Description
The consumption of cultural goods and services is an important indicator of tastes and preferences, and from these, the influence of global exchange flows between nations.
The consumption of cultural goods and services is an important
indicator of tastes and preferences, and from these, the influence of
global exchange flows between nations. If the international per-
formance of cultural industries goes together with more general eco-
nomic performance on the international markets, this makes these
cultural industries an issue that goes beyond culture alone.
In a context where these international exchanges – economic on
the one hand, artistic and cultural on the other, are developing and
presenting a new geographic configuration, these challenges must
be taken up. Certain countries, such as China and India in fact
occupy an increasing place in economic exchanges, and exchanges
of services, after those of goods, continue to increase every year.
Examining the determining factors for cultural exchanges and the
links between them and the notion of cultural proximity between
countries is the objective of the research presented here to encour-
age all the levers affecting exchanges of cultural goods and services
being taken into account.
To analyse questions of distance and cultural proximity in
exchanges, the study is based on the so-called “gravity” model, cur-
rently used in the international economy. This methodology allows
factors in exchanges of cultural goods to be determined at both
global and sector levels (see box p. 23). It also makes the best use
of all the available databases to obtain the same type of results for
cinema and cultural services.
It seeks to identify any effects arising from the existence of bor-
ders in cultural exchanges, to see how much cultural exchanges could
constitute indicators of cultural proximity in both commercial
exchanges and in other important aspects of international economic
relations: Foreign investments and migratory fluxes.
1. This summary has been prepared by François Rouet (Study manager at DEPS), from research undertaken within CEPII by Thierry Mayer
(CEPII, University of Paris I, Paris School of Economics and CEPR), Anne-Celia Disdier (INRA, UMR public economics), Silvio Tai (Uni-
versity of Paris I and Paris School of Economics) and Lionel Fontagné (University of Paris I, Paris School of Economics and CEPII).
2007-2
182, rue Saint-Honoré, 75033 Paris cedex 01
01 40 15 79 13 – 01 40 15 79 99 Available to download from:http://www.culture.gouv.fr/deps
International trade flows
in cultural goods and services:
Issues and determining factors
1
cul ture
studies
PRODUCTION, DISTRIBUTION AND MARKETS
Publishing director: Philippe Chantepie, Head of the Département des études, de la prospective et des statistiques (DEPS) 2007-2 – September 2007
Head of publications: Jacqueline Boucherat
Foreword
Without sources and weighting for
exchange factors, the measurement
and understanding of international
exchanges of cultural goods and serv-
ices are difficult and consequently
rarely undertaken exercises. However,
it is essential to identify the determin-
ing factors for the exchange of cultural
products and services if we want to out-
line a strategy for introducing French
players into international exchanges
and more widely, to defined conditions
for cultural diversity. The results of
research undertaken by a team of
researchers associated with CEPII
who made use of econometric methods
on expertly evaluated sources, consti-
tute a solid base for the definition of
such a strategy from the determining
factors in international exchanges of
cultural goods and services.
P. C.
Secrétariat général
Délégation
au développement
et aux affaires
internationales
Département
des études,
de la prospective
et des statistiques
CULTURAL EXCHANGES:
DISTANCE AND DETERMINING FACTORS
The gravity model allows the examination of
questions of distance and cultural proximity in
exchanges and the determination of factors affect-
ing them at both global and sector levels (see box
p. 23).
Distance and cultural proximity
questions
Different forms of proximity are thus likely to
exist between countries and to affect international
trade. Among them, the fact of sharing a language
and former colonial links, have been the subject of
a large number of research projects.
Linguistic proximity tends to favour trade
between countries: Having a common language
increases the flow of exchanges by about 65%.
Attempts to measure the linguistic distance sepa-
rating two commercial partners – from the absence
of a common language to perfect linguistic simi-
larity – have shown that the greater this distance,
the more trade decreases. However, it is noted that
the influence of this distance has reduced by a third
between 1971 and 1985: All other things being
equal, trade between two countries that are com-
pletely dissimilar from the linguistic point of view
reached 28% of the trade between two perfectly lin-
guistically similar countries in 1971, whereas in
1985, the difference was 48%.
Relations created between countries in the
course of history, in particular in the colonial
period, can also encourage international trade, com-
mercial relations established by colonizing coun-
tries with the colonized countries often persisting
after the colonial period. In addition, with the col-
onized country often having adopted the institu-
tional framework of the colonizing country, the
resulting similarity of legal rules and administrative
systems allows an increase in the security of trans-
actions and a reduction in communication costs,
which tends to favour international trade. We thus
understand why a former colony can retain a higher
than expected
2
level of trade with its former colo-
nizer, but also why two countries, which had the
same colonizer, could also have more developed
bilateral exchange flows than expected. Other
things being equal, the existence of a colonial rela-
tionship in the past multiplies bilateral trade by a
factor of 5.75, whereas the fact of having had the
same colonizing country increases bilateral
exchanges between two countries by 80%. How-
ever, this influence itself is clearly reducing with
time.
These preferred relations between countries,
favouring exchanges, can arise from the existence
of network effects on trade, whether these are busi-
ness networks or social networks. In this respect,
the volume of migratory flows can influence bilat-
eral trade. In fact, immigrants take part in different
ways in the development of trade between their
country of origin and their host country: By retain-
ing, at least in part, a preference for the goods pro-
duced in their country of origin, which leads to
additional demand; by distributing these goods
within the host country; by being particularly able
to establish commercial relations with their coun-
try of origin because their mastery of the language,
the culture and legal practices allows them to be
more in touch with opportunities and needs, and
better connected to the business networks. Finally,
more generally, we have been able to bring out that
the more the citizens of a country felt confident
about the citizens of another country, the greater the
economic exchanges between the two countries,
and in lower proportions, investments in the other
countries.
All these factors are likely to have a dual influ-
ence on exchanges: First in reducing costs, but also,
more indirectly, in influencing cultural preferences.
From which we have the idea of using the
exchanges of cultural goods in which these cultural
preferences are particularly expresses as a measure
of the cultural proximity of the countries and an
approximation of these bilateral preferences (see
“Cultural exchanges: Cultural proximity indicators
and a factor explaining all exchanges”, p. 20).
Determining factors in exchanges
of cultural goods…
At global level
For all goods, bilateral exchanges are propor-
tional to the respective gross national products
(GNP) of the two countries and are larger the higher
the level of development. Colonial links, contiguity
and a common language all have a positive effect,
but the effect of geographic distance itself is nega-
tive. The results for the subset of the seven major
18
2. “Expected” in the sense of “predicted” by the gravity model meaning given the respective economic sizes of the commercial partners and
the distance separating them.
Focuses on cultural topics: economic approaches
2007-2 19 cul ture studies
cultural goods sectors (heritage, books, newspapers
and periodicals, other printed matter, recorded
media, visual arts, audiovisual media) show that the
fundamental determining factors for exchanges of
these goods are imperfectly taken into account by
simple gravity, in particular by GNP and GNP/head.
But if we introduce “fixed effects”, which take
account of the specific features (supposed constant
over time) of each country importing and exporting
cultural goods, the estimate is greatly improved,
which proves the importance of these interrelations
and the impact of specific features between coun-
tries. The impact of distance is however lower for
cultural goods than for all goods, which can be
explained by a larger “cultural” separation that does
not compensate for possibly lower transport costs.
The fact that transaction costs have a lower neg-
ative impact on trade for cultural goods than for
other assets appears to be a very solid result. In con-
trast, the GNP of the exporting country has a posi-
tive impact on cultural trade, with colonial links
increasing bilateral trade by 33%, a common bor-
der by 85% and a common language by 56%, with
this last impact being greater for cultural goods.
At the sector level
An attempt to explain the volume of bilateral
cultural exchanges has also been tried for each of
the seven main categories of cultural goods. We
note that the “distance” variable plays a negative
role in all cases, but that, in contrast, the develop-
ment level of the partner countries measured
through GNP/head does not play a significant role
in explaining the volume of bilateral exchanges,
except for the heritage goods sector.
Two influences – that of GNP and that of vari-
ables related to cultural proximity (colonial links,
common border and common language) – differ
between sectors: Whereas colonial links positively
influence exchanges in the heritage goods, news-
paper and periodical and visual arts sectors, their
influence is, rather surprisingly, negative for audio-
visual media. For the “books, newspapers and peri-
odicals and other printed matter” sectors, the shar-
ing of a border and/or a language encourages trade.
Unlike recorded media, the visual arts and audiovi-
sual media are not influenced by sharing an official
language.
A common border has a positive impact on the
flows in the visual arts and audiovisual media sec-
tors but does not influence the recorded media sec-
tor. The common language is the main factor in the
penetration of cultural goods with written content
– this factor multiplies the flow of books by over
4.5 – and the colonial has certainly shaped the pref-
erences of consumers in the importing countries for
heritage goods where it multiplies the flow by 3.74.
For the “distance” variable, the calculation of the
average distance separating partners in the
exchange according to the types of goods shows
that cultural goods are exchanged over smaller dis-
tances than non-cultural manufactured goods
(respectively 6,005 km against 7,271). In addition,
if we apply this calculation to each of the seven cat-
egories of cultural goods, it appears that it is for her-
itage goods that the partners in the exchange are on
average the furthest apart from each other
(6,616 km): Essentially comprising irreproducible
artistic goods, this category depends less on dis-
tance than other cultural goods and is thus less sen-
sitive to transport costs and more generally to trans-
action costs. The fact that the influence of a com-
mon border is not significant for this category of
goods adds strength to this observation.
The cinema case
The Unesco data for the period 1970-1999, cover bilateral imports
of films from the main producing countries (United States, France,
Germany, Russia, Japan, India, Hongkong, United Kingdom and
Italy) but also for national production. They allow us to establish
that distance plays the same role as in the previous estimates:
The flows of films are less affected by distance than flows for the
rest of manufactured goods, and confirm the fundamental weight
of a common language and a colonial link. However, specific
national features also come into play and generate “fixed effects”
that can be assessed relatively: So, some importing countries,
with a special taste for the cinema, import many films of all ori-
gins in contrast to others that are poorly equipped with cinemas
or distribution networks, due, for example, to very low density and
large spatial dispersal of the population.
The fact of having a common language or a colonial link increases
the market share of the exporter concerned by 70% in countries
that share one or the other. In this respect, the French speaking
community has a real economic advantage in terms of cinemato-
graphic exchanges: France, like Great Britain and the United
States in particular, benefits in general from links created by his-
toric population movements, the similarity of institutions and cul-
tural practices. If the effects of the “colonial link” and “linguistic link”
variables can be very clearly separated for France, they are cumu-
lative in most cases: In fact we note that in a former French colony
where French is still spoken, the market share of French films is
multiplied by three compared to a country where the language is
not French and that has never been a French colony.
… and cultural services
As far as dematerialized services and exchanges
in general are concerned, which do not need phys-
ical contact between the consumer and the producer
and do not give rise to transport costs, what does
the concept of “distance” mean? However, despite
this, most studies show a significant and relatively
large negative effect of distance on the flow of serv-
ices. The explanation is perhaps that distance meas-
ures similarities in consumer preferences, similari-
ties that are larger as the countries are closer on the
geographic level.
At the start of the 2000 decade, the flow of serv-
ices represented more than 20% of international
exchanges, a share that should continue to grow
when the development of ICT is taken into account.
Despite their significant and increasing weight in
exchanges and whatever the services – cultural or
not -, these transactions are poorly recorded and we
have very few statistics
3
due to the lack of a stan-
dardized nomenclature. In world exchanges of serv-
ices, “audiovisual and related services” and “other
royalties and licence fees” still only represent a
minimal but strongly growing share.
If the effect of the “distance” variable is still neg-
ative for these two types of services, it is however
weaker than for all services. In addition, all the
other “gravitational” variables (see above) have the
expected effect: The size of the country has a pos-
itive influence on exchanges of cultural services,
although less strongly, here too, than for all
exchanges of services; the flows also increase with
the countries’development level evaluated by
GNP/head; colonial links and sharing a terrestrial
border strongly favours trade in cultural services
whereas this positive and significant influence of
colonial links is not observed for all services;
finally, sharing an official language does not appear
to influence the flow of services, whether in terms
of all services or for cultural services.
As far as the GNP and GNP/head of the export-
ing country are concerned, their time variations do
not appear to influence exchanges of cultural serv-
ices. In contrast, those of the importing country do
have an impact on exchanges, positive for the first
and negative for the second. In fact it seems that,
following an improvement in their development
level, countries import fewer cultural services, the
development of local production tending progres-
sively to replace imports. Exchanges of cultural
services, contrary to what happens for services in
general, tend to reduce when a common border is
shared; but they nevertheless remain positively
influenced by a the sharing of an official language.
The common language thus seems to be one of the
most robust influencing variables for cultural
exchanges. This variable has an almost systematic
very clear positive impact when the specific features
of the importing and exporting countries are taken
into account.
CULTURAL EXCHANGES:
CULTURAL PROXIMITY INDICATORS
AND A FACTOR EXPLAINING
ALL EXCHANGES
Cultural exchanges appear to constitute, through
the explanation of crossing a border on exchanges,
indicators of the cultural proximity between coun-
tries. In addition, cultural exchanges constitute a
particularly strong explanatory factor for the dif-
ferent aspects of international exchanges (goods and
services, foreign investments, migratory flows).
Border effects for cultural goods
and services
Another way of envisaging international
exchanges consists in asking the question of “the
border effect”, meaning the effect of crossing
national borders on exchanges. Research on the
subject of borders, both between Canada and the
United States and within the European Union sug-
gest fairly high border effects despite the process of
economic integration and sometimes a common
language. These effects are only partially explained
by formal commercial barriers to exchanges (cus-
toms duties…), there thus remains a place for cul-
tural differences that drive individuals to grant a
certain preference – price being equal – to national
products. Measuring border effects on cultural
goods and services partly allows measurement of
these cultural differences. If the impact of national
borders comes from cultural differences between
countries, we must thus find a border effect in
exchanges of cultural goods, which can then
become a sort of “tracer” for cultural preferences.
The Eurostat AUVIS database, which supplies
the number of admissions achieved by films pro-
duced by one country in each of the other countries
in the period 1980-2001, allows the border effect to
be estimated subject to defining a measure of the
20
3. Several databases (OCDE, Eurostat, CEPII) supply data from the same source: the balances of payments of different countries.
Focuses on cultural topics: economic approaches
“internal” distance within the country
4
. We observe
that the border effect is very important: On average,
admissions for national films are nearly ten times
as high as the expected value. This demonstrated
the fact that despite the development of interna-
tional exchanges and the rise in power of the Amer-
ican cinematographic industry on the world scale,
the “domestic bias”, meaning the preference for
national production in consumption of the cultural
good that the cinema comprises, is high. The
hypothesis in which cultural differences are one of
the probable causes of the existence of the effect of
national borders on general trade in goods is thus
found to be reinforced.
This border effect, in view of the results of pre-
vious research, appears very comparable to that for
non-cultural goods taken as a whole and even, par-
adoxically, clearly less than that estimated from
food goods or drinks, non-cultural goods that nev-
ertheless include a large cultural component. The
explanation, in the cinema case, could be that the
market, even if it is far from being completely inte-
grated, is much more so than for other cultural
goods.
Finally, the relative preference of French audi-
ences for national films, or put another way, the
level of “closure” of the French market, is not
greater than for other countries. On the contrary, the
French market is actually slightly more open than
other markets. This can be surprising on first
encounter, in that France is one of the rare countries
where the domestic cinema’s market share remains
large, but this result is explained by the fact that
production is large, which largely justifies the
respectable part that French films occupy in the
internal market. Having regard to a border effect for
France relatively close to that obtained for the
United States, we can consider that public support
policies do not lead to distortions in the consump-
tion of films by French audiences. The existence of
a high volume of production encouraged by the sup-
port systems does not imply a lower level of foreign
film imports but rather a higher total level of film
consumption, from which foreign films benefit.
Can exchanges of cultural goods explain
commercial flows?
How far do exchanges of cultural goods go
together with all exchanges, for which they never-
theless represent, on average, only 1.4% of the total
for the period 1988-2004? Use of the data in the
United Nations Comtrade database allows an esti-
mate that trade in cultural goods positively and sig-
nificantly influences exchanges of all goods. The
use of this explanatory factor as a direct measure of
cultural proximity between countries, also leads to
rendering the influence of the factors usually used
less significant (sharing a common language, dis-
tance, GNP, development level…).
If we consider the impact of all seven categories
of cultural goods individually, we observe that the
flow of five of them exercises a significant positive
effect on total trade (with the exception of the flow
of heritage goods which has a negative influence
and exchanges of other printed matter which does
not have a significant effect). In total, an increase
of 10% in exchanges of cultural goods increases
trade in traditional goods from 3.25 to 4.25% and
exchanges of visual arts alone, in increasing by
10%, increase total trade by nearly 2%. If we limit
ourselves only to bilateral exchanges of films or
cultural services
5
, the positive influence on all trade
is found again even if it is less.
The introduction of cultural goods exchange
flows thus modifies the influence of the traditional
variables relating to cultural proximity (colonial
links, common border, common official language)
in addition to the influence of the “distance” vari-
able. This signifies that there is a consequent inter-
dependence between these variables and the flow
of cultural goods. We can demonstrate, for exam-
ple, that geographic distance, taken as an explana-
tory factor for trade as a whole, partly translates the
cultural distance separating the countries. These
cultural exchange flows explaining the global
exchanges well, this means both that they depend
in part on the cultural proximity of the countries but
also that they take better account of the influence
of cultural proximity on total trade of a country than
the traditionally used measurements (language,
colonial link…).
2007-2 21 cul ture studies
4. The distance between two countries is calculated as the mean of the distances between the main cities of these countries, each of these dis-
tances being weighted by the size of the towns in question. The measurement of the distance “within” a country is calculated in an analogous
manner, by using the distances between the main cities of the country. The distance between countries leads to cultural separation, and dif-
ferences in commercial contexts that tend to reduce exchanges. Conversely, within a country, proximity between producer and consumer pro-
vides greater mastery of the audience’s expectations and tastes and of the distribution networks, which can, for example, contribute to explain-
ing the success of domestic films.
5. Respectively a Unesco source for films and a Eurostat source for cultural services.
The influence of the flows in each of the cultural
goods categories, taken in isolation for all trade, is
systematically positive and significant. Distance
and the existence of a common border retain a sig-
nificant influence. It also appears that each category
of cultural goods taken individually captures part of
the effect of the other categories, which testifies to
the existence of a relationship between sectors.
Again, the influence of visual arts flows is clearly
superior to that of the other six categories.
Finally, we can put forward “memory” or hys-
teresis effects, meaning a positive and significant
influence of past cultural exchanges on present
global exchanges, and this for each of the seven cat-
egories of cultural goods considered. This witnesses
to the weight of habit, even of addiction phenom-
ena in cultural preferences and consumption.
Cultural exchanges, foreign investments
and migratory fluxes
We can also examine the influence of exchanges
of cultural goods on foreign direct investment flows
(FDI), an influence that can possibly be explained
by the fact that the latter supply information,
notably on distribution networks within the coun-
tries and on consumer preferences, which condition
businesses’decisions in investment matters. Use of
the data supplied by the OCDE on bilateral FDI
flows for the period 1980-2001, covering 59 coun-
tries, shows that distance has a negative influence
whereas flows of cultural goods have a positive
influence on them, however, this is slightly weaker
than that previously recorded for cultural
exchanges. We also observe that cinema film
exchanges have a positive influence on them, but
less significant. In contrast, the impact of cultural
services trade on FDI is not significant.
If we consider the combined but individualized
effect of the flows of each cultural goods category,
only three of them have a positive and significant
effect on FDI: Newspapers and periodicals,
recorded media and visual arts. For the first two cat-
egories the particularly intense effect is greater than
the effect on trade flows. Taken individually, the
flows of each of the cultural goods categories have
a positive and significant impact with the exception
of heritage and audiovisual media goods whose
influence is not significant.
Finally, following work undertaken in other dis-
ciplines, in anthropology for example, the same
type of gravity approach can be applied to the
analysis of migratory flows by examining how the
flow of immigrants depends on the economic size
of the country, the bilateral distance, the sharing of
a border, a language, the existence of colonial links,
etc. The use of Eurostat data for the period 1985-
2004 shows that colonial links and the sharing of a
language have a significant positive impact. The
negative impact of distance on migratory flows
seems to be less robust than that observed on trade
and investment flows. As for the impact of
exchanges of cultural goods on migratory flows, it
appears to be weaker than that observed for trade
and FDI flows. The separate introduction of flows
for the seven cultural goods categories confirms this
lower impact because only newspapers and period-
icals have a significant positive influence.
Applying models and methods used to study
global exchanges to exchanges of cultural goods
and services thus appears to be both possible and
fruitful. The results obtained stimulate thought on
the way in which the cultural interferes with the
economic in a context of continuing opening of
exchanges and economic globalization but also of
the perpetuation of specific features and the main-
tenance of diversity.
22 Focuses on cultural topics: economic approaches
2007-2 23 cul ture studies
Status of research
Until now, in the rare works devoted to exchanges of cultural
goods and services, certain international economic
researchers have used the usual explanations of interna-
tional trade to explain the exchanges of reproducible cultural
goods – recorded music, books, films for example – which,
a priori, do not behave differently from other goods in inter-
national exchanges. Nevertheless, these exchanges will
depend on acculturation and cultural affinity phenomena that
are constructed over time between the countries, with the
result that past flows positively influence current flows. This
tends to reinforce the position of the countries that currently
dominate exports. Sharing an official language strongly
encourages exchanges because, in this case, all other things
being equal, the intensity of the flows is multiplied by a fac-
tor of 4.38. Similarly, a positive effect for language, educa-
tion and religion on the export of American films can be
demonstrated.
More recently, theoretical studies have been interested in the
impact of free exchange of cultural goods and services in
terms of social well being
1
. Cultural identity is considered as
the result of interaction between individual consumption deci-
sions. The differences in consumer preferences combined
with differences in the posture of countries with regard to the
production of goods, supposing large fixed costs can lead to
a reduction in cultural diversity when exchanges are opened,
and, in the case of cultural goods, compensate for the ben-
eficial effects of free exchange. Trade barriers can therefore,
under certain conditions, increase the social well being of
the two partner countries in the exchange and so find a the-
oretical economic basis.
Two main reasons explain the rarity of work:
– the definition of the cultural field, still in question. Follow-
ing a recent Unesco
2
report, a list of seven categories of
goods and services that could be identified in the main
international trade and production classifications was
selected (see p. 20 here);
– the lack of detailed data for sufficiently long periods for
many countries. This led to the use of vary varied sources
(The United Nations COMTRADE database, the Unesco
database, the AUVIS-Eurostat database), with the real ori-
gin of the products taken into account to a greater or lesser
exent
3
.
These reasons add to the question: “Is it possible to explain
cultural exchanges in the same way as non cultural
exchanges?” The new theory of international trade consti-
tutes an explanatory framework for the trade in reproducible
cultural goods: These goods, like many others, are charac-
terized by the presence of economies of scale and notable
product differentiation. This is the framework that has been
selected here.
Questions of method
4
To perform the research whose main results are presented
here, CEPII has made use of the latest developments in
econometric techniques as generally applied to large data-
bases, the latter sometimes being unsatisfactory when only
cultural exchanges are observed. The results are based on
the economic estimation of relations between variables
explained and explanatory variables, relations that are con-
sidered as statistically plausible – referred to as “significant” –
and whose direction (positive, negative, or neither one nor
the other) and size can be determined. This has allowed both
individualization of the effects by reasoning that “all other
things are equal” and an assessment of the explanatory
power of the variables with respect to each other. However,
these probable effects, impacts and influences always need
to be explained and interpreted
5
.
The “gravity equation”
The gravity equation, so called because it is based on an
analogy with Newton’s law – the attraction between bodies
is proportional to their respective masses and inversely pro-
portional to their distance –, is a standard tool used for sev-
eral decades by economists to study the determining factors
in exchange flows.
The hypothesis is: exchanges between two partner countries
depend on their respective economic sizes and the geo-
graphical distance separating them, a distance that serves
as an approximation to transport costs and more generally
to the trade costs that affect exchanges.
Other variables are generally added to take account of spe-
cific features of the bilateral relationship such as the sharing
of a land border, a common language, one or both partners
STATUS OF RESEARCH AND QUESTIONS OF METHOD
1. In economics, traditionally, social well being is measured by the sum of “producers surplus” (The difference between what they obtain for the goods and serv-
ices that they produce and the cost involved in offering them) and the “consumer’s surplus” (The difference between what they are disposed to pay for the exist-
ing goods and services and what they have to pay to benefit from them in reality).
2. Unesco, Échanges internationaux d’une sélection de biens et services culturels 1994-2003 [International exchanges of a selection of cultural goods and serv-
ices 1994-2003], décembre 2005.
3. Taking outsourcing phenomena into account.
4. Certain methodological aspects, such as evaluation of the sources and the selection of regression techniques, that have held an important part in this research
are not described in the present document.
5. Several assertions in the text correspond to the results of academic work for which references are given in the complete text of the report. Seehttp://www.cul
ture.gouv.fr/deps.
24 Focuses on cultural topics: economic approaches
belonging to a preferential trade agreement, the existence of
former colonial links… These variables, which account for
the different dimensions of cultural proximity between coun-
tries, are considered as of little importance for exchanges in
general because they are likely to have a significant effect
on exchanges of cultural goods and services.
This modelling is based on a representation of the economy
and exchanges where each producer offers a unique variety
of a goods that competes with all other varieties of good pro-
duced in the world, knowing that consumers want lesser
quantities of high priced goods but globally show a certain
taste for diversity. This taste comes from the individual
behaviour of consumers who wish to diversity their con-
sumption but also from the diversity of individual preferences.
So each country wants to consume all the varieties produced
in the world, the exact proportion of this demand being a
function of the size of production in the country of origin: Pro-
duction giving rise to increasing gains from scale, the largest
countries in terms of population will be able to offer a larger
number of varieties. The volume of exchanges of an export-
ing country will also depend on the bilateral transaction cost,
understood in the widest sense. The distance, common lan-
guage and other variables also come into play here, which
is shown by the different access costs for varieties from dif-
ferent countries: For example it will be easier for English con-
sumers to see an American film than a Mexican film, this will
not only be due to translation costs; it will also be easier to
see a Spanish film than an Indian or Chinese film, because
the cultural references will, in general, be closer. Finally, con-
sumers being held to assign a relatively constant part of their
consumption to the goods in question, exports will be pro-
portional to the size of the destination market and the pur-
chasing power and so to the development level of the differ-
ent importing countries.
doc_460464418.pdf
The consumption of cultural goods and services is an important indicator of tastes and preferences, and from these, the influence of global exchange flows between nations.
The consumption of cultural goods and services is an important
indicator of tastes and preferences, and from these, the influence of
global exchange flows between nations. If the international per-
formance of cultural industries goes together with more general eco-
nomic performance on the international markets, this makes these
cultural industries an issue that goes beyond culture alone.
In a context where these international exchanges – economic on
the one hand, artistic and cultural on the other, are developing and
presenting a new geographic configuration, these challenges must
be taken up. Certain countries, such as China and India in fact
occupy an increasing place in economic exchanges, and exchanges
of services, after those of goods, continue to increase every year.
Examining the determining factors for cultural exchanges and the
links between them and the notion of cultural proximity between
countries is the objective of the research presented here to encour-
age all the levers affecting exchanges of cultural goods and services
being taken into account.
To analyse questions of distance and cultural proximity in
exchanges, the study is based on the so-called “gravity” model, cur-
rently used in the international economy. This methodology allows
factors in exchanges of cultural goods to be determined at both
global and sector levels (see box p. 23). It also makes the best use
of all the available databases to obtain the same type of results for
cinema and cultural services.
It seeks to identify any effects arising from the existence of bor-
ders in cultural exchanges, to see how much cultural exchanges could
constitute indicators of cultural proximity in both commercial
exchanges and in other important aspects of international economic
relations: Foreign investments and migratory fluxes.
1. This summary has been prepared by François Rouet (Study manager at DEPS), from research undertaken within CEPII by Thierry Mayer
(CEPII, University of Paris I, Paris School of Economics and CEPR), Anne-Celia Disdier (INRA, UMR public economics), Silvio Tai (Uni-
versity of Paris I and Paris School of Economics) and Lionel Fontagné (University of Paris I, Paris School of Economics and CEPII).
2007-2
182, rue Saint-Honoré, 75033 Paris cedex 01
01 40 15 79 13 – 01 40 15 79 99 Available to download from:http://www.culture.gouv.fr/deps
International trade flows
in cultural goods and services:
Issues and determining factors
1
cul ture
studies
PRODUCTION, DISTRIBUTION AND MARKETS
Publishing director: Philippe Chantepie, Head of the Département des études, de la prospective et des statistiques (DEPS) 2007-2 – September 2007
Head of publications: Jacqueline Boucherat
Foreword
Without sources and weighting for
exchange factors, the measurement
and understanding of international
exchanges of cultural goods and serv-
ices are difficult and consequently
rarely undertaken exercises. However,
it is essential to identify the determin-
ing factors for the exchange of cultural
products and services if we want to out-
line a strategy for introducing French
players into international exchanges
and more widely, to defined conditions
for cultural diversity. The results of
research undertaken by a team of
researchers associated with CEPII
who made use of econometric methods
on expertly evaluated sources, consti-
tute a solid base for the definition of
such a strategy from the determining
factors in international exchanges of
cultural goods and services.
P. C.
Secrétariat général
Délégation
au développement
et aux affaires
internationales
Département
des études,
de la prospective
et des statistiques
CULTURAL EXCHANGES:
DISTANCE AND DETERMINING FACTORS
The gravity model allows the examination of
questions of distance and cultural proximity in
exchanges and the determination of factors affect-
ing them at both global and sector levels (see box
p. 23).
Distance and cultural proximity
questions
Different forms of proximity are thus likely to
exist between countries and to affect international
trade. Among them, the fact of sharing a language
and former colonial links, have been the subject of
a large number of research projects.
Linguistic proximity tends to favour trade
between countries: Having a common language
increases the flow of exchanges by about 65%.
Attempts to measure the linguistic distance sepa-
rating two commercial partners – from the absence
of a common language to perfect linguistic simi-
larity – have shown that the greater this distance,
the more trade decreases. However, it is noted that
the influence of this distance has reduced by a third
between 1971 and 1985: All other things being
equal, trade between two countries that are com-
pletely dissimilar from the linguistic point of view
reached 28% of the trade between two perfectly lin-
guistically similar countries in 1971, whereas in
1985, the difference was 48%.
Relations created between countries in the
course of history, in particular in the colonial
period, can also encourage international trade, com-
mercial relations established by colonizing coun-
tries with the colonized countries often persisting
after the colonial period. In addition, with the col-
onized country often having adopted the institu-
tional framework of the colonizing country, the
resulting similarity of legal rules and administrative
systems allows an increase in the security of trans-
actions and a reduction in communication costs,
which tends to favour international trade. We thus
understand why a former colony can retain a higher
than expected
2
level of trade with its former colo-
nizer, but also why two countries, which had the
same colonizer, could also have more developed
bilateral exchange flows than expected. Other
things being equal, the existence of a colonial rela-
tionship in the past multiplies bilateral trade by a
factor of 5.75, whereas the fact of having had the
same colonizing country increases bilateral
exchanges between two countries by 80%. How-
ever, this influence itself is clearly reducing with
time.
These preferred relations between countries,
favouring exchanges, can arise from the existence
of network effects on trade, whether these are busi-
ness networks or social networks. In this respect,
the volume of migratory flows can influence bilat-
eral trade. In fact, immigrants take part in different
ways in the development of trade between their
country of origin and their host country: By retain-
ing, at least in part, a preference for the goods pro-
duced in their country of origin, which leads to
additional demand; by distributing these goods
within the host country; by being particularly able
to establish commercial relations with their coun-
try of origin because their mastery of the language,
the culture and legal practices allows them to be
more in touch with opportunities and needs, and
better connected to the business networks. Finally,
more generally, we have been able to bring out that
the more the citizens of a country felt confident
about the citizens of another country, the greater the
economic exchanges between the two countries,
and in lower proportions, investments in the other
countries.
All these factors are likely to have a dual influ-
ence on exchanges: First in reducing costs, but also,
more indirectly, in influencing cultural preferences.
From which we have the idea of using the
exchanges of cultural goods in which these cultural
preferences are particularly expresses as a measure
of the cultural proximity of the countries and an
approximation of these bilateral preferences (see
“Cultural exchanges: Cultural proximity indicators
and a factor explaining all exchanges”, p. 20).
Determining factors in exchanges
of cultural goods…
At global level
For all goods, bilateral exchanges are propor-
tional to the respective gross national products
(GNP) of the two countries and are larger the higher
the level of development. Colonial links, contiguity
and a common language all have a positive effect,
but the effect of geographic distance itself is nega-
tive. The results for the subset of the seven major
18
2. “Expected” in the sense of “predicted” by the gravity model meaning given the respective economic sizes of the commercial partners and
the distance separating them.
Focuses on cultural topics: economic approaches
2007-2 19 cul ture studies
cultural goods sectors (heritage, books, newspapers
and periodicals, other printed matter, recorded
media, visual arts, audiovisual media) show that the
fundamental determining factors for exchanges of
these goods are imperfectly taken into account by
simple gravity, in particular by GNP and GNP/head.
But if we introduce “fixed effects”, which take
account of the specific features (supposed constant
over time) of each country importing and exporting
cultural goods, the estimate is greatly improved,
which proves the importance of these interrelations
and the impact of specific features between coun-
tries. The impact of distance is however lower for
cultural goods than for all goods, which can be
explained by a larger “cultural” separation that does
not compensate for possibly lower transport costs.
The fact that transaction costs have a lower neg-
ative impact on trade for cultural goods than for
other assets appears to be a very solid result. In con-
trast, the GNP of the exporting country has a posi-
tive impact on cultural trade, with colonial links
increasing bilateral trade by 33%, a common bor-
der by 85% and a common language by 56%, with
this last impact being greater for cultural goods.
At the sector level
An attempt to explain the volume of bilateral
cultural exchanges has also been tried for each of
the seven main categories of cultural goods. We
note that the “distance” variable plays a negative
role in all cases, but that, in contrast, the develop-
ment level of the partner countries measured
through GNP/head does not play a significant role
in explaining the volume of bilateral exchanges,
except for the heritage goods sector.
Two influences – that of GNP and that of vari-
ables related to cultural proximity (colonial links,
common border and common language) – differ
between sectors: Whereas colonial links positively
influence exchanges in the heritage goods, news-
paper and periodical and visual arts sectors, their
influence is, rather surprisingly, negative for audio-
visual media. For the “books, newspapers and peri-
odicals and other printed matter” sectors, the shar-
ing of a border and/or a language encourages trade.
Unlike recorded media, the visual arts and audiovi-
sual media are not influenced by sharing an official
language.
A common border has a positive impact on the
flows in the visual arts and audiovisual media sec-
tors but does not influence the recorded media sec-
tor. The common language is the main factor in the
penetration of cultural goods with written content
– this factor multiplies the flow of books by over
4.5 – and the colonial has certainly shaped the pref-
erences of consumers in the importing countries for
heritage goods where it multiplies the flow by 3.74.
For the “distance” variable, the calculation of the
average distance separating partners in the
exchange according to the types of goods shows
that cultural goods are exchanged over smaller dis-
tances than non-cultural manufactured goods
(respectively 6,005 km against 7,271). In addition,
if we apply this calculation to each of the seven cat-
egories of cultural goods, it appears that it is for her-
itage goods that the partners in the exchange are on
average the furthest apart from each other
(6,616 km): Essentially comprising irreproducible
artistic goods, this category depends less on dis-
tance than other cultural goods and is thus less sen-
sitive to transport costs and more generally to trans-
action costs. The fact that the influence of a com-
mon border is not significant for this category of
goods adds strength to this observation.
The cinema case
The Unesco data for the period 1970-1999, cover bilateral imports
of films from the main producing countries (United States, France,
Germany, Russia, Japan, India, Hongkong, United Kingdom and
Italy) but also for national production. They allow us to establish
that distance plays the same role as in the previous estimates:
The flows of films are less affected by distance than flows for the
rest of manufactured goods, and confirm the fundamental weight
of a common language and a colonial link. However, specific
national features also come into play and generate “fixed effects”
that can be assessed relatively: So, some importing countries,
with a special taste for the cinema, import many films of all ori-
gins in contrast to others that are poorly equipped with cinemas
or distribution networks, due, for example, to very low density and
large spatial dispersal of the population.
The fact of having a common language or a colonial link increases
the market share of the exporter concerned by 70% in countries
that share one or the other. In this respect, the French speaking
community has a real economic advantage in terms of cinemato-
graphic exchanges: France, like Great Britain and the United
States in particular, benefits in general from links created by his-
toric population movements, the similarity of institutions and cul-
tural practices. If the effects of the “colonial link” and “linguistic link”
variables can be very clearly separated for France, they are cumu-
lative in most cases: In fact we note that in a former French colony
where French is still spoken, the market share of French films is
multiplied by three compared to a country where the language is
not French and that has never been a French colony.
… and cultural services
As far as dematerialized services and exchanges
in general are concerned, which do not need phys-
ical contact between the consumer and the producer
and do not give rise to transport costs, what does
the concept of “distance” mean? However, despite
this, most studies show a significant and relatively
large negative effect of distance on the flow of serv-
ices. The explanation is perhaps that distance meas-
ures similarities in consumer preferences, similari-
ties that are larger as the countries are closer on the
geographic level.
At the start of the 2000 decade, the flow of serv-
ices represented more than 20% of international
exchanges, a share that should continue to grow
when the development of ICT is taken into account.
Despite their significant and increasing weight in
exchanges and whatever the services – cultural or
not -, these transactions are poorly recorded and we
have very few statistics
3
due to the lack of a stan-
dardized nomenclature. In world exchanges of serv-
ices, “audiovisual and related services” and “other
royalties and licence fees” still only represent a
minimal but strongly growing share.
If the effect of the “distance” variable is still neg-
ative for these two types of services, it is however
weaker than for all services. In addition, all the
other “gravitational” variables (see above) have the
expected effect: The size of the country has a pos-
itive influence on exchanges of cultural services,
although less strongly, here too, than for all
exchanges of services; the flows also increase with
the countries’development level evaluated by
GNP/head; colonial links and sharing a terrestrial
border strongly favours trade in cultural services
whereas this positive and significant influence of
colonial links is not observed for all services;
finally, sharing an official language does not appear
to influence the flow of services, whether in terms
of all services or for cultural services.
As far as the GNP and GNP/head of the export-
ing country are concerned, their time variations do
not appear to influence exchanges of cultural serv-
ices. In contrast, those of the importing country do
have an impact on exchanges, positive for the first
and negative for the second. In fact it seems that,
following an improvement in their development
level, countries import fewer cultural services, the
development of local production tending progres-
sively to replace imports. Exchanges of cultural
services, contrary to what happens for services in
general, tend to reduce when a common border is
shared; but they nevertheless remain positively
influenced by a the sharing of an official language.
The common language thus seems to be one of the
most robust influencing variables for cultural
exchanges. This variable has an almost systematic
very clear positive impact when the specific features
of the importing and exporting countries are taken
into account.
CULTURAL EXCHANGES:
CULTURAL PROXIMITY INDICATORS
AND A FACTOR EXPLAINING
ALL EXCHANGES
Cultural exchanges appear to constitute, through
the explanation of crossing a border on exchanges,
indicators of the cultural proximity between coun-
tries. In addition, cultural exchanges constitute a
particularly strong explanatory factor for the dif-
ferent aspects of international exchanges (goods and
services, foreign investments, migratory flows).
Border effects for cultural goods
and services
Another way of envisaging international
exchanges consists in asking the question of “the
border effect”, meaning the effect of crossing
national borders on exchanges. Research on the
subject of borders, both between Canada and the
United States and within the European Union sug-
gest fairly high border effects despite the process of
economic integration and sometimes a common
language. These effects are only partially explained
by formal commercial barriers to exchanges (cus-
toms duties…), there thus remains a place for cul-
tural differences that drive individuals to grant a
certain preference – price being equal – to national
products. Measuring border effects on cultural
goods and services partly allows measurement of
these cultural differences. If the impact of national
borders comes from cultural differences between
countries, we must thus find a border effect in
exchanges of cultural goods, which can then
become a sort of “tracer” for cultural preferences.
The Eurostat AUVIS database, which supplies
the number of admissions achieved by films pro-
duced by one country in each of the other countries
in the period 1980-2001, allows the border effect to
be estimated subject to defining a measure of the
20
3. Several databases (OCDE, Eurostat, CEPII) supply data from the same source: the balances of payments of different countries.
Focuses on cultural topics: economic approaches
“internal” distance within the country
4
. We observe
that the border effect is very important: On average,
admissions for national films are nearly ten times
as high as the expected value. This demonstrated
the fact that despite the development of interna-
tional exchanges and the rise in power of the Amer-
ican cinematographic industry on the world scale,
the “domestic bias”, meaning the preference for
national production in consumption of the cultural
good that the cinema comprises, is high. The
hypothesis in which cultural differences are one of
the probable causes of the existence of the effect of
national borders on general trade in goods is thus
found to be reinforced.
This border effect, in view of the results of pre-
vious research, appears very comparable to that for
non-cultural goods taken as a whole and even, par-
adoxically, clearly less than that estimated from
food goods or drinks, non-cultural goods that nev-
ertheless include a large cultural component. The
explanation, in the cinema case, could be that the
market, even if it is far from being completely inte-
grated, is much more so than for other cultural
goods.
Finally, the relative preference of French audi-
ences for national films, or put another way, the
level of “closure” of the French market, is not
greater than for other countries. On the contrary, the
French market is actually slightly more open than
other markets. This can be surprising on first
encounter, in that France is one of the rare countries
where the domestic cinema’s market share remains
large, but this result is explained by the fact that
production is large, which largely justifies the
respectable part that French films occupy in the
internal market. Having regard to a border effect for
France relatively close to that obtained for the
United States, we can consider that public support
policies do not lead to distortions in the consump-
tion of films by French audiences. The existence of
a high volume of production encouraged by the sup-
port systems does not imply a lower level of foreign
film imports but rather a higher total level of film
consumption, from which foreign films benefit.
Can exchanges of cultural goods explain
commercial flows?
How far do exchanges of cultural goods go
together with all exchanges, for which they never-
theless represent, on average, only 1.4% of the total
for the period 1988-2004? Use of the data in the
United Nations Comtrade database allows an esti-
mate that trade in cultural goods positively and sig-
nificantly influences exchanges of all goods. The
use of this explanatory factor as a direct measure of
cultural proximity between countries, also leads to
rendering the influence of the factors usually used
less significant (sharing a common language, dis-
tance, GNP, development level…).
If we consider the impact of all seven categories
of cultural goods individually, we observe that the
flow of five of them exercises a significant positive
effect on total trade (with the exception of the flow
of heritage goods which has a negative influence
and exchanges of other printed matter which does
not have a significant effect). In total, an increase
of 10% in exchanges of cultural goods increases
trade in traditional goods from 3.25 to 4.25% and
exchanges of visual arts alone, in increasing by
10%, increase total trade by nearly 2%. If we limit
ourselves only to bilateral exchanges of films or
cultural services
5
, the positive influence on all trade
is found again even if it is less.
The introduction of cultural goods exchange
flows thus modifies the influence of the traditional
variables relating to cultural proximity (colonial
links, common border, common official language)
in addition to the influence of the “distance” vari-
able. This signifies that there is a consequent inter-
dependence between these variables and the flow
of cultural goods. We can demonstrate, for exam-
ple, that geographic distance, taken as an explana-
tory factor for trade as a whole, partly translates the
cultural distance separating the countries. These
cultural exchange flows explaining the global
exchanges well, this means both that they depend
in part on the cultural proximity of the countries but
also that they take better account of the influence
of cultural proximity on total trade of a country than
the traditionally used measurements (language,
colonial link…).
2007-2 21 cul ture studies
4. The distance between two countries is calculated as the mean of the distances between the main cities of these countries, each of these dis-
tances being weighted by the size of the towns in question. The measurement of the distance “within” a country is calculated in an analogous
manner, by using the distances between the main cities of the country. The distance between countries leads to cultural separation, and dif-
ferences in commercial contexts that tend to reduce exchanges. Conversely, within a country, proximity between producer and consumer pro-
vides greater mastery of the audience’s expectations and tastes and of the distribution networks, which can, for example, contribute to explain-
ing the success of domestic films.
5. Respectively a Unesco source for films and a Eurostat source for cultural services.
The influence of the flows in each of the cultural
goods categories, taken in isolation for all trade, is
systematically positive and significant. Distance
and the existence of a common border retain a sig-
nificant influence. It also appears that each category
of cultural goods taken individually captures part of
the effect of the other categories, which testifies to
the existence of a relationship between sectors.
Again, the influence of visual arts flows is clearly
superior to that of the other six categories.
Finally, we can put forward “memory” or hys-
teresis effects, meaning a positive and significant
influence of past cultural exchanges on present
global exchanges, and this for each of the seven cat-
egories of cultural goods considered. This witnesses
to the weight of habit, even of addiction phenom-
ena in cultural preferences and consumption.
Cultural exchanges, foreign investments
and migratory fluxes
We can also examine the influence of exchanges
of cultural goods on foreign direct investment flows
(FDI), an influence that can possibly be explained
by the fact that the latter supply information,
notably on distribution networks within the coun-
tries and on consumer preferences, which condition
businesses’decisions in investment matters. Use of
the data supplied by the OCDE on bilateral FDI
flows for the period 1980-2001, covering 59 coun-
tries, shows that distance has a negative influence
whereas flows of cultural goods have a positive
influence on them, however, this is slightly weaker
than that previously recorded for cultural
exchanges. We also observe that cinema film
exchanges have a positive influence on them, but
less significant. In contrast, the impact of cultural
services trade on FDI is not significant.
If we consider the combined but individualized
effect of the flows of each cultural goods category,
only three of them have a positive and significant
effect on FDI: Newspapers and periodicals,
recorded media and visual arts. For the first two cat-
egories the particularly intense effect is greater than
the effect on trade flows. Taken individually, the
flows of each of the cultural goods categories have
a positive and significant impact with the exception
of heritage and audiovisual media goods whose
influence is not significant.
Finally, following work undertaken in other dis-
ciplines, in anthropology for example, the same
type of gravity approach can be applied to the
analysis of migratory flows by examining how the
flow of immigrants depends on the economic size
of the country, the bilateral distance, the sharing of
a border, a language, the existence of colonial links,
etc. The use of Eurostat data for the period 1985-
2004 shows that colonial links and the sharing of a
language have a significant positive impact. The
negative impact of distance on migratory flows
seems to be less robust than that observed on trade
and investment flows. As for the impact of
exchanges of cultural goods on migratory flows, it
appears to be weaker than that observed for trade
and FDI flows. The separate introduction of flows
for the seven cultural goods categories confirms this
lower impact because only newspapers and period-
icals have a significant positive influence.
Applying models and methods used to study
global exchanges to exchanges of cultural goods
and services thus appears to be both possible and
fruitful. The results obtained stimulate thought on
the way in which the cultural interferes with the
economic in a context of continuing opening of
exchanges and economic globalization but also of
the perpetuation of specific features and the main-
tenance of diversity.
22 Focuses on cultural topics: economic approaches
2007-2 23 cul ture studies
Status of research
Until now, in the rare works devoted to exchanges of cultural
goods and services, certain international economic
researchers have used the usual explanations of interna-
tional trade to explain the exchanges of reproducible cultural
goods – recorded music, books, films for example – which,
a priori, do not behave differently from other goods in inter-
national exchanges. Nevertheless, these exchanges will
depend on acculturation and cultural affinity phenomena that
are constructed over time between the countries, with the
result that past flows positively influence current flows. This
tends to reinforce the position of the countries that currently
dominate exports. Sharing an official language strongly
encourages exchanges because, in this case, all other things
being equal, the intensity of the flows is multiplied by a fac-
tor of 4.38. Similarly, a positive effect for language, educa-
tion and religion on the export of American films can be
demonstrated.
More recently, theoretical studies have been interested in the
impact of free exchange of cultural goods and services in
terms of social well being
1
. Cultural identity is considered as
the result of interaction between individual consumption deci-
sions. The differences in consumer preferences combined
with differences in the posture of countries with regard to the
production of goods, supposing large fixed costs can lead to
a reduction in cultural diversity when exchanges are opened,
and, in the case of cultural goods, compensate for the ben-
eficial effects of free exchange. Trade barriers can therefore,
under certain conditions, increase the social well being of
the two partner countries in the exchange and so find a the-
oretical economic basis.
Two main reasons explain the rarity of work:
– the definition of the cultural field, still in question. Follow-
ing a recent Unesco
2
report, a list of seven categories of
goods and services that could be identified in the main
international trade and production classifications was
selected (see p. 20 here);
– the lack of detailed data for sufficiently long periods for
many countries. This led to the use of vary varied sources
(The United Nations COMTRADE database, the Unesco
database, the AUVIS-Eurostat database), with the real ori-
gin of the products taken into account to a greater or lesser
exent
3
.
These reasons add to the question: “Is it possible to explain
cultural exchanges in the same way as non cultural
exchanges?” The new theory of international trade consti-
tutes an explanatory framework for the trade in reproducible
cultural goods: These goods, like many others, are charac-
terized by the presence of economies of scale and notable
product differentiation. This is the framework that has been
selected here.
Questions of method
4
To perform the research whose main results are presented
here, CEPII has made use of the latest developments in
econometric techniques as generally applied to large data-
bases, the latter sometimes being unsatisfactory when only
cultural exchanges are observed. The results are based on
the economic estimation of relations between variables
explained and explanatory variables, relations that are con-
sidered as statistically plausible – referred to as “significant” –
and whose direction (positive, negative, or neither one nor
the other) and size can be determined. This has allowed both
individualization of the effects by reasoning that “all other
things are equal” and an assessment of the explanatory
power of the variables with respect to each other. However,
these probable effects, impacts and influences always need
to be explained and interpreted
5
.
The “gravity equation”
The gravity equation, so called because it is based on an
analogy with Newton’s law – the attraction between bodies
is proportional to their respective masses and inversely pro-
portional to their distance –, is a standard tool used for sev-
eral decades by economists to study the determining factors
in exchange flows.
The hypothesis is: exchanges between two partner countries
depend on their respective economic sizes and the geo-
graphical distance separating them, a distance that serves
as an approximation to transport costs and more generally
to the trade costs that affect exchanges.
Other variables are generally added to take account of spe-
cific features of the bilateral relationship such as the sharing
of a land border, a common language, one or both partners
STATUS OF RESEARCH AND QUESTIONS OF METHOD
1. In economics, traditionally, social well being is measured by the sum of “producers surplus” (The difference between what they obtain for the goods and serv-
ices that they produce and the cost involved in offering them) and the “consumer’s surplus” (The difference between what they are disposed to pay for the exist-
ing goods and services and what they have to pay to benefit from them in reality).
2. Unesco, Échanges internationaux d’une sélection de biens et services culturels 1994-2003 [International exchanges of a selection of cultural goods and serv-
ices 1994-2003], décembre 2005.
3. Taking outsourcing phenomena into account.
4. Certain methodological aspects, such as evaluation of the sources and the selection of regression techniques, that have held an important part in this research
are not described in the present document.
5. Several assertions in the text correspond to the results of academic work for which references are given in the complete text of the report. Seehttp://www.cul
ture.gouv.fr/deps.
24 Focuses on cultural topics: economic approaches
belonging to a preferential trade agreement, the existence of
former colonial links… These variables, which account for
the different dimensions of cultural proximity between coun-
tries, are considered as of little importance for exchanges in
general because they are likely to have a significant effect
on exchanges of cultural goods and services.
This modelling is based on a representation of the economy
and exchanges where each producer offers a unique variety
of a goods that competes with all other varieties of good pro-
duced in the world, knowing that consumers want lesser
quantities of high priced goods but globally show a certain
taste for diversity. This taste comes from the individual
behaviour of consumers who wish to diversity their con-
sumption but also from the diversity of individual preferences.
So each country wants to consume all the varieties produced
in the world, the exact proportion of this demand being a
function of the size of production in the country of origin: Pro-
duction giving rise to increasing gains from scale, the largest
countries in terms of population will be able to offer a larger
number of varieties. The volume of exchanges of an export-
ing country will also depend on the bilateral transaction cost,
understood in the widest sense. The distance, common lan-
guage and other variables also come into play here, which
is shown by the different access costs for varieties from dif-
ferent countries: For example it will be easier for English con-
sumers to see an American film than a Mexican film, this will
not only be due to translation costs; it will also be easier to
see a Spanish film than an Indian or Chinese film, because
the cultural references will, in general, be closer. Finally, con-
sumers being held to assign a relatively constant part of their
consumption to the goods in question, exports will be pro-
portional to the size of the destination market and the pur-
chasing power and so to the development level of the differ-
ent importing countries.
doc_460464418.pdf