Diminishing marginal value as delay discounting.

Kennedy (1992) wrote a theory paper. He asked: why does the second candy bar taste worse than the first? Economists call this 'diminishing marginal value.' The author said three behavioral rules explain it.

You get full after too much. You value now over later. You always pick the best option on the menu. Put those together and extra units lose their shine.

The paper shows the economic law is just behavior analysis in disguise. Delay discounting does the heavy lifting. The longer you wait, the less the next unit is worth.

The idea matters for addiction. Each extra hit of a drug is worth less, but people still choose it because the delay to natural rewards is too long.

Green et al. (2019) extends the story. They warn: don't call clients 'impulsive.' Steep discounting is tied to the moment, not a trait. Their data show many substance users discount money like controls. The 1992 rule still holds; the context just shifts.

Stancato et al. (2020) and Miller et al. (2024) both test the theory on college drinkers. Stancato finds splitting magnitude from delay sharpens prediction of alcohol problems. Miller goes further: a 2-minute 'willingness-to-wait' task beats classic discounting at flagging risky drinking. The core principle survives; the yardstick got better.

Clarke et al. (2003) puts the theory to work with brain-injured adults. Adding a simple hand-open task makes delayed rewards win. It shows you can hack the delay curve in real time.

When you see a client over-eat, over-game, or over-drink, think 'delay + satiation + best option.' Shorten the wait for healthy rewards or add small tasks that make the delayed choice feel closer. You now have a 30-year trail of studies showing the rule holds from pigeons to college kids to rehab patients.