Summers stumbles

by John Q on February 9, 2022

There’s been a lot of debate lately about whether tightening of anti-trust legislation might be a useful response to inflation. Underlying this question is that of the relationship between monopoly and inflation more generally. The dominant view among mainstream/neoclassical economists seems to be that there is no such relationship. That view is stated by one of the most prominent mainstream theorist, Larry Summers as follows

There is no basis in economics for expecting increases in demand to systematically [cause] larger price increases for monopolies or oligopolies than competitive industries.

Summers goes on to describe the opposite view as ‘anti-science’.

Readers of this blog will be devastated to learn that Summers is dead wrong. It’s quite straightforward to show, in a simple neoclassical model, that imperfect competition amplifies the inflationary effects of demand shocks. Here’s a paper I’ve just written with my colleague Flavio Menezes which makes this point using the concept of the strategic industry supply curve. The same result can be presented, less elegantly in our view, using the standard tools of comparative statics to be found in any intermediate microeconomics test.

We also show that, contrary to a suggestion by Elizabeth Warren, imperfect competition is likely to dampen the impact of cost shocks. There isn’t, however, any equivalence here. Warren’s background is in law, and she isn’t making a claim just observing that monopoly power might be a problem. The distinction between cost shocks and demand shocks is unlikely to have been relevant to her, whereas it should have occurred immediately to a leading theorist like Summers.

I’m not sure about the lessons from all this. For me, it’s to think carefully before making dogmatic statements from authority. If all experts agree on something, we should say so, but be careful to make sure we are right.

{ 32 comments }

1

nastywoman 02.09.22 at 11:05 am

and isn’t this Summer-Dude the dude who already was so entirely NOT right in
‘Social Network?

2

Martin Holterman 02.09.22 at 11:06 am

Yes, I was surprised by Summers’s statement too. But then, I thought the literature showed that it was perfectly feasible to have cost pass through of more than 100% under all sorts of market structures.

https://assets.publishing.service.gov.uk/government/uploads/system/uploads/attachment_data/file/320912/Cost_Pass-Through_Report.pdf

3

eg 02.09.22 at 2:49 pm

Shouldn’t the title be “Summers stumbles yet again”? It’s gotten to the point where the proper response to any utterance of his is a Constanza-like “assume the opposite.”

4

Brett 02.09.22 at 3:42 pm

Definitely a bizarre statement from Summers.

Although it is rather weird to see anti-trust now being utilized in an argument for lowering prices. For a long time, the big complain from the New Antitrust folks was that US antitrust policy was too focused on low prices, and they needed to take other criterion as well.

5

TM 02.09.22 at 4:54 pm

I don’t understand the sentence quoted, its grammatical structure doesn’t parse.

6

oldster 02.09.22 at 5:09 pm

“There is no basis in economics for expecting increases in demand to systematically larger price increases for monopolies or oligopolies than competitive industries.”

I cannot parse that as a piece of English syntax, other than by inserting a verb, such as eg

“There is no basis in economics for expecting increases in demand to [lead to] systematically larger price increases for monopolies or oligopolies than competitive industries.”

Or, “produce” would work just as well in place of “lead to.”

Have you mis-pasted his quote, or was he using some kind of Econo-speak in which that sentence counts as grammatical? And if so, what’s the verb?

7

CJColucci 02.09.22 at 6:59 pm

Is there a verb missing from the Summers quote?

8

John Quiggin 02.09.22 at 7:42 pm

The verb “cause” is indeed missing. I mentally filled it in.

9

oldster 02.09.22 at 8:16 pm

“The verb “cause” is indeed missing. I mentally filled it in.”
Do you mean:
“Summer’s own statement contained the verb “cause,” but when I first transcribed the statement I unconsciously omitted the verb”? Because if the verb is in Summer’s original statement, then you don’t need to put it in square brackets. (Indeed, should not put it in square brackets.)
Or do you mean,
“Summer’s own statement omits the verb, but I understood it as though it contained the verb “cause” in this position”? Because then that makes me wonder about his grasp of English, and also makes me wonder whether your guess about the necessary supplement is the right guess. I mean, maybe you’ve made the right guess. But it would still be your word rather than his.

10

Alex SL 02.09.22 at 9:05 pm

How bizarre. Why does he think people conspire to create monopolies and oligopolies so often?

11

Fake Dave 02.10.22 at 1:26 am

I don’t think it makes much sense to call the elite Wall Street/Chicago economists “mainstream” anymore. They may have dominated the field at large, but I’m not sure they ever represented the average number cruncher and since people like Summers now routinely contradict the textbook models without even apparently realizing it, it makes more sense to treat them as a deviant subculture than leaders in the field.

12

Rapier 02.10.22 at 1:26 am

The old reprobate and neo liberal stalking horse Milton Freidman was right about one thing. Inflation is always a monetary phenomena.

13

Fake Dave 02.10.22 at 1:44 am

Also, as a non-economist, it occurs to me that there’s an obvious common sense illustration why demand empowers monopolies. If someone has a monopoly on something people don’t really need, like Halloween costumes for cats then consumers (with self control) can simply choose not to buy it if the price is too high. If it’s something people desperately need like insulin or clean water, then consumers will pay. Only in bad classical models and ad industry fever dreams can “consumer demand” ever be separated from “human needs.” No sensible store owner would ever forget that. So I’m not sure why we let “mainstream” theorists get away with it.

14

marcel proust 02.10.22 at 2:55 am

I’m a bit late to this issue, now that our blogger has made a correction, but better late than never.

https://www.gocomics.com/doonesbury/1980/01/23

15

Timothy Scriven 02.10.22 at 3:12 am

Thanks for the paper, it looks cool. I agree that a theoretical connection makes sense. My concern though is that most of the people I’ve seen vehemently objecting to the claim that monopoly is inflationary (e.g. Noah Smith) have not pitched their argument on theory (the theoretical argument for a possible relationship with inflation seems clear) but on empirical studies.

For me, at least, this seems like the right battleground, I distrust dueling models in economics at this scale, because it seems like there are too many degrees of freedom and too many heroic assumptions.

I’m curious A) As to whether you have any empirical resources on this and B) Whether you think this kind of question is best resolved empirically or theoretically.

16

John Quiggin 02.10.22 at 8:19 am

Oldster @9 Summers omitted the word, but it’s clear from the context that it should be “cause” or similar.

Timothy @15 As Martin says @2, excess pass-through is theoretically possible. But there’s lots of empirical stuff supporting partial pass-through of cost increases, and good theoretical reasons why this should imply larger responses to demand increases for monopolies.

17

banned commenter 02.10.22 at 11:38 am

Monopolistic industry will have higher prices than a competitive one, that’s obvious.

But that’s not what the quote is about. It seems to me that logically indeed increases in demand might lead to larger price increases in competitive industries.

You see, competitive industries are probably highly optimized for the existing level of demand, have little excess capacity. With increase in demand, distributors, now facing insufficient supply, will raise prices. (Will prices come down later? Hmm. Who knows.)

This is unlikely to happen when an inefficient monopoly, with (presumably) significant excess capacity and warehouses full of product, owns the market.

18

MisterMr 02.10.22 at 12:29 pm

There is a distinction between:

“monopolies push up prices” (which means they are inflationary), which I suppose Summers agrees to, and
“when there are inflationary pressures, monopolies make it so that these inflationary pressures are bigger”

Suppose that in country A the price of a sock is 1$, whereas in country B, because of monopolies, said price is 1.3$.
Then because of other inflationary pressures prices increase by +50%: in A now a sock costs 1.5$, and in B it costs 1.95$.

In this case monopolies in B did not increase inflation, they just kept prices higer both before and after the inflation period.

I think this is what Summers means, however even in this case anti-monopoly measures applied after the inflationary period would lower the prices, and this would look like a lower inflation.

19

MisterMr 02.10.22 at 12:46 pm

Also, is there an estimate of how much inflation is supply side and how much demand side? I see that in the USA there is 7.something inflation whereras in EU and UK it’s 5.something.

Suppose that 5% of this is suppli side; but the relevant part of inflation for monetary policy is demand side, in this case this would be 0% in EU/UK and 2% in USA, too low in EU/UK and just enough in USA; on the other hand if supply side inflation is just 3% it means that inflation is normal in EU/UK but 4% in USA, and therefore a bit too high (but not really that high).

I also fail to understand what people mean for high inflation or too high inflation: when I was a kid Italy was coming out from a period of 20%+ inflation, so when inflation went lower than 10% it meant that it was low, but now what is low inflation? 1%? 2%? 3%? 7%? Is there any objective parameter to say if inflation is high or low or is this just scaremongering?

20

oldster 02.10.22 at 1:42 pm

Thanks John.
After I posted that last comment, I realized that I was verging on grammar-troll territory, so I apologize for that.
I’ll let you get back to discussing economics.

21

Seekonk 02.10.22 at 3:52 pm

Professor Q’s paper refers to market power, sometimes known as pricing power, the ability or goal of a seller to gain sufficient share of a market to be able to increase prices.

This phenomenon is long-recognized and particularly abhorrent when it involves essential goods or services such as medicine or housing. For Summers to deny its existence is further evidence, if we needed any, that Summers is a scoundrel.

22

Lobsterman 02.11.22 at 6:46 am

“How can you tell if Larry Summers is lying?”

23

Andrew 02.11.22 at 4:25 pm

Isn’t the entire point of a monopoly (or something close to it) that it brings pricing power? Does one really need a paper for this? If you’re the only one selling chicken and people want chicken, you can charge what you want.

24

banned commenter 02.12.22 at 1:13 pm

Pricing power is not necessarily a point of monopoly (let alone its only point). It could be, for example, standards of quality and compliance. Check out NH liquor stores.

If a lot of people are getting sick from salmonella in chickens, it could make sense to funnel all chicken sales (or, perhaps, even production) through one company (or two) and hold it responsible.

25

LFC 02.12.22 at 5:59 pm

That competitive markets, and firms within those markets, are more sensitive — in the sense of being more price responsive — to fluctuations in consumer demand is straight neoclassical Econ 101, at least as I was taught it as an undergrad a long time ago. There’s more price competition in genuinely competitive markets, at least that’s what I recall having to regurgitate on the exams. Oligopolies, by contrast, tend to compete more by “differentiating” their products and less by lowering prices or creating so-called loss leaders, for ex. E.g. when airline X in the oligopolistic U.S. airline industry raises or lowers its fares, the others tend to follow suit. (There are low-cost carriers but I’m talking about most of the major airlines, which operate in an oligopolistic market dominated by a handful of carriers.) Instead they compete in other ways, or attempt to.

Hence, all other things being equal, I think neoclassical Econ 101 would lead one to expect that a sharp increase in demand might cause sharper price rises in competitive markets than in oligopolistic markets or, if that statement is too strong, there’s at least no obvious reason why sharp demand increases should lead to higher price rises in oligopolistic markets than in more competitive ones (just as, when demand falls, you’d expect prices to come down more in competitive markets than in oligopolistic ones).

I’m not doubting that JQ and his colleague have shown that “imperfect competition amplifies the inflationary effects of demand shocks,” I’m just saying that’s not what I would have expected based on the way I was taught, or mis-taught, intro ‘mainstream’ Econ about 45 yrs ago.

26

John Quiggin 02.13.22 at 6:50 am

Andrew @23 Monopolies can charge whatever they like, but if they charge too much, people won’t buy (as much of) their products. They charge the price that maximizes their profits. The greater the demand, the higher the profit-maximizing price.

LFC @25 I think you are misremembering. “More responsive to fluctuations in consumer demand” would typically be taken in the sense of “more quantity responsive” and not, as you suggest, “more price responsive”. That’s particularly true if you allow for entry.

27

Barry 02.13.22 at 3:03 pm

Banned Commenter: “This is unlikely to happen when an inefficient monopoly, with (presumably) significant excess capacity and warehouses full of product, owns the market.”

Why would they/waste money on that? The whole point is that people have to buy from you.

28

anon/portly 02.13.22 at 3:58 pm

There’s been a lot of debate lately about whether tightening of anti-trust legislation might be a useful response to inflation. Underlying this question is that of the relationship between monopoly and inflation more generally.

So right away, what I want know is, does JQ think then that “tightening of anti-trust legislation might be a useful response to inflation?”

Also, does this mean that running a higher rate of inflation, over time, increases real profits in industries with more pricing power? Would this have been an argument against (or drawback of) Australia running a higher level of inflation back during the “Great Recession” times? (As I recall Aus got a lot of credit for having a higher inflation target and setting some sort of record for going a long time without a recession). And is this an argument for keeping inflation lower, maybe even for disinflation?

29

banned commenter 02.13.22 at 10:19 pm

@27,
They would waist money on that because they don’t need to count every penny to survive. They can afford to build and maintain adequate, stable production capacity.

@26,
“The greater the demand, the higher the profit-maximizing price.”

Hmm. Isn’t it, rather, elasticity of demand that would affect the optimal price? Mathematically optimal, that is. There have to plenty of political factors involved too.

30

John Quiggin 02.13.22 at 10:28 pm

Anti-trust takes years, so it’s not a macro policy response to a temporary burst of inflation. But if inflation is conceived of as prices rising faster than wages, anti-trust is one way of reversing that,

The second point is too complex for a comments thread.

31

MisterMr 02.14.22 at 12:35 am

“But if inflation is conceived of as prices rising faster than wages”

This is sort of the main point though: is inflation conceived as prices leaving wages behind or is it supposed to be the result of a wage price spiral?
My understanding is that classic keynesian demand-led inflation presupposes a wage-price spiral, which is a very different thing from prices leaving wages behind.

32

John Quiggin 02.14.22 at 9:28 am

Banned commenter @29 An additive increase to a linear demand curve reduces the elasticity of demand at any given price. Do the math for yourself. Elasticity is -(P/Q) dP/dQ, so if Q increases, and nothing else changes (by linearity of demand), elasticity falls. Trust me, I do this stuff for a living.

MisterMr @31 There are lots of different ways of conceiving inflation. But the one that has the most political purchase is “higher cost of living” which translates, to “prices rising faster than income”, or more precisely “my income”

Comments on this entry are closed.