The price of religiosity: Enticing young Haredi men into secular academic studies

Abstract

Standard economic theory cannot explain why so few Haredi (ultra-orthodox) men attain college degrees in Israel, despite the significant economic returns to such degrees. In addition to economic variables, this article introduces a combination of social and behavioral characteristics, such as religious identity, into the individual choice process. This, in turn, enables us to evaluate a possible trade-off between economic benefits associated with a college degree and the corresponding loss of religious identity in the decision of young ultra-orthodox Israeli men to attain a college degree. In the language of standard economics, we simply ask: What is the economic price Haredi men are willing to pay for their religious beliefs? Utilizing case-control sampling and analysis techniques, we collected retrospective data on Haredi men who study in academic higher education institutes (N = 410) and matched them to Haredi men who do not attend college (N = 310). Our logit econometric model indicates that Haredi men, even those with extreme orthodox beliefs, respond to economic incentives. However, our model indicates that the size of the incentives required to entice academic studies increases with the intensity of religious beliefs.

Keywords

Educational choice religiosity retrospective case control ultra-orthodox (Haredi) Jews

Introduction

Haredi Judaism, synonymous with ultra-orthodox Judaism, developed in Europe in the 18th and 19th centuries as a reaction to modernity (i.e. emancipation, enlightenment, and secularization) processes (Caplan, 2003; Friedman, 1991).¹ The Haredi sought to mitigate the influences of modernity by segregating and distinguishing themselves from modern society. To date, Israeli Haredi Jews (roughly 900,000, about 11% of Israel’s population) reside in segregated neighborhoods and communities and distinguish themselves from the secular population by adopting strict behavioral and appearance codes and norms, among which is a rejection of secular education (Friedman, 1991). Thus, for example, in 2013, only about 9% of Israeli Haredi men aged 15 years and over had a college degree, compared to more than 35% of secular Jews aged 15 years and over (Regev, 2013).

Haredi norms pose a challenge to standard economic theory, according to which the behavior of an individual depends on economic variables (Berman, 2000). These models cannot explain why so few Haredi men attain a college degree in Israel, despite the significant economic returns to college degrees. In fact, about 80% of Haredi men who have acquired academic education participate in the labor market (a rate that resembles that of the general population) and their remuneration is about 80% higher than that of their Haredi peers, who lack higher education (Regev, 2013). In light of this apparent paradox, we examine in this article the economic compensation required to entice Haredi men to acquire secular academic education at the expense of their religious identity. Put differently, we explore the economic price Haredi men might be willing to pay for their orthodox religious beliefs. We do so in the spirit of a theoretical model posited by Akerlof and Kranton (2002), wherein we introduce a combination of social and behavioral characteristics, such as religious identity, into the individual choice process. This feature enables us to evaluate the trade-off between the economic benefit associated with acquiring higher secular education and the loss of Haredi identity entailed by such a move.

The next section presents the general literature on factors affecting the choice of higher education, followed by a review of the unique characteristics of the Haredi society. Section four then presents an econometric model that assesses the trade-off between monetary incentives and ultra-orthodox religious identity in the decision of Haredi men to go to college. This is followed by a section that presents the data, measures, and methodology. The last section of the article summarizes the findings and discusses them in light of the literature.

The choice of higher education

Until about two decades ago, the economic literature dealt mostly with the influence of factors such as the return on academic education and employment opportunities on the choice of education and educational track (Berger, 1988; Polachek, 1978). Numerous empirical studies have supported the human capital model developed by Becker (1964) and Mincer (1962) and show that more education is correlated with higher income (Card, 1999). According to the human capital model, educational choice derives from cost-benefit calculus, associated with the loss of current income during the time an individual acquires education and the prospect of higher income once the educational process is completed.

The empirical economic literature rarely addressed the interaction between educational decisions and social factors. In contrast, the sociological literature demonstrates a strong correlation between educational choices and social and family background, including “significant others” (Sewell et al., 1970), social class (Breen and Goldthorpe, 1997; Levy and Massalha, 2010), family income (Becker and Hecken, 2009; Hofferth et al., 1998), cultural capital (De Graaf et al., 2000), and parents’ education (Behrman, 1997; Chevalier et al., 2010; Dryler, 1998). These studies indicate that the choice of children from higher social classes is more ambitious than that of children from lower social classes, even when ability is controlled (Jackson et al., 2007), and find strong positive causal relationship between parents’ and children’s education. Cultural capital is also found to influence educational choice, with the degree of children’s exposure to books and reading being a dominant factor influencing these choices (De Graaf et al., 2000).

However, it appears that cost-benefit calculus provides only a partial explanation for choices in the areas of educational and occupational careers. Specifically, rational economic cost-benefit calculus fails to explain, inter alia, the tendency of high-skilled students to choose educational tracks with low returns (Humlum et al., 2012). Similarly, the ability of social and cultural characteristics to explain these educational choices is relatively weak when economic variables are not taken into account. Explanations that combine the economic-rational approach with the cultural-social approach are quite rare in the literature (but see, Kroneberg et al., 2010).

In this context, Akerlof (1997) presents an economic model that combines social influences on individual choice considerations and describes the interaction between personal aspirations (such as the desire to increase economic well-being) and social ones. In this model, individuals tend to adapt their behavior to the norms which are accepted in the social category to which they belong. As a result, individuals are willing to give up the realization of personal aspirations in order to stay in the group. The idea is further expanded in the work of Akerlof and Kranton (2000), who added to the original model the variable “self-identity.” According to this new model, individuals derive utility from material benefits but also from the extent to which their own characteristics comply with those most preferred by the group to which they belong. Thus, the model finds that individual decisions depend also on the person’s self-identity. Empirically, the social identity component has been shown to be very dominant in the choice considerations among US high school students (Humlum et al., 2012).

This study applies the Akerlof and Kranton (2002) approach to quantitatively examine why so few Haredi men go to college in Israel, and to explore the possibility that both self-identity (ultra-orthodox religious identity) and monetary incentives are important determinants in the secular educational attainment process of Haredi men in Israel. In order to understand the context, we first present the complexity of the Haredi society in Israel and its attitude toward education and the labor market.

The Haredi community in Israel

The Haredi concept of religious devotion is characterized by a deep commitment to Halacha—Jewish law—according to the religious tradition that developed in Eastern European Jewish communities, coupled with a tendency to favor the stricter rulings in the Halachic literature (Friedman, 1991). In its ideology and lifestyle, Haredi Judaism presents itself as a counterculture to the surrounding society (Caplan, 2003). This is translated into daily life in various and broad aspects: the status and size of the family and the individual in the community; the attitude toward work and religious studies; the attitude toward gender difference; the choice of place of residence (Cahaner, 2009); and men’s allocation of time to religious activities such as Torah study, prayer, and charity (Berman, 2000).

The ability of the Haredi community in Israel to sustain its lifestyle while also segregating itself from the general society and economy is largely related to its political role in the Israeli parliamentary system, where the Haredi parties are the linchpin of practically any coalition government. This position enables the Haredi community to preserve a minimal, yet acceptable, standard of living for Haredi men who devote their time exclusively to religious studies by providing them state support. As a result, many of these men are not exposed to secular education and do not participate in the formal economy (Berman, 2000; Gal, 2014).

In the absence of an official statistical identification of Haredi persons, researchers of the Haredi community have devised other ways to characterize the population and assess its size. These have included, for instance, categorization according to the educational institution attended by the children; voting for Haredi parties in elections for the Knesset, and self-definition (Central Bureau of Statistics, 2011; Portnoy, 2007). Other researchers use specific indexes to identify Haredi persons, such as those who describe themselves as fully adhering to the directives of rabbis in their daily lives, those who describe themselves as eating only food certified as kosher by Haredi bodies rather than the state-sanctioned Chief Rabbinate, those whose children go to Haredi educational institutions (“Heder” for young children, and “lower Yeshiva” and “higher Yeshiva” for teenagers and young adults, respectively), and those who report they do not possess a TV (Cahaner, 2009). Most of the researchers who have studied the Haredi population estimate it at 8%–11% of the Israeli population, that is, 600,000–900,000 persons (Cahaner, 2009; Central Bureau of Statistics, 2011; Gottlieb, 2007).

Haredi society is not homogeneous and is divided into numerous sects that represent different social and ideological proclivities, with the main division being between Hasidic: 38%, Litvish: 27%, and Sephardic: 30% (Degani and Degani, 2000; Friedman, 1991).

Hasidic

In the religious sense, Hasidism emphasizes spiritual worship out of joy and fear of God and designates its community leaders as “our master and teacher” or “Rebbe,” as a holy man who intermediates between the Hasid and his Creator. The lives of the Hasidic community surround the courtyard of the rabbi, who, in addition to his spiritual function, also serves as a significant cohesive force in daily life. The Hasids, who conduct most of their economic and social relations within the community and marry within the community, maintain small businesses or are employed in such businesses as salaried employees, and maintain a high level of integration and mutual responsibility (Flint et al., 2010).

Litvish

Unlike the Hasidic, the Litvish sect functions around institutions of Torah study, rather than a rabbi and courtyard, and considers religious studies as the highest value. The Litvish community constitutes a religious elite, and its leaders set the accepted norms for all streams of the Haredi society. The social value of a Litvish man is measured by the time devoted to Torah study and his excellence in these studies (so-called “Torah scholar”). A Litvish family usually lives off the wife’s earnings and a state subsidy that the Litvish students receive from the kolel—the Haredi educational institution for married men (Flint et al., 2010).

Sephardic

This sect includes Haredi persons whose roots are in Islamic countries, the old Sephardic community in Palestine, and the Ladino-speaking community. This group differs from the other groups in its ethnic, cultural, and institutional background, and is rejected by the Ashkenazi (of European origin) Haredi community and its educational institutions (Flint et al., 2010). This rejection led to the creation of Haredi institutions for the Sephardic community that enjoy government support (Friedman, 1991). Despite the establishment of Sephardic Yeshivas and educational institutes, there is currently a growing community of Sephardic persons who consider themselves Litvish in terms of their lifestyle and the centrality of religious studies in men’s life (Flint et al., 2010).

Attitudes of Israeli Haredi Jews to general education and employment

Over the past several decades, all Haredi sects have tracked their children to pursue full-time Torah study for many years and to shun secular studies. That tracking was made possible by the political power of Haredi parties, enabling them to recruit state budgets to maintain a “community of learners” (Friedman, 1991). Friedman (1991) opines that the success of that community also contains the seeds of a constant erosion of its fortitude and economic and social stability, because the rapid demographic growth and cultural isolation of the Haredi community is destabilizing it. Furthermore, he argues that the maintenance of time-intensive religious activities limits individual choices and causes the willful destruction of resources (see also Berman, 2000).² Thus, for example, the exclusion of the Haredi from the labor market has extremely severe economic consequences that are reflected by high poverty rates. According to the National Insurance Institute (2014), the rate of Haredi Jews whose income is beneath the “poverty line” is 52.1% of the total Haredi community, compared to 11.6% among the non-Haredi Jewish public.

To counter some of these forces, the Israeli government has established as of 1999 special programs providing academic and professional training to the Haredi public, especially targeting men. By 2003, there were only about 450 students in these programs (Malach, 2014). Since 2004, however, there has been a sharp rise among Haredi men who seek academic education, and by 2010, the number reached 2000 (Bank of Israel, 2010; Malach, 2014).³ According to Malach et al. (2016), the average growth rate in the number of Haredi male students in the years 2011–2014 was 14% (from 2132 in 2011 to 3227 in 2014), and in 2016, the number of Haredi men enrolled in higher education institutes rose to 3850 (an average annual growth of 10%). As mentioned above, despite the impressive numerical growth in the last decade, the proportion of Haredi academic degree holders out of the total number of working-age Haredi men still remains very low, at about 9% (Regev, 2013).

The growing tension between the desire to belong to the community and the existential need to acquire education creates serious dilemmas for young Haredi men and draws resistance from some of the leaders of the Haredi community against integration in the secular society. On this backdrop, this article examines the extent to which the Haredi identity as against economic incentives influence the willingness of Haredi men to go to college.

The theoretical model

As noted above, in this article, we rely on a theoretical model developed by Akerlof and Kranton (2002). The model illustrates the impact of the identity variable on choices in the area of education and offers a possible answer to the question as to why individuals with high abilities might choose occupations with low income, or educational tracks with low return on investment. In the context of the Haredi society, such a model might explain why many Haredi men choose to exclude themselves from higher education despite the contribution of such education to their earning potential.

In the spirit of this model, the study focuses on the choice of a Haredi man to pursue higher education given his personal background, the particular Haredi sect to which he belongs, the perceived communal costs associated with secular college education, his careerist identity, and his ultra-orthodox identity. The latter identity might be related to the lifestyle typical to that community (studying Torah, strictly following dietary laws, and so on), the extent to which secular culture is avoided (such as surfing the Internet, reading secular press, and the like), and attitudes toward integration in the secular society. According to this model, a more salient and robust ultra-orthodox identity entails, ceteris paribus, a lifestyle which is removed from the secular one and thus reduces the likelihood of choosing secular academic studies. Conversely, a relatively weak ultra-orthodox identity entails, ceteris paribus, a more modern lifestyle that also encourages the realization of personal aspirations and hence increases the likelihood to go to college.

The econometric model

Based on the considerations detailed above and in the spirit of the Akerlof and Kranton (2002) model, we assume that an individual who belongs to the community of Haredi men derives benefit from income and from his ultra-orthodox identity. By assumption, there is a conflict between these two attributes, implicitly due to the trade-off between time allocated to religious activities and time allocated to secular studies (cf. Berman, 2000). Accordingly, a young Haredi man faces the choice between adhering to his religious identity, in which case his income will be low, or choosing to acquire secular training which will reduce his religious attributes but increase his income.

Following McFadden (1978), we assume that a generic individual’s incremental utility, derived from choosing to study rather than not, is given by

U = β_{0} + \sum_{k = 1}^{m} α_{k} Z_{k} + β S + γ I + δ C + ε Y + u

where

$Z$ is a vector of personal background characteristics, S is the individual’s social category (the Haredi sect), I is the individual’s social identity, $C$ is the social cost of choosing to acquire academic education, $Y$ is the expected monetary reward for academic studies, and $u$ is a random disturbance term reflecting unobserved preference shocks.

Clearly, the value of U is not observed, but the decision whether to study (D = 1) or not (D = 0), is. Specifically, the former would be the case if U > 0, and the latter if U < 0. If u Logistic, it can be shown that

P r (D = 1) = \frac{1}{1 + e x p (- β_{0} - \sum_{k = 1}^{m} α_{k} Z_{k} - β S - γ I - δ C - ε Y)}

leading to the logistic specification for the odds that the individual chooses to opt for academic studies:

l o g i t (D) = l o g (\frac{P_{(D = 1)}}{1 - P_{(D = 1)}}) = β_{0} + \sum_{k = 1}^{m} α_{k} Z_{k} + β S + γ I + δ C + ε Y

which enables us to estimate the parameters $β_{0}, α_{k}, β, γ, δ, and ε$ . The data and the variables are described in detail in the consecutive section.

Data, variables, and indicators

We collected data purposely for this study. Probing the motivation of Israeli Haredi men to attend college is best seen as a study of rare events, since the dependent variable (“attending college”) would not be easily identified in a random sample of Haredi men in Israel. A solution to this problem can be achieved by the use of retrospective samples—also known as case-control samples (Agresti, 1996; Lacy, 1997; Manski, 1995; Manski and Lerman, 1977; Varese and Yaish, 2000; Xie and Manski, 1989). Our data contain information extracted from questionnaires answered by a sample of Haredi men who study in academic higher education institutes (the “case”), matched to a sample of their peers who do not attend college (the “control”).

The “case” sample consists of students sampled from all known public higher education institutes with dedicated classes or campuses for Haredi men that opened in the years 2010–2013.⁴ This sample consists of 410 students sampled at random during 2014, in accordance with their proportions in the different educational institutes. The “control” sample of non-student Haredi was recruited through the group of students. In total, 100 of the sampled students were selected at random and asked to find five Haredi men from their community and age group, who live in the same building or nearby. The purpose of selecting neighbors was to match the case and the control samples, as required by this technique (Agresti, 1996), on personal, familial, and community background characteristics.⁵ Furthermore, the cooperation of the sampled students facilitated the cooperation of their “neighbors.” In total, 310 questionnaires were handed to the group of non-students, so that a total of 720 questionnaires were collected. However, the statistical processing of the empirical study is based on only 445 questionnaires (251 students and 194 non-students), because some of the respondents did not define themselves as Haredi and some did not respond to all the questions relevant to the study.

The most appropriate method to adopt in the analysis of these data is the method of case-control samples. Accordingly, the response variable (i.e. the dependent variable) is first identified and sampled, while the independent variables (which we present below) are the random variables in the analysis. As noted above, we apply logistic regression, which is based on odds ratios, to estimate the retrospective effects of the independent variables on the response variable. Since this method is often characterized by samples that do not necessarily represent the entire population (i.e. we do not have information on the true marginal distribution of the response variable in the population), we are unable to weight the data. While this may cause some problems in interpreting the intercept in our models, the coefficients for the effects of the independent variables are nonetheless interpretable. That is to say, we lose the predictive power of the models, but we can learn about the effects of various variables on the outcome we are interested in (Agresti, 1996; Varese and Yaish, 2000; Xie and Manski, 1989).

Variables and indicators

The background variables include: social category, measured through the response to the question: to which Haredi sect do you belong? The respondents could choose the Hasidic, Litvish, or Sephardic sects. The variable “mother’s education” was measured by years of schooling. Focusing on the mothers reflects the fact that Haredi girls are not prevented from acquiring secular education, and therefore, secular education is more prevalent among them. The variable “exposure to secular culture” is based on a positive response to one of the two following questions: “During your childhood, have you read books by secular authors?” and “Have you listened to secular singers?” The variable “childhood standard of living” is the average of normalized scores for 12 consumer items available at the parental home (see Appendix 1).

At the core of the theoretical and econometric model are social (identity) and economic (incentive) variables. The identity indices are derived from a factor analysis from the answers concerning characteristics of identity, behavior, attitudes toward higher education, employment, and salary. We have applied a principal component with a varimax rotation on the items tapping the Haredi lifestyle (e.g. studying Torah and strictly following dietary laws), the extent to which secular culture is avoided (such as surfing the Internet, reading secular press, and the like), and attitudes toward integration in the secular society. This procedure yielded three factors, with 48% cumulative variance explained, representing ultra-orthodox identity, careerist identity, and social costs associated with attending college. We use the factor score to measure these three indices, implying that they are standardized with mean zero and unit standard deviations. The results are presented in Table 1.

Table 1.

Item, loadings, and reliability of key factors generated by factor analysis.

Variable	Item	Factor loading	Reliability (α)
Ultra-orthodox identity	Dietary laws observed in accordance with stringent Jewish law (extra strict Kosher)	0.571	0.658
	Keeping daily Torah study	0.619
	Surfing the Internet (not for study and work)	0.597
	Reading secular newspapers	0.673
Career identity	Yeshiva students with economic difficulties should work	0.663	0.632
	Haredi Jews can go to college to study profession	0.530
Expected social costs (perceived)	Is a single man from your congregation who is in college a less desirable marriage match?	0.804	0.718
	Is a single man from your congregation who serves in the army a less desirable marriage match?	0.820

Finally, the economic incentive for college education, anticipated market premium for higher education, is calculated by computing the wage gaps between individuals with academic education and those who lack it, based on the two following questions:

In my opinion, the average monthly wage of a Haredi man who has a bachelor’s degree is NIS___________.

In my opinion, the average monthly wage of a Haredi man who does not have a college degree is NIS___________.

The answers are used to compute the anticipated rate (percentage) of change in the wage of a Haredi man due to academic education, rounded off into intervals of 10 percentage points.

Findings

Table 2 shows the averages (and standard deviations) of the research variables, divided into the groups of students and non-students, and an indication of statistically significant differences between the groups. As can be seen, the background variables do not distinguish between the two populations, implying that the matching of the two groups on these background variables (mother’s education, exposure to secular culture, childhood standard of living, and social category—except the Hasidic sect), was successful, thereby validating the sampling method.

Table 2.

Descriptive statistics (and standard deviation) of the variables.

Variable	All	In college	Not in college
Hasidic	0.37 (0.48)	0.42** (0.49)	0.29 (0.46)
Litvish	0.36 (0.48)	0.33 (0.47)	0.40 (0.49)
Sephardic	0.27 (0.46)	0.24 (0.43)	0.31 (0.46)
Mother’s education	0.75 (0.43)	0.74 (0.44)	0.77 (0.42)
Exposed to secular culture as a child	0.35 (0.48)	0.36 (0.48)	0.32 (0.47)
Standard of living (childhood)	−0.01 (0.45)	0.00 (0.46)	−0.02 (0.48)
Ultra-orthodox identity	0.26 (0.65)	0.14** (0.61)	0.43 (0.66)
Career identity	0.00 (0.87)	0.18** (0.71)	−0.25 (0.99)
Expected social costs (perceived)	0.03 (0.87)	−0.05* (0.95)	0.13 (0.75)
College premium (perceived)	4109 (3578)	4762 (3847)**	3263 (3003)
N	445	251	194
%	100	56.4	43.6

Significant differences between the college and non-college Haredi: *p < 0.05; **p < 0.01.

As expected, statistically significant differences exist between the two populations in the social and economic variables. Compared to the students, members of the non-student group display stronger ultra-orthodox identity, have relatively lower career aspirations, and perceive the social sanctions involved in digressing from the Haredi lifestyle to be more severe. As seen in Table 2, even though both groups anticipate an income increase with academic education, there are essential differences between the two groups with respect to the corresponding expected premium for higher education.

A closer look at the anticipated market premium for higher education, as shown in Table 3, points to a number of interesting patterns. First, Haredi students, just like the non-students, estimate the monthly wages of a Haredi man without an academic degree at NIS 5550, a little above the Israeli minimum wage. Second, and in contrast to that, there is a major and statistically significant difference between the students and non-students in their estimate of the wages of a Haredi man with a college degree. Respondents from the students’ group anticipate that a Haredi man with a college degree should expect to earn an average of NIS 10,312, whereas respondents from the non-students’ group anticipate that the average wage of such a Haredi man would be only NIS 8812. The difference of 1500 NIS between the anticipated wage for someone with a degree and someone without one is the perceived premium for academic education, as shown in the third row of Table 3. More importantly, Table 3 shows that the market premium for a college degree expected by Haredi students is 86%, very close to the findings of Regev (2013), who reports that the wages of Haredi men with academic degrees working full-time were 80% higher than the wages of Haredi men without a degree.

Table 3.

Subjective evaluation of Haredi monthly earnings with and without a degree.

	All	Not in college	In college	Δ
Expected monthly earnings for Haredi men without a degree	5550 (1826)	5549 (2099)	5550 (1587)	1
Expected monthly earnings for Haredi men with a degree	9658 (3764)	8812 (3212)	10,312 (4026)	1500**
Expected pay premium for degree	4109 (3578)	3263 (3003)	4762 (3847)	1500**
Average expected premium rate for degree	86% (97.02)	70% (68.73)	98% (112.8)	28 percentage points**

p < 0.01.

The econometric model

As indicated above, the econometric model uses a logit procedure to estimate the effects of the different dependent variables on the odds of choosing academic training. Table 4 shows the estimates of a number of preliminary models (Models 1–3) as well as the full econometric model (Model 4). The coefficients presented in Table 4 indicate a change in the odds ratio that derives from a unit change in the value of an independent variable in the model, so that coefficients larger (smaller) than 1 indicate a positive (negative) effect on the probability of studying. Having made these clarifications, we can embark on the analyses.

Table 4.

Odds ratios (and standard errors) from logistic regression models predicting college participation.

	1	2	3	4
Litvish	0.562* (0.132)	0.698 (0.173)	0.536** (0.128)	0.661 (0.166)
Sephardic	0.496** (0.126)	0.479** (0.131)	0.493** (0.127)	0.478** (0.131)
Mother’s education	0.804 (0.192)	0.783 (0.197)	0.798 (0.194)	0.789 (0.200)
Exposed to secular culture as a child	1.293 (0.284)	0.996 (0.232)	1.252 (0.279)	0.979 (0.231)
Parental standard of living	1.050 (0.234)	0.979 (0.226)	1.074 (0.242)	1.000 (0.233)
Ultra-orthodox identity		0.568** (0.0992)		0.572** (0.101)
Career identity		1.516** (0.192)		1.488** (0.191)
Expected social costs (perceived)		0.829 (0.106)		0.859 (0.112)
College premium rate (perceived)			1.049** (0.0173)	1.040* (0.0171)
Constant	2.095** (0.549)	2.602** (0.734)	1.476 (0.422)	1.919* (0.585)
N	445	445	445	445
Pseudo R²	1.8%	7.8%	3.8%	9.1%
AIC	611	580	601	574
BIC	635	617	629	615

AIC: Akaike information criterion; BIC: Bayesian information criterion.

p < 0.05; **p < 0.01.

The base model (Model 1) in the table reports the impact of background variables and the social category of the respondents that are predetermined by the time they make their choices. Similar to Table 2, here too, only the variable “social category” has a statistically significant effect on the odds ratio. In columns 2 and 3, the social variables and the economic variables were added in turn. As expected, these factors have a meaningful and statistically significant impact on the odds of a Haredi man to go to college. Columns 2 and 3 show that compared to the base model, the contribution of the social variables to the explained variance is higher than that of the economic premium. Model 4 combines the social and economic factors. Even though the contribution of the social variables to the explained variance is stronger than the contribution of the economic one, Model 4 indicates that their combined contribution is stronger than that of each of them separately. This means that the econometric model—and thereby the underlying Akerlof and Kranton’s (2002) model—receives a significant reinforcement from the data. Further evidence that Model 4 is the preferred one can be found in the model fit statistics (Bayesian information criterion (BIC) and Akaike information criterion (AIC)), which are presented in the table.

Like the previous models, Model 4 also indicates that among the background variables, only the social category (i.e. Haredi sect) has a statistically significant effect on Haredi men’s odds to go to college. Again, this is not surprising and is an expected outcome of the sampling method, according to which the control sample (the “non-students”) was matched to the case sample on personal, familial, and community background characteristics.

The estimates of the econometric model are consistent with our theoretical expectation that social identity plays an important role in the decision to go to college. Thus, for example, a rise in ultra-orthodox identity by a standard deviation (with all other characteristics remaining fixed) reduces the odds that a Haredi man will go to college by about 60%. The model also indicates that a rise in career identity by a single standard deviation increases the odds of a Haredi man to go to college by 50%.⁶ Although the expected social cost does not reach the standard statistical significance level of 0.05 (probably due to lack of statistical power associated with the relatively small sample size), it is still negatively associated with the likelihood of Haredi men to go to college, as expected. Model 4 also shows the effects of the economic variables. As expected, economic incentives do exert a statistically significant effect on Haredi behaviors, such that raising the perceived market premium for higher education to the next 10 percentage point category increases the odds ratio of studying by about 4%. This is in line with the Akerlof and Kranton (2002) theoretical model, where the trade-off between identity and material factors plays a central role.

In conclusion, we find that Haredi identity has a stifling influence on the odds of acquiring academic education, while the economic premium goes in the opposite direction. In light of these findings, we next quantify the trade-off between the economic incentives required to offset the negative impact of Haredi identity in the pursuit of academic education.

The monetary incentive to pursue academic studies

To quantify the monetary equivalent of the Haredi identity, we exploit the conflicting effects that the market premium and the identity variables have on the odds of pursuing academic education. Specifically, we hold the odds ratio fixed at 1 and compute the trade-off between these two factors.

For the purpose of this calculation, we define a “typical individual” as a Haredi man characterized by the following properties: (1) his mother has post-secondary education; (2) he was not exposed to secular culture in childhood; (3) his family standard of living is average; (4) he has a median careerist social identity; and (5) has an average level of expectations as to social cost.⁷ Given that characterization, we turn to each Haredi sect and compute the combinations of the Haredi identity and the expected earnings premium from academic education that are required to induce an odds ratio of 1, at different levels of Haredi identity. Since the latter are specific to each of the Haredi sects, the resulting trade-offs are computed separately for each sect, while the base wage is identical to all, taken to be NIS 5500, the average anticipated wage of a Haredi man without an academic education.

Table 5 and Figure 1 present the results of this analysis, clearly indicating that monetary incentives work also on Haredi men. That is, standard economic theory works, though in the case of Haredi men, monetary compensations should be adjusted also to the level of religious identity and sect. Thus, for example, Table 5 reports that enticing a Hasidic man to choose academic education with a probability of 0.5 requires a positive monetary incentive only if his Haredi identity is very high. In contrast, for a Sephardic man, inducing the same choice probability requires paying NIS 4627 on top of the base income, even if his Haredi identity is just at its median value.

Figure 1.

Monthly earnings premium required to entice Haredi men to attend college (with probability 0.5), by ultra-orthodox identity and sects.

Table 5.

Exchange between ultra-orthodox identity and earnings premium, as determinants of college attendance with probability 0.5, by sects.

Ultra-orthodox identity (in z scores)		Monthly earnings premium required (base level = NIS 5500)
Ultra-orthodox identity (in z scores)		Hasidic	Litvish	Sephardic
Very low	−0.75	0	0	0
	−0.50	0	0	0
	−0.25	0	0	693
	0.00	0	0	2660
Medium	0.25	0	58	4627
	0.50	0	2025	6595
	0.75	0	3993	8562
	1.00	121	5960	10,529
Very high	1.25	2089	7927	12,496

Figure 1 provides a graphic presentation of the identity–premium combinations that generate a choice probability of 0.5 for academic studies. The three curves represent these combinations for each of the Haredi sects. All curves have positive, but finite, slopes. This indicates that monetary rewards can induce that choice probability at any level of religious identity, albeit at amounts that increase with religious fervor. Thus, a Litvish person, whose religious normalized identity is 0.25, requires no additional incentive to induce a choice probability of 0.5, but if his identity is 1, the required additional compensation is double the base wage. The location of the curves reflects the sect-specific degree of religiosity. The leftmost, representing the Sephardic group, clearly implies the toughest degree of religiosity, as it implies the highest amount of added compensation required to hold the choice probability constant at 0.5 at any level of Haredi identity. For the Hassidic group, at the right, that compensation is the lowest, which is consistent with that group’s characterization provided earlier. The ordering between the Litvish group and the Sephardic one is less obvious, but may reflect the latter’s attempt to emulate the former and prove its adherence to the strictest rules of ultra-orthodox Judaism as manifested by their great unwillingness to exchange their Haredi identity against material rewards.

Conclusion

Ultra-orthodox Jews live in Israel in extremely segregated communities, and study in religious schools that do not prepare (men) to go to college or to the labor market. The ideal Haredi man is a religious scholar, fully dedicated to the Torah, subsidized by his relatively more educated wife and the State. A consequence of this ideal is wide-spread poverty among Haredi families. This situation poses a challenge to standard economic thinking, as it would appear that positive and negative economic incentives do not seem to lure Haredi men to college in order to shun poverty. This article sought to contest this conclusion by introducing monetary as well as psychological and social determinants into a Haredi men’s choice to attain secular academic education.

The idea that economic incentives alone (i.e. economic cost-benefit calculus) determine individuals’ behaviors and actions has been challenged time and again by explanations that combine the economic-rational approach with cultural-social approaches (Akerlof, 1997; Akerlof and Kranton, 2000; Kroneberg et al., 2010). In this article, we adopted Akerlof’s (1997) and Akerlof and Kranton’s (2000) insight and formulation whereby individuals derive utility from material benefits but also from the extent to which their own characteristics comply with those most preferred by the group to which they belong. Put differently, individual decisions depend also on self-identity. Applying this model to study the underlying motivations of Haredi men to avoid/attain college education, we introduced a combination of social and behavioral characteristics, such as religious identity, into the individual choice process in order to evaluate a possible trade-off between economic benefits associated with college education and the loss of Haredi identity entailed by such a choice.

Our results can be summarized fairly straightforwardly: Haredi men do respond to economic incentives, but the size of the incentives needed to entice them to significantly change their behavior depends on the strength of their religious identity. Should we be surprised by this result?

If one is a student of standard economic theory, the answer is: not at all. According to this theory, economic variables determine human behaviors and actions, and the results of this study support this view as they illustrate that religious identity and ideology can be overcome by (sufficiently large) economic incentives. But also students of Israeli Haredi ideology and rituals should not be surprised by these findings.

There is already ample evidence to suggest that ultra-orthodox Jews in Israel are well responsive to external economic incentives. Thus, for example, labor force participation rates of ultra-orthodox men aged 35–54 years in Israel plummeted from about 85% in 1979 to as low as 40% in 2005 (Regev, 2013: Figure 1, page 3), in conjunction with growing government welfare policy benefiting the Haredi lifestyle (e.g. child allowance favoring big families and state support for Torah studies). Just as important, when the then Finance Minister Netanyahu abruptly cut welfare budgets in 2003, Haredi men responded to the challenge and their proportion in the labor force increased by about 25%, from 40% to nearly 50%, in less than a decade (Regev, 2013: Figure 1, page 3).

The point to be made here is that also in as sheltered, secluded, and ideological community as that of the ultra-orthodox Jewish community in Israel, the logic inherent in the homo economicus paradigm works—with an additional cost factor related to their religious fervor and communal adherence.

Footnotes

Appendix

Appendix 1.

Variables in the standard of living index (α = 0.669).

Items	% have
Second kitchen (for passover)	0.32 (0.47)
Paid housekeeper	0.36 (0.48)
Dishwasher	0.30 (0.46)
Drier	0.77 (0.42)
Air conditioner	0.87 (0.34)
Computer	0.71 (0.46)
VCR/DVD	0.37 (0.48)
Radio	0.75 (0.43)
Stereo audio system	0.84 (0.37)
Apartment or house	30.71 (0.93)
Car	10.66 (0.59)
Travel abroad	10.38 (0.72)

Acknowledgements

The authors thank the editor of Rationality and Society and two anonymous reviewers for their valuable comments and suggestions. An earlier version of this article was presented at the Sociology Departmental seminar at the University of Haifa. The authors thank participants in this seminar for helpful comments.

Funding

The author(s) received no financial support for the research, authorship, and/or publication of this article.

Notes

ORCID iD

Meir Yaish

References

Agresti

(1996) An Introduction to Categorical Data Analysis. New York: John Wiley & Sons.

Akerlof

(1997) Social distance and social decisions. Econometrica 65(5): 1005–1027.

Akerlof

Kranton

(2000) Economics and identity. The Quarterly Journal of Economics 105(3): 1167–1201.

Akerlof

Kranton

(2002) Identity and schooling: some lessons for the economics of education. Journal of Economic Literature 40(4): 1167–1201.

Bank of Israel (2010) Bank of Israel annual report–2010. Jerusalem. Available at: https://www.boi.org.il/en/NewsAndPublications/RegularPublications/Pages/eng_app10e.aspx

Barak-Corren

(in press) Reexamining the evidence on ultra-orthodox attitudes and gender separation in academia. Mishpatim [In Hebrew].

Becker

(1964) Human Capital: A Theoretical and Empirical Analysis with Special Reference to Education. New York: National Bureau of Economic Research.

Becker

Hecken

(2009) Higher education or vocational training? An empirical test of the rational action model of educational choices suggested by Breen and Goldthorpe and Esser. Acta Sociologica 52(1): 25–45.

Behrman

(1997) Mother’s schooling and child education: a survey. Penn Institute for Economic Research, Working paper no. 97. Philadelphia, PA: University of Pennsylvania, 15 May.

10.

Berger

(1988) Predicted future earnings and choice of college major. Industrial and Labor Relations Review 41: 418–429.

11.

Berman

(2000) Sect, subsidy, and sacrifice: an economist’s view of ultra-orthodox Jews. The Quarterly Journal of Economics 115(3): 904–952.

12.

Breen

Goldthorpe

(1997) Explaining educational differentials: towards a formal rational action theory. Rationality and Society 9(3): 275–305.

13.

Cahaner

(2009) The Development of the Spatial and Hierarchic Structure of the Ultra-Orthodox Jewish Population in Israel. PhD Thesis, University of Haifa, Haifa, Israel.

14.

Caplan

(2003) The study of the Haredi society in Israel: characteristics, achievements, and challenges. In: Sivan

Caplan

(eds) Israeli Haredim: Integration without Assimilation? Tel-Aviv, Israel: The Van Leer Jerusalem Institute; Hakibbutz Hameuchad Publishing House Ltd, pp. 224–278.

15.

Card

(1999) The causal effect of education on earnings. In: Ashenfelter

Card

(eds) Handbook of Labor Economics, vol. 3. Amsterdam: Elsevier Science, pp.1801–1863.

16.

Central Bureau of Statistics (2011) Measurement and estimates of the population of ultra-orthodox Jews. Technical paper no. 25, September. Jerusalem, Israel: Central Bureau of Statistics.

17.

Chevalier

O’Sullivan

Walker

(2010) The impact of parental income and education on the schooling of their children. Discussion paper series, Geary WP/201032, July. Dublin: UCD Geary Institute.

18.

De Graaf

de Graaf

Kraaykamp

(2000) Parental cultural capital. Sociology of Education 73: 92–111.

19.

Degani

(2000) The demand for housing in the Haredi sector. Report no. 3039, September. Ministry of Construction and Housing, Institute for Spatial Analysis LTD.

20.

Dryler

(1998) Parental role models, gender and educational choice. The British Journal of Sociology 49(3): 375–398.

21.

Flint

Alfasi

Benenson

(2010) Between friends and strangers: micro-segregation of Haredim in Jerusalem. Israeli Sociology 12(1): 81–110.

22.

Fong

Isajiw

(2000) Determinants of friendship choices in multiethnic society. Sociological Forum 15(2): 249–271.

23.

Friedman

(1991) The Haredi (Ultra Orthodox) Society–Sources, Trends and Processes. Jerusalem, Israel: The Jerusalem Institute for Israel Studies.

24.

Gal

(2014) Ultra-Orthodox Jews in Israel: A Status Report, 2014. Haifa, Israel: Samuel Neaman Institute for National Policy Research.

25.

Gottlieb

(2007) Poverty and Labor Market Behavior in the Ultra-Orthodox Population in Israel. Jerusalem, Israel: The Van Leer Jerusalem Institute.

26.

Hofferth

Boisjoly

Duncan

(1998) Parents extrafamilial resources and children’s school attainment. Sociology of Education 71(3): 246–268.

27.

Humlum

Kleinjans

Nielsen

(2012) An economic analysis of identity and career choice. Economic Inquiry 50(1): 39–61.

28.

Jackson

Erikson

Goldthorpe

, et al. (2007) Primary and secondary effects in class differentials in educational attainment: the transition to A-level courses in England and Wales. Acta Sociologica 50: 211–229.

29.

Kroneberg

Yaish

Stocke

(2010) Norms and rationality in electoral participation and the rescue of Jews in WWII: “an application of the model of frame selection.” Rationality and Society 22(1): 3–36.

30.

Lacy

(1997) Efficiently studying rare events: case-control methods for sociologists. Sociological Perspectives 40(1): 129–145.

31.

Levy

Massalha

(2010) Yaffa: a school of their choice? British Journal of Sociology of Education 31(2): 171–183.

32.

McFadden

(1978) Modeling the choice of residential location. In: Karlqvist

, et al. (eds) Spatial Interaction Theory and Residential Location. Amsterdam: North Holland, pp. 75–96.

33.

Malach

(2014) A “Kosher” Degree: Academic Studies in the Haredi Sector. Jerusalem, Israel: Floersheimer Studies.

34.

Malach

Cahaner

Regev

(2016) The Five-Year Plan of the CHE-PBC for the Ultra-Orthodox Population for the Years 2011-2016. Evaluation Study and Recommendations for the Next Five-Year Plan (2016-2020). Jerusalem, Israel: Council for Higher Education.

35.

Manski

(1995) Identification Problems in the Social Sciences. Cambridge, MA: Harvard University Press.

36.

Manski

Lerman

(1977) Estimation of choice probabilities from choice-based samples. Econometrica 45: 1977–1989.

37.

Mincer

(1962) On-the-job training: costs, returns, and some implications. Journal of Political Economy 70(5): 50–79.

38.

National Insurance Institute (2014) Annual report: 2013. National Insurance Institute, Jerusalem, Israel.

39.

Polachek

(1978) Sex differences in college major. Industrial and Labor Relations Review 31(4): 498–508.

40.

Portnoy

(2007) Characterization of the Jewish population by level of religiosity based on linkage to educational institutions. Technical paper no. 19, July. Jerusalem, Israel: Central Bureau of Statistics.

41.

Regev

(2013) Education and employment in the Haredi sector. In: Ben-David

(ed.) State of the Nation Report: Society, Economy and Policy in Israel: 2013. Jerusalem, Israel: Taub Center for Social Policy Studies in Israel, pp. 119–162.

42.

Sewell

Haller

Ohlendorf

(1970) The educational and early occupational status attainment process: replication and revision. American Sociological Review 35(6): 1014–1027.

43.

Varese

Yaish

(2000) The importance of being asked: the rescue of Jews in Nazi Europe. Rationality and Society 12(3): 307–334.

44.

Xie

Manski

(1989) The logit model and response-based samples. Sociological Methods & Research 17: 3283–3302.