Julian Emami Namini

1
International trade with horizontal and vertical
product differentiation and heterogeneous firms
Julian Emami Namini
Ricardo A. López
Erasmus University Rotterdam
Indiana University, Bloomington
IU Microeconomics Workshop, 09 January 2008
1 of 20
2
1 Introduction
1.1 Theoretical literature on firms export
behavior

• 1. Melitz (2003), Econometrica
• exporters more productive than nonexporters

• symmetric countries exporters export to each
  country

• implicitly

asymmetric countries
only more productive firms export to smaller
market
2. Eaton/Kortum/Kramarz (2005), Working Paper
hierarchy of markets
more productive firms export to more markets
2 of 20
3
1 Introduction
1.1 Theoretical literature on firms export
behavior ctd.

• 3. Bekkers (2007), Working Paper
• exporting firms higher quality higher price
• identical quality for each destination market

• implicitly

asymmetric countries
only higher quality firms export to smaller market

• 4. Raff/Stähler/VanLong (2007), Working Paper
• R Ddecision by firms
• productivity gains with exposure to trade
• more R D by exporting firms
• implicitly

asymmetric countries
only higher productivity firms export to smaller
market
3 of 20
4
1 Introduction
1.2 Empirical literature on firms export behavior
This paper
/ Lawless (2007), Working Paper
1. on average productivity exporters gt
productivity nonexporters
Melitz (2003)
data source Annual National Industrial Survey,
National Institute of Statistics, Chile
19901999
4 of 20
5
1 Introduction
1.2 Empirical literature on firms export
behavior ctd.
This paper
/ Lawless (2007), Working Paper ctd.
2. however many nonexporters more productive
than exporters
? Melitz (2003)
data source Annual National Industrial Survey,
National Institute of Statistics, Chile
19901999
5 of 20
6
1 Introduction
1.2 Empirical literature on firms export
behavior ctd.
This paper
/ Lawless (2007), Working Paper ctd.
3. of export destinations many firms export
to limited number of countries
? Melitz (2003)
data source Annual National Industrial Survey,
National Institute of Statistics, Chile
19901999
6 of 20
7
1 Introduction
1.2 Empirical literature on firms export
behavior ctd.
This paper
/ Lawless (2007), Working Paper ctd.
4. market 1 market 2 market share firm 1 gt
(lt) market share firm 2
? Melitz (2003)
? Melitz (2003)
5. less productive firms may export to smaller
market
1.3 Theoretical contribution of this paper
Theoretical model to explain additional empirical
evidence on export behavior
7 of 20
8
2 This model preliminaries
2.1 Households
CES utility function over N varieties of
differentiated good
s 2 for simplicity
quality level of firm f
firm index
2.2 Countries

• only labor, numéraire good

• goods? Partial equilibrium setup
• analyzed sector IRS fixed costs

• countries differ in size

8 of 20
9
2 This model preliminaries
2.3 Firms

• exante uncertainty about MC

1. market entry sunk costs technology unknown
2. draw of technology parameters

• serving domestic/foreign market fixed costs

• decision for each market high / low tech

• high (low) tech ? high (low) fixed costs

• DixitStiglitz monopolistic competition between
  firms

9 of 20
10
2 This model preliminaries
2.3 Firms ctd.

• per unit costs

choice variable some influence on
technologies high / low tech aH lt aL
k MC for zero quality output
random variables
choice variable quality level
c MC for each unit quality

• variable profits

• profit maximizing quality level

10 of 20
11
2 This model preliminaries
2.3 Firms ctd.

• profit maximizing quality level

profit maximizing price level
demand
random variable
cf?
kf?

• identical pf
• quality ?
• market share ?

• pf ?
• quality ?
• market share ?

random variable
11 of 20
12
2 This model preliminaries
2.3 Firms ctd.

• aH / aL? High / low tech?

Assumption

• firm chooses high tech if

high tech profits gt low tech profits
low tech
high tech
technology separation line country specific
12 of 20
13
2 This model preliminaries
2.3 Firms ctd.
isorevenue curves?
high tech
low tech
isorevenue curve low tech
isorevenue curve high tech
13 of 20
14
2 This model preliminaries
2.4 Course of events
market entry sunk costs fE
draw of random variables c k
decision production technology
decision market exit
production
random shock market exit
time
14 of 20
15
2 This model preliminaries
2.4 Success of market entry
random variable
exit
zero profit condition
low tech
production

• variable profits ?

variable profits ?
high tech
random variable

• production after entry only if

variable profits
fixed costs

• zero profit condition

15 of 20
16
3 Open economy equilibrium
3.1 Productivity and export behavior (1)
random variable
Ø nonexporting firm
zero profit condition domestic market
zero profit condition foreign market
Ø exporting firm
technology separation line both markets
random variable
Result 1
per unit costs Ø exporting firm
lt
per unit costs Ø non exporting firm
16 of 20
17
3 Open economy equilibrium
3.1 Productivity and export behavior (2)
random variable
firm 1
zero profit condition large foreign market
firm 2
zero profit condition small foreign market
technology separation line large foreign market
technology separation line small foreign market
random variable
p ?
, but only firm 2
exports to small foreign market.
Result 1
firm 1 has lower per unit costs (? higher
productivity)
17 of 20
18
3 Open economy equilibrium
3.2 Market share and export behavior (1)
isorevenue curve large foreign market
random variable
zero profit condition large foreign market
isorevenue curve small/large foreign market
zero profit condition small foreign market
firm 1
firm 2
technology separation line large foreign market
technology separation line small foreign market
random variable
Result 2 if firms have identical market share
in large country , they must have
identical export behavior w.r.t. small foreign
country!
if firms have identical market share in large
foreign country
Result 2
18 of 20
19
3 Open economy equilibrium
3.2 Market share and export behavior (2)
random variable
random variable
large foreign country
small foreign country
isorevenue curve high tech
isorevenue curve low tech
technology separation line
firm 1
firm 1
firm 2
firm 2
technology separation line
random variable
random variable
Result 3 large forgein country market share
firm 2 gt market share firm 1 small foreign
country market share firm 2 gt market share firm 1
Result 3 large forgein country market share
firm 1 gt market share firm 2
Result 3
19 of 20
20
4 Conclusions

• actual export behavior of firms more complex
  than predicted by Meltiz (2003) and others

• less productive firms may export to smaller
  market

• ? of export destinations not related to
  productivity

• ranking of firms w.r.t. market shares differs
  between countries

• theoretical setup

• so far theoretical results in line
  with new empirical evidence on firms export
  behavior

20 of 20