Spend 45 years arguing for rejoining the EU if you like, but we must leave first

I wrote a whimsical piece for CapX on Tony Blair’s pointless speech on trying to get the British public to rise up against Brexit. It seems to have created rather more excitement than its mischievous nature and its banal statements of the obvious justified.

I pointed out that the debate on whether we should leave the EU is over. It’s been over since June 23rd, 2016. A remarkable number of folk seem to want to keep fighting the EU referendum. It’s almost as if arguing about whether we should leave the EU has become like some addictive computer strategy game or some slot machine targeted at problem gamblers. Somewhere between January 2013 and June 2016, many folk argued about Brexit one too many times and crossed over into a place where they now have a compulsion to keep doing it.

I don’t need an argument that the debate about whether we should leave the EU is over. The result of the EU referendum was a fact not a debating conclusion. Many of you say: “We’ll keep arguing for 45 years, like those who wanted to leave the EU did.” Well, first, many of the most high-profile folk did not argue that we should leave the EU for 45 years. Many of them believed we were better off in until around 2012. And no, that wasn’t a matter of their changing their minds and deciding they’d been wrong all along. It was a matter of the world changing. It is perfectly possible that the UK gained from being an EU member in 1989 and will gain from leaving the EU in 2019.

But, setting that point aside, there’s nothing to prevent folk from spending 45 years arguing we should rejoin the EU after we’ve left. Some people say: “Advocates of leaving the EU didn’t accept the 1975 result”. But in fact virtually all of them did. I’ve never heard of anyone insisting on paying the Common External Tariff voluntarily to the EU authorities when they purchased something from France. They accepted that the result of that referendum meant we continued as members until such time as they won some vote – e.g. a General Election or another referendum – that would legitimize leaving.

So feel free to keep arguing that we should rejoin the EU for the next 45 years. But we must leave first before rejoining. That was what the EU referendum decided. If folk can’t get what they want by voting for it and winning, what do you suppose will happen?

Suppose there were a General Election and Corbyn’s Labour Party won, but then bureaucrats and former MPs got together and said “Corbyn’s a bad idea, so he can’t be in govt even though he won”. What do you suppose would happen then? I’ll tell you: insurrection.

If voting for things and winning doesn’t get those things done people will decide that voting for things isn’t the way to bring about the changes they want and they’ll seek other, distinctly less amiable means to do it.

If Corbyn were to win, we’d have to put up with the disaster until his government collapsed, but we’d have to have that government first. If you think Brexit will be a disaster, you’re entitled to your view and if life outside the EU proves unattractive there will be those making a case to rejoin. But we must leave first.

Anything else would constitute an utter catastrophe for our whole political system, probably bringing about its destruction. That so many smart folk can’t see that is extraordinary.

Who will pay a US tariff on Mexican imports?

Many folk who know a little bit of economics will be familiar with the gist of the following result (if not every detail). Suppose I were a small country – say New Zealand. Suppose that as we began, no-one in the world imposed any tariffs on anyone else’s products. And suppose that in my country my firms sell some of their product abroad and some domestically, with none of it being a large proportion of world output, and there is no product for which my country’s consumers constitute a large proportion of world demand.

Then in such a case all products sold in my country will be being sold at prices set by world supply and demand – let’s call those “world prices”. There is no product for which my producers selling less or more would affect world prices, and ditto for my consumers demanding less or more.

Next, suppose I introduce a tariff on exports from abroad. What will happen? Well, world prices aren’t going to change. So anything imported will now cost the world price plus the import, but anything produced domestically won’t. If my domestic producers compete against each other and have costs at the international frontier (i.e. are the most efficient firms in the world), then domestic prices will rise fractionally, giving my firms just enough incentive to sell domestically (for slightly higher prices) than on the world market (at the old world price). In that case, no tariffs will be paid (since imports fall to zero).

On the other hand, for any markets where domestic production is not perfectly competitive or where domestic firms are not the most efficient in the world, prices will rise. If they rise to a level between the world price and the world price plus the tariff, there are still no imports and no tariffs paid. My domestic firms may gain slightly, as with the same total capacity they can now sell to the domestic market at a higher price and to the world market at the world market price. But there will be a net loss, because of the losses to my consumers.

Lastly, for some markets there might be no domestic production (or insufficient capacity to produce) at prices below the world price plus the tariff. In such cases prices rise by the level of the tariff, imports still occur, and consumers buy (less than before, but still some) imports at the world price plus the tariff.

In such a case, any tariffs that end up being paid are paid by domestic consumers. Foreign producers lose nothing, because their prices are fixed by world supply and demand and the reduction of New Zealand demand because of the tariff is not sufficient to shift the world equilibrium price.

This is the kind of case folk implicitly have in mind when they say that “domestic consumers pay tariffs”. But does this kind of case work when we think about a US tariff on imports specifically from Mexico to pay for Trump’s wall? Can we say, as many folk have said today: “Trump is wrong to say that will be Mexico paying for the wall – instead it will be US consumers”?

The first thing to note about this sort of tariff is that it is not a tariff on all imports, only imports from a particular country. By itself, that may not make much difference. Think of our New Zealand case again. Suppose New Zealand imposed a 20% tariff on imports just from Chile and that there was no market in which Chilean products constituted a large part of world supply (ie no market where removing Chilean output might leave a supplier from another country as a monopoly supplier into New Zealand).

In that case all that will happen if there is a tariff on Chilean imports is that Chileans will sell nothing to New Zealand but more to the rest of the world, and suppliers from the rest of the world will sell more to New Zealand, exactly replacing the Chilean output. No tariffs are paid.

But now let’s take another case. Suppose that the country was not a small one like New Zealand, but, instead, a large one like the United Nations, and that the country on which the tariffs were imposed were a small one, like North Korea. In that case, all of North Korea’s exports go to the United Nations and if North Korean firms tried to charge extra, United Nations competitors would undercut them. So in that case, the tariff is not paid by United Nations consumers in higher prices but is, instead, paid by North Korean firms and workers in lower wages and lower profits.

Which of these cases more strongly resembles a US tariff on imports from Mexico? Well, the US is far and away Mexico’s largest export market, taking around three quarters of its exports. And the US is a very large market, with many domestic firms. It is plausible that for a wide range of Mexican products, Mexican exporters would not have scope to switch to exporting to another country or to raise their prices in the US market materially, and certainly not all the way to offset a 20% tariff. Rather, what would happen is that those Mexican producers, if they wanted to continue to export at all, would have to pay the tariff and bear making lower profits or, to the extent they could, pass on the tariff to their workers in the form of lower wages and perhaps lower prices paid to other suppliers (though it is worth pointing out that some of those suppliers might be US firms – e.g. to the extent that Mexico is assembling cars from components produced in the US).

Eventually, the Mexican economy would adapt, presumably by radically reducing its export dependence on the US. But for some time, Mexican exporters probably would be paying the tariffs and US consumers wouldn’t. My guess is that, by and large, Trump will be able to say that a differential tariff, applied specifically to Mexican imports, does indeed result in extra money paid from Mexico to the US that Trump could then use, for good or ill, to fund his wall.

Partial Determinism (1995)

Dear Edmund,

You have heard me discuss partial determinism a number of times. Now I wish to explain it in some more detail. In this letter I propose, in the main, to lay out what the theory is, rather than come up with arguments to recommend it other than its power and elegance. Of course the power and elegance of a theory are always its most persuasive features, rather than, pace you, any empirical evidence in its favour. Furthermore, in many cases a view does not stand in need of defence, but attack. For example, it is not an idea that needs defending than snow is white. If you wish to disagree with this, you do not merely need to attack any reasons I might put forward in its favour. Almost the whole burden of proof lies on your side, to display somehow that snow is not white. That is, that snow is white is not a view that needs defending, but a view that needs attacking, if one wishes to hold a different view. Now in the matter of partial determinism I do not suggest that as a whole the view is one that needs attacking, not defending, but rather that all its most controversial elements are of this nature. This may seem obscure at the moment, but I hope all will become clear shortly.

What is partial determinism? Briefly, partial determinism is the claim that at any particular time the truth values of some propositions in the future tense are defined, and those of other propositions are not. Propositions about the future would include: “The sun will rise within the next twenty-four hours”; “I shall never talk to you again”; “The Labour Party will win the next General Election”. If the truth value of a proposition is defined then either the proposition is true, or the proposition is false.

Now to be clear: I am making an ontological claim, not an epistemological one. I am not saying that there are some truth values that it is impossible for us to discover, and hence that they are ‘undefined to us’. That is, I am not saying that although all well-formed propositions about the future have defined values, there are some, indeed possibly all of them, that we can never discover. No. I am saying that some propositions about the future as yet have no truth value. Furthermore, because the theory is partial determinism, this is not because it is inherently impossible for a proposition about the future to have a truth value, since I am making the additional claim that some propositions about the future do have a defined truth value. Thus on the ontological question there are three possible positions (among those who believe in bi-valued truth functions for propositions): Every proposition about the future has a defined truth value; No proposition about the future has a defined truth value; Some propositions about the future have a defined truth value and some do not. The last is my view.

Here is a statement for which I suspect the truth value is fixed: “The sun will rise within the next twenty-four hours.” Here is another: “The Liberal Democratic Party will not win an overall majority in the 1997 British General Election.” Please note that the first case is a scientific statement, and implies that it is not merely a remote possibility, but in principle impossible, that random destruction of matter (which is possible) should destroy the sun before dawn tomorrow. Thus it is a strong claim about the future, and one which some physicists would dispute. Nonetheless I suspect they would be wrong. The second statement involves human decisions. Some people would like to think that millions of people could yet change their minds and vote for the Liberals. I am saying that this is not merely unlikely, but impossible. I would not make the same claim about an election after, say, 2001. Now note immediately a property of statements about the future under partial determinism. I claim that the truth values of these two statements are fixed. That is a strong claim. I am not merely predicting that the sun will rise within the next twenty-four hours. I am not just saying that this is overwhelmingly likely. I am saying that this is already true.

Now to some statements which I suspect are not fixed: “There will be a hurricane in North Korea in the year 3240 AD.” and “Bill Clinton will have porridge for breakfast on December 4 1999.” Now I could be wrong about either of these. Perhaps I need the year 324444 AD for the North Korea Hurricane question to be undefined. Perhaps Bill Clinton is allergic to porridge and never eats it. I don’t know. But these are the sort of statements about the future that I suspect are undefined. Again the first statement is a scientific one. What I am suggesting is that quantum events whose resolution is as yet fundamentally undefined (as shown by the Bell Theorem) could, given long enough, through minor perturbations create the well-known butterfly effect upon the weather. Thus the truth value of this macro weather statement is, as yet, fundamentally undefined. I do not believe that the same can be said of hurricanes which will strike tomorrow, or even next week. (There are other mechanisms which might make the weather undefined quite separately from quantum events, of course, but quantum events suit my purposes here.) The second case again involves human choices. Perhaps Bill Clinton will be offered a genuine choice for breakfast that morning, or perhaps his breakfast will be imposed by purchasing decisions the previous week (e.g. they might have run out of all other cereals and he just won’t feel like eggs that morning) but either way I am suggesting that what he will have for breakfast is not yet defined.

Now to something rather pleasing one can do with partial determinism. Consider prophecy. Some people claim to be psychics or seers or prophets, and to know the future. Such people are often mocked by being asked questions like “So who’s going to win the 3.20 at Newmarket?” or “So why haven’t you won the Lottery, then?” Even when prophecies seem to be fulfilled (e.g. much of the book of Daniel) there is a decidedly vague sense about the prophecies. Answers often given about this vagueness include “Well, that is all God revealed” or “To tell people more would be dangerous” or “The disciplines of my occult art forbid me from dabbling in the Lottery.” Partial determinism offers a rather more straightforward solution. Prophecy will always be vague because the future is only vaguely defined. Even the greatest prophet might not be able to win the Lottery because its result may not be defined.

What of God’s Omniscience? Well, under partial determinism an Omniscience God would know everything there is to be known, and so would know the value of every truth function and know the likelihood of each resolution for all undefined truth functions.

This leads us into one of the classic arguments against partial determinism (or against total indeterminism). Some people believe (or at least believe that they believe) that God is outside time and can observe our future as if it were our past. He can see everything that will ever happen, or has ever happened. Since God can see all the truth values for every proposition throughout time, all their values must be defined.

Now I don’t believe that any part of this is coherent, but for now I shall restrict myself to showing that it is irrelevant. In the argument above it was stated that God is outside time. But ‘is’ is a word in the present tense. So presumably, since God is not supposed to be inside time, and ‘God will be outside time’ and ‘God was outside time’ don’t seem any more promising, something else must be meant. If something else entirely is meant, so the argument is not supposed to be that God, at the present, can see all the truth values, then that’s enough for me now. I just want to claim that at the present, some truth values are defined and some are not. Perhaps someone will claim that what is meant is that ‘God was outside time, God is outside time, and God will be outside time’. Inspiring though I am sure this statement is, it does not get us anywhere. To say that God was outside time, just means that at some point in the past, it was correct to say ‘God is outside time’, which, if interpreted temporally, is just a flat contradiction. Now this may seem unfair. There is a plausible sense in which one could use the predicate ‘is outside time’. For example, one might say ‘Love is outside time’, which is presumably to say that Love does not depend on any particular cultural or temporal context. Love is Love is Love, as it were. Or Love is eternal. Similar sentiments might be expressed about Courage or Wisdom. Now if that is what someone means by ‘God is outside time’, then fine. However, it should be observed that this makes it entirely the wrong sort of statement to use in arguing against partial determinism. As I said above, I just want to claim that at the present, some truth values are defined and some are not, and if the argument is not supposed to be that God, at the present, can see all the truth values, then it is not an argument against me.

Someone has argued against me that the question of whether the truth values of propositions about the future are all defined is a separate matter from whether it is the current state of things which defines them. That is, to accept that all the truth values of propositions about the future are defined is not to believe in determinism. I accept that logically there are two questions, but I reject the claim that they are separate questions. If it is not the current state of things that gives the propositions about the future their truth values then what is it? Perhaps some people reserve the term ‘determinism’ for the view that the state of the universe as understood by current physics (and some contiguous development of it) determines all future states. But I would use ‘determinism’ in a slightly wider sense, for the claim that it is the current state of things, including the physical state of the universe, the nature and decisions of gods, and anything else which has happened so far, which defines all futures states of things. Thus, for example, ‘determinism’ would include ‘predestination’. But what is it supposed to mean to claim that the truth value of a proposition about the future is defined but it is not the current state of things which defines it? In what sense is this a truth value?

What of the other two possible points of view? Some people seem to feel that Determinism is obviously correct, in fact so obviously correct that it is not a theory in need of defence. Thus the elegance or power of any alternative is irrelevant. What would be required is an argument against Determinism, and not just an argument which might offer some undermining of Determinism, but one which is so persuasive and obvious that it is irresistible. Thus such people are claiming that the truth of Determinism is as obvious as the whiteness of snow or the redness of blood. I find this view most peculiar and can only conclude that its adherents are suffering from a diet of too much Classical and Relativistic Physics and too little casual living. How can anyone claim that it is overwhelmingly obvious that no-one ever makes a genuine decision, that nothing random ever happens, and that the future has no status different from that of the past? Surely even if these suggestions are coherent (which they most emphatically are not!) then they are profoundly counter-intuitive and as far from obvious as it is possible to be! In any event, Determinism is analytically false. That is, it does not have the status of a theory, since it not the sort of thing that could be proved by any evidence. It is not merely unlikely. It is completely impossible.

Surely the most likely start point is actually that nothing in the future is defined, and the interesting thing to do is to show how things could be otherwise. I consider it a great wonder that the tentacles of the present can reach into the future, but I believe that it is likely that they do. Why do I consider that the natural (though incorrect) intuition is that the future is totally undefined? At its most basic, the instict is that what has happened cannot be changed, but the future is precisely what can be changed. Time is an ordering on events. One attractive ordering relation is through causality (with essentially an unrestricted version of Einstein’s definition of time). Event B precedes event C but comes after event A if event B could, in principle, influence event C, but could not, in principle, influence event A. If events X and Y could not, even in principle, influence each other, then it is tempting to say that they are simultaneous. But this last claim seems to result in the rejection of partial determinism. Suppose I believe that the sun’s rising tomorrow is defined. Then nothing that happens today (outside fixed elements in the causal chain which leads to the sun’s rising) could, even in principle, change that. But when the sun rises tomorrow that event will not be able to change anything that happens today. So it seems that if partial determinism were true then everything that happens today would be, in some sense, simultaneous with the sun’s rising tomorrow. Thus the sun’s rising would not be an event in the future which is already defined today, but rather an event which is somehow already occurring today.

Is this argument flawed? or is it a powerful argument against partial determinism? In fact, neither! I take this argument on the chin, and claim instead that this is the great paradox of partial determinism: some events in the future have, in a sense, already happened. That is precisely what partial determinism is claiming. But can we go one step further, and make the claim of Determinism: that all events in the future have, in some sense, already happened? No! For this is to reject the existence of the future entirely. The future is precisely that set of times where not every well formed proposition has a defined truth value. To accept Determinism is to say that everything has already happened, but that is to put everything in the past and reject the existence of the future, which is obviously nonsense.

That will do for now. Any comments?

Andrew Lilico

10 Predictions for 2017

  1. UK GDP will grow by around 2%, confounding fears not only of recession but also of any material slowdown in growth. The pound will appreciate, on a trade-weighted basis, across the year, not depreciate. Inflation will exceed 2%, but only modestly so.
  2. Article 50 will be triggered on schedule by the end of March, without there being a General Election first. Article 50 triggering will not lead to any material market turbulence or GDP growth slowdown. It will be almost a non-event.
  3. Ukip will come within a whisker of winning the Copeland by-election, but Labour will hold on. Opinion polls will show Labour frequently under 25%. A poll showing below 21% will be written off as a blip (but will actually be the most accurate poll conducted). The Conservatives will consistently poll above 40% and frequently above 45%. The Lib Dem vote share will not recover materially.
  4. Political debate will be utterly dominated by Brexit, swamping all other issues. At around the time of Article 50 triggering, the government will state formally that it has no intention of or interest in remaining in the Single Market, Common Agricultural Policy or Common Fisheries Policy, but that it does not regard there as being one “customs union”, indicating instead some interest in exploring the possibility of keeping a customs union for cars. There will be very little publicly agreed in the Brexit negotiations themselves beyond some initial mutual recognition of the future status of EU citizens in the UK and UK citizens in the EU.
  5. The French, Dutch and German elections will be dogs that didn’t bite, returning establishment governments. Greece-triggered and Italian-related issues will lead to market nervousness, but no country will leave the euro next year.
  6. In what will be hailed as a surprise move, the governments of Poland, the Czech Republic and Sweden will call for renewed public debates about whether they should join the euro, including an intention to hold a referendum on the issue within the following two years.
  7. Trump will abolish NAFTA, but maintain the US trade deal with Canada. Trump will offer the UK a trade deal, seeking to have the principle agreed (albeit not all the details) in 2017 or early 2018. The UK will say we are unable to negotiate at this time. Trump will shrug and move on and the UK will miss out on a big strategic opportunity. We will not finally secure that deal with the US until the mid-2020s. In the meantime Trump will significantly raise US tariffs on imports from the rest of the world (including the UK). This will start to significantly affect UK exports to the US in 2018.
  8. Putin will be left to return order to Syria. Western leaders will express disappointment at his “brutality” and call for his actions to be “measured, proportionate and humane”, but will be secretly pleased that they don’t have to do it.
  9. Trump will announce a new venture to land humans on Mars during the 2020s. In his speech he will make a joke about building one of his hotels there.
  10. 2017 will see the first significant sales of self-driving cars in the UK. Uber will trail the forthcoming launch of a driverless car product. The public will become aware of the concept of self-driving freight lorries. Virgin Galactic will announce a concrete date for its first commercial passengers launch (but that date won’t be in 2017). Progress will be reported on the project to re-program cancer cells, but finalisation will still be some years away.

If I must…


Andrew Marr wrote a piece. It didn’t pretend to be a work of genius, but I quite liked it nonetheless. As I read it, I took it as saying that although Marr had been a Remain voter (or perhaps abstained), since Brexit is now inevitable he wanted to have a go at thinking about how one could make the best of it from a left-leaning point of view (it was, after all, published in the New Statesman). I, as you would expect, disagreed with many of the left-leaning suggestions in there. There were also a couple of things Marr put rather loosely, with a Marr-esque flourish, but I understood that as style. When newspaper or political articles try too hard to be absolutely technically precise on every point at the expense of the odd rhetorical flourishing hand-wavy “you-know-what-I-mean” they quickly become dull and unread.

What I liked about it, and what I took to be almost the whole point of the piece, was its spirit. Its message was: “Brexit is going to happen, whether we like it or not, so we should think about how we can turn it into an opportunity – and here are a few ideas about how a centre-leftie might try to do that.” Good for him.

Alas, in certain parts of the Remain movement, this sort of constructive thinking is utterly verboten. To actually make some suggestions for a way forward post-Brexit! How dare you! That suggests you are accepting that it’s actually going to happen! We can’t allow that!

One standard tactic in such cases is to (probably deliberately) completely miss the point of such an article and instead to purport to point out “technical errors” that supposedly render any attempt to think constructively about Brexit impossible. It is not permissible to suggest that there could be anything, of any positive nature, that could be done post-Brexit that could not already be done better without leaving the EU.

Even if that were true – which it manifestly isn’t – it’s now irrelevant, because we are going to leave the EU.

Jon Worth wrote a (sadly rather long) piece which he claimed “took apart” Marr’s “argument”. I hadn’t been aware of Marr “arguing” anything other than that we ought to try to make the best of Brexit since it’s inevitable. Does that really count as an “argument”? What alternative could there be? Making the worst of it? And how does one “take apart” someone urging you to make the best of a situation whether you like it or not?

Jon was very pleased with his piece, and so offended that Iain Martin and I didn’t find it to be a masterwork demonstrating that Brexit is a terrible idea after all and we should call the whole thing off that he decided to write another piece complaining about us. To be fair to him, he’s not the only person to admire his piece. Many of the smart Remainer folk seemed to have liked it as well.

I wouldn’t have bothered to respond to his piece, since I don’t think it said anything much, but I need something to do for a few minutes to distract me from my cluttered twitter notifications timeline which is full of people complaining at me for not responding to Jon. So here goes…

By my count Jon gets 1,836 words into his post before he says anything that is even arguable, let alone correct. Most of the first part is taken up with his saying we might yet not leave the EU (tedious), that we might yet stay in the Single Market (good luck with that) and that Marr is obviously wrong to suggest that May’s establishing a Department for International Trade and saying in numerous speeches that she is seeking new trade deals means that May’s government is seeking new trade deals. Jon refers to this as “among the weirdest and most deluded parts of Marr’s piece”. What am I supposed to say to that?

Eventually, two full long newspaper articles-length into his piece, Jon finally gets to something amounting to a claim. Marr said that rail renationalisation is “back in play”. I suspect that Marr means both that it’s politically feasible in a way it wasn’t before and that the implementation of renationalisation will be easier than it would have been if we’d remained in the EU.

Ah, says Jon, but EU rules didn’t forbid nationalisation of the railways! “As for rail, this is an old chestnut – all the EU requires is the separation of network from operations, but both of these can be in the hands of the state.” Well, yes…in a sense. But the EU does require that operations compete with each other and are exposed to cross-border competition as well. So within the EU, “nationalisation” wouldn’t mean what Jeremy Corbyn would think it meant. And no political party in the UK would support nationalisation of the railways if it meant competing nationalised franchises competing with foreign franchises. So Marr’s at least arguably right in practice. But we can give Jon an “ever so clever, if not very useful” debating point here if he wants one.

Then more guff, then some pointless stuff attacking Marr for wanting a trade deal with the US (which Jon doesn’t actually oppose), then some stuff trying to appear clever about trade deals, then some nonsense about the FCO. Finally we come to another half-point: Marr made a passing remark about the funding of schools. Jon says: “Funding of schools? The EU has nothing to do with the funding of schools – education is a national competence.” I have no idea what Marr was referring to. I have a suspicion he just felt he should mention education somewhere as lefties like that sort of thing. Let’s give Jon that one (he sorely needs it).

Next Marr had said some pretty anodyne things about countryside management. Jon doesn’t like it, even though it’s blatantly obvious that when we leave the EU we will leave the CAP and that that will mean some changes to our rural policies. So Marr’s obviously right on the big picture. If Jon can’t see that, I can’t help him.

Similarly, we’ve obviously going to need some kind of replacement for the CFP. I don’t think Marr advanced thinking on that terribly much, but then I don’t think he was intending to. He was pointing out that there’s an issue there and floating a few thoughts to get us going. Why Jon felt that was something to “take apart” I have no idea.

Similarly, once more, we’re going to have control over our own VAT rate post-Brexit. Marr’s not wrong about that. But Jon appears so excited that he can spell “recupel” that he seems unable to grasp what Marr is saying.

Thus, but my count, in his epic 4,798 word vent – a good ten times the recommended length of an article on Iain Martin’s excellent website – Jon Worth manages one actual point (the EU doesn’t materially restrict the funding of schools, as far as I am aware) and one half-point (EU rules don’t strictly prevent renationalisation of the railways in theory; they merely do so in any plausible UK practice). In delivering this cornucopia, he magnificently fails to engage with, or even in any obvious way grasp, what Andrew Marr’s thoughtful and constructive (albeit certainly not world-shaking) piece was about, namely that Brexit is definitely going to happen; we’re going to be outside the Single Market and the Customs Union; so if centre-left thinkers want to be relevant to the actual UK debate then they need to come up with some thoughts on how to make the best of that.

On transitional post-Brexit deals

Brexit Kremlinology & rune-reading has a new object: a transitional deal. Apparently such a deal is “absolutely essential”, or “politically unthinkable”, or “both if we’re going to do Brexit at all so let’s not Brexit”.

Instead of a transitional deal, let’s instead think about transitional deals across various sectors, and what they might or might not be worth. I say we should think of four such sectors:

  • Agriculture. Here we’ll almost certainly get no final deal with the EU and no transitional deal, and wouldn’t want a transitional deal either.
  • Goods. Here we should be able to cover almost everything as part of the Brexit deal itself, so the issue of transitional deals doesn’t really arise.
  • Financial services. Here there will obviously have to be some kind of transitional deal, as much for the EU’s sake as for ours. The very minimum that could cover would be issues such the transition from being firms located in a member state of the EU to being firms located outside the EU who have applied for some kind of third country access approval and are awaiting approval. As things stand, firms applying from outside the EU for certain third country approval have some waiting period (e.g. 180 days), but one can’t apply to be recognised as being in a non-EU state whilst still in an EU state (which the UK would be until it left). A more extensive “transitional deal” in financial services might be some arrangement, to be reviewed after say 10 years, whereby UK financial services firms that want to stay subject to EU rules can opt into being so and thus stay in the Single Market for Financial Services. In practice, we’d probably find that after 10 years everyone was happen to extend the arrangement for another ten years and so on ad infinitum.
  • Other services. Here there could be an argument for a bridging arrangement to cover the period between when the UK leaves the EU and when the EU and UK ratified a services trade agreement (or services schedule added to the EU-UK good trade agreement). Services are unlikely to be included in the Brexit deal itself, because then the overall deal would probably require individual chamber ratification and we saw from CETA that that could end badly. So we might not get a services deal until 2023, say. Such a bridging agreement would have an obvious purpose, but not that much value. The services aspects of the Single Market aren’t believed to boost UK-EU services trade all that much and it’s not clear that things would change that much for most services firms if there were no trade agreement with the EU whatever. For example, at present there’s no services agreement with the US, but does that stop US advertising firms or design agencies or software programmers from offering services to EU firms?

I’m not convinced that any of these transitional arrangements (excluding the Financial Services ones, which aren’t really “transitional deals” in the relevant sense of being a temporary or bridging deal) is really all that worthwhile. One that might just be worth considering would be the UK staying in the EEA (and thus continuing to accept free movement) until just before the General Election – so, around one year after the end of the Article 50 process. The main advantage of that, apart from adding one extra year to negotiating time for post-Brexit deals, would be that it could provide a pretext for the UK continuing to contribute to the EU’s budget until 2020. Otherwise, if the UK were to withdraw early, that could create all kinds of problems for the EU in 2019/20. Without the UK’s contribution, would EU Member States have to receive less in structural funds, or would other programmes be cut, or would the Germans be asked for extra money? No-one will want to know that. Better if the UK stayed around giving money for one more year to smooth all that over. Provided free movement was at an end before the General Election, the government should be able to get away with it.

The other variant of that would arise if Theresa May is “forced” into calling a General Election in early 2017 – say, because the House of Lords rejected her Article 50 bill after the Supreme Court said it was necessary. Then (sob!/snigger!), armed with her 120 seat majority, May could do what she liked until 2022 – including, for example, staying in the EEA until then. Free movement need only be gone by the end of the Parliament that commenced before Article 50 were triggered. Staying in the EEA until 2022 would also give us three years post-Brexit – so, five years all up including from Article 50 triggering – to agree and ratify a services trade agreement. That might well be feasible.

Watch this space.

Rational expectations and the short- vs long-term impacts of Brexit

Imagine a very simple economy, growing over the long-term at 2.5%, in real terms, each year. Let’s suppose that everyone expects it to continue to grow at that rate in the future. Let’s suppose that the way people decide how much to consume each year and how much to save means that they seek to smooth their consumption perfectly over their eighty year lifetimes. And let’s suppose that the population is stable, with 1/80th being born and 1/80th dying each year.

Now, let’s next suppose that something unexpected happened that meant that instead of the economy actually growing at 2.5% over the next 15 years, as everyone had assumed up to now, it switched to growing at 2.0%, reverting to its 2.5% rate thereafter. Then people’s expected lifetime incomes would be lower. So they would consume less each period. By the end of that period, consumption would be 6% lower. Roughly speaking, that’s the kind of order of magnitude of effect most of the economic models of the long-term impact of Brexit produced during the EU referendum campaign suggested.

But here’s the thing. If consumers really expect to have lower incomes in the future, they will start cutting back on their spending today. We shouldn’t over-literalize that word “expect”. No-one producing economic models thinks that means that if you did a survey of households they would tell you that yesterday they expected their lifetime incomes to be X and today it’s Y. What it means is that the economy functions as if they knew that – that, embedded or implicit in the vast complexity of prices and quantities and employment levels and investment costs in an economy, there are a set of “expectations” about the future. And those implicit expectations embedded in prices, quantities etc are the most-nearly-correct ones, regarding the performance of the economy, that it is possible to have. (This assumption is called “rational expectations”.)

So, if some shock occurs that means the economy will now grow less in the future than expected, that means consumption ought to fall today, adapting to the new reality. In fact, relative to today’s economy, the drop in percentage terms will only be a bit less than the whole drop over the period. So, for example, in the simple model I describe above, the drop in consumption today will only be a little more than half the drop in the future in absolute terms, but today’s economy is also smaller, so that is still a drop of about 5% in today’s economy.

If we were to introduce a few more realistic assumptions, so that people did not smooth consumption quite so perfectly, we’d quickly see that 5% drop today turn into a 3% drop, but the basic principle is the same: in an economy with rational consumers and investors, where the economy was reasonably competitive and efficient and financial markets worked reasonably well, if GDP would be 6% in 15 years’ time, we’d expect that to trigger a drop in consumption (and hence GDP) today, with the impact being felt immediately and the scale of that impact being quite material (perhaps something like a 3% drop in consumption).

That sort of effect was one of the main drivers of the Treasury’s predictions about “the immediate economic impact of leaving the EU” (its short-term Brexit impact study). It was the effect the Treasury’s report quoted first – calling it “the transition effect”. The Treasury state: “HM Treasury analysis: the long-term economic impact of EU membership and the alternatives demonstrated that the UK would become less open, less productive and poorer as a country in the long term following a vote to leave the EU.

The effect of this would start to be felt immediately. Businesses would start to reduce investment spending and cut jobs in the short term, consistent with lower external demand and investment in the future. This transition effect would also lead to lower incomes, reducing household spending.”

It’s also worth remembering that this was what the Treasury forecast would be the consequence of a vote to leave the EU (p7): “a vote to leave the EU would have a damaging effect on both the demand side and supply side of the economy”. Similarly, on p45 we read: “Two years after a vote to leave the EU, GDP would be around 3.6% lower in the shock scenario than following a vote to remain”.

I have no problem with the chain of reasoning here. It seems entirely consistent with orthodox macroeconomic analysis to believe that forwards-looking economic agents would bring forward anticipated long-term future losses into lower consumption and (perhaps to a less extent) lower investment today. And it also seems entirely proper for the Treasury to have expected that the key moment would be the vote to leave, rather than the actual leaving. After all, it’s the vote to leave that changes one’s expectations about the long-term future.

But here’s the rub of this particular article. I was talking with Tim Harford and Chris Giles the other day about what we know of the impact of Brexit so far. I shan’t tell you exactly what we said – you’ll have to wait for Tim’s doubtless excellent show to find that out. But in reflecting on what I said, I think I failed to state something important. The fact that the economy has not reacted negatively to the vote to leave the EU – the fact that the growth rate of consumption and business investment have not fallen, let alone there being any absolute drop of the sort the Treasury predicted – tells us that, embedded implicitly in all those consumers’ and businesses’ and investors’ decisions, the economy’s expectations of the impact of Brexit are for precious little long-term impact either way.

If the economy had expected GDP to be 6% lower in 2030 as a consequence of the Brexit vote, consumption and investment growth should have dropped in the short-term – indeed, probably consumption and investment should have dropped in absolute terms, triggering a recession of the sort the Treasury predicted. But we can see from the economy’s reaction to Brexit that it does not believe that the Brexit vote changes much, either way, about long-term growth prospects. The economy might be wrong about that. But that’s what it thinks. And that’s something we know now that we did not know before the referendum.

What unilateral guarantee for EU citizens could be offered?

During the referendum campaign, Vote Leave said that, in the event Leave won, the UK government should guarantee the right of EU citizens who were here on June 23rd to remain and work indefinitely. The same position was subsequently adopted by Nigel Farage.

There has been no such guarantee offered. The government’s position has been that the position of EU citizens must be negotiated as part of a deal with other EU member states encompassing the rights of UK citizens living and working elsewhere in the EU.

A number of newspapers have been highly critical of the government. It has few defenders on this point. However, one of those defenders, Jonathan Portes (who is almost always worth listening to even if, as I do, you very frequently disagree with him), has raised a point that on the surface might seem to have some merit: What exactly would the government be guaranteeing? He suggests that, since we have no record of which EU citizens were in the country, and under what auspices, prior to June 23rd, we have no way of knowing who was here then as opposed to arriving later, and thus have no way to guarantee that we will not deport anyone who was here legally on June 23rd.

I have previously explained one problem with that objection, namely that we are not requesting a cut-off date of June 23rd, with anyone arriving after that date being subject to risk of deportation but anyone arriving before then being immune. But that is a “negative” point, saying what is wrong with Jonathan’s case. Here, by contrast, I want to offer a “positive” suggestion, namely to respond to the challenge: “Well, if you are not requesting a cut-off date of June 23rd, with anyone arriving after that date being subject to risk of deportation but anyone arriving before then being immune, what guarantee are you seeking?”

Here goes. The UK will be ending free movement of persons with the EU at some point. That means that, as of some date, anyone who arrives in the UK from the EU after that date may not have the same rights to work, to live, to vote and to receive benefits as they would have had before that date. Let us assume we offer no guarantees, at this stage, in respect of voting or benefits receipt. That is to say, we do not, at this stage, guarantee that any EU citizen will, in respect of being an EU citizen (as opposed to, say, having subsequently acquired British citizenship) have the right to vote in UK elections or to receive benefits. There might be such rights or there might not – that would be a matter of negotiation. We might also not offer any guarantees in respect of those that are convicted of crimes that would not, at present, lead to expulsion from the UK but which we might decide should trigger expulsion later.

The guarantee would be in respect of the right to remain and work of those that do not commit crimes. It would be this. We guarantee that, whatever cutoff date is in due course adopted, after which EU citizens entering the UK (and not committing crimes) will not have the same rights to indefinite leave to remain and indefinite rights to work, that cutoff date will be no earlier than 23 June 2016.

This is a guarantee that the UK government has it fully in its power to make, unilaterally, because it is a guarantee wholly about the policy that the UK government itself will adopt. It is not a guarantee that no individual will be deported. Bureaucrats could make errors, individuals could commit crimes that lead to their deportation, and various other complicated scenarios. None of that is relevant to this commitment. This is a commitment wholly about UK policy and thus wholly within the power of UK policymakers to give.

We should do it. We should do it unilaterally. And we should do it now.

Guaranteeing the rights of EU citizens here before June 23rd to stay is not about setting a cut-off date

During the EU referendum campaign, Vote Leave said that in the event of a Leave vote, the government should guarantee the right to remain and work to EU citizens here legally before 23 June 2016. Since the referendum, many commentators have urged the government to pursue this path. According to British Future the vast majority of Leave voters, as well as the vast majority of the population as a whole, favours this being done. Almost the only person that does not appears to be Theresa May. She says that she hopes and anticipates that EU citizens here before June 23rd will be permitted to stay (i.e. not deported) but that that must be a point for negotiation.

I’ve written repeatedly about why May’s policy is a bad one – nay, an outrageous, shameful and offensive one. Here, though, I want to respond to some points made by Jonathan Portes of NIESR. His view is that it is not feasible for the government to guarantee that those here legally before 23 June 2016 will be entitled to reside and remain in the UK post-Brexit. He says we have no way of knowing who such people are, and that even if we did know there would be many borderline cases.

I say: that’s irrelevant and is based on a misunderstanding of what those of us urging a different policy upon Mrs May want. Jonathan’s objections would have some validity if the proposal were that those not here legally before 23 June 2016 will not be entitled to reside and remain. Then we would have to know either who had arrived after that date or who, amongst those here before that date, was not residing here or not residing here legally. In that case saying: “we have no way of identifying them” might be of some interest.

But that’s not the proposal at all. I’ve not said that anyone who arrived after 23 June 2016 should be subject to risk of deportation. Ultimately, if we are planning to end free movement, there will have to be some date after which not everyone arriving from the EU will have a right to reside and work here. We do not need to have decided what that date is in order to guarantee that, whatever it is, it will be later than 23 June 2016.

For example, one way to go might be to say that anyone who is an EU citizen can move to live and work in the UK indefinitely for as long as the UK is part of the EU citizenship area (i.e. part of the EU). That might be seen as most formally according with our obligations under the Treaties. So the actual cut-off date might be 1 April 2019 (if, say, we trigger Article 50 on 31 March 2017, as if widely speculated). In order to make that date enforceable, we might need to do various preparations of the sort Jonathan identifies. Perhaps someone might object that some of the things he suggests cannot be done whilst we remain an EU member. OK, in that case perhaps the cut-off date would have to be later than 1 April 2019.

We do not need answers to all of these questions in order for the government to say, now, that whatever the final cut-off date will be, it will be later than 23 June 2016. The points Jonathan raises, interesting in their own way though they be, are simply irrelevant to that.