When the Frame Shapes the Answers
The Polls Don’t Lie — But Their Questions Do
A new poll by YouGov asked Americans two separate things: first, what behaviors pose pregnancy risks, and second, what factors contribute to autism. On the surface, both look neutral. They’re straightforward lists of options — smoking, pesticides, pregnancy medications, vaccines, genetics. Respondents simply checked boxes and percentages were tallied. After all, who wouldn't want to know what Americans believe about autism? But beneath the surface, the poll design itself tells a story — one that treats autism as a disease outcome rather than a way of being, a normal human neurological variant.
When the frame assumes autism is something people develop — like cancer or diabetes — the answers can only reflect that assumption. It’s not that Americans came up with deficit thinking; it’s that deficit thinking was built into the poll questions themselves.
Autism as Risk, Not Reality
In one block, autism was folded into a list of “pregnancy risks” that also included smoking, raw fish and unpasteurized cheese. In another, respondents were asked which factors contribute to autism. These design choices matter. They collapse autism into the same category as fetal harm. They position autistic people not as subjects of a life worth supporting but as objects of prevention.
Nearly half of Americans said pregnancy medications pose some risk of autism. A quarter still named vaccines. Seventy-four percent said genetics. Each of these answers looks like data — until you notice the only options on offer were framed as hazards. No one was asked whether autism is a natural form of human variation. No one was asked what helps autistic people thrive. That omission wasn’t an error of execution — it was a deliberate boundary of design.
Manufactured Misunderstanding
This is how a loop gets reinforced:
Scientists frame autism as pathology or risk.
Pollsters design questions that mirror this logic.
Respondents echo the frame back in their answers.
Journalists report the results as “public opinion.”
The cycle gives deficit framing the weight of consensus. What Americans “believe” about autism is actually what institutions keep scripting into the surveys themselves.
The Political Amplifier
This isn’t happening in a vacuum. The poll followed a press conference where Donald Trump and Robert F. Kennedy Jr. told pregnant women to avoid Tylenol, claiming it raises autism rates. Republicans in the survey were more than twice as likely as Democrats to say acetaminophen poses a high risk. That divide didn’t appear by chance. It’s the echo of a narrative pushed from the podium.
Polls don’t just measure opinion. They also reflect how quickly misinformation gets absorbed. Autism, cast as a tragic outcome, becomes the perfect vessel for political messaging. This survey reflects not what autism is, but what powerful people want it to symbolize.
Who Benefits, Who’s Missing
The beneficiaries here are obvious. Politicians gain headlines by naming a new “threat” to unborn children. Polling firms gain credibility by offering numbers to cite. Media outlets gain a ready-made story about cultural divides.
Who’s missing? Autistic people. Not as respondents — but as framers. At no point did the survey ask us how these narratives affect our lives. At no point did it recognize autism as identity rather than pathology. The poll uses us as risk objects while silencing us as subjects.
Methods That Matter
It’s easy to dismiss survey design as technical detail. But details define meaning. If you ask “what causes autism,” you’ve already denied that autism is present from birth. If you ask “what contributes to people developing autism,” you’ve smuggled in the idea of onset. If you ask “what pregnancy risks might lead to autism,” you’ve equated autism with birth defects.
These are not neutral choices. They’re epistemic ones. They set the boundaries of what answers are allowed to exist. And they ensure that the results, however precise the percentages, will only ever circle back to deficit logic. This isn’t semantics — it’s structural framing, and it shapes how entire publics come to understand autism.
Better Questions, Different Answers
Imagine if YouGov had asked:
What kinds of support help autistic people live full lives?
How should schools and workplaces adapt to autistic needs?
What policies would reduce barriers for autistic adults?
Those answers would measure public opinion too. But they would measure attitudes toward support, not prevention. They would frame autism as reality, not risk.
The Real Lie of Polling
The lie is not in the numbers. It’s in the premise. When autism is cast as a pregnancy hazard, the survey has already decided what autism is: a mistake, a tragedy. Respondents can only confirm or deny degrees of fear. The percentages feel scientific, but they’re staging a script written long before the first question went live.
Until autistic people write the questions, polls will keep telling the same story. Not about us, but about the anxieties projected onto us. The polls don’t lie. But their questions do — and that lie is what keeps deficit thinking alive.