Episode Summary
Executive Summary: Rob Wiblin argues that voting can have meaningful expected value in close, high-stakes elections, especially where policy differences are large and the voter is in a tipping-point district or state. He weighs this against the risks of being misinformed and the opportunity cost of research, and also discusses campaign donations and get-out-the-vote efforts as ways to multiply impact.
Main Topics: Probability that one vote changes an election: Using statistical modeling and polling, the episode estimates the odds of a single vote affecting outcomes in national elections, emphasizing that the chance is tiny but not zero and rises in close, competitive districts or states. Value of winning and policy stakes: The discussion shows that even small chances of affecting elections can have large expected value because governments control huge budgets and make decisions on war, regulation, immigration, and other high-impact issues. Objections: epistemic modesty and voter error: Rob considers the argument that if elections are close, voters should distrust their own judgment. He responds that average voters are often uninformed, and a moderately informed altruistic voter may still be better than most. Time cost and opportunity cost of voting: The episode argues that voting itself is cheap, but the research needed to vote well can be substantial, so for some people the same time may be better spent on high-impact work or earning to give. Campaign donations and get-out-the-vote efforts: Beyond voting, the episode evaluates how much it costs to generate additional votes through personal persuasion, canvassing, phone banking, and campaign spending, noting that these can be impactful but vary by context. When voting is altruistically worthwhile: The conclusion identifies conditions where voting is likely to be a strong altruistic action: important issues, clear candidate differences, and a competitive race in a pivotal region.
Key Arguments: A vote matters most when an election is close nationally and locally, especially in a tipping-point state or seat. Statistical modeling is necessary because real elections rarely produce enough exact ties to estimate vote impact from history alone. The expected value of a vote can be large because small chances of influencing government decisions are multiplied by the scale of public spending and policy consequences. Safe states/seats offer near-zero leverage, while marginal regions can raise the probability of decisive influence by orders of magnitude. The difference between candidates can be huge on issues like war, aid, immigration, regulation, and climate policy, increasing the value of being decisive. Epistemic modesty reduces confidence in one’s vote, but ordinary voters are often less informed than a motivated reader who has researched the issues. Even if voting is valuable, the research required to vote well may be too costly for people whose time could create more impact elsewhere. Campaign donations and direct voter outreach can sometimes buy votes or turnout for a relatively modest cost, though estimates are uncertain and context-dependent.
Data Points: Florida 2000 margin: 537 votes - Enough to change the US presidential outcome from Bush to Gore Florida 2000 margin percentage: 0.009% - Margin in Florida in the 2000 election New Mexico 2000 margin: 366 votes - Even closer state margin in the 2000 election Chance vote flips US presidential outcome in a swing state: about 1 in 5 million - Approximate estimate discussed for close US presidential elections Chance vote flips outcome in a close tipping-point state: between 1 in 1 million and 1 in 10 million - Range for competitive elections in competitive districts/states US federal government spending per 4 years: $17.6 trillion - Used to illustrate the scale of government decisions Expected spending influence at 1 in 10 million odds: $1.76 million - Expected value of influencing a presidential election outcome Annual government tax/spend share in rich countries: 25% to 55% of GDP - Rough proportion controlled by governments US federal spending per person over 4 years: $53,000 - Derived from total spending US federal spending per vote in 2000: $129 - Derived from federal spending and number of votes cast Foreign aid example expected value: $17,000 - If policy differences are 0.5% vs 0.3% of GDP and vote odds are 1 in 10 million Nuclear war example value: $25,000 - Expected value to citizens from a vote affecting nuclear-war risk US citizen willingness to pay to avoid nuclear war: $1 million - Illustrative assumption used in the example Chance of nuclear war under one candidate: 1 in 1,000 - Illustrative estimate for one president over four years Chance of nuclear war under another candidate: 1 in 500 - Illustrative estimate for another president over four years Nuclear war mortality: 80% of the US population - Used to show catastrophic stakes Campaign cost per vote via canvassing/phone banking: $30 to $100 - Typical estimate from get-out-the-vote studies for effective methods 2020 paper TV ad cost per vote: $100 to $1,000 - Suggested as a more typical recent estimate Biden and Trump allied group funding per voter in tipping states: about $30 per voter - Potential funding level in the 2020 general election Biden campaign primary funding per voter: $3 per voter - Through Super Tuesday in 2020 primaries UK and Australia marginal-seat funding per voter: about $10 per voter - Compared with US general-election funding levels Public knowledge example: House control: 38% - Americans who knew Republicans controlled the House before the 2014 election Public knowledge example: Senate control: 38% - Americans who knew Democrats controlled the Senate before the 2014 election Public knowledge example: federal spending priorities: 20% - Americans who knew Social Security exceeded foreign aid/transportation/debt interest Foreign aid mistaken belief: 33% - Americans who thought foreign aid was the biggest federal spending item Soviet Union/NATO knowledge: 38% - Share aware in 1964 that the Soviet Union was not in NATO Contact from someone known increasing turnout: 10 percentage points - One experiment cited for interpersonal mobilization
Pivotal Quotes: "the expected value of casting a vote is often far higher than anything else you would be able to do in the same amount of time" — Rob Wiblin: Summarizing the core case for voting in close elections "we shouldn't make the perfect the enemy of the good" — Rob Wiblin: Conclusion on acting even amid uncertainty about which candidate is best "if you're in a close district in a close election, and also there's a noticeable difference in the desirability of different candidates winning" — Rob Wiblin: Conditions under which voting is clearly altruistic
Implications: Voting can be a high-leverage altruistic act in pivotal races, but not all elections justify the same effort. Listeners should focus on close, high-stakes contests, consider opportunity costs, and use persuasion or donations strategically.