Lib Dems tipped for most gains in May’s elections in independent prediction

The Liberal Democrats are tipped to be the big winners from May’s local elections according to the predictions made by local government election experts Professors Michael Thrasher and Colin Rallings.

Their predictions are made each year, based on the results in council by-elections in the run up to the big May round of elections. For 2017 they are predicting the following in England:

Liberal Democrats +100
Conservatives +50
Labour -50
Ukip -100

The big losses predicted for Ukip are not a surprise as Ukip’s vote share in local council elections has been sinking year after year since its breakthrough successes in 2013 (details here – including how to make accurate year on year calculations).

As for vote share, the predictions are (with again vote change calculations done on a like-for-like basis):

Conservatives 31% (-1% on 2016, +5% on 2013)
Labour 29% (-4% on 2016, +0% on 2013)
Liberal Democrats 22% (+8% on 2016, +9% on 2013)
Ukip 10% (-2% on 2016, -12% on 2013)

If that is indeed the Liberal Democrat result, it would be the best since the 2010 general election, though not yet quite back to the sorts of figures that were the norm in earlier years.

But, of course, a question with any predictions from a team with a long track record should be ‘how accurate are these predictions?’. A question oddly rarely asked about the Thrasher/Rallings data despite all the media love of talking about data and statistics when, for example, referring to US politics.

So, lucky reader, you have chosen well to come to this blog… because I have the only public set of records on how previous Thrasher and Rallings predictions have turned out:

• Conservatives: on average the result is within a point of the prediction, but with a wide spread around the average
• Labour: on average the result is around 1 point lower than the prediction but with a spectacular miss in 2009
• Liberal Democrats: on average the result is 2-3 points lower than the prediction

(If the theory floated as an option by Matt Singh that the Liberal Democrats have started doing disproportionately well in council by-elections turns out to be true, then the Lib Dem underperformance against the prediction this year should be larger than average given the prediction is by-election based.)

You can help make the Liberal Democrat prediction accurate by donating to the ALDC Local Election Appeal.