Ceris-Cnr, W.P. N° 14/2005

very different, so should be carefully examined.
If we consider the representative consumer
approach, à la Dixit and Stiglitz (1977), the
results show that assuming free entry, constant
marginal costs with a fixed cost of production,
and a CES utility function (which implies that
products are treated symmetrically), price and
output are independent of market size. Thus, as
market size increases, there are now more
profitable opportunities and new entry will
occur until profits are driven to zero. However,
the number
of
firms
will
increase
proportionately with market size, as the
elasticity of substitution is constant.6
In locational models, à la Hotelling, on the
other hand, competition is localized. Shaked and
Sutton (1987) show that as market size
increases, and assuming the distribution of
consumer tastes is constant, market shares can
always be made arbitrarily small by entering
between existing products and capturing a
positive market share. However, if the number
of firms increases in a particular segment, the
existing firms now find that there is more
competition (the elasticity of substitution has
increased). This implies that as market size
increases, the increase in firm numbers is less
than proportionate. However, in both
approaches
to
understanding
horizontal
differentiation, the inverse structure-size
relationship is a general result.
Thus, in Type 1 industries, we can construct a
lower bound to concentration that tells us the
lowest level of concentration we would expect
to observe for a given market size.7 Depending
on the strategic choices that firms make within
specific industries as well as external factors
6

7

As there are no perceived cross-effects by assumption,
strategic interactions such as product positioning cannot
be analyzed because all products are generalized
substitutes. Intuitively, it is as if firms are assigned a
product randomly on entry; there is no notion of
‘neighbourhood’ relative to the other products in the
market.
In an industry characterized by horizontal differentiation,
a range of equilibrium outcomes becomes possible. The
same group of firms might enter all submarkets leading
to high concentration with a few multi-product firms or,
on the other hand, small niche firms may enter each
submarket, yielding a more fragmented structure.

such as government regulation, many industries
will lie above the lower bound.8 Overall, the
Sutton approach, although general and robust,
cannot tell us much about what we expect to
observe above the lower bound to concentration,
and this is where the resource-based view of the
firm plays a very important role.

1.2 Market Structure and Market Size: Type 2
Industries
In an industry characterized by vertical product
differentiation, every consumer has the same
ranking of quality. Thus, if all products were
sold at the same price, all consumers would
choose the highest quality product (Mussa and
Rosen, 1978; Gabszewicz and Thisse, 1979;
Shaked and Sutton, 1982, 1983, 1987). Products
can be sold at different prices either because
consumers have differing levels of income or
relatively less intense preferences for quality
(i.e., sensitivity to quality).9
In Type 2 industries, firms not only compete
in price but also in advertising and/or R&D to
increase the (perceived) quality of the product.
Intuitively, as market size increases, the
incentive to gain market share through the
escalation of advertising and/or R&D
expenditure also increases. The consequent rise
in overhead costs has a countervailing effect on
market structure by increasing the degree of
economies of scale. The basic notion is that
although there appears to be room for more
firms in a larger market, 'the escalation
mechanism' raises fixed costs per firm, possibly
even to such an extent that the negative
structure-size relation breaks down (Sutton,
1991) as shown in Figure 1.
8

9

Also consider whether the market is in disequilibrium. In
the short run, it may take quite a long time before firms
merge or exit, particularly if economies of scale are large
relative to market size, as exit will tend to be slower, the
more sunk (industry specific) is the plant.
Such an industry may be labeled a ‘natural oligopoly’.
This differs crucially from horizontal differentiation
where if all products were sold at the same price, every
product would have a positive market share, i.e.
consumers vary in their rankings of product
characteristics.

7