Wednesday, May 11, 2011

Ethnic Diversity and the Size of Government

In recent posts I formulated the Sanandaji Principe, which stipulates that due to the left-leaning voting patterns of unskilled immigrants, we can at most have two out of three of Open Borders, Libertarianism and Democracy.

Open Borders and Democracy will inevitably lead to a welfare state, as non-libertarian immigrants sooner or later become the majority of the voters and vote themselves benefits.

One objection that people such as Swedish libertarian Economist Niclas Berggren made is that mass migration causes native voters to turn against redistribution. The reason is that economists believe that solidarity is diminished in ethnically heterogeneous societies. According to this theory voters care more about people with the same race and ethnicity as themselves, and are less willing to help the unfortunate if they have a different skin color. This theory is most prominently suggested by Harvard professors Alesina and Glaeser.

Some libertarians want to rely on this mechanism to tear down the welfare state through open borders and the ethnic tensions they believe that migration will cause.

My first reaction if that is the price of limiting the welfare state, is that I would oppose it. Milton Friedman famously stated that he would oppose reducing the welfare state unless the public was convinced in a democratic fashion that this was in their best interests. I understand that some free-marketers have turned against the very notion of "solidarity", because the left has exploited the term so much. However this should not let us lose sight of the fact that solidarity and national cohesiveness are social goods, not something that we should want to destroy through an immigration shock doctrine.

Leaving my preferences aside, I also believe that Berggren and other libertarians and liberals who rely on the Alesina-Glaeser theory are substantively wrong. Ethnic diversity overall tends to expand the welfare state, not reduce it. While the research only focuses on one effect of unskilled immigration (reduced fellowship), there are at least three effects that go the other way. Here are the main effects of increasing the share of low income minorities:

1. Solidarity is diminished and social ties are wakened, so that the majority population becomes less willing to pay taxes to help "the other". This limits the size of government. The ethnic-diversity-and-redistribution-literature has almost entirely focused on this sole effect.

2. Increasing the share of low income individuals increases the welfare state through a mechanic effect. This means even if you don't vote for any changes to the welfare state, the use of preexisting welfare programs such as unemployment insurance and public health care increases.

For instance, 71% of Hispanic immigrant households in the U.S use at least one form of public welfare, compared to 39% of native households. In Sweden, according to the latest figures around 40% of all unemployed individuals are immigrants.

Even if you don’t make unemployment insurance more generous, having groups with a higher unemployed rate automatically expands the size of government.

3. More disadvantaged citizens increases the need for a welfare state. To the extent that the welfare state reflects a desire to reduce social problems, having more deprived individuals increases the demand for more government to solve problems. The welfare state exists largely because the middle classes and the rich feel sorry for the poor. The left is not stupid or irrational, they rarely demand government intervention where there are few problems.

As immigration increases poverty and social problem, demands for government intervention grow. Note that this is consistent with lower solidarity across ethnic lines, as long as solidarity is not zero (If the new poor immigrants were your co-ethnics, voters would be even more inclined to help them).

To give you a recent example, the majority of the long term uninsured in the United States are ethnic minorities. (Long term uninsurance is a better measure, since many uninsured are just between jobs.)

According to the Medical Expenditure Panel Survey, Hispanics "represented 42.8 percent of the long-term uninsured for the period 2005-2008"

The media does not understand and will not tell you this, but the long-term uninsurance rate of non-Hispanics whites’ above 25 in the United States is merely 3%. This is incidentally one explanation why the white Tea Party activists don’t like President Obama's health care reform, they and their families already have health insurance.

The American uninsurance ”crisis” would likely never had arisen without a high percentage of minorities with extremely high long term uninsurance rates.

Similar, in Sweden the social problem currently most emphasized by the Social Democrats is child poverty. As I explained, 65% of poor children in Sweden are immigrant children (interestingly about two thirds of poor children in the United States are minority children).

Without immigration, there would be no child poverty "crisis" in Sweden for the left to mobilize politically against.

4. Though ignored by proponents of the ethnic-diversity-and-redistribution, minorities also get to vote, and they vote overwhelmingly for the left. This effect is dominant when we are discussing free migration, because with open borders in a world where 700 million people have told Gallup they would like to migrate right now, sooner or later the immigrants will become the majority of voters and make the political preferences of the natives irrelevant.

Pew recently conducted a large survey with lots of questions on economic and social issues. It shows as all other polls that African Americans and Hispanics minorities are far to the left of whites. While 12% of Non-Hispanic whites in America have views that Pew classifies as Libertarian, only 3% of American minorities are libertarian. As America becomes increasingly minority, it will become less libertarian.

The proponents of the Alesina-Glaeser theory tend to focus entirely on point one and ignore points 2, 3 and 4.

It is difficult to test the theory empirically. I will however give you two pieces of suggestive evidence. I am not going to claim that this is definitive proof, just that it is consistent with my view that the net overall effect of diversity is bigger government.

Libertarians like the Alesina-Glaeser theory, because it tells them with more immigration they can reduce willingness to pay for the welfare state. Liberals similarly love the theory because it quite explicitly states that the main reason Americans deny themselves the benefits of a European style Social Democratic system is the racism of Republican voters.

First, I plot the vote share of Obama among non-Hispanic whites with the share of non-hispanic whites in each state. The Alesina-Glaeser theory would predict that whites in states with lots of minorities should vote less Democrat, because of racially motivated lack of support for leftist policies.

In fact, there is no such overall trend. The correlation is not statistically significant, and if anything goes in the opposite direction as their theory would predict.

Sure, there are states with high share of minorities in the South - such as Georgia and Alabama - where whites came out strongly against Obama. Similarly, some very white states in New England went solidly for Obama.

On the other hand, other lily-white states such as Wyoming, Kentucky, West Virginia, Utah and Idaho voted against Obama. Similarly whites in minority states such as Maryland, New York, Nevada, New Mexico and California strongly supported Obama.
A more parsimonious explanation which corresponds better with the observed pattern than the ethnic-diversity-and-redistribution literature is that whites in conservative states voted against Obama, and whites in liberal states voted for him, with little connection to the racial makeup of the state.

A second graph plots per capita spending State and Local spending in 2007, from U.S Census State and Local Government Finances, with the share of state population that are non-Hispanic whites.

Contrary to the prediction of Alesina-Glaser, the overall effect appears to be that states with more minorities spend more per capita.

Thus minority states such as D.C, California, Maryland, New York, Illinois, New Jersey and even Louisiana and Mississippi stand out as spenders, whereas white states such as New Hampshire, South Dakota and Idaho spend the least.

While this is not definitive evidence, I believe points 2-4 tend to dominate point 1, so that the net effect of more diversity is bigger government and less solidarity. At the very least, points 2-4 should be taken into account in the ethnic-diversity-and-redistribution-literature.


A reader suggested I include my RSS-Feed. I belive it is:


  1. Given Tiebout sorting of public services and diversity, I'm not sure these graphs tell us anything. The ability to move easily across states might well mean that there is selection going on of who lives in what state with respect to both diversity and the level of welfare state.

    Given the difficulty of moving across countries, I'd be more convinced by cross-country panel data on ethic diversity and the welfare state.

  2. There is not enough variation across country. Except the United States, all OECD-countries are 90% or more white.

  3. "For instance, 71% of Hispanic households in the U.S use at least one form of public welfare, compared to 39% of native American households."

    This sentence is highly misleading. I'm sure you were just being a little careless when you wrote this, but you may want to pay more attention to the details. Your data refers to "Hispanic immigrant households" rather than "Hispanic households" and "native households" rather than "Native American households"...

  4. Fabian:

    Thanks, I changed it now. Native American certainly was misleading.

  5. "My first reaction if that is the price of limiting the welfare state, I would oppose it"

    Me too! I was for open borders until I heard a talk by Bryan Caplan explain this very point. A society of hate and distrust is not tempting.

  6. I do think there is strong evidence that racial homogeneity leads to larger welfare states. It seems like the small and racially homogeneous populations of the Scandinavian Countries have created an environment that is far too collective and cooperative.

    This seems to have led to more of a willingness on the part of the population to expand the welfare state (for the same reason that you are more willing to give money to someone in your immediate family than a distant friend).

    I know I am just repeating a theory that you already know, but I disagree that the evidence is stacked against it. The plot given for states is not weighted for population or the particular minorities in each state.

    California, for example, has a large minority population, however a large portion of these are Asians, who have higher incomes than Whites. Even though this could lead to the same effect on solidarity. It could also lead to more of a demand for a welfare state from Whites, because of their wish to keep Asians from getting too far ahead.

    Likewise, the anti-Immigrant sentiment is spread accross the entire nation, not just racially diverse states. Social Programs like Social Security, Medicare, Welfare, and Medicaid are paid for federally (mostly, at least).

    As a Kentuckian, I can say that, despite being a very White State, Kentucky Voters partially vote Republican on a Federal Level to keep Federal Tax Dollars from funding Social Programs for Immigrants.

    Perhaps this has kept Social Programs smaller in America relative to some European Countries. Perhaps other factors have played a larger role.

    I don't see solidarity as a plus. Hong Kong is hardly a classic Democracy, yet their government has preserved strong pro-Market and low corruption policies that have contributed to a high level of prosperity. The same is true, to a lesser degree, in Singapore. The point is that having a united population is a vice if they oppose policy that respects the rights of all an leads to a higher level of overall prosperity.

  7. Jim:

    I acknowledge that being homogenous increases willingness to pay for the welfare state, but I point out several other effects that go the other way.

    In your own example, you use Hong Kong. Minimum government Hong Kong is far more ethnically homogenous than Scandinavia, where 10-15% of the population are now immigrations. So is low tax Japan.

    Similarly the OECD countries with smaller welfare states, such as Germany, England, Austria, Holland, Ireland and Australia are *just as ethnically homogenous as Scandinavia*
    (basically more than 90-95% white).

    If we go back to around 1980, where the welfare states were already built, all European countries were at least 98% white, and usually with one nationality. Yet some had large welfare states, some has smaller welfare states.

    You can't just cherry pick Scandinavia=homogenous=welfare state and claim that a homogenous population causes a welfare state. Hong-Kong and Japan are homogenous too. Belgium consists of two nationalities, yet and has the highest taxes after Sweden and Denmark.

    France has the biggest non-white population in Europe and one of the largest welfare states.
    In Scandinavia currently the welfare state is used much more by immigrants and electorally supported by immigrants than by native Scandinavians.

    Similarly the United States is different from Europe in many important ways, not just by having a large African American population.
    Fundamental differences include:

    * Strength of religion
    * Never having had an aristocracy or feudalism
    * Historically having had lots of land and relatively high wages, which build legitimacy for capitalism
    * Never having had a strong labor movement and a strong Socialist/Social Democratic party
    * The Constitution and the bill of rights
    * Very different culture, with roots in the ideology of the Anglo-protestant settlers

    Alesina and Glaese can't just take one aspect where the U.S is different from Europe and attribute all differences in policy to that.

  8. "he plot given for states is not weighted for population"

    I redid the estimate weighting for population.

    The negative correlation between share whites and per capita spending becomes *stronger* (a little), and remains strongly statistically significant.

  9. It is definetley true that there are effects going both ways. I didn't really think of France, which is ethnically diverse (for Europe) and as socialistic as the Scandinavian states.

    As you said, people are more willing to pay for welfare states in more homogeneous communities. At the same time, however, there is less of a need for them without such a large immigrant population.

    It seems like the preference for Social Democracy or Capitalism is based less on the racial makeup and more on the philisophical roots of whatever country it is. Anglo Saxon Countries tend to favor free markets more than Scandinavian Countries regardless of immigrant population.

    I do see a growing welfare state as a major problem in America and Europe. You said that there we can only have two of these three things: Democracy, Open Borders, and Limited Government. Even if we could only have one of those, I would easily choose Lmited Government.

  10. As you see Google-owned Blogger crashed and removed all comments.

  11. Fascinating stuff, really interesting analysis. I never hear this kind of stuff in media discussions about economic policy.

  12. It is interesting, but I suspect that the results are heavily driven by DC, which probably ought not be included in the analysis.

  13. frank:

    First, why shouldn't it? It is the biggest non-white entity in the United States, and these are local and state costs, and do not include federal costs.

    D.C has the highest spending per pupil in public schools, for instance, precisely because of the high share of disadvantaged minority students.

    At any rate, removing D.C doesn't change the results. Also, weighting by population (which makes D.C a small contributor) makes the results slightly *stronger*.

  14. It looks as if in your 2nd graph that the richer states are above the regression line. What would happen to your correlation if you controlled for state per capita income?

  15. Well, in Orange County California Santa Ana and Anaheim which have high level of illegal immigrants have about a 90 percent usage of free and reduce lunch programs versus more whiter schools in Orange County that ranged from 6 to 30. The 30 percent also tend to have Hispanic immigrants. Orange County Ca where whites have a low poverty shows that Hispanics and some Asian groups have more welfare. Radical Libertarians make the wrong assumption that Hispanics just stay a few years and don't have offspring. Hispanics have more children out of wedlock and more welfare use for native born children than Whites in both California and Texas. Hispanics come from countries that might have low welfare states but the government own's the oil company and so forth. Hispanics also support gun control more than whites since Mexico goes or other Latin America countries.

  16. Not Really, Orange County California has more Asians than San Diego and votes more Republican. Asians don't mean the whites in California support the welfare state. Most whites in California that vote Democratic are not heavily welfare state types but are into the environment, for abortion rights and so forth. A hardcore left group among the occuyper bunch but the social conservative Republican does not appeal to whites in Los angeles, there are few libertarians in So Calif that are good on taxes and illegal immigration since the Hispanics population has the highest usage in Calif and the only whites that vote Democratic more than Hispanics are in the Bay area which is traditionally liberal and left.


Google Analytics Alternative