ASi" kUCTURE FlOR DEVELOPMENT
ASi" kUCTURE FlOR DEVELOPMENT
ASi" kUCTURE FlOR DEVELOPMENT
Create successful ePaper yourself
Turn your PDF publications into a flip-book with our unique Google optimized e-Paper software.
Columns 2 thrugh 7 give the shares of popula- two changes from previous editions: GDP has been<br />
tion quintiles and the top decile in total income or replaced by GNP; and PPC (purchasing power of<br />
consumption conre =- hav expenditure bee copie for 45 low- frmtw.am97 and miiddle- currencies) by FF1'. FP1P is the term commonly<br />
income countries and 20 high-income countries. The used to refer to the parities com'puted for a fixedt<br />
data sets for these countries refer to different years basket of products, even though theoreticaly these<br />
betweent 1978 and 1992 and are drawn mostly from are more apprprately labeled PPC The data innationally<br />
representative household surveys.<br />
dude (a) resuldts of the International Comparitson,<br />
'The data sets for the low- and midddle-income<br />
countris havebeen cmpiledfrom to main<br />
Programme (ICP) Phase VI for 1990 for OECD coun-<br />
tries extrapolated backw.ard to 1987 (b) results of<br />
sources: government statistical agencies (often ICP Phase V for 1985 for non-OECD countries exusing<br />
published reports) and the World Bank trapolated to 2987; (c) the latest available results<br />
(mostly data originating from the Living Standards from either Phase IV for 3980 or Phase EII for 1975<br />
Measurement Study and the Social Dimensions of extrapolated to 1987 for countries that participated<br />
Adjustment Project for Sub-Saharan Africa). Wihere in the earlier phases only; (d) a World Bank estimate<br />
the original unit record data from the household for China and the economies of the form-er Soviet<br />
survey were available, these have been used to cal- Union; and (e) ICP estimates obtained by regression<br />
culate direcly the income (or expenditure) shares of for the ren aining countries that did not participate<br />
differenit quantiles; otherwise, the latter have been in any of the phases. Economies whose 1987 figures<br />
estimated fr-om the best available grouped data. For are extrapolated from regression estimates are footfurther<br />
details on both the data and the estimation noted.<br />
methodology, see Chen, Dalt and Ravallion 1993. The blend of extrapolated and regression-based<br />
The data for the high-income OECD ecoDnomies are 1987 figures underlying column 8 is extrapolated to<br />
caeh tie refe to<br />
based on information from the Statistical Office of 1992 using Bank estimates of real per capita ONP<br />
the European Union (Eurostat), Thze Luxembourg it- growth rates and expressed as an index (US = 100)<br />
come Study, and the OEC'D. Those for other high-in- in column 9. For coutrides that have ever particicome<br />
countries come from national sources_. paled in the ICP' as well as foDr China and the<br />
There are several comparability problems across economies of the FSU, the latest available PPPcountries<br />
in the underlying household surveys, based values are extrapolated to 1992 by Bank esti-<br />
These problems are diminishing over time as survey mates of growth rates and converted to current "inmethodologies<br />
are both imnproving and becom-ing temnationtal dollars" by scaling all results up by the<br />
more standardized, particularly under the initia- US. inflation rates. The blend of extrapolated and<br />
tives of the United Nations (under the Household regression-based 1992 estimates is presented hin col-<br />
Survey Capability Program) and.the World Bank urnn 10. Economries whose 1987 figures are extrapo-<br />
(under the Living Standard Measurement Study lated from another year or imnputed by regr'ession<br />
and the Social Dimensions of AdjustmentProject for are footnoted accordingly- The adjuistments do not<br />
Sub-Saharan Africa). The data present-ed here take account of changes in the terms of trade.<br />
should nevertheless be interpreted with caution. In The ICP recasts traditional national accounts.<br />
particular, the following three sources of noncompa- through special price collections and disaggregation<br />
rability ought to be noted. First, the surveys differ in of GDP by expendifture components. 10P details are<br />
using income or consumption expenditure as the prepared by national'statistical offices, and tlhe reliving<br />
standard indicator. For 28 of 'the 45 low- and stilts are coordinated by the U.N. Statistical Divisio n<br />
middle-income countries, the data refer to con- (UNSTAT) with support from other. international<br />
sumnption eqxpnditure. Typically, income is more agences, particularly Eurostat and the OECD. The<br />
unequally distributed thian consumption. Second, World Bank, the Econormic Commission for Europe,<br />
the sresdiffer in using the household or the in- and the Economic and Social Commission for Asia<br />
dividual as their unit of observation; in the firt and the Pacific (ESCAP) also .contribute to this exer-<br />
cs,tequantie reeopercentage of households cise. For Nepal, which participated -in the 1985 ex-*<br />
or per capita, rather than percentage of persons. ercise, total GD? data were not available, and corn-<br />
Third, the surveys diffeer according to whether the parisons were made for consumption onlye<br />
units of observation are ranked by household or per Luxembourg and Swaziland are the only tvo<br />
capita income (or consumption). The footnotes to economies with populations under I mifllon that<br />
the table identify hese differences for each country. have participated in the ICp; their 1987 results, as a<br />
The 1987 indexed figures on PPP-based GNP per percentage of the US. results, are 81u 1 and 15.0, recapita<br />
(US= 100) afe presented in column Note spectively. The next round of ICP surveys, for 199,<br />
245