Saturday, December 21, 2013

Whining about price gouging

One time at a dinner, I was assigned to sit next to and entertain George Akerlof. He asked me what I thought of the theory of supply and demand. I told him that it was good locally, but globally it was hard to estimate either curve, so it wasn't very helpful for large price changes. He seemed a little disappointed in my skeptical answer.

But I think I was right. Here's a picture of supply and demand curves:

How on Earth could you know the whole shape of either of these curves? If the theory is right, we're currently at or near the intersection point. So how can we know the rest of the curves?

A demand or supply curve is a hypothetical thing - it describes how customers or suppliers would react if they observed a certain price change. But to measure either curve, we need to identify shifts in the other curve; if we observe what we know is a supply curve shift, we can use that to map out some piece of the demand curve. Of course, identifying pure supply shocks or pure demand shocks is hard. But more importantly, really big shifts are really rare. So in practice it's usually impossible to know much about demand and supply curves far away from the equilibrium point. The data just isn't there. Demand could be highly responsive to huge price changes, or not very responsive at all. We just won't know until one happens.

Which brings me to price gouging. One of the few times we do see big sudden price changes is when there is a big negative supply shock - for example, a flood may cause an abrupt scarcity of fresh water, or a snowstorm may cause an abrupt scarcity of taxis. At these times, prices rise steeply, and consumers often complain of "price gouging".

Most economists and economics writers defend price gouging against the complainers. After all, it's just supply and demand. For example, the Washington Post's Neil Irwin defended the "surge pricing" that taxi service Uber implemented during a recent snowstorm.

Most econ writers and economists seem to think of anti-price-gouging complaints as being about government policy - in other words, they interpret the complaints as a populist call for the government to implement price controls during negative supply shocks. And some complainers probably really are trying to influence policy. But I think that there may be another reason for the complaints: Signalling.

Because demand curves are hard to estimate far from the equilibrium point, companies don't really know how willing people would be to pay very high prices in the event of a negative supply shock. They may make a mistake in the event of a negative supply shock, and set prices too high for a while, and thus miss a lot of profit opportunities.

But people themselves may have an idea of their own willingness to pay. It would behoove people to communicate their demand curves to the companies. So they may try to signal to companies that they would not be willing to pay very high prices, by complaining. This is probably "cheap talk", rather than a costly signal (because companies wouldn't tend to respond to only one person's complaints). Cheap talk can allow buyers and sellers to coordinate their actions better. A flood of complaints about surge pricing may help companies know not to set their surge prices too high.

In which case, complaining about the complainers would constitute interfering in the healthy working of the economy.

But anyway, I don't feel like thinking about this much more deeply at the moment, so if any of you game theorists out there want to take a crack at making a cheap-talk model of anti-price-gouging complaining, go ahead. Actually, it probably already exists.


  1. In the time it takes for most big moves along a curve, the curve has changed.

    1. Anonymous11:40 PM

  2. The answer probably is going to differ depending on whether you treat this as a one-time or an iterated game. The standard analysis is one-time, but if it is iterated, reputation effects come into play. You could see, for example, the griping about price gouging as an attempt to change the cost curve for the company in raising prices and make it less profitable to raise prices (because a short-term price increase will then have significant long-term costs).

    This probably deserves a larger discussion on the topic, as some results that differ between behavioral economics and traditional economics often come down to whether you consider repeated interactions between the actors or not. This may be another case where price gouging is fine, so is complaining about price gouging, but price gouging may have higher costs than normally discussed if you consider repeated interactions.

  3. Or not such cheap talk, not that they are unwilling to pay more during a shortage, but a means of buyers coordination, letting sellers know that whatever profits they may expect to gain by doing so will be met with losses afterwards as they shun those that do. Not entirely for selfish reasons either, but to make it clear that the creation of artificial shortages to increase profits won't be rewarded.

  4. Hi, do you remember why was Akerlof disappointed?

    I remember one similar anecdote from I think Animal Spirits, but it could be from some other piece of his writings (for non-economists): He mentioned to his colleagues that he had a friend that could sell his house. He was told his friend was stupid because all he had to do is lower the price of his house more.

    Also, I like you point that economists often assume implicit calls for government intervention and try to preemptively deal with that.

  5. To me an interesting question concerns the mechanism of price movements. For example, in a negative supply shock I am familiar with--price of plywood after major storms--to what extent are suppliers setting higher prices versus consumers bidding up prices?

  6. Why does such talk (the price gouging complaints) have to be rationalized in some way? Why isn't the following a sufficient explanation: people don't like other people taking advantage of their suddenly weakened bargaining position to get more out of them, so because they dislike this behavior and it's against their interests, they complain about it?
    Acting like it needs instead to be explained as part of some grand strategy to influence government policy or to signal that they won't in fact pay the higher amount, basically seems like a parody of how economists think, to this non-economist.

    1. I think Noah is talking about the supply-side of the market. Of course, people (demand side) hate to be exploited, so they complaint. Agreed. Now, why does the suppliers fall on that strategy of excessive prices, if finally it can fire back? One option is that, as Noah says, they are adjusting prices, and the complaints can help to adjust them.

      Now, the other option is that they do that because it do not fires back, neither now in demand reduction nor in the future for bad publicity. So they just exploit the opportunity (or the people, been clear).

      But anyway, even so it may be interesting see this situation as a rule of thumb for business to find something similar in real life to a demand elasticity or demand curve. So is a useful excercise to question our own formation as economists.

    2. It seems to me that he is talking about "consumers" throughout, and wondering why they complain about price gouging. So I agree with David Mathers.

  7. You make an interesting argument, but usually the complaint isn't "This is a rip-off! I'm not paying that!", but "It's just awful of them to jack up prices right when we really need it!".

  8. There's a reason there's such pushback (in economic circles) about people complaining about price-gouging: Price-gouging exposes the essential injustice of the capitalist system. Suddenly confronted with the naked operation of supply and demand, the populace instinctively recoils against it. And that scares economists.

    1. The populace recoiling is nothing to how the capitalists recoil, collude, obfuscate & change the rules when forced to negotiate wages in a tight market.

  9. Looks like Krugman was (again) inspired by your post, but turned it into an interesting take on sticky wages, and yet another chapter in the micro foundations debate.

  10. Blissex11:17 AM

    There are much bigger issues with the graph above than those:

    * The market demand curves can only be that smooths if the individual demand
    curves are smooth and the same, that is if all buyers have exactly the same
    tastes. Otherwise demand curves have plenty of kinks. A point made often by
    Steve Keen.

    * As Sraffa showed, supply curves can have, even if fully neoclassical situations.
    kinks too, depending on rate of interest and technology.

    * In practice, as Mankiw showed, there can be very significant "menu" effects.

    * In practice, whichever shapes they have, the supply and demand schedules
    are ribbons, not lines, and a lot of political lobbying and power get used to
    ensure that vendors get the position they want in the region of their intersection.

    Therefore neoclassical theory is full of contradictions and gross mathematical
    mistakes to "work around" the big problems with its imaginary demand/supply
    curves and still deliver its central verity, that income distribution depends solely
    on productivity.

    1. Blissex wins the thread.

  11. If GA had asked me what I thought of supply and demand curves, I would have said they were a meme that infects promising young peoples' minds and turns many of them into bright dolts who from that point on think in severely stunted ways about microeconomic phenomena, especially in product~service markets.

    Why? Because the vast majority of trade-offs buyers make along their margins are not How-many-more-of-the-selfsame-thing? trade-offs but Which? trade-offs (the nicer one or the not-as-nice one?) A demand curve completely ignores Which? trade-offs. Models that provide for Which? trade-offs don't look like single slopes. They look like mountain ranges.

    Thinking both How Many? and Which? vastly changes a lot of product-service micro. GA seems to have some understanding of that. Do you?

  12. In economic theory, prices are signals of scarcity, or more precisely of an excess of demand over supply. Which then focuses minds on the profit to be had by increasing supply. But when (as with taxis in bad weather, or food supplies in a bad year) there is no possibility of increasing supply, then price simply becomes a means of rationing. This is ok for luxuries, but not ok for essentials, as it then signals a divided - and hence vulnerable - community. Can people tell the difference? in my experience, yes. Can economists? Only with real difficulty.

    1. I think comes down to confusing price and value. Usually there is not a great difference between the two - but in emergencies a huge one.

    2. o. nate10:29 AM

      Interesting comment, Peter T. So is an Uber car during a snowstorm an essential or a luxury? It seems like a luxury to me. If you don't want to get stranded during a snowstorm, it seems like common sense to not put yourself in a situation where you're going to need a car to get around.

    3. @Peter T - I agree except that while Uber may have overstated the size of the effect, raising their prices does increase supply. And I think most people would be OK with this if it were a 1.5 or 2x increase in prices. It was 7 and 8x surge pricing that pissed people off. Without transparency into their algorithms or driver supply network, it's impossible to know, but I'd assume there are steeply diminishing returns (in term of # of new drivers on the road) to each 100% price increase, which is what causes people to feel as though they're being "exploited"

  13. All curves are local Noah, your complaint that we don't know much about the curve far away is a sort of existential rendering of Xeno's paradox. Tomorrow never comes, and when we see the far away bit of the curve it isn't far away anymore.
    That doesn't mean we don't know about the curve out there, any more than we won't know about tomorrow, tomorrow, or are incapable of arriving home for Christmas as we are locked in a sequence of half-steps.

  14. "But anyway, I don't feel like thinking about this much more deeply at the moment, so if any of you game theorists out there want to take a crack at making a cheap-talk model of anti-price-gouging complaining, go ahead. Actually, it probably already exists."

    Allow me to introduce you to my bro Al:,_Voice,_and_Loyalty

    1. And here's a really good example. There's a bar in DC called Meridian Pint. There is also a campaign afoot in DC to expand the existing mandatory paid sick day law to restaurant workers. During this campaign, the main owner of Meridian Pint came out against the expansion. Rather than simply take their business elsewhere, furious regulars (keep in mind, a craft beer bar in DC is going to have a pretty left-leaning patronage just by default) rapidly appended their names to an angry petition. And lo and behold:

      Not only did he reverse course on supporting the law, he also unilaterally began offering paid sick days to his staff.

      Moral of the story? The regulars of the bar could have simply taken their dollars elsewhere in disgust. That's the "market" solution economists love. Why didn't they?

      Because they freaking love Meridian Pint. It has one of the top three beer selections in the whole metro area, plus delicious food at reasonable prices and a fun atmosphere at a great location. It's awesome. It's my favorite. I'm there all the time. And yes, I did sign that petition. And it allowed the restaurant to very quickly change course and adopt new policies that allowed it to continue to thrive and allowed existing patrons to continue their patronage. Had they "exited" rather than "voiced" the signal would have been gradual, slow, and ambigous - while talk is cheap, in this case it was extremely effective. This is more about practices than pure pricing, per se, but I think the same general logic applies.

    2. Anonymous9:46 AM

      Left wing? In DC? You must visit Scotland one day.

  15. Anonymous1:32 PM

    There's a difference between "agree to pay" and "happily pay" and models of behavior that don't exhibit that difference are missing something that is essential to real-world performance of businesses.

    There's very little distance between this situation and a mixed-strategy equilibrium in a repeated game where the equilibrium is supported by the threat of punishment. You can go ahead and deviate for short-term gain, but if the game doesn't end you will eventually regret the lowered future returns. It's just that the real world is a complex interaction of changing pay-offs, instead of a single, static ruleset repeated indefinitely.

    Let's take a simple game - I can choose a number from 1 to 10, you can choose to lose the amount I pick (giving me that amount), or to lose N (giving me nothing), and you can pick after you see my choice. You know what N is, but I don't. It's easy to analyze how this game will play out if I have to set my strategy in advance and know only the probability of repetition. However, what if you have 3 options - lose the amount I offered, lose the amount I offered plus K to get a shot (with probability P) of reducing N, or lose N? This, I'd argue, is what's happening here - by complaining the customer is not only signalling to Uber that it shouldn't raise prices as much next time - he or she is signalling to other companies a way to improve upon what Uber is offering. Experience (and perhaps evolved responses) shows that signals like this sometimes succeed in reducing N and thus are worth investing in when the company is taking advantage of its leverage. This is why a lot of pricing calculations fail, IMO - there is no such thing as a completely defined game, even when you think you have all the players specified, the humans will act as if there is a possibility that a new player will show up and change the game.

  16. "But I think that there may be another reason for the complaints: Signalling."

    Much simpler explanation: in the very short run supply is inelastic. That means that price deviations from equilibrium do not cause dead weight loss/inefficiency, they're just all about transferring the surplus from one group to another. People - in their clumsy, non-economic, but intuitive way - understand that, so they get pissed.

    Big blizzard in a small town --> demand for snow shovels goes up --> amount of shovels available is limited to what's at the hardware store --> any price at or below equilibrium results in same quantity trade. By the time more shovels arrive, snow melts.

    That's basically what David Mathers is saying, just put into economese-speak.

  17. I don't think griping is the interesting issue here.

    The interesting debate is, should Uber customers be pissed enough over surge pricing to boycott the company?

    Some people won't get upset at all. It makes sense to pay drivers some extra for the stress and danger of driving in a snowstorm. You can think of it as motivating them not to take the day off. Some people who griped a lot during the surge will quickly get over it. Their griping won’t matter much.

    However, others will be seriously upset and will disdain Uber for longer or forever. I can see two main motives.

    The rider could suspect the driver of arbitrary pricing. Most people want predictable taxi prices and hate having to deal with taxi drivers who can ask whatever they want, which forces you to haggle before every ride or pay a large premium. That's why taxi pricing is so often regulated. If you tried to launch a taxi company in the US that routinely charged variable prices depending on supply and demand, and not just on rare extreme days, Americans would hate it. They would hate the unpredictability and would not trust their driver or dispatcher to be telling them the real computed supply-demand price. Many people would try to haggle.

    Other people would be upset for moral reasons. They would feel that increasing prices during snowstorms more than reasonable “hazard pay” is an immoral exploitation of people in distress. Some people would boycott a surge-pricing taxi company for that reason alone.

    As economists we might think we’re super clever pointing out that if taxis aren’t rationed by price they will be rationed by the de facto lottery of first come, first served. I think almost everybody intuitively understands that, and yet some people will prefer the lottery and see it as more “the right thing to do”.

    A somewhat similar situation exists in the entertainment industry. Popular stars and sports teams often price their tickets below demand, forcing their fans into a first-come, first-serve lottery or into the merciless hands of scalpers. Why? Morals and long-term fan loyalty. They don’t want their fans to be disillusioned by “gouging” ticket prices.

  18. Talbert9:11 PM

    Part A to your crazy blog: Are you crazy? No no and no. I come from the gulf coast, the deep south, and I've experienced this crap almost annually. Price gouging is basically "producers" seriously destroying social capital in order to exploit the ignorance of their customers. There are plenty of essentials - water, food, clothing, etc - always around. Most people don't know how to utilize the resources they DO have because they have built relationships with various institutions and come to trust those institutions in securing their daily needs.

    As a result of this convenience, they don't bother with looking into alternative means of meeting their needs. Should some random event disrupt the chain of supply then specific goods do become scarce. If the situation were perceived to be permanent NO ONE would bother engaging the now insanely overpriced goods and would seek alternative resources (exploiting nearby natural resources to create clean water/food/medicine/clothing what have you). Because such a situation is explicitly known to be impermanent people will simply wait till the natural chain of supply resumes. Unless there is a need to permanently change people's modes of behavior, price gouging is disastrous because in such a situation maintaining normalcy where you can is critical. Not only will people feel that the situation has gotten even MORE chaotic after a disaster, in which it is not uncommon in the least that some of their friends and neighbors have lost someone in the tragedy, but that the very institutions that they trust enough to invest not just their time and money in, but their daily lives, are not above kicking them when they are down. If you can keep something as business as usual, you do - even if it's prices - the little things that make up our daily lives are taken for granted precisely because they are little, but because they are regularly present they become huge when they are gone. The prices of basic goods and services that we depend on in our daily lives are most certainly in that category. Trying to preserve the semblance of normalcy in a disastrous situation is critical because temporary appearances are reality.

  19. Talbert9:13 PM

    Part B to your crazy BS:
    Furthermore, like most economists you idiotically (and yes this is beyond stupid, ignorant and bigoted of you) think that price affects everyone the same. Raising prices won't stop hoarding, you will only cut off this option from those most affected by the disaster in the first place - the poor. Those who are middle class and up will grumble about having to pay higher amounts but that doesn't mean they won't hoard if they think they need to, you're assuming the situation has kept the elasticity of the goods in question in check. The reality is that not only does the situation decrease supplies but it increases demand, but not the rate of consumption, thus the raising the price will be like raising the price of water/gas/or food - people will pay it if A: they have to or B: they think they have to. This also goes way beyond any "customer loyalty BULLSHIT", studies have shown that after a disaster social solidarity, informal social controls, and altruistic behavior actually increases. People come together, this is something you want to encourage and maintain. Someone who's life has been turned upside down, CANNOT be shown a SDGE chart and told that price gouging is for their own good and not want to slug the jackass holding it. Even if the proprietor of the business was truly interested in adjusting demand down to better allocate resources the many people who are short changed in this situation will get VERY pissed at the ones gouging the prices. To quote Hegel: "Against nature man can claim no right, but once society is established, poverty immediately takes the form of a wrong done to one class by another." What you seem too clueless and stupid to grasp is that some individuals in either situation will lack the desired goods. In the non-price gouging situation the individuals blame the natural disaster, I.E. chance and circumstance. In the situation you propose, upon which no amount of vitriol is enough, those individuals experience shortages will blame other INDIVIDUALS, as nature doesn't set prices.

    The fact that economists overwhelmingly have either lukewarm feelings or outright support shows what disasters you shmucks really are. You need to retract this and issue an apology for hurling insults at survivors of tragedy.