The Lord Adonis guide to predicting elections: The best leader wins – nothing else matters

The Lord Adonis guide to predicting elections: The best leader wins – nothing else matters

The LAB peer and former cabinet minister, Andrew Adonis, has a fascinating essay in the latest edition of Prospect on the best guide to election forecasting. His conclusion is encapsulated in the headline above – the party with the leader perceived to be best wins and nothing else matters.

He opens by recalling a Guardian article by Jonathan Freedland when Brown was PM and when many in the Labour party were demanding a new debate on “the issues”

Freedland cautioned that “people do not believe in ideas: they believe in people who believe in ideas.” The moment I read those words, a penny dropped, and my conviction has become stronger with each passing year I have spent in politics, that the battle of ideas in politics—indeed in life—cannot be comprehended separately from the people who hold and espouse those ideas.

This is a view I have long espoused noting that the leader ratings in the elections where the polls were out on voting intention (1992, and 2015 for instance) were a better guide to the eventual outcomes.

Adonis has produced the above ratings on every UK and US election since 1944. The numbers for each main party leader/presidential are derived not from polling but his personal assessment. He notes:-

“. Instead, “leadership points” are given to the two individual leaders contending for power on a 15-point scale. Up to 10 points are awarded for raw leadership talent, and up to another five for fitting with the times. The “winner” is the candidate with most points.”

His ratings appear broadly about right though, no doubt, many would quibble.

So his overall verdict that only in one of all the US and UK election since 1944 did the person perceived as not the best leader win is derived from his own numbers. That does not make the analysis or the main point wrong.

Elections are about choosing those we wish to lead us and if we are broadly satisfied with someone on that count then we are generally more willing to accept policy positions that we are less happy about. We see this in those polling tests when we can observe changed levels of support for an policy if we link a leader’s name to the proposal. Thus in the run-up to GE2015 the EdM plan to bring in energy price controls attracted greater levels of support when his name was not linked to the plan.

Mike Smithson


Comments are closed.