Introduction

Geographical mobility plays a significant role in shaping demographic and social change, and it is a critical aspect of many people’s lives. The study of its social impacts is, therefore, of great importance, as highlighted by the latest research (see Vidal & Lersch, 2021). Despite the long-standing recognition of the geographical dimensions of social inequalities in classical research on social stratification (e.g., Blau & Duncan, 1967),Footnote 1 there is a surprising scarcity of studies examining how geographical mobility affects occupational success and the perpetuation of social inequalities across generations (Ballarino & Panichella, 2021; Bernard, 2023). As recognized by the most recent research on geographical mobility, studies considering the effects of geographical relocation from the point of view of social stratification are scant, despite the substantive importance of the phenomenon (Huinink et al., 2014, p. 1).

This article investigates the effect of South-to-North internal migration on occupational status in Italy, a country characterized by significant geographical disparities (Felice, 2013) and high levels of internal mobility (Panichella, 2014).Footnote 2 This migration flux experienced significant growth from the 1950s to the 1970s, involving more than nine million southern Italians (Impicciatore & Strozza, 2016). Moreover, the South-to-North internal migration shared similarities with current international migration trends, including unidirectional movement, integration challenges, and social conflicts (Panichella, 2018).

The study aims to determine the impact of geographical mobility on occupational success for both men and women, and to measure the migration benefit or disadvantage. In other words, it seeks to determine whether the choice to migrate from the South to central or northern regions has been more advantageous compared to the decision to stay in the place of origin.Footnote 3 Moreover, by adopting a gender lens, the article delves deeper into examining whether the impact of geographical mobility on occupational status is influenced by other significant social factors, including family dynamics such as marital status and parenthood, and the social class of origin. While there is consistent literature on the interrelation between mobility and family issues (Mulder, 2018; Impicciatore & Panichella, 2019), which are the basis of the contrasting effects of mobility on the occupational careers of women and men (Boyle et al., 2009; Cooke et al., 2009; De Jong & Graefe, 2008), to the best of our knowledge, only a few studies considered how the outcomes of geographical mobility change according to the movers’ social background of origin. Is the migration benefit/disadvantage stronger for individuals with lower origins or those with higher origins?

The findings of this study are expected to contribute to a deeper understanding of the complex interplay between geographical mobility and social stratification, shedding light on the ways in which human mobility shapes individual life chances and perpetuates social inequality across generations. By considering both the family dynamics and the social background of origin, this research provides a nuanced perspective on the effects of geographical mobility on occupational success, highlighting the multiple and intersecting factors that influence the outcomes of mobility for men and women in Italy.

This article is organized as follows. The second section provides a review of relevant theories and research on the occupational consequences of geographical mobility, focusing on the role of human capital, gender, and family dynamics. In the third section, we examine how geographical mobility interacts with social background, outlining four different scenarios for the intergenerational reproduction of social inequality. The fourth section describes our data and variables, whereas the fifth focuses on our research strategy and methodology. In the sixth section, we present the results of our empirical analysis. The article concludes with a discussion of the implications of our findings for the study of geographical mobility.

Exploring the Intersection of Family and Gender on the Occupational Outcomes of Geographical Mobility

The economic approach to migration research views both the decision to move and its consequences on life chances through the lens of human capital (Sjaastad, 1962). Mobility is thus seen as an investment with benefits, risks, and costs, resulting in a process of social selection (Bernard & Bell, 2018; Chiswick, 2000). Numerous studies support this perspective and show that those who move geographically (the movers) tend to be more educated, motivated, and ambitious than those who remain in their original society (the stayers or non-movers) (Panichella, 2014). Moreover, mobility also has a positive impact on occupational status, as it expands the job search area and enables movers to escape the limitations of their original occupational structure and take advantage of job opportunities elsewhere (Ballarino & Panichella, 2021), despite the potential disruption to social networks and separation from family and friends (i.e., the social capital is ‘location-specific’, see DaVanzo, 1981). Both internal and international migration are thus seen as an instrument of occupational achievement (van Ham, 2003, p. 6), as it provides access to better occupational positions when job opportunities are unevenly distributed across regions (Huinink et al., 2014).Footnote 4 Human mobility intersects with geographical disparities as people tend to move toward regions offering better employment and social mobility opportunities (Deutscher & Mazumder, 2020; Corak, 2020; Eriksen & Munk, 2020; Bell et al., 2023; Buscha et al., 2021). Extensive research has shown that certain regions serve as escalator regions, enabling individuals who migrate to these areas to experience higher levels of upward mobility and socioeconomic advancement compared to their original or other destination areas (Champion, 2012; Fielding, 1992; Van Ham et al., 2012).

Similarly, both demographic and sociological literature view internal migration as a work-related phenomenon that generally benefits occupational outcomes (Ballarino & Panichella, 2021; Mulder & Van Ham, 2005). These studies also focus on the interplay between geographical mobility and factors such as age (Clark & Withers, 2002) and family events (Impicciatore & Panichella, 2019; Kulu & Steele, 2013; Vidal & Lersch, 2019; Vidal et al., 2017), with a focus on gender differences (Brandén and Haandrikman, 2019). Specifically, these studies have shown that while men who move tend to have better occupational outcomes than those who do not, this is not always the case for women, whose occupational achievement is often negatively impacted by geographical mobility, particularly when it occurs after forming a union and transitioning to parenthood (Boyle et al., 2009; Cooke et al., 2009; De Jong & Graefe, 2008).

This disparity between men’s and women’s experiences of geographical mobility and its impact on their occupational outcomes is related to the traditional gender roles still prevalent in many European societies, where men are expected to be the main breadwinners and women to focus on childcare (cfr. Cantalini, 2020). As a result, geographical mobility is seen as a means of career advancement for men, but for women it is often a response to their partners’ movements rather than a pursuit of occupational opportunities (Taylor, 2007).

This perspective, which posits a subservient role for women in family migration decisions, is not at odds with the economic approach to mobility. The ‘tied migration’ argument (Mincer, 1978) claims that one partner (the first/lead migrant) moves and starts to gain benefits from the relocation in terms of income and/or occupational stability if these benefits exceed the costs of migration, while the other partner (the tied mover) follows her/his partner later, despite potential penalties on her/his career. The tied mover, therefore, moves when the family utility from migration is positive, even when her/his utility is negative. Although this theory is gender-neutral, the gender division in households often results in women being more frequently tied-movers, while men are more frequently first movers (Ballarino & Panichella, 2018).Footnote 5

The first aim of this study is thus to analyze the effect of the South-to-North migration on occupational status in Italy, focusing on the migration benefit/disadvantage, i.e., the effect of internal migration from the point of view of the sending society. By comparing internal migrants (movers) to those who stayed in Southern Italy (stayers), the study evaluates if the decision to migrate to central or northern regions has been beneficial in comparison to staying in the place of origin and if this benefit (or disadvantage) varies based on gender. The comparison between movers and stayers allows for a more accurate evaluation of the migration experience by considering the non-random selection of migrants from the general population of the sending society (Chiswick, 2000). To control for the selectivity of migrants based on both observed (e.g., education, age, marital status, etc.) and unobserved characteristics (e.g., ambition, career orientation, etc.), this study employs statistical matching techniques that compare the occupational status of an individual if he/she had moved from Southern to Northern Italy to the occupational status of the same individual if he/she had not moved. We expect internal migration to have a positive effect on occupational status for men (H1) and a negative effect for women (H2).

Moreover, this work aims at examining the interplay between geographical mobility and family dynamics, which are at the basis of the (possible) gendered effects of internal migration on occupational status. While there should not be relevant differences according to marital status and number of children in the migration benefit among men (H3), we hypothesise a larger migration disadvantage for women moving after the transition to union and parenthood, according to the ‘tied migration’ argument (H4).

Geographical Mobility and Cumulative (dis)Advantages: four Different Scenarios

Previous studies have only recently started to examine the impact of geographical mobility on social stratification and inequality. The question remains whether it increases or decreases the transmission of social inequality from one generation to the next. To address this, it is important to examine how geographical mobility affects occupational success differently among individuals of different social backgrounds, taking into account their varying likelihood of mobility. Empirically, this issue should be analyzed through a three-way interaction between geographical mobility, social background of origin, and occupational outcomes, asking whether the effect of mobility on the occupational career is stronger for those with a lower or higher social origin. The aim is therefore similar, mutatis mutandis, to that of social stratification research studying how different factors of advantage/disadvantage interact with each other according to the ‘cumulative advantage’ pattern (DiPrete & Eirich, 2006). Are the effects of geographical mobility the same for everyone, or do they change according to the social background of origin?

Research on geographical mobility has not given an exhaustive answer to these questions, and the research focusing on the interrelations between geographical mobility and social inequality has come to different, and often contradictory, results, which can be systematized in four different scenarios (see Table 1).

Table 1 Theoretical scenarios on the interrelations between geographical mobility and social inequality
Full size table

The first two scenarios assume that geographical mobility has a positive effect on occupational and social mobility pathways, and this is what studies on this topic have typically found among men (see above). The first scenario posits that those from better-off families can benefit from geographical mobility, which can accumulate over time to increase their advantage. This can be part of a more general process of cumulative advantage, where the advantage of one individual or group over another grows (i.e., accumulates) over the life course (DiPrete & Eirich, 2006, p. 272). Geographical mobility is part of this process if its positive effect cumulates with the family-related one, providing an ‘additional boost’ to those individuals with a better social origin and thus increasing the social distances among groups. From this perspective, internal migration can contribute to widening differentials between groups, even though it may have a positive impact on occupational success for all individuals, as its positive effect is more pronounced for those from more advantaged families (Ballarino & Panichella, 2021).

The second scenario views geographical mobility as a way for individuals with disadvantaged social background to compensate for the disadvantage related to the poor resources of the family of origin, with a balancing effect on social inequality.Footnote 6 This perspective thus considers geographical mobility a ‘strategy’ to improve employment and mobility opportunities, since it allows to enlarge the occupational opportunities of those individuals who have access to limited resources in the area of origin.

In the third and fourth scenarios, geographical mobility has a negative effect on occupational careers. If the negative impact of migration is greater among higher classes, geographical mobility may lead to a leveling-down process, making the group of movers more homogeneous and similar to those who do not move. This third scenario sees geographical mobility placing all movers in the lower strata of the occupational structure, regardless of their individual characteristics such as gender, education, skills, etc. This dynamic reduces inequalities within the group of movers by forcing them to share similar poor occupational paths in the destination region. This scenario aligns with what has been observed among international migrants, particularly in certain contexts like Italy, where highly educated foreign workers from more privileged backgrounds are the ones who face the most significant labor market penalties (Panichella et al., 2021).

In the fourth scenario, the negative effects of geographical mobility are greater for lower social classes. This scenario is rooted in the disruptions caused by the move, such as the costs and difficulties faced by the movers in finding housing and employment, and the loss of social ties. These difficulties are likely to be more disruptive for individuals with a disadvantaged social background as they have fewer resources to cope with them, with negative consequences on their working career. To overcome these costs and difficulties, these individuals may be forced to enter the labor market quickly, even accepting low-skilled jobs with limited career advancement opportunities. Consequently, the disadvantages (i.e., the disruption) implied by the geographical movement cumulate with the disadvantages given by social origin, in contrast to the cumulative advantage described above.

The last aim of our study is thus to explore the impact of geographical mobility on the intergenerational reproduction of social inequalities. Since the literature analyzing the interrelations between geographical mobility, social origin and occupational success is scarce, it is difficult to identify specific research hypotheses concerning the Italian case. For this reason, the empirical analysis will be conducted in an ‘explanatory way,’ analyzing which of the abovementioned scenarios is more plausible for the Italian case. To this aim, we interacted the geographical mobility with the social class of origin. If the South-to-North migration gives stronger benefits to individuals with a high social origin, then geographical mobility cumulates its advantage to the one related to the socioeconomic background. If the positive effect of migration is stronger among those with a low social origin, then it follows a compensation pattern. If internal migration has higher negative consequences for individuals from the higher classes, then mobility fosters a leveling-down process. Conversely, if these negative consequences are stronger among individuals from the lower classes, then the effect of geographical mobility is consistent with a scenario of disruption.

Data and Variables

Data

In this study, we utilize data from the Italian Household Longitudinal Survey (IHLS), which is a panel survey conducted in five waves between 1997 and 2005. The first wave of the survey, conducted in 1997, gathered information on all significant events that occurred between the interviewees’ birth and the date of the interview. The subsequent surveys updated this information by recording significant events that occurred between the previous interview and the date of the subsequent interview.

The IHLS data provide in-depth information on geographical mobility (at the municipality level), educational and occupational careers, and family and reproductive behaviors of a representative sample of the Italian population. The use of panel data in this analysis is particularly useful as it contains time-varying information on the populations of origin and destination, which allows for the examination of the selective nature of spatial mobility, its impact on individuals’ outcomes, and whether it leads to structural changes in the populations of origin and destination (Brimblecombe et al., 2000; Vidal & Lersch, 2021). Furthermore, information on the municipality was matched with administrative data that contain geographical and demographic features, such as the number of inhabitants and altitude of the municipality of residence.Footnote 7 This allows for the matching and comparison of individuals with the same characteristics and who reside in similar geographical environments (see below).

Our analytical sample consists of all individuals born between the years 1925 and 1975. To accurately estimate the effect of migration on occupational status, only those individuals who lived in the South at the age of 15 were considered. The window of observation was 15–60 years of age or from 15 to the age at interview if the case was right-censored, covering the years 1940 to 2005. After a listwise deletion of missing values, the analysis is based on the monthly individual history calendars of a total of 2,800 individuals (1323 males, 1477 females).

Variables

The dependent variable in our analysis is the Standard International Occupational Prestige Score (SIOPS) (Ganzeboom et al., 1992). This score represents a rank of groups of occupations (i.e., occupational categories) based on their internal homogeneity in terms of material and symbolic advantages and their external heterogeneity compared to other groups. In our sample, the SIOPS score ranges from 9 (for occupations such as assembly laborers and miners) to 90 (for professions such as civil engineers, dentists, and CEOs in large firms) with an average score of 36.0 (standard deviation of 18.8). This average score represents occupations such as customer service clerks and welders.

The independent variable in the analysis is geographical origin, which is divided into two categories: southern Italian migrants and southern non-movers. This variable is treated as time-varying. At the time of the first observation, when the individual was 15 years old, they were considered a non-mover if they lived in the South. If a southern resident later moved to Central-Northern Italy, they became an internal mover. This meant that they were originally living in the southern regions at age 15, but they moved at least once to the Central-North area within the observation window.

The IHLS recorded 623 instances of mobility, and by the end of the observation period, 18.6% of individuals from the South had moved at least once to the North. Permanent settlement in the North is the preferred option for those facing limited employment opportunities in the South, with 61% of southern movers settling permanently in the North (see Impicciatore & Panichella, 2019 for details). Return migration to the South and commuting are included in residual categories, but their results are not reported due to the high level of uncertainty in the estimates.

In order to control for potential confounding effects and increase the validity of our results, we also included a set of time-varying variables in our models. These variables are age, period, marital status (single, married or cohabiting, separated or divorced, and widowed), education (lower secondary, upper secondary, tertiary), and the number of children (childless, one child, two children or more).

Methods and Empirical Strategy

A combination of Fixed Effects (FE) models and matching techniques was used on panel data to estimate the effect of geographical mobility on occupational outcomes. This approach allows us to overcome most limitations of observational studies and to create a quasi-experimental design. By controlling for all relevant characteristics before the movement and observing the intra-individual changes over the life course, we are able to simulate a controlled environment, reducing the risk of selection bias and improving the accuracy of our results.

‘Pruning’ Observations with the Coarsened Exact Matching (CEM)

Before estimating the statistical models, we employed the Coarsened Exact Matching (CEM) procedure, which is a monotonic imbalance bounding matching method that balances the treated (movers) and control groups (non-movers) with a set of pre-treatment variables (Iacus et al., 2012). We matched movers and non-movers according to sex (male or female), birth cohort (1930–39, 1940–49, 1950–59, 1960–69, 1970–74), and social class of the parents (non-manual workers, urban petit bourgeoisie, urban working class, and agricultural classes). We also matched the two groups according to the type of municipality of residence at age 15 (city, plain, or hill), thus combining information on the number of inhabitants and the altitude of the municipality and comparing movers and stayers who shared a similar geographical environment of origin.

CEM emulates a fully blocked randomized experiment with the goal of reducing the imbalance in the empirical distribution of pre-treatment confounders between the treated and control groups by ‘pruning observations’ from the data (King & Nielsen, 2019). CEM temporary coarsens each explanatory variable (X) into substantively meaningful groups, and then it applies the method of exact matching to the coarsened data to determine the matches and to ‘prune’ unmatched units. In other words, after coarsening, the CEM procedure creates a set of strata, each with the same value of X. Units in strata that contain at least one treated and one control unit are retained, whereas the others are excluded (i.e., ‘pruned’) from the sample.

CEM requires to use a weight in the analysis, which adjusts for the different stratum sizes (i.e., different numbers of control units are matched to each treated unit across strata). Specifically, to each matched unit i in stratum s, the following weight is assigned:

$${w}_{i}=\left\{\begin{array}{c}1,\\ \frac{{m}_{C}}{{m}_{T}}\frac{{m}_{T}^{s}}{{m}_{C}^{s}}\end{array}\right. \genfrac{}{}{0pt}{}{i{\in T}^{s}}{i{\in C}^{s}}$$

where \({T}^{s}\) and \({C}^{s}\) denote the treated and control units in stratum s, respectively, whereas \({m}_{T}^{s}\) and \({m}_{C}^{s}\) refer to the number of treated and control units in the stratum, respectively. The number of matched units are \({m}_{T}={\cup }_{s\in S}{m}_{T}^{S}\) for the treated and \({m}_{C}={\cup }_{s\in S}{m}_{C}^{S}\) for the controls. Therefore, unmatched units receive weights \({w}_{i}=0\).

By ‘pruning’ the observations included in those strata in which there is not (at least) one treated and one non-treated, and weighting all the other observations based on the number of observations of the two groups in their stratum, this procedure allows to obtain more efficient estimates of the effect of geographical mobility on occupational status, lowering the degree of unbalance, reducing model dependence, and improving the statistical power of the analysis (King & Nielsen, 2019). In general, the CEM procedure enables to get an exact balance on the observed covariates (in our case, sex, birth cohort, etc.) and an on average balance for the variables that are not measured (motivation, intelligence, skills, etc.).

The Fixed Effects Panel Model

We combined the CEM procedure with the estimation of fixed effects (FE) panel models, a popular approach in research examining the impact of geographical mobility on outcomes such as employment status, income, and subjective well-being (Vidal & Lersch, 2021). By focusing only on within-person variation (i.e., changes in the individual over time), FE models can provide unbiased estimates even in the presence of self-selection based on (time-constant) unobservable characteristics (such as ability, personality traits, or motivation), which can affect both the likelihood of moving and occupational status. For example, if individuals with higher occupational status are more likely to move (perhaps due to greater career orientation or ambition), the data may be biased by person-specific unobserved heterogeneity unless FE models are estimated. These models discard between-person variation and only exploit within-person variation (Brüderl & Ludwig, 2015).

The basic model is as follows:

$${Y}_{it}= \beta \left({GM}_{it}\right)+ {\sum }_{k=1}^{40}{\theta }_{k}\left({A}_{it}\right)+\sum_{j}{\lambda }_{j}({X}_{jit})+ {\varepsilon }_{it},$$

where \(\beta\) captures how the outcome, in our case the occupational status (\({Y}_{it}\)), changes after an individual change in the main independent variable, in our case geographical origin (\({GM}_{it}\)), net of individual changes in the observed time-varying covariates (\({X}_{jit}\)) and in age (\({A}_{it}\)). Hence, the model focuses on intra-individual changes over time, estimating the effect of geographical mobility on achievement within the individuals’ life course and considering only time-varying variables. The restrictive assumption of this model is that the unobserved covariates are time-constant.

In this study, we employ a three-step empirical approach to examine the effect of geographical mobility on occupational status. First, we estimate two linear FE models, one controlling only for age and period effects and the other including two time-varying variables (marital status and number of children). The models are estimated separately for men and women. Second, we estimate a model with an interaction term between geographical mobility and the family condition (based on marital status and the presence of children) at the moment of migration. This model allows us to test the hypothesis that women may face a greater penalty when they are in a union and have children at the time of mobility.Footnote 8 Finally, we examine the effect of geographical mobility on occupational status over different social classes of origin and by sex. To do this, we estimate full FE models separately for each social class of origin and sex, which provides insight into how geographical mobility contributes to social stratification and inequality.

Empirical results

Uncovering the Characteristics of South-to-North Internal Migrants in Italy: An Analysis Using the Coarsened Exact Matching (CEM) Method

Figure 1 displays the bounds in which matched comparisons are performed through the CEM procedure, providing clarity on the portion of the dataset that is used in the analysis. For example, the stratum defined as men born in the 1940s with non-manual occupation parents and residing in hill regions (indicated by a cross in the second panel of the figure) comprises 2 control units and no treated units, and thus, these units are ‘pruned’ from the analysis. This is also evident in Fig. 2, which shows the distribution of treated units across the strata determined by the CEM procedure. The larger the circle representing a stratum, the higher the number of treated units in that stratum. This figure also sheds light on the selection process behind South-to-North migration in Italy, which was primarily dominated by male individuals born in the 1930s and 1940s, coming from agricultural backgrounds and rural areas. Most of this migration took place during the 1950s and 1960s, where the ideal South-to-North migrant was a low-educated male from rural Southern Italy moving to an industrial city in the North and joining the working class (Panichella, 2014).

Fig. 1
figure 1

Note: The CEM procedure creates a set of strata, each with the same value of the explanatory variables of interest, then it retains in the analyses only those strata that contain at least one treated and one control unit, excluding (‘pruning’) all the others. The figure shows the strata containing units that are retained (circles) and those that are pruned (crosses). For instance, units in the stratum defined by men, born in 1931–40, from agricultural classes, living in hills are kept in the analysis, as it contains both movers and stayers: these units are thus exactly matched according to those variables. Conversely, units in the stratum defined by men, born in 1931–40, from non-manual classes, living in plain are excluded from the analysis, as it does not contain both movers and stayers

CEM Procedure Results: Comparison of Matched and Unmatched Layers by Sex, Birth Cohort, Social Class of Origin, and Type of Municipality (represented by Circles and Crosses respectively) Source: Own elaboration on Italian Household Longitudinal Survey (1997–2005).

Full size image
Fig. 2
figure 2

Source: Own elaboration on Italian Household Longitudinal Survey (1997–2005). Note: The circles correspond to the number of treated units (movers) found in each stratum defined by the CEM procedure (see also Fig. 1). Each circle is weighted according to the number of movers in the stratum, i.e., the larger the circle, the larger the number of movers in the stratum. Circles, by definition, are not presented for those strata that do not contain at least one mover

Distribution of Treated Units (Movers) Among Strata Defined by CEM Procedure: A Stratification Based on Sex, Birth Cohort, Social Class of Origin, and Type of Municipality.

Full size image

However, despite a larger migration of low-skilled individuals driven by the growth of Fordist industries in northern regions, internal migration also included a substantial number of highly educated and skilled individuals (Panichella, 2012). This is a persistent aspect of Italian internal migration, both during the post-war economic boom and in recent decades (Panichella, 2014), as evidenced by the significant number of treated units (i.e., movers) from upper-class families in cities (Fig. 2). Furthermore, the demographic and employment characteristics of internal migrants have become increasingly diverse over time, reflecting changes in the economy characterized by a shift toward the service sector and a decline in Fordist industries. Among recent migration cohorts, individuals from socially privileged backgrounds in urban areas of the South have become a more prominent group, particularly among men.

Internal Geographical Mobility, Family Dynamics and Occupational Status

Table 2 highlights the impact of internal geographical mobility on male and female occupational status using the SIOPS scale. Model 1, which includes only age and period control variables, reveals contrasting effects of South-to-North migration for men and women. The analysis shows that male migrants experience a 2.70-point increase in occupational status, while female migrants experience a 0.61-point decrease. Despite being statistically significant, these effects are relatively small in magnitude, particularly for women.

Table 2 Geographical mobility and occupational attainment. FE models. Beta coefficients and 95% confidence intervals
Full size table

The results from Model 2, which takes into account civil status and number of children, show no significant changes in the coefficient for men’s geographical mobility (\({\beta }_{Male}=\) 2.60), confirming the effect of internal movement not to be mediated by family variables. The same conclusion applies to women, although their negative effect of geographical mobility is more pronounced (\({\beta }_{Female}=-\)1.23). The results suggest that after accounting for family events, men’s occupational status would increase from 37.1 to 39.6 points in the SIOPS scale after internal migration, while women’s status would decrease from 38.0 to 36.5, confirming a migration benefit for the former and a migration penalty for the latter.

Despite a small positive effect of being married on occupational status among women, coefficients related to family events—primarily, the birth of children—are consistent with the vast literature on motherhood penalties and fatherhood premia, pointing to (slightly) positive impacts for men and (slightly) negative impacts for women (Cantalini, 2019). Moreover, the parenthood effects are small for women because the largest penalty related to motherhood occurs in terms of employment and labor market participation in Italy (Cantalini, 2020, 2022), where many mothers tend to interrupt their careers temporarily or permanently after having children. As a result, mothers who remain employed are likely to be positively selected, making the motherhood penalty in occupational status less noticeable. In general, this evidence confirms, also for the South-to-North internal migration in Italy, that a positive effect of geographical mobility on occupational outcomes is confirmed only as regards men, whereas the effect of mobility is negative for women. This outcome was somewhat expected, as migration is a gendered phenomenon.

To further understand the contrasting impact of South-to-North internal mobility across genders and its relationship with family dynamics, Fig. 3 examines how the effect of geographical mobility varies based on marital status and the number of children at the time of migration. The movers are divided into three groups: those who moved alone, those in a union without children, and those in a union with at least one child. The size of the circles represents the percentage of movers in each group among men and women computed on the whole sample (total percentages). This information provides a clear understanding of which groups of migrants contribute more to the migration benefits or disadvantages and highlights the importance of family dynamics on the impact of mobility on occupational status.

Fig. 3
figure 3

Source: Own elaboration on Italian Household Longitudinal Survey (1997–2005)

Geographical mobility and occupational status, by marital status and number of children at migration. FE models. Beta coefficients and 95% confidence intervals.

Full size image

Among men, South-to-North migration has a positive effect on occupational status regardless of marital status. This means that moving to the central-northern regions increases occupational status for both single men and those in a union, with or without children. However, the positive impact of geographical mobility is significantly greater among men in a union with children, who experience a 5.1-point increase in their occupational status on the SIOPS scale. This type of mobility, which often involves economic and psychological costs (Impicciatore & Panichella, 2019), including a negative impact on the employability of the female ‘tied mover’ partner, is undertaken when men perceive good employment prospects in the destination region. For example, this type of migration results in an occupational benefit similar to moving from farm laborer (23) to papermaking plant operator (28) or from receptionist (38) to travel consultant (43). However, the large benefit for this group does not contribute much to the male migration benefit at the aggregate level (\({\beta }_{Male}=\) 2.60, see above), since this type of internal mobility is much less common than movements involving single men.

The impact of geographical mobility on female occupational status varies greatly based on their marital status and presence of children, highlighting the close relationship between migration, couple formation, and parenthood for women. Mobility has a positive effect only for women who were single and without children at the time of migration, while it has no effect for women who moved in a union and a negative impact for women who moved after marriage and having children. Therefore, women can reap the benefits of migration only when they move alone, without being in a relationship and, particularly, without having children. This finding echoes what was observed among men: the geographical mobility of a family is often aimed at improving the employment opportunities of men, even though it has a detrimental effect on the ‘tied’ women. Additionally, it is important to note that even in cases where geographical mobility has a positive impact on women who move alone, their benefit is still lower than that of men who migrate alone (\({\beta }_{MenAlone}=\) 2.5; \({\beta }_{WomenAlone}=\) 1.5), suggesting that there are additional mechanisms of disadvantage for women that go beyond family dynamics and relate to their integration into the northern Italian labor market.

These findings demonstrate the prevalence of ‘tied-migration’ among women, where internal migration does not result in an improvement in occupational status when it occurs after the formation of a union or after becoming a parent. The impact of family dynamics is also amplified by the unequal distribution of men and women in the three categories. Men are more likely to migrate before marriage (80.6%), while migration after union formation is less common, both with (15.6%) or without children (3.6%). Among women, the distribution between the three categories is more balanced: 49.4% of women migrate alone, but a significant proportion also migrate after forming a union (34.5% without children and 16.1% with children).

Internal Mobility and Social Inheritance

We now turn to the third part of the empirical strategy, focusing on the relationship between geographical mobility and the reproduction of social inequalities. Figure 4 presents the effect of South-to-North migration on male and female occupational status based on their social class background. This interaction enables us to determine whether the impact of geographical mobility is uniform for all individuals, regardless of their social class, or if it has varying effects depending on their family’s resources.

Fig. 4
figure 4

Source: Own elaboration on Italian Household Longitudinal Survey (1997–2005)

Geographical mobility and occupational attainment, by social class of origin. FE models. Beta coefficients and 95% confidence intervals.

Full size image

As in Fig. 3, the size of the circles is proportional to the percentage of movers in the three categories among men and women, and it confirms that this type of internal migration primarily involved individuals from the lower classes (see also Fig. 2).Footnote 9

Results indicate that there is a consistent difference in the impact of South-to-North migration on male occupational status based on their social class of origin. Men from upper-class backgrounds experience a much higher benefit from the migration compared to those from medium or lower classes. After migration, men from upper-class backgrounds see an increase in their occupational status by 6.75 points on the SIOPS scale, whereas those from lower classes only see an increase of 1.72 points. This highlights a cumulative advantage scenario, where individuals from higher social classes reap greater benefits from migration. In other words, the benefit experienced by those from upper classes is akin to advancing from a role as a production manager to that of a legislator or psychologist. However, for those from lower classes, the impact of mobility is more limited, corresponding to a transition from being a bricklayer to a metal melter.

Therefore, South-to-North migration appears to play a role in the process of cumulative advantage, providing an extra advantage to men from higher social strata and potentially widening the gap between social classes. Despite the limited number of individuals involved in this type of migration, as indicated by the size of the circles in the figure, it can still have a relevant impact on the intergenerational transmission of social stratification. Our results suggest that those from upper classes are more likely to benefit more from migration, which could be due to factors such as access to trans-regional social networks and the higher likelihood of moving for education opportunities in a different region (Panichella, 2013; Tosi et al., 2019).Footnote 10

There is not a clear gradient associated with social origin among women. The negative impact of migration is observed for women from both upper (\({\beta }_{FemaleUpper}=-\)1.99) and the lower classes (\({\beta }_{FemaleLower}=-\)1.56), whereas it is negligible for the children of the middle class (\({\beta }_{FemaleMiddle}=-\)0.33). Hence, differently from men, there is limited variability in the negative effect of geographical mobility among individuals with different social classes, as no clear patterns have emerged to align with the scenarios described in the previous paragraph. Instead, our analysis reveals that the effect of geographical mobility on occupational attainment for women is influenced by family dynamics. It is therefore mainly within the family that geographical mobility can play a role in reproducing social stratification between generations.

Conclusion

The birthplace or place of residence of an individual has a significant impact on their occupational opportunities and social mobility. This is because the geographical division of labor shapes the occupational structure(s) and the characteristics of the local labor market, thereby determining the opportunities available for individual careers (Moretti, 2012; Chetty et al., 2014). However, individuals are not limited to their place of origin, as they can choose to move to another location with better opportunities, despite the costs they may incur in doing so that could affect their chances of success.

This work studied the effect of South-to-North internal migration on occupational status in Italy using a combination of fixed effects linear regression panel models and coarsened exact matching (CEM). By considering internal migration as a treatment and adopting an empirical strategy that allowed us to control for the selection processes behind such geographical movement, we aimed at investigating if male and female occupational status benefited from South-to-North migration, and if the (possible) migration benefit or disadvantage changed according to the family status and the social class of origin.

In answering our first research question—whether moving to central-northern regions offers an occupational advantage compared to staying in the southern regions—our findings revealed gender-specific effects of this type of internal migration, consistent with previous research on geographical mobility and job outcomes (Cooke, 2008; Kulu & Milewski, 2007). Our results showed that only men benefit from migration, with an average increase of 3 points in the prestige scale for their occupational status (H1 supported), while women face a slight disadvantage, experiencing a decrease of over 1 point in their average occupational prestige (H2 supported). Female penalization actually occurs only for those women moving after the union formation and the transition to parenthood (H4 supported). On the other hand, women who move alone may still benefit from migration, although the benefit is lower than that of men. This suggests the existence of additional disadvantages for female movers, unrelated to family dynamics and possibly rooted in structural macro-factors surrounding South-to-North migration in Italy. In fact, studies on the ‘great internal migration’ of the 1950s and 1960s showed that men mainly entered the large Fordist industry, benefiting from a consistent improvement in employment compared to the agricultural work they carried out in the South (Reyneri, 1979), while southern women mainly entered in the most marginal, unstable, and unskilled areas in the labor market, such as in the services and care sector (Panichella, 2014). Additionally, the migration benefit for men is not impacted by family events, though it is larger for those who are in a union with children at the time of the migration (H3 partially supported). However, it should be noted that only a small proportion of men migrated internally after marriage and parenthood, so the larger benefit for this group may be due to selection bias.

These results confirm that the interrelation between mobility and family dynamics is crucial only for women and explains a substantial part of their occupational disadvantage in the region of destination. In the Italian society, indeed, traditional gender norms prevail, which consider men as the main breadwinners and women as the main caregivers (Cantalini, 2020). Women’s geographical movements are thus often a response to their partners’ movements, as a result of a subordinate role of women in family relocation decisions, and they are more frequently tied-movers, being occupationally penalized after geographical mobility.

Our study also found that geographical mobility has a differing impact on men based on their social class of origin, thus contributing to the reproduction of social inequalities. Indeed, men from upper classes experience greater migration benefits than those from middle and lower classes, leading to a cumulative advantage scenario. This means that geographical mobility adds to the already existing advantages of upper-class men, widening the social gap between classes. Conversely, no clear gradient was found between social origin and the impact of mobility on women, who face equal disadvantages regardless of their class background. These findings complement those of Bernard (2023), albeit through a distinct approach, thereby underscoring the need to advance the analysis of the interrelation between geographical mobility and social inequality in forthcoming research. Additionally, our results suggest that the interrelations between migration and family dynamics appear to be more important than those between migration and social origins for women. Further studies using large-scale longitudinal data should examine this issue more thoroughly, including the possibility that women from better-off families may escape the ‘tied movement’ disadvantage.

In conclusion, our findings indicate that geographical mobility plays a limited role on intergenerational reproduction of social stratification. Firstly, the impact of mobility varies based on social origin among men, while family dynamics play a crucial role among women. This aligns with previous research on the occupational integration of female international migrants, who are disadvantaged due to the interplay between migration and family (Taylor, 2007). Secondly, while geographical mobility does result in a cumulative advantage scenario for men from upper classes, the limited number of such individuals among movers reduces its overall impact on the reproduction of inequalities. However, this conclusion is specific to the Italian context where low-educated individuals from rural areas primarily engage in internal mobility. In contexts with higher social selectivity among movers, the cumulative advantage pattern could have a greater impact on reproducing inequalities.

To further explore the role of geographical mobility on social inequality, this paper should be expanded in two key directions. On one hand, it is crucial to encompass more intricate and diverse forms of geographical mobility, such as commuting or returning to one’s place of origin. Return migration, for instance, might arise from either a failed movement or a deliberate strategy (Gillespie et al., 2021), both of which have a crucial impact on life trajectories and the reproduction of social disparities. On the other hand, it is necessary to compare the relationship between internal migration and social stratification across different institutional contexts. By doing so, future research can contribute to a more comprehensive understanding of the role of geographical mobility in reproducing social inequalities.