The question I propose to explore today is 'How big should government be?' I believe this topic will be one of the central issues in politics around the world as we go into the next century.

At the beginning of this century, governments in the advanced countries typically spent around 10 percent of national income and taxed their citizens accordingly. On this basis it could be said that the average worker spent one day a fortnight working for the government. In those days the church thought the tithe was quite a good idea.

This was an era of small or limited government, not just in the spending operations of governments but in their roles as regulators of economic activity and as owners of businesses and providers of services as well. New Zealand conformed with this pattern, and at the turn of this century average income levels in New Zealand were among the highest in the world.

By contrast, the twentieth century has increasingly been one of big government. As our century ends, the broad picture is that government spending accounts for around 20 percent of gross domestic product (GDP) in the Asian region, a third or more in the English-speaking economies and around 50 percent in Western Europe. In New Zealand, the average worker spends two days a week working for the government. We have also seen the vast expansion of the modern regulatory state and of central and local government ownership of many commercial activities. Only in the past two decades, with the collapse of the centrally planned economies and the worldwide moves towards privatisation, have these trends been partially reversed.

The debate over how big government should be has raged throughout this century and will continue. The 1960s and 1970s marked the high water mark of socialist and Keynesian ideas. Subsequently we have seen moves in the direction of economic liberalisation in all the OECD countries and many others, but in the economic domain and in other walks of life government continues to remain large or to grow.

I want to focus on two contributions to the debate about the size of government that have recently been made in New Zealand. The first is the work of Gerald Scully and the Inland Revenue Department exploring the size of government and the associated tax burden that maximises economic growth. The second is a speech by Simon Upton to a National Party conference last year which he entitled 'Focusing government on what it does best'. The conclusions of both contributions point in the same direction, but they reach them by somewhat different routes.

Professor Scully has surveyed data across a range of countries, including New Zealand, and has found that as tax burdens rise economic growth suffers. He concludes that the growth-maximising size of government is found with taxes around 20 percent of GDP.

I have a number of problems with the methodology of Professor Scully's study as I shall indicate, but its general findings are in line with much other economic research. For example, a study for the Joint Economic Committee of the US Congress earlier this year found that:

IMF research has come up with similar findings. The New Zealand Business Roundtable published a study in 1994 highlighting the deadweight costs of taxation in New Zealand. The important point which all these studies bring out is that government spending is the true tax burden. If the ratio of general government spending to GDP is, say, 40 percent, as is roughly the case today in New Zealand, the true average tax rate will be roughly 40 percent. If you want a lower tax rate, you must argue that the government spending ratio should be reduced.

Ted Sieper of the Australian National University has made a number of major criticisms of Gerald Scully's work, although he supports the view that smaller

government and lower taxes would improve eocnomic performance. My position is similar. I would have four general reservations about Professor Scully's approach.

First, his data cover the recent decades of high government spending and tax burdens. If he had surveyed earlier decades, he would doubtless have come up with a growth-maximising tax rate well below 20 percent.

Secondly, although several countries have achieved rapid growth with even smaller government, Professor Scully nowhere explains why additional government spending and transfers would add to growth. As Ted Sieper puts it:

Thirdly, the focus of the Scully study is on economic growth, whereas governments rightly have other objectives such as the provision of a social safety net and protection of the environment. Would-be supporters of smaller government will be alienated if such concerns are not taken into account. And even maximising growth is a limited objective compared with the broader concept of maximising welfare – some people will prefer to trade off work and income for leisure.

Finally, trying to determine the proper size and shape of government by looking for a 'growth-maximising tax rate' is a slightly odd and roundabout procedure. Moreover, the optimal size of government clearly depends on a country's circumstances at a given time. If I had been an Israeli at the time of the Yom Kippur war I might well have been happy to see half the country's income devoted to spending on defence. The role and size of government should be determined by a more fundamental analysis of what we need governments to do at a particular time and place.

Nevertheless, whatever the deficiencies of some individual studies, there is now a large and robust body of research pointing to the damaging effects of high levels of taxation on economic performance, particularly in a world of open borders and mobile capital. These were largely ignored in university teaching a generation ago and are still routinely denied by people such as the University of Auckland duo of Susan St John

and Tim Hazledine. But they are simply out of touch with modern economics. For example, a leading contemporary economic textbook, Principles of Economics by Gregory Mankiw, has a whole chapter on 'The Costs of Taxation' which brings out the pervasive effects of taxes on economic decisions.

Focusing on what the core role of government should be, however, is in my view a much more satisfactory way of thinking about the shape and size of government. This task is more firmly grounded in the principles of economics and public policy analysis. It is also the approach that has been followed in New Zealand during the reform period. The goal has not been 'minimalist' government or downsizing per se; it has been to find a proper basis for government intervention in the economy and to 'rightsize' government. Those in the reform camp have asserted a positive role for government – they have emphatically endorsed the need for governments to undertake the things that only governments can do, but have argued that governments should not undertake other activities that are best left to the private sector, including families and voluntary associations.

This is the approach taken by Simon Upton in the speech I referred to earlier. He gave a list of what he regarded as the essential functions of government. It comprised the basic machinery of democratic government, running the legal system including the courts and the police, producing statistical and other information, undertaking basic scientific research, protecting the environment and taking responsibility for foreign relations, immigration and defence.

Mr Upton's list is not definitive, as he acknowledged. One could add to it the role of maintaining a stable currency and keeping markets open to competition, for example. But the key point is that the essential functions all involve genuine public goods: services which cannot be provided adequately, if at all, simply through market transactions in the private sector. We want governments to ensure such public goods are provided and to do that job well. If they don't, we end up with a less efficient economy, and economic growth – and more broadly welfare – will not be maximised.

By contrast the job of providing private goods – those that can be bought and sold in the marketplace or organised through voluntary activities – should be left to the private sector. The list of public goods does not include things such as running Lotto, television stations, coal mines, ports, airports, electricity businesses, property companies, forestry operations, accident insurance schemes and a whole raft of other activities that central and local governments in New Zealand engage in. Because, on average, government ownership of such enterprises leads to inferior performance compared with the disciplines of private ownership, it harms economic growth. It should also be noted that with changing technologies, a number of goods and services such as roading which were once public goods are increasingly capable of being provided and charged for as private goods.

Simon Upton estimated that the core public goods functions which he identified accounted for around 5 percent of GDP. If we were only concerned about maximising economic growth, the role of government would essentially be confined to those public goods functions. That approach, rather than searching for a growth-maximising tax rate, would be the way to determine the optimal size of government. But governments rightly have other goals as well, which largely motivate their involvement in areas of social policy. Mr Upton went on to point out that he had not included in his basic list any social expenditure, not because he did not believe the government has a role in that area but because it is subject to much more debate.

What can be said about social expenditures? All of us want people to have access to services such as health and education, but we need to distinguish between means and ends. Mr Upton made two basic points. First, he noted there is no particular need for the government to be the provider of, say, hospital services any more than it needs to take over the role of private GPs and dentists: its duty is rather to underwrite access to a minimum standard of health care for all.

Secondly, it is very debatable whether it makes sense for the government to be the funder of services such as health and education for the majority of people who could afford to pay for them themselves in a lower tax environment. Such 'middle class welfare' merely involves collecting large amounts of revenue from taxpayers and churning it back to largely the same people through costly bureaucracies which are under little pressure to perform. At present levels of government spending it probably costs average income earners around $1.30 (having regard to the deadweight costs of taxation, including administration and compliance costs) to obtain health or education services that they could buy directly for one dollar. There is a strong argument for targeting taxpayer funding on those who really need help to obtain social services.

Social spending could also be reduced substantially by a new attack on all the obstacles to employment that have been created in recent years and by welfare reform initiatives of the kind that are being implemented in the United States and many other countries. It is paradoxical, to say the least, that although living standards have increased enormously since comprehensive state welfare programmes were introduced – so that, for example, 96 percent of New Zealand households today have a colour television and around 80 percent or more have a microwave, a fridge/freezer and a video – the share of national income devoted to welfare is at an all-time high. Most people would be able to cover most lifetime contingencies through savings and insurance given lower taxes. But even if we were to add back as much as a third of health, education and welfare expenditures to core government spending, those functions that need to be undertaken by the state would still only amount to 14–15 percent of GDP.

I am not suggesting that this is necessarily the right size of government or that we should move to such a level overnight. But there are good arguments for pressing ahead with privatising activities which governments do not need to be involved with as providers, and for moves away from universal funding of many social services. Even more importantly, the ratio of government spending to GDP can be brought down by policies that promote economic growth. This happens even without actual spending cuts if increases in government spending are kept below the growth rate of the economy. It was firm fiscal discipline of this kind that was largely responsible for the fall in the ratio of central government spending in New Zealand from over 40 percent of GDP in the early 1990s to its low point of around 35 percent.

A strategy along theses lines was advocated in the report Moving into the Fast Lane produced by the New Zealand Business Roundtable and the Auckland and Wellington chambers of commerce in early 1996. It outlined a comprehensive programme for reducing government spending to below 20 percent of GDP by 2005. Ruth Richardson has suggested a similar goal, which Jim Bolger dismissed in his recent book as unrealistic. Some politicians never learn. Mr Bolger dismissed a number of fiscal goals which the Business Roundtable put forward in the early 1990s as "too ambitious", yet key ones were achieved even earlier than we had proposed.

Since Ruth Richardson was dumped, subsequent Bolger administrations embarked on a spending programme of 'Think Big' proportions, now totalling some $7 billion. They also failed to press on with other initiatives to 'rightsize' government. Jenny Shipley has inherited this legacy and the collapsing rate of growth, current account blowout and rising unemployment that have come with it. The early plans formulated under the Fiscal Responsibility Act had the government spending to GDP ratio falling to around 30 percent this year. It is now back up to 36 percent and rising. Yet there are many who still haven't understood the disastrous results of such policies – allegedly less 'harsh' and more 'caring' – and are calling for yet more government spending.

By raising its expenditure to GDP ratio, New Zealand now appears to be set on a course which may be unique in the OECD. After almost a century of uninterrupted expansion, big government may be on the decline worldwide. The government share of national income in OECD countries peaked at 41.3 percent of GDP in 1993. This year it is forecast by the OECD to have dropped back to 38.6 percent and to continue falling. Over a slightly longer timescale, some of the falls have been dramatic. In Ireland the ratio has declined from 50.7 percent in 1986 to a projected 33.5 percent next year, largely due to rapid economic growth and fiscal discipline. This huge fall in Ireland's tax burden is unquestionably a major factor in its recent strong economic performance.

Outside the OECD the picture is little different. There have been few trends towards bigger government in Asia. The major exception is Japan, where the doubling of the government sector share in its economy over the past three decades – from 17.5 percent in 1960 to 36.9 percent in 1996 – is a major factor in its present difficulties which is seldom noted. In China, government spending appears to have come down from around 31 percent of the economy when Deng Xiaoping's reforms began in 1978 to around 10 percent today, or perhaps twice that figure if the spending of quasi-government entities is also counted. New Zealand may be competing for investment and trade with many small-government economies as it goes into the next century.

Some observers have noted the recent elections of centre-left governments in Western Europe and suggested that the trends towards economic liberalisation and smaller government may be reversed. But this interpretation may miss the point. Such policies have been pursued over the past 20 years by governments of all political persuasions: the government in Beijing is not usually described as right wing. The Economist recently tabulated the policy programmes of 13 centre-left governments in the European Union. It found that all but two of them were clearly committed to policies of privatisation and all but one to reductions in welfare. Certainly the Labour Party in Australia stood in the recent election on a platform that would have seen a return to protectionism, industry policies, labour market regulation and greater state ownership, but the electorate appears to have voted for ongoing economic reform. Even President Clinton has said that "the era of big government is over".

So where does all this leave us? My response would be that, as the US congressional study I quoted earlier put it:

New Zealand governments have moved far beyond any plausible definition of their core functions, and our economic performance has suffered accordingly. Unless the recent upsurge in government spending is reversed and spending and tax burdens resume a downward path, I believe New Zealand is unlikely to achieve strong growth and may stay mired in its current malaise. I think the countries that are most likely to prosper in the next century are those that understand the benefits of limited government.

No doubt there will be some ebbs and flows in the tide of future events. But at the end of a century of big government, we may be returning to the insights of some of the great thinkers on the subject of government who wrote at the dawn of the democratic era, and New Zealand might do well to heed them. One such was Adam Smith, whose study of the wealth of nations led him to conclude that:

Another was Thomas Jefferson, whom the American columnist George Will has called "the man of the millennium". In his first inaugural address, Jefferson asked: "what more" (after the industry of citizens and the blessings of Providence) "is necessary to make us a happy and prosperous people?" His answer was: