Making anecdotal evidence, statistical

It is one of the cardinal errors in poverty alleviation programmes that it is assumed the poverty reduction process is irreversible i.e those who make to the other side of poverty line do not slide downward or revert back to status quo ante. On the basis of figures derived from BBS's HIES-2016 it can be concluded that the poverty reduction rate, either for the extreme poor or the normal poor, has not gained any momentum through acceleration. Rather, to judge by the figures based on BBS data, there has been stagnation in the rate of poverty reduction, writes Hasnat Abdul Hye

It is a truism that has become almost passe' that there are always exceptions to the rule. But sometimes the exceptions are so glaring and rooted in experience that they turn the rule on its head.  Statistics, cynically designated as the worst of lies, after lies and damn lies in popular parlance, has its occasional redemption. The recent disclosure in the Household Income and Expenditure Survey (HIES)-2016 made by Bangladesh Bureau of Statistics (BBS) about growing inequality and reduction of poverty at a certain rate have become part of conventional wisdom. While the first, the growing income inequality, has received wide and serious attention, the second, stagnation of poverty reduction rate, has not so far raised any eyebrow or caused any concern. This article focuses on the second of the disclosure though the admission about stagnation of poverty reduction rate has not been made explicitly by BBS and has to be read between the lines of their statistical report on HIES-2016.

It is the contention in this paper that the stagnation in the rate of poverty reduction is a matter of concern because even while poverty of the extreme or ordinary kind appears to be reducing over the years the number of poor is increasing because of new entrants in the poor population. Using the benchmark figure of poor, say for 2005 as the HIES-2016 does, the absolute number of poor taking account of new entrants as a result of population growth is missed in the survey by BBS. The obvious evidence of poor increasing in number is anecdotal, in the absence of statistical survey to capture them. The poor can be seen in increasing number in plain sight in all public places, rural and urban. As mentioned earlier, though palpable, this increase in number is not quantified or measured by any standard in a survey but it is so visible and obvious that it can be seen by anyone with discerning eyes in the course of daily life. The poor in the crowd or in isolation do not wear any identifiable badge or a tag but the tell-tale signs of their poverty status like health conditions, clothes they wear, the work they do to eke out a living and the measly food they take are enough of indication about which economic group of the population classified according to income they belong to. No sophisticated definition or elaborate measurement of poverty is required to identify the poor and realise their ubiquity when they are visible to naked eye everywhere. Official statistics of BBS fails to record their presence, particularly when they are new entrants or dropouts from poverty-focused programmes because they are not included in the benchmark data on their total number at a particular year and uses selected households which do not reflect this `economic-demographic change' (new entrants plus dropouts). If the households surveyed by BBS in HIES were changed to capture the new entrants and dropouts perhaps their findings would be more reliable but for the sake of consistency in the use of sample this is not done. As a time-series study their approach is hallowed in academic tradition. But almost inadvertently the HIES-2016 released by BBS has revealed the weakness and shortcoming in this approach. How? The next paragraph discusses this with figures provided by BBS in their report.

The HIES-2016 report released a few days back by BBS showed that the country's poverty rate as per the upper poverty line came down to 24.3 per cent in 2016 which was 40 per cent in 2005 (benchmark year). In the same time poverty line in 2016 as per the lower poverty line was 12.9 per cent which was 25.1 percent in 2005. According to the growth elasticity model used by BBS the poverty rate as per upper poverty line was 21.8 per cent in 2016 while the same figure was 11.3 per cent in 2005. This shows that poverty reduction for the poor defined by upper poverty line has been 1.5 per cent annually during a period of 11 years (2005-2016). For the poor defined by the lower poverty line the annual rate of reduction has also been the same i.e 1.5 per cent for the period 2016-2018. On the basis of these figures derived from BBS's HIES-2016 it can be concluded that the poverty reduction rate, either for the extreme poor or the normal poor, has not gained any momentum through acceleration. Rather, to judge by the figures based on BBS data, there has been stagnation in the rate of poverty reduction. This would not be a matter of concern or appear worrisome had the pool of poor people been static and fixed in number. But there has been addition to the pool of poor people as population has increased year by year. On the basis of conventional wisdom, it is common knowledge that population growth contributes more to the poor segment than to the others. To this incremental number of poor due to population growth should be added the poor who had earlier graduated out of poverty but slid back to poor status because of illness, natural disaster, setbacks to income earnings due to market conditions and other socio-economic causes. It is one of the cardinal errors in poverty alleviation programmes that it is assumed the poverty reduction process is irreversible i.e those who make to the other side of poverty line do not slide downward or revert back to status quo ante. In one of the oft-quoted sentence in books of economics Alice in Wonderland said, `You have to run faster and faster to be in the same place'. In poverty reduction Alice's dictum is not good enough because to sustain the momentum of poverty reduction, running has to be even faster so that the runner does not end up being at the same place.

What should be the rate at which poverty reduction should take place? The question may be asked. A simple and straightforward answer can be: the rate may be worked out by taking the benchmark figure on poor population into account and adding the number of people who join the existing poor at a particular year because of population growth. Over and above this, at least 10 per cent of those who graduate from poverty but are at risk of sliding down the income scale should also be added to the rank of poor. Whatever the rate is determined, to be realistic it cannot be a constant rate. The most important thing to remember and emphasise is that the challenge before statisticians working out figures on poverty reduction is to make anecdotal evidence, statistical. How they will go about it is best known to them. To me, it appears as a common sense exercise and a compelling task

[email protected]