The Current State Of the Public Sector: an Economist’s View

The 11th Annual Public Sector Finance Forum. 10 September, 2007    

Keywords: Macroeconomics & Money; Regulation & Taxation; Statistics;   

It has been my lot to be asked to give two papers to this Public Sector Finance Forum. Today’s paper might be called the ‘macroeconomics’ paper, in which I look at the size of the public sector and its impact on the economy as a whole. I shall not pay much attention to the quality of the public sector expenditure which is the focus of the second paper, with its microeconomic focus.   

While in principle the papers cover separate topics, they are connected by a similar theme. Too frequently public discussion misuses financial data. I am not blaming those who created the data for this misuse, but it is vital that they understand how their technical constructs can be distorted into public ignorance and misleading political rhetoric. In my view, those who construct data have some responsibility to ensure there is an adequate public understanding of it – a position I hold for my professions of econometrics and social statistics.   

Even so, I recognise that there are forces outside our professions which lead to distorted use of the data we construct. Nevertheless we must try to resist them, and at the very least ensure there is an honest account available for the serious and diligent. .   

Today I am going to look at three macroeconomic issues.   

1. How big is the government in the economy?   

2. What are the implications for taxation?   

3. What is the net fiscal impact of the government on the economy?   

My short conclusion is that these are important questions to ask, but difficult to answer. Too often the answers proffered to the public involve the misuse of financial data for purposes of political exaggeration. I’ll explain too, how this exaggeration can lead to policy difficulties.   

How Big is the Government in the Economy?   

A frequent political debate in New Zealand is about the size of government. One side argues that the government spending is too large, that it is weakening economic performance, and that it should be reduced. The other side argues that there are serious social deficits, that they can be addressed by additional government spending, and it should therefore be increased.   

If asked how big is the government sector, my guess is most people would talk about a figure of 30 or 40 or even 50 percent of GDP. They would probably be referring to the estimates prepared by Treasury and reported in the fiscal outlook. For the record two figures are provided 

* Total Crown Expenses are expected to be $67.9b for the year to June 2008, or 41.2% of GDP; 

* Core Crown Expenses are expected to be $52.8b for the year to June 2008, 32.0% of GDP.   

That two figures are provided are an indication that we are in the murky world of definitions which matter. The difference between the two is that the Total Crown Expenses includes spending by Crown Entities and State Owned Enterprises. This means very different types of expenditure get added together. For example, along with ordinary state spending such as by hospitals and schools, total spending includes such things as expenditure by the lotteries commission on winnings and charities, and on the running state owned power companies. Note that neither figure includes Local Government except where it is funded by the Crown. The DHBs are in, but rates funded expenses are out.   

Are we comparing like with like when the expenses are a proportion of GDP? After all the denominator is a measure of production, while the numerator is a measure of outlays, quite conceptually different notions. A better denominator might be GNE, Gross National Expenditure, which currently exceeds GDP because the country is borrowing to fund its investment and even some of its consumption. Perhaps better statements would be 

* Total Crown Expenses are expected to be 38.0% of GNE, for the year to June 2008; 

* Core Crown Expenses are expected to be 29.5% of GDP, for the year to June 2008.   

But even that does not see quite right since some of the expenses of government are actually spending by individual. For instance a pensioner using New Zealand Superannuation to purchase bread appears as a government expense on these measures. If we subtract out subsidies and transfers we get 

* Total Crown Expenses (excluding transfers and subsidies) are expected to be $48.4b for the year to June 2008, or 27.1% of GNE; 

* Core Crown Expenses (excluding transfers and subsidies) are expected to be $35.2b for the year to June 2008, 21.3% of GDP. {Exact figures are not published. These are estimates.]   

It also makes sense to separate out publicly provided services to individuals from collectively provided services. For instance, if you get treated at a public hospital, that appears in Crown Expenses even though you are the only direct beneficiary. That is quite different from military spending, where the benefits are to the community collectively, and not to separate individuals.   

How much is this individual directed spending? We dont know, but if we assume that 80 percent of public spending on health, education is on individuals, we get a figure close to half of all core government spending for this purpose, so again the proportion looks smaller. 

To give some final indication of the complexities of the measure I turn to another data source, the National Accounts, which use different definitions, but which include local government in their totals. They are currently available only up to the March 2007 year.   

In the 2007 March year, General Government Final Consumption Expenditure (that is including all levels of government spending for both collective and individual purposes, but excluding transfers and the activities of state owned enterprises) amounted to 18.6 percent of GNE. To that you might want to add a further 3.0 to 3.5 percent for government capital formation, including by its producer enterprises. In total, then we are talking of the total government use about 22 percent of national spending.   

Another measure available in the National Accounts is the size of the government administration and defence sector, which treats the government as a production sector, and so may be properly compared to GDP. It is only available in 1995/6 prices, but in the most recent year the government production sector contributed only 4.5 percent to GDP. What happened to the rest of it? Some government spending is in other sectors. Education and health services are in Personal and Community Services, which is a much bigger sector. And the government buys inputs from the private sector – from farmers, manufacturers and so on – which appear in its spending but not in its production.   

Had I the time, I could deluge you with further measures – I have not touched on the employment ones for instance. But I hope the message is clear. Those who talk of the government taking over half the economy – or even a third – are misleading themselves and their audiences. The most likely figure is between a fifth and a quarter of the economy.   

Moreover, when we look at the content of that spending, we conclude that the actual government administration is a small proportion of that total – probably about 5 percent of total production, certainly much less than 10 percent. It is true that there is administration associated with the remaining spending, but that administration occurs when the product is delivered by the private sector too. It is said that 25 percent of the American health budget – one of the most private driven health systems in the rich world – is spent on its administration.   

One question you might want to ask is how does the proportion compare with other countries. Say other OECD countries, because they are more like us and the statistical base for comparisons is better.   

The latest OECD data is available for 2004. Ranking by general government spending – roughly comparable to Total Crown Expenses but including local government – as a proportion of GDP, the OECD reports New Zealand as 23rd of its 30 members. We have the same ranking by General Government Consumption Expenditure: 23rd out of 30. (Ranked by Social Security transfers, which is one of the main differences between the two, New Zealand is 22nd out of 30.)   

(The remainder of the paper uses GDP as the denominator as a scaling device.)   

International comparisons are treacherous. New Zealand boosted itself in the rankings by grossing up New Zealand Superannuation and social security benefits to before tax incomes and then treating them as taxable for income purposes. That means that money which never leaves the government accounts appears in its spending measure. Moreover government expenses include the full cost of civil servants pensions accruing as part of the personnel costs, whereas most countries only report that cost when it is paid out to pensioned off staff.   

There may be good reasons for such practices, but the point here is since other countries do not do it, international comparisons are distorted. However, we may have accounting practices which lower our measured spending compared to other countries.   

But in summary, I think it safe to conclude is that the New Zealand government is not a big spender by international standards. Then why do some people go on about the size of government spending, often making exaggerated claims as to its size and significance?   

What are the Taxation Implications?   

One answer is that they consider that New Zealanders pay too much tax. In the end taxation covers government expenses – I’ll talk more about this equation in the last third of this paper. Are New Zealanders highly taxed?    I’ll not go through the sort of calculations I have just done for government spending, but let me remind you definitions matter. For instance we could markedly reduce the apparent level of taxation by defining benefits and the like as negative taxes. A smaller reduction would be to drop GST from government spending, and not to gross up transfers.   

On existing definitions the OECD ranks New Zealand as 20th out of 30 in terms of total government revenue, which includes non-tax revenue sources and local authority rates. Remember too, the effect of grossing up transfers. Thus it is hard to argue that New Zealand is particularly highly taxed in comparison to other rich countries.   

Even so, there has been extraordinary effort to demonstrate that New Zealanders are highly taxed and that the economy would perform better were tax rates lower, often using almost fraudulent statistical procedures. The alacrity with which some advocates seize upon incompetent analysis to justify their views suggests a debate more driven by enthusiasm than competence.   

Despite the vigorous lobby which argues that the levels of taxation are too high and are inhibiting economic growth the evidence is more ambivalent. Certainly there is research which concludes that high taxes affect economic performance, but there is also research which concludes it does not. The pragmatist is likely to conclude that below certain limits the impact of tax levels on the economy, if any, is of uncertain sign (it may be positive, it may be negative) but in any case it is very small. What that maximum level is, is unclear. One cannot but observe high tax European countries which are among the most prosperous in the world. By their standards New Zealand is not a high tax country so it seems reasonable to conclude it is well below the threshold where taxation damages economic performance.   

Moreover, the efficiency of the tax system matters. The average tax rate measure takes no account of the design of the tax regime, treating the clumsy Muldoon personal income tax system as equally effective as the elegant Douglas one.    Nor does the research evaluate whether it matters what the tax revenue is used for. The public may want the collective services supplied or the different pattern of output that the pattern of taxation and spending may generate. They may want more education, more environment, more culture and heritage, more preventative health care. Some public expenditure may boost economic performance, some may add to social coherence. It may be spent efficiently or wastefully. Nor should we forget that some government spending displaces private expenditure: in the case of health services. public funding may be more efficient that compulsory private insurances.   

At this point we are past the grand rhetoric of more or less government spending, and the higher or lower taxation which goes with it. Instead we are looking at individual spending programs and asking whether it is worth more than the tax which has to raised to fund it. While there is some technical elements to any such judgements, in the end they are political ones. A group of politician has to say we want that public expenditure but not that public expenditure and that will involve raising taxation to this level but not beyond it. That is a practical exercise which, in my opinion, our cabinet system does pretty well.   

It ends up with about half the population grumbling they want more spending and the other half arguing they want lower taxation (or, very often, more than half wanting both of these). The government cannot win, particularly when an opposition political party announces it supports lower taxation and greater government spending, sometimes on the same day.   

The rhetoric is really about the political judgement of the benefits of higher or lower taxation and the balance of public and private spending. Each of us entitled to an opinion: economists’ opinions are no more valuable than any other persons, even if some try to hide their political prejudices behind bad technical economics. The economist’s task is to insist there is a tradeoff – there is no free lunch – while the braver of us may press those who want to cut taxation by asking them to identify which public spending they want to cut, and equally to press those who want to increase public spending by asking what taxes they want to increase.   

There may seem to be an easy option of claiming that tax cuts come from efficiency gains. In tomorrow’s paper I shall argue that obtaining such gains in government is harder than the rhetoric of opposition. The other easy option is to claim that one can have tax reductions and government spending by government borrowing. It is this topic which makes up the last third of the paper.   

The Government Deficit, and the Impact of the Government on the Economy   

Let me begin by making the simple point that a government deficit is a tax on future generations – it’s a timing effect. The ability to borrow to increase expenditure over revenue does not break the connection between tax revenue and government spending. The connection is now over a longer period since debt has to be serviced and repaid so there is less available for spending in the long run.   

It is lesson we should have learned from the 1980s, when the government borrowed heavily to provide high public spending while revenue was weak as top income tax rates were lowered and the economy stagnated. In the 1990s, tax rates had to be held high while we paid for the profligacy. Often the balance was at the expenses of social spending incurring a social deficit. Much of today’s government spending is dealing with that social deficit and the infrastructural backlog from the 1970s and 1980s when governments balanced its books by cutting back on necessary capital investment.   

Perhaps I am passionate about such matters, beyond the technical analysis I usually offer. That is because I judge it immoral to solve our immediate problems by loading debt onto our children and grandchildren.    I

t is not a question of left-right politics in the way that spending-taxation often is. You will recall that many in the caucus of the first Labour Government thought there was no government budget constraint and that additional spending could be financed by Reserve Bank credit. Today there remain people on the left who hold similar views. On the right both the Ronald Reagan and George Bush junior governments generated enormous government deficits, but George Bush senior, as well as Bill Clinton, took measures to restrain and eliminate the deficits.   

On the other hand, New Zealand has been marvellously lucky that it has not had the fiscal irresponsibility of Bush junior and Reagan over the three-quarters of century since it had a Reserve Bank. There have been some close runs. I have already mentioned the social credit wing in the Labour caucus in the 1930s, while the fiscal rhetoric of Norman Kirk was not very disciplined although he offset that with the doughty Bill Rowling as his Minister of Finance.   

The sheer quality of our Ministers of Finance has been one of the highlights of New Zealand’s political record. Fortunately they have been able to resist the spending predations of other cabinet ministers and parliament, in part because they have been backed by their prime ministers.   

The major exception was Rob Muldoon who knew the importance of fiscal discipline, but got away with offsetting large deficits by inflation reducing the value of government debt (reminding us that inflation is a form of taxation).    Muldoon had already been fiscally compromised, when he promised to make National Superannuation, as it was then, markedly more generous. He said he could do it at a cost of .2 percent of GDP. In fact the fiscal cost was closer to 2.5 percent of GDP. Basically Muldoon confused a couple of incomparable numbers, thinking that the immediate impact of his scheme could be compared by the full impact of the Labour scheme fifty years later. The resulting surge in public spending without an offsetting nominal tax hike had the economy struggling through the late 1970s and beyond. It was paid by fiscal drag, as taxes rose faster than inflation, and yet the fiscal drag fuelled inflation as income groups fought to maintain their income relativities. That is why inflation was high and endemic under Muldoon.   

An almost as disastrous election promise was made in 2005, when the National opposition promised substantial income tax cuts together with increases in some government spending. The boost to national expenditure would have been over 2 percent of GDP, comparable to the Muldoon promise.   

Suppose National had been elected to office in 2005 and implemented its promise. It had no room to make substantial spending cuts, although no doubt some programs would have suffered. I note too that it may have hoped to shift some of the investment spending off-balance sheet by pubic-private-initiatives. But whatever the merits they may have in improved efficiency, they do nothing to reduce the fiscal impact of the investment.   

The promised fiscal inject would have had a dramatic impact. Assuming that the Reserve Bank would have resisted the inflationary pressures, interest rates would have had to be even higher than they are today, as would the exchange rate. Many people’s interest bill would have risen by more than their tax cut. Imports would have boomed. Exporters would have been struggling even more. Once the current international financial turmoil commenced our financial system would have been in far greater trouble than we are today.   

Over the last 15 years our Ministers of Finance – Ruth Richardson, Bill Birch, Winston Peters, Bill English and Michael Cullen – have been assiduously building a strong government balance sheet in preparation for such an external shock. We cannot avoid the consequences of international financial turbulence but we can prepare for it. No other rich economy may have as robust a public balance sheet although that could not be said for the New Zealand private sector, which is carrying too much debt. Implementation of the Opposition 2005 election promises would have reduced that robustness, and compromised the economy. We were exceptionally lucky the policies were not implemented, given today’s international monetary turmoil.   

The point of raising the topic today is to draw attention to the fact the policy arose from a misrepresentation of an item in the Crown Accounts. At the time this aggregate was called OBERAC – operating balance excluding revaluation and accounting policy changes – but following the revised GAAP standards introduced in July, the item is replaced by OBEGAL – operating balance before gains and losses.   

In election year 2005 Operating Balance out-turned at $11.5 b (or 7.3 percent of GDP). (In that year the OBERAC was $8,486m or 5.4% of GDP.) Being seen as a surplus it was widely thought it indicated that there was considerable opportunity for major tax cuts. The public rhetoric focusses upon this particular amount, although it is clear that the vast majority who quote it do not have the slightest idea what the aggregate means. They are unaware the extent to which most of the OBERAC/OBEGAL surplus is subsequently used for public investment including in State Owned Enterprises and hospitals, roads, and schools, on funding student loans and the New Zealand Superannuation Fund. In other words, the money that OBEGAL represents is not available for tax cuts, because it is already spent. Nevertheless, journalists and commentators persisted in the belief that the OBEGAL surplus justified income tax cuts.   

This was nicely illustrated earlier this year when a week before his budget speech, Minister Cullen, gave his regular speech to the Canterbury Manufacturers Association, setting out his budget analysis. Cullen scrupulously kept to discussing the cash surplus, roughly the amount left after all the additional spending items are deducted from OBEGAL. His message was that the fiscal position was deteriorating and there was little room for increases in spending or income tax cuts. (There is expected to be a cash deficit of about a billion dollars this year, not a surplus.) One journalist accurately reported what Cullen said, but could not resist adding his own interpretation which involved leaping from the cash position to the OBEGAL surplus and demanding income tax cuts.   

Clearly the journalist did not understand what he was writing about – which may be no surprise. I am frequently struck by the sheer stupidity of the public rhetoric. Does anyone seriously think that had there been room for a tax cut in 2005, the incumbent Minister of Finance would not have given it? Cullen is one of the cleverest and most politically astute Ministers of Finance we have had, and generally they are well above the average of their parliamentary colleagues and journalists. Had he the opportunity, would he have forgone income tax cuts in election year? If you think ‘yes’, you must assume he is as stupid as those who think OBEGAL is a valid measure of the government’s ability to cut taxes.   

OBERGAL is but one of a number of indicators of the government’s fiscal position. Highlighting it to justify tax cuts is irresponsible and incompetent – which does not mean that people wont do it.   

Unfortunately there is no single measure which unambiguously indicates how big a tax cut can be justified. Even retrospectively there are a variety of indicators including those in the 2002 Treasury Working paper on the topic, Indicators of Fiscal Impulse for New Zealand by Renee Phillip and John Janssen. A wider set of fiscal indicators is reported in the Budget documents in a section titled Indicators of the Government’s Fiscal Performance (pages 109 to 114 of the 2007 Budget)”   

The section reports at least nine indicators. If you wanted a single measure I would probably use the prospective cash position. It is expected to be negative for the next few years – in a certain sense the government is a net borrower, although the crown debt to GDP ratio is not expected to rise. My sense is that the fiscal position is at the top of the acceptably prudent range although I might go looser if the world economy turns to custard. That is a matter for a thoughtful discussion. The one thing you can be sure is that the OBEGAL measure will hardly appear in it.    So we have here an example of how a technical measure designed for a particular purpose has transmuted into a popular measure with a quite different meaning, and how that transmutation might have led to policies which could have wrecked the economy. And that is why those who constructed the measure should vigorously participate in the public debate to ensure that users understand what it is they are using – even journalists and politicians.   

Let me finish this section by saying I have been tough on the National Party’s 2005 fiscal policy, for the simple reason that there would have been an economic disaster had it been applied. We are very fortunate it was not implemented. My reading of the National Party’s current fiscal intentions, is that they are prudent, intelligent and common sense – as we would expect from the current Opposition spokesman on finance, Bill English.   

The Gross Debt Target 

I want to finish with one further illustration of how a measure can inadvertently lead to sub-optimal policy. The government currently has a gross debt target, which restricts its ability to borrow in one currency and invest in another since that would increase gross debt even though net debt would be the same. Many would argue that we do not want the government speculating in financial markets. I agree. However the government has a large financial portfolio and it needs to manage it to minimise the cost of Crown borrowing.   

As it happens, without the gross debt target, the government could reduce its borrowing costs by about 60 basis in points, without risk, by borrowing in New Zealand currency, investing the receipts in a foreign currency, and purchasing foreign exchange cover so there is no currency risk. In essence the Crown would be offsetting the carry trade, using its advantage as a sovereign borrower to reduce its costs. One of the consequences of leaving this financial market anomaly is that it has made monetary policy more tense than has been necessary. Essentially the Reserve Bank has had to push harder on the short term interest rate (the Official Cash Rate), in order to attempt to reduce liquidity in the mid term range in which the anomaly occurs.   

Thus the effect of the government’s gross debt target is to raise the cost of Crown borrowing, while putting additional pressure on interests rates and probably the exchange rate. Once more we have an example of how a measure designed for one purpose can lead to inferior policies if it is misused.   


To conclude then the technicians who create and evaluate such measures have a responsibility to enure the public understands them. That is the theme of this paper.   

<>Go to top