An Asian Development Bank (ADB) study, Global Food Price Inflation and Developing Asia, says a 10-percent rise in domestic food prices, ceteris paribus, could push an additional 0.55 million people in Nepal below the poverty line of US$1.25 a day (based on 2005 purchasing power parity), or increase the poverty incidence (percentage of people living below the poverty line) by 2 percentage points. Using the POVCAL database of the World Bank, the study finds that the poverty impact of food-price inflation is the highest in rural India (2.9 percentage points for a 10-percent rise in domestic food prices). While the impact is 2.1 percentage points for urban India, the lowest impact is estimated for Sri Lanka, at 1.2 percentage points. The figures for Bangladesh, Pakistan, Nepal and Bhutan are, respectively, 2.5, 2.2, 2 and 1.8 percentage points. It is noteworthy that despite having more or less the same growth rates, the impact of a 10-percent rise in domestic food prices is lower in China (2.2 percentage points in rural China and 0.2 percentage points in urban China) than in India. The difference is particularly marked when one compares urban China with urban India.
Here are some observations on the study, mainly with regard to the estimates for Nepal.
1. According to the inflation estimates of Nepal Rastra Bank, the year-on-year food inflation, as measured by the food and beverage component of the National Consumer Price Index, was 17.3 percent in mid-March 2011. Going by the estimates of the ADB study, this would suggest that during the one-year period to mid-March 2011, more than 0.55 million people were pushed into poverty due to rising food prices. But one cannot say so for sure because the price index (or any of its components) is an "average" concept and may not capture the prices faced by the poor or those just above the poverty line and vulnerable to relapse into poverty.
2. As the study says its estimates are based on the "latest POVCAL database", and the most recent household survey, required to estimate the poverty incidence, is the 2003/04 Nepal Living Standards Survey II (as is used at http://iresearch.worldbank.org/PovcalNet/povcalSvy.html), it appears that the ADB study based its simulation of the impact of food-price inflation on NLSS II. To the extent the headcount poverty incidence has declined since 2003/04, the impact of food-price inflation may well be higher than what has been estimated, unless those who have climbed out of poverty in this period have increased their incomes so much that a 10-percent increase in food prices cannot push them back into poverty (but there is no a priori reason to assume this to be the case). Although this may sound contradictory at first blush, the reason is simple: as more people come out of poverty (a reduction in poverty incidence), some (or most) of them will join those living just above the poverty line, and when there is a rise in food prices, all those living at a certain margin just above the poverty line will be thrown back below the poverty line – with the result that the increase in poverty incidence will be higher than if there had not been a decline in the incidence earlier.
3. The study does not report the change in poverty gap ratio (which measures how far below the poverty line is the average poor person/household) for Nepal due to a rise in food prices. This would have given an idea of how more poor the poor would be due to food-price inflation.
4. The study appears to assume the impact of food-price inflation on the poverty incidence is almost linear, that is, if food-price inflation doubles, the impact on poverty also doubles. This is evident from the estimates of poverty impacts in the alternative scenarios of 20- and 30-percent food-price increases, which happen to be double and treble, respectively, of the estimates for a 10-percent increase in food prices – in all 25 developing countries in the sample. Thus, if 0.55 million people are estimated to be pushed into poverty by a 10-percent increase in food prices in Nepal, the number doubles to 1.1 million for a 20-percent inflation and trebles to 1.65 million for a 30-percent inflation. While the possibility of people above the poverty line in Nepal being distributed in such symmetry as to yield such linearity is there (this scribe does not know what the actual distribution looks like), it is doubtful that this is the case with all the 25 countries. Hence, the linearity of impact is likely to be the result of the specification of the model employed by the study. In reality, it may be the case that as food-prices increase, the poverty incidence (the headcount ratio) increases but non-linearly, probably at a decreasing rate up to a certain point.
5. The ADB study considers the US$1.25-a-day, or international, poverty line. Using the NLSS II data and the US$1.25-a-day poverty line (or about Rs 27 per day using the PPP exchange rate of 2005), the World Bank estimates headcount poverty ratio at 55.12 percent for 2003/04. The poverty rate using the same data but considering the national poverty line (Rs 7695.7 per year, or about Rs 21 per day at 2003/04 prices) is 31 percent for 2003/04. There is thus a difference of 24 percentage points in poverty incidence using the two different poverty lines. This means that an estimation of the poverty impact of food-price inflation using the national poverty line (which is less stringent than the international poverty line, though) may show an even greater number of people being pushed into poverty by food-price inflation. The impact of food-price inflation on the poverty gap ratio would have been illuminating.
No comments:
Post a Comment