A response to Humberto's thoughtful post -- I don't mind the rambling 
at all, and ask your indulgence in some of my own.

You are quite right about the problems associated with having judges 
assign property rights on this basis.  The next item on my syllabus 
after Coase in my economics and law course is the Socialist 
Calculation Debate, which is specifically about the problem of 
identifying efficient outcomes from the center in the absence of 
functioning markets in the relevant property rights, and this is 
followed by Calabresi and Melamed, which is about how certain kinds 
of exchanges, specifically those between tortfeasors and their 
victims, are governed by liability rules because transaction costs 
are too high for this exchange to be organized in markets.  I believe 
that despite the enormous informational problems associated with 
efficient allocation from the center, and despite the power of 
Hayek's arguments about them, the law understood as an economic 
system can only be rationalized as an institutional means to allocate 
a certain kind of right (what C&M called an entitlement) in 
circumstances where markets cannot achieve this because transaction 
costs are high.

The interesting questions are thus not "What is the efficient 
allocation of legal entitlements?" but rather "In the face of the 
Hayekian division of knowledge, what kinds of institutional 
mechanisms does the law provide to encourage efficient exchange by 
individuals where this is feasible, and try to replicate these 
outcomes in environments where exchange in markets is impossible? 
How do these mechanisms work, and how do they change over time?" 
This is precisely the approach taken by C&M, who saw the common law 
of property and tort as evolved institutional mechanisms to encourage 
and govern the exchange of entitlements in markets (property rules) 
or the "centrally planned" judicial allocation of them in cases where 
markets cannot function (liability rules).  Their view is closely 
related to, indeed is one of the canonical sources of, the law and 
economics of "governance," an approach represented by the Coase of 
1937, Herbert Simon, Ian Macneil, Oliver Williamson and others that 
would, had it been widely adopted, have continued to engage legal 
scholars and thus pushed the economics of law in a very different 
(and far superior) intellectual direction from the one it has in fact 
taken.  For at the same moment (1972), Posner's first edition was 
published, and I believe that with this, law and economics began to 
fly off the rails.  Where Coase and Calabresi proposed an 
institutional, evolutionary, qualitatively and historically oriented 
economics of law, Posner, by fixating on the first of the questions 
above, opened the field to the neoclassical economics of optimizing 
and equilbrating and the econometric approach to social science, 
which having conquered the institutionalists (and everyone else) in 
the economics departments, quickly colonized the economics of law as 
well.  The result was a generation of increasingly mathematical and 
econometric economics of law, and a consequent estrangement of the 
entire field from the the lawyers, who (like attorneys Calabresi and 
Melamed) understand that only the institutional approach, or some 
variant of it, can hope to be of use in understanding so complex an 
evolutionary social system as "the law."

This institutional alternative would certainly agree with Humberto 
that "Transaction costs should be lowered," though perhaps it would 
express the proposition as "Necessarily imperfect institutions of 
governance (firms and judicial systems, among others) do indeed 
evolve as so as to enable or facilitate, by institutions other than 
the market, the exchange of entitlements where markets would fail 
because transaction costs are prohibitive."  Liability rules are not 
a way to reduce transaction costs in markets, because they govern 
situations in which those costs cannot be overcome and thus in which 
free exchange is not possible.  They are an institutional alternative 
to markets, one that requires a judge and jury to assess, however 
imperfectly, the costs of an involuntary entitlement transfer in the 
absence of market values and impose this cost back on the entitlement 
taker.  They This seems to me to be exactly where Coase 1960, who was 
after all the same fellow as Coase 1937, was pointing.

Rich Adelstein