Abdul Bayes

Bangladesh enjoyed two decades of relative success in reducing rural poverty from 62 per cent in 1988 to 44 per cent in 2004. During this time, two developments paved the way for poverty reduction:  a monotonic fall in food prices (especially rice) and a shift from farm to non-farm employment facilitated by human capital accumulation.  

However, Bangladesh was caught in a quagmire following global crises in commodity markets during 2007-2008 period. The international institutions and non-governmental organisations (NGOs) voiced concern that higher food prices would dramatically increase poverty and insecurity among the world's most vulnerable populations. There were some models measuring the impacts of theoretical price change but the actual impacts would depend on the magnitude of actual price change, household response to those price changes, and potentially macroeconomic considerations.

  In a rare attempt, a research paper ( by Mahabub Hossain, et al) deals with the impacts of rising commodity prices on rural poverty dynamics.  There are many such studies but the paper is a departure in the sense that it uses panel data from longitudinal surveys. The authors attempt to assess the effects of the dramatic rise in agricultural commodity prices during 2007-2008 period on income dynamics and poverty among rural households in Bangladesh drawing upon repeat sample surveys. The authors use longitudinal data from rural households to document actual changes in household income and poverty during 2004-2008, compare those changes to trends during 1988-2004, and identify the household characteristics and market environments that either mitigated or exacerbated the impact of the food crisis on household income and rural poverty. The data were drawn from a repeat survey of nationally representative sample of rural households conducted to assess changes in rural livelihood systems.  

The benchmark survey was carried out in 1,240 households from 62 villages spread over 57 out of 64 districts in Bangladesh. The sample was drawn using a multi-stage random sampling technique. A census of all households in the selected villages was conducted to stratify the households by the size of land ownership and tenure. A random sample of 20 households was drawn from each village in such a way that each stratum is represented by its probability proportions. The same villages were revisited in 2000, 2004 and 2008 in order to survey the original households and their offshoots, as well as additional households to address the sample attrition problem. The total sample size in the second wave of the survey (in 2000) was 1,880 households comprising 30 households from each of the 62 villages. The total sample size in the third wave of survey (in 2004) was 1,930 covering the households in the first two waves and their offshoots. The total sample size in the final wave of survey (in 2008) was 2,010 covering the households present in the first three waves and their offshoots. The 1988 and 2008 waves offer a wide window of 21 years allowing an understanding of long-term dynamics, while the more recent waves of 2000, 2004, and 2008 permit an analysis of  the shorter-run poverty dynamics.

 The survey instrument is a semi-structured questionnaire designed to collect information on multiple aspects of rural economy and livelihoods , including resource endowments, farm and non-farm activities, income and expenditure, employment and commodity prices, poverty, gender, and government welfare programmes. The researchers also use econometric models of household income in order to assess the determinants of income and Foster et al's measure of poverty.

The analysis reveals that the trends in poverty incidence and depth reversed course, increasing to pre-2000 levels. It is estimated that an additional 13 million people fell into poverty in Bangladesh during that period. They also find that the increase in poverty incidence is caused by a decrease in upward poverty mobility as well as an increase in downward poverty mobility. This suggests that anti-poverty measures should target not only the poor, but also the non-poor households which are vulnerable to becoming poor.

The panel estimates of the income equation largely confirm previous cross-sectional work, with some important modifications. For example, a key finding is that farm income stood at 43 per cent of total income, virtually unchanged from the 2000 and 2004 surveys, but down from 58 per cent of 1988; an additional overseas migrant worker raises household income by approximately 30 per cent, and an additional hectare of land raised household income by approximately 20 per cent in 2008 indicating that land is the key asset for poverty reduction in rural Bangladesh. The authors also shed important insights into chronic poverty. While the rest of the rural population in Bangladesh has dramatically increased its holdings of non-agricultural capital and diversified income away from agriculture, the chronic poor have concentrated their assets and labour in agriculture. The policy implication is that agriculture still matters for the extreme poor, if not for other segments of the society.

Quantifying the effects of various household characteristics on the probability of being poor, the authors reveal a few important caveats absent in earlier analysis. For example, an additional overseas migrant worker reduces the probability of falling into poverty by 2.0 per cent in 2004 and by 22 per cent in 2008. Again, a great dependence on agricultural income increased the likelihood of becoming poor in 2004, but significantly decreased the likelihood of being poor in 2008. It is because of falling prices during 1988-2004 and rising prices since 2004. Thus while high food prices during 2007-2008 caused higher cost of living for all households, it appears that farm income served as a hedge against high food prices, either directly through higher commodity prices or perhaps indirectly through own consumption.

Again, the return to an additional agricultural labour is zero for the non-chronic poor but positive for the chronic poor while the return to additional non-agricultural workers and migrant workers is positive for the non-chronic poor and statistically indistinguishable from zero for the chronic poor. These results suggest that differences in the observed household resource allocation between the chronic poor and the non-chronic poor may be rational; the chronic poor are better off allocating labour to agricultural production, while the non-chronic poor are better off sending marginal labour into non-agricultural employment.

At the very least, these results suggest a great deal of caution in prescribing policies based on still-developing causes of poverty and on imperfect forecasts of economic conditions, especially agricultural prices. These, of course, remain fruitful topics of important research further.

The writer is Professor of Economics at Jahangirnagar University

source: the financialexpress