Introduction

Economic instability is a salient issue for families because it is directly and indirectly related to both family processes and child development (Solantaus et al. 2004). Living in poor socioeconomic conditions may shape parents’ abilities to promote their children’s development and guide their children in becoming effective members of society (Lansford 2012). In addition, economic pressures are related to a wide range of family challenges including marital problems and parental depression (Robila and Krishnakumar 2005). Conger and colleagues’ family stress model proposes that economic difficulties are indirectly related to child outcomes through parents’ emotional states, marital conflict, and parenting (Conger et al. 1990, 1994, 2002, 2010). This model has been adapted and tested with diverse samples in the United States (Conger et al. 1994, 2002; Parke et al. 2004), as well as in other countries (Forkel and Silbereisen 2001; Hraba et al. 2000; Robila and Krishnakumar 2005, 2006; Solantaus et al. 2004). A few studies have examined Romanian mothers and their adolescent children (Robila and Krishnakumar 2005, 2006); however, research testing Conger’s model using reports of mothers and fathers and younger Romanian children is lacking.

At the outset, it is important to understand the potential stressors facing families in Romania. Romania is an eastern European country, controlled by communists until 1989. In 2000 the average marriage age in Romania was 23.6 years for women and 26.9 years for men. Age at marriage in Romania is the lowest in Europe, and similar rates are seen in the Ukraine, Moldova, Belarus, and Russia (Mihai and Butiu 2012). Cohabitation and divorce rates are low compared to other European countries (Robila 2004). Although Romanian fertility rates are comparable to other countries in Europe (Eurostat 2012a), in Romania, parents with lower educational levels and those living in rural areas tend to have more children (Robila 2004).

During the rule of communism in Romania, the regime degraded the economic, social, and moral life of the people (Calafeteanu 2006). Romania has since worked toward Western ideals of democracy, capitalism, and personal freedoms, but the transition of political and economic systems has often been erratic and chaotic. Since the fall of communism, Romania has experienced rising unemployment and high inflation (Robila 2004). In 2000, 41 % of the Romanian population felt they did not have enough to afford bare necessities, and 39 % felt they had only enough for bare necessities (Zamfir et al. 2001). In 2003, 75.2 % of the population reported they had only enough income to make a minimum living (Robila 2004).

As a result of inflation and unemployment, many Romanian families live under conditions of economic uncertainty and poverty. Most families have relatively few economic reserves and those they have accumulated are not sufficient to meet their current or future economic needs (Robila 2004). Furthermore, economic difficulties and the accompanying stress have not hit all areas of Romania equally. Almost half the population (44.9 %) live in rural areas of Romania (Herman 2012b), and the rural poverty rate is over 70 %, the highest rate in the European Union. This has resulted in one of the largest gaps between urban and rural areas in living and social standards (The World Bank 2012). Many rural jobs do not provide enough income to sustain livelihoods, and rural areas lag behind urban areas in labor productivity, economic performance, and education levels (Herman 2012a).

Consequently, Romania is a unique context in which to examine the effects of economic hardship on individuals and families because of the extreme circumstances of the country’s past, as well as the economic challenges families have experienced in recent years (Robila 2002). Economic hardship can impact many aspects of both parents’ and children’s family life; thus, it is particularly important to examine how economic hardship, as related to family functioning, might be related to a child’s behavior in the peer group.

One type of peer group behavior is social withdrawal. Social withdrawal has been defined as the consistent (across situations and over time) display of all forms of solitary behavior when encountering familiar and/or unfamiliar peers (Rubin et al. 2002a). Within the broad construct of social withdrawal, researchers have identified different forms of solitary behavior with each one carrying different motivations, origins, correlates, and outcomes with some being more indicative of risk than others. Specifically, during the preschool years (4–5 years of age), reticence appears to be a form of observed withdrawal that appears to place children at risk.

Reticence is a subtype of social withdrawal (Coplan and Rubin 2010) characterized by frequent observed displays of unoccupied behaviors and on-looking when a child is among a group of peers (Coplan et al. 1994). Socially reticent children appear to want peer interaction, but find that entering social situations results in anxiety and a need to avoid interaction (Hane et al. 2008). During early childhood, this form of withdrawal has been found to be associated with anxious-fearful and hovering behaviors, peer rejection, negative emotion regulation, low self-perceptions, and internalizing disorders (e.g., Coplan and Rubin 1998; Hart et al. 1993, 2000; Nelson et al. 2009; Rubin et al. 1995). Reticence appears to place children at risk of difficulties in the peer group as well as internalizing problems. Consequently, it is important to examine the ways in which economic hardship might work with family functioning to impact children’s behaviors in peer settings, in particular, reticence, because of the potential risk it presents for healthy development.

Economic hardship and family functioning also might be related to children’s shy, reticent behaviors because, conceptually, economic hardship most likely would foster conditions in the home that would particularly be problematic for the development of reticent behavior. For example, numerous studies have shown that the development of reticent behaviors is more likely to occur in the presence of parental over-control (e.g., Rubin et al. 1999, 2002b). Hence, factors such as economic hardship that promote the parental use of control might, in turn, be related to the development of shy, reticent behaviors in children. Indeed, Rubin et al. (2003) have theorized that parental dispositional characteristics and family relationships might contribute to the development of children’s withdrawn behaviors. They hypothesize that parental feelings of helplessness or frustration due to lack of financial resources may result in less than optimal childrearing. Marital discord or dissatisfaction can also impact child-rearing practices. Consequently, stressful personal or economic circumstances, marital conflict, and/or overcontrolling or over-involved parenting all may contribute to a child’s shy, reticent behaviors.

This theorizing regarding possible pathways to child reticence fits in nicely with Conger’s model in that previous research based on the family stress model provides considerable evidence that economic stress can be detrimental to families and to child outcomes (Conger et al. 1994, 2002; Parke et al. 2004). Economic hardship (low per capita income and lack of employment) affects family functioning and individual wellbeing indirectly through every day family economic pressures (i.e., parents’ perceptions of their inability to pay bills and being unable to make end meets). Because economic pressures reflect parents’ assessments of their economic circumstances (Hraba et al. 2000), these pressures “give meaning” to objective economic hardship (Conger et al. 2010, p. 690). Additionally, when parents are less educated, they may be less able to secure adequate employment, which also may contribute to increased economic pressure.

Economic pressure, in turn, can lead to parental depression (Conger et al. 2002; Hraba et al. 2000) and marital conflict (McLoyd 1998). Conger et al. (2002) posit that economic pressures promote negative parental emotions such as anxiety, depression, anger, and frustration. Parents bear the consequences of insecure economic circumstances and if their adaptive capacities are challenged, they may become depressed (Solantaus et al. 2004). Economic pressure can also contribute to marital conflict, including angry interactions, aggressive responses of marital partners, insensitivity, defensiveness, and criticism (Conger et al. 2002).

Economic pressure also may be indirectly related with marital conflict. In Czech families (Hraba et al. 2000), economic pressure has been shown to be indirectly associated with marital hostility through maternal depression. Similarly, economic pressure was indirectly related with marital interactions through parental mental health in a study of Finnish families (Solantaus et al. 2004). Research with mothers of adolescents in Romania (Robila and Krishnakumar 2005) documents a direct relationship between economic pressure and maternal depression as well as an indirect relationship between economic pressure and marital conflict through maternal depression.

Economic pressure also might be indirectly related to psychologically controlling parenting through parental depression or marital conflict. Conger and Donnellan (2007) suggest that parental depression associated with perceptions of economic pressure diminishes the quality of parenting. However, some studies (e.g., Conger et al. 2002; Cummings et al. 2005) have not identified a direct link between depression and parenting once marital conflict is taken into account. Thus, it is also possible that marital conflict may indirectly link the effects of depression on parenting. It may operate in this way. It may be more likely that depressed parents are less caring, supportive, and affectionate, and more impatient, irritable, and hostile towards their spouse, which might contribute to marital conflict. Then depression and marital conflict may spill over into how they parent their children (Solantaus et al. 2004). Indeed, it is possible that if parents are tense, tired, preoccupied, and anxious because of marital conflict and/or depression driven by their financial situation, they may engage in fewer parenting behaviors requiring more energy, and consequently, use psychological manipulation as a way to interact with their child (Stone et al. 2002).

Thus, it is possible that both marital conflict and parental depression may be indirectly related with child reticence through psychological control. As noted previously, there are a number of studies that point to control as playing a particularly problematic role in the development of anxious and reticent behaviors. For example, longitudinal research with children ages 2–8 years in Canada has shown that a number of factors including maternal depression and overprotective parenting were related to higher levels of child anxiety. Results indicated that when maternal overprotection was high, child anxiety increased (Laurin et al. 2015). An example of the link between parental over-control and reticence, specifically, can be seen in a study that found that the association between inhibition at age 2 and reticence at age four was only significant for those children whose mothers at age two displayed high levels of intrusive control and/or derisive comments (Rubin et al. 2002a, b). Finally, although not identifying anxious, reticent behaviors specifically, research with mothers and adolescents in Romania reported that maternal depression was indirectly related with adolescent internalizing problems through psychological control (Robila and Krishnakumar 2006). Taken together, there is evidence that parental depression and marital conflict would be indirectly related to child reticence through parental psychological control. Research also shows that higher levels of marital conflict are directly associated with higher levels of psychologically controlling parenting of younger children (Cummings et al. 2005) and adolescents (Doyle and Markiewicz 2005; Stone et al. 2002).

Psychologically controlling parenting is of particular concern because of its association with child internalizing problems (Barber 1996; Mills and Rubin 1998; Olsen et al. 2002), which are closely related to reticent behaviors in young children (Coplan and Rubin 1998). Because psychologically controlling parents often manipulate and control their children by invalidating feelings, using love withdrawal, and by constraining verbal expression (Barber 1996; Barber and Harmon 2002), this excessive control can disrupt the development of children’s personal autonomy (Mills and Rubin 1998) and contribute to both social withdrawal and internalizing problems (e.g., Barber 1996; Rubin et al. 1998).

In the present study we replicate previous research on the family stress model performed in the United States and other countries and extend past work in Romania by investigating the relationship between economic stress and child reticence, specifically examining the mediating roles of parental depression, marital conflict, and psychological control. Based on our review of the literature and Conger’s family stress model we hypothesized that (1) factors associated with economic hardship (per capita income, parental employment) and parental education would be directly related to economic pressure, (2) economic pressure would be directly related to parental depression, indirectly related to psychological control, and both directly and indirectly related to marital conflict, (3) parental depression would be directly related to marital conflict, indirectly related to child reticence, and both directly and indirectly related to psychological control, (4) parental marital conflict would be directly related to parental psychological control and indirectly related to child reticence, and (5) parental psychological control would predict child reticence.

Method

Participants

Participating families were all ethnic Romanian, two-parent families. Most families (94.30 %) had an annual income of less than US $7000, somewhat lower than the annual income of EUR 5891 (US = $7692.83) reported by Eurostat (2012b). Five families earned between US $7000 and US $15,000, and only one family earned more than US $15,001. The mean number of children in each family was 1.83 (SD = .99). The sample included 61 male children (50.4 %), and the average age of children in the sample was 4.83 years (SD = .60).

Average age of mothers was 31.39 years (SD = 4.54) and the mean age of fathers was 33.57 years (SD = 4.86). Mothers averaged 11.79 (SD = 3.10) years of education, and the mean years of education for fathers was 11.68 years (SD = 3.02). Mothers worked on average 43.76 h per week (SD = 12.61; median = 40.00), with 28.9 % reporting they were unemployed. Fathers worked on average 51.07 h per week (SD = 15.73; median = 48.00), with 24.4 % reporting they were unemployed.

Procedure

After receiving approval from a university’s institutional review board, a school-based study was conducted in 2006 in Romania. Families of children from an urban (N = 78) kindergarten in Iasi, Romania participated, as well as families with children in a rural (N = 46) kindergarten located in Cosecosteşti, Romania, a small village approximately 60 km north of Iasi. Seventy-five (96 %) of the urban families and all 46 rural families who were approached about participating in the study returned completed packets (total sample = 121 families; response rate = 97.6 %). The researcher gave each kindergarten teacher packets for both parents containing questionnaires, as well as consent forms which communicated to parents that they had the right to not participate or withdraw from the study without any jeopardy to their family or their child’s standing at school. After consenting to participate in the study, mothers and fathers independently completed self-report questionnaires in their homes. Teachers read the questions to rural parents who had difficulty reading. Consent forms and questionnaires regarding child social skills for each student were also completed by teachers. Families received 10 RON (approximately US $4.00) for participating, and teachers received 3 RON (approximately US $1.20) for each child questionnaire. Average gross monthly public wage for individuals living in Romania at the time of the study was approximately 1400 RON per month or US $355.00 (International Monetary Fund 2012).

Measures

Measures were translated from English to Romanian by experts fluent in both languages. Using translation-back translation, all instruments were translated from English to Romanian and back translated to English to check for changes in meaning. The researcher was consulted on items that were difficult to translate. Back translations were comparable to English instruments.

Economic Hardship

Economic hardship was assessed with two variables, per capita income and parent employment. Mothers indicated which of six categories corresponded to their family monthly net income (Range: under US $7000 to over US $50,000). Categories were replaced by their mean and divided by the number of family members living in the household; 12 % of the data for this variable was missing. The parent employment variable was coded from 0 to 2 (0 = no parents working full or part-time; to 2 = both parents working full or part-time).

Parent Education

Mothers reported the number of years of education completed by both mothers and fathers.

Economic Pressure

Each parent assessed economic pressure using two items adapted from Conger et al. (1994). Spouses reported on his or her perception of how much difficulty they have paying bills each month (1 = No difficulty, 5 = A great deal of difficulty) and whether they have money left over at the end of the month (1 = More than enough, 4 = Not enough to make ends meet).

Depression

Parental depression was examined with the Center for Epidemiological Studies Depression Scale (Radloff 1977), a 20-item self-report scale designed to measure depressive symptoms in the general population. Mothers and fathers independently assessed how frequently they experienced depressive symptoms (e.g., “felt sad;” “restless”) over the past week (1 = Rarely or none of the time, 4 = Most or all of the time). Items were summed and a mean score calculated. Scores of 36 or higher were indicative of potentially serious levels of depression (Radloff 1977); 57.9 % of mothers and 55.4 % of fathers had scores higher than 36. Psychometric properties have been well established, including test–retest reliability (Radloff 1977). Adequate validity has been demonstrated in Romanian samples (Vrasti et al. 1986, as cited in Robila and Krishnakumar 2006). Internal consistency reliabilities with this sample were acceptable for both mothers (α = .90) and fathers (α = .91).

Marital Conflict

Parents reported on marital conflict with an adaptation of the O’Leary–Porter Scale (Porter and O’Leary 1980). Husbands and wives independently answered 10 questions about how often various forms of marital hostility (e.g., quarrels, sarcasm, physical abuse) were observed by their children (1 = Never, 5 = Very Often). This scale has demonstrated adequate validity with US populations (Cummings et al. 2005). Cronbach’s alpha for this sample was .93 for mothers and .94 for fathers.

Psychological Control

Mothers and fathers independently reported on psychologically controlling parenting using items developed by Barber (1996) and adapted for use with preschool-aged children (Olsen et al. 2002). Eight items assessed how often parents exhibit certain behaviors with their child (1 = Never, 5 = Always). The items represent dimensions of psychological control including invalidating feelings, love withdrawal, personal attack, and constraining verbal expression. Cronbach’s alpha for the scale was .57 for mothers, and .60 for fathers.

Child Reticence

Teachers assessed child reticence using the Social Skills Constructs for Preschool and Kindergarten Teachers (Hart et al. 2000). Teachers rated their subjective impressions of the frequency of reticent withdrawn behavior displayed by kindergarten children using three response options (0 = Never to 2 = Often). Eight items measured aspects of children’s reticent behavior, including “Is off task and preoccupied” and “Is very shy.” Teachers of preschoolers in US samples have used these measures, demonstrating adequate validity as well as good test–retest reliabilities (Hart et al. 2000). The Cronbach alpha coefficient for this scale was .65.

Data Analyses

T-tests and Chi-square analyses were first conducted to identify if differences existed for parents living in rural and urban locations. Next, means, standard deviations, and correlations between all variables were calculated. A confirmatory factor analysis was used to determine factor loadings for indicators on each of the latent variables in the model. Then unstandardized and standardized beta coefficients were calculated to determine the strength of the direct and indirect paths in the structural model. Indirect effects were also calculated.

Results

The sample was drawn from both rural and urban locations, so we performed preliminary analyses to determine if there were significant differences between individuals and families in the two locations (Table 1). Chi-square tests showed there were no significant differences in the number of male children living in rural and urban locations. Significantly more mothers and fathers living in rural locations stated they had lower incomes and were unemployed.

Table 1 Differences between urban and rural locations
Full size table

T-tests indicated no significant differences by location in children’s ages and in the average ages and number of hours worked by fathers and mothers. Both mothers and fathers from rural locations had significantly fewer years of education than those from urban areas. Fathers and mothers living in rural locations also scored significantly higher on being unable to pay bills, being unable to make ends meet, and marital conflict.

T-tests were also calculated to identify if mother and father ratings differed on study variables. T-tests showed that mother ratings of depression, marital conflict, and psychological control were not significantly different than father ratings.

Table 2 presents correlations for the study variables. Correlation analyses revealed that living in a rural location (urban = 0, rural = 1) was significantly associated in the expected directions with all the economic hardship and economic pressure variables, as well as with marital conflict. Correlations between economic hardship and pressure, depression, marital conflict, psychological control, and reticence were in the expected directions.

Table 2 Correlations among study variables
Full size table

We estimated a structural equation model using the Mplus software program, Version 7 (Muthén and Muthén 2010) using full information maximum likelihood estimation to address missing data (Enders 2010). Standardized and unstandardized beta coefficients were calculated to determine the strength of the relationships between variables in the model. We first estimated a measurement model to examine factor loadings for each construct in the model. Because of high multi-collinearity for mothers’ and fathers’ ratings of depression (r = .80) and marital conflict (r = .89), we created latent variables called parental depression and parental marital conflict and used fathers’ and mothers’ ratings as indicators. We also created a latent variable for parental psychological control. We used the indicators, difficulty paying bills and difficulty making ends meet, to create a latent variable for economic pressure. We averaged mother and father reports to create these two indicators because mother and father reports were highly correlated (bills: r = .87; ends meet: r = .75). Thus, economic pressure, parental depression, parental marital conflict, and parental psychological control were examined as latent variables in the measurement model, with each of these showing acceptable factor structures (all factor loadings in the measurement model were above .60). For both the measurement and structural models to be considered a good fit to the data, χ2 values should be nonsignificant; the comparative fit index (CFI) should be above .95; and the root mean square error of approximation (RMSEA) should be less than .08 (Kline 2010). The measurement model fit indices indicated adequate fit of the model to the data: χ2 (10, N = 121) = 16.246, p = .09; CFI = .987; RMSEA = .072.

We next estimated the structural model. Because there were a number of significant differences by location (urban = 0; rural = 1), we controlled for location when estimating the structural model. In the structural model, parent employment and father and mother education were allowed to covary. Our major interest was to replicate Conger’s family stress model which hypothesizes that economic pressure is indirectly related to child outcomes through parental depression, marital conflict, and psychological control. The structural model estimated indirect paths from economic pressure to the outcome variable, child reticence.

The estimated model is shown in Fig. 1. Estimated paths in the model that were nonsignificant are dashed. The control variable and covariances are not shown in the model. The χ2 was not statistically significant [χ2 (59, N = 121) = 69.016, p = .175], indicating adequate fit. Other model fit statistics also indicate acceptable fit with the data (CFI = .989; RMSEA = .037).

Fig. 1
figure 1

Structural model (standardized coefficients in parentheses); a p < .001; b p < .01; c p < .05; nonsignificant paths dashed; χ2 (59, N = 121) = 69.016, p = .175; CFI = .989; RMSEA = .037. The covariances and the control variable, location (urban = 0; rural = 1), not shown

Full size image

In partial support of Hypothesis 1, after controlling for location, higher levels of mother education were associated with lower economic pressure. In partial support of Hypothesis 2, economic pressure was significantly and directly related to parental depression, but was not directly related to marital conflict. Economic pressure was indirectly related to marital conflict through depression (standardized indirect effect = .15; p < .05). In partial support of Hypothesis 3, parental depression was directly related to marital conflict. It was not related to psychological control, nor was it indirectly related to child reticence. Parental depression was indirectly related to psychological control through marital conflict (standardized indirect effect = .18; p < .01). In support of Hypothesis 4, marital conflict was directly associated with psychological control and indirectly related with child reticence through psychological control (standardized indirect effect = .08, p < .01). In support of Hypothesis 5, psychological control predicted child reticence.

Discussion

Preliminary analyses showed that a number of differences were identified for families living in rural and urban locations. Both parents in families living in rural locations had lower levels of education, and significantly fewer rural mothers and fathers were employed full or part-time. Furthermore, rural families had lower per capita incomes, were less able to make ends meet, and pay their bills. Results support studies indicating that compared to Romanian families living in urban areas, families living in rural locations experience higher levels of poverty and are more likely to be under-employed (Herman 2012a).

Parents living in rural areas also experienced higher levels of marital conflict than those in urban areas. In addition to higher levels of economic pressure and lower levels of per capita income in rural areas which might contribute to marital conflict, over half of the fathers and almost two-thirds of the mothers in rural areas reported being unemployed. It could be that rural parents spend more time together because they are not away from the home working, resulting in more contact with each other and more opportunities for conflictual or strained interactions (Walper and Silbereisen 1994). In addition, rural mothers and fathers had completed less education than urban parents. Past research has documented the relationship between education levels and marital satisfaction and marital problems (Conger et al. 2010). Parents with less education may have fewer financial resources, less lucrative employment, and a less well-developed repertoire of conflict resolution skills, all of which might also contribute to higher levels of marital conflict.

Similar to results found in research in the Czeck Republic (Hraba et al. 2000), the family stress process appeared not to begin with objective economic hardship conditions such as parent employment or per capita income, but with the family’s assessment of economic pressure. Although mothers’ educational level was related to economic pressure, parental employment and income per capita were not. Almost all families in our sample (94.3 % of reporting families) had yearly incomes less than US $7000, resulting in little variability. Consequently, it may be that our measures of per capita income were not sufficiently sensitive; however, it is also likely that the psychological stress experienced by parents due to their appraisals of their uncertain economic conditions may have been more salient to parents than their objective economic conditions (Hraba et al. 2000).

Other pathways in the model generally support research hypotheses as well as previous tests of the family stress model (Conger et al. 1994, 2002). Similar to past research assessing mothers of Romanian adolescents (Robila and Krishnakumar 2005), economic pressure was directly related to parent depression and indirectly related to marital conflict through parental depression. Apparently, Romanian parents’ subjective assessments of economic pressure contribute to higher levels of parental depression, which then spills over, influencing their levels of marital conflict.

Although past research with Romanian mothers and adolescents (Robila and Krishnakumar 2006) demonstrated that higher levels of maternal depression were directly related to higher levels of psychological control, we found that parental depression was indirectly related to parental psychological control through marital conflict. Similarly, other research that examined other parenting variables (Conger et al. 2002; Solantaus et al. 2004) has shown that depression, as well as the inability of families to pay bills and make ends meet, contribute to higher levels of conflict between spouses, which also spills over into less than optimal parenting. As Stone et al. (2002) have suggested, it is possible that engaging in marital conflict may result in parents becoming more tense and tired, which could lead to the use of parenting strategies that require less energy and vigilance. Parents may turn to psychological control as a way to get their children to comply because it requires less energy than consistent discipline, inductive reasoning, or monitoring.

We also found that psychological control predicted child reticence. The link between psychological control and child and adolescent internalizing behaviors is well-established in past research in the United States (Barber 2002), and in research with mothers and Romanian adolescents (Robila and Krishnakumar 2006). The findings from the current study, however, are particularly informative because they begin to elucidate the ways in which economic hardship may be related to children’s specific maladaptive behaviors. In particular, it appears that as family factors work to increase the use of psychological control, children tend to exhibit more reticent behaviors among peers. Given the links between reticent behaviors and peer rejection and problems of an internalizing nature (e.g., Coplan and Rubin 1998; Hart et al. 1993, 2000; Nelson et al. 2009; Rubin et al. 1995), it is disconcerting to see the ways in which economic hardship might be related to children’s maladaptive behaviors. In particular, this work adds to the growing body of literature that control, in its many forms including psychological control, place children at risk for the development of shy, reticent behavior in peer settings. Thus, it is important to identify factors that might promote parental use of control, meaning that the results of this study make additional significant contributions by identifying the ways in which economic hardship might impact depression and marital conflict in ways that lead some parents to use psychological control to the detriment of their children. Taken together, these results provide another source of support for the family stress model with Romanian families of young children.

Our research is limited in some respects. It illustrates a series of family processes without considering reciprocity, which likely exists. For instance, depression may affect the way economic pressure is perceived, and child reticence may influence psychological control. In addition, the family system functions with many other unmeasured influences such as social support, which has been identified as an important mediating influence in families of Romanian adolescents (Robila and Krishnakumar 2005). Furthermore, this sample included family groups in Romania who lived in both urban and rural communities, but the small samples may not be representative of all Romanians (Stevenson-Hinde 1998). Finally, rural parents had difficulty reading, so a teacher read the survey questions to them. This could affect the way some parents responded, particularly in under-reporting negative behaviors.

Future research could expand on these findings by including a larger number of rural and urban families that would allow group comparisons of the patterns of relationships among variables. Longitudinal research would aid in better understanding directional effects, as well as family processes over time in Romania. Qualitative studies could assist researchers in better understanding the meanings family members bring to economic hardship and how this impacts family processes. Research could also investigate the cultural beliefs and values that influence the parenting of Romanian mothers and fathers, as well as the microenvironments and developmental niches of Romanian children (Harkness and Super 2002). Likewise, a more nuanced investigation of the socio-demographic environment (not just urban and rural location), focusing on household size, number of children, levels of formal education, and age at first birth, could help researchers to better understand the cultural milieu of Romanian families (Keller 2012). Finally, research could continue investigating how rural and urban families in Romania cope with the economic pressure they experience (Forkel and Silbereisen 2001), focusing on interventions targeting coping with economic hardship, family processes, and children’s emotional and social health.

Future intervention work should continue investigating how families in Romania manage the extreme economic hardship they experience, focusing particularly on interventions targeting family processes and children’s emotional and social health. This is especially appropriate considering Romania’s admission to the European Union and resulting increases in economic pressure. Findings from this study and others (e.g., Cummings et al. 2005; Leinonen et al. 2002) suggest that targeting marital relationship quality could be a beneficial starting point for helping family processes become healthier, and in turn foster the development of socially healthy behaviors in young children—especially in the face of economic hardship. It is also important to continue investigating the contributions fathers make in the family system.

Policy makers interested in promoting positive family processes and enhancing children’s socio-emotional outcomes would do well to consider the recommendations of Anghelescu and Lliescu (2007). These researchers, after coordinating a study to examine a representative sample of parents of children 8 years old and younger in Romania, proposed a number of recommendations for public policy in Romania at both the national and local levels. At the national level they recommended better collaboration and correlation between departments and ministries involved in child protection and education, as well as the creation of a national strategy in the field of parenting education. To support families living in rural (but also urban) locations, researchers suggested (a) involving local government leaders in promoting parenting education programs; (b) providing funding at local and national levels to publish educational materials to promote positive parenting practices, healthy marital relationships, and positive social skills in children; (c) providing funding for services to support families in challenging situations: financial crises, diminished parental well-being, marital conflict, or disrupted parenting; (d) establishing resource centers to support parents and young families; (e) and establishing support groups, workshops, and informal meetings to help parents develop parenting competencies (Anghelescu and Lliescu 2007). Clearly, these are ambitious recommendations for public policy, which if implemented could serve to enhance the future well-being of Romanian parents, children, and families.