Voxplainer on Scott Sumner & market monetarism

I have to admit that I usually dislike Vox. The twitter parody account Vaux News gets it kinda right in my opinion—they manage to turn anything into a centre-left talking point—and from the very beginning traded on their supposedly neutral image to write unbelievably loaded "explainer" articles in many areas. They have also written complete nonsense. But they have some really smart and talented authors, and one of those is Timothy B. Lee, who has just written an explainer of all things market monetarism, Prof. Scott Sumner, and nominal GDP targeting. Blog readers may remember that only a few weeks ago Scott gave a barnstorming Adam Smith Lecture (see it on youtube here). Readers may also know that I am rather obsessed with this particular issue myself.*

So I'm extremely happy to say that the article is great. Some excerpts:

Market monetarism builds on monetarism, a school of thought that emerged in the 20th century. Its most famous advocate was Nobel prize winner Milton Friedman. Market monetarists and classic monetarists agree that monetary policy is extremely powerful. Friedman famously argued that excessively tight monetary policy caused the Great Depression. Sumner makes the same argument about the Great Recession. Market monetarists have borrowed many monetarist ideas and see themselves as heirs to the monetarist tradition.

But Sumner placed a much greater emphasis than Friedman on the importance of market expectations — the "market" part of market monetarism. Friedman thought central banks should expand the money supply at a pre-determined rate and do little else. In contrast, Sumner and other market monetarists argue that the Fed should set a target for long-term growth of national output and commit to do whatever it takes to keep the economy on that trajectory. In Sumner's view, what a central bank says about its future actions is just as important as what it does.

And:

In 2011, the concept of nominal GDP targeting attracted a wave of influential endorsements:

Michael Woodford, a widely respected monetary economist who wrote a leading monetary economics textbook, endorsed NGDP targeting at a monetary policy conference in September.

The next month, Christina Romer wrote a New York Times op-ed calling for the Fed to "begin targeting the path of nominal gross domestic product." Romer is widely respected in the economics profession and chaired President Obama's Council of Economic Advisors during the first two years of his administration.

Also in October, Jan Hatzius, the chief economist of Goldman Sachs, endorsed NGDP targeting. He wrote that the effectiveness of the policy "depends critically on the credibility of the Fed's commitment" — a key part of Sumner's argument.

But read the whole thing, as they say.

*[1] [2] [3] [4] [5] [6] [7] [8] [9] [10] [11] [12] [13] [14] [15] [16]

Video: Adam Smith Lecture 2014 by Scott Sumner

Scott Sumner's Adam Smith Lecture, delivered last week, is now online. For my money it's the best outline of his view of the Great Recession and the correct role of monetary policy that I have yet seen. Scott's presentation can be downloaded here, and Scott wrote about his trip for TheMoneyIllusion here.

What's a neutral monetary policy?

The Federal Reserve Bank of Richmond alerted me to a newish paper from one of my favourite economists, Robert Hetzel, entitled "The Monetarist-Keynesian Debate and the Phillips Curve: Lessons from the Great Inflation"—needless to say it's highly interesting and informative. One bit in particular prompted me to write this screed on neutrality in central banking and monetary policy.

In the Keynesian tradition, cyclical fluctuations arise from real shocks in the form of discrete shifts in the degree of investor optimism and pessimism about the future large enough to overwhelm the stabilising properties of the price system and, by extension, to overwhelm the monetary stimulus evidenced by cyclically low interest rates.

In the quantity theory [monetarist] tradition, cyclical fluctations rise from central bank behaviour that frustrates the working of the price system through monetary shocks that require changes in individual relative prices to reach, on average, a new price level in a way uncoordinated by a common set of expectations.

In the real-business-cycle [new classical] tradition, cyclical fluctuations arise from productivity shocks passed on to the real economy through a well-functioning price system devoid of monetary non-neutralities and nominal price stickiness.

From each of these perspectives, we can derive some sort of definition of monetary/central bank neutrality, as well as an idea of what policy the central bank should operate. It strikes me that only one view is plausible, but before I make the case for that view, I will make the case for a particular theory of "meta-neutrality", i.e. a way that we should think about neutrality, whatever our perspective. I think this is something that everyone should be able to agree on, but I think that once we've agreed on it one view becomes inescapable.

Nothing is neutral with respect to everything. In one of my favourite ever essays, Scott Alexander makes a very similar point about "safe spaces" (nothing can be a safe space for everything—safe spaces for, e.g. a safe space for a disadvantaged group cannot also be a safe space for no-holds-barred rational discussion). In the same way, a monetary policy that is neutral with respect to real interest rates might conceivably have to achieve this by non-neutrality with respect to say, exchange rates. So the interesting question is what economic variables monetary policy must be neutral with respect to for us to call it "neutral" with no qualifiers.

But what we really want to be neutral to is the microeconomic working of the price system and markets generally, which is a bit more complex than any particular macroeconomic variable we could point to. One way around the question is by thinking about what might be non-neutral to the workings of the price system. One answer is: menu and shoe-leather costs, typically associated with high inflation, but more accurately linked to high aggregate demand (nominal GDP) growth.

Both impose restrictions on price adjustment, especially if they are unexpected and hence not "priced in".Menu costs will stop firms re-pricing things as often as would be optimal, impeding price adjustment, whileshoe-leather costs (from the high nominal interest rates associated with high inflation and high NGDP growth) will stop people from holding as much cash as they otherwise would, distorting their consumption decisions.

On the other side, unexpectedly low NGDP growth, combined with "money illusion" in borrowers ("sticky debts") and workers ("sticky wages"), could cause other microeconomic problems—markets won't clear until people's information, expectations and plans have adjusted, i.e. until people realise that the fall in prices/wages is not a relative price adjustment but a fall in overall prices/NGDP.

Overall this suggests we should call a policy neutral without qualifiers not when it is perfectly neutral (which is impossible) but when it is the "neutrality maximising policy". In the words of David Beckworth "neither too stimulative nor too contractionary and is pushing the economy toward its full potential" or in the words of Alan Greenspan one that "would keep the economy at its production potential over time".

That is, one that balances the distortionary costs of high (particularly unexpectedly high) NGDP growth with the costs of low (particularly unexpectedly low) NGDP growth. Empirically, menu cost and shoe leather problems have never been large in the USA and UK when NGDP is ticking along at about 5%. By contrast, NGDP growth less than 2.5% is almost always consonant with stagnation, while NGDP growth of less than zero always means a recession—much bigger costs. This suggests policy, far from being unprecedentedly easy in the lacklustre post-recession recovery, was if anything on the tight side of neutral.

Two crucial final points:

1. Identifying the conditions that we'd want to see in the macroeconomy for a (relatively) undistorted microeconomy does not mean endorsing a particular monetary arrangement or regime. Whether we have a central bank or not, we'd want stable NGDP growth.

2. This 5% level is contingent on society-wide expectations. If long-term expectations held by borrowers, lenders, firms, consumers and workers were for 0% NGDP growth (e.g. the 19th Century), then 0% NGDP growth would be more likely to be the neutrality-maximising monetary policy.

A question for market monetarists

Market monetarism, as propagated most prominently by Scott Sumner's (excellent) blog The Money Illusion, argues that recessions come about due to a collapse in demand. This is a problem because prices cannot adjust downwards quickly. Instead of a costly adjustment period we can simply boost demand by announcing a target and credibly committing to do the necessary quantitative easing (buying gilts to inject money in the system) to achieve that target.

This makes a lot of sense. Markets are finding it hard to clear; we boost AD to put the situation back where it was; now markets find it easier to clear. But lots of the best market monetarists, including Scott, Lars Christensen and many others, argue that right now what we need is more stimulus, because the economy is still in a bad shape, and it is still due to a shortfall of demand.

Last Tuesday Professor George Selgin delivered an extremely interesting lecture at the Adam Smith Institute making the case for productivity driven deflation. He said he agreed with the market monetarists that there is "bad deflation"—the sort that means nominal rigidities stop markets from clearing—but there is also "good deflation", from productivity improvements—and this is not associated with unemployment, stagnant or falling GDP, or any other cyclical issue.

After the talk I quizzed him on whether he agreed with the market monetarists that even though the ideal is a rule-based system, as opposed to the current discretionary way policy is set, right now the best discretionary policy is more easing, because that's probably what the ideal rule would require.

Prof. Selgin disagreed, arguing that we didn't need easier policy, and if you look at the graph above there's at least apparent reason to agree with him. Nominal GDP—aggregate demand—is not only well above its pre-recession peak in the US, but is growing at an apparently steady rate, roughly in line with its long-term trend. If the high unemployment in the US is down to insufficient demand combined with nominal rigidities then why hasn't a long period of higher-than-pre-crisis demand brought unemployment back down.

According to Selgin, policy uncertainty and pro-cyclical strictness in enforcing regulations (particularly risk-weighted lending rules that rate Greek bonds as zero but loans to small business at 100%) are holding firms back from investing their cash piles in capital and it is this that is stopping the robust recovery. He made the point very convincingly and despite trying hard to argue against it I couldn't find a good reason to disagree, except that I hadn't seen a good measure of the importance of these two factors so it was hard for me to compute how big their influence really was.

But many market monetarists—along with New Keynesians and most others—seem very sure that insufficient demand is the overriding factor holding back recovery, in the US as much as the EU, UK and Japan (where NGDP growth is further below trend). So my very genuine question is: upon what arguments and/or evidence do they rest this belief?

Monetary policy still has teeth

A storm has erupted over the past few days in a lot of the economics blogosphere, over an article by Mike Konzcal, backed by Paul Krugman, which claimed that current economic developments were evidence that monetary policy wasn't all-powerful and boosting demand sometimes required fiscal intervention. The claim faced convincing push-back from Scott Sumner, Matt Yglesias, and Ryan Avent. Before I look at the specifics of the claim, I’ll outline a (heavily oversimplified but still broadly true) model of the economy to help us to understand the debate.

In economists' simplest model of the macroeconomy, aggregate demand (AD) and aggregate supply (AS) interact to produce the price level. At the onset of the financial crisis and recession, AD crashed. Usually a crash in demand would produce a movement along the supply curve until price and supply have both fallen to produce a new equilibrium.

But the biggest market in the economy is the labour market, and many nominal prices (especially one of the most important prices, wages) are sticky-downwards. This means that prices do not fall enough to equilibrate the market, and output stays far below where it could be (this is what economists call the output gap).

This is where fiscal and monetary policy come in. Since prices are stuck, we need extra AD to get back where we were before, at the original pre-recession level of output. In theory, both monetary policy (here I will just look at interest rates) and fiscal policy (cutting taxes or boosting spending) can have the same effect on AD.

As far as I know, pretty much every mainstream economist agrees with everything I've said so far. But a key Keynesian claim – which the Konzcal article was arguing for – was that monetary policy is ineffective in special situations. Nominal interest rates can only go to zero – beyond that point savers will simply stash their cash in their mattress. Real interest rates (taking into account inflation) can only go to zero minus inflation. This is called the zero lower bound.

Konzcal said the Federal Reserve’s ‘Bernanke-Evans Rule’ – which promises to keep interest rates low until unemployment falls below 6.5 per cent – has failed to outweigh the $85bn (£55bn) federal cutbacks as part of ‘sequestration’. His evidence is Friday’s GDP release, showing that the US economy grew 0.6 per cent in the first quarter (2.5 per cent sped up as if the quarter were a year) below expectations it would expand 0.7 per cent. Konzcal said this GDP report showed the US economy was stagnating, and that the B-E rule failed to outweigh the sequestration.

But his critics pointed out that this was a big jump in growth over the previous quarter, when government spending hadn’t been cut and the Bernanke-Evans rule had barely taken effect – and growth was under 0.1 per cent (or 0.4 per cent on an annual basis). They point out that by creating inflation – and future expectations of inflation – a central bank can boost AD with monetary policy even when at the bottom bound, reducing real interest rates even when nominal rates can’t fall further. And they argue that monetary policy is dominant; it can always overrule fiscal policy.

If Konzcal's critics are right, it has at least two major implications for the UK. One is that Ed Balls’ plan to slow the pace of austerity further while keeping the Bank of England’s two per cent inflation target would only shift output to the government sector, not boost growth. In fact, if he pressed the Bank into actually achieving their target (consumer price inflation has been above target for 39 successive months) it would mean lower demand and perhaps even a triple-dip recession, as they would have to roll back QE and hike interest rates.

A second upshot is that spending cuts have not harmed growth (though the distortions from tax hikes may have). This is because any fiscal austerity has been offset by the central bank. If the government had not cut spending, the central bank would have had to rein in inflation with less quantitative easing (electronic money printing that can be rolled back) – unless it wanted inflation even further above target.