman et al., 2005; Vogt Yuan, 2008). Children growing up in poverty and social exclusion are more likely to show disadvantages in cognitive, language and socio-emotional development early in life (Feinstein, 2003; Magnuson et al., 2004; Melhuish et al., 2017). Importantly, parents’ well-being may contribute substantially to the association between economic hardship and children's well-being and development (e.g. Gutman et al., 2005; Vogt Yuan, 2008). Parental stress may be caused, in part, by feeling unable to cope with daily demands – demands that can be related to finances, relationships, work or other aspects of daily life. This struggle to cope relates to a lack of material, personal or social resources that can be mobilized when needed (Lazarus & Folkman, 1984).

Background

Social exclusion

Social exclusion and poverty are closely linked; both having negative impact for families by limiting access to resources, including basic services like healthcare, education, and housing. Importantly, the conceptualization of social exclusion captures aspects of people's participation and processes that deprive people of the resources required for participation in basic political, economic and social activities, which are usually available to the majority of society (Levitas et al., 2007; Pierson, 2009). The concept of social exclusion thus focuses on systematic processes that allow some, but deny others, to accumulate advantage (Dean, 2016; Room, 1995). Seen in this way, poverty is conceptualized in relative terms, with reference to the distribution of resources in society (Townsend, 1979).

Social exclusion has been described as a multi-dimensional concept that includes economic, social and political aspects of life. While there is not one clear definition of social exclusion, it is commonly underpinned by poverty or deprivation. Most quantitative measures of social exclusion examine material and social deprivation to index an individual's level of social exclusion. Measures of material deprivation collect multidimensional information on people's lack of material resources that influence their lifestyle and living conditions, and restrict community participation (Townsend, 1979). Thus, measures of material deprivation index the degree to which needs are met, including resources such as housing, possession of durables, and capacity to afford basic requirements (see for example, EU-SILC, 2006). Other dimensions of deprivation that are drivers of social and economic exclusion are lack of access to the labour market, reduced opportunity due to low education or lack of skills, experiences of crime, lack of services, or limited social networks (Levitas et al., 2007; Myck et al., 2019).

Cumulatively, these factors form barriers to social inclusion, creating a sense of isolation, and these factors are mutually reinforcing. Thus, social exclusion can be seen as a dynamic process, where one deprivation factor leads to another, with many factors interconnecting and mutually reinforcing, and jointly contributing to people's experiences of exclusion (Paugam, 1996). On the other hand, resources in one area can off-set risk from lack of resources in other areas. Supportive social networks, in particular, have the potential to compensate for the negative effects of economic risks (Putnam, 2007; Van Tuijl & Leseman, 2013). Research on parenting stress has shown that any type of social support can contribute to reducing maternal stress (Armstrong et al., 2005; Crnic & Greenberg, 1990; Parkes et al., 2015). When investigating parent and child outcomes in the context of deprivation, it is therefore important to capture social integration or social support and to determine their contribution to wellbeing (Ackerman et al., 2004).

Life satisfaction

In addition to objective indicators of well-being, subjective perceptions also need to be considered in the context of inclusion and exclusion processes (Bellani & D’Ambrosio, 2011; Eurostat, 2017). Deprivation has been found to be associated with lower levels of subjective wellbeing, as people living in deprived areas or experiencing deprivation may experience greater stress, anxiety, and negative emotions, which can have a negative impact on their overall sense of happiness and life satisfaction. Subjective well-being can be measured in terms of life satisfaction, which captures how people evaluate their life as a whole, including their satisfaction with different aspects of their life, such as work, relationships and health (Diener et al., 1985; Diener et al., 2002). Quantitative measures of wellbeing commonly involve individuals rating their level of life satisfaction on a Likert-type scale (Pavot & Diener, 2008). This can be used to compare levels of wellbeing across different populations over time. Life satisfaction is considered a useful measure of wellbeing because it reflects an individual's overall evaluation of their life, rather than just one particular aspect of their life such as income or health. It is also a measure that has been used across countries and can be easily understood and interpreted by individuals from diverse backgrounds (Diener & Diener, 1995).

Studies have found that individuals who experience material or social deprivation are likely to report lower levels of life satisfaction (Diener et al., 2015). Life satisfaction has been found to be moderately stable over time. Nevertheless, life events can have a strong influence on long-term levels of subjective well-being. Unemployment has been found to have a substantial detrimental effect going beyond loss of income (Lucas et al., 2004; Winkelmann & Winkelmann, 1998). For example, non-monetary factors such as personal values and life goals, shaped by cultural factors need to be considered (Schwartz & Sortheix, 2018). Additionally, factors such as health, feelings of self-worth, self-efficacy and self-control have been identified as important contributors to subjective well-being (Erdogan et al., 2012), as well as the quality of participation in social relationships (Mellor et al., 2008; Winkelmann, 2006) and local neighbourhood characteristics (Gao et al., 2017; Lee & Yoo, 2015).

Finally, educational attainment and basic skills competencies are central to understanding social exclusion. While having fewer educational qualifications can be related to higher levels of distress and anxiety, and lower physical well-being, it is less clear what associations exist between education and subjective well-being (Oishi et al., 2007; Ross & Van Willigen, 1997). Nevertheless, such factors need consideration because lower educational qualifications and poor literacy are associated with unemployment and reliance on benefits, and lower quality living conditions (Bynner & Parsons, 2006; DfES, 2003).

Importantly, contextual factors including deprivation can have significant impact on parental life satisfaction, which can influence parenting practices and child outcomes. Understanding how aspects of deprivation relate to parents’ life satisfaction can provide guidance for policies designed to improve parent wellbeing, which can have long-lasting negative consequences for children, perpetuating existing inequalities in society and contributing to an intergenerational cycle of deprivation.

Perceived discrimination

The impact of deprivation on subjective well-being can be compounded by other factors such as discrimination and social inequality. When individuals who already experience material and social deprivation also perceive discrimination, this can further limit their ability to access social and economic resources and can create a cycle of social exclusion and isolation contributing to long-term negative effects on well-being and perpetuating inequalities within society (Social Exclusion Unit, 2004).

Perceived discrimination refers to an individual's subjective perception of unfair treatment based on their characteristics such as race, gender or religion. Discrimination is generally based on perceptions of the ‘otherness’ of particular groups with values or characteristics that are perceived as different, alien and inferior to one's own (Harma et al., 2013; Wolfensberger & Thomas, 1999). Such perceptions shape attitudes and behaviour towards people and can lead to segregation and marginalization. Perceived discrimination can reinforce disadvantage, by affecting the social status, mental and physical health, self-perception, self-esteem and self-confidence (Kessler et al., 1999; Krieger, 2014). When individuals perceive discrimination directed towards them, they may feel excluded from certain networks or communities, and may have a reduced sense of belonging and shared values (Schmitt et al., 2017). Importantly, perceived discrimination may lead to barriers to seeking and obtaining needed services, access to care, and greater poverty and social marginalization (Martins, 2008; Thornicroft et al., 2007). In many countries, discrimination in service delivery experienced by socially excluded or disadvantaged families has been linked to adverse outcomes (Van Boekel et al., 2013).

Ethnicity, gender, identity or disability are commonly investigated as drivers of discrimination (Bird & Bogart, 2001; Rivenbark & Ichou, 2020), but poverty may also be a cause of discrimination (Bradshaw et al., 2015; Skosireva et al., 2014). The poor often have reduced access to credit or pay higher prices for services, and they commonly experience social exclusion and marginalization (Killeen, 2008). They feel that others’ perceptions of their poverty lead them to be treated differently, which can influence access to services and their feelings of support (Blewett et al., 2008). It is reported that professionals and agencies are likely to treat them differently, showing a lack of understanding of the lives of poor people (Davies, 2008).

Aims and research questions

Data from a large-scale interview study (Inclusive Education and Social Support to Tackle Inequalities in Society; ISOTIS) are used to explore associations between indicators of social exclusion and life satisfaction and perceived discrimination in disadvantaged families. The study was carried out across seven countries in Europe and purposefully sampled disadvantaged families with children at preschool and primary school age. Data included multiple indicators of social exclusion at individual, family and neighbourhood levels. This study focuses on disadvantaged families who had been in their respective societies for several generations, enabling exploration into drivers of discrimination that are associated with poverty-related aspects of deprivation, rather than ethnicity or immigrant status. This is important because poverty and deprivation are less commonly investigated as drivers of discrimination. Recruitment of survey participants focused on primary caregivers and resulted in a sample of 97.6% mothers. Small numbers did not allow an analysis specific to fathers, and analysis in this paper focuses on participating mothers only. Research questions are:

RQ1. What is the relationship between social exclusion indicators and maternal life satisfaction?

RQ2. What is the relationship between social exclusion indicators and maternal perceived discrimination?

In all cases other demographic and country factors are considered.

Data and methods

Sample

Data derive from a large-scale structured interview study with parents with a disadvantaged background in 10 European countries (Broekhuizen et al., 2018), and was designed within the framework of the EU funded project ISOTIS. Analysis focused on data collected from a sub-set of the total sample, namely non-immigrant mothers of young children at pre-school and primary school age. Non-immigrant status is defined as the respondent and both her parents being born in the country of residence. This included 1227 mothers living in seven different countries with two regions per country selected to reflect local contexts (e.g. political, economic or cultural differences). The countries included are Czech Republic (N = 197; Brno, Ústí nad Labem), England (N = 131; Greater London, North West England), Germany (N = 124; Berlin, Bremen), Greece (N = 89; West Athens Sector, West and East Atica Region), Italy (N = 223; Milan, Turin), Poland (N = 227; Warsaw, Łódź), Portugal (N = 231; Lisbon, Porto). Children were in two age groups with half in the younger age group (3- 6-year-olds, N = 610, M = 4.5, SD = 1.23), and half in the older age group (7–11-year-olds, N = 617, M = 9.72, SD = 1.32). Just over half (54%; N = 657) were girls and 46% (N = 566) were boys. Recruitment focused on deprived neighbourhoods, and interviews took place in schools or centres with a high proportion of families receiving benefits indicative of family disadvantage (e.g. access to free school meals). For more information, see Broekhuizen et al. (2018).

Interview questions focused on mothers’ experience related to their child's education and up-bringing. Many interview questions were specific to a particular child. If a parent had a child in both age groups, parents were asked to focus on the child attending the centre where the parent was recruited (e.g. a preschool). If this was ambiguous (e.g. community centre), the target child was chosen purposefully to achieve balanced numbers for each age group.

Data collection was through face-to-face structured interviews. Interview questions were programmed in an online survey software (Lime Survey) and parent responses were entered in real time wherever possible, and after an interview where necessary (sometimes parent responses were captured with pen and paper). Parents received an incentive for participating in the interview (voucher of €5-10 or participation in a raffle). Data collection for the interviews ran from December 2017 to July 2018.

Dependent variables

Maternal life satisfaction

Life satisfaction was measured by mothers rating the applicability of three statements (I am satisfied with my life; If I could live my life over, I would change almost nothing; In most ways, my life is close to my ideal) on a scale ranging from disagree [1] to agree [5]; Verkuyten, 2008). Responses to the three questions were combined into one measure of life satisfaction by creating the sum across the three items (α = .76).

Maternal perceived discrimination

Perceived discrimination was measured by asking mothers ‘How often do you feel discriminated or unfairly treated because of your background or situation by … ’ (5 items; people in their neighbourhood, parents in the (pre)school of their child, teachers in the (pre)school of their child, people working in healthcare, statements in social or public media) with responses ranging from never [1] to often [4]). Responses to the questions were combined into one measure of perceived discrimination by creating the sum across the five items (α = .68).

Independent variables

Social exclusion indicators

Material deprivation. A continuous measure of material deprivation was created, with a possible range from 0 to 13. To assess material deprivation, 13 questions considered whether a parent experienced difficulties affording a range of items or activities, e.g. car or internet connection, meals, keeping the home warm, leisure activities, getting together with friends, family holidays, replacing furniture or worn clothes and shoes. Answer options were ‘yes’ or ‘no’(Guio et al., 2016; α = .83).

Workless households. Participants were asked to report if they and their partner currently had a paid job (yes/no). A single dichotomous measure of employment deprivation was created, indicating if a household was characterized as workless/unemployed, or whether one or both parents were earners (single earner or dual earner family).

Lower maternal education level. Questions on maternal educational level were coded to indicate ISCED levels in three categories (low [ISCED 0-2], medium [ISCED 3-5], high [ISCED >5]; reverse-coded; ISCED 2011).

Perceived literacy difficulties. Assessed by asking three questions on whether the parent experienced difficulties when talking, reading, or listening to the national language (ranging from never [1] to always [5]), combined into one measure by creating the sum (3 items, α = .77).

Perceived neighbourhood disorder. Indicated by asking three questions on what happens in the neighbourhood (3 items; frequency of violence or crime against people, crime involving property, and general nuisance – e.g. litter, abandoned cars (ranging from never [1] to often [4], Barnes, 1997); combined into one measure by creating the mean (3 items, α = .75).

Fewer avalable neighbourhood play areas. Indicated by asking about the availability of safe places for children to play outdoors and nearby (Barnes, 1997). Coded to compare two or more play areas vs. one and no play areas.

Perceived lack of social support. Indicated by asking six questions related to the companionship, assistance or support available when needed (six items; someone to take you to the doctor, someone who shows you love and affection, someone to share your mist private worries and fears with, someone to turn to for suggestions about how to deal with a personal problem, someone to do something enjoyable with; all ranging from never [1] to often [4], reverse-coded. Selected from the MOS Social Support Survey: Sherbone & Stewart, 1991); into one measure by creating the sum (6 items, α = .83).

Country and demographic characteristics

Location of data collection (country), level of diversity in neighbourhood (portion of population with immigrant background in the neighbourhood – ranging from (almost) none [1] to (almost) all [5]; Laurence et al., 2018), child age group (3–6; 7–11 years), child gender (female/male), mother age, number of children in the household, reported religious affiliation (no/yes), perceived importance of religion (ranging from not important [1] to very important [5]).

Analysis strategy

Analyses focussed on understanding which social exclusion factors might influence mothers’ life satisfaction and perceived discrimination. Models were fitted for the two dependent variables taking into account demographic factors and all the independent variables. In a first step, regression models were fitted. Dependent variables were normalized to have a mean of 0 and a standard deviation of 1. Continuous co-variates were standardized to have a mean of 0 and standard deviation of 0.5, so that the corresponding model coefficients represent the change in the normalized outcome corresponding to a change of 2 standard deviations in the continuous covariate, and thus make comparison of effects associated with continuous and categorical variables more equivalent (Tymms, 2004).

A linear mixed-effects regression model of outcome Perceived discrimination was fitted in terms of all covariates, including Life satisfaction. A linear mixed-effects regression model of outcome Life satisfaction was fitted in terms of all covariates, including Perceived discrimination. Models were fitted to complete cases data. A pseudo-R-squared coefficient of determination was calculated using the method of Nakagawa et al. (2013).

Subsequently a structural equation model (SEM) was fitted to the data. The two outcome variables ‘maternal perceived discrimination’ and ‘maternal life satisfaction’ were regressed on all covariates. The categorical variables were coded as follows: country – all other countries vs. Portugal; neighbourhood immigrant diversity – around half or more vs. around a quarter or less; child age group – 7–11 vs. 3–5-year-olds; number of children – more than one vs. one; workless vs. working households; availability of neighbourhood play spaces – one of fewer spaces vs. two or more play spaces; sex – male vs. female; reported religious affiliation – yes vs. none/not reported; importance of religion – important or very important vs. moderately important or less; maternal education – low vs. medium or high.

For the SEM, it was desirable to have the covariate effect sizes comparable with the conditional correlation fitted between the two dependent variables. To achieve this, both dependent variables and covariates were standardized to have mean zero and standard deviation 1. To ensure comparability of effect sizes between the continuous covariates and the binary covariates, the latter were coded 0/2 rather than 0/1 (Gelman, 2008). Thus, all the coefficients in the SEM should be approximately comparable. However, the coefficients in the SEM will be approximately half those in the regression models.

Results

Descriptive statistics

The summary statistics for the country and demographic factors are given in Table 1.

Table 1. Summary statistics – country and demographic factors.

Variable	Level	%	N	Min	Max	Mean	SD	N missing
Country	Portugal	18.8	231
	Poland	18.5	227
	Italy	18.2	223
	Czech Republic	16.1	197
	England	10.7	131
	Germany	10.5	129
	Greece	7.3	89
	Missing	0.0	0
Level of diversity in neighbourhood	(Almost) none	42.3	519
	Around a quarter	26.2	322
	Around half	19.0	233
	Around three-quarters	8.1	99
	(Almost) all	3.1	38
	Missing	1.3	16
Child age group	3–5 year-olds	49.7	610
	7–11 year-olds	50.3	617
	Missing	0.0	0
Child Gender	Female	53.5	657
	Male	46.1	566
	Missing	0.3	4
Mother age			1208	19.00	63.00	36.24	6.85	19
Number of children in the household	1	30.9	379
	2	36.2	444
	3	19.7	242
	4 or more	11.7	143
	Missing	1.5	19
Reported religious affiliation	None/not reported	36.8	451
	Yes	63.2	776
	Missing	0.0	0
Perceived importance of religion	Not important	32.2	395
	Slightly important	15.4	189
	Moderately important	16.4	201
	Important	18.9	232
	Very important	15.3	188
	Missing	1.8	22

The summary statistics for the maternal outcome variables and indicators of social exclusion are given in Table 2.

Table 2. Summary statistics – maternal outcomes and social exclusion indicators.

Variable	Level	%	N	Min	Max	Mean	SD	N missing
Maternal outcome variables
Maternal life satisfaction			1223	1.00	15.00	9.54	4.12	4
Maternal perceived discrimination			1219	1.00	21.00	3.93	4.06	8
Social Exclusion Indicators
Material deprivation			1213	0.00	12.00	4.04	3.16	14
Workless household	Working household	70.3	862
	Workless household	29.3	359
	Missing	0.5	6
Lower maternal educational level	High (ISCED >5)	13.0	159
	Medium (ISCED 3-5)	47.5	583
	Low (ISCED 0-2)	38.9	477
	Missing	0.7	8
Perceived literacy difficulties			1227	1.00	15.00	3.05	2.85	0
Perceived neighbourhood disorder			1225	1.00	4.00	2.45	0.90	2
Fewer available neighbourhood play areas	Two or more safe play spaces	26.6	326
	One safe play spaces	55.0	675
	No safe play space	18.4	226
	Missing	0.0	0
Perceived lack of social support			1226	1.00	26.00	6.58	6.07	1

Life satisfaction scores were (on average) in the medium range (M = 9.45; SD = 4.12). Mothers’ perceived discrimination was low across all items (M = 3.93; SD = 4.06); scores were highest for social and public media (M = 1.57, SD = 0.98), followed by being from people in the neighbourhood (M = 1.50, SD = .89), followed by parents at the pre-school/school (M = 1.31, SD = 070).

Statistical modelling

The dependent variables were analysed with linear mixed effect regression models in terms of the predictor variables. After establishing the model for life satisfaction in terms of these background variables, perceived discrimination was added as a predictor to the model. Similarly, after establishing the model for perceived discrimination in terms of these background variables, life satisfaction was added as a predictor to the model.

The results for the final regression models are shown in Table 3.

Table 3. Results of multiple linear regression analysis predicting maternal life satisfaction and maternal perceived discrimination.

Predictor variables		Maternal outcomes
		Life satisfaction		Maternal perceived discrimination
		Beta	p	Beta	p
Country and demographic factors
Country	Portugal	(reference)
	Poland	−0.463	0.003	+0.446	0.053
	Italy	−0.390	0.013	+0.263	0.221
	Czech Republic	−0.646	<0.001	+0.323	0.137
	England	−0.111	0.382	+0.480	0.050
	Germany	−0.562	0.005	+0.804	0.013
	Greece	+0.061	0.637	+0.409	0.088
Level of diversity in neighbourhood	(Almost) none	(reference)
	Around a quarter	+0.037	0.606	+0.067	0.381
	Around half	+0.056	0.516	+0.204	0.027
	Around three-quarters	+0.034	0.759	−0.043	0.720
	(Almost) all	+0.052	0.753	+0.134	0.439
Child age group (7-11 vs. 3–6 years)		−0.120	0.036	+0.113	0.061
Child gender (male vs. female)		−0.038	0.478	+0.016	0.768
Mother age		−0.071	0.242	−0.015	0.816
Number of children in the household	1	(reference)
	2	+0.147	0.024	−0.016	0.811
	3	+0.137	0.074	+0.058	0.470
	4 or more	+0.277	0.002	+0.140	0.143
Reported religious affiliation (yes vs. none)		+0.035	0.700	+0.088	0.363
Perceived importance of religion	Not important	(reference)
	Slightly important	+0.013	0.897	+0.089	0.382
	Moderately important	+0.028	0.798	+0.122	0.283
	Important	+0.082	0.445	+0.114	0.312
	Very important	+0.068	0.538	+0.184	0.115
Social exclusion indicators
Material deprivation		−0.368	<0.001	+0.349	<0.001
Workless household		−0.193	0.004	+0.115	0.105
Low maternal educational level	High (ISCED >5)	(reference)
	Medium (ISCED 3-5)	−0.077	0.393	−0.099	0.299
	Low (ISCED 0-2)	−0.052	0.595	−0.093	0.374
Perceived literacy difficulties		−0.076	0.169	+0.149	0.011
Perceived neighbourhood disorder		−0.191	0.001	+0.244	<0.001
Non-availability of neighbourhood play areas	Two or more	(reference)
	One	−0.143	0.032	+0.026	0.711
	None	−0.109	0.211	+0.082	0.373
Perceived lack of social support		−0.430	<0.001	+0.096	0.121
Maternal outcome variables
Maternal perceived discrimination		−0.218	<0.001
Maternal life satisfaction				−0.238	<0.001
Sample size		N = 1134		N = 1134
Model fit (R^2)		R^2 = 0.255		R^2 = 0.162

After controlling for a wide range of demographic and background characteristics, higher material deprivation and higher perceived neighbourhood disorder were associated with lower life satisfaction and higher perceived discrimination. In addition, living in a workless household, fewer available neighbourhood play areas, and perceived lack of social support were associated with lower life satisfaction. Perceived literacy difficulties were associated with higher perceived discrimination. Only one of the indicators of social exclusion, lower maternal educational qualifications, did not have a significant association with either mothers’ perceived discrimination or life satisfaction. In terms of other demographic variables included as control variables, associations were significant for life satisfaction mainly with significant associations for child age group, having more children in the family, and country (four of six countries scoring lower than Portugal).

The regression models for perceived discrimination revealed essentially the same effects for background variables whether or not life satisfaction was included as a predictor in the model. Similarly, the regression models for life satisfaction revealed essentially the same effects for background variables whether or not perceived discrimination was included as a predictor in the model. This pattern of results indicates that, independent of background variables, there is a significant association between perceived discrimination and life satisfaction. However, adding life satisfaction as a predictor for perceived discrimination greatly improved the model, and vice versa.

Such a pattern of results implies the model shown in Figure 1 is appropriate to represent the pattern of relationships between variables. Structural equation modelling was used to test the fit of this model with the data. The results indicate that this model had a good fit with the data (RMSEA < 0.001; CFI > 0.995). Only statistically significant effects are included in the diagram.

Figure 1. Diagram showing results of structural equation modelling. Only statistically significant effects are shown in the diagram.

Findings in the SEM model (Which show and take account of the significant negative association between life satisfaction and perceived discrimination) reflect findings from the regression models with some additional significant associations identified – all in the expected direction. Here, workless households, perceived literacy difficulties, lack of social support and having the target child in the older age group is significantly associated with both outcomes, predicting lower life satisfaction and higher perceived discrimination. In addition, religious affiliation is positively associated with perceived discrimination.

Discussion

The present study investigated potential links between indicators of social exclusion and perceived maternal discrimination and maternal life satisfaction amongst low-income, non-immigrant mothers of young children in seven European countries. The focus was on families who had been in their respective societies for several generations, and the aim was to investigate associations between non-ethnically motivated discrimination and life satisfaction associated with socio-economic disadvantage.

This study showed that lack of resources in multiple dimensions of social exclusion was linked to less satisfaction with life and higher perceived discrimination. In line with previous research (Diener & Biswas-Diener, 2002; Helliwell, 2001), this study indicates that objective life condition variables associated with social exclusion (material deprivation, unemployment, poor neighbourhood quality) are important in contributing to well-being, but that other more relational aspects also have to be considered when addressing causes and consequences of poverty. These findings point towards the importance of social support and neighbourhood characteristics when considering policies and interventions that aim to increase social inclusion and well-being for families who face barriers to participation in society.

An important contribution of this study is the exploration of variables indicating deprivation in education and skills as contributors to subjective well-being and perceived discrimination. These are relevant because they link to employability, income, and living conditions. Whilst no significant association was found between maternal education levels and either life satisfaction or perceived discrimination, literacy related difficulties with the native language (including listening, reading and conversation skills) were associated with higher perceived discrimination, and (in the SEM model) also life satisfaction. This indicates that adult literacy and language competencies are important dimensions for policy consideration in the context of disadvantage. These findings are in line with other studies showing that language and literacy difficulties can create barriers to participation in employment, access to services, social relations, and civic activities (Gele & Harslof, 2012; Stewart et al., 2011). Thus, this study further supports the importance of language and literacy related competencies as key indicators of social inclusion (Bird & Akerman, 2005). It draws attention to the fact that shortcomings in language and literacy skills are not only an important consideration for families with immigrant and other language backgrounds, but also need to be addressed when supporting families who do not have to learn the national language as a second language.

Taken together, the results confirm that parents who are deprived and excluded at multiple levels are likely to be less satisfied with their lives and to perceive more discrimination. Life satisfaction and perceived discrimination were found to be associated, indicating that both dimensions need to be considered in relation to subjective well-being. While research has drawn attention to the fact that parent well-being contributes to child well-being, little research has examined the importance of perceived discrimination or life satisfaction for parenting behaviour. More research needs to be done to close this gap.

While perceived discrimination levels were relatively low across all countries studied, perceived discrimination was linked to measures of social exclusion and well-being, which has important implications. Relatively little is known on resources available to families to negotiate stigma attached to being socio-economically disadvantaged. Identification with ones ‘group’ has been found to be a protective mechanism in the light of discrimination (Branscombe et al., 1999), but there is doubt that community identity or belonging associated with social class can be supportive in the same way, leading to suggestions that in this context, experiences of discrimination can lead to withdrawal and decreased levels of belonging, thus accelerating marginalization (Bradshaw et al., 2015).

Data for this study was drawn from a survey carried out as part of the ISOTIS project (Broekhuizen et al., 2018), including a large, cross-national, European sample. The present analysis is based on data collected from mothers of preschool-and primary-school-aged children in families with no recent immigration background, living in relatively deprived areas in seven European countries. A strength of this study is the fact that it focuses on a population which is traditionally under-researched and challenging to engage in research (Levitas et al., 2007). However, it has to be acknowledged that purposive sampling was applied, which is reflective of cross-national research, and such targeting populations who are living in deprived circumstances. In the absence of population weights this sample cannot be created as representative of the population. We therefore refrain from a cross-county comparison, and treat the population of this study as one diverse group of disadvantaged families without recent immigrant background, living in different areas and contexts in several European countries. Nevertheless, models in this study showed associations between the co-variate ‘country’ and maternal outcomes which were significant over and above neighbourhood, family and child-level variables. This indicates the importance of macro-level factors when investigating life satisfaction and perceived discrimination. Future research needs to further examine and explain these differences, including macro-level variables such as national governance strategies, employment policies and educational and family support practices.

It is important to note that this study is focusing on maternal outcome variables only.

This is a particular shortcoming of this study; fathers play an important yet und-researched role in nurturing and shaping their children's growth, and there is a gap on research evidence on those factors that facilitate outcomes for fathers. Further limitations of the study relate to the fact that the outcome measures used for this study are complex, yet their measure in this study is based on a limited number of survey questions; more questions may better have better captured a wider range of experiences that that contribute to each construct.

Nevertheless, by investigating how contextual factors relate to maternal life satisfaction and perceived discrimination, this study contributes to the understanding of the different aspects and implications of poverty and social exclusion on the wellbeing of families, and therefore the lives of children. This is vital for the planning and delivering of policies and services that not only address factors that contribute to deprivation, but also compound its impact on wellbeing – including for example factors such as discrimination and social inequality. To conclude, this study contributes research findings that reaffirm the importance of combatting social exclusion. There are implications for public policy, emphasizing the importance of joined-up policies that tackle different forms of exclusion, and increasing commitment to services which focus on strengthening families’ sense of belonging and supportive inter-group interactions, ‘bridging social capital’ as described by Putnam (2000) in the context of diverse communities. In addition, the results concerning the source of perceived discrimination have implications for guidelines for the media in terms of how people from low-income deprived backgrounds are represented, and potentially for professional development for those working with low-income deprived families in supporting them in negotiating stigmas attached to their statues.

Acknowledgements

This study would not have been possible without the efforts of the researchers in each participating country and the collaboration of the wider ISOTIS team. ISOTIS researchers involved in this study were: Lenka Kissová, Zuzana Lenhartová, Jana Obrovská, Ladislav Zilcher (Masarykova Univerzita), Edward Melhuish, Katharina Ereky- Stevens, Pinar Kolancali (University of Oxford, England), Katrin Wolf, Hande Erdem, Beyhan Ertanir, Yvonne Anders (Freie Universität Berlin), Konstantinos Petrogiannis, Ioanna Strataki (Hellenic Open University), Sophie Sarcinelli, Cecilia Rubiolo (University of Milano-Bicocca), Olga Wysłowska, Kamilla Wichrowska (Uniwersytet Warszawski), Inês Ferreira, Cecília Aguiar (Instituto Universitário de Lisboa). More information on the large-scale structured interview study can be found in Broekhuizen et al. (2018). The authors would like to thank all institutions, schools and professionals who supported us in reaching out and contacting participant families and/or providing the conditions for conducting some of the interviews with families. We are very grateful to all mothers and other primary caregivers who participated in this study, by sharing their experiences and perspectives.

Disclosure statement

No potential conflict of interest was reported by the author(s).

Additional information

Funding

The research leading to these results is part of the ISOTIS project (www.isotis.org). The work was supported by the European Union's Horizon 2020 Framework Programme for Research and Innovation under grant agreement No. 727069.