Margins of Adjustment: Inflation Edition

June 29, 2022
Price Inflation
Here’s how my mom knows something’s up:
notion image
Here’s the only slightly-more-sophisticated way economists know something’s up:
notion image
(Yes, I recognize there’s not overwhelming consensus among professional economists regarding the source of our price-inflationary woes. At least in this instance, that strikes me as an indictment of the state of macroeconomic thinking).
What I want to suggest is that, as significant as this price inflation is, it likely does not capture the full story regarding the magnitude of dislocation that monetary inflation has caused. As with other interventions, monetary inflation allows producers many different margins of adjustment. Price is not the only margin that adjusts.
To the extent that producers have other margins available to them, a myopic focus on prices will understate the true magnitude of inflation. In other words, but for these other margins of adjustment, prices would have risen even more than they have.
Economists do not currently possess a good understanding of what all these margins are or how they interact with price increases. That more research has not been done on this topic may reflect a wholly artificial divide between micro- and macro-theory.
Economic Goods
Goods, said Carl Menger, satisfy human wants.
It turns out, though, that there’s more to goods than meets the eye. Goods are bundles of attributes, a point Yoram Barzel develops in a classic 1982 paper titled “Measurement Costs and the Organization of Markets.” What he means is that goods are a collection of characteristics that yield valuable services to the humans who consume them.
Even a simple good like an orange illustrates this point. Oranges differ in their weight, size, color, juiciness, freshness, expiration date, sweetness, seededness, probability of transmitting disease, fertilizer used to grow them, ease of “opening” (how thick is the skin?), and capabilities in satiating hunger or slaking thirst.  That’s quite a few attributes for something as simple as an orange.
notion image
Condensing Barzel’s brilliant paper quite a bit, sellers sort over some of these attributes, while buyers sort over others. When you go to the orange bin, there’s one price listed, but you don’t begin blindly grabbing oranges at random. You spend time sorting, that is measuring the oranges’ attributes, and selecting those which you judge to be a good choice given the (uniform) price posted above the bin. Typically, sellers have already pre-sorted to some degree (there aren’t usually putrid, rotting oranges in the bin), but it’s clear they leave some sorting to the buyer too because oranges in the bin aren’t perfectly homogeneous.
Consumers tend to prefer attribute homogeneity, other things equal. In other words, given some “average” level of attributes, buyers would prefer the distribution to be tighter around the mean. The reason is simple. With a tighter distribution, they do less searching, searching takes time, and time has alternative uses.
notion image
Attributes are a Choice Variable
Producers aren’t “prisoners” to a good’s attributes. Producer choices—ultimately aiming at satisfying consumer preference—influence the mix of goods’ attributes. For example, a choice to extend the growing season could result in oranges that are riper and larger.
In earlier work, Barzel argued that per-unit taxes cause consumers to shift out of a good’s taxed attributes and into attributes that are untaxed. In a nutshell, an excise tax (say, on cigarettes) taxes the quantity of cigarettes a consumer buys, but not the quality. Other things equal, an excise tax causes consumers to alter their composition of purchases so that higher-quality cigarettes comprise a larger share of cigarette purchases than before the tax.
notion image
Now consider two adjustments to this theory of adjustment. Imagine inflation instead of an excise tax and imagine producer adjustments instead of consumer adjustments. Inflation is, among other things, an “invisible tax” (on cash-holders and savers). Here, I’m using the “Austrian” definition of inflation—an increase in the money supply—because I think it’s the most satisfactory definition for understanding real-world phenomena. An increase in the money supply increases demands for goods, which in turn may raise their prices.
Of course, sellers raising their prices is an important—maybe the most important—part of any inflationary story.
But now imagine other possible adjustments for sellers. Sellers have far more margins of adjustment available to them than a myopic (though important and necessary) focus on prices would suggest.
For example, it’s well-known that sellers will sometimes maintain the prices of the goods they sell, all while reducing the size of the product, so-called “shrinkflation.” Size, mass, weight—these are all attributes of a good. Whether sellers facing inflation will a.) maintain a good’s attributes but increase its price or b.) increase the price while changing a good’s attributes or c.) some combination of the two is determined like everything else is. Namely, what’s profit-maximizing?
But “shrinkflation” shouldn’t get all the attention.
A Universe of Attributes
Consider just how many margins of adjustment there are. Suppose that oranges are such that visual inspection does not easily reveal bruising. Suppose further that the quality of the orange falls when it’s manhandled. A consumer only learns of this orange abuse after she removes the skin. Suppose further that there are many ways to grow, process and transport oranges, some using more labor, and some less. Lastly, imagine that the process using more labor ends up being gentler on the oranges.
One way for a producer to adjust would be to fire some of these workers (i.e. cut costs) and offer a roughed-up orange whose price hasn’t changed. That’s a margin of adjustment that may be available. (Horticulturists, forgive me if the example is strained).
Think about the distribution of attributes within the orange-bin too. We might think of this distribution as a “meta-attribute” of the oranges with, remember, consumers preferring more homogeneity to less, other things equal. Measuring and sorting over these attributes is costly. A seller could cut costs by simply engaging in less pre-purchase sorting.
The oranges that then arrive in the bin exhibit a higher variance over the attributes consumers care about and so consumers spend more time sorting and measuring, while still paying the old money price. Measurement costs for buyers have risen, even if money prices have not. Buyers, rather than sellers, must now bear this measurement cost. Inflation by another name.
Here’s shrinkflation portrayed visually:
notion image
A Variation Mystery
Good applied economics is largely about explaining variation across time and space.
What explains this variation in producer choices? Why do some producers raise prices, while others alter attributes? Why do some do a mixture of both? Is there variation among producers selling roughly the same sort of good? My intuition would suggest no, since a good’s attributes likely play a large role in determining the profit-maximizing way to adjust. Among those that change attributes, why do some select one attribute, while others choose something else?
There do seem to be some patterns that emerge. For instance, perusing Wikipedia’s article on “shrinkflation”, I’m struck that the phenomenon tends to afflict processed consumer food items disproportionately. At least, that’s the lion’s share of examples that the article offers.
Providing more rigorous microfoundations for how the inflationary process plays out seems a worthwhile, and very challenging, goal. If someone knows of existing research that seeks to explain the variation I mention above, I’d love to know about it.