Once subdivision of the cake is allowed, political favoritism becomes the most general solution, and it can be shown that the market solution is a subset of the political favoritism solution. It is not as efficient as the market solution because individuals can no longer be assumed to have enough power to determine their own needs for cake relative to other things, because constant rights to exchange can no longer be assumed - those rights are now dependent upon political contest, and a political (i.e. group) authority can shape individuals' preferences for cake, which means that the resulting allocation is necessarily less efficient than a market allocation where people have unchecked power to determine their needs relative to what they have.
Markets are social constructs. Individuals can never be assumed to have enough power to determine their own needs and preferences independent of group authority and political context (nevermind right to exchange...). Someone always has "market power," because markets are a social construct.
Economic models ignore this very basic fact about markets at the peril of talking nonsense.
- Jake If you only spend 20 minutes of the rest of your life on economics, go spend them here.
That's where the work of Stiglitz (hardly a pro-market economist) is important. While showing that market-based policy frameworks are probably greatly flawed, even in efficient allocations to say nothing of equitable allocations, because of the information asymmetry problem he also shows that none of the other policy frameworks for allocating resources thus far observed in any society will be able to be more efficient than a market-based solution in general. (There are very specific exceptions, such as health care, for example.) Why? Because instead of better solving the information asymmetry problem they tend to exacerbate it by removing more sources of truth from the system rather than adding more sources of truth to it.
Take, for instance, the scenarios above in martingale's cake distribution game: in scenarios b,c,and d, the information asymmetry problem is worsened, not improved, so the solution MUST be less efficient, except by dumb luck. In b, the political solution without providing for a market to determine who really needs the cake will only determine what the political authority determines the need is, through some non-exchange contest of some kind. In c, information is also restricted by assuming that everyone needs one piece of cake. Likewise in d, where information is restricted entirely to chance. Only in a market solution that provides for aportioning the cake can efficiency be better ensured than any other system.
Markets fail to exist when trust is sufficiently impaired, as the recent financial market problems have proven. So the fact that markets do exist in various commodities is evidence that a degree of reflexivity and honesty has already become well developed.
Among the market participants. Many people who are part of an industrial monetary production economy are not a part of any market, under this definition. On the labour "market," for instance, individual workers cannot be considered "market actors" - their relationship to their place of employment is more akin to that of feudal serfs to their landlord than it is to a buyer or seller at an auction. Which is to say nothing of the people who are touched by the externalities of market actions.
None of these people need to trust the market actors in order for markets to form, so market actors do not need to show them honesty and reflexivity.
Markets are not at all perfect or optimal because there are always some stakeholders excluded. However, markets are still able to include MORE stakeholders and provide more truthful information about the stakes than other means of determining who gets what in society, which means they're more efficient.
Regarding labor markets, I think you're being too harsh on life of serfs. There's actually little basis for saying that life is any better today for commoners than it was then for commoners in feudal periods, except in absolute material goods (which, for many, is everything of course). As the eminent economic historians Fernand Braudel and Karl Polanyi have written about at length, serfs may have been more or less tied, on the one hand, to the land they were born to (and with immigration restrictions today, has that really changed?), but on the other hand, state-level political authorities had almost no role, least of all economic, in their lives which revolved almost entirely around communal and family-level production and distribution. State power intervened in trade, not in production, most of which was agricultural.
Quite true about externalities and market failures. However, that only addresses what's wrong with markets, not what's right about any other system. Externalities are also a problem in ANY system of figuring out who gets what in society, and, because of the information asymmetry problem, a system that restricts sources of true information can almost never be better at accounting for externalities than a market-based system.
Bluntly put, Joe Q. Median has a vote in a parliamentary system. In a great many parts of the political economy, he has no vote under a market system, because Joe Q. Median is not a market actor. So for the majority of the population accountable democratic government is superior to market economics simply on the count that the latter disenfranchises them while the former does not.
A lot of successful regulation is about cutting the big market players down to a size where Joe Q. Median can become a genuine market player. But this requires continual political vigilance, because markets are better at forcing market actors to cease being market actors than they are at making room for new market actors.
And then of course there are the sectors where there can be no "market" because the logistics of the production won't permit it. That could be due to the returns on economics of scale being sufficiently high that any gains from market forces would be offset by the cost of maintaining sufficient numbers of separate systems to have a meaningful market. Or it could be due to the risk of cascading failures making the independence of market actors impossible. Or it could be due to the ability of large market actors to take the political system that maintains the market hostage. Or some combination.
