2024 Annual Chair and CEO dinner: How to Predict an Election and What This Means for Business

Lord Finkelstein began by highlighting the relevance of statistics to political analysis, emphasising the need to avoid over-correlating expert opinion as well as distinguishing between signals and noise. When considering the general election or analysing political press coverage, general knowledge and awareness of politics varies greatly between individuals. Lord Finkelstein outlined his three rules of predicting an election: the ‘perceived pain’ rule; the ‘jaws’ rule and the ‘contrast’ rule.

The ‘perceived pain’ rule

How constituents feel, alongside their own firsthand experiences, will affect their political decisions. Taking medical procedures as an example, if asked to describe the pain of surgery, most patients will recount the duration of the procedure multiplied by the amount of pain at any given moment. Patients most easily recall the peak of the pain and the end of the procedure, meaning the amount of pain remembered is less than experienced. As such, a lengthier procedure with lower pain levels is perceived as more tolerable than shorter, more painful surgery.

Politics works in a comparable way. If the question is asked ‘are you better off now than under previous governments?’, it is unlikely for the general public to frame their own lives in a way that fits this form of political analysis.

When making a calculation about the general election, the first point for business to consider is how well-off people feel generally, the second is how well-off they feel following economic decisions made by the current government. When predicting what will happen in the general election relative to the present, it is crucial to consider whether the economy improves people’s personal disposable income before the election. Should economic circumstances continue to improve, people may feel better off and be more inclined to be sympathetic to the government currently in power.

The ‘jaws’ rule

In New Jersey between 1912 and 1916, a number of people were killed by shark attacks resulting in the closure of beaches, upsetting the local economy. In the subsequent election in 1916, Woodrow Wilson, who was running for President, performed worse in the polls in all the areas which suffered shark attacks compared with those that did not. While not connected, voters did not produce a long chain of reasoning as to the fault of these attacks and began to blame Woodrow Wilson. In the current political climate, the longer a party is in office, the more ‘shark attacks’ occur, and more support tends to be lost.

The ‘contrast’ rule

Most voters make judgments by contrast. Comparing perceptions of whether a political leader is more or less popular than another explores the notion that by simply establishing what someone is not, you can establish what they are. To be successful in politics, the contrast needs to be in the party leader’s favour and is crucial to winning a general election on both economic and leadership fronts.

Where does AI sit in the elections debate?

During the discussion, the subject of AI targeting individuals and mobilising social media was not believed to have a significant impact, due to opposing parties having broadly similar technical campaigning abilities.

Political parties can use AI to gain a competitive advantage in understanding of the electorate. Where AI is deployed as a technique alongside increased election campaign spending it can bring disruption and voting shifts.

The 2024 annual Chair and CEO dinner proved to be a thought-provoking and insightful event. Valuable business insights were shared on national energy and infrastructure strategies, financial market pressures and the importance of governance and regulation. The Odgers Berndtson team would like to thank Lord Finkelstein warmly for his questions, address and then answering and responding to the discussion.

Get in touch. Follow the links below to discover more, or contact our dedicated leadership experts from your local Odgers Berndtson office here.