Dissmarket Newsletter — Beehiiv Edition | May 2026
TL;DR: The polling industry has a 94% nonresponse rate, elections keep surprising us, and prediction markets only reflect what a narrow slice of traders think. Dissmarket measures both signals — and the gap between them is where the real insight lives.
Quick stat for you: telephone surveys in the 1990s had 30–40% response rates. Today? Pew Research Center puts it at roughly 6%.
That means 94 out of every 100 people a pollster contacts just... don't answer.
The entire polling industry depends on the assumption that the 6% who do respond can be weighted to represent everyone else. That's a big ask — and recent elections suggest the weights aren't holding.
The result: Polls missed the mark in 2016. Again in 2020. And again in 2024. Not because pollsters are incompetent — because the methodology has a structural flaw that no amount of statistical sophistication can fully fix.
You've probably seen the hype around prediction markets. Polymarket alone processed billions in trades during the 2024 election. Major outlets now cite market odds alongside poll numbers.
And prediction markets are powerful. They update in real time. They incentivise accuracy (your money is on the line). And the research backs them up — they consistently outperform polls on election forecasting.
But here's what the prediction market boosters don't tell you:
Here's where Dissmarket comes in.
We don't pick sides in the polls-vs-markets debate. We measure both — and we track the gap between them.