Parental Involvement and School Choice in Hungarian Primary Schools

ABSTRACT

In this study, we examine the features of school choice by comparing parental involvement between denominational and public schools. In Hungary, denominational schools reemerged after the fall of communism in 1989, and their share increased further after 2010. The survey data, we employ, refer to primary school children’s parents and were focused on parental involvement at home and at school. Our hypothesis is based on the work of Epstein and Coleman and suggests a prominent role for denominational schools in parental involvement. However, our results show that Hungarian denominational schools differ in terms of parental involvement and school choice from what theory or practice in other western countries suggests. School choice of Hungarian parents is primarily driven by the intention to ensure a more culturally and economically homogeneous and safer environment for their children. Parental involvement tends to be stronger at home and at school due to higher parental religiosity, regardless of the school sector. The novelty of our findings is, that opposite to the literature, mostly not the school type but parents’ religiosity increase both the home based and the school based parental involvement.

KEYWORDS:

• Denominational schools
• school sector
• school choice
• parental involvement
• home based and school based involvement
• centralized and standardized schooling

Introduction

Studies on the added value of schools and education have highlighted the importance of schools’ ability to involve parents in the care for their children. Families’ school choice, aiming that the child should attend a “good school” in some sense, is a simply goal-oriented rational action, which is taken for granted for the elite. However, school choice is also value-oriented as parents choose the school where their children are most likely to encounter a culture similar to their family socialization. Our research question is whether Hungarian schools are able to create a school culture that is open to parents, and whether church-run primary schools are more effective than other schools in this respect.

School choice is a particularly important issue for Hungarian families because the Hungarian school system was characterized by a high degree of centralized regulation (prescribing the curriculum, exams, and outputs), considerable rigidity, and multiple parallel educational paths (tracking) during the years of socialism. This situation changed somewhat with the political changes after 1990: tracking increased, but centralization first decreased and then grew again after 2010.

As in many other European countries, denominational schools in Hungary are state-funded nonpublic institutions, operating on a noncommercial basis and following a state-mandated curriculum. They do not represent a high level of participation by European standards but are under constant attack by political organizations and experts who hold views similar to those of the Democrats in the United States (Jenkins, 2003; Maranto & van Raemdonck, 2012). Critics point out that the establishment of denominational schools in itself leads to undesirable local competition between schools, to an increased propensity to choose and change schools, and thus to greater inequalities in school composition. This argument assumes that denominational schools encourage parents to separate their own children from children of low-status or underclass groups who are consequently left behind. Parents who exercise their right to free school choice are thus interpreted as creating and sustaining segregation in schools in order to promote their children’s status attainment, with the goal of achieving upward mobility, maintaining their current favorable position, or avoiding a status loss.

Parents in favor of school choice insist on choosing the right school environment for their children, claiming that they have picked a denominational school because of the worldview and culture that characterize church-run institutions. Parents expect to encounter greater discipline, more rigorous academic requirements, and a strong religious and moral education in denominational schools, as international research has shown (Coleman & Hoffer, 1987; Shakeel & DeAngelis, 2018). The culture in denominational schools involves the development of an intergenerational school community, an intergenerational closure, whereby the school sees itself as a place where children are educated in accordance with the principles of family socialization (Maranto & Shakeel, 2021; Pusztai, 2011). Public schools, in contrast, offer a so-called neutral education, which does not require parents to commit to a worldview, limits the impact of the school to the curriculum, and does not interfere in matters related to upbringing. As expected, this leads to considerably different patterns of parental involvement.

Drawing on the literature, we distinguish the neutral, bureaucratic, client-like approach of public schools from the functional value-based communities of denominational institutions (Coleman & Hoffer, 1987). We consider parental involvement as an indicator of conscious parenting which requires responsible decision-making and subsequent action. We assume that parents who choose an institution other than a public school are more conscious and have a fundamentally different attitude toward their child’s education and schooling. The concept of parental empowerment implies that parents choose schools in order to achieve the best education for their children. Greater parental involvement at school and at home may be a manifestation of personal responsibility, of which the choice of a non-conventional school is only one indicator. In addition to school choice, the level and quality of investment parents make in their children’s education also reflect a significant parental decision. This is because parents’ involvement in their child’s education is weaker if they consider schooling as a mere bureaucratic act driven by the state. In our study, we compare parents in public and church-run schools in Hungary in terms of their involvement in child-rearing at school and at home.

Pluralism of providers and school choice in the Hungarian education system

The Hungarian education system has historically been part of the European continental model. At the end of the 18^th century, the so-called enlightened absolute monarchs, as part of their efforts to modernize the countries of Central Europe, organized a centralized education system. They drew on the well-established, centuries-old educational traditions of the Catholic Church and followed the German model. In Hungary, the state introduced compulsory education in the second half of the 19^th century and established eight-year elementary schooling before World War II. Schools had to teach centralized curriculum with a state-controlled examination system – in line with Allmendinger’s (1989) typology on the German school system.

After World War II, under the Soviet influence, church-run schools were banned in Central and Eastern Europe and state schools were obliged to transmit party ideology as set out in the central curriculum. After the collapse of socialism in the 1990s, the Hungarian education system broke with the rigid continental traditions and adopted the less centralized Anglo-Saxon model. Schools were handed over to local governments, the right to establish schools was opened up, resulting in a wide range of schools run by foundations, private organizations, and churches. This was the first wave of establishing denominational schools in Hungary (Pusztai, 2006). Moreover, local curricula were allowed, and the textbook market was liberalized. Several types of secondary schools (lasting 6 and even 8 years) came into existence, leading to a further increase of tracking in the public education system. This has reinforced the importance of school choice but also intensified inequalities, as families often made such decisions when their children were very young (Horn, Keller, & Róbert, 2016).

In the 2000s, the costs of the marketized education could no longer be covered by local governments with limited tax revenues. Educational institutions in lagging regions and municipalities became impoverished, and the development of local curricula severely reduced interchangeability and the chances for further education. The disadvantaged, rural, uneducated population received a poor quality education, while the advantaged, urban, better educated groups received a significantly better quality education (Pusztai & Chrappán, 2012).

Citing the growing inequality of educational opportunities in the 2010s, the Hungarian government reasserted state control in education by reducing the number of school providers, setting out uniform curricula, and providing free state-regulated textbooks. The state took over the funding and management of municipal schools, and churches were also encouraged to take over struggling municipal schools to reduce the burden on the state as school provider. Since the state offered more incentives to churches to take over schools in rural areas, denominational expansion was more pronounced there (Pusztai, 2013). This was the second wave of the establishment of church-run schools in Hungary, further expanding the choice of alternative schools.

Despite the centralized system in Hungary, the free establishment of schools and free school choice persist. The enrollment of primary school children in their own district is guaranteed by the state, but parents may apply for their child to be enrolled in another district. Under the Act on the Right to Freedom of Conscience and Religion, parents have the right to decide on the moral and religious education of their children, and under the Act on National Public Education, parents are entitled to opt for schooling in a denominational or private institution. The state carries out its duty to ensure public education by entering into contracts with other providers, including churches or foundations. As in most European countries, the public funding of education is conditional on parents choosing a school that follows the uniform state curriculum and has entered into an agreement with the state.

In Hungary, denominational schools follow the state curriculum. However, the law allows their pedagogical program to prescribe elements appropriate to their religious and ideological commitment, compulsory religious education, rules of conduct and appearance, duties, rights, and religious activities related to the teaching of the religious community, which are applicable to not only the children but also the teachers (Pusztai, Bacskai, & Morvai, 2021). Hungarian legislation also protects the rights of those who do not wish to attend denominational schools. If the parents of a child do not wish to choose a church-run school available in the place of residence, the local government must ensure that they are not disproportionately burdened if they choose a public school. The churches themselves rarely found new schools because the Hungarian school system is extensive and fragmented with many small schools. However, switching from a state provider to a church provider is more common. This process is usually initiated by local authorities because they do not have sufficient resources. Alternatively, local communities ask churches to take over the school. An institution can only be given denominational status if over half of the parents support it.

In the 2010s, the share of church-run schools has increased rapidly, particularly at primary school level (see Figure 1). However, the share of the church-run sector in Hungary is medium by international standards, lagging behind the proportion of denominational institutions in France, Spain, the Netherlands, and Belgium. For primary schools, both the share of church-run school sites and the share of children attending them are about 15% in the proportion of all students. As for secondary general education, churches run 29.7% of the school sites, with 25.3% of the students, again in the total of the whole school system. Among secondary vocational schools, denominational schools account for 15%, instructing 11% of all students in vocational education. The extent of the independent private sector is somewhat more modest, while the public school sector is dominant in Hungarian school system (KSH, 2020).

Figure 1. Number of students attending church run schools (2001–2020).

Source. KSH (2020).

Religious people are over-represented among those who opt for a church-run school. In the 2011 census in Hungary, one in two adults reported to belong to a religious community (KSH, 2012), but just over one in ten people reported to regularly and actively participate in religious community practice following the teaching of a church (Pusztai, 2013). Regional difference in social composition plays a role in explaining school choice. The population of rural regions is characterized by lower status and fewer religious people. At the same time, religious parents are more likely to choose denominational schools as opposed to public ones (Pusztai, 2013), as they seek schools where religious culture dominates (Martin, White, & Perlman, 2003; Uecker, 2009). In the capital and large towns, people of high status are more religious and choose church-run schools in significant proportions. However, church-run schools also educate a significant proportion of children with a non-religious background. In denominational schools, two thirds of teachers are strongly religious and follow the teaching of the church. In Catholic schools an additional 20% of teachers claim to be believers without belonging to a church, while this rate in Protestant schools is 31% (Pusztai et al., 2021).

Parental involvement: theoretical background and previous research

Our paper is largely based on the work by Joyce Epstein, who developed the methodology for improving parental involvement and school participation, and on the policy report by James Coleman, that bears his name (Coleman et al., 1966). In relation to parenting at school and at home, Epstein stressed that parenting and school cannot be rigidly separated, as the seemingly distinct sets of home-based and school-based activities actually intersect. Parenting at home may at first glance appear to be part of the private sphere, but it is in fact a social investment. In this sense, the parent is an external supporter of school education who assists and complements the educational processes. This parental behavior also affects children’s performance at school (Epstein & Sanders, 2002).

Parental involvement includes, on the one hand, parents’ visibility at school, participation in events, programs, activities organized by the school, and volunteering at school (Schneider & Coleman, 1993), and on the other hand, attending parent education workshops and teacher-parent meetings (Epstein & Sanders, 2002; Hornby & Lafaele, 2011). Another aspect of parenting is manifested in activity at home (Kim, 2009), often referred to as parental support, and this includes e.g. supervising homework or discussing what has happened at school. Parental support activities do not necessarily need to have a direct, immediate impact on children’s performance at school. Nevertheless, it contributes to the development of children’s specific skills and to the transmission of values, norms and expectations that lay the foundations for good academic performance (Sy, Gottfried, & Gottfried, 2013).

The dependent variable in our empirical analysis relates to the meta-analysis by Tan, Lyu, and Peng (2020), in which 98 studies on parental involvement were reviewed, and six indicators were found to increase student achievement. The three dimensions of home involvement were parental support for child learning at home, discussions at home about school matters, and parents and children reading together. The three elements of school involvement included parental academic expectations, parental participation in school work, and parent-teacher communication. However, it is also important to note that parental involvement at school may also be higher due to problems with the child in the school (Boonk, Gijselaers, Ritzen, & Brand-Gruwel, 2018).

As mentioned before, parental involvement, an investment in children, their education, and their future is a good way to describe the process of school choice. Parental influences are still quite strong in primary school, while the peer group plays a greater role in school-related processes and decisions at older ages. In this sense, the empirical part of this study, which examines the determinants of school and home involvement, relates to school choice.

Greater involvement, associated with a more conscious, thoughtful, and attentive school choice, is status specific. In terms of factors influencing involvement, parents’ favorable social background is the driving force (Epstein & Sheldon, 2006; Lareau, 1987; Markström, 2013; Souto-Manning & Swick, 2006). However, the positive effect is mainly due to the educational attainment of the parents and not to the financial situation of the family (Fleischmann & de Haas, 2016; Vinopal, 2018). This is because highly educated parents communicate more easily with teachers and are more likely to interact with them due to their similar cultural dispositions (McNeal Jr, 1999). Parents with low educational attainment are often unable to help with school tasks at home as they lack the necessary skills and capabilities (Ronak, Rashmi, & Srivastava, 2021).

As regards the first wave of expansion of denominational schools in Hungary after 1990, the study by Dronkers and Róbert (2004) showed that students in church-run secondary schools were more successful, both in terms of academic performance and their chances of entering into higher education. This may be due to informal parent-teacher relationships, but it can also be assumed that parents in these schools are more involved in their children’s academic progress at home or at school (Coleman, 1988; Coleman & Hoffer, 1987). In addition to the type of school, religiosity itself may be a driving force for parental involvement, especially in case of care and education at home (Jorgensen, Mancini, Yorgason, & Day, 2016).

In terms of family structure, investigating the existence and strength of ties, Coleman (1988) drew attention to the positive impact of stability as an indicator of social capital. He also considered the number of siblings another indicator of social capital and tried to predict the unit of parental attention and care per child. Siblings can indeed be seen as competing actors who share the parental resources (time, financial means, attention, and emotions) and limit each other’s access to social capital (Parcel & Dufur, 2001).

As a further indicator of family structure, researchers stressed that single mothers were less involved in school life, compared to when the father was also present (Mencarini, Pasqua, & Romiti, 2019). However, when single mothers remarried (resulting in a patchwork family), their children’s academic achievement increased, although mainly for mothers from favorable social backgrounds (Wagmiller, Gershoff, Veliz, & Clements, 2010). In developed countries, mothers are usually more involved than fathers both at home and at school, fathers’ involvement appeared to be slightly higher at school than at home (Crouter, McHale, & Bartko, 1993; Muller, 1998).

In addition to individual and family factors, contextual influences are also important. In Central Europe, the spatial disparities in the socio-economic composition of society are more pronounced than in Western or Northern Europe. The underclass and the elite are significantly segregated, and there are extreme differences in the socio-economic composition across both state and denominational schools. In high-status neighborhoods, schools operate with a favorable student composition, while in low-status neighborhoods, those who choose to attend alternative schools may not be different from families who accept the public school. Thus, student composition and size of schools can also influence involvement at school.

Based on the literature, the main hypothesis of our research is that the degree of parental involvement at home and especially at school differs depending on whether the school is church-run or not; and if so, whether it is a traditional denominational school or an institution taken over by a church only recently. We control for this effect using different variables that reflect the characteristics of families and schools. Of these, we assume stronger involvement if the parents are religious and/or have a higher level of education.

Data, measurements, methods

The data comes from a 2019/2020 Hungarian survey of parents of children in the fourth grade of primary schools (N = 1.156). Multi-stage cluster sampling was used: schools were selected in the first stage and classes in the second, aiming to survey the selected classes completely. Data collection took place using a paper-based self-administered questionnaire. Sampling intended to represent the Hungarian fourth-grade children in the selected schools and classes in terms of region, type of settlement of the place of residence, and social composition. Although the proportion of church-run schools in the sample reflects the corresponding national proportion, the final sample mostly includes relatively more advantaged church-run schools. The sample size has not allowed to differentiate between schools run by different denominations (Catholic, Protestant, Lutheran, Calvinist, Baptist, or Jewish).¹

Dependent variables

The two dependent variables in our study were parental involvement at home and at school. According to the meta-analysis by Tan et al. (2020), student achievement is enhanced by six variables measuring parental involvement. However, the constraints of the database led us to slightly modify this list by omitting certain factors (e.g., parental academic expectations, on which we had no data) and adding other factors to the analysis.

In measuring involvement at home, we asked questions about whether parents help their children with schoolwork. In line with a broader understanding of childrearing at home, we included variables on the parent performing general parenting tasks, and three items on the parent building a support network around the child. Thus, twelve items measured home involvement on a four-point scale, from which we constructed the dependent variable using principal component analysis. (See Table A1 in the Appendix.)

As for involvement at school, in addition to the frequency of teacher-parent contact, eleven activities measured whether the parent visited the teacher. As the twelve indicators were measured on different scales, they were standardized before creating the composite indicator. Again, principal component analysis was used to generate the dependent variable. (See Table A2 in the Appendix.)

Explanatory variables

The main explanatory variable in our analysis was the school provider. This variable took three values: church-run schools established in the first wave (1990–1995); church-run schools established in the second wave, after 2010; and non-denominational schools (run by state-controlled school districts, foundations, or minority self-government entities) as the reference category.²

Explanatory dummy variables included information on the parents such as: self-reported religiosity; homogeneous as well as heterogeneous graduate backgrounds with non- graduates as the reference category; the respondent is the mother of the child or otherwise. The analysis also incorporated additional explanatory variables such as: the respondent’s age in years; the number of children in the family; a binary variable on the family structure (single-parent family or otherwise); and finally the place of residence (urban or otherwise).

Two additional school-level variables provided contextual information, which were extracted from the 2019 National Assessment of Basic Competencies database and could be linked to our database using the unique identifier of the school. These variables were an index for the social composition of the school; and two binary variables for school size, whereby we distinguished between small and large schools based on the number of children in the 6^th grade (the reference category was medium-size schools).

Methods

Given the two dependent variables in our study, we carried out stepwise linear regression analysis. The explanatory variables described above were included in the regression model in four steps:

1. two binary variables for school provider;

2. the respondent’s self-reported religiosity; the number of children in the family; single-parent family; the place of residence;

3. whether the respondent was the mother of the child; the respondent’s age; two binary variables on parental tertiary educational attainment;

4. the school-level measures: social composition and school size.

Results

Before presenting the results of the multivariate analysis, we briefly describe the types of schools examined. Non-religious parents were overrepresented in non-denominational schools, while religious parents were overrepresented in church-run schools established after 1990. There were no significant differences in the proportion of single-parent families or the number of children by school provider. However, the proportion of parents with a degree differed. In non-denominational schools, there were more parents without a tertiary degree. In traditional church-run schools, typically only one parent had a tertiary degree, whereas in the newly established church-run schools both parents had a degree, the homogeneous graduate background was over-represented.

The settlement type of the child’s place of residence also differed significantly by school provider. While in non-denominational schools rural residence was overrepresented, traditional and new church-run schools tended to be more urban. Further differences by provider may be due to the fact that the sample included relatively advantaged church-run schools. In terms of school size, small schools are overrepresented in non-denominational schools, while both traditional and new church-run schools had a much higher share of medium-size schools. The social composition of schools also varied according to the school provider. While non-denominational schools were the most disadvantaged, new church-run schools had an average social composition, and the most favorable background was observed in traditional church-run schools.

Multivariate analysis

For parental involvement at home, our data revealed no significant difference by school provider (Table 1). In accordance with the literature, home involvement turned out to be higher when the respondent was religious, there were fewer children in the family, and the respondent was the mother. In addition, heterogeneous graduate parental background increased involvement at home. Home involvement was slightly more pronounced for younger parents, but this effect disappeared when controlled for the school-level measures. These variables, school’s social composition and size had no significant effect on parental involvement at home. Similarly, place of residence and, contrary to expectations, single-parent families were not significant predictors either.

Table 1. Estimates from linear regression analysis for parental involvement at home (N = 702).

	B(SE)	B(SE)	B(SE)	B(SE)
Traditionally church-run school	0.213 (0.17)	0.187 (0.171)	0.146 (0.168)	0.185 (0.177)
New church-run school	0.017 (0.106)	−0.033 (0.108)	−0.033 (0.106)	0.000 (0.109)
Respondent is religious		0.220 (0.075)**	0.162 (0.075)*	0.171 (0.077)*
Number of children		−0.11 (0.038)**	−0.095 (0.038)*	−0.09 (0.039)*
Single parent		0.009 (0.123)	0.030 (0.122)	0.038 (0.128)
Place of residence		−0.009 (0.076)	−0.034 (0.077)	0.012 (0.097)
Respondent is the mother			0.515 (0.114)***	0.534 (0.119)***
Respondent’s age			−0.014 (0.007)#	−0.012 (0.008)
Homogeneous graduate background			0.121 (0.096)	0.152 (0.104)
Heterogeneous graduate background			0.198 (0.092)*	0.216 (0.094)*
Social composition index				−0.014 (0.01)
Small school				0.208 (0.128)
Large school				−0.005 (0.097)
Adjusted R-squared	0.000	0.016	0.056	0.062

*** p < 0.001, ** p < 0.01, * p < 0.05, #p <0 .1

For involvement at school, however, significant differences were found by the type of school (Table 2). In Model 1, school involvement was slightly stronger for schools that were taken over by a church after 2010. While this effect was no longer significant when controlling for individual and school-level composition effects, it slightly reduced parental involvement at school in steps 2–4 for traditional church-run schools.

Table 2. Estimates from linear regression analysis for parental involvement in school (N = 760).

	B(SE)	B(SE)	B(SE)	B(SE)
Traditionally church-run school	−0.262 (0.174)	−0.341 (0.175)#	−0.363 (0.175)*	−0.309 (0.181)#
New church-run school	0.168 (0.101)#	0.082 (0.104)	0.078 (0.104)	0.107 (0.105)
Respondent is religious		0.162 (0.071)*	0.141 (0.073)#	0.152 (0.074)*
Number of children		0.014 (0.035)	0.012 (0.036)	0.018 (0.036)
Single parent		−0.084 (0.116)	−0.101 (0.117)	−0.146 (0.12)
Place of residence		0.182 (0.073)*	0.155 (0.075)*	0.129 (0.093)
Respondent is the mother			0.149 (0.114)	0.232 (0.117)*
Respondent’s age			0.012 (0.007)#	0.017 (0.007)*
Homogeneous graduate background			−0.007 (0.095)	−0.022 (0.103)
Heterogeneous graduate background			0.12 (0.091)	0.140 (0.092)
Social composition index				−0.018 (0.009)*
Small school				−0.108 (0.119)
Large school				0.02 (0.094)
Adjusted R-squared	0.005	0.015	0.018	0.033

*** p < 0.001, ** p < 0.01, * p < 0.05, #p < 0.1

Involvement at school was also increased for religious respondents and mothers, but the latter effect was only revealed in the last step when we controlled for the effect of school-level variables. The role of age was also significant, but the sign of the effect was opposite to that of home involvement, it was stronger when the parent was older. Contrary to the literature, our data showed that school involvement was unaffected by a graduate background, the number of children in the family, and single-parent situation. However, it mattered whether the family’s residence was urban, although the effect disappeared after adding school-level variables. Parental involvement at school was stronger when the social composition of the school was less favorable, but school size had no effect on the extent of parental involvement at school.

Discussion and conclusion

This study addressed the European perspective of school choice using the example of Hungary. The Hungarian case is relevant for school choice because the country’s educational system operates with considerable tracking, centralization, and standardized curricula and educational outcomes where parental decisions matter more. The historical background to this lies in the school system, traditionally following the German model and the socialist educational policy. Our analysis focused on church-run schools as such schools were again allowed to operate in Hungary after 1990, opening up a new dimension for school choice. As we have shown in the context of newly established church-run schools, in the first wave of the process after 1990, a kind of historical justice prevailed, when traditional church-run schools appeared again. The second wave of establishing church-run schools came after 2010, with a change of provider (from the state to a church) often for budgetary reasons.

School choice and parental involvement are linked in several aspects; both involve parental investment in the child’s success and better performance. Our data covered parental involvement at home and at school and related to school choice indirectly but allowed for a good comparison of public and denominational schools. Parental involvement is considered as an attribute of parents, but it is clear that schools can support families’ involvement in many ways. The aim of this study was precisely to investigate whether Hungarian church-run primary schools are more effective than other schools in this respect, or instead the differences in the parental communities of these schools result from parents’ different perceptions of their role and responsibilities based on religiosity. As our research targeted parents of fourth-grade primary school children, the data included the result of a previous school choice based on family background, especially religiosity.

Our results show that church-run schools in Hungary did not increase parental involvement as much as would be expected from Epstein’s or Coleman’s research. Parental involvement at home did not differ by school provider. The findings for parental involvement at school were closer to expectations, but did not reflect the hypothesis of stronger involvement in traditional, first-wave church-run schools than in new denominational schools or public schools. In fact, parental involvement was lower in traditional church-run schools than in public schools.

To understand this relationship, it is worth looking more closely at the questions used to measure parental involvement at school (Table A2). The proportion of affirmative responses regarding parents seeking contact with teachers was particularly high for two items: “when teachers ask to talk to me” (78.5%) and “if my child complains about school matters” (70.8%). This suggests that the main reason for parental involvement at Hungarian primary schools is some sort of problem with the child. Denominational schools may have fewer problems with children, especially in traditional institutions, which explains parents’ low involvement. The advantage of such denominational schools lies primarily in their favorable selection, i.e. school choice. This can create a problem-free atmosphere, which is not the same as the social capital environment Coleman describes. Thus, parental involvement in Hungarian primary schools is not of the nature what Epstein’s or Coleman’s theory would suggest; it is not aimed at favorable outcomes such as better school performance, but rather at avoiding potentially adverse environmental conditions for the child at school.

Nevertheless, we found that religious parental background was associated with high parental involvement both at home and at school, in line with our hypothesis. This result is largely in line with a range of previous findings regarding the impacts of religion on family life and social live more broadly (McCleary & Barro, 2019). In addition, our data showed a significantly higher proportion of religious parents in church-run schools, which suggests that parental involvement and school choice are indeed related, rather thanks to parents. At the same time, the evidence did not support other hypothesized effects, such as a higher number of siblings or a highly educated background, which were significant only for home involvement. The absence of the expected link that high-status (graduated) parents are more involved in school activities may also be due to the fact that parental involvement in Hungarian primary schools means something different. The mechanisms that would follow from Epstein’s or Coleman’s theory are not valid in either public or denominational schools in contemporary Hungary, due to the prevailing centralized and standardized expectations.

Some other specific features are also worth mentioning. In Hungary, church-run schools have denominational diversity regardless of the denomination of the provider, e.g. children of different religions attend separate religious education classes. Usually, parents do not belong to the same denomination and therefore do not attend the same church, nor do they form the kind of community described by Coleman. As our sample is weighed toward large urban church-run schools, parents come from a wide area, and even if they belong to one denomination, they do not necessarily participate in the same church community. Moreover, schools that have been taken over from the state by a church in the last ten years have mostly retained the previous teachers. As we know from previous investigations, teachers do not necessarily take on the extra work of involving parents, so the school may not (yet) have the atmosphere that one would expect from a denominational school (Pusztai et al., 2021). Teachers’ salaries in Hungary are blatantly low by European standards, according to Eurydice (2021), and teachers in church-run schools are not paid significantly better. Moreover, centrally prescribed curricula mean a higher teaching workload in Hungarian schools compared to the European average.

In this study, by linking school choice and parental involvement, we have sought to contribute to the understanding of motives behind school choice and the social mechanisms of the process. Our findings lead us to conclude that many Hungarian parents enroll their children in church-run (primary) schools to solve their parenting problems. They intend to avoid the conflicts arising from the social composition of public schools, where there is a high degree of cultural heterogeneity, partly due to the lower proportion of families following religious norms, partly because the average social status of families is lower, and children come from more heterogeneous social backgrounds. In poorly paid, dual-earner Hungarian families, parents are often overburdened with work, have neither the time nor the energy to get involved in school matters and leave their children to the care of the school, which they visit only when absolutely necessary.

The education system of contemporary Hungary features free school choice, which means that parents try to choose a better social environment for their child but have little chance of getting the child into a better educational environment. It seems that teachers in neither the public nor the denominational sector strive to create a welcoming school culture for parents and to support parenting but more research is needed to find further evidence on this. Since our database did not contain data on the community and volunteering factors of the Epstein model, we could not test here whether these dimensions are significant in any school sector. At any rate, the advantage of church-run schools is rooted in the higher proportion of religious parents, which positively influences parental involvement both at home and at school.

Acknowledgments

The paper is based on a research implemented by the MTA-DE-Parent-Teacher Cooperation Research Group, with support from the Research Program for Public Education Development of the Hungarian Academy of Sciences.

During the preparation of this study, Péter Róbert was supported by the project ‘Children in school: well-being and beyond,’ Grant number: NN 125715.

Disclosure statement

No potential conflict of interest was reported by the authors.

Notes

1. According to the census, two-thirds of those who identify themselves as belonging to a church are Roman Catholic, one-fifth are Calvinist. There are also smaller denominations, Greek Catholics, other Protestants (e.g. Lutherans, Baptists) and members of the Jewish community. Denominational schools belong to the these churches or religious organizations in the same proportion (Pusztai, 2013, 2022).

2. As for the reference category, the non-denominational schools, the fully state-run, socialist education system began to change in 1990 (Dronkers and Róbert (2004). Figure 1). The denominational sector seemed to have reached its maximum expansion by 2000, when primary schools accounted for 5%, secondary schools for 10%, and there were a few denominational vocational schools. As described and explained above, after 2010, there was again a significant increase across all schools types in the denominational sector, despite the demographic decline, while the public sector shrank (cf. Pusztai, 2006, 2013).

Appendix

Table A1. Descriptive statistics (row percentages) and factor loadings of items of parental involvement at home (N = 988).

In the past month, the respondent …	Never	Once or twice	Often	Daily	Factor loadings
Activities assisting child in learning
discussed school experiences with the child	0.2%	2.6%	11%	86.3%	0.531
studied together with the child	2.2%	15%	34%	48.8%	0.543
read tales, literature, or popular science texts together with the child	8.8%	34.7%	34.5%	22%	0.579
General parental tasks
discussed experiences from the Internet with the child	9.8%	29.9%	42.5%	17.9%	0.626
discussed general topics of life and the world with the child	3.3%	14.3%	48.5%	33.9%	0.667
discussed topics of religion and psychology with the child	9%	29.6%	39.6%	21.7%	0.605
planned common leisure activities with the child	2.3%	20.8%	62.9%	14%	0.672
played together with the child	2.6%	21.2%	54.2%	22%	0.671
and the whole family dined together	4.4%	10.6%	32.5%	52.5%	0.359
Building a supportive network for the child
talked with the parents of the child’s friends and peers	11.1%	42%	38%	8.9%	0.591
invited the child’s friends and peers	17.3%	46.9%	31.2%	4.7%	0.489
taken part in an extended family event or visit with the child	13.6%	34.9%	42.8%	8.8%	0.531

Note. The principal component measure had a minimum of −4.1, a maximum of 2.42, a mean of 0, and standard deviation of 1, explained 33.5% of the variance of the input variables, and had an eigenvalue of 4.02.

Table A2. Descriptive statistics (row percentages) and factor loadings of items of parental involvement at school (N = 1090).

In the past month, the respondent …	Never	Once or twice	Often	Daily	Proportion of affirmative responses (%)	Factor loadings
talked with the child’s teachers (personally, on the phone, electronically)	9.3%	43.9%	38.3%	8.4%		0.266
The respondent contacts the child’s teacher …
if the child’s behavior has changed					67%	0.585
if the child complains about school matters					70.8%	0.504
when picking up the child from school, almost always					24.5%	0.215
if deemed necessary based on discussions with other parents					36.9%	0.414
if problems arise in the child’s academic progress or the electronic administration system					63.3%	0.539
to show interest in new activities and events					40.6%	0.584
when teachers specifically ask to					78.5%	0.561
to offer parental help					58.2%	0.648
to say thank you for something					66.9%	0.675
to ask for the teacher’s help before a major event (e.g., moving house, birth of a sibling)					31.5%	0.522
to discuss a newly discovered ability of the child					27.7%	0.580

The principal component measure had a minimum of −2.36, a maximum of 1.8, a mean of 0, a standard deviation of 1, explained 27.6% of the variance of the input variables, and had an eigenvalue of 3.32.