Theory of Economic Integration
Dynamic effects
De-fragmentation and industrial restructuring
Gravity model
Katarzyna Śledziewska
Major dynamic effects:
1. Reaping benefits of economies of scale and
learning effects
2. Reducing the monopoly power
3. Reducing levels of x-inefficiency
4. De-fragmentation and industrial restructuring
Gravity model
Liberalization, defragmentation and industrial
restructuring
• Europe’s national markets – separated by a whole hst of
barriers
• Tariff and quotas – until 1968
• Technical, physical and fiscal bariers – until 1992
– When barriers – firms can be dominant in their home market –
market fragmentation
• Reduces competition
• Raises prices
• Keeps too many firms in business
• Tearing down intra-EU barriers
– defragments the markets
– produces extra competition
• The pro-competitive effect squeezes the least effecient
firms – industrial restructuring,
• Europe’s weaker firms merge or get bought up
Economic Logic Verbally
•
•
•
•
•
liberalisation →
de-fragmentation →
pro-competitive effect →
industrial restructuring (M&A, etc.)
RESULT: fewer, bigger, more efficient firms facing
more effective competition from each other.
Theory
•
•
•
•
•
•
•
•
Economic logic: background
BE-COMP diagram
Details of COMP curve
Details of BE curve
Equilibrium in BE-COMP diagram
No-trade-to-free-trade integration
State Aids
Collusion
• We start with the simplest form of imperfect competition:
monopoly, duopoly, oligopoly
Economic logic: background
•Monopoly case
Demand
Curve
Price
P’
P”
C
Marginal Revenue
Curve
Marginal
P*
Cost Curve
A
B
Price
Demand
Curve
D
Marginal
Cost
E
Q’ Q’+1
Sales
Q*
Sales
Economic logic: background
• Monopoly
–
–
–
–
–
Easy case (instructive)
Avoids strategic interactions
The only restraint – the demand curve
Consumers – price takers
The trade off between prices and sales depend only on
the demand curve
Economic logic: background
•Duopoly case, example of non-equilibrium
price
price
Firm 1’s expectation of
sales by firm 2, Q2
Firm 2’s expectation of
sales by firm 1, Q1
Demand
Curve (D)
p2’
Residual Demand
Curve firm 1 (RD1)
p1’
A1
MC
x1’
Firm 1 sales
Residual Marginal Revenue
Curve firm 1 (RMR1)
Demand
Curve (D)
Residual Demand
Curve firm 2 (RD2)
A2
x2’
Residual Marginal Revenue
Curve firm 2 (RMR2)
MC
Firm 2 sales
Economic logic: background
• Duopoly
– Most European firms faced competition as firms have
the same marginal cost curves
– No equilibrium – the outcome not consistent with
expectations
– The easiest way – assumption – symmetry of firms,
each firm sale the same amount
Economic logic: background
•Duopoly & oligopoly case, equilibrium outcome
price
Typical firm’s expectation
of the other firm’s sales
p*
price
Typical firm’s expectation
of other the other firms’
sales
D
D
p**
RD
RD’
A
MC
A
RMR
x*
Duopoly
MC
RMR’
2x* sales
x**
Oligopoly
sales
3x**
Economic logic: background
• Duopoly & oligopoly
– More firms competing in the market
– The residual demand curve facing each one shifts
inwards
– Number of firms continues to rise
• Lower prices and lower output per firm
BE-COMP diagram
Mark-up (µ)
µmono
µduo
BE (break-even) curve
µ’
COMP
curve
n=1 n=2
n’
Number
of firms
BE-COMP diagram
• The impact `of European integration on firm size
and efficiency, number of firms, prices
• Price – cost gap
– „mark-up” of price over marginal cost curve
Details of COMP curve
Mark-up
price
µmono
p'
A’
µduo
p"
B’
D
Monopoly
mark-up
Duopoly
mark-up
MC
COMP
curve
R-D (duopoly)
B
Marginal cost
curve
A
R-MR
xduo
MR (monopoly)
xmono
Typical firm’s
sales
n=1
n=2
Numbe
r of
firms
Details of BE curve
euros
price
po=µo+MC
AC>po
Mark-up
(i.e., p-MC)
Home market
BE
Demand curve
A
ACo=po
µo
po
AC<po
B
A
B
AC
MC
Sales
per firm
x’= Co/n’
x”= Co/n”
xo= Co/no
n” no
Co
Total
sales
n’
Number
of firms
Details of BE curve
• The positive link between mark-up and the breakeven number of firms
• A – firms are not covering their fixed cost, there
would be the tendency for some firms to exit the
industry (mergers and bankruptcies)
• B – firms are making pure profits, more firms to
enter the market
Equilibrium in BE-COMP diagram
euros
Price
Mark-up
Home market
Demand curve
E’
p’
p’
BE
E’
µ'
E’
AC
COMP
MC
n’
x’
Sales
per firm
C’
Total
sales
Number
of firms
Equilibrium in BE-COMP diagram
• The COMP curve – firms would charge a mark-up of µ’
when there are n’ firms in the market
• The BE curve – n’ firms could break even when the markup is µ’
• Let us determine the equilibrium number of firms, markup, price, total consumption and firm size (all in one
diagram)
No-trade-to-free-trade integration
euros
price
Mark-up
Home market only
Demand curve
BE
BEFT
p’
p”
E’
p’
E”
p”
E’
µ'
E’
E”
E”
A
pA
1
µA
A
AC
COMP
MC
x’ x”
Sales
per firm
n’
C’ C”
Total
sales
n”
2n’
Number
of firms
No-trade-to-free-trade integration
euros
price
Mark-up
Home market only
Demand curve
BE
BEFT
p’
p”
E’
E’
p’
E”
p”
C
µ'
E’
E”
E”
A
pA
1
µA
A
AC
COMP
MC
x’ x”
Sales
per firm
n’
C’ C”
Total
sales
n”
2n’
Number
of firms
No-trade-to-free-trade integration
• Reduction of trade barriers
• Assumptions:
– H & P identical
– We focus on H’s market
• The immediate impact:
– Second market of the same size
– Double the number of competitors
– Lower µ
• More firms, BE curve shifts out (to point 1)
– At any given mark-up more firms can break even
No-trade-to-free-trade integration
• Pro-competitive effect:
– Equilibrium moves from E’ to A: Firms losing money (below BE)
– Pro-competitive effect = markup falls
– short-run price impact p’ to pA
• Industrial Restructuring
–
–
–
–
–
–
A to E”
number of firms, 2n’ to n”.
firms enlarge market shares and output,
More efficient firms, AC falls from p’ to p”,
mark-up rises,
profitability is restored
• Result:
– bigger, fewer, more efficient firms facing more effective competition
• Welfare: gain is “C”
Empirical evidence
• Little direct evidence in Europe
• More direct evidence linking market size with efficiency
and competition
– Campbell Jeffrey R., Hugo A. Hopenhayn. 2002. „Market Size
Matters”. NBER Working Paper No. 9113
• The impact of market size on the size of distribution of firms in retailtrade industries across 225 US cities
• In every industry examined – establishment larger in larger cities
• Competition is tougher in larger markets and this accounts for the
link between firm-size and market-size
State aid (subsidies)
•
2 immediate questions
–
–
•
“As the number of firms falls, isn’t there a tendency for the
remaining firms to collude in order to keep prices high?”
“Since industrial restructuring can be politically painful, isn’t there
a danger that governments will try to keep money-losing firms in
business via subsidies and other policies?”
The answer to both questions is “Yes”.
State aid (subsidies)
•
1.
2.
3.
Profit losing firms to leave the industry:
Can be bought out
Merge with other firms
Go bankrupt
–
–
Job losses
Reorganization – workers change job or locations
•
•
Painful
Governments seek to prevent them (firms government owned,
trade unions)
Economics: restructuring prevention
Mark-up
BE
BE
µ'
E’
FT
1
E”
µA
A
COMP
n’
n”
2n’
Number
of firms
Economics: restructuring prevention
•
•
Consider subsidies that prevent
restructuring (in H&P)
Specifically, each governments make
annual payments to all firms exactly
equal to their losses
–
–
•
i.e. all 2n’ firms in Figure from slide 28
analysis break even, but not new firms
Economy stays at point A
This changes who pays for the
inefficiently small firms from
consumers to taxpayers.
Mark-up
BE
BEFT
µ'
E’
1
E”
µA
A
COMP
n’
n”
2n’
Number
of firms
Economics: restructuring prevention
• The too-many-too-small firms problem
• Firms continue to be inefficient
• The subsidies prevent the overall improvement in
industry efficiency
• Do nations gain?
restructuring prevention: size of subsidy
euros
Price
Mark-up
COMP
Demand curve
AC
A
a
b
A
A
pA
c
MC
Sales
x’
xA= 2CA/2n’per firm
FT
E’
E’
p’
pA
BE
2n’
C’ CA
Total
sales
Numbe
r of
firms
restructuring prevention: size of subsidy
•
•
•
Pre-integration: fixed costs = operating profit = area “a+b”
Post-integration: operating profit = b+c
ERGO: Breakeven subsidy = a-c
–
NB: b+c+a-c=a+b
euros
Price
Mark-up
COMP
Demand curve
pA
A
a
b
A
A
pA
c
MC
Sales
x’
xA= 2CA/2n’per firm
FT
E’
E’
p’
AC
BE
2n’
C’
CA
Total
sales
Numbe
r of
firms
restructuring prevention: welfare impact
•
•
•
•
Change producer surplus = zero (profit is zero pre & post)
Change consumer surplus = a+d
Subsidy cost = a-c
Total impact = d+c
euros
Price
Mark-up
COMP
Demand curve
pA
A
a
pA
b
d
A
A
c
MC
Sales
x’
xA= 2CA/2n’per firm
FT
E’
E’
p’
AC
BE
2n’
C’
CA
Total
sales
Numbe
r of
firms
Only some subsidise: unfair competition
•
•
•
•
•
If Foreign pays ‘break even’ subsidies to its firms
All restructuring forced on Home
2n’ moves to n”, but all the exit is by Home firms
Unfair
Undermines political support for liberalisation
EU policies on ‘State Aids’
•
•
1957 Treaty of Rome bans state aid that provides firms with an
unfair advantage and thus distorts competition.
EU founders considered this so important that they empowered
the Commission with enforcement.
Anti-competitive behaviour
•
Collusion is a real concern in Europe
–
•
Collusion in the BE-COMP diagram
–
–
•
dangers of collusion rise as the number of firms falls
COMP curve is for ‘normal’, non-collusive competition
Firms do not coordinate prices or sales
Other extreme is ‘perfect collusion’
–
–
–
Firms coordinate prices and sales perfectly
Max profit from market is monopoly price & sales
Perfect collusion is where firms charge monopoly price and split the sales
among themselves
Economic effects
Mark-up
BEFT
Perfect
A
collusion
µmono
B
pB
p”
E”
Partial
collusion
COMP
n=1 n” nB
2n’
Number of
firms
Economic effects
•
collusion will not in the end raise
firm’s profits to above-normal
levels.
–
–
•
Mark-up
2n’ is too high for all firms to break µmono
even.
Industrial consolidation proceeds as
usual, but only to nB. Point B Zero
pB
profits earned by all.
p”
BEFT
Perfect
A
collusion
B
E”
Partial
collusion
prices higher, pB> p”, smaller
firms, higher average cost
COMP
n=1 n” nB
2n’
Number of
firms
Economic effects
•
The welfare cost of collusion (versus no collusion)
–
four-sided area marked by pB, p”, E” and B.
price
Mark-up
Demand curve
p
mono
pB
FT
BE
Perfect
A
collusion
µmono
B
B
E”
E”
p”
Partial
collusion
COMP
n=1 n” nB
CB
Total
sales
Number of
firms
EU Competition Policy
•
•
To prevent anti-competitive behavior, EU policy focuses on two
main axes:
Antitrust and cartels. The Commission tries:
–
–
•
to eliminate behaviours that restrict competition (e.g. price-fixing
arrangements and cartels)
to eliminate abusive behaviour by firms that have a dominant position
Merger control. The Commission seeks:
–
to block mergers that would create firms that would dominate the market.
Other dynamic effects
• The polarization effect
– Benefits of trade creation becoming concentrated in
one region
– An area may develop a tendency to attract factors of
production
• The influence on the location and volume of real
investment
Remarks
• Dynamic effects include various and completely
different phenomena
• Apart from economies of scale, the possible gains
are extremely long term
Major dynamic effects:
1. Reaping benefits of economies of scale and
learning effects
2. Reducing the monopoly power
3. Reducing levels of x-inefficiency
4. De-fragmentation and industrial restructuring
Gravity model
Gravity equation
• often used as an instrument to measure different
aspects of trade effects.
• In the standard gravity model we assume
– economic power of trading partners
• can be measured as GDP
– trade costs
• can be measured as distance between them
• the key variables to explain the volume of trade.
The theoretical application
• Helpman (1987)
– Helpman’s theorem proclaims that the volume of trade
relative to GDP will be proportional to the relative size
of countries.
• can explain the expectations:
– bigger and more similar in terms of size countries tend
to trade more intensely with each other than the
smaller and different ones.
Gravity equation
• “the workhorse for empirical studies” in
international economics
– Eichengreen, Irwin 1997
– responsible for the eruption of the empirical works
Gravity equation
• Attractiveness
– a possibility to obtain the transparent answer to most
important questions about the determinants of
bilateral trade
– strong fit to the data and the possibility to test a
variety of hypothesis by adding proxies of trade costs.
• in order to evaluate the trade effect of economic integrations,
can be added
– dummy variables for membership in particular agreement
The traditional version of a gravity model
• value of export is a function of bilateral trade for
pair of countries, their GDPs and the distance
between them
ln EXPORTijt = β 0 + β1 ln(GDPi t ) + β 2 ln(GDPjt ) + β 3 ln GDPpcit − GDPpctj +
+ β 4 ln dist ij + ε ijt
EXPORTijt
- exports from country i do j, time t
GDPi t
- nominal GDP of country i
GDPjt
- nominal GDP of country i
GDPpcit − GDPpc tj
distij
- difference of GDP per capita between i and j
- distance between country i and j.
The gravity equation & theory
• can be derived from a variety of theoretical
models based on
– neoclassical or monopolistic competition approaches
– for homogenous and differentiated goods
– with the representation of the role of
• technology,
• factor endowments
• demand differences.
The gravity equation & theory
• Anderson (1979), Bergstrand (1985, 1989),
Helpman and Krugman (1985), Deardoff (1998),
Anderson and van Wincoop (2001) Eaton and
Kortum (2001)
– have given the theoretical background for this popular
tool for measuring the trade effects.
• Anderson (1979)
– a theoretical foundation for the gravity model based
on constant elasticity of substitution (CES) preferences
and goods that are differentiated by the region of
origin.
The gravity equation & theory
• Bergstrand (1989, 1990) and Deardoff (1998)
– have preserved the CES preference structure and
added monopolistic competition or a Hecksher-Ohlin
structure in order to include the specialization effect.
• Anderson and Wincoop (2001)
– provided the theoretical explanation of how border
effects effect trade.
• Bergstrand (1989)
– the first to derive the gravity equation including per
capita incomes as independent variables.
The gravity equation & variables control the
impact of regionalism on exports
• PSA - dummy variable indicating whether both trading countries are
the members of a partial scope agreement, data obtained from WTO
database
• PSA&EIA - dummy variable indicating whether both trading countries
are the members of a partial scope agreement and economic
integration agreement, data obtained from WTO database
• FTA - dummy variable indicating whether both trading countries are
the members of a free trade area
• FTA&EIA - dummy variable indicating whether both trading countries
are the members of a free trade area and economic integration
agreement
• CU - dummy variable indicating whether both trading countries are
the members of a customs union, variable controls the impact of
regionalism on exports
• CU&EIA - dummy variable indicating whether both trading countries
are the members of a customs union, variable controls the impact of
regionalism on exports and economic integration agreement
Gravity modeling of RTAs
The gravity model
• The choice of proper estimation method
• to adopt one of the typical panel data based
estimators
– fixed or random effects approach.
• the main disandvantage of the fixed effects
approach is the unavailability of parameter
estimates on the variables that are constant over
time for
– example of this kind of variables is a distance between
a reporter and its trade partner.
The gravity model
• follow most authors and assume exogeneity of
the regressors, without testing it with some
particular test
• one solution to be applied
– the Hausman-Taylor estimation method
• it allows for the use of both time-varying and time invariant
variables
– it is allowed that some of them can be endogeneous in the
sense of correlation with individual effects, but still exogeneous
with respect to idiosyncratic error term.