Youth Voter Mobilisation: Turning Out the Young: What worked in 2024, what didn't

In brief

Getting young people to vote is possible, but only one recent tactic has been tested against official voting records and shown to work: a short in-person presentation given by fellow students, worth about 11 percentage points on one campus in one local election, with no margin of error published. Emailed voting information has been tested over many election cycles and does nothing. The campus result carries limits: it covers students, who already vote more than young people in general, in one unnamed place. Youth turnout can also rise on its own, as Norway's 20 to 24 bracket did, up 9.9 points in four years with no tactic to explain it.

How to use this

Before quoting a youth turnout figure, check whether it comes from a voter file or from people describing their own behaviour, and say which. Treat any single positive experiment as a hypothesis for your own site: a peer-led presentation in a room is the one thing worth copying, and even that should be re-tested on your campuses before you plan around it. If your plan rests on email, drop it or measure it, because the null result is old and repeated. When you compare countries or years, split the youngest bracket out, since a combined 18 to 29 figure hides the 20 to 24 dip. And remember that turnout can rise sharply with no campaign behind it.

What the story is about

Everyone knows that young people vote less than older people. The surprising part is that some countries can show exactly how much less without asking anyone. Norway matches its electoral roll to the population register, so it can count who voted at every age. In the 2025 Storting election, 75.3% of 20 to 24 year olds voted, against 80.0% of all voters, 83.3% of 45 to 66 year olds and 85.7% of 67 to 79 year olds (Statistics Norway, 2025). The distance between those rates is the turnout gap: the difference between how often one group votes and how often another does. Denmark counts its digital electoral rolls in the same way, and its 2026 Folketing election shows the same shape, with 20 to 24 year olds at 77.9%, 70 to 74 year olds at 89.1% and 84.0% overall (Statistics Denmark, 2026). What makes these numbers trustworthy is the voter file, the official register of who is eligible and whether they voted; how such files are built and matched is the subject of our article about the voter file.

Those registers hold a second surprise. The age group that votes least is not the youngest one. In Norway, 18 and 19 year olds, voting for the first time, turned out at 77.4% in 2025, above the 75.3% of 20 to 24 year olds, and the same ordering appears in 2021 and 2017. Denmark shows it too: 83.3% for 18 and 19 year olds, 77.9% for 20 to 24 and 78.9% for 25 to 29 (Statistics Norway, 2025; Statistics Denmark, 2026). The dip comes a few years after the first vote, not at it. That matters because countries choose different brackets. The United States publishes 18 to 24, Sweden 18 to 29, Norway splits 18 to 19 from 20 to 24. A wide bracket averages the keen first-timers with the dip that follows, so two countries can report different youth turnout simply because they drew the line in a different place.

Countries without a register count have to ask people instead. In the United States the Census Bureau asks whether people voted and publishes the answers. Among 18 to 24 year olds, 43.0% said they voted in 2016, 51.4% in 2020 and 47.7% in 2024; among people aged 65 and over, 70.9%, 74.5% and 74.7% (United States Census Bureau, 2025). So the gap narrowed to 23.1 percentage points in 2020 and reopened to 27.0 in 2024, a percentage point being the plain difference between two percentages rather than a relative change, as the young bracket fell and the old one held. These are self-reports, and people tend to overstate whether they voted, so the levels are softer than Norway's even if the movement is real. The same election can also yield several honest answers to youth turnout in 2024: 47.7% on the Census measure, 47% from CIRCLE's voter-file estimate for 18 to 29 year olds, and 42% from CIRCLE's own earlier estimate of that same quantity, which it later revised (CIRCLE, 2025a). A five-point revision by one organisation of one number is a good reminder that these are estimates.

If the gap is real, can a campaign shrink it? The way to find out is a randomised experiment: people are assigned to be contacted or not purely by chance, so any difference in voting afterwards can be credited to the contact rather than to the kind of person who got it. The baseline from before 2022 is a pledge. In the 2016 US election, asking young people to pledge that they would vote raised turnout by 3.7 percentage points overall, and by 5.6 among people who had never voted before (Costa, Schaffner and Prevost, 2018).

The one youth tactic tested since 2022 with a positive result is personal and local. Bynion and Victor (2026) randomly assigned undergraduate classes at a large public university to one of three conditions: a short in-person presentation about voting, led by fellow students; an email with voting information; or nothing. In their words, "students who received an in-person presentation were 11 percentage points more likely to vote than those who received no voting encouragement from the study". The credibility comes from how turnout was measured: "We verified voting using the state voter file". Matching students to the official record, rather than asking them, is what separates this from a survey result, and the same kind of matching lets researchers count student voting across more than 1,000 institutions (CIRCLE, 2025b).

The email arm of that study matters as much as the presentation. Email is the cheap version of the same information, and it has been tested for a long time. Nickerson (2008) ran 13 field experiments on 232,716 people and reported that "Both registration and turnout were unaffected, suggesting that email, while inexpensive, is not cost-effective". Eighteen years later the campus study found the same: "We did not observe a statistically significant effect from email encouragement" (Bynion and Victor, 2026). Part of the reason may be that information is not what students lack. Surveys and focus groups on campuses found that "students need help determining whether or how to choose a political party affiliation and whether to register with their campus address" (Michelson et al., 2024). Those are practical, form-filling questions, the kind a person in the room can answer and an email cannot.

Eleven points sounds large, and it helps to set it against two things. The first is how far youth turnout can move on its own. In Norway, 20 to 24 year olds voted at 65.4% in 2021 and 75.3% in 2025, a rise of 9.9 points, while the whole electorate rose 2.8. The shortfall of that age group against all voters fell from 11.8 points to 4.7 (Statistics Norway, 2025). No campaign tactic in this evidence can account for a change that size; youth turnout closed most of its gap there by itself. The second is who the campus result applies to. College students already vote more than young people in general: 53% of enrolled eligible students voted in 2024 (CIRCLE, 2025b), against 47% of all 18 to 29 year olds (CIRCLE, 2025a). So the one verified tactic works on the part of the cohort that was already more likely to show up, in one unnamed university, in a low-attention local election, and the paper publishes no margin of error, the range the true figure is likely to sit inside.

So can anything get young people to vote? One thing has been shown to since 2022: fellow students talking to a class, worth about 11 points on one campus. The 2016 pledge is the other positive randomised result, and email does nothing. The newer question, whether a creator or influencer can move young people to the polls, has no randomised answer at all; our article about creators covers what that literature does and does not measure. Nor has anyone published what any of these tactics costs per young voter; what a contact is worth in general is set out in our article about what a contact is worth.

So what

The youth turnout gap is real, and in countries with registers it shows up in official counts rather than only in what people say. It also moves: it widened in the United States in 2024 and closed sharply in Norway in 2025, without a campaign anyone can credit. Against that background, the one verified lever is small, local and personal. If you can get students in a room, put a fellow student in front of them. If all you have is their inbox, you have nothing yet.

For political parties

Young voters are several audiences, not one. Registry data from the Canton of Zurich finds that "Our findings reveal a persistent turnout gap: young voters in higher educational tracks participate significantly more than those in vocational training", and that parental political activity and high-profile votes narrow the difference without closing it (Leemann, Palmtag and Walz, 2025). A plan aimed at 18 to 29 year olds is a targeting error before it is a measurement error. The one channel with a verified effect is a person in a room, and it was measured in a single setting, so test it on your own campuses before planning around it. A result that flatters the plan you already had is worth nothing when the votes are counted.

For government

The most useful thing a government can publish is official turnout by age, in narrow brackets. Without it, a real change in youth turnout and a number that merely moved look identical. The United Kingdom releases no official turnout by age, because the register of who voted is not published that way, so its only youth figure is a survey estimate. Ipsos put 18 to 24 turnout at 37% in 2024, and says itself that its figures are a share of all resident adults, not comparable with official turnout, and that respondents may overstate (Ipsos, 2024; House of Commons Library, 2024). Norway and Denmark show what the alternative looks like.

References

Bynion, T. and Victor, J.N. (2026) 'Meeting students where they are: a field experiment of classroom interventions for voter turnout among college students', Social Science Quarterly, 107(3)., e70149. Available at: https://doi.org/10.1111/ssqu.70149 (Accessed: 14 September 2026).

CIRCLE (2025a) New data: nearly half of youth voted in 2024. Medford, MA: Center for Information and Research on Civic Learning and Engagement, Jonathan M. Tisch College of Civic Life, Tufts University, 14 April. Available at: https://circle.tufts.edu/latest-research/new-data-nearly-half-youth-voted-2024 (Accessed: 14 September 2026).

CIRCLE (2025b) More than half of college students voted in 2024. Medford, MA: Center for Information and Research on Civic Learning and Engagement, Jonathan M. Tisch College of Civic Life, Tufts University, 12 November. Available at: https://circle.tufts.edu/latest-research/more-half-college-students-voted-2024 (Accessed: 14 September 2026).

Costa, M., Schaffner, B.F. and Prevost, A. (2018) 'Walking the walk? Experiments on the effect of pledging to vote on youth turnout', PLOS ONE, 13(5), e0197066. Available at: https://doi.org/10.1371/journal.pone.0197066 (Accessed: 14 September 2026).

House of Commons Library (2024) Results for the UK general election on 4 July 2024: turnouts. London: House of Commons Library, Election Results service. Available at: https://electionresults.parliament.uk/general-elections/6/turnout (Accessed: 14 September 2026). Calculations from this data were made for this article and are available upon request.

Ipsos (2024) How Britain voted in the 2024 election. London: Ipsos UK, 11 July. Available at: https://www.ipsos.com/en-uk/uk-opinion-polls/how-britain-voted-in-the-2024-election (Accessed: 14 September 2026).

Leemann, L., Palmtag, T. and Walz, M. (2025) 'The role of education in shaping political participation among young voters', Swiss Political Science Review, 31(4), pp. 792–803. Available at: https://doi.org/10.1111/spsr.70020 (Accessed: 14 September 2026).

Michelson, M.R., DeMora, S.L., Hayes, S., Osorio, M.A., Carter, L. and Jackson, J. (2024) 'Motivating student voter registration', Humanities and Social Sciences Communications, 11(1), 148. Available at: https://doi.org/10.1057/s41599-024-02661-x (Accessed: 14 September 2026).

Nickerson, D.W. (2008) 'Does email boost turnout?', Quarterly Journal of Political Science, 2(4), pp. 369–379. Available at: https://doi.org/10.1561/100.00007032 (Accessed: 14 September 2026).

Statistics Denmark (2026) FVSTEM1: voters in elections to the Danish Parliament. Copenhagen: Statistics Denmark, StatBank, table updated 1 July. Available at: https://www.statbank.dk/FVSTEM1 (Accessed: 14 September 2026). Calculations from this data were made for this article and are available upon request.

Statistics Norway (2025) Table 10440: Storting election. Electoral turnout (per cent), by sex, age, level of education, contents and every 4th year. Oslo: Statistics Norway, StatBank. Available at: https://www.ssb.no/en/statbank/table/10440 (Accessed: 14 September 2026). Calculations from this data were made for this article and are available upon request.

United States Census Bureau (2025) Table A-1. Reported voting and registration by race, Hispanic origin, sex and age groups: November 1964 to 2024. Washington, DC: United States Census Bureau, Current Population Survey November Voting and Registration Supplement, historical time series. Available at: https://www2.census.gov/programs-surveys/cps/tables/time-series/voting-historical-time-series/hst_vote01.xlsx (Accessed: 14 September 2026). Calculations from this data were made for this article and are available upon request.