There is a change of guard at the World Bank and while many have objected to Mr Wolfowitz's credentials to run the largest NGO in the world, many more do not realise the self-interested bureaucracy that the World Bank has become.

It will be an interesting match to watch: Mr Wolfowitz attempting to change the Bank and the certain attempt by Big Brother Bank to change him. The bets must be on the latter.

However, poverty, the raison d'etre of the Bank, cannot be eliminated without hope, and I do hope that Mr Wolfowitz will be able to change the vision, and purpose, of the Bank.

Who are the Bretton Woods Sisters?

But a change would imply that the poverty objective has largely been achieved; a conclusion not reached by the World Bank and its junior partner in the fight against poverty, the IMF: "the donors needed to at least double their aid budgets over the next five years if there was to be any chance of meeting the UN poverty reduction and development goals".

Note the linking of another NGO, the UN, in the World Bank-IMF declaration. Sisters stick together. Poverty removal is big business and all the super NGOs want a large piece of the action.

Small pieces are to be left for the regional poverty removal Banks -- Africa, Latin America and Asia.

But really, has there been only a little dent in world poverty, despite growth in the poor world? The first task for Mr Wolfowitz should be an independent, objective, non-ideological, non self-interest group appraisal of what has happened to poverty removal in developing countries, and the role the Big 3 sisters have played.

If he were to do that, he will be shocked at both the large poverty decline that has actually taken place, and the little role that the Big 3 played.

The poverty decline has taken place primarily in India and China, and both countries will argue that besides some good technical advice, the growth, and poverty reduction, was achieved largely by domestic efforts.

However, this conclusion is debatable, and its premise that poverty has declined by large amounts in the Big 2 economies is also a result avoided by the World Bank and the affiliated sisterhood of self-interest.

Part of the reason for the unnecessary debate is because these organisations would like the world to believe that poverty calculation is a complicated exercise, when in reality it is mere accounting.

What is shocking, however, is that the flaws in simple accounting can easily be identified, as I hope to do in the paragraphs that follow. Rather than concentrate on world poverty, I will document the findings of the two heavily researched and large countries, India and China.

If World Bank (WB) pverty numbers for these two countries are suspect, then Mr Wolfowitz may have a chance of helping the genuine poor in the world.

According to data on the World Bank website, http://iresearch.worldbank.org/PovcalNet/jsp/index.jsp, the number of poor people in 1990 was 1,047 million, with China and India contributing about 360 million each.

In other words, world poverty was essentially an India-China story. A growth decade later (2001), the importance of India-China is slightly reduced (370 and 210 million poor, respectively) but they still comprise about 60 per cent of the world's poor.

Given the importance of these countries, one would expect, in normal circumstances, an objective set of poverty numbers for India and China.

Despite all the transparency (data on the website, large set of researchers in its stable, etc.), this most manifestly is not the case. Let me document this reality.

In India, given our own poverty line, the fraction of the poor (head count ratio) in 1993-94 was 37 and 33 per cent, rural and urban areas, respectively.

The corresponding World Bank figures are 49 and 22 per cent. There is no reason for the two figures to match, unless the poverty lines are the same.

But they are. The World Bank exchange rate of 1993 rupees to 1993 PPP dollars is 7.02, and the urbanisation rate in that year was 26.2 per cent.

The official poverty lines for 30-day per capita consumption are Rs 206 and Rs 286, rural and urban areas, or an all-India average of Rs 227 per capita per month. Voila; the much "discovered" $1.08 a day poverty line is exactly the Indian poverty line obtained by the simple arithmetic of (227/30 = 7.57/7.02 = PPP1993$1.08)

This makes the comparison of World Bank and Indian government estimates very simple -- the two should be identical. But they are wide apart.

For 1993-94, the World Bank estimate of all-India poverty is 42 per cent; the government of India estimate for the same year, using the same NSS data, is a considerably lower 36 per cent; on a population of 900 million, that is an extra 63 million people deemed Indian poor by the World Bank.

The story for 1999-00 is worse. Again, for the same poverty line, the World Bank estimates 36 per cent poor, in comparison with the GoI estimate of 26 per cent -- that's an extra 100 million poor manufactured by the World Bank.

Apart from problems with estimates for India at any point of time (i.e. the non-matching with the official data), there are problems with understanding India's poverty numbers relative to the WB estimates for our neighbouring countries.

For example, in 2001, the World Bank's estimate of poverty in Pakistan is only 12 per cent of the population, a massive decline from 48 per cent in 1990.

Appreciate this statistic: Pakistan had a higher level of poverty than India in 1990. Because of economic reforms, growth occurs in India, and poverty declines at a reasonable pace.

Pakistan, by most non-World Bank accounts, is in a deep slump during this decade but apparently benefits enormously from the presence of the Mujahideen, and Bin Laden, and military expenses in Afghanistan.

So brilliant are the policies (worthy of emulation?) that in terms of poverty reduction, again with the World Bank stamp of quality, Pakistan achieves a world record -- 36 percentage point reduction in absolute poverty, in only 11 years, and with an average growth rate in per capita GDP of less than 1.3 per cent per annum!

There are other problems with the World Bank data on country, and world, poverty. Another World Bank poverty flagship country is Nepal -- there the reduction in poverty is from 44 per cent in 1990 to 14 per cent in 2001.

Did you know that, according to the authoritative WB, the average Nepalese in 2001 was almost twice as rich as the average Indian? Not so artificially lucky is Bangladesh.

This country, according to the World Bank, despite rapid growth (3 per cent per annum since 1990), has not been able to reduce poverty at all -- the head-count ratio stays the same, at 33 per cent in both 1990 and 2001 (but note, still lower than India).

This is not the first time I am pointing out the blinding flaws in World Bank estimates of country and world poverty. Something is clearly rotten with these guesstimates of poverty.

But how can such miscalculations happen? They do, because unlike firms faced with corporate governance lawsuits on the products they sell, the World Bank and its sister organisations are immune to charges of corporate misgovernance on the products they sell -- namely research, knowledge, interpretation, and advocacy of policy.