Obliquity . . .

We do have a few advantages, perhaps the greatest being that we don’t have a strategic plan

Warren Buffett

Obliquity . . . or indirectness of means is a subject to which it turns out I’ve given a lot of thought over the years going back at least to Charles Lindblom’s attacks from the 1960s on upon what he called the ‘synoptic’ method of policy decision making. This approach thinks of the ideal policy process as a kind of giant cost-benefit analysis.  I had a (rather inarticulate) crack at some of the issues here.  I’m pleased to see that the very terrific John Kay has now written a book about it using a wealth of illustrations from his extensive knowledge of history of various corporates with which he’s been engaged, or which he’s studied.

Anyway there’s a problem here for people on John Kay’s side of the fence – like me.  The problem is that one spends one’s time poking holes in propositions which seem like the very acme of reason. You see John Kay has a jaundiced view of ‘evidence based’ policy. This is not, dear reader, because he or I think that we should have ‘faith based’ policy, or that we shouldn’t do away with evidence in our decision making.

Rather we think – or I think he thinks – that the current fad for evidence based policy, for strategy in which one determines one’s ends first and then goes hunting for the right means is not a good description of how we do make policy decisions, or could ever do so.  The descriptions of policy making in which things are all described as if the process were a linear process of getting one’s priorities sorted and then rationally going about achieving them are fantasies and so serve more to disorient strategic thinking and discussion than they serve to enlighten. Here’s an example of this kind of thinking.  It all has the ring of commonsense about it, and to the extent that it’s commonsensical who could object.  But I think it conceals more than it reveals.

Andrea DeMaio, for whom I have a very high regard, also engages in this kind of – let me be bold and call it –  ‘cheap rationalism’ when he says that Government 2.0 is just a set of tools, a means, not an end in itself.  Well yes, there’s some sense in saying that, but then there’s some sense in quibbling with it too.   This was my response on Adrea’s post.

It’s easy to say that one should adapt one’s means to the ends one wants, but you’ve already said that you don’t know what the shape of the ends will look like. Which raises a problem don’t you think?

Often means are important and it’s important that they conform to certain values. Government is structured in this way since we understood the importance of separation of powers. Throughout government decisions are not just means to ends, it’s important that they ‘make sense’ in and of themselves – particularly in a world full of unintended consequences.

Government is shot through with this – and it should be. Judges make decisions based on a mix of values, principles, consequences etc. A freedom of information act isn’t just a means to an end, the power of its values is one of its driving forces.

So I think Government 2.0 is also powered by its appeal to shared values. That’s one of the ways we try to navigate when – as you’ve indicated in your post – we don’t really know where we’re going – we’ve just got hunches.

Kay has written about obliquity before, and says that his interest goes back to 1996.  Now he’s published a book on the idea, from which the Financial Times has an abstract.  I look forward to reading it, but here are a couple of key paras from his article in the FT.

[T]he idea that good decisions are the product of orderly processes – is more alive than ever in public affairs. The success of the physical sciences has encouraged us to believe there might be a science of decision-making. With its aid, all kinds of problems could be managed objectively. Such a procedure would lead every conscientious person to the same answer. As a result, both political and personal disputes could be resolved by the collection of evidence and the pursuit of rational discourse.

There is not, and never will be, such a science. Our objectives are typically imprecise, multifaceted and change as we progress towards them – and properly so. Our decisions depend on the responses of others, and on what we anticipate these responses will be. The world is complex and imperfectly known, and this will remain true however much we analyse it.

We do not solve problems in the way the concept of decision science implies because we can’t. The achievement of the great statesman is not to reach the best decision fastest, but to mediate effectively among competing views and values. The achievement of the successful business leader is not to articulate visions of the far future, but to match continuously the capabilities of the firm to the changing market environment. The test of financial acumen is to navigate successfully through irresolvable uncertainties. . . .

Our approaches are iterative and adaptive. We make our choices from a limited range of options. Our knowledge of the relevant information and of what information is relevant, is imperfect. Different people make different judgments in the same situation, not just because they have different objectives, but because they observe different options, select different information, and assess that information differently; even with hindsight it will often not be possible to say who was right and who was wrong. . . .

It is hard to overstate the damage recently done by leaders who thought they knew more about the world than they did – the managers and financiers who destroyed great businesses in the pursuit of shareholder value; the architects and planners who believed that cities could be drawn on a blank sheet of paper; and the politicians who believed they could improve public services by the imposition of targets. They failed to acknowledge of the complexity of the systems for which they were responsible and the multiple needs of the individuals who operated them. . . .

Successful decision-making is more limited in aspiration, more modest in its beliefs about its knowledge of the world, more responsive to the reactions of others, more sensitive to the complexity of the systems with which it engages. Complex goals are generally best achieved obliquely.


This entry was posted in Economics and public policy, Philosophy, Political theory, Politics - international. Bookmark the permalink.

19 Responses to Obliquity . . .

  1. JJ says:

    Nice piece and very sensible. I remember the fad for ‘zero based budgeting’ where you pretended you could wipe the slate clean and work out where the money should really go. Metaphors for real policy making would be interesting to collect. And yet talking about values and principles is hard to do in a bureaucracy.

  2. derrida derider says:

    “Our approaches are iterative and adaptive” – indeed, yes, because that’s exactly what policy makers mean by “evidence based policy”. The evidence forms part of the classic policy cycle, of the sort taught in every MPA course, that is intended to be iterative and adaptive.

    Kay (someone for whom I, too, have great respect) seems to me to have two distinct propositions here – that evidence based policy is not what we do, and that evidence based policy is not what we should do. On the first, well that’s democracy for you. On the second, I’ve seen too much of the alternative – faith-based policy often supported by policy-based “evidence” to be convinced.

    In my experience its not the those persuadable by evidence who lack humility but those who “know” a priori what needs to be done. Twain’s line about “its not what you don’t know that hurts you, its what you know that just aint so” comes to mind.

  3. Paul Frijters says:

    Like Derrida, I am also hesitant about the conclusion that we should forget about planning simply because life is too complicated. Where is the recognition that many of the most beautiful cities have exceptionally carefully planned designs (like Canberra, or the central boulevarts of Paris, or many towns in the US)? Where is the recognition that many of our most successful tax-and-spend institutions are planned to the hilt, such as the system of separation of powers, the geographic spread of services, the national curriculum of many countries, etc.

    I would tend to a very different proposition to the one espoused here. Yes, there is incompetence, hubris, and an evolution that is mainly marked by one failed plan after the other. But the wish to be rational, to have ex ante plans, to design institutions and thoughts that use the available resources efficiently, is an important element in our current success. If only due to the law of large numbers, some plans do get it right and those will then become a model to follow for others. Recognising the mistakes that have been, are, and will be made is all well and good, but this shouldnt be confused with the idea that planning is silly and has no real role in politics or decision making. The only question of actual interest is the one of the optimal mix: is there too much planning and too little react-as-we-go? The answer to this question should recognise that we educate our politicians, businessmen and civil servants to be able to make plans and follow processes. The question whether we should teach them something else is a question about the counter-factual and nothing above seems to even address that question. Where is the evidence that societies with less planning oriented education and policy structures but more ‘trust me, I know what I am doing’ discretion do better?

  4. Nicholas Gruen says:

    Well guys, you’re certainly confirming the implicit prediction I made in the piece. Do you think JK or I think that we shouldn’t make plans, or that we shouldn’t base policy on as much evidence as possible?

  5. Gummo Trotsky says:

    The Rudd “Education Revolution” with its standardised nation tests of learning and all the other technocratic accessories – the My School, computerised longitudinal tracking of student progress – is a fine example of “evidence based policy” in action. The whole program is a crock and its proponents continue to ignore mounting evidence that it won’t work. Now that’s faith based policy.

  6. conrad says:


    you can add Teach for Australia to that also — the program with the brilliant idea of putting the least experienced teachers in the hardest to deal with places.

  7. Gummo Trotsky says:


    Let’s not divert the thread too early. I was merely presenting an example of “evidence based policy” – I’m confident we’ll be presented with more after the next Federal election. And that on examination they’ll turn out to be faith based.

  8. Paul Frijters says:


    perhaps you can explain better what it is you would like to change about our education practise and our policy practise? The difficulty I am having is that I completely agree with many of your observations about the hubris of planning but am, like Derrida, mindful of the other side (ie. the usefulness of planning, if only as a habit in an evolutionary path). I guess I dont see what your agenda is, i.e. what the proposed alternative is.

    For instance, one alternative to the ‘illusion of control’ that dominates much of our teaching and thinking is to educate people very broadly, to equip them with many different perspectives and to essentially make them historical pattern recognisers who are armed with a vast database of historical mistakes and successes as well as loose rules of thumb to recognise the position they find themselves in. People taught in wisdom, not specialised planning knowledge. That would mean, for instance, that we should teach more (and more informed) history. We should teach more languages, and less plan-oriented subjects (management theory, maths, engineering?). More sociology and psychology (which is all about the particular and the futility of certainty), and less economics (where the mainstream is full of the illusion of control).

    If I think of societies which have less illusion-of-control teaching, I dont think of societies that do better at all. Societies that train individuals in highly plan-oriented fields like engineering, economics, and even law, seem to do very well. Despite the obvious fact that an illusion of control is an illusion, there seems no aggregate harm done in having it widely taught and mimicked.

    Hence, what kind of alternative do you have in mind?

  9. Nicholas Gruen says:


    As I intimated at the outset of the post, it’s actually not all that easy to explain all that clearly what I mean. I’m not proposing some shining new (or old) method of solving problems. I am pointing to the hubris of the way in which strategy is 1) spoken about and 2) delivered.

    So it’s a critique more than a proposal. But the most substantial thing I said in the post, the most substantial example I think is my critique of the meme that we just want to sort out our ends and then adapt the means to them. Something which is then amplified in the link I provided.

    Of course in the right place that’s exactly what you want to do. But dealing with complex systems, and dealing with people the value of that kind of cheap rationalism is – I’m asserting – much more constrained than people assume it is.

    I’m sure that DD could point to some programs that have been responsibly and well developed according to what evidence can be mustered, (and we can all point to ones that are not). If they’re well done, that will be pretty clear to most reasonable people. That won’t mean that such programs mightn’t get closed down for capricious reasons, but reciting the ‘evidence based’ mantra (while refusing to do cost-benefit studies on multi-billion dollar projects and shovelling billions out to the lest competitive of one’s industries) won’t do much to protect them.

    I’ve never attended a strategy retreat that I didn’t come away from thinking was a more or less total farce. I used to think that this was because of certain characteristics of some of the important attendees. But I’ve since done them in other circumstances. Yet every time we go away, and everyone gets to have their say, and we work on what I call ‘religious activities’ which is to say you come up with pithy little summaries of what the agency’s ‘vision’ and ‘mission’ are, and virtually no insights come even remotely in view.

    I doubt you’re unsympathetic to the examples I’m giving – so we agree on the policy, and don’t really agree on how to describe how it is arrived at or how it should be arrived at. But of course other things being equal, more evidence is better than less, and policy should be as evidence based as it reasonably can.

    If all this seems strange to you, and you’re still wondering what we’re arguing about, fair enough, but remember, John Kay is as much of an ‘evidence guy’ as anyone I know. So what’s he on about?

  10. Paul Frijters says:


    no problems with any of this. As an observation you are spot on and I have my own experience of retreats and even UN conferences to know exactly what you are saying about the strong religious aspects of what goes on at such places. Much of the value of the illusion of control is pure psychological consumption. I dont disagree with your description of how policy is arrived at.

    The next step after critique is of course a doable alternative (how it should be done), and the only major alternative that came to my mind didnt seem all that great on closer inspection. Perhaps you can come up with something better since this is such a long-term thing of interest to you?

  11. Don Arthur says:

    The descriptions of policy making in which things are all described as if the process were a linear process of getting one’s priorities sorted and then rationally going about achieving them are fantasies and so serve more to disorient strategic thinking and discussion than they serve to enlighten.

    One of the reasons these descriptions are fantasies is that policy makers don’t always know what they want. And when they do, often the different policy makers involved in the process don’t agree.

    In his 1971 paper ‘The Technology of Foolishness‘ James G March wrote:

    The argument that goal development and choice are independent behaviorally seems clearly false. It seems to me perfectly obvious that a description that assumes goals come first and action comes later is frequently radically wrong. Human choice behavior is at least as much a process for discovering goals as for acting on them.

    March argues that people develop their goals and values by doing things. This seems right to me.

    And I can’t see how it’s reasonable to assume that policy makers have a stable and consistent set of preferences. As philosopher Isaiah Berlin observed, rational people can have values that conflict with each other.

    Policy makers frequently devise programs first and come up with objectives second. If the objectives aren’t met, they’ll often choose to revise the objectives rather than change or abandon the program.

    It seems to me that value pluralism is a defensible normative position. To me, monistic theories like utilitarianism seem entirely unconvincing.

    Of course this makes some decision tools (particularly those popular with economists) unusable in many circumstances. But that’s not an argument against value pluralism — it’s just frustrating for people who are good at using those tools.

  12. Gummo Trotsky says:

    To me, monistic theories like utilitarianism seem entirely unconvincing.

    I’m going to grab the opportunity to seize on this remark and either amplify it or spin off on a bizarre tanget. (you decide – but keep in mind that if you actually tell me that what I’m going to go on to say below is too bizarrely tangential to be taken seriously, I might get all huffy and argumentative).

    Recently it occurred (or recurred) to me that one of the (many IMHO) deficiencies of classical microeconomic theories of consumption is their one sided emphasis on utility, at the expense of situations where the issue is not the consumers desire to maximise utility, but the desire to minimise disutility in the form of pain and suffering.

    What prompted this thought is the issue of health care and all the commentary that’s been flying around about Obama’s health plan and today’s debate between Kevin Rudd and Tony Bugner. If you have appendicitis, for example, you’ll experience a lot of pain and discomfort – disutility. But once you’ve had one appendectomy, there’s absolutely no utility (except maybe in cases of Munchhausen’s Syndrome) in getting another.

    Marketers and advertisers have been onto the commercial uses of disutility for years – that’s why they discovered “pester power”.

    I think I’ve just produced an example of a decision tool (classical microeconomic theories of consumption) that’s unusable in many circumstances. So if you’re going to argue that this comment is bizarrely irrelevant…

  13. Gummo Trotsky says:

    Awaiting moderation? Que?

  14. Nicholas Gruen says:


    Don has stolen my thunder by saying something much more apposite, something that takes the discussion much further than I could have.

    But to answer your question, I was offering a critique of what was going on right now. This wasn’t to satisfy some craving to complain but because it seems to me that so much is wrong with the way in which these things are discussed, and that those things are wrong in quite commonsensical ways, and in ways that people should be able to recognise fairly easily and identify improvements on the spot – ie by not being so silly in the way they’re proceeding.

    That is the point of it. In many disciplines, this kind of critique is what much of the ‘theory’ of the discipline is composed of. It’s a kind of reaching for self-awareness. Sometimes some technique might come along which seems useful, but critique is seen as valuable in itself.

    If someone says that the key to policy making is for it to be more ‘evidence based’ – they could add that it should be sustainable, equitable and lots of other things pretty obviously my complaint is not that we should reverse those values. It is rather to say that these are slogans which anyone can mouth.

    It’s a bit like a coach saying that the key to the game is for his side to kick more goals than the other side. If I then say he’s a nitwit you could say “well do you have a better idea – like kicking fewer goals than the other side?” In fact the point is just to make people aware that he’s a nitwit, that his contribution doesn’t get us anywhere and that we’ll need to do something cannier than that.

    So the problem is that a lot of discussion about policy is silly in ways identified in Don’s comment. It’s unrealistic to the point of being fanciful. The idea that someone critiquing the current fad calling for evidence based policy should assume that the critics of such slogans want to get out the Ouija board makes me laugh.

    I’m afraid my ‘solution’ is just more modest, more commonsensical description of policy development, one that’s more aware of the foibles of some One True Method, less faddish.

  15. Paul Frijters says:

    Don, Nick,

    I agree with all these observations, but I am still a fan of the mainstream economic position that pretends the world can be altered towards desirable ends if one thinks hard enough about it (‘the plan’). Why do I say this is the mainstream position? Think of the social planner in macro models, the ‘instrument’ literature, the ‘Homo Economicus’ view of the world, the institutional design and regulation literatures: all the mental models in these literatures are very much predicated on the idea that we should formulate goals and plan our way towards these goals. Of course there are dissenting voices within economics. Hayek and the other Austrians were more of the ilk ‘we dont know where we are, where we are going and what we want, but it seems to going ok so far so lets just keep going’, but mainstream economics has definitely bought into the illusion of control. I see many advantages to that attitude.

    For one, the illusion of control is really popular with the students who become enthusiastic about economics because they can pretend to themselves that they will solve the big issues of their day. This energizes them to think hard about these problems and it makes them think about what their own goals and means are: it shapes them. The illusion of control is hence a very uplifting and constructive attitude without which it is hard to motivate the students. If you like, the competition for hearts and minds of the next generation is being won with the use of an illusion. Unless you think economics on balance is a force for bad then that’s a good thing.

    Secondly, the ‘step back and think about what we want to achieve attitude’ is a centring activity, i.e. it creates goals and openness: it is inclusionary in that it allows others to work out whether they might have better ideas about how to reach the same goals, it allows one to more easily find partners who share the same stated goals, etc. In this sense, economics is prescriptive about the kind of decision making it wants to see happen, rather than a description of how it actually happens. Economists are thus merely being part of the landscape by voting for what they would see as rational decision making. Yes, there are plenty of economists who are unaware of the fact that what they are describing as reality is in actual fact moral doctrine, but such blindness is true of any ism and not by itself a reason to disagree with it. It is up to the individual self-aware economists whether they want to be part of moralising story, and many economists are in favour of it. I certainly am: I much prefer an open discussion about goals and means to the alternative in which there is nothing solid to debate because goals are fluid and means are opportunistic.

    Thirdly, the illusion of control is a constraining and self-correcting exercise. When reality becomes so out of line with the stated goal that the opposite is no longer a tenable story, plans or goals have to change. This constrains the politicians. If the stated goals are conspicuously not met, then new plans have to be drawn up, lessons have to be learned, etc. Having stated goals thus itself leads to a self-correcting process whereby the whole system gradually learns from its many mistakes.

    Fourth, the illusion of control is a very optimistic attitude which allows one to try grand experiments and to dream big. Without the illusion, it becomes much harder in a political situation to justify big changes or big experiments because the mantra ‘we dont really know what we are doing so lets not change anything’ will then blow the advocates of change out of the water. The misguided self-belief that the illusion is true gives the changers the mental fortitude to really go for their plans and to ignore the sceptics. In this sense, the illusion of control is needed at the macro-level in order to generate experiments from which the more conservatively inclined on-lookers can then draw their conclusions as to what works and what doesnt.

    On reflection hence, the illusion of control and the planning mentality that goes with it has a lot to be said for it in social choice situations, even though it is perfectly true that it makes a very poor description of political reality. The real question is, once again, whether we can do even better, i.e. whether we can have our cake and eat it, by, for instance, having the majority buy into the illusion of control whilst explicitly educating a minority into the way things actually go. That is perhaps a question of executive training versus rank-and-file training. However, such experiments too are up to market forces. It needs someone passionate enough about the futility of the illusion of control to push it…

  16. Don Arthur says:

    Paul – I’m intrigued by your suggestion that an illusion or false belief can be a good.

    Are you aware of any literature on the economics of false beliefs? I wondered about this in a Troppo post a couple of years ago.

  17. Tel says:

    For one, the illusion of control is really popular with the students who become enthusiastic about economics because they can pretend to themselves that they will solve the big issues of their day. This energizes them to think hard about these problems and it makes them think about what their own goals and means are: it shapes them. The illusion of control is hence a very uplifting and constructive attitude without which it is hard to motivate the students.

    Maybe they should be putting the energy into Engineering or Physics? Admittedly, those have their own illusions but for the most part less than Economics does.

    Thirdly, the illusion of control is a constraining and self-correcting exercise. When reality becomes so out of line with the stated goal that the opposite is no longer a tenable story, plans or goals have to change. This constrains the politicians. If the stated goals are conspicuously not met, then new plans have to be drawn up, lessons have to be learned, etc. Having stated goals thus itself leads to a self-correcting process whereby the whole system gradually learns from its many mistakes.

    Oh very well then, they should be studying Cybernetics. The first year students who can demonstrate a knack for controlling tank levels, furnace temperatures and conveyor belt speeds are selected to move to a higher grade attempting to control more complex situations. I get annoyed with Economists who seem to maintain the illusion of control, but can’t actually explain what a PID controller is.

    I’m plugging through the James G March stuff on foolishness and my first feeling is that March confuses rationalism with reductionism. Probably a fair call back in 1971 but the reductionists have come to an amicable settlement with the systems theorists in more recent years and are happy to embrace each other under the very large vaulted ceiling of rational thought.

    Also, there’s nothing irrational about intuition. Human brains are known to be high power statistical correlation engines so if you have one handy might as well use it for what it is worth. Note that the human brain’s intuitive statistics also have notable weaknesses (e.g. Bayesian reasoning) so be advised to know the advantages and limitations of your tools (egats! rational meta-knowledge).

  18. Paul Frijters says:


    yes there is such a literature, following a much larger literature on this in psychology where they have essentially found that only the socially inept and manic depressives go through life without a whole pantheon of self-promoting illusions. If you drop me an email I will send you an ARC proposal by a set of economists that looks into this, as well as the ensuing papers that have come out of that project sofar.
    The argument as to whether you can realistically expect truth in public debates was already handled here: http://clubtroppo.com.au/2007/06/28/can-you-handle-the-truth-ii-does-everybody-lie-and-does-it-matter/

  19. Nicholas Gruen says:


    Having read your comment 16, I’m not really sure what to say. It certainly reminds me of the joke in which someone gets rooms and rooms of monkeys typing because if he does it with enough monkeys for long enough one of them will eventually type out Shakespeare’s Richard III.

    His whole life he gets more and more monkeys, more buildings, more typewriters, and eventually as an old man, he sees one of the millions of monkeys in one of the massive buildings he’s funded take out a clean sheet of paper and type “Now is the winter of our discontent made glorious summer by this son of York” and for days and days the monkey types to growing excitement around the world. One clean sheet of paper after another. On was is clearly going to be the final page, he types “A horse, a horse, my kingdom for a hor&cse%@wea1fwq? eca !w;asdf##sdfad”

Comments are closed.