Joe Bain Is 'Junk Science'?

[charles mueller <cmueller@metrolink.net>]

        One of our members recently proposed to demonstrate for us that the
economic work of the late Professor Joe S. Bain (Harvard/Berkeley) was and
is 'junk science' and hence should not be allowed into an American courtroom
in a monopoly case.  

        Bain authored a large body of work on industrial organization
(antitrust) economics, including his BARRIERS TO NEW COMPETITION (Harvard
Univ. Press, 1956) and INDUSTRIAL ORGANIZATION (Wiley 1959).  An empiricist
(like Adam Smith), his research gave considerable emphasis to 2 central
issues in antimonopoly policy, namely, (1) what degree of structural
concentration marks the onset of monopoly power in American industrial
markets, and (2) how much of that concentration/monopoly is essential to the
achievement of 'efficiencies' (scale economies)?  

        Bain's empirical findings here, as I recall, were generally that (1)
top 8-firm shares of roughly 70% mark the real-world threshold between
effective competition and effective monopoly/oligopoly, e.g.,
higher-than-competitive profits/prices, and (2) only a couple of American
industries (e.g., typewriters) require a single-firm share of even 10% to
exhaust all economies of scale, i.e., to produce at the lowest unit cost
permitted by the existing technology.  (In autos, for example, the figure is
reportedly 7.5%, which would of course permit 13 U.S. car manufacturers--all
of maximum efficiency.)  

        His 8-firm-70% threshold of monopoly power translates approximately
into a top 4-firm share of about 50% (and a top single-firm share of roughly
12%).  

        The U.S. courts have certainly been persuaded that Bain was
wrong--have accepted the argument that consumer prices, for example, DON'T
start rising above the competitive level until an industry has been stripped
of all but, at most, 2 firms--one of which has a single 'dominant' share of
at least 60%, more often of 70% or greater.

        Judicial acceptance, though, does not, of itself, good science make.
Joe Bain launched the critical empirical research in the field of
industrial-organization (antitrust) economics and many a lance has been
broken since in a (generally fruitless) effort to unhorse his major
findings.  If our ambitious young member knows how to do it, it seems to me
that we should encourage him, not belittle his efforts.  I have long been
impressed with that vigorous equine principle which holds that, 'If you have
a a horse who thinks he's Man-O-War, do you really want to talk him out of it?'

        The note below is a post of mine on this issue on another list.

        Charles Mueller, Editor
        ANTITRUST LAW & ECONOMICS REVIEW
        http://webpages.metrolink.net/~cmueller
   
                                                    ******

        In regard to the discussion of economics as science, there's an
interesting example of how it performs (or doesn't perform) the function of
dispensing scientific findings for society's benefit.

        One of the branches of economics--industrial organization (which has
some 2,000 AEA members)--is of course the repository of society's knowledge
of which types of industry structure and behavior turn in the best economic
performance for society.  Is monopoly/oligopoly  more efficient that
competition?  More innovative?  If competition yields a better performance,
what do we mean by "competition," i.e., how many firms do you have to have
to make a market "competitive"?  

        Well, economic "science" was clear on this back in the '60s and
'70s:  It reported (through its industrial-organization branch) that there
was a systematic positive relationship between an industry's or market's
structure (concentration, etc) and its price/profit level, with the
threshold of noncompetitive prices being a top-4-firm share of around 40%
(equivalent to a l-firm share of about l2%).  This meant, of course, that
mergers, for example, should be stopped if the acquiring firm would, via the
merger, go over that l2% market share threshold.  (If one firm can merge its
way to a 12% share, all others have the same right, so you get a 4-firm
share of 48%--and prices that exceed the competitive level.)

        The policymakers in Washington and the courts accepted this research
finding of economic science and set national policy accordingly:  Mergers
yielding shares above l2% were stopped.  Firms wanting to grow larger had to
do it via internal growth.

        Then, in the mid-'70s, the neoclassical school of economics
(Chicago) began to gain increasing influence in the industrial-organization
branch and, under President Reagan, its members were put in charge of
antitrust/merger policy in Washington.

        Suddenly, economic "science" reported that the previous numbers were
all wrong.  Oligopolies do not, as previously reported, charge
higher-than-competitive prices.  Only "dominant" individual firms do
that--firms with individual shares of at least 60% and more often 70% or
higher.  The policymakers in Washington--and the courts--dutifully adopted
the new scientific findings of economics and the merger rules, for example,
were accordingly revised:  Only a merger yielding a single-firm share of 60%
or more would be challenged.

        Today, that's still the policy standard and the mergers currently
being approved by the government in such key industries as health-care and
telecommunications, for example, are being approved on the basis of the
economic view that it takes such a 60+ market share to impose
supracompetitive prices (and thus injure the consuming public). 

        Now, how did economic "science" get from a 12% to a 60% threshold
for identifying monopoly power?  Was this radical change of the profession's
collective mind the result of great progress on the empirical front?
Massive, incontrovertible new empirical studies demonstrating that it only
takes 2 firms (instead of 10 or so) to create all the "competition" it takes
to assure consumers "competitive" prices?  Or could it have been something
so crass as "politics" that rearranged the "scientific" landscape in
economics, e.g., the "conservative" resurgence in Washington?    Heaven forfend!

        If economics is "science," does it have a responsibility to give the
policymakers good numbers--the kind that don't change with the election returns?

        Charles Mueller, Editor
        Antitrust Law & Economics Review
        http://www.metrolink.net~cmueller/