Why macroeconomists, not bankers, should set interest rates

More thoughts on the idea that interest rates ought to rise because of the possibility that the financial sector is taking excessive risks: what I called in this earlier post the BIS case, after the Bank of International Settlements, the international club for central bankers. I know Paul Krugman, Brad DeLong, Mark Thoma, Tony Yates and many others have already weighed in here, but - being macroeconomists - they were perhaps too modest to draw this lesson.

To most macroeconomists, the theory of monetary policy is pretty straightforward. Interest rates should be set at a level which closes the output gap, which can be defined as the level of output and unemployment that will keep underlying inflation constant. We can call this real interest rate the Wicksellian natural rate. The difficulty is not in the concept, but in the practice of putting numbers to this concept when inflation is noisy, the output gap is hard to estimate, there are lags in the system etc etc.

But, respond those putting the BIS case, wasn’t that what monetary policymakers thought they were doing in 2007, and look what happened next. Monetary policy cannot afford to ignore the financial sector, and the risk of excessive lending and bubbles that subsequently blow up the economy. There are signs, they say, that what happened in 2007/8 may be happening again now, so we need to raise rates to prevent another crash, even though there is still a negative output gap and inflation is below target.

Which might seem plausible, until you notice what is going on here. The implication is that a financial crisis only happens because interest rates are set at the wrong level. The Great Recession was all the fault of the Fed, who kept interest rates too low after the 2001 recession. The gradual deregulation of the financial sector in the decades before? - not an issue. The widespread misselling of subprime mortgages? - these things happen. All the other examples of misselling and fraud? - boys will be boys. An industry that profits from a massive implicit public subsidy? - we see no subsidy. Classifying subprime products as AAA? Massive increases in bank leverage in the 00s? - all the result of keeping interest rates too low.

When those putting the BIS case tell you that macroprudential controls (a.k.a. financial regulations) are ‘untested’ and ‘uncertain in their impact’, what they are really saying is that the financial system cannot be regulated to make it safe when interest rates are low. There is no evidence for that proposition, and a lot of history that says otherwise. We do not have to accept a deregulated financial sector which has the power at any moment to derail the real economy. But of course most working in the financial sector hate regulation. They have an interest in perpetuating different stories about the Great Recession. If you spend too much time around bankers, there is a danger that you come to believe these self-serving stories.

But, you might say, what harm would a modest increase in interest rates do? Again, basic macroeconomics, which I have not seen anyone putting the BIS case address. Raising rates implies in current circumstances a larger negative output gap, which will reduce inflation further below its target. As happened in Sweden, and accurately predicted by macroeconomist Lars Svensson. Two things could then happen. First, interest rates come back down again (in Sweden’s case by outvoting the governor for the first time since it gained its independence in 1999), but the cost of lost resources and higher unemployment created in the meantime can never be redeemed. Second, interest rates stay high for long enough that the public will conclude that the inflation target has in reality been revised down, and we risk converging to a deflationary steady state (technical discussion here), or in non-technical terms a Japan-like lost decade or more of low output and deflation.

To see clearly why this makes no sense, consider the symmetric case. Suppose someone argued, when inflation was above target, that we should not raise rates, but instead allow the output gap to be positive. I suspect those currently making the BIS case would scream disaster – it is the 1970s all over again. So why is that wrong but doing the same thing in reverse OK? In fact it is worse than that. If long run expected inflation rises, a central bank can always signal its true inflation target by sharply raising rates. In the opposite case it may not be able to, because of the Zero Lower Bound.

I like to praise the current UK government when I can. In setting up a Financial Policy Committee that is separate from the Monetary Policy Committee they did exactly the right thing. This formalises an assignment: macro prudential policy to control financial sector excess, and interest rates to control demand and inflation. Most macroeconomists know this makes sense. But the financial sector has a pecuniary interest in pretending otherwise. Those that get too close to that sector should be kept well away from setting interest rates. 

The financial instability argument for raising rates

There is a nice juxtaposition of recent articles in the Economist. This one, by P.W. (it is weird this convention they have for signing with just their initials), puts the “case against maxing out monetary policy” (a.k.a. raise rates now). The Fed, Bank of England and ECB “argue that the priority is to restore growth and to do battle against low inflation. But [they] grievously misread the risks before the financial crisis, which weakens their claim to be reading them correctly now.” Let’s call the proposition that we should raise rates now to avoid financial instability the BIS case, after the Bank of International Settlements who have been making this argument ever since the recession began. In contrast Ryan Avent writes that there are two big problems with this argument. I want to expand on his discussion, and be a little less polite.

I want to begin by conceding a point. Suppose, as a monetary policymaker, you believe a financial crisis is possible, and that by raising rates you may be able to prevent it. Assume, crucially, that there is nothing else you can do to help prevent the financial crisis. In that case, you will consider raising rates, even if inflation is below target. If you have just one instrument (interest rates) and two targets (inflation and preventing a crisis) you will be influenced by both targets. If you want this point expressed more formally, see this post and the paper by Mike Woodford it discusses. 

 

However that is not the end of the story. If you raise rates to prevent financial instability when inflation is below target, inflation will remain below target or may fall even further. You cannot ignore that. So if interest rates are raised today to head off a financial crisis, they will have to be lower in the future to deal with the lower inflation or even deflation you have caused.

This is not just what macroeconomic theory says. In mid-2010 the Swedish central bank started raising interest rates (from 0.25% to 2%), despite forecasts that inflation would stay below target and with unemployment well above its natural rate. They did this explicitly because they were worried about the build up of household debt and a possible housing bubble. Inflation began to fall, and since 2013 it has been at or below zero. As Lars Svensson has pointed out, even on its own terms this is not a very clever policy, because with lower inflation the real value of debt is higher than it would otherwise have been. But the key point for the current discussion is that now interest rates are coming down again (currently 0.75%), because you cannot ignore inflation being over 2% below target.

Some of those making the BIS case understand this. What they hope is that if interest rates are raised to, say, 2% and stay there, that will still give us enough monetary stimulus to eventually get inflation back up to target. It clearly was not correct in the Swedish case, and with Euro inflation still at 0.5% it looks pretty improbable there too. But maybe it could be correct for the US and UK. So by raising rates by a modest amount today we might prevent financial instability, but at the cost of delaying the recovery.

I want to make two observations that follow from the BIS argument. The first is that they are in effect saying that the Zero Lower Bound (ZLB) constraint on monetary policy is even more severe than we thought, because if we leave interest rates at the ZLB for too long this generates an unacceptable risk of financial instability. That in turn must strengthen arguments (pdf) for raising inflation targets above 2%. Strangely, I do not hear advocates of the BIS case also arguing for higher inflation targets. The second is that, the more severe the ZLB constraint is in practice, the more compelling the case for using fiscal stimulus when we hit the ZLB. (Fiscal policy has become more expansionary in Sweden.) Again, this is something you do not hear BIS advocates argue for – in fact they often push austerity.

As Ryan Avent says, we can avoid all these difficulties by adding an extra instrument, which is macroprudential regulation. If parts of the financial system appear prone to instability because people are taking insufficient account of risks, bring in controls (or maybe taxes) of various kinds to stop this happening. Now, as R.A. notes, those taking the BIS position counter that such measures are untested and may not be effective. Here is a typical example in the FT, where it is stated that “macroprudential policies will fail to stop investors taking irrational risks”.

So we must raise interest rates, and delay the recovery, because nothing else can stop some in the financial system taking excessive risks. To which I can only say, summoning all my academic gravitas, what audacity, what impudence! Not only have we had to suffer the consequences of the Great Recession because of excessive risk taking within a largely unregulated financial system, we now have to cut short our main means of getting out of that recession because they might do it again. I do not know what planet these people are on, but if its mine, can they please get off and play their games elsewhere.

The intellectual bankruptcy of the austerians

It is both amusing and tragic to watch the advocates of fiscal austerity try and deal with the fact that the thin intellectual foundations for their approach have crumbled away, while at the same time the empirical evidence of their folly accumulates. I say thin not for dramatic effect: the economic textbooks tell us that this policy was foolish in the first place, so the austerians were always arguing something like ‘forget the accumulated wisdom of the past century, on the basis of two or three papers we know better’.[1]

So we have the latest annual report from the Bank of International Settlements, which is telling central banks to stop all this unconventional monetary policy stuff and get back to the serious business of eliminating... no sorry, ‘controlling’ inflation. Ryan Avent has an excellent commentary on the same report from last year, most of which could equally apply to this year's edition. Here I just look in detail at one of the report’s six chapters on fiscal consolidation. (As Mervyn King once said, central bankers’ obsession is fiscal policy rather than inflation.) In a section on the costs and benefits, after a first paragraph outlining the case against, we have this:

“There are reasons to be sceptical about all these arguments. First, even if the short-term adverse effects of fiscal policy on output (or fiscal multipliers) are somewhat greater than in the pre-crisis period, there is considerable uncertainty about their magnitude and no compelling evidence that they are large enough to render fiscal consolidation more difficult (or actually self-defeating).”

Just look at the language here: ‘sceptical’, ‘somewhat greater’, ‘considerable uncertainty’, ‘no compelling evidence’. You could write a paragraph like this about virtually any economic idea. Notice also the target and how it changes: it starts off in the wrong place, by implicitly assuming that if multipliers were smaller the policy would be costless, and then at the end the target is the much stronger position that fiscal consolidation is self-defeating.

It gets worse. In the next paragraph we have:

“Second, other factors almost surely contributed to unexpectedly weak growth. Especially in the euro area, investors’ worries about fiscal sustainability and liquidity drove up sovereign bond yields, putting a strain on bank and sovereign balance sheets and leading to more restrictive credit conditions.”

The first sentence is a non-sequitur. And then we get the euro periphery countries, but no mention of how this crisis was brought to an end not by fiscal consolidation but by the actions of the ECB. You would think the BIS would want to give credit to a successful central bank policy! Next paragraph:

“Third, larger multipliers do not necessarily undermine the case for an early or relatively fast adjustment. The argument for back-loading or slowing the pace of fiscal consolidation relies on the expectation that fiscal multipliers will decrease in the future or that economic growth will rebound significantly. However, if these expectations do not materialise, shifting the bulk of fiscal consolidation to the future would mean greater debt and higher debt servicing costs, making future adjustment even more costly and prolonged.”

Notice the language again: “do not necessarily”. The only way to make sense of this paragraph is that we are going to be stuck at the zero lower bound forever.

In the next paragraph we get that old chestnut, credibility:

“The case for back-loading fiscal adjustment also relies on the credibility of fiscal plans. Current governments will have to make commitments on behalf of future ones.”

This is just wrong, unless you are only thinking about the Eurozone periphery again. For ‘backloading’ fiscal consolidation to work, current governments do not have to commit to anything. You just need to believe that future governments, once the recovery is assured, will focus on stabilising and then reducing debt, rather than allowing debt to explode. As they invariably do. [2]

Don’t worry, its coming to an end:

“Finally, the impact of fiscal consolidation on growth extends beyond the short run. By restoring sound financial conditions, eliminating the risks associated with high debt and reducing the resources needed to service the debt, consolidation will lead to higher sustainable economic growth. As a result, its long-term benefits will more than offset its short-term costs.”

This is more like a recitation of an article of faith than an argument. It first says reducing government debt is good for long term growth. Let’s accept that: the main arguments against austerity have never involved suggesting that long term high government debt is good for you. It’s all about when is the best time to stabilise and reduce government debt. It is the final sentence that just does not follow. Because there are long run benefits to reducing government debt, it must be the case that the sooner we start the better? No. Exercise is good for you, but you don’t start when you are down with the flu.

This kind of thing is amusing sport for academics who are trying to put off getting down to serious work on a Monday morning. But of course it is tragic because the austerians, even though they have been soundly defeated on the battlefield of ideas and evidence, continue to call the shots on policy nearly everywhere. What to do about this tragedy? Just keep plugging away at the arguments I guess, although a bit of ridicule also helps convey the poverty of the austerian argument. But it is also up to others with more effective voices to point out that this emperor has no clothes. 

[1] There may also be an element of Keynes famous academic scribbler in this too. (“Madmen in authority, who hear voices in the air, are distilling their frenzy from some academic scribbler of a few years back.”) There was a period - in the 1980s - when those economists studying at masters/PhD level might have been taught that the undergraduate textbooks were wrong, and that Keynesian economics was doomed. That proved to be a temporary aberration, Keynesian economics rejuvenated itself and in most departments it was restored to its rightful place at the centre of macroeconomic wisdom. But perhaps some of those now running policy did their PhDs during the 1980s, and never kept up with what happened subsequently.

[2] This is where obsession comes in. Occasionally a government fails to prevent debt exploding, and forces the monetary authority to monetise the debt, leading to hyperinflation. This is the central bankers’ nightmare. If you want to avoid this remote possibility at all costs, then you might be willing to sacrifice anything to do so, including the health of the real economy.