Indeed, all other estimated results are consistent with what we found above about
determinants of different types of migration: a small increase in household wealth (land
owned) and human capital endowment (highest level of education) lowers the propensity to
participate in temporary and permanent migration and increases the household propensity to
participate in international migration48. Therefore, wealthy households able to overcome
‘entry barriers’ to international migration will on turn be more likely to employ modern
farming technology, thereby achieving higher productivity. Asset-poorer households, on the
other hand, are unable to support migration costs and fall back on domestic migration, which
does not help them to overcome financial or risk constraints locking them into poor
productive performance. This, in turn, raises questions on the potential role of international
migration in alleviating poverty (and inequality) at national and global level.
6. Conclusions
According to the NELM approach, the typical migrant is part of a rural extended family who
dispatches members to other places of employment to generate capital and to obtain new
investment opportunities (e.g. change of technology) for the family farm.
The main idea underlying this study is that if on the one hand imperfections in capital and
insurance markets constitute a motivation to migrate - as stated by the NELM hypothesis, on
the other hand they also may represent a constraint to do it; this is so because migration is a
form of lumpy investment, especially onerous for such households as those living in poor
rural areas of Bangladesh. Therefore, determinants of migration simultaneously shape the
economic impact of having a migrant member on farm households left behind. This has
important implications while seeking to understand the complex linkages between migration
opportunities and economic development in local communities.
Assuming that higher initial asset holdings make it less likely that liquidity constraints bind,
our empirical evidence shows that household’s wealth-related capital (mainly in the form of
land) is crucial in shaping heterogeneous migration behaviour towards different typologies of
migration. Asset-poor farm households are more likely to enter into domestic migration with
low entry costs, and low returns. Entry into high-return migration (i.e. international
migration), in which most households would probably like to engage in a ‘first-best’
perspective, is restricted to richer and large-holder households. In particular, throughout a
multinomial logit model estimation, we find that at low level of wealth, an increase in asset
ownership reduces, at different rates, the propensity to migrate either temporarily,
48 Also owing cattle has the same impact as found in the first part of inferential analysis? Here it might be not
strictly exgenous, though, but the exclusion of it does not change much estimation results.
34