An earlier version of this paper was presented at the 20th Annual Convention, International Studies Association, Toronto, March 1979. This research was supported by a grant administered by the Paterson Centre for International Programs, Carleton University. The authors wish to thank Ari Abonyi, Abby Dann, Dianne Pothier, Ivan Sylvain, Elizabeth Smythe, and Hans Van Beek who assisted in the data compilation, and the School of International Affairs at Carleton, for providing research assistance.

¹ See Duvall, R., “Dependence and Dependencia Theory: Notes Toward Precision of Concept and Argument,” International Organization 32(1978): 51–78.CrossRefGoogle Scholar

² See Ray, J. and Webster, T., “Dependency and Economic Growth in Latin America,” International Studies Quarterly 22 (1978): 409–434.CrossRefGoogle Scholar

³ This does not, of course, detract from continued efforts to devise coherent theories of the dependencia variety. However, to the extent that the use of the concept dependence provokes the a priori dismissal of the otherwise interesting and reliable findings of the body of research on economic linkages and development, there may be cause to drop the concept altogether until alternative theoretical formulations are available.

⁴ See the following studies: Bornschier, V., Chase-Dunn, C., and Rubinson, R., “Cross-National Evidence of the Effects of Foreign Investment and Aid on Economic Growth and Inequality: A Survey of Findings and a Reanalysis,” American Journal of Sociology 84 (1978): 651–83; PCrossRefGoogle Scholar. McGowan, and Smith, D., “Economic Dependency in Black Africa: An Analysis of Competing Theories,” International Organization 32 (1978): 179–235CrossRefGoogle Scholar; Ray, and Webster, , “Dependency and Economic Growth in Latin America”Google Scholar; Walleri, R. D., “The Political Economy Literature on North-South Relations: Alternative Approaches and Empirical Evidence,” International Studies Quarterly 22 (1978): 587–624.CrossRefGoogle Scholar

⁵ Bornschier, , Chase-Dunn, , and Rubinson, , “Cross-National Evidence.”Google Scholar

⁶ Stoneman, C., “Foreign Capital and Economic Growth,” World Development 3 (1975): 11–26.CrossRefGoogle Scholar

⁷ See the studies cited in note 4, above.Google Scholar

⁸ See Caporaso, J., “Dependence, Dependency, and Power in the Global System: A Structural and Behavioral Analysis,” International Organization 32(1978): 21–22.CrossRefGoogle Scholar

⁹ Kuznets, S., “Quantitative Aspects of the Economic Growth of Nations, VIII: The Distribution of Income by Size,” Economic Development and Cultural Change 11 (1963): part 2.CrossRefGoogle Scholar

¹⁰ The primary sources for this discussion are: Snider, D. A., Introduction to International Economics (Homewood, Illinois: Irwin, Inc., 1971 and 1975)Google Scholar; and Cline, W. R., “Distribution and Development: A Survey of Literature,” Journal of Development Economics 1 (1975): 359–400.CrossRefGoogle Scholar

¹¹ Snider, , Introduction to International Economics, 1971, p. 426.Google Scholar

¹² Ibid., p. 441.

¹³ Two criticisms made by conventional economists are that even though growth spreads across sectors, this requires time, and it is possible that growth in the export sector will continue to exceed that in domestic sectors. These arguments suggest a proposition opposite to that presented.

¹⁴ Cline, W. R., “Distribution and Development.”Google Scholar

¹⁵ Moran, T., “Multinational Corporations and Dependency: A Dialogue for Dependentistas and Non-Dependentistas,” International Organization 32 (1978): 79–100.CrossRefGoogle Scholar

¹⁶ Cline, , “Distribution and Development,” p. 372.Google Scholar

¹⁷ Ibid.

¹⁸ Bornschier, , Chase-Dunn, , and Rubinson, , “Cross-National Evidence.”Google Scholar

¹⁹ The analysis was also conducted for the period 1967 to 1975, with no substantive change in results.

²⁰ This measure was also weighted by GNP, with no change in results.

²¹ The authors do not specify the time period employed.

²² Cyprus and Malta are excluded from the regional analysis.

²³ The selection of the 1970–1973 time period provides a more appropriate lag between independent and dependent variables than was the case in the study by Bornschier, et al. The source and form of all indicators are listed in Appendix B. For the investment and growth indicators, the values for two outliers were reduced to two standard deviations from the mean in order to meet the assumptions of regression analysis.

²⁴ In addition to flows, this difference indicator includes reinvested earnings and valuation adjustments.

²⁵ Expressions that imply causality are difficult to avoid. While the phrases “impact on” and “effects on” have some causal connotations, they are used here in the statistical context of regression equations which contain a priori dependent and independent variables. The language used is not intended to imply causal statements.

²⁶ The bivariate relationship between rate of growth and stock of d.f.i. is approximately zero (r =.02, N = 66), but stocks are quite strongly associated with flows, savings, and level of GDP per capita (r =.62,.44 and.60, respectively). These latter three variables are positively related to rate of growth (r =.28,.20 and.12, respectively). Thus when the effects of flows, savings, and GDP per capita are controlled, the partial b of the stock of d.f.i. becomes strongly negative.

^* Statistically significant at ≤.05 level (two-tailed test).

²⁷ Another possible explanation for the positive relationship between d.f.i. flows and economic growth is that investment is directed to countries experiencing growth; that is, the actual relationship is reciprocal.

²⁸ Moran, T., “Multinational Corporations and Dependency,” p. 91.Google Scholar

²⁹ Except that in the latter test, the F statistic for the stocks estimate is slightly below the.05 level. Substantively, as the lag between the period of stocks accumulation and growth is shortened, the negative impact of investment on economic growth declines.

³⁰ In light of the results for equation (1), the aid and trade variables are excluded from this analysis.

³¹ Such unequal sample sizes make comparisons more difficult. However, this procedure was preferable to taking the mean of the sixty-six countries in the equation because the ninety-one countries for which GNP per capita data were available almost constitute the population of underdeveloped countries in 1965. To use the mean of the sixty-six would have placed some countries that were above the real mean into the sample of poorer countries.

³² As a check on this result, a reanalysis was conducted employing only the two investment and two control variables used by Bornschier, et al.; the R² remained at.20 (see Appendix C). This difference in variance explained is likely due to the inclusion in the Bornschier, et al. study (p. 680) of several European states.

³³ Galtung, J., “A Structural Theory of Imperialism,” Journal of Peace Research 2 (1971): 81–118.CrossRefGoogle Scholar

³⁴ Walleri, D., “Economic Imperialism as a Cause for Retarded Development in the Third World.” Paper delivered at the annual meeting of the APSA, San Francisco, 09 1975. “Apparently” because the Walleri paper was unavailable; we do not, therefore, know the extent of correspondence between our respective independent variables.Google Scholar

³⁵ Bornschier, , Chase-Dunn, , and Rubinson, , “Cross-National Evidence”, p. 664.Google Scholar

Ibid., p. 677.Google Scholar

³⁷ Appropriate data for income distribution are unavailable.