Democracy Symposium: Going Marginal



Murali did his undergraduate degree in molecular biology with a minor in biophysics from the National University of Singapore (NUS). He then changed direction and did his Masters in Philosophy also at NUS. Now, he is currently pursuing a PhD in Philosophy at the University of Warwick.

Related Post Roulette

23 Responses

  1. Avatar david says:

    What domains of public policy are left after we have done all of this? Is there any other role at all for democratically elected bodies? And is what is left so terribly important that delegating it to some other independent technocratic body or constitution or for that matter, the market would be a mistake? Can such a form of government where so little is left for democratic bodies to handle rightfully be called democratic?

    Well, taking separability to its logical conclusion. There’s technocratic policy as in, “the most efficient ways to achieve some goals are X, Y, and Z”, and politics as in “our goals should should be A, B, and C”. What’s the easiest way to legitimize a claim that some aggregate goals, which may necessarily be non-Pareto-improving goals, are the deserving ones?Report

    • Avatar david says:

      Hmm. Murali works in Singapore.

      A couple of local political scientists have commented – I think Dr Cherian George was one of them, but I can’t really be sure – that the state ideology in Singapore has a notion that the only legitimate and respectable state goal is maximizing material economic wealth. It’s not really explicit, it’s just that anyone who forwards other goals from the left gets mocked as impractical, and from the right as endangering the fragile ethnic peace.

      This lends itself to drily efficient technocracy quite well, and (let’s be fair) the incumbent regime has done a pretty good job of efficient technocratic pursuit of economic growth. So the ideology is one which validates the regime ex post. It’s not quite the explicit acquiescence-for-rapid-growth of the PRC, rather some very successful entrenchment of the values of the incumbent regime amongst its people; prescribed cultural identities are accepted.

      But of course, many other polities have other notions of goals, particularly when identity politics is legitimized as fundamental to the advancement of human liberty instead of ferociously condemned as destabilizing. Then the politicization of distribution between identities cannot be avoided.Report

    • Avatar James K says:

      This is one of the reasons I’m intrigued by Futuarchy. The values government is to pursue are decided democratically, but the method of achieving those values is left up to prediction markets.Report

      • Avatar david says:

        That really just squirrels the hard bit of technocratic judgment into how the security used as a proxy for future outcomes is designed.

        The neoclassical instinct moves in favour of prediction markets, but the game-theoretic instinct tells me that designing auctions to work well is really, really hard, and designing auctions where the outcome could have a massive impact on asset values elsewhere and thus permits investors to complicate their incentives in a way that drags in all the hell that is general equilibrium in Arrow securities is… well, not something I’d assume to be possible without proof.Report

        • Avatar James K says:

          Prediction markets don’t use auctions, they work via exchanges. As far as manipulating them goes, all the research I’ve seen indicates they’re effectively impossible to manipulate, and are in any event much harder to influence than elections.

          I don’t think futuarchy is ready for prime time yet, but I think it merits more investigation.Report

          • Avatar david says:

            In the economic sense of an auction, of which securities exchanges are generally a subset.

            I reiterate the point that the difficulty of policy design is not solved, just hidden. Generic futures markets have the nice element that we assume that participants don’t really have an opinion about (say) what the future value of pork should be, we assume that they just want to maximize their portfolio of pork and non-pork assets, whatever those may be. A decision market starts blurring the boundaries here. A decision market whose price outcome itself influences the values of the securities reflecting each decision raises all the !!!MULTIPLE EQUILIBRIA!!! alarms.Report

            • Avatar James K says:

              No, I’m really not seeing the problem here. Decision market securities converge on best estimates of probabilities very reliably. Plus, unlike other securities, the value of decision market securities is constrained – they can’t rise above an implicit probability of 1.Report

              • Avatar david says:

                How do those properties help?

                Here is an easy decision market: bet on the weather tomorrow. There are lots of methods of collecting data out there, and nobody has any way of influencing the weather that we know of. In other words: the people with relevant information face liquidity constraints in the same ballpark as other people with relevant information, and furthermore the outcomes being bet on are exogenous to the values of the decision-market securities.

                That’s not true once you actually make policies depend on that decision market outcome. Then you have the possibility of (1) agents who are prepared to make a loss in order to influence a decision in some way that will make them a profit elsewhere (and Real Life has liquidity constraints, so such agents can operate) (2) multiple equilibria, where the relevant probabilities are endogenous to the policy being made, which messes up arbitrageurs who are supposed to fight (1) to begin with.

                You can’t just offer Arrow securities on the values of some social-welfare-function of possible policies and then pick the policy with the highest SWF as assigned by the decision market; policies do not exist in a vacuum and investors adjusting their portfolio affect the SWF in other ways. e.g., if I DM’ed a choice of more rail lines or more bus lines, an investor isn’t just going to buy Arrow securities, they’re also going to buy carriage companies or bus factories, which affects the real benefits of rail vs. bus, and this makes the value of the security feed back into the probabilities over the outcomes.

                And the technocratic problem of designing the SWF over outcomes is, yes, buried, not eliminated… Arrow’s theorem prevents this from being democratically decidable in any obvious sense.Report

              • Avatar James K says:

                That’s not true once you actually make policies depend on that decision market outcome. Then you have the possibility of (1) agents who are prepared to make a loss in order to influence a decision in some way that will make them a profit elsewhere (and Real Life has liquidity constraints, so such agents can operate) (2) multiple equilibria, where the relevant probabilities are endogenous to the policy being made, which messes up arbitrageurs who are supposed to fight (1) to begin with.

                Ah I see, thank you for your patience.

                As it understand it point 1 has been tested experimentally and it turns out that it doesn’t work. Attempts to manipulate the market inevitably fail – undermined by arbitrage. Point 2 on the other hand, I could see how that could cause problems. It’s definitely a weakness of futuarchy, but I suspect it is manageable in practice. You would need to make sure the SWF was built on a large number of high level macro variables, and ideally not on the direct outputs of other financial markets. I don’t think it’s a fatal problem, but it’s definitely a weak point.Report

              • Avatar david says:

                You can’t do that, either, for the same reason that an investment market dominated by index funds will produce index funds that do badly. The goal of arbitrage is not for arbitrage’s sake; the goal is to reward individuals who believe they have superior information for revealing this information, and a high-level SWF bundles too many extraneous conditions. You are demanding that these individuals purchase a considerable amount of risk to garner any benefit. They could almost certainly do better just buying ‘index funds’.

                And really, liquidity constraints matter for real-life decisions. The decision which democracies make most badly are precisely those with a concentrated group of special interests who monopolize valuable information regarding policy consequences. Unfortunately, the conditions where futures markets perform undeniably well are also the conditions where polling people also does well.Report

  2. Nob Akimoto Nob Akimoto says:

    War making is definitely the most important thing a state is capable of doing and perhaps the one area where technocratic expertise should be kept to the background as much as possible. The Best and the Brightest aside, there’s certainly plenty of other situations where letting a bureaucracy run itself didn’t really end well when force was involved.Report

    • Avatar Kolohe says:

      “War making is definitely the most important thing a state is capable of doing and perhaps the one area where technocratic expertise should be kept to the background as much as possible”

      Though the state has pretty much made war the central focus of its technocratic expertise since Sun Tsu.

      (and curiously, the most recent use of military force that the American government has undertaken – Odyssey Dawn – was rather completely divorced from the political process)Report

  3. Avatar Don Zeko says:

    Quis custodiet ipsos custodes? I think you are vastly overrating the effectiveness of non-democratic forms of decision-making. I simply don’t see how prediction markets or panels of experts will be proof against corruption or capture. To take the most obvious example, suppose we follow your advice and create a panel of economists to set fiscal policy (as an aside, how many on this panel? 5? 50? 500?). How will we select these economists? Who even determines who is or is not a credentialed expert in the field? It seems that all this suggestion does is take us from a world in which we send tremendous amounts of time and money influencing elections to a world in which we spend tremendous amounts of time and money trying to influence the selection of the unaccountable experts who will sit in judgment of us. I, for one, don’t look at the modern Supreme Court confirmation process and think “gee, wouldn’t it be great if we made all of our decisions like this?”Report

    • Avatar Murali says:

      How about selection of the chairman of the fed? How much heartache goes into that? A large part of the BS that goes into supreme court confirmation is culture wars idiocy.Report

      • Avatar Don Zeko says:

        Are you really confident that a board empowered to decide the funding level of Medicare or whether taxes rise or fall would look more like the Fed than the SC? I would suggest that the Fed is a bit of an outlier; organized American politics is strangely uninterested in monetary policy. Put fiscal policy on the docket and that de-politicization evaporates.

        Also, pointing to the Fed doesn’t really address the other problem of this idea. How exactly are we credentialing or empowering experts? Who decides who is a member of these bodies and how do we ensure that their decisions are viewed as legitimate and binding?Report

        • Avatar Murali says:

          Actually, I was thinking more in terms of a permanent civil service which would tend to regulate itself and choose its own members. They would get bonuses everytime their policy outperformed their targets. So naturally, they would choose people with PHDs in the relevant fieldsReport

          • Avatar david says:

            Who staffs the agency that measures these targets?

            And that’s the easy question. Here comes the hard one: who makes the judgment calls when a conflict-of-interest exists for senior civil servants? e.g., If I own a lot of land, I would object to a land value tax if I thought it would be shown, technocratically, to be a good way to acquiring revenue to do whatever (like cut taxes elsewhere, perhaps). And so it would not be in my interest to hire the best tax economists I can find. Even if better economists would increase my bonus, I would lose my land rent.

            And of course, material interest aside, senior civil servants can have prejudices or ideological commitments too – commitments that they might be prepared to sacrifice some income for.Report

    • Avatar James K says:

      All the research on prediction markets I’ve read says that prediction markets are much harder to manipulate then elections – in fact it appears they are effectively impossible to manipulate.Report

      • Avatar david says:

        But you go ahead and have the election anyway. You don’t go “welp the market says Gore will be the winner, so let’s save ourselves the trouble and appoint Gore as High Technopriest, yay to whoever bet on him!”

        See the difference?Report

        • Avatar James K says:

          The proposed model for futuarchy doesn’t work like that. It looks at policies rather than leaders, and it evaluates the outcomes of policies against a democratically-determined social welfare function, not just by predicting what option is most popular.Report

  4. Avatar Roger says:


    Great post as usual. I think I agree, though I still need to follow your links to get more of an idea of how technocratic experts and prediction markets would perform.Report

    • Avatar Roger says:

      One immediate concern is that if being defined as an expert has power privilege, then we will no longer have experts. We will have people brokering for power in the guise of experts as revealed in the extremes of environmentalism (both sides) and creationism.

      Even the IGM board of economist experts are weighted far to the left of mainstream society. Though I pretty much agree with the consensus on just about everything, so that is a point in your favor.Report

    • Avatar Murali says:

      technocratic experts crafted Singapore’s healthcare policy, which is fairly close to your own ideal system.Report