Lessons from failure: fiscal policy, indulgence and ideology.

Wren-Lewis, Simon

Macroeconomic theory clearly suggests that at the zero lower bound, fiscal contraction will reduce output and slow any recovery. Yet in 2010 the focus of fiscal policy in many countries switched from supporting activity to reducing debt, despite the fact that the recovery from recession often appeared weak. While high levels of public debt can explain this switch in some countries, it does not provide a satisfactory account in others. In addition, the possibility of using balanced budget fiscal expansion or tax switches that bring forward spending remain largely unexplored. This paper suggests that policy has been influenced by an opposition to countercyclical fiscal policy which has ideological roots.

Keywords: Countercyclical fiscal policy; government debt; deficit bias; balanced budget fiscal expansion

JEL Classifications: A14; E62; E63: E65; H63

I. Introduction

"The tragedy of our current economic mess is that the solution to our problems is not, in fact, mysterious--basic macroeconomics, macroeconomics that has worked quite well in the last two years, shows the way. But the men in suits have decided that they know better. "

Paul Krugman, 13/09/10

Over the past thirty years, macroeconomics has made tremendous progress in knowing how to deal with modest macroeconomic shocks. The great moderation period demonstrated that an active monetary policy, coupled with explicit or implicit inflation targets and operated by independent central banks, could stabilise both output and inflation. (1) Unfortunately what is now becoming increasingly clear is that policymakers are not able to deal with large negative macroeconomic shocks.

During the first phase of the current recession, the prognosis about how macroeconomic policy responds to a large negative shock looked more optimistic. Monetary policy cut interest rates rapidly and as far as possible (although with hindsight US policymakers did rather better than those in Europe), and fiscal policy was actively used both in the US, the UK, China and other countries. Lessons from the 1930s appeared to have been learnt. However the use of discretionary countercyclical fiscal policy was opposed by many. Partly as a result, fiscal action was insufficient to prevent large falls in output across the globe.

In 2010 the view of policymakers (the 'men in suits' referred to in Krugman's quote above) changed decisively towards the need for austerity to reduce public debt. The likely consequence is that in many countries the recovery will not close the output gap for a number of years. This will waste a huge amount of resources, and cause widespread unhappiness. So we have a failure of macroeconomic policy: not on the same scale of the 1930s, but a failure of substantial proportions nevertheless. This paper is about why that has occurred.

For many economists policy failure can be put down to the inability of politics to control public debt during the period before the recession: what is often called deficit bias. This is the indulgence referred to in the title. It is clearly the case that past deficit bias has made fiscal expansion more difficult, and for a few countries it has become impossible. However, is high debt a sufficient explanation for why fiscal expansion has been abandoned? This paper makes two points which suggest it is not.

1) Not all countercyclical fiscal policy needs to involve an increase in debt. A failure to use balanced budget fiscal expansions has to be explained some other way.

2) The debt situation has not been a critical problem (a problem that has to be addressed immediately) for many countries. In those countries fiscal expansion today, followed by austerity tomorrow, is technically both feasible and probably optimal from a macroeconomic point of view. It may not be feasible politically, but those who oppose it have not generally focused on this political constraint. Instead they have argued either that fiscal expansion is unnecessary or ineffective.

Both of these points suggest another factor beside past indulgence that could be playing some part in explaining the current failure of macroeconomic policy to ensure a rapid recovery from recession. This paper suggests it is ideology. In the 1970s, when macroeconomic theory was ruled by 'competing schools of thought', it would have been commonplace to link different macroeconomic models with ideological and political positions. More recently that situation appeared to change. The microfoundations programme provided a unifying device, which, although not eliminating disagreement, at least meant that macroeconomists by and large talked the same language. However, the current debate suggests that this does not mean that policy discussion among macroeconomists has become ideologically free.

The paper begins, in Section 2, by attacking the proposition of 'irrelevance', which states that no fiscal action is required because recovery is underway. The point that policy should not be content with modest growth because this reduces the output gap rather slowly is straightforward but appears to require repeating. Section 3 discusses indulgence, both in terms of what causes deficit bias, how much this has prevented more extensive countercyclical action, and whether the macroeconomics discipline has focused enough on this problem. Section 4 looks at other arguments against expansionary fiscal policy. The section is titled 'ignorance', because such arguments appear to contradict the basic consensus model that is the basis of modern macroeconomics. It is sometimes said that those arguing against the effectiveness of fiscal policy are assuming a world of price flexibility, but this paper will argue that this argument is not tenable when we hit a zero lower bound for interest rates. Section 5 suggests that there is a strong ideological element both in the opposition to countercyclical fiscal policy and the denial of its effectiveness. Section 6 concludes.

2. Irrelevance: it will be all right in the end

In the UK, the recent move to fiscal austerity at the low point of the recession has a distinct feeling of deja vu about it. In 1981, a Conservative government tightened fiscal policy at a very similar point in the business cycle. Then, as now, the majority of UK economists publicly protested against this move. There are of course important differences between the two periods: in 1981 the problem was primarily inflation, whereas now it is large budget deficits. Yet to the extent that there is a received wisdom about this period, it is depressing. The view of many seems to be that the concerns expressed by 364 economists in their famous letter to The Times were unwarranted, because the economy expanded quite rapidly from about 1981 onwards. (See Smith, 2009, for example.)

I would certainly agree that the famous letter itself made little intellectual sense. For those who remember the intellectual chaos that was macroeconomics at the time, getting that number of economists to agree on anything that was half sensible would have been impossible. The key issue, however, is not the details of the letter, but whether the government was right to tighten fiscal policy at the low point of the recession. UK unemployment remained persistently high through most of the 1980s. There is no evidence that the fiscal contraction in 1981 increased growth thereafter, or that the economy could not have recovered more rapidly than it did. Instead, as Nickell (2006) writes "By ensuring that subsequent output growth was below trend for a number of years, it did indeed deepen the depression just as predicted. Furthermore, it was unnecessarily tight in the sense that a somewhat looser policy would still have raised unemployment far enough above its equilibrium level to bring inflation down over a reasonable period.'' (2)

This sense of deja vu is intensified by reading some of the commentary about prospects looking ahead. Staying with the UK economy, an IMF report (IMF, 2010) published in November 2010 contains a section entitled 'To Tighten, or Not to Tighten--UK Fiscal Policy in the Public Debate'. The report gives four reasons in favour of current tightening. Thankfully none argue that fiscal expansion would not work in increasing output in the short run; the IMF had earlier encouraged governments to use fiscal action to support the economy during the initial stages of the recession. Two of the arguments relate to debt, which will be discussed later. But the other two are that (a) the recovery is underway, with the expectation of 'solid growth' ahead, and (b) we do not know how much spare capacity there actually is. These two arguments appear very weak.

What is meant by solid growth? The main text of the report says "growth is projected to rise gradually to 2 1/2 per cent in the medium term, only slowly closing the output gap." The last point is crucial: solid growth so defined means a drawn out recovery with persistently high unemployment. In the UK and elsewhere the rise in unemployment is concentrated amongst the young; youth unemployment is now at 20 per cent. The dangers of long-term unemployment among those without work-based skills are well known, in part because of the UK experience in the early 1980s.

It is certainly true that the extent of spare capacity in the goods market is unclear. (In the UK there are some unusual puzzles, but these look idiosyncratic, and are not present in the US.) However the one clear indicator of spare capacity we have is unemployment, and that has increased substantially and remains high. (3) There is no evidence that this increase in unemployment is predominantly either a voluntary choice by workers or a sudden emergence of skill mismatch. (Of course, the longer this unemployment continues, the risk that unemployment will become structural rises.) More generally, there is always a risk that we will hit capacity constraints sooner than expected. Yet it is incredible to argue that we are currently at full capacity, and a 'solid recovery' that has growth only moderately above the growth in potential implies a drawn out recovery with large economic and social costs. So there remains a clear prima facie case for using fiscal policy to increase the pace of recovery rather than the opposite.

The key problem with arguments that point to the risks that we will hit capacity constraints sooner than is expected is that they ignore monetary policy. Although the zero bound prevents conventional monetary policy from assisting the recovery, it can still be used very quickly to dampen things down if a recovery does get out of control, and core inflation starts to rise. (4) A key difference between now and the 1970s is that monetary policy is in the hands of central banks rather than governments, and in most cases there is little reason to worry that central banks will be shy about raising interest rates if core inflation does begin to rise above target. (Indeed, as we will note in section 4, it is in part their reluctance to contemplate inflation moving above target that has blunted the effectiveness of 'unconventional' monetary policy at the lower bound.)

Hopefully the move from fiscal expansion to austerity that occurred in many countries in 2010 may not result in a lost decade for the world as a whole. However, this is not the relevant yardstick. There is no reason why a recovery from a large negative shock should not be as rapid as possible, thereby minimising the distress that the downturn creates. It certainly makes no sense to take action that is bound to slow that recovery from what it otherwise would have been, unless there is some overriding reason to do so. Yet that is exactly the effect of the fiscal policy switch that occurred in 2010, and this paper seeks to examine why that switch happened.

3. Indulgence

The near default in Greek government debt in early 2010 was decisive in changing the consensus among policymakers from tentative fiscal expansion to austerity. Yet it is important to recognise that, in the case of Greece, the sharp rise in debt caused by the recession was not in itself the problem. Instead it brought to light an underlying fiscal position which was not sustainable. Equally, the fiscal position in Ireland became problematic largely because of the decision to guarantee the deposits of Irish banks. The recession may also have exposed how the previous housing boom had flattered the Irish government's fiscal position, but this again suggests the true problem predated the recession.

This point is important because it shows that problems of debt default have no implications for the possibility of countercyclical policy in principle, but have everything to do with the cyclically adjusted state of the public finances. Default becomes possible when there is a significant chance that government may be unwilling to impose the level of taxes required to service its debts. (5) A recession will raise debt but, once the recession is over, tax receipts should return to their pre-recession levels. So the increase in debt caused by the recession and any countercyclical policy that went with it will only push the economy into default territory if the economy was already close to that position (or if the government guarantees the deposits of a large insolvent banking sector).

Debt default, and the high risk premiums that go with that possibility, are not so much a result of recession as the consequences of indulgence before it. In the period from the mid-1970s to the mid-1990s, OECD debt rose from around 40 per cent of GDP to around 75 per cent of GDP. There is no apparent economic justification for this increase, and for this reason it is often labelled 'deficit bias'.

Deficit bias: causes and cures

A number of explanations for deficit bias have been proposed in the literature (see Calmfors and Wren-Lewis (2011) for a detailed discussion and literature references). One possibility is very pessimistic: that the current generation does not care enough about future generations, and so tries to take resources away from them. This explanation is consistent with other mechanisms that do the same, like climate change, unfunded social security schemes or rising house prices. It is pessimistic because it is unclear why a popularly elected government would ever have an incentive to do anything about it.

Another possibility is that deficit bias is a strategic response of political parties (and their supporters) to elections. A government in power knows that it may lose a future election. It then may be optimal for it to build up debt, which will then constrain the actions of the opposition if they gain power. Here it might be possible to get both parties to agree not to use this mechanism just before an election with uncertain outcome, but it would require some commitment mechanism to enforce it once one party had won.

A similar but conceptually different story arises if governments are more impatient because politicians may lose power. In these circumstances, they will put too little weight on the long-run consequences of fiscal decisions, which may imply that there will be a tendency for debt to rise inexorably over time. If the electorate is unable to punish governments of this type (because, for example, elections are fought over many issues) then we require some other device that might restrain impatient politicians.

Common pool theory provides both a plausible and testable account of why deficit bias might occur. The idea is that individual tax cuts or spending increases benefit small groups a great deal, but have a much smaller cost to everyone else. As a result, these groups lobby for these fiscal benefits without internalising the overall cost. The idea can be mirrored within a cabinet, where each spending minister pushes his own programmes, and only the finance minister is in a position to prevent excessive overall spending. In this account, deficit bias is more likely in governments that are more fragmented, and there is empirical evidence to support this. Here deficit bias can be counteracted by some device that increases the power of the finance minister in any negotiations.

Finally deficit bias may arise because of a lack of information available to the electorate. Governments may deliberately exploit this, for example by falsely suggesting that fiscal benefits announced shortly before an election are consistent with long-term fiscal responsibility. Alternatively governments may be overoptimistic about the consequences of their policies on long-term growth, and therefore future tax receipts, and the public is unaware of this over-optimism. Here an institution that could provide the public with unbiased information might mitigate this source of deficit bias.

The device that is normally prescribed to avoid deficit bias is some form of fiscal rule. Unfortunately international experience with fiscal rules over the past twenty years or so has not been that successful. There appears to be an inherent problem in trading off enforceability and optimality. No simple fiscal rule (like a balanced budget) is likely to be optimal. As we will note below, there are good reasons for allowing debt to act as a buffer for fiscal shocks, and furthermore that this buffer should correct only very slowly. However complex rules that might approximate optimal fiscal policy are likely to be easy to fudge. A clear example is cyclical correction, which is a calculation that is both important to implement and easy to distort.

More recently, a number of countries (with the encouragement of international bodies like the IMF, OECD and European Commission) have set up 'fiscal councils' to work alongside fiscal rules. Fiscal councils are independent institutions set up by the state whose task it is to monitor the government's finances. New fiscal councils in countries including Canada, Sweden, and the UK have now joined more longstanding bodies in the US, Netherlands, Denmark, Belgium and elsewhere. As Calmfors and Wren-Lewis (2011) note, these institutions differ widely in size, structure and scope, and it is too early to say how successful they will be. However, they do have the potential to mitigate some of the causes of deficit bias outlined above. They can provide information that might expose fiscal bribes and over-optimistic forecasts. They can apply political pressure that may restrain impatient politicians or bolster the power of a finance minister over his or her colleagues.

All of the fiscal councils that currently exist are advisory bodies only. If these bodies establish a reputation for sound advice, then one possibility would be to give them some formal power. Of course individual tax and spending decisions have to remain in democratic hands, but it is possible to imagine a fiscal council prescribing the overall budget deficit that a government has to aim for. This would amount to fiscal policy delegation that was similar in extent to existing monetary policy delegation to central banks. (The comparison between monetary and fiscal policy delegation is explored in detail in Wren-Lewis, 2011.) Whether this possibility is ever put into practice may depend on how effective advisory fiscal councils turn out to be in tackling deficit bias.

Why deficit bias might, or might not, prevent fiscal expansion

There are two main arguments that link deficit bias with the undesirability of fiscal expansion. Both may be valid for some countries, but do not necessarily apply to all. The first argument notes that for those countries suffering a debt crisis, fiscal expansion is impossible. As we noted above, this crisis reflects past deficit bias (or the socialisation of banking debts). However deficit bias (and potential banking problems) appears to have been present in most countries to a greater or lesser extent. As a result, it is suggested, most countries with high debt are potentially vulnerable to a market-led debt crisis, and this prevents any possibility of additional countercyclical fiscal policy in those countries.

An alternative to default, when most government debt is in nominal domestic currency terms, is inflation. This is not an option for an individual member of the Euro Area, because it would require inflation across all Euro Area countries. However, for countries like Japan, the US and the UK, it is probably a preferable and more palatable alternative to default. Yet there is as yet no sign in the markets in those countries that significantly higher inflation is anticipated. These countries also appear at the moment to be able to borrow at very low interest rates. Instead, it may be that the Euro Area countries currently in trouble are more of a special case.

Another problem with the 'fear of default is just around the corner' argument is that it seems especially inapplicable to the world as a whole. Globally the evidence points to a shortage of safe assets rather than a glut, which is exactly what we would expect following a worldwide recession. If the current recession reflects too little demand (as, in section 4, I argue that hitting the zero lower bound for interest rates surely implies) it means we have too much private sector saving, so the public sector in the world as a whole should be saving less not more. It is a rather strange world where there is a global excess demand for government debt, but each individual government may be about to lose market confidence in their debt.

A second argument linking deficit bias with the inadvisability of fiscal expansion is much more political. As Martin Feldstein wrote in the Financial Times, (6) "If the timing of the fiscal consolidation were just a technical economic problem, the right policy would be to enact a multi-year budget that starts with little or no deficit reduction for the first two years, followed by a rapid return to budget balance. The slow start would be particularly appropriate in those countries where aggregate demand is now very weak. But such a gradual adjustment strategy cannot work politically in countries where voters are sceptical about government promises of future deficit reductions. Immediate action is necessary to make future deficit cuts credible."

This argument concedes that the debt problem is a long-term issue, which requires government action over the long run. The key initial requirement is that the government implements a programme which will eventually stabilise debt. To minimise the risk of default, debt then needs to be reduced to a level that can be serviced even if the economy is hit by adverse shocks. Beyond that, debt may need to be reduced further, depending on an assessment of what is the optimal long-run level of debt. (What this might be is discussed below.) However, a result that appears very robust is that any adjustment in debt should be slow, and that during this process debt should be allowed to vary in response to economic shocks (see Leith and Wren-Lewis (2000), Kirsanova and Wren-Lewis (2010) and Marcet and Scott (2008) for example, and the discussion below).

Countercyclical fiscal policy, in contrast, only requires temporary increases in government spending. Indeed, as we shall note in Section 4, temporary increases in spending are much more effective at raising demand than permanent increases. It is technically possible, therefore, to combine stimulus now with austerity later, and furthermore this would be the optimal policy given the lack of current demand. Unfortunately, so the argument goes, this optimal policy combination is not politically tenable, because governments cannot be trusted (and cannot be forced) to implement austerity in the future. Under this view, any political opportunity to implement austerity has to be taken, even if it means that the recovery from recession will be slow. A slow recovery is less of an evil than a failure to control debt, and political constraints prevent us dealing with the short-term problem in the short term, and the long-run problem in the longer term.

The key element in this argument is that the desire to reduce deficits in the future, after a fiscal expansion, will be weaker than the incentive for austerity today. Underlying this idea might be a view that deficit bias itself reflects the fact that politicians, left unconstrained, will find it too attractive to spend money and/or cut taxes. Only when the public or markets perceive the problem of debt to be sufficiently serious will politicians be forced to do something about it. The danger in delaying austerity is that recovery from recession will make the debt problem appear less serious (because taxes will be higher and transfers lower), and so the incentive to tackle the deficit will decline.

This view requires public perceptions to be pretty superficial. (7) In cyclically corrected terms, the debt problem once the recovery has been completed will be worse than at the bottom of the recession, particularly if debt financed expansion has been conducted to assist the recovery. It does not take much analysis to understand that the debt problem has not gone away after the recovery. Even if the public's view of the fiscal position is very naive, the first best policy must be to try and improve the public debate through institutional changes like establishing a fiscal watchdog. After all, a 'solution' to the problem of deficit bias that relies on the occasional debt crisis is not much of a solution.

Neither of these arguments against fiscal expansion suggests that it would fail to stimulate the recovery in principle. Instead they imply that deficit bias in the past either leads to a significant chance of a debt crisis that should be avoided, or means that governments cannot be trusted to reduce debt once the recession is over. The fact that neither deny that countercyclical fiscal policy will work is important for two reasons. First, for countries where levels of debt are not a problem, or where governments can commit, it implies that fiscal expansion is both possible and desirable if the recovery is slow. Second, it is important to note that in some situations fiscal contraction may also be appropriate.

For economies with flexible exchange rates there are good reasons for believing that monetary policy should always be preferred to fiscal action as a means of preventing excess demand leading to rising inflation (see Kirsanova et al., 2009). There is no upper bound to interest rates! However, for economies that are part of a fixed exchange rate zone or a currency union, it is well understood among economists that only fiscal policy can manage demand in the face of idiosyncratic shocks. Yet this point seems to have been lost on many Euro Area governments. The original Stability and Growth Pact may have been partly to blame for this. With its emphasis on deficit ceilings, it became too easy for governments that were experiencing booms, and that therefore were not breaching this deficit ceiling, to believe that their aggregate fiscal position was appropriate. In reality, inflation in countries such as Ireland was systematically above inflation in Germany, and so their real exchange rate was moving away from its sustainable level. These countries should have been counteracting their domestic booms with contractionary fiscal policies. This is exactly what some economists were advising them to do at the time, (8) but their arguments were easier to ignore given the Euro Area's fiscal framework.

The lack of research on optimal debt

What is the optimal long-run level of government debt? Given the potential importance of this issue, one might imagine that this would be a well worked area of macroeconomic research. Furthermore, if this research suggested that current levels of debt were probably above optimum levels, then this might have provided additional pressure on governments to tackle deficit bias.

Unfortunately, the little research that has been done on this question comes to what might be regarded as an unhelpful conclusion. Suppose agents are free to borrow and lend as they wish, and they also care about their children in such a way that they in effect live forever. There are some public goods that it makes sense for the government to provide, but these have to be paid for by distortionary taxes. In these circumstances, if we could choose the level of government debt for this economy from scratch (i.e. we ignored what debt we inherit), then it would make sense for the government to hold assets, not debt. The reason is that we could use the interest from these assets to pay for public goods, and then we could do away with all distortionary taxes. In other words, the optimal level of government debt in the absence of history would be negative.

Now suppose, more realistically, that we inherit some positive level of debt. Does it follow that we should then steadily reduce it? The answer is no, using an argument based on Barro's tax smoothing hypothesis. To get debt down, we need to raise taxes. This would have the benefit of reducing debt and therefore taxes in the long run, but at the cost of raising taxes in the short run. Tax smoothing implies that a better policy would be to keep taxes constant, and taxes can only be constant if debt is kept at its historic level. In this sense, the optimal level of debt is not negative, but just the level of debt we inherit. (9)

This theory, which is sometimes called the random walk steady state debt (hereafter RWSSD) theory, means that we cannot say that our inherited level of debt is either too high or too low, because the costs of changing it always exceed the benefits. So, although as economists we might deplore deficit bias after the event, following this theory implies that we cannot argue that deficit bias should be undone, but only that it should not occur in the future.

The RWSSD theory relies on using a model with the features I have described. The assumption about consumers noted above is crucial because they imply a key ingredient in the result, which is that the real rate of interest in the economy tends towards the rate at which agents discount their utility i.e. their impatience. With consumers who are effectively infinitely lived, this is exactly what we find in steady state, because only then will consumption be constant over time. (This follows from the consumer's Euler equation.) This economy also implies that Ricardian Equivalence holds; consumers will save all of a cut in lump sum taxes. So we could call this a 'Ricardian Economy'.

The RWSSD result is highly problematic for two reasons. First, it is what is often called a 'knife edge' theory, which means that a very small deviation from one of its key assumptions will lead to very different implications. Second, it embodies an ethical position that is difficult to defend. Let us take each in turn.

What would happen if the rate of interest was slightly above the rate of impatience? Would that make the optimal level of long-run debt slightly less than its historic level? Not at all. It would in fact reintroduce the idea that the optimal level of debt in the long run was negative; the government should eventually strive to own positive net assets, so that it can do away with distortionary taxes. (See Aiyagari et al. (2002) for example.) The reason is simple. The benefits of cutting debt now just outweigh the short-term costs of raising taxes, so we should gradually reduce debt. In that sense, the RWSSD proposition is a knife edge result; the rate of interest has to exactly equal the rate of impatience for it to hold, and if it does not, we get a very different result. The reason for this odd knife edge is that tax smoothing is really about the speed at which debt should approach a long-run target (which, as we noted earlier, should be slow), and not really about what that target should be. If the rate of interest is only just above the discount rate, it is optimal to move towards the long-run (negative) debt target very slowly indeed, and it is also optimal for debt to rise temporarily in response to shocks to the public finances. At the limit, when the rate of interest equals the discount rate, the speed of adjustment is zero: hence RWSSD. For this reason, RWSSD is not at all robust.

The Ricardian economy is also problematic because it discounts the welfare of future generations. At first sight the assumption that today's consumers care about their children appears to avoid any ethical judgment about intergenerational welfare, because consumers internalise the problem. But they only do so on the current generation's terms, which involves impatience. In effect, therefore, the utility of future generations is being (heavily) discounted because of the impatience of the current generation. At the very least, this ethical judgement is not obviously correct. (We discuss it further in section 5) Yet it is critical to the optimality of the RWSSD result. The costs of higher taxes today only exceed the benefits to our children of reducing debt because we are discounting our children's utility in this model. (Barrell and Weale, 2010, not only stress the role of intergenerational equity for debt policy, but also examine how it could be used to offset intergenerational redistributions caused by other factors such as rising house prices.)

If the RWSSD result is so problematic, why has work on optimal long-run government debt not abandoned it? There may be many reasons, but one could be that the Ricardian model on which it is based has two advantages from a particular ideological point of view. The first is that the assumption about internalising the welfare of future generations appears to avoid the need to deal with how to compare welfare between generations. There is a longstanding strand in economics that strives to avoid making any interpersonal utility comparisons. (10) Second, the Ricardian economy, when combined with a simple form of production and competition, implies a strong welfare result, which is that the level of capital produced by the market economy is also the optimal level of capital. In that case, there is no reason for government intervention to either encourage or discourage saving or investment. In both respects, therefore, the Ricardian economy diminishes the need for any state intervention.

4. Ignorance

Some macroeconomists (and many others) argue that fiscal expansion would not work, even if high levels of debt were not a constraint. (Some examples are given at the beginning of the next section.) This section aims to show that such claims contradict basic macroeconomic theory, while the next asks why they are nevertheless made.

We can split the issue of whether expansionary fiscal policy can be effective at raising output into two parts. The first asks whether it will tend to raise demand, while the second moves from demand to output. The key question here is one of effectiveness, and not optimality. There are good reasons why, in normal times, monetary policy is preferred to fiscal policy as an output stabilisation tool (see Eser et al., 2008). The question the current situation poses is one where conventional monetary policy cannot do this job. (11)

Fiscal expansion and demand

Everyone who has done first year undergraduate economics knows that in the IS/LM model expansionary fiscal policy works in raising demand. But pretty much everyone who does a masters course in economics knows that this is not the core model in current macroeconomic analysis. At the heart of macroeconomic analysis today is the intertemporally optimising consumer. If we assume that these consumers also internalise the utility of future generations, and that there are no credit constraints, then we get the Ricardian Equivalence result that tax cuts are completely saved. This is the 'Ricardian economy' discussed in the previous section. In this model, fiscal expansion that involves cutting lump sum taxes with unchanged government spending would have no impact on demand, because the tax cut would be saved.

However, exactly the same model implies that a temporary increase in government spending will increase demand. In fact, if Ricardian Equivalence holds, it makes no difference if this additional government spending is financed by issuing debt or tax increases. The intertemporal consumer smoothes the impact of current or future tax increases, but the increase in government spending is not smoothed. Note that for a positive multiplier we only require consumers to smooth the impact of tax changes to a significant extent; complete Ricardian Equivalence is not required. We have not had to put any structure on any other part of the macromodel. As long as consumers smooth, the only way a temporary increase in government spending could not raise demand is if consumers did not believe changes in government spending could be temporary.

What about an open economy with flexible exchange rates: could movements in the exchange rate crowd out the additional demand? The first point to note is that if the increase in government spending is temporary, in the long run aggregate demand has not increased, so there is no reason why the 'equilibrium' exchange rate should change at all. Second, if the long-run exchange rate does not change, then Uncovered Interest Parity suggests the exchange rate in the short run will only change if real interest rates rise. But if nominal interest rates are stuck at the zero lower bound (hereafter ZLB), real interest rates will only rise if inflation expectations fall following a fiscal expansion, which is hardly likely. So there will be no crowding out via the exchange rate. (12)

Of course there is much that we might want to change in these very simple accounts of how the economy works. However, the basic ideas they contain seem very robust (see Woodford, 2011). Consumers will tend to smooth tax changes, and if they cannot, then tax increases can be delayed by issuing debt. As a result, the government (which has no reason to smooth) can alter the pattern of aggregate demand. This logic can be expressed in many different ways. One is to say that as the demand problem arises because the private sector is saving too much (and we cannot counteract this tendency by lowering real interest rates), then the public sector needs to save less.

Demand and output, and "demand denial'

Will an increase in demand generated by fiscal expansion lead to an increase in output and a reduction in unemployment? The standard answer to this question is yes, if demand is deficient and prices are sticky. An equally standard argument against fiscal expansion is that prices are sufficiently flexible to eliminate any demand deficiency, so no stimulus to demand is required. (13) This view is I believe untenable when monetary policy targets inflation and we hit the ZLB.

Keynesian economics is all about the economics of aggregate demand, and it is generally taught that Keynesian economics requires some form of price rigidity. By implication, if price adjustment is rapid enough, demand deficiency will be 'self correcting', and we then ignore aggregate demand and focus on supply if we want to know what output and employment will be. But bow much does this statement rely on certain assumptions about monetary policy?

The first, uncontroversial, point to make is that in standard models it is the real rate of interest that moves demand to equal supply. Imagine an economy without capital, so output is produced by labour alone. For simplicity, assume for the moment that labour supply is fixed. Given an aggregate production function, this gives us one number for total output. We also know that, if consumers make optimal intertemporal decisions subject to no constraints beyond their lifetime budget, then variations in the real rate of interest will alter the current level of consumption. Let us therefore assume that there exists a real interest rate that equates consumption (demand) to the output that could be produced by workers (supply). The key question is whether price adjustment can ensure that this real interest rate is achieved. This is where we have to talk about prices, but we also need to talk about monetary policy. Real interest rates are nominal interest rates less expected inflation.

Suppose from a position of full employment, some negative shock reduces consumption. Consumers save by hoarding money, so aggregate demand falls. Lower demand will lower output, and some workers will be laid off. (14) Firms may cut their prices, and unemployment may lead to cuts in nominal wages. To make things simple, let us assume all firms are monopolistic competitors, and that the production function is linear. In that case prices will only fall if nominal wages fall (the mark-up is independent of demand and output), but nominal wages will surely fall as unemployed workers try to get back into work. In the textbook model, if unemployment persists, nominal wages will continue to fall until they reach zero, unless the process of falling nominal wages and prices in itself increases consumption. (15)

In the macroeconomics of first year text books, the monetary authorities fix the nominal stock of money. In this case we can say two things. First, lower prices will increase real money balances, which leads to lower nominal interest rates. Second, the long-run neutrality of money implies that once output returns to full employment prices will return to their original level. If prices are falling now, then at some point they will start rising, implying an increase in expected inflation. So we have combined price adjustment and real interest rate adjustment to give us a story of why demand will eventually adjust back to supply.

Of these two mechanisms, the decrease in nominal interest rates is straightforward. The second, involving inflation expectations, is more complex. It may not work, for example, if inflation expectations are static (based on recent inflation), because falling inflation leads to lower expected inflation, higher real interest rates etc. However if inflation is forward looking, such as implied by a New Keynesian Phillips curve, then price setters will note that although there may be a negative output gap today, at some stage the output gap will have to become positive to generate positive inflation to restore the price level if we return to equilibrium. As a result, at some point inflation will be positive. This will be anticipated, and so at some point real interest rates may fall.

The speed of self correction depends on how rapidly prices adjust. If price adjustment is slow, adjustment can be quickened by an active monetary or fiscal policy. This is the neoclassical synthesis. There exists a self-correction mechanism, but if price adjustment is slow, there is also a role for an active policy. The quicker is price adjustment, the quicker demand adjusts to supply. Keynesian demand based macroeconomics and price inflexibility are clearly linked.

The role of inflation expectations becomes critical if nominal interest rates hit the ZLB. Now we have to rely on falling prices raising inflation expectations to get us to the real interest rate required to bring demand up to supply. If the money supply is fixed, and we return to full employment, then at some point prices will have stopped falling and started to rise. But to expect this presumes we get back to full employment. An alternative might be that inflation expectations do not rise, we stay stuck at a real interest rate that is too high, a nominal rate that is zero, and permanent involuntary unemployment. (16)

The discussion so far has assumed a textbook world where nominal money is fixed. Nowadays, for good reasons, we tend to think about monetary policy involving the choice of short-term interest rates in the context of an inflation target. This modifies the discussion in one crucial respect. We can no longer argue that falling prices today must, if full employment is returned to, lead to inflation expectations rising above the inflation target tomorrow. Instead, policy will aim to return inflation to its target level from below. This does not matter if we rely on nominal interest rate movements to get us to the appropriate real interest rate. It is the achievement of New Keynesian macroeconomic theory and the great moderation to understand that this is the job of monetary policy. But it matters a great deal if we hit the ZLB.

At the ZLB any self-correction of deficient demand would have to involve an increase in expected inflation, because the real interest rate has to fall. But if we have central banks committed to preventing inflation rising above some target level, then the combination of zero nominal interest rates and inflation at target may still give us a level of real interest rates that is insufficient to raise demand enough. If inflation expectations are rational and central banks are effective, real interest rates cannot fall by enough to solve the problem of insufficient demand.

A key point to note is that this conclusion holds however rapid price adjustment is. Self-correction can only occur at the ZLB through expectations of rising inflation reducing real interest rates. While this may be conceivable under a price level or money target, it becomes more difficult to imagine under inflation targeting.

There is a link here with one particular version of unconventional monetary policy. This is the idea that monetary policy can still be effective at the ZLB by promising higher (than target) future inflation. Eggertsson and Woodford (2004) show that this comes close to a policy of pursuing a price level target, which in turn is close to the fixed money supply case already examined. (17) Yet central banks around the world have said they would not pursue such a policy--they would not allow inflation above target. Given that this unconventional monetary policy is seriously time inconsistent (once the recession is over, it is too tempting to abandon the plan to have excess inflation), such denials have to be taken seriously.

With inflation targeting, once interest rates hit the ZLB there is unlikely to be a mechanism by which demand deficiency is self-correcting. This in turn means that, at this lower bound, it is no longer legitimate to argue that aggregate demand can be ignored because prices are flexible. This is critical to the fiscal policy debate. As we have argued earlier, temporary increases in government spending will raise demand in the standard model. If it is also the case that output is constrained by deficient demand because interest rates are at the lower bound, then fiscal expansion must raise output and employment.

The proposition that we can always ignore aggregate demand in macroeconomics, that some have termed 'Demand Denial', is not a new phenomenon. Keynes also asked why the dominant theory at the time (call it 'classical theory') refused to acknowledge the potential importance of aggregate demand, and wrote the following in the General Theory:

"That it [Classical Theory] reached conclusions quite different from what the ordinary uninstructed person would expect, added, I suppose, to its intellectual prestige. That its teaching, translated into practice, was austere and often unpalatable, lent it virtue. That it was adapted to carry a vast and consistent logical superstructure, gave it beauty. That it could explain much social injustice and apparent cruelty as an inevitable incident in the scheme of progress, and the attempt to change such things as likely on the whole to do more harm than good, commended it to authority. That it afforded a measure of justification to the free activities of the individual capitalist, attracted to it the support of the dominant social force behind authority."

There are many explanations for demand denial in this quotation. Some would seem to have less relevance today. For example New Keynesian theory can be grafted on to a Real Business Cycle framework, which suggests that the 'vast and logical superstructure' argument does not apply, unless you take your microfoundations very seriously (Wren-Lewis, 2011). The next section suggests that the last sentence from this quote is a more promising avenue to explore.

5. Ideology

In the 1970s and 1980s it would have been natural to talk about ideology and macroeconomics. The discipline was fragmented into opposing 'schools of thought', which appeared to have quite different models of how the economy worked. Each school was identified with a clear ideological position. However this appeared to change in what some have called the new neoclassical synthesis (Goodfriend and King, 1997). This was the idea that old debates between New Classical and Keynesian economists had been resolved in a new grand bargain. Keynesian economists agreed to work with models that used rational expectations, and had an RBC framework at their core, on the understanding that New Classical economists agreed that it was legitimate to add sticky prices within that framework, and that such models would then exhibit Keynesian characteristics. The glue behind this consensus was the acknowledgement that macroeconomics required solid microfoundations, coupled with a microfounding of price stickiness itself. (See some of the papers in Arestis, 2007, for example).

One of the implications of adopting this new consensus model is that monetary policy appears the natural first choice for macroeconomic stabilisation. But, as the previous section showed, it also suggests that if monetary policy cannot do this job, fiscal policy can take its place. Countercyclical fiscal policy works in this model. Yet, in both the US and UK, it seems to have become an article of faith in the major parties of the right (Republican and Conservative) that countercyclical fiscal policy does not work. It is true that in both cases this view was born out of opposition to a government that was attempting to apply such policies to mitigate the impact of the recent recession. In a sense it is natural to oppose. Nor have these parties always taken this view. The G.W. Bush presidency, for example, used Keynesian arguments to justify their tax cuts in 2001. However, it must be worrying when such an important part of political opinion appears to take a view that is contradicted by mainstream macroeconomic theory.

But do politicians and the public know that it is contradicted by mainstream macro theory? Consider the following quotes, all from professors at Chicago University: (18)

"The problem is simple: bailouts and stimulus plans are funded by issuing more government debt. (The money must come from somewhere!) The added debt absorbs savings that would otherwise go to private investment. In the end, despite the existence of idle resources, bailouts and stimulus plans do not add to current resources in use. They just move resources from one use to another."

"Every dollar of increased government spending must correspond to one less dollar of private spending." "But, if we do build the bridge by taking tax money away from somebody else, and using that to pay the bridge builder--the guys who work on the bridge-then it's just a wash. It has no first-starter effect. There's no reason to expect any stimulation. And, in some sense, there's nothing to apply a multiplier to. (Laughs.) You apply a multiplier to the bridge builders, then you've got to apply the same multiplier with a minus sign to the people you taxed to build the bridge. And then taxing them later isn't going to help, we know that."

The first two are just a crude Say's Law. As DeLong (2010) notes, even Say eventually realised that his law did not hold. The existence of (outside) money is sufficient to establish this. The third directly contradicts the Ricardian model considered in the previous section. Of course quotes of this kind are never fully contextualised; perhaps in the last case it was presumed that the bridge building was permanent. But they are indicative of the fact that the policy debate in the US has involved discussion not just about how effective fiscal policy may be (a debate over the size of multipliers), but a debate about whether it is effective at all (about the existence of multipliers). How is that possible, when every first year macro textbook teaches the Keynesian multiplier, and the more modern model outlined earlier also tells us that temporary increases in government spending will raise demand?

Part of the answer may reflect the unequal nature of the new neo-classical synthesis. For Keynesian economists it meant a great deal; New Keynesian theory was established on a real business cycle base. However, for Classical economists there was nothing to enforce the idea that New Keynesian theory had to be taken seriously. Graduate textbooks would begin by outlining the Ramsey model, and only later consider various 'imperfections', of which sticky prices was just one. As one academic who had taught graduate macro at a freshwater US department told me, they ran out of time before being able to cover New Keynesian theory.

However, this explanation cannot be the whole story, because it neglects one crucial fact. Monetary policy almost without exception involves the use of Keynesian theory of one form or another. To those conducting monetary policy, the overwhelmingly dominant view is that they are in the business of demand management. It is therefore rather odd to teach macroeconomics without addressing the body of knowledge that monetary policymakers routinely apply.

The explanation explored here is that ideology is influencing what at least some macroeconomists argue and teach. The ideological view in question is that state intervention should be minimised. However deficient demand due to sticky prices or the ZLB represent a generic market failure, which requires in pretty well every country the active involvement of an arm of the state in macroeconomic affairs. For those whose political instincts are pro-market, and against government intervention, this is something of an embarrassment.

One way of avoiding this embarrassment is to not teach New Keynesian theory at graduate level, and to ignore it in public debate. Even at undergraduate level, this ideological slant may help explain one feature of how macroeconomics is taught that at first sight appears odd. Pretty well every undergraduate macroeconomics text uses IS/LM. This of course assumes that the monetary authorities fix the money supply. Yet in the US and UK the last time the monetary authorities attempted to pursue a money supply target regime was at the beginning of the 1980s. I am sure I am not alone in finding it embarrassing to teach undergraduates a core model part of which applied only during a brief period before they were born, and which was generally regarded as a failure. Now this strange state of affairs could in part simply reflect inertia induced by the economics of textbook publishing, coupled with the fact that for many problems assuming fixed money gives answers that are roughly right (Wren-Lewis, 2009). However, as we saw in the previous section, sometimes the difference between inflation targeting and money targeting can be crucial.

An alternative explanation for the persistence of the LM curve in undergraduate macroeconomics teaching is that it downplays the role of monetary policy in avoiding demand deficiency. This is because fixing the money supply can be (and generally is) portrayed as a default, do nothing policy, that is natural and involves no intervention. With this default policy in place, demand shocks will be 'self-correcting' through the process of price adjustment. The irony here is that fixing the money supply is hardly a do-nothing policy, as both the UK and US found out when they tried to implement it.

Contrast this with what most central banks actually do today, which is to set interest rates so as to minimise deviations from an implicit or explicit inflation target and the natural rate. There is no default or do nothing policy here that allows the economy to be self-correcting: fixing the nominal interest rate could well be destabilising for well known reasons. As a result, if and how quickly the economy eliminates the impact of demand shocks will depend at least as much on the performance of monetary policy as the speed of price adjustment. In this context it is impossible to describe the economy as self correcting as long as monetary policy does nothing. State intervention, in the form of a monetary policy that has reasonable properties, becomes essential to the proper functioning of the economy.

When it comes to fiscal expansion at the ZLB, acknowledging that such a policy could work not only involves recognising this essential role for monetary policy, but it also requires admitting that other forms of state intervention may be required when this fails. But what solid evidence do we have that these ideological concerns are actually important in the current emphasis on fiscal austerity rather than expansion in many countries? Surely concerns about increasing debt are central. In Section 3 I discussed why such concerns may be less important for some countries than others, and why abandoning fiscal expansion in a recession may be a second best policy compared to institutional reforms. However, one clear indication of why concern with debt is not the only important factor behind abandoning fiscal stimulus is the absence of discussion of balanced budget fiscal expansion.

Balanced budget fiscal expansion

If the objection to expansionary fiscal policy in the current situation was simply that it would raise debt to unsustainable levels or prejudice a commitment to future austerity, then a natural reaction would be to emphasise the possibility of using fiscal policy in an expansionary manner without raising debt. (19) As we noted in the previous section, the standard macro model involves Ricardian Equivalence, which suggests temporary increases in government spending would still be effective even if they were paid for by raising taxes and keeping debt unchanged. To put the same point another way, the way to reduce debt today with as little negative impact on demand as possible would be to raise taxes rather than cut government spending. Yet those who argue for austerity today hardly ever make this argument.

Could it be that this policy is not explored because it is thought that a significant body of consumers are credit constrained and not Ricardian? Two points suggest not. First, the key issue is the extent of consumption smoothing. Even credit constrained consumers may smooth a tax increase over a greater period than any temporary increase in government spending, and any non-constrained consumers would smooth at least over their lifetime. Second, if there were thought to be a significant and identifiable body of consumers who were credit constrained, then a tax switch from the constrained to the unconstrained would be expansionary. These considerations would lead to a discussion of which types of balanced budget fiscal expansion were more effective, rather than a debate over whether any were effective at all.

Of course it is not difficult to see why either type of balanced budget fiscal expansion would run into political objections from the right. Increasing government spending increases the size of the state, albeit temporarily. Credit constrained agents are more likely to be poor, while the rich will almost certainly be unconstrained. So a tax switch that increases demand is likely to involve tax increases for the rich.

The key point is that a balanced budget, temporary increase in government spending is almost certain to be expansionary, and so if debt is the key constraint on bond financed fiscal expansion, this is an obvious alternative. If the concern is over the temporary nature of the additional spending, then the focus should be on choosing public investment projects that are inherently short term. Yet this is not the route taken by those who typically oppose debt financed fiscal expansion.

There are further possible balanced budget fiscal expansions that do not have straightforward political difficulties for those on the right. One is to temporarily and unexpectedly cut sales taxes (financed by a temporary increase in income taxes). The prospect of a future rise in sales taxes will stimulate consumers to bring spending forward. A temporary VAT cut was implemented by the Labour UK government as part of its expansionary fiscal policy to tackle the recession (see Barrell and Weale, 2009). There are a host of other specific fiscal measures that might provide incentives for firms or consumers to bring forward their spending. In effect this is fiscal policy trying to mimic monetary policy, by changing the intertemporal terms of trade.

If debt was the overriding constraint preventing fiscal expansion, then we might expect more of a focus on measures such as these. Instead, the debate tends to either ignore the possibility of balanced budget fiscal expansion altogether and focus on the dangers of high debt, or assert that fiscal policy will not work. This would be consistent with an aversion to acknowledging the extent of market failure, and furthermore to deny that state intervention may be required to correct that failure.

Pursuing these ideas, it becomes clear that talking about the effectiveness of 'fiscal policy' in a blanket way makes little sense. There are many fiscal instruments, operating on many different margins. It would be strange indeed if none of them had any impact on demand. To make the blanket claim that all fiscal policy is ineffective either requires 'demand denial', or an ideological imperative, or both.

Parallels with climate change

Nick Stern (Stern, 2006) has argued that climate change represents the largest externality ever faced by mankind. There are undoubtedly some climate change sceptics who genuinely dispute the science. There are others who may not dispute the science, but genuinely believe that we can afford to do little now to mitigate its effects because future costs can be discounted heavily. However, there is a clear correlation between climate change denial and a pro-market ideology that is instinctively against any state intervention (see Conway and Oreskes, 2010).

If the views of policymakers and economists on climate change were governed mainly by the question of how much future costs should be discounted, then we might tend to find those arguing for fiscal austerity rather than stimulus today (which will benefit future generations compared to the present) would also argue for strong action to tackle climate change. Equally, those who were relaxed about the need to tackle climate change would also be relaxed about the level of public debt. The reason is as follows.

One of the key issues in the debate between economists on climate change is how utility across generations should be discounted. US economists tend to want to discount the utility of future generations in much the same way as individuals appear to discount their own utility over time (as in the Ricardian economy). The Stern report argued that there should be no discounting of future generations per se. The ethical issues involved are discussed in Broome (1992). The more we discount future generation's utility, the less action is needed now to tackle climate change. As we noted in Section 3, if we base our debt policy on a Ricardian economy we are relatively relaxed about debt increasing over time as a consequence of a string of negative shocks. We also noted in the previous section that this model, although it implies Ricardian Equivalence, also suggests that temporary government spending increases will be effective at stimulating demand.

So, from this perspective, a view that downplayed the need to take action against climate change might be expected to be correlated with a view that downplayed the problems of high debt and was quite relaxed about countercyclical fiscal policy. However, among politicians in the US, and the media in the UK, the correlation goes the other way. One straightforward explanation for this is ideological. There is a reluctance to acknowledge market failure, and a distrust of state intervention as a means to correct it. This is what links climate change denial with a denial that countercyclical fiscal policy can work.

If this is the connection between the two, then a further point follows. With ideological roots, antagonism towards countercyclical fiscal policy should persist even in circumstances where debt is not seen as a major problem. It will not be enough to argue that high debt prevents the use of fiscal expansion (as in Section 3), because this will imply that if debt was not a constraint, such action could be used. Instead, ideology will require an attack on the effectiveness of fiscal policy per se. That, as we have seen, does appear to characterise at least part of the debate.

The climate change debate may also be revealing in what it tells us about how academic views are translated into policy positions and media debate. Scientists who dispute climate change are a tiny minority. This minority receives political support from those who have ideological motives for disputing the science. This in turn encourages the media to portray issues where there is a broad academic consensus as one that is instead controversial. This makes it difficult for both the public and uncommitted policymakers to establish that there is an academic consensus. (Conway and Oreskes, 2010, extend this point to other areas like the link between smoking and cancer. Krugman, 1994, looks at how extreme views about the Laffer curve became public policy during the Reagan period.)

Making ideological positions transparent

For those old enough to remember the Keynesian, Monetarist and New Classical debates, the idea that macroeconomics might be influenced by ideology will seem straightforward. However during that period the ideological associations of different schools of thought were transparent and widely discussed. Macroeconomists today are much more reluctant to suggest that policy positions taken by their colleagues have ideological roots.

In part the reasons for this are understandable. The new neoclassical consensus model based on microfoundations has brought a degree of unity to macroeconomic analysis, and has given the discipline more of a claim to be a science. It is precisely the existence of this common core of theory that helped establish, in the previous section, that countercyclical fiscal policy clearly worked, particularly at the ZLB. However, there is also a danger in conducting academic debates that ignore the elephant in the room.

This is particularly true in the case of fiscal expansion following the recession. While Stern may be correct in describing climate change as the largest externality ever faced by mankind, there is a good case for describing the recent recession as a massive financial market failure. To then argue that fiscal stimulus should be rejected because it represents undesirable state intervention in an otherwise well functioning market system would indeed be a hard sell.

6. Conclusions

Most views of macroeconomic policy in the 1930s are not very charitable. While policy following the recent recession has undoubtedly improved on this performance, I believe it should still be judged a failure. In many countries the recovery from recession will not be as rapid as it could have been. This policy failure primarily reflects an inability to use countercyclical fiscal policy aggressively enough during the recession, and the switch to fiscal austerity once the modest recovery had begun. In part this failure is a result of past fiscal indulgence--the deficit bias that was observed during the last quarter of the 20th century. However this paper argues that all the blame cannot be laid at this door.

If excessive levels of debt were the only problem, we should have seen much more effort to explore various types of balanced budget fiscal expansion. We should have seen more fiscal expansion in countries, like the US, that appeared to be able to borrow at very low rates. We should not have seen a debate on whether countercyclical fiscal policy was effective, particularly when macroeconomic theory clearly tells us that it can be. I draw an analogy with the debate over climate change. Although certain views are held for strong political or ideological reasons, the most effective method to advance those views is to suggest the science is controversial when in fact it is not.

No doubt some of the explanation for the switch from fiscal expansion to austerity reflects simple political motives, like a desire to reduce the size of the state, or to get as much fiscal pain out of the way as possible before elections. However I have tried to suggest that it also has ideological roots, reflecting a view that the market economy is inherently self correcting and state intervention is unnecessary and undesirable. This ideological view continues to influence how macroeconomics is taught and discussed.

It is important that we recognise how ideology still influences what and how macroeconomics is taught, but it is equally important that we acknowledge this influence in contributing to the failure of macroeconomic policy following the recession. To put it very simply, this paper argues that without this ideological influence within the discipline as well as the public debate, the recovery from recession could have been significantly faster, achieving a substantial improvement in social welfare.

REFERENCES

Aiyagari, S.R., Marcet, A., Sargent, T.J. and Seppala, J. (2002), 'Optimal taxation without state-contingent debt', Journal of Political Economy, 110(6), pp. 1220-54

Arestis, P. (ed.) (2007), Is There a New Consensus in Macroeconomics?, Palgrave Macmillan.

Barrell, R. and Weale, M. (2009), 'The economics of a reduction in VAT', Fiscal Studies, 30, 1, pp. 17-30.

--(2010), 'Fiscal policy, fairness between generations, and national saving', Oxford Review of Economic Policy, 26, pp. 87-116

Bi, H., Leith, C. and Leeper, E. (2010), 'Stabilization versus sustainability: macroeconomic policy tradeoffs', mimeo.

Brendon, C. (2010), 'Demand-deficient unemployment in equilibrium', mimeo, Oxford University.

Broome, J. (1992), Counting the Cost of Global Warming, Cambridge, The White Horse Press.

Calmfors, L. and Wren-Lewis, S. (2011), 'What should fiscal councils do?', CESifo Working Paper Series no. 3382 and Economic Policy (forthcoming).

Conway, E.M. and Oreskes, N. (2010), Merchants of Doubt, Bloomsbury Press, New York

DeLong, B. (2010), 'Is macroeconomic hard?', blog, June 28.

--(2011), 'Is economics a discipline', blog, March 20.

Eggertsson, G.B. and Woodford, M. (2004), 'Optimal monetary and fiscal policy in a liquidity trap', NBER International Seminar on Macroeconomics.

Eser, F., Leith, C. and Wren-Lewis, S. (2008), 'When monetary policy is all you need', Oxford Economics Department Discussion Paper No. 430.

Goodfriend, M. and King, R. (1997), 'The new neoclassical synthesis and the role of monetary policy', NBER Macroeconomics Annual, 12, pp. 231-96.

IMF (2010), 'Staff Report for 2010 Article IV Consultation', October.

Kirsanova, T. and Wren-Lewis, S. (2011), 'Optimal fiscal feedback on debt in an economy with nominal rigidities', Economic Journal (forthcoming).

Kirsanova, T., Leith, C. and Wren-Lewis, S. (2009), 'Monetary and fiscal policy interaction" the current consensus assignment in the light of recent developments', Economic Journal, 119, pp. 482-96

Krugman, P. (1994), Peddling Prosperity, Norton.

Lane, P. (1998), 'Irish fiscal policy under EMU', Irish Banking Review, Winter.

Leith, C. and Wren-Lewis, S. (2000), 'Interactions between monetary and fiscal policy rules', Economic Journal, 110, pp. 93-108.

Marcet, A. and Scott, A. (2008), 'Debt, deficits and the structure of bond markets ', Journal of Economic Theory, 144 (2), pp. 473-501.

Nickell, S (2006), 'The Budget of 1981 was over the top' in Booth, P. (ed.), Were 364 Economists All Wrong?, Institute of Economic Affairs.

Smith, D. (2009), '364 Economists can be wrong', ESRC Society Today.

Stern, N. (2006), The Economics of Climate Change, Cambridge University Press.

Woodford, M. (2011), 'Simple analytics of the government expenditure multiplier', American Economic Journal of Macroeconomics, 3, pp. 1-35.

Wren-Lewis, S. (2009), 'Bringing undergraduate macroeconomics up to date', in Fontana, G. and Setterfield, M. (eds), Macroeconomic Theory and Macroeconomic Pedagogy, Palgrave Macmillan.

--(2011), 'Internal consistency, nominal inertia and the microfoundation of macroeconomics', Journal of Economic Methodology (forthcoming).

NOTES

(1) Some may argue that this period was just lucky and demonstrated nothing, or worse still contained the seeds of its subsequent destruction. However, whatever truth there might be in those arguments, it seems highly unlikely that we will go back to how monetary policy was practised and discussed in the 1980s.

(2) At a meeting of Treasury economists after the budget chaired by Sir Terry Burns I argued that the only way the budget made sense was as a way to further weaken organised labour. Perhaps, as one of the small number of economists working on the economic effects of the budget in the Treasury at the time, I was trying too hard to rationalise what was being done. However I would still contend that the view I expressed at the time makes more sense than the idea that fiscal contraction was required to produce a better macroeconomic outcome.

(3) Once again there are differences between countries, with remarkably little additional unemployment in Germany, an increase below what might have been expected given the decline in GDP in the UK, and something pretty standard in the US.

(4) Although UK inflation is currently well above target, this can largely be explained by past exchange rate movements, higher commodity prices and tax increases. UK wage inflation is low and real wages are falling.

(5) There is a maximum tax revenue implied by the Laffer curve (see, for example, Bi, Leith and Leeper, 2010), but it seems likely that the political constraint will bite first.

(6) 23/7/2010.

(7) A similar argument might lead a government itself to favour austerity now rather than later. If the electorate has a short memory, and is influenced by GDP growth as much as its level, then it might be politically expedient for a newly elected government to cut sooner rather than later, even when this was not justified from a macroeconomic point of view.

(8) In the case of Ireland, see Lane (1998) for example.

(9) We can generate exactly the same result if government spending rather than taxes is the fiscal instrument, and the costs of being away from the optimal level of government spending are convex.

(10) In macroeconomics we can see this in the literature on dynamic inefficiency in OLG models. All the emphasis is on the possibility that interest rates are too low, because only then can Pareto improvements be made. However if interest rates are too high, the current generation may in effect be exploiting future generations.

(11) Unconventional monetary policy in the form of commitments to raise future inflation above target are discussed below.

(12) Fiscal policy will also be effective in a fixed exchange rate regime for well known reasons.

(13) I take it as obvious that the recession represented a large negative demand shock. It is possible to make an argument that it can be considered as a negative productivity shock, to the extent that finance is seen as necessary for working capital required for production. However there are also obvious reasons why a lack of finance will reduce the demand for consumption and investment goods, which is precisely what we have seen.

(14) In this sense, output always follows demand. The question is whether there is some mechanism for ensuring that demand follows supply.

(15) It would obviously be wrong to infer that because wages are not zero, there must exist such a mechanism. An alternative reason for positive wages when there is excess supply of labour is that something stops wages falling. After all, the problem in this simple model is not that real wages are too high: the real wage is set by the constant mark-up of firms.

(16) If expectations are forward looking, and we impose the transversality condition that prices reach their target, then reaching an equilibrium in which excess demand is eliminated is more likely, although not inevitable. But by imposing this transversality condition are we not assuming the result we are trying to demonstrate? Suppose instead that long-run inflation expectations move with nominal interest rates. In that case we can have multiple equilibria where output is demand constrained. (For further analysis, see Brendon, 2010.)

(17) They also show that even if this unconventional monetary policy worked, it would still be inferior in welfare terms to the use of countercyclical fiscal policy; the unconventional monetary policy is only effective because inflation in the future is above target. For this reason it seems legitimate to lay the blame for the current weak recovery at fiscal rather than monetary policy's door.

(18) These come from DeLong (2011), who laments the extent to which Chicago macroeconomics today appears to sideline the macroeconomics of Milton Friedman.

(19) Students often wonder why we cannot simply have a money financed fiscal expansion. The problem here is that the money created by the central bank buying government debt will be destroyed once the economy recovers. (If it is not, then printing money does become inflationary.) That means selling the government debt back to the private sector, and as the debt problem is long term, nothing will have been achieved on this account by temporarily parking that debt with the central bank.

Simon Wren-Lewis, Economics Department and Merton College. University of Oxford. e-mail: simon.wren-lewis@economics.ox.ac.uk. I am grateful to Charles Brendon, Paul Levine, Peter McAdam, Joe Pearlman, David Vines, and two anonymous referees for helpful conversations and comments, but responsibility for the views expressed here is entirely mine.

doi: 10.1177/0027950111420919