Does exposure to books foster a taste for spatial mobility? Home library size in childhood and adult migration intentions

Abstract

Does exposure to books in childhood encourage adult migration? We contend that it does—through reading, which provides a simulative experience and arouses readers’ curiosity, as well as raising their openness to new experiences, awareness of opportunities in other places, and confidence in social situations, all of which foster migration intentions. Using data from the Life in Transition-III Survey, we find that the number of books in a childhood home is positively related to adult migration intentions. This relationship holds after controlling for a set of socio-economic characteristics, including individual and parental education. We also show that the book effect increases with the respondents’ level of education.

1. Introduction

The more that you read,

the more things you will know.

The more that you learn,

the more places you'll go.

    Dr Seuss. (1978)

How much truth is in this quote from the popular American children’s book? While its first part is unquestionably true, the second one does not seem that obvious. Can childhood reading indeed positively influence adult travel and, in particular, migration intentions? It is widely acknowledged that reading books ‘broadens horizons’—but does this mental horizon-broadening capacity also apply to spatial horizons? While cultural socialization within a family, including being taken to libraries and encouraged to read books, has been shown to promote upward social mobility (Scherger and Savage 2010), and the number of books in individuals’ childhood homes was shown to predict their educational and occupational attainments and skills later in life (Brunello, Weber, and Weiss 2017; Sikora, Evans, and Kelley 2019; Breznau, Sauter, and Salikutluk 2020; Heppt, Olczyk, and Volodina 2022), it is less clear whether reading books may influence an individual’s spatial mobility. The fact that the link between social and spatial mobility is unclear (Savage 1988; Favell and Recchi 2011; Fortunati and Taipale 2017; see also Hage 2005) does not make things easier.

We contend that books foster mobility aspirations, among other things, in that they make their readers more aware of opportunities in other places, more curious and open to new experiences, more confident in social situations, and can ‘transport them into a narrative world’ (Van Laer, De Ruyter, and Wetzels 2012), and offer a ‘simulative experience’ (Mar and Oatley 2008). We argue that narratives conveyed by books can inspire not only tourist visits (Wong, Lee, and Lee 2016) but also migration. At the same time, childhood reading is interdependent with a child’s family background—including, among other things, parental education level, type of work performed by them, ways of spending free time, and wealth (also affecting expenses for books). These considerations motivate our key research hypothesis: Childhood reading is positively related to adult migration intentions, net of an individual’s educational and socio-economic background.

Verification of this hypothesis is not easy given the potential difficulty of recalling one’s reading practices from childhood in a retrospective study and the associated recall bias. The number of books in childhood home may be perceived as a more tangible indicator and, given that a book-rich home environment was found to be positively related to the frequency of engagement in recreational reading (Merga 2015),¹ can serve as a proxy measure for childhood reading. In this study, we thus ask whether a large home library in childhood fosters future adult migration aspirations, which find expression in higher migration intentions among people who come from more ‘bookish’ homes. We hypothesize that the number of books in a childhood home is positively related to one’s migration intentions in adulthood.

To address this question and verify our hypothesis, we use multi-country cross-sectional data from the Life in Transition-III Survey, which included questions on both household library size in childhood and current migration intentions. In terms of geographical coverage, the survey focused primarily on post-socialist countries in Eurasia. Our findings demonstrate that the number of books in a childhood home is indeed positively related to adult migration intentions, net of individual and parental education, and other socio-economic characteristics of individuals. We argue that this independent influence of childhood home library size on current migration intentions points to the role of childhood reading in shaping one’s spatial aspirations. In addition, we find that the link between the size of a childhood library and migration intentions strengthens with the respondent’s level of education.

This article brings together research on determinants of migration intentions (for a review, see Aslany et al. 2021) and the formative importance of reading (Djikic, Oatley, and Moldoveanu 2013; Freestone and O’Toole 2016; Mar et al. 2006). Demonstrating the positive relationship between exposure to books in a childhood home and adult migration intentions, it adds to this literature in two ways. First, it deepens the knowledge of determinants of migration intentions, an important antecedent of migration behaviour (Tjaden, Auer, and Laczko 2019). In particular, it highlights the importance of accounting for unconventional, intangible (Hagen-Zanker, Hennessey, and Mazzilli 2023) factors in predicting migration intentions, which adds to a more holistic understanding of determinants of spatial aspirations in general (Schewel and Fransen 2022). Second, it, indirectly, contributes to the literature on the long-term effects of reading, by pointing to the unexpected consequences of childhood reading.

The remainder of the article is structured as follows. We first outline which constructs books at home may serve as a proxy for, and briefly review empirical studies that examined how these constructs, alongside books themselves, relate to migration intentions. We then explain in more detail why we expect childhood reading to positively influence adult migration intentions. Next, we describe the data, measures, and methods used. After reporting the results of our statistical analyses, we discuss our findings, pointing to the limitations of our study and suggesting directions for future research.

2. Books at home and migration intentions

While we assume that the number of books a family owns reflects its members’ reading activities, household library size may also be indicative of other factors potentially influencing migration. Books at home have typically been viewed as a tool of human and cultural capital formation, as well as a cultural capital artefact, signalling a high level of cultural capital and/or the high status of a family (Sieben and Lechner 2019). They may also capture family economic resources and parental care (Brunello, Weber, and Weiss 2017), educational household investments (Skrbis, Woodward, and Bean 2014), in addition to indicating socio-economic status (Heppt, Olczyk, and Volodina 2022) more broadly. That is why in the empirical analysis it is important to control for individual characteristics related to status.

In studies predicting individuals’ educational and occupational attainments with the number of books at home, the explanations for the positive book effect go beyond the operation of human and cultural capital. According to Breznau, Sauter, and Salikutluk (2020), this effect is supposed to come both from parents and ‘in a perchance form’, given the very presence of books at home. Books in the childhood home are thus supposed to reflect both parents’ commitment to scholarly culture, manifesting itself with parental support for the practice of reading (Merga 2015), and access to books as such. This is in line with Bourdieu’s (1986) view of books as an objectified form of cultural capital, whose properties are however revealed only in the presence of cultural capital in its embodied form (i.e. long-lasting dispositions, such as attitudes and preferences).

While migration theories generally predict education, as a form of human capital, to be positively related to migration at the individual level (Schewel and Fransen 2018), existing empirical studies have yielded mixed findings, with some studies showing that schooling increases one’s propensity to migrate (e.g., Ivļevs and King 2012; Gevrek, Kunt and Ursprung 2021) and others providing inconclusive or more nuanced results as to the role of education (e.g., Schewel and Fransen 2018; for a review, see Aslany et al. 2021). The role of cultural capital in influencing migration has been studied less extensively. In the US context, cultural capital, operationalized as interest and participation in the arts in high school, has been found to increase the frequency and distance of residential moves (Pettit 1999). Research addressing the link between books at home and migration specifically has been even scarcer. Access to books in the family home (treated as a measure of cultural capital) was found to stir a desire to be mobile—both within the country and internationally—among young people in Australia (Skrbis, Woodward, and Bean 2014). While migration was not the main outcome of interest in their study, Brunello, Weber, and Weiss (2017) noted that rural boys coming from homes with larger libraries were more likely to engage themselves in internal migration into cities (considered as means of reaping the economic returns from their education). Chapela (2022) included the ‘library size at home’ among control variables in a model predicting lifetime migration in Spain, concluding that ‘having more than ten books at home at the age of ten tends to reduce the likelihood of ever migrating’; his results were not statistically significant, however. Importantly, none of these three studies provided a convincing theoretical argument on why childhood home library size should affect adult migration intentions and behaviour. By considering such mechanisms, we attempt to fill this gap with our study.

3. Why do we expect reading to influence one’s migration intentions?

We argue that reading books, similarly to reading content on the internet, affects ‘people’s conceptions and awareness of other places’ (Thulin and Vilhelmson 2017). By increasing knowledge of the conditions and opportunities available in other locations, books, just like the internet, expand one’s spatial horizons (‘opens up worlds’, Thulin and Vilhelmson 2014), ‘broadening potential action space’ (Thulin and Vilhelmson 2016). By bringing the previously unknown and potentially geographically remote places closer, placing them within an individual’s ‘attainable reach’ (Thulin and Vilhelmson 2016), books influence individuals’ dreams, aspirations, intentions, and ultimately their decisions. They contribute to the development of ‘geographical imaginations’ that influence migration decision-making—both the propensity to migrate and the choice of destination (Thompson 2017). Since books are less likely than the internet to be a source of concrete, practical, and up-to-date information, their influence is more likely to manifest itself in the early, inspirational phase of the pre-migration process than in its later stages of migration decision-making (cf Grubanov-Boskovic et al. 2021). When the focus is on books read in childhood, this influence is expected to be even more likely to concentrate itself in this initial stage (although at times it may also shape later plans, e.g., by affecting the choice of destination by ‘two-step decision makers’, Brunarska 2019). By developing the awareness that people live differently in other parts of the world, it is expected to inspire curiosity and a general interest in other places, and to stimulate a desire to discover them. This interest may later stir adults’ desires and aspirations, influencing their early considerations of the possibility of moving elsewhere. In particular, fiction literature may be consequential for these initial thoughts about moving, since it not only produces mental images of living in specific places but also offers a ‘simulative experience’ (Mar and Oatley 2008), by ‘transporting readers into a narrative world’ (Van Laer, De Ruyter and Wetzels 2012). If an inability to picture the opportunities lost by not moving abroad is one of the main reasons for non-migrating (McKenzie 2024), then books should foster migration by helping people to visualize their life in a new place. These arguments are in line with Koikkalainen and Kyle’s (2016) proposition that people contemplating going abroad engage in cognitive migration (i.e., tend to visualize themselves in a future time and a specific place) before making an actual move. We argue that reading books make this ‘mental time travel’ more likely through providing the simulative experience and, in the long run, through the development of one’s imagination.

Intrapersonal factors are believed to be salient in the pre-contemplation phase of the migration process (Tabor and Milfont 2011). Importantly, one of the personality traits that has been shown to be consistently (positively) associated with a propensity for residential mobility is openness to experience (Jokela 2009; Canache et al. 2013; Tabor, Milfont, and Ward 2015; Fouarge, Özer, and Seegers 2019; Jokela 2021; Shuttleworth et al. 2021). This association may be explained by the fact that open individuals—who may be described as creative, imaginative, curious, active, broad-minded, and adventurous—are more inclined to both aspire to migrate and bring their desires into action. Acknowledging that personality is mutable through various experiences gained over the course of a lifetime, we assume that reading in childhood, especially since it is likely to involve fiction (Djikic, Oatley, and Moldoveanu 2013), contributes to open-mindedness, particularly open-mindedness in terms of ‘conceivable places to live and work’ (Thulin and Vilhelmson 2014). Given Canache et al.’s (2013) finding that the effect of openness is conditional on a respondent’s level of education, we additionally test for a conditional effect of books on migration intentions.

Apart from fostering migration aspirations, reading books equips readers with certain skills and knowledge that enable them to feel more confident in various social situations and consequently lowers the self-perceived psychical costs of adapting to new conditions awaiting prospective migrants abroad. This concerns, for instance, the ease of forming and maintaining weak social ties (Pettit 1999), which facilitates social integration abroad.

4. Data and methods

4.1 Data

We use data from the Life in Transition-III survey (LiTS-III), administered by the European Bank of Reconstruction and Development and the World Bank in 2015/2016. The survey covered twenty-nine post-socialist countries of Central and Eastern Europe and Central Asia (including Mongolia but excluding Turkmenistan), as well as Turkey, Greece, Cyprus, Germany, and Italy. The nationally representative samples consisted of ca. 1,500 face-to-face interviews in each country. Households were selected according to a two-stage clustered stratified sampling procedure. In the first stage, the framework of primary sampling units was established using information on local electoral units. In the second stage, a random walk fieldwork procedure was used to select households within the primary sampling units. The interviewees within households were automatically selected by using Computer-Assisted Personal Interviewing software.²

4.2 Variables

Our key independent variable captures the number of books in a childhood home and draws on the survey question ‘About how many books were in your childhood home? Do not count magazines, newspapers, or school books’. Respondents were offered five answer options: (1) ‘None or very few (0–10 books)’, (2) ‘Enough to fill one shelf (11–25 books)’, (3) ‘Enough to fill one bookcase (26–100 books)’, (4) ‘Enough to fill 2 bookcases (101–200 books)’, and (5) ‘Enough to fill 3 or 4 bookcases (200+ books)’. In some specifications, we treat it as a continuous variable, taking values from 1 to 5, which correspond to the five possible responses. In other specifications, we treat it as categorical in order to check for non-linearities, incorporating four dummy variables in our models. We also create a joint dummy variable for the ‘Don’t know’ and ‘Not applicable’ answers, which amount to 4.5 per cent of the sample (4 per cent and 0.5 per cent, respectively). While we are able to include the ‘Don’t know/not applicable’ dummy alongside the four number-of-books dummies, and thus avoid the potential bias from removing specific answers from the analysis, in specifications using the continuous number-of-books variable we have to exclude the ‘Don’t know’ and ‘Not applicable’ answers.

Our dependent variable—migration intentions—is based on the question ‘Do you intend to move abroad in the next 12 months?’, with two possible responses: ‘yes’ (1) and ‘no’ (0). A non-negligible proportion of respondents (2 per cent) provided a ‘don’t know’ answer, which we also code as 0. In two robustness checks, we (1) exclude the ‘don’t know’ answers from the analysis and (2) include them as a separate outcome of the dependent variable. The results of the corresponding binary and multinomial logit regressions (provided in the Supplementary Tables S4–S7) are consistent with our main results.

In line with previous literature on the determinants of migration intentions, we include the following control variables: age, gender, three levels of education (primary, secondary, and tertiary), marital status, employment status (proportion of time employed in the 12 months prior to the interview), self-perceived position on a 10-step wealth ladder, attitudes to risk on a 10-point scale, belonging to an ethnic minority, self-assessed health level, living in an urban area, and international migration experience (a dummy variable for having lived in another country or being born in another country or completing secondary school in another country).³ Crucially, the LiTS-III survey contains information on the education levels of the mother and father (primary, secondary, and tertiary), which we also use as controls. We also controlled for internet access to account for the fact that it can independently influence migration intentions through similar theoretical mechanisms. For the survey questions used and basic summary statistics for all variables included in the analysis, see Supplementary Table S1.

4.3 Estimation strategy

Given the binary nature of the dependent variable, we estimate the relationship between the availability of books in a childhood home and migration intentions using a binary logit model with country-fixed effects. We include the latter (dummy variables for countries) to account for the observed and unobserved country-level influences that might influence people’s migration intentions. This allows us to focus on within-country individual-level relationships, without the risk of them being confounded by country-level characteristics. As a robustness check, we additionally estimate corresponding multilevel models. Their results (consistent with our baseline estimations) can be found in the Supplementary data (Supplementary Tables S8–S9). Additionally, given the issues with the inclusion of interaction terms in non-linear models (Ai and Norton 2003), to test for the continuous-by-continuous book by education interaction, we use the linear probability model.

5. Results

Table 1 reports the country-level means of the variables of interest: intentions to migrate and the size of the childhood library (captured by the mean of the continuous number of books in the childhood home variable as well as by the proportions of people reporting the highest and lowest categories of childhood library size). Intentions to migrate range from 0.33 per cent in Turkey to 10.94 per cent in Armenia. We also observe considerable country-level heterogeneity in the childhood library size. Both these observations warrant the utilization of either country-fixed effects or a multilevel approach.

Table 2 reports the means of all variables included in the analysis—for the whole sample, for people with and without intentions to migrate, and for those reporting the lowest and the highest size of the childhood library. We notice that people with migration intentions had larger childhood libraries than people without migration intentions (people with the largest childhood library size are also more likely to report migration intentions than people with the smallest childhood library size), which provides initial support for our main hypothesis. Overall, prospective migrants are younger and more likely to be male, not married, in work, healthy, risk-loving, live in urban areas, belong to an ethnic minority group, have higher levels of individual and parental education, previous international migration experience, and have internet access. We also observe stark differences in the socio-demographic characteristics of respondents having the smallest and largest childhood library size.⁴

Table 3 reports the results of multivariate regression models. Specifications 3.1 and 3.2 include only the childhood library size variables and country-fixed effects. The results suggest that an additional unit on the 1-to-5 library size scale increases the likelihood of reporting migration intentions by 0.8 percentage points (P < .001)—a substantial increase relative to the proportion of people reporting migration intention (3.37 per cent). Treating the childhood library size as a set of five dummy variables (specification 3.2) confirms the positive monotonic relationship between the variables of interest: relative to people with none or very few books in childhood, those reporting a number of books enough to fill one bookcase are 1.6 percentage points more likely to reports migration intentions, two bookcases—2 percentage points, and three or four bookcases—3 percentage points.

The inclusion of standard control variables (specifications 3.3 and 3.4) reduces the magnitude of the marginal effects of the childhood library size variables by approximately 40 per cent, highlighting the correlation between the library size variables and some of the controls.⁵ Further inclusion of parental education variables results in somewhat smaller and less statistically significant marginal effects of the childhood library dummies, although the overall relationship between the variables of interest is still visible (specifications 3.5 and 3.6). Finally, controlling for internet access further reduces the size and statistical significance of the associations of interest.⁶ In the model with a full set of controls (specifications 3.7 and 3.8), an additional unit on the 1-to-5 library size scale increases the likelihood of reporting migration intentions by 0.4 percentage points (P < .001); alternatively, relative to people with no or very few books in childhood, those with a number of books enough to fill one bookcase are 0.7 percentage points more likely to report intentions to migrate (P < .05), two bookcases—0.9 percentage points more likely (P < .05), and three or four suitcases—1.7 percentage points more likely (P < .001).

To test if the relationship between childhood library size and migration intentions depends on the respondents’ level of education, we estimated our models for the three different levels of education (specifications 4.1–4.6 of Table 4). The childhood library size variables are statistically non-significant in the sample of respondents with primary education (specifications 4.1–4.2). For respondents with secondary education, the marginal effect of the continuous library size variable is statistically significant at the 0.01 level and, in terms of magnitude, similar to the full sample estimate (specification 4.3). In the model with the library size dummies (specification 4.4), the respondents educated to a secondary level with the biggest childhood library size are 1.3 percentage points more likely to report migration intentions than their counterparts with the smallest library size (P < .05). These effects more than double in magnitude and become statistically significant at the 0.001 level for those educated to a tertiary level (specifications 4.5–4.6). Taken together, these results show that the positive link between the childhood library size and migration intentions strengthens with the respondent’s level of education (see also Fig. 1).

We further check for the moderating effect of education level by running a full-sample estimation with an interaction term between the continuous childhood library size variable and the ‘continuous’ education variable (that is assigned values 1, 2, and 3 for primary, secondary, and tertiary education). The results, reported in specification 4.7 of Table 4, confirm the moderating effect of education: the interaction term is positive and statistically significant (P < .01).

Moreover, Fig. 1 suggests that the more educated people with the largest home libraries are, the more likely they are to show migration intentions. This suggests that books alone do not affect intentions; they need to be coupled with formal education.

6. Discussion

In this article, we examine whether exposure to books can influence individual spatial mobility. Specifically, we are interested in the effect of childhood reading on adult migration intentions. While current or recent reading habits may be measured based on questions on the frequency and quantity (duration and efficiency) of one’s reading behaviour, in a retrospective study based on adult recollections of childhood reading practices such questions would surely yield a substantial proportion of ‘don’t know’ answers and would be subject to considerable recall bias. This is where the question of home library size—in which case the problem of recall bias is likely to be smaller as it is easier to re-imagine bookshelves in childhood home than to recall and assess one’s reading practices from the past—comes in handy. We thus focus on the effect of exposure to books in the childhood home on adult migration intentions. Offering several theoretical motivations, we hypothesized that the number of books at home in the past is positively related to one’s current migration intentions. We argue that this positive link occurs due to childhood reading, which is encouraged by the presence of books at home, and reading’s formative role in providing a simulative experience and making readers more aware of opportunities in other places, more curious and open to new experiences, and more confident in social situations.

Based on the Life in Transition-III data, we showed that the number of books in a childhood home was indeed positively related to adult migration intentions, net of individuals’ socio-economic characteristics, including individual and parental education, and self-perceived wealth. By controlling for wealth and education levels, we attempted to isolate the independent effect of books (e.g., through the openness to experience induced by them) from the effects of knowledge and status. We argue that childhood library size fosters childhood reading, which shapes one’s future spatial aspirations and results in higher migration intentions.

Our analysis demonstrates, furthermore, that the book effect increases with the respondent’s level of education—it is not discernible in individuals with a primary level of education but becomes positive for people with secondary and especially tertiary education. We can speculate that children who originated in bookish homes but failed to acquire higher education levels had not read books that had been available in their home libraries (and hence the effect of books is not visible in their case). An alternative explanation may be that the effect of books is intensified by education, since the former fosters migration aspirations, while the latter additionally increases migration capabilities/abilities (Carling and Schewel 2018; De Haas 2021), and only their combination results in stronger migration intentions.

The observed moderating effect of education on the relationship between childhood library size and adult migration intentions is in line with Canache et al.’s (2013) findings, who found that the effect of openness to experience on emigration intentions is conditional on education level. Like us, they found a complementing effect, with openness predicting the intention to emigrate only for the highly educated and having no effect where the level of education is low. This consistency of both our and Canache et al.’s results provides indirect support for the proposition that the underlying mechanism behind the positive book effect can be related to openness. Interestingly, Canache et al. view and interpret their interaction from the other perspective—stating that whether people act on the opportunities offered by their higher education level may depend on their personality (among other things, openness).

Our study has some limitations, which indicate potential avenues for future research. First, the number of books at home in childhood is only a proxy for past actual reading activities and a rough one at that. It does not allow us to directly capture the influence of reading books (focusing on owning books instead). While existing studies point to the positive association between owning and reading books, treating the former as a proxy for the latter disregards, for example, the possibility of borrowing books from a library or family and friends, or the possibility that some people like to stock up books, while others prefer to pass them to other potential readers. Moreover, this measure does not make a distinction between children’s and adults’ books, whereas the number of children’s books may matter over and above the number of books belonging to the parents (Heppt, Olczyk, and Volodina 2022). It also does not differentiate between types of books, while it has been shown that exposure to fiction and non-fiction may have divergent effects on social outcomes (Mar et al. 2006; Djikic, Oatley, and Moldoveanu 2013). Moreover, it does not specify the period in the childhood that a respondent should think of when recalling their childhood library size. This may be consequential in particular for respondents whose childhood coincided with the period of political transformation, in which case the reply to the library size question could most strongly depend on whether they thought of the late 1980s or 1990s.⁷ We nevertheless assume there existed a certain continuity of reading habits and the ownership of books within families. Finally, such an empirical approach disregards the fact that household library size may be also indicative of other variables that are related to both the number of books at home and migration intentions, for instance, parents’ social networks. Household library size can also be linked to the consumption of other cultural goods (magazines, newspapers, movies, cartoons, theatre, radio and TV shows, exhibitions, etc.), which potentially influence migration intentions through similar theoretical mechanisms. Unfortunately, we cannot test for these factors as we lack information on the respondents’ consumption of these cultural goods during childhood.

Second, the available data do not allow us to test empirically for the role of openness to experience in shaping the relationship between reading books and migration intentions. Studies on the ‘book effect’ on educational and occupational achievements often face the problem of endogeneity, as not only members of households with larger libraries fare better, but those performing better accrue more books (Engzell 2021). While it is theoretically plausible that parents accrue books with a view towards their future migration plans and transmit these plans to their children, this problem is less salient in our study. There is, however, another potential endogeneity issue concealed in our theoretical reasoning: what if inborn openness to experience makes one read more in the first place? While it is reasonable to assume that more open-minded children are more likely to read more, as has been shown in a study on adults (Trapp and Ziegler 2019), this personality trait does not necessarily directly translate into the number of books at home (unless it is genetically inherited, assuming that the agency lies with the parents when it comes to buying books). Nevertheless, the reverse causality—that pre-existing characteristics are being maintained through reading—cannot be entirely ruled out.

Third, we focus on migration intentions in our empirical study, while in theoretical terms one may expect to find larger effects in a study focusing on migration aspirations as an outcome, since they concern hypothetical ideal circumstances, while intentions already account for constraints (Huber et al. 2022). Thus, to study the impact of exposure to books on an individual’s migration—or even more broadly: mobility—aspirations throughout the course of a lifetime would be a worthwhile endeavour for future research. While a stronger book effect is expected to be found with respect to aspirations than intentions, the discrepancy between aspirations and intentions may be greater in the case of books than in the case of other predictors—it remains to be established whether children reading more books are more likely to become ‘dreamers’, that is, willing to migrate but never transforming their aspirations into intentions and being able to fulfil them. Acknowledging that determinants of migration behaviours, intentions, and aspirations may diverge (see Brzozowski and Coniglio 2021), a study on the influence of exposure to books on migration behaviours would be another valuable direction for future research.

Acknowledgments

We wish to thank two anonymous reviewers for their valuable feedback on an earlier draft of this paper.

Supplementary data

Supplementary data are available at Migration Studies online. Replication code for the analyses, allowing the recreation of all the operations on the original survey data, is available at the osf.io platform (doi: 10.17605/OSF.IO/VGEPC).

Conflict of interest statement

None declared.

Funding

This work was supported by the Polish National Agency for Academic Exchange under the Bekker NAWA Programme [BPN/BEK/2021/1/00226] and the University of Warsaw under the Microgrant of the ‘Excellence Initiative—Research University’ programme [BOB-661–1109/2022].

Data availability

The survey data that support the findings of this study are available from the EBRD website (https://www.ebrd.com/what-we-do/economic-research-and-data/data/lits.html).

Notes

References