Socioeconomic Status, Family Processes, and Individual Development

Abstract

Research during the past decade shows that social class or socioeconomic status (SES) is related to satisfaction and stability in romantic unions, the quality of parent-child relationships, and a range of developmental outcomes for adults and children. This review focuses on evidence regarding potential mechanisms proposed to account for these associations. Research findings reported during the past decade demonstrate support for an interactionist model of the relationship between SES and family life, which incorporates assumptions from both the social causation and social selection perspectives. The review concludes with recommendations for future research on SES, family processes and individual development in terms of important theoretical and methodological issues yet to be addressed.

We begin this report by considering the economic changes families have experienced during the period from 2000 to the present. Unfortunately, this review demonstrates that many of today’s families face significant financial problems as a result of the current crisis in the economy. Following our discussion of economic conditions, we turn our attention to evidence for the association between socioeconomic status (SES) and (1) satisfaction and stability in adult romantic relationships, (2) the quality of parent-child relationships, and (3) the personal adjustment of adults and children. To guide our efforts we used these and related keywords in our search for information through Sociological Abstracts, PsycINFO and direct inspection of major family and developmental journals such as Journal of Marriage and Family. From our review of research and theory during the past decade, we conclude that the relationship between SES and the quality of family life is less simple than once assumed. Although we find support for the notion that a family’s socioeconomic position affects the life course development and interrelationships of family members, we also find evidence that individual differences appearing during childhood and adolescence help shape economic and educational achievements and also competence as a parent and romantic partner during the adult years. We conclude this review by considering the implications of these observations for the development of theory and research during the next decade.

Because of space constraints, we were selective in our use of reference materials and incorporate only those with the greatest relevance for the purposes of this report. In addition, because one article in this decade-in-review issue considers economic influences on family formation (Smock & Manning, 2010), and another by Edin and Kissane (2010) describes the work on poverty during the past decade, including the influence of poverty on family processes and child development; we do not address these issues in this report. Rather, we focus our attention on research which measures socioeconomic status or social class as a continuum rather than as a set of categories such as poor versus not poor. Thus, the focus of this review is primarily on quantitative analyses inasmuch as most of the qualitative work on families and social class or SES is concerned with the issue of poverty.

Especially important, our primary interest concerns possible theoretical explanations for the association between SES and family life. Research dating back to the depression years of the 1930s has confirmed that families often suffer when faced with economic hardship or low SES (e.g., Angell 1936). A crucial and continuing issue concerns the mechanisms that might account for the relationship between SES and family processes, as well as factors that might moderate that relationship. In this review we are particularly interested in theoretical progress during the past decade that improves understanding of the nature of the relationship between SES, family processes, and individual well-being. After reviewing evidence for sometimes opposing theoretical perspectives, we evaluate important new directions for future theoretical developments in this area of research. We also address the methodological requirements for adequately pursuing these theoretical issues. Before turning to these matters, we describe economic trends during the past decade, consider the measurement of SES and social class, and review underlying assumptions of various theoretical approaches.

The Economic Climate of the New Millennium

The first decade of the new millennium (i.e., 2000-2009) has been one of uncertainty and instability. Economic growth has averaged slightly over 2% per year since 2000, compared to 3% per year during the previous two decades and 4% in the 1960s (U. S. Department of Commerce, 2009). Following the mid-1990s, housing prices soared, increasing on average nearly 50% after two decades of stability. Since then, the twelve-month change in nominal house prices has turned negative nationwide for the first time since the Great Depression and mortgage loan foreclosures have soared (OECD, 2008), underscoring the significant economic distress in the U.S. as the decade draws to a close. A number of trends further demonstrate these adverse changes in the economy.

For example, the total unemployment rate among those aged 16 years and over rose from 4% in 2000 to 9.7% during June and July of 2009. By the end of 2009, the national unemployment rate was over 10% and the underemployment rate was around 16% (U.S. Bureau of Labor Statistics, 2009); these figures reflect the highest level of unemployment since the severe recession of the early 1980s (Gomstyn, 2009; Irwin & Shin, 2009). Although unemployment increased across all racial and ethnic groups during the 2000s, unemployment was experienced disproportionately by African Americans and Hispanics. Problems with employment are reflected in levels of family income. From 1995 to 2000, median family income for all families increased from $56,971 to $63,430, followed by a decrease to $61,976 in 2005, and then another decrease in 2008 to $61,521 (in 2008 dollars; U.S. Bureau of the Census, 2009). However, these trends have varied by race and family structure. Whereas White families saw median income decrease from $70,317 in 2000 to $70,070 in 2008, African American families saw a decrease from $42,105 to $39,879 in the same period, and Hispanic family wages decreased from $43,063 to $40,466 (in 2008 dollars; U.S. Bureau of the Census, 2009).

In 2008, mother-headed (single-parent) families, father-headed (single-parent) families, and traditional male-as-breadwinner families also had median incomes that were less than they had been in 2000 (in 2008 constant dollars; U.S. Bureau of the Census, 2009). Married-couple families with both spouses working earn more than any other family structure. In contrast, the median income of mother-headed families was less than half that of married couple families throughout the 2000s. Father-headed households fared slightly better. The median income for families with and without children under 18 was very similar in 2000 ($63,478 and $63,388, respectively); however, since then the wages for families with children have dropped, while families without children reached a high in 2007 of $65,940 (in 2008 dollars; U.S. Bureau of the Census, 2009). Also important, the 2000s saw a continuation in the structure of income distribution established in the 1990s, which strongly benefited the upper classes. The top five percent’s share of income grew from 14.6% in 1980 to 20.5% in 2008. At the same time, the lowest quintile’s share has fallen from 5.3% in 1980 to its current share of around 4% (U.S. Bureau of the Census, 2009). Most remarkable, the top 20% of families account for almost 50% of aggregate income in the U.S.

Race and gender continued to influence the economic circumstances of individuals and families during the past decade. Although the gender gap between women’s and men’s median annual earnings reached an all time low, with women making 77.8% of men’s income for full-time workers in 2007 compared to 71.6% in 1990, a significant gender gap continues to exist. Moreover, racial and ethnic minorities have not seen gains to the same degree as women. African American family income, for instance, was 56.0% of White family income in 1990 compared to 56.9% in 2008 (U.S. Bureau of the Census, 2009). The statistics for families with children under 18 living in poverty are similarly dismal. All racial groups for which there are complete data showed increases in the poverty rate from 2000 (12.7%) to 2008 (15.7%), yet the contrasts between groups are stark. In 2000, 7.7% of White families with children lived in poverty, while 25.3% of African American and 23.3% of Hispanic families with children had incomes below the poverty line. By 2008, this had grown to 9.3% of White families, 29.6% of African American families, and 26.8% of Hispanic families (U.S. Bureau of the Census, 2009). For the first decade of this century, then, it appears that almost all families have suffered economically, but ethnic minority families have suffered the most.

Measurement of SES or Social Class

As noted, the economic decline during the last decade has placed significant pressures on many families in terms of financial distress, reduced employment opportunities, and fewer resources to help family members pursue their educational goals. As we discuss in this section, these dimensions of economic, occupational, and educational experience represent important markers of social class or socioeconomic status. In this report, we use the term social class interchangeably with socioeconomic status (SES) as is typically done in quantitative analyses of class effects (e.g., Haas, 2006; Scott & Leonhardt, 2005). SES is a construct that captures various dimensions of social position, including prestige, power, and economic well-being (Hoff, Laursen, & Tardiff, 2002; Oakes & Rossi, 2003). Most contemporary investigators agree that three quantitative indicators provide reasonably good coverage of the domains of interest: income, education, and occupational status (Bradley & Corwyn, 2002; Ensminger & Fothergill 2003).

Despite the fact that these indicators of social position are positively correlated (Ensminger & Fothergill; 2003), Duncan and Magnuson (2003) suggested that each of these markers of social status demonstrates different levels of stability across time and differentially predicts family processes and child adjustment. Education is one of the most widely used indicators of SES and is considered by many to be the canonical element of SES because of its influence on later income and occupation (Krieger, Williams, & Moss, 1997; Mueller & Parcel, 1981). Thus, income, education, and occupational status are sometimes used together as indicators of SES and families, and 26.8% of Hispanic families (U.S. Bureau of the Census, 2009). For the first decade of this century, then, it appears that almost all families have suffered economically, but ethnic minority families have suffered the most.

Measurement of SES or Social Class

As noted, the economic decline during the last decade has placed significant pressures on many families in terms of financial distress, reduced employment opportunities, and fewer resources to help family members pursue their educational goals. As we discuss in this section, these dimensions of economic, occupational, and educational experience represent important markers of social class or socioeconomic status. In this report, we use the term social class interchangeably with socioeconomic status (SES) as is typically done in quantitative analyses of class effects (e.g., Haas, 2006; Scott & Leonhardt, 2005). SES is a construct that captures various dimensions of social position, including prestige, power, and economic well-being (Hoff, Laursen, & Tardiff, 2002; Oakes & Rossi, 2003). Most contemporary investigators agree that three quantitative indicators provide reasonably good coverage of the domains of interest: income, education, and occupational status (Bradley & Corwyn, 2002; Ensminger & Fothergill 2003).

Despite the fact that these indicators of social position are positively correlated (Ensminger & Fothergill; 2003), Duncan and Magnuson (2003) suggested that each of these markers of social status demonstrates different levels of stability across time and differentially predicts family processes and child adjustment. Education is one of the most widely used indicators of SES and is considered by many to be the canonical element of SES because of its influence on later income and occupation (Krieger, Williams, & Moss, 1997; Mueller & Parcel, 1981). Thus, income, education, and occupational status are sometimes used together as indicators of SES and also as separate variables in data analyses so that investigators can evaluate their unique, additive contributions to family characteristics and human development. In some cases, earlier education can even be used as a predictor of later income and work success. In the following review, we focus on research during the past decade that has examined the relationship between income, occupation, and/or education and family processes.

Underlying Assumptions in Studies of SES and Family Life

An a priori assumption of most research on SES, family functioning, and human development is that social position influences families across time, and that socioeconomic disadvantage has negative consequences for adults and children (e.g., Conger et al., 2002; Haas, 2006). This underlying tenet represents an instance of the social causation perspective which assumes that social conditions lead to variations in health and well-being. Other theoretical models assume that the relationship between SES and family processes is explained by individual differences in the personal characteristics of family members that affect both their SES and their family relationships. This view represents the social selection perspective which assumes that the traits and dispositions of individuals influence both their social circumstances and their future emotions and behaviors (e.g., McLeod & Kaiser, 2004).

Theories based on the social selection perspective have been offered as a serious challenge to the presumption that social disadvantage has a causal influence on families and children, and we consider these theoretical arguments in this review. It is important to note that neither the social causation nor the social selection views are theories in their own right; rather, they represent underlying principles upon which specific theories are based. After reviewing the research evidence generated from theoretical arguments related to both of these perspectives, we conclude that the causal dynamics between SES and family relationships may be more complicated than previously imagined and that new types of theory and research will be required to fully understand these complexities.

SES as a Predictor of the Quality and Stability of Romantic Relationships

In this section we consider theoretical perspectives and empirical findings related to the social causation view that social class or SES should affect romantic relationships. Studies of marriage and other adult romantic relationships typically focus on two primary outcomes: relationship quality (e.g., couple ratings of their relationship happiness or satisfaction) and relationship stability (e.g., divorce, separation, or couple reports of relationship commitment as opposed to thoughts or plans related to divorce). Past research has shown that marital quality or satisfaction is one of the primary predictors of relationship stability or instability (e.g., Karney & Bradbury, 1995); thus, we consider these two interrelated concepts together in this report.

Empirical Findings: SES and Couple Relationships

Research evidence from 2000 to 2009 clearly supports the idea that higher SES both reduces the risk of separation and divorce and also increases the level of satisfaction and happiness in romantic unions (cf., Karney & Bradbury, 2005). For example, research reports during the past 10 years have shown that higher levels of educational attainment are associated with greater marital stability (e.g., Heaton, 2002; Martin, 2006; Orbuch, Veroff, Hassan, & Horrocks, 2002). Similarly, a number of reports have shown that greater income and financial resources are positively associated with marital stability (e.g., Orbuch et al., 2002; Popenoe, 2007; South, 2001; Stanley, Amato, Johnson, & Markman, 2006). In a report on an experimental intervention program designed to enhance the incomes of poor families, Duncan and his colleagues even provided preliminary evidence that families experiencing greater economic gains demonstrated greater marital stability (Duncan, Huston, & Weisner, 2007).

With regard to marital quality, two recent studies found that greater educational attainment was positively related to marital satisfaction (Dakin & Wampler, 2008; Rauer, Karney, Garvan, & Hou, 2008). In addition, a number of reports have shown that low income, financial instability, or economic problems are associated with lower levels of marital quality (e.g., Amato, Booth, Johnson, & Rogers, 2007; Cutrona et al., 2003; Dakin & Wampler, 2008; Falke & Larson, 2007; Karney, Story, & Bradbury, 2005; Rauer et al., 2008; Stanley et al., 2006). These studies have often used measures more indicative of economic pressure or financial strain than income level, however. In that sense they are open to concerns raised by White and Rogers (2000) about the use of subjective measures of economic status. Worth noting, though, is the report by Amato and his colleagues (2007) who found that income was directly related to marital quality as well as indirectly through economic pressure.

Especially important for present purposes, Amato et al. (2007) showed that lower levels of income, educational attainment, and occupational prestige were associated with higher rates of marital problems, less marital happiness, and greater instability. These investigators reported on two broad surveys conducted with couples, one in 1980 and the other in 2000. As part of their analyses, they created a typology of couples based on their social class standing. The types of relationships included couples described as: (a) disadvantaged, young, single earners, (b) working-class, young, dual earners, (c) working/middle-class, traditional single earner, (d) middle-class, dual earner, egalitarian, and (e) upper middle-class, prosperous, mostly dual earner. In their analyses, they demonstrated that the dimensions of marriage discussed earlier conformed in expected ways with SES. In terms of reported divorce proneness, for example, the two most disadvantaged groups reported the greatest marital instability whereas the most prosperous couples reported the lowest probability of risk for divorce. The upper middle class couples also reported the lowest levels of marital conflict and relationship problems. In addition, the most prosperous group reported the greatest marital happiness and the two most disadvantaged groups reported the lowest levels of happiness with their unions. These findings capture very well the basic message from related research during the past decade. On average, higher economic, educational and occupational status is associated with greater marital stability and quality.

An important question, however, is whether these same types of predictions hold for unmarried, cohabiting couples. Unfortunately, little research during the past decade has focused on the relationship between SES and the quality or stability of cohabiting unions. Investigations during the past ten years have demonstrated that cohabiting compared to married couples tend to be less well educated, have lower incomes, lower occupational prestige, less trusting relationships and greater male violence (McLanahan, 2009; McLanahan & Percheski, 2008). The findings did not indicate, however, whether the greater problems in cohabiting compared to marital unions were a result of differences in SES (e.g., Hsueh, Morrison, and Doss, 2009). In a rare study, Wu and Pollard (2000) examined the role of economic circumstances in the stability of cohabiting couples. They found that economic deprivation contributed to relationship instability in cohabiting couples, consistent with previous research on married couples. An overall decrease in household economic circumstances increased the risk of separation of the cohabiting couples they studied. However, increases in personal earnings (rather than household) also increased the likelihood of separation, suggesting an “independence effect” for cohabiting couples that previous research on married couples has not found.

Simply put, there is good evidence that social class or SES is positively related to marital quality and stability and, more tentatively, to the same outcomes for cohabiting partners. Despite findings consistent with the notion that SES predicts relationship outcomes, it needs to be noted that most of the studies reviewed here have both significant methodological strengths, such as large representative samples (e.g., Amato et al., 2007), as well as noteworthy design limitations, such as reliance on a single informant and the absence of information across time (e.g., Stanley et al., 2006). We return to these issues in later comments on directions for future research. We next turn to a central concern of this report; that is, the possible explanation for the positive association between SES and romantic relationship quality and stability. In our review of research during the past decade, we found that a majority of theoretically driven studies on this issue drew in whole or in part on predictions from the family stress model (FSM).

The Family Stress Model (FSM) of Romantic Relationships

The term “family stress model” was first proposed by Conger and Elder (1994). The FSM predicts that economic problems will lead to deterioration in marital relationships and increase risk for marital instability. Although the FSM’s focus is on economic conditions, we suspect it also captures much of the influence of limited educational or occupational achievement. For example, education is an important predictor of income across the life course (Krieger et al., 1997) and there is a strong positive association between occupational prestige and income (Treiman, 1976). Given these established relationships, it is reasonable to expect that much of the influence of educational or occupational status on marital functioning will be indirect through variations in economic well-being. Thus, we assume that findings related to the economic predictions from the FSM likely reflect educational and occupational differences in SES as well.

Conger and his colleagues developed the FSM to help explain how financial problems influenced the lives of rural families going through the severe downturn in the agricultural economy during the 1980s (see Conger & Conger 2002, Conger et al., 2002). Figure 1 illustrates predictions from the FSM to the quality and stability of marital and other romantic relationships beginning with the proposition that economic hardship leads to economic pressure for partners in romantic unions. Markers of hardship include low income, high debts relative to assets, and negative financial events (e.g., increasing economic demands, recent income loss, and work instability). These hardship conditions are expected to affect couples primarily through the economic pressures they generate including: (a) unmet material needs involving necessities such as adequate food and clothing, (b) the inability to pay bills or make ends meet, and (c) having to cut back on even necessary expenses (e.g., health insurance and medical care). According to the FSM, the experience of these kinds of pressures or strains gives psychological meaning to economic hardship (Conger & Conger, 2002; Conger et al., 2002). Contrary to the suggestion by some researchers that economic pressure involves subjective impressions (e.g., White & Rogers, 2000), the FSM holds that these are tangible events in an individual’s life that can significantly impact family functioning.

Figure 1.

The family stress model of romantic relationships.

The model predicts that when economic pressure is high, romantic partners are at increased risk for emotional distress (e.g., depression, anxiety, anger, and alienation) and for behavioral problems (e.g., substance use and antisocial behavior) (Conger et al., 2002). This proposition regarding the role of economic pressure in exacerbating maladjustment derives from Berkowitz’s (1989) reformulation of the frustration–aggression hypothesis. Berkowitz demonstrated that many stressful, frustrating, punishing, or painful events and conditions are lawfully related to increased emotional arousal or negative affect that ranges from despondency to anger in both humans and other animal species. For the FSM, economic pressure is a construct that reflects the kinds of painful or frustrating experiences hypothesized by Berkowitz to increase emotional distress and behavioral disturbances. According to the FSM, angry responses to economic pressure are expected to increase couple conflict and despondency which is expected to lead to withdrawal of supportive behaviors and reductions in pleasurable interactions. As shown in Figure 1, this economic stress process is hypothesized to lead to decreases in relationship quality and stability.

Several specific tests of this aspect of the FSM during the past decade have produced results consistent with the model. For instance, in a study of rural and urban African American couples, Conger et al. (2002) found that economic hardship predicted economic pressure which, in turn, exacerbated emotional distress for both partners in the relationship. As expected, emotional distress increased conflict in these relationships. Two other studies, one conducted in Finland (Solantaus, Leinonen, & Punamäki, 2004) and the other in southern California (Parke et al., 2004), also evaluated predictions from hardship to pressure to distress to conflict and generated similar results. Worth noting is the fact that the California study specifically recruited separate subsamples of European American (N = 111) and Mexican American (N = 167) couples and found that predictions from the FSM for romantic relationships held for both ethnic groups. In a more extended evaluation of the Finnish study of 608 couples, Kinnunen and Feldt (2004) found support for all of the hypothesized pathways in the FSM and also showed that economic pressure sometimes had direct as well as indirect effects on marital adjustment. The presence of a direct effect from economic pressure to marital problems has also been reported for studies in Korea (Kwon, Rueter, Lee, Koh, & Ok, 2003) and Turkey (Aytac & Rankin, 2009). Thus, while hypotheses from the FSM have been supported both in the U.S. and in other countries, there is some evidence that economic pressure is more likely to have a direct as well as a mediated impact on marital outcomes outside the U.S. This finding may relate to different cultural and/or economic traditions in other societies that are not fully understood at this point in time.

To summarize, there is replicated evidence with diverse populations suggesting that the FSM provides a reasonable account of the stress processes through which economic hardship is related to marital functioning. Part of this stress process involves the concept of economic pressure or strain, which we argue is not simply a subjective impression but a real descriptor of the aversive events that occur in people’s lives when they are under financial duress. Despite evidence during the past decade which supports the FSM model of SES and romantic relationships, this research has both significant methodological strengths as well as noteworthy limitations. A major limitation is that all of these studies employed cross-sectional research designs, thus leaving unresolved the question of the temporal ordering of the processes described in the model. This concern is mitigated somewhat by earlier research demonstrating that economic pressure leads to changes in partner emotional distress and marital conflict over time (Conger, Rueter, & Elder, 1999). Strengths of several of these reports include the use of representative community samples (e.g., Solantaus et al., 2004) and the use of multiple reporters for assessing study variables, including the use of trained observers of couple interactions (e.g., Conger et al., 2002). We return to these issues later in the discussion of future research directions.

Before turning to consideration of the role of SES in child development, it is important to note that the FSM interpretation of the relationship between SES and the quality of romantic relationships is consistent with the broad field of research on health disparities, the observation that socially and economically disadvantaged adults and children are at increased risk for both physical and mental health problems (e.g., Berkman & Kawachi, 2000; Herd, Goesling, & House, 2007; Kim & Durden, 2007; Link, 2008; Oakes & Rossi, 2003; Wickrama, Conger, Lorenz, & Jung, 2008). This body of work demonstrates that low SES adults in general are at greater than average risk for developing diverse types of health problems. A noteworthy limitation of the FSM, though, is that it describes a simple recursive process from SES to the quality of family relationships. Once one changes the frame of reference from the proposed influence of SES on adult psychological functioning and the quality of romantic relationships to adult development in general, a much more complex, interdependent dynamic is involved. Returning to Figure 1, for example, a more fully articulated account of the hypothesized family stress processes would include feedback loops from relationship dynamics to adult psychological functioning and to economic problems.

For instance, various studies have shown that low marital quality is related to psychological distress and is one of the main reasons that people seek counseling for emotional and behavioral problems (e.g., Berscheid, 1999; Overbeek et al., 2006). We would expect, then, that over time families under economic pressure will not only experience the types of problems described by the FSM but that these relationship problems will feedback to further disrupt adult psychological well-being and the capacity to deal successfully with economic difficulties (see Conger & Conger, 2002). Moreover, interparental discord and relationship dissatisfaction are likely to spill-over into parent-child relationships, thus increasing conflict and chaos in the family environment more generally and posing additional risks for adult psychological and economic problems (e.g., Cui, Donnellan, & Conger, 2007; Nelson, O’Brien, Blankson, Calkins, & Keane, 2009). The most extreme relationship outcome described by the FSM involves actual separation or divorce. For both cohabiting and married couples, such relationship disruptions are related to physical and psychological difficulties and to economic dislocations, especially for women (e.g., Avellar & Smock, 2005; Blekesaune, 2008; Liu & Umberson, 2008; Williams & Dunne-Bryant, 2006: Zhang & Hayward, 2006). These findings underscore the types of reciprocal processes expected to operate in an extension of the FSM across the life course (Figure 1). In addition, Amato and Cheadle (2005) have shown that divorce in one generation has continuing negative consequences for psychological functioning, educational attainment, and the quality of family relationships across multiple generations, again demonstrating the importance of relationship disruption for adult development and the future functioning of family members.

SES as a Predictor of Parenting and Child Development

One of the exciting innovations in research on SES and children during the past decade has come in the form of randomized experiments involving intervention programs for low income families. In these studies, families are randomly assigned to either an intervention or a control group and comparisons are made between the groups after intervention. These experimental tests provide the best evidence for a causal relationship between income and child development. Results from these studies have produced evidence that these programs can have a positive influence on parents’ well-being and on developmental outcomes for children and adolescents. Although the findings are quite complex and tend to be contingent on a number of factors, such as the age or gender of the child, there is a growing body of evidence suggesting that improvements in family income may have beneficial effects for parents and children (e.g., Huston et. al., 2005; Leventhal, Fauth, & Brooks-Gunn, 2005; Morris, Duncan, & Clark-Kauffman, 2005). Consistent with such findings, Costello and her colleagues reported results from a quasi-experimental study which demonstrated that, after a casino opened in a poor community, increases in parental employment and family income were associated with decreases in behavioral problems for children in the study (Costello, Compton, Keeler, & Angold, 2003).

Given this array of evidence regarding a possible causal connection between SES and child development, the search for the family mechanisms that may account for this association becomes even more important. In this regard, recent research on SES and children’s development has primarily been guided by two theoretical frameworks, each of which focuses almost exclusively on family wealth or income (see Conger & Donnellan, 2007; Gershoff, Aber, Raver, & Lennon, 2007; Yeung, Linver, & Brooks-Gunn, 2002). The first theoretical paradigm involves the extension of the family stress model (FSM) from the relationships of couples, including parents, to the relationships between parents and children and how they may be adversely affected by family financial difficulties (Conger & Conger, 2002). The second perspective, the investment model (IM), proposes that economic resources increase the investments parents make in their children’s development, thus promoting a wide range of academic and social competencies that accrue to the benefit of the child (Bradley & Corwyn, 2002; Mayer, 1997). We evaluate the empirical research on social class and children during the last decade within the context of these two theoretical frameworks, beginning with the FSM.

The Family Stress Model (FSM) and Children’s Development

Figure 2 illustrates the extension of the FSM from the couple relationship, shown in Figure 1, to the lives of children: this model predicts that economic hardship primarily influences the development of children through the lives of parents. Notice that family economic hardship and pressure are related only indirectly to children’s adjustment through their influence on the behavioral and emotional functioning of parents. Earlier we addressed only the circumstances of romantic partners; however, the model also applies to single-parent families inasmuch as the custodial parent living only with children would also be expected to be at increased risk for emotional or behavioral problems when economic pressure is high.

Figure 2.

Extension of the family stress model to the lives of children.

According to the model, parents’ economically influenced maladjustment predicts problems in relationships between parents, consistent with Figure 1, and also problems in parenting such as harsh, uninvolved and inconsistent childrearing practices (Conger & Conger, 2002; Conger et al., 2002). Research indicates that the prediction of conflict and withdrawal is valid not only for biological parents but also for stepparents, cohabiting unmarried romantic partners, and other caregiver relationships such as daughters and mothers raising a child together (Conger et al., 2002). The FSM also proposes that interparental conflict and relationship problems will be directly related to disruptions in parenting. The primary hypothesis is that disrupted parenting will mediate or explain the influence of parental distress and interparental conflicts on child development, including both decrements in competent functioning (e.g., cognitive ability, social competence, school success, and attachment to parents) and increases in internalizing (e.g., symptoms of depression and anxiety) and externalizing (e.g., aggressive and antisocial behavior) problems. For single-parent families, caregiver conflicts with one another may be omitted from the model or conflicts with an ex-spouse or current romantic partner might be substituted, as economic problems are expected to affect these relationships as well (Conger et al., 2002). Although elaborations of the FSM include factors that promote resilience or exacerbate vulnerability to these mediating pathways, the model in Figure 2 provides the basic tenets of this theoretical framework for the lives of children (Conger & Conger, 2002).

During the past decade an impressive number of studies provided support for predictions from the FSM using the same or quite similar labels for constructs as described in Figure 2. All of these studies reported their findings since 2000 and they represent wide ethnic, geographic, and structural variations. The child adjustment outcomes in these studies were quite varied and included an array of measures reflecting either competent development (e.g., academic performance) and/or emotional or behavioral problems. Parenting behaviors reflected either positive care giving (e.g., warmth and involvement) and/or negative childrearing (e.g., harsh and inconsistent). Findings consistent with the FSM were reported for: (a) 422 two-parent or two-caregiver (e.g., daughter and grandmother) African American families with fifth grade children (Conger et al., 2002); (b) 419 primarily minority (57% African American, 28% Hispanic) and single-parent (83%) families with children from 5 to 12 years of age (Mistry, Vandewater, Huston, & McLoyd, 2002); (c) a nationally representative sample of 753 families and children ages 3 to 5 years (Yeung, Linver, & Brooks-Gunn, 2002); (d) 527 two-parent families in Finland with a focal child between 12 and 13 years of age (Solantaus, Leinonen, & Punamäki, 2004); (e) a community study of European American (N = 111) and Mexican American (N = 167) families of fifth graders who lived in urban areas of Southern California (Parke et al., 2004); (f) 444 Chinese American families with an early adolescent child living in northern California (Benner & Kim, in press); and (g) a study of 493 White and African American families and preschool-aged children (Linver, Brooks-Gunn, & Kohen, 2002).

The degree of replication across these diverse studies is quite remarkable. For example, four of the studies specifically recruited 2-parent families from distinct racial or ethnic groups: African American (Conger et al., 2002), Finnish (Solantaus et al., 2004), Mexican American (Parke et al., 2004), and Chinese American (Benner & Kim, in press). The results from each of these studies provided at least some support for each of the hypothesized pathways in the FSM. In structural equation models, all four studies found positive and statistically significant paths from: (a) indicators of economic hardship to economic pressure, (b) economic pressure to parent emotional distress, (c) parent emotional distress to conflicts between parents, (d) conflicts between parents to disruptions in effective parenting behaviors, and (e) disruptions in parenting to child maladjustment. The biggest exception to this pattern of findings occurred for the Mexican American families in that interparental conflict in this ethnic group directly predicted child maladjustment.

In addition to these studies that involved direct replications of all or parts of the FSM, other studies have reported interesting extensions or variations. For example, Gutman, McLoyd, and Tokoyawa (2005) found results consistent with the FSM for 305 low income African American families with early adolescent children living in inner city neighborhoods and also extended the model by examining neighborhood stress in their analyses. Gershoff and her colleagues (2007) found support for FSM predictions using data from a nationally representative sample of over 20,000 parents or parent figures and their 6 year old children in a study that included constructs similar to the FSM but labeled in different ways. For example, economic pressure was conceptualized as “material hardship” which included indicators of insufficient resources to meet basic material needs and pay monthly bills. Two recent reports with Mexican American families and children also support predictions from the FSM but add cultural factors to the model to evaluate their contributions to the economic stress process (Behnke et al., 2008; White, Roosa, Weaver, & Nair, 2009). Leinonen, Solantus, and Punamäki (2003) also showed that instrumental and emotional support protected Finnish mothers from the adverse influence of economic pressure on their parenting behaviors. In addition, Sobolewski and Amato (2005) demonstrated that the FSM predicts to children’s psychological adjustment into the adult years.

Mistry and her colleagues have pursued a number of interesting extensions of the basic FSM. For example, using data from 1,363 families with children from 6 to 36 months of age, they showed that the mediated relationship between income and child social and cognitive outcomes through family processes was stronger for poorer families (Mistry, Biesanz, Taylor, Burchinal, & Cox, 2004). In a study of 516 primarily minority and single-parent families with children aged 6 to 15 years, Mistry, Lowe, Benner, and Chien (2008) extended the basic FSM by demonstrating that both unmet desires for economic “extras” as well as economic needs play an important role in the economic stress process. They also found evidence that certain financial management strategies and informal sources of economic support temper the adverse influence of economic hardship. Finally, Mistry, Benner, Tan, and Kim (2009) hypothesized that adolescent perceptions of economic problems would be predicted by parent reports of economic pressure and that adolescent reports would be associated with school achievement. Based on a study of 444 Chinese American families with early adolescent children, their results were consistent with their extended model; however, they did not include other FSM mediators such as family conflict or parenting in their analyses so it remains unclear whether adolescent perceptions would be directly related to economic pressure with these variables in the study. Earlier research suggests that they might not be (Conger, Conger, Matthews, & Elder, 1999).

Taken together, these studies suggest that the FSM provides a reasonably good heuristic model for understanding how economic hardship influences family members, socialization processes, and the positive or problematic adjustment of children and adolescents. Also noteworthy, however, are some of the limitations in the research to date. For instance, Barnett (2008) suggested that the FSM and similar models need to be extended to more complex social arrangements that go beyond the traditional nuclear family. McLanahan and Percheski (2008) recently proposed a model very similar to the FSM that may provide a useful means for addressing this issue. They noted, for example, that some family structures are associated with fewer economic and parental psychological resources and proposed that these limitations will affect child outcomes through their influence on parenting behaviors, a hypothesized process very similar to the FSM. Another methodological limitation of the studies discussed here is that most of them are cross-sectional. On the other hand, there are many methodological strengths in these investigations including the use of representative community samples (e.g., Solantaus et al., 2004), reports by multiple family members (e.g., Parke et al., 2004), and the use of observational data describing family interactions (e.g., Mistry et al., 2002). We return to these methodological issues in later discussion of future research directions.

The Investment Model (IM) and Children’s Development

The IM is primarily concerned with the advantages that accrue to the developing child because of family wealth and financial prosperity. The IM proposes that families with greater economic resources are able to make significant investments in the development of their children whereas more disadvantaged families must invest in more immediate family needs (Bradley & Corwyn, 2002; Duncan & Magnuson, 2003; Linver et al., 2002; Mayer, 1997). These investments in children involve several dimensions of family support including (a) parent stimulation of learning both directly and through support of advanced or specialized tutoring or training, (b) the provision of adequate food, housing, clothing, and medical care, and (c) living in a more economically advantaged neighborhood that fosters a child’s competent development.

In general, the IM predicts that economic well-being will be positively related to child-rearing activities expected to foster the academic and social success of a child. Although the IM does not consider the influence of education or occupational status on these types of parental investments, Conger and Donnellan (2007) have proposed that these markers of SES should have an influence similar to family wealth or income. That is, better educated parents with higher than average occupational status should place a priority on the development of a child’s human capital in the form of good health and educational success. Several studies have confirmed the two most basic propositions of the investment model; that is, that family income during childhood and adolescence is positively related to academic, financial and occupational success during the adult years (Bradley & Corwyn, 2002; Mayer, 1997; Teachman, Paasch, Day, & Carver, 1997) and that family income promotes parental investments that are expected to foster the social, emotional and cognitive well-being of their children (Bradley & Corwyn, 2002; Mayer, 1997).

Most central to the purposes of this review, however, are findings during the past decade that relate to the IM hypothesis that parental investments will explain the association between SES and child development. In a paper published in 2002, Linver and her colleagues provided clear evidence for the set of empirical relationships proposed by the model. They used information from several hundred families participating in a large-scale, multiethnic study of children from birth to 5 years of age at the time of their analysis. Consistent with the IM, the association between family income and child cognitive development at ages 3 and 5 years was significantly reduced when the investment mediator (i.e., language stimulation, teaching colors and numbers, providing books and other learning materials, and exposing the child to learning experiences outside the home) was introduced into the analyses. A test of indirect effects also supported the conclusion that the influence of family income on cognitive development was partially the result of these parental investments. The investigators also found that the measure of parental investment completely mediated the association between income and child behavior problems at 3 and 5 years of age. In a similar set of analyses using data from a nationally representative sample of 753 families with children ages 3 to 5 years, Yeung et al. (2002) found evidence that family income had an influence on child outcomes at least partly through parental investments in the competent development of children. Interestingly, both the Linver and Yeung studies examined predictions from both the IM and FSM. Each study found that family stress processes were better predictors of behavioral problems whereas parental investments were better predictors of cognitive development.

Finally, using data from a nationally representative sample of over 20,000 parents or parent figures and their 6 year old children, Gershoff and her colleagues (2007) also conducted a joint test of both the FSM and the IM. As already noted, they found evidence consistent with predictions from the FSM. They also demonstrated support for expectations derived from the IM. That is, they showed that family income predicted parental investments such as the purchase of cognitively stimulating materials for the child and these investments, in turn, largely accounted for the association between income and child cognitive development. Consistent with the two studies just discussed, they also found that the causal pathways proposed by the IM primarily predicted cognitive development of the young children in the study and the causal pathways proposed by the FSM predicted social emotional development. These findings are in need of further investigation in future studies.

To summarize, a few studies during the past decade have found support for the meditational processes proposed by the IM from income to parental investments to child competent development. Further research on this hypothesized mediated pathway is crucial for gaining additional understanding of how family income or wealth conveys developmental advantage across the life course. As with studies related to the FSM, however, these investigations suffer from the fact that they were primarily cross-sectional and did not provide information over time. Their strengths include large, representative and diverse samples and multi-informant measurement of theoretical constructs. We return to these methodological issues in later discussion of future research directions.

Predicting from the Individual to SES

The theories and findings reviewed thus far assume that SES should have a major influence on family functioning and the lives of individual family members, consistent with the social causation perspective. An alternative line of reasoning, however, assumes that the characteristics of individuals will shape both their socioeconomic attainments and the quality of their family relationships. According to this view, connections between SES and family functioning result from a process of social selection (e.g., Lerner 2003; Mayer 1997; Rowe & Rodgers, 1997). To better understand this viewpoint, it is useful to first conceptualize SES as a constellation of developmental outcomes that are potentially influenced by individual differences in traits such as intelligence and personality. In theory, positive individual characteristics like persistence in the face of adversity both facilitate the accumulation of social and economic advantages and also are transmitted from parents to children. The mode of transmission could be genetic (e.g., Rowe & Rodgers, 1997), or it could involve social learning of particular dispositional styles across generations.

What is critical is the proposition that the observed associations among SES, family processes, and child and adolescent development result from their common dependence on personal traits and dispositions rather than from causal connections among them. In the case of child development, for example, Mayer (1997, pp. 2-3) proposed that, “parental characteristics that employers value and are willing to pay for, such as skills, diligence, honesty, good health, and reliability, also improve children’s life chances, independent of their effect on parents’ income. Children of parents with these attributes do well even when their parents do not have much income.” And although they did not link their arguments to the role of SES in these processes, Asendorpf (2002) and Lykken (2002) have made similar arguments about the importance of personality and genes in the development of romantic relationships. Interestingly, McLanahan and Percheski (2008, p. 265) suggested the possibility of similar selection processes with regard to the influence of family structure on children’s development when they noted that “parents’ interpersonal skills likely affect both their family structure and their parenting skills. Children of parents with poor interpersonal skills who experience a lower quality of parenting likely have worse outcomes than other children, regardless of whether they live with both parents.”

If these arguments are correct, then the FSM and IM may not provide valid causal accounts of the role that SES plays in marital relationships and child development. For instance, returning to Figure 2, the social selection argument proposes that positive characteristics of parents, such as those described by Mayer (1997), will reduce exposure to economic hardship and pressure, decrease the likelihood of parent emotional distress and interparental conflict, foster nurturing and involved parenting, and lead to greater child well-being. This proposition leads to the statistical expectation that the connections among the economic variables, family stress processes, and child development predicted by the FSM will be greatly reduced or eliminated once these third variables are included in data analyses. The same arguments would apply to the connections among SES, parental investments, and child developmental outcomes as proposed by the IM.

Consistent with the assumptions of social selection, research reported during the past decade has demonstrated that individual traits and dispositions during childhood and adolescence predict to later status attainment. There is evidence from longitudinal studies, for example, that early emerging individual differences in personality, aggressiveness, and cognitive ability predict SES-relevant outcomes in adulthood such as income, occupational status, economic stress and bouts of unemployment (e.g., Donnellan, Conger, McAdams, & Neppl, 2009; Feinstein & Bynner, 2004; McLeod & Kaiser, 2004; Shiner, Masten, & Roberts, 2003). Donnellan and his colleagues (2009) also demonstrated that traits such as neuroticism, conscientiousness, and agreeableness measured during adolescence predicted to the quality of romantic relationships and parenting practices during adulthood even after controlling for SES in the family of origin.

Simply put, there is some evidence from research during the past decade that supports the social selection assumption that individual characteristics from childhood and adolescence affect later SES and family relationships just as there is evidence consistent with the social causation assumption that SES plays a role in family dynamics and the life course development of family members. Conger and Donnellan (2007) proposed that a more complete understanding of the relationship between SES and family life could be obtained by combining the social causation and social selection approaches into what they termed an interactionist model. We now turn our attention to this conceptual framework and then consider important theoretical and research priorities for the future.

The Interactionist Perspective

Given the evidence just reviewed, a comprehensive model that assumes that processes of both social causation and social selection will operate in producing an association between family dynamics and SES may hold the most theoretical promise. Research during the past decade suggests the utility of such an approach. For example, using data from two national birth cohort studies in Britain, Schoon et al. (2002) showed that low SES in a child’s family of origin predicted lower academic achievement and continuing life stress across the years of childhood and adolescence, consistent with the idea of social causation. Children’s lower academic competence and higher life stress, in turn, were associated with lower SES when the children reached their adult years, consistent with the idea of social selection. Similarly, Wickrama and his colleagues (2008) found that low SES in the family of origin increased risk for both mental and physical health problems during the transition to adulthood which, in turn, predicted economic problems and poorer social circumstances during the early adult years. Very similar results were reported by Haas (2006). Thus, consistent with the interactionist perspective, these studies suggest a reciprocal process in which early SES predicts personal characteristics of children that influence their SES in adulthood.

Building on these ideas and findings, Figure 3 depicts our interactionist model of SES, family interaction processes, and child development. The model, which extends an earlier perspective proposed by Conger and Donnellan (2007), systematically incorporates social selection and social causation processes into an overarching framework. Consistent with the social causation view, the model proposes that the exogenous variable, the SES of the first generation parents (G1), will have both a direct impact on the second generation child’s (G2) traits and dispositions during the first two decades of life as well as an indirect influence through the types of family dynamics discussed earlier in relation to the FSM and IM. G1 family dynamics include family stress processes like economic pressure, marital conflict, and parenting practices as well as specific parental investments as proposed by the IM. Also important, the model proposes intergenerational continuity from G1 to G2 SES and from G1 to G2 family dynamics. Consistent with the social selection approach, the model hypothesizes that G2 traits and dispositions will predict G2’s SES, family dynamics, and the adjustment of G2’s children (G3) during G2’s adulthood. Notice, however, that G2 SES and family dynamics also are expected to affect the G3 child, consistent with the FSM and IM and with the assumption of social causation.

Figure 3.

The Interactionist Model.

Two recent reports have found preliminary evidence consistent with predictions from the interactionist model. Using a community sample of over 200 adolescents who had been studied from early adolescence to young adulthood and who had become parents, Schofield and his colleagues (in press) demonstrated that G1 SES and warmth and support to G2 during G2’s adolescence were positively related to G2 personality characteristics involving low neuroticism, high sociability, and a conscientious, persevering work ethic. These characteristics, in turn, predicted greater G2 SES during adulthood and greater caring and concern for children. They also predicted less economic pressure, G2 emotional distress and interparental conflict. Also consistent with the model, these family interaction processes in G2’s family of procreation predicted G3 positive adjustment (i.e., secure attachment, vocabulary skills, academic competence and prosocial behavior in the family) in the expected directions and in most instances completely accounted for the relationship between G2 personality and G3 adjustment. The findings also showed that G1 SES directly predicted G2 SES and G1 family processes predicted G2 family processes, consistent with the model in Figure 3.

Using data from the same longitudinal study, Martin and her colleagues (in press) took a different approach to evaluating the interactionist model. Instead of examining positive traits of G2, they focused on G2 problem behavior during adolescence. Their findings showed that SES in the first generation was negatively related to the antisocial behavior of G2 during adolescence. They also demonstrated that G2 problem behaviors were negatively associated with later G2 SES and investments in children but positively related with family stress processes. These social characteristics of G2 families during adulthood predicted in the expected directions to G3 antisocial behavior. Thus, both the Martin and Schofield studies provide preliminary support for the interactionist perspective. That is, personal traits prior to adulthood predict to SES and family processes during the adult years, consistent with the assumption of social selection, and these social and economic characteristics predict the adjustment of the next generation of children, consistent with the assumption of social causation.

Concluding Comments and Future Directions

This review demonstrates that the study of SES and family life has advanced both theoretically and empirically during the past decade. Theoretical developments have moved beyond earlier assumptions about the singular direction of effects to a new perspective regarding the interplay among individual differences, SES, and family relationships. In addition, empirical studies during the past 10 years have provided important evidence supportive of these theoretical advances. Given this progress, the question for the moment concerns the theoretical and research priorities that will similarly advance this work during the coming decade. The following comments draw from the literature just reviewed to propose several promising directions for the field over the next 10 years. These recommendations relate to: (a) the quantitative assessment of SES and its influence on families and family members, (b) needed improvements in research methods, (c) the elaboration and extension of current theoretical frameworks, and (d) the introduction of genetic information into the study of SES and family life.

Recommendation #1: The Quantitative Assessment of SES

The research reviewed here demonstrates that, during the past decade, most empirical and theoretical work in this area has focused on the economic aspects of SES. The two other usual quantitative markers of SES, education and occupational status, have largely been ignored, used as control variables, or combined with income to construct an overall index of SES. During the next decade, it will be important to disaggregate these indicators of SES and examine their unique and combined roles in relation to family life. As noted earlier, education is a primary determinant of both income and occupational status; thus, rather than being static indicators of a single construct, there is a dynamic relationship among these variables. It will be important for researchers to consider, for example, the influence of educational attainment on family income and on occupational achievement. In addition to acting as a driver of these other SES variables, education may act as an important personal resource that buffers against the potentially damaging impacts of reductions in income during downturns in the economy. In future examinations of specific theoretical models like the FSM, IM and interactionist model, it will be important to isolate both the unique influences of education, income, and occupational status in the hypothesized causal processes as well as specific mediators that likely explain their effects (see Conger & Dogan, 2007; Duncan & Magnuson, 2003).

Recommendation #2: Improvements in Research Methodology

The single most glaring methodological flaw in the research reviewed here involves the use of cross-sectional data to evaluate models proposing causal relationships among theoretical constructs. Although there has been significant replication of predictions from the FSM and IM using cross-sectional studies, this methodology can only lead to very weak inferences regarding hypothesized causal relationships. During the next decade it will be important to conduct tests of these models using data collected over time. In that way it will be possible, for example, to determine if changes in income lead to changes in the adjustment of parents and children and in the quality of their family relationships. This approach will also allow consideration of feedback loops such as the possibility that the emotional distress of parents will reduce their capacity to cope successfully with economic hardship, thus increasing family financial problems across time (see Conger & Conger, 2002). Indeed, the current downturn in the economy creates a natural experiment for examining how families are influenced by both increasing economic problems and, hopefully, the return of better economic fortunes. In addition to these types of passive longitudinal research designs, investigators should take advantage of changes in local economies that create opportunities for quasi-experiments of the type conducted by Costello and her colleagues (2003) when a new casino opened in a community. There should also be more attention paid to theoretically driven experimental interventions like the New Hope project (see Huston et al., 2005). Other methodological improvements relate to better measurement of family characteristics using multiple family informants and trained observers of family interactions (e.g., Lorenz, Melby, Conger, & Xu, 2007; Mikelson, 2008;). Also important will be additional studies that evaluate SES and family functioning for different ethnic, racial, and national groups.

Recommendation #3: Elaboration and Extension of Current Theoretical Frameworks

This review found evidence consistent with hypotheses generated by the family stress (FSM), investment (IM), and interactionist models. Important next steps related to these models include: (a) evaluation of their relative contributions to the relationship between SES and family life, (b) elaborations that add potentially important new mediating variables, and (c) consideration of personal, social, and cultural factors that might modify (moderate) the effects of proposed causal pathways in the models.

Joint tests of explanatory models

With regard to the first issue, almost no research has been done that specifically compares predictions from the FSM and the IM. Some studies have compared limited aspects of these models in empirical tests, but large elements of the FSM (e.g., interparental conflicts) have been omitted from these evaluations of the models (e.g., Gershoff et al., 2007; Linver et al., 2002; Yeung et al., 2002). During the next few years, it will be important to conduct combined tests of the full FSM and IM to see whether they are truly competitive explanatory frameworks or simply complementary aspects of a single, comprehensive model. In addition, joint tests of the models are crucial for clarifying whether family stress processes provide a better explanation of child and adolescent social emotional development whereas parental investments included in the IM provide a better explanation of cognitive development as suggested by some of the studies reviewed earlier in this report (e.g., Gershoff et al., 2007). Finally, combined model tests also need to incorporate information about personal characteristics of parents to determine whether proposed selection effects in the interactionist model alter the causal connections proposed by the FSM and the IM.

New mediating pathways

The work reviewed earlier by Mistry and her colleagues (2008, 2009) suggested new explanatory pathways in the stress process proposed by the FSM. For example, they found that adolescent as well as parent experiences of economic pressure may have an adverse affect on child development. The report by Gutman et al. (2005) also suggested an additional mediating pathway from income to neighborhood stress to family stress processes and adolescent adjustment. We suspect that new research that looks beyond the family and targets potential mediators involving peers, schools, and community characteristics may add to the explanatory power of the FSM, IM and interactionist models. For example, both current family economic circumstances (social causation) and personal traits of parents (social selection) should affect the characteristics of the peers, schools, and neighborhoods to which parents and children are exposed and, in turn, the psychological well-being of family members.

Tests for moderation

Research reviewed earlier suggested several different factors that may moderate predictions from the FSM, IM and interactionist models. During the next decade, new research is needed that will extend this earlier work. Especially important will be new studies to determine if predictions from these models vary by child age. For example, because research on the IM has primarily involved younger children and research on the FSM has primarily involved older children and adolescents, it could be that family stress processes have been shown to be most salient for social and emotional problems because the greatest risk for these difficulties occur during adolescence whereas the IM predicts best to the cognitive development of young children because cognitive development and preparation for entry into school are major tasks during early childhood. Additional studies that test the FSM with young children and the IM with older children and adolescents will help determine whether there are real age differences in the developmental influences proposed by these two theoretical frameworks. Also important will be future research that considers factors that might buffer the adverse effects of SES-related family stress; such as social support (e.g., Leinonen et al., 2003), personal dispositions that promote effective adult role performance (e.g., Schofield et al., in press), or cultural differences that may modify the way these models work with different ethnic groups (e.g., Parke et al., 2004).

Recommendation #4: Consideration of Genetic Variability

There is growing evidence that environmental stress joins with certain genetic characteristics to affect individual adjustment (e.g., Guo, Roettger, & Cai, 2008; Moffitt, Caspi, & Rutter, 2006). For example, variations in genotypes related to the processing of serotonin and dopamine can either increase or decrease risk for psychopathology in response to stresses involving interpersonal relationships, financial difficulties or personal adversities of the types incorporated in the FSM (Moffitt et al., 2006). During the next 10 years, these promising lines of inquiry should be added to empirical and theoretical developments related to SES and family life. We believe that attention to these four recommendations will help improve understanding of the association between SES and family functioning and in that process help promote better policies and programs for the welfare of families and family members.