How Well Do Political Practitioners Actually Persuade the Public?

A recent paper by Broockman, Kalla, Caballero, and Easton showed that “political practitioners and laypeople both perform barely better than chance at predicting persuasive effects.”

That means, when our campaigns write two different persuasive messages, they can’t accurately guess or intuit which one will be more effective. Or even if either will be effective!

This has several possible implications for campaigners:

Most obviously, if campaigners can’t reliably predict which messages are persuasive, testing is crucial for effective messaging!

In fact, the authors of the paper say:

This suggests that gathering experimental data to learn what works, rather than relying on political experts’ limited ability to forecast what is persuasive, may be most useful for political practitioners.


They point out that this kind of data-driven campaigning is on the rise, but that only 53% of the practitioners they surveyed said that they often use rigorous testing to inform their messaging.

But even among those who are testing regularly, if we don’t have great intuition about which messages change minds, writing persuasive messages to test is also likely to be a challenge!

We should aim to diversify the messages we test, even writing some that our intuitions tell us won’t be effective. Message authors could try involving members of the target audience in message creation, writing messages based on what they hear from our audiences in their own words, and should generally aim to take inspiration from a wider range of sources than our own brains and intuition.

In short, the study tells us that if we want to be better at persuading people, we need to use data and testing to guide us. And when we’re making messages, we should keep an open mind and try lots of different approaches.