The role of remittances in Bangladesh's growth and development needs no emphasis. Direct and indirect impact of foreign as well as local remittances on reduction of poverty is amply documented in the available literature arguing that poverty would have increased in rural areas in the absence of migration - local or international. Thus, migration is essential, inevitable and potentially beneficial component of economic and social life of population in every region and country.
While there is a lot of research papers on migration, very few of them originate from household surveys, and perhaps a few from panel dataset. The recently completed survey on households in 62 villages (called Mahabub Hossain Dataset or MH Dataset) provides useful information about migration from rural Bangladesh. Data have been generated using a multi-stage stratified random sampling technique in repeat sample surveys undertaken in 2000, 2008 and 2014. For the moment, however, the presentation below would focus on descriptive statistics rather than econometric exercise.
How many households in rural Bangladesh do receive remittances from their members working inside or outside the country? The MH Dataset reveals that about one-thirds of sample households had at least one migrant member in 2014. This implies that one in every three households appears to have received remittance as opposed to one in four in 2008 and one in five in 2000. Arguably then, migration has been on a rise over time which is also reflected in various official documents. Of the total migrants, inside or outside, roughly one-thirds are destined for foreign countries, about half for other districts and one-sixths within districts in Bangladesh. In other words, domestic migration accounts for 70 per cent of the total migration taking place.
Who are the migrants? A priori reasoning would suggest that members from relatively richer households migrate outside the country while the poorer ones migrate inside the country. It is interesting to observe that about half of the foreign migrants in 2014 constituted members of poor households - defined as owning up to 0.2 ha of land (or 50 decimals). They are functionally landless. The proportion of this group has doubled within a decade as the share was just 25 per cent in 2000. It could further be observed that more than four-fifths of the foreign migrants in 2014 came from poor and small households who were reported to own up to 1 ha (about two and a half acre) as against two-thirds in 2000. On the contrary, migrants from relatively solvent groups (1 ha +) constituted about one-thirds of all foreign migrants in 2000; the share dropped to only 14 per cent in 2014. The upshot is that: (a) between 2000 and 2014, most of the migrants overseas were from relatively poor families, and (b) the share of relatively solvent families declined over time. The scenario is almost the same for domestic migrants. Remittance thus is contributing to growing resilience of households in shocks.
Where are they going? About half of the total migrants from Sylhet and Chittagong seem to leave the country. This is in contrast to three fourths of migrants from Rajshahi, Barisal, Rangpur and Khulna who are destined for domestic migration. Again, 70 per cent of migrants from these districts migrate outside respective districts. Apparently, higher wage and better amenities of life overseas pulled the former while pervasive poverty pushed the latter to migration. Among foreign migrants, about 70 per cent went to Middle East countries, 16 per cent to Asia, and 12 per cent to Europe. Of those who went to the Middle East, 70 per cent are reported to be secondary dropouts, and 26 per cent have crossed Secondary School Certificate (SSC)/Higher Secondary Certificate (HSC). For Asia, the proportion is the same. In Europe, only one-thirds of migrants are reported to be secondary school dropouts. Among foreign migrants, 80 per cent have no training whatsoever and the trend is the same across continents. Among them, trained drivers, electricians, welders and carpenters constitute the lion share. It is thus no wonder that Bangladeshi unskilled migrants earn very low wages. The necessity of pre-migration training, even if at certain costs, could be welcome.
Migration has costs - both monetary and physical. Since non-monetary costs are very difficult to measure, we consider the monetary cost of migration. In 2000, an overseas migrant had to spend on average $ 2,406 that rose to $3,805 in 2014. What is to be noted is that the maximum amount was $16,000 in 2000 and $23, 000 in 2014. This indicates the level of desperation for migration. However, against the cost of $3,805 in 2014, foreign migrants sent home on average about $2,413 back home. This means possibly that migration cost could be recouped within two to three years. However, the average remittance from domestic migration was $610. Data collected on migration show that in 2014, two-thirds of foreign migrants left the country for services and one-fourths went for wage labour against 52 and 11 per cent respectively in 2000.
How is the remitted money channelled into the country? There was a time when transaction took place mostly through 'hundi', an unofficial channel. We have no comparable data but in 2014, four-fifths of cash transfer from foreign countries was made through banks followed by only 6 per cent through hundi. This is a remarkable achievement as funds are now mostly flowing through official channels. No less remarkable is the change in the mode of cash transfer on the domestic front. There was a time when post office or by hand mode - taking much longer time - dominated the transaction in domestic migration. But the most recent data show that about 60 per cent of the transactions are being made through mobile phones followed by roughly 20 per cent delivered through hand or relatives. This 'revolutionary' development helped mostly the poor segment who migrated to cities to eke out a living. They can send money within a few minutes to relatives at home.
How is the remitted money being spent? Interesting insights can be found in expenditure pattern. Food expenses account for 43 per cent of total expenditure out of domestic remitted income compared to 27 per cent overseas and 31 per cent for all. About 12-14 per cent of the expenses (both domestic and foreign migrants) are earmarked for agricultural pursuits such as buying inputs, machines and land leasing. Within agriculture, foreign migrants spend most of their earnings on buying land (90 per cent). Construction/repairing and maintenance of households claim 15 per cent for domestic money and 20 per cent for foreign money, education and medical roughly 10-11 per cent each. Investment in business, government securities and stock market is very negligible. However, foreign migrants have 10 per cent bank savings compared to 3 per cent domestic migrants. By and large, remitted money seems to be used very rarely for productive investments.
The writer is a former Professor of Economics at Jahangirnagar University.