Disagreement on poverty trends

The extent of poverty and its trend have become a hot topic in Pakistan. Looking at the conditions on the ground, most of the independent analysts believe that standards of living, particularly of the poor, have deteriorated or remained almost stagnant.
While economic managers of the government, citing certain household surveys, claim that poverty levels have declined sharply in the recent past and attribute this success to superior economic management of the present government. In the latest Economic Survey for the year, 2005-06, the government revealed the results of a Household Integrated Economic Survey (HIES) that showed that the percentage of population living below the poverty line had fallen from 34.5 percent in 2001 to 23.9 percent in 2004-05, showing a decline of 10.6 percentage points over this period.
The percentage of population living below the poverty line in urban and rural areas came down from 22.69 percent and 39.26 percent to 14.90 percent and 28.10 percent respectively. The Survey covered 14706 households - 5808 households in urban areas and the rest in rural areas.
The poverty line was based on 2350 calories per adult equivalent per day. Adjusted for inflation, the poverty line for 2004-05 was Rs 878.64 per adult equivalent per month - up from Rs 723.40 in 2001. While the government has been claiming credit for sharply reducing poverty in such a short period and bragging about it, the results of the HIES have been questioned by many quarters.
In a draft on "Pakistan promoting rural growth and poverty reduction", the World Bank has given its own estimates of poverty level, arguing that poverty reduction between 2001-02 and 2004-05 has been only 5.2 percentage points, coming down from 34.4 percent to 29.2 percent at national level and from 39.1 percent to 34.0 percent for rural households.
The World Bank's estimates were different from the government of Pakistan's due to variation in the inflation rates used to determine poverty level. According to the World Bank, estimates of poverty in Pakistan were particularly sensitive to differences in methodology because a high percentage of rural households have per capita expenditures very close to the official poverty line.
Some 10.9 percent of rural households in 2001-02 had per capita expenditures within (+/-) five percent of the official poverty line; in 2004-05, 8.95 percent of rural households were within (+/-) five percent of the Planning Commission's official poverty line.
Therefore, small changes in calculated real income (expenditures), whether due to actual changes in expenditures, price deflators or other methodological issues related to updating the poverty line, can lead to misleadingly large changes in poverty estimates.
The disagreement on the level of poverty in the country during a certain period, in our view, is a sad reflection on the ability of the government to design a universally acceptable methodology to measure the poverty level in the country and convince others about its genuineness. If this was the only problem, it perhaps would not have been too difficult to solve.
The real problem is that all governments in Pakistan, past and present, try to present figures that give a very rosy picture of various indicators which, are invariably somewhat removed from reality. The biggest casualty of such manipulation is usually the credibility of the government's own statistics.
In the instant case, there does not seem to be disagreement on the reduction in poverty as such, that is due to improvement in agricultural production and high growth rate during the years under review, but the economic managers of the country in their eagerness to show an improved situation may have been tempted to depict an unusually brighter image of the results of their policies.
Such a conclusion seems plausible due to the reality on the ground and the sensitivity of the Survey to slight changes in per capita expenditures. We say this because over 10 percent reduction in poverty could have made a lot of difference to the living conditions of the poor people but such an improvement is not visible on the ground or in the streets.
Authorities are talking about increased sale of cars, air conditioners, refrigerators, motorcycles etc as a proof of prosperity, but this is not the stuff of the poor. Another problem in the case of Pakistan is the high density of per capita income and expenditure around the poverty line which means that a slight variation in the measurement methodology could yield vastly different results.
It is particularly surprising that divergence between the government and World Bank figures which was small in 1998-99 and 2001-02, increased to 5.4 percentage points in 2004-05. The reasons behind such a large divergence need to be explained in detail. Keeping all these factors in view, we feel that the government should devise an appropriate methodology, in consultation with independent economists and the multilateral institutions, to measure the poverty level in the country accurately and then stick to such a formula in the years to come for a proper comparison.
People in general are now confused because of the difference in figures for whatever reasons. The motive behind the proposed exercise should be the acceptability of the government statistics and its greater credibility at all levels. After all, data is compiled and released not to exaggerate but to reflect a certain situation faithfully.