It is hard to imagine how Populus could have made their poll, given prominent coverage by the Times, on the budget tax changes less reliable. They make just about every mistake in the book.
Misleading questions
Populus found support for tax increases on "large, so-called gas-guzzling" cars; the poll accepts the accuracy of the Chancellor's spin as a given.
The very term "gas-guzzling" is clearly leading as it carries undertones of greed and wastefulness. Even if the question had been framed in a more neutral manner, though, that isn't what we had in the Budget. The Telegraph reports the findings of our research on this issue:
"Analysis shows that over the next two years, millions of drivers will face soaring bills as road tax on some popular family models doubles.
However, duty on Rolls Royces and Porsches will rise by less than the typical family saloon - undermining claims that the taxes are meant for gas guzzlers."
Put that in your poll and see what result you get...
Polling before the measures in the Budget were really understood
All the fieldwork for this poll was done on Wednesday evening, not long after the Chancellor's speech and before the morning papers. At that stage the main news source will have been the BBC, which largely accepted the Chancellor's presentation of his VED changes. Since then the true picture has been emerging. With the most popular broadsheet and tabloid newspapers slamming the changes and exposing how misleading their initial presentation in the Budget was do we seriously think that public opinion won't have changed?
Small sample size
The last TaxPayers' Alliance annual conference poll (PPT) by YouGov surveyed 2,162 people. By contrast, this new Populus poll only asked 596 people. That means its results are far less reliable. This is true for all of the results that the Times cites so confidently but particularly when they focus on particular social classes. The number of, for example, "professionals and managers" in the survey will be very low.
No past vote weighting
Mike Smithson of the respected PoliticalBetting.Com website notes that:
"The survey did not include a voting intention question and was almost certainly not past vote weighted. This is likely to have produced a smallish but significant nevertheless pro-Labour sample."
This will probably mean people look less favourably on the Budget changes than this poll suggests.