Here’s the full report of the opinion polling industry’s post-mortem into its own mistakes at the 2015 general election.
Three points are particularly of note.
First, the pollsters were by no means the only industry to get the election badly wrong. The media might blame the pollsters for misleading them, but it’s striking how willing the polling industry has been to examine its mistakes compared to the relative absence of media post-mortems.
Second, the report contains an important set of recommendations to improve the already good transparency of the political polling industry provided by the British Polling Council. Compared to, for example, the US, polling transparency in Britain is streets ahead and these further sensible reforms will reinforce that welcome position.
Third, the key finding about the polling errors is that the pollsters were not sampling truly representative samples of the population. Lots of detail on why other explanations for their errors don’t really stack up and on what caused the sampling problems are in the full report, so here it is:
Sturgis, P. Baker, N. Callegaro, M. Fisher, S. Green, J. Jennings, W. Kuha, J. Lauderdale, B. and Smith, P. (2016) Report of the Inquiry into the 2015 British 0general election opinion polls, London: Market Research Society and British Polling Council.