Author's note: I wish to thank Jane McCormick for the graphics, and Jenai Gardner for the typing of the manuscript.

¹ Three of the countries without large oil deposits—Morocco, Tunisia, and Jordan—were significant exporters of phosphates, and so benefited from the fourfold increase in phosphate prices which occurred between 1973 and 1974.

² GNP is ihe market value of all final goods and services produced by factors (land, labor, and capital) owned by nationals of the country in question. GDP is the value produced by factors located within the country's boundaries. Before the recent nationalizations, GDP exceeded GNP by an appreciable margin in the Middle Eastern oil states, since a large fraction of the income generated within these states was paid as profits to the foreign owners of oil facilities. National income measures the factor income paid to nationally owned factors of production, and therefore equals GNP minus depreciation and indirect taxes.

³ Wilfred, Beckerman, In Defense of Economic Growth (London: Jonathan Cape, 1974), p. 85.Google Scholar

⁴ This is at least my own conclusion on what is admittedly a controversial issue in development economics. For one piece of supporting evidence among many that might be cited, see Gary, S. Fields, “Who Benefits from Economic Development? A Reexamination of Brazilian Growth in the 1960s,” American Economic Review, 67 (09 1977), 570–582.Google Scholar

⁵ The repricing estimates for 43 of these countries are discussed in Irving, B. Kravis, “A Survey of International Comparisons of Productivity,” Economic Journal, 86 (03 1976), 1–44.Google Scholar Estimates for the remaining four countries are discussed in Irving, B. Kravis et al. , “Real GDP Per Capita for More Than One Hundred Countries,” Economic Journal, 88 (06 1978), 215–242.Google Scholar

⁶ The derivation of the Iranian estimate is reported in detail in Irving, B. Kravis et al. . International Comparisons of Real Product and Purchasing Power (Baltimore: Johns Hopkins University Press, 1978). Two repricing estimates of per capita Iranian GNP in 1970 are given by Kravis et al.. one based on the geometric mean method described above and the other using a specially constructed set of “international prices” to value output. The two estimates are, respectively, $1,089 and $1,236 (in dollars of 1976 purchasing power), both of which are significantly higher than my estimate of $887, discussed below. The estimates of the Kravis group are based on output quantities reported by the Bank Markazi, several of which seem exaggerated. To cite one example, it is difficult to believe that per capita construction of office buildings in Iran was as high as 70 percent of the American level.Google Scholar

⁷ For the original development of this estimation technique, see Wilfred, Beckerman and Robert, Bacon, “International Comparisons of Income Levels: A Suggested New Measure,” Economic Journal, 76 (09 1966), 519–536.Google Scholar

⁸ Robin, Barlow, “A Test of Alternative Methods of Making GNP Comparisons,” Economic Journal, 87 (09 1977). 450–459.Google Scholar

⁹ Among the 41 reference countries. 37 possessed data on all three indicators. Equation (1) was estimated by ordinary least squares from data for these 37 cases, and yielded an R² of 0.97 (i.e., the equation explained 97 percent of the variance in log X₁).

¹⁰ Barlow, , “Test of Alternative Methods,” p. 455.Google Scholar

¹¹ Equation (2) was estimated from data for 38 of the 41 reference countries and yielded an R² of 0.94.

¹² The 1976 edition of the World Bank's World Tables gives estimates of Qatar's GNP which are implausible in the sense that the level of oil output in Qatar implies a GNP roughly twice as large as what the World Bank reports. See n. 13 for more details on the relationship between oil output and GNP. The 1980 edition of World Tables gives no estimate of Qatar's GNP, either before or after 1972, and neither do the various editions of the U.N.'s Yearbook of National Accounts Statistics (at least through the 1979 edition).

¹³ National oil output is government oil revenues divided by the posted price of crude oil (Arabian Light f.o.b. Ras Tanura). This particular indicator of oil activity was chosen as being appropriate for estimating the growth of real GNP over time, since it reflects changes in output volume and in profit shares between government and concessionaire (changes which affect real GNP), but does not reflect changes in oil prices or terms of trade (changes which do not affect real GNP). One might note as an aside that a measure of real products which did reflect changes in the terms of trade (the ratio of an export price index to an import price index) would be superior to real GNP for estimating changes in welfare over time, but such measures are not generally available for Middle Eastern countries before 1960. One of the few examples is the series on “real gross national income” estimated by Mead for Egypt (Donald, C. Mead, Growth and Structural Change in the Egyptian Economy). Equation (3) is based on 45 observations of per capita dollar GNP, as estimated by equation (2), and per capita national oil output, of which 13 pertained to Kuwait (for the years 1960–1972), 13 to Saudi Arabia (1960–1972), II to Libya (1962–1972), 6 to Oman (1967–1972), and I each (for 1972) to Bahrein and the Emirates. Sources of data on national currency GNP in these six countries, needed for the estimation of dollar GNP via equation (2), are given in n.j of the Appendix. Data on oil revenue and prices were obtained from OPEC, Annual Statistical Bulletin, 1973. R² for equation (3) was 0.96.Google Scholar

¹⁴ Discrepancies of this magnitude or even larger have been found for some countries outside the Middle East where detailed repricing studies have been conducted. The ratio between the level of per capita dollar GNP determined by such studies and the level resulting from official exchange rate conversion is 250 percent for Columbia and 335 percent for India. See Kravis, et al. , International Comparisons, p. 10.Google Scholar

¹⁵ See, for example, Time (01 10, 1977), p. 50: “The foreign money entering Israel spurred growth that has given Isarel a G.N.P. only a shade smaller than that of Egypt, though its population, at 3.3 million, is less than a tenth the size.”Google Scholar

¹⁶ Except for the pre-oil years in the Emirates, when another estimation method described in the appendix below was used.

¹⁷ Equation (1) itself could not be used in these cases, since for the years and countries in question there were no data on newspaper circulation. When there was information on both energy consumption and telephones, the following equation was used:

log X₁ = 1.807 + 0.290 log X₂ + 0.294 log X₃(4)

When nothing more than telephone data was available, the following was used:

log X₁ = 2.544 + 0.552 log X₃(5)

Equations (4) and (5) were each estimated from data for 37 of the 41 reference countries mentioned above. R² was 0.96 in (4) and 0.92 in (5).

¹⁸ With a semilogarithmic scale, a steepening (flattening) of the growth curve from one year to the next indicates an increase (decrease) in the percentage rate of growth.

¹⁹ For the purpose of these regional estimates of per capita GNP, the Middle East is defined as the 23 countries whose growth curves are shown in figure 1, plus Gaza. As indicated in the appendix below, per capita GNP in Gaza in 1966 (expressed in Jordanian dinars) was estimated to be 23.2 percent of per capita GNP in Jordan. In making the regional per capita GNP estimates, the same ratio between per capita GNP in Gaza and in Jordan is assumed to prevail in all other years of the period between 1950 and 1972.

²⁰ It should be noted that the level of real per capita GNP in Saudi Arabia, quite apart from its rate of increase, is highly uncertain. This is because there is much disagreement about the size of the population. The Saudi government claims a 1972 population in excess of six million, but some observers think that the real figure is only about half as large as that. In this paper one of the smaller estimates of the Saudi population has been used.

²¹ Also in these years the posted price of crude oil started to rise significantly, after more than twenty years of stability. But as noted above, this development, involving changes in the terms of trade, has no effect on real GNP as conventionally defined.