Remittances and household expenditure behavior in Sri Lanka
Loading...
Date
2016-07-28
Authors
Ekanayake , P. O.
Dorabawila, S.
Journal Title
Journal ISSN
Volume Title
Publisher
University of Peradeniya, Sri Lanka
Abstract
Introduction
International remittances are money and goods received by the home communities which are transmitted by people working abroad. It is one of the most visible developmental effects of migration. In 2009 the average per capita remittance to Sri Lanka amounted to US$ 164 compared with South Asia’s average of 35 (Ministry of Finance and Planning).
There are few studies conducted on how remittance money is utilized in Sri Lanka. Prabal (2012) using Sri Lanka Integrated Survey (SLIS) data shows that Sri Lankans move up the income ladder using remittances. Caritas (2012) finds instances of husbands mismanaging and squandering remittances sent by female migrant workers. The current research expects to contribute to the literature by investigating how household expenditure changes with the receipt of remittances. Particular attention was paid to the creation of human capital1 arising from remittances. Expenditures on education and health directly help create human capital, by increases access to higher quality education and creating a healthier population, respectively. Many researchers have shown a positive link between expenditure on health and economic growth.
The research objectives were to examine whether:
* The gender of remitter makes an impact on the household budget allocation
* On average, the receipt of remittances makes an impact on the household budget allocation
Methodology
Data source for this research is Sri Lanka Integrated Survey (SLIS). All the figures found in SLIS are in monthly figures. The study sample consisted of 403 remittance-receiving households and 517 non- remittance receiving households, adding up to a total of 920 households. The analysis was carried out in 2 steps. First, it measured the gender effect and remittance effect, keeping in mind that the functional form should be able to accommodate a number of expenditure categories and that non-random treatment of migration could cause biases. (Migrant and non-migrant may households differ in terms of many demographic variables causing endogeneity (Guzman et al. 2006)).
An adjusted Working-Leser model was chosen. In addition, Propensity Score Matching (PSM) was carried out to test the consistency of the Working-Leser model. The dependant variable reflects categories of household expenditures which are bound between 0 and one. The model is a standard fractional Logit model, as shown below.
< mathematical formula > 1
< mathematical formula > 2
where
Wᵢⱼₕ - Fraction of the budget devoted to expenditure category i by household h and gender j.
GEN - Dummy variable which equals unity for those households having a male remitter and zero for households that have a female remitter.
REM- Dummy variable that equals unity for households receiving remittance and zero for households not receiving remittance.
Xₕ - Total household expenditure
n ₕ- Number of people in the household
𝛿ₕ - Vector of household characteristics that may affect household behavior
The aforementioned independent variables are the same treatment independent variables that were used in propensity score matching.
< mathematical formula > 1
< mathematical formula > 2
Equation 3 was used to test the effect of gender impact and equation 4 was used to test the remittance impact. The treatment is expressed through a dummy variable Dj equal to one if household j receives remittance from a male and zero if remittance is received from a female. Kj is equal to one if household j receives remittance from a female and zero if household j does not receive remittance. Let Yij1 and Yij0 indicate the outcome variables representing the allocated amount of budget in good i for household j. This applies both to the presence and the absence of treatment. In this case the treatment effect is measured which is the difference in the outcome for unit j with the treatment and the one without the treatment.
Results and discussion
Results from the first model indicate that the gender of the remitter does impact on the household budget allocations. Allocation for food decreases by 1.5 per cent in male remitter households compared with female remitter households. Share of household budget for education rises by 1% more in male remitter households. The other highlight is liquor which shows a 0.8% decrease in male remitter households. Allocation for clothing is benefitted by remitter being male by an increase of 0.46%.
Results from Model 2 suggest that expenditure categories like food and housing are affected by the treatment of migration. The impact on housing category stands out. Fraction of housing expenditure is 1.7 per cent higher in remittance receiving households which is significant at 1 per cent. Food expenditure appears to drop by 0.7 per cent in remittance receiving families. The results from propensity score matching show consistency with the Working-Leser curve results.
Conclusion
(As per Model 1) the gender of the remitter matters in household budget allocations. It influences expenditure categories which are vital for the creation of human capital (education) in Sri Lanka. In addition, clothing and ceremonial expenditure rises with a male remitter. Arguably it shows that a male remitter spurs conspicuous expenditure. In male remitter households the fraction of the budget allocated for liquor and food fall. These results do not provide conclusive evidence on which gender would benefit a family to increase the quality of human life. If the emphasis is on education, encouraging males to migrate is advisable instead of female migration. That would also help to reduce the expenditure on liquor which generally causes adverse effects on a household’s welfare. On the other hand female migration can be encouraged if the expenditure on food is planned to be increased.
The main financial motive for migration appears to be housing which includes housing rent, household maintenance and repairs. This is highlighted in Caritas (2010) where they identify construction of houses as the main financial motive behind migration. Rather than increasing expenditure on food, education or health which directly influences the creation of human capital, the remittance receiving families allocate more on housing expenditure. In the decision on how to allocate funds for health, liquor, education, and clothing, remittances are regarded as any other source of income. Model 2 suggests that the remittance receiving family’s needs are more tangible ones rather than funding for education and health. Interestingly, budget allocation for food is less in remittance receiving families. Plans should be made to divert remittance money into expenditure categories which are more productive and help to create human capital. It seems that the negative effect of a male remitter made on the budget share of food is greater than the positive effect made by a female remitter.
Description
Keywords
International remittance
Citation
Proceedings of the International Conference on the Humanities and the Social Sciences (ICHSS) -2016 Faculty of Arts, University of Peradeniya. P. 12- 15