THURSDAY 23 SEPTEMBER 2004
CONFERENZ 8TH ANNUAL PUBLIC SECTOR FINANCE FORUM
The Fiscal Responsibility Act: A Stocktake
ROGER KERR
EXECUTIVE DIRECTOR WELLINGTON
NEW ZEALAND BUSINESS ROUNDTABLE 23 SEPTEMBER 2004
THE FISCAL RESPONSIBILITY ACT: A STOCKTAKE
Introduction
This year marks the tenth anniversary of the passage of the Fiscal Responsibility Act 1994 (FRA). A bill currently
before parliament, the Public Finance (State Sector Management) Bill, will both repeal the FRA and integrate its
provisions into the Public Finance Act 1989. These amendments will be accompanied by changes that are intended to make
the Fiscal Strategy Report (which accompanies each budget) the main vehicle for setting out high-level fiscal strategy.
The focus of the Budget Policy Statement is to be narrowed so that it is more on the priorities for the coming budget
and less on high-level fiscal strategy issues.
Ten years is a reasonable period of time to permit an assessment of the strengths and weaknesses of the FRA, and to
consider possible improvements. Next month the Business Roundtable will be releasing a major report on experience under
the FRA by Dr Bryce Wilkinson. The report has four main sections:
- a review of fiscal policy and outcomes over the past decade and in a longer-term perspective;
- a discussion of the relationship between the size and quality of government spending and economic growth;
- the case for introducing explicit tax and spending limitations into the FRA; and
- other options for constraining government spending.
This paper adopts a similar structure.
Fiscal policy and performance under the FRA
Fiscal policy is an important part of government operations. With government spending at all levels (central plus local)
amounting to around 40 percent of gross domestic product (GDP), it has a major impact on the economy. The taxation
needed to finance spending programmes has an independent impact, determined in part by the efficiency of the tax
structure. Operating deficits and surpluses have additional effects, partly through the channel of monetary policy and
interest rates. In the absence of significant asset sales or other balance sheet changes, the cumulative tally of
deficits and surpluses largely accounts for the stock of public debt and the level of Crown net worth, which have
intergenerational implications and influence the country's credit rating.
The quality of fiscal policy deteriorated during the 1970s and "80s. Government spending mushroomed; the tax system
became distorted and inefficient with, among other things, excessive reliance on income tax and a punitive top tax rate
of 66 percent; and governments ran large deficits which resulted in a heavy debt burden and credit rating downgrades.
The parlous state of the government's books was a factor in the economic crisis triggered by the 1984 election. The
incoming Labour government did much to improve fiscal policy, especially through sound tax reforms, but it struggled to
achieve fiscal discipline and bequeathed its successor a deficit problem that necessitated the spending measures of the
1991 budget.
In response to this haphazard record, the architects of the FRA, notably finance minister Ruth Richardson, sought to
bring a greater focus in annual budgets to issues of fiscal prudence and longer-term strategy. In submissions on recent
Budget Policy Statements, the Business Roundtable has assessed the record as follows:
- The focus on fiscal prudence has paid dividends in the form of sustained operating surpluses, improvements in Crown
net worth and, most recently, an AAA credit rating for Crown foreign currency debt by one rating agency.
- The FRA has not been successful in constraining government spending and the aggregate tax burden. Governments in the
1990s failed to achieve their long-term objective of reducing spending below 30 percent of GDP and subsequent
governments have lifted the target to 35 percent. The OECD has observed that the contribution of the FRA to expenditure
restraint has been limited.
- Nor has the FRA imposed any meaningful discipline on the quality of government spending.
- Greater transparency of the public finances has been achieved under the FRA. The pre-election "opening of the books'
has avoided nasty surprises and ongoing trends are highlighted. The system of improving accountability by requiring the
minister of finance and the secretary to the Treasury to sign a statement of responsibility appears to have worked well.
- Contrary to expectations, the FRA has not led to meaningful debate about the value for money of government spending
and the link between budget policy and economic growth. In particular, there has been no dialogue with governments about
(1) the gap between their targets for economic growth and forward projections; (2) the contribution of lower tax rates,
privatisation and deregulation to economic growth; and (3) the need for systematic reviews of the quality of government
spending.
The proposed changes to the FRA do not address these concerns.
Government spending and economic growth
The government has portrayed itself as a prudent economic manager. However, this claim is only plausible with respect to
its record of maintaining operating surpluses and ongoing debt reductions. Its spending and taxing record has been
profligate. The decision to lift the long-term objective for spending under the FRA to 35 percent of GDP effectively
sought to appropriate another twentieth of national income for the public sector. In the 2004 Budget the government
announced additional spending plans totalling a massive $14 billion over the next three years, mostly in forms that
would be classified as unproductive. The tax take has risen by nearly 40 percent since the government came to office.
Most of the increase is due to a period of solid economic growth. It is difficult to avoid the impression that the
government is spending the "growth dividend' simply because it is available, and is unable to entertain the idea of
giving surplus revenue back to taxpayers.
The problem with this strategy is that it is inconsistent with the government's stated top priority of increasing the
economy's growth rate. No Organisation for Economic Cooperation and Development (OECD) member country has sustained
growth in real per capita GDP of 4 percent a year or more - the growth rate the government is targeting - with total
government spending equal to 40 percent of GDP.
No competent analyst would argue that smaller and better government is a sufficient condition for faster growth. The
modern economic growth literature suggests that openness to trade and investment, respect for property rights and the
rule of law, flexible labour markets, lighthanded regulation of the productive sector and relatively non-distorting
taxes are also prerequisites. But on the basis of OECD experience, a lower government spending ratio is a necessary
condition for sustained economic growth of 4 percent per annum or more.
There have been some attempts to dispute this proposition.
A point made by some critics is that New Zealand's spending ratio is not among the highest in the OECD - indeed it is
below the average. The point remains, however, that at around 40 percent it is too high for fast growth. To repeat, none
of the OECD countries with government spending at or above that ratio has sustained 4 percent per annum per capita
growth. It is implausible to suggest New Zealand could be an exception.
Another response is that a number of high income countries have high levels of government spending, suggesting that the
two are not incompatible. However, countries like Germany became wealthy before their governments became big - after
World War II Germany was known as the free market economic miracle. The better performance of Ireland - formerly with a
government spending ratio of over 50 percent - has been related to large reductions in the share of government spending
in the economy and lower taxes. Currently the large welfare states of the European Union are falling behind the growth
performance of the freer Anglo countries such as the United States, the United Kingdom and Australia.
A third response is that the quality of government spending is more important than the overall size. The secretary of
the Treasury, John Whitehead, expressed this view in a speech earlier this year.
These are false alternatives. Quality matters but so does size. Few analysts would argue that economies with government
spending ratios of 60 or 80 percent are capable of fast growth. There is a strong presumption that much of this spending
would be wasteful and that the deadweight costs of taxation - which rise roughly with the square of the marginal tax
rate - would be enormous. As an empirical matter, a 40 percent ratio appears to limit growth to below the government's
target range, contrary to John Whitehead's open-ended assertion. A more mainstream view on this point has been expressed
by the OECD:
The main conclusion from the literature is that there may be both a "size" effect of government intervention as well as
specific effects stemming from the financing and composition of public expenditure. At a low level, the productive
effects of public spending are likely to exceed the social costs of raising funds. However, government expenditure and
the required taxes may reach levels where the negative effects on efficiency and hence growth start dominating.
Some level of government spending is obviously necessary. There is a need for government involvement in the provision of
public goods and a welfare safety net (which includes underwriting access to services such as health and education). In
a New Zealand context a generous allowance for such spending (including by local government) would put it at no more
than 20-30 percent of GDP. Work by International Monetary Fund researchers suggests that much of what governments want
to achieve could be realised with government spending at around 30 percent of GDP. On this basis it would be quite
feasible to envisage a reduction of total government spending in New Zealand from around 40 percent to 30 percent of
GDP.
In a study for the Business Roundtable, Winton Bates estimated that reducing the most ill-justified government spending
by 10 percentage points could add 0.5 percent per annum to the growth rate for a 10-25 year period. This estimate is not
out of line with findings endorsed by the OECD and Arthur Grimes. Few policy changes offer gains of that magnitude. This
motivates the search for more effective mechanisms to constrain government spending.
The case for tax and spending limitations
Since the FRA has been relatively ineffective in curbing growth in government spending, an obvious question to ask is
whether more explicit disciplines could be introduced into it. The OECD has favoured supplementing deficit and debt
targets with measures aimed at limiting spending and taxation.
Voters know about spending caps from their personal lives. Most families cannot buy everything they want and must limit
their spending to what they can afford. They have to prioritise. Governments need to do the same. Tax and expenditure
limits (TELs) also have the merit of giving politicians a level of protection against vested interest groups and may
encourage decisions that are more closely aligned with the public interest.
There is nothing undemocratic about such limitations. They could be self-imposed by a government, or alternatively
decided or ratified through referenda. A subsequent government would be free to repeal or modify their provisions, or to
seek endorsement for modifications through a referendum.
TELs have been a feature of fiscal management in a number of countries. At a national level, Hong Kong, Japan, the
Netherlands, Spain, Sweden, Switzerland and the United States have had various rules aimed at capping spending. Over the
last 25 years, 27 US states have had TEL rules on their books, and there have been variants at the state level in
Australia and Canada.
Evidence from the United States suggests that TELs have been effective in constraining spending. Through their effect on
increasing economic freedom, TELs have been found to be conducive to economic growth. The difference in per capita
incomes between US states with the highest and lowest levels of economic freedom is as high as US$7000.
The research also finds that the most effective rules have:
- been constitutional rather than statutory;
- capped spending at the rate of inflation plus population growth;
- provided for immediate refunds of surplus revenues to taxpayers; and
- been linked to other rules, such as balanced budget requirements.
Provision can be made for such rules to be set aside when necessary, but a super-majority requirement for voting to do
so is likely to be desirable if the rule is to be effective.
Colorado is generally regarded as having one of the best arrangements currently in place. The Colorado Taxpayer Bill of
Rights accords well with these criteria. Along with other states, Colorado did, however, experience a period of fiscal
stress following the dotcom crash. This has led some observers to propose modifications to its rules.
The Colorado Taxpayer Bill of Rights also requires voter approval for any new taxes, tax rate increases, extensions of
an expiring tax, or tax policy change that directly increases net revenue. Voters defeated measures to raise taxes or
increase spending every year from 1993 to 1999, but in 2000 they did approve an amendment that increased state aid to
public education. There is a case for a super-majority requirement for tax increases because of the risk that a simple
political majority may act in a predatory fashion on either a minority of the population (eg high income earners who pay
the lion's share of tax) or on a relatively unorganised majority of taxpayers. Such a super-majority provision, which
could apply both to citizen referenda or parliamentary decisions on tax, has parallels in the Companies Act which
requires a major transaction to be approved by (at least) a 75 percent majority of shareholders.
The Business Roundtable report proposes limits on both spending and taxes. Such a TEL rule would fit readily into the
FRA or its successor. A rule that held spending growth to population growth and inflation would allow economic growth to
reduce spending (and thereby average tax rates) as a percentage of GDP without necessarily requiring cuts in real
government spending per head. With per capita GDP growth of 2.5 percent per annum (the average of the last decade), such
a rule if applied successfully to overall government spending would lower it from 40 percent of GDP to 30 percent in
about 12 years.
Other options for constraining government spending
A TEL rule is inferior in principle to a "bottom up' rule that spending be cut where the costs to the nation exceed the
benefits. Just as in a household context, however, such a limit can be justified as an overall constraint, particularly
in protecting taxpayers at large against spending pressures from special interests. A key virtue of a TEL is that it
shifts the power to increase taxes and spending from politicians to voters. It would be difficult to exaggerate the
desirability of asserting the principle that politicians should not be spending taxpayers' money without their general
agreement. It is not just the endless reports of waste and corruption in government spending of relatively small amounts
that is offensive to good government. It is also unconscionable that politicians are spending an amount equivalent to
around 40 percent of GDP year in and year out without establishing that even major spending programmes are providing
value for money for taxpayers at large.
Because a TEL rule largely ignores the issue of the quality of base spending, it needs to be supplemented by other more
direct measures. The OECD has noted that base spending, 95 percent of the total in New Zealand, is not properly
reviewed.
The key problem here appears to be a lack of political will. Assessing value for money is not rocket science. It is
simply a matter of demanding convincing answers to basic questions. For example, a report from the OECD secretariat
recently commended the following questions for evaluating value for money:
OECD proposals for evaluating value for money
1. Does the programme still serve a clearly defined public purpose that matters?
2. Is this an appropriate role for government?
3. Would we establish the programme today if it did not already exist?
4. Is it desirable to maintain it at its current level?
5. Can it be delivered more effectively or efficiently? Have there been changes (in the service environment,
infrastructure, technology, etc) since the programme's inception that would now permit an alternative means of achieving
its objective with greater economy, efficiency, or effectiveness?
Sources: Canadian Office of the Auditor General and Finance Canada.
The OECD report suggested that an outcome-based review of social assistance would probably offer the greatest gains. It
advocated a focus on welfare administration, benefit levels, the use of market mechanisms and recourse to user charges
to limit demand, and noted the benefits of competition, choice and diversity.
Spending discipline would also be improved by taxation that was more closely linked to benefits. An old tax principle
was that "every man is bound to contribute to the public revenue in proportion to the benefits he receives from the
public protection". A flat or proportional income tax might accord with this principle, as well as having desirable
properties of efficiency, fairness and simplicity. People on higher incomes would still pay more with such a tax, but
all would be required to contribute if it were applied from the first dollar earned. This would encourage greater
scrutiny of spending programmes and make it harder for a majority of voters to exploit a minority of taxpayers. Both the
2001 McLeod Tax Review and the Treasury have advocated moves towards a flat or flatter tax structure from a growth
perspective. Because faster rates of growth are likely to provide the best route to reduced poverty and greater wealth
for all, such a tax policy reform is arguably more equitable as well. The increase in the top tax rate to 39 cents and
the government's Working for Families package in the 2004 budget, which increases effective marginal tax rates on
average, go in the opposite direction.
Supplementary expenditure rules could also be used to highlight the proper role of government. Government activities
should be principled and limited - limited to dealing with serious problems inherent in voluntary arrangements, such as
the provision of public goods, and a social safety net. This would rule out government involvement in most commercial
activities, allowing end users to get the benefits of competitive private supply. The conflicts of interest associated
with government ownership distort competition, and a large body of evidence has built up indicating that private
ownership and operation, on average and over time, is more efficient than public ownership. The current government's
opposition to sales of commercial enterprises regardless of where the national interest lies is essentially ideological
and stands in stark contrast with policies in other OECD countries. There should also be a ban on subsidies to
businesses, or at the very least a strong presumption against such spending and lending.
Consideration could be given, independently of a TEL rule, to allowing citizens to vote in a binding referendum to
strike down current and proposed spending programmes and taxes.
Other proposals are canvassed in the Business Roundtable report, including measures to improve the accountability of
select committees to citizens, the accountability of the executive to parliament, and the delegation of parliament's
power to spend.
Conclusions
The provisions in the Fiscal Responsibility Act 1994 have served New Zealanders well to date, except for the lack of
real government spending discipline. New Zealand's experience is in line with the findings of international research
that deficit limitation rules do not restrain growth in spending and taxation through time.
Current arrangements allow the government to treat the additional revenue from economic growth as being available to
spend as a windfall. There is no burden on governments to prove that the spending is justified and no presumption that
the money should be returned to taxpayers unless the government can show that additional spending is in the national
interest.
Based on the experience of the United States, tax and expenditure limitation rules backed by citizen referenda and
super-majority rules have the best chance of improving spending disciplines. This is the main conclusion of the
forthcoming Business Roundtable report.
Regardless of whether a TEL is put in place in New Zealand, there is much that can and should be done to improve the
quality of government spending. Super-majority requirements for new taxes, increases in existing tax rates, or increases
in the tax base would assist. So would allowing citizens to strike down such legislation by referenda. Under all
scenarios there should be a systematic review of base spending, applying the OECD's checklist.
There are other options that should be considered to help discipline government spending and taxation in New Zealand.
Many of them are complementary rather than mutually exclusive.
ENDS