Monopolistic competition

1
Monopolistic competition

• Session 3
• International trade and MNES
• M1

2
INTRODUCTION
3

• Monopolistic competition has been during two
  decades the work horse of international trade
  theory and empirics.
• Provides a framework for IIT-H
• Fits well the gravity equation
• Elegant way of introducing increasing returns
• Can easily be combined with comparative advantage
• Fits well in CGE models
• Convergence of IO and international trade
• Limits
• In the canonic version, the dynamics of the
  market structure is poor.
• Firms are homogenous

4
3.1/ Some basics

• Demand side
• Supply side

5
DEMAND SIDE

• How to represent the demand for variety?
• Two possibilities
• Representative consumer valorising varieties per
  se, varieties all symmetric
• Heterogeneous consumers, looking for the ideal
  variety. The latter is a bundle of
  characteristics. Every variety is a different
  consumption technique, linearly combining these
  characteristics.
• Corresponding respectively to
• The SDS-Krugman type models
• The Hotelling-Lancaster type models.
• SDSK is more tractable in particular for
  empirical purposes.

6

• Imposes to model preferences of the consumer with
  a two stage budgeting approach.
• In a first stage, choice between categories of
  products.
• In a second stage, choice among the varieties of
  a given product category, for categories with
  differentiated products.
• Dimensionality can be increased beyond 2
  industries in applied models. Principles remain
  the same.
• Often, nested CES, or CES nested in a Cobb
  Douglas. We will use below an even simpler
  version of the latter choice.

7

• Lets replace the utility function
  UU(c1,c2,cm), where cj are the quantities
  consumed of goods j (1,2, m)?
• ByUUU1(c11,c21,cn1), , Um(c1m,c2m,cnm)
• We now have n varieties (i) in each of the m
  industries ( j)?
• This decomposition of utility is subject two an
  assumption

8

• Then we have the following nesting

w
S0
C
C2
Cm
C1
...
C2m
Cnm
C1m
...
C11
C21
Cn1
...
9
SUPPLY SIDE

• Monopolistic competition
• 2 characteristics
• A- monopoly on each variety Fixed cost of
  production in each variety
• B- many firms competitive aspect
• Pricing strategy Marg. RevenueMarg. Cost (Cm)?
• Equilibrium pricing P(Mark-up).Cm
• Hypothesis of free entry in the long run
• null profits and P AVC
• Openness translates into additional economies of
  scale, reduced mark ups and real income gains

10
3.2 The synthesis between standard and new
theories of trade (Krugman, 1981)?
11
Motivation

• 1- much of world trade is between countries with
  similar factor endowments.
• 2- A large part of trade is intraindustry
  (two-way trade in similar products).
• 3- Finally, much of the expansion of trade in the
  postwar period has taken place without sizable
  reallocation of resources or income distribution
  effects.

12
3.2,1 Closed economy Demand

• 2 industries (1, 2) each producing a horizontally
  differentiated good
• We must nest the two levels of utility
• sub-utility attached to the consumption of the
  different varieties of a given good
• utility attached to the consumption of the two
  goods.
• ULog

13

• Other assumptions
• Within each industry, varieties are imperfect
  substitutes for consumers.
• On the supply side, Labour will be specific to
  each industry.
• Each country will have a potentially different
  (exogenous) repartition of the labour force
  within the two sectors.
• Li1 is the labour force engages (and stuck) in
  industry 1.
• Li2 respectively for industry 2.
• Hence wage will be different in the two
  industries in a given country.
• Production functions are the ones used
  previously, with now a subscript indicating the
  industry j.
• Industries differ only by the quantity of their
  specific factor fixed and marginal costs are
  identical.
• Lets normalise the labour stock to 2 and define
  a variable d indicating the distribution of
  labour among the two industries.

14
(No Transcript)
15
(No Transcript)
16

• Firms are of the same size in the two industries.
• Since the two industries account for the same
  share in overall demand, and in absence of pure
  profit, relative return to the factor is simply
  determined by its relative specialisation in the
  industries.
• Here, two variables are key the relative
  endowment in specific factor d and the
  substitutability alpha between varieties of the
  same good.
• For d1 there is no specialisation of the
  specific factor and the relative wage is equal to
  1. For low values of d, the economy is
  specialised in industry 1 and L2 will receive a
  relative high return.
• The weakest the alpha, the more differentiated
  are the products and the less important are the
  economies of scale.

17

• Openness to trade
• Zero transport cost
• Same technology in the two countries
• Symmetrical economies (mirror image
  assumption) the only difference between the two
  countries is the repartition of labour between
  the two industries.
• When the economic distance is maximal d 0
• When the economic distance is nil d 1
• d is an inverse proxy of the economic distance.
• In order to determine trade flows, one has to
  keep in mind that the consumer splits her income
  in two equal parts and then among the nj
  varieties.

18
Thanks to the symmetry, we conclude that
19

• We can easily check that trade is balanced.
• We also check that bilateral trade sums to Y see
  remark on gravity above.
• The latter result does not depend on d, which is
  on the contrary specific to this model here
  economic distance does not imapct the volume of
  trade.
• We observe that IIT (as measured by the GL ratio)
  is inversely related to economic distance.
• Actually, GLd.
• When the two economies are not specialised (d1,
  the labour force is equally distributed in the
  two activities in the two economies), trade is
  100 intra-industry.
• When the labour force is quite exclusively
  employed in industry 1 in the domestic economy (d
  0) and in industry 2 in Foreign, then trade is
  purely inter-industry and GL 0.

20
4.2.2 Stolper Samuelson effects

• We now tackle distributive effects.
• Since we are considering a framework where trade
  is of intra-industry nature, without additional
  economies of scale, gains accruing to the economy
  are associated with variety only.
• Two types of workers.
• Need to tackle the change in their real income.
• 2 different numeraires and 2 industries where one
  can be employed. This leads to 4cases. Lets
  define wkj, the real return, in numeraire j, of
  factor L employed in industry k.

21

• Each consumer will demand a quantity cij of each
  variety i of each good j
• Leading to a utility of
• We will now calculate the utility of workers
  employed in one industry, taking into account
  their real income.
• This is done first in autarky and then under free
  trade.
• One specific factor will gain in variety and lose
  in real return
• The other one will gain on both fronts.
• We will concentrate on the former, who needs to
  compensate a Stolper Samuelson loss by a Krugman
  type gain.

22

• The autarkic utility levels of workers employed
  in the two industries are
• For Home, the abundant factor is L1. Hence, we
  know that the specific factor L2 will be the
  victim of the Stolper-Samuelson effect. We
  concentrate on L2.

Stolper-Samuelson
Krugman
23

• The utility change for L2 is
• Note that the real wage of L2 in terms of
  numeraire 2 is not affected. Thus, we can
  simplify the former expression as

24

• The change in w2 is inversely proportional to the
  initial distribution of income (w2/w1).
• And we know that w1/w2d/(2-d).
• Thus the Stolper-Samuelson effect is
• If d1 the loss is log10. Hence when the
  economic distance between the two trade partners
  is nil, there is only intra-industry trade and
  the Stolper-Samuelson effect vanishes. Every
  factor will gain (in variety).

25
4.2.3 Stolper Samuelson effects and Krugman-type
compensation

• Lets now concentrate on the latter gain.
• Under free trade the two sets of variety (Home,
  Foreign) are available to the consumer.
• Hence, after simplification
• Summing Stolper-Samuelson and Krugman effects,
  and rearranging the terms, we get

26

• We now consider all cases where d lt1 in order to
  have a Stolper Samuelson effect.
• For the first term is always
  gt0.
• Regarding the third term, as long as d lt 1 we
  have
• In order to have the second term gt 0 we need to
  have
• Lastly, for the first term is
  always compensating the third one.
• Hence, is a sufficient condition
  to have the scarce factor not losing.

27

• The former result is easily interpreted the
  differentiation will be large if alpha is low.
• And the variety gain will be large enough to
  compensate for the Stolper Samuelson loss.
• Now, what about situations where
  ?
• There exists accordingly a threshold d0 such as
  the scarce factor always gains to trade. And
  this threshold is increasing in alpha.

28

• Interpretation in order to have the two factors
  gaining to trade, a lesser differentiation of the
  products will have to be compensated by a more
  limited economic distance.

d
d0
29
ECONOMIC DISTANCE
CONFLICT
MUTUAL INTEREST
0
0.5
1
DIFFERENTIATION
30
4.2.4 Standard theory and SDS-Krugman reconcilied

• We now conclude by reconciling Krugman type
  models and the standard theory of trade (of a HOS
  type), using the integrated equilibrium.
• As proposed by Helpman and Krugman (1985).
• Consider two industries, each producing a
  horizontally differentiated product, with two
  different production functions.
• Varieties of good 1 are produced with a more
  capital intensive technique.
• As long as the zero profit condition is assumed,
  the integrated equilibrium framework can be
  mobilised easily, and we can rely on the net
  factor content of trade.
• Economic distance will be defined here as the
  relative capital intensity of the two countries.
• Identical countries will trade on a purely IIT
  basis, why different countries will combine inter
  and intra-industry trade.

31
L
0
K
D
C
K
L
0
32

• Home will export and import varieties of both
  goods, but will be a net exporter of varieties of
  good 1 and a net importer of varieties of good 2.
• The net factor service content of trade is the
  one predicted by HOV.
• IIT has no impact in Stolper-Samuelson terms.
• The largest the economic distance
• The lowest the GL index.

X1
M1
X2
M2