Globalization and Food Industry Consolidation: What Does Industrial Organization Have
to Say About Market Access?
Steve McCorriston (University of Exeter, UK)
(e-mail: [email protected])
Ian Sheldon (Ohio State University, USA)
(e-mail: [email protected])
Invited paper for First Biennial Research Conference
Food System Research Group
June 26-27, 2003
University of Wisconsin-Madison
JUNE 2003
Globalization and Food Industry Consolidation: What Does Industrial Organization Have
to Say About Market Access?
Abstract: This paper focuses on the interface between the industrial organization of the food
sector in developed countries and the issue of market access for developing countries. Noting
that the food sector in developed countries can be characterized as a successive oligopoly, we
explore the issue of market access where some developing countries can export high-value
processed goods and others export raw agricultural commodities. Tariffs apply on the imports of
both these goods such that we have the possibility of tariff escalation, an issue often highlighted
by trade and development economists as a barrier to developing countries wishing to up-grade
their export profile. We outline in the paper a theoretical framework in which we explore aspects
of market access when the domestic food sector is successively oligopolistic. Employing a
market access rule that market access concessions should not be discriminatory, we show that
reducing tariffs on each of these goods by equivalent amounts will result in biased market access.
Conversely, if market access concessions between developing countries are to be neutral, tariff
reductions should be differential and tariff escalation should increase. This is in contrast to
traditional views on the tariff escalation issue but arises due to the feedback effects tariff
reductions at different stages of the marketing chain have in this vertically related set-up. The
overall implication of the paper is that to provide appropriate analysis on traditional policy
issues, researchers should take into account the industrial organization of the markets on which
they focus.
Introduction
In recent years, there has been a considerable amount of research on market structure issues in
food and related markets, this literature being comprehensively summarized in Sexton and
Lavoie (2001). Research on the impact that market structure issues in the food sector have on
traditional policy analysis carried out by agricultural economists has, however, been relatively
thin which is particularly true if one were to focus on the international consequences of domestic
market structure. While those interested in market structure issues have largely focused, at least
implicitly, on domestic anti-trust implications, a corresponding criticism can be leveled at those
primarily interested in agricultural trade policy analysis, the models used here seldom departing
from the competitive benchmark. Taken together, agricultural economics research at the
interface between industrial organization and trade, or globalization issues more generally, has
largely been unexplored. Of course, this is not to say that such research has been non-existent,
but the predominant focus for agricultural economists interested in industrial organization has
been on domestic issues while agricultural trade economists have largely ignored issues relating
to domestic market structure.1
Yet issues at the interface of trade and industrial organization in food and agricultural
markets are potentially relevant both in relation to trends in agricultural trade and for
understanding current trade policy issues. For example, a large proportion of international trade
in food and agricultural products, particularly between developed countries, is in processed food
products thus implying the direct involvement of the food industry in trade. The food industry is
1
In the context of the mainstream economics literature, there is a substantial body of research on industrial
organisation and trade. For example, the international economics literature is replete with papers where oligopoly
theory has become widely used. Much of this research has focused, however, on aspects of strategic trade policy
and, more recently, on the impact on anti-dumping policies. Industrial organization economists have also begun to
deal with issues relating to the international consequences of domestic anti-trust policy.
1
also highly concentrated in many developed countries both at the retailing and processing levels,
with consolidation in the industrial food sector having increased in recent years. This
consolidation has also reflected aspects of globalization as a significant part of this consolidation
has taken the form of cross-border mergers and acquisitions. Taken together, the underlying data
would suggest increasing trade in processed food products involving an increasingly
concentrated food sector.
As far as current trade policy issues are concerned, the current Doha Round of trade
negotiations of the World Trade Organization (WTO) has been labeled the ‘development round’
a key aim of which is to increase developing countries’ access to developed country markets.
The distinction between raw commodity and processed food exports has also surfaced here with
development economists, and policy advisors more generally, advocating that developing
countries should diversify their exports away from reliance on raw agricultural exports, at least
in terms of ‘up-grading’ their export portfolio to export more processed or higher value food
products.2 In this regard, however, several barriers are perceived to exist. First, developing
country exporters often face the problem of tariff escalation where tariffs on the processed good
are higher than those on the raw commodity. In this context, the traditional trade policy literature
suggests that reducing tariffs on the raw commodity while keeping tariffs on the processed good
unchanged, or reduced by a lesser amount, will increase the level of effective protection and thus
act as a disincentive to developing countries exporting higher value processed food products.3
2
We use the term ‘policy advisors’ rather broadly to include organizations such as the World Bank and UNCTAD
as well as academics.
3
Effective protection is the extent to which value-added is protected. This will depend on the level of value-added
produced by an industry, i.e., the share of raw or intermediate inputs purchased from other sectors domestic and
foreign, as well as the level of tariffs on final and intermediate goods. Tariff escalation refers to the extent to which
the tariff on the final good exceeds the tariff on the raw or intermediate good.
2
Second, recent research on developing country issues, not necessarily confined to work done
by economists, has recognized additional barriers faced by exporters keen to access developed
country markets. Using a somewhat more eclectic ‘value chain’ framework, this paradigm
explicitly recognizes that issues relating to market power at various stages of the food chain, not
just levels of tariffs or the problem of tariff escalation, matter in addressing market access
opportunities for developing country exporters.4 However, while the ‘value chain’ approach is
useful for highlighting various trade and market structure issues facing developing country
exporters, it lacks a formal structure that industrial organization or trade economists would
readily recognize, nor does it give clear policy insights beyond somewhat vague pointers as to
what developing country exporters should perhaps do to access downstream food markets in
developed countries.
The objective of this paper is to explore tying some of these issues together. More explicitly,
we consider the issue of market access to developed country markets where food industry
structure is characterized by imperfect competition at both the retailing and processing stages,
i.e., we have successive oligopoly. Developing countries export raw commodities and processed
food products, though we assume below that these are different countries and where tariffs are
applied on both goods, i.e., the issues of tariff escalation and effective protection form a key
ingredient in this analysis. Setting the issue of market access, imperfect competition and
effective protection in a formal framework leads to important insights. First, in the context of a
model of successive oligopoly, an equal reduction in tariffs on the processed food product and
raw commodity are not equivalent in guaranteeing increases in market access for raw agricultural
and processed good exporters respectively. Second, the extent to which this is true depends on
4
See Kaplinsky (2000) for a good discussion of these issues.
3
the nature of competition in the developed country markets. Third, to the extent that the
processed food exporter and the raw commodity exporter are different countries, tariff reductions
that maintain the same level of effective protection are likely to be discriminatory in terms of
market access considerations. This is in contrast to the general perception of trade policy that
reducing tariffs by the same amount is necessary to avoid increasing the disincentives to
processed good exporters. If the primary focus is on market access, which we argue below
should be the appropriate focus of trade negotiations, varying the level of tariff reductions will be
necessary to avoid discrimination between developing country exporters. This in turn has a
political economy consideration: developing country negotiators should not focus solely on
market access commitments offered by developed countries, but also be aware of what each
developing country receives in terms of market access contingent on their export profile, if
negotiated market access outcomes are to avoid being discriminatory.
The paper is organized as follows. In section 1, we outline some trends in trade flows in raw
commodities and processed food products, and the role therein played by developing countries.
In section 2, we report some data relating to market structure of the food industry in developed
countries, focusing primarily on the US and European Union (EU). In section 3, we summarize
some recent concerns facing developing country exporters concerning the issues of tariffs and
tariff escalation. A more formal modeling approach is outlined in section 4 which is used to
explore the issues of tariff concessions and market access when the importing country’s food
industry structure is characterized by successive oligopoly. Key results that arise from this
theoretical model are presented in section 5. In section 6, we summarize and conclude and
consider some avenues for future research on this issue.
4
1.
Trade in Food and Agricultural Products
In this section, we provide some data on exports of food and agricultural products by developing
countries, with particular emphasis on their exports to developed countries, by way of
background to the issues relating to developing country access to developed country markets. In
general, the key characteristic of trade in this sector, defined as all food items in the 3-digit SITC
categories 0, 1, 22, and 4, is that developing countries account for a considerably smaller share of
the value of world exports of food and agricultural products as compared to developed countries.
As is shown in Table 1, on average, the developing countries accounted for 31 percent of the
value of world exports of all food items over the period 1990-2000. In contrast, over the same
period, the developed countries accounted for 66 percent, on average, of the value of world
exports of all food items.
With respect to trade in specific food and agricultural products, a very clear pattern also
emerges. As shown in Table 2, the developing countries have significant shares of the value of
world exports for those raw commodities in which they have a traditional comparative
advantage, averaging 65, 57, 77, 66, 85, and 78 percent respectively for rice (042), sugar (061),
coffee (071), cocoa (072), tea (074) and spices (075) over the period 1990-2000. In contrast, the
developed countries dominate the share of the value of world exports in high-value and
processed food and agricultural products. The obvious exceptions to this are fresh and frozen
shellfish (036), prepared and frozen fish (037), and other fixed vegetable oils (424), where the
developing countries on average accounted for 66, 54, and 82 percent respectively of the value of
world exports over this period. In addition, as shown in Table 3, the developed countries also
typically dominate the share of the value of world imports in high-value and processed food and
agricultural products, whereas the developing countries are major importers of non-processed
5
grains such as wheat (041), rice (042), barley (043), maize (044), and wheat flour (046) and other
cereal meals and flour (047), averaging 61, 67, 46, 48, 75, and 60 percent respectively of the
value of world imports over the period 1990-2000.
In summary, this overview of trade in food and agricultural products shows that, in
aggregate it is dominated by the developed countries. In addition, when trade is disaggregated to
specific product groups, the developed countries account for most of the value of world trade,
both exports and imports, in high-value and processed food and agricultural products, whereas
the developing countries account for a large proportion of the value of exports in traditional
unprocessed commodities such as rice, sugar, coffee, cocoa and tea, most of which have
continued to suffer from declining world prices over this period (UNCTAD, 2000). It is against
this background, that economists and policy advisors have advocated that developing countries
upgrade their exports into the more highly processed food sectors.
2.
Market Structure of Developed Country Food Processing and Retailing
Having noted the relatively low market share in developed countries of developing country
exports of food and agricultural products, we now focus on the extent to which food processing
and retailing in developed countries is concentrated, with specific focus on these sectors in the
United States and the EU. This sets the background for some of the wider concerns relating to
market access by developing countries and the theoretical model presented below.
(i)
Food Processing
The food-processing sector in both the United States and the EU is highly concentrated
(Cotterill, 1999). In the United States, a small number of large firms dominate the sector, with
the top-20 food and tobacco-manufacturing firms accounting for over 52 percent of the sector’s
6
value added in 1995.
If food manufacturing is separated from beverage and tobacco
manufacturing, the top-20 food manufacturing firms accounted for 37 percent of value added in
1997, while the top-20 beverage and tobacco-manufacturing firms accounted for 79 percent of
value added (US Census Bureau, 2001). In Table 4, we list those specific food products where
the 4-firm concentration ratio was over 60 percent in 1997, the average being just below 76
percent.
Turning to food manufacturing in the EU, the data in Table 5 show that typically at the
country level, average seller concentration is higher than in the United States, ranging from an
average 3-firm concentration ratio of 55 percent in Germany to 89 percent in Ireland, with an
average 3-firm concentration ratio across 9 EU countries of 67 percent. As in the United States,
these averages hide some high levels of seller concentration for specific products in each EU
country, most notably baby foods, canned soup, pet food and coffee.
It should be noted,
however, that while seller concentration at the product level is high in many individual EU
country markets, there are few examples of firms that dominate sales across EU countries as a
whole (Cotterill, op. cit.).
(ii)
Food Retailing
In the case of food retailing, there are quite important differences between market structure in the
United States and that in the EU. As Table 6 reports, 5-firm seller concentration in food retailing
at the national level is much higher in EU countries than it is in the US, with average 5-firm
seller concentration in the former being 65 percent, compared to 35 percent in the latter.
However, at the EU-wide level, 5-firm seller concentration is much lower at 26 percent (Hughes,
2002). In addition, in the US, it is important to examine concentration in food retailing at the
local and regional level. Cotterill (op. cit.) reports that in 1998, 4-firm seller concentration
7
averaged 74 percent across the top 100 US cities, while across major US regions 4-firm seller
concentration averaged 58 percent.
As well as the high levels of concentration in US and EU food retailing, it is important to
recognize, that several firms in this industry, which were previously national in origin, are now
becoming international in scope. Hughes (op. cit.) reports that in the 1980s, food retailers in the
EU, such as the French firm Carrefour, began expanding beyond their national base, while the
US-based firm Wal-Mart expanded into Canada and Mexico. This phenomenon continued in the
1990s, with EU-based retailers such as Royal Ahold and Sainsbury expanding into the US
market (Cotterill, op.cit.), Carrefour and Royal Ahold expanding into various developing country
markets in Central and Latin America (Chavez, 2002; Farina, 2002; Gutman, 2002), and USbased Wal-Mart expanding into the EU (Hughes) and into Central and Latin America (Chavez,
op. cit.; Farina, op. cit.).
In summary, the food manufacturing and retailing sectors in the United States and EU are
highly concentrated, such that the vertical structure of the food marketing system in developed
countries can appropriately be characterized as one of successive oligopoly. This of course
means that the vertical structure of the food chain as a whole matters in addressing aspects of
market access for exporting countries.5
3.
Access by Developing Countries to Developed Country Markets
In this section we consider some of the issues involved when developing country exporters are
faced with the problem of market access in the context of a vertically related market. We
5
One issue that we do not discuss here relates to the nature of vertical ties between each stage of the vertical food
chain. The model presented below assumes arm’s length pricing whereby it is the horizontal dimension of
competition at each stage that matters. Some discussion of the relevance of vertical contractual arrangements can be
found in Sheldon and McCorriston (2003).
8
consider two issues: first, the levels of tariffs and the problem of tariff peaks facing developing
country exporters; and, second, the problem of tariff escalation.
(i)
Tariffs and Tariff Peaks
The traditional focus of trade models is on explicit trade barriers such as tariffs and quantitative
restraints. Following the Uruguay Round of GATT, there was a process of tariffication whereby
a range of non-tariff barriers including quantitative restraints were converted into tariff
equivalents. A first step, therefore, in considering market access issues facing developing
countries is to focus on the level of the tariffs they face when exporting to developed country
markets.
A recent survey by USDA (2001) indicates that on average, world tariffs in the food and
agricultural sector stand at 62 percent for bound, most favored nation (MFN) rates. However,
average food and agricultural tariffs for WTO members by region vary from 25 percent for North
America and 30 percent for the EU to 113 percent for South Asia. This compares with average
developed country MFN tariffs of 5 percent across all sectors (Hoekman, Ng, and Olarreaga,
2002). It should also be noted that tariff levels in developing countries are also high, in many
cases higher than those in developed countries, such that developing country access to other
developing countries, may actually be more restricted than to developed countries.
While the level of average tariffs is in some way informative, it should be noted that
products in this sector are often characterized by tariff peaks in the developed countries, i.e.,
where the tariffs on some imported products far exceed the average level. For example, as
reported by Hoekman, Ng and Olarreaga (op. cit.), the United States, the EU, Japan and Canada
have respectively 48, 290, 178, and 85 tariff peaks for food and agricultural products, with the
maximum tariff rates being on butter (Canada, 340 percent), edible bovine offal (EU, 250
9
percent), raw cane sugar (Japan, 170 percent) and peanuts in the shell (US, 120 percent). As
noted by Hoekman, Ng and Olarreaga, many of these tariff peak products are of interest to
developing country exporters such that market access may be more limited than what the average
tariff rates imply.
However, an additional aspect of market access relates to preferential treatment. Many
developed countries also provide limited preferential access for food and agricultural products
under both the Generalized System of Preferences (GSP), and reciprocal trade agreements such
as NAFTA. For example, the EU has a myriad of preferential access agreements that cover a
large number of developing countries with some developing countries being more favored than
others. In Table 7, average MFN tariffs by Harmonized System 2-digit food and agricultural
products are listed for the EU and the United States, along with margins for less developed
country (LDC) preferences. Across all products, the average MFN tariff on food and agricultural
products is 38.3 percent in the EU, and 30 percent in the United States, with developing
countries getting preferences that imply on average they pay duty up to 50 percent of the MFN
tariff. However, there are some clear tariff peaks in the EU for products such as meat and edible
offal (02), cereals (10), and oil seeds (12) where preferential access for the developing countries
is small, and likewise in the United States for oil seeds.
In conjunction with the trade data, the tariff data suggest that developing country access to
developed country markets for food and agricultural products could be improved through trade
liberalization, particularly in the case of products that exhibit tariff peaks in developed countries
and limited preferential access beyond MFN tariffs. This, however, is not the only consideration.
10
(ii)
Tariff Escalation
For developing countries attempting to diversify and up-grade their exports from raw agricultural
commodities to processed food products, one of the most often-mentioned difficulties is that of
tariff-escalation. Tariff escalation occurs when tariffs on imports of processed goods are higher
than the tariffs on the corresponding raw commodity. This issue has been well-known from the
work of Balassa (1965) and Corden (1971). UNCTAD (2002) has recently cited this issue as one
of the main problems facing developing country exporters in diversifying their export profile.
The recent evidence on the extent of tariff escalation is rather mixed. For many agricultural
commodities supported by government intervention in the developed countries, the tariff on the
raw commodity is often exceptionally high. For example, USDA (op. cit.) report higher levels of
tariffs on grain compared to grain products in several developed countries including the United
States and the EU. Nevertheless, tariff escalation is still perceived to be a major issue facing
developing country exporters. In Table 8 we report the highest post-Uruguay Round tariff
escalation estimates for a series of commodities for the US, Japan and the EU. The estimates
show high levels of tariff escalation across all three countries. The table also highlights that the
level of tariff escalation has decreased following the Uruguay Round with some of the
commodity groups facing the highest levels of tariff escalation also being the ones exhibiting the
highest levels of reduction.
UNCTAD (2002) also reports high levels of tariff escalation for products exported solely
from developing countries. For example, for coffee, tea and spices, the level of tariff escalation
in Japan and the EU rose from averages of 0.11 per cent and 1.63 per cent for raw material
imports in these two countries respectively to 8 per cent and 20 per cent in the case of the final
product. Taken together, and in spite of the decline in tariff escalation following the Uruguay
11
Round, tariff escalation remains a problem for developing countries diversifying their exports
and attaining market access for the processed good.
4.
(i)
Theoretical Framework
Schematic Outline
It is perhaps useful to outline the modeling framework by representing the scenario we are
addressing in Figure 1. In this market there are two domestic firms at the retail stage, and two
domestic firms at the processing stage, i.e., there is successive oligopoly. The two firms at each
stage do, however, differ in terms of where they buy their inputs. At the processing stage, firm 1
buys inputs from the domestic agricultural sector while firm 2 buys the inputs of the raw
commodity from a foreign supplier. Subsequently, these two firms compete and sell the
processed good to the firms in the downstream retail sector. In the downstream retail sector,
there are again two firms that compete at this stage, the distinction between the two firms being
from where they source their inputs. Firm 1 buys from the domestic upstream stage while firm 2
sources its inputs from the world market. The retail firms may or may not add value and may be
involved solely in distribution. Since this stage is closest to the consumer, we assume that the
import by firm 2 is of a relatively high-value, processed good. At each stage, firms take the
input prices, whether sourced from the world or domestic market, as given.
As far as the supply of the imported good is concerned, we assume that the suppliers of the
raw commodity and the higher value good are different developing countries, say due to past
investment in the food sector or lack of it. This avoids detailing any specific strategy of what
combination of goods to produce. However, market access at each stage is affected by tariffs,
with a tariff imposed on the processed import and on the raw commodity import. Thus tariff
12
escalation and effective protection can be characterized in this model. If the tariff on the
processed good is higher than on the raw commodity, there is tariff escalation and the level of
effective protection afforded to the downstream firm 1 exceeds the level of nominal protection.
Taken together, the set-up of the model as presented in Figure 1 captures most of the issues
referred to in sections 2 and 3 of the paper. First, developing countries may export raw
commodities or more highly processed products but face the problem of tariffs imposed at each
stage and where tariff escalation may persist. Corresponding to the characteristics of the food
sector in the developed countries, we have imperfect competition at each stage of the vertically
linked food chain. Hence, we have successive oligopoly. Finally, we have a framework more
directly appropriate for considering the impact of trade reform than standard trade models
employ as the model explicitly recognizes the relevance of market power and the role of tariffs
on raw commodities and processed goods in determining market access at each stage of the food
chain. With this model, we can ask questions about ensuring equivalent market access for raw
commodities and processed goods if trade reform leads to commitments on tariff reductions
particularly on products of interest to developing country exporters.
(ii) The Model
Assumptions
As noted above, we have a model of successive oligopoly, i.e., both the upstream (processing)
and downstream (retailing) stages are imperfectly competitive. We assume that the technology
linking each stage is one of fixed proportions. Formally, x1=φxU, where x1 and xU represent
output of downstream firm 1 and the upstream stages respectively, and where φ is the constant
coefficient of production. To ease the exposition, φ is set equal to one in the framework outlined
below.
13
In terms of the game-theoretic structure of the model, the timing of the firm’s strategy
choice goes from upstream to downstream. We outline the model in general strategy space since
we wish to capture the nature of competition in influencing the market access outcomes.
Specifically, given costs and the derived demand curve facing the upstream sector, upstream
firms simultaneously choose their strategy to maximize profits, which generates Nash
equilibrium at the upstream stage.
The processed good prices are taken as given by the
downstream firms that, simultaneously chooses their strategy to maximize profits, thus giving
Nash equilibrium at the downstream stage. In terms of solving the model, equilibrium at the
downstream stage is derived first and then the upstream stage.
Equilibrium in the Downstream Market
The model is written in general form following Bulow, Geanakopolos and Klemperer (1985). s1
is the strategy choice of the firm buying from the domestic upstream stage and s2 is the strategy
choice of the firm importing from the world market. The revenue functions can be written as:
R1 ( s1 , s 2 )
(1)
R2 ( s1 , s 2 ) .
(2)
We assume downward sloping demands and substitute goods.
Given (1) and (2), the relevant profit functions are given as:
π 1 = R1 ( s1 , s 2 ) - c1 s1
(3)
p
π 2 = R 2 ( s1 , s 2) - c2 s 2 − t s 2 ,
(4)
where c1 and c2 are the downstream firms’ respective costs. Firms’ costs relate to the purchase
of the intermediate input. Note that firm 2 also faces a per unit tariff on the processed import
which is given by t p .
14
The first-order conditions for profit maximization are given as:
(5)
R1,1 = c1
R 2, 2 = c 2 + t
p
,
(6)
Equilibrium in the downstream stage can be derived by totally differentiating the first-order
conditions (5) and (6):
 R1,11 R1,12   ds1  dc1 
.

 
 =
p
 R 2, 21 R 2 , 22   ds 2  dc 2 + dt 
(7)
The slopes of the reaction functions are found by implicitly differentiating the firms’ firstorder conditions:
R1,12
ds1
= r1 = R1,11
ds 2
(8)
R2, 21
ds 2
= r2 = .
R2, 22
ds1
(9)
With this set-up, we can deal with both strategic substitutes and strategic complements
where the variable of interest is the cross-partial effect on marginal profitability, i.e., sign ri =
sign Ri,ij. Consequently, with reference to equation (8) and (9), if Ri,ij <0, then ri < 0. In this case,
we have the case of strategic substitutes, and the reaction functions are downward sloping.
However, if Ri,ij > 0, the reaction functions are upward sloping and we have strategic
complements. The distinction between strategic substitutes/complements relates to the
“aggressiveness” of firm’s strategies. With strategic substitutes, firms’ strategies are less
aggressive than those associated with strategic complements. Following Bulow et al. (ibid.),
15
assume that the strategy space relates to quantities. With strategic substitutes (complements), an
increase in the output of firm 1 would be met by a decrease (increase) in that of firm 2.
Given (7), the solution to the system is found by re-arranging in terms of dsi and inverting
where ∆ is the determinant of the left-hand side of (7):
 R 2 , 22 - R1,12 
 ds1
-1

 = ∆ 
ds 2 
 - R 2 , 21 R1,11 
 dc1 
dc + dt p  .
 2

(10)
To simplify the notation re-write (10) as:
 ds1
a 2
-1
 = ∆ 
ds 2 
 b2
b1   dc1 
,
a1  dc 2 + dt p 
(11)
where,
a1 = R1,11 a 2 = R 2, 22
b1 = R1,12 b2 = R 2, 21 .
For stability of the duopoly equilibrium, the diagonal of the matrix has to be negative, i.e., ai
< 0, and the determinant positive, i.e., ∆ = (a 1 a 2 - b 1 b 2 ) > 0. Given these conditions, further
comments can be made about the reaction functions. ri = -(bi)/ai from (8) and (9). If ai < 0, then
for strategic substitutes, b i < 0, in order to satisfy r i < 0, and b i > 0 in order to satisfy ri > 0 for
strategic complements. The expression for ri can be substituted into (11) in order to make the
comparative statics easier to follow:
 a 2 a1 r1  dc1 
 ds1
-1
.
 
 = ∆ 
p
 a 2 r 2 a1 dc 2 + dt 
ds 2 
(12)
Equilibrium in the Upstream Market
Given the fixed proportions technology and φ = 1, total output in the upstream sector is given by
xU(= x1), where the combined output of the two upstream firms x1U + x U2 = x U . The processed
16
good is assumed to be homogeneous so that downstream firm 1 is indifferent about the relative
proportions of x1U and xU2 used at retail. Assuming that the downstream firm faces no costs other
than the price paid for the processed good, the inverse derived demand function facing firms in
the upstream sector can be found by substituting piU for ci in (5) and (6) where superscript
U
denotes the upstream sector. Firms’ profits in the upstream sector are, therefore, given by:
U
U
π1 = R1U ( s1 , s 2) - c1 s1U
(13)
r
πU2 = RU2 ( s 2 , s 2) - cU2 s U2 − t s 2 ,
(14)
where c1U and cU2 are the upstream firms’ costs respectively. Again note that the firm that imports
the raw commodity from the world market also faces the cost associated with the tariff.
Given this, following the outline above, equilibrium in the domestic upstream market is
given by:
 a 2U
ds1U 
U −1
 U  = (∆ )  U U
a 2 r2
ds 2 
5.
a1U r1U   dc1U 
.

a1U  dc UB + dt r 
(15)
Market Access with Successive Oligopoly
In terms of thinking about the issues facing developing country exporters of raw commodities
and processed goods when the home market is characterized by successive oligopoly, it will be
useful to focus more directly on market access issues. As Bagwell and Staiger (1999) note,
GATT/WTO rules are fundamentally about market access commitments. What comes out of the
model employed here is that in the context of successive oligopoly, reducing tariffs on the
processed good and raw commodity by equivalent amounts may hinder market access for one of
the developing country exporters. Consequently if market access is the principal focus of the
17
Doha Round of trade negotiations, to avoid discriminatory market access in favor of one
developing country at the expense of another, equivalent market access commitments for each
exporter can only be sustained by reducing the tariff on each import by differential amounts. As
we show below, there are several factors that determine the appropriate reductions in tariffs
including the incidence of upstream tariffs on the downstream sector, the strength of the ‘backshifting effect’ as tariff reductions on the processed good have an impact on the derived demand
facing the upstream sector and, hence, the demand for the imported raw commodity. The nature
of competition between firms at each stage of the vertically linked food chain also matters.
Consider the following hypothetical scenario. Trade negotiators in the Doha Round are
intent on increasing developing countries access to developed country markets. But they
recognize that some developing countries export more highly processed goods while others
export raw commodities. Reducing tariffs is the principal way of encouraging market access. But
they also recognize that when each export enters the importing country’s food chain, reducing
tariffs at one stage has an impact on imports at the corresponding upstream or downstream stage.
They may also know that the nature of competition at each stage influences the outcome.
Consequently, if they were to reduce tariffs on the raw commodity keeping the tariff on the
processed good unchanged, imports of the processed good will fall as the competitiveness of the
downstream firm that purchases its inputs from the domestic sector has now improved. Hence
market access would be biased in favor of the raw commodity exporter at the expense of lower
imports from the processed good exporter. Similarly, if the trade negotiations lowered the tariff
on the processed good keeping tariffs on the raw commodity import unchanged, this would
increase imports of the processed good at the expense of imports of the raw commodity. This
arises because the increase in imports of the processed good reduces the sales of firm 1 in the
18
downstream sector, which shifts back the derived demand in the upstream sector, thereby
reducing imports of the raw commodity. Potentially therefore, developing countries have
conflicting interests depending on the type of good they export to the developed country market.
To avoid such conflict, therefore, the trade negotiators consider a market access rule to
avoid any conflicting interest between these developing countries. If tariffs on imports from
developing countries are to avoid any discrimination in market access, then the net change in
market access following tariff reductions on raw and processed goods should be unchanged.
Specifically:
∆MA r (dt r ) + ∆MA p (dt p ) = 0 ,
(16)
where ∆MA r is the change in market access of the raw commodity import which depends directly
on, the change in the raw commodity tariff (dt r ) ,and where ∆MA p is the change in the market
access of the processed good which depends directly on the change in the tariff on the processed
good (dt p ) . Since we noted above these tariff changes can have offsetting effects in the
corresponding upstream or downstream sector, we set the net change that the trade negotiator
would aim for, in order to avoid any discriminatory effects, equal to zero. Of course, one may
argue that changing tariffs on raw commodities and processed goods by equal amounts would
result in an equivalent increase in market access. As we show below, in the context of successive
oligopoly, this will not be the case. In turn, if the focus is on market access considerations
without discriminating between developing country suppliers, the reduction in tariffs for imports
at each stage will not be equal.
We amend the trade negotiator’s rule as given above. We set it such that we consider what
would be the appropriate reduction in the tariff on the processed good ( APT ) that would offset
the reduction in the tariff on the raw commodity ( dt r ) while keeping the change in market access
19
the same at both stages. As the market access rule relates directly to import volumes, we assume
that the strategy space is in quantities. Replacing s i with xi in the upstream and downstream
industries respectively, we can re-write our market access rule as6:
[(dx 2 / dc1 )(dc1 / dt r )]dt r + APT (dx 2 / dt p ) = 0 .
(17)
The first argument is the effect of the tariff on the raw import on imports in the downstream
market: clearly the change in the raw commodity tariff ( dt r ) changes upstream prices, which
then affects competition in the downstream sector.
By changing the competitiveness of
downstream firm 1, this affects the level of imports of processed goods by downstream firm 2,
x2 . The second argument is the effect of the tariff on imports of the processed good. Setting the
rule equal to zero and solving for APT, gives the appropriate amendment in the processed good
tariff to keep the change in market access in both the upstream and downstream markets the
same for a given change in the tariff on the raw agricultural import. Re-arranging (17) we have:
APT =
− (dx 2 / dc1 )(dc1 / dt r )dt r
.
(dx 2 / dt p )
(18)
However, it should be noted that while the upstream tariff has an impact on the downstream
market via the change in the downstream firm’s costs, the change in the processed good tariff
will also have an effect on the upstream price via the ‘back-shifting’ effect. Therefore, we need
to expand the denominator to give:
APT =
− (dx 2 / dc1 )(dc1 / dt r )dt r
.
[(dx 2 / dt p ) + (dx 2 / dc1 )(dc1 / dt p )]
(19)
6
It should be noted that the distinction between strategic substitutes and complements can still arise in quantity
space with strategic complements being an outcome if the demand function is sufficiently convex.
20
Suppose for the moment we ignore the ‘back-shifting’ effect. Note that to keep market access the
same we should not necessarily reduce the processed tariff by the same amount as the raw tariff.
This is due to the possibility that the incidence of the upstream tariff on upstream prices, i.e., the
downstream firm’s costs may change by less than 100 percent as given by the term (dc1 / dt r ) .
The direct effect of the processed good tariff affects imports of the processed good but this is
offset by the ‘back-shifting’ effect. The ‘back-shifting’ effect reduces upstream prices, because
the inverse derived demand function shifts back. In turn, this reduces the downstream firm’s
costs, which in turn reduces imports. So while the reduction in the processed good tariff
increases processed good imports, the ‘back-shifting’ effect serves to ameliorate the effect to
some degree. So the amendment in the processed good tariff may have to be greater to offset this
‘back-shifting’ effect as long as the incidence is less than the direct effect on processed good
imports.
Expression (19) can be re-written to derive explicit results. Specifically, we have7:
APT =
− ∆−1 a 2 r2 (dc1 / dt r )
dt r ,
p
−1
−1
∆ a1 + ∆ a 2 r2 (dc1 / dt )
(20)
which can be simplified to (assuming a1 ≈ a 2 ):
APT =
− r2 (dc1 / dt r ) r
dt .
1 + r2 (dc1 / dt p )
(21)
Result 1: If we have strategic substitutes, a reduction in the tariff on the raw agricultural import
should likely be matched with a reduction in the tariff on the processed good. If we have
strategic complements, a reduction in the raw agricultural good tariff should be matched with an
increase in the tariff on the processed good.
7
To conveniently separate the effects, we do not write out the tariff pass-through and the back-shifting effects in
explicit form.
21
The sign of r2 is the key to this result. With strategic substitutes, r2 < 0 , so the numerator
is positive. Also the denominator is positive as long as r2 {dc1 / dt p } < 1 which will likely be the
case under most reasonable circumstances.8 Therefore, a tariff reduction at the upstream stage
should be matched by a tariff reduction at the downstream stage. In the strategic complements
case, the tariff on the processed good imports should be raised if market access is to stay the
same as the numerator will be negative. This may seem an unlikely result, but it can be
explained. In this multi-market set-up, the tariff reduction in one market has an effect on the
related market. In the strategic complements case, competition is pretty aggressive such that
reducing tariffs on raw commodity imports would also increase processed good imports. Given
our rule aimed at keeping market access constant for both types of goods, this suggests
increasing tariffs on processed good imports to offset the pro-competitive externality associated
with the tariff reduction in the upstream stage. Note that this did not arise in the case of strategic
substitutes as the appropriate APT was to match a reduction in a tariff on the raw commodity
with a reduction in the tariff on the processed good as the multi-market effect here was negative.
In the strategic complements case, this multi-market effect is positive thus changing the sign of
the APT .
Result 2: In the strategic substitute case, the reduction in tariffs on processed good imports
should likely be less than the reduction in tariffs on raw imports. This is because the absolute
value of r2 is less than 1. The denominator will also be less than 1 under most reasonable
circumstances but as long as r2 (dc1 / dt p ) is not ‘too large’, the APT will be less than dt r . In the
strategic complement case, the increase in the tariff on the processed import should be less than
the reduction in the tariff on the raw good.
Note closely what this result implies. Consider first the strategic substitute case, which is the
more intuitive result. What it implies is that to avoid any discriminatory effect between
8
If there was ‘over-shifting’ of the back-shifting effect and r2 was sufficiently large, then this would reverse the
result for the strategic substitutes case.
22
developing countries, which export food and agricultural products to different stages of the food
chain, tariffs on raw commodities and processed goods should be reduced by different amounts.
But note that the tariff reduction on the processed good should be reduced by less than the tariff
on the raw commodity. In the context of the traditional trade policy literature, tariff escalation
should therefore increase to avoid discriminatory effects in terms of market access. In the
context of successive oligopoly, with strategic substitutes, tariff escalation is per se not a bad
thing as long as both tariffs are reduced. However, in the strategic complements case, tariff
escalation should also be an outcome, this time with the reduction on the tariff on the raw
commodity matched by an increase in the tariff in the processed good.
5.
Summary and Conclusions
The aim of this paper has been to consider the implications of the industrial organization of the
food sector in developed countries for market access for developing country exporters. To do so,
we have attempted to bring together some discordant themes in the literature into a unified
framework. Specifically, we have explicitly accounted for the fact that the food sector in
developed countries is more appropriately characterized by successive oligopoly. We have also
noted that in the context of on-going trade negotiations, that market access considerations for
developing country exporters are high on the trade negotiating agenda. In addition, we have also
endeavored to accommodate the fact that developing countries export raw and processed goods
and that they are constantly being encouraged to up-grade their export portfolio and concentrate
less on exports of raw commodities. In this context, we have also accommodated the possibility
of tariff escalation in this framework.
23
Based on a (hypothetical) market access rule aimed at avoiding discriminatory market
access between different developing country exporters, we have generated the following results:
(i) changes in tariffs at different stages not only affect tariffs directly at the stage at which they
are applied but also affect the imports at the preceding or subsequent stage; (ii) even if tariffs on
imported raw commodities and processed goods are reduced by the same amount such that the
measure of effective protection stays the same, the impact on market access on imported raw
commodities and processed goods will differ; (iii) the extent to which the impact of tariff
reductions at each stage differ will depend on the nature of competition, specifically whether we
have the strategic substitute or strategic complements case.
Taken together, the results suggest that tariff escalation may be a desirable outcome if
discrimination between developing country exporters is not to arise. This result seems at odds
with the traditional literature and observations made by commentators of the trade negotiating
process. But the result here comes from the fact that we have explicitly accounted for the
interaction of tariffs at different stages in the context of a successive oligopoly where policy
changes at one stage will impact on the market equilibrium at another. The overall moral of this
story is that appropriate consideration of the industrial organization of food markets should be
made when considering the outcome of policy reforms. Simple benchmarks that ignore market
structure issues may lead to misguided policy outcomes.
Of course, the model presented here is itself, simplified and was employed to highlight a
specific issue. There are clear possible extensions in this regard. We mention two in the context
of the model we have presented. First, in our model of successive oligopoly, competition occurs
‘horizontally’ and we make no allowance for the range of vertical contracts that link the two
stages together. These cover various forms of vertical restraints through to vertical integration.
24
As is well known to industrial organization economists, these vertical contracts can have pro- or
anti-competitive effects and thus, may affect the outcomes presented above. Second, the
importers of the raw commodity and processed goods took world prices for these imports as
given. There are two additional issues worth exploring here. For example, the foreign importing
firm may vertically integrate with the developing country supplier, an issue highlighted by the
value-chain approach. Alternatively, the importing firm may have some degree of monopsony
power such that terms of trade effects may arise. Each of these may have an impact on
determining market access considerations in this set-up. No doubt there are other possibilities,
our view here is that exploring the interface between industrial organization and traditional
policy issues, in our case here, trade policy will be an area of research deserving further
attention.
Finally, the analysis presented here was on the premise of a market access rule consistent
with the recent work of Bagwell and Staiger (op. cit.), specifically that trade negotiations are
primarily about market access issues. As set up, the rule was aimed at avoiding discriminatory
market access for developing countries in the context of trade negotiations. However, market
access is often discriminatory in nature reflecting either the bargaining power of certain countries
in trade negotiations or even the nature of current trade arrangements. For example, as shown in
Table 7, preferential access to developed country markets varies by commodity group. To the
extent that the range of preferences varies by country, some developing countries will have more
favored market access than others. As such, considering variations to the non-discrimination rule
may be of interest in future research, even if confined to fully analyzing the impact of
preferential trade arrangements in the context of successively oligopolistic markets.
25
References
Bagwell, K. and R. Staiger “An Economic Theory of GATT Think,” American Economic
Review 89 (1999): 779-795.
Balassa, B. “Tariff Protection in Industrial Countries,” Journal of Political Economy
73(1965): 573-594.
Bulow, J.I., J.D. Geanakopolos, and P.D. Klemperer (1985) “Multimarket Oligopoly:
Strategic Substitutes and Complements,” Journal of Political Economy 93, 488-511.
Corden, M. The Theory of Protection, Oxford: Clarendon Press, 1971.
Cotterill, R.W. Continuing Concentration in Food Industries Globally: Strategic Challenges
to an Unstable Status Quo, University of Connecticut, Food Marketing Policy Center,
Research Report No. 49, 1999.
Chavez, M. “The Transformation of Mexican Retailing with NAFTA,” Development Policy
Review 20(2002): 503-513.
Farina, E.M.M.Q. “Consolidation, Multinationalization, and Competition in Brazil: Impacts
on Horticulture and Dairy Products Systems,” Development Policy Review 20(2002):
441-457.
Gutman, G.E. “Impact of the Rapid Rise of Supermarkets on Dairy Products Systems in
Argentina,” Development Policy Review 20(2002): 409-427.
Hughes, D. “Grocery Retailing in Europe and Emerging Routes to the Consumer,”
EuroChoices 1(2002): 12-16.
Hoekman, B, Ng, F. and Olarreaga, M. “Eliminating Excess Tariffs on Exports of Least
Developed Countries,” The World Bank Economic Review 16(2002): 1-21.
Kaplinsky, R. “Spreading the Gains from Globalization: What Can be Learned from Value
Chain Analysis?” Journal of Development Studies 37(2000): 117-146.
Lindland, J. “The Impact of the Uruguay Round on Tariff Escalation in Agricultural
Products,” Food Policy, 22(1997): 487-500.
McCorriston, S. “Why Should Imperfect Competition Matter to Agricultural Economists?”
European Review of Agricultural Economics 29 (2002): 349-372.
Sexton, R. J.and N. Lavoie “Food Processing and Distribution: An Industrial Organization
Approach.” Handbook in Agricultural Economics, Volume 1B, B. Gardner and G.
Rausser, eds. Amsterdam: Elsevier North Holland, 2001.
Sheldon, I.M. and S. McCorriston ‘Globalization, Food Industry Consolidation and Market
Access: Implications for Developing Countries. Paper presented at conference on
‘Globalisation, Agriculture and Rural Livelihoods’ Cornell University, April 2003.
UNCTAD Strategies for Diversification and Adding Value to Food Exports: A Value Chain
Perspective,
United
Nations,
Geneva,
UNCTAD/DITC/COM/TM/1;
UNCTAD/ITE/MISC.23, 2000.
UNCTAD Export Diversification, Market Access and Competitiveness. United Nations,
Geneva, UNCTAD/TD/B/COM.1/54, 2002.
UNCTAD Handbook of Statistics, Geneva: United Nations, 2003.
USDA. Profiles of Tariffs in Global Agricultural Markets, AER-796, Washington D.C.:
Economic Research Service/USDA, 2001.
US Census Bureau Concentration Ratios in Manufacturing, Washington D.C.: US
Department of Commerce, US Census Bureau, 2001.
Table 1: Share of World Exports of All Food Items1, 1990-2000 (%)
Country Grouping
1990
2000
Developing
29.2
32.9
Developed
67.7
64.7
EU
44.7
40.7
United States
12.9
12.5
Source: UNCTAD (2003).
1
Defined as SITC codes 0+1+22+4
Table 2: Share of World Exports by SITC, 1990-2000 (%)
SITC
001 Live animals for food
011 Meat, fresh, chilled, frozen
012 Meat dried, salted, smoked
014 Meat prepd, prsrvd nes, etc
022 Milk and cream
023 Butter
024 Cheese and curd
025 Eggs, yolks, fresh, prsrvd
034 Fish, fresh, chilled, frozen
035 Fish salted, dried, smoked
036 Shell fish fresh, frozen
037 Fish etc prepd, prsrvd nes
041 Wheat etc, unmilled
042 Rice
043 Barley, unmilled
044 Maize (corn), unmilled
045 Cereals nes, unmilled
046 Wheat etc, meal or flour
047 Other cereal meals, flour
048 Cereal etc preparations
054 Vegtb etc fresh, simply prsrvd
056 Vegtb etc prsrvd, preprd
057 Fruit, nuts, fresh, dried
058 Fruit prsrvd, preprd
061 Sugar and honey
062 Sugar preps non-chocolate
071 Coffee and substitutes
072 Cocoa
073 Chocolate and products
074 Tea and mate
075 Spices
081 Feeding stuff for animals
091 Margarine and shortening
098 Edible products, preps nes
111 Non alcoholic beverages nes
112 Alcoholic beverages
121 Tobacco, unmanufactd, refuse
122 Tobacco, manufactured
222 Seeds for soft fixed oils
223 Seeds for other fixed oils
411 Animal oils and fats
423 Fixed vegetable oils, soft
424 Other fixed vegetable oils
431 Procesd animl and veg oil, etc
Source: UNCTAD (2003).
Developed Countries
1990
2000
57.17
79.03
83.59
81.29
97.40
98.07
69.94
65.37
93.58
85.90
94.09
92.79
97.39
94.38
83.64
79.08
68.36
57.45
84.74
77.82
31.59
32.19
44.91
37.18
91.21
83.16
40.55
28.24
98.24
92.54
87.13
71.45
81.23
87.63
85.22
68.30
70.84
63.82
90.17
82.69
67.47
67.44
62.66
59.16
60.03
56.12
54.82
55.31
38.16
41.40
82.73
67.01
19.95
23.99
25.56
36.30
94.50
86.47
15.26
14.22
21.68
18.85
60.47
61.48
73.11
61.99
84.06
77.25
80.54
76.81
92.54
85.28
47.14
40.93
86.62
77.63
62.54
60.97
42.95
55.34
69.39
84.92
63.12
58.00
17.70
16.97
54.82
54.35
Developing Countries
1990
2000
18.57
12.86
1.82
22.01
3.54
1.89
1.44
11.97
28.49
14.90
65.72
47.62
7.73
57.51
0.86
11.42
14.95
11.63
27.66
8.58
29.87
35.29
39.31
41.45
59.76
14.45
79.91
71.51
4.95
84.48
76.31
37.94
15.74
10.47
19.11
5.53
51.87
13.21
30.77
24.69
5.27
33.87
81.90
42.41
16.06
15.86
1.67
29.33
8.61
3.49
2.62
15.97
40.27
20.36
67.37
60.89
14.39
71.60
2.54
26.21
10.02
27.09
31.95
14.97
30.44
36.63
42.77
39.31
54.12
29.12
74.25
61.33
8.24
84.77
79.71
36.51
32.59
20.02
20.21
12.77
57.36
20.14
34.23
36.94
13.62
36.63
82.95
44.56
Table 3: Share of World Imports by SITC, 1990-2000 (%)
SITC
Developed Countries
Developing Countries
001 Live animals for food
011 Meat, fresh, chilled, frozen
012 Meat dried, salted, smoked
014 Meat prepd, prsrvd nes, etc
022 Milk and cream
023 Butter
024 Cheese and curd
025 Eggs, yolks, fresh, prsrvd
034 Fish, fresh, chilled, frozen
035 Fish salted, dried, smoked
036 Shell fish fresh, frozen
037 Fish etc prepd, prsrvd nes
041 Wheat etc, unmilled
042 Rice
043 Barley, unmilled
044 Maize (corn), unmilled
045 Cereals nes, unmilled
046 Wheat etc, meal or flour
047 Other cereal meals, flour
048 Cereal etc preparations
054 Vegtb etc fresh, simply prsrvd
056 Vegtb etc prsrvd, preprd
057 Fruit, nuts, fresh, dried
058 Fruit prsrvd, preprd
061 Sugar and honey
062 Sugar preps non-chocolate
071 Coffee and substitutes
072 Cocoa
073 Chocolate and products
074 Tea and mate
075 Spices
081 Feeding stuff for animals
091 Margarine and shortening
098 Edible products, preps nes
111 Non alcoholic beverages nes
112 Alcoholic beverages
121 Tobacco, unmanufactd, refuse
122 Tobacco, manufactured
222 Seeds for soft fixed oils
223 Seeds for other fixed oils
411 Animal oils and fats
423 Fixed vegetable oils, soft
424 Other fixed vegetable oils
431 Procesd animl and veg oil, etc
1990
71.98
79.55
85.28
76.62
52.00
60.27
88.13
75.85
85.56
76.67
91.36
88.22
27.24
31.86
49.29
45.93
62.74
16.64
38.04
74.07
81.40
75.52
84.01
87.71
35.96
72.15
88.43
87.81
83.85
35.44
57.14
57.21
57.00
68.90
75.26
86.03
77.16
51.26
78.26
69.33
51.42
49.42
48.60
63.08
1990
24.07
14.81
6.38
13.62
46.34
25.08
11.13
23.08
13.93
18.40
8.55
9.75
57.21
63.55
32.41
37.50
33.58
79.65
61.56
25.43
17.74
18.61
12.69
8.67
32.46
20.46
8.16
7.40
14.00
40.66
37.11
16.45
36.23
24.46
23.02
11.11
19.20
37.13
18.86
24.82
45.38
46.50
48.20
32.68
Source: UNCTAD (2003).
2000
71.55
71.65
92.11
80.72
50.18
69.94
84.26
66.33
78.75
66.07
86.98
86.94
30.62
27.40
32.83
39.53
46.62
20.35
36.54
70.27
76.71
79.42
75.60
83.94
41.89
69.57
82.59
80.90
77.68
43.98
54.99
63.17
41.27
62.15
71.60
84.87
59.14
59.17
50.46
67.76
44.33
45.84
40.53
54.34
2000
27.06
24.56
7.54
16.47
47.43
26.31
14.01
31.42
18.65
33.19
12.86
11.66
65.18
69.91
60.56
57.68
45.82
71.03
58.07
25.71
20.00
16.64
19.03
12.59
48.24
24.44
10.86
11.48
17.39
44.37
40.30
31.51
50.57
33.21
25.81
12.41
28.66
35.75
48.27
28.03
50.47
48.79
56.52
39.73
Table 4: Product Concentration Ratios in US Food Manufacturing1, 1997
Product
CR4 (%)
Dog and cat food mfg.
63.4
Malt mfg.
66.5
Wet corn milling
73.7
Soybean processing
73.4
Other oilseed processing
72.7
Breakfast cereal mfg.
86.7
Sugar cane mills
61.8
Cane sugar refining
96.4
Beet sugar mfg.
82.7
Chocolate and confectionary mfg.
86.6
Condensed/evaporated dairy mfg.
68.8
Cookie and cracker mfg.
64.6
Snack food mfg.
63.0
Brewing
90.7
Distilling
64.8
Cigarettes
98.0
Average
75.9
1
Source: US Census Bureau, 2001. Share of value added
accounted for by the 4 largest firms.
Table 5: Concentration Ratios1 by Product in EU Countries
Product
Baby food
Canned soup
Ice cream
Coffee
Yoghurt
Chocolate
confectionary
Pet food
Breakfast
cereals
Tea
Snack foods
Carbonates
Butter
Pasta
Frozen meals
Wrapped
bread
Biscuits
Canned fish
Mineral
water
Fruit juice
Canned
vegetables
Average
Ireland
98
100
-91
69
95
Finland
100
85
84
72
83*
74
Sweden
100
75
85
71
90
--
Denmark
99
91
90
70
99*
39
Italy
96
50
73*
60
36
93
France
93*
84
52
100
67
61
Spain
54
-84
-73
79
UK
78
79
45
74
50
74
Germany
86
41*
72
67
76
--
Average
91
87
76
75
70
74
98
92
80
--
84
52
40
70
64*
88
73
70
53
82
77
65
87
67
79
73
96
72
85
-83
-85
90
70*
50
-97
-44
63
80
62
-82
63
47
64
78
-100
61
-59
80
71
60
-51
90
80
82
50
69
32*
57
62
70
62
56
79
-65
39
96
52
73
55
65
37
39
58*
55
48
60*
30
49
65
9
72
68
71
65
65
62
59
83
---
73
70
100
51
72
74
44
49
70
55
68
37
61
43*
--
53
33
31
42
43*
14
50
-22
58
55
50
---
70
68
50
47
65*
50
62
36
26
29
38
--
35
--
46
--
48
47
79
69
69
67
63
61
56
55
68
89
Source: Cotterill (1999).
1
3-firm concentration ratios, except * which are 2-firm.
Table 6: Seller Concentration in US and EU Food Retailing, 1990s
Country
CR5 (%)
Austria
79
Belgium-Luxembourg
57
Denmark
78
Finland
96
France
67
Germany
75
Greece
59
Ireland
50
Italy
30
Netherlands
79
Portugal
52
Spain
38
Sweden
87
UK
67
EU
26
United States
35
Source: Cotterill (1999), McCorriston (2002), and Hughes (2002).
Table 7: Tariff Peaks and Preferential Access in the EU and United States
EU
HS 2-digit Product
Average
MFN tariff (%)
LDC
preference1
38.2
71.0
18.7
59.1
25.4
20.2
16.0
75.6
38.2
74.4
17.8
56.0
23.5
37.6
24.0
34.1
26.1
19.2
35.7
56.2
38.3
0.06
0.08
1.00
0.12
0.79
0.66
0.50
0.06
0.17
0.15
1.00
0.60
0.68
0.14
0.25
0.37
0.88
0.95
0.71
1.00
0.50
01 Live animals.
02 Meat and edible meat offal
03 Fish & crustacean, mollusk nes
04 Dairy prod; birds' eggs; honey
07 Edible vegetables and roots & tubers
08 Edible fruit and nuts; melons
09 Coffee, tea, mate and spices
10 Cereals.
11 Prod mill indust; malt; starches
12 Oil seed, oleagi fruits; misc grain
13 Lac; gums, resins & other veg
15 Animal/veg fats & oils & prod
16 Prep of meat, fish or mollusks
17 Sugars and sugar confectionery
18 Cocoa and cocoa preparations
19 Prep of cereal, flour, starch/milk prod
20 Prep of vegetable, fruit, nuts prod
21 Miscellaneous edible preparations
22 Beverages, spirits and vinegar
24 Tobacco and manufactured
Average
Source: Hoekman, Ng, and Olarreaga (2002).
1
United States
Average
LDC
MFN tariff (%)
preference1
-19.2
-20.9
20.6
16.7
--16.3
77.9
-19.9
---16.8
28.7
19.8
-73.5
30.0
-0.00
-0.38
0.88
0.80
--1.00
0.00
-0.50
---0.84
0.55
0.74
-0.14
0.53
A value of 1.00 implies imports enter duty free.
Table 8: Levels of Tariff Escalation by Highest Group Post-Uruguay Round
Commodity
Group
Processing Stage
Level of Tariff
Escalation (%)
Reduction in Tariff
Escalation Post-Uruguay
Round (%)
2nd Stage
39.5
-7.0
1st Stage
33.6
-6.1
1st Stage
31.2
-4.9
2nd Stage
27.7
-1.1
3rd Stage
15.6
-2.6
2nd Stage
84.8
-17.7
4th Stage
37.2
-2.6
2nd Stage
34.4
-17.3
1st Stage
19.8
-11.2
1st Stage
14.1
-23.1
2nd Stage
160.1
-32.8
1st Stage
82.2
-14.8
1st Stage
50.3
-10.8
3rd Stage
1st Stage
30.0
29.1
-30.0
-7.7
United States
Dairy and Egg
Products
Dairy and Egg
Products
Sugar Products and
Sweeteners
Sugar Products and
Sweeteners
Dairy and Egg
Products
EU
Fruit Products
Sugar products and
Sweeteners
Dairy and Egg
Products
Root and Tuber
Products
Tobacco and
Pyrethrum
Japan
Dairy and Egg
Products
Sugar Products and
Sweeteners
Root and Tuber
Products
Hides and Skins
Dairy and Egg
Products
Source: Lindland (1998)
Figure 1: Schematic Outline of the Model
Retail Demand Curve
Firm 1
Firm 2
Imports (semi-)
processed good
from LDC
Firm 1
(Upstream)
Domestic
Agricultural
Sector
Firm 2
(Upstream)
Imports raw
commodity from
LDC