Remittances and direct impact on households
The impact of remittances on development is still an area open for debate and further analysis but most experts agree that remittance flows in the form of direct transfers to households raise household incomes and have the potential to contribute to overall poverty reduction.
However, it also must be noted that due to high costs, most labour migrants are from lower middle-income groups, instead of the poorest levels of society (Kapur 2004). Several surveys have shown that many migrant households already owned a house before one of its members migrated (Arif 2010, IOM 2009, Siddiqui 2004). Nevertheless, unemployment and perceived poverty seem to be the main drivers to migrate. In contrast, migrants from urban areas often cite unemployment as the main reason to migrate, while those from rural areas tend to migrate to achieve a specific purpose, such as to build a house (Arif 2010).
A significant portion of remittances is used to repay loans taken to cover the cost of recruitment. During the first period of migration, which can last from four months to two years, a migrant’s entire income often goes to repay loans. Consequently, research on development impacts of remittances must extend beyond the initial payback period in order to draw accurate results. Moreover, the cost of migration also determines whether and how soon there will be development impacts of remittances.
Surveys from a number of locations in South Asia (Bangladesh, Pakistan, Sri Lanka and the Indian state Kerala) reveal that migration improves the economic status of migrant households. Depending on the method used, some surveys found that migrant households are generally better off than non-migrant households, while others indicated after comparing households before and after migration, that the majority of them improved their economic status after migration. The gains in economic status were even greater for households with professional or educated workers abroad (Arif 2010).
Studies indicate that in most cases, households use remittances to establish new income sources, which sustainably improve their income. As an example, a representative household survey taken in Bangladesh not only found that the majority of migrant households used remittances to establish new income sources (IOM 2009) but that they were confident about being able to sustain the increase in household income after remittances ceased to come in. A survey from Sri Lanka, meanwhile, revealed that migrant households tended to have more income sources than non-migrant households. The authors concluded that this was because they were able to invest remittances in income-generating activities (Arunatilake, Jayawardena and Weerakoon forthcoming).
Surveys from Pakistan and Bangladesh showed that households with a long-term migrant abroad (10 to 15 years) received higher yearly remittances than households with a short-term migrant (Arif 2010, IOM 2009). This can be attributed to higher salaries for migrants with more experience abroad and reduced costs of migration for repeat-migrants. At the same time surveys revealed that households with long-term migrants abroad were less able to create alternative income sources. One possible explanation for this might be that many of the migrants abroad were forced to become repeat migrants due to the failure the households to generate other income sources.
Use of remittances
The development impacts of remittances also depend on how remittances are used. Many experts claim that these impacts are limited as remittance flows are mostly used for consumption. But, research shows that although the majority of remittances are indeed used for consumption, this consumption improves the fundamental well-being of households by providing added income to afford better housing or more nutritious food.
Research from Bangladesh also indicates that remittances were spent more productively in the 1990s than in the previous decade when migration was a relatively new phenomenon (Siddiqui 2004). This suggests that the experience of households in handling remittances, as well as government policies to encourage productive investment of remittances, can lead to more positive development impacts.
Surveys conducted in Bangladesh and Pakistan show similar patterns in the use of remittances, although it is difficult to compare these survey results due to the application of different methods and categorization. In the survey taken in Pakistan, about 22 per cent of remittances were spent on real estate and agricultural machinery, 18 per cent on food and about 17 per cent on costs related to marriage.8 Of note, in that survey, the use of remittances to pay for marriage was significant even though only 0.6 per cent of the respondents mentioned “earning money for marriage” as the main reason to migrate. In a survey taken in Kerala, 50 per cent of the cash remittances were used for “subsistence”, such as food, 24 per cent for education and 8 per cent to repay debt (Rajan and Zachariah 2007).
A survey in Bangladesh conducted in 2004 (Siddiqui 2004) showed similar spending patterns of remittances to a study on Pakistan conducted by Arif (2009). However, significant differences did appear with regard to savings. In Bangladesh, respondents saved only 3.4 per cent of the remittances, compared to 21 per cent in the Pakistan sample.
It also must be noted that according to the survey in Bangladesh, a significant portion (7 per cent) of the remittances was spent to finance the migration of other family members, which respondents considered as an investment for further enhancing family income. However, while the families see this as an investment, it in fact entails additional spending on recruitment costs. Taking into account the point noted earlier that the first months of migrant workers’ income usually are spent on repaying loans taken to cover recruitment costs, using remittances to cover the recruitment cost of another family member further increases the proportion of remittances spent on the cost of migration and reduces the potential development impact.
A comprehensive household survey undertaken in Bangladesh did not assess the percentage of overall remittances spent on each item, but instead determined how many households spend remittances for each item. According to that survey, 81.2 per cent of remittance-receiving households used the funds to pay family expenses, 39.1 per cent to repay debts, 38.2 for celebrations of religious festivities, 22.3 per cent for medical treatment and 21.3 per cent for the education of children (IOM 2009).9
In Turkey, remittances are more commonly used for investment purposes instead of being sent directly to households. Research often attributes this to weakening family ties of the Turkish population abroad as in most cases overseas migrants left the country with their families and many of them are now second and third generation migrant families (Sayan and Tekin-Koru 2004). The improvement of financial products targeting migrants may have also played an important role in channelling remittances into investments instead of direct transfers to households.
Impact of remittances on health, education and housing
Remittances help improve health and education outcomes through direct spending. Moreover, they provide households with a stable source of income and thus social protection, which, in turn, reduces vulnerability to income shocks. With this kind of social protection, households are more willing to invest in education, especially for girls (Kapur 2004).
The studies of Bangladesh and Pakistan show that relatively small proportions of remittances are spent on education (between 3 and 5 per cent) and health (between 3 and 4 per cent) (Arif 2010, Siddiqui 2004). However, other research conducted in these countries, as well in Sri Lanka, indicate that remittance-receiving households generally spend more on health and education than non-remittance-receiving households, with the higher expenditure often used to pay for higher-quality services (Arif 2010, Siddiqui 2004, Arunatilake, Jayawardena and Weerakoon forthcoming). For example in the survey conducted in Pakistan, 79 per cent of the migrant children between 5 and 15 years in the sample were enrolled in school, a figure that exceeded the national average school enrolment rate.10 In addition, many households used remittances to pay for a better quality of education, which is reflected in increased enrolment in private schools (Arif 2010).
The study from Kerala, the state where the majority of migrants from India originate, did not find a significant difference between remittance-receiving and non-remittance receiving households in terms of elementary and secondary school enrolment. But, it did show that remittance-receiving households were more often able to afford private self-funded colleges. An explanation for this may be that the public school system in Kerala is relatively good and basic educational needs are already met. Therefore, households would rather invest remittances in tertiary education, which can also explain the higher use of remittances for education in the Kerala sample than in the studies from Bangladesh and Pakistan (Rajan and Zachariah 2007).
Remittance-receiving households are more likely to be able to afford better quality health services as supported by research on Bangladesh, Pakistan, Sri Lanka and the Indian state Kerala (Arif 2010, Rajan and Zachariah 2007, Siddifui 2004). Research on Sri Lanka even observes higher concentrations of private health services in areas with larger concentrations of migrant households (Arunatilake, Jayawardena and Weerakoon forthcoming).
Remittances can also help improve the quality of housing. Surveys reveal that, for example, in Pakistan, house ownership and the quality of housing was already relatively good before migration, but it improved after migration. Before migration, 59 per cent of the migrants owned a house made from cement and other concrete materials, while after the migrant returned home, 74 per cent of the households owned such a house. The Pakistan survey also indicates that 15 per cent of the migrant households owned a house with five or more rooms before migration, with the percentage increasing to 37 per cent after migration (Arif 2010). In household surveys undertaken in the Indian state Kerala, remittance-receiving households had significantly better housing than non-remittance receiving households; a total of 58 per cent of the remittance-receiving households owned a house made of concrete that had at least two bedrooms, attached bathrooms and tiled floors, while only 17 per cent of the non-remittance receiving households could afford housing of that quality (Rajan and Zachariah 2007).
8 In a Muslim cultural context, this mainly consists of the dowry (to be paid to the wife), cost for housing, clothing for the bride and cost of various celebrations.
9 Percentages add to more than 100 per cent due to multiple answers provided by respondents.
10 According to the ESCAP Statistical Yearbook for Asia and the Pacific 2009, primary net enrolment in primary education in Pakistan was 67.2 per cent.