Tax cuts vs spending vs helicopters

Some people still seem unable, or maybe unwilling, to understand the basic New Keynesian (NK) model. Should it be surprising in this model that cutting taxes on wages at the Zero Lower Bound (i.e. when nominal interest rates are fixed) are contractionary? Of course not. The basic NK model contains an intertemporal consumption function that implies Ricardian Equivalence holds, so consumers save all of the extra income they get from a tax cut. But cutting taxes increases the incentive to work, thereby increasing labour supply, which through a Phillips curve decreases inflation. With a fixed nominal interest rate that implies higher real rates, which are contractionary. QED.

Now the main thing not to like here is the consumption function and Ricardian Equivalence. Empirical evidence points strongly to a significant income effect, with a marginal propensity to consume around a third rather than zero. There are good theoretical reasons why you might get this result, even with totally rational consumers. But the implication that cutting taxes will lead to some increase in labour supply seems reasonable, and that will put some downward pressure on inflation. This is why pushing ‘structural reforms’ that expand the supply side in a liquidity trap can be counterproductive in the short term. (Things are more complex when you have a fixed exchange rate.)

Now you may quite reasonably believe that in the real world a positive income effect from a tax cut will raise demand by more than any increase in supply, so inflation will rise and real rates will fall. But it remains the case that as a stimulus measure directly raising demand through higher government spending does not generate this supply side offset. That the NK model has this feature seems like a virtue to me. The only point I have to add is that because helicopter money, as traditionally envisaged, is a lump sum transfer (everyone gets an equal amount, so it is independent of wages), you do not get this offsetting supply side effect. So for that reason helicopter money is more effective as a stimulus instrument in a liquidity trap than cutting taxes on wages.

Evidence for monetary offset

As Tony Yates among others has observed, antagonism towards using fiscal policy for macroeconomic stabilisation seems to be an essential part of market monetarism. However their argument is not that fiscal policy will have no impact on demand and therefore output, but rather that monetary policy can always offset this impact. This can be called the ‘monetary offset’ argument.

As I have noted before, the idea of monetary offset is actually a key part of Keynesian objections to austerity in a liquidity trap [1]. In a liquidity trap monetary policy’s ability to offset fiscal austerity is severely compromised, but at other times it can be done. It therefore makes much more sense to postpone austerity until a time when monetary offset is clearly possible. So the idea that monetary offset can happen is common ground. What is in dispute is the extent to which a liquidity trap (or almost equivalently the fact that nominal interest rates cannot become too negative) prevents complete monetary offset.

If empirical evidence could be found that complete monetary offset has operated during a liquidity trap that would be powerful support for the market monetarist case. Scott Sumner recently presented (HT Nick Rowe) some analysis by Mark Sadowski which he said did just that. Taking the cyclically adjusted primary balance as a measure of fiscal policy, it showed that there was no correlation between this and growth in nominal GDP in the single period from 2009 to 2014 for those countries with an independent monetary policy.

There are tons of problems with simple correlations of this kind, some of which I discuss here, which is why quite elaborate econometric techniques are nowadays used to assess the impact of fiscal policy. But there is a rather simpler problem with this correlation. As far as I know, no one had expressed a concern about fiscal austerity because of the impact this will have on nominal GDP. The issue is always the impact on real activity, for reasons that are obvious enough.

So what happens if we relate fiscal policy to real GDP growth, using Sadowski’s data set? Here is the answer.

There are two obvious outliers here: at the top Singapore, and to the right Iceland. Exclude those and we get this.

There is a clear negative correlation between the extent of fiscal tightening and the amount of real GDP growth. Strange that Sumner gave no hint of this :)

Do I think this is definitive evidence? No, for two reasons. First, the obvious problems with simple correlations of this kind noted earlier. Second, this sample includes quite a few countries where interest rates over this period have averaged over 2% (Australia, Norway, New Zealand, and Korea) and so are unlikely to be subject to a liquidity trap. Others may only have been in a liquidity trap for a part of this period. What we can say is that these correlations are perfectly consistent with the view that austerity reduces growth in countries with an independent monetary policy. [2]

If you were to conclude that we just do not have enough data to know to what extent monetary offset can operate in a liquidity trap, I think you would be right. If you then went on to say that therefore the data cannot discriminate between the two sides in terms of policy, you would be wrong. What market monetarists want you to believe is that there is no need to worry about fiscal austerity in a liquidity trap, because an independent monetary policy can and will always offset its impact. This is wrong, precisely because the empirical evidence is so limited. We know, both from theory and the great majority of econometric studies, that fiscal contraction has a fairly predictable impact in reducing GDP. We have virtually no idea how much unconventional monetary policy is required to offset this impact. Given lags, that means trying to achieve monetary offset in a liquidity trap is always going to be hit and miss. The moment you think about uncertainty, the market monetarist argument for not worrying about austerity in a liquidity trap falls apart. 

[1] It is not the only reason why fiscal austerity in a severe recession might be a bad idea. There is a lot of empirical evidence that the impact of austerity is greater in recessions than when the economy is stronger, and there are other theoretical reasons besides monetary offset why that may be the case. This is of some importance for individual economies in a monetary union.

[2] If the coefficient on fiscal policy was lower for this sample than for Eurozone countries (I’ve not tried), would that at least be evidence for some monetary policy offset? The trouble here is that some of the countries driving the EZ results were also suffering from an overvalued real exchange rate as a result of earlier excess demand, and so this might bias upwards the coefficient on fiscal policy in those regressions.

Telling lies

What do you do when a well known macroeconomics blogger says you have made a claim which you have never made? You have in fact clearly said the opposite, and the claim you are supposed to have made is obviously silly. Ignore it maybe? But then you get comments on your own blog expressing surprise at how you can make such a silly claim. There is only one thing you can do really - write a post about it.

Some background which is important because it makes it clear why this is no simple misunderstanding or mistake. I had been reading stuff about how US growth in 2013 refuted the Keynesian position on austerity. 2013 was the year of the sequester, when many economists had voiced concerns about how a sharp fiscal contraction could derail US growth. Growth in 2013 turned out to be modest, and this led some to argue that this modest growth had refuted Keynesian economics.

So I thought I’d do a simple calculation, discussed here. I took the data series for government consumption and investment (call it G), and computed what growth would have been if we assumed an instantaneous multiplier of 2 and no austerity. If the 2013 experience really did refute the Keynesian position, then my counterfactual calculation of growth without austerity would have given some implausibly large number. I choose a multiplier of 2, because that rather large number would give the Keynesian analysis a real test. I did the same for earlier years. The counterfactual number I got for 2013 growth was 3.7%, rather than actual growth of 2.2%, with similar growth rates for earlier years. Hardly an implausible number for a recovery from a deep recession with interest rates still at zero, so no obvious refutation.

Scott Sumner then wrote a post where he said three things in particular.

1)    “Simon Wren-Lewis also gets the GDP growth data wrong”

2)    “He claims that RGDP growth was 2.3% in 2012 and 2.2% in 2013”

3)    “austerity began on January 1st 2013”

(1) and (2) were a rather strange way of saying that I should have used Q4/Q4 growth rather than annual growth. On (3), I wrote a new post simply plotting the data series I had used, which shows a pretty steady fiscal contraction starting in 2011 and continuing in 2013. [1]

Which brings us to his latest post. He writes, about those two original posts:

“He [sic] second claim is to deny that austerity occurred in 2013.”

He goes on to say that this claim is absurd, which of course it is. The only problem is that I never made it, or anything like it. In fact I obviously thought the opposite.

Could this be a simple misunderstanding? There are two reasons why not. The first was that my counterfactual with no fiscal contraction had raised growth from 2.2% to 3.7% in 2013. That would not happen if there had been no austerity in 2013. The second was that I had reproduced the data which clearly shows continuing austerity in 2013! So this was no misunderstanding, or even exaggeration. It is difficult to know what else to call it other than a straightforward lie.

[1] Sumner also says in this latest post that I’m using the wrong variable: rather than G I should use a more comprehensive measure including taxes and transfers, because G is not the variable used to measure austerity in the Keynesian model. But he must know that macroeconomists use both G and some measure of the deficit to look at short term fiscal impacts, for a simple reason. Consumers can smooth the impact of tax and transfer changes, while the impact of G is direct. So equating a $ worth of cuts in G with a $ tax increase, as a deficit measure would do, is wrong: particularly if timing is important, which in this case it is.

So using G is a pretty standard thing to do. In this case, however, it does not seem to make much difference. Here is the IMF WEO series for the US structural deficit. It shows a very similar pattern to G. Austerity starts in 2011, and continues thereafter. 

Speak for yourself, or why anti-Keynesian views survive

“The evidence for the Keynesian worldview is very mixed. Most economists come down in favor or against it because of their prior ideological beliefs. Krugman is a Keynesian because he wants bigger government. I’m an anti-Keynesian because I want smaller government.”

Statements like this tell us rather a lot about those who make them. As statements about why people hold macroeconomic views they are wide of the mark. Of course there is confirmation bias, and ideological bias, but as the term ‘bias’ suggests, it does not mean that evidence has no impact on the views of the majority of academics.

The big/small government idea makes no theoretical sense. Why would wanting a larger state make someone a Keynesian? Many Keynesians, and most New Keynesians, nowadays acknowledge that monetary policy should be used to manage demand when it can. They also know that any fiscal stimulus only works, or at least works best, if it involves temporary increases in government spending. So being a Keynesian is not a very effective way of getting a larger state.

It is also obviously false empirically. In the UK and US a large majority of economists appear to hold Keynesian views. I think it rather unlikely that a similar majority want a large state, and I can think of some notable Keynesians who clearly do not. Central bank models are typically Keynesian. Does that mean central banks want a larger state? No, it means the evidence suggests Keynesian economics works.

Russ Roberts says more recently:

“The evidence is a mess leaving each of us free to cherry-pick what sustains our worldview be it ideological or philosophical or just consistent with our flavor of economics.”

Ryan Bourne of the Institute of Economic Affairs goes further:

“when the facts change, the Keynesians don’t change their minds.”

To illustrate their belief that Keynesians ignore awkward facts both the authors above use the example of US growth following the 2013 sequester. (In my experience anti-Keynesians tend to shy away from data series, and especially econometrics, and prefer evidence of the ‘they said this, and it didn’t happen’ kind - particularly if ‘they’ happens to be Paul Krugman.) The problem is that this episode actually illustrates the opposite: that anti-Keynesians are so keen to grasp anything that appears to conflict with Keynesian ideas that they fail to do simple analysis and ignore others that do.

In this post I just looked at the data and did some simple arithmetic to show that this episode was quite consistent with Keynesian fiscal policy analysis. I’m sure others have done the same. But such analysis just gets ignored: they have a superficially good story, and that is all that matters. (Read this post to see how Scott Sumner in response to my work dug himself an even deeper hole.)

Why do we have to go over, yet again, that the clear majority of studies show that Obama’s stimulus worked. Why do we have to keep going over why UK growth in 2013 does not prove austerity works? Why do these people never mention the meta studies that confirm basic Keynesian analysis of fiscal policy? Because they want to believe that the “evidence is a mess” so they can carry on holding their anti-Keynesian views.

Parts of the political right have always had a deep ideological problem with Keynesian analysis. As Colander and Landreth describe, the first US Keynesian textbook was banned. New Classical economists, for all the many positive contributions they brought to macro (in the view of most mainstream Keynesians), also tried to overthrow Keynesian analysis and they failed. 

When anti-Keynesians tell you that support or otherwise for Keynesian macroeconomics depends on belief about the size of the state, they are telling something about where their own views come from. When they tell you everyone ignores evidence that conflicts with their views, they are telling you how they treat evidence. And the fact that some on the right take this position tells you why anti-Keynesian views continue to survive despite overwhelming evidence in favour of Keynesian theory.

Faith based macroeconomics

When you just know something is true, like fiscal policy never matters much and NGDP targeting would have avoided the Great Recession, everything becomes about proclaiming your faith in the most effective way possible. It becomes a debating contest. The best example I know of someone like this is Scott Sumner. Here is what he had to say about something I wrote recently.

“Simon Wren-Lewis also gets the GDP growth data wrong, in a way that makes austerity look worse. He claims that RGDP growth was 2.3% in 2012 and 2.2% in 2013 (the year of austerity in the US.) But that’s annual y-o-y data, and since the austerity began on January 1st 2013, you need Q4 over Q4 data. In fact, RGDP growth in 2012, Q4 over Q4, was only 1.67%, whereas growth in the austerity year of 2013 nearly doubled to 3.13%.”

The italics are mine. When you read that someone got the data wrong, or that they claim the data is whatever, you expect to find that they made an excel error, or used old data. But Sumner is not using ‘wrong’ and ‘claim’ in their ordinary sense. He is in debating mode. What he means is that by choosing to use the (correct) annual data, I’m (accidentally, deliberately?) hiding something important. He then quotes two figures that supposedly prove his case. No analysis, no graphs – it’s a debate.

Well here is a graph of US real government consumption expenditure and gross investment, taken from FRED.

According to Sumner “austerity began on January 1st 2013”. Now look at the graph.

It gets worse. Tyler Cowan quotes from Scott’s post with approval, I guess because the guy shares the faith.

Now do you really want to follow those whose macroeconomics is so faith based that they do not even need to check the numbers? Do you want to follow someone who says (earlier in the post) “it would be useful to do a more systematic study of fiscal austerity”. What about the many studies that have already been done (e.g. here, here or here). Do they not count because they generally find that fiscal policy can matter a lot, and so fail to accord with the faith? Do you want your macroeconomics derived from faith or from careful academic analysis?  

Unconventional Monetary Policy versus fiscal policy

In a previous post I explained why, in a very simple setting, it was best to use lower interest rates to stimulate demand, but that both tax cuts and increases in government consumption could do this job as well, with welfare costs that were minor compared to the cost of inadequate demand. So, to use a bit of jargon, cutting interest rates is first best, but if that first best was not available because nominal rates had hit zero then fiscal policy should be used. If there was a financial constraint on the size of the stimulus, government spending was generally more effective than tax cuts.

What about unconventional monetary policy? There are two main kinds: forward commitment to above target inflation (and a positive output gap) in the future, and printing money to buy various kinds of assets (QE). In each case I want to compare the welfare costs of these policies with the costs of using fiscal policy. However there is also the issue of uncertainty of impact: we need to know how much of a policy measure to apply: this uncertainty issue was not critical in the previous post because we have a lot of evidence about the impact of conventional monetary and fiscal policy. I will consider each type of unconventional monetary policy in turn.

One way of stimulating demand when interest rates are stuck at zero is to promise a combination of higher than ideal inflation and higher than ideal output in the future. (This can be done either explicitly or implicitly by using some form of target in the nominal level of something like nominal GDP. For those not familiar with how this works, see here.) The cost of this policy is clear: higher than ideal future inflation and output. Once again, these costs can be worth it because of the severity of the current recession, which is why nominal rates are stuck at zero. Whether these costs are greater or less than the cost of changing government spending is debatable: a paper by Werning that I discussed here suggests optimal policy may involve both.

One issue that arises with this particular policy is the problem of time inconsistency. The central bank may promise to raise inflation above target in the future to help reduce the recession today, but once the recession is over will it keep to its promise? Will the public let it? If people think it might not then the policy will be less potent, which increases the uncertainty associated with the policy’s effectiveness. This is one reason why it may be useful to hardwire the policy by means of some nominal target. [1]

The other unconventional monetary policy is QE: printing money to buy assets. Now it could be that this policy is doing nothing more than signal forward commitment to lower interest rates in the future, which moves us back to the previous discussion. Suppose it is more than that. I think a largely unresolved problem is how distortionary this policy is.

For example, in one of the most popular models that has explored the effectiveness of QE by Mark Gertler and Peter Karadi, the central bank makes loans or buys government debt. In doing this it reduces a risk premium, which is welfare improving. This raises the obvious question of why QE is not permanent. The authors get around this problem by assuming that the central bank is less efficient than private banks in knowing which assets to buy. However I’m not sure whether anyone, including the authors, has any idea what these efficiency costs might be.

Perhaps these distortions are quite small. However this discussion illustrates a more serious problem with QE, which is that we still have no clear idea of its effectiveness, or indeed whether effects are linear, and what the best markets to operate in are. Announcements about QE clearly influence the market, but that could be because it is acting as a signalling device, as Michael Woodford has argued. Jim Hamilton is also sceptical. This strongly suggests that the uncertainty associated with the impact of QE is far greater than any uncertainty associated with either conventional monetary policy or fiscal policy.

Thinking about it this way, I cannot see why some people insist that unconventional monetary policy is always preferable to fiscal policy. In a comment on a recent Nick Rowe post, Scott Sumner writes “My views is that once the central bank owns the entire stock of global assets, come back to me and we can talk about fiscal stimulus.” What this effectively means is that it is better for one arm of the state (the central bank) to create huge amounts of money to buy up large quantities of assets than to let another arm of the state (the Treasury) advance consumers rather less money to spend or save as they like. This preference just seems rather strange, but maybe Lenin would have approved! 

[1] If a temporary increase in government spending is in fact believed to be permanent, its effectiveness at stimulating the economy largely disappears, but this is not a problem of time inconsistency. Another difference is that governments are increasing and decreasing spending all the time, whereas it is much more unusual for an advanced economy central bank to deliberately create a boom.

Comments on Comments

                I number of people have asked me why I have not replied directly to Scott Sumner’s criticism of this and subsequent posts of mine, particularly as he keeps claiming that I made some mistake. Well, for the record, I do sometimes make mistakes, and when they are pointed out I acknowledge them. However on this occasion my writing on this issue seems pretty consistent to me and as far as I’m aware error free. So, why no reply?

                Well, I did in fact leave a comment on Scott’s second post. Scott then wrote another post (rather than comment on my comment). I stopped at this point, partly because the subject matter appeared to be moving away from what Cochrane and Lucas said to other issues which were not obviously relevant to my original point. I think Brad DeLong nails it here. This is one of the problems with the ‘you were inconsistent here, and you have forgotten this here, but you are a professor at Oxford so I’ll give you the benefit of doubt’ sort of exchange. I think it can muddle rather than clarify an issue.

                So instead I wrote a few self contained posts which tried to throw light on some of the issues, but which also made sense on their own. The original quotes I looked at appeared to suggest that if taxes went up, consumption would immediately fall by the same amount (“it’s just a wash”). I pointed out in my original post that this will not happen because of consumption smoothing. What I had not anticipated is that some people might think that lower saving would automatically lead to an equal fall in spending on capital goods without any change in income (another wash). That is why I wrote the savings equals investment post, which explained why this would not happen. Some of the comments to my original post said hey, these guys are just assuming full employment, so I wrote this on that general issue. There also seemed to be some confusion in the debate on the difference between behavioural responses and equilibrium relationships, which Paul Krugman and subsequently Brad DeLong discussed, and which Chris Dillow brilliantly anticipated. As the debate went on, I thought I could clarify a point about multipliers and consumption smoothing (or ‘Old Keynesian’ and New Keynesian models), so I wrote this. I’m glad to see that John Cochrane is now less dismissive of fiscal stimulus, which leads Noah Smith to make observations about politics and macro that have some similarities to those in my original post.

                While I’m on the subject of comments, I should say something about comments on my own posts. I had not anticipated so many people reading my stuff, and therefore so many comments, and if I tried to answer them all I would have to neglect the day job. However I do read them all, and if there is a common theme that I would like to say something on, I’ll write a new post on it (like ‘Demand Denial and Ideology’). One exception is where someone points out an error in what I wrote, or something where in retrospect I think I have been misleading or unclear, in which case I think it is sensible to recognise that immediately by replying to the comment. So thank you to those who have left comments, as I do find them useful.