Author Archive


AIT so far

Last August, I did a blog post suggesting that in order for the Fed’s new average inflation targeting policy to be successful the PCE price level needed to be roughly 135.207 in January 2030, which represents a 2% annual growth rate over the January 2020 price level (110.917.) The most recent PCE data is for April 2021, and shows the price level at 114.075. Thus inflation is averaging 2.27% during the first 15 months of the 2020s. The Fed needs PCE inflation to average 1.96% for the remainder of the decade in order to hit their AIT target.

Let’s hope they are serious.

Feedburner is ending

I’m not entirely sure what Feedburner is, but I’m told that it is important to many of you. The program is ending on July 1st and will be replaced by Feedblitz. You can sign up to Feedblitz here:

Basil Halperin on sticky wage models

Some of the earliest New Keynesian models featured wage stickiness. By the 1980s, NKs switched to price stickiness, which remains the standard assumption even today. Basil Halperin has an excellent essay that explains why wage stickiness is a more useful assumption for macro models. This portion of his essay caught my eye:

1. Identification: the source of the shock matters!

Recessions caused by tight monetary policy should cause real wages to increase and be too high, leading to involuntary unemployment. Recessions caused by real supply-side shocks should cause real wages to fall and nonemployment to rise.

If the economy experiences a mix of both, then on average the correlation of real wages and recessions could be anything.

Maybe in 1973 there’s an oil shock, which is a real supply-side shock: real wages fall and nonemployment rises (as in the data). Maybe in 2008 monetary policy is too tight: real wages spike and unemployment rises (as in the data). Averaging over the two, the relationship between real wages and unemployment is maybe approximately zero.

This view was around as early as Sumner and Silver (1989) JPE, where they take a proto-“sign restrictions” approach with US data and find procyclical real wages during the real shocks of the 1970s and countercyclical real wages during other recessions.

But: while Sumner-Silver was published in the JPE and racked up some citations, it seems clear that, for too long a time, this view did not penetrate enough skulls. Macroeconomists, I think it’s fair to say, were too careless for too long regarding the challenge of identification.

My sense is that this view is taken seriously now: e.g. in my second-year grad macro course, this was one of the main explanations given. At the risk of overclaiming, I would say that for anyone who has been trained post-credibility revolution, this view is simply obviously correct.

The paper I did with Steve Silver is my first published use of “never reason from a price change” (NRFPC). My subsequent research also relied heavily on that maxim. I’m not sure why it didn’t “penetrate enough skulls”, but perhaps it had something to do with the fact that we both taught at Bentley College. In any case, I’m glad to hear that it has now penetrated more skulls. This blog has been bashing people over the head with NRFPC for 12 years.

As Pissarides (2009) ECMA pointed out, it doesn’t really matter if the wages of incumbent employed workers are sticky. What matters is that the wages of new hires are sticky.

Why is this? Suppose that the wages of everyone working at your firm are completely fixed, but that when you hire new people, their wages can be whatever you and they want. Then there’s simply no reason for involuntary unemployment: unemployed workers will always be able to be hired by you at a sufficiently low real wage (or to drop out of the labor force voluntarily and efficiently). If new hire wages were sticky on the other hand, that’s when the unemployed can’t find such a job. . . .

And the best evidence from Hazell and Taska does argue for sticky wages for new hires from 2010-2016 – in particular, sticky downwards.

It’s hard to image a world where existing workers have sticky wages and new hires do not. In that case, a company could announce that it was firing 100% of its workforce at 5pm on Friday and hiring back 100% of its workforce the following Monday morning at 10% lower wages. Surely wage stickiness must be more deeply embedded in labor markets than suggested by the new hire/existing worker distinction. Something more is involved.

Halperin summarizes a great deal of empirical and theoretical evidence and reaches this conclusion:

[I]f you think of involuntary unemployment as being at the heart of recessions, you should start from a sticky wage framework, not a sticky price framework.

And here’s the policy implication:

More importantly, this should affect your view on normative policy recommendations:

1. Sticky prices – when modeled via Calvo – prescribe inflation targeting: in the simplest setup, stabilize aggregate inflation so that the stickiness of prices need never affect anything. (This provides an intellectual foundation for the policy of inflation targeting used by most developed central banks today.)

2. Sticky wages on the other hand prescribe stabilizing nominal wages: in the simplest setup, stabilize an index of aggregate nominal wages, so that the stickiness of wages need never affect anything.

Halperin illustrates the advantage of nominal wage targets with this graph:

The aggregate nominal wage series (blue line) correctly signals a mild recession in early 2008 and a major recession in late 2008. The price series (red line) does not suggest any sort of problem in the first half of 2008.

PS. I hope David Beckworth is correct:

PPS. I interpret the market reaction to the Fed’s recent announcement as reflecting a growing realization that the Fed is serious about 2% AIT. They removed some tail risk of high inflation, and thus TIPS spreads fell about 10 basis points. Otherwise policy did not change, and hence there wasn’t much reaction in the stock market. So far, so good.

The new “election fraud”

Here’s Bloomberg:

“We are witnessing the greatest election fraud in the history of the country, in my opinion in the history of any democracy,” Israeli Prime Minister Benjamin Netanyahu said this week, echoing Trump as he smeared an unusual political coalition formed to unseat him. Though he is also mired in a corruption trial, Netanyahu said he won’t recognize a new government. Vitriol among his supporters has prompted Israel’s domestic security service to warn of escalating, possibly lethal, violence.

Brazilian President Jair Bolsonaro faces re-election next year and has already lashed out at his country’s courts and made baseless claims of voter fraud. He has refused to condemn the Jan. 6 siege in the U.S. because, he said, there were “a lot of reports of fraud.” Should he “have problems” in Brazil, he said he would deploy the military to solve them.

Netanyahu and Bolsonaro have absorbed a valuable lesson from Trump: If you co-opt the imagination and intentions of enough members of your own party and your voters, you can persuade them to buy into your lies and rise up on your behalf when power slips from your grasp. 

Claims of election fraud have been around for decades. So why do I call this the “new” election fraud? What’s new about it?

Correct me if I’m wrong, but most of the claims of widespread election fraud during the 20th century were made by the candidate that was challenging the incumbent. After all, the government controls the electoral process, and thus is in a position to assure that it is reasonably fair. What’s new is not claims of election fraud—candidates challenging authoritarian leaders have been complaining about fraud for decades—what’s new is that there are now lots of incumbents crying fraud.

PS. Yes, this argument is a bit less applicable to the US where the election apparatus is controlled by state governments. But there were lots of claims of anti-Trump election fraud even in states controlled by the GOP. Weird.

PPS. I recently appeared on the Bob Murphy Show. We discussed monetary policy during the Great Recession.

Vermont and Wyoming

Vermont has the nation’s highest vaccination rate — 72.4%

Wyoming has a rate that is well below average — 38.2%

Why such a big difference? These states rank number 49 and 50 in population. Both states are lacking in big urban areas. Both states are overwhelmingly white, with relatively few minorities. At the same time, there are obviously huge cultural differences between the two states.

It’s also worth noting that Vermont has the lowest Covid fatality rate among the lower 48 states. Wyoming’s fatality rate isn’t that bad, but is 2 1/2 times worse than Vermont, and also far worse than states like Utah, Washington and Oregon.

I’d be surprised if Vermont’s high rate of vaccination was unrelated to its low Covid fatality rate. Perhaps both facts are related to Vermont’s culture. Perhaps Vermonters are risk averse people who “trust the science”. Here are two counterarguments for my claim:

1. Vaccination rates are high throughout the Northeast, and yet many northeastern states have high Covid fatality rates.

2. Perhaps culture doesn’t explain the low vaccination rates, rather in places with higher infection rates there is less vaccination because many people already have have natural immunity from previous exposure. They don’t believe they need to be vaccinated.

Of course, the second point in is tension with the first. Lots of people in Massachusetts have been exposed to Covid (far more than in Wyoming), so why such a high vaccination rate in Massachusetts?

In the Northeast, the worst Covid wave was during the spring of 2020, before many people realized the severity of the pandemic. In addition, those areas are densely populated. (Even indoors. I find that shopping in Mission Viejo is much less crowded than back in Boston.)

My best guess is that cultural differences explain the Vermont/Wyoming discrepancy in vaccination rates. This FT story has “nudge” suggestions for encouraging vaccination from behavioral research.

PS. Biden got 66.1% of the vote in Vermont, highest of any state in the country. In 1936, Alf Landon got 56.4% of the vote in Vermont, highest of any state in the country.

Vermont is weird.

PPS. Here’s a Yahoo story on Wyoming:

Some public health officials have resigned themselves to the reality that many in their community will not budge on shots.

In Wyoming’s Sweetwater County, population 44,000, authorities are at a loss for what else they can do to achieve herd immunity.

Sweetwater carries the unfortunate distinction of being the county with the steepest increase in infections in the state with the most new infections per capita in the country. Only a quarter of its residents are fully vaccinated, and public health officials don’t see the number budging much higher.

Jean Stachon, Sweetwater County’s health officer, said officials held mass clinics, brought vaccine doses to employers and churches and accept walk-ins at the public health office. They have sacrificed extra doses in a vial to vaccinate at least one person. But demand is minimal, even as the virus still looms in the community. Two people died of covid-19 in the last week. Eight emergency room patients were diagnosed with coronavirus in one night.