The Value of the Alternative

I · Waller’s Rule

In 1938 the American sociologist Willard Waller published a book about the family, and buried in it was a sentence that has outlived everything else he wrote. Waller had spent the previous years studying what he called the rating and dating complex among college students, an arrangement he described with the sourness of a man who found it distasteful. Out of that work came a proposition he stated as a rule: in any pair, the person who cares less about the relationship continuing has the greater power over whether it does. It is now known as the principle of least interest, and it is the closest thing relationship sociology has to a law.

This magazine published an essay elsewhere in this issue called The Shortlist, which examined the proposition that women tend to accumulate suitors and that the accumulation resolves into a choice between two. It concluded, on four literatures, that the queue a woman holds is a consequence of who does the messaging rather than evidence of a collecting strategy. The reporting in it stands. The argument was too quick.

What that essay did was count acceptances. It found that women, offered more candidates, did not take more of them, and it concluded from this that no maximizing was occurring. But acceptances are the wrong quantity. Waller’s rule says that what determines a person’s position is not what they have taken but what they could take instead, and that is a quantity nobody in the speed-dating literature was measuring. An option you decline is not a wasted option. It is the entire source of your standing in the arrangement you keep.

Set out plainly, the case is this. A person’s power inside a relationship depends on the quality of what lies outside it. The evidence for that is not folklore; it is the best-validated model of romantic commitment in the field, tested on nearly twelve thousand people. Anyone who would prefer not to be the more-invested party therefore has a standing reason to want alternatives, whether or not they ever intend to use one. And in the heterosexual market as it is actually organized, alternatives arrive at one address far more often than the other, which means the position is available to one sex more or less by default. None of this requires anybody to be vain, calculating, or cruel. It requires only that people prefer not to be the one who cares more.

That is an argument, and this essay will make it. It will also, before it finishes, present the bill, which is larger than the people who like this thesis usually admit.

II · The Variable With a Name

Waller’s rule sat for decades as an aphorism with an author. What turned it into something testable was a model built in the nineteen-eighties by the psychologist Caryl Rusbult, which proposed that commitment to a relationship is produced by three things and not one. The first is satisfaction: whether the relationship is good. The second is investment: what one has already put in that could not be recovered. And the third, which is the one that matters here, is the quality of one’s alternatives — how good the next best thing looks.

In 2003 the psychologists Benjamin Le and Christopher Agnew pooled the tests of that model: fifty-two studies, sixty independent samples, 11,582 participants. All three variables correlated significantly with commitment, and together they accounted for close to two-thirds of the variance in it. Commitment, in turn, significantly predicted whether couples broke up. The quality of one’s alternatives is not a footnote to the theory of commitment. It is one of its three legs.

Three years later the psychologists Susan Sprecher, Maria Schmeeckle, and Diane Felmlee took Waller’s rule to a longitudinal sample of dating couples, some of whom went on to marry, and asked whether the less involved partner really does hold the power. They found that the less emotionally involved partner perceived themselves as having more control over whether the relationship continued. They also found that perceptions of unequal involvement were common and that they were stable over time. This is not a transient condition that couples grow out of. It is a standing feature of a substantial share of relationships, and both parties can see it.

Put the two together and the mechanism is complete. Commitment falls as alternatives improve. Power rises as commitment falls. Therefore power rises as alternatives improve — not as a metaphor, but through a chain each link of which has been measured. A person with a good next-best-thing is, by the arithmetic of the model, less committed and by Waller’s rule more powerful, in a relationship they may have every intention of keeping.

It is worth saying clearly what this does and does not establish, because the temptation to overread it is considerable. The model shows that alternatives predict commitment. It does not show that anybody goes out and acquires alternatives in order to lower their own commitment, and no study located for this essay shows that. What the model licenses is a claim about incentive, not about intention: it tells us what a person in that position has reason to do, which is a different and weaker thing than telling us what they did.

III · Insurance, Not Appetite

The word that wrecks this conversation is want. To say that women want to maximize suitors sounds like a claim about appetite, and appetite invites a moral response: greed, vanity, cruelty, the usual vocabulary. But nothing in the mechanism above runs through appetite. It runs through exposure.

Consider the position on its own terms. A person entering a relationship is being asked to become dependent on somebody whose future conduct they cannot observe. The available protections are few. One can invest less, which damages the relationship. One can demand guarantees, which are unenforceable. Or one can maintain a plausible outside position — not a plan to leave, simply the standing capacity to — which costs nothing to hold and changes the balance of the arrangement without a word being spoken about it. That is not appetite. It is the only insurance on the market.

And it is not a female instrument. Anybody would want it. The reason the pattern presents as a sex difference is that the insurance is not equally available: the alternatives have to come from somewhere, and in the heterosexual market they arrive overwhelmingly at one address. The messaging records that The Shortlist reported are the relevant fact here, read forward instead of backward. When more than eighty per cent of opening approaches are sent by men, women are handed the raw material of the outside position and men are not. One sex must manufacture its alternatives; the other has them delivered.

Two objections arrive here and both can be answered. The first is that this describes a cynic, and most people are not cynics. But the mechanism requires no cynicism, because it does not require the option to be held deliberately. A woman who simply does not close down every conversation she is not pursuing has an outside position whether or not she has ever thought of it that way; the investment model does not ask whether alternatives were sought, only whether they exist. The second objection is that this amounts to leading people on. It does not, or not necessarily: nothing in the mechanism requires misrepresenting one’s intentions to anybody. The value of the outside position comes from its existing, not from anybody being told about it, which is precisely why it can be held by somebody behaving impeccably.

This is why the thesis is worth arguing even though its usual proponents argue it badly. The interesting claim is not that women are acquisitive. It is that the market issues an asset to one side and not the other, that the asset is valuable in a way the best model of commitment can quantify, and that anyone holding it would be irrational to discard it. A woman who keeps two or three conversations alive while seeing somebody is not, on this reading, doing anything to him. She is holding the only instrument available to a person who would rather not be the one who cares more — the same instrument he would hold if the market had handed it to him.

IV · The Same Finding, Read Forward

Which brings this essay to its disagreement with its predecessor, and to a result both pieces rely on. Between 2002 and 2004 the economists Raymond Fisman, Sheena Iyengar, Emir Kamenica, and Itamar Simonson ran fourteen speed-dating sessions with Columbia graduate students, varying the number of partners each participant met without telling them in advance. They found that male selectivity was invariant to the size of the pool while female selectivity rose strongly with it, and that female subjects were no more selective than males when the pool was small.

The Shortlist read that as decisive against the thesis. Give a woman more suitors, it observed, and she does not take more of them; she raises the bar. Her list stays the same length. Whatever that is, the essay concluded, it is not maximizing.

It is, though — of the only quantity that matters. Raising your threshold as your alternatives improve is not a refusal to maximize. It is precisely what maximizing looks like when the thing being maximized is the value of the option set rather than the length of the list. An economist would call the moving line a reservation value, and its rising with the quality of what is available is the defining behavior of somebody whose outside position is doing exactly the work this essay claims for it. Fisman and his colleagues said as much in their own language, without drawing the conclusion: women’s benefits appeared more concave, or their costs more convex, in the number of dates — which is to say that the value to a woman of the next acceptance falls away fast, while the value of the standing she gains from having had the choice does not.

The two essays are counting different things. Count acceptances and women look like satisficers with a rising standard. Count option value and the same behavior is a person managing a portfolio with discipline: taking few positions, repricing constantly, and holding the rest as capacity. Neither reading is dishonest, and the data cannot adjudicate between them, because no study in this literature measures what a person holds — only what they accept. That gap is where the whole disagreement lives, and this magazine should have said so the first time.

It is worth saying what would settle it, since the two readings are not permanently irreconcilable. Somebody would have to measure the option set rather than the choices made from it: how many live possibilities a person is actually carrying at a given moment, what they believe those possibilities are worth, and whether that belief moves their behavior inside the relationship they are in. The instruments exist — the investment model already asks people to rate the quality of their alternatives — and nobody appears to have pointed them at a courtship market and counted. Until somebody does, this is a disagreement about bookkeeping conducted by people looking at the same ledger, and honesty requires saying that neither side is entitled to declare it closed.

V · The Version This Essay Does Not Need

There exists a stronger and far more contentious form of this argument, and intellectual honesty requires naming it rather than quietly borrowing its momentum. In 2004 the psychologists Roy Baumeister and Kathleen Vohs proposed what they called sexual economics: that a heterosexual community can be analyzed as a marketplace in which men seek sex and offer other resources for it, with women in the structural position of sellers and men of buyers, and with a price set by supply, demand, and competition among sellers. On that account, everything in this essay follows trivially. A seller in a thin market with many buyers holds pricing power, and holding suitors is simply inventory management.

The theory is contested, and not politely. The psychologist Laurie Rudman, with Janell Fetterolf and in a subsequent exchange with its authors, has argued that sexual economics theory is patriarchal in construction, that it makes women responsible for suppressing sexuality in order to protect a price, and that a growing body of evidence contradicts it. That criticism is substantial and it is not obviously wrong.

What the theory gets right, and what its critics have not dislodged, is narrower and duller than either side tends to say: that when one side of a market does most of the approaching, the other side acquires the properties of a scarce good, whatever the good is taken to be. That observation does not require sex to be a commodity, or women to be sellers, or anybody to be suppressing anything. It requires only an imbalance in who initiates, which is measured.

This essay declines the theory, and it is worth being explicit about why the argument survives without it. Sexual economics requires a specific and disputable premise: that sex is a female-held commodity exchanged for male-provided resources. The argument made here requires no such thing. It requires only that commitment falls as alternatives improve, which is measured across sixty samples, and that power rises as commitment falls, which is Waller’s rule and has been tested. Those hold whatever anyone is exchanging, and they would hold in a market with no sex in it at all — which is why the next part of this essay is about employment law.

VI · Why It Rests at Two

Grant the mechanism and the second half of the thesis becomes tractable. If alternatives are valuable because of the standing they confer, then the question of how many to hold is not a question about desire at all. It is a question about how many are required to do the job, and the answer is small.

One alternative does not do it. A single outside possibility is an anecdote: it may be a fluke, a crush, an aberration of one person’s taste, and it can be dismissed as such by the holder herself as easily as by anybody else. It does not establish that she could replace the arrangement she is in; it establishes only that one particular man liked her once. Two is different in kind rather than in degree. Two independent instances stop being events and start being a rate. They demonstrate a capacity to attract rather than an incident of having attracted, and it is capacity, not incident, that sets a reservation value.

At the other end, the number is capped by what a person can actually hold in mind. The psychologists Alison Lenton and Marco Francesconi, working across eighty-four speed-dating events with 1,868 women and 1,870 men, found that as the option set grew, choosers shifted their attention to characteristics that can be assessed instantly — height, weight — and away from those that take time to elicit, such as occupation and education. A large option set is not held; it is triaged. Whatever the inbox contains, the number of alternatives a person is actually pricing at any moment is the number they can keep in view, and that is not forty.

So the process is squeezed from both ends. Below two, the outside position is not credible enough to be worth anything. Much above two, it cannot be maintained with the attention available. Two is where a maximizing process comes to rest once one stops counting names and starts counting the value of the position — and it is, not coincidentally, the arrangement that every story about this subject describes. The story is not distorting the market. It is reporting the equilibrium.

The credibility half of that argument is this magazine’s own and is offered as argument. No located study measures how many alternatives a person holds at once, or reports any distribution clustering at two; the attention constraint is evidenced and the credibility constraint is reasoned. A reader who wants to reject the second half is entitled to, and should know that the first half stands without it.

VII · The Bill

An honest brief presents the costs of its own position, and this one is expensive in a way its usual advocates never mention.

Return to Sprecher, Schmeeckle, and Felmlee, because their paper does not stop where the flattering half of it ends. Yes: the less involved partner perceived more control. But the same study found that equal emotional involvement was associated with greater relationship satisfaction and greater stability, and that unequal involvement was associated with negative emotions and with lower satisfaction after the relationship ended. The asymmetry that confers the power degrades the thing the power is held over. One is buying leverage in an arrangement and paying for it in the quality of the arrangement.

Nor is the leverage particularly usable. What the less involved partner reports is perceived control over continuation — over whether the relationship persists. That is the power to end something, which is the only power a good outside position actually confers. It is not the power to make somebody love you better, and there is no evidence here that it converts into anything a person would want on an ordinary Tuesday.

And the position is stable, which the paper reports and which is the least comfortable finding in it. Unequal involvement did not tend to correct itself over time. Whoever cared less at the beginning tended still to be caring less later. Insurance that never lapses is also a condition one never leaves.

There is a further cost that the studies do not measure and that the argument implies. An instrument whose value comes from being unused has to stay unused to keep its value, which means the alternatives that confer the standing are, by construction, relationships that do not happen. Whatever is on the other end of those conversations — two people who might have suited each other, on a different week, in a different order — is the premium being paid. It is paid quietly, by parties who never learn they were paying it, and nothing in the literature attempts to price it.

So the argument of this essay is not that the maximizing party is winning. It is that she is doing the rational thing, that the rational thing has a price, and that the price falls on her as much as on anybody. That is a considerably less satisfying conclusion than either side of this argument normally reaches for, which is a point in its favor.

VIII · The Available Exit

The value of an alternative depends on whether it can actually be chosen. New York’s many introductions and visible profiles make the idea of an outside option vivid, but the presence of another name on a screen does not make a relationship with that person available. The research earlier in this essay distinguishes perceived alternatives, attention to them, and a decision to leave. The city is a setting in which those distinctions are easy to forget.

Imagine two people who both know that the app could be reopened tonight. The knowledge may make an unhappy arrangement easier to leave. It may also make a workable one harder to evaluate, because every inconvenience is compared with a stranger whose inconveniences are still invisible. Neither outcome is a measured New York effect. The evidence supports a more careful claim: an alternative changes decisions through its perceived quality and attainability, not through a raw count of faces.

That is why the distinction between an option and an appetite matters. A person can keep one possible path in view without wanting to take it. The city can supply abundance without proving that anyone is maximizing a list of partners. The question is whether a particular alternative would improve a particular life after the costs of leaving and beginning again have been counted.

IX · The Value of the Alternative

The thesis this magazine was asked to argue was that women tend toward maximizing the number of their suitors, and that the tendency comes to rest in the two-suitor arrangement. Here is the version of it this essay is prepared to defend.

Power in a pair goes to whoever is less committed; that is Waller’s rule, and the less involved partner does in fact report more control over whether the thing continues. Commitment is set in substantial part by the quality of one’s alternatives; that is the investment model, pooled across sixty samples and nearly twelve thousand people, explaining close to two-thirds of the variance in commitment. Therefore an outside position is the only instrument available to somebody who would prefer not to be the one who cares more — and it is valuable precisely when unused, which is why counting acceptances, as this magazine did last week, measures the wrong thing entirely. The market issues that instrument to one side far more than the other, four opening messages in five running in a single direction. And it settles at two because one alternative is an anecdote and forty cannot be held in mind.

What that argument does not establish should be said as plainly. It shows what a person in this position has reason to do; it does not show what anybody did, and no study located here measures intent, or counts alternatives held, or finds a distribution clustering anywhere. It runs on samples of students and platform users. Its most confident published version, sexual economics, is contested by serious people and has been declined here. And it ends in a place its advocates rarely want: with the finding that the asymmetry conferring the power also predicts lower satisfaction, worse feeling afterward, and a condition that tends not to correct itself.

Which returns us to Waller, in 1938, watching a courtship system he disliked and writing down the rule that explained it. He did not think he was describing an appetite. He thought he was describing an arithmetic — one that would operate in any pair, of any composition, wherever two people wanted something from each other in unequal degrees. The modern argument about whether women maximize suitors is mostly an argument about character, conducted by people who have not noticed that Waller settled the mechanical question ninety years ago and left only the accounting open. On the accounting, this essay says: count the options, not the acceptances. On the character, it says nothing at all, which is the correct amount.

Notes on Sources

Waller, W. (1938). The Family: A Dynamic Interpretation. Origin of the principle of least interest. Waller, W. (1937). The rating and dating complex. American Sociological Review, 2(5), 727–734.† (The 1938 formulation is theory rather than data, and the dating study describes one American student culture of its period. The attribution and dating of the principle were read in reference works rather than in the first edition.)

Sprecher, S., Schmeeckle, M., & Felmlee, D. (2006). The principle of least interest: Inequality in emotional involvement in romantic relationships. Journal of Family Issues, 27(9), 1255–1280. Longitudinal sample of dating couples, some subsequently married. Source of the finding that less emotionally involved partners perceived more control over continuation, that perceptions of unequal involvement were common and stable over time, that equal involvement was associated with greater satisfaction and stability, and that unequal involvement was associated with negative emotions and lower satisfaction after dissolution.† (Correlational and self-reported; it measures perceived control, not bargaining outcomes.)

Le, B., & Agnew, C. R. (2003). Commitment and its theorized determinants: A meta-analysis of the Investment Model. Personal Relationships, 10(1), 37–57. Fifty-two studies, sixty independent samples, 11,582 participants; satisfaction, quality of alternatives, and investment size each correlated significantly with commitment and together accounted for close to two-thirds of its variance; commitment significantly predicted breakup. The model is Rusbult’s.† (Correlational. It establishes that alternatives predict commitment; it does not establish that anybody acquires alternatives in order to reduce their own commitment, and the essay says so in the text.)

Bruch, E. E., & Newman, M. E. J. (2018). Aspirational pursuit of mates in online dating markets. Science Advances, 4(8), eaap9815. Source of the finding that more than eighty per cent of first messages come from men. Four US cities, January 2014.† (One platform, one month; it measures messages, not relationships.)

Fisman, R., Iyengar, S. S., Kamenica, E., & Simonson, I. (2006). Gender differences in mate selection: Evidence from a speed dating experiment. The Quarterly Journal of Economics, 121(2), 673–697. Source of the elasticity result and of the authors’ own reading that women’s costs appeared more convex, or benefits more concave, in the number of dates.† (Part IV reinterprets this published result; it adds no measurement. The reading in this magazine’s companion essay remains available, and the two readings count different quantities, which the text states.)

Baumeister, R. F., & Vohs, K. D. (2004). Sexual economics: Sex as female resource for social exchange in heterosexual interactions. Personality and Social Psychology Review, 8(4), 339–363. Rudman, L. A. (2017). Myths of sexual economics theory: Implications for gender equality. Psychology of Women Quarterly; and Rudman, L. A., & Fetterolf, J. C. (2014), with the subsequent exchange with the theory’s authors.† (Contested at the level of both theory and evidence. Named here and expressly declined; the essay’s argument does not rest on it.)

Lenton, A. P., & Francesconi, M. (2010). How humans cognitively manage an abundance of mate options. Psychological Science, 21(4), 528–533. 1,868 women and 1,870 men across 84 speed-dating events; source of the shift toward quickly assessed characteristics as the option set grows.

† marks findings limited by design, sample, or contested status. Figure 2 is a qualitative editorial diagram; Figure 1 assembles two measured relations into a chain that neither source paper reports. The credibility argument in Part VI is this magazine’s own: no located study measures how many alternatives a person holds at once, and none reports a distribution clustering at two. Part VIII is an editorial application to New York, not a measured city effect. This piece dissents from “The Shortlist,” published in the same issue; both remain in print.

Drafted with AI (Claude Opus 5) as a private research synthesis, September 1, 2026; edited, and answered for, by humans. Commissioned as an argument and written as one: the thesis is defended rather than surveyed, every mechanism it runs on is a published finding, the most convenient supporting theory is named and declined because it is contested, the companion essay’s reading is stated fairly before it is answered, and the bill for the position argued is presented in full. Set, with affection and without affiliation, in the manner of a certain Manhattan weekly.New York revision drafted with AI on October 12, 2026; editorial review pending. ◆ ◆ ◆