Proportional democratic weight: votes that track how long the voter lives with the result

A constitutional referendum decided this year will bind a newborn for seventy-five years and a 75-year-old for perhaps fifteen. The franchise gives the pensioner a full vote and the newborn nothing at all. Nobody ever argued for that ratio. It is an inheritance from a period when lives were shorter and most decisions resolved inside a generation.

Two changes would bring vote weight closer to stake. The first is a modest age-based multiplier that scales with how long a voter has left to live with the result. The second is a custodial vote for every child from birth, held in trust by parents and handed to the child at the franchise age. One principle runs through both. The weight of a vote should track how long the voter lives with the outcome.

A country that will not invest in its own future

This is a pattern across decades, not a single decision. Britain has repeatedly chosen not to spend on the people who will live here longest.

Consider what a political system oriented towards its own future would prioritise, and compare it to what this one does. At the start of 2026, approximately one million people aged 16 to 24 in the UK were not in education, employment or training.^[1] The figure had not passed a million since 2013, and the trend has been upward since the pandemic. A cohort is being detached from the labour market at the point of entry. No party has made a credible answer to that a serious part of its offer.

Housing tells the same story over a longer horizon. The median home in England now costs around 7.6 times median full-time earnings, against roughly 4 times on the same measure in the late 1990s.^[2] The binding constraint for younger buyers is the deposit, inflated by rents that consume the income those buyers might otherwise save. Meanwhile the stock of outright ownership has grown. Outright ownership rose from 30.8% of households in England and Wales in 2011 to 32.8% in 2021, roughly 900,000 more households, and around three-fifths of outright owners are households headed by someone aged 65 or over.^[3] The asset base stays concentrated in the generation that already holds it while the one beneath is locked out. Housing policy has not shifted this in decades. The median voter already owns a home, and is not trying to buy one.

University funding shows the intergenerational transfer at its starkest. A graduate in England now leaves with average debt of between around £47,700 and £53,000 depending on which loan plan they entered under, against a tuition cap of £9,790 a year for 2026/27 that rises to £10,050 in 2027/28.^[4] The political generation that designed and voted for this regime was, overwhelmingly, educated for free. Those who received a public good voted to charge their successors for it. The next disruption is already visible. Artificial intelligence appears to be cutting into the number of entry-level jobs, the rung the young most depend on. Its regulation remains light and contested. The evidence is early and so far American. Payroll data there shows a 13% relative employment decline for 22- to 25-year-olds in the most AI-exposed occupations, against no measurable displacement across the economy as a whole.^[5] The people most exposed to that shift have the least say over how it is regulated.

These are policy choices, not accidents, and they share a cause. A system competing for the median voter has less and less reason to spend on people who vote in smaller numbers, vote later, or cannot vote yet. Parties do not fight over the interests of the young because there are not enough young voters to make it worth the effort.

The loop is demographic, and it tightens on itself. As an electorate ages, the median voter's interests shift from investment in future capacity towards consumption of accumulated entitlement. Pensions, healthcare, and asset protection rise in priority; schools, childcare, job creation, and housing supply fall. UK fiscal allocation already shows the pattern. The state pension triple lock is treated as politically untouchable while child benefit has been frozen and cut in real terms, per-pupil school funding fell 9% in real terms between 2009-10 and 2019-20^[6], and housing supply has been constrained in part by older homeowners protecting asset values through the planning system. Children gain most from long-term investment and carry no electoral weight at all, while the age groups that gain from cutting it hold a growing share. The effect compounds as the population ages.

Short-termism is unfair to the young, and it is also a structural disadvantage for the country. China's rise is usually credited to sustained infrastructure build-out, to industrial strategy held across decades, and to long-run investment in energy and skills. All of it needs planning horizons that electoral democracies struggle to match, and a democracy working to the next election has little pulling against its own impatience. Giving more weight to voters with longer to live moves policy towards the patient investments a short electoral cycle starves: education, housing, research, and national capability. The proposal is fairer, and it would make the country better at the things that take time.

Brexit is the most-discussed single instance of the same mechanism. In 2016 the UK voted to leave the European Union by 52% to 48%, a margin of 1.3 million votes. Voters aged 18 to 24 backed Remain by roughly 73%; those over 65 backed Leave by roughly 60%.^[7] By the end of 2025 the National Bureau of Economic Research's decade-on study put the realised cost at 6 to 8% of gross domestic product, while the Office for Budget Responsibility's forecast assumes a 4% reduction in long-run productivity, about a third of which has so far shown up in the data.^[8] A permanent, hard-to-reverse decision was settled by the cohort with the shortest remaining stake in it, over the objection of the cohort that would live with it longest. By Jun 2026, 57% of Britons judged the decision wrong against 30% who maintained it was right, and on demographic estimates the original majority has since reversed through mortality and replacement alone.^[9] The outcome stands regardless, because nothing in the design lets a changed public mind reopen it.

Weighting by remaining stake

The proposal is a continuous age-based multiplier applied to vote counts in national elections and referenda, anchored at 1.0x for the median adult lifespan cohort and scaling modestly upward for younger voters and downward for older ones.

The pivot sits at the 45 to 54 band, centred on age 50. UK adult life expectancy from age 18 is a further 63 years or so, taking women and men together, which puts the midpoint of adult life at around 50.^[10] Bands adjust by 0.1x per decade either side, and the maximum spread runs from 1.3x to 0.7x.

The multipliers are deliberately small. Turnout rises steeply with age. At the 2024 general election, on Great Britain figures, it ran at 37% among 18- to 24-year-olds against 73% among the over-65s, so the oldest band supplied just under a third of the votes cast while making up under a quarter of the adult population.^[11]

The table models the age weighting alone and excludes custodial votes. It runs from 18 because it uses United Kingdom general election data, where the franchise starts there. The two oldest bands are shown combined at a population-weighted 0.75, because turnout is not published separately for them, and shares are rescaled to sum to 100, which is why the pivot band moves slightly despite its weight of 1.0.

Adding the gaps across the six bands, ignoring their direction, the votes cast differ from the adult population by about 20 percentage points in total. The schedule cuts that to about 5, on these estimates and against a population baseline; against an eligible electorate the correction would be smaller.^[12] A quarter of the gap remains. Mirroring the population exactly would need 1.5x at the youngest band and 1.35x at the next. The schedule under-corrects on purpose. The aim is to move electoral weight towards the population it is drawn from.

The effect on a general election result would rarely be decisive. It matters most on constitutional votes, which are hard to reverse and bind the young for longest.

A vote for every child, held in trust

Age weighting corrects the imbalance among adults who can already vote. It does nothing for children, who will live with the result longest and carry no electoral weight at all.

The proposal is that every child carries a vote from birth, exercised on their behalf by their parents until they reach the franchise age. The vote is the child's own, held in trust until they can cast it themselves.

Each child's vote splits in two, half to each parent or legal guardian. Where parents agree, the two halves are cast together and the child's full weight is expressed. Where parents disagree, the halves cancel towards neutral, which is the correct outcome when a household genuinely differs about the child's interest. For looked-after children, the corporate parent casts on the child's behalf. The resident parent or guardian is the default in cases of dispute or sole custody.

The weight transfers at the franchise age, which is 16 for Scottish Parliament, Senedd and Scottish and Welsh local elections, and 18 for a United Kingdom general election. On the child's 16th or 18th birthday, according to the election, the custodial weight passes from the parents to the young citizen, who now exercises it directly at the 1.3x band. While the child is a minor their vote is worth 1.0, expressed as two custodial halves of 0.5. On majority they enter the highest-horizon band and vote at 1.3x in their own right.

The step from 1.0 to 1.3 is deliberate. A child's stake is expressed through two custodians; a young adult's is expressed directly, in the band with the longest remaining horizon. The effect is to sharpen the incentive for politicians to compete for young citizens at the point those citizens gain their own voice.

Take two parents in the 35 to 44 band with three children. Each parent holds their own vote of 1.1 plus three custodial halves of 0.5, totalling 2.6. The household casts 5.2 weighted votes across five people, of which 3.0 represents the three children's own stake held in trust. As individuals the parents hold their band weight of 1.1 each and nothing more; the rest is their children's representation, carried until those children can carry it themselves.

A newborn will live with a decision longer than anyone and has no voice at all. Extending representation below the franchise age completes the principle the rest of the design rests on.

What the law says now

Plural voting was abolished by the Representation of the People Act 1948, and the United Kingdom has taken on obligations since that bear on the franchise. Article 25 of the International Covenant on Civil and Political Rights guarantees universal and equal suffrage, and the Human Rights Committee has read that as requiring one person, one vote. Article 3 of Protocol 1 to the European Convention on Human Rights guarantees free elections, and is given domestic effect by the Human Rights Act. The Venice Commission lists equal voting power among the principles of Europe's electoral heritage.^[13]

Parliament can legislate this. It abolished plural voting by statute and can revisit what it settled. Doing so would put the question in front of the courts, and what that involves is set out below.

Objections

It creates two-tier citizenship. The same rule applies to everyone. The weight varies by age, which every citizen moves through, and by nothing else. A voter can look up their own weight and see what it will be at every later stage of their life.

Why age, and not sex or anything else? Life expectancy differs between women and men, so a scheme that tracked remaining years precisely would weight them differently. This one does not. It uses a single averaged figure and applies on age alone, because the aim is to track how long a person lives with a decision without sorting people by anything else about them. Age is already the only line the franchise draws, at 18, and everybody crosses it.

It breaches equal suffrage. It may not. The rule applies identically to every citizen, and a voter weighted at 1.3 at twenty is weighted at 0.7 at eighty on the schedule everyone else is on. On that reading equal suffrage protects universality and non-exclusion, not identical weight at a single moment, and no court or treaty body appears to have ruled on weighting by remaining stake. If that reading fails, the proposal is a change to the settlement rather than an application of it. A weighted franchise would be litigated, a declaration of incompatibility under the Human Rights Act is a realistic outcome, and the prisoner voting cases show such a standoff running more than a decade before settling narrowly.^[14] Anyone pursuing this needs that timeline in front of them at the outset. Voters already old at commencement carry the lower weights without ever having held the higher ones. That is a cost of moving between arrangements, and it falls away as the first cohorts pass through.

The formula is arbitrary. The pivot comes from life expectancy data and moves with it. The formula should be set in statute, with revision tied to the decennial Office for National Statistics life tables, so the calibration is not left to political discretion.

Parents cannot be trusted to vote in a child's interest. Some will not. The franchise already tolerates that at the other end of life, where a person does not lose the vote at the first sign of cognitive decline. It operates on categories rather than case-by-case fitness tests, and accepts that any workable universal system contains some imperfectly exercised votes.

The ballot is secret, so nobody can check how a proxy was used. True, and it is the one real difference between this and every other authority parents hold. Medical consent, schooling, and residence are all supervised, reviewable, and governed by a best-interests standard. A proxy vote cannot be. The design accepts that, on the ground that an unsupervised vote for a child is better than none, and that supervision is impossible under a secret ballot whatever the scheme.

It rewards having children. Nobody takes on the cost of raising a child to gain half a vote for eighteen years. If anyone did, the country has a birth rate below replacement and would be better for it.

It is an attack on older voters. The child's weight is not added to any adult's franchise. Each parent's own vote stays at their band weight, and the extra they carry is the child's, surrendered at majority. The adults carrying it are also the ones caring for their own ageing parents, and they see both ends of the picture.

Wisdom comes with age, and discounting it is a mistake. Experience is valuable and the design does not dispute it. Experience and stake are different things, and only the second is being adjusted. An informed 75-year-old is not a worse citizen than a distracted 25-year-old. The older voter has less of their life remaining inside the consequences of today's structural choices, and that alone sets the difference in weight.

It is a slippery slope away from equal votes. Two features hold it in place. Every citizen passes through every band, so no permanent class of advantaged voters can form. The weighting also makes no judgement of worth, only of remaining stake.

Referenda are the wrong place to start. One specific obstacle sits there. The Northern Ireland Act 1998 requires that a border poll be decided by a majority of those voting in it. On the face of that provision a weighted poll would not meet the test, so applying the scheme to referenda would need it amended.^[15] The case for starting there anyway is that constitutional decisions are the ones hardest to reverse and longest lived, which is exactly where the mismatch between who decides and who lives with it matters most.

It has been proposed before and never adopted. It has. Parental proxy voting is known as Demeny voting, and it was defeated in the German Bundestag in 2003 and 2008 and dropped from Hungary's 2011 constitutional process. No country has implemented it.^[16] That record is part of the diagnosis, not a refutation of the proposal. Proposals to give the young more weight do not fail on their merits so much as on the absence of anyone with an electoral reason to carry them.

The same result could come without changing the vote. It could. If parties began competing seriously for young voters, and building policy on a horizon longer than a parliament, the outcome this proposal aims at would arrive without any change to the franchise. Raising turnout, lowering the voting age to 16, sunset clauses on irreversible decisions, and a commissioner for future generations would each move in that direction. None of them changes the underlying arithmetic, which is that a party has little electoral reason to act for people who cannot yet vote. A serious argument about that arithmetic would be worth having whether or not it ends in a weighted franchise.

It is politically impossible. It would be difficult. The franchise was extended to men without property, then to women, then to 18-year-olds, and each extension met this objection first. Any party whose coalition skews young, or which wants to build one, has a reason to take it up.

The pathway

• Commission an independent actuarial review to formalise the age-weighting multipliers against Office for National Statistics life expectancy data.
• Define the custodial allocation rules in statute, including defaults for sole custody, separated parents, and looked-after children.
• Introduce the reform as a manifesto commitment or Private Member's Bill by a party with, or seeking, a young electoral coalition.
• Apply it initially to constitutional referenda, where the consequence-horizon imbalance is most acute and most defensible.
• Extend it to general elections following a review period.
• Tie all formula and threshold revisions to decennial Office for National Statistics updates, removing ongoing political discretion.

The principle underneath

The current system gives every adult identical weight and every child none, however long each of them lives with the result. For routine policy that can be undone later, that works well enough. For decisions that cannot be undone, it hands the outcome to the people with least at stake.

Vote weight should track how long a voter lives with the result. Age weighting applies that principle among adults, and custodial representation extends it to children. Nobody is disenfranchised, and no class, region, or affiliation gains an advantage.

A 75-year-old counts as 1 and a newborn as 0, on a decision one will live with for fifteen years and the other for seventy-five. Nobody has defended those two numbers on their merits.