Is inflation a distributional phenomenon?

Probably not.

A few days ago, Olivier Blanchard shocked (get it?) the macro world by tweeting that inflation is essentially a distributional phenomenon - it is the result of conflict between different players in the market. This was then backed up by Ivan Werning, who posted a model where two players keep increasing prices, leading to an inflationary wage-price spiral. On slide 5 it asks if this means that inflation isn’t a monetary phenomenon. And now we have a modern classic Nick Rowe thread which points out that the New Keynesian model is flawed because it literally does not include money. And really, this is what allows the Werning model to have a simple wage-price spiral, firms and workers in the models do not have to worry about there not being enough money to buy their goods!

Olivier Blanchard @ojblanchard1

1/8. A point which is often lost in discussions of inflation and central bank policy. Inflation is fundamentally the outcome of the distributional conflict, between firms, workers, and taxpayers. It stops only when the various players are forced to accept the outcome.

(I mean, imagine if you priced your good at $100 in an economy where there’s only a $10 bill, clearly that is not the profit-maximizing choice).

But this discussion is getting silly, so let me try and go over the distributional theory of inflation in my own way. We will discuss another Tweet by Blanchard proposing a possible wage-price spiral, and analyze how it could happen in the real world under different policy regimes and market structures.

Olivier Blanchard @ojblanchard1

1/2. On distributional conflict and inflation, and some tweets. A clarification. Conflict can be the original cause of inflation. E.g. a government legislating a large increase in wage. Which leads to increase in prices, which leads to further increases in wages, and so on.

OK, my first problem here is that we’re not talking about the central bank at all. Perhaps that’s the point (Blanchard is saying what will happen if the central bank does literally nothing), but it’s still problematic. Let’s go over the logic:

Scenario 1: Supply Shock under Perfect Competition

Assume a competitive labor market (we will relax this assumption later). The government legislates an increase in wages. Since we assumed that the labor market is competitive, this will cause a rise in unemployment (firms will not be able to pay higher wages for everyone). The economy experiences a negative supply shock - unemployment rises, less goods are produced, the price level goes up. What’s strange to me is Blanchard’s claim that the spiral will continue. Wages will not continue to rise under any plausible model, the price level only rose because supply decreased, which in turn happened because of higher wages - firms here are not able to increase wages further for workers.

But the above is assuming that the central bank keeps the money supply constant. What if they just keep nominal interest rates constant? In this case, the policy change will also cause inflation expectations to rise (because of the impending rise in the price level). Won’t this cause an inflationary spiral? Probably not. Remember that inflation expectations cause persistent inflation through their effect on money creation. People borrow more, so more money is created, so prices keep going up. But if inflation expectations rise from a supply shock then your ability to pay off debt goes down (because your real income is decreasing). In this scenario, a rise in expectations is not always inflationary (see my other post on this issue).

So Blanchard here is basically just wrong, the supply shock he describes will not cause a spiral. Unless, of course, workers convince the government to raise their wages again. In that case we’d just see more and more negative supply shocks, until no one has a job anymore. Not sure if that’s what anyone has in mind.

Scenario 2: Supply Shock under Monopsony

But this was assuming a competitive labor market! What if this was a monopsony model. Well, in that case prices don’t rise at all! Government-mandated wage increases in the monopsony model (up to a point) are getting us towards the efficient market outcome. Because the firm’s marginal costs decrease, they do not raise prices! In fact, if you think about it, this is deflationary. More people get jobs through a positive supply shock (remember that a mandated higher wage increases employment in a monopsony model), which causes prices to fall.

And if you’re confused because we now have more people employed with higher wages, which seems to imply that more people will spend money on things, remember that their wages rose because the firm’s profits fell! Workers will indeed spend more, but firm owners will spend less, the overall price level won’t rise (and, because of increased production, will likely fall).

In both these scenarios, we discuss the implications of Blanchard’s thought experiment. Depending on the market structure, we find that, even if the central bank does nothing, it does not result in a wage-price spiral. We do, however, see a change in the price level. In one case, the price level rises and this results in redistribution between firms and job-losers, to job-keepers. In the second scenario, the price level falls and we get redistribution between firm-owners and workers.

But really, the only way we could have gotten an inflationary spiral was if the central bank made it happen. Otherwise, inflation would simply not have been profitable.