Steve Williamson’s scepticism about empirical and saltwater macro

Steve writes wide-rangingly, taking issue with several aspects of empirical macroeconomics.

Steve goes for the modern VAR literature, which seeks to identify monetary policy shocks, and measures that these shocks have effects on real variables, and that their effects on all variables take time to their full effects.

He rightly points out that we might be concerned how VAR researchers identify those shocks. But, can we dismiss all of them?

The methods I know about comprise;  short-run timing restrictions embedded in the VAR [eg Christiano-Eichembaum-Evans];  long-run restrictions [Blanchard-Quah];  sign restricitions [Uhlig and many others];  narrative methods of identifying monetary policy shocks [Romer and Romer];  monetary policy surprise measures [Rudebusch; heavily criticised by Sims himself, but still] constructed from the gap between rate expectations and outturns;  external instruments [Stock and Watson, Mertens and Ravn…].

With the exception of the Uhlig work on sign restrictions, [contradicted by other, later work in a similar vein] I thought it was the case that this work found non-neutrality of monetary policy shocks.  And that there was a reasonable consensus about the lags.

We can quibble with probably all of the papers in this large literature individually, but I find the overall conclusions persuasive.

Moreover, all of the work tries to use theory to some extent to identify those shocks, and that ought to reassure someone like Steve.  For example, the early Blanchard-Quah work uses the insight that satisfactory models have the property that ‘nominal’ shocks should be neutral on real output in the long run.  Narrative measures are explicitly motivated by the theoretical distinction between expected, systematic changes in policy, and the rest.  Even timing restrictions are theoretically motivated – though perhaps most contentious.  And sign restrictions are restrictions that come from theoretical models.  Rarely are data allowed simply to ‘speak for themselves’.

Steve also seems to question whether the shocks that VAR researchers crave are interesting:  the proper concern for us being the consequences of a particular systematic rule for monetary policy.

In reply to this:  the unsystematic shocks, which if central banks had been doing their jobs properly in history would simply not be there, since they have no reason to be there [caveat:  save for reasons of experimentation], are the unfortunate accidents that allow us to identify the parameters of the underlying structural model [including those in and out of the policy rules in place in the past].  Once this is done, we can then evaluate what the effects of alternative systematic policies would be, and try to work out what the best one to adopt is.

Another part of Steve’s pessimism about empirical macro connects the irrelevance of ‘long and variable lags’ in a world where there is open and communicated forward-planning by policymakers, and private agents that may form expectations about future central bank actions.

I don’t see a fundamental problem here.

For a start, provided we accept that history is nonetheless full of monetary policy shocks, [with randomness in early vintages of data, why would it not?] we can characterise all this forward-lookingness in a model [for the sake of argument, say the NK model], and see what a shock does in that model, and compare it to the counterpart in the data [the VAR].

Second, if we think think that there are unsystematic announcements pertaining to future events – what the literature terms ‘news shocks’, we can try to identify those too.  And several have.

 

Advertisements

Spencer Dale on the new economics of oil

Duncan Weldon alerted Twitter to this very nice think piece by my old boss, now at BP.

One of the points he makes is how new reserve discoveries have been outpacing oil consumption.  And he uses that to question whether  the old Hotelling model of oil as an ultimately exhaustible resource is now not right for the job.

I wonder.

First, Spencer is making a practical point.  In principle, oil is an exhaustible resource nonetheless, given the pace at which current biomass generates it, relative to current consumption.

Second, as Spencer discusses, we might presume that sooner or later the world will get its act together to make sure that not all the carbon-generating fuel that we have is consumed, to limit climate change.  In which case, new discoveries don’t add to the amount of oil that can and will be consumed.

There’s a lot of uncertainty about how soon and how effective collective action will bite.  And there is the possibility that clean technologies may expand what’s burnable.  But right now, a rational oil price-setter/discoverer would surely start, at least from a precautionary principle, with the view that we won’t find a way to burn all of that oil.

But the science of climate change would seem to make Hotelling’s model of exhaustion, or some version of it, as relevant as ever.

Discovering more oil whose burning we ban would be like discovering oil in the middle ages.  During those times there were no doubt frequent discoveries of seeping oil and gas, but since they had few known uses, they could not relieve the energy scarcity at the time, which would have involved the quantity of burnable wood.

Jeremy Warner on the Charter, and sanctions

Two comments on this piece by Jeremy Warner in the Telegraph.  Jeremy remarks:

“there is no purpose to law without sanctions, and this one appears to have none beyond the extra capacity for embarrassment if it is broken.”

First, I suppose it’s hoped by the charter’s authors that successfully adhering to it will lower the cost of borrowing, other things equal.  So the sanction is higher taxes [fewer votes, more restless right wing MPs] for the same amount of public spending.

I’m not so sure the Charter can or will be adhered to, and whether trying might not lead to more uncertainty, rather than less, but leave that aside for the moment.

Second, the reputational costs Jeremy mentions are not to be sniffed at.  These seem to have served us reasonably well in the case of the legislation isolating the Bank of England’s monetary policy decisions from political interference.

Why the fiscal charter is not right

Ultimately I am against this charter for the same reason that I am against moves to legislate that the Fed should follow, or be monitored relative to a monetary policy rule.

Macroeconomic fiscal policy is, crudely, about weighing the demands of vigorous application of the stabilisers (automatic and otherwise) today, against the need to preserve room for doing the same at some point tomorrow.   Where this requires keeping the ratio of public debt to GDP from going above some limit beyond which the cost of debt finance might be expected to spiral, and inflation stability and the capacity to deficit finance to counter recessions would be undermined.

Most of the ingredients here are a matter of at least some controversy.  The likely safe limit for the debt-GDP ratio.  The size and frequency of recessions and financial crises.  The robustness of monetary policy objectives in the face of an otherwise necessary, extreme fiscal stimulus.  Crucially, the path of potential output – even the appropriate concept of potential output – about which the purpose of fiscal policy here regulated is to stabilise actual output.

For that reason, I don’t see how any sensible legislation could frame a fiscal rule like this.

The government have tried, but it falls down on many counts.  To give one example, the definition of ‘normal times’ when the surplus rule applies, is framed in terms of the annual growth rate of GDP of 1%.  Yet one thing almost all economists would surely agree on, is that the appropriate definition would be in terms of the difference between actual output and potential.  I can see why the legislation does not involve this concept, because it is such a conceptually and empirically elusive thing.  But I am not convinced that the solution to legislate in terms of growth rates is better than pure discretion.

Current circumstances seem to be exempt.  But imagine a repeat, 10 years hence.  After seven years more, interest rates are pressed against or close to the zero bound.  The economy is growing again, consistently, but many think that output is still below potential, and inflation is substantially below target.  A surplus in such times would be hazardous, to say the least.

To press on with this single example, its conceivable that a surplus might be needed even if the growth rate of output is less than 1%.  There is no economic law that potential output is an inevitable straight line process inexorably enriching us by 2% per year.  Sustained falls in output can be caused by falls in potential output, and during these times it may be necessary to have tight fiscal policy, where the loose fiscal policy allowed by this charter would be futile, inflationary and counter-productive.

As many others have pointed out, the shift from the previous version of the Charter to include public investment is perplexing.  It’s possible to sympathise with a political economy view that democracies, in which the unborn do not vote, may have a tendency to ‘invest’ without real expectation of future returns, simply as another form of covert, current spending.  But the Office for Budget Responsibility’s old terms of reference seemed adequate defence against that.  Likewise, it’s possible to sympathise with the view that public investment decisions should be made on a long-term basis, and that such investment should not rise and fall frequently with the tide of the cycle.  However, on occasion, the demands of macroeconomic crises will trump that desire.  And it may be reasonable now to expect real finance costs to be low for a very long time [look at long-term real rates now for instance], and that deficit finance to take account of this bounty, on projects that generate genuine future returns [meaning future tax revenues] would be justified for a long time to come.

All this is not to say I am against legislative constraints on the macroeconomic aspects of fiscal policy.  On the contrary, I am in favour.  But I think they should, [as they did until now] follow the model of the Bank of England.  Delegate, as far as democratic niceties allow, to an expert, independent body, a set of objectives, and give them the discretion to pursue them, rather than tie them to inappropriate and, therefore, ultimately non-credible rules.

In this case that could mean expanding the role of the OBR to comment on the appropriateness of the fiscal stance.  Not just in so far as it bears on long-term sustainability, its current mandate.  But whether, given a government defined objective, it weighs appropriately those competing demands, the demand to stabilise vigorously today, vs the demand to preserve the capacity to do so in the future.  In support of that it would form a view about the appropriate debt/GDP ratio, the output gap, the evidence on the likely size and frequency of recessions and crises [and what strain on finances fiscal stabilisers such crises would put].  Perhaps even and independent view of the efficacy of monetary policy tools that are the alternative.  Such legislation could even go so far as to stipulate that in the extreme event that interest rates are constrained at zero, the fiscal stance could be decided, or at least discussed openly with the Bank of England’s Monetary Policy Committee.

As some have pointed out, this Charter should not be considered a left-right issue.  Even fiscal hawks should be concerned not to legislate in this fashion, just as monetary policy hawks, frightened about inflation stability, should not try to tie a central bank to a monetary policy rule that is not foolproof.

More Koning on future marginal liquidity services

JP comments on my last post commenting on his post, and this post responds.

He notes that while a technological innovation that reduces the need for money, or the liquidity and convenience services that come from holding money, would have no effect on the actual marginal liquidity services enjoyed at a point when interest rates are zero, [because zero rates means that people are sated with money], nevertheless, if rates are expected to rise in the future, something will happen to future money demand.  But what?

Without doing it properly [read with pencil, paper and Matlab], I can’t know for sure.  But, at a guess, not much in the NK model.  Technological innovation will reduce real money demand at all levels of interest rates.  The central bank will follow through with any previously computed optimal trajectory for nominal interest rates following [eg] a shock to the natural rate, but will note that this rate leads to lower money balances throughout.  Those lower money balances would have no other consequence in this model, however.

 

Is monetary impotence due to monetary innovation? On JP Koning.

JP poses this question in yet another thought-provoking post.

I’ll offer one thought, from the standard New Keynesian model of money, interest and prices.  In this model, the fact that money confers liquidity services is captured, rather crudely, by assuming that we are all like coin collectors and get pleasure from holding real money balances.  JP’s thought experiment is undertaken in this model by wondering what happens if you slowly reduce the amount of pleasure all of us coin collectors get.

Woodford explains the title of his textbook ‘interest and prices’, in particular, the virtual absence of money, in his opening salvos, by pointing out that central bank control over interest rates, and thereby over the whole economy, does not depend on this ‘pleasure’ being of a certain magnitude.  In fact, the simple version of the model he studies is one in which this pleasure is conjectured to be infinitely small.  Monetary policy’s leverage here depends on money’s role as a unit of account.

Looked at this way, the tendency for the economy to get stuck in a liquidity trap in response to shocks to the natural rate of interest (for example) should not depend on the liquidity convenience (pleasure!) we get from money.

Another way we can interpret JP’s question, though is this.

Suppose that a shock to the natural rate comes along, requiring an extended period of interest rates at the zero bound [and perhaps other stimulus too], to return inflation slowly to target.  But then on top of that there is some monetary innovation lowering the liquidity services to money.  [Apple Pay?!]  What would happen then?  Looked at through this model, nothing at all.  Because the fact that interest rates are already at zero shows that these liquidity service benefits have already long been exhausted.

Here I have assumed that the monetary reform does not change the asymptote of these liquidity service benefits as real balances get very very large.  All that happens is that the monetary reform changes the profile of those marginal service benefits at finite values of real balances.

What if the reform that happens mid-way through responding to the natural rate shock did change the asymptote, however?  For example, what if the reform involved abolishing higher denomination notes, which increases the costs of managing cash?  I’m not absolutely sure here, without actually doing this properly, but I hazard this guess.  Nothing happens, much, if central banks hold interest rates where they are, except that they have to allow real balances to decline via a reduction in nominal money quantities.   The reform, however, will create an opportunity for the central bank to lower rates if it wants to, in case the old floor to rates [previously zero, now lowered] implied an amount of stimulus that was less than optimal.

The more interesting question – in fact the question that JP Koning actually posed – is what happens in the real world, of course.  But you’ll have to go elsewhere for answers to questions like that.

Is there really a credibility problem in the face of deflation?

Tim Young, in typically combative style, doubts in a comment on this blog that there really is a credibility problem in the face of deflation.  I accept that historically, and theoretically, we have focused on the difficulty of persuading the private sector that we will not generate too much inflation, thereby eroding the real value of nominal government debt, nor nominal wage contracts.

However, there are two genuine doubts that observers of a central bank might have.

First, there is the doubt that, whatever they say about how confident they are, central banks may not have potent or reliable tools to fight deflation in the face of the zero bound.  The Bank of England has reassured us that ‘we have the tools’.  But perhaps they haven’t.  We know that they have a vested interest in reassuring us because in so doing, they help anchor inflation expectations, and thus inflation itself, and thus make their lives easier.

Second, there lurks the suspicion in the popular mind that central banks are actually inflation nutters and would really like 0 inflation, rather than the 2 per cent actually mandated.  Sure, we had a history of governments in the UK and elsewhere generating too much inflation.  But monetary policy was then supposedly passed on to these central bank types to solve that problem.  Perhaps this was overdone?  If you think this is far-fetched, note that several official remarks about our sub-target inflation, and remarks by prominent commentators, have been approving of the benefits of 0 inflation.

So, for those reasons, I think Marvin Goodfriend was right to warn that there is a credibility issue in fighting deflation, or at least sub-target inflation, and, accordingly, a motive for monetary policy to lean towards the vigorously stimulative.