Why the First Option Gets Picked

Why We Act · 12 October 2026

Ballots, wine tastings and menus often show the first option winning more than its share. The edge is real when people know little and choose fast, and it fades or flips when the list is long or the chooser is expert.

Key takeaways

In New York City's 1998 Democratic primaries, Koppell and Steen found candidates won a larger share when listed first in 71 of 79 contests.
California's randomised ballot lottery showed major-party candidates gaining one to three points from first place in primaries, and no detectable effect in general elections.
Dayan and Bar-Hillel's 2011 menu study found items at the start or end of a list up to twice as popular as the same items in the middle, though only one category passed the test for chance.

In a list or a sequence, the first option usually starts with an edge it did not earn. The edge is largest when people know little and choose quickly, and it can shrink or reverse when experts judge options one by one. Before trusting a choice made from a list, find out who set the order.

Check whether the ballot order was fixed

Elections give the cleanest evidence, because some places rotate candidates' names between precincts, which turns each contest into a natural experiment. Jonathan Koppell and Jennifer Steen used the 1998 Democratic primaries in New York City, where names rotated in this way. In 71 of 79 contests, candidates won a higher share of the vote when listed first. In seven of those contests the advantage of the first option on the ballot was larger than the winner's margin.

Joanne Miller and Jon Krosnick had found a similar pattern in the 1992 elections in three large Ohio counties. Across 118 races, name order made a significant difference to vote totals in nearly half, and in almost every such case the candidate listed first gained. The effect was largest where voters had least to go on, in nonpartisan races with no party label to guide them and in contests with little press coverage.

The strongest test comes from California, which sets ballot order by a randomised alphabet lottery. Daniel Ho and Kosuke Imai analysed statewide elections there from 1978 to 2002 and found a more limited effect. In general elections, ballot order helped only minor-party candidates, with no detectable effect for the major parties. In primaries, being listed first helped everyone: major-party candidates generally gained one to three percentage points, and minor-party candidates could double their vote share. Across all elections, the largest gains, about three points, came in nonpartisan races.

Between them, the three studies agree on where the first option wins. It wins when the voter arrives without a reason to prefer anyone, and it fades when a party label supplies one.

It wins when the voter arrives without a reason to prefer anyone, and it fades when a party label supplies one.

Test the first glass against the last one

A ballot is usually seen all at once. Many choices arrive one at a time instead, and there the record is mixed. In 2009 Antonia Mantonakis and colleagues reported in Psychological Science a large tasting experiment in which people sampled two, three, four or five wines in sequence and then chose the one they liked best. The first wine had a large advantage. Participants who knew a lot about wine behaved differently in the longer sequences, with a pull towards the last glass, which psychologists call a recency effect.

Dana Carney and Mahzarin Banaji found the first option favoured even when the gap between options was a matter of seconds. In three experiments published in 2012 under the title "First is best", people showed an implicit preference for the first salesperson they met and the first team they encountered, and were more likely to choose the first of two consumer items shown in quick succession. The preference held even when the first option was a criminal. The authors note that earlier research on order had produced inconsistent results.

Some of that inconsistency comes from judgements made over a long run. Wändi Bruine de Bruin studied the Eurovision Song Contest and found that ratings rose with each later slot in the running order, whether juries scored at the end or after each song. The same rising pattern appeared in figure skating. When the list is long and every entry is scored, the last performer tends to do better than the first.

Read the first word as setting a direction

The oldest experiment on first impressions explains why order can matter even when nothing is forgotten. In 1946 Solomon Asch read two groups the same six traits about an imaginary person: intelligent, industrious, impulsive, critical, stubborn and envious, or the same list reversed. The group that heard intelligent first was far more likely to pick favourable words to describe him, choosing humorous 52 per cent of the time against 21 per cent for the reversed list, and generous 24 per cent against 10.

Asch's account was that "the first terms set up in most subjects a direction which then exerts a continuous effect on the latter terms." He was explicit that this was not a memory effect of the kind Ebbinghaus studied: "It is not the sheer temporal position of the item which is important as much as the functional relation of its content to the content of the items following it."

The Choice Factory, Richard Shotton's 2018 book of 25 biases for marketers, uses Asch as the anchor for its tenth chapter, Primacy Effect, at pages 71 to 75. The book asks whether a 1946 result about people still holds for products, and tests it. Shotton told 500 consumers about an invented brand, Black Sheep Vodka, listing the same attributes in opposite orders, and those who heard the strongest attributes first rated the vodka 11 per cent higher. The test has not been published, so its method is known only from the book's account of it.

The chapter offers two explanations and draws different advice from each. If primacy reflects limited attention, a brand should lead with its strongest argument, cut secondary messages and buy the first slot in an advertising break. If Asch's account is right, the first attribute sets a direction that colours the rest, so a brand should lead with the attributes it can most easily own and spend heavily at launch, because first impressions are hard to overturn. It also recommends invented brands for tests like this, because real ones carry feelings that contaminate the result.

Look at both ends of the list

The menu evidence complicates the advice to always go first. In 2011 Eran Dayan and Maya Bar-Hillel moved items around restaurant menus, first in a laboratory study with 240 students and then in a Tel Aviv café, where they counted 951 orders over 15 days. Items placed at the beginning or the end of their category were up to twice as popular as the same items placed in the middle. The pull was towards the edges, and the first option shared it with the last.

The café result is less tidy than the headline. Only 20 of roughly 60 items changed position, within three categories: alcoholic coffees, soft drinks and desserts. Taken one category at a time, only the soft drinks showed a difference large enough to rule out chance, and orders placed without looking at the menu were not recorded.

On this evidence, the first option has a real advantage when choices are quick, when people know little about the options, and when nothing else marks one as better. That is consistent with Asch's account, though low attention would explain it too, and it fits the book's case for leading with the attribute a brand can own. The advantage shrinks or reverses when options are judged one by one over a long sequence, when the chooser is an expert, or when the list is long enough for its end to stand out.

Against that, the clean evidence comes mostly from settings where little is at stake, such as ballots for minor offices and drinks orders, so the effect may be smallest where decisions matter most. We accept that, and still think it worth correcting for, because most everyday choices are made with little information.

The same pull operates in other places. The first option on a form often doubles as the default, and the first price you see sets the frame for everything after it. A few habits follow from the research, none of them tested as a set:

Check whether the order was random, rotated or chosen by someone who benefits from it.
Count the options. With two or three, expect the first to be favoured; with a long run judged one at a time, expect the last.
Ask what you would pick if the list ran backwards, and look again at whatever sits in the middle.
Write down what you want before you open the menu, the ballot or the comparison page.
If you run the list, test the order by rotating it and measure what changes.

Most lists are ordered by habit, by the alphabet or by whoever built the page, and only occasionally by someone who gains from the result. The research does not say which of those set the list in front of you, so the habit worth keeping is to treat position as information about the list rather than about the options. Give the middle of the list a second look.

Common questions

Why does the first option get picked more often?

Solomon Asch argued in 1946 that the first item sets a direction that colours everything after it. Carney and Banaji's 2012 experiments found people preferred the first option even when it came seconds earlier. The effect is strongest when choices are quick and people know little about the options.

Does being listed first on a ballot win elections?

Sometimes. In New York City's 1998 Democratic primaries, Koppell and Steen found candidates did better when listed first in 71 of 79 contests. In California, Ho and Imai found major-party candidates gained one to three points in primaries but showed no detectable effect in general elections.

Sources

Koppell, J. G. S., & Steen, J. A. (2004). “The effects of ballot position on election outcomes.” The Journal of Politics, 66(1), 267-281. doi:10.1046/j.1468-2508.2004.00151.x. Abstract read (Montclair State University repository): 1998 New York City Democratic primary, names rotated by precinct; 71 of 79 contests showed a higher share when listed first; in 7 the advantage exceeded the winner's margin.

Miller, J. M., & Krosnick, J. A. (1998). “The impact of candidate name order on election outcomes.” Public Opinion Quarterly, 62(3), 291-330. doi:10.1086/297848. Institutional summary read (Ohio State University news release); the journal abstract and full text were not accessible: 1992 elections in Franklin, Cuyahoga and Hamilton counties, 118 races, name order significant in 48 per cent of races, first-listed candidate advantaged in nearly all cases, effects largest in nonpartisan and low-coverage races.

Ho, D. E., & Imai, K. (2008). “Estimating causal effects of ballot order from a randomized natural experiment: the California alphabet lottery, 1978-2002.” Public Opinion Quarterly, 72(2), 216-240. doi:10.1093/poq/nfn018. Abstract read: in general elections only minor party candidates affected; in primaries major party candidates gain one to three percentage points and minor party candidates may double their vote shares; nonpartisan races show about three points.

Mantonakis, A., Rodero, P., Lesschaeve, I., & Hastie, R. (2009). “Order in choice: effects of serial position on preferences.” Psychological Science, 20(11), 1309-1312. doi:10.1111/j.1467-9280.2009.02453.x. Abstract read: choice sets of two to five wines; large primacy effect; knowledgeable participants showed a recency effect in longer sequences.

Carney, D. R., & Banaji, M. R. (2012). “First is best.” PLoS ONE, 7(6), e35088. doi:10.1371/journal.pone.0035088. Abstract read: three experiments; implicit preference for first salesperson and first team; first of two consumer items more likely chosen; held for an evaluatively negative target (a criminal); notes previous inconsistencies in the empirical record.

Bruine de Bruin, W. (2005). “Save the last dance for me: unwanted serial position effects in jury evaluations.” Acta Psychologica, 118(3), 245-260. doi:10.1016/j.actpsy.2004.08.005. Abstract read: Eurovision Song Contest ratings increased with serial position under end-of-sequence and step-by-step procedures; replicated in World and European Figure Skating Contests.

Asch, S. E. (1946). “Forming impressions of personality.” Journal of Abnormal and Social Psychology, 41(3), 258-290. doi:10.1037/h0055756. Full text read: Experiment VI, Series A intelligent to envious (N = 34) and Series B reversed (N = 24); Table 7 check-list: generous 24 against 10 per cent, humorous 52 against 21 per cent; both quotations verbatim from pp. 271 to 272.

Dayan, E., & Bar-Hillel, M. (2011). “Nudge to nobesity II: menu positions influence food orders.” Judgment and Decision Making, 6(4), 333-342. doi:10.1017/S1930297500001947. Full text read: Study 1, 240 students, hypothetical choices; Study 2, Tel Aviv café, 951 orders over 15 days, 20 of about 60 items reordered in three categories; items at beginning or end up to twice as popular as in the centre; only soft drinks significant at category level.

Shotton, R. (2018). The Choice Factory: 25 behavioural biases that influence what we buy. Harriman House. Read in full by this publication, with a reliability audit. Bias 10, Primacy Effect, pp. 71 to 75. The Black Sheep Vodka test (p. 73, 500 consumers, 11 per cent higher rating) is unpublished.