Scott Sumner  

My naivete about government officials

PRINT
I'm 90 Percent American and 10... Henderson on Piketty, Part 4...

Back in the 1980, I believed the Fed needed to target inflation or NGDP, and if they did so the problem of high inflation could be solved. These views were widely mocked by people on both the left and the right, for different reasons:

1. On the left they mocked Milton Friedman, who they claimed was obsessed with the view that money was the key determinant of inflation. They thought it obvious that the guns and butter policy of the 1960s ignited inflation, and that the peak periods of inflation represented supply shocks. They believed that the only way to control inflation was to control budget deficits. (That was before inflation plunged in the 1980s, during a period of very high peacetime deficits.)

2. On the right they talked about the "time inconsistency" problem with fiat money. Inflation was the inevitable result when you had governments facing re-election that had to worry about the unemployment rate. Of course the past 5 years provided a near perfect test---we had a government very worried about unemployment, which faced re-election in 2012. And inflation stayed low.

In the 1980s the government did bring down the rate of inflation, although even in the late 1980s it was still running in the 4% to 5% range. Then in the 1990s they focused on a 2% inflation target, and choose the PCE inflation index as their target. So I decided to compute the actual average rate of PCE (headline) inflation over the past 24 years. And it turned out I was wrong; the Fed did fail to hit their inflation target. Here's the actual PCE inflation rate over the past 24 years:

Screen Shot 2014-10-25 at 9.02.37 AM.png
Now what would the old Keynesians of 1980 say? Here are some possible explanations:

1. Since 1990, the Newt Gingrich/Nancy Pelosi/John Boehner regime has expertly steered inflation at roughly 2%, by skillfully adjusting the fiscal deficit.

If you've stopped laughing, let's proceed to the next explanation:

2. Luck.

The Fed got very lucky, and even more amazingly, the central banks of other major countries were also able to luck out at almost exactly the same time.

In contrast, those on the right seem to think 1990-2014 didn't happen. Here's something from the comment section of my previous post:

This is a pretty naive view of self control. Libertarians like to stress the imperfection and inevitable ignorance of man right up until their favored policy choices require Herculean efforts of virtue and restraint. At that point people are suddenly near perfect.
Silly me, for believing that central banks could improve their performance.

People on the right are stuck in the "bumblebees can't fly in theory" rut, and liberals don't even understand that central banks are steering the nominal economy. Liberals should have been bashing the Fed for tight money since 2008 and conservatives like Charles Plosser (who think it's the Fed's job to control inflation, not unemployment) should be smiling like the cat that swallowed the canary.

I'd also like to clarify a few issues that were raised in the comment section.

1. Some people like to compare the classical gold standard with the fiat money period. That makes no sense. Either compare the entire gold standard to the entire fiat money period, or leave out the bad gold standard of 1918-33, and leave out the (bad) pure discretion period of 1968-90. The price level has been more predictable (both short and long term) under the inflation-targeting regime since 1990 than it was during even the classical gold standard (1879-1914). If you had forecast the September 2014 price level back in September 1990, under the assumption the Fed would hit its 2% inflation target, you would have been less than 1% off. That was not true under the classical gold standard, when the price level was roughly a random walk.

2. People get dates wrong. The classical gold standard began in the US in 1879, and went up to 1914. The interwar gold standard lasted from 1918-1933. The quasi-gold standard lasted from 1934 to 1968. Nothing meaningful happened in 1971. We had already completely left the gold standard in 1968, when the market price of gold was allowed to float and foreigners could not longer redeem dollars for gold.

[Broken formatting fixed.--Econlib Ed.]


Comments and Sharing


CATEGORIES: Monetary Policy




COMMENTS (17 to date)
ThomasH writes:

I think that too many liberals believed that the Fed was and will be too cowed by inflation hawks to execute policies needed to stimulate the economy in any significant way. (To wit: it has not even been able to prevent inflation from falling below its target.)

And perhaps they think that if governments only invested in long term projects with positive net present value at low inflation-induced interest rates nothing beyond "passive" acceptance by the Fed would be necessary. This leads them to focus more of their fire on the errors of fiscal policy (which is supposed to be more responsive to public opinion) than the errors or monetary policy. (They correctly assess that the chances of changing the mind of inflation hawks is close to zero.)

Andrew_FL writes:

Scott-this is a comment I wanted to leave here but earlier was unable to.

Good to see it looks like comments are working again, I was worried there for a second.

James writes:

If the Fed can reliably hit an inflation target plus or minus some X%, the long term cumulative deviation from target will be the product of X, the square root of time and the autocorrelation of X. Getting to within 2 bps of the target over 24 years can't be explained by either skill or luck for any plausible value of X.

When you go to traffic court for a speeding ticket, the judge will not let you provide your own speedometer reading as evidence that you were only driving as fast as you were supposed to. We should be similarly skeptical when the government supplies measures of how quickly it debases a currency.

Even if the Fed could reliably hit whatever target they have, so what? History furnishes plenty of examples where some economic activity was organized by a committee developing a plan to hit some target. Given how that stuff usually turned out, I can't see why anyone would still put their trust in such a process. Maybe it's time to consider that having government agencies set targets is just a bad idea.

Sam Haysom writes:

Exactly James. The government chose the PCE index precisely because it is the most manipulatable, eschews fix based prices, and omits energy and most food costs. In other words, those staples were inflation shows up first and causes the most chaos in economies.


I'll also add that I think it is strange that both right-wing politics and support of the gold standard are attributed to me on the basis of a post on moral theory and human nature. I guess I just don't see how my point is rebutted. I never argued that the government couldn't control inflation. What I argued was that is that if libertarians think this is true then they need to rethink their opposition to an activist government on Hayekian lines. If they can hit the sweet spot on inflation, let them take another shot at price controls. Maybe Nixon-era struggles with price controls weren't normitive. they were simply the result of a government that wasn't as talented as it is today. After all the Fed couldn't manage a low inflation economy in the 1970s but it can now. Let's stop using the 1970s experiment when making the argument against price controls today.

This would be a horrible politcal development in my mind but I can't see any objection the author could really provide.

Bob Murphy writes:

Scott, did you rely mean to say that the price level from 1879 - 1914 follows a random walk?

Andrew_FL writes:

@Sam Haysom-I think Scott is criticizing the Gold Standard because I said some positive things about it, not you.

But you seem to be confused about the way in which the Government is being said to control inflation. Price controls don't control inflation except in the sense that you can't measure a fever if you break the thermometer.

This is the CPI during World War II. The period in red is from April of 1942 to June of 1946. What is the significance of those dates? In late April of 1942, the General Maximum Price Regulation restricted most prices from rising above their highest levels in March of that year. Controls were briefly lifted in June of 1946 because Truman vetoed a bill that would have had them end in 9 months. Their re-imposition was not as extensive.

You might conclude that, indeed, one can fight inflation by just imposing nigh economy wide controls on prices. And in some sense that'd be true: you'd just get shortages.

Lorenzo from Oz writes:

The power of political narratives.

It even fits into Arnold Kling's three languages in a way -- if you see conservatives as regarding inflation as barbarism and liberals as seeing unemployment as oppression and government spending as social-justice-in-action.

Political narratives get in the way of making government agencies accountable. But, between folk who think government-is-good and those who think it is congenitally hopeless, I guess work-a-day improving performance is not the game they want to be concerned with.

I would also argue that the US has a much bigger political narratives problem than folk realise:

Sam Haysom writes:

I don't think price controls fight inflation Andrew_FL. The point I am trying to make is that libertarian political theory is based at least in part on Hayek's critique of centralized decision making. Price controls are a notorious example of a failed policy implemented in the 1970s. However, concomitant with that the government also failed to controlled inflation something it can now apparently do technocratically.

One take away from this could be that sophisticated computers and date mining make price controls feasible. You are assuming that price controls cause shortages because in the past all price controls caused shortages. Well in the past governments printing money at a non-stop clip to bolster the stock market created inflation as well. You can't say "oh well the government has figured out the money printing game" only to turn around and keep insisting that price controls don't work because look at history. What I am taking issue with is Sumner's rejecting the historical approach to inflation, but wanting to maintain the historical interpretation of price controls and the "Hayekian" skepticism of centralized decision making.

Roger McKinney writes:
The price level has been more predictable (both short and long term) under the inflation-targeting regime since 1990 than it was during even the classical gold standard (1879-1914)

Maybe, but that's trivial. Check out Financial Stability Paper No. 13 – December 2011 Reform of the International Monetary and Financial System. Here's some quotes:

Net capital flows tended to be large under the Gold Standard (Chart 1). However, passive domestic monetary policy responses meant that they were not accompanied by large cross-country policy inconsistencies and so did not pose the same threat to global financial stability as those of today. Table A below, which presents a range of summary statistics on the performance of different IMFS regimes, shows for example that the incidence rate of banking and currency crises in the Gold Standard was much lower than in today’s system.
The direction of net capital flows during the Gold Standard seemed broadly consistent with the efficient allocation of capital across countries. In particular, these imbalances were associated with ‘downhill’ flows of capital from the older, advanced economies in Europe to more productive opportunities in the younger, fast-growing economies in Asia and the Americas (Kenwood and Lougheed (1999)). Further, private sectors played the dominant role in these capital flows, which is consistent with the notion that economic fundamentals were at work.
Overall, the Gold Standard appeared to perform reasonably well against its financial stability and allocative efficiency objectives, while the internal balance objective was of secondary importance.
Andrew_FL writes:

@Sam Haysom-I'll let Scott speak for himself on what his own views are, however:

You are assuming that price controls cause shortages because in the past all price controls caused shortages.

No, I'm not assuming anything based on history. I know that price controls cause shortages because that's the direct implication of the laws of supply and demand.

There's no way to computer program your way out of that one. It's not a matter of the price controls just not being "done right." Price controls have bad results because they are done at all. It's not a matter of the government not knowing the right price to set, it's a matter of the government setting the wrong price on purpose.

Price controls, in short, are not something libertarians even need Hayek's points to criticize. All they need is to explain elementary economics.

Of course, there are plenty of libertarians who would not agree with the assertion that "the government has figured out the money printing game" but putting that aside, you are confusing two very issues.

The other thing you seem to be confused about is the fact that the rhetoric of the Presidents and Congress in the 70's was of fighting inflation, and that events showed they couldn't control inflation because the rate was higher than the rate they said they wanted:

This ignores the fact that 1. The policy they pursued to fight inflation was price controls-which is a wrong headed approach. and 2. The rhetoric didn't match the actions. They allowed NGDP growth to accelerate, or even encouraged it, while claiming to fight inflation. Well, actually, the Fed didn't claim to be fighting inflation. As I recall they denied any responsibility whatsoever.

Scott Sumner writes:

Thomas, Actually the Fed is far easier to influence than the GOP in the House. The Fed did do stimulus to offset the austerity of 2013, the opposite is much harder to imagine.

James, I find those conspiracy theories to be utterly implausible. Who is this "government" that you refer to? Obama? The Fed? Congress? Who is manipulating the PCE, and why? And why do private price indices show similar inflation rates?

Sam, You said:

"I'll also add that I think it is strange that both right-wing politics and support of the gold standard are attributed to me on the basis of a post on moral theory and human nature."

It would be strange if it were true.

And your claim about the PCE index is just completely inaccurate. It does include food and energy.

Hayek favored NGDP targeting, and so do I.

I don't understand the price control issue. The reason we can control inflation now has no bearing on whether we can make price controls work, or fly to Alpha Centuri. They are unrelated issues.

Bob, Yes, that is what research shows. Check out Barsky's paper.

Lorenzo, I agree.

Roger, As far as the US is concerned, banking and currency crises were more common under the gold standard. But look, we had big problems with our banking regulations back then, and still do today (albeit different problems.) I'm not a fan of our banking system.

Sam Haysom writes:

"By contrast those on the right" followed by a block quote of something I said. You are right no imputation at all.

James writes:

"James, I find those conspiracy theories to be utterly implausible."

You are arguing from a fact about your mind to a claim about the world. I was arguing from facts about math. The difference is pretty important.

Scott Sumner writes:

Sam, You were not able to find any evidence that I suggested you support the gold standard, because there wasn't any.

If I had been confident you were on the right I would have used your name. Instead I simply provided a link so that any reader could find out for themselves what you believe.

James, OK, I find your "facts" about math to be unpersuasive, and indeed unintelligible.

As far as this sentence is concerned:

"You are arguing from a fact about your mind to a claim about the world."

No, I did not. If I had, you'd presumably have found a sentence you could cite to prove it.

James writes:

Scott,

I didn't realize you had actual evidence against corruption in the production of inflation figures. All you provided was a rhetorical question.

It doesn't matter if you find math persuasive or if you understand it. Variance still scales with the square root of time and autocorrelation. That includes the variance of actual PCE growth and target PCE growth.

And again, so what if the Fed could hit some target? Why should anyone favor any arrangement where government agencies work on hitting economic targets? Those types of arrangements have pretty lousy track records relative to their alternatives.

A writes:

James, why is the relationship between a time scale and the standard deviation over that time scale an argument against achieving a given mean over that time scale?

Glenn writes:

"It doesn't matter if you find math persuasive or if you understand it. Variance still scales with the square root of time and autocorrelation. That includes the variance of actual PCE growth and target PCE growth."

So what?

"I didn't realize you had actual evidence against corruption in the production of inflation figures."

Sorry; you make an outrageous and indeed by its very design untestable claim, and then attack another for not disproving it?

The simple fact is that there is extraordinary empirical evidence in favor of the claim that the Fed can and does effectively target inflation levels. To suggest that this is evidence of some massive collusion and fraud between numerous policy makers (all of whom are otherwise at policy odds with one another, and have competing interests in exposing the illegitimacy of their political adversaries) strikes me as, frankly, absurd.

A (weakly) better argument would be that price-level stability is incidental to Fed policy; this at least has the appearance of rationality (in that it can logically follow from a set of plausible premises). You - and I'm not trying to be deliberately antagonistic here - just appear to be raving.

Comments for this entry have been closed
Return to top