The changing landscape of international migration: evidence from rural households in Bangladesh, 2000–2014

ABSTRACT

Evidence suggests that the poor cannot readily take part in overseas employment, largely because of high placement costs and poor access to job information. Using unique data on rural households in Bangladesh for 2000, 2008 and 2014, this study explores the socioeconomic characteristics of households which decide to send family members to work abroad and evaluates their changes over time. Analysis shows that the education of household members, asset holdings and social networks have been positive correlates of migration, particularly in the past. More recently, however, less educated and poorer households with weak social networks have begun taking part slowly as entry barriers have decreased. The possible causes for such change include the strong presence of recruitment agencies, persistent demand for low-skilled workers in major destination countries, growing domestic labor demand favoring educated workers, and better access to non-collateral loans and grants to finance migration.

KEYWORDS:

• Asia
• Bangladesh
• human capital
• migration networks
• migration costs

Acknowledgments

This work was supported by the Japan Society for the Promotion of Science (JSPS) KAKENHI Grants Number JP15J11506 and Number 25101002.

We thank late Dr. Mahabub Hossain for granting access to data for this research. We also wish to thank Kazushi Takahashi and Tomoya Matsumoto for their insightful comments and suggestions. All errors remain ours.

Disclosure statement

No potential conflict of interest was reported by the authors.

Correction Statement

This article has been republished with minor changes. These changes do not impact the academic content of the article.

Notes

1. This paper is developed from the work presented in the author’s doctoral thesis (Kikkawa., 2017) and a previous GRIPS discussion paper (Kikkawa & Otsuka, 2016).

2. SDG 10 on reducing inequality within and among countries includes two relevant clauses on migration: (1) facilitating orderly, safe, regular and responsible migration and mobility of people, including through the implementation of planned and well-managed migration policies; and (2) by 2030, reducing to less than 3% the transaction costs of migrant remittances and eliminating remittance corridors with costs higher than 5%.

3. The Latin American Migration Project (https://lamp.opr.princeton.edu/) contributes to generating longitudinal household panel data of migrant source communities.

4. Overseas employment also incurs physical and mental costs that emanate from harsh work environments and separation from family and home (Chen et al., 2019), as well as the risk of being cheated (Das et al., 2018). Our survey data, which will be described in a later section, show that a total of 31 households (1.3% of the total households surveyed in 2014) reported cases of fraud relating to overseas jobs, with the damage ranging from Tk6,000 to Tk900,000 ($81 to $12,922).

5. $1 = Tk69.65 based on the 2010 official exchange rate from World Development Indicators, and all prices mentioned are price-adjusted at the 2010 level using the consumer price index.

6. For example, the Bangladesh Rural Advancement Committee, the country’s largest microcredit institution and Probashi Kallyan Bank, a state bank, provide such loans.

7. Our survey data also show that availability of non-collateral loans increased over time.

8. Except for highly skilled workers, who are allowed to bring families and have the right to residency in some destinations, for example, Singapore.

9. It selects one representative village per district and selects household samples within each village.

10. Household members is defined here as the those who share residence and daily meals. Migrant members are those who belonged to the household in the past but left the residence at the time of the survey and lived elsewhere (within or outside of Bangladesh). For our study, we only focus on households with international migrants who work abroad.

11. This is comparable to the result of the government’s Household Income and Expenditure Survey of 2005, which showed that 9% of households have migrant(s) abroad (Raihan et al., 2008).

12. Average land owned by any household declines as population increases (0.6 hectares in 2000 to 0.4 in 2014), which may partly explain the narrowing gap between the two groups.

13. See next section for more discussion on these network variables.

14. Almost all international migrants in the survey sample were male, with only four women in 2009 and eight in 2014. The low participation of women was consistent with the government departure statistics (Bureau of Manpower Employment and Training (BMET), Bangladesh, 2018). Most movement was a temporary migration of the worker himself and did not involve the worker’s entire family.

15. Between 2000 and 2013, mean schooling years among adults in Bangladesh increased from 3.7 to 5.1 years (United Nations Development Program (UNDP), 2016).

16. We also estimated the equation with a probit model clustering the errors at village level and found comparable results (Appendix Table A1).

17. We also fitted a squared term of land asset (McKenzie & Rapoport, 2010; VanWey, 2005), but the coefficient was insignificant.

18. We did not use the household head education level to represent the household education level, because migrant households often designate female spouses or parents as heads.

19. Education level can also be endogenous if the migrant job is unskilled, because migrant candidates are pulled out of school to work abroad (McKenzie & Rapoport, 2010).

20. This specification removes 507 households from our sample. Inclusion of these households in our analysis as an additional reference group does not significantly alter our main result.

21. The average years of education among sample households in those years range from 3 to 4 years (Table 1).

22. We also conducted a sensitivity check replacing the owned land size with that of inherited land, but our results remain unchanged.

23. The analysis was conducted using samples for 2014 only due to data availability.

24. Analysis using the year 2000 sample shows that migrant households are more educated than non-migrants and that government statistics show that large portion of migrants during this period headed to the Middle East and East Asia (Bureau of Manpower Employment and Training (BMET), Bangladesh, 2018).

25. Since 2010, the Bangladeshi economy has grown more than 6% annually, and unemployment has declined from 5% in 2009 to 4.3% in 2014 (World Bank, 2017; Zhang et al., 2013).

26. The instrumental variables are the number of workers per household and lagged international and domestic migration networks.

27. Includes costs of obtaining travel documents, domestic and international travel, and agent/handling surveyed in 2000, 2008 and 2014.

28. See Appendix.

29. Those households which accessed non-collateral loans are poorer than other migrant households or non-migrant households when the size of owned land or education level is considered.

Additional information

Funding

This work was supported by the Japan Society for the Promotion of Science [25101002, JP15J11506].

Notes on contributors

Aiko Kikkawa

Aiko Kikkawa is an Economist at the Economic Research and Regional Cooperation Department, Asian Development Bank. Her research interest includes migration and development and population aging.

Keijiro Otsuka

Keijiro Otsuka is a Professor of Development Economics at the Graduate School of Economics, Kobe University, and a Chief Senior Researcher at the Institute of Developing Economies in Tokyo. His research interests are in agricultural and industrial development of Asia and Africa.