CAIRN-INT.INFO : International Edition
Volume
75
format_quote Cite share Share Switch to French

Full-text

Where there is poverty, internal or international migration is often a means to improve families’ living standards. These movements are challenging to observe and understand, requiring updated data collection instruments; for example, multisite surveys at places of origin and destination, or the use of geolocation. Here, the authors use the GPS coordinates of individuals’ places of departure and arrival in 2006 and 2012 to study internal mobility in Senegal. They observe highly gendered patterns in migrants’ reasons for departure, distance travelled, and places of destination.

1Gender equality and women’s empowerment are consequential for economic development. While much attention has been paid to the importance of equal-gender access to education (Abu-Ghaida and Klasen, 2004) and healthcare (Bloom et al., 2014; Stenberg et al., 2014), and to improving women’s bargaining power within their household (Duflo, 2012), the determinants and impacts of gender differences in access to migration remain to be investigated extensively, particularly for sub-Saharan Africa. Though there may not be much difference between genders in the probability of migrating, women may be more constrained than men if they do not move for the same reasons or under the same conditions. Dating to the early 1990s (Chant et al., 1992), the pioneering works on female migration suggested that, in most developing countries of that time, it was largely associated with family events and had few connections with labour market participation, unlike male migration. Is this still the case? The question appears particularly relevant because unequal access to and variations in motives for migration may have adverse effects on women’s educational investment and empowerment.

2This paper studies the gender-specific patterns and determinants of internal migration and distance travelled in Senegal. It contributes to the literature in several ways. First, avoiding the limitations and constraints of using administrative units to define migration (Bell et al., 2015), we use individual panel data from a nationally representative survey with GPS coordinates to track individuals within the country’s boundaries. Migration distance is calculated precisely, and individual mobility is mapped with cartographic tools. Secondly, we highlight gender differences by documenting the demographic and economic variables that correlate with migration decision-making and distance, at individual, household, and regional levels.

I – Literature and context

3Migration has long been recognized as necessary for economic development. While early works by economists (Lewis, 1954; Todaro, 1969) insisted on migration as resulting from an individual’s search for labour, subsequent research emphasized migration as a means to diversify family income sources and to smooth consumption against negative income shocks (Stark, 1980; Stark and Lucas, 1988). But migrating for work may not be the only way to do so. Seeking to explain marriage-related mobility patterns, Rosenzweig and Stark (1989) showed that, in rural India, families may use marital arrangements as insurance schemes when insurance markets are incomplete and risks are spatially covariant. Families with more variable profits from agriculture tend to marry their daughters to more distant partners as part of an implicit interhousehold contract aimed at mitigating income risks and smoothing consumption through remittances.

4Gender differences in access to and forms of migration have been increasingly documented over the last decades (Donato et al., 2006). Many studies from various disciplines have stressed the limited geographical mobility of women, explained by gender roles or family constraints (Kanaiaupuni, 2000; Assaad and Arntz, 2005; Massey et al., 2006; Chort, 2014). In developed countries, the rationales for migration differ depending on distance travelled. Whereas short-distance mobility is associated with housing and life-cycle motives, long-distance migration is driven by employment motives (Cordey-Hayes and Gleave, 1974; Clark and Huang, 2004; Niedomysl and Fransson, 2014).

5Africa has been under-represented in this literature. Moreover, a strong focus on international migration may have somewhat overshadowed more modest—and maybe more feminized—internal moves. Since the 1980s, the internal migration of women has intensified in Senegal and in other sub-Saharan countries (Antoine and Sow, 2000). Although in the region’s patrilocal societies, female internal mobility is high and largely associated with marriage (Kudo, 2015), studies focusing on the recent period have presented a more nuanced picture. For Mali, Lesclingand and Hertrich (2017) found an almost continuously increasing proportion of women migrating for work, starting with cohorts born since 1954, such that they now outnumber men. For Ivory Coast, Comoe (2013) showed that women’s mobility is almost exclusively driven by family motives, which puts into question women’s alleged autonomy in migration decision-making. Similarly, for Senegal, Duboz et al. (2011) demonstrated that family motives are predominant among women migrating to Dakar. Vause and Toma (2015), in a comparative analysis of Senegal and the Demographic Republic of Congo, found that Senegalese women are less likely to move abroad than Congolese women, and less likely to migrate independently, possibly due to the more rigid patriarchal system prevailing in Senegal.

6Contributions to this literature have generally relied on retrospective data to reconstruct migration trajectories (Lesclingand and Hertrich, 2017; Vause and Toma, 2015; Herrera and Sahn, 2013). These data are subject to recall and selection biases, or they pertain to a single region (Dakar for Duboz et al. (2011) and Vause and Toma (2015)). Moreover, most surveys and censuses collect only place of birth and current (and sometimes previous) residence, which can obscure movements due to step [1] or circular migration. [2] A second problem is that the ‘places’ reported may not be sufficiently specific; for instance, the region or department (département) may be given, but not the village (Bell et al., 2015). This lack of detail impacts the measurement of flows because those who move within the same administrative unit remain unnoticed.

7Avoiding these problems by using nationally representative survey data and GPS coordinates to measure migration, this paper thus provides insight into the dynamics of migration and distance within the context of a developing country (for an overview, see Lucas, 2016), thereby contributing to the sociological and anthropological literature on gender differences in access to migration in sub-Saharan Africa.

II – Data

1 – The Poverty and Family Structure individual panel survey

8Our data come from the two waves of the Poverty and Family Structure survey (PSF), conducted in Senegal in 2006–2007 and 2010–2012. [3] The first wave’s sample is nationally representative and includes 1,750 households (14,450 individuals) from 150 randomly drawn census districts. All individuals surveyed in the first wave were tracked, except when abroad, forming an individual panel. The attrition rate between the two waves is 11.6%. [4] We address attrition-related issues in Section 3.2.

9The PSF surveys are particularly suited to studying internal migration since they provide individuals’ exact locations through GPS coordinates. With this information, we can calculate great-circle distances, i.e. the shortest distance between two points on a sphere, between first- and second-wave locations. While the geography of Senegal makes it complex to use great-circle distances, due to the position of The Gambia along the Gambia River, this method of measurement is the most appropriate and accessible way to compute internal migration distances (Bell et al., 2002). In addition, most observed mobility from and to the area south of The Gambia (the Casamance, which includes the Ziguinchor and Kolda regions) is connected to Dakar, as shown in Figure 1. Dakar and Ziguinchor have been reconnected by ferry since 2005 (after the dramatic sinking of Le Joola in 2002), and few travellers choose the land route. The great-circle distance thus seems to be a relevant proxy for the travel distance even from and to the regions of Senegal south of The Gambia.

10Furthermore, the PSF data contain rich information on individual and household sociodemographic characteristics, which allows us to finely document the determinants of internal migration. Specifically, consumption data are collected for each household subgroup, or cell. Cells are semi-autonomous consumption units comprising a cell head and all her dependents (in particular, her children, foster children, and widowed mother or father). The average number of cells per household is 2.5. [5] We are thus able to account for consumption at both the household and cell levels. In all our regressions, we include variables for household and cell sizes and for the consumption share of the individual’s cell relative to total household consumption.

11Local communities in the PSF data are classified as urban or rural, based on the categories used by the Senegalese national statistical agency. We use these categories to provide a finer description of mobility patterns. We also used data from a 10% sample of the 2002 Senegalese census to calculate indicators of poverty and inequality [6] at the department level, based on 2006 administrative boundaries. [7]

Figure 1

Individual moves by gender between PSF survey waves, Senegal

Figure 1

Individual moves by gender between PSF survey waves, Senegal

Interpretation: Each line stands for one individual; dots represent destinations.
Note: Sample size is 556 (326 women, 230 men).
Source: Poverty and Family Structure survey, 2006–2007 and 2010–2012; authors’ construction.

2 – Descriptive statistics

12Our analysis focuses on individuals aged 15 and older, as younger individuals’ mobility is more likely to be determined by their parents or linked to child fostering. Therefore, the initial database is reduced to a panel of 6,941 individuals (or 8,356 including attritors, individuals deceased between the two waves, and those who moved abroad). To avoid the problems inherent in using administrative geography, our definition of internal migrants is based on the great-circle distance between the two locations calculated from recorded GPS coordinates. We use 5 km (approximately an hour’s walk) as a lower bound for internal migration because this distance, in Senegal, is sufficient to incur costs. However, results are robust to using a 10-km cut-off. Based on great-circle distances and excluding intra-Dakar migrants, 556 individuals moved more than 5 km between the first and second waves. [8] We construct an alternative measure of distance using tools provided by Google to calculate road distance between first- and second-wave locations. The two measures are strongly correlated, [9] and regression results using road distances (available upon request) are very close to our main results using great-circle distances.

13Internal migrants account for 10.4% of individuals tracked in the panel. In our sample, they represent 63.7% of all migrants (556 internal vs. 317 international), making internal migration a significant phenomenon. Table 1 reveals that internal migrants are more likely than non-migrants to be women (59% vs. 55%), tend to be better educated (27% have a secondary education or higher vs. 21%), and are younger (27 years old on average vs. 35).

14Regarding geography, mobility within the region of Dakar is especially high and related to urban expansion (Beauchemin and Bocquier, 2004). Our data demonstrate that Dakar is an important starting point and destination. Indeed, 20% of internal migrants came from Dakar, and 26% moved there between the two survey waves, thus representing 46% of all internal migrants within a metropolitan area that accounted for 20% of the Senegalese population in 2002 (ANSD, 2006).

15Figure 1 represents male and female mobility between the two survey waves, based on GPS coordinates. The numerous lines converging towards Dakar illustrate its appeal for both men and women. However, female migration patterns are more diverse and include more short-distance and rural-to-rural moves, whereas male migration is almost exclusively directed towards Dakar.

16Figure 2 represents the distribution of migration distances over 5 km by gender. The distributions are significantly different at the 5% level according to a Kolmogorov–Smirnov test. Women are more likely to move over short distances, while men are over-represented in long-distance moves.

Table 1

Descriptive statistics of non-migrants and internal migrants in Senegal

Table 1
(1) (2) (3) (4) Non-migrant Internal migrant Total t\|² diff(1) - (2) Individual characteristics Male (%) 45 41 44 Age (years) 35.12 27.01 34.47 *** Fostered (%) 14 16 14 * First-born (%) 34 31 33 Ethnicity (%) *** Wolof/Lebu 42 38 42 Serer 12 17 13 Fulbe 27 23 27 Diola 5 7 5 Other 13 16 14 Relationship to household head (%) *** Head 21 9 20 Spouse/parent 23 11 22 Child/sibling 37 51 38 Other/non-relative 19 29 19 Education (%) None 37 38 38 Primary 25 24 25 Secondary/university 21 27 24 Quranic 16 10 13 Occupation (%) *** Agricultural worker 13 11 13 Self-employed/employer 15 8 14 Salaried worker/intern/trainee 13 10 13 Family worker 7 9 7 Student 6 12 6 Inactive/unemployed 29 25 29 Missing 18 26 19 Household characteristics Equivalized consumption (XOF, in thousands) † 496.55 667.86 510.3 Household size 11.31 10.91 11.27 Cell size 4.32 4.64 4.35 ** Cell/household expenditure 0.46 0.51 0.46 *** Community of origin characteristics (%) *** Dakar 35 20 33 Other cities 24 26 24 Rural 41 54 42 Department of origin characteristics Headcount poverty (Dakar/urban/rural) 0.45 0.49 0.45 *** Gini index 0.44 0.43 0.43 *** N observations 6,385 556 6,941 6,941

Descriptive statistics of non-migrants and internal migrants in Senegal

† Consumption per adult equivalent (619 XOF [West African CFA franc] ≈ 1 USD).
Note: To facilitate comparison between groups, all means are calculated at the individual level because households may include both migrants and non-migrants. Large households thus tend to be over-represented in the reporting of descriptive statistics for household variables.
* p < .10, **p < .05, *** p < .01.
Source: Poverty and Family Structure survey, 2006–2007 and 2010–2012.
Figure 2

Internal migration distance by gender, Senegal

Figure 2

Internal migration distance by gender, Senegal

Note: Sample size is 556 (326 women, 230 men).
Source: Authors’ calculations based on the Poverty and Family Structure survey, 2006–2007 and 2010–2012.

17Figure 3 shows the distribution of reasons for entering new households in Wave 2 across distance travelled for all internal migrants. It reveals a striking contrast between women’s and men’s reasons for moving. Regardless of distance, 45%–65% of women’s mobility is explained by marriage and patrilocality, whereas employment and education are marginal motives and appear only for medium to long distances. Some individuals move due to personal difficulties, such as illness and family problems. By contrast, 40%–66% of male migration over 50 km is explained by either education or employment; for short distances, 20%–30% returned to their households of origin after seeking employment somewhere else.

III – Empirical approach

1 – Empirical models

18We explore the role of individual variables such as gender, age, education, and socio-occupational category. Emphasizing the household dimension of the decision to migrate, Stark and Bloom (1985) and Rosenzweig and Stark (1989) considered individuals’ relative position in the household. Likewise, we control for the relationship to the household head, but also for the birth rank among siblings because research has shown that elders in Senegal are more likely to migrate due to remittance expectations (Chort and Senne, 2015). We also investigate relative deprivation as a potential driver of migration (Stark, 1984) at two levels. First, at the household level, we exploit the rich data on consumption disaggregated at the cell level to proxy for the relative economic status and/or bargaining power within the household. Secondly, we explore the impact of inequality at the department level, which we expect to be positively related to out-migration.

Figure 3

Reasons for migration by gender and distance, Senegal

Figure 3

Reasons for migration by gender and distance, Senegal

Note: Sample size is 556 (326 women, 230 men).
Source: Authors’ calculations based on the Poverty and Family Structure survey, 2006–2007 and 2010–2012.

19We estimate a probit model for the decision to migrate, in which our dependent variable is a dummy equal to 1 if individual i migrated internally between the two survey waves. Internal migrants are defined as individuals who, at Wave 2, had moved more than 5 km from their first-wave location. Intra-Dakar migrants are considered non-migrants. Lacking exhaustive retrospective information on individual migration trajectories, we cannot account for temporary mobility between the two waves, i.e. individuals who left and moved back to their household of origin.

20Explanatory variables include individual, household, community, and department characteristics at Wave 1. Individual variables are gender, age, education, ethnicity, dummies for having been fostered before age 15 and for first-borns, relationship to the household head, and socio-occupational category. Household and cell controls include household size, a consumption index per adult equivalent, cell size, and the ratio of cell consumption to that of the household. Community determinants include controls for the environment (urban or rural). Last, two variables are defined at the department level: a measure of poverty (headcount) and a measure of inequality (Gini index). [10]

21We estimate our model on both the pooled sample and separately for men and women, as we expect migration determinants to vary by gender.

22Finally, we investigate the determinants of distance travelled. Our dependent variable is the log of the great-circle distance between the locations of an individual at Waves 1 and 2, computed from the recorded GPS coordinates.

2 – Treatment of attrition

23Attrition is a critical issue when using panel data to study migration. Of the individuals surveyed in the first wave, 89.4% were tracked and surveyed again in the second wave in their new location. Many individuals not found in the second survey wave presumably left their household of origin and moved far enough away that surveyors could not find them. [11] Internal migration thus probably contributes to attrition. However, when attritors’ observed characteristics are compared to those of four distinct groups—internal migrants (i.e. who moved more than 5 km), deceased individuals, international migrants, and short-distance movers (less than 5 km)—attritors’ characteristics may be closer, on average, to those of international migrants than to any other group’s. This similarity is particularly true for education, socio-occupational category, and household wealth, which suggests that international migration is probably a much stronger driver of attrition than internal migration. Nonetheless, to analyse the sensitivity of our results to attrition, we estimate a probit model for internal migration, treating attritors as internal migrants. [12]

24Additionally, attrition may be correlated with migration distance, but the sign of the correlation is ambiguous. Individuals moving far away from their original location may be less easily tracked, but even short-distance mobility within urban areas may also generate attrition due to weaker neighbourhood networks that might otherwise help to track ‘lost’ individuals. The success of tracking those who had moved relied predominantly on the accuracy of information collected by the second-wave fieldwork controller either from other household members who remained at the first-wave location or from neighbours when the whole household had moved. Since our estimates of the determinants of migration distance may be biased by selective attrition, we follow Senne (2014) and estimate a two-step Heckman (1979) selection model, using second-wave fieldwork-controller dummies as excluded instruments. [13] We expect their individual characteristics to have an impact on tracking outcomes and thus on attrition, but not on migration distance.

IV – Empirical findings

1 – Determinants of migration

25A gendered pattern clearly emerges: women are more likely to be internal migrants (Table 2). This highly feminized dynamic should be qualified because the model accounting for attrition reveals that men, though not all of whom can be considered internal migrants, are more likely to have been lost to observation between the two waves. The only common feature of male and female internal migration is that being a child or sibling of the household head is associated with a significantly greater probability of migrating.

26Ethnicity is correlated with female migration only. Women belonging to the Diola and, to a lesser extent, Serer ethnic groups are more likely to migrate than members of the Wolof, Senegal’s largest ethnic group (compare Brockerhoff and Eu (1993) on the migration of Serer and Diola women to urban regions for domestic work).

27As for education, men with at least some primary education are more likely to become internal migrants than those with none at all, while for women no significant differences in migration propensities are observed.

28Household income, proxied by consumption per adult equivalent, is positively correlated with the probability of migrating for men. In addition, men are more likely to be internal migrants when their cell enjoys a larger share of household expenditures. These two findings suggest the existence of higher costs to male migration, consistent with descriptive statistics pointing to the different motives of migration for men (mainly employment search and education) and women (marriage). Men living outside Dakar are more likely to move than women, and often to Dakar (as shown in Figure 1). This confirms Dakar’s importance as a destination for male migrants seeking better economic opportunities.

Table 2

Probit model of internal migration in Senegal

Table 2
All Women Men Individual characteristics Male −0.020*** (0.007) Age in years −0.001 (0.001) −0.002 (0.002) −0.000 (0.002) Age squared 0.000 (0.000) 0.000 (0.000) −0.000 (0.000) Fostered 0.015 (0.009) 0.013 (0.013) 0.015 (0.013) First-born 0.003 (0.007) 0.009 (0.010) −0.006 (0.010) Ethnicity (Ref.: Wolof/Lebu) Serer 0.011 (0.010) 0.016 (0.014) 0.007 (0.015) Fulbe −0.010 (0.008) −0.004 (0.012) −0.015 (0.012) Diola 0.018 (0.015) 0.040** (0.020) −0.008 (0.022) Other 0.014 (0.010) 0.026* (0.014) 0.002 (0.015) Relationship to household head (Ref.: Head) Spouse/parent −0.011 (0.014) −0.020 (0.021) 0.000 Child/sibling 0.049*** (0.013) 0.051** (0.023) 0.045*** (0.017) Other/non-relative 0.064*** (0.014) 0.038* (a) (0.023) 0.098*** (a) (0.019) Education (Ref.: None) Primary 0.003 (0.009) −0.014 (a) (0.013) 0.025* (a) (0.014) Secondary/university 0.019* (0.011) 0.003 (a) (0.016) 0.039*** (a) (0.015) Quranic 0.002 (0.010) 0.006 (0.013) −0.000 (0.014) Occupation (Ref.: Salaried worker) Agricultural worker −0.009 (0.014) −0.005 (0.023) −0.009 (0.018) Self-employed/employer −0.005 (0.014) −0.024 (0.024) 0.009 (0.017) Family worker 0.008 (0.016) 0.022 (0.024) −0.008 (0.022) Student 0.024 (0.015) 0.023 (0.025) 0.025 (0.019) Inactive/unemployed 0.001 (0.012) 0.008 (0.019) −0.001 (0.018) Missing 0.025** (0.012) 0.039** (0.019) 0.008 (0.015) Household characteristics Consumption index (adult equiv.) 0.008* (0.004) 0.003 (0.006) 0.015** (0.006) Household size −0.001 (0.001) −0.001 (0.001) −0.000 (0.001) Cell size −0.002 (0.002) −0.002 (0.003) −0.002 (0.003) Cell/household expenditure 0.036* (0.019) 0.025 (0.027) 0.049* (0.027) Community and department of origin characteristics (b) Community of origin (Ref.: Dakar) Other cities 0.035*** (0.012) 0.013 (a) (0.016) 0.061*** (a) (0.017) Rural 0.062*** (0.013) 0.028 (a) (0.018) 0.110*** (a) (0.020) Headcount poverty (Dakar/ urban/rural) 0.078* (0.041) 0.121** (0.056) 0.016 (0.059) Gini index −0.128 (0.080) −0.129 (0.108) −0.156 (0.118) N observations (c) 6,874 3,825 3,029

Probit model of internal migration in Senegal

(a) The coefficients for women and men are significantly different at the 10% level based on Student tests on the coefficients of interaction terms in a fully interacted model where all variables are interacted with the gender dummy.
(b) Department characteristics were computed based on the 2002 Senegalese census. Headcount poverty and Gini index were generated with PovMap2 (World Bank, 2009).
(c) The number of observations used in the regression differ slightly from the number of individuals in our sample (6,941) due to missing values for consumption data.
Note: Marginal effects are reported. Standard errors in parentheses. The dependent variable is the log of the distance (in km).
* p < .10, ** p < .05, ***p < .01.
Source: Poverty and Family Structure survey, 2006–2007 and 2010–2012.

29Finally, we find a positive effect of poverty on female migration only, indicative of the persistence of marriage-related female migration in the poorest and remotest areas of the country (discussed below). At the regional level, poverty thus seems an important ‘push factor’ for women. However, including the poverty rate at destination in our regression does not enable us to test whether they moved to wealthier areas because we do not have enough variation in destination characteristics given the high share of internal migrants to Dakar. [14]

Robustness checks

30Results may be biased due to the possible correlation of attrition with internal migration (Section 3.2). To assess such bias, we treat attritors as internal migrants (see Appendix Table A.1). [15] Contrary to what is reported in Table 2, women are no longer found to be more likely to migrate than men because men are more likely than women to be lost to observation between the two waves. That female migrants are more easily tracked suggests they are less likely to sever links with their household of origin. Female migration thus appears less autonomous than male migration. In addition, the share of attrition due to the loss of the entire household is lower for women (45.9%) than for men (52.3%). This difference is partly due to the frequent loss of urban, male-headed one-person households characterized by higher mobility and lower insertion in local networks.

31However, apart from the different gender composition of the attrited and migrant populations, most results from Table 2 do not change when treating attrition as internal migration. For women, attrition is very likely driven in part by the high mobility of young, salaried domestic workers (maids or petites bonnes), as was documented by Lesclingand (2011) for Mali. This interpretation is supported by the positive correlation between attrition, on the one hand, and the salaried-worker dummy and household income variable, on the other hand, because only relatively wealthy households can afford to employ domestic workers. By contrast, male attrition may be partly attributable to international migration, as male spouses of the head are more likely to be lost to observation, as well as individuals from wealthier households.

32Because decisions to move within and outside the country are probably interconnected, we estimated a multinomial logit model with three options: to stay (the reference), to migrate internally, or to migrate abroad. Results (available upon request) are remarkably close to those presented in Table 2, with only marginal differences in the coefficients of the ethnicity dummies.

2 – Migration distance

33Table 3 shows the second-stage estimates of a two-step Heckman selection model, described in Section 3.2, with dummies for the second-wave fieldwork controller as excluded instruments to correct for selection biases due to attrition. [16] Consistent with our descriptive statistics, women do not move as far away as men from the household of origin, but the coefficient is not significant. Regressions by gender reveal different determinants of migration distance for men and women.

Table 3

Determinants of migration distance in Senegal, Heckman model estimation (second step)

Table 3
All Women Men Individual characteristics Male 0.048 (0.176) Age in years 0.021 (0.023) 0.028 (0.026) −0.012 (0.033) Age squared −0.000 (0.000) −0.000 (0.000) 0.000 (0.000) Fostered −0.141 (0.164) −0.194 (0.205) −0.203 (0.208) First-born −0.116 (0.132) −0.276* (a) (0.158) 0.089 (a) (0.166) Ethnicity (Ref.: Wolof/Lebu) Serer −0.196 (0.207) −0.171 (0.250) −0.082 (0.245) Fulbe −0.104 (0.166) −0.116 (0.198) 0.116 (0.205) Diola 0.390 (0.274) 0.409 (0.334) 0.242 (0.333) Other −0.304 (0.201) −0.133 (0.234) −0.456* (0.257) Relationship to household head (Ref.: Head) Spouse/parent −0.005 (0.289) −0.437 (0.416) Child/sibling 0.860*** (0.285) 0.172 (0.412) 0.776* (0.406) Other/non-relative 0.592** (0.285) 0.038 (0.414) 0.550 (0.401) Education (Ref.: None) Primary 0.399** (0.201) 0.089 (0.210) 0.367 (0.312) Secondary/university 0.136 (0.221) 0.238 (0.252) −0.200 (0.332) Quranic 0.314 (0.197) 0.046 (0.222) 0.434 (0.269) Occupation (Ref.: Salaried worker) Agricultural worker −0.451 (0.282) −0.133a (0.432) −0.845*** (a) (0.294) Self-employed/employer −0.145 (0.269) −0.478 (0.419) 0.023 (0.269) Family worker −0.100 (0.303) −0.131 (0.412) −0.142 (0.347) Student −0.014 (0.282) 0.045 (0.399) −0.076 (0.313) Inactive/unemployed 0.035 (0.225) 0.047 (0.320) 0.080 (0.304) Missing 0.174 (0.230) 0.152 (0.332) 0.040 (0.242)

Determinants of migration distance in Senegal, Heckman model estimation (second step)

Table 3

(cont‘d). Determinants of migration distance in Senegal, Heckman model estimation (second step)

Table 3
All Women Men Household characteristics Consumption index (adult equiv.) 0.053 (0.083) −0.020 (0.111) 0.147 (0.091) Household size 0.034** (0.016) 0.027 (0.019) 0.024 (0.019) Cell size −0.068* (0.037) −0.106** (0.047) −0.043 (0.039) Cell/household expenditure 0.700** (0.338) 0.604 (0.436) 0.533 (0.401) Community and department of origin characteristics (b) Community of origin (Ref.: Dakar) Other cities −0.405 (0.254) −0.763*** (0.273) −0.170 (0.385) Rural −0.384 (0.345) −1.180*** (a) (0.325) −0.093 (0.569) Headcount poverty (Dakar/urban/rural) 2.849*** (0.807) 2.921*** (0.957) 1.551 (1.012) Gini index 5.163*** (1.669) 5.953*** (1.811) 6.356*** (2.418) Constant −1.938 (1.225) −0.038 (1.395) −0.963 (1.713) Lambda (inverse Mills ratio) 1.579*** (0.427) 0.988** (0.400) 0.732 (0.591) N observations (including attritors) 1,300 678 622 N observations (without attritors) (c) 514 306 208

(cont‘d). Determinants of migration distance in Senegal, Heckman model estimation (second step)

(a) The coefficients for women and men are significantly different at the 10% level based on Student tests on the coefficients of interaction terms in a fully interacted model where all variables are interacted with the gender dummy.
(b) Department characteristics were computed based on the 2002 Senegalese census. Headcount poverty and Gini index were generated with PovMap2 (World Bank, 2009).
(c) The number of observations used in the regression differ slightly from the number of individuals in our sample (556) due to missing values for consumption data.
Note: Marginal effects are reported. Standard errors in parentheses. The dependent variable is the log of the distance (in km).
* p < .10, ** p < .05, *** p < .01.
Source: Poverty and Family Structure survey, 2006–2007 and 2010–2012.

34Reflecting the gender roles in Senegalese society, women do not migrate as far as men when they are the eldest of their siblings. Not only do eldest daughters play an important parental role for their younger siblings, they are also an object of financial transaction in marriage (a bride-price paid to their family), which in turn is expected to provide their younger brothers with the resources to get married. This financial dependency of the household, common to many other African societies (Horne et al., 2013; Trinitapoli et al., 2014), may explain in part why eldest daughters move less far, as their households of origin seek to maintain close links.

35On the total sample, we find a positive correlation between distance and both household size and the share of the migrant’s cell in total household expenditures (Table 3). Poverty at the department level is positively correlated with migration distance for women only. Female internal migrants from outside Dakar move near their households of origin, consistent with the predominance of short-distance moves related to marriage.

36A feature common to male and female migration is the positive relationship between distance travelled and income inequality in the department of origin. This finding is in line with Stark’s (1984) relative deprivation motive whereby individuals living in areas with higher income inequality may move to improve their relative situation. This result is also linked to the geography of Senegal and to Dakar’s attractiveness, since the remotest regions in the south-east of the country are also characterized by the highest levels of inequality.

37Additionally, we take into account the rural–urban dimension of internal migration, which reveals different patterns for men and women (Appendix Table A.3). We find that women are more likely than men to move to rural areas, which, in accordance with marriage-driven migration, is especially true of first-born women (Quisumbing and McNiven, 2010; Herrera and Sahn, 2013). Educational achievement has opposite effects on the decision to move to an urban or rural destination. Compared to individuals with no education, those with a primary or secondary education are more likely to move to urban areas and less likely to move to rural ones, which is also the case in other countries. Ackah and Medvedev (2012), for example, found that in Ghana more-educated individuals are more likely to move to urban settings from areas where education and health services are lacking. Our results suggest that the correlation between education and rural–urban mobility is particularly significant for women—positive with migration to urban areas, negative with migration to rural areas—which may reveal unequal opportunities and different reasons for moving depending on women’s educational achievement. The relationship to the household head highlights different trajectories of sons and daughters. Whereas male children or siblings of the household head are more likely to move to urban centres, female children and siblings have a higher probability of migrating to rural areas. Unsurprisingly, individuals from wealthier households tend to migrate to urban areas.

38Finally, we find that being born in a rural area has a positive effect on the probability of migrating to a rural area, which is significant for women only. Additional specifications decomposing migration by both destination and origin provide similar evidence that women are more likely than men to experience rural-to-rural moves.

Conclusion

39This article studies the determinants of internal migration using rich individual panel data from a nationally representative survey conducted in Senegal in 2006–2007 and 2010–2012. The tracking of all individuals who remained in the country by using GPS coordinates at both dates allowed us to identify and map gender-specific mobility patterns between the two survey waves. Unique for this region of the world, these features enable us to make an important contribution to the literature on internal migration and gender in sub-Saharan Africa. Using distance rather than administrative boundaries has the advantage of providing a less arbitrary and more homogenous definition of migration. Had we defined internal migration based on changes in department of residence, we would have missed 19% of our sample, three-quarters of whom are women.

40Our empirical analysis reveals clear gender differences. Women were more likely than men to have migrated between the two survey rounds; however, women tend to do so over shorter distances, which is explained by their characteristics. The decomposition of the decision to migrate to rural and urban destinations shows that women are more likely than men to move to rural areas, especially when originating from them. The analysis of motives for migrating reveals that the primary reason for women is marriage, whatever the distance travelled. For men, labour and education are the most cited motives.

41These findings reveal the durability of gendered migration patterns and motives analysed since the 1990s (Chant et al., 1992). These findings echo those of Comoe (2013) for Ivory Coast. They are also consonant with Duboz et al. (2011), who showed that family motives are prevalent among Senegalese women who moved to Dakar, and with Vause and Toma (2015), who found little independent international migration of Senegalese women.

42We find that education increases the likelihood of migrating to urban areas, especially for women, and could thus be an effective channel for promoting women’s access to independent migration. However, the question of the actual benefits of migrating for both women and men remains open for future research.

Acknowledgements

We are grateful to Christophe Guilmoto, Joachim Jarreau, Robert E. B. Lucas, Karine Marazyan, Jean-Noël Senne, and Sorana Toma for their helpful comments and suggestions.

Appendix

Table A.1

Probit model of internal migration including attritors living outside Dakar at Wave 1

Table A.1
All Women Men Individual characteristics Male −0.001 (0.009) Age in years −0.002 (0.001) −0.004** (0.002) 0.001 (0.002) Age squared −0.000 (0.000) 0.000 (0.000) −0.000 (0.000) Fostered 0.018* (0.011) 0.026* (0.015) 0.004 (0.016) First-born 0.003 (0.008) 0.018* (0.011) −0.016 (0.012) Ethnicity (Ref.: Wolof/Lebu) Serer 0.032*** (0.012) 0.042*** (0.016) 0.021 (0.018) Fulbe −0.002 (0.010) 0.010 (0.013) −0.017 (0.015) Diola 0.017 (0.018) 0.027 (0.024) 0.004 (0.026) Other 0.020 (0.012) 0.034** (0.016) 0.004 (0.018) Relationship to household head (Ref.: Head) Spouse/parent −0.004 (0.015) −0.001 (0.024) 0.101 (0.073) Child/sibling 0.061*** (0.015) 0.075*** (0.026) 0.046** (0.020) Other/non-relative 0.089*** (0.015) 0.066** (0.026) 0.128*** (0.021) Education (Ref.: None) Primary −0.024** (0.011) −0.029** (0.014) −0.020 (0.016) Secondary/university 0.002 (0.012) −0.007 (0.018) 0.005 (0.017) Quranic −0.017 (0.011) −0.009 (0.015) −0.028* (0.016) Occupation (Ref.: Salaried worker) Agricultural worker −0.020 (0.017) −0.041 (0.027) −0.007 (0.021) Independent/employer −0.009 (0.016) −0.046* (0.027) 0.016 (0.021) Family worker 0.005 (0.018) 0.002 (0.027) 0.003 (0.026) Student 0.037** (0.018) 0.024 (0.029) 0.045* (0.024) Inactive/unemployed −0.005 (0.014) −0.011 (0.021) 0.005 (0.021) Missing 0.030** (0.014) 0.031 (0.022) 0.017 (0.019) Household characteristics Consumption index (adult equiv.) 0.015*** (0.005) 0.008 (0.007) 0.024*** (0.007) Household size −0.002** (0.001) −0.003** (0.001) −0.001 (0.001) Cell size −0.006** (0.002) −0.005 (0.003) −0.006* (0.003) Cell/household expenditure 0.042* (0.022) 0.014 (0.031) 0.071** (0.033) Community and department of origin characteristics (a) Community of origin (Ref.: Dakar) Other cities 0.131*** (0.014) 0.095*** (0.018) 0.173*** (0.021) Rural 0.122*** (0.016) 0.079*** (0.021) 0.183*** (0.025) Headcount poverty (Dakar/urban/rural) 0.095** (0.047) 0.121* (0.063) 0.063 (0.069) Gini index −0.111 (0.089) −0.141 (0.118) −0.100 (0.133) N observations 7,210 3,983 3,227

Probit model of internal migration including attritors living outside Dakar at Wave 1

(a) Department characteristics were computed based on the 2002 Senegalese census. Headcount poverty and Gini index were generated with PovMap2 (World Bank, 2009).
Note: Standard errors in parentheses.
* p < .10, ** p < .05, *** p < .01.
Source: Poverty and Family Structure survey, 2006–2007 and 2010–2012.
Table A.2

Determinants of migration distance in Senegal, Heckman model estimation (first step)

Table A.2
All Women Men Individual characteristics Male −0.443*** (0.090) Age in years 0.014 (0.015) 0.026 (0.020) −0.029 (0.025) Age squared −0.000 (0.000) −0.000 (0.000) 0.001* (0.000) Fostered −0.074 (0.107) −0.178 (0.148) 0.112 (0.164) First-born 0.040 (0.088) −0.066 (0.124) 0.218* (0.132) Ethnicity (Ref.: Wolof/Lebu) Serer −0.239* (0.128) −0.348** (0.174) −0.012 (0.203) Fulbe −0.144 (0.107) −0.244 (0.151) 0.046 (0.162) Diola 0.104 (0.187) 0.356 (0.271) −0.017 (0.284) Other −0.189 (0.130) −0.198 (0.182) −0.157 (0.200) Relationship to household head (Ref.: Head) Spouse/parent −0.186 (0.177) −0.460 (0.288) −6.874 Child/sibling 0.315* (0.162) 0.043 (0.301) 0.591*** (0.217) Other/non-relative 0.246 (0.165) −0.031 (0.302) 0.546** (0.224) Education (Ref.: None) Primary 0.428*** (0.115) 0.312* (0.161) 0.696*** (0.182) Secondary/university 0.393*** (0.126) 0.275 (0.186) 0.655*** (0.189) Quranic 0.261** (0.126) 0.237 (0.182) 0.424** (0.196) Occupation (Ref.: Salaried worker) Agricultural worker 0.114 (0.184) 0.812** (0.316) −0.267 (0.243) Independent/employer −0.022 (0.166) 0.096 (0.279) −0.070 (0.213) Family worker 0.273 (0.204) 0.808*** (0.300) −0.205 (0.303) Student 0.050 (0.184) 0.289 (0.289) −0.281 (0.251) Inactive/unemployed 0.112 (0.139) 0.459** (0.202) −0.285 (0.227) Missing 0.201 (0.140) 0.482** (0.210) 0.043 (0.204) Household characteristics Consumption index (adult equiv.) −0.022 (0.054) −0.072 (0.086) 0.056 (0.076) Household size 0.019* (0.011) 0.023 (0.016) 0.016 (0.015) Cell size 0.016 (0.025) 0.011 (0.037) 0.023 (0.036) Cell/household expenditure 0.195 (0.219) 0.327 (0.333) 0.156 (0.322) Community and department of origin characteristics (a) Community of origin (Ref.: Dakar) Other cities 0.363** (0.150) 0.316 (0.202) 0.482** (0.240) Rural 0.900*** (0.168) 0.716*** (0.227) 1.166*** (0.266) Headcount poverty (Dakar/urban/rural) 0.614 (0.512) 0.592 (0.713) 0.446 (0.797) Gini index −2.155** (1.076) −1.859 (1.448) −3.155* (1.750) Controller dummies Yes Yes Yes Constant −0.694 (0.723) −0.542 (1.019) −0.942 (1.167) N observations 1,300 678 622

Determinants of migration distance in Senegal, Heckman model estimation (first step)

(a) Department characteristics were computed based on the 2002 Senegalese census. Headcount poverty and Gini index were generated with PovMap2 (World Bank, 2009).
Note: See Table 3 for the second step of the model. Here, the dependent variable is a dummy variable equal to 1 for non-attritors. Standard errors in parentheses.
* p < .10, ** p < .05, *** p < .01.
Source: Poverty and Family Structure survey, 2006–2007 and 2010–2012.
Table A.3

Determinants of internal migration in Senegal by origin and destination

Table A.3
All Urban origin Rural origin Urban destination Rural destination Urban destination Rural destination Urban destination Rural destination Individual characteristics Male −0.002 0.006 −0.018*** 0.005 −0.010 0.006 −0.008 0.006 0.007 0.010 −0.039*** 0.010 Age in years −0.001 0.001 0.000 0.001 −0.000 0.001 0.001 0.001 −0.002 0.002 −0.001 0.001 Age squared 0.000 0.000 0.000 0.000 0.000 0.000 -0.000 0.000 0.000 0.000 0.000 0.000 Fostered 0.005 0.007 0.012** 0.006 0.001 0.008 0.011* 0.006 0.013 0.013 0.013 0.011 First-born −0.003 0.006 0.006 0.005 0.002 0.006 −0.005 0.005 −0.009 0.010 0.020** 0.008 Ethnicity (Ref.: Wolof/Lebu) Serer 0.017** 0.008 −0.005 0.007 0.012 0.010 −0.006 0.010 0.012 0.012 −0.010 0.012 Fulbe −0.005 0.007 −0.006 0.006 0.011 0.008 0.008 0.006 −0.031*** 0.013 −0.023** 0.011 DIola 0.013 0.010 0.003 0.011 0.017 0.011 0.009 0.011 −0.002 0.021 −0.018 0.023 Other 0.006 0.008 0.010 0.006 −0.003 0.010 0.002 0.008 0.007 0.013 0.008 0.011 Relationship to household head (Ref.: Head) Spouse/parent −0.011 0.012 0.005 0.012 −0.007 0.013 0.010 0.014 −0.006 0.021 −0.013 0.020 Child/sibling 0.028*** 0.010 0.030** 0.012 0.017 0.012 0.021 0.014 0.050** 0.019 0.033* 0.020 Other/non-relative 0.031*** 0.011 0.037*** 0.012 0.035*** 0.012 0.043*** 0.014 0.032 0.020 0.013 0.021 Education (Ref.: None) Primary 0.031*** 0.007 -0.026*** 0.007 0.013 0.009 −0.022*** 0.007 0.040*** 0.011 −0.020 0.012 Secondary/university 0.041*** 0.008 −0.024*** 0.008 0.033*** 0.009 −0.024*** 0.008 0.031** 0.015 0.005 0.016 Quranic 0.009 0.009 −0.005 0.006 0.004 0.012 0.001 0.007 0.008 0.013 −0.013 0.010 Occupation (Ref.: Salaried worker) Agricultural worker −0.022* 0.012 0.015 0.010 −0.002 0.018 0.003 0.017 −0.024 0.020 0.022 0.025 Independent/employer −0.001 0.011 −0.007 0.012 0.006 0.011 −0.002 0.010 −0.008 0.024 −0.055 0.045 Family worker −0.011 0.013 0.019* 0.010 −0.002 0.027 0.018 0.015 −0.005 0.021 0.027 0.025 Student 0.014 0.011 0.015 0.012 0.009 0.012 0.017 0.011 0.040* 0.021 0.013 0.029 Inactive/unemployed −0.001 0.009 0.005 0.009 −0.004 0.010 −0.001 0.008 0.013 0.018 0.023 0.025 Missing 0.010 0.009 0.017* 0.009 0.023** 0.010 0.017** 0.008 −0.002 0.018 0.018 0.024

Determinants of internal migration in Senegal by origin and destination

Table A.3

(cont’d). Determinants of internal migration in Senegal by origin and destination

Table A.3
All Urban origin Rural origin Urban destination Rural destination Urban destination Rural destination Urban destination Rural destination Household characteristics Consumption index (adult equiv.) 0.010*** 0.003 −0.004 0.003 0.008** 0.004 −0.002 0.003 0.011* 0.006 −0.007 0.006 Household size −0.001 0.001 −0.001 0.001 −0.002* 0.001 -0.000 0.001 0.000 0.001 −0.002* 0.001 Cell size −0.003* 0.002 0.002 0.001 −0.003 0.002 0.000 0.001 −0.003 0.003 0.004* 0.002 Cell/household expenditure 0.043*** 0.015 −0.017 0.014 0.023 0.017 −0.009 0.014 0.074*** 0.026 −0.037 0.027 Community and department of origin characteristics (a) Community of origin (Ref.: Dakar) Other cities 0.059*** 0.010 −0.021*** 0.008 0.034*** 0.010 −0.009 0.007 Rural 0.084*** 0.012 −0.015*** 0.009 Headcount poverty (Dakar/urban/rural) 0.014 0.033 0.053* 0.028 0.097** 0.038 −0.026 0.029 −0.112* 0.062 0.232*** 0.072 Gini index 0.014 0.059 −0.156*** 0.057 0.124 0.085 −0.251*** 0.094 −0.001 0.092 −0.191** 0.087 N observations 6,873 6,873 3,954 3,954 2,919 2,919

(cont’d). Determinants of internal migration in Senegal by origin and destination

(a) Department characteristics were computed based on the 2002 Senegalese census. Headcount poverty and Gini index were generated with PovMap2 (World Bank, 2009).
Note: Standard errors in parentheses.
* p < .10, **p < .05, *** p < .01.
Source: Poverty and Family Structure survey, 2006–2007 and 2010–2012.

Notes

  • [1]
    Step migration is a process by which the migrant accumulates the capital necessary to reach her preferred destination (Paul, 2011). For instance, rural migrants wishing to reach their country’s capital may first choose to stay in intermediate-sized towns, or international migrants may first look for a job in the capital to finance their move abroad.
  • [2]
    Circular migration may be defined as the process of leaving and then returning to one’s place of origin (Newland, 2009). However, the duration of stay in the destination has to be short, and the movement between origin and destination has to be repeated.
  • [3]
    The survey was conducted by a team of French researchers and researchers from the national statistical agency of Senegal (Agence nationale de la statistique et de la démographie); it is described in detail in De Vreyer et al. (2008).
  • [4]
    Individuals were tracked whatever their relationship to the household head. The 11.6% of individuals that could not be recontacted moved between waves, and information on their new place of residence could not be collected among family members or neighbours. A further 4.8% of first-wave individuals died between the surveys.
  • [5]
    See Lambert et al. (2014) for a more detailed definition of cell.
  • [6]
    Both measures were obtained using PovMap2, a software package developed by the World Bank (Elbers et al., 2003; Zhao and Lanjouw, 2009).
  • [7]
    Senegal was subdivided into 34 departments and 11 regions in 2006. Following administrative reforms in 2008, the country is now subdivided into 45 departments and 14 regions.
  • [8]
    We treat intra-Dakar migrants (115 individuals) as non-migrants because mobility within Dakar, even though on a distance larger than 5 km, might not be considered migration.
  • [9]
    The correlation coefficient between the two measures of distance is 0.96.
  • [10]
    The poverty headcount is the proportion of the population whose consumption is below a given threshold. The Gini index is a measure of inequality between individuals that can take a value between 0 and 1. The higher the index, the higher the inequality: 0 is the hypothetical situation where everyone has the same level of consumption, while 1 corresponds to the situation where one person consumes all resources, and all other members of the population consume none.
  • [11]
    To limit attrition when whole households moved between the two waves, surveyors were instructed to collect information from neighbours or reference persons in the community.
  • [12]
    This model is not our main specification; it aims to provide an upper bound for the bias resulting from attrition because the assumption that all attritors are internal migrants leads to an overestimation of the number of internal migrants.
  • [13]
    Ten teams of two to three surveyors were managed on the field by one controller each, who was in charge of checking compliance with procedures and of verifying that accurate tracking information had been reported by surveyors.
  • [14]
    Although there is variation in the poverty rate within Dakar between poor areas like Guédiawaye and rich ones such as Almadies, the 10% extract of the 2002 census that we could exploit does not allow us to construct a poverty measure at a level of disaggregation finer than that of the department. The same limitations apply to our department-level inequality measure, included here as a ‘push’ factor only.
  • [15]
    To avoid overestimating internal migration and to remain consistent with our definition of internal migrants, which excludes intra-Dakar mobility, we exclude from this regression sample the attritors surveyed in Dakar during the first wave.
  • [16]
    The results of the first step are shown in Appendix Table A.2.

Abstract

English

This study explores internal migration patterns in Senegal using individual panel data from a nationally representative survey collected in 2006–2007 and 2010–2012. The data are unique in that they contain the GPS coordinates of individuals’ locations in both waves. We are thus able to calculate distances and map individual moves, thereby avoiding the problems involved in using administrative units to define migration. Our results reveal highly gendered mobility patterns and confirm their durability over the last decades. Women are more likely to migrate than men, but to rural rather than urban destinations. While education increases the likelihood of migration to urban destinations, especially for women, female mobility is mostly linked to marriage, whereas labour mobility is more frequently observed for men.

  • internal migration
  • gender inequalities
  • rural–urban migration
  • Senegal
  • geolocalized data
Français

Mobilité genrée au Sénégal

Cet article examine les modèles de migration interne au Sénégal à l’aide de données individuelles provenant d’une étude représentative sur le plan national réalisée en 2006-2007 et 2010-2012. Ces données sont uniques dans la mesure où elles contiennent les coordonnées GPS des personnes enquêtées lors des deux vagues. Il est alors possible de calculer les distances et de cartographier les déplacements individuels en évitant les problèmes posés par l’utilisation des unités administratives pour définir les migrations. Ces résultats mettent en lumière des comportements de mobilité très différents selon le sexe et confirment leur persistance pendant les dernières décennies. Les femmes sont plus susceptibles de migrer que les hommes, mais vers des destinations rurales plutôt qu’urbaines. Bien que l’instruction augmente les probabilités de migration vers les villes, surtout chez les femmes, la mobilité féminine est essentiellement liée au mariage, tandis que les migrations de travail concernent plus souvent les hommes.

Español

La movilidad de género en Senegal

Este artículo examina los modelos de migración interna en Senegal, gracias a datos individuales provenientes de un estudio representativo a nivel nacional, realizado en 2006-07 y 2010-12. Estos datos son únicos pues contienen las coordenadas GPS de las personas interrogadas en las dos olas. Ello permite calcular las distancias y cartografiar los desplazamientos individuales evitando los problemas que plantea el uso de las unidades administrativas para definir las migraciones. Los resultados ponen en evidencia comportamientos de movilidad muy diferentes según el sexo y confirman su persistencia durante las últimas décadas. Las mujeres tienen más probabilidades de emigrar pero más bien hacia zonas rurales que urbanas. Aunque la instrucción aumenta las probabilidades de emigración hacia las ciudades, sobre todo en las mujeres, la movilidad femenina está esencialmente ligada al matrimonio, mientras que las migraciones de trabajo conciernen sobre todo los hombres.

Outline

  1. I - Literature and context
  2. II - Data
    1. 1 - The Poverty and Family Structure individual panel survey
    2. 2 - Descriptive statistics
  4. III - Empirical approach
    1. 1 - Empirical models
    2. 2 - Treatment of attrition
  6. IV - Empirical findings
    1. 1 - Determinants of migration
      1. Robustness checks
    3. 2 - Migration distance
  8. Conclusion

References

References

  • OnlineAbu-Ghaida D., Klasen S., 2004, The costs of missing the millennium development goal on gender equity, World Development, 32(7), 1075–1107.
  • OnlineAckah C., Medvedev D., 2012, Internal migration in Ghana: Determinants and welfare impacts, International Journal of Social Economics, 39(10), 764–784.
  • Agence nationale de la statistique et de la démographie, 2006, Rapport national de présentation des résultats définitifs. Résultats définitifs du troisième recensement général de la population et de l’habitat, 2002, Dakar, ANSD.
  • Antoine P., Sow O., 2000, Rapports de genre et dynamiques migratoires: le cas de l’Afrique de l’Ouest, in Bozon M., Locoh T. (eds.), Rapports de genre et questions de population, Vol. II, Paris, INED, 143–159.
  • OnlineAssaad R., Arntz M., 2005, Constrained geographical mobility and gendered labor market outcomes under structural adjustment: Evidence from Egypt, World Development, 33(3), 431–454.
  • OnlineBeauchemin C., Bocquier P., 2004, Migration and urbanisation in francophone West Africa: An overview of the recent empirical evidence, Urban Studies, 41(11), 2245–2272.
  • OnlineBell M., Blake M., Boyle P., Duke-Williams O., Rees P., Stillwell J., Hugo G., 2002, Cross-national comparison of internal migration: Issues and measures, Journal of the Royal Statistical Society, A, 165(3), 435–464.
  • OnlineBell M., Charles-Edwards E., Kupiszewska D., Kupiszewski M., Stillwell J., Zhu Y., 2015, Internal migration data around the world: Assessing contemporary practice, Population, Space and Place, 21(1), 1–17.
  • OnlineBloom D. E., Kuhn M., Prettner K., 2015, The contribution of female health to economic development (Working Paper No. 21411), Cambridge, MA, National Bureau of Economic Research.
  • OnlineBrockerhoff M., Eu H., 1993, Demographic and socioeconomic determinants of female rural to urban migration in sub-Saharan Africa, International Migration Review, 27(3), 557–577.
  • Chant S. (ed.), 1992. Gender and migration in developing countries, London, Belhaven Press.
  • OnlineChort I., 2014, Mexican migrants to the US: What do unrealized migration intentions tell us about gender inequalities? World Development, 59(C), 535–552.
  • OnlineChort I., Senne J.-N., 2015, Selection into migration within a household model: Evidence from Senegal, The World Bank Economic Review, 29(suppl. 1), S247–S256.
  • OnlineClark W., Huang Y., 2004, Linking migration and mobility: Individual and contextual effects in housing markets in the UK, Regional Studies, 38(6), 617–628.
  • OnlineComoe E. F., 2013, Femmes et migration en Côte d’Ivoire: le mythe de l’autonomie, African Population Studies, 20(1), 89–117.
  • OnlineCordey-Hayes M., Gleave D., 1974, Migration movements and the differential growth of city regions in England and Wales, Papers in Regional Science, 33(1), 99–123.
  • De Vreyer P., Lambert S., Safir A., Sylla M., 2008, Pauvreté et Structure Familiale, pourquoi une nouvelle enquête? Statéco, 102, 5–20.
  • OnlineDonato K. M., Gabaccia D., Holdaway J., Manalansan M., Pessar P. R., 2006, A glass half full? Gender in migration studies, International Migration Review, 40(1), 3–26.
  • OnlineDuboz P., Macia E., Gueye L., Boetsch G., Chapuis-Lucciani N., 2011, Migrations internes au Sénégal. Caractéristiques socioéconomiques, démographiques et migratoires des Dakarois, Diversité urbaine, 11(2), 113–135.
  • OnlineDuflo E., 2012, Women empowerment and economic development, Journal of Economic Literature, 50(4), 1051–1079.
  • OnlineElbers C., Lanjouw J. O., Lanjouw P., 2003, Micro-level estimation of poverty and inequality, Econometrica, 71(1), 355–364.
  • OnlineHeckman J., 1979, Sample selection bias as a specification error, Econometrica, 47(1), 153–161.
  • OnlineHerrera C., Sahn D. E., 2013, Determinants of internal migration among Senegalese youth (Cornell Food and Nutrition Policy Program Working Paper No. 245), Ithaca, NY, Cornell University.
  • OnlineHorne C., Dodoo F. N.-A., Dodoo N. D., 2013, The shadow of indebtedness: Bridewealth and norms constraining female reproduct ive autonomy, American Sociological Review, 78(3), 503–520.
  • OnlineKanaiaupuni S. M., 2000, Reframing the migration question: An analysis of men, women, and gender in Mexico, Social Forces, 78(4), 1311–1347.
  • OnlineKudo Y., 2015, Female migration for marriage: Implications from the land reform in rural Tanzania, World Development, 65(C), 41–61.
  • OnlineLambert S., Ravallion M., Van De Walle D., 2014, Intergenerational mobility and interpersonal inequality in an African economy, Journal of Development Economics, 110(C), 327–344.
  • OnlineLesclingand M., 2011, Migrations des jeunes filles au Mali: exploitation ou émancipation? Travail, genre et sociétés, 25(1), 23–40.
  • OnlineLesclingand M., Hertrich V., 2017, When girls take the lead: Adolescent girls’ migration in Mali, Population, 72(1), 63–92.
  • OnlineLewis W. A., 1954, Economic development with unlimited supplies of labour, The Manchester School, 22(2), 139-191.
  • OnlineLucas R. E. B., 2016, Internal migration in developing economies: An overview of recent evidence, Geopolitics, History, and International Relations, 8(2), 159–191.
  • OnlineMassey D. S., Fischer M. J., Capoferro C., 2006, International migration and gender in Latin America: A comparative analysis, International Migration, 44(5), 63–91.
  • Newland K., 2009, Circular migration and human development (Human Development Research Paper No. 2009/42), United Nations Development Programme.
  • OnlineNiedomysl T., Fransson U., 2014, On distance and the spatial dimension in the definition of internal migration, Annals of the Association of American Geographers, 104(2), 357–372.
  • OnlinePaul A. M., 2011, Stepwise international migration: A multistage migration pattern for the aspiring migrant, American Journal of Sociology, 116(6), 1842–1886.
  • OnlineQuisumbing A., Mcniven S., 2010, Moving forward, looking back: The impact of migration and remittances on assets, consumption, and credit constraints in the rural Philippines, The Journal of Development Studies, 46(1), 91–113.
  • OnlineRosenzweig M., Stark O., 1989, Consumption smoothing, migration, and marriage: Evidence from rural India, The Journal of Political Economy, 97(4), 905–926.
  • OnlineSenne J.-N., 2014, Death and schooling decisions over the short and long run in rural Madagascar, Journal of Population Economics, 27(2), 497–528.
  • OnlineStark O., 1980, On the role of urban-to-rural remittances in rural development, Journal of Development Studies, 16(3), 369–374.
  • OnlineStark O., 1984, Rural-to-urban migration in LDCs: A relative deprivation approach, Economic Development and Cultural Change, 32(3), 475–486.
  • Stark O., Bloom D., 1985, The new economics of labor migration, The American Economic Review, 75(2), 173–178.
  • OnlineStark O., Lucas R. E. B., 1988, Migration, remittances, and the family, Economic Development and Cultural Change, 36(3), 465–481.
  • OnlineStenberg K. Axelson H., Sheehan P., Anderson I., Gülmezoglu A. M., Temmerman M., et al., 2014, Advancing social and economic development by investing in women’s and children’s health: A new Global Investment Framework, The Lancet, 383(9925), 1333–1354.
  • Todaro M. P., 1969, A model of labor migration and urban unemployment in less developed countries, The American Economic Review, 59(1), 138–148.
  • OnlineTrinitapoli J., Yeatman S., Fledderjohann J., 2014, Sibling support and the educational prospects of young adults in Malawi, Demographic Research, 30(art. 19), 547–578.
  • OnlineVause S., Toma S., 2015, Is the feminization of international migration really on the rise? The case of flows from the Democratic Republic of Congo and Senegal, Population, 70(1), 41–67.
  • Zhao Q., Lanjouw P., 2009, Using PovMap2: A user’s guide, Washington DC, The World Bank.

Authors

Isabelle Chort
Universite de Pau et des pays de l’Adour, E2S UPPA, CATT, Bayonne, France.
8 Allée des Platanes, CS 68505, 64185 Bayonne Cedex, France.
isabelle.chort@univ-pau.fr
Philippe De Vreyer
Université Paris-Dauphine, PSL Research University, CNRS, IRD, LEDa, [UMR 8007-260], DIAL.
Thomas Zuber
Department of History, Columbia University.
This is the latest publication of the author on cairn.
This is the latest publication of the author on cairn.
This is the latest publication of the author on cairn.
Uploaded on Cairn-int.info on 04/12/2020
Cite
Distribution électronique Cairn.info pour I.N.E.D © I.N.E.D. Tous droits réservés pour tous pays. Il est interdit, sauf accord préalable et écrit de l’éditeur, de reproduire (notamment par photocopie) partiellement ou totalement le présent article, de le stocker dans une banque de données ou de le communiquer au public sous quelque forme et de quelque manière que ce soit.
keyboard_arrow_up
Chargement
Loading... Please wait