Public economics in an age of austerity1

A B Atkinson, Nuffield College, Oxford

1. Public economics and austerity

Wondering how to start this Lecture, I remembered that one year the Norwegian Statistical Yearbook (Statistisk Årbok) contained a table in which there is a cell with only one person – and that person is Agnar Sandmo. (I
should hasten to say that I am not divulging any secrets gained from my work in the Research Department of Statistics Norway, since he told me this himself.)
The fact that Agnar stands alone seems very appropriate. That he is a singular
economist is widely recognised. He stands out as a commanding figure for his manifold contributions to economics over a wide variety of fields. He is always turning his fertile imagination and creative powers to new topics. What is
more, and what is extremely rare among economists, is the fact that he can write. He has a stylish pen, nowhere more evident than in his recent Economics evolving: A history of economic thought (Sandmo, 2011).
One field has been Agnar‟s particular concern: public economics. Hence my choice of subject today. I first encountered the term “public economics” when a student in the mid-1960s through reading the book with that title by Leif Johansen, whom I got to know a few years later, and who generously agreed to join me in 1972 in founding the Journal of Public Economics. In
preparing this lecture, I was looking back at the first issue of that journal, forty years ago. In addition to noting that it contained an article by one A Sandmo, on a topic relevant to my talk – public infrastructure (Sandmo, 1972) - I was struck by the extent to which the authors in that first issue have consistently been concerned with the application of their analysis to the pressing problems of current policy. The authors included one future chair of the US Council of Economic Advisers and one future Governor of the Bank of England.
It is with the application of public economics to policy that I am concerned, specifically its application to the current austerity programmes and fiscal consolidations. This is certainly topical. In 2010, Merriam-Webster's Dictionary named the word "austerity" as its "Word of the Year" based on the number of web searches this word generated. According to the UK Prime Minister, we have entered a new age of austerity, which he characterized as putting an end to “excessive government spending”. All around the world,
there are governments adopting or are being required to adopt "austerity programmes”. But these programmes are highly controversial. They divide

1 Revised text of the fifth Agnar Sandmo Lecture given on 12 January 2012. I would like to thank the Department of Economics, Norwegian School of Economics and Business Administration, for their invitation to give this Lecture and for their hospitality.

politicians, they divide countries, and they divide governments from many of their citizens, as shown by riots and demonstrations in the street.
There is therefore much to debate. In this debate, there are many economists clamoring for the ear of the public and of policymakers. One hears a great deal from macro-economists and from economists specializing in financial markets. Yet one hears relatively little from experts in public economics. This seems to me surprising. Is not public economics concerned with deficits and taxes and government spending? Is not the national debt one of the essential topics in any course on public finance? Should not students be
taking courses in public economics? To take just some examples, cannot public economics help us think about the following questions:
A. The balance between tax increases and spending cuts is typically
25%/75%. Are these the right proportions?
B. Tax rises in a number of countries take the form of raising VAT rates.
Should we also consider broadening the VAT tax base?
C. Past fiscal consolidations have fallen disproportionately on cutting government capital investment. Does this simply pass on the burden to the next generations in a different form?
D. Should we cut back on personal services for the elderly or on youth services?
E. …

The … indicate that there are many more.

The first theme of this Lecture is that public economics can indeed contribute to this debate. I shall argue that we can learn lessons for the design of austerity programmes from a number of pieces of public economic analysis – some old and some recent. One of the recent pieces is the major review of taxation in the UK, carried out by the Institute for Fiscal Studies and chaired by Sir James Mirrlees, to which I shall be making particular reference (Mirrlees,
2010 and 2011). At the same time, I shall also be arguing that there are
significant limitations to the analysis. The second theme of this lecture is that current public economics is too blinkered, and that new departures and new ways of thinking are required.
I start in section 2 with classical public finance, epitomized by A C
Pigou, whose book A study in public finance (1928; 1947) can still profitably be read today, asking how such an analysis of the optimal design of taxation and public spending is affected by the onset of austerity – something that Pigou did in fact consider. A crucial conclusion, developed in section 3, is that there are
alternatives. Not all austerity packages are the same. But this is only part of the story. In section 4, I suggest that the underlying economic framework is seriously deficient, and that the policy recommendations may flow as much from the way the model is specified as from the economic logic. The models employed allow precision but at the cost of leaving out important considerations, such as imperfect competition discussed in section 5, the

2

return of inheritance discussed in section 6, and the non-steady-state dimension of intergenerational justice discussed in section 7. As I conclude in section 8, I am both advocating the application of the principles of public economics to the design of austerity programmes and challenging its exponents to go further.

2. Classical public finance

The classic public finance design problem – posed explicitly by Pigou to Frank Ramsey in the 1920s – is to choose the tax and spending parameters to maximize social welfare subject to the government budget constraint. This is set out formally in the box. Vi refers to utility of person i, where this utility depends on the taxes and spending, as well as other variables, such as their income, not shown explicitly. For Pigou, social welfare was constituted by the sum of the individual utilities, since he was a thorough-going utilitarian. In the slide, I have kept my options open, and have therefore written a function W{}
of Vi, allowing other possibilities apart from simply adding utilities (I return to this at the end of the Lecture).

Social welfare is maximized subject to the government budget constraint. Revenue from taxes has to be enough to cover the costs of government spending on goods and services (roads, schools, defence, etc.) and on social transfers (pensions, child benefits, etc.), and to service the national debt. If we assume that the government aims to keep a constant ratio of debt to national income, and if national income is growing at 2 per cent, then it can increase debt by 2 per cent, so that the amount needed to service the debt is the rate of interest less 2 per cent.
The government budget constrain now preoccupies the media. The
public finances are said to be out of control. Many people are bemused by this, and wonder how we have got into this situation. There are in fact several reasons. First, there is the effect of the world recession. A decline in GDP has the effect of triggering automatic stabilisers: tax receipts fall and social transfers increase. We tend to forget that one of the purposes of the welfare

3

state is to cushion household incomes, and they have in fact proved quite effective. If one looks at the euro area, then from 2004 to 2008, when economies were expanding, total household disposable income grew less rapidly than GDP, but then, when GDP fell by some 5 percent after 2008, household income remained broadly stable for 2 years. But this meant running deficits, and this increased the size of the debt. At the same time, the fall in
the growth rate (or negative growth) meant that less new debt could be issued, so the debt service burden increased. What is more, there were other factors leading to the rise in government debt, notably bailing out distressed financial institutions.
These macro-economic forces mean that the government budget constraint has become tighter and that we face an immediate need for fiscal consolidation. They also mean that we are less optimistic about the economic future. Most OECD economies are below the long-run path they expected to follow. In particular, the outlook for household per capita incomes looks bleak, since they cannot be expected to grow at the same rate as GDP. This has implications that I take up in section 7. I begin however with the immediate pressure on the government budget.
The fact that the government budget constraint has become tighter in many countries means that the marginal cost of public funds has risen. As a result, the optimal level of government spending may be expected to fall. This argument was, interestingly, discussed by Pigou, who said that, after a war, “certain government expenditures, which it used to be worthwhile to undertake, a country may no longer be able to „afford‟” (1947, page 32). This is a genuine “austerity” argument for cuts in public spending. However, the marginal value of tax revenue has also risen, so that there is a case for sharing the adjustment between spending reductions and tax increases.
So a choice has to be made. The government budget can be brought into balance either by cutting spending or by raising taxes. In the UK, the Coalition Government has chosen to achieve fiscal consolidation mainly through spending cuts, as may be seen from Figure 1. The tax line hardly rises, while
the spending line falls sharply. The Office for Budget Responsibility (2011)
figures show a 8.4 percent of GDP improvement in the primary balance between 2009-10 and 2015-16 being achieved by 7.3 per cent cut in spending and 1.1 per cent tax rises. In other countries, the ratio of spending cuts to tax increases is lower, but the Financial Times has referred to a 3 to 1 ratio as being representative (1 December 2011, page 34). Fiscal consolidation is being achieved by a mix of 75 per cent spending cuts and 25 per cent tax rises.
Why a 75/25 ratio? Why are austerity programmes so weighted towards spending cuts? In considering possible explanations, it is important to distinguish those concerned solely with austerity from those that introduce
other considerations. For example, it is possible that the cuts are seen as short-

4

term, and that spending is more flexible than taxation. Certain public spending programmes can be postponed. This would be a purely austerity-based argument.
On the other hand, governments may argue that there is now a reduced willingness to pay for public services. Such an argument introduces another element apart from austerity: it postulates a shift in political preferences. Equally, governments may be arguing that public goods expenditure is carried out for redistributive reasons, justified by the greater weight given to benefits going to the poor, and that they are no longer concerned with making such a redistribution. Such an argument again has nothing to do with austerity.

It is important therefore to understand the reasons behind the adoption of an austerity programme, and the extent to which choices are being made between alternative routes to deficit reduction.

Figure 1 UK Fiscal adjustment 2009-10 to 2015-16

50

45

Spending

40

Taxes

35

30

2009-10 2010-11 2011-12 2012-13 2013-14 2014-15 2015-16

Source: Office for Budget Responsibility, 2011, Chart 4.3

5

3. There are alternatives

Indeed, the classical public finance analysis makes clear that there are alternatives. Not all austerity packages are the same. A ratio of 75/25 can be questioned. A ratio closer to 50/50 would, however, raise the further question as to which taxes should be raised. What is the scope for tax increases?
To illustrate this, I take one of the proposals of the UK Mirrlees Review, which is to broaden the base of VAT. In the UK, VAT is not paid at present on food, children‟s clothing and other important items such as books. The Mirrlees Review proposed bringing these within the tax net. Since the immediate objection is that such a VAT extension would hit disproportionately those on low incomes, the Review plans to use more than half the revenue to raise the income tax threshold and to increase income-tested transfers (and adjust tax rates). They conclude that the overall package would raise an additional £11 billion (0.7 per cent of GDP), substantially increasing the contribution of increased taxation to deficit reduction, in a way that would be distributionally progressive (Mirrlees, 2011, page 227).
This package illustrates the importance of looking at the full range of policies, and not just considering one part of the problem. One has to look at indirect and direct taxes together. It also demonstrates the contribution of public economics. The logic of the Mirrlees Review proposal derives from results in optimal tax theory which turn on the range of tax instruments at the disposal of the government – see Table 1 (line 1). If indirect taxes are expected to do all the work, then the design of the indirect tax structure has to balance both equity and efficiency objectives. In loose terms, one wants on equity grounds to tax more heavily goods and services that are luxuries, and on efficiency grounds to tax more heavily goods and services that do not respond much to taxation (their demand is inelastic). These may point in opposite directions and a balance has to be struck.
But if one allows the government also to employ the simplest direct tax, which is a uniform poll tax, then the role of indirect taxes changes (line 2 in Table 1). Apart from migration in and out of the country, the poll tax does not affect behaviour – it is purely lump sum. From an efficiency point of view, this is how revenue should be raised. It is only if the government is concerned about equity that it should use indirect taxes, which have a higher efficiency cost, since they distort consumer choices. By raising revenue from indirect taxes on luxuries and reducing the poll tax, the government is benefitting the poorer households. Distributional concerns are central. In many areas of economics it is assumed that all consumers are identical – much of macro- economic theory is based on the assumption of identical representative agents

– but such an assumption would make no sense in public finance. The key

6

issues in public finance arise only when we allow for the fact that people have different capacities to pay.
It is the differing capacities to pay, specifically differing earning power, that are the basis for the theory of optimal taxation developed by Bill Vickrey and Jim Mirrlees. In the top two rows in Table 1, I have allowed in effect for a linear income tax, with a threshold and a uniform tax on all consumption (leaving aside for the moment the issue of savings). If we allow for a general non-linear income tax, with rates that vary over different tranches of income, then this allows us to achieve more redistribution. Indeed, under certain conditions, it allows us to achieve all the desired redistribution, and there is no need to use indirect taxes for equity objectives. A uniform rate of indirect tax is optimal. Hence the proposal by the Mirrlees Review for the extension of the VAT base to cover food and other necessities. One caveat which is entered is that efficiency may require taxes to be levied, or subsidies paid, on goods and services that are socially costly or beneficial – or so-called externalities. For
this reason the Mirrlees Review recommends keeping the excise taxes levied on alcohol, tobacco and petrol (2011, page 217). I mention this for two reasons. The first is that the optimal taxation of goods generating externalities is a subject to which Agnar Sandmo has made key contributions (Sandmo, 1975). The second is that the notion of externalities has a wider pertinence than the reference to excise taxes would suggest, and I come back to this later.
It is the result shown in Line 3 of Table 1 that underpins the proposal to move to a broader base for the VAT. Since the result is one that was proved by Joe Stiglitz and myself, back in the mists of time (Atkinson and Stiglitz, 1976), I am naturally pleased to see that it has current relevance. At the same time, as one of its parents, I would like to caution against placing too much weight on the theorem. The “certain conditions of separability” determine a boundary, and we may find ourselves on one side or the other, not just on the boundary line itself. We saw the case of separability (between leisure and all
consumption goods) as providing a reference point, not as describing the real world.

7

Table 1 Optimal indirect taxes under different assumptions about possible direct taxes

4. BUT: There is a need for a more T-shaped economics

This brings me to the BUT part of my talk: the limitations of the analysis. Reflecting more generally on the failures of economics that have
been widely discussed following the financial crisis, I was struck by the parallel with the criticisms that have been made in England of the performance of their rugby team in the 2011 World Cup. One influential criticism is that the coaches were insufficiently T-shaped. They had become specialists, focusing on the
backs or the forwards, or on kicking, or on running, without being able to see the whole picture. They could not join up the back play with the forwards; they could not develop a whole strategy.
It has seemed to me for some time that much the same applies to economics. We have, as Nicholas Stern (2010) argued in his Presidential Address to the European Economic Association, become too specialized. We
have made great progress in deepening – moving the vertical further down – but have tended to lose sight of the wider context and to ignore important
considerations that are not in the model. The different branches of economics are increasing our understanding, but they are not being joined up. We need not just specialists in public economics but also economists who are aware of developments in other fields and who can integrate these developments into their public finance analyses.
Looked at this way, there are obvious reasons to question the economic model on which are based policy recommendations such as that discussed from the Mirrlees Review. The underlying model is essentially one of competitive

8

exchange: it is the Arrow-Debreu economy taught in general equilibrium
theory. Production barely appears. There is perfect competition, so there is no monopoly power and no bank is too big to fail. Information is perfect, so that all prices are known. All risks can be fully insured. No one has to search for a job. It is a world that abstracts from most of the concerns that people have.
The worries that keep people awake at night are largely assumed away.
Now, what of course we have to ask is, not whether these real-world elements are missing, which they certainly are, but whether their absence affects the conclusions drawn. This is where the horizontal knowledge is important. In what follows, I give three examples of where missing elements could affect the conclusions drawn. They are only examples and they are necessarily treated rather briefly, but I hope they are sufficient to demonstrate that one needs to think more about the context and about missing dimensions.
5. Imperfect competition
The Mirrlees Review proposes extending VAT to food. In most countries, the supply of food is dominated by supermarkets, and within supermarkets there is a high degree of market concentration. Surely, the non-economist would ask, the dominance of the supermarkets in grocery sales has implications for the major change in taxation proposed. Is not the argument affected by
the fact that food is largely supplied by oligopolists, with market power, rather than by millions of small shops acting perfectly competitively?
At the very least, market structure may affect the incidence of the tax. The Mirrlees Review states explicitly (although only in a footnote) that “it is assumed that the incidence of the VAT reform is fully on retail prices” (2010, page 301n). Such an assumption is valid where there is perfect competition and constant costs of production, but ceases to be so when these conditions do not hold. A rise in the tax on food may be shifted backwards onto producers. It is also possible that oligopolistic supermarkets may raise their prices more than the tax. There may be either under- or over-shifting, and that in the latter case it is possible that profits may actually increase with the tax rate, as has been shown by Jesus Seade (1985). (For further references, see Myles (1995, pages
361-3).
It would not therefore be surprising if the existence of imperfect competition were to affect the optimal design of indirect taxation. This question has been examined by, among others, one of the authors of the Mirrlees Review, Gareth Myles. As he showed (Myles, 1989), the conditions for optimal indirect taxation (the first line in the earlier table) now include terms which depend on the degree of shifting, and how it varies across industries, and on the extent to which profits are taxed. As had been pointed out by Austin
and Joan Robinson at the time of the monopolistic competition revolution (see

9

Chapter 13 of Economics evolving), the tendency for imperfectly competitive firms to charge more than marginal cost creates a distortion that leads to households consuming less than they otherwise would. This distortion can be corrected by a subsidy – i.e. taxing the good less. The situation is just like that of an externality, which is why I highlighted this aspect earlier. As explained by Alan Auerbach and James Hines, the condition for an optimal tax on an imperfectly competitive industry

“carries precisely the interpretation offered by Sandmo for the [optimal] tax conditions in the presence of externalities. Intuitively, the „externality‟ in the case of imperfect competition is the outcome of the oligopolistic output selection, resulting in the extra mark-up” (2003, page 15).

Seen this way, externalities are much more widespread than the case of the excises on alcohol, tobacco and petrol. In the real world, they arise, to
differing degrees, across many industries. In the Mirrlees Review, the issue is mentioned only in a footnote on page 156 of the Report, to be dismissed. In my view, this is too hasty.
The important conclusion is that the design of the tax structure has to take account of the industrial structure and conditions of production. There needs to be a closer integration between public economics and industrial organization (IO). This applies in both directions. Public economics needs to draw on the developments in IO, moving beyond the Cournot-type models used in the literature just cited. IO needs to take account of the impact of corporation and other taxes. It is evident for example that taxes affect entry and exit, and hence the degree of industrial concentration.
Closer integration across different branches of economics would make our subject more T-shaped. I chose IO as the example, but the same point could be made regarding other fields. A fine example is provided by the paper by Mats Persson and Agnar Sandmo, where they note that the labour market

“deviates from the standard competitive model in a number of

important respects. The latter insight has so far had little effects on the theory of taxation, which still relies heavily, both for positive and normative studies, on competitive assumptions” (2005, page 558).
They investigate optimal income taxation in a “tournament” model where wages are determined not by productivity but by one‟s productivity relative to other workers. As they note, such a model is particularly relevant to the salaries of top executives. It may therefore be a more suitable framework within which to examine the optimal top tax rate – one of the issues addressed by the Mirrlees Review (2011, pages 108 to 110).

6. The return of inheritance

10

The changes in indirect taxes may affect profits. Who receives this capital income? Who are the shareholders in supermarket chains such as Tesco or REMA1000? The answer is that they vary a great deal. In some cases, there are large shareholdings; in some cases, there are many small shareholdings; in many cases there are institutional holdings by pension funds and investment funds, where the beneficial owners are pensioners or small savers.
There are big differences in capital ownership, but are such differences adequately captured by the earlier analysis? The assumption in the model employed was that people only differ in their earning power. There is simply one dimension along which people differ. If savings were only made for life- cycle smoothing purposes, and everyone has the same preferences, then wealth differences for any given cohort will reflect earnings differences, and we do indeed have a one-dimensional problem. The holdings of shares, either directly or indirectly via pension funds, would mirror the distribution of wage
income. However this is not the only way in which people acquire shares. Some people inherit them. This is a second dimension along which people‟s endowments vary – a second dimension along which there is initial inequality. And this second dimension is becoming more important again.2
After a long period when inheritance was declining, it is now increasing again in significance. In France, this has been shown in an interesting recent study by Thomas Piketty (2011) – see Figure 2. He examines the amount of wealth transferred each year, expressed as a fraction of national income. A hundred years ago, this was about 20 per cent, but the amount inherited fell over the twentieth century until it was some 3 per cent in the 1950s and 1960s. Estates and gifts inter vivos then began to rise and total transmitted wealth as
a proportion of national income has been “multiplied by a factor of about 3–4 between the 1950s and the 2000s” (Piketty, 2011, page 1073). Together with Facundo Alvaredo of Oxford, I am working on similar calculations for other countries. The historical record in the UK, looking only at estates – that is
wealth transferred at death – shows that the decline was remarkably similar, although it continued longer – until the end of the 1970s. Since then there has been a rise, although smaller than that found in France (Karagiannaki, 2011).
Has inheritance returned simply because people are thinking more kindly about their children? They may be, but one important part of the story – one not often discussed – is the rise in the ratio of personal wealth to personal income. This is relevant to another possible tax base – an annual wealth tax. In the UK, the introduction of a wealth tax was considered seriously by the Labour Government in the 1970s, but it was rejected. Indeed, wealth taxes have been going out of favour for some time, having been abolished in Austria, Denmark,

2 Agnar has reminded me that the original version of our paper on the taxation of savings (Atkinson and Sandmo, 1980) contained a section on inheritance, but that this was removed following the editorial comments.

11

Finland, Germany, Sweden and Spain. Although, we should note that France has swum against the tide with its Solidarity Tax on Wealth.
Returning to the UK, we should note that the situation has changed very much since the 1970s. The ratio of personal wealth to personal income has more than doubled. At the very least, this changes the arithmetic of introducing a wealth tax. Alternatively, we should – as the Mirrlees Review suggests – look at wealth transfer taxes. The return of transmitted wealth means that we need to look again at its revenue-raising potential. In 1900, estate taxes in the UK were an important source of revenue. Today, as the MR says “the current UK system does not stack up terribly well against any reasonable set of principles for the design of a tax on inherited wealth” (2011, page 360). For this reason, they look to the introduction of a comprehensive lifetime wealth transfer tax. I welcome this, since I proposed precisely that in my book Unequal shares, published 40 years ago (Atkinson, 1972).
Here I want to make a different, and more theoretical, point. With the introduction of inherited wealth, we are moving to a multiple-characteristic model. We are seeking to determine the optimal tax function T(y,z), where y is earned income and z is inherited wealth. In this context, it is useful to ask
how, say, the optimal income tax rate is changed when individuals now differ in the second characteristic. If there is greater inequality in inherited wealth, does this lead us to tax inheritance more heavily and earned income less
heavily? But the multi-dimensionality also introduces a new element, which is the extent to which we can operate separate progressive taxes on earned income and on inherited wealth, or whether they have to be jointly taxed?
My conjecture is that the answer to these questions depends, among other things, on the degree of correlation between the two dimensions. It might be natural to start from the case where they are independent, but there is no reason to suppose this to be realistic. In the classical class model where
some people, workers, have wages and no capital, and others, capitalists, have capital but no wages, the correlation is minus 1. At the other extreme is the case where the amount inherited is determined by the parent‟s earnings and earning power is perfectly correlated across generations. The most probable case is one where inherited wealth and earnings are correlated but less than perfectly. In that case, in order to tax them appropriately we may need to take account of both dimensions at the same time. This gives another sense to the separability issue referred to before: the ability to assign particular
instruments depends not only on the separability in the utility function (as in the Atkinson-Stiglitz Theorem) but also on the degree of linkage between the
marginal distributions.

12

Figure 2

Source: Piketty (2010, Part II, Table B1, column 12).

13

7. Inter-generational justice

The taxation of inherited wealth raises the obvious question – on whom does the incidence fall? This brings us to the important issue of inter- generational justice, an issue very relevant to the design of austerity programmes. Much of the rhetoric of fiscal consolidation is concerned with the national debt as a burden on future generations. Many years ago, President Eisenhower said that

“I do not feel that any amount can be properly called a „surplus‟

as long as the nation is in debt. I prefer to think of such an item

as „reduction on our children‟s inherited mortgage‟” (State of the

Union Message, January 1960).
Yet in his next, and last, State of the Union Message in 1961, he recorded proudly that he had been responsible for the “largest public construction

programme in history” (the Inter-state highway system) and many other major

public investments. The children and grandchildren are driving along those roads today. It is not just the national debt. The next generation may be worse off because we leave them less in the way of public assets. Austerity programmes may reduce debt but also mean deterioration of the public infrastructure.
One lesson of the public economics literature on the national debt is that we have to look at the full picture. We pass on to the next generations:
National debt;
State pension liabilities; Public financial assets;
Public infrastructure and real wealth; Private wealth;
State of environment; Stocks of natural resources.
We need to look at the overall balance sheet, where assets as well as liabilities are taken into account. This does not however mean that the position is a healthy one. If we consider the difference between the assets of the state and the national debt, expressed as a percentage of the total national wealth, then in the 1950s the net worth of the state was negative, but it was becoming less negative, and turned positive in the 1960s (Atkinson, 2011). But we need to take account of other liabilities. I was taught by James Meade that the state needed to have a positive net worth, measured this way, since no allowance had been made in this series for the obligations to pay state pensions. If we allow for the estimated value of these liabilities – that is the rights already earned, the net balance was close to zero at the end of the 1970s. (Even at
that time, though, the balance was not positive when account is taken of the pensions of public sector workers.)

14

The direction of change since the 1970s has however been in the wrong direction – and it is the long-running decline in the non-pension net worth that is the source of the UK‟s fiscal problem. In effect the process of privatisation, with the proceeds used largely to fund tax cuts, transferred wealth from the state to the personal sector. We saw that it was at the end of the 1970s that personal wealth began to rise faster than income. The worsening of the public balance sheet is the other side. Personal wealth has risen faster than national wealth since the 1970s because, in effect, assets have been transferred from the public to the private sector. We are passing on more privately to the next generation but less publicly.
Reversing this pattern can be achieved not only by reducing the national debt, but also by increasing public assets. This brings me back to Agnar‟s paper in the first issue of the Journal of Public Economics and the subject of intermediate public goods. Public infrastructure investment tends to suffer particularly in times of austerity. In their analysis of 32 episodes of lasting and significant budget consolidation, Balassone and Franco (2000) found that in 25 cases the ratio of public investment to GDP had decreased, and that in 23 cases investment fell more than current outlays (not including debt interest). This is certainly the case in the United Kingdom. When I began studying economics in
1964 the UK was investing about 10 per cent gross, and around 6 per cent net, of its GDP in public investment. This continued until the mid-1970s when in the
1976 Sterling Crisis the UK government was forced by the IMF to make major
cuts in its public spending. At the time, there was much debate about the extent to which fiscal adjustment should be made via higher taxes or by lower spending. It was spending that bore the brunt, and capital spending was particularly affected. This, apparently short-term, measure was followed by a long-run decline in public investment. In part, but only part, this decline was associated with the denationalization of a number of industries such as electricity, gas and telecommunications, and with reduced investment in housing. But a significant part reflected reductions in investment in core public services, such as health, education and transport (Clark, Elsby and Love, 2001). The state still needed to invest, but by 1997 the level of net investment was close to zero. The Blair-Brown Government reversed this trend and net investment reached 3½ per cent in 2009, but under the cuts initiated by Labour and taken further by the Coalition public investment is due to fall back to
under 1½ per cent. A difference of 2 per cent, which is over a quarter of the total fiscal adjustment, would build each year 60 major hospitals, refurbish
2,500 schools, or build a thousand miles of motorway.
It is of course possible that hospitals, schools and motorways contribute less to economic growth than a reduced national debt, but this is a debate that needs to be held. There is no presumption that cutting back on public sector capital formation is the right choice.

15

Moreover, the pattern of current spending can have significant inter- generational effects. A major criticism of the UK austerity programme is that the cuts fall more heavily on the young than on the elderly. As has been shown by John Hills, the plans imply major changes in the composition of spending, with increases in the share of health and pensions but a decrease in the share of education and non-pension transfers (Hills, 2012, Table 4). There have been major reductions in spending on youth programmes, the abolition of the
educational maintenance allowance, which financed young people to stay on at school, and large rises in university fees. At the same time, there have been smaller cuts in care for the elderly and, as yet, no scaling back of the programme for free bus travel for those over 60.
The assessment of the fairness of such a pattern of cuts depends on one‟s value judgments regarding inter-generational equity. That is true, but it also depends on the expectations that one holds with regard to future living standards. Here it is important to remember that we are not in a steady state. Different cohorts alive today have had very different experiences. The oldest lived through the Great Depression and the world war (some even two world wars). In contrast, my working lifetime has been affected neither by war nor by depression. It is not clear that my children‟s generation will be so fortunate, and for those entering the labour market today face even more gloomy prospects. As I said at the outset, I believe that we have to take a less optimistic view about future consumption growth. Even if growth of GDP is resumed at past rates, a substantial part of the additional resources is going to be required to meet the needs of an ageing population and to offset environmental damage. Household real spendable incomes are likely to
increase at a slower rate than GDP.
This means that future generations are going to be less well-off in terms of consumption than we anticipated in the past. This in turn affects our view of inter-generational equity. This may be seen immediately in terms of the social discount rate, where a major component is the expected growth rate of consumption per head (in the formula δ + εg, where δ is the catastrophe risk ,

g is the growth rate and ε is the elasticity of marginal utility). A slower expected rate of growth means that the discount rate is smaller, and we give greater weight to the future. For example, using the kind of parameters applied in the UK by HM Treasury (2011), a growth rate of 2 per cent per year implies that we give a weight of around a third to consumption in a generation‟s time, compared with consumption today, whereas with a growth rate of 1 per cent, we would give a weight close to one half. It makes a big difference.

The implications for the design of austerity packages is that, whatever weight we applied in the past to the benefits of different generations, the fact that we now have to take a less optimistic view about the future means that
we should shift the weight in favour of younger cohorts. Two immediate, and

16

quite obvious, consequences are that we need to prioritise services for young people, from early years interventions to youth programmes, and that public investment for the future should be safeguarded. These are obvious – one might say blindingly obvious – conclusions, but are the reverse of what is being done
in the austerity programme in the UK, where public investment is scheduled to
fall by 2 per cent of GDP and it is services for the young that are being cut.
A third implication concerns the tax side of the equation: there should be higher taxes on better-off members of the older generations. Here there seems to be a direct conflict with the Mirrlees Review in the UK, which has actually proposed cutting the social security contributions paid by older workers, by ending them at age 55 rather than at pension age (65 for men), and by raising the tax threshold for the same age group. The rationale is “to improve economic efficiency by minimizing work disincentives for those most responsive to them” (2011, page 121). Such an argument has been made favoring lower taxes for both younger and older workers by Kremer (2001), but he notes that the elasticity argument must be tempered by distributional considerations, where the earnings of older workers are likely to be more highly correlated with their permanent income than in the case of younger workers. In the same way, the recent literature on age-dependent taxes, by Weinzierl (2011) and others, points to the optimal tax rate at any income level
rising with age. To these arguments, I am adding a further consideration: inter- cohort justice. The case for higher taxes on older persons today becomes even
stronger when we take account of the inter-generational inequalities in a non- steady-state world. The shift in our expectations regarding future living standards means that there should be a shift towards taxing more heavily the better-off members of the older cohorts. I would go in the reverse direction from the Mirrlees Review. It seems to me unjustified, for example, that people, like me, who continue working after 65 are exempted from social security contributions. I would remove this exemption.

8. Exclusions and conclusions

There is much that I have not covered in this lecture. I have earlier criticized public economics for not having taken sufficient account of advances in other branches, such as industrial organization and labour economics; in the same way I am concerned that public economics has taken insufficient account of developments in other disciplines, notably philosophy and psychology. In particular, public economics remains too rooted in utilitarianism. Utility plays a dual role in the literature on the optimal design of public policy. The government budget constraint depends on the impact of taxes and spending on the decisions of households and firms. The decisions of the former are typically assumed to be based on maximizing a utility function. It is the same utility function that enters the social welfare function. Both of these are open to question.

17

The modeling of individual decision-making has been particularly questioned in the recent literature on behavioural public economics. As it has been well put by Peter Diamond,

“In standard modelling, we assume consistent behaviour across

economic environments, captured in preferences that are defined only in terms of commodities acquired (absent externalities). One of the key messages of behavioural economics is that context
(also referred to as situation) matters in ways that are not recognized in standard modelling” (Diamond, 2008, page 1859).
This is an active and very interesting area of research, and one that is likely to have a significant impact on how we model the responses to public policy.
For the social welfare function too there are alternatives. Public economics has not been blind to differences in objectives. My own response to reading the original paper by James Mirrlees on optimal income taxation (Mirrlees, 1971) was to ask how the results would change if the utilitarian objective were replaced by one concerned, in Rawlsian fashion, with the least advantaged. The Benthamite solution could be contrasted with that of the more egalitarian John Rawls. The optimal choice of tax and spending can be parameterised in terms of the degree of inequality aversion (taking a concave
function of the individual Vi). However, this parameterization falls way short of meeting the concerns that have been expressed about the limitations of utilitarianism. In particular, many have questioned whether we have the right evaluative basis. As Amartya Sen has proposed, we need to consider
alternatives such as capabilities (see, for example, Sen, 2009).
A second important topic that I have not discussed is the global reach of the analysis. In recent years, Agnar has made several interesting contributions on the theme of global public finance (Sandmo, 2006 and 2007), and many of the actions of national governments do indeed have implications for the citizens of other countries. We have therefore to consider who exactly is included in the social welfare function. In the box at the beginning of the Lecture, I indicated a sum over i = 1 to n, but where does n stop? Not presumably at the borders of the UK or Norway. This is particularly important in times of austerity, when there is a temptation to downgrade international commitments.
Turning to the overall conclusions regarding austerity programmes – my main focus in the Lecture - I believe that, despite the omissions just signaled, we can say something substantively about the design of austerity programmes. Perhaps most importantly, not all austerity packages are the same. There are alternatives, and public economics can contribute to the analysis of their strengths and weaknesses.

18

The mix of 75% spending cuts/25% tax increases (or even more biased towards spending cuts in the UK) is open to question; a reasoned case has to be made why spending should bear the brunt. How far is this mix a response to the macro-economic crises, and how far is it a long-term shift towards a smaller state? It is not as though tax rises are impossible. The UK Mirrlees Review proposed a base-broadening for the VAT which would fill a significant part of the deficit. I have suggested other possible ways of raising tax revenue: a lifetime capital receipts tax, and a periodic wealth tax. Within any agreed mix of spending cuts and tax increases, there are choices to be made, and these choices determine who bears the burden of fiscal consolidation. I have suggested that time is of the essence and that inter-generational justice suggests shifting spending cuts towards older generations, protecting public capital formation – unlike past budgetary adjustments - and raising taxes on older generations.
The Lecture also has implications for public economics. I have tried to demonstrate that one can learn a lot from the literature, but I have also argued that there are serious limitations. There is a risk that the subject has become imprisoned in a particular model of the economy, and we need to consider important features of the real world such as imperfect competition and labour markets where pay is not simply a reflection of one‟s productivity.
For this, we need to draw on developments in other fields of economics (this is also a two-way process). We have to recognize that the real world is changing
in important ways. I have stressed two of these: the return of inherited wealth, and the differing experiences of different cohorts. I have just touched on the need to move beyond utilitarianism, to consider alternative explanations of behaviour and alternative ethical principles, and on the challenge of global public finance. Challenge is indeed the word on which to end – there is a lot of exciting research to be done!

19

References
Atkinson, A B, 1972, Unequal shares, Allen Lane, London.
Atkinson, A B, 2011, “Income distribution and social change after 50 years”, Titmuss Lecture at LSE.
Atkinson, A B and Sandmo, A, 1980, “Welfare implications of the taxation of savings”, Economic Journal, 90: 529-549.

Atkinson, A B, and Stiglitz, J E, 1976, “The design of tax structure: direct versus indirect taxation”, Journal of Public Economics, vol 6: 55-75.

Auerbach, A and Hines, J R, 2003, “Perfect taxation and imperfect competition” in S Cnossen and H W Sinn, editors, Public finance and public policy in the new century, MIT Press, Cambridge.

Balassone, F and Franco, D, 2000, “Public investment, the Stability Pact and

the „golden rule‟”, Fiscal Studies, vol 21: 207-229.

Clark, T, Elsby, M and Love, S, 2001, “Twenty five years of falling investment?”, IFS Briefing Paper, London.

Diamond, P A, 2008, “Behavioral economics”, Journal of Public Economics, vol

92: 1858-1862.

Hills, J, 2012, “The changing architecture of the UK welfare state”, Oxford

Review of Economic Policy, forthcoming.

HM Treasury, 2011, The green book, HM Treasury, London.

Karagiannaki, E, 2011, “Recent trends in the size and the distribution of

inherited wealth in the UK”, CASEpaper 146, London School of Economics.
Kremer, M, 2001, “Should taxes be independent of age?”, working paper,
Harvard University.
Mirrlees, J A, 1971, “An exploration in the theory of optimum income taxation”, Review of Economic Studies, vol 38: 175-208.
Mirrlees, J A (chairman), 2010, Dimensions of tax design, Oxford University
Press, Oxford.
Mirrlees, J A (chairman), 2011, Tax by design, Oxford University Press, Oxford.

20

Myles, G D, 1989, “Ramsey tax rules for economies with imperfect competition”, Journal of Public Economics, vol 38: 95-115.

Myles, G D, 1995, Public economics, Cambridge University Press, Cambridge. Office for Budget Responsibility, 2011, Economic and fiscal outlook, November

2011, Cm 8218, HMSO, London.

Persson, M and Sandmo, A, 2005, “Taxation and tournaments”, Journal of

Public Economic Theory, Vol 7: 543-559.

Pigou, A C, 1928; 1947, A study in public finance, third edition, Macmillan, London.

Piketty, T, 2010, “On the long-run evolution of inheritance: France 1820-2050, Data Appendix”, Paris School of Economics.

Piketty, T, 2011, “On the long-run evolution of inheritance: France 1820-2050”,

Quarterly Journal of Economics, vol 126: 1071-1131.

Sandmo, A, 1972, “Optimality rules for the provision of collective factors of production”, Journal of Public Economics, vol 1: 149-157.

Sandmo, A, 1975, “Optimal taxation in the presence of externalities”, Swedish

Journal of Economics, vol 77: 86-98.

Sandmo, A, 2006, “Global public economics: Public goods and externalities”,

Economie Publique, Numbers 18-19.

Sandmo, A, 2007, “The welfare economics of global public goods”, NHH
Department of Economics Discussion Paper 35/2007.
Sandmo, A, 2011, Economics evolving, Princeton University Press, Princeton.

Seade, J, 1985, “Profitable cost increases and the shifting of taxation”,

Warwick Economics Research Papers 260.
Sen, A, 2009, The idea of justice, Allen Lane, London.
Stern, N H, 2010, “Imperfections in the economics of public policy, imperfections in markets, and climate change”, Journal of the European Economic Association, vol 8: 253-288.
Weinzierl, M, “The surprising power of age-dependent taxes”, Review of

Economic Studies, vol 78: 1490-1518.

21