How to predict an election | Andreas Graefe | TEDxMünchen

Summary

The speaker reflects on the unexpected outcomes of the 2016 U.S. presidential election, emphasizing the challenges in election forecasting. Despite correctly predicting Hillary Clinton's popular vote win, the forecasts failed to anticipate her loss in the Electoral College, primarily due to underestimating Trump support in key states. The talk explores the limitations of polling, expert predictions, and alternative forecasting methods like prediction markets and political economy models. It concludes by stressing the unpredictability of elections and the importance of voter participation, encouraging individuals to shape the future rather than rely solely on forecasts.

Takeaways

• 😀 The speaker experienced shock at the unexpected outcome of the 2016 U.S. presidential election despite initially predicting a Clinton victory in the popular vote.
• 😀 While national polls were accurate, significant errors in swing states like Pennsylvania, Wisconsin, and Michigan led to incorrect Electoral College predictions.
• 😀 A major challenge in polling is obtaining a representative sample, particularly as certain demographic groups may not respond to polls, skewing results.
• 😀 Experts in political forecasting often misjudged the election, influenced by biases and a lack of acknowledgment of opposing evidence.
• 😀 Prediction markets, which allow betting on election outcomes, performed poorly in 2016, reflecting a narrow perspective from a highly educated participant pool.
• 😀 An effective method for forecasting is simply asking voters who they believe will win, which has historically proven to be accurate.
• 😀 Political economy models that assess the incumbent's performance and economic indicators provided surprisingly accurate predictions for the election.
• 😀 The speaker highlights the importance of combining multiple forecasting methods to mitigate errors and improve accuracy in predictions.
• 😀 There is a need for continued research to understand failures in election forecasting and to improve methodologies for future elections.
• 😀 Ultimately, the speaker encourages voter participation, emphasizing that predictions should not replace active engagement in the democratic process.

Q & A

What was the initial expectation for the 2016 U.S. presidential election results?

The initial expectation was that Hillary Clinton would win both the popular vote and the Electoral College, based on the forecasts by many analysts.

What were the actual outcomes of the 2016 election compared to predictions?

Hillary Clinton won the popular vote but lost the Electoral College, leading to Donald Trump's victory, which contradicted most forecasts.

What issues did the speaker identify with polling accuracy in the 2016 election?

The speaker noted that while national polls were relatively accurate, there were significant errors in key swing states, particularly Pennsylvania, Wisconsin, and Michigan.

What demographic group was underestimated by pollsters according to the speaker?

Pollsters underestimated the turnout for Trump among white voters with college education.

What method does the speaker advocate for improving election predictions?

The speaker advocates for combining forecasts from multiple sources and methodologies rather than relying on a single prediction.

How did experts' forecasts compare to actual outcomes in the 2016 election?

Experts tended to overestimate Clinton's chances, believing that the polls underestimated her support, which resulted in higher errors than in previous elections.

What role do prediction markets play in forecasting elections?

Prediction markets involve individuals betting on election outcomes, providing insights based on real money at stake; however, their predictions in 2016 were significantly off due to demographic biases.

What is the most accurate method for predicting election outcomes according to the speaker?

The most accurate method mentioned is asking voters who they think will win the election, rather than who they will vote for.

What lesson does the speaker draw about forecasting methods from the 2016 election?

The speaker emphasizes that forecasting methods that worked well in the past may not work in the future and that many methods had similar errors in 2016, leading to overestimations of Clinton's support.

What final message does the speaker convey regarding predictions and voting?

The speaker encourages individuals not to rely too heavily on forecasts but to actively participate in shaping the election outcome by voting.