Contribution to the 6th Conference of the International Association of Sport Economists,

ATINER, University of Pireaus, Athens, Greece, 31 May - 2 June, 2004.

Final revised and corrected version

International trade in sporting goods

Wladimir Andreff

Academic economists have not yet investigated very much the area of international trade in sporting
goods. Although data is available at a macroeconomic level in developed countries, neither the old
economic theory of international division of labour (Ricardo) and international specialisation (Heckscher-
Ohlin-Samuelson - HOS), nor the new international economics (Krugman-Helpman), focusing on intra-
industry trade, have been seriously tested as regards to the sporting goods industry. They have hardly
been used or referred to in this context (two exceptions are Andreff, 1989 and Harvey & Saint-Germain,
2001). One problem is that trade in sporting goods can only be depicted in using the most detailed SITC
classification for which data is not available or published in many countries, namely among developing
countries. Another hindrance is that, even with the most detailed figures, the same country often appears
to be both importer and exporter of the same sporting good, for a number of SITC categories, so that one
faces intra-product trade within an overall intra-industry trade. Moreover, due to widespread
subcontracting with outward-processing trade and foreign direct investment in the sporting goods
industry, a share of international trade is simply an intra-firm transfer of products. The scarcity of
microeconomic studies about such transfers operated within transnational corporations (TNCs), such as
Nike, Adidas, Reebok, etc., does not provide an overall view of intra-firm trade in this industry. Since no
macro- and micro-economic data base covering the world trade in sporting goods has been set up to date,
researchers can basically rely on national statistics that do not allow extensive comparison between
different countries. Thus, the whole topic has remained widely unheeded until now. It is one of the most
promising areas for research in the economics of sports in the future. The present contribution gathers the
little non-systematic knowledge existing in the literature.

Global trade in sporting goods

An overview of the world trade in sporting goods results from a research by Harvey & Saint-Germain (2001) based
on the data coverage of 28 countries whose detailed SITC figures are available in the UN commodity trade
statistics, from 1974 to 1994. These countries represent 75% of global trade in sporting goods and encompass the

1
three NAFTA countries (Canada, Mexico, USA), the fifteen EU countries (as of 1995, after the fourth
enlargement) and ten South-East Asian countries (China, Hong-Kong, Indonesia, Japan, Malaysia, Philippines,
Singapore, South Korea, Taiwan, Thailand). The 25% missing share of global trade is concentrated in Switzerland,
Eastern Europe, some Asian countries (Pakistan, India, Vietnam), Maghreb (Morocco, Tunisia) and Latin
American (Argentina, Brazil) countries. Among the sampled countries, in 1994, the ten major exporters of sporting
goods were the USA, China, Hong-Kong, France, Austria, Korea, Japan, Italy, Germany and Canada ; the ten
major importers were the USA, Japan, Germany, Hong-Kong, Canada, France, the UK, Italy, Netherlands and
Spain.

Table 1 – Global trade in sporting goods by trading blocs *

Trading bloc
Trading bloc Year NAFTA EU Asia ten Other countries Total
NAFTA 1974 40.6 21.9 26.1 11.5 100
1984 56.4 13.4 21.2 9.1 100
1994 57.2 9.5 27.9 5.4 100
EU 1974 20.0 49.5 10.7 19.8 100
1984 16.0 52.4 16.9 14.8 100
1994 11.8 50.2 20.7 17.2 100
Asia ten 1974 51.9 19.3 18.5 10.2 100
1984 28.9 14.5 50.4 6.2 100
1994 31.0 12.1 50.0 6.8 100
* (X+M)/2 %; X = export; M = import Source: Harvey & Saint-Germain (2001).

The concentration of global trade in sporting goods by trading areas (Table 1) exhibits a tendency of
developed (NAFTA and EU) countries to primarily trade together. About two-thirds of NAFTA sporting
goods trade is with other NAFTA and EU countries ; nearly two-thirds of EU sporting goods trade is with
other EU and NAFTA countries. Thus, trade in sporting goods displays a geographic concentration on
developed countries just like most of the manufactured products whose global trade concentrates, in the
range of two-thirds, on North-North trade. The new international economics emerged in view of
explaining such a trade. The same contention can be extended to the ten sampled Asian countries since
some of them are developed (Japan) or newly industrialised countries. However, the share of trade with
NAFTA and EU in their overall sporting goods trade decreased from 1974 to 1994, because intra-area
trade across Asian countries has skyrocketed meanwhile. In 1994, the intra-area trade was 50% of overall
trade in sporting goods in Asia (50.2% in EU and 57.4% in NAFTA). The share of the intra-area trade in
sporting goods has not significantly changed in EU while it has increased in NAFTA. In 1994, 69.1% of
Canadian and 82.8% of Mexican sporting goods trade were with NAFTA (only 19.6% in the US case) ;
on the other hand, 52.8% of Italian, 41.8% of the UK, 40.8% of German and 35.2% of French sporting
goods trade were carried out with EU countries. Therefore, the second tendency is one of
‘regionalisation’ of the sporting goods trade into continental blocs.

2
Foreign trade of major countries in the sporting goods industry

Calculating an export to import ratio r = [X/M].100 shows whether a country is a net exporter (r > 100) or
a net importer (r < 100) of sporting goods. In 1990, among the ten European countries sampled in a study
for the Council of Europe (Andreff et al., 1994), only Italy was a significant net exporter while Belgium,
Finland, France, Germany, Hungary, Portugal, Sweden and the UK were net importers of sporting goods. Table 2
confirms that developed countries were net importers rather than net exporters of sporting goods, and this is partly
due to the relocation of the sporting goods industry in some developing countries (see below). All NAFTA
countries were net importers in 1974-1994 whereas only five out of fifteen EU countries were net exporters, the
most successful being Austria, one of the strongest exporters of winter sports goods, together with Switzerland.
Belgium, Sweden and the U.K. were net exporters in 1974 but ceased it to be later on. Finland, France, Ireland and
Italy still were net exporters in 1994. Within NAFTA, the USA accentuates her net importer position while Canada
and Mexico reduce their position as net importers.

Table 2 - Ratio of sporting goods export to import, major trade partners

Country 1974 1989 1994 Average 1974-94

Canada 44.7 49.9 61.7 51.3
Mexico 13.1 54.7 91.9 95.8
USA 57.9 42.8 43.3 48.4
NAFTA 38.6 49.1 65.6 65.2
Austria 384.0 390.3 301.1 381.1
Belgium-Lux. 130.1 56.1 47.0 74.7
Denmark 36.9 72.0 86.7 58.6
Finland 195.1 113.4 174.0 196.4
France 340.0 109.4 130.2 182.6
Germany 71.3 67.7 50.8 66.5
Greece 17.4 2.5 5.7 6.8
Ireland 118.6 151.9 194.7 157.4
Italy 95.2 67.7 141.2 109.1
Netherlands 25.4 45.9 39.6 41.3
Portugal 13.8 55.2 23.6 50.7
Spain 76.1 34.7 40.2 71.0
Sweden 137.7 63.5 59.3 83.5
United Kingdom 158.6 54.1 52.8 91.6
EU 128.6 91.7 96.2 112.2
China n.a. 696.5 799.0 747.8
Hong Kong 99.2 109.1 124.8 101.9
Indonesia 7.5* 72.2 2082.1 540.9
Japan 94.5 28.7 23.2 63.0
Malaysia 5.4 78.4 104.8 47.8
Philippines 67.9* 132.3 248.2 213.3
Singapore 18.9 57.5 67.4 45.5
South Korea 831.4 1100.2 326.4 844.6

3
Taiwan n.a. n.a. n.a. n.a.
Thailand 21.3 156.9 570.0 153.5
Asia nine 178.5 270.2 482.9 260.0
* In 1979 Source: Harvey & Saint-Germain (2001).

All Asian countries are net exporters, except Japan, Malaysia and Singapore, including those non-sampled
countries such as Pakistan, India, Sri Lanka, Vietnam (according to scattered information). The very high
value of the export/import ratio in South Korea, Indonesia, China, Thailand and Philippines is the other
side of the coin with regards to the relocation of the sporting goods industry from developed countries.
Maghreb countries are net exporters as well, since they are a privileged location for outward-processing
trade in the textile-clothing and footwear-leather industries, namely as far as the production of sportswear
and sporting footwear is concerned.

‘Equipment-intensive’ versus ‘trite’ sporting goods trade

When it comes to international specialisation in sporting goods trade, one can only notice the absence of
inter-country comparative studies. Since the crux of the matter is to analyse how countries specialise in
the intra-industry trade – all the twenty-eight countries are both importing and exporting sporting goods -,
the issue is to go deeper into the most detailed SITC products classification. Then, it is crystal clear that,
for a country, the advantage of exporting (importing) winter sports goods on the one hand, and balls or
sportswear on the other hand, has not the same economic value and the same impact on its trade balance.
Looking at their unit value in foreign trade, skis, ski boots, sailing boats, windsurfs or golf equipment
cannot be categorised as the same sort of sporting goods as, say, sportswear, tracksuits, balls, swimsuits,
sporting footwear. The former group contains goods with a high unit value, due to a significant value
added in the production process, a rather sophisticated and evolving technology and know how whereas
the latter group consists in cheaper goods (per unit) with a lower value added, which are produced with a
mature technology and an easily transferable know how. Moreover, high unit value sporting goods are
usually required for the practice of specialised equipment-intensive sports such as, for instance, sailing,
winter sports, surfing, motor sports or golf. Let us coin these sporting goods ‘equipment-intensive’. Low
unit value sporting goods are less specialised and can be used in a wider range of sport practices
(gymnastics, walking, body building, keep fit, team sports and track and fields) or even on leisure time
without any sport practice (ex. : sportswear, tracksuits, sporting footwear). Let us classify them as ‘trite’
sporting goods (Andreff, 1989). With this categorisation in mind, one can switch to a more detailed
qualification of product specialisation within the international intra-industry trade in sporting goods.

4
Unfortunately, until now, such a detailed research has only been done at a country level, examining the
foreign trade in sporting goods of one country, without international comparison.

International specialisation in the sporting goods trade

The differentiation of sporting goods in the intra-industry trade has been studied in detail in the French
case. Inter- or intra-product specialisation is measured together by the export to import ratio, product by
product, and the calculation of an intra-industry trade index for each detailed product. Several indexes are
available in the literature, the simplest one being the Balassa index :
Bi = [(Xi – Mi) / (Xi + Mi)].100,
where usually i stands for an industry. Here i will stand for one product or a product group smaller than
the entire sporting goods industry. When Bi = 100, a country is exclusively exporter and when Bi = -100
it is exclusively importer of the sporting good i. This product is typically a ‘pure’ Heckscher-Ohlinian
good and the country exhibits an inter-product specialisation as regards this good in the sporting goods
industry. When Bi = 0, a country exports exactly as much as it imports of a sporting good i; one can coin
it a ‘pure’ Balassa good and the country shows a Krugmanian intra-product specialisation as regards this
good in the sporting goods industry. Economists usually conclude, when -30 < Bi < +30, that one
observes an intra-industry (here intra-product) trade, corresponding to the international specialisation
across developed countries on imperfect markets with increasing returns (a Krugmanian specialisation).
When Bi < -30 and Bi > +30, trade is considered as inter-industry (here inter-product), in tune with a
traditional HOS international specialisation.

Table 3 - International specialisation in the sporting goods industry: France, 1981-2001

Some major products 19 81 19 86 19 92 19 99 200 1

X/M Bi X/M Bi X/M Bi X/M Bi X/M Bi
Equipment intensive goods
sailing boats, yachts 436 63 717 76 185 30 430 62 456 64
skis & accessories 496 76 897 80 742 76 598 71 440 63
ski boots n.a n.a. 409 61 231 40 114 6 104 2
windsurfs & accessories 184 30 114 6 * * * * * *
golf equipment 19 (-68) 10 (-81) 51 (-33) 66 (-20)
Gymn. & sport. equipment 7 (-87) 27 (-58) 26 (-59) 102 1 95 2
Trite sporting goods
Swimsuits 268 46 108 4 84 (-9) 57 (-27) 57 (-27)
Sports footwear 79 (-12) 68 (-19) 44 (-39) 29 (-55) 19 (-69)
Skates 44 (-39) 15 (-73) 46 (-37) 58 (-26) 55 (-29)
* Included in sailing boats; X/M: ratio of export to import; Bi: Balassa
index.
Sources: Andreff (1989), STAT-Info (Ministry for Sports).

5
Table 3 shows that, in the long run, France is specialised as an exporter of ‘equipment-intensive’ sporting
goods such as sailing boats, yachts, windsurfs, skis and accessories, and (less and less) ski boots; she
improves its net importer position in gymnastics and other sports equipment and in golf equipment. On
the other hand, at least since 1981, France is a net importer of ‘trite’ sporting goods such as skates and,
increasingly, sports footwear while she has switched from a net exporting to a net importing position in
swimsuits (as well as in other sportswear, not in the table). A conclusion can be derived, to the extent that
France is representative, which is that developed countries tend to be net exporters of high value added
and high-tec ‘equipment-intensive’ sporting goods whereas they are net importers of ‘trite’ sporting
goods. The next question will be: where from? The second conclusion arising from Table 3 is that French
trade in various sporting goods exhibits an inter-product specialisation, in particular in ‘trite’ goods such
as sports footwear and skates. However, in ‘equipment-intensive’ goods, an intra-product trade is
observed for ski boots and golf equipment, in the recent years ; whether France is a net exporter or
importer of these products, she imports a significant volume of these items from other developed
countries (since developing countries produce nearly no ski boots and golf equipment). French trade in
‘equipment-intensive’ sporting goods is rather representative of North-North intra-industry and intra-
product trade in high value-added manufactured goods, which grows in a context of imperfect
(monopolistic or oligopolistic) competition. The size and the large world market share of French firms
such as Salomon and Rossignol in ski and ski boot production or Bénéteau in the production of sailing
boats are in tune with previous observations.

In order to complete the analysis of a country’s international specialisation in sporting goods trade, some
information is needed about where exports are flowing to and where imports are coming from. In the case
of France, the major trade partners are:
. the USA, Japan, Germany, Italy, Switzerland, Belgium, Sweden, Canada, Austria, the UK as regards
major exports of ‘equipment-intensive’ sporting goods ;
. Italy, Austria, Switzerland as regards major imports of ‘equipment-intensive’ sporting goods ;
. Eastern and Southern European, South Asian and Maghreb countries as regards major imports of ‘trite’
sporting goods, namely Morocco, Tunisia, China, Thailand, Pakistan, South Korea, Hong Kong, Taiwan,
Indonesia, Philippines, Malaysia, Hungary, Poland, the Czech Republic, Romania, Croatia, as well as
Italy, Spain, and Portugal ;
. French exports of ‘trite’ sporting goods are geared towards European developed markets, namely
Germany, Italy, Belgium, and the UK.

6
The outflow of market-seeking exports crosses the inflow of market-seeking imports in ‘equipment-
intensive’ sporting goods. On the other hand, most of imported ‘trite’ sporting goods come from countries
with a lower unit labour cost, and they are backed by an efficiency-seeking (or cost-reducing) rationale,
sometimes linked to production relocation in the Third World and Eastern and Southern Europe.
However, this specialisation comes out with an overall French trade deficit in sporting goods, since the
late nineties, insofar as the net export of ‘equipment-intensive’ goods now is lower than the net import of
‘trite’ goods. For the same reason, in 1990, different European developed countries exhibited a trade
deficit in sporting goods (Andreff et al., 1994) such as Germany (PPP$ 1,065 million), the UK (PPP$ 536
million), Sweden (PPP$ 72 million), Belgium (PPP$ 69) and Finland (PPP$ 29 million) while Italy had a
trade surplus (PPP$ 468 million), being a net exporter of both ‘equipment intensive’ and ‘trite’ sporting
goods. Another consequence of this specialisation pattern is that a number of sporting goods, such as
balls, sportswear, sporting footwear, bikes and rackets, which were produced in France in the sixties and
the seventies, have been crowded out by imported products. For instance, in sporting footwear, the ratio
of import to domestic demand has increased from 50% to over 80% in the late eighties, while the French
domestic production has halved. Thus, France was allowed by the European Community to restrict, from
1988 on, the imported sporting footwear from Korea (a 2 773 000 quota of sport shoes pairs in 1991) and
Taiwan (a 778 000 quota in 1991), after she argued that Korean and Taiwanese sport shoes imports were
accountable for 14,000 redundancies in the French industry.

Production relocation: outward-processing trade and foreign direct investment

Facing competition from developing and newly industrialising countries that enjoy lower unit labour
costs in the production of ‘trite’ sporting goods, North American and European firms embarked on
relocating their production in the Third World and Eastern and Southern Europe. For an American or
European firm, it is worth relocating its production when:
wh / qh – wf / qf > ci + cj + ti + tj + g - e
where wh stands for the wage cost in the firm’s home country, qh for the labour productivity in the home
country (so that wh / qh is the unit labour cost in the home country), wf / qf for the unit labour cost in a
foreign subsidiary (or subcontractor) in a Third World country, ci for the transportation cost of inputs
manufactured in country h to the country f, cj for the transportation cost of the relocated output from
country f to country h (or any other developed customer country), ti and tj for the tariffs on the previous
international flows of input and output, g for the governance costs of the subsidiary (or subcontractor)
located abroad, and e for the transaction costs saved on the firm’s exports substituted by the relocated
production.

7
The first strategy, adopted by Nike and Reebok, was one of subcontracting with local producers and
trading inputs and output under the benefit of outward-processing trade regulation. This strategy
sometimes had gone so far that Nike became a hollow corporation with no longer any production unit in
the USA. The second strategy is foreign direct investment with setting up subsidiaries in low unit labour
cost countries. As a result, nearly all the global production of balls concentrated in Pakistan, India and
Taiwan, most of the global bike wires production was relocated in Malaysia, 90% of the global sporting
footwear, 80% of all tennis rackets and over 90% of tennis balls were manufactured in Korea and Taiwan
while the great bulk of sportswear was produced in Italy, Portugal, Eastern Europe, and Third World
countries. All these products then started to be imported by developed countries, either in the framework
of outward-processing trade or in the intra-firm trade of the major TNCs of the sporting goods industry
that had settled subsidiaries in developing countries. All the Nike and Reebok sporting footwear now is
manufactured by Asian subcontractors, as well as 80% of Mizuno sport shoes, whereas Adidas has
relocated 70% of its sporting footwear production in Asia, Tunisia and Hungary. In a second wave of
production relocation, Asian producers of sporting goods have relocated their plants in lower labour cost
Asian countries such as the Taiwanese Kunnan (Kennex) in Thailand, the Korean Tae Hwa in Indonesia,
and others in China, Philippines, Malaysia and Vietnam. Therefore, there is an obvious globalisation of
the sporting goods industry in both trade and production. For instance, in 1998, 41% of Nike’s global
sales were carried out outside the USA while 45% of Adidas’ global sales were outside Europe (Bourg &
Gouguet, 2001).

Child labour in the relocated sporting goods industry

Relocating production in cheap labour countries is not without its problem to TNCs. The major issue is
child labour in the factories where the production of ‘trite’ sporting goods is relocated, either in a TNC’s
foreign subsidiary or more often in a local subcontractor’s plant. A well-known example is Nike. In
Indonesia, 160,000 workers were involved in the production of sporting footwear for the Nike trademark.
In the Bogor plant, the daily wage was half a dollar and a glass of milk in 1998 while the thirteen
members of the Nike’s board of directors were earning an annual income over $5 million each (not
including their stock options), twice the amount of the overall wage bill of 6,600 workers employed to
produce for the Nike trade mark in the Djakarta area. On each shoes pair sold in developed countries, the
Nike’s subcontractor worker got 10 cents of a dollar (0.2% of the selling price) while each shareholder got
40 cents. Nike’s subcontractors in Indonesia are located in special (closed) trade zones where they are

8
supervised by waged-armed guards and trade unions are not allowed. The Sialkot assembly line of soccer
balls in Pakistan was sadly infamous and publicised for resorting to mass child labour (Riddle, 1997).

After such a negative advertising for its industry, the World Federation of Sporting Goods Industry
(WFSGI) was so much concerned with phasing out child labour that it convened a conference to look at
the economic and social accountability of the sporting goods industry in developing countries where final
products are manufactured and assembled. A task force on global manufacturing practices worked out an
assessment of the extent and scope of child labour in the soccer ball industry. A meeting with ILO
(International Labour Organisation) and lasting negotiations with Pakistani producers (subcontractors)
came out with an industry-wide programme to eliminate child labour in soccer ball stitching. The problem
is that this programme is voluntary, not compulsory. ILO intends to prolong practical action to phase out
child labour in this industry (Tucker, 1997). Finally, the WFSGI adopted, by end of 1997, a Model Code
of Conduct for global business practices that addresses working conditions (child labour, forced labour,
wages, the length of the working day, the right of unionisation, etc.). It is a gentleman agreement or a
moral code rather than a demanding economic regulation. However, due to the bad global image created
by child labour, most TNCs in the sporting goods industry now claim their zero tolerance and have taken
some initiatives against this practice in developing countries.

The scarcity of microeconomic data: transnational corporations in the sporting goods industry

There is no detailed database about TNCs in the sporting goods industry and, until now, researchers can
only rely on case studies. However, in many sporting goods industry the global market structure is
typically an oligopoly with a handful of big TNCs and, in each developed country, a number of
competing small and medium enterprises. The global sporting footwear market is precisely representative
of this market structure with twelve oligopolistic TNCs covering an 83.4% market share, in 1998 (Table
4).

Table 4 - The global sporting footwear market: firms with more than a 1% market
share

Firm 1998 sales* Percentage of sales in: World market

the USA rest of the world share
Nike 4880 61 39 32.9
Adidas 2510 34 66 17.0
Reebok 1730 56 44 11.7
Fila 620 34 66 4.2
New Balance 560 55 45 3.9
Asics 415 29 71 2.8

9
Converse 400 40 60 2.7
Puma 325 9 91 2.2
Keds 250 73 27 1.7
Vans 180 76 24 1.2
Brooks 160 31 69 1.1
K-Swiss 150 90 10 1.0
Diadora 150 8 92 1.0
Total 12330 50 50 83.4
World sales 14790 46 54 100%
* Converted into million Euro

The global strategy of a sporting goods TNC means (Andreff, 2003) that it has a world outlook of
competition, it has a good knowledge about its oligopolistic competitors, it concentrates its activity on the
Triad countries (North America, Europe and Japan), it behaves as a global player of the world economy, it
looks for innovation on a global scale, it locates its operations where they are the most profitable
according to the comparative advantages of different host countries, it co-ordinates the network of all its
subsidiaries, plants, laboratories with the help of the new information and communication technologies
(global networking). In addition, a global TNC in the sporting goods industry includes in its strategy
transborder take-overs such as Adidas over Salomon, Pony, Arena and Le Coq Sportif. A TNC strategy
more specific to this industry is global sponsoring 1 : Adidas, Nike, Reebok are the sponsors of a number
of international sport events, national teams, and famous high level athletes (advertising and
communication expenditures reach about 13% of Adidas sales). The global market for sport sponsorship
was estimated at Euro15 billion, in 1998.

Conclusion

International trade in sporting goods and the role of transnational corporations in their production remain
among the most unheeded areas of research in the economics of sports. These topics deserve and require
more empirical investigation that could be used as a rocket pad for a more elaborated economic analysis.

References :

Andreff W. (1989), ‘L’internationalisation économique du sport’, in W. Andreff, ed., Economie politique du sport
(Paris : Dalloz), 203-236.
Andreff W. (2003), Les multinationales globales, Repères n° 187, 2nd ed. (Paris: La Découverte).

1
See also “ Sponsorship ” in this volume.

10
Andreff W., J.-F. Bourg, B. Halba, J.-F. Nys (1994), The Economic Impact of Sport in Europe : Financing and
Economic Impact, Background document, 14 th Informal Meeting of European Sports Ministers, Council of Europe,
Strasbourg, 28-29 April.
Andreff W., J.-F. Nys (2002), Economie du sport, Que sais-je ? n° 2294, 5th edition (Paris : Presses Universitaires
de France).
Bourg J.-F., J.-J. Gouguet (2001), Economie du sport, Repères n° 309 (Paris : La Découverte).
Donagu M.T., R. Barff (1990), ‘Nike Just Did It : International Subcontracting and Flexibility in Athletic Footwear
Production’, Journal of the Regional Studies Association, 24, 537-551.
Harvey J., M. Saint-Germain (2001), ‘Sporting Goods Trade, International Division of Labor, and the Unequal
Hierarchy of Nations’, Sociology of Sport Journal, 18, 231-246.
Riddle J. (1997), ‘Sports Industry Tackles Child Labor Issue. Ball Manufacturers Spearhead Effort, WFSGI News
Bulletin, January-February.
Tucker A. (1997), ‘Child Labor Issue at the Top of the Global Agenda’, WFSGI News Bulletin, August.

11