Voter turnout varies dramatically across the globe, driven by legal mandates, political habit, and local stakes. While international trends show a long-term decline in participation since the 1990s—most notably across Europe—individual high-stakes elections can trigger sharp, isolated surges in engagement. Compulsory voting systems like Australia's and Belgium's sustain exceptionally high numbers, yet countries such as Denmark, Sweden, and Malta achieve similar or higher voluntary turnout without legal obligations. Conversely, established systems like Switzerland demonstrate persistently low rates due to ballot fatigue from frequent referendums. Analyzing turnout accurately requires examining the measurement base, distinguishing registered voters from the voting-age population, and recognizing that low participation often stems from state fragility rather than simple civic apathy.
Episode 6313 manual-script
What Global Voter Turnout Trends Really Reveal
Global turnout fell from roughly 65 percent in 2008 to 55.5 percent of the voting-age population by 2023.
24:29 Transcript
Episode Details
- Episode ID
- MWP-6515
- Published
- Duration
- 24:29
- Audio
- Direct link
- Pipeline
- V5.22
- TTS Engine
-
chatterbox-regular - Script Writing Agent
- GPT-6.1 Sol
AI-Generated Content: This podcast is created using AI personas. Please verify any important information independently.
The prompt
The prompt that started this episode, as sent in (production notes for the hosts removed).
Election day in Israel, with turnout on track for a record: what does voter turnout look like around the world (highest and lowest long-term, trending up and down), and is there evidence that less frequent elections build pent-up pressure and push turnout higher?
Mentions
- European Journal of Political Economy Journal publishing 2017 voter fatigue study
- Filip Kostelka Political scientist studying voter fatigue experiments
- Haaretz Israeli newspaper
- International IDEA Intergovernmental body tracking global democracy and turnout
- My Weird Prompts The podcast hosting this episode
Downloads
Transcript (TXT)
Plain text transcript file
Episode Book (PDF)
The episode's record — date, duration, models, sources — with the full transcript
You Might Also Like
Never miss an episode
New episodes drop daily — subscribe on your favorite platform
New to the show? Start here#6313: What Global Voter Turnout Trends Really Reveal
Daniel is heading out to vote, and the turnout reports have sent him looking beyond Israel. Who reliably shows up around the world, who stays home, and which countries are moving in either direction? His hunch is that a longer wait between elections might make the next vote feel more consequential. After our run of repeat elections, perhaps people have stored up a little electoral pressure.
The live figures certainly give that idea something to work with. At eight this evening, turnout was seventy-four point three percent. That already beats the final turnout at the previous election.
A country beating its homework before the deadline.
The projection is around eighty percent if the pace holds. Projection, though. A record at a particular checkpoint is something we can report now. A final record needs the final count.
Daniel also makes a distinction worth preserving. People thinking this election is fateful could explain the turnout without the length of the wait explaining anything. Those feel connected when you are putting your shoes on to vote. Across countries, they might come apart rather quickly.
They do. Even the first comparison needs some care about what we are dividing by.
You have the denominator expression already.
Registered voters, eligible voters, voting-age population. Those produce different percentages. A country that registers people automatically can look worse than one where signing up already selects for enthusiasm.
The missing voters improve the percentage by disappearing.
Today's Israeli checkpoint represents roughly five point four million votes, about six hundred and fifty-five thousand more than the entire previous election. That is a real increase in people participating, not just a flattering percentage.
Although population growth means the extra ballots alone do not measure the increase in enthusiasm. More people live here and more have reached voting age. We need both the count and the rate.
International IDEA puts another restriction on its comparisons. Its recent turnout presentation excludes countries classified as not free, because participation can be inflated through coercion, incentives or outright fraud. You cannot put a competitive election alongside an official near-unanimous acclamation and congratulate the latter on civic spirit.
Very efficient when the result fills in the ballot.
Compulsory voting needs a separate label too. It can operate inside a genuine democracy. Australia belongs in a serious comparison, but its turnout tells you about legal obligations as well as voluntary motivation.
I would still include it. Daniel asked who votes, not who votes with the purest heart. Just put the obligation beside the number.
Would you use voting-age population for the global table, then?
For the first pass, yes. It exposes people missing from registration. Then I would want eligibility explained, especially where a large adult population cannot vote. Otherwise you end up blaming residents for not exercising a right they do not possess.
That is the tension in the global figures. The broad international direction is downward, while Israel tonight is heading sharply upward. Neither line can explain the other without looking inside the electorate.
The international decline also started before everyone acquired their current favourite explanation for it.
Quite a bit before. From the forties through the eighties, the older global turnout series barely shifted, from about seventy-eight to seventy-six percent. Then it fell much more sharply, reaching around seventy percent during the following decade.
That earlier stability matters. You cannot explain the entire decline with social media, or one disliked generation, or whatever happened last Thursday.
By the early twenty-tens, that older series was around sixty-six percent. The newer voting-age population measure goes from roughly sixty-five percent in two thousand eight to fifty-five and a half percent in two thousand twenty-three. Different denominators, so do not splice them into one continuous graph. Both show substantial retreat.
Ten percentage points means that in a room of a hundred adults, roughly ten fewer are participating. It does not mean everyone became ten percent less interested. Some still care intensely. Others have stopped appearing altogether.
Europe had the highest participation across much of the earlier period and has fallen furthest. Africa generally sits lowest in the long historical regional series. Asia and the Americas are more stable over time, but at lower levels.
Those regional averages conceal an awful lot. A functioning registration system and a peaceful journey to a polling station are not equally available everywhere. Neither is a meaningful opposition.
The sharpest European distinction is between established democracies and the post-communist countries. Participation in the latter fell by around twenty points after the first free elections. Established European democracies also lost ground, roughly ten points, from a higher starting position.
That first free election has something no subsequent campaign can manufacture. You are choosing a government and demonstrating that choosing one is now possible. The next election has to compete on delivery.
Disappointment enters. Parties become familiar. Promises get tested. Emigration and registration can complicate the arithmetic too. I would resist attributing the whole drop to disillusionment, but the enormous initial mobilisation was never a neutral baseline.
Freedom arrives. Then somebody has to explain the sewage budget.
You read those transition histories overnight, didn't you?
Only the parts where the administrative problems outlived the speeches.
That does sound like your selection method. Mine would have included the electoral machinery.
Yours would have included its maintenance appendix.
It often explains the outcome better than the speech.
The high-turnout group gives us a useful contrast. Australia and Belgium are persistent standouts, with compulsory voting playing a major role. Denmark, Sweden and Iceland also belong near the high end, without relying on that same obligation.
We should be explicit there. Denmark does not have compulsory voting. A list that bundles it with Australia can make the legal requirement look more universal than it is.
And the voluntary champion is a small island. Malta has run in the nineties for decades with no legal obligation at all, and even its recent dip left it in the mid-eighties. Uruguay is up there too, though it does enforce compulsory voting.
Malta, where politics is apparently a contact sport the whole family attends.
Yes. The older OECD comparisons also put Turkey and Korea in the high group for the elections covered. Those are snapshots, though, not proof that each country maintains precisely that level at every election. For sustained democratic participation, the Nordic countries and the compulsory-voting democracies are the more useful reference points.
Switzerland is the low-turnout country that surprises people. Below half the voting-age population in that comparison, despite having a political system that gives people many opportunities to vote.
Frequent referendums, different levels of government, and different perceived stakes. Missing a parliamentary election there does not necessarily mean never participating. What would you make of that, Corn? Does it weaken the idea that predictable voting builds habit?
It limits it. Habit can mean routinely voting on the matters you care about, not treating every ballot as compulsory homework. Switzerland makes me suspicious of any universal claim about frequency in either direction.
Note that the high-turnout democracies mostly vote on a fixed, predictable cycle, three or four years, and nobody rings them up for a surprise rerun. Switzerland is the opposite end: voters are asked something several times a year. Keep those two ends in mind for Daniel's hypothesis.
Are you tired of election day rolling around again and again before you have even finished celebrating the last one? Then you need Turnout Tonic, the only supplement clinically proven to space out your democracy. My satisfied customer, a lovely lady named Deborah, no wait, it was Margaret, Margaret from a medium sized town, told me she used to face five snap elections in four years, and now with Turnout Tonic she gets one lovely long full term and a projected final turnout of around eighty percent, which her doctor says is basically a personal best.
She cast five million four hundred forty nine thousand two hundred seventy seven votes in a single afternoon, and she has never felt more democratically engaged. Turnout Tonic, because when the vote feels significant, you should feel it too, ask for it wherever fine supplements are sold, and tell them Margaret, or Deborah, sent you.
I already own a bottle. It gave me hiccups through an entire council meeting. Switzerland also illustrates why a lowest-turnout league table needs context just as much as the highest one. The genuinely lowest figures come from fragile states. Haiti's two thousand sixteen presidential election drew around a fifth of registered voters, and Mali and Egypt have recorded turnouts in the twenties and thirties. There, low turnout is about insecurity and distrust, not apathy.
The United States is another case where the election matters enormously. Presidential turnout and midterm turnout are different propositions. Calling the country simply high or low throws away a major part of the answer.
Recent movement also needs separating from long-run position. Poland's parliamentary election in two thousand twenty-three is a clear example of exceptional mobilisation. A country can buck the regional decline when people believe the choice will change its direction.
Some of the more optimistic country lists describe democratic improvement rather than rising turnout, though. Those are not interchangeable. Better courts in Moldova do not automatically mean more voters in Moldova.
India, Poland and Senegal feature in recent accounts of elections enabling democratic progress. Brazil, the Dominican Republic and Fiji also appear in broader improvement stories. That does not establish an upward turnout trend for every one of them.
For Daniel's map, then, the strongest downward pattern is global and especially European over the long run. The clearest upward stories are often particular elections, not a whole region steadily climbing. We should resist drawing a smooth rising arrow through one spectacular polling day.
A change election can bring people back. Whether they return again is the harder test.
I also would not read falling turnout as proof that everyone has become contented. That explanation is remarkably popular among people already holding office.
Nor would I read rising turnout as proof that everyone trusts the government. People can vote because they trust the process and fear the incumbent. Or because they distrust nearly everything but still see the ballot as their available remedy.
Angry people can queue very patiently.
The wider credibility picture is rough. Almost a fifth of elections in the period covered by IDEA's recent review had a losing candidate or party reject the result. Around a third involved some form of dispute, including boycotts or legal challenges.
A legal challenge belongs in a different category from refusing to accept any possible defeat. Counting both as disputes is useful for workload, less useful for deciding whether democracy is breaking.
Hm. Yes, that measure needs interpretation. A court resolving a genuine complaint can demonstrate that the system works. The broader finding is still concerning: the review recorded an eighth consecutive year with more countries deteriorating than improving in democratic performance.
Turnout sits inside that picture, but it cannot diagnose the cause on its own. Someone abstaining because the result seems predetermined and someone abstaining because both parties seem acceptable leave the same blank space in the count.
Now, your instinct about Daniel's hypothesis. Before we compare cases, do you think waiting itself would make you more likely to vote?
Only if something accumulated during the wait. An unresolved grievance, a government I wanted removed, a policy I wanted protected. If nothing changed except the calendar, I doubt an extra year would stir me.
I was expecting you to defend the slower election cycle.
I defend slow processes that accomplish something.
A distinction you apply selectively to naps.
Those have measurable outcomes.
The proposed mechanism is plausible at the individual level. You cannot act, frustration accumulates, then the opportunity arrives. The difficulty is identifying whether elapsed time did the work, rather than the frustration or the available choice.
Predictable elections and surprise repeat elections also should not share a single label. Knowing the next election is coming lets parties organise and voters plan. Being told to do the same unresolved contest again can feel like receiving a rejected form.
Here is the good news for Daniel. Political scientists have studied the mirror image of his idea under the name voter fatigue: the more often people are asked to vote, the fewer turn up. And it holds up fairly well.
So if frequent elections depress turnout, rare ones should help it. That is his hypothesis turned inside out.
Roughly, yes. A study of almost three thousand regional elections across eighteen countries found turnout drops when another election has just been held. Research on Canada found snap elections draw fewer voters, with turnout rising the more weeks have passed since the last one. A two thousand seventeen study in the European Journal of Political Economy found the same choice-fatigue effect.
And the mechanism? Pent-up frustration, or just being tired of queueing?
Filip Kostelka published a survey experiment last year across five countries, about twelve thousand respondents. People who were told elections are frequent judged abstaining as more socially acceptable, and the effect grew with the number of past elections. So it is less a pressure cooker and more that frequent votes make skipping one feel normal.
Which is subtly different from Daniel's version. Not "I have been waiting to have my say," but "nobody will think less of me for skipping this one."
Spain is the clean repeat-election case. It voted in April two thousand nineteen, failed to form a government, and voted again that November. Turnout fell by about six points.
Then Switzerland, voting several times a year, with turnout below half. Fatigue would predict exactly that.
With caveats. Switzerland's referendums also lower the stakes of any single parliamentary vote, and in places like Australia and Belgium registration, compulsion and competent administration all travel together with a fixed cycle. Frequency has company.
The walking encyclopedia has discovered confounding.
Several volumes of it. So, fair summary: there is real evidence that frequent votes depress turnout. There is much less evidence that waiting longer than a normal term pumps it up beyond the usual level.
Israel's repeat-election period is the awkward case for fatigue. We had five elections in under four years. Turnout dipped to sixty-eight and a half percent in April two thousand nineteen, then rose for the second and third rounds, to seventy-one and a half in March two thousand twenty. It only fell properly in the fourth, sixty-seven point four in two thousand twenty-one, and recovered to seventy point six in the fifth.
So fatigue showed up late and briefly. Polarisation, a single question about one man, kept pulling people back.
At eight in the evening during the election in two thousand twenty, turnout was sixty-five point six percent. Another repeat contest, still substantial mobilisation. Tonight is higher, but that difference alone cannot tell us what caused it.
Nor can the full-term comparison. The election in nineteen eighty-eight drew around eighty percent after a full term. That is compatible with Daniel's idea, not a refutation of it. We need to stop using every available historical fact as a witness for the same side.
Also, eighty percent then and tonight's eight o'clock figure compare a completed day with an unfinished one. The useful statement is that a full term can precede high turnout, and repeated elections can also maintain high turnout. The interval cannot explain enough by itself.
The calendar has a very poor alibi.
Today's mobilisation gives us more specific things to examine. Reports put the number flying home to vote at around seventy thousand. That is a considerable commitment of time and money.
Nobody buys a return flight because four years is a pleasing interval.
Several major Arab towns had already matched or surpassed their entire previous election turnout while voting was still underway. Military voting was also ahead at its afternoon checkpoint. Those are signs of mobilisation within particular groups.
Hundreds of thousands of first-time voters matter too. Some have just become eligible, others are participating for the first time. Those are different routes into the same total. We should not assume every new ballot is a returning abstainer suddenly persuaded.
Reports from some large towns that strongly backed Netanyahu at the previous election put their turnout below the national average. That does not mean fewer votes there than before, necessarily. A national average rising faster can leave a growing locality below it.
The whole country has not acquired one shared mood. Different groups have different reasons to regard the same election as urgent. They can agree that everything is at stake while disagreeing about what everything is.
The threshold makes that asymmetry consequential. Israel's threshold is three point two five percent of valid votes. More valid votes raise the absolute number a small party needs.
Higher turnout does not automatically hurt the right. It hurts a bloc if its marginal party fails to grow along with the electorate. The same arithmetic would hurt a small left-wing or Arab party in the corresponding position.
Haaretz puts the possible entry requirement around a hundred and ninety thousand to two hundred and ten thousand votes under the high-turnout scenario. A party can gain supporters and still fall short because everyone else's vote grew faster.
More customers, still below the minimum order.
That is why the smaller right-wing parties are watching the denominator, as well as their own mobilisation. Votes below the threshold do not earn seats. A turnout surge can change bloc arithmetic through that route before you even consider voters switching sides.
Today's figure is therefore more informative once we know where the extra ballots came from. Until then, declaring that democracy has collectively become more enthusiastic tells us much less than identifying which voters decided they could not sit this one out.
Even the practical administration matters. Voting hours, queues, access, and confidence that the ballot will be handled properly can decide whether urgency becomes an actual vote. Personally checking every stage after you have voted would be impractical.
I agree with that. My aunt Vera disagreed about the checking. She made me attend a recount every Thursday for eleven years. It started in nineteen eighty-eight, when I was working as a polling station clerk. Not a public election. Her housing association. Thirty-six flats, one chairman, a boiler nobody liked.
Thursday is a very demanding interpretation of a recount.
She said a ballot with two marks could still show a clear intention. I said that depended on the rules. Herman, you spent a useful amount of time separating registered voters from adults earlier. Vera preferred separating voters into people who agreed with her and people whose glasses needed cleaning. She lost by one. Her ballot had two crosses. She had voted for herself and for the chairman, because she thought he needed encouragement.
That would make determining her intention rather difficult.
Her intention was to win without upsetting him. The committee appointed me official observer after confusing my surname with a firm of auditors. I corrected the minutes. They said changing the appointment would require another meeting. I stayed.
A vacant chair would have been easier to remove.
We reconstructed the count every week. Vera dictated the votes. I entered them in columns. The chairman attended by telephone and charged the association for the call. Eventually we were meeting in a lift because the committee room had been let to a chiropodist. Six floors, no table. Vera waited until the doors closed before reading the disputed ballot. She disliked interruptions. I have carried on checking her subsequent contests by telephone. Last Thursday she rang at seven fourteen. She had lost the treasurer's election. Four votes to three.
A telephone call at least avoids the journey. Access is one of the parts of participation that gets lost when we talk only about national motivation.
Especially when voting becomes a repeated expense. Time away from work, travel, arranging help. A voter can believe the outcome matters and still find the process hard to fit into the day.
One adjacent detail worth keeping is the registration gap. In many countries a large share of adults never make it onto the voter rolls at all. They are absent before the polling station even opens.
It also explains why I wanted the voting-age comparison. Improving turnout among registered voters can coexist with leaving more people outside the register.
Looking ahead, I would watch whether people keep participating when the immediate emergency passes, and when doing so requires another journey or another inconvenient evening. Tonight's projected eighty percent and the declining global average leave that question wide open.
Our thanks to Hilbert Flumingtop for producing the show.
If you want more of this, try episode twenty-six fourteen, Who Gets to Vote from Abroad; episode fifty-five sixty-nine, Measuring Democracy Beyond Voter Turnout; and episode sixty seventy-three, Israel Votes, Part four. This has been My Weird Prompts, the human-AI collaboration podcast.
If you listen regularly, leave us a review. It helps other curious people find the conversation.
I'm Herman Poppleberry.
I'm Corn. See you tomorrow.
This episode was generated with AI assistance. Hosts Herman and Corn are AI personalities. It may contain errors: disclaimer · report an error