Opportunity cost and marginal thinking

Price a decision by what it displaces rather than by what it costs.

The cost of anything is what you gave up to get it. That sentence sounds like a slogan and it is actually a procedure, because it tells you which number to look for and the number is usually not on any invoice. Everything else in this lesson is that procedure applied twice: once to what a choice costs, and once to how much of it to do.

The cost is the alternative

A free concert ticket costs whatever you would have done with that evening. A four-year degree costs the tuition plus four years of the salary you did not earn, which for most degrees is the larger of the two. Money sitting in a checking account costs whatever it would have earned in a savings account, and at 4% on $10,000 that is $400 a year for doing nothing.

None of those costs appear anywhere. They are all real, they all change decisions, and the habit of looking for them is most of what economic thinking is.

ConventionAccountants and economists mean different things by profit, and the difference is exactly this. Accounting profit is revenue less the costs somebody invoiced you for. Economic profit also subtracts what your time and money would have earned elsewhere. A business with an accounting profit of $70,000 and an economic profit of negative $20,000 is a business that is quietly costing its owner money while looking successful in every document.

How much, rather than whether

Most real decisions are not yes or no. They are how many: how many hours to work, how many stores to open, how much to study. For those, the question to ask is not whether the whole thing is worth it. It is whether the next one is.

A coffee cart below. Each hour brings in less than the last, because the morning rush comes first. Each hour costs a little more, because tired people work slower. Find the closing time.

what the hour brings in what the hour costs

Hours open6

That last hour, on its own

Profit for the day

$135

Average per hour

$23

Best you can do

6 hours, $135

Hour 6 paid for itself and hour 7 would not. That is the whole rule.

Count the $15 an hour you could earn elsewhereCount the $200 cart rental, which you owe either way

Then use the two switches. One counts the wage you gave up to be there, which is arriving in the middle of a marginal decision. The other counts the cart rental, which you owe whether you open or not, and which will refuse to move the answer no matter what you do to it.

Worked example

Three ways to get it wrong

The cart takes $60 in its first hour and $5 less each hour after. It costs $20 in the first hour and $2 more each hour after. Six hours is the right answer, making $135.

Stopping where the average is best gives one hour. Profit per hour peaks at $40 in the first hour and falls from there. Following it makes $40 instead of $135. Averages are a summary of what already happened and they are silent on what to do next.
Carrying on while profitable gives ten hours. The day is still in profit at hour seven, at hour eight, and beyond. Each of those hours lost money on its own, and the total went down every time. Being profitable overall is not a reason to do the next thing.
Counting the rent changes nothing. Add a $200 daily rental and the best answer is still six hours, now losing $65. Six hours and losing $65 beats closed and losing $200. A cost you owe either way cannot tell you what to do, which is the most practically useful thing on this page.

The figures come from the model on this page and are checked by tests, including one that confirms across 36 parameter settings that the profit-maximizing hour count is always the last hour that pays for itself.

Why the wage you gave up is the hard one

Counting the rental is easy because somebody sends a bill. Counting your own hour is hard because nobody does, and it is the one that changes the answer.

At $15 an hour of alternative work, the cart should close after four hours instead of six. Nothing about the cart changed. No cost went up in any account. The right decision moved by a third because the comparison finally included the thing being compared against.

What breaks

Opportunity cost is the next best alternative, not all of them added up. Spending an evening reading costs you the single best thing you would otherwise have done with it. It does not cost you the sum of every possible evening, which would make every choice infinitely expensive and the concept useless.

Sunk costs are irrelevant to the decision and not to the lesson. The $80 ticket should not affect tonight. It should absolutely affect whether you buy non-refundable tickets in future. Ignoring a sunk cost when choosing and learning from it afterwards are both correct.

Marginal thinking assumes you can adjust in small steps. Some decisions only come in one size. You cannot open a factory for six hours of the day, and you cannot hire a third of an engineer. Where the increments are large, the marginal rule turns back into a yes or no question about one large lump.

What to do with this

Two questions, and they cover most of it. When somebody quotes you a cost, ask what it displaced. When somebody asks how much of something to do, ask what the next one adds and what it costs, and stop caring about the total.

The second one is the harder habit, because everything around you reports totals and averages. Your bank shows a balance, your business shows a profit, your report card shows an average. None of those answer the question you are usually asking, which is what to do next.

Test yourself

01You paid $80 for a concert ticket. On the night you would rather stay home, and the ticket cannot be resold. What should you do, and what is the $80 doing in this decision?

Stay home. The $80 is gone whichever you choose, so it cannot distinguish between the options and has no business in the comparison. The only question left is whether the evening at the concert beats the evening at home, and you have already said it does not.

Going anyway is the sunk cost trap, and the reason it feels compelling is that not going seems to waste the money. The money was spent when you bought the ticket. All that is left to decide is how to spend the evening.

02A friend says their business made $70,000 last year, so it was a good year. What is missing?

What they could have earned doing something else, and what the money tied up in the business could have earned somewhere else. If the alternative was a $90,000 job, the business cost them $20,000 to run last year and the accounts will never show it. Accounting profit counts what was paid out. Economic profit subtracts what was given up, and that is the number that answers whether it was a good year.

03Why is 'we are still making money, so keep going' a bad reason to keep going?

Because it answers a question about the whole using a rule that only applies to the next step. A business can be profitable overall while the next hour, the next store, or the next hire loses money. The right test is whether the marginal one pays for itself, and the total tells you nothing about that.

A tutor that knows this lesson. It asks before it explains, and it will not tell you what to do with your own money.

Educational material, not investment or policy advice. Figures are cited where they come from a filing or a statistical series, and labelled as illustrative where they do not.