Winner of the New Statesman SPERI Prize in Political Economy 2016

Monday, 19 September 2016

In the long run ... our children are adults

One of the annoying aspects of the Brexit debate is that every piece of macroeconomic news, every survey or data point, is interpreted as evidence one way or another about the economic costs of Brexit. The problem with this is partly that Brexit has not happened yet, but more fundamentally the important costs of Brexit were always long term. The Treasury’s analysis of the permanent costs of Brexit looked 15 years ahead, because that is the kind of time period over which the full impacts will be felt.

That has another annoying implication, which is that it will be very difficult to ever know what the actual costs of Brexit will have been. GDP being 6% lower after 15 years (the Treasury’s central estimate based on a bilateral trade agreement) will have a noticeable impact on economic growth, but who knows what the counterfactual is? As I have noted many times, the trend growth in UK GDP per capita was a remarkably steady 2.25% until the financial crisis, but since then productivity growth has collapsed, so who knows what it might have been without Brexit.

It is true that with rational expectations the future will have an impact on the present. That is why the exchange rate fell sharply on news of the vote. As I keep pointing out, unless those in the markets change their minds, that depreciation makes every UK resident poorer, perhaps forever. Equally if everyone anticipated that their future income will be lower they should reduce consumption now. But one reason the vote went the way it did is that many people did not believe that Brexit will have a long run negative impact on their standard of living.

We can make the same point in a more concrete way by thinking about the problem facing the OBR as it makes its forecast before the Autumn Statement in November. Those expecting to see something dramatic in these forecasts may be disappointed, partly because Brexit is likely to actually occur in the middle of the OBR’s 5 year forecasting period. Where the OBR will have to come clean about their view on the long term impact of Brexit will not be until next summer, when it does its 50 year ahead projections.

While these long term costs were always what really mattered, I have the impression (see also Paul Krugman) that the campaign spent much more time talking about short run impacts. As I have argued, these could be significant but are much more uncertain because they are more complicated. (How much will the depreciation boost net exports, for example?) So why was there so much focus on the short term in the campaign, and why did some (mainly politicians) start talking about Brexit creating a short term crisis?

I suspect (and this is just a guess) that one reason is that talk about long run costs had little impact in the media and on voters. (This is what the polls suggest: voters seemed more prepared to agree that there might be short run costs to Brexit.) To an economist that seems odd, because the economics behind the long term impact is much more solid than what might happen in the short run. Of course Keynes had a famous phrase about all being dead in the long run, but as Simon Taylor points out he made that comment to counteract a tendency for some to dismiss problems like unemployment caused by recessions as unimportant because it will disappear in the long run.

One suggestion I have seen links the lack of traction over long run costs to the fact that Leave voters tended to be older, and therefore that they did not care too much about the long run. I think this is unfair: most of those older voters also have children who they care about.

I suspect the problem came from a basic misunderstanding that was deliberately encouraged by the Leave campaign, which is to see all economic analysis as an unconditional macroeconomic forecast. The retort ‘who knows what will happen in 15 years time’ resonates if that is what you are familiar with. Too many people who should have known better, or perhaps chose not to know better, failed to make the distinction between conditional and unconditional forecasts. We had the ridiculous charge that the Chancellor should not have said people will be worse off in 15 years time, because with normal growth in absolute terms they probably will not be.

To see how nonsensical this framing is, think about the advice any doctor will give you that by smoking you will be worse off. Society does not collectively shrug that off by saying who knows what will happen in 10 or more years time. Except of course some teenagers do say this and come to regret it. Nor, by the way, do people tell medics that they failed for decades to predict that smoking would kill people so why should we take any notice now. We are completely familiar with doctors giving us conditional forecasts, but for some reason some in the media kept trying to view any analysis of long term Brexit costs as another unconditional macroeconomic forecast. [1]

One implication of this is that the consequences of Brexit may never become obvious, particularly to those who voted to Leave. Of course economists will do the best they can with the data, but I doubt very much that their analysis will get through to most people. One of the many sad aspects of the Brexit decision is that those who helped make it possible will never be held responsible for their actions.

[1] Note also that these long term Brexit costs essentially came from empirical studies with fairly common sense theoretical content well grounded in evidence. 


  1. Thank you for your refutation of Keynes in so few words. After all, it was he who said

    "In the long run, we shall all be dead."

    1. If you think that is a refutation I've a bridge in Brooklyn to sell.

    2. Consider the context. Just before, he wrote that economists set themselves too easy and useless a task if all they can do amidst a storm is to sit and claim someday the sun will shine again.

      In my humble opinion, his meaning was that waiting for the sun may prove too long and damaging -- in other words, the long run may be very long.

      Professor Wren-Lewis didn't refute Keynes. He just tapped into a popular kind of word play, although must concede that quote from Keynes might not have been a good choice. It's like Marx' statement about religion being the opium of the people -- Marx wasn't calling people dumb, but saying religion is a distraction that prevented people from seeing how miserable they were (in his opinion). Those two quotes, Keynes' and Marx', are very often misused. It is unfortunate given that there is some value in reading those texts seriously, at least in part and at least once.

    3. Stéphane,

      You are quite right, but I couldn't resist the joke. Actually, it was a compliment to SWL for the conciseness and pithiness of his wording.

      The people he refutes are economists and politicians who use the quote as a respectable version of "après nous le déluge" and believe that governments should keep on piling debt without considering the tradeoffs.

  2. The short term consequences also depend on the political response. The Tories had promised to call article 50 the morning after the referendum. That would have caused a much larger adjustment shock than the current situation where article 50 might be called some time in a few years. I still expect that that will never happen; that is a very different situation.

  3. A small twisted part of me is hoping that we get the swift, hard Brexit with no transition mechanism that some of the headbangers are hoping for.
    Just so the costs get front-loaded in a very obvious way.

    1. I doubt that salience is a fundamental problem. When was the last time you heard someone saying the point estimates are just too small. If only studies repported bigger numbers, they'd listen... Or not.

  4. (aggregate) Productivity growth had stopped long before the financial crash.
    The crash was the realisation of what had happened over the previous decade. A catastrophic moment of impact is the inevitable result of a prolonged fall. To understand the impact you need to consider what was happening beforehand.

    1. «trend growth in UK GDP per capita was a remarkably steady 2.25% until the financial crisis, but since then productivity growth has collapsed,»
      «(aggregate) Productivity growth had stopped long before the financial crash. The crash was the realisation of what had happened over the previous decade.»

      For an alternative view I like this rather graphical article by F Coppola:

      She shows that the aggregate productivity drop is mostly due a drop in production of oil and energy, in other words to the large and continuing fall in scottish oilfield output.

      Obviously describing changes in the fertility of land as changes in the labour productivity makes very little sense, but since JB Clark expunged "land" and natural resources from Economics (as well argued by M Gaffney) that is the situation.

      F Coppola also points out that productivity in finance has also decreased, which may related to «what had happened over the previous decade».

      B DeLong has made a very explicit argument about the fall in productivity in the USA in finance:
      «In finance we now pay some 8% of GDP—2% of asset value per year on an asset base equal to 4 times annual GDP. We used to 3% of GDP —1% and change of asset value a year for assets equal to 2.5 annual GDP. It does not seem to me that our corporate control or our allocation of investment is any better now than it was then. Certainly people now are trading against themselves more, and thus exerting a lot more price pressure against themselves. They are making the princes of Wall Street rich. Is there any increase in properly-measured real useful financial services that we are buying for this extra 5% of GDP? Paul Volcker does not think so. And I agree.»

  5. I also have thought about how people underrate the long run, or very long run. In investing, in anything, not just financial assets; global warming insurance, public investment in scientific and medical research,…, the long run for what you care about can be very long. If you have young children in grade school, like I do, you care about their health and happiness and security when they are in their 50's, a half century from now. And when they are in their 50's, they will be a lot happier knowing that their children will be secure and happy in their 50's. So that alone, plus much else, means it's important to try to help the lives of your grandchildren when they will be in their 50's, as well as when they retire. This means a horizon of over a century! just for these important reasons, let alone others.

  6. «GDP being 6% lower after 15 years»
    «think about the advice any doctor will give you that by smoking you will be worse off.»

    Voters may well believe that 6% over 16 years is going to happen, but that is very different from personally dying of throat and lung cancer 10-20 years earlier than otherwise, and still people smoke.

    Besides who knows how the 6% opportunity cost will be distributed; perhaps "Leave" voters think that lower immigration will mean better wages and job security for themselves, and most of the cost will fall on business and property owners.

    As our blogger wrote recently: «I’ve been told of one meeting where the response to the argument that EU membership had increased GDP was ‘maybe your GDP but not my GDP’.»

    «talk about long run costs had little impact in the media and on voters. [ ... ] To an economist that seems odd, because the economics behind the long term impact is much more solid than what might happen in the short run.»

    So «an economist» thinks that the discount rate is 0% for average people with a finite lifetime horizon and much personal risk and uncertainty? My impression is that a lot of people behave as if their discount rate was phenomenally high.

    There are some interesting polls that show that in the UK and the USA 40% of households would have difficult covering an unexpected expense of a few hundred pounds, and someone expects "hoi polloi" to give a lot of weight to 6% over 15 years? :-)

  7. «GDP being 6% lower after 15 years»
    «talk about long run costs had little impact in the media and on voters»

    I'll mention wise whiggish JM Keynes:

    «But at any given time facts and expectations were assumed to be given in a definite and calculable form; and risks, of which, though admitted, not much notice was taken, were supposed to be capable of an exact actuarial computation. The calculus of probability, though mention of it was kept in the background, was supposed to be capable of reducing uncertainty to the same calculable status as that of certainty itself.
    Actually, however, we have, as a rule, only the vaguest idea of any by the most direct consequences of our acts»
    «Thus the fact that our knowledge of the future is fluctuating, vague and uncertain, renders wealth a peculiarly unsuitable subject for the methods of the classical economic theory. By uncertain knowledge, let me explain, I do not mean merely to distinguish what is known for certain from what is merely probable»
    «The sense in which I am using the term is that in which the price of copper and the rate of interest twenty years hence, or the obsolescence of a new invention are uncertain. About these matter there is no scientific basis on which to form any calculable probability whatever.»

    IIRC our blogger even mentioned this passage previously...

    Now things have improved and some modelers are confident about their models, and their range of 3-8% over 12-18 years, but for the average person what will be the impact of "Leave" on *their personal "GDP"* is simply "uncertain", regardless of the amazingly high discount rate they may apply to what will be in 2030.

    As to me, I was and stay for "Remain" or a close approximation, but to me that range and timeframe sound plausible and worth taking into account, but other people with different circumstances may well have other views.

  8. Also, in the long run, junior doctors are consultants!

    1. That's a very good (I hope sarcastic) point with interesting implications.

  9. I think this analysis is seriously wrong. Most people who voted to leave accepted that there would be an economic cost. Nevertheless that cost was worth it to bring sovereignty back and to have some control of our borders. Every time Junker talks on TV I think the Brexit camp wins 1% of the UK electorate. In the words of Niles Fraser from the Guardian: Democracy should not be for sale. The EU federalist project has failed.

  10. Skidelsky's latest Project Syndicate post tries to deal with the issue of the elderly and debt.

  11. In the long run...I think you over-estimate the virtue of older voters.

  12. The Treasury analysis of UK trade shows that UK exports to non-EU has been increasing over the past few years and is now higher than exports to EU. We have already left.

  13. The limitation of economic forecasts over the medium-long run is that they ignore the volatility of a world in which the pace of change invalidates the assumed stability of parameters, or even the model itself.

    To get a sense of how dramatic change can be over a 15-year period such as that used for the Treasury Brexit predictions, just consider such intervals over the past century:
    1911: no Europe-wide war for nearly a century and 40 years since the Franco-Prussian clash; world economy expanding and globalising; tensions between colonial powers or in the Balkans look manageable; fall of the antiquated Manchu dynasty.
    1926: aftermath of a cataclysmic war; Hohenzollern, Hapsburg and Ottoman empires have vanished; Bolsheviks in power in Russia; but signs of stabilisation after post-war slump with sterling back on the gold standard, ‘roaring twenties’ in the US and relations between European powers improving, although war debts still unresolved.
    1941: Nazi Germany dominates Europe with France having fallen and Barbarossa bringing German armies to the gates of Moscow; much of China under Japanese occupation; attack on Pearl Harbour brings US into global war.
    1956: early stages of post-war boom in developed economies with expansion of trade under Bretton Woods agreement; Treaty of Rome; cold war stabilising; rapid Soviet growth and de-Stalinisation but invasion of Hungary; Suez fiasco; India independent and China communist.
    1971: late stages of post-war boom with resurgence of social conflict in advanced economies; decolonisation and imminent US defeat in Vietnam have shifted balance of world power; Soviet stagnation; Cultural Revolution still underway in China.
    1986: world economy recovering from stagflation; OPEC and Iranian revolution have shifted power in Middle East and commodity markets; early stages of new Chinese economic model; Thatcher and Reagan in power; Soviet power faltering despite Gorbachev; EU Single Market.
    2001: Soviet Union has collapsed; China joins WTO; internet boom ends in bust but central banks appear in control with inflation moderate; Euro launched; trade expansion and globalisation; climate change recognised as scientific fact; 9/11 attack on New York.
    2016: unresolved aftermath of financial crash; EU in “existential crisis” (Junker) with Euro failing, refugee crisis and Brexit; Middle East in flames in aftermath of Iraq war and ‘Arab spring’; China soon to overtake US GDP but growth declining and emerging markets spluttering; austerity.

    Now go forward to 2031. Will we have escaped from ‘secular stagnation’? Will the EU (or at least the Euro) still exist, and if not what will replace it? Will the Middle East crisis ease or deepen? Will Trump have become president and if so, then what? Will conflict be managed successfully in the South China Sea, Kashmir and Korea? Will Russia continue its return to global politics? Will China have transitioned from its current economic and political model? Will South Asia, Africa and Latin America be growing or not? What should we expect from artificial intelligence or bio-technology? How bad will global warming be? Will advanced economies have retreated into protection and nationalist populism? What are the “unknown unknowns”?

    Against this record or history and future opportunities and risks, predictions such as that made by the Treasury on the presumption of a stable world are bland. Of the periods above, only those from 1956 to 1971, and perhaps 1986 to 2001, look anywhere near stable enough for macroeconomic predictions to have been made with any conviction and in both cases fundamental underlying changes were soon to bring that to an end.

    The ‘gravity model’ argument that restricting trade between close neighbours with economic and cultural similarities will be damaging to growth is sound. But in the Brexit debate economists overreached by making quantified predictions and ignoring politics, and the public knew it. “Who knows what will happen in 15 years’ time” is a more profound reading of history than anaemic models.

    1. Point taken about the uncertainty and the problem of unconditional forecasts but I think your conclusions are too strong.

      Your argument reinforces the point that economists should be making conditional predictions. No reason why we cannot say that GDP will be x% lower than it would otherwise be. That's not over-reaching. Of course, there are implicit conditions such as that there won't have been some cataclysmic event (nuclear holocaust, or major volcanic eruption) that upsets everything but I don't see that as making it meaningless to make quantified conditional predictions, especially if they come with a range of values.


Unfortunately because of spam with embedded links (which then flag up warnings about the whole site on some browsers), I have to personally moderate all comments. As a result, your comment may not appear for some time.