Asymmetry, Entropy and Policy: Reply

• Published date: 01 December 1975
• Date: 01 December 1975
• DOI: 10.1177/0003603X7502000405
• Subject Matter: Article

ASYMMETRY. ENTROPY AND POLICY:

REPLY

by

J. L.

HEXTER·

and

JOHN

W.

SNOW··

Walter

Heller! recently commented

that

economists tend

to live contentiously with one another and

that

out of this

state our knowledge is increased.

vVe

agree and welcome

Professor Ramsay's comments pointing out certain limita-

tions in our paper. The purpose of this response is to pre-

sent

our

thinking more clearly and to suggest some limita-

tions to his comments.

Ramsay is correct

that

relative entropy

for

an entire in-

dustry

is not an

appropriate

measure of asymmetry

for

that

market when the number of firms is variable. However, it is

an

appropriate

measure of asymmetry

for

asubset

of

firms

where the number of firms in

that

subset is fixed. Our posi-

tion may be

better

understood if viewed from aslightly dif-

ferent perspective,

that

of conditional entropy.

Conditional entropy is equal to minus the weighted aver-

age of the logarithms of conditional

market

shares of the

dominant firms in an industry. Traditionally, industrial or-

ganization economists have measured market power in terms

of the share of the market held by the top four firms." The

conditional market share means the proportion of the share

of the top four held by the

ith

firm. Hence, conditional

entropy is

written:

•Professor of 'Finance

and

Economics,

Kent

State University,

Kent, Ohio.

••

Deputy Under Secretary of Transportation, Department of

Transportation, Washington, D.C.

1W. W. Heller, "What's Right with Economics," American Eco-

nomic Review (March 1975), pp. 1-26.

2There is nothing sacrosanct about

that

number,

but

the four

firm concentration ratio has become a widely accepted

standard

in

the industrial organization literature.

769