« What do universities do? | Main | Limits of social mobility »

August 12, 2010


Luis Enrique

I'm a bit confused ...surely if you have a big welfare state, it's not that you don't need to "engage in big Keynesian counter-cyclical policy" it's that such a policy is built in ('automatic stabilizers')

here's a post about that:

Still, that doesn't say why the discretionary variant should entail a larger increase in debt. That scatter plot looks pretty reliant on the presence of Iceland and Ireland. I can't help suspecting that conditional on the right control variable, the relationship between size of welfare state and public debt might evaporate (but heck, you can say that about any correlation)


Err, in the graph, ireland and iceland are apart from the rest. If you drop them there's no correlation worth speaking of. So the question is: is there anything special about these two countries that means they can't be compared with the rest?

Wonder what it could be?


If you exclude Ireland and Iceland, the correlation between social spending and the change in the debt ratio falls from -0.41 to -0.17. Whether this sways you or not depends on the strength of your priors about the mechanisms I've described.
But should we exclude the two? I mean, were the easy money policies that encouraged boom and bust really unrelated to the weakness of their welfare states? Was it really a coincidence that the worst banking crises hit the economies with the weaker welfare states?
Also, even if we do exclude the two, the negative correlation between welfare spending and the level of debt in 2010 remains: -0.22 vs -0.27 if we include them.
I'll grant that the correlations become insignificant if we exclude Ireland and Iceland but include Korea.
I don't expect you to be swayed by the statistics. I include them only for suggestive value, to suggest that there might be something in Rajan's point.
Also, however you cut the data, we're a long way from significant positive correlations. The hypothesis "welfare states are responsible for government debt" has much less empirical support than the Rajan thesis.

Jimmy Hill

Could the negative correlation between the size of a welfare state and debt be down to the different attitudes towards tax in the respective countries?

Perhaps in countries with high welfare spending the populace is more relaxed about tax increases (indeed, this might be why they have a big welfare state). Whereas, in those countries with more stingy welfare states perhaps the populace are more negative about tax rises.

This would mean that politicians, when faced with a population sceptical of tax increases, have a big incentive to borrow rather than tax to facilitate more government spending.

Gerry J

This is just about the dumbest retionalization I've ever heard. You're advocating big government, nanny states, permanently high government spending to avoid big government spending for short periods.

Big government spending has never lead to prosperity and never will. The socialists are in command in Europe and have been for decades and they're in power in the US now. Obama's plan has failure written all over it. Obama is advocating more government spending when even the socialists in Europe are curtailing spending. And the fact that European Nations have spent less than the US in actual dollars (Euros or whatever) is not proof that it's worked.


Some evidence please Gerry J otherwise your silly generalisations merely make you sound like a teabagger.

Dain (Mupetblast)

Anyone read the book The Narcissism of Minor Differences? It suggests that the EU and US are more similar than dissimilar on social indicators, size of government, etc.


welfare of society needs fund, recession hits the state badly and the people also

David Friedman

It's an interesting argument, but you seem to be ignoring a more straightforward argument that would imply the opposite conclusion. A recession in a country with a large welfare state causes larger increase in government spending—for welfare—than it would in a country with a more modest welfare system, which increases debt more in the former than in the latter.

That doesn't, of course, explain your data, but you have already agreed that the data are only suggestive, since the results depend to a considerable degree on what countries you include.

The comments to this entry are closed.

blogs I like

Blog powered by Typepad