Don’t Judge a Nation by the Size of Its Government

February 21, 2013

A recurring issue in the political debate is that of big or small government. From the political right, big government is invariably seen as an objectively bad thing. It breeds corruption and it lulls hard working citizens, and especially minorities, into a false sense of entitlement. It leads to public overspending and fiscal disaster. The left are not crazy about framing the debate in the big versus small dichotomy. They could say that big government means social security, public schools, infrastructure and so on. But since most people are naturally suspicious of the government, the term “big government” will never have a good ring to it. It just sounds too much like Big Brother (and sometimes it is). Recently president Obama even went as far as to say the following in his State of the Union speech,

It is not a bigger government we need, but a smarter government that sets priorities and invests in broad-based growth.

And he is a big government guy!

This got me thinking. Is the size of the government really objectively good or bad? There are no official measures the government size (that I know of at least) but I think total tax revenues as a percentage of the GDP could be a reasonable proxy.  I selected 38 countries that are comparable in terms of development, that is most countries in Europe and USA, Canada, Australia, and the richer countries in East Asia. I omitted really small countries since they would affect the result disproportionally. That gave me a list of 38 countries listed below, in order of size (biggest first),

  1. Denmark 49.0
  2. Sweden 47.9
  3. Belgium 46.8
  4. France 44.6
  5. Finland 43.6
  6. Norway 43.6
  7. Austria 43.4
  8. Italy 42.6
  9. Germany 40.6
  10. Netherlands 39.8
  11. Slovenia 39.3
  12. Hungary 39.1
  13. United Kingdom 39.0
  14. Spain 37.3
  15. Argentina 37.2
  16. Portugal 37.0
  17. Russia 36.9
  18. Israel 36.8
  19. Czech Republic 36.3
  20. New Zealand 34.5
  21. Poland 33.8
  22. Estonia 32.3
  23. Canada 32.2
  24. Australia 30.8
  25. Ireland 30.8
  26. Latvia 30.4
  27. Greece 30.0
  28. Slovakia 29.5
  29. Switzerland 29.4
  30. Japan 28.3
  31. United States 26.9
  32. Korea, South 26.8
  33. Croatia 26.6
  34. Belarus 24.2
  35. Lithuania 20.9
  36. Chile 18.6
  37. Singapore 14.2
  38. Hong Kong 13.0

Next, I looked up some key indicators of how well a country is doing:

1. GDP (PPP) per capita (IMF 2010-11)

2. Gross government debt as a percentage of GDP (IMF 2011)

3. Corruption Perceptions Index (CPI) from Transparency International (2012).

4. Murder rate by country according to UNODC, most recent year.

I chose murder rather than crime in general since it’s a more robust measure than lesser crimes that will be reported in varying degrees in different countries.

And then I did the scatter plots. Starting with GDP (PPP) per capita (IMF 2010-11),

Government size/tax burden on the x-axis and GDP (PPP) per capita on the y-axis.

Government size/tax burden on the x-axis and GDP (PPP) per capita on the y-axis.

As you can see there is practically no correlation at all, a negligible 0.04. Next the public debt, does big government lead to fiscal irresponsibility?

Government size/tax burden on the x-axis and gross government debt as a % of GDP on the y-axis.

Government size/tax burden on the x-axis and gross government debt as a % of GDP on the y-axis.

Again, the correlation is negligible at 0.05. But what about countries that are in real trouble? If we look at the  the ten countries with the largest debts, they have an average size/tax burden of 33.3, which is very close to the overall average of 34.0. Next we look at the Corruption Perceptions Index,

Government size/tax burden on the x-axis and Corruption Perceptions Index on the y-axis.

Government size/tax burden on the x-axis and Corruption Perceptions Index on the y-axis.

As you can see from the regression line, now we have something, a small correlation of 0.16. But note that the CPI measures lack of corruption so a higher score is better. This means that corruption actually decreases as the size of the government increases. This seems counterintuitive since big government should provide more bureaucracy for corruption to thrive on. But it’s a well-known fact that big government countries like those in north western Europe are among those with the least corruption in the world. Finally crime, or murder to be precise,

Government size/tax burden on the x-axis and murder rates per 100K inhabitants on the y-axis.

Government size/tax burden on the x-axis and murder rates per 100K inhabitants on the y-axis.

Here we a have also have a slight but negative correlation, -0.19. This is more what you’d expect since a small government means less social security and less social control, and that should translate to more crime.

So what’s the verdict? With such small correlations it’s hard to say that the data gives much support for big or small government. It may seem to lean towards bigger is better but those correlations are just too low to make much of. The conclusion is rather that government size is a matter of preference. If you value personal freedom and feel that everyone should take more responsibility for themselves, then a small government is the way to go. If you value a society that takes care of all its citizens even if it stands in the way of your personal freedom, then a big government is the answer.


A Few Thoughts on Gun Control and Media-Driven Politics

January 21, 2013
Making a difference?

Making a difference?

Obama’s recently declared plan on gun control in response to mass shootings like that in Sandy Hook aims at reducing this sort of violence by stricter legislation. Some of the main points are better background checks, a ban of assault rifles, and limiting magazines to 10 rounds.

This sounds pretty reasonable but will it really have an effect? To get this problem in proportion we need to look at some basic stats. According to FBI, homicides involving three or more victims are very rare. They account for less than one percent of all homicides, and they have done so since 1980, which is as long as they’ve kept records on this type of crime. In 2008 homicide with four victims were at 0.2 percent and five or more victims at 0.1 percent. It’s not that I want to diminish the fact that children in elementary school were gunned down but according to the Bureau of Justice Statistics some 500-700 children under the age of five have been killed annually since 1975 in America, hardly anyone of them were victims of mass shootings. Are they somehow less dead?

Also, does less guns mean less gun violence?  There was a ban on assault weapons from 1994 to 2004 with no discernable impact on gun violence. Another way to look at it by comparing how armed a state is in comparison to its firearm deaths. The Daily Beast made ranking based on background checks. If we compare that to death by firearm (provided by Statemaster) there is not much of a connection either. Using a calculator from easycalculation.com, I typed in every state (except Kentucky which has a different policy making their numbers completely different from the other states). It turns out, Washington DC, the least armed state, had the most number of deaths by firearm.  And heavily armed states, like Utah and the Dakotas, were under the average in deaths by firearm. The overall correlation was a meagre 0.246. So the old argument that guns don’t kill people makes sense.

As for background checks, well that actually also makes sense. It’s not a controversial claim that the prisons are full of boys and men with personality disorders. Banning them from owning guns might actually accomplish something. But all in all, Obama is handling this pretty much like any politician: he focuses on problems that make headlines and that make him look good. Around 600 children under the age of five will be killed in America this year. They will not be making any headlines or creating photo-ops, so they won’t be a priority for the president,


%d bloggers like this: