Monday, February 24, 2014

Is income inequality a result of diversity?

I have long suspected that income inequality is less about market 'failures' in the private economy than a simple reflection of how diverse a country is.  The more diverse, the more unequal because groups who are very different or live in very different places confront different opportunities and issues which must ultimately be reflected in their economic outcomes.  So I decided to investigate the relationship between ethnic homogeneity - the proportion of the population that is part of the dominant race and culture and income inequality as measured by pretax, pretransfer Gini. I plotted each of the 50 states plus the District of Columbia.  I also plotted the five 'Nordic' countries that have a sterling reputation for income equality:  Iceland, Norway, Sweden, Finland, Denmark.  And here's what I found:


Source:  US Census Bureau, Wikipedia Gini Tables.


As you can see, there is a significant negative relationship between ethnic and racial homogeneity (in percent on the X axis) and income inequality (gini scores on the Y axis) - the US State Correlation Coefficient is -.60 which is moderately strong. When I add the five nordic countries (in red) and rerun the analysis they plot very consistently with US states and the correlation coefficient including them rises to -.63 which suggests that almost 40% of the variation in inequality can be explained by the level of racial homogeneity.   Finland, at the high edge of the dispersion is nordic but not ethnically scandanavian while Iceland at the very bottom is perhaps more homogenous than any other country in the world.  All of them perform similarly to their American peers.  I say peers because these countries have smallish populations and are far more akin to a mid-sized state like Missouri or Minnesota (or Wyoming if you're Iceland) than the USA.

It turns out that the various US states and nordic countries' economies deliver similar levels of income inequality depending upon the level of  'diversity' that they have.  The smaller the 'dominant' race is (and in all cases except Hawaii, that race is 'White, European') the greater inequality is.  So economy-wise the Nordics aren't naturally 'fairer' than their North American cousins. 

You may also notice that every single US state has a lower level of income inequality than the United States as a whole - the exception is District of Columbia. This makes sense because the entire US is much more diverse than any single state, just like it is much more diverse than a nordic country.  This is the reason why it makes no sense to compare these types of statistical analyses between countries with very different circumstances.  This type of league table jockeying inevitably creates more heat than light.

Now the clever chaps in the booming academic field of Envy Studies will point out that the Nordics intervene in their economies significantly more than the US does to equalize outcomes and it results in much lower inequality.  And that is a Good Thing.  So I plotted post tax post transfer Gini.  Note:  there is evidently no post tax/transfer data at the state level.  I made a macro adjustment for all 51 data points, using the US nationwide reduction in Gini from government intervention which was .11: from .49 to .38 and plotted the nordic numbers.

Source:  US Census Bureau, Wikipedia Gini Tables.

Unsurprisingly four of the nordic countries redistribute more income and reduce their (comparable) pre tax ginis substantially more: ranging from .16 (Norway) to .21 (Finland) points.  Iceland does not, only moving .08 points which I suspect is because Iceland is small and already so homogenous that elaborate redistribution just doesn't matter as much. No doubt if US states had the same control over their taxing and social welfare policies that the Nordics have we'd see some states intervening substantially more to redistribute income. However, the US has a national redistribution policy that is driven by a very complex and diverse electorate and consistent with public choice theory, such a conglomeration is less willing to accept redistribution and collective solutions than the homogeneous nordics.  (Incidentally, public choice Theory also predicts that as homogeneous countries become more diverse, redistribution and acquisition of public goods will fall which is precisely what has happened to the most 'diverse' nordic:  Sweden who has seen its welfare state trimmed and top marginal tax rates fall as it added more immigrants)

But Gini says nothing about well being.  Because there are real trade offs between more government intervention to ensure equal outcomes and other metrics like productivity, output per person, disposable income not to mention liberty and autonomy.  The nordics are more equal which requires them to be more coercive which contributes to them being poorer for both mean and median household disposable income.  And nordic citizens have many choices made for them that for better or worse Americans are trusted to choose for themselves.


Source:  OECD Better Life Index               

So what do I think I know now?   First, US income inequality when looked at at a relevant state level is what you would expect it to be given our levels of racial and ethnic diversity.  The 'best', most 'equal' countries don't do any better relative to their racial and ethnic diversity.  Second, I have only looked at one component of diversity:  racial diversity and it explains about 40% of the inequality variation.  Other differences such as geography, climate, religion, economics and form of government also drive inequality up and I believe would explain substantially more of the variation.  Third, given our national diversity we don't do as much redistribution as homogeneous Nordics do which is what you would expect and probably won't change for Public Choice reasons.  Fourth, if you want income equality it's best to make sure everyone looks and thinks like you.  It turns out that diversity is hell on equality.  Which makes sense:  the more I differ from you in heritage, religion, ethnicity, race, geography, weather the more my economic results are going to differ.  Fifth, the more you intervene to level incomes the poorer your country will be - there is dead weight cost to an economy of coercive income redistribution.  It discourages work and productivity both at the top and the bottom of the distribution. Something that apparently doesn't happen with voluntary redistribution:  Utah in real terms is among the richest of states despite having far higher levels of charitable giving which is just another word for income redistribution that doesn't include guns and handcuffs. Sixth,  perhaps the only way that US proponents of a 'fairer' or more accurately flatter income redistribution regime will be successful (with 'successful' meaning more redistribution without nasty social and economic repercussions) is if taxing and spending policies are set at the state level rather than dictated nationally as they are now.  There are many states with the homogeneity of  nordics and a similar culture who might end up at the same place as a Sweden, or at least closer.

But when one looks at inequality at a reasonable level - meaning the state level, the US  does fine.  It simply has made the choice to intervene less and thereby trade the abstract concept of 'fairness' for the real dollars and cents of wealth, innovation and productivity.  There's nothing particularly immoral or shameful about not aping the snow elves choices.  It's just that we value liberty, autonomy and wealth more and sameness less. Which is probably one reason why the Norwegian and Swedish American's ancestors moved here in the first place.


No comments:

Post a Comment