The Marginal Utility

Marginal utility

Marginal utility is the change in benefit from consuming an additional unit of a good or service. Diminishing marginal utility means that successive units provide less additional benefit, holding the relevant circumstances fixed. It is a common pattern, not a guarantee for every activity and situation.

Suppose you can afford two beers, two sandwiches, or one of each. You might prefer the combination because a second beer or sandwich adds less benefit than the first. That choice is not automatically best: it depends on your preferences, the prices, and the circumstances.

The source refers to Graph 8, Total Utility vs Marginal Utility; the graph is not included here.

A marginal decision compares the benefit of an additional unit with what must be given up to obtain it. Equating marginal benefit and marginal cost is useful when both are expressed comparably and the choice permits such a balance. Utility and a monetary price cannot simply be equated without specifying how they are measured.

A farmer with surplus food and worn-out clothes may benefit from trading with a tailor who has spare clothes but needs food. Each gives up something valued less at the margin for something valued more. A mutually acceptable exchange does not require their subjective valuations to be identical.

Time is one opportunity cost. The farmer could learn tailoring, and the tailor could farm, but both would sacrifice time otherwise spent on their existing skills. Specialization can make trade attractive. Economies of scale are a related but distinct possibility: spreading fixed costs across more output or sharing administrative resources may reduce average cost. Neither benefit follows merely from becoming larger.

When comparing purchases within a budget, marginal utility per unit of price is one useful concept. Here, "utility" means the benefit represented in the model. The ratio MU to P compares extra benefit with price, subject to the assumptions needed for that comparison. Under suitable conditions, allocating spending across goods can bring these ratios into balance.

Translating money into hours of work can make a tradeoff more concrete, but it does not make time the sole measure of value. Enjoyment, relationships, uncertainty, and what else a person could do remain relevant. Those limitations become clear in the source’s proposed comparison of dating and paid sex.

A cost model and its limitations

The model makes a "3 dates before sex" assumption and assumes that the man pays and earns $10 per hour. These are stipulated assumptions, not a rule about relationships or an entitlement to sex. Its first coffee date costs $20 and takes 2 hours. At $10 per hour, earning the money and attending the date takes 4 hours.

The dinner date costs $50 and takes 3 hours. At $10 per hour, the combined time cost is 8 hours.

On the third date, the woman cooks; the model counts 3 hours spent together and reports a total time cost of 3 hours. It leaves her work and other costs outside the calculation.

The account then reports a total of 17 hours, or $170 at $10 per hour. That total does not follow from the preceding rows: four plus eight plus three is fifteen. A conclusion based on the stated total therefore needs correction before any comparison is made.

The alternative uses an unresearched Craigslist price of about $150 per hour, counting 1 hour together and 15 hours of work to pay for it. This is a historical assumption in the example, not a verified price or a complete account of the choice.

The objections expose a deeper problem than arithmetic: "The man got to enjoy himself on the dates as well." Another is, "He may be able to have sex with the woman he is dating more than once, thus, being able to allocate the cost of the initial dates to more sexual encounters bringing down the cost." Enjoyment does not become irrelevant because it is hard to quantify. Nor does a relationship have a guaranteed output over which an initial expense can be allocated. An opportunity-cost comparison also depends on what the person would actually have done instead.

The source’s next calculation changes its assumptions again. It proposes three more encounters lasting 1 - 2 hours, mentions a cost of $20 per encounter, then divides $170 by three to obtain about $57 and adds $15 of time cost to reach $72. These steps do not establish a consistent average: three additional encounters would not be the same denominator as three encounters in total, and costs already included in the initial calculation must not be counted twice. The resulting figure cannot support the claimed comparison.

Repeated activity can spread a fixed cost across more occasions. Whether sex with the same partner 3 times provides the same benefit as experiences with different partners, however, is a question about preferences and circumstances, not something that follows from dividing a cost. Consent and the other person’s preferences are also outside a model that treats the choice as one person’s expenditure alone.

Applying the idea

Time spent on one activity is unavailable for another, so opportunity costs help compare choices. But minimizing effort is only useful after identifying what matters in the outcome.

The source calls its illustration the "Hookers vs Dating" problem and refers to a "same night close" as a shortcut. Those labels do not solve the valuation problem. If companionship, mutual enjoyment, or a long-term relationship matters, a calculation aimed solely at a sexual encounter answers a different question.

For an ongoing relationship, initial search costs may become a small part of a long shared life. That observation does not turn later experiences or obligations into interchangeable units. Marginal analysis is useful when its units, alternatives, and assumptions fit the decision being made.