Majority of Brits want a soft Brexit compromise, but Leave voters don’t

With the battle over Brexit returning to the House of Commons with the Internal Market Bill, it is easy to lose sight of the bigger picture: what the British public wants in terms of a longer-term relationship with the European Union. To try to find out, Simon Hix (LSE), Clifton van der Linden (McMaster University) and Mark Pickup (Simon Fraser University) conducted a survey experiment with a random sample of British voters, where they asked them to choose between hypothetical “package deals”. This forced voters to have to make trade-offs across key issues. When faced with such choices, British voters overall prefer a “softer” form of Brexit: where the UK applies EU regulatory standards in return for quota-free and tariff-free access to the EU’s single market. However, a majority of Leave voters prefer a much “harder” trade-off: of regulatory sovereignty but restrictions on UK exports. Reconciling this difference will continue to plague British politics.

With the end of the Brexit transition period, on 31 December 2020, rapidly approaching, the likelihood that the UK and the EU fail to reach a deal is increasing by the day. While the two sides are haggling over fishing rights, state aid rules, customs checks, and other minutiae of the bargain, the preferences of the British public on the bigger trade-offs in a longer-term “package deal” – such as quota and tariff-free access for UK producers to the EU single market versus the UK having to apply EU regulatory standards – have receded into the background.


To try to understand where the British public stands on the key trade-offs at this point – after the UK formally left the EU on 31 January 2020 – we conducted a survey of 5,593 UK residents in two separate waves drawn from the Vox Pop Labs online respondent panel and fielded between 14 March and 13 May 2020. The sample was weighted against marginal distributions in population demographics, to render nationally representative estimates.

We could have simply asked people what they thought about each of the key issues separately, such as their attitudes towards market access, applying EU regulatory standards, the free movement of people, payments into the EU budget, and so on.  his approach would have allowed us to gauge the strength of support for individual issues in the negotiations, but would not have enabled us to identify how people felt about the difficult trade-offs that will inevitably have to be made in a compromise “package deal”.

So, what we did was conduct a survey experiment using a method known as “conjoint analysis”, which is a technique originally developed in market research to understand consumers’ preferences about different aspects of a product. How this worked was as follows. First, we identified four dimensions of a Brexit deal and two possible options for each of those dimensions. To ease understanding, we tried to keep the options and the descriptions as simple as possible, as Table 1 shows.

Table 1. Softer and Harder Choices on 4 Dimensions of Brexit


Second, from these two choices across these four dimensions, we created all the possible potential “packages”: 16 in total.  We then presented respondents with a series of pairs of randomly chosen packages and asked them to say which package they prefer. From this simple set up, we can identify two things: (1) how popular each package is relative to every other package; and (2) the relative popularity of the individual options.

On the first of these, Figure 1 shows the overall probability that a package was preferred in a pairwise comparison, for all voters and broken down by whether someone voted Leave or Remain in the 2016 Brexit referendum. We also indicate which packages we consider to be “viable” (from the EU’s point of view), in that these packages would involve either (a) full EU market access in return for the UK applying EU regulatory standards (two green squares in the first two columns), or (b) tariffs and quotas on UK exports in return for full UK regulatory sovereignty (two orange squares in the first two columns).

Overall, the two most preferred packages for all voters (the top 2 rows in the figure) are not viable – or, put another way, they are variations of “having our cake and eating it”. Interestingly, though, the most preferred viable package amongst all voters would be a relatively “soft” form of Brexit (the third package down in the figure): with full market access (no quotas or tariffs on exports) in return for the UK applying EU regulatory standards, plus free movement of high skilled labour, and payments into the EU budget to participate in research collaboration and education exchanges.

Figure 1. Package Preferences: All Voters, and Leave vs. Remain


However, there is an important difference between Leave and Remain voters, in that while this package is popular amongst Remain voters (who preferred it 89% of the time), Leave voters prefer (61% of the time) a much harder form of Brexit amongst the viable packages: tariffs and quotas on UK exports to the EU in return for UK regulatory sovereignty, plus free movement of high skilled labour, and no payments to the EU to secure scientific collaboration (the 10th package down in the figure, the “orange-orange-green-orange” package).  Meanwhile, a “no-deal” Brexit (orange across all 4 boxes) is not popular amongst either group: it was preferred only 11% of the time by Remain voters and even among Leavers, only 53% of the time.  In contrast, Remain voters choose the Leave voters most-preferred viable package 25% of the time.

When comparing how voters in England and Scotland respond to these choices (Figure 2), some interesting differences emerge.

There is no difference between English and Scottish voters in terms of which overall package is most preferred or in terms of which of the viable packages is most preferred.

However, Scottish voters much more strongly prefer the “softest” viable package to English voters: this package wins 76% of the time amongst Scottish voters as opposed to 61% of the time amongst English voters.  And, at the other end, a “no-deal” outcome (the row of 4 orange squares) is far less popular in Scotland than in England (23% compared to 35%), and in fact, is the least preferred package amongst all 16 amongst Scottish voters. This reinforces the view that a No-Deal Brexit would threaten the future of the Union, as this outcome would be highly unpopular North of the border compared to many other possible Brexit outcomes.

Figure 2. Package Preferences: All Voters, and England vs. Scotland


In terms of the relative preferences for the individual options, figure 3 plots the average popularity of all packages that contain a particular option. This reveals how important each individual option on each dimension was in determining whether a package was chosen over another one.

These results reveal that on two dimensions Leave and Remain voters prefer the same option. On the market access options, both Leave and Remain voters prefer packages with few restrictions over those with significant restrictions (averaging over all other options for all other dimensions), but Remain voters feel more strongly about this than Leaver voters. And, on the movement of people, both Leave and Remain voters prefer an exception for highly skilled workers over a complete end to the free movement of people, although Remain voters again feel more strongly about this than Leave voters.

Figure 3. Average Marginal Effects of an Option on a Respondent Preferring a Package


On the other two dimensions, Leave and Remain voters prefer different options. On regulatory alignment, Leave voters prefer the UK to be free to decide its own standards, while Remain voters prefer that the UK align with the EU’s standards.  Leave voters feel more strongly about this than Remain voters. And on payments to the EU, Leave voters prefer the UK to not make any payments and be excluded from scientific research and educational exchange programs, while Remain voters want the UK to make some payments to continue to be involved in these programmes. Remain voters feel more strongly about this than Leave voters.

Finally, on three of the four dimensions, respondents from England and Scotland prefer the same option but on one dimension (regulatory alignment), Scottish and English respondents have different preferences. Specifically, English respondents prefer the UK to be free to decide its own regulatory standards, while Scottish respondents prefer the UK to align its standards with the EU.

Overall, our findings about what the overall majority of British voters want in terms are consistent with other research that has asked British citizens to consider trade-offs between different Brexit dimensions. For example, in September 2017, the most prominent and professionally organized citizens’ assembly on Brexit concluded that the UK should stay in the single market (and so accept EU regulatory standards) if a free trade agreement that guaranteed quota-free and tariff-free access to the single market could not be agreed.

A problem for British politicians and policy-makers, however, is that the majority of Leave voters – the core supporters of the Johnson administration – prefer a completely opposite package deal; complete regulatory sovereignty for the UK is a top objective, and are willing to pay the price of quotas and tariffs on UK exporters to the EU to achieve that.

Whatever deal the UK and the EU manage to agree to, either before 31 December 2020 or in the weeks and months following that deadline, this deal will not be the final endpoint of the Brexit process. The tension between market access and regulatory freedom will continue to be at the heart of the ongoing relations between the UK and the EU. And, the gap between the views of the majority of the British public (who place more emphasis on market access) and the majority of Leave voters (who place more emphasis on regulatory freedom) is likely to plague attempts to resolve this tension for quite some time.

This article gives the views of the authors, and not the position of LSE Brexit, nor of the London School of Economics. Image: CC-BY-NC-4.0 Fotograf: Tom Nicholson

This is a companion discussion topic for the original entry at