Table 2.

OLS Regressions of the Return to Migration from Norway to the United States

	Dependent variable = ln(earnings)
	Population (1)	Match 1 (2)	Match 2 (3)	Match 3 (4)	Weighted (5)	Match 1 Iowa data (6)	Add penalty (7)
In US	0.609 (0.017)	0.606 (0.009)	0.644 (0.009)	0.572 (0.015)	0.641 (0.024)	0.554 (0.010)	0.466 (0.009)
N	122,620	17,501	33,641	7,596	14,647	17,352	17,501

Notes: Standard errors are reported in parentheses. All regressions control a quadratic in age. The first column contains a representative sample of the population of Norwegian-born men between the ages of 38–50 in 1900 from the 100 percent 1900 Norwegian census and 1 percent 1900 US census sample (IPUMS). Column 2 reports estimates from the first matched sample, which is based on an iterative matching strategy that searches first for an exact match and then for matches in a one- or two-year age band. Column 3 uses the second matched sample, which allows men to match in Norway by name, age, and province of birth. Column 4 reports estimates from the third matched sample, which instead requires that matched observations be unique within a five-year age band. Columns 5 through 7 return to the first matched sample. In column 5, US migrants are assigned earnings from the 1915 Iowa census (appropriately adjusted for inflation). We lose 157 observations whose occupations do not match categories in the Iowa census. In column 6, we reduce the Cost of Living earnings by 13 log points in each occupation based on the earnings penalty for Scandinavian migrants reported in Hatton and Williamson (1994). Column 7 weights the matched sample to reflect the urban status, asset holdings and occupational distribution of fathers in the full population. We lose 2,905 observations because of missing information (primarily missing data on fathers’ occupations).