What makes a Margin?
What makes one cost “marginal” and another “fixed”? When economists discuss fixed costs, we often invoke capital intense long term projects, but at one point in time, a human being decided to make those investments. And at that moment in time, those investments were driven by incentives. One way to think about margins is that they are decisions about the differences between two close alternatives. On the margin, do you buy 4 apples or 5 when you go to the grocery store? You compare the cost of the fifth apple to the benefit you get from eating a fifth apple that week. Maybe you like apples, but buying 5 means you are eating an apple every single week day. The marginal utility of the last apple is lower, so maybe you’ll get one banana or one orange instead. But sometimes, decisions are lumpy. Once a decision is made, it can often not be reversed. At the moment you make the decision, you need to rely on your estimation of the future and the incentives you’ll face then. That’s when the incentives switch from marginal to not marginal. When someone can no longer alter their choice, the incentives cease to be marginal. That’s what makes sunk costs sunk. You can hope for better yesterdays, but it won’t do you any good.
Over time, more incentives become marginal, as people get more opportunities to adjust to them. The cannonical example is gas prices. If the gas price doubled tomorrow, people’s consumption of it might change by only 10% at most. It’s pretty inelastic. But if the gas price changed permanently to $7 or so, people would reduce their consumption of gas over the next 5 years quite dramatically. They would buy more fuel efficient cars, they would arrange to carpool more, they would more closer to work, etc. Likewise, once a worker has learned the skills of a particular industry, it is hard for them to change to another. A dramatic change in wages will translate to a dramatic change in welfare. But over time, fewer workers will enter that field and the economy will rebalance. As a final example, a tax on past fixed capital investment or land will initially have low deadweight loss, but in the long run, all costs are variable. Firms will reduce their upkeep on their capital and stop replacing it when it is broken. A tax on land will reduce incentives to develop and allocate it properly. In the long run, all incentives are marginal.
Garett Jones on average vs marginal