Paper abstract:

Is one dollar of remittance income equal to one dollar of any other source of income? If it is not, then migrant money transfers may cause behavioral changes at the household level, and their development impact can be huge. According to the World Bank, Kyrgyzstan is the number three country in the world by its share of personal remittances in GDP (28.5% in 2019), which makes its economy highly dependent on the money transfers by its labor migrants. On the other hand, there are frequent claims in Kyrgyz media that remittances sent by migrants are often used to finance celebratory events, which are both numerous and expensive in relation to the incomes of households. This paper tests the assumption of fungibility of income and explores whether there are significant differences in the patterns of spending remittance versus other sources of income by Kyrgyz households, using five years (2010-2013, 2016) of “Life in Kyrgyzstan” dataset.

In terms of the methodology, the Working-Leser expenditure model is extended to incorporate a constant elasticity of substitution function, which allows explicitly calculating elasticities of substitution between remittance income and other household income. In addition, to account for the possibility of endogeneity of remittance income due to selection bias (i.e. remittance-receiving households may be different from other households), the control function approach in the context of non-linear estimation is employed. Using a system of demand equations with separate parameters for remittance income and other sources of income allows for both accounting for different patterns of spending by households, as well as differentiating between various types of goods (normal, luxury, and inferior) according to the income source used. To allow for the possibility that households of varying income levels may view different expenditure categories and income sources differently, the analysis is extended to study the distributional effects.

The results indicate that remittance income and other income are in fact not fungible and thus, mental accounting matters. Besides, there are significant differences in how remittance income and other income are spent by families in Kyrgyzstan. Contrary to the optimistic view of remittances in the literature, households spend most of their remittance income budget on consumption goods. In addition, they spend more of their remittance income on celebrations, funerals, and rituals compared to income from other sources. The findings suggest that the massive remittance inflows into the country do not have a positive development impact that they potentially could.