Pundits love political horse races, analyzing the latest polls to predict who might win an election. And in the run-up to the US presidential election, these forecasts can influence markets and shape public opinion and policy.
But as America heads to the polls in November, new search warns forecasters to be careful with a lot of data: past demographic trends in voting. By analyzing census figures and U.S. presidential and congressional election results since 1952, researchers find that demographic factors considered reliable determinants of an election’s outcome, such as age, race, gender and voter education, may be less useful than forecasters think.
“We find that even guessing a 50-50 level is just as effective, if not better, than using some population forecasts”
How bad are demographic trends in predicting election results? So much so that, in some cases, a forecaster would be better off flipping a coin to determine a likely winner. “We find that even guessing a 50-50 level is just as effective, if not better, than using some demographic forecasts,” says Vincent Pons, the Byron Wien Professor of Business Administration at Harvard Business School and one of the authors of the article. .
The main reason is that party and candidate positions on key issues change subtly or not-so-subtly from one election to the next in response to changes in voter sentiment, thereby undermining demographic trends. Political strategists hoping to develop long-term plans based on voter demographics may need to rethink their approach.
“After an election, the electorate always seems to revert to the other party within a fairly short period of time,” says co-author Richard Calvo, a former research associate at HBS and now a doctoral student at the University of California at Berkeley. “Any prediction that there will be a long-term advantage for one party will inevitably be wrong.”
Pons and Calvo teamed up with Jesse Shapiro, George Gund Professor of Economics and Business Administration at HBS, on the National Bureau of Economic Research study, “The pitfalls of demographic forecasts for US elections.”
The appeal of demographic forecasts
To be clear, the authors are not claiming that demographic factors are unimportant in elections.
In fact, demographic groups – whether defined by age, race, gender, income, education level, or other factors – have clearly had very distinct voting patterns over the decades . Black Americans, for example, have traditionally voted overwhelmingly for Democrats, while rural and working-class white Americans have increasingly thrown their support toward Republicans in recent elections.
Demographic data and corresponding election data appeal to forecasters precisely because the available numbers are broad and precise. It is possible to predict long-term trends influenced by aging, migration, fertility and mortality. Even projections of the future size and composition of various demographic groups are often considered reliable, according to the document.
However, assuming that demographic voting trends are monolithic and impervious to small, or even large, changes from one election to the next is where population forecasts consistently falter, the authors conclude. And the further out election predictions go – say eight, twelve or more years – the more dubious they become.
Studying nearly 70 years of voting records and results
The team analyzed voting from the American National Election Study and demographic information from government sources dating back to 1952. From there, the researchers:
- Creating a predictive model linking a person’s vote to their age, gender, race, income, education, and the type of region they live in.
- And used demographic information to predict individual vote choices in upcoming elections.
“We looked at how demographic groups might have changed in size and how they had voted, say, eight years earlier,” Calvo says. “Then we (calculated) the expected election results based on these numbers.”
The results were not good for forecasters, even after the authors statistically “stacked the deck” to favor demographics-based election projections:
- Predicting up to five elections in advance is like guessing that the outcome will be the same as today.
- Someone using demographic data to predict election winners would be on average 22% worse off than someone who assumes the next election will always be a tie — a 50-50 split between Democrats and Republicans.
- Even a forecaster with full knowledge of future demographic trends would have mispredicted the upcoming election.
“Using previous voting results and demographic trends and applying them to future elections is simply not a reliable way to predict future elections,” says Shapiro.
Think Trump on abortion and Harris on immigration
The results reflect the constantly evolving political situation and voter behavior. In a separate piece of new searchPons examines the range of barriers to voter participation that have changed in recent years, including election laws, voting procedures, and registration requirements, and highlights issues that merit future study.
And when it comes to demographic-based election predictions, both in models and in reality: voters’ and parties’ positions on issues are not static from one election to the next.
Ultimately, political parties and candidates regularly change their positions to accommodate the changing opinions of different voting groups. Therefore, demographic voting patterns can change from one election to the next. Even the slightest change in a demographic group’s vote percentage from one election to the next can spoil a seemingly well-thought-out forecast, the authors conclude.
Such changes are evident in the race between former President Donald Trump and Vice President Kamala Harris.
Trump, a Republican, has recently tried to soften his anti-abortion stance amid signs he may be losing women’s support. Meanwhile, Democratic candidate Harris has recently hardened her stance on immigration and border controls, aligning with what polls say many Americans are saying.
“Candidates and parties are clearly paying attention to polls to understand what issues are important to voters,” says Pons. “Trump realizes it on abortion and Harris realizes it on immigration. These candidates are trying to adapt to voters’ concerns.”
You might also like:
Any feedback or ideas to share? Email the Working Knowledge team at hbswk@hbs.edu.
Image credit: HBSWK with assets from AdobeStock/Carolina, Rossarin and Custom Scene
