Size does matter - at least to the IMF

Economics

Just in case you hadn't noticed, on the authority of the International Monetary Fund at least, the UK is a "small government" country.

Surprised? Actually, that is a bit of a cheat. IMF staff have been studying* the relationship between public spending and economic and social performance and for the purposes of the exercise they divided countries into three groups: those where public spending was more than 50 per cent of GDP in 1990, those where it was between 40 and 50 per cent and those where it was less than 40 per cent. The big government group include the Nordic countries, plus a few European ones like Belgium and the Netherlands; the medium government group were most of the rest of Europe; and the small comprised the UK, the US, Japan, Switzerland and Australia. But we only squeaked in at just under 40 per cent, while Japan and the US were way down in the 30s, and we are now back above 40 per cent again.

Still, even when allowing for the apparent crudity of the classification, it is an interesting exercise for anyone who genuinely wants to see how high public spending might affect both economic and social performance, for it brings out some rather surprising results.

You might suppose that big government countries had rather worse economic performance but scored rather better on social indicators. But this is not so. Big government countries do score a little worse on some economic indicators, but by no means all. On the other hand they do not do much better on social indicators: in fact in quite a lot of them come out far worse.

Some examples. On economic growth, both small and medium government countries do rather better than large, but small do not do any better than medium. On investment there is nothing in it: all are the same. On inflation, small governments come out worst, the opposite of what one might expect. There are, however, three measures on which small government countries are clearly best: unemployment, the number of patents (a measure of inventiveness) and unsurprisingly, the size of the black economy. On that last point, the estimated size of the black economy for the big government countries is more than 11 per cent, against a still-large six per cent for the small ones.

On the social side, the real surprise is how similar the various countries are. All three categories have similar life-expectancy, similar infant mortality, and similar divorce rates. Inequality of income, as measured by the share of income of the lowest 40 per cent, is a little higher in the small government countries, but the difference is not enormous. Only in the size of prison population do the small government countries come out worse, and that result is skewed by the US - take that out and they come out the same as the medium-sized countries and not that much worse than the big government ones.

The authors of this study also look at the historic pattern of the growth in government, and at the social performance of the more advanced newly- industrialised countries such as Singapore and South Korea, both of which have small governments but nevertheless achieve high scores on social indicators.

Their broad conclusion is that since most of the progress in social welfare in the industrial world took place before 1960, when governments were much smaller, and the NICs score well despite smaller government, there is not much of a case on either social or economic grounds for public spending higher than about 30 per cent of GDP.

Their further conclusion is that there should be considerable scope for cutting public spending in the developed countries without compromising social welfare. And they suggest that this needs a combination of a well- working private market and appropriate regulation by government.

Doubtless the government tax and spending take could be reduced in this way, but naturally there would be enormous political resistance to any such programme. The graph on the left shows what has happened to public spending in a number of countries between 1980 and 1994 and, as you can see, it has risen in almost all of them. The only countries in which there was a fall were Belgium, the Netherlands, Ireland, the UK (just!) and New Zealand.

So it can be done. But does it need to be done? If people want to have a large proportion of a country's output handled through the state, vote in governments which advocate such a policy and pay the necessary taxation, why should they not do so?

There seem to me to be three powerful reasons why at least some shading- down of the size of the public sector has to take place. The first is that there is some economic penalty carried by countries with big governments. The differences in growth may not be large, but they do exist, and over the years this will result in large differences in wealth. Singapore is now richer than Sweden. Further, there does seem to be a link between large government and high unemployment, driven presumably by the size of the tax wedge between the employer and the employee.

Second, the larger the government the larger the black economy. This may not matter in purely economic terms, but it is surely alarming in social terms. It is profoundly corrupting for one-tenth or more of an economy to take place outside the law, and I suspect (though I cannot prove it) that this corruption has a knock-on effect on people's behaviour in matters other than taxation.

Third, present spending levels are simply not sustainable because in practice, whatever they say, voters will not provide the revenue. The graph on the right comes from the latest IMF World Economic Outlook, just out last week. As you can see until the mid-1970s (actually until the oil shock of 1973-74) taxation and public spending in the main industrial countries broadly kept pace with each other. Since then, taxation has fallen hopelessly behind. The result of these deficits is to transfer the burden of taxation to future generations, a transfer which is even larger when you add in the unfunded pension liabilities of these countries' governments, and the rise in the dependency ratio which will take place as populations age.

If all this sounds rather glum there is some good news too. The authors of the new World Economic Outlook note that countries that do cut their deficits, particularly those that do so by cutting public spending rather than increasing taxes, often enjoy a faster rate of growth. Naturally faster growth then boosts tax revenues and makes the whole adjustment considerably less painful.

The left in Britain, as elsewhere in the world, is pretty suspicious of IMF remedies, and the idea that it is possible to cut government spending and still achieve the social objectives desired by most people, not just those on the left, may appear unrealistic. There is certainly no evidence from that graph on the right of any downward trend in public spending, even if the rise in revenues does seem to have tailed off.

But the countries towards the bottom of the table on the left, plus the East Asian tigers, outperform the countries at the top of the table, then expect small government to become much more fashionable. The trick for governments of the left will be to find different ways of achieving their social objectives without having to increase substantially the size of the state machine - as new Labour is doubtless very well aware.

* `The Growth of Government and the Reform of the State in Industrial Countries', Vito Tanzi and Ludger Schuknecht, IMF Working Paper.