Why the Taylor Rule isn’t a rules-based approach to monetary policy

From Gavyn Davies:

…these three different interpretations of the Rule were expected by Ms Yellen to differ by as much as 2 per cent in the appropriate level for the Fed Funds rate from 2012-15. Given these wide discrepancies, any of which could presumably be chosen by the Fed under the proposed legislation, it seems pointless to try to force a rule-based system on the FOMC just for the sake of it.

Furthermore, the Rule does not really help with several key problems faced by the FOMC today. The first is how and when to reduce the size of the Fed’s balance sheet, and how that decision should relate to the appropriate level of short rates. Next is how to determine the right relationship between economic objectives and financial stability when setting short rates. Based on their views on these two issues, the FOMC might decide that short rates should be either much higher, or much lower, than suggested by the Rule.

The Rule is also largely silent on another of the Fed’s main headaches right now, which is whether to treat the official unemployment rate as a good indicator of the amount of slack in the labour market. Many members of the FOMC, including the Chair, have argued that the amount of slack is greater than implied by the unemployment rate, because the labour participation rate has been temporarily depressed by the recession. The use of the Taylor Rule does not solve this debate, it simply treats it as if it does not exist.

From the FT there is more here.  You don’t have to regard any of those points as arguments against a “Taylor Rule.”  But it is disingenuous to think that peddling the Taylor Rule as a monetary option counts as a rule for those who have general reasons for believing in monetary rules over discretion.  The Taylor Rule is lucky enough to be called a “rule,” and besides, any reaction function can be described as a rule of sorts.  In those two senses it is indeed a rule.  But the Republicans who are behind this are fooling themselves if they think this will yield the (supposed) traditional benefits of monetary rules, namely stability, predictability, non-ambiguity, transparency, and so on.

Addendum: Nick Rowe has some comments.


the fed, by doing a good job, deserves its independence

And vice versa? God I hope not.

Good point. But Taylor's heart is in the 'rule' place, and for a long time (last 5 years?) no one else had the heart to root-for-rules of any kind. It was all ludicrously macho fireman hosing stuff with helicopters whirling overhead, a la Krugman pace Stiglitz. Taylor said useful things about Hayek to support the rules-based approach. His last book remains very important. For a long while I could find no one else other than Taylor (maybe Cochrane too) to tweet about favourably on the subject of right-principled monetary policy. But really, it's true, a rules approach is not about setting particular numbers anyway. It's more about anti-planning and anti-engineering. Anti-plumbing economics. Rules replace panic buttons. Probably need to look to Buchanan for the general constitutional principles of monetary 'generality'.

"no one else..."

You obviously haven't been following MR carefully. Tyler has said favorable things about Scott Sumner's NGDP targeting based on NGDP futures markets. That is a non-discretionary rule with less wiggle room than the Taylor Rule.

Sorry, not convinced. Sumner policies look to me like central planning veiled by neoliberal ideological leanings in tangential areas. Monetary policy, industrial policy, often the same thing in substance. Economists of narrow technical training who forget or don't understand what "non-discretionary" means really need to read their James M. Buchanan at the very least. Or you could read my book.

In what sense is NGDP targeting central planning? Is an inflation target central planning?

In short, yes, it is central planning. That this question is being asked is amazing to me, but demonstrates how a lot of central planners are blind to what they do.

1. So if I'm reading you correctly then all fiat money regimes are central planning. Then it's still really not clear why you reserve special vitriol for NGDP targeting that you don't for inflation targeting.

2. Is it really that amazing to you that people don't consider gov't-provided currency to be central planning? It's something gov'ts have been doing for thousands of years, it's part of English legal tradition and is written into the US Constitution. For most people, providing a currency is simply not at all the same thing as, say, nationalizing the entire steel industry. I think it's a sign of your own obtuseness that you fail to see the distinction.

I don't care whether you call it central planning or not. That's a meaningless label. It's not an argument against NGDP level targeting as opposed to the Taylor rule.

Anything a libertarian doesn't like is "central planning". Nothing new to see here.

Try harder.

IIRC, Taylor has been critical of Sumner's idea, recalling its last heyday ca. 1983, and has said it is fundamentally discretionary.

The tortured or tortuous relationship of market monetarism to Buchanan was revealingly laid out in full detail by Lars Christensen here :- http://marketmonetarist.com/2013/01/12/forget-about-hawks-and-doves-what-we-need-is-a-monetary-constitution/

I don't think Buchanan would be eager to get on board with the smoke-and-mirrors about being on the same 'constitutional' side... not in his right mind. As for myself, aside from the reckless presumption of knowledge at the heart of such ambitious targets, I think it simply boils down mainly to how monetary easing (the more the better and the longer the better say some market monetarists) sends out a tidal wave of wrong signals. Not to mention that it lets politicians and economists put supply side structural reforms and reforms of the welfare state on the back-burner. It's fundamentally a bad idea to smother real economy signals and delay the day of reckoning. At best you get a fake or undesirable kind of recovery, with none of the underlying distortions eliminated. At worst you take a staggering risk with society's welfare on the wing of a test-tube experiment with no precedent. That's my humble opinion, but my preferred kind of economics, the Schumpeter kind, has itself been smothered by the profession in the past 6 years, strange when you remember how much respect was shown to it before the crisis.

Here's where I make my break with Austrians. What have the last six years been but the most wrenching of most people's lives in this country? How much of your beloved Schumpeterian creative destruction can we take?

I guess it takes a subtle mind to understand that aggregate demand is not a meaningless concept, even if much of left-wing economics is wrongheaded. That crazy central-planner Milton Friedman got this.

If you have a state with taxation power, then it is better for everyone concerned that the state has a clear monetary policy which people can use for their calculations. If the state has a clearly announced monetary policy, being targeted by a publicly available market, then all business calculations will be done taking that into account. There will be no smothering of real economy signals. Any gov that tried to put too much of its efforts into non-productive activities will see the inflation component increasing and the real component shrinking.

the right relationship between economic objectives and financial stability when setting short rates.

First we were told that low interest rates would bring about inflation.

Then we were told that QE did nothing and that it was dangerous (at the same time).

Then we were told that financial stability leads to instability.

Then we were told that the size of central bank's "balance sheet" (aka "the money supply") is a cause for concern because ... ?

The more I read the crap economists have to say, the more I wish the Fed was replaced by a computer tasked solely with keeping wage inflation on track.

The Taylor rule isn't a bad idea, but there are troubles with its implementation.

First of all, the CPI is no good measure of inflation (and I don't mean Shadowstats.)

Over the last 15 years, average annual inflation was 3.6% rather than 2.4%. Here's more:


Then I calculated a simple unweighted average out of the seven series

If you're trying to prove you're an idiot, you're succesful.

There are lots of ways to be wrong. Like not including house prices, for example. NOT that anyone would do that.

If you want to know what I think, the whole idea of measuring prices is idiotic. The inflation you want to look at is WAGE INFLATION, since it's sticky wages that cause the business cycle.

I was just pointing out that our idiotic friend's claim of shocking news is the result of an UNWEIGHTED average. Which is just plain dumb.

Hmmm...what you suggest is so far removed from what the government does are you sure his answer isn't closer than theirs?

Your answer is so far removed from making sense are you sure it's what you meant to say ?

Okay, let's try it this way.

1. Does the Fed target wage inflation?

Keep in mind, I'm working with your standard, which I consider dumb too.

And now I consider you an asshole. But since you are an asshole, I don't expect you to care. I'm not sure you are an idiot yet. An idiot would be an asshole and care that people think he's an asshole.

Wage inflation did not cause the recent recession. The housing bubble did.

ANY inclusion of house prices, weighted or otherwise, anything other than jawboning constantly that there was no housing bubble, would have helped.

And go fuck yourself.

Wage inflation did not cause the recent recession. The housing bubble did.

That's what morons like you think.

The reality is - there was no housing bubble


And wage DEFLATION caused the recession. And pretty much all business cycles in US history.

And your big argument is that the Fed doesn't currently target wage inflation ? So what ?
They didn't target price inflation. Until they started to.

Now fuck off, moron.

You seem like a complete idiot. There is no evidence that sticky wages had any effect during the Great Recession, let alone other recessions. Please return to Reddit where your kind of arguments fit in better.

There is no evidence that sticky wages had any effect during the Great Recession, let alone other recessions.

I guess being an idiot comes naturally to some people.

Just out of curiosity - what would it look like if sticky wages were the cause of recessions ?

Perhaps the Taylor Rule should be called the Taylor "historical regression of past Federal Reserve judgement-based reaction functions over some fairly arbitrary time period and then locking it in forever for no particular reason other than it's a rule" Rule?

@Kyle, good point, kind of like Friedman's "increase money supply by 2% forever" mechanical 'rule'.

Truth is, only when we go back to a fixed money supply, where only human emotions can blow it up or cause it to crash, can we have true stability. Like a koan, that is true. Bring back the Panic of 1837! Better to let the free market cause crashes rather than bungling bureaucrats.

Spot-on satire, bro.

Probably is satire, but of course, the world ended in each of the panics of the 19th century.

Fiat money has a much better track record the commodity money.

Bring back the farthing!

The panics of the 19th century did not affect GDP growth over the long run, but I doubt today's docile population would accept no Keynesian shock absorbers. Conceptually, the area under the curve of GDP over time was about the same if not slightly greater during the Gold Standard era of the 19th century, but it was jagged (peaks and troughs) due to panics, whereas today's 'smooth growth' is less volatile. But Americans love insurance, so I think they enjoy Big Government for the same reason.

Sorry to hear humanity has failed to rise to your expectations.

Taylor rule is not a rule because, "equilibrium interest rates" and "potential output" are (absolutely) non-observable.

Our understanding of the transmission mechanism of monetary policy is as deep as that of impact of carbon dioxide concentration on temperature. Climate and monetary sensitivity or "total factor productivity" are well defined rates whose determinants depend on not well known (ultra) complex mechanisms.

The Fed does not have to reduce the size of its balance sheet.

One, by year end they will quit buying bonds so it will stabilize and decline as some of the portfolio matures each year.

Second, and more importantly as the economy growth a stable balance sheet will decline as a share of the economy.

that is what happened to the WW II debt, it was never paid off.

Thanks for the marvelous posting! I actually enjoyed reading it, you might be a great author.

I will always bookmark your blog and will come back
someday. I want to encourage you to ultimately continue
your great work, have a nice holiday weekend!

Comments for this post are closed