Macroeconomists seem to be pre-occupied with sticky prices (the idea that prices adjust slowly to “shocks”). In fact, the existence of sticky prices is the main difference between the real business cycle model I discussed in my initial post and the New Keynesian model that serves as the workhorse of a lot of monetary policy research. However, most macroeconomic theories resort to ad-hoc explanations of sticky prices. In fact, macroeconomists have such a hard time understanding sticky prices that they have resorted to asking people why they do not adjust their prices.
Regardless of whether the slow adjustment of prices is important for macroeconomics, the idea that prices do not respond quickly to fluctuations in demand is an issue that is important for microeconomic analysis.
In today’s post, I would like to explore why prices are so slow to adjust to fluctuations in demand at the microeconomic level.
In the process of thinking about sticky prices, an important lesson about price gouging emerges. The lesson is not the typical defense of price gouging. Economists often make the argument that we shouldn’t place arbitrary limits on prices via price gouging laws because doing so precludes the efficient allocation of resources. I will instead argue something quite different, that price gouging is unlikely to occur.
Thus, in the true spirit of Economic Forces, I am going to explain why prices are sticky and why price gouging isn’t a problem — but not for the reasons you might think.
As most of you reading this are aware, I work at a university. Like most universities, the student union at my university has a place where people can buy food from various chain restaurants. If you walk through this area at various times during the day, you see very different scenes. Walking through the student union around noon reveals very long lines, especially at places like Chik-Fil-A. Yet, if one walks around the same exact spot just two hours later, it is very unlikely that anyone is waiting in line.
College football is also popular at my university. It is so popular that when the university’s football team has a home game, thousands of people come from various other places around the country to see the game. On the nights before the football games, these visitors crowd the local restaurants. Those who arrive at a restaurant on the Friday night before a football game are met with a long waiting list to get a table.
I could go on with similar examples, but the point seems evident. The demand for Chik-Fil-A is greater at lunch time than it is at 2:00 p.m. The demand for a table at a restaurant is higher on Friday nights before football games than a typical Friday night. Yet during these periods of high demand, these restaurants choose to ration their meals by having their customers wait rather than by increasing their prices.
If you are a staunch advocate of using supply and demand to understand behavior, this presents a challenge. These increases in demand are completely predictable. Why do firms choose to ration by having their customers wait rather than by using the price mechanism? This question is further complicated by the fact that on the Friday night before football games, some restaurants do raise their prices relative to a typical Friday night. However, they still have long waiting lists to get a table and the prices do not respond to the length of the list.
Armen Alchian suggests that the answer to this question revolves around search costs. He uses the example of a newsstand (remember newspapers?!). He points out that a newsstand tends to have a lot of business in the morning when people are on their way to work. These people often have to wait for the newspaper because there is excess demand at the current price. Later in the day, there is less traffic at the newsstand. Nonetheless, the price is the same at both times of day. So what does this have to do with search costs? If one shows up to buy a paper and finds that there is no line, but the price is four times higher than usual, the customer might walk to another newsstand to check the price. However, leaving and going to another newsstand is costly. One has to give up the time to find another newsstand and there is no guarantee that the price at the other newsstand will be any lower. Nonetheless, if there is uncertainty about the price, customers might search several newsstands each morning to see who has to better price. The fact that we do not see newsstands vary the price of newspapers with fluctuations in demand is evidently because the cost of waiting in line to buy the paper at a fixed price is lower than the search costs.
This might be evidently true, but why?
As Yoram Barzel points out, the amount of time spent waiting depends on the consumer’s willingness to pay. Rationing by waiting implies that the marginal buyer will pay an amount less than or equal to his/her willingness to pay in terms of both monetary costs and the time cost of waiting. This latter cost, however, is not captured by anyone. In other words, it is well known that when the price is below the equilibrium price, the total surplus from trade at that below-equilibrium price is smaller than at the equilibrium price. As Barzel points out, however, the decline in the surplus at this lower price is actually bigger than a casual observer might think. The reason is that some of the buyer’s surplus is “spent” waiting in line. This implies that there is some amount of the surplus from trade that goes to no one (note that this is over and above the deadweight loss from having “too low” of a price). Barzel argues that in these scenarios, firms have an incentive to try to capture some of this lost surplus.
Beyond that, Barzel’s argument isn’t really helpful here. The reason is that Barzel is referring to a scenario in which the monetary price is zero. In the examples that I gave, the non-price rationing is voluntary. The restaurants could capture more of the surplus by adjusting their prices.
Nonetheless, Barzel’s insight might show us the way. David Haddock and Fred McChesney draw on Alchian’s work to explain “contrived shortages” created by firms, like those of the restaurants I described. What they point out is that firms often invest in intangible capital, like their “brand.” One way to develop a reputable brand is to produce a high-quality good. However, competitors will also invest in producing a high-quality good. What consumers care about when comparing across brands is the price-to-quality ratio.
Before consumers make an initial purchase, they likely search around to compare the price of the good relative to its perceived quality. Once these consumers have identified a good they want to consume based on its price-quality ratio, the firm can foster brand loyalty by maintaining not only the quality of the brand but also the price. The reason is that the consumer is content with the price given the quality. Rather than searching every time they consume, the consumer will prefer to buy the same quality brand at the same price and avoid costly search.
Changes in the price of the good will potentially disrupt this loyalty. The price change might induce the consumer to search anew. In fact, the very nature of moving to the new equilibrium at the higher price necessarily implies that the quantity of trade will decline. This more than likely manifests in fewer customers, at least to some degree. If other firms have not changed their price, the consumer might choose one of these other brands and switch their loyalty as well.
Now, let’s think about how this might apply to our examples. Consider the average restaurant in Oxford on the eve of a football game. Some of the people who are going to that restaurant have been there before. They have a sense of the quality of the food and the price, based on previous experience. There are other restaurants that they have never visited. They do not know the quality and the prices of the food at those places. Going to a bad restaurant is costly. You can often observe the prices before you go in, but you cannot (perfectly) observe the quality of the food without tasting it. Part of the cost of search is discovering bad meals. A constant price might therefore build brand loyalty among the frequent visitors since they know what price and what quality to expect.
However, if these frequent visitors arrive and they observe a dramatically different price than the last time, this might make them wonder why and encourage them to search for other restaurants. If they expect that the costs of search plus the monetary costs of the meal will be lower than the new, higher monetary cost of the meal, then they will choose to search. Furthermore, new visitors might be turned off by a relatively high price for food of unknown quality.
Competition might motivate some restaurants to have different pricing policies. If one restaurant adjusts its price to clear the market without a waiting list, another might keep its price constant and maintain a waiting list. Those who observe the high prices at the busy restaurant might be led to search for a better deal and might be more likely to settle on a restaurant with the lower price. This is almost certainly true for new visitors and is also quite possible for recurring visitors suddenly met with higher than expected prices. If the new visitors find the quality acceptable at the lower price at the alternative restaurant, they might never try the original restaurant that had much higher prices during the busy dinner hours. The recurring visitor might switch their loyalty.
Ultimately, what this seems to suggest is that firms care not just about the level of demand at a moment in time, but also the present-discounted value of future demand from loyal customers. Raising prices to capture the extra surplus during a busy Friday night might lower future demand from customers who would have tried the restaurant at a lower price and become a loyal customer or from customers they lose to search. One might therefore think of the foregone surplus today as an investment in customer loyalty that produces a higher present discounted value of future demand.
So this seems like a possible explanation of why prices are sticky. But what does this have to do with price gouging?
The simple answer is that this theory of sticky prices seems to provide a prediction about how firms will behave when we experience sudden shortages and natural disasters. During times when there is a sudden shortage or a natural disaster, there is excess demand for particular goods. Firms could eliminate this excess demand by raising prices. However, they often do not. In fact, like Chik-Fil-A at lunchtime, they tend to keep their prices constant, despite the shortage.
A few years ago, residents of southeast Michigan and northwest Ohio were told not to drink the water because of contamination in Lake Erie. I happened to be in the area and visiting family when this occurred. I gave an account of what happened to the Jim Tankersley, then of the Washington Post. In short, what occurred was a substantial increase in the demand for bottled water. Despite higher demand, the big retailers in the area did not increase the price of bottled water. Instead of increasing the price, most retailers quickly announced that they were diverting water shipments from other locations to meet the increased demand.
The one exception in terms of pricing behavior seemed to be independent convenience stores, some of which were charging very high prices for cases of bottled water.
Both the behavior of the big retailers and the independent convenience stores are consistent with the theory I outlined above. Big retailers have a much more significant brand name than the independent convenience store. Furthermore, while the independent convenience store is likely to have loyal customers, cases of bottled water are unlikely to be something that loyal customers regularly purchase. Thus, for the big retailers, there is a cost associated with raising prices if one’s competitors do not because previously loyal customers might switch to another big retailer if they can buy water at that retailer at the typical price. The expected loss of future demand is sufficient for them to maintain their typical price. Nonetheless, to prevent their loyal customers from defecting to another brand, the store needs to communicate that the shortage is only temporary, which will discourage a costly search. This is exactly what these firms did.
Finally, although stores were initially caught off guard by the increased demand for bottled water, when the new shipments started, these stores started imposing quantity limits on customers. These quantity limits might prevent people from getting the quantity that they demand, but they allow more customers to satisfy some of their demand. This allows them to satisfy at least some of the demand for a larger number of their loyal customers than if they stuck with the first-come-first-served allocation.
This argument is different from the often standard response that one gets from an economist about price gouging. Typically, economists will argue that price gouging laws unnecessarily restrict prices from coordinating supply and demand and therefore actually makes things worse. My argument is completely different. I think that price gouging laws are largely irrelevant when firms have brand names.
Maybe you still doubt this argument that prices are sticky because of search costs. If so, you might also doubt my claim about price gouging. How can I convince you search costs are the right explanation? The theory has other predictions that we can look at. For example, if search costs are the underlying cause of sticky prices, then we should expect prices to be more flexible when the search costs are low.
Casual observation suggests that this is correct. Gas stations, for example, change their prices quite frequently. A natural disaster or some other type of supply disruption often leads to an increase in the current demand for gasoline due to the expectation of higher future prices. Despite this being a temporary increase in demand, however, gas stations raise prices immediately. Within the context of this search cost paradigm, the reason is simple. The search costs associated with gasoline prices are low. One can observe gasoline prices simply by driving by a gas station. Since this can be done when consumers are already driving for some other purpose, the marginal cost of search is near zero.
In this scenario, contrived shortages seem unwise. The marginal willingness to pay is above the market price. Customers care not only about the amount of monetary expenditures, but also the amount of time they spend waiting in line. A customer who observes low prices, but long lines, might be willing to continue searching to see if the total cost (monetary and non-monetary) is lower since search costs are very low.
We at Economic Forces are ardent advocates of using supply and demand to explain all sorts of behavior. Contrived shortages seem like something that would challenge the usefulness of supply and demand. Not so fast.