The Better Man

I · Just Say Yes

On the twelfth of August, 2001, HBO broadcast the twelfth episode of the fourth season of a comedy about four women in Manhattan. Carrie Bradshaw is being courted by a furniture maker named Aidan, who intends to propose. Before he does, she runs into the financier the show calls Mr. Big, and tells him so. Big informs her that she is not the marrying kind. Aidan proposes anyway; she accepts; the episode is called “Just Say Yes.” Over six seasons the two men will alternate in and out of the position without ever quite occupying it at the same time, and the show will end by awarding it to the one who said she was not the marrying kind.

That arrangement—one woman, two men, and a choice presented as her sovereign act—is the most durable love story New York tells about itself. It is not a New York invention, but the city has done more than anywhere else to furnish it: the apartment, the two telephone numbers, the friends offering counsel over brunch, the sense that the decision is a referendum on what kind of life the chooser wants. Ask an American to picture romance at its most flattering and a version of this tableau will very often be what arrives.

The tableau makes three claims, and they are worth separating before any of them is tested. The first is descriptive: that being wanted by two people at once is a thing that happens, and happens often enough to be the default shape of a love story. The second is evaluative: that it is the best available romantic position, better than being wanted by one person, because a choice made against an alternative is worth more than a choice made against nothing. The third is prescriptive, and it is the one addressed to men: that a man who wants somebody is expected to enter, to be visibly in the running, and that a man who declines to be measured against another man has, by that refusal, removed himself from consideration.

Each of those claims can be checked, more or less. The first is a question about a mechanism, and the mechanism has been measured. The second is a question about what the position is actually like, and the city has a century of testimony on that subject, most of it unhappy. The third is the interesting one, because the evidence for it is thinner than almost anybody assumes, and because the sanction it describes turns out to be levied by an audience rather than by the woman the story appoints as judge.

Two warnings about what follows. The research is overwhelmingly American, heterosexual, and collegiate, and where a study reached beyond students it will be said so; the New York material is cultural and historical, and no study measuring how often two men actually compete for one woman, in this city or any other, was located. And the essay is not a brief against competition or for it. Men compete for things; so does everyone. The question is narrower: whether the tournament described above is a description of how New York works, a demand it makes, or a story it enjoys.

II · Six Per Cent

Underneath the story is a real psychological claim, and it has a name. Mate choice copying holds that the probability of accepting a mate rises when that mate is seen to have been chosen by somebody else already. It was first demonstrated in other animals—guppies, quail—and then in people, and it is the reason the two-suitor plot feels intuitively correct rather than merely dramatic. If a rival’s interest genuinely raises a man’s value, then a contest is not a contrivance. It is a valuation mechanism.

In 2018 the biologists Amany Gouda-Vossos, Shinichi Nakagawa, Barnaby Dixson, and Robert Brooks pooled the human studies and asked how much the effect is actually worth. Their meta-analysis separated two designs: studies that add a member of the opposite sex beside the target, and studies that augment the added person’s apparent quality. In the first design, across seventeen effect sizes, women rated men as about six per cent more attractive or desirable when the man was pictured beside a woman than when he was pictured alone.

Six per cent. It is not nothing, and it is not a rounding error, but it is a very long way from the weight the culture puts on it. Nor is it evenly distributed. Across seven effect sizes for the mirror case—men judging women who were pictured beside another man—the pooled estimate was slightly negative and not statistically significant. The effect that props up the two-suitor plot appears when women assess men and largely vanishes when men assess women. In the augmentation studies the sex difference disappeared: both sexes rated a target higher when the person beside them was more desirable, on five and two effect sizes respectively. What the copying literature is describing, at its most defensible, is that other people’s apparent judgments are information, and that both sexes use them.

The authors then did the thing that separates a meta-analysis from an advertisement. Heterogeneity across the seventeen studies ran to eighty-eight per cent, meaning the studies disagree with each other far more than sampling error can explain. And their sensitivity analysis found what the abstract calls a moderate publication bias in favor of positive reports. Under one of the three trim-and-fill estimators, three studies were missing from the distribution; when they were filled in, the corrected effect fell to about three per cent and stopped being statistically significant. The authors’ own reading is that the bias is probably modest and the key finding survives. That is a fair reading. It is also fair to notice that a plot device carrying this much cultural freight rests on an effect that one standard correction can halve.

III · Three Hundred and Seventy Names

New York’s habit of organizing courtship as a ranked competition is much older than the television, and for most of the city’s history it ran in the opposite direction. On the sixteenth of February, 1892, the New York Times published a list supplied by Ward McAllister, the social arbiter who managed the calendar of Mrs. Astor. He produced it under pressure: a rival paper had printed its own accounting of who mattered, putting the number at a hundred and fifty, and McAllister wished to correct the record. His own figure was four hundred, a number he had derived from the capacity of Mrs. Astor’s ballroom. The list as published carried somewhat fewer names than that, which is the sort of detail the city has always found funnier than the list itself.

The Four Hundred was a marriage market with published standings, and in it the competition was among women and among families. Rank was the scarce good and men above a certain line held it; daughters were the entrants; the season was the tournament. Edith Wharton, who grew up inside that arrangement, wrote its two great examinations. In The House of Mirth, published in 1905, Lily Bart is beautiful, well born, and without money, and the novel is the record of what happens to a person whose entire value is her standing in an appraisal she does not control. In The Age of Innocence, 1920, the machinery is gentler and just as total.

This matters for the modern story in a specific way. If being competed over were simply the natural and best romantic position, one would expect the culture to have placed women there consistently, since it is women whom the current story places there. It did not. For most of the city’s documented social history the person being competed over, in the sense of being appraised against alternatives by a market with published criteria, was as often the woman, and the experience is recorded almost entirely as an ordeal. The two-suitor tableau is not the discovery of a natural arrangement. It is a recent reassignment of an old one, and the older testimony about what that position feels like is not flattering.

What is continuous is the city’s taste for legibility. New York likes a ranking, likes it published, and likes to know where the line is drawn. A romantic tournament is that taste applied to private life: it converts an unmeasurable question—whether two people should be together—into a measurable one, which is who came first. The conversion is what makes it satisfying, and it is also what makes it wrong.

IV · Whom the Contest Is With

The story assigns three roles: two competitors and a judge. The scientific literature on how human mating competition actually works has trouble filling the third one. In 2010 the anthropologist David Puts published an argument that has been contested ever since, and its target was the field’s own consensus. Evolutionary psychology had settled on mate choice—principally female choice—as the primary mechanism of sexual selection in humans. That conclusion, Puts argued, conforms neither to theoretical predictions nor to the available evidence.

His reason is uncomfortably simple. Contests override choice. If one competitor can exclude another by force or the threat of it, then choice, and every other selective mechanism, is foreclosed before it can operate: you cannot select from a set that a rival has already emptied. From the two-dimensionality of the human environment, the clustering of mates and competitors in space and time, our phylogeny, and a series of anatomical considerations—the male-female differences in upper-body strength and in the equipment for delivering and absorbing blows—he concluded that contest competition, not female choice, was probably the primary mechanism of sexual selection operating on men.

This is an argument, not an experiment, and it should be read as one. It also licenses nothing whatever about how anyone should behave; a claim about the Pleistocene is not a permission slip. But it does relocate the judge’s chair. On the contest account, the crucial audience for male competitive display is other men, because the function of the display is to establish a position among them, and the woman’s assessment operates downstream of an arrangement she was not asked to referee.

The catalogue of what men actually do points the same way. In 1988 the psychologist David Buss assembled a taxonomy of mate-attraction tactics and then had them rated for how often people performed them. In the undergraduate sample of a hundred and four, close friends reported that men displayed resources more often than women did—0.80 against 0.37 on the study’s frequency scale—and bragged about resources more often, 1.03 against 0.59. Men were also reported higher on displaying strength, displaying athleticism, acting promiscuous, and showing off. Almost every item on that list is a performance requiring spectators.

Buss’s paper is also unusually honest about what did not work, and the failures are informative. Wearing sexy clothes, acting provocative, and flirting had been predicted to be women’s tactics, and none of the three produced a significant sex difference: 0.67 against 0.71, 0.81 against 0.88, and 1.89 against 1.67. The abstract concedes several predictive failures and unanticipated findings. What survived was the display cluster, which is precisely the part of male behavior that needs an audience and does not need an adjudicator.

V · The Entry Decision

Whether men are expected to compete is a question about expectation, which is hard to measure. Whether men enter competitions when offered one is a question about behavior, which is not. In 2007 the economists Muriel Niederle and Lise Vesterlund published the cleanest available answer, in a setting with no romance in it at all. Their subjects performed a real task—adding up sets of numbers—first for a piece rate, then in a tournament, and were then asked which scheme they wanted for a further round.

Seventy-three per cent of the men chose the tournament. Thirty-five per cent of the women did. There was no gender difference in performance. The gap was produced by two things: men were more overconfident about where they stood, and men were simply more willing to perform in a competition, independent of what they expected to earn from it.

The second half of the paper’s title is the half that gets dropped in retellings, and it is the half that matters here. Men do not merely compete more; on the paper’s analysis they compete too much. A substantial number of the men who entered the tournament would have earned more under the piece rate. Entering was not, for them, a strategy. It was a preference, and it cost them money.

Carrying this into courtship is an analogy and must be labelled as one. Nothing in the experiment concerns romance, the subjects were students, and adding numbers for cash is not the same as wanting somebody. But the analogy is precise about the mechanism it names, which is the entry decision considered separately from the prize. A tournament format recruits entrants by being a tournament. It does not first check whether entering is good for them. If courtship is presented as a contest—and in the New York story it is presented as nothing else—then one should expect men to enter it at high rates, to overestimate their standing in it, and for a meaningful fraction of them to be worse off for having entered.

VI · The Manufactured Rival

There is a version of this subject that blames women for staging the contest, and it should be handled carefully, because a real behavior sits underneath a false generalization. Deliberate jealousy induction exists and has been modelled. In 2005 the communication researchers Amy Fleischmann, Brian Spitzberg, Peter Andersen, and Scott Roesch analyzed it into goals and tactics. The goals divided into two: seeking relational rewards—attention, commitment, evidence of being valued—and seeking relational revenge. The tactics divided into three, and one of them is called the flirtation façade, which is the manufacture of an apparent rival who is not a real one.

That is the taxonomy. What this magazine could not find is a prevalence: no figure it could verify in an accessible primary source for how many people, of either sex, actually do this. The best-known early study on the subject was not available for checking, and no number from it is repeated here. So the honest statement is narrow. Manufacturing a rival is a recognized enough behavior to have been given a name and a factor structure by researchers; how common it is, this essay does not know, and neither does anybody quoting a percentage at a dinner party.

The more useful correction runs the other way, and it concerns what the contest is actually rewarded for. In 2008 the psychologists Jeffrey Snyder, Lee Kirkpatrick, and H. Clark Barrett ran three experiments to separate two things the literature had been mixing: dominance, meaning the use of coercion and intimidation to subordinate rivals, and prestige, meaning recognition by one’s peers for skill and achievement. The received wisdom was that women prefer dominant men. In the first study, the women preferred the high-prestige target to the high-dominance one. In the second, with the two manipulated independently, they preferred high prestige to low—and preferred low dominance to high. In the third, dominance was preferred only when displayed in an athletic competition, and only in ratings of attractiveness and short-term desirability.

These are vignettes rated by college women at two American universities, which is stated preference and not behavior, and the article claims nothing more for them. But the pattern is specific enough to be worth having. The thing the tournament story trains a man to perform—visible ascendancy over a named rival—is the narrow case that these participants rewarded least, except inside a sporting frame. What they rewarded was standing earned in front of peers. A man optimizing for the story is often optimizing for the wrong variable.

It is worth adding what the contest does to the two men, since the story tracks only the winner. A tournament with one prize converts a relationship between two people who may have nothing against each other into a zero-sum arrangement in which each is the reason the other cannot have what he wants. The display evidence suggests men will accept that arrangement readily and perform inside it. Nothing in the evidence suggests it is a good use of them.

And the position of prize is worth one further word, because the story flatters it and Wharton did not. To be the one choosing between two suitors is to have one’s own preferences converted into a public verdict, and to acquire a permanent obligation to have chosen correctly. It is a position with an audience, which is not the same as a position with power.

VII · The Cost of Declining

So what happens to the man who will not do it—who wants somebody, says so, and refuses to be measured against a rival, to escalate when a competitor appears, or to treat another man’s interest as a summons? The honest answer is that this magazine could locate no study measuring a penalty for declining to compete. Not a small penalty; none at all. The question appears not to have been asked, which is itself worth reporting, given how confidently it is answered in bars.

What can be said is where the sanction plainly does come from, and it is not from the chair the story assigns. It comes first from the narrative, which has no role for a non-competitor: a man who does not contest is not a character, he is the part of the plot that resolves early. It comes second from other men, which is what the display evidence in Part IV predicts—a repertoire built for spectators is enforced by spectators. And it comes third from the man himself, in the form of the entry preference that Niederle and Vesterlund measured, which does not need any external enforcement to operate and is expensive without it.

It also helps to notice that “refusing to compete” names at least three different acts, which the phrase merges and which carry different costs. There is declining to escalate: a rival appears, and the man does not raise his bid, contest the ground, or start behaving differently because somebody else is in the room. There is declining to display: he does not perform the resource-and-strength repertoire that Part IV found men performing for spectators, and so registers as absent from a contest he never joined. And there is declining to pursue at all, which is not a refusal of the ideal but simply an absence of interest, and which the story misreads as the first two because from outside they look identical. This taxonomy is an editorial synthesis of the preceding evidence rather than a measured distinction, but the merging it corrects is where most of the confusion on this subject lives.

The three also differ in who can see them. A man who declines to escalate does so in front of an audience that was expecting escalation, and pays immediately, in the currency of how the episode gets described afterward by other people. Whatever the market charges him, by contrast—if it charges him anything—arrives silently and unattributably, as an outcome that would have been different and cannot be inspected. The visible penalty is therefore social and the invisible one is romantic, which is a reliable recipe for a folk belief: the cost everyone can see gets credited to the cause nobody can.

The one thing that is regularly asserted—that women reliably punish the man who will not fight for them—has no located evidence behind it, and the nearest relevant evidence points mildly the other way. The dominance experiments found coercive ascendancy over rivals to be the least rewarded of the traits tested outside a sporting context. That is not proof of the opposite claim, and it should not be inflated into one. It is a reason not to make the original claim confidently.

Something should be said about the widely reported retreat from dating, because it is usually offered as the consequence of exactly this refusal, and the numbers do not cooperate. In a survey of 4,860 American adults conducted in October 2019 and published by the Pew Research Center in August 2020, half of single adults said they were not currently looking for a relationship or dates. That figure is routinely presented as a portrait of men withdrawing. In the same survey, single men were far more likely than single women to be looking: sixty-one per cent against thirty-eight. Whatever the retreat is, it is not principally a male retreat, and it is national rather than a fact about this city. Being on the market is also not the same as competing on it, and the survey does not ask about competing at all.

VIII · The Manhattan Contest

The tournament story has an arithmetic requirement that nobody checks: it needs enough rivals for a contest to feel ordinary. In the Department of City Planning’s 2017 profile, Manhattan had fewer adult men than adult women. The figures do not settle the question for people of dating age; women’s longer lives widen the all-adult gap, and this essay has no age-specific count. They do show how carelessly a familiar New York story can be mistaken for a demographic law.

More decisive than the head count is the script. The romantic contest makes two men visible, assigns them opposite qualities, and treats the woman’s decision as a verdict on which quality deserves to win. New York supplies plenty of stages for that story, from the crowded party to the dating app, but no city can turn another person’s choice into a judgment on one’s entire worth. The experiments in this essay measure choices in constrained settings. They do not instruct a man to keep entering contests.

The alternatives are quieter and more useful: decline to escalate a rivalry, decline to perform one for an audience, or decide that a particular relationship is worth pursuing on its own terms. These are editorial distinctions, not measured New York categories. The city is large enough for a person to leave one contest without leaving the possibility of connection.

IX · The Better Man

The phrase the tournament exists to produce is the better man, and the trouble with it is grammatical before it is anything else. Better is a comparative, and a comparative needs a second term. The story supplies one by putting another man in the room. What it cannot supply is the thing the comparison is supposed to establish, because whether two particular people should be together is not a quantity on which two candidates can be ranked. The contest answers a question adjacent to the real one and hands back the answer with great confidence.

The three claims, then. Being wanted by two people at once does raise a man’s rating, in photographs, by about six per cent, with heterogeneity high enough to embarrass the estimate and a standard bias correction that can halve it. Being the person who chooses is a position with an audience rather than a position with power, and New York has a hundred years of its own testimony saying so. And the expectation that a man enter the contest is enforced by the narrative, by other men, and by his own taste for entering—a taste measurable enough that at equal ability seventy-three per cent of men and thirty-five per cent of women will choose the tournament, and that some of the men lose money by it. No study locatable for this essay shows women imposing that expectation, and the nearest evidence finds them rewarding standing among peers rather than ascendancy over a rival.

Which returns us to an apartment in Manhattan in August of 2001, and to a detail the tableau usually hides. In the episode that gave this essay its opening, the two men never meet. There is no contest. There is a woman who is not sure, a man who asks, and another man who does not, and the whole apparatus of rivalry exists in the space between them rather than in anything either man does. The most famous two-suitor story New York ever told stages almost no competition at all. The audience supplied it, which is, on the evidence assembled here, roughly what the audience has always done.

Notes on Sources

Sex and the City, season four, episode twelve, “Just Say Yes,” HBO, first broadcast August 12, 2001; season four ran June 3, 2001 to February 10, 2002. Source of the opening and closing scene.† (Read in the episode’s published synopsis of record rather than a transcript; the essay paraphrases and quotes no dialogue. A television plot is evidence about a narrative form, not about behavior in the city.)

Gouda-Vossos, A., Nakagawa, S., Dixson, B. J. W., & Brooks, R. C. (2018). Mate choice copying in humans: A systematic review and meta-analysis. Adaptive Human Behavior and Physiology, 4(4), 364–386. Model i, female choice with the addition of a female cue: 17 effect sizes, lnRR = 0.058, p < .05, reported by the authors as men rated 6.01 per cent more attractive or desirable; I² = 88.38 per cent. Model ii, male choice: 7 effect sizes, −0.022, p = .377. Models iii and iv, cue augmentation: 0.113 and 0.1293. Trim and fill under the R0 estimator: three studies filled, corrected estimate 0.032 (3.3 per cent), not significant.† (The abstract reports a moderate publication bias in favor of positive results; the authors judge its impact probably modest. Photographs and vignettes, not real rivals.)

Ward McAllister’s list of “the Four Hundred,” published in the New York Times, February 16, 1892, after a competing list in the New York World; the number is attributed to the capacity of Mrs. Astor’s ballroom and the published list carried somewhat fewer names.† (Read in an encyclopedic account of the 1892 publication, not on the newspaper page.) Wharton, E. (1905). The House of Mirth; Wharton, E. (1920). The Age of Innocence. Literary evidence of how the New York marriage market was understood by someone raised inside it.

Puts, D. A. (2010). Beauty and the beast: Mechanisms of sexual selection in humans. Evolution and Human Behavior, 31(3), 157–175. Source of the contest-overrides-choice argument and of the claim that the mate-choice consensus conforms neither to theoretical predictions nor to available evidence.† (A reinterpretation of comparative, anatomical, and phylogenetic evidence, contested within its own field, and not an experiment. It licenses no behavior.)

Buss, D. M. (1988). The evolution of human intrasexual competition: Tactics of mate attraction. Journal of Personality and Social Psychology, 54(4), 616–628. Act nominations from 113 college students; Study 1 close-friend observer reports, undergraduate sample N = 104; Study 2, 214 individuals in 107 newlywed couples. Table 1 values: display resources 0.80 against 0.37, t = 3.85; brag about resources 1.03 against 0.59, t = 3.38; wear sexy clothes 0.67 against 0.71, act provocative 0.81 against 0.88, and flirt 1.89 against 1.67, none significant.† (The paper’s own abstract concedes several predictive failures and unanticipated findings; observer and self-reported frequencies in American student and newlywed samples.)

Niederle, M., & Vesterlund, L. (2007). Do women shy away from competition? Do men compete too much? The Quarterly Journal of Economics, 122(3), 1067–1101. Source of the 73 and 35 per cent tournament-entry rates, the absence of a performance difference, and the over-entry result.† (A laboratory experiment on compensation schemes with student subjects; its transfer to courtship is an analogy about the entry decision and is labelled as one in the text.)

Fleischmann, A. A., Spitzberg, B. H., Andersen, P. A., & Roesch, S. C. (2005). Tickling the monster: Jealousy induction in relationships. Journal of Social and Personal Relationships, 22(1), 49–73. Source of the two goals, the three tactics including the flirtation façade, and the three partner responses.† (A factor structure, not a prevalence. No verifiable prevalence estimate for deliberate jealousy induction was located in an accessible primary source; White, G. L. (1980), Inducing jealousy: A power perspective, Personality and Social Psychology Bulletin, 6, 222–227, is the classic reference but was not accessible for this piece, so no figure from it is quoted.)

Snyder, J. K., Kirkpatrick, L. A., & Barrett, H. C. (2008). The dominance dilemma: Do women really prefer dominant mates? Personal Relationships, 15(4), 425–444. Three experiments with college women at two US universities; source of the preference for prestige over dominance, the preference for low over high dominance when the two were manipulated independently, and the athletic-context exception limited to attractiveness and short-term ratings.† (Written vignettes measuring stated preference, not behavior.)

Pew Research Center, “Nearly Half of U.S. Adults Say Dating Has Gotten Harder for Most People in the Last 10 Years,” by Anna Brown, released August 20, 2020; survey of 4,860 U.S. adults conducted October 16–28, 2019. Source of the 50 per cent of single adults not currently looking for a relationship or dates, and of the 61-against-38-per-cent gap between single men and single women who are looking. National figures, not New York’s.

New York City Department of City Planning, table DP05, “ACS Demographic and Housing Estimates,” 2017 American Community Survey 1-Year Estimates, New York City and Boroughs. New York City: 4,112,539 men and 4,510,159 women of 8,622,698; sex ratio 91.2 per hundred women overall and 88.1 among those 18 and over. Manhattan: 90.1 overall, 88.4 among adults, 68.8 among those 65 and over. The 2017 profile is the most recent published at this address; the Census Bureau’s data interface now requires a key that was not obtained for this piece, so no newer or age-specific figure is cited.

† marks findings limited by design, sample, publication bias, or contested interpretation. The three-way distinction between declining to escalate, declining to display, and declining to pursue is an editorial synthesis, not a validated taxonomy. Figure 2 is a qualitative editorial diagram; Figures 1 and 3 carry published values. The Manhattan sex ratio is an all-adult figure and is not a young-adult ratio. No located study measures how often two men compete for one woman or what declining to compete costs.

Drafted with AI (Claude Opus 5) as a private research synthesis, September 1, 2026; edited, and answered for, by humans. Every study named above was read at its source or its published record before it was cited, the two figures carrying numbers carry only published ones, a meta-analysis’s own bias correction is reported beside its headline result, and the two questions for which no evidence could be found are named as unanswered rather than answered anyway. 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. ◆ ◆ ◆