-President George W. Bush's "Tax Plan"
In my last post, I proposed a metric for evaluating how generally left-leaning (in the economic sense) a country is so that we can make informed comparisons between countries' attributes and their economic policies. I have adjusted that metric somewhat, and I now apply the following formula:
Taxes, Progressivity, and Redistributive Spending - 45%
30%: Tax burden/GDP
7.5%: Income tax/Taxes on goods and services
3.75%: Social spending as a % of GDP
3.75%: Social spending as a percentage of tax receipts
Labor - 25%
15%: Union density
5%: Employee dismissal protections (scaled average of OECD rating)
5%: Annualized minimum wage/GDP per capita
Health Care - 15%
7.5%: Government health care expenditures as a % of GDP
7.5%: Government health care expenditures as a % of total healthcare expenditures
Education - 10%
5%: Government education expenditures as a % of GDP
5%: Government education expenditures as a % of total education expenditures
Note that for health care and education, I gave equal weight to the amount of money the government spends (as a % of GDP) and the relative prominence of the government role in that area (by taking government expenditures/total expenditures). I think doing it this way captures the two distinct aspects of government involvement in these policy areas - one is how much government taxes and spends to provide certain services, and two, how much control the government wields over the overall "market" for that service. Similarly, for social spending I took the ratio of social spending to GDP and social spending to total tax reciepts. Here are the new results:
SDI 2.0
It's not surprising that SDI is strongly correlated (R = - 0.731) with lower income inequality:
SDI and Income Inequality
But what about economic performance? If Bush is right, countries with high SDI ratings should perform relatively poorly economically. One way to evaluate this claim is to look at the correlation between SDI and GDP per capita. It turns out that SDI is moderately-to-strongly positively correlated (R = + 0.481 where 0.5 is considered "strong") to GDP per capita.
SDI and GDP per capita
But this could simply mean that countries that have already acquired great wealth tend to take on more social democratic characteristics later on. In other words, it does not demonstrate that social democracy is the cause of these countries' economic success. So consider, instead, the average annual increase in GDP per hour worked from 1980-2006 (i.e., the rate of productivity increase since the beginning of the Reagan/Thatcher era). For the OECD 30 (excluding Turkey, for which data was unavailable), SDI is moderately correlated (R = +0.368) with faster productivity growth.
Note: the relevant data for former Eastern Bloc countries is not available for the Communist years, so productivity growth figures for the Czech Republic, Slovakia, Hungary, and Poland do not reflect the years before the fall of the Soviet Union.
SDI and Average Annual GDP per Hour Worked Growth, 1980-2006
Interestingly, this relationship becomes significantly stronger when you restrict the sample set to reflect only more "mature" economies like the U.S. and Canada, "Old Europe," (the non-ex-Communist countries, excluding the recently relatively poor Western European countries like Greece, Spain, Ireland, and Portugal) and Oceania (Australia and New Zealand). If we only consider this set of countries - an arguably more appropriate set for comparison given the greater economic similarities between these countries - the R = +0.527 (strong). This indicates a marked positive correlation between social democratic policies and recent produtivity growth in mature economies.
SDI and Average Annual GDP per Hour Worked Growth, 1980-2006 for "Mature" Economies
Once again, it is theoretically possible that this association reflects some alternative mutual cause, but no such cause springs to mind. The correlation is strongly suggestive of a positive relationship between SDI and economic performance, and it certainly gives the right (who do not seem to shy away from using comparative data nor suggesting that correlations directly imply causation) some serious explaining to do.
In later posts, I will try to use more data to shed some light on why it might be that high-SDI countries enjoy such strong economic performance and productivity growth.
