My interest today is not in what happened or in the responses of MSNBC and the people it insulted, but rather in the substance of the claim: that people on the right side of the political spectrum would be opposed to mixed-race families including an adopted child. This follows a somewhat similar MSNBC kerfuffle over jokes implying that an African-American grandchild was somehow out of place in a Romney family photo.
I thought perhaps a little academic light might clear up some of the press’s ignorance about adopted and step-families in America (the Romney case).
The General Social Survey, which is the best-run of the omnibus surveys (getting about a 70 percent response rate as opposed to a 1-25 percent response for more typical opinion polls) has data on family makeup. I was able to analyze their 2000 through 2012 data on 723 households in which any family members were adopted children or step-children of the head of the household or of the spouse of the respondent. For almost all of these households, the GSS reports both political views and whether the household is mixed race, as opposed to a white household, an African-American household, or whatever.
Not surprisingly, there is no statistically significant left-right political differences in the proportion of adopted or step-families that are in mixed race households. Indeed, among families with step-children or adopted children, 11 percent of conservatives were living in mixed race households compared to 10 percent of liberals living in mixed-race households.
Similarly, 9.4 percent of Republicans living in step- or adopted families were in mixed-race households, compared to only 8.8 percent of Democrats in such families. (Again, this small advantage for Republicans is not large enough to be statistically significant).
If one breaks things down further by both party and political orientation, only 7.7 percent of liberal Democrats and 3.6 percent of moderate Democrats lived in mixed-race adopted or step-households, compared to an insignificantly different 10.6 percent of conservative Republicans.
Thus, there is no evidence in the GSS data that Republican, conservative, or conservative Republicans who were living with step-children or adopted children were less likely to live in mixed-race households than Democrats, liberals, liberal Democrats, or moderate Democrats in adopted or step-families. Indeed, in each instance the point estimates for living in a mixed-race household were insignificantly higher for the right side of the spectrum than for the left side.
As Matt Welch suggests at Reason, when some in the press think they are attacking bigotry, they are instead spreading it themselves.
UPDATE (Friday, 3:40pm, ET):
My original post dealt with the adopted and step-child situation because that had been on my mind since the Romney family portrait dispute. After scheduling the post last night for publishing this morning, I began thinking about a family situation closer to that of the latest MSNBC contretemps.
I went back to the most recent General Social Survey (2012) to look at those who lived in households that included a child of the head of household or a child of the spouse of the head of household (774 respondents). Families with biological, adopted, and step-children were included (essentially variables RELHH1-RELHH14 & RELSP1-RELSP14 with codes 4 & 41-43). I also broke the data down by minority v. non-Hispanic white (non-Hispanic whites were those who were white and non-Hispanic on all the variables RACE, RACECEN1, RACECEN2, RACECEN3, and HISPANIC).
On this analysis of families with children, again the differences are not significant:
- 11.9% of conservatives live in mixed-race families compared to 11.4% of liberals.
- 9.5% of Republicans live in mixed-race families compared to 11.2% of Democrats.
What was interesting was the breakdown by race:
- 26.1% of conservative minorities live in mixed-race families compared to 21.2% of liberal minorities.
- 2.0% of non-Hispanic white conservatives live in mixed-race families compared to 2.4% of non-Hispanic white liberals.
- 26.6% of Republican minorities live in mixed-race families compared to 17.2% of Democratic minorities.
- 2.8% of non-Hispanic white Republicans live in mixed-race families compared to 0.7% of non-Hispanic white Democrats.
While again the differences were not significant, the pattern of point estimates was interesting: while overall, 9.5% of Republican families live in mixed-race households compared to 11.2% of Democrats, when you break down by race, the point estimates were higher for both Republican minorities and Republican whites. These results open up the possibility that in some databases Democrats could indeed live in more mixed-race families with children. But what would be driving that result – if it were found – was simply that more minorities identify as Democrats and the percentage of mixed-race households is higher among minority respondents. Given that, it is perhaps a little bit surprising that Democrats in families with children are not more likely to live in mixed-race households.
On the conservative-liberal dimensions, the pattern was less interesting. The point estimates for conservatives were insignificantly higher both overall and for minorities. That’s probably because—though African-American Republicans are very rare—African-American conservatives are not. A few years ago, I looked at that question using several data sources (including Pew and the GSS) and found that among African-Americans, about 25% identified as liberal and 25% identified as conservative. After Barack Obama’s election, that has shifted, but just a little. Among African-Americans in the 2012 GSS, while only 2% identify as Republicans, 22.6% consider themselves conservatives compared to 27.7% considering themselves liberals. Among all minorities (those not non-Hispanic whites), 25.8% consider themselves conservatives compared to 27.4% identifying as liberals (just a 1.6% difference).