Table 5.

Sensitivity of OIS-ACS difference to sampling error, assumptions about emigration, and assumptions about coverage error (see notes below)

		Difference is greater than can be explained by plausible assumptions about:
	% Differencea	Sampling Errora	Emigrationb	Coverage Errorc	Any of the 3
All Countries/Regions	7	Y	.	Y	.
Mexico	31	Y	Y	Y	Y
Central American/Caribbean	6	Y	.	Y	.
Asia	1	.	.	.	.
Other Regions	8	Y	.	Y	.
Men, <5 years of U.S. Residence
Mexico	2589	Y	Y	Y	Y
Central American/Caribbean	1404	Y	Y	Y	Y
Asia	403	Y	Y	Y	Y
Other Regions	369	Y	Y	Y	Y
Women, <5 years of U.S. Residence
Mexico	1727	Y	Y	Y	Y
Central American/Caribbean	1128	Y	Y	Y	Y
Asia	214	Y	Y	Y	Y
Other Regions	371	Y	Y	Y	Y
Men, 5+ years of U.S. Residence
Mexico	43	Y	Y	Y	Y
Central American/Caribbean	5	Y	.	Y	.
Asia	−2	.	.	.	.
Other Regions	3	.	.	.	.
Women, 5+ years of U.S. Residence
Mexico	13	Y	Y	Y	Y
Central American/Caribbean	0	.	.	.	.
Asia	−1	.	.	.	.
Other Regions	8	Y	.	Y	.
Mexican Men
18-29	29	Y	Y	Y	Y
30-39	29	Y	Y	Y	Y
40+	105	Y	Y	Y	Y
Mexican Women
18-29	2	.	.	.	.
30-39	11	Y	.	Y	.
40+	34	Y	Y	Y	Y
Non-Mexican Men
18-29	4	.	.	.	.
30-39	0	.	.	.	.
40+	5	Y	.	Y	.
Non-Mexican Women
18-29	−1	.	.	.	.
30-39	3	Y	.	Y	.
40+	8	Y	.	Y	.

^aassuming moderate emigration (SSA series) and no coverage error (these estimates come from Table 1, last column for the top panel; Table 2, 3^rd and 6^th columns for the middle panel; and Table 3, 4^th and 8^th columns for the bottom panel)

^bACS estimate (assuming no coverage error) falls significantly outside range of OIS estimates while varying emigration assumptions

^cOIS estimate (assuming moderate emigration) falls outside range of ACS estimates while varying coverage error assumptions

Y = yes