What's all this tell us? Not a great deal and not with great confidence.

Hmm. From Do Some Countries Discriminate More than Others? Evidence from 97 Field Experiments of Racial Discrimination in Hiring by Lincoln Quillian, Anthony Heath, Devah Pager, Arnfinn H. Midtbøen, Fenella Fleischmann, and Ole Hexela.

Having grown up internationally and lived and worked internationally as an adult, Americans often regard me as quixotic when I argue from experience that aversion to the outsider is a universal human trait and that only a handful of countries (usually from the Anglophone) have a pretty strong culture of acceptance (as measured by the World Values Survey). Contrary to the story some Americans tell, America is pretty open to outsiders and much of the argument about racism is ill-informed, unfounded, or unsophisticated. It is ideological, not empirical.

Hence my interest in this research. It confirms my experiential priors.

Comparing levels of discrimination across countries can provide a window into large-scale social and political factors often described as the root of discrimination. Because of difficulties in measurement, however, little is established about variation in hiring discrimination across countries. We address this gap through a formal meta-analysis of 97 field experiments of discrimination incorporating more than 200,000 job applications in nine countries in Europe and North America. We find significant discrimination against nonwhite natives in all countries in our analysis; discrimination against white immigrants is present but low. However, discrimination rates vary strongly by country: In high-discrimination countries, white natives receive nearly twice the callbacks of nonwhites; in low-discrimination countries, white natives receive about 25 percent more. France has the highest discrimination rates, followed by Sweden. We find smaller differences among Great Britain, Canada, Belgium, the Netherlands, Norway, the United States, and Germany. These findings challenge several conventional macro-level theories of discrimination.

On the other hand, they are using a research technique (call-backs - they create identical resumes but with ethnic affiliated names and then see who gets a call-back) as a measure of discrimination which is popular but known to be riddled with methodological issues, particularly when used across boundaries.

For example, in America there is a wealth of call-back research but most of it is motivated research and the flaws are readily identifiable. About 15% of blacks in America are of foreign birth - Jamaica, Nigeria, Haiti, Brazil, South Africa, Somalia, etc. We know that each of these groups have markedly different success rates. Nigerians do better on most socioeconomic measures than do white Americans and dramatically better than African Americans. Similarly, but to a lesser extent, with Jamaicans and Haitians.

American academics have been markedly uninterested in these differences and yet they are central to whether we are talking about discrimination and or success based on race or ethnicity or culture or class, etc. It has long been my contention that what gets classified in America as racial discrimination is what would elsewhere be readily identifiable as class discrimination.

American academics, when they design call-back research, use African American names as black names when in fact those names are markers for African Americans rather than markers for black. It would be interesting to see the call-back research if they used characteristic names from other black groups such as Nigerians to see whether the responses are being triggered by race or by culture.

The point is that academics are using a highly flawed methodology to reach a conclusion not actually supportable because of the very flaws of the methodology.

So, my inclination is to appreciate the research for making the point I have long argued - America is not systematically racist, it is not uniquely racist, and it does a better job of policing invidious discrimination than most. At the same time, I substantially discount the research because of the flawed methodology. One step forward, one step back.

But for the sake of argument, let's pretend that the call-back approach does have some indicative power.

Another issue that lurks in the background is the absence of context by country. How are racial attitudes in Britain, where they have seventy-five years experience of material Asian immigration, different from those of Norway where it is much more recent. And how are those attitudes affected by within Asia origin (i.e. in Britain, Asia is primarily South Asia Hindu/Muslim whereas in Norway it might be East Asia Sinotic)?

Then there is the real issue that cultural origin indicators can often carry significant information which is accurate. It might be indicative of recency of immigration (and therefore an indicator of how much training might be involved), quality of education (a college degree from Oxford is different from a college degree from University of Angola), degree of native origin misogyny (a possible harbinger of future sexual harassment lawsuits), etc. Setting all that aside as well.

Here are the results.

Click to enlarge.

Each box is the results for a country. From left to right are black, white, MENA, Hispanic, Asian. The discrimination ratio is the ratio of callbacks for white natives to the indicated minority group. The line is the confidence interval. The number below the line is the number of studies.

Let's discard any results with fewer than three supporting studies. The authors acknowledge that they are underpowered.

First let's do an average of the discrimination ratio for each country based on the groups for whom there are three or more studies in that country, and excluding white immigrants. Raw discrimination rankings:

Germany - 1.24
Netherlands - 1.25
USA - 1.29
Norway - 1.35
Belgium - 1.49
Canada - 1.53
Britain - 1.54
Sweden - 1.65
France - 1.9

US uniquely racist and discriminatory? Doesn't look unique and doesn't look strongly discriminatory compared to others. It is in the top third for least discrimination.

Now let's look at the degree of discrimination which might be attributed solely to presumed immigrant status. Across the nine countries, the average discrimination ratio against whites is 1.12. So for all the numbers, accepting that there are employer costs of hiring someone unfamiliar with the country's laws, customs, language, etc., you would logically have to subtract .12 which is the discrimination premium attached to immigration. USA would become 1.17. Beginning to look kind of impressive.

Now let's look at discrimination against blacks. Surely the US has a uniquely endemic racism against blacks? Civil War, KKK, all that? Well, No. For those countries with measurements for discrimination against blacks. Remembering that in the US there are black immigrant groups who do much better than African Americans and even average white Americans.

Netherlands - 1.19
USA - 1.36
Britain - 1.49
Canada - 1.65
France - 2.02

What's all this tell us? Not a great deal and not with great confidence. But it is entirely consistent with my experiential observations.