Ideas to Action:

Independent research for global prosperity


Views from the Center


Andy Sumner and I recently wrote about the fact that the number of low income countries in the world is rapidly shrinking –which is great news because it suggests poor countries are getting richer.  But how much does graduating to ‘middle income’ mean?  Here’s how the original income classification came about, according to the World Bank’s website:

The process of setting per capita income thresholds started with finding a stable relationship between a summary measure of well-being such as poverty incidence and infant mortality on the one hand and economic variables including per capita GNI estimated based on the Bank's Atlas method on the other. Based on such a relationship and the annual availability of Bank's resources, the original per capita income thresholds were established. Thereafter, the original thresholds have been updated every year to incorporate the effect of international inflation....

The economies whose per capita GNI falls below the Bank's operational cutoff for "Civil Works Preference" are classified as low-income economies; economies whose per capita GNI is higher than the Bank's operational threshold for "Civil Works Preference" and lower than the threshold for 17-year IBRD loans are classified as lower-middle income economies; and those economies whose per capita GNI is higher than the Bank's operational threshold for 17-year IBRD loans and lower than the threshold for high-income economies are classified as upper-middle income economies.

Got that?  In short: the World Bank found that there’s a close link between GNI per capita and broad-based development (hem, hem), the institution needed a system to help apportion limited funding and set borrowing rules, and so it chose income thresholds to meet that need.  The current income thresholds have nothing whatsoever to do with a particular status of countries themselves –something that a country just over the $1,000 middle-income threshold has that a country just under that threshold doesn’t have. Fair enough –but the way the classification was created suggests that the World Bank would be completely consistent in setting new income thresholds because of the changing availability of Bank resources or new views about which countries need civil works preference in procurement.

Meanwhile, my colleagues Ben Leo and Todd moss have looked forward and forecast what the increasing wealth of formerly poor countries means for the International Development Agency –the soft loan arm of the World Bank which lends to countries classified as low income and a little bit richer.  They predict that there will only be a billion people left in IDA-eligible countries in a few years’ time.  “This drastically altered client base will have significant implications for IDA's operational and financial models,” Todd and Ben conclude.  But, once again, a simpler course would just be to reset similarly arbitrary IDA thresholds. Here is how IDA goes about deciding who gets soft loans, from a 2001 paper:

The ceiling for IDA eligibility (currently called the historical cutoff), initially set at $250 per capita in 1964, has been revised to account for inflation, reaching $1,445 in 2000.  In the early eighties, the availability of IDA resources was not adequate to fund programs for all countries below this eligibility ceiling.  As a result, IDA ceased to lend to countries at the upper end of the notionally eligible per capita income scale.  This created a second and lower “operational cutoff” which was formally recognized by IDA donors in IDA8.  The operational per capita income cutoff has been reaffirmed by the donors in each subsequent replenishment, and now stands at $885 [in 1999 US$].  The historical cutoff is used only for determining preference in civil works procurement.

The current operational cutoff for IDA eligibility for FY11 is $1,165 (2009 GNI per capita), while the historical ceiling is $1,905 (2009 GNI per capita).   Once again, eligibility was determined by how much money there was to go around rather than any particular feature of countries poorer than $250 per capita in 1964.  And then eligibility was re-determined based on the availability of IDA resources to about half that amount adjusted for inflation in the 1980s.  So why not just return to the historical cutoff?  Or double it, if there’s enough money in IDA?

A different approach would be to attempt to make the income and IDA thresholds actually reflect something about the nature of countries independent of their relationship to the World Bank and its arcane concerns with civil works preference.  Here’s one idea: countries that can’t wipe out absolute poverty (people living on less than $1.25 a day) relying on their own resources alone are considered low income and IDA eligible.  Martin Ravallion of the World Bank suggests that most countries with an income over $4,000 (PPP) could plausibly end absolute poverty without outside assistance using a tax on rich people within the country –so there is the new cut-off.  Countries with incomes around $4,000 PPP have average Atlas GNI’s per capita of close to $2,300.  Give or take, then, that doubles the current income threshold.  As it happens, it is also pretty close to the ‘historical ceiling’ for IDA.

As to high income, Lant Pritchett suggests that a plausible upper bound poverty rate is $10 a day.  Let’s say any country where average incomes are five times that counts as rich –that comes out to around $18,250 (PPP) –countries near that level have an average Atlas GNI per capita of about $11,800 –pretty close to the current high-income cutoff of $12,276.  Why five times a $10/day level?  To be honest, just because it comes out close to the current cutoff.  So I’m open to better ideas!


CGD blog posts reflect the views of the authors drawing on prior research and experience in their areas of expertise. CGD does not take institutional positions.