The Socialist Calculation Debate, with Prolegomena to Any Future Metamarkets. Part I: Really Hard Math.

Jason Kuznicki

Jason Kuznicki is a research fellow at the Cato Institute and contributor of Cato Unbound. He's on twitter as JasonKuznicki. His interests include political theory and history.

Related Post Roulette

111 Responses

  1. Mike Schilling says:

    What all this I hear about meat markets?Report

  2. zic says:

    I think no matter how virtuous, intelligent, and well-informed the allocators are — and let us not doubt for a second that the very best people always serve in the federal government, incorruptibly and to the best of their abilities — still, the socialist calculation problem remains insoluble.

    This is so tiresome. The notion that the allocators are some other group, separate from you and me, some government over there that has nothing to do with you or I, except to dictate to us, is just ludicrous.Report

    • Jason Kuznicki in reply to zic says:

      This is so tiresome. The notion that the allocators are some other group, separate from you and me, some government over there that has nothing to do with you or I, except to dictate to us, is just ludicrous.

      I don’t understand. If they form the plan, and if I don’t form the plan, and if I just have to accept passively whatever they offer — then how are we and they not separate?

      And if the state is instead democratic, and if I’m outvoted — then, once more, how are these two entities not separate?

      What we have here is a principal-agent problem, compounded by the fact that each individual is a principal in the problem and the agents aren’t going to be able to make everyone happy.

      I’m sorry that you find this tiresome, but finding it tiresome doesn’t mean that you’ve solved the problem.Report

      • BlaiseP in reply to Jason Kuznicki says:

        I’m sorry that you find this tiresome, but finding it tiresome doesn’t mean that you’ve solved the problem.

        It means you haven’t defined the problem. Where are these Nefarious Plans? And who’s making them? The Gummint?Report

      • Patrick in reply to Jason Kuznicki says:

        You have the principal-agent problem in market actors, too, though, Jason. It’s not like this is a problem that only infects the public sphere.Report

        • Jason Kuznicki in reply to Patrick says:

          Freely granted. But I never go around saying, “Don’t you see? Your insurance company is you, man!”

          I notice that people do that with the government all the time. It’s just plain false. That’s the point here.Report

          • James Hanley in reply to Jason Kuznicki says:

            JK’s point here is more on point than mine below. (Mine is also correct, just not as on point.)Report

            • Gaelen in reply to James Hanley says:

              I don’t know how persuasive that analogy is, in large part because for most of the products we buy the principal-agent problem is about advice and information not actual decisions.

              I think the ‘government is us’ argument is much closer to saying that ‘the shareholders are company X,’ a proposition that I don’t have much trouble with. The shareholders are the people in whose interests the decisions are, at least nominally, to be made, and who have some control, however small it may be.

              If not the shareholders, then who is the corporation (or is that even a worthwhile question)? While the managers and directors make the decisions, I find it hard to say that the agent, rather than the principal, is the company (or government)Report

          • Chris in reply to Jason Kuznicki says:

            This is one point on which you and I agree completely. I find the “the government is us” bit to be a sad indictment of whichever political group is uttering it (and members of both of the major parties in this country use it when it suits them).Report

        • James Hanley in reply to Patrick says:

          But less so. My agent acts for me, and not for a multitude, and while the agency problem necessarily exists* I can fire him the moment I get suspicious. When my agent represents a multitude simultaneous and I have no direct authority to hire or fire, the problem is compounded immensely.

          *One of my favorite examples in agency problems in the markets is with real estate agents. A study a year or two ago showed that they sold their own houses more quickly than their clients’ houses–with their own they were unwilling to bear the costs of keeping it on the market longer, while with their clients’ they were focused on the size of their commission.Report

          • Kimmi in reply to James Hanley says:

            That’s not the analysis I’ve seen, on real estate agents. Might be that they’re employing “off the books” selling… Might simply be that they’re more accurately evaluating prices and setting fair ones. There is NO incentive for a selling agent to give you an accurate valuation, and quite a bit for them to set the price high (so that you dont’ just choose the swindler next door, who will tell you your house is worth so much more).

            And real estate agents you sign a contract with, for a year. You lose substantial value from firing the person. And they do not have your best interests in mind.


            • James Hanley in reply to Kimmi says:

              Indeed, I got it backward. They persuade clients to sell more quickly (because the incremental gain in their commission isn’t worth the wait), but hold their own longer to sell for higher prices. Bad memory on my part. Here’s the NBER paper.

              And real estate agents you sign a contract with, for a year.
              Eh, maybe where you live, but not as a general rule.Report

      • Kimmi in reply to Jason Kuznicki says:

        ” If they form the plan, and if I don’t form the plan, and if I just have to accept passively whatever they offer — then how are we and they not separate?”

        You have many levers to pull.

        For one thing: you can propose your own plan. Now, I know you’re a good writer, Jason. And you might not be as good a psychologist as some of my friends, but you aren’t as horrid at that as I seem to be.

        So why haven’t any of your plans been acted upon????!?

        Are you failing to put enough effort into proposing plans? (we can’t actually expect every single idea to be acted upon…)

        Are you really that bad with numbers, that your plans wouldn’t work?

        I write this from a perspective of knowing someone who’s gotten multiple plans enacted (and helped veto a few at higher levels of gov’t).Report

    • Jaybird in reply to zic says:

      Have you ever met your Congressman?

      I have seen mine, once.

      I had numerous interactions with my previous Congressman’s staff… I talked to the INS Liason every week for a few months there and then, on our 6th Anniversary, we went over to the office to talk to the INS Liason for a minute or two (sort of a “hey! Remember me! I’m still married!” conversation that I thought might brighten her day) and I shook the hand of my Congressman’s Chief of Staff. (As we were walking back to the car, I pointed out that “that guy had the softest hands I’ve ever touched” and Maribou said “YEAH I KNOW!” (and, please, keep in mind: I’m in IT).)

      Anyway, all that to say, I’ve had more interaction with my Congressman’s office than… yeah, I’m willing to say more than anybody I know. And I’ve seen my Congressman, in the flesh, only once.Report

      • zic in reply to Jaybird says:

        Yes, I have met my Congressional representatives; both state and federal.

        They return my phone calls. They’ve contacted me directly to ask me to participate in various public efforts.Report

        • Jason Kuznicki in reply to zic says:

          And did they give you exactly the healthcare system you wanted?Report

          • Mike Schilling in reply to Jason Kuznicki says:

            Did the free market?Report

          • zic in reply to Jason Kuznicki says:

            Politics is the art of compromise, Jason. Just like marriage, family, work: we give something and we get something and not everybody is happy. I never expected to get ‘the system that I wanted,’ I expected to get a system that served more people better then the old system did; and I expected that there were very many ways to get to that point.

            Which is my point. I know my congressional representatives (and Senators) because I communicate my wants/needs to them in national debates. They are not some ‘other,’ they, like me, are citizens. The people who work to execute the laws they pass, like you and me, are also citizens, often working hard to do a good job in difficult circumstances.

            You have many worthwhile insights; I value your thoughts and opinions. But your anti-government bias is sometimes no different then any other bias to my mind; it leads you to presumptions and assumptions that I don’t think are merited.Report

            • James K in reply to zic says:

              I don’t think anything you’re saying is contradicting what Jason is saying. You compromise at work, but no one says you are your boss. Equally, you comprise at home and no one says you are your spouse and children.

              Jason is merely saying that while government exists to serve the will of the voters (this is the agent part of principal-agent), its incentives are not perfectly aligned with the voters and it therefore it needs to be modelled as a separate entity, and not a simple extension of the voters’ will.Report

        • Jaybird in reply to zic says:

          How representative is your experience within your circle? Because, lemme tell ya, I am a *HUGE* outlier in mine.Report

          • Chris in reply to Jaybird says:

            I met my congressman every year for a few straight years when he showed up to the opening day of my son’s Little League season. I shook his hand, had him take a picture with my son, and said thank you each time. Unfortunately, I still don’t have the yacht I want the government to give me.Report

            • Jaybird in reply to Chris says:

              Did the free market give you a yacht?

              Anyway, the point I think I was trying and, yeah, failed to make is that my interactions with the government are pretty much limited to voting and taxes and I’m one of the people who has interacted with my representative *MORE* that pretty much anybody I know.

              It’s not surprising for me to hear that people think that the allocators are some other group. For me, that is exactly what they are… and similar is true for pretty much everybody I know.Report

              • Kimmi in reply to Jaybird says:

                And your interactions with most corporations are about as context-free, are they not? Pay money, get services?Report

              • Jaybird in reply to Kimmi says:

                My interactions with corporations feel a lot more voluntary. For example, when it comes to my dinner tonight, there are literally dozens of options… from delivery to picking something up to eating out to going to one of several different grocery stores (madness!) and picking out from any of hundreds of pre-packaged options to any number of countless combinations of fresh ingredients.

                When it comes to my gummint?

                It feels a lot less choosy. I’m stuck between “being here” and “Somalia”.Report

              • Kimmi in reply to Jaybird says:

                I feel a lot more free about my government than I do about which cell phone plan I’m on…. or half a dozen other big corporations that I purchase stuff from.

                (one, i was even involuntarily pushed into a business relationship with!)

                Combined, I spend a lot more on them than I do on taxes.Report

              • greginak in reply to Jaybird says:

                And if Applebee’s and 7-11 could provide for cops, health care, public goods etc etc etc than that might be the start of something. Since they can’t well…Report

              • greginak in reply to Jaybird says:

                I have briefly met my Rep. I’ll never come within an unguarded mile of anybody involved in the LIBOR rip off. I think LIBOR had some not insignificant effect on the economy.Report

              • Chris in reply to Jaybird says:

                Yeah, I was just being silly, saying “I’ve met mine more than you’ve met yours! Nah nah nah boo boo!”

                I think asking whether you have the health care system you want from the government or the market is pretty silly, because it doesn’t actually get to any point about how either of them works, and it really says nothing about whether I am either (I’m not either; and I’d like to do away with both!).Report

          • Kimmi in reply to Jaybird says:

            Yaknow, If I had bothered to show up to NetrootsNation, I could have met gobs of different senators etc….
            I was unemployed at the time, and the fee ($200) was a mite bit pricy.

            I hear they destroyed the Warhol with excessive partying. Andy would have been proud.Report

      • I’ve met the entire Congressional delegation from Maine. It helps to come from a small state and donate to high-profile causes.Report

        • I find it easy to conclude that the commentariat/bloggertariat at the LoOG is extraordinary.

          Is meeting one’s congressperson that common out there?

          I’m beginning to feel ripped off.Report

          • North in reply to Jaybird says:

            I shook Al Franken’s hand. We talked for a while about defeating the SSM amendment. Technically though he’s a Senator so I’m not sure that counts.

            Full disclosure, I was impressed, barring him doing some kind of astonishingly bad thing Franken’s got my support for Senator for as long as he cares to go after the job.Report

          • Kimmi in reply to Jaybird says:

            I know someone who routinely vets candidates. He’s kvetched at Obama in person.Report

    • Patrick in reply to zic says:

      That’s not very charitably framed, but it’s a valid point to some extent. It also points out the weakness in comparing free markets to socialized distributions… because in reality we’re not really doing that, which is my objection to the piece.

      We’re not typically comparing “a completely free (theoretical) market” to “a completely planned (theoretical) economy”.

      We’re comparing a “mostly free market” to a “somewhat regulated economy”, and both of those things are well beyond an epsilon neighborhood of their starting points. Hell, they’re a lot closer to each other by several orders of magnitude than they are to either of their starting points.

      The allocators (such as they are) are appointed by people, and there are representative problems with that, and efficiency problems with that… but they’re also not planning a centralized economy, either.

      They’re attempting to correct for some market failures (which I hope we all agree exist) by instituting some corrective measures that admittedly cause some deadweight losses when compared to a direct market allocation. But given that a lot of those corrective measures are compensating for externalities, it’s not clear to me that this isn’t really a deadweight loss in all cases, either.

      Back to the OP:

      But the remedy for that is to assign ownership, and thus to extend rather than contract the market order.

      Jason, this sounds to be overly optimistic. In some cases, I’ll agree, this works out (provided “ownership” comes with “responsibility for externalities”… which it typically does not historically speaking), but I don’t know that it’s a useful universal approach (see previous parenthetical).Report

      • North in reply to Patrick says:

        Yeah Pat, heck I’m bemused that in my uninteresting comment I pretty much outlined that very response in advance. Yes the libertarian solution to commons is extend ownership to the commons and where the commons can’t be practically or literally apportioned to private owners the solution is to LOOK PINK LAMA! *handwave*handwave*

        That said the tragedy of the commons is incidental to the point of either his original post or this one. Jason’s certainly not claiming infallibility for libertarian-ism.Report

        • Jaybird in reply to North says:

          Let’s just hammer this out: where the commons cannot be practically or literally apportioned, the solution is to assert ownership? “You can’t fish here. On pain of getting your damn boat sunk. Yeah, I’m serious.”

          That’s how I’m reading what you’re implying. Am I misreading?Report

          • North in reply to Jaybird says:

            I don’t think I’m implying any such thing, Jay. I’m puzzled where you’re getting it from. Even if you could, maybe, assert by force ownership of, say a fisheries (though good luck on the really big ones), even that wouldn’t work for bigger commons, say the atmosphere.Report

            • Jaybird in reply to North says:

              I may have been drinking when I wrote that comment, originally.

              I was thinking that the other “solution” to the tragedy of the commons usually tends to be legislation of some sort. Usually of the form “don’t do that”.Report

              • North in reply to Jaybird says:

                Well yes, there’re non-libertarian answers about the commons. Your standard statist one if to have a polity that encompasses the commons in question and then prevent the commons from being debauched via legal force/regulation.

                But that’s somewhat limited to, when you think about it. States can only effectively enforce their writs over commons they have control over/encompass (or over commons that multiple states encompass if they can all agree). So Iceland can preserve the fish stocks within their territorial reach* but Iceland cannot preserve the ozone by itself. You need some kind of global government to effectively apply statist solutions to global commons. Then again (contra anarchists) wouldn’t you need a global government to recognize/protect property rights over global commons to allow the most common libertarian solution to commons as well?

                *though mostly by enforcing kindof an ownership/market mechanism through owned catch quotas in their case.Report

              • Jaybird in reply to North says:

                I see more and more nudges toward a trans-national system. We need western-style regulations of the third world, after all. The intercommunication and interdependence demonstrated by the internet, of course. And, yes, the commons that we won’t be able to fully address until our jurisdiction is recognized.Report

              • North in reply to Jaybird says:

                I’m pretty skeptical about the former, not clear on the second but the latter seems valid but also nowhere near dire enough (until something catastrophic and dramatic and visible starts happening) to overcome the institutional bias against such a thing.Report

              • Jaybird in reply to North says:

                It’s not that tough. Just look at any progress made by the 3rd World since, oh, the 90’s.

                Then tell them “you didn’t build that”. At the same time, bribe the crap out of the elites.

                You’ll be amazed at how things fall into place.Report

    • Brandon Berg in reply to zic says:

      This is so tiresome. The notion that the allocators are some other group, separate from you and me, some government over there that has nothing to do with you or I, except to dictate to us, is just ludicrous.

      This is entirely beside the point. A collective is no better at making these decisions, and arguably much worse, than a group of elite planners.

      The problem isn’t that the wrong people are making the decisions—it’s that they’re being made at the wrong scale. Decisions that should be made at the individual or firm level, and tailored to their specific circumstances, are being made on an economy-wide level.

      Suppose that we were to take a national vote on what to have for dinner each day. No representatives, just direct democracy. Sure, the outcome would be a product of collective preferences. And a plurality of voters would get exactly what they wanted for dinner. But we still end up with the majority of the voters eating something other than what they want, for no good reason. Political decisionmaking works very badly, and should only be undertaken when individual decisionmaking simply isn’t an option.

      One example is air and water pollution. It’s not possible to decide on an individual basis how to manage the trade-offs between industrial production and clean air and water, so it makes sense to make those decisions politically. It’s a sure bet that the political process will make a complete and utter hash of it, but there’s no real alternative.Report

  3. Chris says:

    I was one of the ones who raised the “something could replace it someday” point in the last thread, but I didn’t do it as an objection. I merely wanted to put it out there, because there are people, a lot of them, who think that markets as we know them are fundamental to human nature (that is, that we were born as Hayekian creatures, in your terms), and we’ve now arrived at the best and most natural system (or at least an approximation of it, which would be even better if those pesky statists would just get out of the way!). Of course, there are also people who think that markets are wholly unnatural, and they’re equally wrong, but they didn’t seem particularly relevant in the last post. I didn’t mean to include you in either group, really, as I thought your description of the Hayekian creature made it clear that you didn’t belong in them.Report

    • Kimmi in reply to Chris says:

      Markets are wholly unnatural. They’re quite made up.
      But humans like make-believe, and we’re good at games.Report

      • Chris in reply to Kimmi says:

        What does wholly unnatural mean, there? I mean, languages are made up too. Are they “wholly unnatural?”Report

        • Christopher Carr in reply to Chris says:

          Good point. What if I’m a naturalist?Report

        • Kimmi in reply to Chris says:

          … bear in mind I’ve a touch of psychophysiological training… A language is made up, but we have certain innate (or nearly so) capabilities that we seem to build language off of.
          Likewise, our eyes capability of distinguishing hues has a great deal to say about what we call different colors (basically, how many we distinguish, and where we draw the lines).

          Colors and linguistic aptitude in general seem to be “made-up” stuff mapped over biological reality.

          I don’t consider a market to be any different than any other game. It doesn’t correspond to a reality, just like a computer game. We make rules for markets, so that we are more likely to participate (as it’s more fun to have a market with more people in it).

          I guess here’s the difference: in wholly unnatural things, we write /all/ the rules, and can change all of them as fancy pleases.

          in more natural, more reality-driven things, we can’t. We can slow down the speed of light (by shining it through an object), but we can’t do so without respecting actual physical laws.Report

  4. Christopher Carr says:

    “We can know this for a fact because a Walrasian equilibrium would leave no goods or raw materials wasted — and yet in the real world, both are constantly wasted. ”

    Objection. This idea itself is a non-sequitur:

    If no goods or raw materials were wasted then there would be no dynamic economy. By definition all goods and services would be already being put to whichever use maximizes utility. Subsequent economic exchange would necessarily make each party worse off.

    By the way, thank you so much for breaking this up. I’m finding myself unable to keep up with all the great posts around here lately, and it helps to think of one thing at a time.Report

  5. Shazbot5 says:

    “As it turned out, the biggest objection seems to have been that markets don’t do much that couldn’t be accomplished by sufficiently intelligent and fair-minded people allocating resources according to scientific calculation: The price system adds little, if anything, to process of simply sharing ideas.”

    This was not the objection that Brian or I raisd in that thread.Report

    • LeeEsq in reply to Shazbot5 says:

      This. The opening sentence doesn’t capture any of the critics of the market that I know about under liberal or socialist thought. The critique is more of a critique of the culture and practices of business in general rather than the markets in particular. A critique of the inefficiencies of the free market in allocating necessary goods like healthcare or education is part of modern liberal or socialist theory but its not the whole of it.

      The disaster in Bangledish this week isn’t really a critque of markets. Its a critique of certain business practices that discount things like adequate safety measures and worker’s rights. None of this has to do with whether a market in clothes is the best way to allocate clothing to the world’s population.Report

  6. Shazbot5 says:

    Brian and I argued that over generations we test different ways of doing things and societies pick the better way slightly more often. This is not a financial market mechanism.

    Markets are good in many cases, but often they fail, e.g. as in healthcare, where socialist systems or markets mixed with regulations and socialism work better.Report

  7. Shazbot5 says:

    I stipulated that maybe, maybe one day 100 or more years from now markets could become unnecessary.Report

  8. Shazbot5 says:


    Did you read the Shalizi post at CrookedTimber that Brian linked to in that last thread?

    It is pretty excellent and may make the argument that you want to make that the math is too damned hard, even for future supercomputers.

    Me, I think predictions about the future about what computers will or won’t be able to do is a mug’s game, just like most futurism. Maybe they will. Maybe they won’t. Humean skepticism and all that.

    That said, I think it is a bit unclear what we disagree about.

    Do you think you can prove a priori (with some definitions about markets and human nature) that markets will always be more efficient or better in some way than socialist systems, not just in the aggregate, but in every area, including (for example) healthcare?

    That’s a lot to prove a priori, IMO,Report

    • Jason Kuznicki in reply to Shazbot5 says:

      I don’t agree that the math is too hard. Please read my post, I think you’ll be surprised. The math appears to be doable, and, as I suggest, something else is the real problem.

      That said, I’ll take a look at the other post…Report

      • Shazbot5 in reply to Jason Kuznicki says:

        I thought you were implying that you would argue that the math is too hard in the next post, because you were saying that there must be something wrong about the 1993 paper by Cockshott (snicker) and Cottrell.

        But my apologies.Report

        • Jason Kuznicki in reply to Shazbot5 says:

          No worries, the next one will be up probably over the weekend. It’s where the really interesting stuff lies, I think.Report

        • Shazbot5 in reply to Shazbot5 says:

          And the phrase “read my post” makes me think I have angered you, so I will stop posting.Report

          • Jason Kuznicki in reply to Shazbot5 says:

            I owe you an apology.

            I admit I’d inferred that you hadn’t read the above post through to the end.

            I did so because the other interpretation had not occurred to me, namely that you expected my next post to produce a reason why the math still doesn’t work after all. On reflection, that’s a perfectly reasonable reading.

            Still, I think the math may very well work. The real problem isn’t that the math can’t be done. It’s that the math doesn’t model reality. For some very interesting reasons that I will soon discuss.

            I’m admittedly not an expert in the area, but there seem to be some experts who say the optimization problems can be solved. That’s not the same thing, though, as saying that these types of problems are good models to use.Report

            • Shazbot5 in reply to Jason Kuznicki says:

              No need for apologies.

              I really like your posts and our arguments. You never take the easy, cheap way out of an argument. And you often find some interesting arguments and ideas. We agree about nothing, apparently, but I respect your arguments,Report

  9. Morat20 says:

    Hmm. The use of “socialism” in “scientific socialists” would be better rephrased as “market socialism”. It makes your point more clearly, as otherwise your use of “socialist” could be alluding to a variety of things — including a regulated free-market economy with a broad social services system, which isn’t the same thing.

    Walras wasn’t talking about, say, England (not even England of the 60s with quite a few nationalized industries) but given the frequently…imprecise..usage of the word ‘Socialism’ and it’s subsequent shift in meaning in casual conversation (at least in America), being more precise is probably a good thing!

    (For instance, it’s pretty obvious Blaise misread it upthread — market socialism is, well, not exactly widely practiced. Or at all, really. I can’t recall the differences offhand, but I remember the big “socialist” economies of today aren’t market socialism (China’s not, despite the big Socialist in the name) in the Walras sense.

    I did have to turn to wikipedia to remember all the jargon: FWIW, apparently European style capitalism is styled “Social Market Economies” (which is very distinct, again, from the market socialism you’re critquing here) and apparently the US is “Anglo-Saxon economy”. Go figure. 🙂

    Anyways, if you’re gonna be talking about this — probably a good idea to either use the formal jargon (so us plebes can go look it up or refresh years old classes) or define, strictly, what you’re talking about.

    After all, what’s on point for “market socialism” migth be entirely baseless for “Social market economies” and it’d be really easy to confuse the two and end up drawing invalid conclusions.Report

    • Jason Kuznicki in reply to Morat20 says:

      I’m using “scientific socialist” really only as a contrast to “utopian socialist” here. The latter thought human nature would change; the former thought we needed to calculate.

      Market socialism is (to me, anyway) more like what Abba Lerner and Oskar Lange wanted — a market for consumption goods, but no market for capital goods.

      There are intersections all over the place, of course. I just needed a convenient handle for the type of socialists who wanted to do planning, as opposed to the type who did not.Report

      • Morat20 in reply to Jason Kuznicki says:

        Yeah, but there’s all these labels for a reason — it might all fall under the heading of ‘socialist’ in the very large, but those labels define what they think should be planned — and to what extent.

        I can see not wanting to get bogged down in the weeds — but if you’re overly general, your points are likely to fail when addressed to specific examples. If you’re overly specific, you’re probably gonna have a forest/trees issue.

        Take three types of technical socialists: The market socialists, who believe everything should be centrally planned and optimimum efficiency can, in fact be computed. (This may or may not be true in theory, but is highly unlikely to be true at the moment). There’s the China and USSR style socialists, who centrally planned only the big economic pillars — a lot of detail work was handled either in a decentralized fashion or even in a more classic market situation.

        And then there are European-style socialists, who tend to only want to plan those aspects of the economy that seem unsolveable by classic markets, or that suffer repeated market failures. (Now there’s a lot of ink that can be spilled over what is and isn’t an aspect that would require such intervention…)

        But if you render a criticism of, say, central planning solutions rendering sub-optimal solutions — that doesn’t even get past first base on European style setups, because they (or claim they do) only address areas of classic market failures. Sub-optimal is better than “failure”, right?

        I think where the rubber really meets the road is right there in the word ‘socialism’ — it’s “social”. Optimum market outcomes and efficiencies might not be socially optimal or acceptable. That doesn’t render the optimal market outcome any-less optimum, just moot — as market outcomes aren’t the only criteria.Report

  10. James K says:

    I really don’t see how the calculations are solvable Jason. For on thing you need the preferences from every person in the market to actually solve the equations, and that information simply isn’t available.

    Secondly even if you solve the equation the first time, people will change their actionsto game your price setting algorithm.

    But perhaps I’m anticipating points you’re planning to make in future parts?Report

    • Morat20 in reply to James K says:

      Nah, you could do it in aggregate. The market doesn’t react to me individually. It doesn’t CARE that I might want this model of car to come in Color X as a base option. It only cares if a LOT of people agree with me.

      Gaming algorithms and dueling algorithms wouldn’t matter anyways, since you’d be looking for stable states.

      Hmm. I bet you could handle it similar to a data mining problem. Evolutionary algorithms would be an ironic (yet quite on point) choice, and you could probably hash out fairly solid answers quickly. You’d never get to 100%, but what if 95% is better than what we’ve got?

      Computationally expensive, but not a direct examination of all cases. You’d need quantum computers to even try that — or have someone prove that NP problems are not, in fact, NP. (FYI: If anyone does develop an algorithm to solve NP problems — like the travelling salesman — in polynominal time, well…hope like hell the government solved it first and the banks quietly upgraded their encryption. It’d be Zombie-apocalypse level chaos).Report

      • Kimmi in reply to Morat20 says:

        Traveling salesman gets solved probatilistically all the time.

        It’s just not solvable the way the mathemeticians want to solve it.

        “It’s not like I’m the best in my field”
        Me: “You’re not?”
        “Well, it’s not unanimous…”Report

        • Morat20 in reply to Kimmi says:

          I’ve solved travelling salesman problems myself. It depends on what you mean by “solved”.

          In the math case, proof that your route is the shortest. In the sense I’ve solved it, I’ve found a route that was a heck of a lot shorter than any of the others. And I did it quickly.

          Getting a 95% efficient answer in 30 seconds versions a 99% efficient answer is 30 years is a good trade-off.

          But the Travelling Salesman problem itself, as an NP problem, can only be solved through the direct examination of all cases. Thus as the number of ‘cities’ grows, the length of time to solve the problem rises exponentially, not polynomially.

          Of course you might be able to solve some NP problems non-deterministically (ie: Not the way computers do) in polynomial time, but some you probably can’t. (Hence NP hard).

          The upshot of proving that NP problems are P (solveable in polynomial time) is “All encyrption but one-time pads become useless” as step one. 🙂Report

          • Jim Heffman in reply to Morat20 says:

            “It depends on what you mean by “solved”.”

            The point is not to try every possible solution and pick the best one.

            The point is to have a single equation where you plug in all the coordinates and the result is the best path.

            It’s like the difference between finding the roots of “Ax^2+Bx+C=0” by trying every possible value of x and finding the roots by using the quadratic equation.Report

            • Mike Schilling in reply to Jim Heffman says:

              Or a procedure that in general finds the best solution in a reasonable amount of time, even if that time isn’t fixed. Quicksort is a very good sorting algorithm, but its worst case is petty bad.Report

      • George Turner in reply to Morat20 says:

        Nah, you could do it in aggregate. The market doesn’t react to me individually. It doesn’t CARE that I might want this model of car to come in Color X as a base option. It only cares if a LOT of people agree with me.

        But that’s just a big averaging of the market. When you’re in the “market” for a car and go down to the car lot, wanting Color X is tremendously important to the price you and the salesman will reach. If each Color X is only 1% of the market, but 100% of the market is made up of people who each want a particular Color X[1..100], then color availability (having the right product selection in the right places) will have a profound effect on the overall market.Report

        • Morat20 in reply to George Turner says:

          True, but again — it’s about scale. Very, very, very few businesses truly care about the individual customer — and those are all custom shops — whether it’s custom motorcyles, bespoke tailoring, or whatnot.

          There’s no point — at least at the moment — approaching economic problems in an agent-based way (that is, simulating discrete individuals) unless you’re modelling very small problems based on small groups. (In which case, it’d be easier to run a study and not bother with the computers).

          Now, if 3D printing technology or nanotechnology moved us to an on-demand, highly individualized economy, maybe that might change. But as it is, “customers, in aggregate, prefer having many color choices to few” is a big, big difference between “I like the color green”.

          99.9% of the economy cares about the former, for there are enough of those to matter, and the remaining 00.1% I probably can’t afford — and they care about the latter (at least beyond making the immediate sale with what’s at hand).Report

          • Roger in reply to Morat20 says:

            No. They WISH they didn’t have to care about the customer. Unfortunately, for them, the customer won’t let them not care, unless of course they have a coercive monopoly.

            That said, you are of course right that it is uneconomical to allow customers to have every color. But the point here is again that customers care about things being too expensive, so businesses serve these tradeoffs as best they can. Mass marketers care as much about customers as custom shops. They are just appealing to different segments.

            Note of course that there is a monopoly which does not have to cater to customers, just to voters, but they can be bought as part of a package deal to protect them from Sarah Palin (and that’s certainly worth something).Report

          • Qub in reply to Morat20 says:


            I’m probably giving Kuznicki’s game away, but have you read Hayek’s “The Use of Knowledge in Society”?

            There are two distinct issues here. One is the impossibility of capturing enough data and analyzing it quickly enough to be able to determine in a real time basis what resources are demanded where. As Hayek noted, all the businessperson needs to know is that the price increased, she doesn’t need to know why, to react in the appropriate way. So there’s a real efficiency in the use of knowledge, there. No great extent of information and grand calculations are necessary.

            The other is that even if we could, through algorithms and such, do satisfactory calculations (c.f., Asimov’s robot brain), it wouldn’t do as well as the market because the market doesn’t just calculate value, it discovers value. To use a silly example, no central calculation device would be able to discover the value of pet rocks. Or more seriously, central calculation would never have moved us from VHS (if it could have even gotten us to VHS) to DVD to Blue Ray, or from vinyl disks (if it could even have gotten us there) to CDs to IPods. A central calculator for allocating resources would be a stagnant system, responding only to demand that exists at the moment, not exploring for new demand that’s not actually predictable.Report

            • Roger in reply to Qub says:

              Plus 1Report

              • Qub in reply to Roger says:

                A lot of Hayek’s critics misunderstood point 2 and focused only on point 1. When Lange provided a theoretical solution to 1, a lot of economists thought Hayek was licked. Even those who thought the solution merely theoretical, and not real-world implementable, thought socialism’s theoretical possibility had been proved. Very few ever really grasped point 2. That may be Hayek’s fault, though, for not really writing it clearly.Report

            • Kimmi in reply to Qub says:

              Nu, you could build in exploration. Pet rocks? Yes, that would be harder. But fads seem to be decently predictable (you can churn out 100 products, and use whichever sells best.)Report

  11. Fnord says:

    As it turned out, the biggest objection seems to have been that markets don’t do much that couldn’t be accomplished by sufficiently intelligent and fair-minded people allocating resources according to scientific calculation…Why, it probably shouldn’t even qualify as a different conceptual layer.[1]

    It was?

    That wasn’t my objection. I agree with you that there’s a layer beyond the Gregorian layer, and I’m at least sympathetic to the idea that markets are part of that layer, I just don’t think markets are the only part of the layer.

    Without speaking for them, that doesn’t match my interpretation of Burt’s or BrianM’s or Rod Engelsman’s objections, even as they did say that they’re skeptical of markets being a layer separate from the Gregorian layer. Nor does it match my reading of GordonHide’s objection. I think it’s not CK Macleod’s objection, who seems to objecting that markets are harmful, not superfluous. Shazbot said a lot of things, including a little bit along those lines, but it doesn’t seem to be the core of the objection; it’s not present in the top-level comment he made.Report

  12. Michael Cain says:

    Walras himself emphasized that markets only approximated his theoretical equilibrium through a process he called tâtonnement—literally, groping.

    I was pleased to see this. If only economists had paid greater attention to it. My biggest complaint about contemporary economics is the assumption that economic agents solve optimization problems. They don’t — but they do behave very much like a limited optimization algorithm. Limited history of how they got to where they are now; limited ability to predict the future; trying to figure out which way is “uphill” towards the minimum at successive points in time. Add in a bit of hysteresis for good measure. And the coefficients of the function to be optimized keep changing.

    I understand why economics went in the other direction: studying big dynamic systems wasn’t really possible until computers became common enough. Whereas small systems of differential equations could be solved, or at least shown to have a unique solution. A fellow student once asked me (as the math major in the graduate econ class), “Why is the prof going through all this stuff?” And the answer was already obvious: “So she ends up by the middle of the semester with ‘maximize a concave function on a compact convex set,’ which we know has a unique solution (ignoring the pathological cases).”Report

  13. Rogue Economist says:

    Market theory itself is a failure. The totally free, unregulated market is anarchy, and the results are plain to see. It creates wild swings in pricing. It creates reasons for hoarding. It creates the necessary upwards mobility of wealth into the hands of only a few, to the point where marketplace oligarchy takes over government or becomes pseudo-government.

    This is the history of “markets.” Runs on the banks, crashes, bubbles, and situations where only those in the oligarchy can hold onto money because they carry an asymmetric amount of access to information in both amount and timeliness.

    Hayekian economic theory is immoral pap.

    On the other hand of the argument is “pure socialism.” It’s what you’ll find practiced in religious communes and societies, monasteries and nunneries. Those who don’t pull their own weight in some way are kicked out, those who try to take more than their fair share are kicked out, and those who are malcontents are kicked out. It’s a great way to run a society as long as you have the safety valve to kick out those who aren’t wanted.

    100% Marxist theory doesn’t scale up to the size of nations well.

    The most efficient markets require regulation, significant and strict and sure. Insider trading is to be curtailed, price fixing agreements between businesses are to be curtailed because they are hoarding behavior, price gouging is to be curtailed whether by a monopolist or in time of emergency. Irresponsible behavior needs to be proactively addressed because the courts are an incomplete and incompetent solution; the situation in West, Texas illustrates the point well. Years of safety violations, years in which the management of the plant simply paid the fines and continued to violate, until finally lives were lost and there’s a smoking crater in the ground.

    Lives can’t be replaced, and the libertarian reliance on torts as the ultimate solution for everything is an illustration of the sort of blind, magical-thinking nonsense that makes libertarianism invalid.Report

  14. Francis says:

    I knew a little about water supply. Californians who don’t have their own well get their water from (i) a city, (ii) a limited purpose public agency, or (iii) a corporate entity subject to pretty strict oversight by the Public Utilities Commission. Those entities in turn either develop their own supplies locally or (far more common) purchase from larger regional water wholesalers and participate in regional projects led by the wholesalers. (The list of State Water Project Contractors is here. All of the Contractors are public agencies.)

    The basic premise underlying each retail agency’s actions is dead simple. They promise to deliver water that is (a) safe, (b) affordable and (c) reliable to everyone in their service agency for as much as each person wants. Demand in excess of supply is managed by a combination of developing new water supplies (including recycling) and reducing demand (thru such things as increased prices for higher uses, lawn replacement, public education, irrigation management [for ag users], etc.)

    Since only government is large enough to develop state-wide systems, and since retail delivery is the purest of natural monopolies, it’s not surprising that there is not a robust market in water supply. (Poseidon’s de-sal plant in Encino being a notable exception.)

    A market could, in theory, develop. But what the pro-marketers (that I’ve read) consistently refuse to understand is that the allocation of water is, first, a social choice. We can, by collective and governmental action, force people to get water from a private party. But doing so is no more natural than the current system.

    Markets can be a powerful tool for allocating goods. But certain political groups seem to worship markets at a level disproportionate to their successes. Markets should serve people, and when they don’t (water / health care) government should step in. Personally, I’m an American before I’m a capitalist. My right to vote on market failures is more important than my right to withhold my spending from a failed market.Report

  15. HankP says:

    I think the math is too hard, but not for the reasons you cite. The fact that x number of simultaneous equations can be solved in y time doesn’t help with the fact that developing the content of those equations is virtually impossible for anything other than basic undifferentiated commodities and staples.

    A market sets a clearing price in a relatively efficient way because it leaves the hard work – the actual computation (actually approximation) of those equations to each individual as they make a decision on what to buy and how much to spend. In effect it’s a giant analog computer and each market participant is a CPU (with it’s own set of data). To try and replicate that in a traditional computer is not difficult because of the amount of calculation that goes on but because of the dispersion and variety of the calculations that go on, and the fact that each calculator may value different criteria at any point in time. Unless you have a giant ongoing polling process you couldn’t do it, and even if you did you’d have to incorporate sentiments that individuals may not be able to put in mathematical form or even into words. That’s where the math is too hard. The only way to do it would be to introduce rather brutal simplifying assumptions, and as we can see from history that tends to make the results of the computation move farther from the real economy as the assumptions get more simplified.

    So I don’t think you’ve disposed of this problem. We’d need several significant breakthroughs in the quantitative aspects of psychology and sociology to even begin to attempt it.Report

  16. Rod Engelsman says:

    My first reaction upon reading this essay was to wonder what the hell the point was eventually going to be. Given that Socialism as defined when the actual Calculation debates were being had–what we would now call Communism–has basically died with only a couple of sad hold-outs (Cuba, N. Korea), the only real remaining targets would be those socialistic elements of contemporary mixed economies. And that would be… what? Healthcare, education, and market regulations? But I’ll allow Jason to develop this in further essays before criticizing more.

    The other thing that strikes me is that Calculation is a fractal problem that manifests at all levels and in any system. If you assume robust computational resources, the Socialist Calculation problem reduces to ignorance of individual needs and preferences as noted by HankP and also by James K. upthread. But a Hayekian market suffers from the opposite problem whereas the computational units (individuals) now have decent access to their own needs and preferences but a very limited access to, and limited interest in, the larger picture. That information is supposed to be transmitted through the network by the price system and therefore the healthy functioning of the system as a whole depends critically on prices accurately reflecting relative scarcities, undistorted by externalities and other market failures.

    So to my mind, the real debate we should be having is how to best use the powers of government to provide accurate pricing signals for those inputs prone to externalities and other market failures. The real calculation problem is GIGO regardless of the level of abstraction.Report

  17. Tim Ellis says:

    I’m sorry this thread got derailed (quite appropriately mind you) because this has been fascinating reading.

    I often discuss politics and economics from a socialist perspective, and I consider myself fairly well informed, but I feel quite happily outclassed by the posts and comments that this thread has generated.

    Consequently, I’m offering this simple comment of thanks instead of injecting my opinion.

    I have some questions, but I’ll save them til the end of the series as I expect most will be addressed.

    Til then – thanks for this.Report