How migrants choose their destination in Burkina Faso? A place-utility approach Prof. Sabine Henry Geography department, FUNDP, Belgium Prof. Richard Bilsborrow Carolina Population Center, Univ. of North Carolina, USA IUSSP, Fuzhou, December 2011 Supported by AUF
Contextual factors of migration Theory Push and pull model (Lee, 1966) The cost-benefit model (Sjaastad, 1962) The place utility approach (Wolpert, 1965) Application Recently, contextual factors at the origin influence the migration decision Very few studies on factors at the destination (mostly in developed countries) Environmental versus economic migrants? 2
Objective of this study Objective To understand how migrants, leaving a village or a secondary town, choose their destination in Burkina Faso. Assumption Push factors = pull factors Hypotheses #1. A favourable natural environment is a pull factor of major importance for migrants. #2. The economic diversification of villages attracts migrants. 3
The study area: Burkina Faso Rainfall conditions Land degradation Population Migration
General background - Sahelian region in West Africa - Largely a rural country :. 80 % of the population living in settlements of less than 10 000 inhabitants. Agriculture is the main source of livelihood (90 % of the population) - Country s economy depends heavily on the agriculture :. 1/3 of GDP comes from subsistence farming and cattle-raising The agro-climatic conditions are critical to rural households, for which agriculture represents the main source of livelihood
Agro-climatic context Four agro-climatic regions (New, Hulme et Jones, 2000) Strong south-north decreasing gradient of annual rainfall
Agro-climatic context context Four agro-climatic regions (New, Hulme et Jones, 2000) Strong south-north decreasing gradient of annual rainfall The Sahel : - Climate :. < 500 mm rainfall a year. High temperatures - Agriculture :. Extensive pastoralism and subsistence farming
North 8
Agro-climatic context context Four agro-climatic regions (New, Hulme et Jones, 2000) Strong south-north decreasing gradient of annual rainfall The South-West : - Climate :. > 900 mm rainfall a year - Agriculture :. Subsistence and cash crops
South-West 10
Agro-climatic context context Four agro-climatic regions (New, Hulme et Jones, 2000) Strong south-north decreasing gradient of annual rainfall Two intermediary regions :. From 500 to 699 mm rainfall a year. Subsistence agriculture (sensitive to rainfall conditions). From 700 to 899 mm rainfall a year. Subsistence and cash crops
Annual rainfall (mm) Rainfall situation (1960-98) High interannual rainfall variations
Population density (1996) Census in 1996 10,3 million people Urban rate: 20% Life expectancy: 53.8 years 13
Migration in Burkina Faso 2.9% of the population moved between provinces in 1996, 4% in 1985 (Source : INSD, 2000). First rural out-migration (1970-98) Duration 4 out of 5 migrations are long-term (>= 2 years) Destination 48% to Rural areas 14 % to Urban areas 38% to Abroad (mostly to Côte d Ivoire) Migration motives Male Economic reasons Female Family (Source: EMIUB, 2000 in Henry et al., 2004) 14
Methods and data
Data Migration dynamics, urban integration and environment survey in Burkina Faso (EMIUB) Individual life history data (2000) National survey, 8644 individuals A complete migration history since the age of 6 1972 1975 Male, born in 1966, farmer House characteristics Migration purpose Matrimonial status etc. 1980 1996 16
Data Migration dynamics, urban integration and environment survey in Burkina Faso (EMIUB) Individual life history data (2000) National survey, 8644 individuals A complete migration history since the age of 6 Community survey (2002) Each village where at least 3 spells of residence 600 villages out of 1800 17
The 600 settlements of the survey Distance to market Presence of primary school Migration history Ethnic groups Crops etc. 6 % of villages in Burkina (< 5 000 people) 60 % of medium-sized villages (5 000-10 000 people) 86 % of towns (> 10 000 people) 18
Data Migration dynamics, urban integration and environment survey in Burkina Faso (EMIUB) Individual life history data (2000) National survey, 8644 individuals A complete migration history since the age of 6 Community survey (2002) Each village where at least 3 spells of residence 600 villages out of 1800 Rainfall data (CRU, 2000) 0.5 x 0.5 Monthly rainfall data extracted at the department level (GIS) 19
Rainfall pixels (0.5 x 0.5 ) Monthly rainfall data extracted at the department level by using a GIS Rainfall in JAN 1960 Rainfall in FEB 1960 Rainfall in DEC 1998 20
1996 1972 1980 1975 Male, born in 1966, farmer House characteristics Migration purpose Matrimonial status etc. Distance to market Presence of primary school Migration history Ethnic groups Crops etc. GIS Rainfall in JAN 1960 Rainfall in FEB 1960 Rainfall in DEC 1998 21
Migration model Multinomial logit model: limit to a few destination choices Use of a random utility model (estimated by a discrete-time logistic regression) An individual is able to evaluate the utility associated to each potential destination and to choose the place that maximizes his utility. The choice of moving from location j to location k depend on: the attributes of locations j and k; the attributes of many other potential locations. k a b j 22
Variables Period 1990-2000 Dependant variable Last male migration (more than 3 months) before the survey (2000) from a village or a medium-sized town to a village or m-s town after age 15 Control factors at the individual-level Age, education, ethnic group and main economic activity (TV) Control factors at the community-level Size of the settlement (TV) Distance to the two main cities Presence of road in the settlement (TV) Sample 1801 males including 217 migrants. 23
Destination choice Migration Intra-provincial move Inter-provincial move 24
Results
#1. A favourable natural environment is a pull factor of major importance for migrants. Mean annual rainfall (1960-98) Drought frequency: time-varying deviation from the mean (3 preceding years) - (TV) Rainfall (mm) 820 800 780 760 740 720 700 680 660 640 620 Migration 1960 1965 1970 1975 1980 1985 1990 1995 Mean annual rainfall (60-98)
Mean annual rainfall at the destination ( agricultural productivity) Attraction of the drier regions (return migration?) Odds ratio 6 5 4 3 2 1 0 The risk to chose this region is 3x higher than for the 900mm region 200-499 mm 500-699 mm 700-899 mm 900 mm and over Mean annual rainfall at the destination (Ref.) Intra-prov. Inter-prov. Significant result 27
1,2 Rainfall deviation at the destination ( drought) Odds ratio 1 0,8 0,6 0,4 0,2 0 If drought: repulsion < 85 % 85 94 % 95 and over Odds ratio 40 35 30 25 20 15 10 5 0 No significant results for longdistance moves < 85 % 85 94 % 95 and over Rainfall deviation at the destination (Ref.) = normal rainfall (Ref.) = normal rainfall Significant result 28
#2. The economic diversification of villages attracts migrants. Presence of income-generating activities: gardening, cash crop, paid farm employment, paid non-farm employment (TV) Modernization of agriculture: plow, tractor, irrigation, use of water-conservation techniques (TV)
The presence of income-generating activities Attraction 2,5 Odds ratio 2 1,5 1 0,5 Repulsion Intra-prov. Inter-prov. 0 Gardening Cash crop Paid agricultural employment Paid nonagricultural employment Presence of (var) at the destination Ref. =absence of (var) in the destination village Significant result 30
Modernization of agriculture 3 Attraction 2,5 Odds ratio 2 1,5 1 Repulsion Repulsion Intra-prov. Inter-prov. 0,5 0 Tractor Plow Irrigation Water-cons. tech. Presence of (var) at the destination Ref. =absence of (var) in the destination village Significant result 31
Example of a waterconservation technique in the north of Burkina Faso Source: Marceau-Rochette, 1989. 32
Conclusions
Conclusions Few studies on pull effects of migration in an African setting Place utility approach (random utility model) #1. A favourable natural environment is a pull factor of major importance for migrants. ONLY for short-distance mig - Repulsion if drought - Attraction of the drier regions #2. The economic diversification of villages attracts migrants. @ Presence of income-generating activities - Attraction of gardening - Repulsion of paid employment @ Modernization of agriculture - Attraction of tractor - Repulsion of irrigation and water-conservation techniques 34
Environmental versus economic migrants? Source: Black R et al. in Nature, 2011, vol. 478. 35
Additional questions Models by categories of migrants (motives, main activity, ethnic group, age group, etc.) but limitation by the size of the sample; Perceptions of characteristics at the destination (transmission of information, groups of migrants, etc.); Comparison with characteristics at the origin (benefit for migrant); Social networks; Interactions between social, economic and environmental factors. Comments? sabine.henry@fundp.ac.be 36