The first glass of water when you are thirsty is worth a great deal. The fourth is worth almost nothing. The seventh is unpleasant.

That is the whole idea, and its reach is much wider than the textbook version suggests, because the curve it describes applies to most things people pursue and almost nobody plans around it.

The shape, and where it bends

The useful question is never whether the curve declines. It is where it flattens, because that is where additional effort stops paying and something else becomes the better use of the same effort.

For money, the flattening is well documented and lower than most people assume. The difference between poverty and a modest stable income is enormous and touches nearly everything — health, relationships, sleep, the ability to absorb a setback. The difference between comfortable and rich is smaller on every one of those measures, and past a certain point the marginal dollar mostly buys optionality and status rather than experience. Where exactly it flattens is argued about; that it flattens is not.

For time, the shape is the same. The first hour of unstructured time in a week is transformative. The fortieth is not, and people with unlimited time frequently do less with it than people with some.

For choice, the curve turns negative. A few options are better than one; thirty options are worse than five, because the cost of choosing and the regret afterward both rise with the size of the set.

What it changes

It makes the second question the right one. Not "is more of this good" — more of most things is good — but "is more of this better than the same effort spent on something I have none of?" A person with a great deal of money and no unstructured time is usually further along the flat part of one curve and at the steep start of another.

It explains why balance is arithmetic rather than a slogan. Pushing one thing while at the flat part of its curve, while another sits untouched at the steep part, is a worse allocation than moving effort across. That is not a moral claim about a well-rounded life; it is the same reasoning that says you should fix the constraint rather than optimize the step that is already fast. See the bottleneck.

It sets a stopping rule. Effort has its own declining curve — the tenth revision of a document is worth less than the second — and knowing that is what makes it possible to stop deliberately rather than from exhaustion. This is what satisficing formalizes.

Where it is doing the work underneath something else

Several ideas people hold separately are this one seen from an angle.

Why the 80/20 shape recurs. If the value of successive units declines steeply, then a small number of the earliest, largest units carry most of the total. The Pareto principle is what a steeply declining marginal curve looks like when you add it up.

Why adaptation is affordable. The hedonic treadmill means the felt value of a gain decays over time; diminishing marginal utility means the value of the next gain was already lower than the last. They compound in the same direction, which is why the third version of the same improvement is so consistently disappointing.

Why insurance and reserves make sense. Losing a dollar when you have few hurts more than gaining one when you have many helps, because the curve is steeper at the bottom. That asymmetry — and not fear — is the rational case for holding a margin of safety.

The exception worth knowing

Not everything declines. Some things compound, where having more makes the next unit worth more rather than less — skill in a field where the pieces connect, a reputation, a network, an understanding deep enough that new facts have somewhere to attach.

Those are rarer than they are claimed to be, and they are the things worth pushing past the point where the ordinary curve would say stop. Telling the two apart is most of what a good decision about where to spend a decade consists of.