¹ For a summary of these changes, see Gallman, Robert E., “The Pace and Pattern of American Economic Growth,” in Davis, Lance E., Easterlin, Richard A., and Parker, William N., eds., American Economic Growth (New York, 1972)Google Scholar.

² Per capita income estimates for 1840 are used to make the 1850 calculation. These are reported in Easterlin, Richard A., “Interregional Differences in Per Capita Income, Population, and Total Income, 1840–1950,” National Bureau of Economic Research, Trends in the American Economy in the Nineteenth Century, Studies in Income and Wealth, vol. 24 (Princeton, 1960), pp. 73–140.Google Scholar For the number of jobs in 1850 we have used decennial census data for the total number of employed free males in each of the states. Income and number of jobs data for 1880 (and subsequent years) are from Lee, Everett S., Miller, A. R., Brainerd, C. P., and Easterlin, R. A., Population Redistribution and Economic Growth, United States, 1870–1950, Vol. I: Methodological Considerations and Reference Tables (Philadelphia, 1957)Google Scholar, Tables L-l and Y-l. In interpreting the meaning of the zero-order correlation between jobs and income, one should not focus on the level of this correlation. What is significant is the change in this correlation over time. The direction of the change is indicative of how labor is being reallocated through time.

³ Gallaway, Lowell E., Vedder, Richard K., and Chapin, Gene L., “The Impact of Geographic Mobility on Regional Wage Differentials: A Test of the Steady-State Equilibrium Hypothesis,” Proceedings, Business and Economics Section, American Statistical Association (Washington, D.C., 1971), pp. 386–94.Google Scholar For an analysis that is somewhat similar although not identical, see Bjork, Gordon C., “Regional Adjustment to Economic Growth: The United States, 1880–1950,” Oxford Economic Papers (March 1968), 81–97Google Scholar.

⁴ Gross flows would still occur but they would exactly cancel one another.

⁵ These are taken from our “Mobility of Native Americans,” this Journal, 31 (Sept. 1971), 613–49Google Scholar.

⁶ By 1960 the mean elasticity for the southern states had increased to a point at which it exceeded that for the northern states.

⁷ The net migration data are taken from Lebergott, Stanley, “Migration within the U. S., 1800—1960: Some New Estimates,” this Journal, 30 (Dec. 1970), 839–47Google Scholar.

⁸ The correction for price level changes in the net mirgration relationship is required due to its linear character. Such a correction would not be necessary in a log-linear function; however, the negative values for some net migration observations preclude the use of a log-linear relationship. To deflate income levels we use a combination of indices. Our deflators are the product of splicing series E-40 and E-52 found in U.S. Bureau of Census, Historical Statistics of the United States, pt. 1 (Washington, D. C., 1975). With 1910–1914 = 100, the deflators are 100.0 for 1880, 91.8 for 1900, and 221.5 for 1920.

⁹ This ratio is the inverse of what some demographers call the migration-efficiency ratio. See Shryock, Henry R. Jr., Population Mobility within the United States, University of Chicago Community and Family Study Center (Chicago, 1964)Google Scholar; Galle, Omer R. and Williams, Max W., “Metropolitan Migration Efficiency,” Demography, 9 (Nov. 1972)Google Scholar.

¹⁰ An argument for the pertinence of the range statistic is given in Gallaway, Lowell E., Manpower Economics (Homewood, Ill., 1971)Google Scholar, ch. 7. The range is measured by the difference between per capita income in the most affluent and the poorest states. All incomes are expressed in 1880 prices using the deflators described in note 8. The source for per capita income data is Lee et al., Table Y-l.

¹¹ The process of upward movement in this ratio has continued, with southern income now being over 80 percent of northern income. In fact, some scholars argue—prematurely, we think—that there no longer is a North-South differential. See, for example, Coelho, Philip R. P. and Ghali, Moheb A., “The End of the North-South Wage Differential,” American Economic Review, 61 (Dec. 1971), 932–37.Google Scholar Two recent pieces bear on the issue of convergence. Coelho, Philip R. P. and Shepherd, James F., “The Impact of Regional Differences in Prices and Wages on Economic Growth: The United States in 1890,” this Journal, 39 (March 1979), 69–85Google Scholar, suggest that interregional comparisons of money wage levels may be misleading due to price level differences. In the same issue, though, Roberts, Charles A., “Interregional Per Capita Income Differentials and Convergence: 1880–1950,” this Journal, 39 (March 1979), 101–12Google Scholar, concludes that interregional price differentials do not invalidate the convergence finding.

¹² This suggests not only North-South convergence of income levels, but a general convergence.

¹³ By 1960, this correlation had become a + .18, confirming the process of continued movement in the directions taken between 1870 and 1920.

¹⁴ The estimates are taken from Gallman, “Pace and Pattern,” with an average value of 0.31 for 1880–1920, and Scully, Gerald W., “The North-South Manufacturing Wage Differential, 1869–1919,” Journal of Regional Science, 11 (Aug. 1971), 238–52CrossRefGoogle Scholar, with an average value of 0.39 for 1880–1920.

¹⁵ Space does not permit us to detail the varying assumptions underlying the analysis. Adjustments, however, are made to abstract from the impact of immigration and differential labor force participation among immigrants.

¹⁶ Gallman, “Pace and Pattern.”

¹⁷ This is quite similar to the estimated impact of railroads on the American economy reported by Fogel, Robert, Railroads and American Economic Growth: Essays in Econometric History (Baltimore, 1964)Google Scholar.