Punters, pundits and the UK election

This column in the Financial Times was published on 30 March 2015. David Cameron duly won the election, against most pundit predictions.

If you accept that the primary interest of investors is to form objective judgements about the outcome of uncertain future events, then the forthcoming UK general election poses a particular challenge. This is set to be the hardest general election in living memory to predict. The range of possible outcomes for who forms the next government is unusually wide. Among other consequences, not one of the mainstream party leaders can be certain of still being in post by the autumn.

Even Nate Silver, the new star of global political punditry, confesses to being unsure of calling this election right. By his own admission his forecasts at the last UK election performed underwhelmingly and were a good deal less successful than his near perfect state-by-state forecast of the 2012 US presidential election. With this election, according to his website, he seems happy to delegate a good deal of the heavy analytical work to three UK academics who follow a similar probabilistic philosophy to his own.

Their most recent forecast at the time of writing, which you can follow at electionforecast.co.uk, had the Tories narrowly beating Labour as the party with most seats (284-278) and the largest share of the vote (34.2% to 32.4%). But the margins for error are significant: at 90% confidence intervals, the Tories and Labour could both do around 4% better or worse in share of vote and 40 or more seats better or worse in terms of seats won.

In fact, the three academics estimate, neither Labour nor the Tories has more than a 4% chance of winning an outright majority, while there is only a 37% chance that any two parties will control a majority of seats in the new Parliament hence the plurality of future government possibilities. If this turns out to be the case, voters in aggregate will have little or no chance of knowing what exactly they are voting for, making tactical voting an even more complex task than in the past.

Of course nearly all electoral forecasts are derived from models that ultimately, with a variety of other tweaks, derive from surveys of voting intentions, in which opinion polls inevitably play a dominant part. And opinion polls, as we all know, can be wildly misleading, as the 1992 UK general election and the recent Israeli vote vividly demonstrated. Most investors share the sensible belief that bookmakers odds are in general a more reliable guide to election outcomes than opinion polls.

Here too the message, while more encouraging for Mr Cameron and the Conservatives, is compatible with the academics assessment. According to Oddschecker, which tracks most leading bookmakers current odds, the bookies have the Tories at around 2:1 on to win both the most seats and the most votes, with Labour roughly 2:1 and 6:4 against respectively. A hung Parliament however remains by a long way the odds on favourite.

When punters and pundits agree, in this respect at least, it might seem foolish to offer an alternative prospectus. Yet the uncertainty surrounding the outcome means the final result could still be materially different from the central forecast case. One obvious reason is that the coalition government, the first in peacetime since the 1930s, has changed the traditional rules and norms of the political process. Another is that the financial crisis of 2008 has produced some atypical economic consequences.

According to a recent study at the University of Essex, the traditional relationship between the state of the economy and political party ratings has not been anything like as strong or as positively correlated since 2008 as it was before the crisis. Economic recovery has not (so far at least) fed through into higher approval ratings for the coalition government to the same extent that it would have done in the past.

That seems to be partly because voters have (sensibly) decided that neither the crisis nor its aftermath can be wholly attributed to the efforts of the UK government and partly because the lopsided and unusual nature of the recovery (employment up, but productivity and household incomes, at least until very recently, weak) means that the benefits of recovery have not been as widely distributed across the population as would normally be the case.

In The Coalition Effect, a recently published compendium of essays that analyses the record of the coalition government, Paul Johnson and Daniel Chandler of the Institute for Fiscal Studies note that whoever wins the next election is effectively committed to introducing a further dose of austerity at least as severe as that which Mr Osborne has presided over in the past five years. Yet you would not know it either from the way that the political debate is being framed, or from the meaningless fiscal forecasts offered by the man parties since the Budget, none of which bear more than a passing resemblance to what lies ahead.

Amidst all this uncertainty, certain traditional trends do still seem to be reasserting themselves, albeit much later in the day than usual. The polling lead of the main opposition party invariably declines as the election approaches. The Government vote almost always declines from one election to the next (but this time the Lib Dems are bearing the brunt of the fall). Parties which are behind on both economic competence and leadership qualities have never won a UK election. Unfortunately for Mr Milliband none of these factors work in Labour s favour and my guess is that the odds will continue to drift in Mr Cameron s direction between now and May 7th, even if the polls fail to show any signs of shifting.