*This post has greatly benefited from exchanges with Chris Garroway.
Mostly thanks to China´s supercharged 2000s growth and the related global development impact, extreme poverty (in its 1.25$_PPP/day/person variety) inside as well as outside China has dropped as a percentage of a growing world population - from 40 to 20% over the past two decades. Consider this back-of-the-envelope calculation: One percent of GDP growth in China has been associated with 0.34% of GDP growth in countries with an annual Gross National Income (GNI, Atlas method) below 1,035$/year that are classified by the World Bank as Low-Income Country (LIC); estimates of poverty elasticity for LICs to growth vary between 1.2 and 3.1 for the 2000s (they are higher than the estimates for the 1990s). Assume a poverty elasticity of 2; then, a percentage point of Chinese growth would lower the LIC poverty headcount by .68%. With roughly 1.1 billion people still in extreme poverty outside China, one percent of China´s growth has translated into 7.7 million people less in extreme poverty year by year[1]. However, note that while the bottom third of the global income distribution have also made significant income gains, real incomes of the poorest 5% of the world population have remained the same even in the past Golden Age of emerging-country growth[2].
Mostly thanks to China´s supercharged 2000s growth and the related global development impact, extreme poverty (in its 1.25$_PPP/day/person variety) inside as well as outside China has dropped as a percentage of a growing world population - from 40 to 20% over the past two decades. Consider this back-of-the-envelope calculation: One percent of GDP growth in China has been associated with 0.34% of GDP growth in countries with an annual Gross National Income (GNI, Atlas method) below 1,035$/year that are classified by the World Bank as Low-Income Country (LIC); estimates of poverty elasticity for LICs to growth vary between 1.2 and 3.1 for the 2000s (they are higher than the estimates for the 1990s). Assume a poverty elasticity of 2; then, a percentage point of Chinese growth would lower the LIC poverty headcount by .68%. With roughly 1.1 billion people still in extreme poverty outside China, one percent of China´s growth has translated into 7.7 million people less in extreme poverty year by year[1]. However, note that while the bottom third of the global income distribution have also made significant income gains, real incomes of the poorest 5% of the world population have remained the same even in the past Golden Age of emerging-country growth[2].
Millennium
Development Goal #1 – to halve extreme poverty by half by 2015 – was thus
reached fairly easily with China as the global poverty reduction machine (and not because leaders signed the
Millennium declaration!). A close corollary to developing-country growth
performance has been LICs´ eligibility for and graduation from concessional
finance. Several countries have graduated from IDA
borrowing
since 1999, including populous countries such as China, Egypt and Indonesia,
with India expected to graduate soon. There are today only 36 LIC countries eligible
for concessional finance – grants and soft loans – by the International
Development Association (IDA) and the soft windows of the regional development
banks[3].
The graduation threshold itself is controversial, however: an outdated LIC/MIC
threshold explains that the share of the poor (and, indeed, of the whole
population) in LICs has declined over time. With China and India having crossed
the threshold, the reduced demand for concessional finance (CF) may partly
reflect a statistical artefact.
What arguably
accounts better than income levels for CF demand is a country´s domestic
capacity of resource mobilization, especially through taxation[4].
Availability of public and private resources for development, coupled with the
fall in global poverty, are said to imply that dramatically more funding is
potentially available for each poor person. And domestic resources, especially
tax receipts, may be mobilized better as result of high raw material income and
better tax administrations.
Looking beyond 2015,
the target year defined by the Millennium Goals, international organizations
and their leaders have increasingly joined a chorus of euphoric We-Can-End-Poverty declarations. Perhaps
the biggest mouthful came from OECD boss Angel Gurria who proclaimed when the
last DAC Report Ending Poverty was presented late 2013 in London: “Ending Poverty Completely and
Forever”, seconded by DAC chair Eric Solheim “Eradicating Extreme Poverty
Completely – ´Yes We Can´”. There are
today numerous websites devoted to the goal to end extreme poverty, such as www.endpoverty2015.org, www.theendofpoverty.com, www.endpoverty.org, or www.stoppoverty.com, which translates a strong belief
in the feasibility of poverty eradication. While macroeconomic observers are now
buzz with secular stagnation, China´s forthcoming crash, or emerging market taperitis, this is not the stuff that most of
the ODA crowd is familiar with on a professional level. So the End-of-Poverty
banner waving seems quite detached now from what can be expected from future
growth and seems to extrapolate a trend of global poverty reduction that may
have been special to the past decade.
The End-of-Poverty
banner waving seems to be based on two influential studies[5],
henceforth called CGD (Moss/Leo, 2011) and ODI (Kharas/Rogerson, 2012) scenarios, which
have recently projected total population in IDA-eligible countries to decline
from 3 bn (2012) to 1 bn by 2025 and the global poverty pool to shrink
dramatically by 2025 as a result of high per capita income growth. According to
these studies, soft CF windows such as the African Development Fund (AfDF), the
Asian Development Fund (ADF), the International Development Association (IDA)
and also some International Monetary Fund (IMF) facilities would likely face a
wave of country graduations by 2025. These studies also foresee a reduced
number of selected low-income, post-conflict and fragile countries, mostly in
Africa, re-established as the main location for CF-eligibility. A drastically
altered client base will have significant implications for the strategic
options as well as operational and financial models of IFIs. The strategic
choices facing the IFI shareholders are at the heart of the future of the
global concessional finance architecture.
Both studies are
penetrated by emerging market optimism that underpins their view that “we” can
successfully eliminate poverty without continued access to concessional finance.
However, there are important shortcomings to both studies on a closer look at
technical detail:
- Both the CGD and ODI studies cited above are essentially based on prolongations of GDP projections provided by the IMF in its World Economic Outlook (WEO). However, the accuracy of IMF-WEO forecasts is dismal as they have been overly optimistic in the past[6]. Projecting growth rates over long horizons is hazardous in any case, especially if rates are held constant and do not build in major occasional disruptions to growth, such as from natural disasters and financial crises.
- Mind the gap: GDP is not the appropriate income concept but GNI, especially in poor countries due to inflows of remittances (that could be used to reduce poverty at home). Disruptions to remittances (e.g. when resource-rich countries send immigrant workers home, as happened recently in Saudi Arabia) can lead to important differences in the growth projections in the two national-account concepts.
- The two studies also understate the Balassa-Samuelson effect (or Penn effect) of growth convergence in poor countries: the increase in price levels that have accompanied the last decade of high growth in emerging countries[7] may have given rise to estimates for CF demand that are excessively low. Higher prices for services (such as real estate rents, transport cost, schooling) lower the purchasing power of nominal incomes and may lock in the poor below poverty thresholds. The CGD study, e.g., holds all calculations in constant $_2009 with the hidden assumption that price levels stay the same. Over a decade-long projection, this leads to massive distortions.
The Penn Effect of Income Convergence
Source: “Appreciating the BRICs”, The Economist, 24-12-2014
- As poor-country growth over the last decade has been based on expansive monetary policy in OECD countries and on unsustainably high growth in China, a return to normalized rates of global money supply and to sustainable growth in China must receive special attention in growth projections to 2025[8]. Both studies were carried out at a time that many observers consider now the peak of emerging-country euphoria. Emerging-country optimism has rapidly ebbed with two major sources – ultralight monetary policy in the US notably and unsustainable growth in China – running dry.
- Finally, there seems to be excessive optimism around about poor countries´ capacity to mobilize their own resources rather than to depend on aid dollars, via redistribution from the “rich” to the “poor” within developing countries.
Marginal Tax Rate Required to Close Poverty Gap
Source: Martin Ravaillon, (2012), ´Should we equally care about poor people wherever they may live?´
Ravaillon shows that there is a positive correlation between domestic capacity for redistribution (as indicated by a low required marginal tax rate to close the poverty gap) and a country´s average per capita income[9]. His measure – marginal tax rates on the ´non-poor by US standards´ required to cover the poverty gap - finds that for most (but not all) countries with annual consumption per capita under $2,000 the required tax burdens are prohibitive—often calling for marginal tax rates of 100 percent or more. By contrast, the required tax rates are very low (1% on average) among all countries with consumption per capita over $4,000, as well as some poorer countries. The tax ratio calculations thus demonstrate that LICs (and even UMICs) have little or no prospect for increasing domestic resources in the medium term to meet these needs.
So there is a mismatch
between the loudness of the End-of-Poverty chorus and the empirical solidity of
the projections on which the official declarations are pegged.
A fascinating survey
just released by Gallup[10]
cautions also against excessive post-2015 optimism. Gallup's self-reported
household income data across 131 countries indicate that more than one in five
residents (22%) live on $1.25 per day or less. About one in three (34%) live on no more than
$2 per day. The World Bank Group recently set a new goal of reducing the
worldwide rate of extreme poverty to no more than 3% by 2030, but Gallup's data
suggest meeting that goal will require substantial growth and job creation in
many countries. In 86 countries, more than 3% of the population lives on $1.25
per day or less.
Extreme Poverty as % of Population
Source: Gallup (2013), More Than One in Five Worldwide Living in Extreme Poverty, 23-12-2013.
Even in the OECD,
“Ending Poverty Completely and Forever” has not been achieved, as the Gallup
global poverty map makes clear. To be sure, OECD leaders have no core competence in
eliminating poverty. They would thus benefit from more modesty in their pronouncements on global poverty. Despite tax, welfare
and transfer systems more extended and per capita income levels much higher
than in developing countries, most OECD countries have witnessed a remarkable
rise in poverty rates during the past fifteen years. In most OECD member
countries, relative poverty - defined
as the share of people living in households with less than 50% of median disposable
income in their country – affected 11% of OECD population in 2010, after taxes
and transfers. This was a marked increase of poverty rates compared to 1995.
Source? OECD[11]!
[1] For China´s growth linkages, see Chris Garroway et al. (2012), “The Renminbi and Poor-Country
Growth”, The World Economy, 35.3, 273-294.
Estimates of poverty elasticity at around 2% have been reported by Ajay
Chhibber and Gaurav Nayyar, (2008) "Pro-poor growth: explaining
the cross-country variation in the growth elasticity of poverty", International Journal of Development Issues, Vol. 7. 2, 160 – 176.
[2] Branko Milanovic (2013), “Global Income Inequality by the Numbers:
In History and Now – An Overview –“ , World Bank: Washington DC, mimeo.
[3] The Inter-American Development Bank and its soft window, the Fund
of Special Operations (FSO) considerably reduced in size.
[4] Francisco Sagasti (2013), ´From graduation to gradation in
international development finance´, London: ODI, developmentprogress.org,
December.
[5] Todd Moss and Benjamin Leo (2011), “IDA
at 65: Heading toward Retirement or a Fragile Lease on Life?” Working Paper
246. Washington, DC: Center for Global Development; and Homi Kharas and Andrew
Rogerson (2012), “Horizon
2025: Creative Destruction in the Aid Industry”, London: Overseas
Development Institute.
[6] Source: Tyler Durden, ´Hilarious
Charts Of The Day: IMF's "Growth Forecasts" Over Time´,
zerohedge.com, 8-10-2013.
[7] Read the brilliant post “Appreciating the BRICs”, The Economist,
24-12-2014. It also shows how actual and projected GDP is lowered by holding
prices and currencies constant (by correcting for the Penn effect).
[8] Large
emerging countries with high external deficits (Brazil, India, Indonesia, South
Africa and Turkey) are supposed to suffer once monetary stimulus is scaled
back; see ´Fragile Five Countries Face Taper
Crunch´, FT,
17-12-2013.
[9] Martin Ravaillon (2012), ´Should
we equally care about poor people wherever they may live?´, World Bank.
[10] Gallup (2013),
More Than One in Five Worldwide Living in Extreme Poverty,
23-12-2013. Results are based on telephone and face-to-face interviews with
approximately 1,000 adults, aged 15 and older, per survey administration.
Income data came from interviews were conducted in 131 countries and regions from
2006 to 2012. Data for each country have been aggregated over multiple
administrations; at least 2,000 interviews are required for a country to be
included in the income data set. For results based on each sample of national
adults, one can say with 95% confidence that the maximum margin of sampling
error ranged from a low of ±1.4 to a high of ±4.7.