Jon Mellon, Chris Prosser, Adam Feldman and Jordan Urban

Now that Parliament has finally dissolved for an election, it’s time to take stock of how well the Conservative party fulfilled the promises it made before the 2017 General Election. Going into the next election, it’s important for voters to know how much they can trust the big promises that parties make during the campaign.

Voters are generally sceptical that parties fulfil their promises, with many citing high profile broken promises such as the Liberal Democrats’ infamous u-turn on tuition fees. However, the academic literature has generally rated governments a lot more sympathetically, finding that parties fulfil most of their campaign promises.

This is also the case for the 2017 Conservative manifesto. We went line by line and found 257 separate promises. Out of these, the Conservatives partially or fully fulfilled 175 of their promises (69%).

But this ignores an important problem, some promises matter more than others. Many of the fulfilled promises are not exactly what the average voter would think of when they consider the government’s platform. Promises such as Use 3rd party identification to make accessing online government services more secure and Update the laws on selling pets might be valuable policies, but they probably weren’t what anyone was thinking of when deciding whether to vote for the Conservatives in 2017.

By contrast, in 2017 the Conservative manifesto also promised to leave the EU, leave the EU’s Customs Union and reduce annual net migration to below 100,000. These promises received a lot of attention, but none of these promises were fulfilled.

Most studies treat the fulfilment (or lack of fulfilment) of each of these promises exactly equally, but intuitively it seems like a bigger failure to deliver on the key immigration and EU promises than to not update GOV.UK’s web interface.

But how can we work out which promises voters thought were important to a parties’ platform in an election? In a new paper we try to answer this question.

We asked 4,908 respondents about how central different sets of these promises are to what the Conservatives promised to do in the 2017 election. Because these combinations are randomly assigned, we can use statistical methods to find how central each promise was seen as being to the Conservatives’ 2017 agenda.

The following plot shows how much this changes our view of what the Conservatives stood for in 2017. The dots show the unweighted percentage of promises about each issue. The arrows show how this changes when we account for how central voters saw each promise as being to the Conservative agenda. The distributions show how much statistical uncertainty we have about the importance of each issue area. Higher peaks mean that there is a higher probability that the percentage of the agenda is more likely to be at that point.

If we just count up promises, 5% were about Europe and 3% about immigration. But these are overwhelmingly the promises that voters saw as defining what the Conservatives stood for in 2017. When we weight the promises according to how important voters thought they were, EU promises make up 35% of the Conservatives’ agenda and immigration promises make up an additional 13% of their agenda. By contrast, the vast collection of small miscellaneous promises such as Review the design of government buildings and Set up a commission to ensure technology advances ethically fall from 59% of the Conservative agenda to 25%.

% of the Conservative manifesto that was about each issue area. Distributions show the probability that an issue took up a certain proportion of the Conservatives agenda after weighting.

The top five Conservative promises were:

  1. Leave the EU (broken)
  2. Reduce annual net migration to under 100,000 (broken)
  3. Leave the EU’s Customs Union (broken)
  4. Introduce a bill making preparations for setting up new trade deals after Brexit (fulfilled)
  5. Leave the EU’s single market (broken)

While the bottom five Conservative promises were:

  1. Keep supporting the Welsh-language TV channel S4C (fulfilled)
  2. Introduce child bereavement leave (fulfilled)
  3. Require schools to have a single point of contact with mental health services (partially fulfilled)
  4. Create Schools Maps to give parents detailed information about local schools (fulfilled)
  5. Hold a ‘Great Exhibition of the North’ to showcase Northern culture (fulfilled)

So how well did the Conservative government perform if we measure completion in terms of weighted promises? In short, not that well. Their weighted completion rate (fully or partially) is just 48%, a lot lower than the simple proportion of promises completed (69%). This difference is even starker if we just look at promises to change the status quo. The weighted completion rate is 40% whereas the unweighted percentage was a considerably better 64%.

% of 2017 promises the Conservatives kept after applying weights. Distribution shows the likelihood of different completion percentages. Arrow shows the difference from the raw percentage.

So what does this tell us for the 2019 election campaign. In general our evidence and others is that parties will generally a high proportion of their promises if they get into office. But it’s clearly not always the case for their most central promises. So perhaps voters are justified in being a little sceptical of parties’ most eye-catching promises.

We will be following up with another round of our study for whichever party wins the next election (using the British Election Study Internet panel) to track just how well they keep their promises.

Note: You can look at all the promises we assessed at GovTracker at

%d bloggers like this: