Public Think Private Healthcare is "Acceptable" for a Prime Minister

Today saw the weekly flurry of Friday polling, and generally it corroborates the broader picture. Techne and Omnisis found Labour leads of 21% and 20% respectively, while PeoplePolling continued their recent pattern of giving Labour relatively higher leads, on 27%. Deltapoll is the only pollster outside of the low twenties range - as they've typically been understating Labour leads relative to other pollsters, currently they give Labour a 14% lead.

It's difficult to specifically account for Deltapoll's divergence (made more so by the limited data included in their tables). It could be partly a result of their turnout weighting, which excludes those who don't put their likelihood to vote at 9-10/10. The recent decline in government popularity might incline disaffected Conservatives to put their likelihood to vote lower, which would see their responses excluded. The pollster also finds a relatively lower vote share for Reform, which could indicate their respondents are choosing to switch from the Conservatives with less frequency - though the fact they also prompt for UKIP vote at first time of asking likely provides a more powerful explanation for this.

Looking deeper into the new polls, PeoplePolling did ask one topical question:

"Do you think it is acceptable or unacceptable for a Prime Minister to have private health insurance (which means they rely less on the National Health Service)?"

53% responded it was acceptable and 27% disagreed. This is a large margin, so it's likely the result would remain consistent with a differently worded question, though the margin could be exaggerated by the framing. The inclusion of the bracketed phrase seems like it might incline a respondent more towards acceptable. Moreover, the decision to ask about "a Prime Minister" generally might also influence responses, especially when Rishi Sunak is already often perceived to be somewhat out of touch.

The choice of the specific wording could also be material. The word "acceptable" provides a less strong endorsement than, say, "right" would, while a different framing asking if voters "approve" might also see lower responses. That being said, it's still a useful question which provides telling insight into an issue that's prominent in political debate. It should just be reported with care.