Poll-driven politics does nobody any favours

Receive free Data Points updates

In his 1985 book Amusing Ourselves to Death, American cultural theorist Neil Postman argues that “the medium is the metaphor” — that is, the dominant ways in which people consume information influence that society’s culture.

He contrasts the era when people got their information from the printing press — a time when the big conversations of the day were detailed and logical — with the age of television, where complexity is avoided and nuance dispensable, rendering America’s national discourse “shrivelled and absurd”.

It’s beginning to feel as if the age of poll-based punditry is having a similar impact on politics.

In the last month alone, more than 100 large-scale representative surveys have been carried out on politics and current affairs in the US and Britain, as political parties, media organisations and advocacy groups probe public sentiment on various issues.

In the UK, Rishi Sunak’s government appears to be trying to poll its way to popularity. Each week brings a new front page story heralding a survey showing voters either love or hate a particular policy, with the implication that by taking a strong line on that issue, the Conservatives will eat into Labour’s lead. But after many weeks and many issues — immigration, clean air schemes, net zero — the Tory vote share hasn’t budged.

This may sound surprising, but it shouldn’t. One-dimensional survey questions are not a good way of measuring what is almost always multidimensional public opinion.

To take one recent example, the Conservatives have identified cutting inheritance tax as a sure-fire vote winner. A poll taken earlier this summer suggests they’re right. Just 23 per cent of Britons say inheritance tax is fair, and 43 per cent say unfair, the biggest net negative rating out of 13 taxes polled. But later in the same survey, only one in seven people say they would prioritise cutting it. And in an in-depth survey run in June this year by the cross-party think-tank Demos, 55 per cent said people should always be able to pass on inheritance tax free. But when asked what the threshold should be above which tax is charged, only 22 per cent said inheritance should always be tax-free.

Polling on immigration raises similar issues. In surveys of Britons and Americans carried out for the FT by the research company Focaldata, 56 per cent of US adults said immigration is too high, as did 60 per cent of Britons. Clear support for policies that would cut numbers, then. But when we asked how people felt about the number of immigrants from 12 different categories depending on job or reason for migration, in every case only a minority said there should be fewer.

Only tiny minorities of Britons and Americans say their country should take fewer healthcare workers, scientists or engineers from overseas. Even when asked about people coming to the country to avoid conflict, fewer than a third say numbers should be reduced. The highest opposition is towards people coming “to seek a better life”, but even here only a minority wants numbers reduced.

To be clear, none of the latter responses invalidate the former. The public is under no obligation to hold coherent policy views. But the emerging pipeline from simple for/against poll to front page story to policy change does neither the public nor politicians any favours. 

It’s not that people are answering the for or against questions dishonestly, but rather that one simple question cannot capture the nuance of public opinion on complex issues. Putting a single number on these topics will inevitably mislead, and produce overconfident predictions of the impact of policy changes.

Perhaps most important of all, political science has shown that for decades now, what voters care about above all else is competence on the key issues facing the country, not differences in position on second-tier concerns.

If a party wants to improve its standing with the public, not to mention improve the material circumstances of the country, its best strategy is to solve problems, not obsess about the polls., @jburnmurdoch

Survey details

Focaldata surveyed 1,009 British adults and 1,027 US adults between 21 and 26 September 2023. Click to download the underlying data tables for the UK and US.

Articles You May Like

Fighting AI fakes with blockchain: Real AI & crypto use cases, No. 4
Judge approves sale of bankrupt, bond-financed Arizona sports venue
Replace-by-fee (RBF), explained
Circle and SBI Holdings partner to boost USDC circulation in Japan
KyberSwap DEX hacker sends an on-chain message: Be nice, or else