The case for NGDP targeting

The case for NGDP targeting. 

By Matthew Yglesias. SlowBoring.com. 

March 30, 2021. 

Federal Reserve Chair Jerome Powell prepares to speak at a House Financial Services Committee hearing on Oversight of the Treasury Department's and Federal Reserve's Pandemic Response in the Rayburn House Office Building on December 2, 2020 in Washington, DC. 

(Photo by Jim Lo Scalzo - Pool/Getty Images)

The Ever Given has mercifully been freed from the clutches of the Suez Canal.

But back when the ship was stuck, I read a CNN article that briefly mentioned the idea that “a spike in prices could force the Federal Reserve to hike interest rates sooner than expected.”

When you think about it, this is a crazy idea.

A ship gets stuck in the Suez Canal.

That means other ships need to detour around the Cape of Good Hope.

That means they burn more fuel, so shipping costs go up.

The higher shipping costs lead to higher consumer prices.

To fight this, the Fed raises interest rates.

The higher interest rates lead to job losses in interest-sensitive sectors (housebuilding, etc).

Higher unemployment leads to cheaper oil (more unemployment equals less commuting) and less consumer demand for durable goods.

Prices settle back at their pre-blockage level.

Why would you react to a ship being stuck in a canal by trying to get construction workers to lose their jobs? It doesn’t make sense on its face. And the good news is I’m fairly certain the Federal Reserve would not actually have done this. But the fact that it even sounded vaguely plausible is reason enough to return to an idea that I like but haven’t written about in a long time — instead of targeting inflation, the Fed should target nominal GDP or nominal income.

It would be temporarily disorienting to switch things up, but in the longer term, it would make it much clearer what kinds of things are and aren’t valid criteria for changing monetary policy. It would also make the appropriate stance of policy much less dependent on contestable issues like how to measure inflation. And it could help us avoid recessions in the future.

What NGDP targeting is

Nominal GDP is a bit of an unfamiliar idea. But to an extent, the fact that it’s unfamiliar is weird. Everyone who follows policy at all has heard of gross domestic product, a measure of the total value of all the goods and services sold in the economy.

But GDP or “real GDP” is a statistical construct, made by first counting up the nominal GDP — the literal dollar value of all the output — and then doing an inflation adjustment.

Trying to measure inflation is important. A 10-year span in which GDP grows by 2% per year and inflation grows by 3% per year is very different in its implications from a decade of 3% growth and 2% inflation. But measuring inflation is also controversial and difficult. You need to measure changes in the quality of goods and services, for example, which is hard.

The nominal figures aren’t perfect or trivial to measure either, but they are simpler. People don’t have big disagreements about NGDP or whether it’s being counted correctly.

But what the Fed has done for a while now is try to target inflation. First, they kind of quietly put it out there that they were targeting 2% inflation, but kept that unofficial. Then under Ben Bernanke, they started explicitly saying the goal was 2% inflation. Now under Jay Powell, they have further clarified that it is an average of 2% inflation, so you might compensate for undershooting the 2% target by overshooting in the future. They keep trying to make small tweaks to this framework, but the constant tweaking reflects an underlying conceptual problem.

Prices change for different reasons

One basic issue here is the Suez Canal Problem.

Inflation is a potential problem with excessively stimulative monetary policy. But prices can rise for all kinds of reasons, potentially including a ship running aground in an Egyptian canal. But if the problem is a ship stuck in a canal, the solution is to dislodge the ship not to raise interest rates.

That’s one reason why Alan Greenspan preferred not to make the 2% target explicit. He wanted to always reserve the right to sort of blow off weird events and ignore them. The more recent view is that being explicit about the target gives the Fed more credibility and helps “anchor expectations,” and that with well-anchored expectations they have more flexibility to blow off weird events. Powell, for example, says the Fed will “look through” transient price increases associated with the reopening of the travel and leisure economy this summer.

In this specific case, I agree with Powell.

But we had a converse situation in 2015 and 2016 when inflation was falling rapidly due to events abroad and a strengthening of the dollar, but Janet Yellen’s Fed deemed it transient and proceeded with plans to raise interest rates. This helped generate what Neil Irwin dubbed the “invisible recession of 2016” — basically a slump in the agriculture, energy, and manufacturing sectors. We might not think too much of a localized slump in the agriculture, energy, and manufacturing sectors that occurred in the year 2016, but it also happens to be the case that Donald Trump narrowly won a presidential election that year in a way that very likely wouldn’t have happened if the manufacturing and agricultural sectors had been doing a bit better.

One interesting thing about that invisible recession is that if you look at the economy in NGDP terms, it’s perfectly visible. Instead of an “invisible recession,” what you get is a clearly visible growth slowdown — not an actual recession year like 2009, but not some kind of undetectable mystery.

GDP percent change by year

That’s part of the virtue of an NGDP framework — you do less guessing about what’s transitory and what isn’t, or what you can look through and what you can’t.

You try to keep a focus on steady NGDP growth. If the economy gets bad news for some non-monetary reason (ship stuck in a canal), you get more inflation and less real growth. But the solution to the non-monetary problem is itself non-monetary (you need to fix the canal). What the monetary authority commits to is avoiding a 2016-style slowdown.

You never need to call for more inflation

Within the confines of the existing Fed framework, I was saying in 2016 that the Fed was making a mistake given the below-target inflation.

But one huge problem is that I think policymakers will always have a difficult time being in a position where they are arguing in favor of more inflation. After all, inflation is bad. People like lower prices rather than higher prices. And even at the time, my view wasn’t really “it would be good if prices were higher.” It was more like “the sudden collapse of inflation was a symptom of an inadequate demand situation.” Lael Brainard, then and now my favorite Fed Board member, tried to explain this — a negative demand impulse from abroad was afflicting the United States and showing up as a strong dollar and low prices, and the Fed needed to try to counteract that impulse.

The NGDP framework offers a much more constructive and plausible way to talk about this — you’d say that spending and incomes were too low and the Fed was trying to generate more [nominal] spending and higher [nominal] income. It’s true that the higher nominal spending and higher nominal income would partially come from more inflation. But it’s unlikely that they all would. And on an aspirational level, the Fed would never be positioning itself as pro-inflation. Inflation would just be a bad thing that sometimes happens.

A better division of labor

Contemporary thinking about central banking is deeply influenced by the experience of the 1970s, which left economists convinced that the natural state of democratic politics is for elected officials to err on the side of excessively stimulative policy.

The solution they came up with was to try to set up central bankers as insulated from day-to-day politics and able to serve as macroeconomic chaperones who’d shut down excessive stimulus and prevent inflation.

I think the experience of the past 20 years suggests that this theory of what went wrong in the 70s is probably mistaken, and we don’t need to design all of our institutions around the fear of replicating a specific set of bad decisions made in 1971-73. In particular, I think it’s notable that while the post-1980 policy consensus was supposed to keep us focused on the primacy of the supply side of the economy, in practice, productivity growth has been worse.

NGDP targeting isn’t a panacea for that.

But I do think it suggests a more constructive division of labor in which the central bank guarantees something like 5% annual growth in nominal income, and then it’s left up to the elected officials to make sure as much of that growth as possible takes the form of real increases in living standards. Right now politicians are always talking about “creating jobs,” which is an ambiguous concept that could refer to either stimulating demand or else deliberately doing things in a low-productivity, labor-intensive way. It would be better for the Fed to promise to keep the stimulus on track no matter what — possibly even generating too much inflation — and tell the politicians to focus on boosting productivity to minimize inflation.

The Fed just completed a review of its monetary policy framework last summer, so I know they’re not going to do another change in the near term. But by the same token, I know there will be another framework review at some point in the future. And when it happens, I hope they will consider that the constant tweaks to the inflation targeting regime show there’s an underlying problem. The Fed wants both a clear rule and the flexibility to respond differently, depending on what’s actually happening.

The way to accomplish that is to have a rule that’s clear, but not about inflation — and that is nominal income.