Introduction

In 2003, the World Bank launched a global round of the International Comparison Program (ICP),aimed at measuring the purchasing power parities (PPP) and corresponding price and volume levels of Gross Domestic Product (GDP) on a comparable basis for more than 100 countries, grouped into 5 regions¹.

Although the comparisons involving GDPs adjusted for purchasing power parities are not yet ready, there are preliminary regional results for the major component of GDP - household consumption. Those results compare ten South American countries whose consumer prices were surveyed in 2005 to compute PPPs and apply them to household consumption for the same year. Household consumption in the South American region accounts on average for two-thirds of GDP. It is therefore unlikely that once estimated, the other components of national expenditure will bring about signifi cant changes in the ranking of GDP per capita or any marked changes in national parities.

Highlights:

• In 2005, Argentina, Chile and Uruguay were relatively better off than the other South American countries

On the basis of the estimated 2005 figures, there are no major surprises in the ranking of the first three countries relatively to the situation ten years ago. Table 1 shows Argentina remaining at the top in terms of per capita expenditures² on household goods and services, at sixty per cent above the regional average. Chile and Uruguay follow respectively with 48 and 43 per cent above average. Paraguay and Bolivia remain at the bottom of the table, the former with two thirds of the regional average and the latter with fifty two per cent. The middle group is less spread out, with Venezuela at the top, slightly above average, and Colombia roughly twenty per cent below. Brazil ranks sixth, almost ten per cent below par.

• In 2005, the most expensive countries to live in were Chile, Brazil and Uruguay; Paraguay and Bolivia were the cheapest

The best known application of purchasing power parities is to determine comparative price levels, or in ordinary language where it is most (or least) expensive to live. Table 2 below shows the newly calculated estimates – expressed in relation to the South American average. If a visitor from outside the region moved from one country to another and bought in each rigorously the same basket of goods and services, he would find that he had spent most in Chile and least, in Paraguay and in Bolivia.

When comparing price levels with

per capita expenditures, it should be noted that countries with the highest price levels are not necessarily those with the highest per capita expenditures. Chile’s price level is higher than Argentina’s, and so is Brazil’s (See Chart 1).

• Argentina spends more than its neighbours on most major household items but Uruguay leads in expenditure on transportation

How do per capita expenditures on specifi c groupings of consumer goods and services compare? Tables 3a and 3b answer the question for some broad categories at a highly aggregate level. Although prices are collected for specifi c goods and services, per capita expenditures rely on both price data (PPPs) and expenditure estimates from the national accounts. However, the most detailed national accounts estimates of expenditure are in fact aggregates, such as food, clothing, etc. Even when some detailed national accounts information is available, it is not always fully comparable between countries. In addition, values for public and private health and education are presented together as the results are not yet robust enough to be shown separately. When the full GDP results are released at the end of the year, private and public education and health will be shown as four separate categories.

• Books are expensive in Chile, Brazil and Venezuela, but cheap in Argentina

While estimates of per capita spending on specific products are not very reliable, the PPPs for some specifi c products are more trustworthy. Accordingly, a few specifi c comparisons can be made across the ten countries.

Table 4 provides some examples. Beef is much more expensive in Chile than in beef- producing countries such as Argentina and Paraguay. Coffee is relatively expensive in Chile and Uruguay, and not surprisingly, relatively cheap in Brazil. Less obviously, pharmaceutical products are expensive in Uruguay and Brazil, but inexpensive in Chile and Ecuador. Motor vehicles are cheap in Chile and in Argentina. However, in Brazil, the largest producing country, prices are not far from the regional average. Books are expensive in Chile, Brazil and Venezuela, but cheap in Argentina. Video equipment and electronics are cheapest in Chile and in Colombia. Dining out is cheap in Brazil, Bolivia, Paraguay and Colombia, while it is expensive in the top three per capita countries (Argentina, Chile and Uruguay).

• The extrapolation of data over a long period yields misleading results and exaggerates the gap between the "haves" and the "have nots"

The results obtained in this round differ significantly from the estimates based on the last round of the ICP, which were extrapolated to the present³ on the basis of a number of indicators, chief among them the national CPIs.

If one compares the present results with those extrapolated on the basis of the PPPs of ten years ago, the ranking of the countries in terms of per capita expenditure is similar. But the difference between the country with the highest spending per capita (Argentina) and the one with the lowest (Bolivia) is greatly exaggerated in the projections. The new estimates suggest that the ratio of one to the other is three to one rather than fi ve to one. Furthermore, the overall variability of the new per capita numbers is only two thirds of that of the extrapolated ones. In general the new estimates show that neither are the poor as poor as conventional statistics make out, nor the relatively rich quite as well off as is sometimes believed.

The extrapolated price levels do not compare as well as the per capita expenditures. Not only do they exaggerate the dispersion around the regional average, but the ranking is fundamentally changed (see Table 6).

The comparison of the results of this ICP round with the data extrapolated from the previous round shows the imperative need to calculate parities regularly. In the absence of a multilateral program that requires statistical offi ces to do so, any view held about which countries are relatively poor and which are relatively rich will be obsolete in a short while. And without recent results, policy makers will fi nd themselves unable to make timely and balanced decisions.

__________________________

1

South America, Africa, the Middle East, Asia, and the Commonwealth of Independent States (formerly Union of Soviet Socialist Republics).

2

Throughout the text, "per capita expenditures" refers to the real per capita expenditures as adjusted by PPP.

3

See World Development Indicators Database