The Shortlist

I · A Month of Opening Lines

For thirty-one days in January of 2014, in New York, Boston, Chicago, and Seattle, two researchers recorded who wrote to whom. The sociologist Elizabeth Bruch and the physicist Mark Newman had obtained the messaging records of a free online dating service, and what they built from them was a map of who was pursuing whom across four American markets at once. Their paper is chiefly about ambition: both sexes, it turns out, write to people above them, men to partners about twenty-six per cent further up the ranking than themselves and women about twenty-three. Aspiration is not a male trait. It is close to universal and almost exactly the same size in both directions.

But the finding that matters for the present argument is a plainer one, sitting in the same data. More than eighty per cent of the first messages were sent by men. Women received more messages than men overall, and were likelier than men to get a reply when they wrote. In four large American cities, in one month, on one platform, the traffic ran overwhelmingly in one direction.

So the observation underneath the thesis this essay was asked to examine is correct, and it should be conceded before anything else is said. A woman on that market ends up holding a queue. Not a metaphorical one: a literal, countable stack of unanswered approaches, arriving faster than any reasonable person could process them. Anybody who has watched this happen at close range—or lived inside it—is not imagining the asymmetry. It is one of the best-measured facts in the sociology of modern courtship.

The thesis goes on from there to two further claims, and this is where it needs testing. The first is that the queue reflects a preference: that women want the queue, seek to enlarge it, and are doing something when they hold one. The second is that this appetite resolves, in practice, into the familiar arrangement of two serious contenders and a choice between them. The first claim confuses an outcome with a strategy, and four separate literatures say so. The second is not a claim about mating at all, though it took this magazine some time to see it.

One caution before the evidence starts arriving, and it applies with unusual force here. The studies below are conducted on students, on speed-dating attendees, and on the users of dating platforms. They report averages inside those samples. A sentence beginning “women in general” is asking for something the data cannot supply, and this essay will not pretend otherwise. What the data can do is tell us whether the specific mechanism the thesis proposes is the one operating, and on that narrower question it is unusually informative.

II · Four Different Things Called Maximizing

The verb is doing too much work. To maximize suitors could mean at least four separate things, which have different evidence and different answers, and the argument turns on keeping them apart.

It could mean maximizing offers received: ending up with the largest possible number of people expressing interest. It could mean maximizing partners accepted: actually taking up more of them. It could mean maximizing options held—keeping a set of possibilities alive without acting on them, which is a portfolio rather than a queue. Or it could mean maximizing in the technical sense the decision literature uses: searching exhaustively for the best available rather than stopping at the first good enough.

Only the first is established. The Bruch and Newman data settle it, and settle it decisively. But the first is also the one that requires the least from the person it describes. An inbox is not an achievement. If one side of a market sends four-fifths of the opening messages, the other side accumulates offers whether it wants them or not; the queue is a consequence of somebody else’s sending behavior, arriving unbidden, and the recipient’s only decision is what to do with it.

The second—maximizing partners accepted—is the one for which the evidence runs most clearly the other way. In 2003 the psychologist David Schmitt, with a large international team, published results from the International Sexuality Description Project, an anonymous survey administered simultaneously in dozens of countries. Across analyses covering 16,288 people in ten major world regions, the desire for a larger number of partners was male-skewed, consistently, everywhere they looked. The interpretation of that result is argued over. The direction is not, and it is the opposite of the thesis.

The fourth sense deserves its own definition, because it is a term of art and the essay will need it later. In the decision literature a maximizer is somebody who searches for the best available option rather than stopping at the first one that clears a threshold; the contrast is with a satisficer, who sets a standard and takes the first candidate to meet it. The distinction is not about ambition. A satisficer may have an extremely high standard. What separates the two is whether the search continues after an acceptable option has been found, and it turns out to be one of the more consequential differences in how people run their lives.

The third and fourth are where the interesting questions live, and the rest of this essay is mostly about them. Holding options is a real behavior with real evidence, and it turns out to be measured in an unexpected direction. Maximizing in the technical sense is a disposition with a well-studied cost. Neither, as it happens, produces the pattern the thesis describes.

III · The Rotation

Everyone knows that women are the choosier sex, and for most of the history of this literature the finding was reported as a fact about women. In 2009 the psychologists Eli Finkel and Paul Eastwick took three hundred and fifty speed daters and altered one arbitrary detail of the format. At speed-dating events, by convention, the men move between tables and the women stay seated. Finkel and Eastwick ran fifteen events: at eight, the men rotated as usual; at seven, the women did.

Rotators were significantly less selective than sitters. The familiar pattern—men saying yes to more people than women do—appeared at the events where men rotated and disappeared at the events where women rotated. The variable was not sex. It was who had to walk across the room.

This is worth sitting with, because it reframes everything downstream. The physical act of approaching somebody appears to increase attraction to them, or at least to lower the threshold for saying yes; and in the ordinary organization of courtship, one sex does nearly all the approaching. What gets recorded as female choosiness is substantially a measurement of the seated position. Put a woman in the moving chair and she becomes, on average, the less selective party.

The authors’ proposed mechanism is worth stating because it is so ordinary. Approaching somebody is an act, and acts have a way of producing the attitudes that would have justified them; a person who has crossed a room to sit down opposite a stranger has already invested something in the encounter, and arrives at the table slightly disposed to find it worthwhile. The seated party has invested nothing and is presented with a series of arrivals to evaluate. That is not a theory about desire. It is a theory about the difference between doing something and having something done in your direction.

None of which proves that no dispositional difference exists. The study altered one norm at one kind of event and found the difference vanished; a different setting might behave differently, and the sample was student-heavy and small by modern standards. But it does establish that the difference is not a fixed property of the people. It is at least partly a property of the arrangement, and the arrangement is not something women chose.

IV · The Elasticity

Between 2002 and 2004, in a bar near the Columbia campus, the economists Raymond Fisman, Sheena Iyengar, Emir Kamenica, and Itamar Simonson ran fourteen speed-dating sessions with graduate and professional students. The design’s virtue was one detail: they varied the number of partners a participant met, without telling participants in advance how many that would be. That gave them something the field rarely has, which is exogenous variation in the size of the pool.

Two results came out of it, and the more famous one is the less interesting. The famous one concerns preferences: women weighted intelligence and ambition more heavily, men weighted physical attractiveness, and men were significantly less likely to accept a woman more ambitious than themselves. The second result is the one this essay needs.

Male selectivity, they found, was invariant to the number of potential partners. Female selectivity was strongly increasing in it. And—this is the sentence that should be quoted more often than it is—female subjects were no more selective than males in small groups. The sexes did not differ in how choosy they were. They differed in what happened to their choosiness when the pool got bigger. As the number of candidates rose, men’s acceptances rose roughly in step; women’s did not. The authors put it in the language of their discipline: women’s costs appeared more convex, or their benefits more concave, in the number of dates.

Translated out of the economics, this is a direct answer to the commissioned thesis, and it is not the expected one. Give a woman in this experiment more suitors and she does not take more of them. She raises the bar. Her list stays about the same length and its entry requirement climbs. That is not maximizing in any sense the word ordinarily carries; if anything it is the behavior of somebody who regards additional candidates as a cost to be managed rather than a stock to be accumulated.

The sample deserves saying out loud, because it is the study’s weakness and the thesis’s, equally. These were Columbia graduate students at events run twenty years ago, deciding whom to see again, in a bar, on a weeknight. It is the single most on-point finding available and it is also exactly the sort of narrow, elite, dated sample from which no honest person derives a claim about women in general. Both things are true at once, and the essay is stuck with both.

V · Who Keeps the Portfolio

There remains the third sense of maximizing, and it is the one the thesis probably means: not accepting more suitors but keeping more of them alive—warm, in reserve, available if the present arrangement fails. The behavior has a name in the research literature, which is a good sign that somebody has counted it. They are called back burners: prospective romantic or sexual partners with whom a person stays in contact against a future possibility.

In 2014 the communication researchers Jayson Dibble and Michelle Drouin asked how many of them people actually keep, and how the number relates to a person’s investment in a current relationship. Electronic back-burner activity was common; the sample averaged about five and a half; and the number correlated positively with how good a person judged their alternatives to be, which is what the investment model would predict.

The framework matters for reading the result. The investment model holds that commitment to a relationship depends not only on satisfaction and on what one has sunk into it, but on the perceived quality of the alternatives—on how good the next-best thing looks. A back burner is that variable made concrete and given a telephone number. Dibble and Drouin found the two moving together: people who judged their alternatives to be good kept more names warm, which is what the model predicts and which tells us that the practice is at least partly a way of managing a relationship one is already in.

And men reported more back burners than women.

That is one sample, using self-reported counts of electronic contact, and it should not be turned into a character sketch of anybody. But it is the only direct measurement this magazine could find of the specific behavior the thesis attributes to women, and it points the other way. If keeping a bench of alternatives warm is maximizing, the evidence available says men do more of it.

It is worth being precise about what this does and does not overturn. It does not show that women never keep options open; obviously some do, and the study found women keeping several on average too. It shows that the practice is not distinctively female, and that a phenomenon offered as an explanation of women’s behavior turns out, when counted, to be at least as much a description of men’s. The thesis has been aimed at the wrong half of the room.

VI · What Abundance Costs the Chooser

Suppose the queue exists anyway—which it does—and ask a different question: what does holding it do to the person holding it? Here the evidence is better than the folklore, and less flattering to the position than either the thesis or its critics assume.

In 2010 the psychologists Alison Lenton and Marco Francesconi assembled data from eighty-four speed-dating events—1,868 women and 1,870 men—and asked how the number of available options changed what choosers attended to. The answer was that abundance did not make people choosier in any useful sense. It made them shallower. With more options in the room, participants leaned more on characteristics that can be assessed in a glance, such as height and weight, and less on characteristics that take time to elicit, such as occupation and education. Human mate choice, the authors concluded, is bounded by cognitive resources; when the list gets long, the criteria get cheap.

Apply that to the inbox and the consequence is uncomfortable in a way that neither the thesis nor its critics tend to notice. A queue of forty is not a richer set of possibilities than a queue of four; it is a set that must be sorted with cheaper instruments. Whatever a person would have found out about the fifth message by reading it slowly—the thing that takes a paragraph to show and a minute to recognize—is exactly the sort of information that goes first when the pile grows. Abundance does not hand the chooser a better decision. It hands her a faster one.

The second finding is about the disposition rather than the situation. In 2006 the psychologists Sheena Iyengar, Rachael Wells, and Barry Schwartz followed graduating students through a job search, having first measured how strongly each tended to maximize—to search for the best available option rather than stopping at one that was good enough. The maximizers did better and felt worse. They obtained starting salaries about twenty per cent higher than the satisficers, and they were less satisfied with the jobs they took, and more miserable throughout the process. The mediating mechanisms were a heavier reliance on external sources of information about what they should want, and a fixation on the options they had not taken.

That is a job market and not a romantic one, and the transfer is an analogy rather than a finding. But it names something the thesis leaves out. If maximizing is what is happening, it is not a comfortable position occupied by somebody enjoying her leverage. It is an expensive one, and the expense falls on the person doing it.

Here the essay owes the reader a correction, because the obvious next move is to reach for the paradox of choice, and the paradox of choice does not survive its own literature. In 2010 the psychologists Benjamin Scheibehenne, Rainer Greifeneder, and Peter Todd pooled sixty-three conditions from fifty published and unpublished experiments, with a combined sample of 5,036, and found a mean effect size of virtually zero, with considerable variance between studies. Too much choice sometimes hurts and sometimes helps and mostly does neither reliably. Anyone building an argument on the jam aisle is building on sand, and this magazine declines to.

VII · Why It Stops at Two

Which leaves the second half of the thesis, and it is the half worth the most. Grant everything the first half claims. Suppose a person really is accumulating suitors, really does want the largest possible set, really is running the search exhaustively. Ask then the question the thesis assumes has an obvious answer: why would that process arrive at two?

It has no reason to. A maximizing search does not have a natural terminus at the second candidate; the whole point of maximizing is that it terminates late and reluctantly. Option value does not collapse when a third possibility appears—if anything it rises. The market is thick, the queue is long, and the arithmetic that produced the queue in the first place keeps producing it. Every mechanism the thesis invokes pushes toward more, and none of them stops at two.

What does produce two is easier to see once one stops looking for it in the market. Two is the smallest number that permits a comparison. Better requires exactly one alternative; supply a third and the sentence becomes a ranking, which is a different and much less interesting kind of statement. Two is also the number a situation must be reduced to before it can be described to a friend, and roughly the number a person can hold in mind while weighing anything against anything. The finite-attention finding from Part VI is the same observation arriving from the other side: a long list is not really held; it is triaged down to a short one, and the short one is short because attention is.

So the convergence the thesis describes is real, and its cause is not the appetite. It is the shortlist: the compression that happens between the inbox and the account a person gives of their own situation. The queue may have forty names in it. The version that can be thought about has two, because two is the largest number that fits in a comparison and the smallest that makes a decision feel like one.

This is the article’s own argument and is offered as one. No located study measures how many suitors a person holds at once, and none reports a distribution clustering at two; the absence is stated here rather than papered over with a plausible number. What can be said is that the thesis explains the wrong thing. It reaches for an account of desire when what needs explaining is an account of attention, and of the shape a person’s situation has to take before they can narrate it.

VIII · The Visible Bid

New York supplies two familiar ways to imagine selection. At an auction, the bids are visible and the price moves until all but one bidder has stopped. In an inbox, the chooser sees messages the senders cannot see together, and the grounds for a reply remain private. Neither is a model of an entire city’s courtship. They are two ways of making a crowd legible, and only one lets a participant watch the rivals.

The two-suitor story borrows the first picture. It puts two men onstage and makes the contest itself appear to determine the outcome. Yet the messaging data with which this essay opened make the second picture more useful: the modal event is an unanswered message, not a rising public bid. Abundance can narrow attention without creating a tournament anybody else can observe.

This is an analogy, not a measurement of New York dating practice. It helps separate the number of people who approach someone from the number whose possibilities that person actively considers. The city can offer a crowded market and still leave each decision private, incomplete, and resistant to the tidy arithmetic of a shortlist.

IX · The Shortlist

The thesis, then, taken in the order it was offered. Do women end up with more suitors than men? On the evidence, plainly yes, and by a wide margin: more than four-fifths of the opening messages in four American cities came from one side. Is that because they are maximizing? Nothing located for this essay supports the attribution, and four separate findings resist it—the rotation experiment, which moves selectivity from the sex to the chair; the elasticity result, in which women were no more selective than men until the pool grew and then raised the bar instead of the count; the back-burner survey, in which men kept more reserves; and the cross-national evidence on wanting more partners, which runs the other way and has for as long as anybody has asked.

What is left is more interesting than either the thesis or a flat denial of it. A structural asymmetry in who initiates deposits a queue on one side of the market. The person holding it does not appear to want it longer; she appears to respond to its length by becoming harder to reach, which is what raising a threshold looks like from outside and what managing a cost looks like from inside. And the abundance itself is not a comfortable perch: it degrades the criteria, pushing attention toward what can be judged in a glance, and where it becomes a disposition it produces better outcomes and worse feelings.

And then it arrives at two, not because the market ends there but because thinking does. Somewhere between the inbox and the sentence a person says out loud, a long list becomes a short one, and the short one is two because better needs exactly one thing to be better than.

Which returns us to that month in 2014, and to the four cities, and to the several hundred thousand opening lines that Bruch and Newman counted. Almost none of them became anything. That is the part the tableau leaves out and the data insist on: the characteristic event in this market is not a woman weighing two devoted candidates but a message that is never answered, sent by somebody aiming twenty-six per cent above himself, into a queue he will never see the length of. The two-suitor story is the rare, legible, tellable end of a distribution that is almost entirely made of silence. It is not a lie about the market. It is the market’s most flattering possible summary, and, like most flattering summaries, it describes the tail and calls it the shape.

Notes on Sources

Bruch, E. E., & Newman, M. E. J. (2018). Aspirational pursuit of mates in online dating markets. Science Advances, 4(8), eaap9815. Messaging records from a free online dating service in New York, Boston, Chicago, and Seattle, 1–31 January 2014. Source of the statement that more than eighty per cent of first messages are from men, that women receive more messages overall and are likelier to receive replies, and of the reported aspiration gaps of twenty-six per cent for men and twenty-three for women against a headline figure of about twenty-five. Desirability is a PageRank score over the messaging network.† (One platform, one month, four cities; it measures messages rather than relationships, and establishes who sends rather than who wishes to receive.)

Schmitt, D. P., et al. (2003). Universal sex differences in the desire for sexual variety: Tests from 52 nations, 6 continents, and 13 islands. Journal of Personality and Social Psychology, 85(1), 85–104. International Sexuality Description Project; analyses reported across 16,288 people in ten world regions.† (Anonymous self-report; the direction is consistent and the interpretation is contested. It concerns desired partners, not suitors received.)

Finkel, E. J., & Eastwick, P. W. (2009). Arbitrary social norms influence sex differences in romantic selectivity. Psychological Science, 20(10), 1290–1295. Three hundred and fifty speed daters across fifteen events; men rotated at eight and women at seven; rotators were significantly less selective than sitters, and the usual sex difference disappeared where women rotated.† (One event format, one arbitrary norm, a student-heavy sample. It shows the norm matters, not that nothing else does.)

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. Columbia graduate and professional students, fourteen sessions, 2002–2004, with randomized variation in group size. Source of the finding that male selectivity is invariant to the number of potential partners while female selectivity is strongly increasing in it, and of the paper’s statement that female subjects are no more selective than males in small groups; the authors read the difference as women’s costs being more convex, or benefits more concave, in the number of dates.† (An elite American student sample two decades old, measuring stated acceptances at an event.)

Dibble, J. L., & Drouin, M. (2014). Using modern technology to keep in touch with back burners: An investment model analysis. Computers in Human Behavior, 34, 96–100. Source of the definition of a back burner, of the finding that men reported more of them than women, of the sample average of about five and a half, and of the positive association with quality of alternatives.† (Self-reported counts of electronic contact in a single sample; reported here because it runs against this essay’s commissioned thesis, not because one sample settles a question.)

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 and away from slowly elicited ones as the number of options rises.

Iyengar, S. S., Wells, R. E., & Schwartz, B. (2006). Doing better but feeling worse: Looking for the “best” job undermines satisfaction. Psychological Science, 17(2), 143–150. Source of the twenty-per-cent salary advantage for maximizers, their lower satisfaction and greater negative affect, and the mediating mechanisms.† (A job search, not a courtship; the transfer is explicitly analogical.)

Scheibehenne, B., Greifeneder, R., & Todd, P. M. (2010). Can there ever be too many options? A meta-analytic review of choice overload. Journal of Consumer Research, 37(3), 409–425. Sixty-three conditions from fifty published and unpublished experiments, N = 5,036; mean effect size virtually zero with considerable between-study variance.† (A near-zero mean with high heterogeneity makes the effect unreliable rather than absent; cited against the folk version of the paradox of choice, not as proof that abundance is costless.)

† marks findings limited by design, sample, source of record, or contested interpretation. Figures 2 and 3 are qualitative editorial diagrams; Figure 1 carries published values. The four-way distinction between maximizing offers, partners, options held, and search effort is an editorial synthesis, not a validated taxonomy. Part VII is this magazine’s own argument: no located study measures how many suitors a person holds at once, and none reports a distribution clustering at two. Part VIII compares two ways of picturing selection; it does not measure New York courtship.

Drafted with AI (Claude Opus 5) as a private research synthesis, September 1, 2026; edited, and answered for, by humans. The thesis was commissioned as a claim and is examined as a question, which is the only way this magazine knows how to take one seriously: the half the evidence supports is conceded in the first part, the finding that most embarrasses the other half is reported in full, the popular paradox invoked to explain it is reported as having failed its meta-analysis, and the question for which no measurement exists is left 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. ◆ ◆ ◆