Arnold Kling  

The Crack Theory of Inflation

Richard Green on the Mortgage ... Robert Solow on Sylvia Nasar...

NPR lets John Taylor make the case against easier money and higher inflation.

He says encouraging inflation is a slippery slope. A central bank may just want to raise inflation moderately, but he says, it inevitably gets translated into higher inflation for longer periods. That's partly because pushing inflation down can cause economic pain, so people resist it.

So, the theory is that inflation is like crack. Take a little of it, and you'll soon become crazy and hooked, and then the withdrawal will be painful.

Point noted, but it's more difficult to make the case that nominal GDP is like crack. So I still favor giving a Sumnerian monetary expansion a try. However, unlike Rogoff, I would shut the experiment down if we were getting 5 percent inflation instead of better real growth.

Comments and Sharing

CATEGORIES: Monetary Policy

COMMENTS (16 to date)
Patrick R. Sullivan writes:

Milton Friedman made the same argument in 'Free to Choose' (iirc), but used heroin; at first it feels good, but the bad things happen later.

Kit writes:

The UK has had a prolonged period of greater than 5% inflation and no growth. And BoE is back on the crack - shutting down the experiment doesn't seem to be an option.

Inflation is creeping up. Today's data from the BLS shows August at 3.8% and rising. As Kit noted, UK is already there.

Matt C writes:

Is there empirical evidence that central banks can target a given inflation level, hit that target, and keep the rate there?

I'm doubtful of this. I don't think central bankers have nearly this level of control. But I would be happy to be wrong.

I also wonder if we have anyone today with nerve comparable to Volcker's, if inflation starts heating up.

I do not understand how targeting nominal GDP is different in an important way than targeting the inflation rate. It's the same process, yes?

Rick Hull writes:

I have trouble considering deliberate monetary inflation to be anything other than subtle, insidious fraud, harming savers in particular, unless the new dollars are distributed uniformly among the populace.

If we instead funnel the new dollars through the banking system, and this process indeed causes real output to go up, I can understand why this might seem like a good tradeoff. But I can think of lots of unsavory infringements that might achieve the same result.

But what if we inflate and real output does not go up? Why would we try fraud when the outcome is so uncertain?

I realize that my position is outside the mainstream and much of Sumner's NGDP-targeting nuance is lost on me.

If everyone is scrambling just to keep up with inflation, we may see increased output. An image of economists dangling a carrot in front of the nose of the donkey turning the mill comes to mind.

Is there a commonsense explanation (think: Hazlitt) of why inflation, Sumnerian or otherwise, should solve any problems we have?

GIVCO writes:

Just a metaphor alert: you can enjoy crack and/or heroin for longer than you think without becoming physically addicted. Don't believe the WOD hype.

Arthur_500 writes:

Inflation worked so well for the economy during the Carter years that I am sure encouraging some inflation will be beneficial.

all those unemployed would have much less buying power with their "stimulus" unemployment checks.

People would be much happier spending their hard earned dollars on break, milk, vegetables and meat at inflated prices.

Think of how many new jobs will be created with money that has lesser value.

And the investors in US bonds will be ecstatic!

I don't know what Bernanke was smoking on that fateful trip to Japan but he obviously shared it with some other economists.

As the passing of Steve Jobs is in the news, compare and contrast "Govco" with Apple. Both raise cash by selling bonds. Both have income: Apple from the sale of things; Govco from the income tax, which is actually a kind of citizen's revolving charge account.

What, then is "inflation" but a single admittedly large market entity borrowing money and making promises to pay later? What if Apple's balance sheet looked like Govco's?

Granted, Apple does not create its own currency, but it could. Disney does. Apple only chooses to use the Federal Reserve Notes because they are convenient. That detail does not change the analysis. Would you buy a Mac or an iPod or an iPhone if Apple's balance sheet looked like Govco's?

It is the patriot's fallacy that the Secretary of the Treasury or the Chairman of the Federal Reserve answers to you. As a shareholder in Apple, you have as much control over who manages the accounting department.

MikeDC writes:

Rick Hull,
I don't think inflation is a complete solution, but it might be a partial solution to 1) nominal rigidities, 2) money illusion. The problem is one needs the "Goldilocks" just right amount of inflation to do this, there's no model to get it right, and there are very big public choice problems - call them temptations - once it starts.

Think of it this way. Most every mortgage and interest rate based contract put into effect over the period up to mid 2008 was predicated on a set of expectations about inflation. Those expectations weren't 0%, so to the extent that we're not meeting those inflation expectations, we're making those contracts that much more difficult to get out from under.

At a really abstract level, I always think of adding money in an economy like adding poker chips in a high stakes game. You want an impartial house to provide more chips to players when the stakes get high. Imagine if the players in the game can't deal with the house and get more chips for the assets they bring in. Instead, they bet with each other by throwing those assets on the table. In movies, a Rolex, or a house, or a night with the other guy's wife make for compelling stakes in a card game. In the real world though, such assets are very hard to value, whereas poker chips are very easy to value. So without enough chips, making deals gets hard. So "printing more money" is both good and necessary to facilitate our trades and measure our wealth in calculable terms as it increases.

Of course, bad things can happen too. Note that this system relies on an impartial house dealing out more chips as needed to make the card game work. If the house isn't impartial, and is, for instance, secretly giving a couple players extra chips every few rounds, you get inflation that way too, but with all sorts of pretty obvious and terrible negative effects.

James writes:

Is NGDP so different from crack? Using inflation to raise NGDP while real output is stagnant is pretty similar to smoking crack in hopes of making a stagnant standard of living more enjoyable. Unfortunately, smoking crack tends to lower one's real standard of living so much that greater and greater amounts of crack are needed to maintain the same level of satisfaction and eventually no amount of crack is enough. How exactly is NGDP targeting supposed to be different?

ColoComment writes:

All y'all are busy discussing the economics of what Mr. Taylor said. I heard the NPR Taylor interview this morning & was simply tickled that NPR had JOHN TAYLOR on the show! What a breakthrough, when their usual econ interviewee is Robert Reich.

Badger writes:

Once you get to 5% there's no "shutdown." That's the entire point made by sensible Taylor, but even Kling, apparently, can't get it.

Bob Murphy writes:

Point, Badger. Sumner's prescription might be right or wrong, but Kling's post here makes no sense. He is basically saying, "I agree with Taylor that trying to get just a little inflation is bad, because it's impossible to put the genie back in the bottle. That's why I recommend targeting NGDP. And if doing that leads to a moderate amount of inflation, well then we'll just put that genie back in the bottle."

Charles R. Williams writes:

There is no evidence the fed can hit an NGDP target. There is no evidence that if the fed succeeds in hitting the target, the economy will grow smoothly with full employment. Some of us remember when the fed tried to manage the money supply. The markets nervously awaited the money supply data so they would know what would happen to interest rates. It was in fact destabilizing. The NGDP targeters want to accommodate shifts in demand for money as well as shifts in supply. Fine. What they don't seem to understand is that the fed has lost control over money and banking. This is the result of both globalization of finance and technology. Under normal circumstances the fed can penalize or subsidize the regulated banking sector in the US. This has at best a tenuous relationship to economic activity.In abnormal circumstances they can function as a lender of last resort - essentially pulling financial intermediation activity in house during a panic.

Joe Cushing writes:

Inflation will help all the people who are overloaded in debt at the expense of those who lent to them. It can clear the debt of some people, which will help turn the economy around. The problem is that after lenders catch on, they raise rates to protect themselves and it becomes hard to take on new debt.

There is a lot of lag in the system. Measured inflation lags monetary inflation; interest rates lag measured inflation. So when the fed cuts the money supply (calls it raising interest rates) rates go up above that which is needed, given the current level of inflation. It's the lag that allows the fed to say it is raising rates when it is taking action that will eventually push them up and lowering rates when it is taking action that will eventually push then down. That's why it's like crack.

To say it clearer; "lowering rates," by creating new money to buy debt, is inflation. Inflation raises rates because lenders want to protect themselves from losses. So the act of lowering rates causes them to go up but with a lag. The Fed feels pressure to push them back down with more inflation, which causes them to keep rising. This can turn into an out of control cycle.

English Professor writes:

A couple of thoughts/questions by someone simply trying to understand the issues.

Like a few others here, I don't see the difference between stimulating inflation and targeting NGDP. Isn't the point to increase NGDP more than real GDP, and isn't that simply done by inflating the currency?

Is there any other purpose for increasing NGDP than to bring down real wages? I have assumed that this is intended as a solution to the sticky wage problem; is there any other benefit?

Kenneth Rogoff recently wrote a piece encouraging moderate inflation (4 or 5 percent). If I remember correctly, he suggested that the current state of the economy is so bad that savers must suffer for the good of all. If we hadn't already bailed out many institutions that deserved to fail, I might approve of this--that is, everyone in turn must do their part. But it seems to me that tax payers have been expected to pay the bills of big-time creditors through bailouts, and now savers are to pay again through inflation. This is heaping fraud on top of moral hazard.

Finally, with the lag in measured inflation, I don't see how the Fed could hit an NGDP target or an inflation target. You take actions now, and the effects appear 18 months or 2 years later. There is no feedback to let you know if you are creating too much or too little inflation. And then, when it will take another 18 months to 2 years to correct the problem, you once again have to shoot in the dark.

Comments for this entry have been closed
Return to top