There are places in England where politics has not so much failed as been quietly abandoned. Not through apathy - that word is too passive for what is actually happening. The people interviewed outside polling stations in two recent northern by-elections were not indifferent. They were decided. They had looked at the available choices and concluded that none of them were worth the walk.
What came through in those interviews was not ignorance or disengagement in any simple sense. It was an active judgement that the political process was not for them - that it was a performance staged by people who did not know them, had never lived near them, and would not be affected by whatever they decided.
Into that space, two things had moved. One was simple withdrawal: people who had voted all their lives quietly stopped. The other was Reform. Not because its policies had been carefully weighed, but because it offered something the mainstream parties had stopped providing: a simple account of why things had gone wrong and who was responsible for it. The immigration narrative, the hotels, the sense of being deprioritised - these were not primarily factual claims. They were moral ordering.
This is the environment that any serious democratic renewal project has to reckon with.
Why standard digital democracy thinking misses this
Most proposals for direct digital democracy assume three things that do not hold here. They assume people are willing to deliberate. They assume people trust the process enough to engage with it. And they assume people believe that voice precedes power - that if they speak, something will change.
None of those assumptions survive contact with the voters described above. More participation mechanisms look, to someone who has already concluded that participation is theatre, like more theatre. A well-designed app with a deliberative forum is still an app. It still asks people to invest time and trust in a system that has repeatedly returned nothing.
The Taiwan model, the citizens' assembly model, the deliberative polling model - all of them have evidence behind them. But all of them worked in conditions where at least residual legitimacy existed, or where a legitimacy crisis had already broken open and created demand for something new. They are not tools for reaching people who have pre-emptively decided the whole exercise is a waste of time.
What has actually cut through
The evidence from comparable situations is instructive. Participatory budgeting worked in Porto Alegre and parts of Europe where communities controlled real money and could see the results in their streets. When it became symbolic - when the amounts were too small or the decisions too constrained - engagement collapsed. Agency came before trust, not after.
In one of the two constituencies mentioned, a party that had barely registered before managed to win. It did so not through a superior digital platform or a more compelling policy programme, but through sustained local presence, visible commitment to specific local grievances, and - crucially - the perception that a vote might actually change something tangible. Whether that result holds or proves replicable is an open question. But the conditions that produced it are worth examining: local, concrete, short feedback loop between action and outcome.
That is not an argument against direct digital democracy. It is an argument about sequencing and design.
What DDD would need to do differently
If direct digital democracy is to reach the people most alienated from politics, it cannot begin with deliberation. It has to begin with control over something real and bounded - a budget line, a local decision, something where the distance between participation and outcome is short enough to be felt.
It has to accept that non-participation is a legitimate response, not a failure to be corrected. A mature model does not penalise abstention or treat it as civic delinquency. It produces outcomes that remain legitimate even to those who chose not to engage.
It has to work offline first. Digital tools can extend reach once legitimacy has been established. Leading with an app in communities where the smartphone is not the problem signals, again, that this is something designed by people who do not live there.
And it has to be honest about what it is offering. Not a revolution. Not a solution to everything that has gone wrong. A mechanism - imperfect, improvable - for giving people more control over bounded decisions that affect their lives. If that sounds modest, it is. But modesty is more credible than promise in places where promise has been broken repeatedly.
The voters who have walked away from politics are not waiting to be inspired. They are waiting to be shown that something is different this time. That is a harder thing to demonstrate than it is to claim.