Nominal rigidities: how often do retailers really change prices?

Martin Eichenbaum interviewed by Romesh Vaitilingam, 30 May 2009

Martin Eichenbaum of Northwestern University talks to Romesh Vaitilingam about his research on nominal rigidities and reference prices, which analyses a new weekly scanner data set from a major US retailer containing information on prices, quantities and marginal costs for every item in over 1,000 stores. The interview was recorded at the American Economic Association meetings in San Francisco in January 2009.

Listen

Unfortunately the file could not be found.

Open in a pop-up window Open in a pop-up window

Download

Download MP3 File (3.53MB)

a

A

See Also

Eichenbaum, M., N. Jaimovich and S. Rebelo (2008),  'Reference Prices and Nominal Rigidity', CEPR Discussion Paper No. 6709.

Transcript

View Transcript

Romesh Vaitilingam interviews Martin Eichenbaum for Vox

January 2009

Transcription of an VoxEU audio interview [http://www.voxeu.org/index.php?q=node/3611

Romesh Vaitilingam: Welcome to Vox Talks, a series of audio interviews with leading economists from around the world. My name is Romesh Vaitilingam and today's interview is with Professor Martin Eichenbaum of Northwestern University. Marty and I met at the American Economic Association's Annual Meeting in San Francisco in January, 2009, where we spoke about his research on nominal price rigidities and their relationship with the monetary transmission mechanism.

Martin Eichenbaum: Macroeconomists have traditionally thought that price stickiness, nominal rigidities, were a central element in understanding how monetary policy affects the economy. Over the last few years, we've gotten lots of interesting microeconomic data that just wasn't feasible to collect or to analyze years ago. And wonderful researchers have shown that prices change all the time at the micro level. That's a huge challenge for people like myself who think nominal rigidities are really important. If prices are changing every couple of weeks, how important can these sticky prices be? And that's almost become a conventional wisdom now.

So, in recent work with Sergio Rebelo and Nir Jaimovich, we've starting looking at drawing a distinction between prices changing and prices being optimized and they're obviously very different.

So, for example, we got access to this marvelous dataset from the largest supermarket chain in the United States, its scanner data, so it was really high quality data. We know the price of every good that they sell and we know the marginal cost. What I mean by marginal cost is what they claim is their marginal cost to replace the item if they sell it.

The first thing you see in this dataset is, yes, consistent with what lots of other folks have shown, prices change on average about every three weeks. But, if you graph the data, the first thing that grabs your eye is what you see is this pattern of the price is one level and it falls and it just comes back up to the old level, so it kind of has a memory.

So, siren like, the old price is saying, "Come back, come back." And then, suddenly the price goes above that default price and then the data says, "Come back, come back, come back."

So we struggled a long time trying to decide how do you quantify what our eye is seeing? We actually screwed up a whole lot in trying to get that right. And at the end, we decided to look at something called just the modal price during, let's say, a quarter. And what you find is that modal price is very inertial, even though prices are changing very frequently. So even though prices change once on average every three weeks, the modal price only changes once a year.

That's kind of really fun because it helps reconcile Alan Blinder did this famous book and survey, where he asked executives how often they changed prices and they said, "Oh, about once a year." And everyone always wondered, how do you reconcile that finding with what we're seeing the micro data?

Well, the answer is, they were probably asking how often does this modal price change, maybe it's called the regular price or they kind of have that in their heads so they were answering questions that we were misinterpreting what they were really...

So, anyway, so this has opened up a whole can of worms, or exciting worms, because now we have these prices and this "normal" price, how does that relate to marginal costs?

Well, we're really, I think, one of the first people to ever get data on marginal costs and you see fascinating patterns. Roughly speaking, you can describe what's going on as these folks have an unconditional markup so they will then say, for the sake of argument, it's not about this company, but for the sake of argument, 20 per cent.

And you see, if they don't change their price, suppose you don't change your price and your markup would really be very far from 20 per cent, there's a big likelihood that you're going to change your price. So these guys are really almost satisfying, if you'd like.

Romesh: So, it's not quite a rational response.

Martin: Well, I don't want to say that, because we can be rational almost by definition if you say, "Well, it would be real expensive if we got the managers to get together," which it probably is. My wife worked for a long time for a very large company and she used to tell me the most expensive thing in the world is getting management together, dwarfs almost anything else. She worked for Kraft cheese was more expensive. But, leaving cheese aside, management time's really expensive. So, would you really want to take your senior manager...

Romesh: To spend a lot of time discussing price changes.

Martin: They're very small. It just wouldn't make any sense. But it does mean that you can get interesting nominal rigidities even though prices are changing all the time. So, that's kind of really fun.

Romesh: So, this is really some kind of micro foundation as to why there might be problems with the monetary transmission mechanism?

Martin: Absolutely and it's trying to reconcile the apparent, well, the view, that there are interesting nominal rigidities with the observation that prices change all the time.

Romesh: Fascinating. Where do you go with this now, Marty? Where does this research program take you next?

Martin: Well, we're working with some folks who have access to the largest pharmacy in the United States and we may be able to do experiments in the stores to try out various hypotheses. Because, I said, we can describe this world, there are all sorts of weird stuff you find out when you actually speak to real managers. So, most economists, certainly, I would have said, it's easier to change price to put something on sale than change the permanent price.

It turns out for these pharmacies, it's exactly the opposite is true. If they're going to put something on sale, they have to know three months in advance. Why? Two things, you better make sure the goods are in inventory, or you'll just really make customers mad and, secondly, you have to advertise. Because if you don't advertise, you're just giving stuff away to people who are already in the store who are probably going to buy it anyway.

So their view is we need to draw new customers with a sale, that whole process of advertising, making sure you're stocked up, that's three months. If it's a regular price change, you do that very quickly because they don't have to advertise it, it's going to be there for a long time.

That really changes your notions about how prices, which prices, respond quickly, why they respond. Ultimately, somehow, we have to get the essence of these, if you like, stories, and ask what's the right abstraction of macro model?

So, I think of this empirical work as the analogue to I'm going to mispronounce this pointillism, where if you're too close, you can't see anything but, if you step back, a pattern emerges. So, that's how I think of all this micro.

Romesh: Nice image. How do you think of this in terms of what's going on now, just anecdotally? You walk out on the streets of San Francisco and you see massive discounting going on, there's a lot of price change is happening right now it seems to me and the American economy is responding very quickly to the downturn.

Martin: Well, there's no question that demand is down, there are lots of sales in the retail industry. It's hard to think of this recession, the first order of business is not price rigidities. This is not what this recession is about. This is about financial markets. I do think it's true that, in part, the traditional Fed response of lowering interest rates is partially motivated by the view, you know, wages haven't started to change very much, so you do see the price of L.L. Bean and all this other stuff coming down, but what about the guys behind the counter?

You don't see wages there being cut very quickly; they will be cut but it takes a while. But I think this is not a crisis primarily about the traditional recession, it's a much more unusual episode.

Romesh: OK. Martin Eichenbaum, thank you very much.

Topics: Monetary policy
Tags: nominal rigidity, reference prices

Ethel and John Lindgren Professor of Economics and Co-Director, Center for International Economics and Development, Northwestern University

Subscribe