In recent days, the possibility that the UK might ‘walk away’ from Article 50 negotiations once again captured headlines and spawned hashtags. On 11 October, Sky News tweeted the results of a poll:

Screen Shot 2017-10-12 at 19.49.50

As one might expect, the head-banging end of pro-Brexit opinion seized upon these ‘data’ (sic) to conclude that “there is big public support for the British government to walk away from a deal with the European Union.” I shall leave it to others to explain in detail exactly why ‘no deal’ (or ‘crashing out’, or ‘WTO Rules’) would be very bad indeed from an economic point of view, and instead concentrate on the logical deficiencies which render this poll useless.

Would you rather?

Before smartphones and Wikipedia, pub conversations after the fifth pint or so were often to be about trivia. For example: was Die Hard originally pitched as a sequel to Commando? Whereas a few friends (male friends, probably) would once have argued the toss for hours, nowadays someone will whip out an iPhone and find the definitive answer  in seconds. (The answer is no, by the way.)

But we can still slur our way through endless arguments about topics which are conditional, contingent, hypothetical or incommensurable. In a fight between Spartans and Samurai, who would win? Could the Nazis have won WWII if the Molotov–Ribbentrop Pact had held, or if the US had been neutral? Would the 1966 England squad be good enough qualify for the World Cup nowadays? And so forth.

Even less susceptible to empirical data is the game of “Would you rather…?” in which a (usually ludicrous) choice is presented. Would you rather have a finger for a tongue, or tongues for fingers? Would you rather only be able to whisper, or only able to shout? Would you rather shit bricks, or puke slugs?

The rational answer, unless you happen to be a weirdo or a masochist, is almost always “neither”, but that’s not playing the game. Yet, at least in this game, both options are intelligible. In the game of “Would you rather get no Brexit deal or a bad Brexit deal”, neither of the variables is clearly defined.

‘No Deal’…

The poll question essentially presents people with two unknown outcomes and asks them to choose between them. How people choose will largely depend upon their own beliefs about what constitutes a ‘bad deal’ and what the likely result of ‘no deal’ would be.

Analogously: would you rather have no government, or a bad government? Your answer is going to depend on at least two things. First of all, how you feel about the prospect of, and likely effects of, having no government. Are you an optimistic Libertarian or do you fear anarchy? Your view might be more nuanced, of course. You might think that a period without government would be better than having a Venezuela-type government, not because anarchy is good in itself, but because it will inevitably lead to democratic change.

Second, how bad is this imagined ‘bad government’? Is it bad in the sense that the current UK government is unimaginative, statist and conspicuously lacking in talent? Is it bad in the sense that an economically incompetent, authoritarian Corbyn government would be? Or are we talking ‘bad’ meaning a government that starves and executes its own citizens? Or might we even be talking about a government that is so wicked and destructive that even Jeremy Corbyn would not feel able to praise and admire it?

To return to Brexit, how are we to define ‘no deal’? In the sense that a lot of politicians and wonks use it, the term seems to suggest ‘no bespoke free trade deal’. But it might also mean no deal, or deals, of any kind. Thus, no agreements about the status of UK/EU citizens, no agreements about European airspace, no agreements about security, no agreements about anything.

For what it’s worth, I think that ‘no (trade) deal’, which is to say, ‘trading under WTO rules’, would be pretty bad; although, as a believer in free markets, I can understand why it is tempting. Even the cheerleaders for this option (like Professor Patrick Minford) admit that it would mostly eliminate manufacturing; which is undoubtedly very bad if your job (or your community) happens to rely on that sector. In the last few days, John Redwood MP has been ‘reminding’ (sic) his 26K Twitter followers:

Screen Shot 2017-10-12 at 21.16.23

Taken alongside his other tweets, the implication is clear: other countries trade with the EU, so ‘no deal’ is just fine. What he fails to mention, however, is that most of the larger economies from the ‘rest of the world’ do have a number of agreements in place. They may not have comprehensive free trade deals in place, but they still have ‘deals’. None of the EU’s top 20 trading partners trades with the EU on WTO Rules alone. Between the EU and the USA, its biggest trading partner, there are no fewer than 135 agreements of various kinds (including 55 which are bilateral), for example.

But whether you believe that ‘no deal’ (WTO Rules) will be a disaster or an unalloyed blessing, you still cannot (reasonably) say whether it would be better or worse than a ‘bad deal’ because that notional ‘bad deal’ is not defined either.

How bad is ‘bad’?

It is obviously the case that ‘no deal’ must be better than a ‘bad deal’ because it is at least logically possible that a ‘bad deal’ might be very bad indeed (just as it is obviously true that ‘Brexit means Brexit’ — it’s just that neither tells us very much). If the EU only agreed to strike a deal in exchange for lives of every first-born British child and 10% of the adult population taken into servitude, then yes, ‘no deal’ would be better. But we are not living in the Book of Judges. The problem with framing the question so vaguely is that it is always possible to imagine something worse.

What constitutes a ‘bad deal’ will depend on your point of view. If, like me, you support Brexit because the EU is a federalist, anti-democratic racket, insensible to the real sufferings of the very people it imagines to be its ‘citizens’, then it’s really all about sovereignty. If, after March 2019, we continue having free movement with the EU, then I’d have no objection. In fact, as a right-winger and lover of free markets, I’m very much in favour of the free movement of labour. Others see it differently, of course. The EU Referendum was not about immigration, despite the best efforts of a curious alliance between xenophobes and EUphiles determined to explain it as such. Still, we cannot doubt that some people, at least, voted for Brexit because they wanted to stop further movement of people from the rest of Europe to the UK.

Whatever they thought they were voting for, what they actually voted for was the proposition on the ballot paper (that the UK should leave the EU), and thus their economically half-witted xenophobia can (and should) be ignored. But when they respond to a poll, these views will colour their answers. At least some of these people would presumably prefer economic catastrophe, so long as it is only the ‘indigenous British’ who are affected by that catastrophe.

The Fallacy of Relative Privation

The logical fallacy at work in this poll and the discussion around it, is one with which we are all familiar, although it is so moronic that it used only to be employed with children before the advent of the internet. It is the fallacy that seeks to establish the goodness of something by comparing it to something worse. We probably all remember it being used on us as children, when we were told that starving children in Africa would be glad of the broccoli/cabbage/turnips that we had refused to eat. How dare a child complain about greens tasting yucky when starving children would devour them gratefully?

In other words, there is always something worse. You may complain about your aches and pains with the ‘flu, but what about people who have cancer? Okay, let’s talk about them… How dare you complain about your treatable cancer when there are people with terminal cancer! Fair enough. So how can you moan about having ‘only’ two years to live, when some people with aggressive cancers have less than 2 weeks? And so forth.

Would you rather scavenge in dustbins for food or die of starvation? One would have to be supremely fastidious (and irrational) to prefer a miserable death to the relative yucky indignity of eating from bins. Would you rather eat human flesh than starve to death? Again, this is an horrific prospect, but one to which many human beings have resorted, in extremis. It’s likely that a majority of human beings would regard consuming refuse, or other humans, as a less bad option in comparison to a lingering death from avoidable hunger. It does not follow, however, that ‘A majority of British people want to eat from bins,’ or ‘Most Brits are in favour of cannibalism’.

Those who say they prefer ‘no deal’ to a ‘bad deal’ might simply be able to imagine a very bad deal indeed. It does not follow that they are hoping for ‘no deal’. It would be rational for me to prefer to be diagnosed with testicular cancer (survival rate up to 98%) than pancreatic cancer (survival rate around 1%), but you cannot infer that I want testicular cancer.

Moreover, the choice given was between “no deal rather than a bad deal” and “any deal rather than no deal”, which is begging the question somewhat. Those who prefer ‘no deal’ to a ‘bad deal’ are behaving more rationally that those who prefer ‘any deal’ to ‘no deal’. This is because ‘any deal’ is an infinite set which must logically include some very bad deals indeed.

Equivocal terms

The problem here is that the term ‘deal’ is used in different ways (or is easily interpreted differently) within the same question. ‘No deal’, as we have seen, either heralds a frighteningly solipsistic short/medium-term future of economic isolation or an exciting future of endless possibilities likely to redefine the terms of international trade for the better, depending on your point of view. A ‘bad deal’ is, by definition, bad. This is why the only rational answer must be to prefer ‘no deal’ — even though no deal would be bad, it is possible to imagine a ‘bad deal’ which is worse.

The manipulative term is ‘any deal’. Logically, ‘any deal’ must include the possibility of ‘bad deals’, but that is probably not how we immediately understand it. The standard definition of ‘deal’ is:

An agreement entered into by two or more parties for their mutual benefit.

Therefore, without the qualifier ‘bad’, it is natural to assume that a deal, by definition, is something good upon which both sides can agree. If you strike a deal with HMRC to repay in instalments the tax you ‘forgot’ to pay last year, then that’s a deal. Ending up in court, having your assets seized and being sent to prison, is not. Similarly, there are some ‘bad deals’ which are so bad that they cannot properly be considered deals at all, but rather vengeful punishments.

Is it a meaningless poll?

The above considerations notwithstanding, it is still possible to infer something from this poll result. The problem, as we have seen, is that it is impossible to infer anything with certainty. The result seems to suggest that a majority of respondents feel, perhaps, that a ‘bad deal’ (and a very bad deal at that) is likely, while a minority of respondents feel so confident of a deal which is good that they forget that ‘any deal’ necessarily includes ‘bad deals’. And it tells us nothing useful about the proportion of respondents who would still prefer ‘no deal’ to a ‘good deal’.

What it might reflect, I suggest, is a snapshot of public opinion regarding our relationship with the EU. The majority is convinced that leaving the EU is necessary and unavoidable precisely because the EU (the other party in the deal) is likely to offer only a ‘bad deal’. A shrinking minority (26% in this poll) is still pro-EU to the extent that they regard the prospect of a ‘bad deal’ unimaginable. But it is only the false dilemma posed in the poll question that allows us to draw this conclusion.

The question asked in the poll is logically equivalent to this:

With respect to Brexit, which statement to you agree with?

A: No deal is better than a deal which would be worse than no deal.

B. Any agreed and mutually beneficial deal is better than no deal.

When stated like this, the false dichotomy is laid bare. The rational person (even one who regards ‘no deal’ as a pretty bad outcome) would be able to assent to both (A) and (B). Of course, this would not allow either side in the Brexit debate to make the claims they would like to. “Most people think a good deal is better than no deal, but no deal would be better than a deal which is cruel and punishing to the UK,” might well be true, but it generates neither hashtags nor headlines.

For a poll of this kind to be meaningful, and useful, the choices (that is, the terms) must be clear and unequivocal. First, there ought to be some greater clarity about what ‘no deal’ actually means. Does it mean trading with the EU like Australia or China, i.e. without a comprehensive FTA but with other agreements in place (e.g. concerning non-tariff barriers to trade)? Or does it mean trading with the EU sui generis, with no agreements of any kind in place?

Second, terms like ‘bad deal’ and ‘any deal’ are too wide in scope and therefore next to meaningless. I can imagine a deal so bad that ‘no deal’ would certainly be preferable. But a serious answer requires a clearer definition of what this notional ‘bad deal’ actually entails. ‘No deal’ might be better than ‘bad deal X’ but worse than ‘bad deal Y’. Until we know what X and Y are, it is an impossible question to answer.

Greater clarity is also likely to mean less unanimity of response. As a thought experiment, let’s take a few (logically possible) ‘deals’:

A1. The UK continues to participate in the Single Market but must be part of the Customs Union.

A2. The UK continues to participate in the Single Market as now so long as Freedom of Movement continues.

A3. The UK continues to participate in the Single Market as now, and while Freedom of Movement must continue in principle, the UK may set limits if it is politically or economically necessary.

A4. The UK continues to participate in the  Single Market with no requirement for Freedom of Movement to continue.

As far as I am concerned, only A1 is a ‘bad deal’. Trading across borders is a function of the Single Market, not the Customs Union, whatever Chuka Umunna might think. Being part of the EU Customs Union, though, would prevent (or at least hamper) the UK striking her own trade deals (as Norway, Iceland, Switzerland and Liechtenstein are able to do). A4 is pure fantasy, but it’s still a deal I’d be happy with. A2 is basically Norway, and A3, you might say, is Liechtenstein, so I’d be happy with those too.

However, there is a minority of Brexit supporters (probably a smaller minority than many suppose) which regards any sort of continued free, or free-ish, movement to be a ‘betrayal’ of Brexit. It’s no such thing, of course — Brexit, leaving the EU, is about no longer being a member of the federalist, supranational, political entity of the EU, not the European Economic Area (as I argue this long and ponderous article). But for some, preventing the citizens of EU countries from coming to the UK to contribute to the economy with their labour and taxes is a top concern. British unemployment for British workers is a price worth paying to ensure that we only have British jobs for British workers.

A4, you might imagine, would be acceptable to almost everyone. It is pretty close, after all, to the politicians’ fantasy of ‘all the benefits with none of the disadvantages’. And yet even here there are those who would regard this as a ‘bad deal’ because they have decided that leaving the EU is not enough. Rather we must purify ourselves from its taint by withdrawing from all of its associated structures. They were convinced, perhaps, by the Prime Minister’s dubious assertion that Brexit ‘cannot mean’ membership of the Single Market.

Brexit is not a thought experiment

This poll, and others like it, have asked people to choose between options which are all but meaningless. The referendum result was close, and the reasons for voting one way or another were varied, but since then there has been a shocking lack of clarity on the part of politicians. UKIP has shown itself to be more of a xenosceptic party than an ‘EU-sceptic’ party. The Liberal Democrats’ only plan is to somehow reverse the referendum result and slip back into full EU membership, despite public opinion moving in the opposite direction. The Labour Party are confused to say the least, and even when the odd MP manages to say something sensible about the importance of participation in the Single Market they tend to betray their misunderstanding of the facts by adding, ‘and the Customs Union’. And the Conservative Party, the party that is actually (just about) in power, is no better.

Mrs May, like all politicians nowadays, likes to use phrases like ‘let me be clear…’ but has been anything but clear. I believe her when she says she wants to strike a ‘good deal’ for Britain, but I do wonder whether her reticence over the specifics is due to her wanting to play her negotiating cards close to her chest, or due to her not having a clue what she thinks. Again, I suspect she honestly believes that ‘no deal’ is preferable to a ‘bad deal’, but she must at least have in mind some definition of a ‘bad deal’ beyond which she is not willing to go. At the moment, the government’s definition of ‘good deal’, ‘bad deal’ and ‘no deal’ are etherial and plastic. ‘Good deal’ means something to Mrs May, presumably, but it probably means something different to me, and something very different to Mr Farage.

As long as Mrs May talks of ‘good deals’ and ‘bad deals’ then no one can disagree. A ‘good deal’, is by definition, good, and a ‘bad deal’ is bad. Only when she begins to spell out the features of an agreement that would make it (in her terms) ‘good’ or ‘bad’ can we begin to agree or disagree. Only then can we begin to form a judgement as to whether this supposed ‘good deal’ or ‘bad deal’ would be better or worse than ‘no deal’.

Imagine someone asks you whether you’d prefer your living room to be painted in ‘Babouche’ or ‘Vardo’? These are real colours on the Farrow and Ball colour chart, but unless you see the chart, or at the very least know that ‘Babouche’ is a sort of orange and ‘Vardo’ is a kind of teal green, then you cannot possible make an informed choice between them. It would be little more than a guessing game.

If the government, and the pollsters, really want to take the temperature of public opinion, then there must be agreement about terms, definitions and even — whisper it — facts. Without this level of basic clarity, then we are engaged in little more than imaginative wordplay; no more than a particularly dull and depressing game of Would you rather…