Winner of the New Statesman SPERI Prize in Political Economy 2016

Friday 23 November 2012

Inconvenient truths, probably

I recently noted the exchange between Jonathan Portes (Director of NIESR) and Jesse Norman (Conservative MP for Hereford) on the Treasury Select Committee. Now reading too much into transcripts of oral evidence can be dangerous: people often do not say quite what they mean, and I’m not a fan of gotcha type political journalism that makes a big deal of such inexactitude. However Dr. Norman, formally a philosophy fellow at UCL, has now clearly set out the reasoning behind his complaint, which I think is rather revealing.

He sums up his views as follows:

1.  Economics is not an exact science.  Its substantive claims can only ever be more or less probable, though that probability can sometimes confer knowledge.
2.  So:  absolutely categorical statements—statements like “all of the interest rate differential was due to low economic activity, and none due to government policy” or “Plan A is failing” cannot be purely economic in character.
3.  Those in positions of public authority should be capable of being held to account.  This applies to politicians of course—but also to civil servants and academics.
4.  As Director of NIESR, Jonathan is in a position of public authority.  Indeed NIESR is an academic institution (NB its website is, and so subject to academic standards of rigour; as well as partly or wholly funded by public money.

There is no disagreement on (3) and (4), although note that his arguments apply to all academics (like me and other academic bloggers) and not just directors of well known public institutions. I would also agree with (1), if claims are about the real world rather than models of the real world. So Dr. Norman’s problem, to paraphrase, is that Jonathan should have said ‘Plan A is probably failing’ rather than ‘Plan A is failing’.[1]

Does this mean that in future posts I should insert a ‘probably’ in any statement that makes a claim about the real world? Take some recent news, like figures released today showing UK earnings rising at 1.4%, compared with inflation over the same period of 3.5%. Instead of writing

“This decline in real wages is due to the high level of unemployment, caused by the recession that followed the financial crisis. Workers are suffering today because of irresponsible behaviour by the banks before the crisis.”  

I should, to preserve academic standards, write

“This decline in real wages is probably due, at least in part, to the high level of unemployment. This in turn may have something to do with the recession, although we cannot rule out that other factors are at play. The recession itself could well have been triggered by a global financial crisis, although we can never know this for sure. Some people suggest that this crisis reflected poor lending decisions by banks, so there just might be some connection between irresponsible behaviour by banks before the crisis and people’s declining wages today. However it is all very uncertain.”

Would you class the second as economics, but the first as a political attack on the banks?

So why is Dr. Norman so keen that I or Jonathan Portes insert all these qualifiers to what we write? Well lets just imagine what might happen if I take Dr. Norman’s advice, and if I am ever asked to appear in front of the Treasury Select Committee again.

Wren-Lewis: “In my view Plan A has probably failed”
Conservative MP: “But you cannot say for sure - you admit it is possible it has worked”
Wren-lewis: “Well of course, but I think the balance of probability is that it has failed”
Conservative MP: “Ah, the balance of probability. Would you care to put a number on that professor?”
Wren-Lewis: “Well that is very difficult to do, but my guess is that the probability that it has failed is over 95%”
Conservative MP: “Your guess! My dear professor, I suggest you stop guessing and come back to this committee when you have something a little better than guesswork. In the meantime, I can tell you that Plan A has clearly worked, and I can say that without qualification.”

What I find revealing about all this is that while I am supposed to say ‘Plan A has probably failed’ as an academic economist, it is OK to say ‘Plan A has failed’ as a political statement. Political statements, apparently, do not need qualification even when qualification is due. Is this because they have no truth value whatsoever? Perhaps this was what was going on when certain well known US economists claimed that “the negative effect of the administration’s ‘stimulus’ policies has been documented in a number of empirical studies.” Here it did not matter that nearly all such studies had found the stimulus policy had been effective, because it should have been understood that here they were making a ‘political statement’, which should not be confused with the ‘economic statements’ they normally make. (Fortunately members of the Conservative Party have not as yet started making ‘political statements’ denying evolution, although climate change denial is doing pretty well.)

The reality of course is that most people now realise Plan A has failed, and those that cannot admit that only have low interest rates to point to as any kind of success. So when macroeconomists point out that interest rates would be expected to be low for an economy that is so depressed, and that the experience of certain Eurozone countries just does not apply to us, these economists become dangerous voices that need either to be discredited as partisan, or intimidated into silence. Such tactics can be quite effective when used against an organisation as timid as the BBC: lets see if it can also work with an institution like NIESR.

[1] The more normal criticism by politicians of economists is the opposite, which is that we are always hedging. ‘Give me a one-handed economist’ President Harry Truman is reported as saying.


  1. It looks to me that this strategy is closely related to the Republican strategy of "compromise." What they try to do, and unfortunately often get, is to have the Democrats (Obama) come up with a compromise as their option, so they can (further) compromise on this. While on their side they certainly don't open the negotiations with a compromise, in fact they tend to add extreme negotiation points to allow "compromising" on (like now putting Obamacare on the table.) So I don't think this is really about political versus economic statements. But rather it is about a certain kind of politician who never wants to compromise, never wants to admit being wrong, and never wants to accept reality over ideology, and who tries to figure out reasons to straight that out.

  2. I have commented before on this blog that economists need to adopt a code of ethics. My previous comment was in the context of whether economics is a science. For that to be the case, economists need to follow the scientific method, and to be seen by others to follow the scientific method. The best way to do that would be to adopt a code of ethics and to introduce sanctions against any economist who broke this code.

    I would add two further justifications here.

    Firstly, I have seen numerous examples, in newspapers and on blogs, where economists accuse each other of being unethical, incompetent or both. This occurs between mainstream/left economists and mainstream/right economists, and also between mainstream economists and heterodox economists. When economists behave in this way towards each other, they can have little comeback when a non-economist adopts the same tactics.

    Secondly, as this example shows, ethical economists need an effective defence against politically motivated charges about their ethics. Most conclusions in macroeconomics have political implications. This is particularly true when the political debate is between free-market and Keynesian policies, and where an economist self-brands himself with one of those labels. Economists are naive if they expect their conclusions to be accepted by ideological politicians who don’t like what they are being told. A code of ethics would provide the best defence against this type of accusation.

    The most ironic aspect of this subject is that many of the Conservative politicians advocating austerity policies, including David Cameron, studied economics at Oxford University. If Keynesian economics is the best path for us to follow then either economics education at Oxford is ineffective in putting across that case or the politicians are being unethical in ignoring their education for political purposes. In the latter case, they would surely have benefited from being introduced to a code of ethics for economists during their education.

  3. I have to say that in all probability this is an extremely enlightening and amusing post, although it is possible that I am mistaken in this regard. It could be asserted that some government supporters simply "don't like it up 'em", but it is perhaps too soon to state this with complete confidence.

  4. Very nice post. What makes science real science? Of course logical constraints are compulsory, like making the difference between a probable interpretation, an established fact and a universal proposition. But you also have to be honest intellectually (I mean really looking for the truth, not just arguing for your own interest); and honesty overtakes logic. Your Conservative MP clearly doesn’t want to know the truth and always can find a logical fallacy in your assessment (because economics is a science of the probable). Each one has to wonder if a given argument is really an argument or just a quibble.

  5. If Jonathan had gushed praise on Osborne and said his policies were working and that they were the way forward. Do you think Norman would accuse him of political bias then? Not a chance,he'd be quoting him every chance he got.
    Grow up Norman and learn to take criticism

  6. Of course, even your second version is not exactly bulletproof. The average earnings statistics have a degree of uncertainty to them, and there are all sorts of questions about the measurement of inflation. So you are already in trouble with Dr Norman when you say "“This decline in real wages". I would suggest "the probable decline in the purchasing power of a statistical construct meant to represent an average employee, which may or may not be relevant to any actual person, in terms of a consumption basket which I believe is probably reasonably representative of the average consumer, which I believe to be outside the normal range of uncertainty of the measuring error". And then the rest goes through.

  7. According to the Jesse Norman principle:

    1. Geology is not an exact science.  Its substantive claims can only ever be more or less probable, though that probability can sometimes confer knowledge.

    2. So:  absolutely categorical statements—statements like “All the evidence is that the earth was created 4.5 billion years ago, and not 4216 years ago” cannot be purely geological in character.

    Er, yes they can. The statement in 2 above treats as false what is evidently false.

  8. Ah – how much public persuasion occurs by way of manipulating the burden of proof? In Australia the Productivity Commission's reports very often gets to their preferred policy answers by asserting that approaches it disapproves of "could" lead to some undesirable outcome. I documented in this piece on automotive industry policy.

    Certainly simplifies the argument for those writing the reports.


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. In addition, I cannot publish comments with links to websites because it takes too much time to check whether these sites are legitimate.