Why do many US restrictionists use “non-Hispanic whites” as the normative comparison group?

While reviewing the page on Hispanic crime and illegal immigration in the United States, I noticed the following paragraph I’d written a while back while preparing the page:

Inclusion of blacks in native groups compared against?: Restrictionists generally compare the crime rate among Hispanics with that among “non-Hispanic whites” (rather than all Americans, which would include blacks). Supporters of Hispanic immigration claim that a better comparison would be against all Americans (including blacks). These two different types of analysis yield quite different conclusions because the crime rate among blacks in the United States is higher than that among all other racial categories (whites, Hispanics, and the numerically much smaller East Asians, South Asians, and other racial minorities).

At the time I wrote this, I didn’t clearly try to understand why so many restrictionists choose to use “non-Hispanic whites” as the normative comparison standard against which to judge Hispanic crime rates (and, by extension, the crime rates of prospective immigrants). [NOTE: If you read the page, you’ll discover that immigrants have lower crime rates than natives, both in total and when compared for specific ethnicities or specific combinations of ethnicity and high school graduation status. Restrictionists, however, prefer to consider the crime rates of Hispanics rather than immigrants, but this is a topic for another day.]

Here are a few random quotes from restrictionists illustrating this.

Tino Sanandaji comments on a blog post here:

Chinese textiles do not commit crime at 261% of the native white rate. [a reference ot Hispanics committing crimes at 261% the native white rate]

Another comment in a blog post on immigration here:

Hows this for a fear inspiring anecdote: The hispanic incarceration rate is 2.9 times the non-hispanic white rate…. oops, that’s a statistic. And yes, it should inspire fear if you have an ounce of rationality in you.

I haven’t been able to locate an explicit explanation from a restrictionist for why this choice was made, so I’ll just include my guesses here. I see two possible explanations for this choice by restrictionists.

Non-Hispanic white normativity as a principled position

One explanation is that the use of non-Hispanic whites as the normative comparison standard is a principled position. Non-Hispanic whites are treated as the “norm” or “normal”. Any individual or group that does as well as the non-Hispanic white norm is considered average. Any individual or group that does better than the non-Hispanic white norm is considered above average. For instance, in the context of height, groups whose average height is greater than that of non-Hispanic whites would be considered “tall” and groups whose average height is lower than that of non-Hispanic whites would be considered “short.”

Critical race theorists use the term “normativity” in conjunction with ideas of privilege and prejudice, but my use here does not connote either privilege or prejudice, though it might on on occasion be linked to these. It could be a form of “centric bias” whereby people believe that their own selves or immediate surroundings are the norm, standard, or prototype. It does not, however, mean that they automatically disparage different things. A person growing up poor may consider low incomes the “norm” but that does not mean disparagement of high incomes — quite the contrary, the person may be more easily impressed by mid-level incomes than somebody who grew up rich. In the same way, non-Hispanic white normativity does not indicate a disparagement of other groups.

Non-Hispanic white normativity as a trade-off

Here’s a more cynical explanation of non-Hispanic white normativity. Restrictionists, when choosing a comparison group to judge immigrants or immigrant ethnic groups against, have to balance two criteria:

  1. The immigrants or immigrant groups should perform clearly worse on the indicator than the comparison groups.
  2. The comparison group should be something that a large number of their readers can identify with.

In an ideal world, immigrant groups would show performance that’s clearly worse than natives on the whole, and hence immigrant groups could be compared directly against natives, or “all Americans” — this would appeal to a maximum number of Americans.

However, there are many cases, such as crime, where immigrants, and immigrant ethnic groups, don’t perform worse than natives on the whole in a clear way. Restrictionists thus need to narrow down the definition of native. At one extreme, the restrictionist could narrow down to “upper middle class college educated whites” or “Ph.D. Ashkenazi Jews” as a comparison group and immigrants/immigrant ethnic groups would perform quite badly in comparison. While this is great for (1), it compromises on (2) — the comparison group is too small and few readers would identify with it. The middle ground of choosing “non-Hispanic whites” or “native non-Hispanic whites” yields a sweet spot that makes immigrant groups look reasonably bad by comparison, and also allows a large number of readers to identify with the comparison group.

The chart featured at the top of this post is a breakdown of the US incarceration rate by race, as of 2006. Authored by the November Coalition and released into the public domain; found on Wikimedia Commons.

2 thoughts on “Why do many US restrictionists use “non-Hispanic whites” as the normative comparison group?”

  1. Vipul,

    First, a few points about the relevance of crime to open borders overall. Crime is only one kind of negative externality people can have, probably less important over time than effects on politics and corruption effects. However, it is nonetheless quite expensive. Economist David Anderson, incorporating rarely sundry costs like opportunity cost of time and private crime-prevention expenditures, estimates an annual cost of crime at $1.7 trillion for the US, around a tenth of GDP. And most of that cost comes from a few demographic minorities who add less than the average to total output. So it is initially plausible that “double GDP” genre claims get seriously clawed back for migrants whose descendants have high crime rates.


    In some ways this is an over-estimate for purposes of immigration: most crime is focused on those near the criminal, e.g. African-American crime disproportionately harms African-Americans, Hispanic-American crime disproportionately harms Hispanic-Americans. A crime victim is just as dead whether in Mexico or the U.S. However, the costs of enforcement and defense are higher in rich countries, and this consideration only applies to part of crime.

    On the other hand, there are large economic effects of crime that are often left out because of modeling difficulties. High crime rates in the U.S. contributed heavily to suburbanization, reducing the agglomeration benefits of dense cities, which many economists estimate as immense. Enormously costly NIMBY regulations in housing and environmental policy are made to reduce local crime rates. Further, making high density feasible is important to enable the effective receipt of large migrant populations.

    With respect to reference populations for discussing crime rates, it seems to me that the key context is that migration policies can be biased towards migrants who more or less resemble different populations. Chinese-Americans have very low crime rates, lower than European-Americans, and there are plenty of potential immigrants available from China. In America they also complete more education, have policy views more resembling economists than the general population, earn more income, pay more taxes, and claim fewer welfare benefits than Hispanic-Americans. China is also poorer than Mexico, and migrants from China gain more political freedom than migrants from Mexico.

    So it looks like for a given level of harm to natives and risk to rich country institutions from negative externalities, you could create more GDP, help more people, and do more good by adopting policies that would have the effect of favoring Chinese migration relative to Latin American migration. Amnesties for US illegal migrants and family sponsorship favor Hispanic migration, while point-based immigration systems that favor immigrants who will contribute a lot to the receiving country, like those in Canada or Singapore, favor Chinese migration.

    So if we are contrasting such systems with open borders it makes sense to use what we know about crime statistics for the descendants of past migrants drawn from different regions. Talking about Chinese-Americans or Swedish-Americans vs Hispanic-Americans makes sense because different migration policies will get very different mixes thereof.

    1. Actually, the $1.7 trillion figure for US annual cost of crime was apparently from an earlier study. The more recent figure including more costs was $3.2 trillion annually.

Leave a Reply