Ritchie on how economics isn’t science

Quoting Chang:

[A]ll .. economic theories are at least debatable and often highly questionable. Contrary to what professional economists will typically tell you, economics is not a science. All economic theories have underlying political and ethical assumptions, which make it impossible to prove them right or wrong in the way we can with theories in physics or chemistry. This is why there are a dozen or so schools in economics, with their respective strengths and weaknesses, with three varieties for free-market economics alone – classical, neoclassical, and the Austrian.

Given this, it is entirely possible for people who are not professional economists to have sound judgments on economic issues, based on some knowledge of key economic theories and appreciation of the political and ethical assumptions underlying various theories. Very often, the judgments by ordinary citizens may be better than those by professional economists, being more rooted in reality and less narrowly focused.

This is of course very convenient for Ritchie. It means that he can never be wrong. For there is no standard of proof or truth to which he can be held.

Except, of course, there are bits of economics which are indeed science, where we have hard answers.

For example, this Laffer Curve thing. We know that there’s an income effect (raise taxes and people will work more to reach their target income) and a substitution effect (raise taxes and people will work less as working ain’t worth it). We also know that both apply to almost all of us at some time or income level or another. And it’s the interplay between these effects that determines the change in tax collected from a change in tax rates.

It’s true that we then have to guess a little at what the rates are where, in general, a change in tax rates leads to more revenue or less.

However, we do have solid empirical evidence in some parts of this. For example, we know very well that the income effect is much weaker in, the substitution effect much stronger, in married women. We really do know that as a matter of fact. Quite why, well the surmise is easy enough, why not look after your own household at zero tax rather than work, get taxed and pay another to do so? But we do know the fact.

Which means that, just imagine, if a retired accountant were to write a Budget briefing paper for the TUC, arguing that a significant rise in income tax rates would lead to an increase in labour being offered on the market, that increase coming from married women where the income effect would be much stronger than the substitution effect, well, we would know that that retired accountant was, simply and factually, wrong.

And note that our knowledge of why he is wrong does not derive from any theoretical assumptions at all. Only from direct empirical observation. We know that when tax rates go up married women are more likely to withdraw from the labour force rather than working more hours.

Or, even if we are to accept that economics is only partially sciencey in those areas where it is Ritchie is still wrong.

Quelle Surprise.
#

41 comments on “Ritchie on how economics isn’t science

  1. Ritchie isn’t keeping up. This “Economics is Bunk” thing has been reversed ever since some French Lefty bloke wrote a book that ticks all the Left’s prejudices on the matter. And you can’t get less scientific than tax accountancy, where all the rules and concepts have underlying political and ethical assumptions. That of course is assuming that the Courageous State has any underlying ethical assumptions at all.

  2. Well yes. Economics isn’t like chemistry or physics. It’s not experimentally verifiable in precise terms. But nor’s Darwin. T Rex is a matter of opinion. And probably wasn’t a socialist & certainly not a semi-retired accountant in Norfolk

  3. All economic theories have underlying political and ethical assumptions, which make it impossible to prove them right or wrong

    This is a classic confusion of the “is” and the “ought”. Economics as a science addresses the “is”, economics used as an ideology confuses the matter by introducing the “ought”. Economics cannot tell you whether it is ethical to provide State healthcare. It can provide an analysis of the potential effects of doing so on the economy. Economics cannot tell you whether to have a socialist or capitalist system. It can tell you what kind of outcomes each choice wil produce.

    Ritchie doesn’t understand the distinction, which is why his economics is crankish.

  4. Chang seems quite embittered. It’s almost as if he’s been laughed at by”professional economists” once too often and had 21 too many blogs take the piss out of him.

  5. “which make it impossible to prove them right or wrong in the way we can with theories in physics or chemistry”. Demonstrates an undergraduate level of error of what science is – not even worth of full pint’s worth of discussion. After all – Newton was proved right for several hundred years, before Einstein;s theory of general relativity. A straightforward misunderstanding of the scientific method.

  6. It’s getting beyond parody.

    I seriously think he needs psychiatric help.

    It’s becoming apparent he is suffering from a delusional disorder, as he fabricates a world around him in which he is the perfect being who, by definition, can do no wrong.

  7. Of course, the layman’s opinion of economics is no longer preferable to the neoclassical expert in the case where the neoclassicist is someone like Scott Sumner who believes in targetting nominal GDP, and the layman believes that the country should be run like a firm, living within its means, and that printing money cannot make everyone richer at the same time.

    So he’s not even consistent in his anti-intellectualism.

  8. In which we can with physics?

    So, Chang can tell us whether string theory or loop quantum gravity is right?

    I’d say that economics is just a young science. Newton spent years tinkering with alchemy. Physics knows better today.

    Give it a few hundred years, people will wonder at the stupidity of our age.

  9. personally I am happy to say economics is not a science, in the sense we cannot refer to established facts replicable in the laboratory (whilst I appreciate even the hard sciences are a bit wobbly here too) because we have to deal with an extremely complicated world and cannot rule out competing explanations and so disagree about most everything. on the other hand we at least do our best to state theories precisely and look for ways to test them in the data and try to proceed in a scientific fashion as best we can given the constraints placed upon us by our subject matter.

    I once read a nice paper that demonstrated that if the quantity of available data is sparse relative to the complexity of the processes you are trying to understand (loosely speaking say we have macroeconomic data on x quantities but have X candidate mechanisms), then individuals with the same information but competing models can validly disagree for eternity.

  10. Tim Almond: Give it a few hundred years, people will wonder at the stupidity of our age.

    First!

  11. “I once read a nice paper that demonstrated that if the quantity of available data is sparse relative to the complexity of the processes you are trying to understand (loosely speaking say we have macroeconomic data on x quantities but have X candidate mechanisms), then individuals with the same information but competing models can validly disagree for eternity.”

    Remind me again. Why do we have a university system? Is it purely to give layabouts somewhere to layabout? That’s a “nice little paper”?. There’s nice little papers come with perforations. Presumably there’s currently a researcher writing a nice little paper on how to use 3 or 4,

  12. Mathematics has underlying assumptions too. You can’t keep going back proving everything ad infinitum, I guess he’s never heard of axioms.

  13. bis

    what a stupid response. That paper shows us, amongst other things, that disagreement about matters such as economics does not imply that anybody is making any errors, lacks any data, or is evidence that economists are “unscientific”.

    It’s actually closer to a paper that a statistician would produce than an economist. But I suppose you’d say research in statistics, learning, information theory etc. is all worthless.

  14. “lacks any data” in the sense that there’s not available data that somebody you disagree with has that they you could show then to change your mind. We all lack data in the sense that insufficient is available to anybody.

  15. “what a stupid response”

    On the contrary: like much academic layaboutry, your nice paper is a statement of the bleeding obvious dressed up to look clever. And we’re not only required to pay for it, but to be grateful too.

  16. it’s not obvious that disagreements can persist absent errors once everybody has access to the same information and more data becomes available over time.

    And what’s with the “layaboutry”? You might not value the product, but an awful lot of hard work has gone into it.

    You could shut down all pure research into maths, information theory, decision theory, learning, economics, network theory etc. etc. One problem doing so is that its hard to predict when pure research will end up having use real world applications.

  17. I’m with Luis here.

    By contrast, I’m not with any academic (and I will call Chang an academic) who writes about ‘science’ when he means ‘natural’ or ‘physical’ science (or does he?) and would describe science as ‘not science’ and condemns it for claims it hasn’t made. An academic who is apprently aiming for precision but waffles on about ‘science’ instead of scientifc methods. An academic who chooses to ignore (because I don’t accept he is ignorant of) centuries of philsophical examination of what is and is not ‘science’.

  18. I’m also sickened that anyone with any academic pretensions could write

    “…. Very often, the judgments by ordinary citizens may be better than those by professional economists, being more rooted in reality and less narrowly focused.”

    This strikes me as the temper tantrum of the professionally humiliated. Maybe Tim hit home harder than he could have imagined.

    All he achieves with this nonsense is to give succour to chancers like Murphy to parade their ignorance as a virtue.

    Let’s extend Chang’s logic to medicine shall we: Perhaps the layman’s “knowledge” “being more rooted in reality and less narrowly focused” might lead us to start bleeding patients.

    This isn’t about science; it is an anti-knowledge manifesto.

  19. @Luis
    “it’s not obvious that disagreements can persist absent errors once everybody has access to the same information and more data becomes available over time.”

    It may not be obvious to academics but it’s obvious to parts of the gaming industry, for a start. And as a non-academic I can explain why different people, working with incomplete data, will come to different conclusions. Because there’s two data sets in operation. One is the common data on the problem. The other’s the experience each person brings to the problem, they use to extrapolate the missing data. Each individual is has different experience, therefore has a different data set., produces a different extrapolation.
    I won’t bother about applying for the research grant. I’ve made enough money out of this, as it is.

  20. bis

    yes it would be obvious if there was private data (each person’s experience) but that’s the bloody point of this paper, disagreements can persist absent that

    similarly, disagreements between gamblers can be attributed to errors, different private histories etc.

    yes I wouldn’t bother applying if I were you

  21. I’m with Luis on this one. And while many academic papers are statements of the obvious, I’m fairly certain that the paper he’s describing does not fall into that category.

  22. ken

    yes my favourite genre of obvious academic paper is applied micro development research summarised by “bad things have bad effects”

    of course somebody actually has to find evidence and estimate their magnitude. sometimes intuition is wrong.

  23. oh and I do not pretend to understand the content of that paper beyond abstract/introduction/conclusion

  24. I wonder if Ritchie and Chang (and the Guardian) will apply something similar to Climate Change (for example), or Racial Discrimination? Women’s rights?

    The day I see the headline:

    ‘Climate Change is too important to be left to the experts’

    as a headline in the Guardian is the day I’ll start taking them seriously. For one thing it could kill off William M. Connolley – ‘the wikipedia Climate fraudster’

  25. “three varieties for free-market economics alone – classical, neoclassical, and the Austrian”?!?

    I think that misapprehends the terms. The classical economists (Smith, Ricardo, the Marginalists) are all dead and the book has closed on them. When, after some upheaval, much of the economics profession came to the conclusion that the classicals had gotten much right about the core of economics, they were called the neoclassicals. While they have some difference, advances even, over the classicals, they are not in any sense a separate “school.” They are, by and large, what the classicals would have been given another few centuries of thought and research.

    In other words, if there is a living economist who will avow that he is of the classical, not that nasty neo-classical, school, he must be pretty lonely.

  26. Hmm. It is a statistician’s paper. There was a time when Econometrica regularly published articles that I was interested in as opposed to things that only interest econometricians. It makes sense and is indeed not a statement of the obvious. (I suspect mathematicians might think it was a statement of the obvious, but I dont have one to ask.)

  27. Economics does n’t cut it as hard science: it does n’t cut it as a humanities discipline either because its working language is too problematic.(Also it has the category of heresy particularly for those who dare to point out that banks create money) But talking about bank” lending” when the banks don’t have the money to lend out has been shown up recently by the B o E “Money Creation in Modern Economy ” Quarterly Bulletin. And as Max Keiser keeps saying, you can’t talk about the inflation rate when you don’t include housing costs. Piketty (whom I doubt anybody here has read all through) has a good chapter towards the end entitled” For a Political and Historical Economics” p573 in which he argues for the return of the term Political Economy.
    He writes “For far too long economists have sought to define themselves in terms of their supposedly scientific methods. In fact, those methods rely on an immoderate use of mathematical models ,which are frequently no more than an excuse for occupying the terrain and masking the vacuity of the content.. The new methods often lead to a neglect of history and of the fact that historical experience remains our principal source of knowledge…the imperfect lessons that we can draw from history and, in particular , from the history of the last century, are of inestimable, irreplaceable value, and no controlled experiment will ever be able to equal them”.
    It is a sobering thought that the Sorbonne Economics students were vehemently complaining about the immoderate use of mathematical models in 2000.

  28. @Luis
    “yes it would be obvious if there was private data (each person’s experience) but that’s the bloody point of this paper, disagreements can persist absent that ”
    You’ve now left me wondering in which world a person can come to a problem with absolutely no prior experience of anything to draw on. Oxbridge PPE is the only one springs immediately to mind.

  29. So historical experience remains our principal source of knowledge? You might as well say sensory perception is our principal source. Both might be true statements but do nothing to help us understand science or economics. Maybe P has left out the internal reflective bit about developing knowledge? I can’t see where economists decided to ignore history.
    Besides which, if P and others are taking account then, by definition, economists are not ignoring it. If P can generalise for “economists” so can the rest of us.

  30. Bis

    You still have the stick by the wrong end. Point is that even if everybody could perfectly communicate all their personal experiences disagreements could persist.

  31. @Luis
    That’s highly arguable.
    It’s a thought experiment in AI. If two entities share the sum of their memory, at that point do you have two entities or two instances of the same entity.

  32. BiS

    The AI’s have two different operating systems, Even with the same data their models give them different views of the universe. (This is a bad analogy, but the best I can do with AIs)

  33. Well, it’s been interesting to read all the comments here.

    And really, one can come to only one conclusion.

    No-one gives a toss that The Dick is bonkers…

  34. Yep. There are various candidate models and no way to say which is more likely to be true. You could think of the two AIs each habe different model ‘in mind’ and having no reason to change even under full information.

    I read this paper because I’m doing a bit on Knightean uncertainty, so was interested in circumstances in which we may have no way of assigning probabilities to different possibilities

  35. ken “(I suspect mathematicians might think it was a statement of the obvious, but I dont have one to ask.)”

    Oh yes you do; I’m a mathematician! While I wouldn’t presume to speak for all mathematicians [1], it seems to me like a proof of something mathematicians would have believed but would not have claimed to know. We have a lot of these; they are also referred to as “things mathematicians believe and physicists (or economists) know”.

    [1] As in the commonly-told joke: three mathematicians walk into a bar, and the barman says “Would you gentlemen all like a beer?” The first replies “I don’t know,” the second replies “I don’t know,” and the third says “Yes, we would.”

  36. “I’m also sickened that anyone with any academic pretensions could write

    ‘…. Very often, the judgments by ordinary citizens may be better than those by professional economists, being more rooted in reality and less narrowly focused.’

    This strikes me as the temper tantrum of the professionally humiliated. Maybe Tim hit home harder than he could have imagined.”

    Ha! Don’t get too high on your horse there, Mister…

    Every hear of Public Choice Theory? How many decades did it take The Profession to figure out that which has be been blindingly obvious to anyone who deals with bureaucracies and bureaucrats since the dawn of time?

    Full props to Buchanan, but he got a Nobel Prize for discovering what any tax accountant, securities attorney or Fortune 500 CEO could have told any economist at any time over the past 75 years… had said economist been interested in listening.

  37. Dennis

    OK, well made. Many econonic observations are articulated by other professions and with hindsight are all obvious and trivial, except…

    How many insights do you really expect to get from a retired small-firm partner in Norfolk? Do you really want to spend your time reading the economic analysis of soemone who says: “nobody is applying the the job at the salary on offer, so there’s clearly no demand, so we can reduce the salary on offer”?

    So, if you’ll excuse me, the high horse remains saddled up.

  38. Meanwhile…

    All hail the Meissen Bison!!

    For being the first to congratulate Ritchie for winning over that tough crowd of highly trained, world weary, 17 year olds, despite them bringing all those years of learning and personal economic experience to bear.

    You’ve earned your Friday evening pint mate. Now take your toungue out of your cheek or you won’t enjoy it.

  39. @Ken
    But one could say each has now shared the memory of the other, thinking as itself. It remembers being the other as well as it remembers being itself. So which is which?

    On disagreement from common data;
    Folk value what they think they know over what they discover & what they think they’ve discovered over what they’re told. In this lies the root of all good cons.

  40. BiS

    The AI’s different programmes “remember(interpret)” the data in different ways and refuse to admit the other’s interpretation is correct.

    Philip Walker
    The reason I tend to believe that it is something that mathematicians might already have proved is that econometrics is littered with ideas pilfered from maths and this sounds like the kind of thing someone might already have proved. For example economists use White 1980* robust standard errors, but this method of correcting for heteroscedasticity actually belongs to a statistician. (Huber IIRC?)

    * Another econometrica paper.

Leave a Reply

Name and email are required. Your email address will not be published.