Transcript Document

Raff
Trade, Heterogeneity,
Intermediation
1
International Trade, Firm Heterogeneity, and
Intermediation
Horst Raff, University of Kiel
Zhejiang University
17-19 May 2011
Syllabus
Raff
Trade, Heterogeneity,
Intermediation
2

International Trade with Heterogeneous Firms
i.
ii.
iii.
Introduction
Trade Model with Monoplistic Competition (Krugman)
Monopolistic Competition with Heterogeneous Firms (Melitz,
Ottaviano)
iv. Reciprocal Dumping Model (Brander)
v. Reciprocal Dumping Model with Heterogeneous Firms (Long,
Raff, Stähler)
vi. Trade and Innovation
vii. Intra-Industry Adjustment to Import Competition

Intermediation in International Trade
i.
ii.
iii.
iv.
Introduction
Buyer Power in International Markets
Imports and the Structure of Retail Markets
Manufacturers and Retailers in the Global Economy
Introduction
Raff
Trade, Heterogeneity,
Intermediation
3
Intermediaries in International Trade
1. Role of retailers/wholesalers in intermediating
2.
3.
4.
5.
6.
international trade
Welfare effects of trade liberalization when there are
big, powerful buyers who may capture rents from
manufacturers and/or consumers
Effect of retail market structure and retail margins on
pass-through of import prices into consumer prices
Effect of retail regulation on trade and pass-through
Structural shifts in employment from manufacturing
into retailing
Retailers as gatekeepers to consumer markets
Buyer Power
Raff
Trade, Heterogeneity,
Intermediation
4
Buyer Power in International Markets
Horst Raff and Nicolas Schmitt
Journal of International Economics 2009
Buyer Power
Raff
Trade, Heterogeneity,
Intermediation
5
Buyer power:
Exercise of significant market power by retailers/wholesalers
What is the impact of buyer power on international markets:
•
•
•
•
the volume of international trade
the terms of trade
consumer prices
domestic welfare/gains from trade?
Buyer Power
Raff
Trade, Heterogeneity,
Intermediation
6
Stylized facts





In many countries, retailers and wholesalers have become
significantly more powerful in recent years compared to
manufacturers.
In EU food retailing, the concentration ratio rose by 20% between
1993 and 1999.
The aggregate concentration ratio is much higher than in
manufacturing:
• the 20 largest retailing firms account for 43% of aggregate EU
retail food turnover
• equivalent number for manufacturing: only 14.5%.
The same phenomenon has been observed in many other markets,
such as apparel and clothing.
In these markets, the retailers impose contract terms on
manufacturers, not the other way round.
Buyer Power
Raff
Trade, Heterogeneity,
Intermediation
7
Wal-Mart which is today the world's biggest
company by sales (US$312.4 billion).
Procter & Gamble employs 200 people who work solely on the
Wal-Mart account. The company, along with many others in
the Fortune 500, set up offices in northwest Arkansas to be
more accessible to Wal-Mart, its largest client.
Raff
Trade, Heterogeneity,
Intermediation
8
Source: Basker and Van (2005), p. 50
Buyer Power
Raff
Trade, Heterogeneity,
Intermediation
9
Why is it important to look at buyer power from a trade
perspective?
1. Retailers with buyer power participate extensively in
international markets:
•
Wal-Mart alone accounted for an incredible 10% of
total US imports from China in 2004 (Basker and
Hoang Van, 2005; Fishman, 2006).
•
Wal-Mart imports more than half of its non-food
products (Smith, 2004).
•
In the apparel market, 12% of the apparel sold by US
retailers in 1975 were imported against 48% in 1993.
Buyer Power
Raff
Trade, Heterogeneity,
Intermediation
10
•
Without buyer power, imports from China and other
developing countries would not have grown as quickly as
they did over the last decade.
-
Pressure from big retailers (Wal-Mart) has forced
domestic suppliers to relocate production abroad.
-
Major retailers also buy directly from low cost
countries. Most major US retailers have overseas
buying offices, especially in East Asia, with contacts
with a large network of suppliers.
-
In 2002, Wal-Mart took over Pacific Resources
Exports (PREL), its exclusive global buyer between
1989 and 2002. PREL lists over 6000 suppliers, 80%
of which are located in China
Raff
Trade, Heterogeneity,
Intermediation
11
Source: Basker and Van (2005), p. 50
Buyer Power
Raff
Trade, Heterogeneity,
Intermediation
12
•
Role of big buyers may be crucial in understanding why,
despite formidable spatial and cultural distances,
countries like Japan, South Korea, Taiwan, Hong Kong,
Singapore, and now China have been so successful and
for so long in exporting to Western countries.
•
East Asian growth in trade may be better explained by
the role of buyer-driven global commodity chains than by
more traditional explanations, such as the role of exportoriented policies (Gereffi, 1999).
•
But buyer power has also been blamed for limiting trade
by making import penetration more difficult than it would
otherwise have been, thanks in particular to the use of
exclusive agreements.
Buyer Power
Raff
Trade, Heterogeneity,
Intermediation
13
Why is it important to look at buyer power from a trade
perspective?
2. Why does buyer power seem to be on the rise?
•
Fundamental reason: greater prevalence of
differentiated products in international markets.
•
These goods require a lot of information and a
good match between the characteristics of buyers
and sellers (Rauch and Feenstra, 1999).
•
Prices are not enough to convey the necessary
information; hence some institutions must emerge
to handle these issues.
Buyer Power
Raff
Trade, Heterogeneity,
Intermediation
14
•
If buyers are in the driving seat, it must because
they know consumers' characteristics better than
segmented and/or distant suppliers. Buyer power
may occur naturally in association with
international markets.
•
Buyer power can have far-reaching effects on
international markets.
•
But it is far from clear whether buyer power should
increase or decrease trade and welfare.
•
Hence, it seems important to start investigating
the role of buyer power for international markets.
Buyer Power
Raff
Trade, Heterogeneity,
Intermediation
15
Buyer Power and Welfare – IO Perspective
 Countervailing power hypothesis (Galbraith, 1952): buyer power
counteracts seller power (second-best solution).
 If sellers have little or no power, increased buyer power
unambiguously leads to higher retail prices and lower welfare.
 Recent IO literature looks at implications for prices and welfare of
different contractual tools big buyers may employ w.r.t. to
manufacturers:
• Marx and Shaffer (2004), Rey and al. (2005), Inderst and Wey
(2004)
• General conclusion: retailers with market power have
considerable scope for anti-competitive behavior.
Buyer Power
Raff
Trade, Heterogeneity,
Intermediation
16
Our approach
• Start from recent IO literature by looking explicitly at
the contractual arrangements between buyers and
sellers.
• Extend the analysis to an international environment
characterized by barriers to trade and asymmetries
in the market shares of manufacturers.
• How does trade liberalization affect consumer
prices and welfare in the presence of buyer power?
• How does this compares to a world in which
producers have market power?
Buyer Power
Raff
Trade, Heterogeneity,
Intermediation
17
Trade Literature on Intermediaries
•
Rauch (2001) on the role of networks in international trade,
•
Feenstra and Hanson (2004) on the role of Hong Kong
intermediaries with respect to Chinese products,
•
Raff and Schmitt (2005, 2006) on the role of exclusive territory
and exclusive dealing in international markets
•
Richardson (2004) on the comparison between exclusivity in
the distribution of domestic products and trade policy to restrict
the market access of foreign producers.
•
Basker and Van (2005): only paper on buyer power in an
international trade context. Focus is different from ours since
they want to explain why, in the presence of economies of
scale in retailing and in the import process, trade liberalization
has led to an explosion of imports by large buyers.
Buyer Power
Raff
Trade, Heterogeneity,
Intermediation
18
Model




Two differentiated retailers, R1 and R2, who distribute
a homogeneous product in the domestic market.
The product can be obtained from a domestic
manufacturer, h, and/or a foreign manufacturer, f.
Costs:
•
•
Marginal production cost: c
Marginal cost of distribution: zero
Preferences:
where q‘s are quantities bought from the retailers.
Buyer Power
Raff
Trade, Heterogeneity,
Intermediation
19
•
Demand faced by retailer i=1,2:
•
Contracts:
-
•
Retailer i offers two-part tariffs to manuf. j, contingent on
exclusivity E or no exclusivity N (j may sell to rival retailer)
Timing of the game
1. Retailers make take-it-or-leave-it offers to manufacturers.
2. Manufacturers accept or reject contracts from one or both
retailers.
3. Contracts are implemented and retailers compete in prices.
Buyer Power
Raff
Trade, Heterogeneity,
Intermediation
20
Approach:
 Derive equilibrium in autarky, where retailers can only by from
manuf. h (Rey, Thal and Verge, 2005)
 Then derive equilibrium in free trade, where both manufacturers
are active.
 Compare equilibrium prices and welfare in the two cases.
 Extend results to the case of positive, non-prohibitive trade
costs.
Two equilibrium outcomes:
 Exclusive contracts (E): manuf. sign exclusive contracts with
one retailer, the other retailer does not sell
 Non-exclusive contracts (N): both retailers are active (either
non-exclusive contracts or each retailer has exclusive contract
with a separate manuf.)
Buyer Power
Raff
Trade, Heterogeneity,
Intermediation
21
Autarky – Exclusive contract equilibria
• Contracts:
- conditional on exclusivity, offer the manufacturer a wholesale
price equal to marginal cost and a fixed transfer equal to the
entire monopoly profit,
- for non-exclusivity, zero transfer.
• Manuf. h accepts one of the exclusive contracts.
• These equilibria always exist - trivially.
• Retailers compete to be the exclusive retailer, and are forced to give
up the entire profit to the manuf.
• Equilibrium prices:
- demand faced by the active retailer: q=1-p
- wholesale price: c
- retail price:
- manuf. profit:
Buyer Power
Raff
Trade, Heterogeneity,
Intermediation
22
Autarky - Non-exclusive equilibria
•
•
Consider the one that is Pareto-undominated for the retailers.
Two equilibrium conditions:
1. Manuf. must be indifferent between accepting one retailer's
exclusive contract and accepting both retailers non-exclusive
contracts.
2. Wholesale price offered by retailer i has to maximize the joint
profit of retailer i and the manuf. given the wholesale price
offered by the rival retailer -i.
•
Equilibrium wholesale price:
•
Equilibrium retail price:
Buyer Power
Raff
Trade, Heterogeneity,
Intermediation
23
•
Equilibrium existence:
-
Necessary condition for existence of a non-exclusive
equilibrium: joint profit of the manufacturer and both retailers in
the non-exclusive equilibrium exceeds the joint profit of a single
retailer-manufacturer pair under exclusive contracts.
-
This is satisfied if b≤0.73205, i.e., when the retailers are
sufficiently differentiated.
-
Only in this case are there enough rents to prevent retailers from
deviating to offering an exclusive distribution arrangement to the
manufacturer.
Raff
Trade, Heterogeneity,
Intermediation
Buyer Power
24
Non-exclusive
0
0,73205
Exclusive
1
b
Buyer Power
Raff
Trade, Heterogeneity,
Intermediation
25
Free Trade – Exclusive contract equilibria

harder to sustain than in autarky, since the active retailer now
has to compensate two manuf. for not selling to his rival.

We show that such an equilibrium can only exist if there is
sufficiently little differentiation between retailers (b≥0.61803).

Equilibrium prices are the same as in the foreclosure equilibrium
in autarky.

Retailers compete to foreclose their rival and in the process
transfer their entire profit to the two manuf.
Buyer Power
Raff
Trade, Heterogeneity,
Intermediation
26
Free Trade - Non-exclusive equilibria
•
There are two possibilities:
1. each retailer deals exclusively with one manufacturer
2. at least one retailer buys from both manufacturers under a
non-exclusive contract.
•
Derive equilibria in four steps:
- Prove that case (2) cannot occur in equilibrium, by showing
that a deviation to dealing with a single manuf. is profitable.
- Calculate equilibrium wholesale prices when each retailer
buys from a single manufacturer.
- Show that proposed contracts are best responses.
- Verify necessary condition for existence.
Raff
Trade, Heterogeneity,
Intermediation
Buyer Power
27
Free
Trade
Non-exclusive
b
0
0,61803
0,67209
1
Exclusive
Non-exclusive
Autarky
0
0,73205
Exclusive
1
b
Buyer Power
Raff
Trade, Heterogeneity,
Intermediation
28
Effects of Trade Liberalization – contracts and prices
Buyer Power
Raff
Trade, Heterogeneity,
Intermediation
29
Effects of Trade Liberalization – social welfare
Buyer Power
Raff
Trade, Heterogeneity,
Intermediation
30
Buyer versus seller power
Buyer Power
Raff
Trade, Heterogeneity,
Intermediation
31
Buyer versus seller power
•
•
Size of the rents accruing to the retailers and to the manufacturers
differs depending on who has the power to make take-it-or-leave-it
contract offers.
More importantly, equilibrium prices and hence the competitive effects
of free trade are different.
Autarky
•
•
•
Prices are at least as high under seller power than under buyer power.
Reason: the domestic manufacturer is able to eliminate the
competition between retailers by setting a wholesale price above
marginal cost and extracting rents through the fixed fee.
Under buyer power, retailers only internalize the effect of their
wholesale price on the manufacturer, but not on the rival retailer.
Hence competition is tougher under non-exclusive contracts.
Buyer Power
Raff
Trade, Heterogeneity,
Intermediation
32
Free Trade
Now compare retail prices in free trade:
• When there are non-exclusive contracts, retail price are the
same under both buyer and seller power.
• But exclusivity, which occurs only under buyer power, leads to
higher retail prices
Buyer Power
Raff
Trade, Heterogeneity,
Intermediation
33
Buyer Power
Raff
Trade, Heterogeneity,
Intermediation
34
Conclusions
•
Opening up markets to the forces of international trade has
traditionally been seen as a means of inducing domestic
industries to become competitive and more efficient.
•
Monopoly rents are diluted, consumers gain from competition.
•
Is this really true or can the rents that manufacturers may lose
be captured by retailers, wholesalers and other intermediaries,
especially once these become unavoidable agents in the
process of reaching consumers?
Buyer Power
Raff
Trade, Heterogeneity,
Intermediation
35
We obtain some surprising results:
1.
Under some circumstances the rents can be completely
captured by manufacturers once free trade is introduced, even if
additional sources of supply are available in free trade and even
if there is (imperfect) competition among retailers.
2.
Hence buyer power does not necessarily mean that retailers
capture the rents in free trade.
3.
Price competition can be lower in free trade than in autarky
because an equilibrium in which some retailers are foreclosed
may be easier to sustain in free trade.
4.
Hence economic integration may lead to a smaller increases in
competition and smaller welfare gains than suggested by
traditional models.
Buyer Power
Raff
Trade, Heterogeneity,
Intermediation
36
What comes next?
•
•
We do not yet have a theory of buyer power in an international
context.
Buyer power in our model is exogenous: the retailers have all
the bargaining power irrespective of the trade environment.
We only spell out the implications of the existence of buyer
power – and find that it matters!
We have nothing to say about the idea that buyer power
might be a by-product of freer trade.
Need to endogenize market structure in retailing and in
manufacturing to capture general equilibrium effects.
Heterogeneous Retailers
Raff
Trade, Heterogeneity,
Intermediation
37
International Trade and the
Structure of Retail Markets
Horst Raff (University of Kiel)
Nicolas Schmitt (Simon Fraser University)
Heterogeneous Retailers
Raff
Trade, Heterogeneity,
Intermediation
38
How does the retailing sector respond to globalization,
especially the increased scope to import consumer
products?
1.
What are the effects of trade liberalization on the structure of retail
markets and retail competition?
2.
How does retail market structure affect the transmission of external
shocks (fall in trade costs, exchange-rate changes) into consumer
prices?
3.
How does retail market regulation affect consumer prices, imports,
pass-through of import prices?
We address these questions in a model of trade with
heterogeneous retailers.
Heterogeneous Retailers
Raff
Trade, Heterogeneity,
Intermediation
39
Three stylized facts:
1.
Retail market structure has changed dramatically
•
•
2.
Strong increase in market concentration due to growth of big
national chains operating large stores.
Retailing is often much more concentrated than manufacturing.
Big retailers handle a massive amount of trade
•
•
•
Wal-Mart accounts for over 15% of US imports from China
(Basker/Van).
Big retailers are three times as likely to import from China as small
retailers.
Expansion of big retailers accounts for 19% of the growth in US
consumer good imports from China.
Heterogeneous Retailers
Raff
Trade, Heterogeneity,
Intermediation
40
3.
Distribution margins account for up to 50% of consumer
prices
•
Changes in the cost structure and competition in the retail sector
will have strong effects on consumer prices and the demand for
imports.
Heterogeneous Retailers
Raff
Trade, Heterogeneity,
Intermediation
41
Policy Issues:
1.
Pass-Through from Import to Consumer Prices
•
•
•
•
•
Determines the gains from trade for consumers.
Monetary policy and the choice of exchange rate regime depend
on how external shocks are passed through into consumer prices.
Pass-through from import into consumer prices appears to be very
low (Exchange-rate disconnect puzzle).
Retail sector plays a central role in explaining the seemingly low
degree of pass-through since retail margins are such a big part of
consumer prices.
Which retail sector characteristics (trade costs, technology,
productivity distribution,…) lead to the low pass-through rate?
Heterogeneous Retailers
Raff
Trade, Heterogeneity,
Intermediation
42
2.
Retail-Market Regulation
•
•
•
•
•
Some countries (France, Japan, Belgium, Italy,…) have a tradition
of restricting the size of retail establishments.
This affects retail market structure, imports and consumer prices.
It limits the pass-through of import prices.
France: consumers complain that their purchasing power is falling
and that they don’t have access to cheap foreign goods.
US: poor consumers benefit from the low-priced goods that WalMart and other big retailers import from China (Broda and Romalis,
2008).
Heterogeneous Retailers
Raff
Trade, Heterogeneity,
Intermediation
43
How does trade liberalization affect retail markets?

Retailers source goods through two channels:
•
•

domestic sourcing – includes domestically produced goods and
goods imported by third parties;
imports.
Only the big retailers choose imports:
•
•
low marginal cost (bypass additional layers of intermediaries, identify
cheapest supplier,…)
high fixed cost (operating buying offices, searching for suppliers,
developing products, training suppliers, monitoring quality,…)
Heterogeneous Retailers
Raff
Trade, Heterogeneity,
Intermediation
44


Hence trade liberalization favors big retailers:
•
•
•
they tend to buy more imports,
grow even bigger as a result,
displace smaller competitors.
Evidence:
•
Retail segments where the share of large retailers is high
correspond to segments where the share of imports is high as
well (Canada 2003: clothing, clothing accessories, footwear,
audio, video, small electrical appliances, toys and games).
•
US imports from China and other less-developed countries in
1997-2002 rose especially quickly in retail sectors with the
largest consolidation into chains (Basker and Van, 2008b).
Heterogeneous Retailers
Raff
Trade, Heterogeneity,
Intermediation
45
Modelling issues
1.
2.
3.
Significant size differences among retailers
•
•
Few big, efficient firms and many small, inefficient firms.
Pareto distribution of productivity provides a good fit (Campbell and
Hopenhayn, 2005).
Fixed cost of direct imports
•
Cost of maintaining buying offices, cooperating with foreign partners
to bypass intermediaries, acquiring information,...
Endogenous mark-ups
•
Important channel for pass-through.
Melitz, Ottaviano, Market Size, Trade and Productivity, REStud 2008.
Heterogeneous Retailers
Raff
Trade, Heterogeneity,
Intermediation
46
Model of firm heterogeneity in retailing

Continuum of retailers selling in the domestic market:
• retail services are non-tradable
• retailers are differentiated (products, retail location or services
differ)

Consumers:
• L consumers each supplying 1 unit of labor
• Each consumer has quasi-linear preferences over goods sold by
retailer i and the numeraire good y:
Heterogeneous Retailers
Raff
Trade, Heterogeneity,
Intermediation
47

Market demand faced by retailer i,
where
N is the mass of consumed varieties,
is the average retail price.
Heterogeneous Retailers
Raff
Trade, Heterogeneity,
Intermediation
48

Technology
•
•
Labor is the only factor of production (price of labor = 1)
Labor requirements:
Heterogeneous Retailers
Raff
Trade, Heterogeneity,
Intermediation
49

Pareto distribution of cost over support

Monopolistic competition in retailing:
• Retailers take the average industry price as given,
• earn zero profit in equilibrium
Heterogeneous Retailers
Raff
Trade, Heterogeneity,
Intermediation
50

Profit maximization by retailer i:
• When sourcing domestically:
•

When relying on direct imports:
Profit maximizing outputs:
Heterogeneous Retailers
Raff
Trade, Heterogeneity,
Intermediation
51

Maximized profits:

Marginal cost at which a retailer is indifferent between domestic
sourcing and direct imports:

Expected-zero-profit condition:
Raff
Trade, Heterogeneity,
Intermediation
Heterogeneous Retailers
52
Equilibrium

Endogenous variables:

How does trade liberalization (marginal decrease in t) affect
the equilibrium values of these variables?
•
Examine the signs of
•
Then determine how the average retail price, number
of active retailers, average output and the variance of
output change with trade liberalization.
and
Raff
Trade, Heterogeneity,
Intermediation
Heterogeneous Retailers
53
Profit functions (gross of sunk entry cost)
Profits
c
Imports
Domestic
Sourcing
Inactive
Raff
Trade, Heterogeneity,
Intermediation
Heterogeneous Retailers
54
Trade Liberalization
Profits
c
Imports
Domestic
Sourcing
Inactive
Heterogeneous Retailers
Raff
Trade, Heterogeneity,
Intermediation
55
Effects of Trade liberalization


More retailers source goods from abroad
Least efficient retailers become inactive
This implies:
• Lower average retail price
and, if the trade cost is sufficiently small,
•
•
Higher average output, lower output variance
Lower mass of active retailers
Heterogeneous Retailers
Raff
Trade, Heterogeneity,
Intermediation
56
1.
Pass-through of import prices
into

consumer prices of imported goods

consumer prices of domestic goods

average consumer prices
Heterogeneous Retailers
Raff
Trade, Heterogeneity,
Intermediation
57
Heterogeneous Retailers
Raff
Trade, Heterogeneity,
Intermediation
58
Pass-through

First term: mark-up (share of cost savings passed on to
consumers) times expected import share
•


Import share depends on trade cost and k (high k = distribution
skewed toward inefficient retailers)
Last three terms (from selection effect):
•
•
•
fall in expected retail cost
lower mark-up for domestically sourced goods
Higher import share/prob. that a good is imported
Overall pass-through rate can be greater than unity.
Heterogeneous Retailers
Raff
Trade, Heterogeneity,
Intermediation
59
2.
Retail Market Regulation



France: “Loi Raffarin”
Japan: Large-Scale-Retail-Store Law
Regulation acts like a constraint on the sales of the most
efficient firms.
How does this affect
• retail market structure,
• consumer prices,
• imports,
• pass-through of import prices?

Heterogeneous Retailers
Raff
Trade, Heterogeneity,
Intermediation
60

Critical value of the marginal cost at which a firm is just
constrained (assuming that the firm for which the constraint
is just binding is an importer):

What are the effects of a tightening of constraint
•
•
•
?
Even less efficient importers will now be constrained.
Constrained retailers raise their prices.
Residual demand for unconstrained retailers rises.
Heterogeneous Retailers
Raff
Trade, Heterogeneity,
Intermediation
61
Profits of unconstrained firms
Profits
c
Direct
Imports
Domestic
Sourcing
Heterogeneous Retailers
Raff
Trade, Heterogeneity,
Intermediation
62
What happens to the average retail price?
 Constrained firms raise their price, and the likelihood that a
firm is constrained is higher.
 Greater probability that inefficient retailers remain active.
 But retailers are more likely to source (cheaper) goods from
abroad.
Heterogeneous Retailers
Raff
Trade, Heterogeneity,
Intermediation
63
Pass-through of import into consumer prices


Constrained firms do not change their retail price when
import prices fall.
Only unconstrained firms lower their price in response to a
fall in import prices.
Heterogeneous Retailers
Raff
Trade, Heterogeneity,
Intermediation
64
Retail Competition and Welfare

Competition: Herfindahl Index

Welfare effect
Heterogeneous Retailers
Raff
Trade, Heterogeneity,
Intermediation
65
Conclusions
1.
Trade liberalization changes retail market structure and
reduces average retail prices:


More direct imports
Small retailers are forced to shut down
Heterogeneous Retailers
Raff
Trade, Heterogeneity,
Intermediation
66
2.
Pass-through of import into consumer prices:


Incomplete and increasing when trade is liberalized
Consistent with empirical evidence
• Pass-through has increased over time (Campa, Goldberg).
• Differences in pass-through rates in the Euro area are
driven by differences in openness to non-Euro-area imports
(Campa, Gonzales Minguez, 2006).
Heterogeneous Retailers
Raff
Trade, Heterogeneity,
Intermediation
67
3.
Retail Market Regulation




Raises the likelihood that inefficient retailers survive
Induces more retailers to source from abroad
Tends to raise average consumer prices
Tends to reduce pass-through
Raff/Schmitt
Raff
Manuf.
Trade,and
Heterogeneity,
Retailers in
the Global
Intermediation
Economy
68
Manufacturers and Retailers in
the Global Economy
Horst Raff (University of Kiel)
Nicolas Schmitt (Simon Fraser University)
Stylized Facts
Raff/Schmitt
Raff
Manuf.
Trade,and
Heterogeneity,
Retailers in
the Global
Intermediation
Economy
69
1.
Increased importance of services in general and
wholesale/retail trade in particular
•
•
•
•
•
US employment fell in manufacturing between 1970 and 1990 but
rose by 71% in wholesale/retail trade;
US retail industry is second largest industry in terms of employment
(10.5% share).
In Canada, retailing accounted for 12% of employment in 2004.
50-60% of retailer expenditure on inputs is on labor.
Retailing contributed 22% to productivity growth in the business
sector between 1990 and 2003.
2. Retailers are becoming bigger, retailing more concentrated
•
•
•
More varieties,
Greater average size,
Higher ratio of square footage to sales.
Stylized Facts
Raff/Schmitt
Raff
Manuf.
Trade,and
Heterogeneity,
Retailers in
the Global
Intermediation
Economy
70
3. Retailers have become gatekeepers of consumer goods
markets.
•
•
•
•
•
Slotting allowances (lump-sum payments made by manufacturers to
retailers to carry their products) have become an important feature
of retailing.
Slotting allowances started in US in the early 1980s.
They exist now in groceries, tobacco, household supplies, health
and beauty aid, textiles, footwear, automotive parts, etc;
Typically explained by retailers choosing among many new
products, many of them likely to fail;
Slotting not used by all retailers in a given segment and they vary a
lot across products.
Research Questions
Raff/Schmitt
Raff
Manuf.
Trade,and
Heterogeneity,
Retailers in
the Global
Intermediation
Economy
71
1.
What explains the reallocation of labor from manufacturing
to retailing, market concentration in retailing, the rise of
slotting allowances?
•
•
Trade liberalization: increase in the number of
imported varieties
Technological change: new technology has made it
cheaper to handle larger variety of goods
Research Questions
Raff/Schmitt
Raff
Manuf.
Trade,and
Heterogeneity,
Retailers in
the Global
Intermediation
Economy
72
2.
How does trade liberalization affect product variety, prices
and welfare when retailers act as gatekeepers?
We address these questions in a general equilibrium
model of manufacturing and retailing.
Model Components
Raff/Schmitt
Raff
Manuf.
Trade,and
Heterogeneity,
Retailers in
the Global
Intermediation
Economy
73
1.
2.
Retailing (Feenstra and Ma, 2008)
•
•
Multi-product retailers choosing assortment and retail prices
Each retailer takes into account the impact of adding a variety on the
demand for the other varieties it sells (cannibalization)
Manufacturing (Krugman, 1980)
•
•
3.
Monopolistic competition
Single-product manufacturers
Wholesale market (Raff and Schmitt, 2009)
•
•
Bargaining between manufacturers and retailers over the wholesale
price
Two-part tariff
Model
Raff/Schmitt
Raff
Manuf.
Trade,and
Heterogeneity,
Retailers in
the Global
Intermediation
Economy
74
I. Model
Key elements:
•
•
Dixit-Stiglitz preferences,
only factor of production: labor with endowment L.
Preferences:
Demand for variety i:
where
L
is aggregate expenditure on differentiated products
Model
Raff/Schmitt
Raff
Manuf.
Trade,and
Heterogeneity,
Retailers in
the Global
Intermediation
Economy
75



R retailers, all carrying different varieties
Retailer 1 carries the first M1 products, retailer 2 carries the
following M2 products….
Total mass of varieties consumed

The CES price index is:
Raff/Schmitt
Raff
Manuf.
Trade,and
Heterogeneity,
Retailers in
the Global
Intermediation
Economy
Model
76



Symmetry: each retailer sets the same retail price for all the products he
sells:
R

1 
CES price index P  
M r pr 

 r 1

Price elasticity of demand 
1
1
yr pr
  (1  sr )  sr
pr yr
where market share of retailer r is:
sr 
M r pr y r
r1 M r pr yr
R

M r p1r 
1
M
p
r1 r r
R
Raff/Schmitt
Raff
Manuf.
Trade,and
Heterogeneity,
Retailers in
the Global
Intermediation
Economy
Model
77
Technology
l  k0  k1M r
r

Retailing labor requirement:

Manufacturing labor requirement to produce variety i:
Model
Raff/Schmitt
Raff
Manuf.
Trade,and
Heterogeneity,
Retailers in
the Global
Intermediation
Economy
78
Wholesale Market

Two-part tariffs
wr , Tr 
w – wholesale price
T – transfer/fixed payment from the manufacturer to the retailer
(slotting allowance if T is positive).

Bargaining:
•
•
Simultaneous bargaining between retailer-manufacturer pairs
Division of surplus of each pair through “efficient” bargaining
Equilibrium
Raff/Schmitt
Raff
Manuf.
Trade,and
Heterogeneity,
Retailers in
the Global
Intermediation
Economy
79
II. Equilibrium

Retailer maximizes:

First-order conditions:
• with respect to retail price leads to:
•
with respect to retailer’s assortment, M:
marginal benefit is reduced by the cannibalization effect
Equilibrium
Raff/Schmitt
Raff
Manuf.
Trade,and
Heterogeneity,
Retailers in
the Global
Intermediation
Economy
80

Total surplus for a manufacturer-retailer pair:

This surplus is maximized for a wholesale price>mc (due to
cannibalization effect):

Externality since each retailer-manufacturer pair ignores the impact of
its deal on other retailer-manufacturer pairs.

The zero-profit condition for manufacturer then determines T:
Equilibrium
Raff/Schmitt
Raff
Manuf.
Trade,and
Heterogeneity,
Retailers in
the Global
Intermediation
Economy
81

Zero-profit condition for retailers

Labor-market clearing condition

We can now solve for the equilibrium values of:
w, T , y, p, R, M
Equilibrium
Raff/Schmitt
Raff
Manuf.
Trade,and
Heterogeneity,
Retailers in
the Global
Intermediation
Economy
82
# of retailers:
Retailer’s assortment:
Retail price:
Output per variety:
Slotting allowance:
Equilibrium
Raff/Schmitt
Raff
Manuf.
Trade,and
Heterogeneity,
Retailers in
the Global
Intermediation
Economy
83
Proposition 1: The equilibrium slotting allowance is
i. increasing in the retailer fixed cost ko and
cost per variety k1
ii. decreasing in the manufacturer fixed cost
iii. increasing in the elasticity of substitution
Equilibrium
Raff/Schmitt
Raff
Manuf.
Trade,and
Heterogeneity,
Retailers in
the Global
Intermediation
Economy
84
III. Second-best Allocation
# of retailers:
Retailer’s assortment:
Output per variety:
Retail price:
Equilibrium
Raff/Schmitt
Raff
Manuf.
Trade,and
Heterogeneity,
Retailers in
the Global
Intermediation
Economy
85
This implies:
Excessive product variety implies that too much labor is
devoted to distributing these varieties as opposed to producing
output of each variety.
Comparative Statics
Raff/Schmitt
Raff
Manuf.
Trade,and
Heterogeneity,
Retailers in
the Global
Intermediation
Economy
86
IV. Product market integration and technological
change in retailing

What are the forces driving the structural changes in retailing and
manufacturing?
• Shift in employment from manufacturing to retailing
• Retail market concentration
• Rise in slotting allowances.

Can we link these changes to market integration or technological
change?
Economic Integration
Raff/Schmitt
Raff
Manuf.
Trade,and
Heterogeneity,
Retailers in
the Global
Intermediation
Economy
87
A. Product market integration with C countries


Market size remains the same for retailers (retail services
are non-tradeable).
Market increases for manufacturers: each can now sell to
(and spread its fixed cost over) C markets.
Per-market fixed cost is equal to

Adjusted labor-market clearing condition:



Rk0  RM  k1    RMy  L
C

Economic Integration
Raff/Schmitt
Raff
Manuf.
Trade,and
Heterogeneity,
Retailers in
the Global
Intermediation
Economy
88
Proposition 4: Product-market integration:
i. leaves the number of retailers unchanged,
ii. raises the product assortment carried by each retailer,
iii. raises the total mass of varieties available to consumers,
iv. raises slotting allowance per variety
v. shifts labor from manufacturing to retailing.


Labor saved through decrease in domestic varieties is
reallocated to retailing, enabling retailers to carry more
imported varieties.
Slotting allowances increase as the quasi-rents earned by
manufacturers fall by less than the fixed cost.
Technological Change in Retailing
Raff/Schmitt
Raff
Manuf.
Trade,and
Heterogeneity,
Retailers in
the Global
Intermediation
Economy
89
B. Technological Change in Retailing
How is the equilibrium affected by a marginal reduction in k1 at
the expense of a marginal increase in ko ?
Proposition 5: Technological change in retailing
i. increases retail market concentration by reducing the
number of retailers and increasing product assortment
of each retailer,
ii. has ambiguous effects on the allocation of labor
between manufacturing and retailing and on the size
of slotting allowances.
Welfare
Raff/Schmitt
Raff
Manuf.
Trade,and
Heterogeneity,
Retailers in
the Global
Intermediation
Economy
90
V. Welfare and Policy Implications


Fundamental distortion in the relationship between multi-product
retailers and manufacturers.
Product-market integration does not reduce this distortion, although
welfare rises: consumers spread their income over a greater variety of
goods.
Retail Market Integration


Equivalent to full market integration: goods are simply traded by retailers
instead of manufacturers.
Greater number of retailers in the C countries so that the market share of
each retailer falls.
Welfare
Raff/Schmitt
Raff
Manuf.
Trade,and
Heterogeneity,
Retailers in
the Global
Intermediation
Economy
91
Proposition 6: Retail market integration
i. moves the economy closer to the second best,
ii. raises social welfare by more than product market
integration.



Retail market integration implies that consumers buy from a
larger number of retailers, each carrying a larger
assortment.
As retailer market share falls, the cannibalization effect
becomes smaller.
The wholesale price moves closer to marginal cost.
Conclusions
Raff/Schmitt
Raff
Manuf.
Trade,and
Heterogeneity,
Retailers in
the Global
Intermediation
Economy
92
1.
2.
3.
4.
Reallocation of labor from manufacturing to retailing and
rise in slotting allowances are consistent with productmarket integration.
Rise in retail market concentration is consistent with
technological change in retailing.
Product-market integration raises social welfare but does
not address the inefficiency created when big, multi-product
retailers deal with small, single-product manufacturers.
Retail market integration moves the economy closer to the
second best, hence yields bigger gains than product-market
integration.
Relation to the Literature
Raff/Schmitt
Raff
Manuf.
Trade,and
Heterogeneity,
Retailers in
the Global
Intermediation
Economy
93

Effects of trade on retail market structure:
Raff, Schmitt (2010): heterogeneous retailers face fixed
cost of importing, trade liberalization raises market
concentration in retailing as the big retailers get bigger
and the small retailers are forced out
Eckel (2010): increase in international product variety
raises retailer fixed costs and leads to market
concentration
Basker/Van (2009): economies of scale in importing
allow a chain store to take markets away from small,
independent retailers.
Relation to the Literature
Raff/Schmitt
Raff
Manuf.
Trade,and
Heterogeneity,
Retailers in
the Global
Intermediation
Economy
94

Retailer (buyer) market power may have negative effects on
the gains from trade:
Raff, Schmitt (2009): gains from trade are generally
smaller in markets with buyer power compared to
markets with seller power and may even be negative if
trade leads to more exclusive dealing.
Eckel (2010): possible losses from trade when greater
market concentration allows retailers to raise their
mark-up.
Relation to the Literature
Raff/Schmitt
Raff
Manuf.
Trade,and
Heterogeneity,
Retailers in
the Global
Intermediation
Economy
95

Growing literature on intermediation in international trade.
Raff, Schmitt (2005, 2006)
Antras, Costinot (2010)
Akerman (2010)
Blum, Claro, Horstmann (2010)
AER P&P (2010)
Raff
Trade, Heterogeneity,
Intermediation
96
Greasing the Wheels of International
Commerce: How Services Facilitate
Firms‘ International Sourcing
Peter Debaere (Virginia)
Holger Görg (Kiel)
Horst Raff (Kiel)
Research Questions
Raff
Trade, Heterogeneity,
Intermediation
97
What role do services play in the international sourcing of
material inputs?
1.
Does better access to services induce more offshoring by
manufacturing firms?
2.
How crucial is access to local services?
1.
Does this effect differ across firms?
We examine these questions both theoretically and
empirically using firm-level data from Ireland.
Observations
Raff
Trade, Heterogeneity,
Intermediation
98
1.
2.
Increased importance of global production networks
•
•
Firms pursue complex offshoring strategies
Imports of intermediates from many foreign suppliers
Global production requires “intermediation” services such
as transportation, financing, insurance,….
• Services grease the wheels of international commerce (Jones and
Kierzkowski, 1990).
3.
Large service sector geared to facilitating offshoring.
• Supply-chain management.
• Third-party logistics firms (UPS, DHL,…) coordinate and integrate a
fragmented value chain.
Literature
Raff
Trade, Heterogeneity,
Intermediation
99

Trade literature on offshoring (e.g. Antras/Helpman) has so
far largely ignored the link between services and offshoring.

Exceptions:
• Jones and Kierzkowski (1990)
• Deardorff (2000)
• Long, Riezman, Soubeyran (2005)

Huge business literature on supply-chain management that
explores how this link can be organized and optimized.
Structure of the Paper
Raff
Trade, Heterogeneity,
Intermediation
100

Model that explains the sourcing decision of individual
firms as a function of:
•
•

Availability of local services
Firm characteristics – labor productivity, size,…
Empirical analysis:
•
•
Plant-level data from Ireland on domestic and international
sourcing of materials and services.
How is the ratio of imported materials to total sales affected by firm
characteristics and the availability of services (number of service
firms in different regions in Ireland)?
Modelling Issues
Raff
Trade, Heterogeneity,
Intermediation
101
1.
Nearly all firms in Ireland source some materials from
abroad.
How much do they import relative to total material input or
total output?
2.


How does this fraction depend on the productivity of the firm?
How does it depend on the availability of local services?
Model:


Melitz, Ottaviano, Market Size, Trade and Productivity, REStud
2008.
Raff, Schmitt, Imports, Pass-Through, and the Structure of Retail
Markets (2009).
Model
Raff
Trade, Heterogeneity,
Intermediation
102
International sourcing by heterogeneous firms
Continuum of final-good producers selling in the local (Irish/European
market):
• Differentiated final products,
• monopolistic competition in final goods, perfect competition in
intermediates.
Consumers:
• With quasi-linear preferences:
Model
Raff
Trade, Heterogeneity,
Intermediation
103
•
Market demand faced by firm i is:
where
N is the mass of consumed varieties,
is the average price.
Model
Raff
Trade, Heterogeneity,
Intermediation
104
Technology
•
•
Sunk cost of market entry:
•
Intermediate x combines a domestic input, z, and an imported
input, m, according to:
•
Labor productivity is Pareto distributed:
Production of a finished good requires c units of labor and one
unit of a composite intermediate good x.
Model
Raff
Trade, Heterogeneity,
Intermediation
105
Two options for importing materials
1.
Use local services to organize supply chain (mode D):
•
•
2.
Variable cost of imported materials (plus services) =
This cost depends negatively on service availability.
Internalize the supply chain or import services (mode I):
•
•
Variable cost of imported materials = 1
Fixed cost of internalization/imports:
Model
Raff
Trade, Heterogeneity,
Intermediation
106

Conditional demands for imported intermediates:

Cost functions for composite intermediate input:

Profit maximization by firm i:
Model
Raff
Trade, Heterogeneity,
Intermediation
107

Profit-maximizing outputs:

Profit-maximizing prices:
Model
Raff
Trade, Heterogeneity,
Intermediation
108

Maximized profits:

Marginal cost at which a firm is indifferent between buying services on
the market and internalizing services:

Expected-zero-profit condition:
Raff
Trade, Heterogeneity,
Intermediation
Equilibrium
109
Profit functions (gross of sunk entry cost)
Profits
c
Mode I:
Internalize
services
Mode D: use
domestic services
Inactive
Equilibrium
Raff
Trade, Heterogeneity,
Intermediation
110

Endogenous variables:
Firm-level observations:



Spending on domestic materials, z
Spending on imported materials, m
Sales revenue: pq
How does service availability affect m/(pq)?
Equilibrium
Raff
Trade, Heterogeneity,
Intermediation
111
Effects of Service Availability

For mode-D firms a thicker service market implies a higher
ratio of imported intermediates to sales:
•
•
Substitution toward imported materials
Lower output price
Equilibrium
Raff
Trade, Heterogeneity,
Intermediation
112
Effects of Service Availability

Mode-I firms are only indirectly affected by service
availability:
•
•
no substitution effect,
price of output falls.
Empirical Implications
Raff
Trade, Heterogeneity,
Intermediation
113
 Ratio of imported materials to sales
• depends positively on availability of local services,
• is higher for more productive (lower c) firms.
 Most productive firms
• internalize/import services,
• and therefore are not directly affected by local
service market thickness.
Empirical model
Raff
Trade, Heterogeneity,
Intermediation
114



Plant i, industry j, region r, time t.
Thickness measure: thickness of services in industry j
and region r in which firm i operates at time t.
Controls X
• Labor productivity of firm i (firm heterogeneity)
• Size, dummies for exporters, foreign and domestic
•

multinationals (firm heterogeneity, internalization)
Industry services intensity, industry intensity of use of local
services
Dummies for three digit industry, region, time
Data
Raff
Trade, Heterogeneity,
Intermediation
115

Plant level data for Ireland: 2000 to 2004
• Annual Business Survey of Economic Impact
• Survey of plants with at least 10 employees
• Plants in manufacturing as well as (internationally
•
tradable) services
Provides information on materials and services purchases,
industry, nationality of ownership

In 2004, 1206 manufacturing plants, of which 343 foreign
MNEs, 108 Irish MNEs, 557 domestic exporters, 198
purely domestic

In 2000 also information on domestic multinationals

3 regions: Dublin, South, Border/Midlands/West
Data
Raff
Trade, Heterogeneity,
Intermediation
116

Standard stylized facts:
• MNEs are bigger, more productive than local firms
• Exporters are bigger, more productive than nonexporters

Material imports/sales revenue:
• 20% on average
• max: 27 % Transport Equipment (NACE 34)
• min: 12 % Food (NACE 15)

Service thickness in each region:
• # of service firms
• # of service MNCs vs. domestic service firms
• “effective thickness”: # of service firms x service use
in industry j
Alternative service availability measure
Raff
Trade, Heterogeneity,
Intermediation
117

variation across industries based on input-ouput linkages:
• Number of service firms in sector s:
• Inputs from service sector s used in manuf. sector j as
as a percentage of output of industry j:
• Number of manufacturing firms in region r:
Data
Raff
Trade, Heterogeneity,
Intermediation
118
Estimation Results
Raff
Trade, Heterogeneity,
Intermediation
119
 All firms vs. subgroups (domestic and foreign MNEs,


local firms, exporters)
Industry vs. firm-fixed effects
Different specifications to deal with endogeneity
Raff
Trade, Heterogeneity,
Intermediation
Estimation Results
120
Table 5: All Plants
Industry-Fixed Eff
Time, Region Eff.
LHS ln(imp/sales)
Services firm thickness 0.098**
Foreign MNC services
firm thickness
Domestic services firm
thickness
Labor productivity
0.008
0.06**
0.007
0.012
0.056**
0.007 .-0.013**
Dummy domestic
MNC
0.018*
Dummy foreign MNC
0.05***
Dummy exporter
0.051***
Firm size
0.007**
Observations
6937
6937
6937
R-squared
0.16
0.16
0.2
Dependent variable: ln(imported materials/sales)
Raff
Trade, Heterogeneity,
Intermediation
121
Table 5:
OLS, all
plants,
industry,
time, region
dummies
Dependent variable: ln(imported materials/sales)
Raff
Trade, Heterogeneity,
Intermediation
122
Table 5: OLS, subgroups, industry, time, region dummies.
Dependent variable: ln(imported materials/sales)
Raff
Trade, Heterogeneity,
Intermediation
123
Table 7: exporters only, firm-fixed eff., time eff.
Robustness checks
Raff
Trade, Heterogeneity,
Intermediation
124



Endogeneity of service availability measure
• Dummies!
• Fixed effects.
• IV estimator using regional housing price index, first
and second lags of availability measures as
instruments. Valid instruments. Cannot reject
exogeneity of variables (Durbin-Wu-Haussman test).
Include more time-varying variables at the region and
industry level to alleviate concern that availability may pick
up only regional time-varying characteristics.
Use alternative dependent variables:
• Imports relative to total inputs
• ln(exports*imports/sales)
Conclusions
Raff
Trade, Heterogeneity,
Intermediation
125
 Availability of services is important for the


decision to import material inputs.
This is especially so for local Irish firms.
Multinationals are not affected by the availability
of local services (may internalize services
provision or import services).
 Evidence that services indeed grease the
wheels of international commerce.
Raff
Trade, Heterogeneity,
Intermediation
Data
126
Year
2000
2001
2002
2003
2004
Imported
materials /
Revenue
mean
0.212
0.200
0.188
0.181
0.190
Std.dev.
0.147
0.151
0.149
0.147
0.146
Services
firm
thickness
mean
135.23
150.96
163.20
175.21
149.16
Std.dev.
1.71
1.77
1.73
1.68
1.64
Raff
Trade, Heterogeneity,
Intermediation
127