But whatever view we take on the appropriate size of the state, we tend to see the issue through the prism of its role in a developed country. That is understandable enough - it is a central issue in our political debate, as we will doubtless see again on Wednesday - but what governments in developed countries do is only one half of a global story. The other half, affecting a much larger number of people, is what is happening to the role of the state in the developing world.
So the choice of the World Bank to use its annual World Development Report to review the state's position is particularly welcome. The size of the state in developing countries tends to be smaller as a percentage of GDP than in developed ones. But whereas in the developed world the quality of public sector management is reasonably homogeneous - the quality of service from a relatively well-run government is not that much better than the quality from a relatively badly-run one - in the rest of the world the quality is much more uneven. Governments in the really badly- run countries are dreadful, while governments in the well-run ones are excellent. In a way, therefore, the quality of government in a developing country matters even more than in a developed one.
Almost everywhere governments have been tending to get bigger, as the graph on the left shows. There is some evidence that a turning point, or at least a plateau, may now have been reached, but not a lot. There are a few developed countries where government is responsible for a smaller proportion of GDP than in 1980 - the UK, Ireland, the Netherlands, Belgium and New Zealand; as you can see from the graph there was some decline between 1985 and 1990 in the overall size of government in developing countries. And I suppose we should also include the shrinking of the state that has taken place in the former Soviet bloc. But in six out of the Group of Seven countries the state is now larger than it was a decade and a half ago.
The graph on the right shows another aspect of the rise of the state: that spending in the G7 countries has run far ahead of the ability of the tax system to cover it. Until the first oil shock the two ran pretty much in line, but since the mid-1970s taxes have lagged. Indeed tax revenues now are only a fraction higher as a proportion of GDP than they were a decade ago. One of the main reasons, maybe the main reason, why there is enormous pressure on governments to hold down spending is the inability to raise revenue any further.
So not larger government, but more effective government - what does this mean? The World Bank has some helpful thoughts here. It makes a distinction between the state's role and its capability; put crudely, the first is what it does and the second is how well it does it. It may seem elementary but it is important, because if the state is not very capable, it should do as little as possible, for when it does intervene it is liable to do so badly. The bank identifies five core responsibilities: establishing a legal framework, running policies which do not distort the economy, building basic services and infrastructure, protecting the vulnerable and protecting the environment.
Few would quarrel with that list. Yet when the bank looked at the actual quality of government, it concluded that a large number of countries were failing in these tasks. For example the "lawlessness syndrome" was evident in many countries, with the former Soviet Union ranking at the bottom of the "credibility index", just below sub-Saharan Africa.
Once one goes beyond these basics, the World Bank argues that the state need not be the sole provider. Once, people tended to think that the state should be the main provider of social security, but now even developed countries are starting to reform their welfare systems. Countries which have not been able to afford them in the past will never now be able to do so, for they are being hit by similar problems of ageing populations. So middle-income countries which are now looking at ways of improving social welfare systems will have to find ways of providing a lower-cost service - though they do not have to dismantle systems they can no longer afford.
In other areas there are models: for example in managing privatisation, there are successful and less successful ways of carrying through the policy; there are good and bad examples of regulation; and effective and ineffective examples of industrial policy. The key is to know the state's limits: capable states can be given more leeway in policy-making; and though the World Bank does not use this sort of blunt language, the ones that make a mess ought to try and do less.
How do you encourage the state to improve its service? The World Bank has a couple of useful approaches. One is to seek ways of invigorating state institutions - for example by establishing rules for and restraints on civil servants, increasing competitive pressure on government bodies, and increasing citizen participation.
All this will seem familiar to anyone living in Britain: there is continuing clear pressure on the state to lift its game, for we now get upset if the quality of service from a government department is noticeably worse than it would be from a private sector provider. But in some parts of the world the link between perceived shortcomings and a response to fix them is very weak. This raises the tough question: why do some countries find a mechanism to improve the quality of government, while others fail to do so?
As with most tough questions there seems to be no clear answer. Some countries carry on with dreadful governments for decades, while others manage to reform. The World Bank notes, I think rightly, the influence of culture and of example. Some countries, for example in east Asia, will, for cultural reasons, not tolerate high inflation, while others, particularly in Latin America, will do so. It also notes that if reforms are to be sustained, losers have to be compensated: there is often a problem when the winners are numerous but the gains small, while the losers, though few, face large losses.
There is a further problem when gains are in the future and losses in the present. Part of the task of politics is to explain reforms and persuade people of their value. But this is tough when politicians are not themselves trusted, or when the benefits of reforms are far into the future and the costs are immediately evident.
The key point that the World Bank makes is that there are no easy solutions, no quick fixes. What surely is helpful is to foster a discussion about governments in a practical way, rather than an ideological one. Consider this. We still accept that the state has a monopoly over a large range of services: that it should be normal, for example, that we in Britain should have to "buy" our national insurance from the UK government, rather than being able to buy the service from, say, the Dutch one, or from a private sector alternative. But why? Why not think of governments as competing for excellence of service, just like everyone else? As the World Bank says: "Good government is not a luxury - it is a vital necessity for development."
"The State in a Changing World - World Development Report 1997", World Bank/OUP.