Absolute poverty measures for the developing world, 1981–2004

Footnotes

• ^aTo whom correspondence should be addressed at:
  World Bank, MSN MC3–306, 1818 H Street NW, Washington, DC 20433.
  E-mail: mravallion{at}worldbank.org

• Author contributions: S.C. and M.R. performed research; S.C. analyzed data; and M.R. wrote the paper.

• The authors declare no conflict of interest.

• This article is a PNAS Direct Submission.

• ↵ b The theoretical arguments for and against this approach are discussed in ref. 2. Relative poverty measures for the developing world can be found in refs. 1 and 3.

• ↵ c Examples include refs. 5 and 6.

• ↵ d For example, ref. 7 attributes up to 40% of the difference between the (higher) growth of gross domestic product per capita and (lower) growth of mean household per capita consumption from household surveys in India to unreported increase in the incomes of the rich. Selective compliance with random samples could well be an equally important source of bias, although the sign is theoretically ambiguous; ref. 8 provides evidence on the impact of selective nonresponse for the U.S.

• ↵ e The use of a per capita normalization is standard in the literature on developing countries. This stems from the general presumption that there is rather little scope for economies of size in consumption for poor people. However, that assumption can be questioned; see ref. 9.

• ↵ f The annual World Development Indicators provides data on a broad set of indicators, including poverty measures, but also measures of health and education attainments (10).

• ↵ g Coverage varies across regions, from 78% in SSA to 98% in Eastern Europe and Central Asia and South Asia.

• ↵ h The poverty gap index is the mean distance below the poverty line as a proportion of the line where the mean is taken over the whole population, counting the nonpoor as having zero poverty gaps.

• ↵ i For further discussion of the difference between these two methods and the bearing on poverty measurement, see ref. 14.

• ↵ j The precise method used varies from country to country, depending on the data available. For more information on the alternative methods found in practice, see ref. 11.

• ↵ k The regression coefficients on time are −0.83 (SE = 0.09) and −0.77 (SE = 0.05) for $1 and $2, respectively. Note that this trend rate of poverty reduction is more than the rate of 0.6% points per year that would be more than enough to halve the 1990 $1-a-day poverty rate by 2015, which is the first of the United Nations' Millennium Development Goals.

• ↵ l The regression coefficients on time are −0.45 (SE = 0.03) and −0.28 (SE = 0.03) for $1 and $2, respectively.

• ↵ m The agrarian reforms that commenced in the late 1970s are believed to have brought a huge reduction in the number of poor over a fairly short period. For further discussion of both the data for China and the various policy reforms impinging on poverty over the 1980s and 1990s, see ref. 17.

• ↵ n The main reason for the sharp reduction in poverty in China in the mid-1990s was probably that the government brought the procurement prices for its foodgrain quotas up to market levels, which entailed a substantial drop in its (implicit) taxation of farmers; for further discussion see ref. 17.

• ↵ o Regressing SSA's share of the poor on time, the prediction for 2015 is 39.4% (SE = 1.2%).

• Abbreviations:

  COL,

  cost of living;

  EAP,

  East Asia and Pacific;

  ECA,

  East Central Asia;

  LAC,

  Latin America and the Caribbean;

  MNA,

  Middle East and North Africa;

  NAS,

  national accounts;

  PA,

  poverty assessment;

  PPP,

  purchasing power parity;

  SAS,

  South Asia;

  SSA,

  sub-Saharan Africa.

• © 2007 by The National Academy of Sciences of the USA