NGDP targeting: Hayek’s Rule

One thing I go on about on this blog is how nominal GDP targeting—a market monetarist policy proposal that has even won over a small group of New Keynesians—is also the kind of policy an Austrian should want in the medium term. Of course, in the long term we’d like to abolish the Bank of England altogether, but even then we’d get, with free banking, something like a stable level of nominal GDP, so it’s a pretty good target to work towards.

The economist Nicholas Cachanosky wrote a paper in the Journal of Stock & Forex Trading about a year ago, which I missed, called “Hayek’s Rule, NGDP Targeting, and the Productivity Norm: Theory and Application” which lays a lot of the Austrian arguments for targeting the level of nominal income in a very clear and cogent fashion. I include some key extracts below:

The term productivity norm is associated with the idea that the price level should be allowed to adjust inversely to changes in productivity. If total factor productivity increases, the price level (P) should be allowed to fall, and if total factor productivity falls, the price level should be allowed to increase. A general increase in productivity affecting the economy at large changes the relative supply of goods and services with respect to money supply. Therefore, the relative price of money (1/P) should be allowed to adjust accordingly. In other words, money supply should react to changes in money demand, not to changes in production efficiency.

The productivity norm was a common stance between monetary economists before the Keynesian revolution. Selgin [14, Ch 7,8] recalls that Edgeworth, Giffen, Haberler, Hawtrey, Koopmans, Laughlin, Lindahl, Marshall, Mises, Myrdal, Newcome, Pierson, Pigou, Robertson, Tausig, Roepke and Wicksell are a few of the economists from different geographical locations and schools of thought who, at some point, viewed the productivity norm positively.

One of the attractive features of productivity norm-inspired monetary policy rules is the tendency of the results to mimic the potential outcome of a free banking system, one defined as a market in money and banking with no central bank and no regulations. Among the conclusions of the free banking literature is that monetary equilibrium yields a stable nominal income.

Throughout Cachanosky distinguishes carefully between an NGDP target and a productivity norm, though I think these are overstated; and between ’emergent’ stability in NGDP and ‘designed’ stability, which he (like Alex Salter) thinks are importantly different (I am not convinced).

Cachanosky believes that the 2008 crisis implies that NGDP growth beforehand was too fast, and led to capital being misallocated, but I still doubt the Austrian theory of the business cycle makes any sense when you have approximately efficient capital markets.

Despite our differences, I think that Cachanosky’s papers are very valuable contributions to the debate, and hopefully they can go some of the way to convincing Austrian economists that the market monetarist approach is not Keynesian.

Nominal GDP targeting for dummies

Nominal Gross Domestic Product (GDP) targeting is a type of monetary policy that people like me think would give us a more stable economy than we currently have. It would replace the Bank of England’s current monetary policy, inflation targeting.

Nominal GDP can be understood as sum of all spending in the economy. Total spending can increase either because of price rises (inflation) or because there’s more stuff to go around (economic growth). If this year inflation is 2% and we have 2% economic growth, nominal spending (nominal GDP) will have risen by 4%.

The current policy of inflation targeting means that the Bank of England tries to control the money supply so that prices rise, on average, by 2% every year. If prices rise by more or less than this, the Bank is judged to have failed in its job.

Nominal GDP targeting would mean that the Bank of England would stop trying to target price rises, and instead try to target the total amount of nominal spending that takes place in the economy. That means that if economic growth was lower than usual, the Bank would have to try to make inflation higher than usual. If economic growth was higher than usual, inflation would be lower than usual.

This system is appealing because it is often the total amount of spending in the economy that matters, rather than inflation per se. Wages are usually set in nominal terms, which means that they do not automatically adjust upwards and downwards according to inflation.

Because of this, a drop in the amount of spending going on can lead to a mismatch between all the wage demands in the economy and the amount of money available to pay them. In other words, there is not enough money in the economy to pay everyone. This has two possible outcomes: either wages can be cut to meet the new level of spending, or people will have to be fired.

Empirically, it seems as if firms prefer to fire some workers than to cut wages across the board. In fact, firms really hate cutting wages, for some reason, and unemployed people are often reluctant to take the same job that they once had for a lower wage. Economists refer to this phenomenon as “sticky wages”.

So the outcome of a fall in total spending is usually unemployment. This is an example of a nominal change having a real effect, and destroys wealth that need not be destroyed, because the previously-profitable relationship between the worker and the firm has now been undone.

When this happens across the economy it can affect economic growth. In fact, this seems to be a very important factor in recessions – when there is a steady level spending taking place, the market is pretty good at finding new ways of using unemployed workers fairly quickly. When there just isn’t enough spending going on, we have to wait for workers and firms to cut wages enough to hire them again, which can take a long time.

Under nominal GDP targeting, the Bank of England would commit to keep the spending level growing even if economic growth dipped. As I’ve said, that would mean more inflation in times of slow growth and less inflation in times of quick growth.

Because inflation is being used to offset the changes in economic growth, negative economic ‘shocks’ like oil crises will translate into higher prices, prompting the market to adjust to take account of new realities, but never creating the domino effect of mass unemployment that we sometimes currently experience. The real economy would still adjust to real shifts in supply and demand, but we’d avoid the chaos that unstable monetary environments can create.

The key is that almost all contracts in the modern economy are set in nominal terms. That means that money that is managed in the wrong way can create a lot of unnecessary destruction of wealth. Nominal GDP targeting would probably give us the most neutral monetary system possible with the government, with the monetary environment kept stable so the real economy can do its work in allocating resources.

Money matters. The 2008 crisis happened because expectations of inflation, and hence nominal spending levels, dropped sharply, causing the ‘musical chairs’ problem of too little money to fulfil all the existing contracts and wage demands, which led to widespread bankruptcies and job losses. Today, the UK and the US have begun to get their spending levels growing at a healthy rate again, and their real economies have begun to grow healthily again too.

The Eurozone is the saddest story. The European Central Bank has been obsessed with fighting inflation (possibly because Germany has not suffered much, and Germans have bad memories of hyperinflation during the 1920s), and as a result nominal spending has grown very slowly indeed. The consequences are easy to see: in the weaker European economies, like Greece, Spain and Italy, unemployment is at historically high levels. It seems likely to stay there for many years.

Many people, myself included, believe that a system where private banks could issue their own notes without a central bank at all would be the best system. This is known as ‘free banking’. One of the best arguments for free banking is that it would keep nominal spending levels steady, because banks would issue more notes during periods of slow growth and fewer notes during periods of high growth. This should sound familiar – nominal GDP targeting is probably the closest we can get to ‘stateless’ money while having a central bank.

Nominal GDP targeting would not prevent all recessions or guarantee growth. The real economy is what determines things like that. But badly-managed money can destroy growth, create recessions by itself, and turn small ‘real’ recessions into extremely bad depressions, as happened in the 1930s and 2000s. Nominal GDP targeting would give us stable, neutral money that avoids these things. We would have been better off with it in 2008, and we would be better off with it today.

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.


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]

The eurozone is in dire need of nominal income targeting

It may well be that, in the US and UK, nominal GDP is growing in line with long-term market expectations.* It may well be that, though we will not bring aggregate demand back to its pre-recession trend, most of the big costs of this policy have been paid. And so it may be that my pet policy: nominal income/GDP targeting, is only a small improvement over the current framework here in the UK or in the US. But there is one place that direly needs my medicine.

As a whole, the Eurozone is currently seeing very low inflation, but plenty of periphery countries are already suffering from deflation. And this is not the Good Deflation of productivity improvements (can be identified because it comes at the same time as real output growth) but the Bad Deflation of demand dislocation. The European Central Bank could deal with a lot of these problems simply by adopting a nominal GDP target.

When it comes to macroeconomics, the best analysis we really have is complicated econometric models on the one side, and highly stylised theoretical models on the other. Both are useful, and both can tell us something, but they rely on suspending quite a substantial amount of disbelief and making a lot of simplifying assumptions. You lose a lot of people on the way to a detailed theoretical argument, while the empirical evidence we have is really insufficient to conclusively answer the sort of questions I’m posing.

In general, I think that very complex models help us make sense of detailed specifics, but that “workhorse” basic theoretical models can essentially tell us what’s going on here. Unemployment is a real variable, not one directly controlled by a central bank, and a bad thing for the central bank to target. But in the absence of major changes in exogenous productivity, labour regulation, cultural norms around labour, migration and so on, there is a pretty strong relationship between aggregate demand and unemployment. Demand dislocation is almost always the reason for short-run employment fluctuations.

Unemployment rose everywhere in 2008-9. But it nudged down only marginally post-crisis in the Eurozone, whereas in the UK and US it soon began to steadily fall toward its pre-crisis rate (the red line, though not on this graph, has tracked the green one very closely). In the meantime the Eurozone rate has risen up to 12%. This is not at all surprising, given the almost complete flattening off of aggregate demand in the Eurozone—this means a constantly-widening gap with the pre-recession trend (something like 20% below it now).

Although intuitively we’d expect expectations to steadily adjust to the new likely schedule, three factors mean this takes a while: firstly the ECB is very unclear about what it is going to do (and perhaps unsure itself), secondly some plans are set over long horizons, and thirdly the lacklustre central-bank response to the 2007-8 financial crisis is unprecedented in the post-war period.

1. We have a huge literature on the costs of policy uncertainty—the variance of expected outcomes has an effect on firms’ willingness to hire, invest, produce, independent of the mean expected outcome.

2. Many firms invest over long horizons. It may have become clear at some point in 2011, when the ECB raised interest rates despite the ongoing stagnation and weak recovery, that the macro planners, in their wisdom, were aiming for a lower overall growth path and perhaps a lower overall growth rate in nominal variables. And so, after 2011 firm plans started to adjust to this new reality. But many plans will have been predicated on an entirely different 2009, 2010, 2011, 2012, 2013, 2014, and so on. And as mentioned before, the gulf between what was expected for the mid-2010s back in 2007 and what actually happened is actually widening.

3. Thirdly, and finally, the period 2008-2010 is unprecedented and will have slowed down firm adjustment substantially. As mentioned above, even if firms set plans with a fairly short-term horizon (a few years) they wouldn’t have been able to adjust to the new normal in 2008, 2009 and 2010 unless they really expected the ECB’s policy of not only not returning to trend level, but not even return to trend rate!

All of these three issues are convincingly resolved by nominal income targeting. It’s very certain—indeed the best version would have some sort of very-hard-to-stop computer doing it. It promises to keep up to trend. And it is very stable over long horizons.

Recent evidence reinforces the view, implicit in our models, that (unconventional) monetary policy is highly effective at the zero lower bound, even through the real interest rate channel (!) All the ECB needs to do is announce a nominal income target.

*This reminds me: isn’t it about time we had an NGDP futures market so we could make claims here with any kind of confidence?