rural-urban migration and its implications for food security in bangladesh rural-urban migration and...

32
Rural-urban Migration and its Implications for Food Security in Bangladesh Institution: Department of Statistics, Shahjalal University of Science and Technology (SUST) Research Team: Dr. Md. Zakir Hossain, PI Professor, Dept of Statistics, SUST Dr. M. Mizanul Haque Kazal, CI Chairman, Dept of Development & Poverty Studies, SAU Mr. Jasim Uddin Ahmed, CI Associate Professor of Economics, Moulvibazar

Upload: elfrieda-louise-hutchinson

Post on 26-Dec-2015

224 views

Category:

Documents


1 download

TRANSCRIPT

Rural-urban Migration and its Implications for Food Security in Bangladesh

Institution: Department of Statistics, Shahjalal University of Science and Technology (SUST)

Research Team: 

Dr. Md. Zakir Hossain, PIProfessor, Dept of Statistics, SUST

Dr. M. Mizanul Haque Kazal, CIChairman, Dept of Development & Poverty Studies,

SAU

Mr. Jasim Uddin Ahmed, CIAssociate Professor of Economics, Moulvibazar Govt.

College

CHAPTER I INTRODUCTIONCHAPTER II METHODOLOGY2.1 Research Methods 2.2 Analytical Tools and TechniquesCHAPTER III LITERATURE REVIEWCHAPTER IV HOUSEHOLD PROFILE AND MIGRATION DIVERSITY 4.1 Profile of Surveyed Household4.2 Diversity of Migration StrategyCHAPTER V STATUS AND PREDICTORS OF FOOD INSECURITY5.1 Perception, DCI, Food Expenditure and CSI Methods5.2 General Coping Strategies to Food Insecurity Situation5.3 Predictors of Household Food InsecurityCHAPTER VI IMPACT OF RURAL-URBAN MIGRATION ON

HOUSEHOLD FOOD SECURITY STATUS 6.1 Consequence of Migration on different Household Indicators and

Fulfillment of Expectations6.2 Impact of Migration on Food Security through NELM ModelsCHAPTER VII CONCLUSION AND RECOMMENDATION

Contents of the Report

Key Research Questions(i) Who migrates? Are the food insecure households more

prone to migration?(ii) What are the different types of migration including child

migration? (iii) Are the rural-urban migration flows concentrated towards

some big cities only?(iv) Is the food security status of migrant households different

from that of non-migrant households and how?(v) What are the Asian evidences and policy implications of

rural- urban migration associated with food security?(vi) What is the impact of migration on food security status at

origin?(vii)What strategies are to be taken to optimize the rural-urban

migration outcome to food security?

Objectives and Expected Outputs

Objectives • To sharpen policy-makers’ understanding of the diversity of rural to

urban migration strategies and their impacts on household and individual food security in sending and receiving areas,

• To provide information on potential interventions to strengthen migrant household food security.

Expected Outputs(i)Identification of migration diversity through exploring the typology

and mapping of its patterns,(ii)Determination of food security status of the migrant households and

impact of migration on their food security, (iii)Review of Asian literature to find out evidence on the association of

food security, rural urban migration and poverty reduction,(v)Formulation of intervention to address food security of migrant

households.

Data and MethodologyThe study has adopted following techniques to gather the primary data:•Household-level survey at origin•Tracer survey at destination•Focus Group Discussion (FGD)•Key Informant Interview (KII)In addition, the study has analyzed the HIES-2010 and Panel data of Dr. Mahbub Hossain to compare the relevant findings of the survey data.

Sample Design for Household Level Survey & Tracer Survey•The study provided the main Indicators in 2 rural domains, according to the East-West divide reported by the World Bank study of Bangladesh

•The sample size determination formula yields that 750 migrant households are required to cover in each domain.

•The study has adopted cluster (PSUs of BBS) sampling and covered 30 clusters in each domain (using systematic PPS)

•Ultimately, 3000 units of analysis have been covered:

- 1500 Migrant and 750 Non-migrant households at origin

- 500 Migrant and 250 Non-migrant households under Tracer survey

Analytical TechniquesStep-1: Measure the flow of migration including selectivityThe flow of rural-urban migration has been explored through

descriptive statistics in different dimensions including flow-mapping.

Step-2: Measurement of the status of food security(i) Perception analysis; (ii) Direct calorie intake (DCI); (iii) Food expenditure; (iv) Coping strategy index (CSI) score

Step-3: Identification of the predictors of food insecurityThe binary multiple logistic regression model (BLRM) has been

applied to identify the predictors of food insecurity.

Step-4: Impact of migration on food security The impact of migration on food security at origin has mainly

been studied using 2-stage and 3-stage NELM models In particular, instrumental variable (IV) regression has been

employed to study the impact of migration on per capita calorie intake.

The similar model has also been used by simultaneous consideration of migration determinants, remittance behaviors and income.

2-Stage NELM Models

3-stage NELM Models

The basic equation for household income, (as a proxy of household food security) according to the NELM hypothesis can be expressed as

k = on-farm, off-farm;

whereTo model migration, this study considers using count regression functional form, particularly the Poisson distribution form, since the number of migrants is non-negative.

kkkkkokk

c ZRMY 321

RRo ZMR 21

MMZfM );(

MMM ZZf )exp();( 10

Stage 1:

Stage 2:

Where FS measures the per capita calorie-intake as a proxy food security status

at the household level; Mig measures the number of migrants per household; X

vector encompasses the household characteristics; Z is a vector of instruments

Concluding Remarks from Review of Asian Evidence

Positive Impact At origin: Positive Impact at Destination:-Poverty reduction & gain of assets- Human resource development- Food security at household level- Women empowerment- Fertility control - Change in family composition

- Urban growth - Availability of workers for urban services-Availability of manpower for industry, particularly manufacturing, development sector

Negative Impact At origin: Negative Effect at Destination:- Labor depletion in a few cases- Loss of agricultural productivity- Family disintegration-Increase of women and children chores

- Unplanned urbanization- Growth of slums- Environment pollution- Ultimately, ill consequences on transport, healthcare & other service sector

Migration Pattern/Diversity/Causes:- Mainly young people strive for migration- Migration is higher from environmentally fragile areas- Increasing trend of temporary and circular migration- Feminization of internal migration- Concentration of migration towards big cities- Major causes of migration are wage differentials, population pressure etc.

About 29.2% study population were below 15 years of age,

about 66% fell 15-64 years and about 4.4% were of 65 years or

above.

Approximately 21% of the adult men earned incomes from

agriculture, about 22% were unemployed, 12.6% engaged in

study and the rest earned from off-farm activities. On the

contrary, three-quarters of the women were engaged in

household work, 9% engaged in study, about 11% were

engaged with agriculture.

About 21% of the study population aged 5 years or older was

found to have no education, about 36% were found to have

primary-level education, about 39% were found to have

secondary/higher secondary level education and only 5%

people attained graduate level education.

The analysis of housing condition, sources of water & lighting

and sanitation facilities as well as asset score of the surveyed

households according to the migration status indicates that the

migrant-households are better positioned than their non-

migrant counterparts in terms of housing condition and

possessing wealth.

General Features of the Study Population for RUM-2012 Survey data

Landholdings, Income & Expenditure Pattern of Migrant and Non-migrant Households for Three Data Sets

Landholding, Income & Expenditure

RUM-2012 (origin) HIES-2010 Panel Data

Migrant Non-migrant Migrant Non-

migrant Migrant Non-migrant

Average Size of Land 114.3 73.9 108.6 76.9 160.9 92.6P-value P<0.01 P<0.10 P<0.01

East (Mean) 88.3 70.2West (Mean) 140.1 77.6Average On-farm Income 53720 46733 47694 55583 61453 50910P-value P<0.12 P>0.10 P<0.10

Average Off-farm Income 115629 89014 86742 77144 55855 33413P-value P<0.01 P<0.10 P<0.01

Average Food Expenditure 73205 68157 72306 71241P-value P<0.05 p>0.10

Avg. Non-food Expenditure

68345 51306 49801 44871

P-value P<0.01 P<0.10

Total (n) 1509 746 427 6622 424 1352

Flow of Migration in Bangladesh

Percentage of households reporting migration of any member by residence

Origin Data Source Period

DestinationTotal

householdsDomestic Abroad Mixed

Rural HIES-2010 2006-2010 4.84 9.25 - 7840

Urban HIES-2010 2006-2010 1.62 6.85 - 4400

Rural RUM-origin 2006-2010 9.29 - - 8033

Rural RUM-origin 2001-2011 22.25 11.47 1.90 8033

Rural-East RUM-origin 2001-2011 19.20 15.70 2.40 4793

Rural-West RUM-origin 2001-2011 27.20 4.70 1.10 3240

Rural-Urban Migration Flow from Study Clusters/Villages in Bangladesh

Migration Rate according to East-West Divide

East 4.38%West 6.86%Total 5.30%

Are the food insecure households more prone to migration?Comparison of some basic indicators between migrant (non-migrant in 2000 and migrated during 2001-2008) and non-migrant households (non-migrant since 2000) using panel data

Basic Indicators Migrant Households

Non-migrant Households

Average Landholdings (in decimals) 149.12 116.81Average Per Capita Income (in Taka) 10673 9009Calorie intake (% of HHs)Extreme poor (1805 K.Cal) 18.7 22.0Moderate poor (2122 K.Cal) 25.1 31.1Non-poor 56.2 47.0Poor on the basis of self perceptionExtreme poor 9.2 13.0Poor 32.2 35.9Vulnerable 35.7 36.2Solvent 23.0 14.9Total Households (n) 283 1133

Selectivity: Age and Education at the time of Migration Age distribution of migrants for survey data and panel data

Educational status at the time of migration for Three Data Sets

Selectivity: Occupation of the MigrantsOccupation Pre-migration period (%) Occupation at Destination (%)

East West Both East West Both

Job/service 5.6 2.5 4.0 32.2 28.5 30.3Business 2.4 1.5 1.9 3.6 2.7 3.1Petty traders/ hawker

4.5 1.9 3.1 7.9 5.5 6.7

Garment worker 1.1 .8 1.0 10.7 14.8 12.8Wage labourer 13.5 7.6 10.5 21.3 17.3 19.3Student 36.6 38.1 37.4 15.6 20.2 18.0Household work 4.9 4.7 4.8 4.6 4.9 4.8Agriculture 10.9 21.5 16.4 1.1 2.0 1.6Unemployed/Others

20.6 21.3 21.0 2.8 4.0 3.3

Total (n) 801 849 1650 801 849 1650

Destination of the migrants by region

Intended Nature of MigrationType of Migration % of Migrants

Permanent 6.7Study 21.7Temporary 66.9Seasonal 4.7

Exploration of the child migration:Comparison of Pre-migration occupation and occupation at

destination of the child migrants

Pre-migration occupation Occupation at destination

Factors for Migration

Push Factors % of MigrantsPoverty/food insecurity 23.9Unemployment 30.4Underemployment/ Demonstration effect 11.4Insufficient education facilities 23.1Others 11.2

Pull Factors % of MigrantsBetter job opportunities 39.9Better schooling 22.8Relatives/friends there 13.8Wage differential 13.6Others 9.9

Extent of Food Insecurity: Perception, DCI and CSI Methods

Perception Method

At Origin At Destination

Migrant HHs (%)

Non-Migrant HHs (%)

Migrant HHs (%)

Non-Migrant HHs (%)

Had been anxious about food in last 3 months (normal insecurity)

21.4 27.2 6.2 9.0

Had been bound to take less than 3 meals in a day (moderate insecurity)

12.8 16.2 2.6 0.8

Had been bound to sleep in hunger (severe insecurity)

5.0 7.0 1.0 0.4

No food insecurity 78.6 72.8 93.8 91.0Extent of Poverty/Food Insecurity by DCI Method% of households below hardcore poverty line

13.5 18.1 5.2 8.2

% of households below absolute poverty line

32.6 38.3 29.7 35.9

Food Insecurity Status by CSI Method Low/Medium CSI Score 12.3 15.0 3.6 5.8High CSI Score 9.1 12.2 2.6 3.1Average CSI Score 30.3 30.4 26.0 27.8Total (n) 1509 746 499 256

Figure 5.2: Percentage of Households using Consumption Coping Strategies

Strategies

a. Rely on less preferred and less expensive foods

b. Undertaking more jobs and/or working longer

hours

c. Borrow food or rely on help from a

friend/relative

d. Purchase food on credit

e. Gather wild food, hunt or harvest immature crops

f. Consume seed stock held for next season

g. Send household members to eat elsewhere

h. Sending children for working

i. Send household members to beg

j. Limit portion size at mealtimes

k. Restrict consumption by adults in order for

children to eat

l. Feed working members at the expense of non-

working members

m. Reduce number of meals eaten in a day

n. Skip entire days without eating

88.8

68.360.5

87.1

10.6 7.4 8.914.1

6.8

49.2

17.510.1

29.5

8.9

0102030405060708090

100

a. b. c. d. e. f. g. h. i. j. k. l. m. n.

% of ho

usehol

d

Adopted coping strategies during food insecurity

Predictors of Food Insecurity: Relative risk against different categories of the covariates for different levels of food insecurity

Variables Model 1:Normal food insecurity

Model 2:Moderate food insecurity

Model 3:Severe food insecurity

LandholdingNone® 1.000 1.000 1.0001-49 0.346*** 0.340*** 0.385**50-99 0.149*** 0.124*** 0.088***100-249 0.094*** 0.085*** 0.060***250 & above 0.048*** 0.088*** 0.050***Occupation of the Household HeadAgriculture Farmer ® 1.000 1.000 1.000

Labor 1.628*** 1.506* 1.171 Service 0.618* 0.591 0.397Others 1.537** 1.518* 1.825Education of the Household HeadIlliterate® 1.000 1.000 1.000Primary 0.792* 0.633*** 0.604**Secondary 0.664*** 0.478*** 0.462***Post-secondary & above 0.315*** 0.283*** 0.289

® Ref. category; *** Sig. at 1% level; ** Sig. at 5% level; * Sig. at 10%

Cont.

Model 1:Normal food insecurity

Model2:Moderate food insecurity

Model3:Severe food insecurity

Debt Status of the HouseholdDid not receive loan® 1.000 1.000 1.000Received loan 2.484*** 2.699*** 2.432***Location (East-west divide)West® 1.000 1.000 1.000East 0.847 1.255* 2.547***Dependency ratio 1.004*** 1.004*** 1.006***Family Size 0.961 0.892*** 0.818***Tilling TechnologyNon-mechanized 1.000 1.000 1.000Mechanized 1.063 1.106 1.094*Sex of the Household HeadMale 1.000 1.000 1.000Female 1.822*** 1.994*** 1.899*Constant 0.733 0.433** 0.135***® Ref. category; *** Sig. at 1% level; ** Sig. at 5% level; * Sig. at 10%

Impact of Migration on Different Household Indicators (Perception-based)

Socio-economic indicators

Impact/Change due to migration (% of households)At Origin At Destination

Positive Negative Positive Negative

Frequency of meal 4.8 0.7 4.4 3.4Quality of food 12.0 3.2 26.5 11.4Food security condition 17.8 5.2 29.7 10.6Educational status of the migrants 14.5 0.0 16.6 0.0

Working hours of the household members

25.7 10.6 35.7 12.2

Land Size (in Decimal) 5.4 70.8 1.8 31.7Type of House 6.1 1.1 58.1 12.2Number of rooms 5.3 0.9 11.2 40.1Type of toilet 6.2 1.4 47.9 10.2Television 9.0 2.5 16.8 17.4Fridge 2.9 0.5 8.4 4.8Economic condition of the HH

13.4 2.6 17.8 4.6

Labor compensation and fulfillment of expectations due to migration

East West BothAnnual cost of labour compensation due to migration (%)Yes 19.8 4.5 12.5Average amount (Tk.) 5760±4741 4157±2976 5481±4521Fulfillment of expectations (based on responses against key migrants)Don’t know 7.1 2.8 4.9Satisfactory 51.5 61.6 56.6Partially satisfactory 33.3 29.8 31.5Unsatisfactory 8.1 5.8 7.0

Involvement of women and children in economic activities due to migration

Increase 27.2Decrease 8.8Constant 59.0

Impact of Migration on Food Security through 2-Stage NELM Models

Dependent Variable: Logarithm of per capita calorie intakeEndogenous Variables: Number of migrants in the householdInstrumental variables: Migration network, Share of male/female in 16-39 age groups at household level

VariablesGMM estimates of IV regression

Coefficient P>zNumber of migrants of Household .021872 0.002Total operative land of Household .0000477 0.034Occupation of HH (Farming) .0265117 0.048Occupation of HH (Labourer) -.0117138 0.401Occupation of HH (Service) .0118429 0.493Education of HH (Years of schooling) .0029068 0.004Age of Household Head .0006258 0.072Region (East=0, West=1) .0020233 0.787Housing (poor quality=0, Good quality=1) .0127442 0.146Religion (Muslim=0, Non-Muslims=1) .0194966 0.080Score of household durables .0000245 0.931Sex of the household head (Female=1) .0304034 0.098

Cont.Variables Coef. P>zNumber of adult male members -.011606 0.024Tilling technology (mechanized=1) .0116239 0.291Distance from commercial centre .0002534 0.197Cropping nature of land (multi-crop=1) .0046391 0.637Constant 7.592765 0.000No. of observations 2255

R-squared centered & R-squared uncentered 0.0375 & 0.9995

F-test 6.23 (P-value=0.000)

Tests of overidentifying restrictions:

Sargan N*R-sq test statistic 1.433 (P-value= 0.2313)

Basmann test 1.422 (P-value= 0.2330)

Hansen-J-Statistic 1.488 (P-value= 0.2225)Test for endogeneity of Number of Migrants (Ho: Regressor is exogenous)

Wu-Hausman F-test; F(1, 2237) 2.968 (P-value= 0.085)

Durbin-Wu-Hausman chi-sq test; chi-sq(1) 2.988 (P-value=0.084)

Impact of Migration and Remittance on Household Income using 3-Stage NELM Models

Dependent Variable: On-farm income; Off-farm income without remittanceEndogenous Variables: No of migrants in the HH; Amount of annual remittanceInstrumental variables: Migration network, Share of male/female in 16-39 age groups at household level, Relation of the migrants with household head

CONCLUSIONS

•In rural Bangladesh, over half of the households are functionally landless and the migrant-households are economically better positioned.•Internal migration flow is higher from the West and international migration is higher from the East.• Food insecure or poor are not more prone to migration. •The migrants tend to concentrate to capital city and district headquarters.•The young people (aged 15-29 years), males and sons/daughters of the household heads are more exposed to rural-urban migration.•Temporary migration dominates over other types of migration and independent/single migration over family migration.•Literate people are more prone to permanent migration and illiterate people are more prone to temporary migration.•Type of migration from food insecure households significantly differs from that from food secure households.•Poverty, unemployment and poor educational facilities are found out as the main push factors. In contrast, better employment opportunities, better schooling, and wage differentials are sorted out as the main pull factors.

Cont…

•About three-fifths of the migrants sent remittances (average amount of Tk.38397) and household heads mainly decide to utilize remittances. •All the estimates of food insecurity indicate that non-migrant households are more vulnerable to food insecurity. The estimates also show that the migrants at destination are significantly less food insecure than their origin counterparts.•Reliance on less preferred and less expensive food items and purchase of food on credit are explored as the top coping strategies of food insecurity.•Landholding, occupation and education of the household head, debt status, location, dependency ratio, family size and household headship are determined as predictors of household level food insecurity.•Perception-based estimates on change pattern of some socioeconomic indicators reveal that migration puts mixed impact with positive net impact.

•NELM-models determine that rural-urban migration exerts significantly positive impact on food security through different dimensions.

•The major findings covered by the review of Asian literature are corroborated by the findings of the present study.

Recommendation•Improve educational facilities for quality education in rural areas, which matches the skill needs of rural labour markets including vocational and training facilities.

•Promote programmes to enhance more and better opportunities for employment and entrepreneurship development in rural Bangladesh to provide alternatives to distress rural-urban migration, especially for youth.

•The Government should also invest in better labour market information systems and job information services so that youth can access to better jobs and undertake migration in a more informed manner.

•Actions to prevent and eradicate child labour, especially its most hazardous forms, with particular attention to unaccompanied child migrants.

•Decentralize both administrative and developmental activities to discourage the concentration of rural-urban migration to capital city in particular and other big cities in general.

•Local government should make proper arrangements for safety and security in rural areas for a sound environment to make rural stay hassle and anxiety free to discourage rural-urban migration and encourage reverse migration.

Overall Message of the Study1) The study explores that different kinds of adversity at origin compel a

stipulated section of rural population to strive for migration to urban end where different kinds of prosperity attract them.

2) Different impact determining analyses based on descriptive, perception-based and model-based estimates identify that impact of rural-urban migration is significantly positive on food security in particular and overall living condition in general.

3) The majority of the findings of the present study are found to be aligned with those extracted from the review of literature, particularly on Asian countries.

4) One final message is that since migration is a revealed preference of the migrants, it cannot be stopped; rather, the policy-makers can make policies to better manage migration flows by providing support to leverage the opportunities arising from migration and remittances.