AUT Policy Observatory May 2016: The original is here.

Contents                                                                                                            Page

1.  Why Government Spending is Necessary                               4

2.  The Public’s Demand for Public Spending                            11

3.  The Long-Term Record of Government Spending                15

4.  Recent Trends in Public Spending                                         27

5.  How Taxation Fits in                                                             36

6.  Should There Be More Government Spending?                    42

Tables

2.1         Spending Preferences                                                                              11

2.2         Responses to Particular Prepositions                                                                                            12-13

5.1           General Government Revenues as a Percentage of GDP (2013)                                                                       40-41

Graphs (Fiscal years unless otherwise stated)

3.1            Total (Current) Government Expenditure: 1876 – 2011              16

3.2          Defence Expenditure: 1880 – 2011                                                     17

3.3           Education Expenditure: 1863 – 2011                                                 18

3.4          Health Expenditure: 1862 – 2011                                                         19

3.5            Law & Order Expenditure: 1876 – 2011                                            20

3.6           Social Security Transfers: 1862 – 2011                                             21

3.6A        Social Security Transfers: 1938/9-2010/11                                      22

3.7           Other Government Expenditure: 1876 – 2011                                23

3.8          Other Government Spending by Components: 1946/7 – 2010/1                                                                         24

4.1           Total Core Crown Expenses                                                                  28

4.2          Big Government Spending Categories                                               29

4.3          Small Government Spending Categories                                          33

Executive Summary

There are good reasons that some services should be funded and provided by the government because private markets do not always give the economic outcomes the community desires. A second consideration is the level of economic inequality generated by the private market may be unacceptable and public transfers are necessary to reduce it. 

There seems to be a substantial public desire for additional government spending in some expenditure areas. The total demand may amount to about an extra $4 billion a year, or about 3 percent of GDP. 

Government spending in appropriate sectors has a long history of increasing as a proportion of GDP. 

There has however been the pattern in recent years with many spending areas falling or stagnating as a proportion of GDP rather than continuing to rise as they have in the past. This almost certainly reflects a reluctance of the government to raise taxes. 

The main source of additional revenue has to be from income taxes, given that there are limited alternatives and widespread public concerns about income (and other economic) inequality.

1. WHY GOVERNMENT SPENDING IS NECESSARY

Introduction

Economic transactions which occur in perfect markets are simple, elegant and efficient. But they require that both transactors have perfect (or at least the same) knowledge of what is going on, that the transaction costs of the deal are negligible, that there is good title, that the purchaser pays out of their own pocket and that the purchaser has adequate means to pay – among other things. 

However, markets are rarely perfect and imperfect markets can be complex, clumsy and inefficient. Sometimes the imperfection of the market can be improved by public interventions. These can be simple — imagine a world in which there was no enforced standard of weights and measures; or they can be very complicated, as we shall see. Sometimes it is judged that the most effective means involves public funding and perhaps also public supply of the service.

Reasons for Public Spending

For the purposes of analysis of the public sector, economists characterise goods and services along two dimensions. One is whether its provision to one person can ‘exclude’ others from using it. The famous example of non-excludability is the lighthouse; every ship may avail itself of its locational light even if it was erected for only a particular ship or group of ships. The other dimension, called ‘rivalry’, is the extent to which consumption by one prevents others from consuming it. An example of non-rivalry is that using a piece of information does not prevent anyone else using it. 

The two dimensions together offer four possible categories. In three cases the private market works reasonably well, sometimes after a little attention. However in the fourth category, of public goods which are non-rivalrous and non-excludable, private delivery is deeply problematic. 

Among the examples of public goods (and services) are fresh air, knowledge, some public infrastructure, national security, the justice system, some education, flood-control systems, and street lighting, as well as lighthouses. Each faces the critical problem that the private market is likely to undersupply it relative to public demand and economic effectiveness. Note that ‘public goods’ are not the same things as ‘the public good’ (a.k.a. ‘common good’, or ‘public interest’), which is beneficial for all or most members of a given community.

While the existence of public goods provides a case for public provision of a good or service, it is not the only reason. Among other salient ones are to generate a better distribution of income (or wealth or wellbeing), to reduce transaction costs, and the incompleteness of markets. 

Economic analysis has never been able to demonstrate that market transactions result in an equitable income (or wealth or wellbeing) distribution. In fact equity is a normative judgement – its various statistical measures are not only imperfect and conflicting but require a normative judgement to evaluate. Assessments of efficiency and other measures of economic effectiveness are of a more positive nature requiring no ethical judgements (although such judgements may be necessary when bringing together a number of the measures). In a democracy the ultimate decider of the appropriateness of the economic distributions is Parliament. In practice every democracy has concluded that market outcomes have to be modified by transfers, taxes and other measures. 

Market solutions can also involve very high transaction costs. For instance, the public-based ACC replaced a private-based employee insurance system, in which 40 percent of the expenses were used for litigation and associated costs, by a public-based one, in which administrative costs are about 10 percent. Nor is it accidental that the most market-based healthcare system of the US has proved to be the most expensive and yet does not perform well on measures of health outcomes for the whole population.

(The case for public spending because of positive externalities — the favourable effect of an activity on an unrelated third party –usually involves a public good with high transaction costs of recovering a contribution from the third party.) 

In the ideal world there are what economists call ‘complete markets’, in particular where there are inter-temporal and stochastic decisions to be made. In particular, there are incomplete markets in insurance and finance; this means that the canniest, most rational decision-maker cannot make decisions to cover all contingencies. As a consequence, catastrophic events – such as earthquakes and financial meltdowns – often involve some government bailouts when they occur. (The size of the bailout can be reduced by requiring economic agents to make prudent decisions for what can be covered by private insurance and informed investment decisions.) Sometimes the government may be involved because the outcomes are so complex that the existing property rights may be inadequate to handle them, except by long litigation involving high transaction costs. 

It does not follow that public provision is necessarily superior to private provision even when it is imperfect. Yet, the record is that rich ‘mixed’ economies use public provision for between 20 and 40 percent of total market supply (although there are definitional problems) with, typically, a range of interventions to improve the imperfect private remainder. 

Deciding the balance between public and private provision requires judgements and predictions of facts (including of how people behave) underpinned by values (including a willingness to tolerate different kinds of imperfect outcomes). 

Some Illustrations

While the reasons for public spending set out earlier in the chapter are separate, in practice the reasons for any spending overlap. Here are some examples of some of the major public spending groups and the reasons why they might exist. They are listed in approximate order of the size of the group.

Health

Healthcare can be divided broadly into two categories: population-based healthcare and person- based healthcare. 

Most of the population-based healthcare is a public good, that is the benefits from it are rarely exclusive to a single person and as, a general rule, one person benefiting from it does not prevent others from benefiting. Examples are sewerage and vaccinations (since one person being vaccinated usually benefits others by reducing the probability of an epidemic). Another group of activities centre on the individual not being fully informed about the health implications of behaviour.

Personally directed healthcare is rarely a public good. However, it would involve very high transaction costs to get the market to work effectively. In practice the decision-maker is, effectively, the health professional, which can result in Supplier Induced Demand (SID), where the supplier uses their privileged position to encourage an individual to take up a treatment they supply which a fully informed consumer would not. 

The inefficiency of the market transaction is compounded where the funder of the transaction is not the consumer (patient). This is inevitable in public supply of healthcare and involves a number of tensions. But it also applies where the funder is a private insurance provider – the need arising because the incidence of the need for healthcare is erratic among individuals.

These failures appear to be so large that nowhere in the affluent world is healthcare provided without considerable public intervention funded by the taxpayer or by compulsory insurance. As a general rule the more transactions are left to the market, the less successful is the health delivery system, well-illustrated by the US one which is the world’s most expensive but, on the standard indicators of health attainment, is below average among its comparators.

Education Services

There are four main reasons why there is public involvement in the provision of educational services. In no particular order, first, the beneficiary is not always the funder. Even under a purely private system the funder would be the parent rather than a child (at least up to secondary level). 

Second the decider is not well-informed about what is to be ‘consumed’; this is well-illustrated by a paper by an economist which required advanced university mathematics to make the optimal decision whether to take up tertiary education. More generally, parents use league tables to make decisions about which school their children should go to, despite the league tables being meaningless for this purpose. (They assess the level of educational attainment of the school’s pupils, but what the parent needs to know is the educational increment the school will add to their children – the two are only the same if all children start off at the same level.)

Third, the cost of schooling is such that without public funding support, it would be unaffordable and most children would receive less education than would be efficient or equitable.

Fourth, there are aspects of education which are public goods – training in citizenship for instance.

Retirement Provision

New Zealand provides for the older population a universal basic income (New Zealand Superannuation) and supplements the cost of residential care (which appears in healthcare spending). Fundamentally this is because most people save insufficiently for their retirement. This is true in every affluent country even where there are compulsory – and not so compulsory – occupational pensions, even if they are subsidised. (Many are not actuarially sound and will have to be supplemented by the central government, one way or another). The less support a country gives its elderly, the more of them are in poverty. Thus the spending is for distributional purposes.

Social Security Spending

As already mentioned, markets do not generate socially equitable income distributions. Moreover, high inequality may lead to social incoherence and disturbances and to economic efficiency insofar as individuals may not be able to develop their talents because of inadequate family support, healthcare or education. Social Security is a means of shifting the disposable income distribution to something which society judges to be more appropriate for reasons both of equity and to increase economic performance. 

Housing

Neither does the market generate a socially equitable wealth distribution. On the whole New Zealanders do not worry about this; the statistics on the wealth distribution are not nearly as comprehensive as those for the income distribution (which are relatively limited anyway). The exception is housing. The analysis of this is complicated (as are all situations involving the wealth distribution interacting with incomes and consumption) but for these purposes it is sufficient to observe that affluent economies have found it necessary to spend on housing to cover for families with inadequate housing.

Defence Spending

Defence spending is almost only a public good (or service). Defend one, defend all. 

Law and Order Spending

Like defence spending, most law and order Spending is also almost only a public good (or service).

Heritage, Culture, the Arts, Recreation and Environment

These are generally public goods (i.e. services). As a result, for a number of complicated reasons the private market by itself under-supplies the quantities which the public desires. 

In the case of heritage and environmental spending there is also an element of intergenerational equity, where the spending benefits future generations (who have no influence on current market demand). Markets do not reflect the demands of future generations (nor obligations to past generations) very well. 

General Economic Services

There is a view that the private sector can be assisted to provide greater output or better quality output. There is some dispute within the economic profession about just how far this approach should go. There has been a major shift in policy views between the Muldoon era and today as to how extensive these interventions should be; the alternative view is to keep business taxation and other industrial costs low.

There would be general agreement that some infrastructure would be undersupplied by the market, because the transaction costs of efficient charging would be too high (e.g. roading tolls). There is also a widespread acceptance that industrial training and research and development would be undersupplied if left to the market. The essence of the two cases is the externality argument that they are a kind of public good which has very high transaction costs to fund privately. 

Core Government Services

Core Government Services (mainly the government’s administrative spending) provide almost only public goods and services. In the 1980s ways of charging for some of them were explored, but on the whole they proved unsuccessful. 

Conclusion

All the above explanations are too brief, but what they indicate is that there are good reasons for public spending in some areas of activity. They do not demonstrate that there should only be public spending for typically the best delivery comes from a mix of the two. Nor do they conclude that they should necessarily be publicly supplied. 

Consequences of Public Spending for Taxation

But the balance also requires funding of public spending, for the larger it is (in a fully employed economy), the smaller is private spending (and vice versa). Since the level of private expenditure is primarily controlled, borrowing aside, by the level of private disposable income (PDI), an increase in public spending requires a reduction in PDI (an increase in saving aside). Taxation is a means of reducing PDI. Thus the role of taxes is to reduce the size of the private sector in order to expand public sector spending. 

As the caveats in the previous paragraph indicate, there are complications from borrowing and saving, but these do not change the main conclusion markedly – taxation is required to influence the balance between private and public spending. (Another caveat is that the public sector has some revenue streams of its own; even so, most of its funding comes from taxation.)

Not only then does funding public provision require taxes – with perhaps some loss of overall production from behavioural responses as higher taxation discourages work and savings efforts — but there is no obvious automatic limit to the amount of public spending to be provided. A private market purchaser is limited by his or her income plus, possibly, what can be borrowed – the budget constraint. Certainly there is a ‘budget constraint’ in the public sector which is the revenue it can raise – borrowing is discussed later – but insofar as taxation revenue can be increased by higher (and more) taxes the constraint is ‘soft’. (Many of the arguments against higher public spending and higher taxation – discussed later – are attempts to make the public budget constraint ‘harder’.) The decision to set the level is a political one – the restraint is the willingness of the polity to reduce its private spending including the total loss of production that higher taxation may cause. 

Designing effective tax systems is not simple. But even the most efficient (and equitable) tax system will, to some extent, change behaviour, typically in the direction of reducing market production (which is taxed) towards non-market activities (which are not). That has the effect of reducing the totality of (market) production available for public and private provision. 

It is true that some public provision will increase effective market production – perhaps by reducing transaction costs (as when weights and measures are standardised and enforced) or increasing productivity by inducing activities which will not be properly funded by private market payments (infrastructure investment, for example).

Notes on Debt Servicing

Debt servicing is the consequence of spending more than other current revenue in the past. One generation can always increase its spending by borrowing, in effect reducing the spending of a future generation. While the data is presented by years (being collected that way) one might think of the principles being set out here applying to an entire deficit and surplus cycle. The case for and against deficit spending is complicated and is not pursued here.

Public Spending and Wellbeing

But at the margin, public spending is for the purpose of private or community consumption. As such it can contribute to people’s wellbeing, although offsetting any increase will be a reduction in their wellbeing from reduced private consumption. But if the public provision is prudent, there should be a net improvement in total wellbeing – or at least that is the judgement the political process attempts to make, as it decides on the level and distribution of public provision and the level and means of taxation required to fund it. 

The ideas in this introduction are complex and necessarily the presentation is dense including alluding only to some important caveats. The rest of the report will elaborate what here has had to be conveyed briefly.

2. THE PUBLIC’S DEMAND FOR PUBLIC SPENDING

What the Public Says

The 2014 Election Study provides an example of surveying adult New Zealanders for their opinions on public spending. The findings are summarised in Table 2.1. They are not very different from other and earlier surveys. 

Table 2.1 Spending Preferences (Percent)

<
> Spending Area More Same Less Net More Health 68 27  2 66 Education 65 30  2 64 Housing 50 37  8 41 Law Enforcement 44 46  5 39 Public Transport 43 44  8 35 Environment 41 47  7 34 Superannuation 32 57  5 27 Business & Industry 31 50 12 19 Defence 19 55 19 -1 Welfare 16 43 36 -20 Unemployment Benefits 13 53 42 -29

Source: http://www.jackvowles.com/SectionC2014.htm

However, as we shall see, there are difficulties interpreting the responses. A second set of questions which involved asking participants to agree or disagree with a series of statements provides some help. Their responses are summarised in Table 2.2 (which includes some statements not germane to this report).

As a general rule, people want more public spending. The exceptions are their broadly neutral response on defence and antagonism to welfare spending (of which more below). However, it is not clear whether they support higher taxation to fund the spending; one would like to think them realistic enough to accept this with such large public spending demands. It is likely that they are inclined to taxing others, though. 

Table 2.2 Responses to Particular Propositions (Percentage)

<
> Proposition Agree Neither Disagree Net Agree Work for the dole 76  8 13 63 Subsidise company R&D 68 15 12 56 Income inequality too large 67 13 15 52 Trade unions necessary to  protect workers 64 12 15 49 Assist export companies 61 17 14 47 Reduce income inequality 64 12 19 46 Assist international sportspersons 58 19 19 38 Big business has too much power 55 16 17 38 SOE privatisation has gone too far 53 15 18 35 Assist film makers 52 19 24 28 Proposition Agree Neither Disagree Net Agree Raise Super age to 67 43 10 39  4 Many on Welfare don’t deserve help 40 14 41 -1 NZ needs CGT excluding family home 34 10 35 -1 Unions have too much power 27 24 30 -3 Lower benefits help people stand on their own feet 39 13 43 -4 Help banks in time of crisis 27 23 42 -15 More immigration 11 38 46 -35 Abortion always wrong 21 15 60 -40

Source: http://www.jackvowles.com/SectionC2014.htm

The one area where participants report they want reduced spending is on welfare. However, more than any other category, this involves ambiguities. (Presumably those surveyed are referring to benefits, although conceivably there may be some doubts about the effectiveness of social work services.)

However, the benefit payments represent such heterogeneity of purposes that it is unclear what the public may mean. Undoubtedly, as both tables show, the public objects to unemployment benefits. Perhaps they are indicating a rejection of unemployment per se. One guesses they may not object to support for the transition between layoff and the next job; presumably the great objection is to people who are permanently (or seem permanently) on unemployment benefits. One assumes that there is little objection to those who are unemployed because of disability or sickness. 

How about solo parents who are not in the paid labour force? There has been some antagonism to the Domestic Purposes Benefit (now Sole Parent support). Again the respondents may not be objecting to the benefit so much as to the existence of a social phenomenon they dislike. Perhaps there is belief that the benefit encourages family breakups (and there is the inequity that a parent without a partner practising childcare receives state support but one doing the same job with a partner does not).

On the other hand, the public wants spending on New Zealand Superannuation transfers to be increased, albeit to a narrower age group. That makes up over half of benefit spending. (In comparison the Jobseeker Support and Emergency Benefit was about 8 percent in 2014/5.) A bigger item is support for children, amounting to around 13 percent, excluding Sole Parent

Support which is another 5.5 percent. The other ‘big’ item is housing support, amounting to just over 6 percent – again housing is favoured for additional spending. 

Curiously while the public wants less spending on welfare, it is divided on whether ‘many (sic) on welfare don’t deserve help’. Perhaps it just wants less on welfare – certainly fewer who are unemployed and on the jobseeker benefit. 

One is left with the uneasy thought that the public’s antagonism to welfare spending is not very informed (and the reflection as to who is at fault for the ignorance together with the thought that a detailed systematic survey would be helpful at unravelling these puzzles).

To add to the puzzlement, over half of those surveyed thought there was a need to reduce income inequality. Presumably they want more tax on those on high incomes, but at the bottom of the distribution those in need are families with children, especially those dependent on benefits. Thus the public’s dislike is probably a signal that there is a dislike of the unemployment on long-term benefits rather than at the whole welfare system. 

There may be other oddities in the responses. There is a little net support for spending more on business and industry, but considerably more enthusiasm for some components of that spending: subsidising company R&D and assisting export companies. 

The public very clearly indicates it wants more spent on health, education, housing, law enforcement, public transport and environment (in that descending order). The support for public transport is especially interesting because only 2.5 percent of urban trips are by it. Supporters may include those using private transport who would like others to use public transport to free up their roads. 

The culture, heritage and recreation group was not included in the survey, but perhaps support for assisting sportspersons and filmmakers may give clues. Perhaps both reflect a demand for the promotion of national identity. 

In summary, and subject to reservations about the ambiguity of the questions, the public seem to want most of the public spending that the New Zealand government provides and to want more of it. Unfortunately, neither this survey, nor most of the others, asks useful questions about their willingness to pay taxes to fund the spending. 

3. THE LONG-TERM RECORD OF GOVERNMENT SPENDING

What Data is Available?

The limited number of long series which reflect the pattern of government spending generally cover only the central government. This chapter brings together that data with some commentary. It compares public spending with nominal GDP. (There are numerous technical issues with the data. A longer version of this chapter which sets out the major ones is available from the author.)

A rise in government spending relative to GDP might be because the price of government services has risen relative to the general price level of the products in nominal GDP – the ‘Baumol’ effect of service industry prices (in both the private and public sector) rising faster than in other sectors because they do not have the same productivity gains. Or it might be because there is an increased demand for the expenditure by the whole (or relevant) population.

The Long Term Series – from the Nineteenth Century

Total Central Government Current Spending (1876- Graph 3.1, next page) (excluding war expenses) is a fairly constant 13.5 percent of GDP through the nineteenth century up to 1918. Thereafter it begins a steady increase rising to about 22.5 percent of GDP in 1966 and a rapid increase thereafter, peaking at about 40 percent in the early 1980s. It began to decline from the mid-1980s bottoming out at about 32 percent. In the late 2000s spending increases again.

<
>

Source: Statistics New Zealand.

The colours in the graph distinguish leftish from rightish governments. They suggest that rises in aggregate spending do not particularly reflect the politics of the government. 

<
>

Source: Statistics New Zealand.

Defence Spending (1880- Graph3.2) has generally been low except for periods of global warfare. In the post-war era there has been a tendency for it to fall.

<
>

Source: Statistics New Zealand.

Education Spending (1862- Graph 3.3) begins as negligible, in part because of the low proportion of children in the population, in part because much educational spending was a provincial responsibility. It rises with the abolition of the provinces and then creeps up to just under 1.5 percent in 1886, flattening out for the next 20 years, and then doubles to almost 3 percent over the next 26 years to 1932. The spending then falls as a proportion of GDP apparently because of Great Depression and the Second World War. After the war it trebles from 2 percent of GDP to 6.5 percent. There is a discernible slowdown in the spending increase from the mid-1970s arising from the relatively slower growth of the young population. There is no evident difference in the rises according to the politics of the government.

<
>

Source: Statistics New Zealand.

Spending on Healthcare (1862- Graph3.4) shows an upward trend similar to educational spending. In the middle of the nineteenth century it was a minuscule 0.3-0.4 percent of GDP because many health outlays were private, charitable or from local bodies paid by rates. Much of the central government spending was on population-based healthcare, rather than for personal care (except for the indigent or Maori). The proportion rose mildly in the early twentieth century until the First Labour government began funding personal healthcare and as new, more expensive and more effective treatments were introduced. Healthcare spending reached 3 percent of GDP by 1949 and continued to grow to almost 8 percent by 2011. As in the case of education, there is a demographic effect, but the ageing of the population worked in the opposite direction. There is no evident difference in the rises according to the politics of the government. 

<
>

Source: Statistics New Zealand.

In the middle of the nineteenth century the central government was spending about the same on Law and Order (1876- Graph 3.5) as on education or health, although outlays on all of them were small. But Law and Order’s spending share of GDP was on a downward track until the 1950s to 0.2 percent. In the late 1960s the downward trend reverses and spending on law and order more than quadruples to over 0.85 percent in 2011. 

<
>

Source: Statistics New Zealand.

Social Security Transfers (1862- Graphs 3.6 &3.6A, overleaf) are difficult to interpret because they are a part of the state income redistributive system (the other major part is taxation) and so they give only a partial – even misleading – picture.

<
><
>

Source: Statistics New Zealand.

Despite the introduction of the Old Age Pension in 1898, there were barely any transfers before the end of the Great War. They lift to close to 2 percent of GDP in the 1920s, rise a little in the Great Depression (but unemployment support was often treated as outside these government accounts).

The introduction of the 1938 Social Security Act revolutionised the spending, initially lifting it by 1.5 percentage points of GDP. It remained at about that level during the war and then lifted another 3.5 percentage points after to over 8 percent of GDP. (The total includes war pensions.) The First National Government was not so energetic in increasing benefit levels and made few extensions. In any case there was little unemployment. Spending sank to below 7 percent of GDP, climbing back to just above 8 percent under the short-lived Second Labour Government. By 1971 spending was below 6 percent of GDP under the second National Government. Irrespective of the colour of the government, social security spending increased dramatically in the decades of the 1970s and 1980s, more than doubling as a result of the implementation of the recommendations of the 1972 Royal Commission on Social Security, the Third National Government’s universal retirement provision (New Zealand Superannuation), the ‘grossing up’ (paying at a pre-tax rate) of social security benefits, and the rise of unemployment.

The Muldoon era was abnormal for the blues. Under the Fourth National (Bolger and Shipley) Government (1990-1999) spending fell back to 12 percent of GDP, a trend continued under the Labour-led Government down to 10 percent of GDP. Factors here included the raising of the age of eligibility for universal retirement provision from 60 to 65 years, increasing benefit levels for inflation but not for rises in real market income and a more benign labour market. 

So, contrary to the pattern for the other spending components, there is a case that the leftish governments tend to raise social transfers more than rightish governments – although the Muldoon and the Clark-Cullen Governments are exceptions.

<
>

Source: Statistics New Zealand.

Other Current Government Expenditure (1876- Graph 3.7) (excluding debt servicing) ran at about 10 percent of GDP in the period to the end of the 1920s, rose sharply during the Great Depression peaking near 18 percent (it included work creation schemes), fell to 6.5 percent in the late 1950s, and then rose sharply to just over 17 percent in the early 1980s (explained in the next chapter), falling back to 7.5 percent in the middle of the first decade of the twenty-first century. 

Post-war Spending: Components of Other Current Government Expenditure (Graph 3.8) 

<
><
>

Source: Statistics New Zealand.

Core Government is the administrative services of the government. It ran at just under 2 percent of GDP in the early post-war years, increased to almost 2.5 percent in the mid-1970s, and then declined to below 2 percent again from the 1990s. There are two deviations from this pattern. Some of the costs of state restructuring in the late 1980s and early 1990s were charged to core government. The spikes at the end are the consequences of writing off of student debt owed to government.

Financial Costs – mainly debt servicing – were typically about 2.5 percent of GDP in the early post-war era, falling to below 2 percent after 1990 and nearer 1.5 percent at the end of 2010. However, between 1974/5 and 1999/2000 finance costs rose to almost 8 percent of GDP in 1987/8, mainly from paying off off-balance sheet guarantees given by the Muldoon Government to private and public businesses. Some of the subsequent decrease in debt servicing occurs because State Owned Enterprises do most of their borrowing on their own behalf rather than through central government. 

Economic and Industrial Services show a slightly rising trend to the mid-1980s, and begin falling after as the government cut back its direct industrial subsidies. They were never large – around 1 percent of GDP. 

There are two series which start in 1971 /2. 

Miscellaneous (Heritage, Culture and Recreation; Primary Services, Housing and Community Development, Other) It is not clear that this is a coherent homogeneous group. In 1995 it was 0.7 percent of GDP doubling in the next fifteen years (carbon credits complicate the interpretation thereafter).

Transport and Communications Services appear to be falling a little from the early 1970s. They include centrally funded (including by transport levies) road building. Again there are definitional problems. 

Conclusion

Government spending compared to GDP has tended to rise over the years. The increase has usually been slow, but there was a sharper increase from the late 1960s to the early 1990s (and in the late 1930s from the Welfare State and War). There has been some cut-back since, but the level of spending is still higher than it would have been had the trend up to the 1960s continued.

The main drivers of the increasing proportion of government spending were, roughly in order of importance

• social security especially from the 1970s;
• healthcare, which has grown faster than if it were just due to the ageing population;
• educational services, which like healthcare, have grown faster that the population they service.
• law and order since the 1950s, which is currently smaller than the previous two;

Not included in this ranking are 

• defence, which has an enormous impact during global wars, but not recently;
• debt servicing, which appears to be low and generally stable in the long run, although financing costs had a serious impact in the 1980s and early 1990s. 

Other spending components are small and relatively stable. 

Public spending seems more influenced by the state of the economy and long-term public demands than by the political stripe of the government. The exception is that Labour is usually more benign towards social security, although even here the Third National Government (Muldoon) and the Fifth Labour Government (Clark-Cullen) were not entirely in character.

Most importantly, the history of public spending in New Zealand shows a public demand for it and a willingness, to some extent, to pay taxes to fund it.

4. RECENT TRENDS IN PUBLIC SPENDING

Government Spending 1995/6-2014/5

Note that the existing data base ends in 2014/5 but there is (almost) another full year of spending since. That and spending plans announced in the 2016 budget will alter the conclusions, but probably not by much.

For recent years, there exist reasonably consistent series of Core Crown Expenses, corresponding to government spending. However, definitions change over time and the categorisations used are not always those best suited for this analysis so that some adaptations have been necessary. They are explained in the appendix to this chapter. 

The New Zealand Economy 1995/6-2014/5

From the mid-1990s the New Zealand economy began expanding through to 2007, when the economy began to turn down. (In 1998 there was a production downturn from the over-reaction of the Reserve Bank to the Asian crisis.) Shortly after the New Zealand economy entered a cyclical downturn (as did the US one), the world economy had the Global Financial Crisis (GFC) which further depressed the economy. The incoming National Government chose to react with income tax cuts which, with the savings the previous Labour-led Government had squirreled away and the opening up of exports to China following a free trade agreement, meant that New Zealand did not experience a major downturn as did many affluent economies. Rather, it experienced a period of five-to-six years of stagnation of per capita GDP. 

Two other events had important fiscal impacts. From late 2010 through 2011, the Canterbury area suffered devastating earthquakes. The government chose to fund its contribution to the recovery out of existing revenue rather than levy a special tax or borrow. Additionally, one of the consequences of the GFC was that many countries, including New Zealand, guaranteed deposits in financial institutions. Although it levied a charge for the guarantee the cost to the exchequer from bailing out depositors in failed institutions exceeded the revenue. (The bailouts appear in the Core Crown Expenses.)

Allowing that a budget in May of year X tends to set government spending from July in year X to June in year X+1 (i.e. fiscal year X/X+1 or just fiscal year X+1), even if there is a change of government towards the end of year X, the political responsibilities for the twenty years can be allocated as follows:

          1995/6-1999/2000: National (Prime Ministers: Jim Bolger and Jenny Shipley;

Ministers of Finance/Treasurers: Bill Birch, Winston Peters and Bill English)

               2000/1-2008/9: Labour-led (Prime Minister Helen Clark; Minister of Finance: Michael Cullen) 

               2009/10 – 2014/5: National (Prime Minister John Key; Minister of Finance: Bill English)

Core Government Spending  

<
>

Source: New Zealand Treasury

As Graph 4.1 illustrates, Total Core Government Spending (excluding losses) was about 31 percent of GDP under the National Government of the 1990s, fell to about 29 percent in the first years of the Labour-led Government in the 2000s and then rose slightly in its latter years to about 31 percent again. In the first years of the latest National Government there was an increase in spending to about 34 percent of GDP, mainly as a result of expenditure not being restrained as much as production stabilised (this was for counter-cyclical purposes), the financial bailouts and the Canterbury earthquakes. From 2011/12 the Key-English Government has been restraining expenditure and by 2014/15 the level was 30 percent of GDP, a fraction lower than when National left office in 1999. 

The next sections show the components of spending in approximate order of size.

<
>

Source: New Zealand Treasury

Healthcare (Graph 4.2)

Healthcare spending increased from 5.0 percent of GDP in 1996/7 to 6.0 percent 2007/8, rose to 6.4 percent by 2010/11 and the share has since been drifting down to 6.2 percent in the last available (2014/15) year. 

We might discount the rise in the early 2010 as a result of government spending not being as restrained as much as the stagnation of production. If we project the trend from 1996/7 to 2007/8, spending on healthcare would have been slightly over 9 percent higher than it was in 2014/15 – say an extra $1.4 billion. 

We might expect spending on healthcare to rise faster than GDP because:

 First, an ageing population requires more healthcare. The over-65s consume more healthcare resources than the under 65s (and the over-85s even more so).

 Second is the ‘Baumol effect’, where the price of services such as healthcare rises faster than that of other sectors, so that even at constant volumes the value share of services in nominal GDP increases. (The Baumol effect applies to most services which the government provides.)

 Third, as they become more affluent people demand more healthcare. That seems a perfectly reasonably decision; as you get richer do you want more knickknacks or a better quality of life from improved health?

However these effects do not seem to have been strong enough in recent years to overcome the government’s demand for spending restraint. 

New Zealand Superannuation (Graph 4.2) 

New Zealand’s universal retirement pension was costing about 5 percent of GDP in 1996/7, fell to 4 percent at the end of the first decade and has begun rising again to near 5 percent in 2014/15. The pattern partly reflects, the raising of the age of eligibility from 60 to 65 years by 2001, some changes in rate of payment relative to wages, and population ageing, with the baby boomers increasingly coming on stream after 2010.

The last point says the population dividend – when maturing age groups contribute to a relative increase in the labour force – is over. The expectation is that, as far as demographers can project, the ageing of the population will continue. Even raising the age of eligibity of the pension to 67, say, will only moderate this effect, not eliminate it in the long run. 

(This spending category does not include spending on the retired’s healthcare nor some supplementary transfers that are classified as social security.)

Other Social Security & Welfare (Graph 4.2)  

Social security transfers were higher at around 7 percent of GDP in the late 1990s but have been falling since and were nearer 5 percent in 2014/15. They show a temporary upturn at the time of the Asian crisis in the late 1990s and a longer one at the time of the GFC but, these aside, there is an undoubted downward trend.

It is out of historical character that spending on social security fell under the Labour-led Government. In part this was due to falling unemployment but more importantly it never restored the benefit cuts which the previous National Government made in 1991, only maintained the real value of benefits, indexing them to prices rather than to, say, wages – in which case beneficiaries would have shared in the prosperity of the era. (There is an argument that the 1991 cuts were overzealous, so that even if benefit levels should not rise relative to other incomes, there is a case for raising the relative level in the short term.)

Social transfers spend as a proportion of GDP stabilises towards the end of Labour’s last term. This is because it introduced the Working for Families transfer for those with children, but beneficiaries were specifically excluded from a significant part of it. Entitlement to the In Work Tax Credit required the recipient to be working at least 20 hours a week and off the benefit. 

Educational Services (Graph 4.2)

Educational Services appear to be modestly rising as a proportion of GDP from 4.8 percent of GDP in 19967/7 to 5.4 percent in 2014/5. There are two complicating effects. One is population shifts which, given the ageing population, suggests that the share should fall. However, like other government services, education is subject to the Baumol effect. 

Additionally, there has been a big extension of funding of Early Childhood Education as a greater proportion of children have been involved. Indeed, three-quarters of the increase in share of GDP can be attributed to this subcomponent (from 0.22 percent to 0.68 percent of GDP compared to a total increase in spending of 0.60 percent). The other big gainer is departmental expenses which increases .021 percentage points. The data base does not indicate why this happened.

A second big effect is the write down of the capital value of student loans due to zero interest policy and defaults in 2005/6, 2008/9 and 2010/10, which causes blips in the underlying trends. These all appear to be related to election promises although this may be timing effect. Arguably, tertiary education is not simply an investment for students but also includes an element of distributional entitlement and a public good in the improved capacity for citizenship. Even so, the practice of primary and secondary schools having to charge private fees to parents did not get a similar treatment; schoolchildren do not vote so there has not been the same electoral targeting of them.

Debt Servicing (Graph 4.2)

Debt servicing fell from 3 percent of GDP in 1996/7 to just above 1 percent in 2009/10 as a result of falling relative debt levels and lower interest rates. After which, the level rises reflecting the additional borrowing to smooth New Zealand through the GFC, flattening out because borrowing was eased back and interest rates continued to fall. In 2014/5 it was 1.6 percent of GDP. 

General Economic Services (Graph 4.2)

General Economic Services cover Transport and Communications, Economic and Industrial Services and Primary Services but exclude Kiwisaver subsidies which are classified with Other (below). Basically they are direct assistance to industry, hovering around 2 percent of GDP. They rose to about 2.5 percent of GDP under the Labour-led Government but have fallen back to 1.8 percent under the subsequent National Government.

The following spending categories are generally smaller and shown in Graph 4.3 (overleaf). Only ‘substantial’ variations are mentioned here. 

The jump in Law and Order in 2006/7 applied to all of Police, Ministry of Justice and Department of Corrections. 

Even after the removal of Tax Write-downs and Official Development Assistance Core Government Services shows some jumps reflecting (mainly) variations in Non-Departmental Expenses. The underlying cost of running the bureaucracy has remained a fairly constant proportion of GDP. 

Defence spending has been a near constant proportion of GDP with perhaps a tendency to slowly fall off. 

<
>

Source: New Zealand Treasury

Heritage, Culture, Recreation and the Environment (from which is excluded the costs of the Emissions Trading Scheme (ETS) – in Other) grew steadily from 0.27 percent of government spending in 1996/7 to 0.52 percent in 2009/10 and then has largely stagnated. The same calculation that was done for Healthcare would conclude that if the trend before the GFC was continued, the group would have had an extra $1b spent on it in 2014/5, or almost three-quarters more. (This may underestimate the squeeze since the sectors have also been in receipt of lotteries monies which are also drying up.) Perhaps about half would be spent on the environment and half on the other items in the group. 

The content of Housing and Community Development expenses seems to have changed over time. It is left here separately to show it is small, indicating, perhaps, that it is not a high government priority despite the public thinking it important. 

The Other category is a rather mixed bag, including for these purposes, Kiwisaver subsidies, tax write-downs, Official Development Assistance and pay-outs under the Emissions Trading Scheme.

Summary

Despite the public’s desire for more government spending there has been little increase in the aggregate level relative to GDP over the last 20 years, except for a slight rise immediately after the GFC followed by a fall. It is now lower that it was at the end of the Bolger and Shipley governments, but a fraction higher than under the early Labour-led Government of Clark and Cullen. 

Insofar as it is slightly higher in 2014/5 than 10 to15 years earlier this is mainly because of rising spending on New Zealand Superannuation and items in the other category such as Kiwisaver and pay-outs under the ETS (both the latter were introduced by the Labour-led Government). 

The decision on how much the government is to spend is one for the political process led by the government of the day. Even so the current government’s decisions seem to contradict the popular preference indicated by surveys.

The next chapter explains how taxation preferences cap the willingness of a government to spend more. To give an order of magnitude of the issue it will consider increasing spending by a further $4 billion a year (or about 1.6 percent of GDP) in 2014/5 conditions. A possible derivation of the figure is as follows:

Additional spending

• Health care $1.4b
• Heritage, Culture, Recreation and the Environment $1b
• Social Transfers $1b. 
• Other expenditure items such as Educational Services $0.6b.

The first two components come from the above calculations, the third is the Child Poverty Action Group’s estimate of the most urgent demand to reduce child poverty, and the fourth is a residual to take the total up to the $4b indicator. Readers are welcome to think of an alternative and scale the following discussion accordingly. 

Appendix to Chapter 4.

The New Zealand Treasury has a standard set of categories for Core Crown Expenses. The ones used here are slightly different reflecting different analytic needs.

1. Health: (as defined by Treasury) 
2. New Zealand Superannuation: (as defined by Treasury)
3. Other Social Security and Welfare: Social Security and Welfare (as defined by Treasury) less New Zealand Superannuation
4. Education: (as defined by Treasury) 
5. Debt Servicing: Finance Costs (as defined by Treasury)
6. General Economic Services: Treasury categories of Transport and Communications, Economic and Industrial Services and Primary Services less KiwiSaver (includes housing deposit subsidy).
7. Law and Order: (as defined by Treasury)
8. Core Government Services: Treasury categories of Core Government Services less Tax Receivable Write-down and Impairments less Official Development Assistance.
9. Defence: (as defined by Treasury)
10. Heritage, Culture, Recreation and the Environment: Treasury categories of Heritage, Culture and Recreation and the Environmental Protection less Emissions Trading Scheme.
11. Housing and Community Development: (as defined by Treasury)
12. Other: The residual including Treasury categories Other, GSF Pension Expenses, Tax Receivable Write-down and Impairments, Official Development Assistance and KiwiSaver.

5. HOW TAXATION FITS IN

Taxation is the converse of Government Spending for its function is to reduce private spending in order to make room for government spending. 

Earlier it was noted that the government could borrow to increase spending temporarily but unless that spending was an investment which gave a direct return – very little of Core Crown Expenses are – then at a later stage the additional borrowing would have to be recouped through additional taxation. Basically a government spending more on its core services than its tax revenue shifts the burden of taxation through time; it does not eliminate it. Note, however, that some government spending – most prominently on healthcare, education and child services – is a social investment any public return of which involves tax payments.

By international standards the burden of general government revenue (which includes revenue of lower tiers of government and other central government revenue as well as that from taxation) in New Zealand relative to GDP among affluent economies is not high. The proportion reported in the appendix to this chapter suggests that the New Zealand rate of 39.7 percent of GDP is below the OECD country average (42.4 percent) in the 2013 year. To give an indication of the spread, recall that the previous chapter suggested that New Zealanders might contemplate an increase of government spending by 1.6 percentage points of GDP. That would put New Zealand still below the OECD average and well below such luminaries as the Scandinavian economies, France and Germany. 

The comparisons involve a number of assumptions which tend to make them superficial rather than exact. For instance, New Zealand could reduce its level by netting off the income tax on New Zealand Superannuation and on social security benefits and the GST on government spending and turning working for families tax credits into negative income taxes. The current conventions exist for very good reasons but other countries’ conventions need not be so scrupulous. An additional factor in comparisons is that the New Zealand tax system with its few exemptions is usually considered more efficient compared to those of many other countries.

An indication of how problematic the measure is, is that a report from Australia’s Grattan Institute claims that if compulsory super contributions are included, Australia’s proportion would be slightly higher than New Zealand’s.^[1] (Because of the depth of the trans-Tasman labour market, New Zealand’s comparison with Australia is especially important. However, it would be wrong to simply compare tax rates, since the scope and quality of government services and the living environment generally are relevant when a family is making a decision to migrate – job opportunities may be the single most important factor.)

Perhaps surprisingly (to this economist anyway) there is no compelling evidence that low rates of taxation are associated with high economic growth among affluent economies. What counts in economic performance is not the size of government, but, rather, its quality.^[2] (The surprise at this conclusion reflects the substantial body of economic theory which suggests that taxes discourage labour and investment incentives. No doubt they are there, but there appear to be many other more important factors in determining the economic growth rate.)

In summary, if New Zealanders want more government spending, a rise in tax levels would not be out of line with similar affluent-country levels, although care would be necessary to ensure tshe new regime did not to have too great a distorting impact. Indeed, there are even measures which could increase its efficiency. 

Options for Raising Taxes: Removing Anomalies

Before considering the standard recipe of increasing income tax and/or GST rates, other possibilities should be considered. Obviously tax loopholes – or poor coverage – should be eliminated whatever. Here are some examples.

While trusts are an integral part of the management of property, they are also used for tax avoidance since their top tax rate (currently 30 percent) is below the top income tax rate on persons (33 percent). An obvious change would be to tax them in a manner similar to that for companies. The trust would pay the top tax rate on all its income but it would pass on credits for tax it has paid to the trust’s beneficiaries when it paid out. The result would be that those beneficiaries on low incomes (and hence tax rates) would have some of the tax reimbursed and they would often receive more net income from the trust, without compromising the use of trusts for other (non-tax) avoidance purposes. 

There is a widely held view among economists that all returns on capital should be taxed at the same rate in order to reduce distortions in investment decisions from the tax system. With some exceptions, capital gains are not taxed. This would suggest that at the very least income tax should be extended to cover capital gains from share transactions and from all investment housing. (The primary residence need not be included because the apparent capital gain made when selling it is offset by the capital loss on purchasing a replacement house.)

The failure to levy GST on low-valued imports of goods and services is partly an issue of compliance costs. Even so, the exemption threshold seems too high. 

While these changes might generate some useful revenue gains in total they would be small in comparison to the size of the demands for additional government spending. 

Options for Raising Taxes: New Taxes

Excise duties on tobacco and alcohol demonstrate that properly targeted, low-compliance-cost duties can contribute to improving the nation’s health. That does not mean that all such levies will be equally effective. 

In recent years the New Zealand government has outlayed an average of $215m a year to purchase carbon credits. The amount is likely to increase as the world takes global warming increasingly seriously. It is not obvious that this should be paid out of general taxation, when a carbon tax would both cover the outlay and encourage reductions in carbon emissions. The economic advice would be that while it be levied on producers it should be targeted at consumers and so also be levied on imports whose production involved emitting carbon, while any tax might well be rebated on exports leaving the importing company to address its consumption. 

There is a strong case for a Financial Transactions Tax. However, it would be pointless for New Zealand to introduce it by itself since it could be easily evaded by shifting the transactions offshore. Some members of the European Unions are currently exploring the imposition of an FTT. If it proceeds, New Zealand should be a fast follower, perhaps within a consortium of non-EU countries which would implement a jointly compatible one. (Some estimates of the revenue gains from an FTT are implausibly large.)

Most members of the 2009 Tax Working Group (TWG) supported the introduction of a lowrate land tax. Their fundamental reason was not very different from that proposed by Henry George and other single taxers. The supply of land cannot change – especially given the concerns of the TWG; it is not globally mobile. Therefore, a well-designed land tax will have a negligible distortionary effect on use incentives. Unfortunately, the value of land is deeply imbedded in a regime in which there is no land tax, and owners have made decisions on that basis, especially borrowing to purchase it. The initial impact of a land tax where there is heavy debt on the land is difficult to evaluate. For instance, the TWG acknowledges that the price of land will fall if a tax is imposed; that would raise the debt-to-land value, in some or many cases, to dangerously high levels. (Astonishingly, the TWG does not discuss this; there was no one with farm expertise on the group.) Given the central role that the farm industry plays in the New Zealand economy it seems wise to proceed with a land tax with caution. It is unlikely to have any effect in the time with which this report is concerned. (There is a proposal to tax land owned by foreigners. Whatever the reasons for doing this, it will not raise a lot of revenue.)

The New Zealand government is already in receipt of some revenue from resource levies but it does not seem to have a comprehensive approach to them. For instance, most water usage is not levied. 

In summary, there are a number of prudent ways that New Zealand government revenue could be increased by extending the tax regime. But collectively they would not contribute the sort of sums the public seems to require for its public spending ambitions.

Are There Spending Areas Which Can Be Cut Back? 

It is easy, and therefore common, to propose cuts in existing government spending. Of course the government should seek to improve the efficiency of delivery of its services. It does; it has been doing so for as long as anyone can remember. There is no reason to believe it is less efficient than private sector equivalents.

It is well to remember that excessive downward pressure on some government spending programmes can compromise the quality of its delivery – especially in the long run. Additionally, it may raise transaction costs and/or increase effective inequality, especially when the effect is to shift costs onto households. 

More relevant is cutting or greatly modifying spending programmes. This can be controversial for a programme valued by some may be resented by others. There may be a case for better targeting of some programmes. This will be easier if the overall system of redistribution is fairer, but there remains a danger of high effective marginal tax rates which will be a disincentive to people improving themselves. (They tend to be a burden especially on the poor.)

A widely held view is that the age of entitlement for New Zealand Superannuation should be raised above 65 years, reflecting the rising longevity of the elderly. Like the earlier increase from 60 to 65 the change should be well signalled and incremental. (Because many affluent economies did not address this issue, they implemented abrupt and unexpected increase in the age of eligibility and reductions in levels of payments during fiscal crises.) There would be some reduction in government spending from such a measure but because it would be phased in, the gains would not great in any immediate term.

The Fundamental Conclusion

Major increases in government spending cannot be solely financed by the above measures. Instead it would be necessary to increase the rates of GST or income taxes. An increase in the GST rate would not change the after-tax income distribution much so it would not address the widespread concerns about income inequality. If the public wants a less unequal income distribution they would want the additional funding to be raised by higher income tax rates – eliminating some of the anomalies/loopholes would help. 

It is not proposed to set out a new income tax structure (including how it might interact with the transfer system). But two points could be made about it.

First, reducing income inequality requires that the tax system be more progressive, that the rate hikes should be higher at the higher income end than at the low income end.

Second, while the rich would pay relatively more if the redistribution system of taxation and transfers is more progressive, much of the burden of additional revenue raised would be paid by those in the middle of the income distribution. For many of the increases in spending they will be the main beneficiaries – especially over their life cycle. This argument is detailed in the case of healthcare in the next (final) chapter.

Table 5.1 General Government Revenues as a Percentage of GDP (2013)

 % of GDP

<
> Denmark 56.0 Norway 55.4 Finland 55.2 France 53.0 Sweden 51.9 Belgium 51.5 Austria 49.6 Italy 48.0 Greece 47.8 Hungary 47.3 Portugal 45.2 Slovenia 45.2 Netherlands 44.5 Germany 44.5 Luxembourg 44.4 OECD Average 42.4 Iceland 42.0 Czechia 41.2 United Kingdom. 39.8 New Zealand 39.7 Turkey* 38.6 Estonia 38.5 Slovakia 38.4 Poland 38.2 Canada 38.0 Spain 37.5 Israel 37.2 Ireland 34.9 Australia 34.0 Japan 33.9 Switzerland 33.6 USA 33.1 Korea 33.1 Mexico  24.5

Source Government at a Glance 2015 (OECD) Last updated: 06-Jul-2015  * Turkey 2012, not included in average.  

6. SHOULD THERE BE MORE GOVERNMENT SPENDING?

Chapter 1 showed there can be good reasons for government funding some economic activities instead of leaving the funding to private decisions. Chapter 2 reported that there seemed to be a substantial public desire for additional government spending in some expenditure areas, while Chapter 3 showed many appropriate sectors where government spending has a long history of increasing as a proportion of GDP. However, as Chapter 4 illustrated, this has not been the pattern in recent years with many spending areas falling or stagnating as a proportion of GDP rather than continuing to rise as they have in the past. This almost certainly reflects a reluctance of the government to raise taxes. Chapter 5 concluded that any emphasis on additional revenue has to be from income taxes, given that there are limited alternatives and widespread public concerns about income (and other economic) inequality.

Because of the way the preferences for government spending are surveyed it is unclear whether individuals who want more government spending are also willing to accept the substantial increase in taxation that their spending preferences require. The current government may well think that a majority does not want to make such a trade-off; it is a matter of record that the public has not really been presented with the trade-off in recent general elections.

Instead, much of the public discussion on areas of government spending has focussed on the demand to increase it in one area without mention of the tax consequences; perhaps the message is that other spending areas should be cut. Conversely, tax policy discussions which favour lower tax rates equally rarely mention that the consequences of the recommendations would be lower government spending.

This report has emphasised that there is no technical argument which favours more or less government spending. That is a political decision. It has been suggested that the reasonable public demands for more government spending may amount to near $4b a year, which would be a 5 to 6 percent increase in current levels of government spending and, perforce, a similar increase in tax revenue. Even so, such an increase in government spending would move New

Zealand’s public spending as a proportion of GDP from just a little below the OECD average to close to it.

It is beyond the scope of this report to argue for additional government spending for every item that might arise. But it might be useful to illustrate such arguments with the case of healthcare. (The Child Poverty Action Group has argued in some detail for substantial increases in child support to reduce child poverty, which would also reduce income inequality.^[3] The vast majority of the measured poor are children and their parents/caregivers. Generally, however, sector advocacy has been less detailed.) 

Chapter 4 showed that in the last few years public spending on healthcare has fallen behind its long term trend relative to GDP; for the last year for which data is available, the annual deficit amounts to about $1.4b. This will affect all spending areas. 

This chapter has little to say about population-based healthcare expenditure, except that it is a pure public good and that a failure to provide sufficient results in higher requirements for personal-based healthcare expenditure.

Personal-based expenditure on healthcare is not a pure public good since each treatment benefits an individual (and their associate); it is neither non-excludable nor non-rivalrous. The case for public provision rests on the fact that private provision can be inequitable, but even if adequate private insurance was available (it isn’t; even the US requires public assisted Medicare for the elderly and disabled) transaction costs would be high and the system would be inefficient.  

Government expenditure on healthcare is about four-fifths of total health spending in New Zealand (it was 83 percent in 2010 – the latest figure available). This may underestimate effective healthcare since some of the private spending is on treatments which medical science would judge as not evidence-based. The proportion has probably fallen since 2010, given the relative fall in government health funding.

There are no measures of the degree to which any relative rise in private funding has resulted in increased inefficiency from, say, private patients choosing less effective treatments or resources being deployed on lower priority treatments. But even where there have not been such decreases in inefficiency, there have been consequences to the health of the nation. 

These can be traced by dividing the population of those who miss out on public healthcare into two categories: those who did not receive treatment and care (either full or partial) and those who bore the cost of the treatment and care privately.

1. If the health care was not provided, the consequence was that individuals in need died earlier or were in greater pain and discomfort than if there had been greater public funding. 
2. If they have purchased the required healthcare themselves – many cannot afford to – then

0%20Our%20Children%20Our%20Choice%202014.pdf

they are poorer.

There can be wider consequences to the government. The failure to provide adequate healthcare for children may not only compromise their health but limit their development and opportunities. Thus what might be attributed to a failure of the education system may, in fact, be due to an inadequate healthcare system. The failure to provide adequate mental healthcare has the consequence that there are additional pressures on corrections facilities increasing pressures on Law and Order spending. Inadequate healthcare can also result in a person being on benefit which would be avoided were there appropriate treatment, adding to social security spending. 

The failures can also fall on the private sector. Inadequate residential care for the needy elderly not only impacts on the quality of life of those without other means of support, but the families of those in need may find themselves under pressure. This is a rising concern given the ageing population. Not only may those in their fifties find themselves having to support ageing relatives but the retired elderly may find themselves in a similar position having to support very elderly family. (There are parents under pressure supporting children in medical difficulties and vice versa.) 

The difficulty with private provision is that, unlike, say, food, the incidence of healthcare need is erratic and highly variable. In principle, it could be met by (voluntary) private health insurance but that has proved unsatisfactory on a number of dimensions where it has been relied upon and has to be supplemented by some public support; the resulting system has proved to be very inefficient. 

This is not to rule out a compulsory national insurance system, but to be effective that requires a ‘unitary funder’ while a levy is a tax by another name. Nor does the analysis rule out private healthcare insurance topping up a high quality public healthcare system. At issue is the adequacy of the public system. The public seems to think the current one is inadequately funded and therefore overall supply of healthcare is inadequate.

Because the demand for personal healthcare is erratic with high variability of outlay, the public places some premium on an across-the-board public delivery. (Note that these characteristics are true at any point in time, but they are also true – to a lesser extent – during the lifetimes of individuals.) In effect, they are expecting the state to provide a system of mutual (or social) insurance.

This has an important implication for the tax regime which funds the health system. There is no requirement for the scheme to be redistributive by income (and practically it would be difficult to be redistributive by age and gender, which would be relevant in a purely private insurance system). Any increase in taxation to fund additional public healthcare should therefore be imposed on all taxpayers. 

This need not be necessarily true for the other items for government spending. If it were to prove that heritage, culture and recreation were demanded more by the affluent than the poor (possible, since there is likely to be a higher income elasticity of demand for many of these services) there would be case for tipping the tax system so the affluent paid a higher proportion of the increment.

This requirement is even stronger when measures to eliminate poverty and reduce inequality are implemented. Making the tax system more progressive would reduce inequality. (It should also be noted that income is not necessarily a good indication of affluence but needs to be adjusted for the number who depend on the income and for other circumstances.) 

Of course, any package funding an increase in government spending would have multiple spending objectives so there will be considerable challenge in the design of the package. However the indications from the above survey and many others are that the public are demanding greater progressivity in the tax system.

One further consideration is that it would be unwise to implement all the additional spending in a single year. For instance, as much as public healthcare is severely constrained from providing for all that is demanded of it, the supply side could not cope with an additional single injection of $1.4 billion a year. It would need to be phased in, in addition to the annual injections already projected.

This need not apply for measures to reduce poverty to the same extent, since additional consumption from the additional income of the poor would offset the reduction in consumption from the lower incomes of those higher in the income distribution. 

Conclusion

There are good reasons that some services should be funded and provided by the government because private markets do not always give the economic outcomes the community desires. A second consideration is that the level of economic inequality generated by the private market may be unacceptable and public transfers necessary to reduce it. 

There is evidence that the public wants more of these services and transfers. To implement their demands they must also accept they will have to pay higher taxes. The designing of a tax regime to meet the public’s desires is not easy. 

However, the first step must be a clear direction that the public’s wishes should be pursued.  

Acknowledgements:  Among those who have commented on the paper are Julienne Molineaux and Susan St John. All have been valued

<
>

[1] http://grattan.edu.au/news/budget-repair-and-the-size-of-australias-government/

[2] I. McAuley & and M. Lyons (2015) Governomics.

[3] Child Poverty Action Group (2014) Our Children, Our Choice.

<
><
> http://www.cpag.org.nz/assets/Publications/1410063-