The Swiss vote roughly four times a year. Between 1995 and 2005, they trudged to the polls (or posted their ballots) 31 times to answer 103 federal questions—never mind the cantonal and municipal referenda piling up alongside. Turnout, however, now languishes around 47 per cent, among the worst in Western Europe. When offered more chances to participate, why do citizens increasingly shrug their shoulders?
Governments worldwide are falling over themselves to ‘reinvent’ democracy through participation. Taiwan has algorithmic consensus-building. Ireland convenes citizens’ assemblies. Paris does participatory budgeting. Estonia lets you vote from your laptop. The pitch always sounds marvellous—democracy 2.0 will heal fractured politics, restore institutional trust, transform passive subjects into engaged citizens. But what if this is mostly an expensive way to avoid the harder task of governing competently?
Taiwan’s clever algorithm
Taiwan’s vTaiwan platform, launched in 2015, deploys Polis software to find consensus in divided communities. The logic is elegant: instead of letting online forums devolve into shouting matches, Polis visualises where agreement exists and surfaces statements that bridge divides. When Uber tried muscling into Taiwan’s taxi market in 2015, vTaiwan corralled 4,000 participants into crafting regulations that kept both taxi drivers and tech evangelists mostly happy. The government duly adopted the recommendations.
Sounds marvellous, except vTaiwan’s shine wore off rather quickly. The platform’s complexity and tech-first approach put people off; simpler alternatives like Join grabbed market share. Covid-19 killed the face-to-face meetings that made the whole thing work, and participation fell. Turns out digital democracy demands sustained volunteer effort, technical literacy, and bottomless enthusiasm—commodities governments overestimate.
vTaiwan has since retreated to safer ground: AI ethics discussions and collaborations with outfits like OpenAI. Policy recommendations remain non-binding, naturally. The platform now thrives on abstract technology policy rather than prosaic governance—a tacit admission that citizen engagement works best when sufficiently removed from anything concrete.
The Irish exception
Ireland offers a brighter story, though it cuts both ways. The 2016-17 Citizens’ Assembly—99 randomly selected citizens—spent five sessions deliberating on abortion, hearing from experts and wading through submissions. When a landmark abortion referendum came in 2018, 66.4 per cent voted to repeal the Eighth Amendment and allow the interruption of pregnancies. Remarkable in a historically Catholic country.
Even more remarkable is that the Citizens’ Assembly had recommended that 64 per cent favoured ‘terminations without restrictions’, nearly identical to the referendum result. So much for the notion that informed deliberation inflames rather than resolves divisions. Politicians who’d spent decades dodging the abortion question suddenly had cover to act.
But there’s a catch. In 2024, Irish voters delivered the government its largest-ever referendum defeat—67 per cent and 74 per cent rejecting proposals on family roles—after politicians watered down what the Citizens’ Assembly recommended. The lesson writes itself. Citizens’ assemblies work brilliantly when governments implement what citizens recommend. Treat them as expensive focus groups to be selectively mined for convenient bits, and you breed cynicism faster than traditional politics ever could.
Porto Alegre’s rise and fall
Porto Alegre in Brazil pioneered participatory budgeting in 1989, letting residents allocate roughly half the municipal budget through neighbourhood assemblies. The early results looked transformative. Between 1988 and 1997, sewer and water connections jumped from 75 to 98 per cent of households. New public housing units sheltered 1,700 people in 1986; by 1989 that figure was 27,000. Schools more than quadrupled. Participation exploded from 1,000 people in 1990 to 40,000 by 1999. Impressive.
Then politics did what politics does. After four Workers’ Party terms, the opposition took over and suddenly had other priorities. Funds allocated through citizen assemblies dwindled. Porto Alegre eventually suspended the programme entirely—the world’s most celebrated experiment in direct democracy withering in its birthplace. Political commitment proved essential; when it vanished, so did the programme.
Participatory budgeting has since spread to over 11,500 municipalities worldwide, though results vary wildly. Continuity depends on political will and resources; when either runs dry, the programmes get cancelled or neutered. There’s also participation fatigue—people who show up to get their street paved lose interest once the tarmac arrives. Democratic renewal can quickly devolve into competitive lobbying for local amenities. Sewage systems instead of tax breaks, but interest-group politics all the same.
The Swiss paradox
Switzerland runs the world’s most extensive experiment in citizen decision-making. Yet in October 2023’s federal election, only 47 per cent of eligible voters bothered—down from 81 per cent in 1933. On average, roughly a quarter of Swiss always vote, another quarter never do, and half pick and choose based on the issue.
Trust is eroding fast. For the first time, Swiss who distrust their government outnumber those who trust it, following some embarrassing calculation errors in pension forecasts. The government lost three major votes in 2024—motorways, pensions, more pensions. Direct democracy, it turns out, doesn’t prevent government cock-ups. It just provides more opportunities to punish them.
Some scholars reckon the Swiss are so satisfied they don’t feel the need to vote—a counterintuitive finding that challenges assumptions about participation. Perhaps maximal democracy breeds complacency rather than engagement. Or perhaps voting four times yearly on everything from motorway expansion to tenancy law simply exhausts people’s capacity for informed decision-making. Democratic overdose.
Estonia’s digital frontier
In March 2023, Estonia became the first country where over half the voters cast ballots online in a national election. The system, running since 2005, builds on mandatory digital ID cards that handle secure authentication and legally binding digital signatures. Estonia has built an internet-voting system widely trusted and used—nearly half of all votes now arrive electronically.
But security worries persist. International experts who poked around Estonia’s system in 2015 found “staggering gaps in procedural and operational security” and warned the system was wide open to cyberattacks that “could alter votes or leave election outcomes in dispute”. By 2024, distrust in e-voting hit 40 per cent following messy 2023 elections and unsubstantiated fraud claims from some political actors. The internet-voting process has become politicised—one party that discouraged online voting (and predictably trailed in the online count) challenged the whole thing in court. Sound familiar?
Studies show internet voting hasn’t boosted turnout where voting was already easy, nor has it created digital divides. It’s just a more convenient delivery mechanism for democracy. Useful, perhaps. Transformative? Hardly. Making voting easier doesn’t address why citizens tune out in the first place.
The listening problem
Direct democracy mechanisms work when governments commit to implementing what emerges (Ireland’s abortion referendum), when political will sustains the institutions (Porto Alegre’s early years), and when citizens trust the system stays secure (Taiwan’s early vTaiwan). They fail when politicians water down recommendations (Ireland’s 2024 disaster), when political priorities shift (Porto Alegre’s collapse), or when security and fairness get questioned (Estonia’s polarisation).
The real problem isn’t insufficient participation channels. It’s insufficient responsiveness to participation that already occurs. Representative democracy’s crisis isn’t that citizens lack chances to make their voices heard. It’s that when they speak—through elections, protests, surveys, traditional lobbying—governments don’t listen, or listen only to those with money and connections.
Even successful participatory budgeting in Porto Alegre lacked mechanisms for getting citizens, especially the poor, into long-term city planning. The initiatives worked brilliantly for small-scale, community decisions—where to build a park, which streets to pave. But they couldn’t give citizens meaningful say over strategic choices: urban development, fiscal policy, economic planning.
Participation theatre
Creating more participation mechanisms without fixing underlying responsiveness produces what might be called ‘participation theatre’—elaborate consultative processes that look like citizen power whilst leaving real decision-making untouched. A World Bank study of 253 Brazilian municipalities found those with participatory budgeting collect 39 per cent more local taxes than similar places without it. So these mechanisms work partly by making citizens more willing to fund government, not necessarily by making government more accountable.
Both Ireland’s Constitutional Convention and Citizens’ Assembly called for direct democracy mechanisms like citizens’ initiatives—69-83 per cent support. Nothing, however, happened. Governments love consultative processes they control but like far less the binding mechanisms that transfer real power.
The fashionable reinvention of democracy through participation risks becoming expensive distraction. It costs money to run citizens’ assemblies, maintain digital platforms, organise participatory budgets. It demands time from citizens already stretched thin. And it often produces recommendations that governments water down or ignore.
True democratic reinvention would be simpler and harder: governments that implement what citizens clearly want, respond swiftly to obvious problems, and deliver competent administration. Swiss political scientist Martina Mousson reckons broader alliances behind projects increase the odds of reform success—in other words, old-fashioned coalition-building and compromise.
The Swiss vote themselves hoarse and turnout falls. Taiwan builds clever consensus algorithms and engagement wanes. Porto Alegre shows what’s possible, then dismantles it. Ireland proves citizens’ assemblies work—when politicians follow through.
Maybe the lesson isn’t that democracy needs reinvention through participation. Maybe democracy needs governments capable of listening to the participation that already exists.
The real question isn’t how to get citizens participating more. It’s why governments consistently ignore the clear signals citizens already send. Adding more channels whilst ignoring those signals isn’t reinvention. It’s just more noise.
Photo: Dreamstime.
