Vi Digital Library - Text Preview
Social Interactions of Migrants and Trade Outcomes
Working paper by Tai, Silvio H. T., 2009
Download original document (English)
Staff Working Paper ERSD-2009-02 January 2009
World Trade Organization
Economic Research and Statistics Division
Social Interactions of Migrants and
Silvio H. T. Tai
World Trade Organization – PhD Program and
Université Paris I Panthéon–Sorbonne
Manuscript date: January 2009
Disclaimer: This is a working paper, and hence it represents research in progress. This
paper represents the opinions of the author, and is the product of professional research. It is
not meant to represent the position or opinions of the WTO or its Members, nor the official
position of any staff members. Any errors are the fault of the author. Copies of working
papers can be requested from the divisional secretariat by writing to: Economic Research and
Statistics Division, World Trade Organization, Rue de Lausanne 154, CH 1211 Geneva 21,
Switzerland. Please request papers by number and title.
Social Interactions of Migrants and
Silvio H. T. Tai **
This paper investigates the social interactions performed by immigrants in France. A framework for
immigrant’s choice of location is based on recent studies on non-market interactions which explains
how migrants concentrate. Applying data on the distribution of immigrants in 95 French provinces,
the social interactions are subsequently estimated. This “social component” of migration is then tested
on international trade, providing a direct measure of the impact of social networks on the economy.
JEL classification: F10, F22.
Keywords: Social Interaction, Non-market interaction, International Trade, Migration
* Thanks are extended to: Amy Bisno, Fabrice Defever, Frédéric Docquier, Christopher Hearle, Arastou
Khatibi, Tobias Müller, Marcelo Olarreaga, Nicolas Maystre, Alissa Petee, Roberta Piermartini, Michele Ruta
and the participants of the “Jeunes Chercheurs” Seminar of University of Geneva
** Université de Genève / CEPR - Marie Curie Research Training Network – TOM, World Trade Organization
– PhD Program, Université Paris I Panthéon–Sorbonne
Disclaimer: This paper represents the opinions of the author. It is not meant to represent the position or opinions
of the WTO or its Members, nor the official position of any of its staff members.
An anecdotic fact can illustrate the role of immigrant networks in their location and economic
activities. I describe a family in Brazil that has been immigrating from China for four generations.
People of each generation were born in China and moved to Brazil. The first generation immigrated to
the city of Vitória in the Southeast Brazil. The second generation has chosen the economically
booming city of São Paulo also in the Southeast Brazil. The network structure established from the
first generation in Vitória facilitated the settlement in São Paulo, 882 km 1 from Vitória. The third
generation divided between São Paulo and Curitiba due to São Paulo becoming saturated and business
opportunities being sparse. In contrast, Curitiba had become a rapidly developing city in Brazil. Some
3230 km separate São Paulo from Curitiba, but the social interactions meant that this location choice
was apt. The fourth generation immigrated at a very young age joining their parents in São Paulo.
Some stylized facts can be drawn from this episode. Firstly, new immigrants depend on the
compatriot's community, since they face barriers of language, habits and culture. Job, loans, associates
and relationships are found in this specific community. Secondly and conversely, as time passes, and
immigrants are more integrated into Brazilian culture, they become less dependent to a specific
community. Generations born in Brazil become further disconnected to the Chinese community as can
be seen in their choice of more diversified professions and lifestyle. Thirdly, the network effect
determining location choice is robust with different economical contexts. Since the first migration in
the beginning of the XXth century, Brazil has been experiencing increasingly diverse situations and
has responded to particular economic, social and political shocks. Nevertheless, migration movements
have remained constant due to an ever existing network of ties, bonds and connections.
These interdependences are reflected in the distribution of people. Immigrants have a tendency to
cluster at higher densities relative to local population. Figure 1 shows the over-concentration of
immigrants in France, in comparison to the French population as a whole: 70% of the foreigner
population is more concentrated than the French population2. While there could be specific reasons
within a region, this cannot provide a complete explanation for this over-concentration of immigrants.
1 Source: Brazilian National Department of Transport Infra-Structure
2 The foreigner population in France have a coefficient of geographical concentration G=0.24. This positive
concentration coefficient equals 0 if foreigners are distributed exactly the same as the total population.
Chiswick, Lee and Miller (2002): "Where the overseas born group has a distribution across regions the same as the total
population, G equal to 0. Where the overseas born group is completely segregated, the upper bound of G will equal (100-
GS), where GS is the percent of the total population accounted for by the specific birthplace group."
Figure 1 – Foreigners share versus French share, by French “département”
0 .01 .02 .03 .04 .05
Share of Overall French Population
This graph compares the total population of foreigners in France to the total French population. The data is from 1999
census. Eg: 8% of all the foreigners living in France resided in Paris whilst only 3% of French population as a whole did so
in this year.
This paper attempts to formalize these kinds of social interactions within a community. Particularly,
this study addresses the question of how social interactions can impact where immigrants choose to
locate, and how these interactions can develop trade networks
Social networks provide a major explanation for the impact of immigrants on international trade.
James Rauch and co-authors (1998, 1999, 2001, 2002, and 2004) have contributed much to this area
of the literature demonstrating theoretical reasons and empirical evidence for this. However, the
conception of social interaction itself is not based on a formal framework, and a direct measure of
these networks is not provided. This study follows and extends the paper of Herander and Saavedra
(2005) which outlines a conception of this immigration network based on states that share a common
Further studies have recently updated the literature on social interactions. Although not treating these
interactions as determinant of an economical effect, Brock and Durlauf (2001, 2002) succeed in
developing a theoretical model that integrates social relations. Head and Mayer (2008) have extended
this work and have also provided empirical evidence which has confirmed that the choices that
parents make in naming their offspring is influenced by the environment in which they are embedded.
The contribution of this paper to the literature is threefold. Firstly, by building on existing literature, it
develops a framework for immigrants' social interactions which is then used on data on the
distribution of immigrants in France. Secondly, it proposes a functional form to the impact of distance
on social interactions. Thirdly, the aforementioned analyses are applied to international trade and the
economic impacts of social interactions are evaluated.
This paper is structured as follows. Section 2 analyses the relevant literature on social interactions as
well as the nexus between migration and trade. Section 3 presents the theoretical framework while
Section 4 describes the dataset and the variables used. Section 5 reports on the findings before some
tentative conclusions are made in section 6.
2. Relevant Literature
2.1 Previous Literature on Social Interactions
Gary S. Becker has contributed vastly to the literature3 on social interactions. The Nobel Prize winner
of 1992 was awarded to him "for having extended the domain of microeconomic analysis to a wide
range of human behaviour and interaction, including non-market behaviour".4 The book jointly
written with Kevin M. Murphy (Becker and Murphy, 2000) provides a synthesis of existing academic
thought in this field and the link between the social and economic realms is explored further. It
acknowledges that the consumption of common goods can be influenced by a social dimension.
More recently, Brock and Durlauf (2001) have developed theoretical models for studying individual
decisions, including social interaction effects, into the private utility. They model how the dynamics
of a group can influence the decisions and actions of an individual. Building on the literature on
discrete choice, Brock and Durlauf investigate for self-consistent equilibriums which presents a
hyperbolic tangent shape such as found in some specific models in physics such as the Curie-Weiss
model of magnetism.
Manski (1993) relates social interaction to the reflection phenomenon. He develops numerous models
which take three main hypotheses: "(a) endogenous effects, wherein the propensity of an individual to
behave in some way varies with the behavior of the group; (b) exogenous (contextual) effects, wherein
the propensity of an individual to behave as in some way varies with the exogenous characteristics of
the group, and (c) correlated effects, wherein individuals in the same group tend to behave similarly
because they have similar individual characteristics or face similar institutional environments." The
first hypothesis generates a social multiplier effect creating mimetic behavior in a group. The second
3 Scheinkman (forthcoming) , Manski (2000) and Glaeser and Scheinkman (2003) provide surveys of
the literature and discuss different approaches to define and model social interactions
hypothesis distinguishes exogenous and independent determinants of behavior whilst the third relates
to correlated determinants of behavior.
Head and Mayer (2008) analyze the extent of non-market interactions and investigate the social
transmission of parental preferences regarding the naming of their children. They argue that the
frequency of existing child names in the neighborhood can influence parent's choice of what to name
their son or daughter. Head and Mayer use data on the geographic distribution of names in France to
explain the popularity of three types of names: Saints names, Arabic names and American names.
They find that the importance of geographic distance declines over time while differences in class and
national origins have increasing explanatory power.
Glaeser et al. (1996) present a model where, after controlling for economical and social conditions,
social interactions explain the high variance of crime rates across cities. An index of social
interactions is constructed using data from the FBI and the New York City Police Department. Where
the crime is more serious such as in murder, social interactions are less likely to have played a major
role in provoking a criminal act.
Furthermore, several empirical studies point out to the phenomenon of immigrant's concentration.
Bartel (1989) finds that U.S. immigrants are geographically concentrated at higher densities compared
to the American population as a whole. Bartel found that education was significant in explaining the
distribution these immigrants, with increased dispersion being associated with higher levels of
education. Moreover, immigrants were found to internally migrate more frequently relative to the
native population. This finding is confirmed by other studies as Chiswick and Miller (2004),
Chiswick, Lee and Miller (2002), Funkhouser (2000) and Gonzalez (1998). These authors find
numerous determinants that explain the reasons for the existence of ethnic enclaves. Furthermore,
these studies show that the inter-generational transmission of culture and lifestyles choices is
strengthened when there are higher densities of immigrant populations. Quoting Funkhouser (2000,
page 489): "The geographic concentration of immigrants is an important part of the assimilation
process, allowing immigrants to maintain some cultural ties to their country of origin". This network
externality causes migrants to follow the culture and habits of their home country rather than the
country in which they presently live.
Theoretical work from the field of New Economic Geography has provided certain explanations for
the concentration of immigrants. The theoretical Core-Periphery model (Fujita et al. 2000 and
Baldwin et al. 2005) describes a self-reinforcing mechanism of agglomeration. On the one hand, firms
in the presence of transport costs and economies of scale, choose regions where the demand for
4 Definition from the Nobel Prize Foundation
consumer goods is higher. On the other hand, individuals look for a wide range of commodities and a
low cost of living due to lower transport costs. Furthermore, the market crowding effect considers a
dispersion mechanism which reflects the fact that imperfectly competitive firms have a tendency to
locate in regions with relatively few competitors. Crozet (2004) and Poncet (2006) apply models from
New Economic Geography models to empirically discover that market access is significant in
explaining the degree of regional immigration. This approach explains the agglomeration by pure
economical reasons for an entire population, but it overlooks the fact that a concentration of
immigrants is related to non-market reasons as well.
2.2 The Impact of Migration on Trade
The impact of migration on trade has been attested by many studies in the literature5. Several
mechanisms have been presented to justify this association, such as preference and network effects.
Most of the literature treats the preference effect as immigrants’ consumption of home-country
products. Nonetheless, Bowles (1998) considers migration as a way to transmit preferences through
the exposure to different cultures. Tai (forthcoming) also endorses the idea of a cultural transmission
effect through migration, and argues that the personal consumption of immigrants does not fully
explain the implied amounts of trade. The relative impact of preferences and networks on trade has
been covered elsewhere (Tai, forthcoming) and is outside the scope of this paper, which instead will
focus on the network effect.
Another major mechanism underpinning the impact of migration on trade is the network effect. It
relies on two fundamental concepts. Firstly, immigrant communities have the potential to deter
violations of informal contracts. Secondly, immigrants benefit from privileged information on the
home-country and host-country markets.
Existing literature commonly uses the quantity of migrants to ascertain the impact of this network.
Rauch (2002) innovatively used the probability that in any two given countries, both individuals will
be Chinese in origin.
This immigrant's network can be an analysis of the interaction of immigrants amongst themselves in
the host-country as well as acquaintances in the home-country, as Rauch (2002) tests. However, what
is neglected in this paper is an appreciation that networks can also operate within the host-country.
Immigrants living in a certain province have the ability to interact with compatriots living in other
provinces located in the same country. Herander and Saavedra (2005) analyze the impact of
immigrant networks on trade within each state of United States of America. For example, for
Colorado, the authors have estimated how the compatriots residing in the neighboring states of
Wyoming, Nebraska, Kansas, Oklahoma, New Mexico and Utah have an influence on the trade of
Colorado. Yet noteworthy is that this impact is not as great as the effect induced by within-state
immigrants and confirms that intra-national networks do have a significant effect. This effect is
reduced by distance, since immigrants who are located further away from each other have a lower
effect on trade. Nonetheless Herander and Saavedra do not consider a framework or any empirical
support for explaining these networks. Moreover, their concept of a network is limited in that it only
accounts for immigrants residing in neighboring states. Using a specific framework and empirical
evidence, this paper attempts to underpin the eminence of immigrants’ networks thereby extending
the overly simplistic concept of a network that only exists between states that share a common border.
3. Social Interaction Framework
For the sake of clarity time subscripts are omitted. In this theoretical section geographic location is
referred to by the word “region”, responding to either a French “département” or region6.
3.1 Social Interactions Term
The potential social interaction of an individual k can be defined by:
ij sS νln (1)
This social component is based on the actual number of resident compatriots at a particular point in
time. Assuming the inverse relationship between distance and social interaction, the framework used
in this paper takes the function ν for calculating the extent of the network. Figure 2 illustrates the
interactions within a network of foreigner for the periods t and t+1 . In this example the stock of
immigrants living in Paris at the period t+1 is determined by a pre-existent network of immigrants
residing in other French “départements” at the period t.
5 See Rauch (2001) and Wagner (2002) for a synthesis of literature in this area.
6 This is necessary because while migration data is available at “département” level, trade is only
available at a regional level.
Figure 2 – Spatial-Time Interactions of Immigrants
Lille Paris Nice Marseille Bordeaux
The social component of utility is obtained by summing together the weight ν of each immigrant. The
choice to use the inverted distance follows the same functional form of many applications in
economics, such as the market potential. Taking this into account, it can be considered that each
immigrant has the potential to interact with his or her immediate surroundings normalized to 1. The
potential level of interaction is spread along a line that measures 2π times the distance from an
immigrant. Figure 3 shows a point (point A) between two immigrants where 1/2πr +1/2πR is the total
sum of possible interactions that can occur at this locality.
Figure 3 – Spatial Interactions of Migrants
3.2 Migration Function
Adapting Docquier's model7, a framework for migration can be conceived by considering self
selection based on utility differences perceived in both home and host countries.
The costs of migration for an individual k migrating from country of origin i to region j is a function
of moving costs ( )ijd , the possibility of social interactions ( )ijS and characteristics of the home-
country ( )iz . The idea is that a new immigrant could lower his or her costs of installation through the
help of a installed social network, such as support offered for housing, language and bureaucracy. I
assume Ckij homogeneous of degree 1 on Sij.
( )iijijkij zSdcC ,,= (2)
where 0<′dc and 0<′Sc
The average probability this individual, k, having an income in the host country is a function of the
possibility of social interactions ( )ijS and the characteristics of the destination region ( )jz . Resident
compatriots offer a wider variety and more extensive opportunities for job and business development
than the national population. I assume Πkij homogeneous of degree 1 on Sij.
( )jijkij zS ,π=∏ (3)
Social impacts have a positive impact:
The utility of an individual k of nationality i living in region j is:
CwU ε+−∏= (5)
7 This model was presented by Frédéric Docquier at the "Seventh Summer School in International and
Development Economics" of the Marie Curie Research Training Network on "Transnationality of Migrants"
(TOM). The original version also accounts for skill differentiation.
Adopting the appropriated distribution of errors, the share of new immigrants choosing a region j is
expressed by a multinomial logit function:
While multinomial logit functions are a conventional way to estimate location choices, I follow
Guimarães et al. (2003) with the application of a Poisson estimator. This procedure provides a more
tractable model, avoiding the problem of non-linearity.
Then, considering the quantity of immigrants from country i to region j:
with ∑= J
The idea is to estimate these quantities using a Poisson model:
( ) iijjij DCwij eqE απ +−= (8)
The Poisson model's likelihood is expressed as:
( ) ( )[ ]∑∑
ijiijjijiijjp nDCwqDCwL ijij
!logexp απαπ (9)
Considering the first order conditions with respect to αg:
Linking equation (10) to equation (9), the Poisson likelihood becomes:
∑∑∑∑∑ ∑ = === = −
1 111 1
The last two terms of equation (11) are constants. The first term is the log-likelihood of the
conditional logit. Estimated coefficients are the same in both models. Estimated regressions are
proceeded to be done using the Poisson model with quantities as the dependent variable. Independent
variables are taken in logarithmic format to provide elasticities.
However substituting the social term Sij into the migration equation leads to a complex function.
Nonetheless, some algebraic manipulation can simplify this expression in order to obtain an estimable
equation. Social terms can be separated from all other terms because of the homogeneity assumption.
Then the estimate equation becomes:
( ) ⎥⎦
− cteMillsdistanceqq itjtij
1 βνβ (12)
Where jtΩ and itΩ are fixed effects for region-time and country-time.
While country of origin i may be any country in the world, destination regions j are all restricted to
one country, in this case France. This implies a selection bias since each immigrant who had chosen
region j had previously chosen France as a destination country. This bias is corrected by a probit
estimator. The variable Mills is the inverse of the Mills ratio which has been estimated from the first
stage probit8 estimator, following the Heckman (1976) method.
The compatriot's network is hypothesized to capture the structure that resident immigrants can offer to
a new one, such as assistance with bureaucracy, language, housing, employment and business
opportunities, access to home produce and leisure. This phenomenon can operate at a distance from
one region to another, assuming that the new location is either deemed to be more attractive or less
8 The selection variable is the distance between the countries of origin and France. Other independent variables of
the first step probit are: the log of origin GDP, the log of origin population, dummies for common border, common language
and colonial links and fixed effects for “départements” and time.
4. Data and Variables Conception
The migration data used in this paper comes from the French census of 1968, 1975, 1982, 1990 and
1999. These provide information on the stock of immigrants living in each French region9 in that
particular year. The total French populations are obtained from Insee (Institut National de la
Statistique et des Etudes Economiques)10.
Using this data, a "Jacobin" model of migration integration can be used to understand how foreigners
interact within a country. The French model of integration particularly reinforces the need that new
immigrants should incorporate French moral values. This is seen to be more important than any
consideration into the demographic characteristics of an individual such as ethnicity, sex or religion.
This ideal contrasts, for example, with the American model which is based on ethnic diversity where
the rhetoric is that of a "nation of nations" (Schain, 2004). Further research needs to be done in other
countries in order to provide comparisons.
Data on French trade at a regional level is available11 from the French Ministry of Ecology, Energy,
Sustainable Development and Town and Country Planning12. This data is available online for a
restricted sample of countries from 2003 to 2004. Namely, these countries are Australia, Austria,
Belgium, Cyprus, Czech Republic, Denmark, Estonia, Finland, Germany, Greece, Hungary, Ireland,
Lithuania, Latvia, Malta, The Netherlands, Poland, Portugal, Slovakia, Spain, Sweden, Switzerland
and the United Kingdom. The year 2004 is chosen for this study as it is close to the year 1999, which
is the last year that migration data is available for France and also it is more of a complete dataset
compared to that of 2003.
Geographical variables such as “common border” (a dummy variable set to 1 for pairs of countries
that share a border) and “common language” (dummies equal to one if both partners share a language)
are extracted from the CEPII database13.
The data on Gross Domestic Product and national population are taken from the World Bank “World
9 The French regions considered are: Alsace, Aquitaine, Auvergne, Basse-Normandie, Bourgogne, Bretagne,
Centre, Champagne-Ardenne, Franche-Comté, Haute-Normandie, Ile-de-France, Languedoc-Roussillon, Limousin,
Lorraine, Midi-Pyrénées, Nord-Pas-de-Calais, Pays de la Loire, Picardie, Poitou-Charentes, Provence-Alpes-Côte d'Azur,
Rhône-Alpes. Corse and non-metropolitan region are not considered
10 French institute of statistics. Source:
12 Ministère de l'Écologie, de l'Energie, du Développement durable et de l'Aménagement du territoire
4.2 A Description of the Instrumental Variables
There are two main problems concerning endogeneity. The first one is due to the use of "stocks of
immigrants" as a dependent variable. As Figure 4 shows, part of this stock (the grey colored square) is
already present at the period t. Then, the network variable based on the stocks of neighbor
“départements” at t has a reverse causality relationship with the dependent vector. This would lead to
over-estimated coefficients for the network variable. This endogeneity is corrected by the instrumental
variable method with the two lags of the network variable. This corresponds to 16 years14 and, based
on the data, there is no intersection between the stock of immigrants at t+1 and the stock of
immigrants at t-215.
Figure 4 – Endogeneity and the Stock of Immigrants
Lille Paris Nice MarseilleBordeaux
14 Migration data is obtained from the French census of 1968, 1975, 1982, 1990 and 1999. Each lagged
variable corresponds to 8 years difference on average.
15 This conclusion is made based on data from the Institut National d'Etudes Démographiques, from
France (http://www.ined.fr/): The sum of immigration flows from 1999 to 2005 is 119 422, the respective
difference in the stocks is 652906. Of the total population of immigrants in 2005 (4 959 000) 31.98% were not
in France in 1999. On average, each year from 1999 to 2005 6% of the population is composed of new
immigrants (immigrants who arrived in the current year).
The second issue is that the network variable does not take into account the stock of immigrants for
the reference “département”. In the Figure 4, the network variable does not sum the immigrants living
in Paris at the period t, which could cause an over-estimation by a missing agent. However, since the
reverse causality is controlled for, this effect captures the impact of immigrants living in the reference
“département” at period t who do not live there at the next period (the white square underneath the
grey square). Instead, these immigrants are part of the resident network and should henceforth be
accounted for. This second issue allows an improvement of the estimation as it partially considers the
network within the region.
Another predicament is that of endogeneity, which is due to not considering explanatory variables that
may have a significant effect. However, this is mitigated by using origin and destination fixed effects,
which can be interacted with time fixed effects.
This section investigates how and to what extent immigrants interact with each other, providing some
significant results. Two outcomes are considered: the location choice of new migrants and the
5.1 Location of immigrants
A first issue in the empirical analysis is to control for agglomeration forces other than social
interactions. For example, Paris is a very centralizing city in France, as confirmed by many studies
cited in section 2. Exogenous economical factors or some amenities could explain a huge
concentration of immigrants in this region. I control it by origin country fixed effects interacted with
years fixed effect.
Regression (1) verifies the positive impact of the Compatriot's Network on the immigrant's location.
Even controlling for all specific geographic effects over time, the quantity and the proximity of pre-
existing people from the same nationality determine the distribution of immigrants. This kind of
agglomeration offers benefits to communication, housing, job and business possibilities, access to
home produce, and leisure. These effects are analyzed by some recent studies (eg. Chiswick and
Miller, 2004) that point to evidence of a large concentration of immigrants in contrast to the national
Table 1 – Location Choice of Immigrants
Regression (1) (2) (4)
Specification PPML Probit IV PPML Probit
Dependent Variable Stock of Immigrants t+1
Ln Compatriot's Network 1.17*** 0.28*** 0.27***
(0.08) (0.03) (0.08)
Ln Distance -0.88*** -0.56***
Nonselection Hazard -0.35 -0.46*** -7.35**
(3.05) (0.11) (3.43)
Country, “département”, year F.E. Yes Yes Yes
Country and “département”
interacted with year F.E. Yes No No
Country-region F.E. No Yes No
Observations 39805 30292 29165
Note: Robust standard errors in parentheses with *, ** and *** respectively denoting significance at the 1%, 5% and 10%
Regression (2) introduces country-“département” fixed effects. This method controls for all effects
that are specific to each pairing of a country and “département”. Geographical fixed effects are not
interacted with time fixed effects in this regression because of technical limitation. The coefficient of
the network measures the impact within each geographical pair considering just the time variation of
the network variable. If the quantity of resident immigrants of a given nationality, living close to a
reference “département”, raise 10% each year, or if the same quantity concentrates 39.8 km each year
(if each immigrant moves 39.8 km closer to the reference “département”), the quantity of immigrants
living in the reference “département” raises 2.8% .
Regression (3) corrects for endogeneity applying lagged16 variables for networks with two steps IV
method. This regression includes the 16 years lagged variable, what is in fact 24 years lagged to the
dependent variable. This period of time is more than enough to control for the endogeneity, as
discussed in section 4.2. The coefficient is smaller than before, but still positive and significant.
Taking column (5), a 10% increase in the network close to a reference “département” implies an
increase of 2.7% in the stock of immigrants of this “département”. These regressions confirm the
existence of social interactions of migrants above any specific effect or endogeneity. An immigrant
counts on the network of compatriots when deciding a location in the host country. This natural
choice implies economic consequences, since the network assistance includes not only help in the
moving process, but also business and job developments. The next section analyzes the trade output
resulting from these interactions.
16 Because data are provided by 1968, 1975, 1982, 1990 and 1990 census, one lag represents 8 years in
Table 2 report results from a trade analysis. A limited sample is available which includes the
following countries: Australia, Austria, Belgium, Cyprus, Czech Republic, Denmark, Estonia,
Finland, Germany, Greece, Hungary, Ireland, Lithuania, Latvia, Malta, The Netherlands, Poland,
Portugal, Slovakia, Spain, Sweden, Switzerland and the United Kingdom.
These regressions follow the last advancements in trade literature. The Poisson Pseudo Maximum
Likelihood estimator is applied (Santos Silva and Tenreyro, 2006) and specific country and region
effects are controlled by fixed effects. Exports and imports are regressed separately.
Regression (1) and (2) estimates the impact of the Compatriot's Network on exports and imports,
respectively. In these regressions, the distance shows an expected coefficient (Disdier and Head,
2008). Immigrants’ network presents a positive impact on trade. Its coefficient is significant at the
99% level of confidence. In addition, as Tai (forthcoming), the impact of migration on imports is
higher than the impact on exports. Following this reference, the reason for this is that France imports
less differentiated products than exports. Immigrants act as a medium for the betterment of the trade
of less differentiated products because they do not need specific knowledge. Alternatively, these
products are easily replaceable and therefore more sensitive to a positive shock in their preferences.
Regression (3) and (4) apply the instrumental variable method. This corrects for a potential
endogeneity between trade flows and migration. This bias is already mitigated by the use of
Immigrant's stock instead of flows. However, the instrumental variable with a lag of 8 years can offer
a more accurate result. Coefficients are positive and significant at the 99% level of confidence.
Magnitudes rarely change and the coefficient for imports remains higher than the coefficient for
Table 2 – Trade Regressions
Note: Robust standard errors in parentheses with *, ** and *** respectively denoting significance at the 1%, 5% and 10%
levels. IV variables are the networks variables lagged once, which corresponds to 8 years averagely.
Regressions (5) to (10) provide a comparison to other studies of the impact of migration on trade.
Regressions (5) to (8) introduce the quantity of compatriots living in the reference region. It is exactly
the approach normally used in this kind of literature: the total quantity of immigrants living in a
country determines the trade of this country. From these estimations it is clear that this kind of
procedure is missing a key part. In columns (5) and (6), the stock of migrants presents a non
significant coefficient when the Compatriot's Network is controlled. This means that the network
variable incorporates more information that the stock variable. Regressions (7) and (8) do not control
for the Compatriot's Network. Even there, the coefficients are smaller than the ones of the network.
Regressions (9) and (10) apply the quantity of compatriots living in regions that share a common
border. This approach is the same of Herander and Saavedra (2005). One can see that this variable is
significant just for imports and at the 95% level of confidence. By contrast, the Compatriot's Network
variable remains significant at the 99% level of confidence for exports and imports. Yet again, the
method for counting immigrants' network proposed in this study seems to give a better explanation on
the impact of these networks on trade.
This paper investigates the extent to which social interactions impact location decisions and develop
trade networks. Empirical evidence strongly supports this conclusion. Data from French census allows
for immigrants to be accounted for by their nationality, in the “département” level (95
Regression (1) (2) (3) (4) (5) (6) (7) (8) (9) (10)
Dependent Variable Exports Imports Exports Imports Exports Imports Exports Imports Exports Imports
Specification PPML PPML - IV
Ln Compatriots' Net 0.67*** 1.12*** 0.85*** 1.66*** 0.97** 1.96*** 0.81*** 1.28***
(0.14) (0.19) (0.24) (0.40) (0.40) (0.72) (0.28) (0.46)
Ln Immig Stk within
Region -0.06 -0.15 0.17** 0.39***
(0.13) (0.24) (0.07) (0.13)
Ln Immig Stk
Neighboring Regions 0.04 0.68**
Ln Distance -0.46*** -0.47*** -0.77*** -0.49** -0.79*** -0.54*** -0.86*** -0.64*** -0.76*** -0.129
(0.08) (0.16) (0.12) (0.19) (0.12) (0.19) (0.12) (0.18) (0.13) (0.25)
Constant 11.46*** 14.34*** 14.78*** 18.22*** 15.73*** 20.66*** 10.24*** 6.18*** 6.77*** 14.36***
(0.76) (1.00) (1.39) (3.13) (2.81) (4.98) (1.39) (2.23) (1.87) (2.69)
Observations 525 523 525 523 525 523 525 523 525 523
“départements”), for five years17. Results prove that the location choice of an immigrant depends
strongly on the residing network from which he or she can benefit, even when destination region,
origin country, and time specific factors are controlled.
A function for the role of the distance on social interactions is presented and empirical outputs show
that these interactions diminish following the inverse of the distance between two immigrants.
Therefore, an immigrant benefits from the compatriot's network of destination region and also from
the network installed in other regions of the country.
Social interactions get ease the immigrant's settlement, but also provide business opportunities that are
verified by a very significant impact of networks on international trade. The trade of a certain region
is determined not only by the social interactions of immigrants within the region, but also by the
social interactions of the whole network of immigrants living in the country. This measure of network
is more robust than those measures in previous research.
AKERLOF, G. A, 1997., “Social Distance and Social Decisions”, Econometrica, 65(5): 1005-1027
AKERLOF, G. A, KRANTON, R., 2000, “Economics and Identity”, Quarterly Journal of Economics,
ANDERSON, J.E. AND E. VAN WINCOOP, 2003, “Gravity With Gravitas: A Solution to the Border
Puzzle”, American Economic Review, 93(1): 170-192.
BALDWIN, R., 2006, “In or out: does it make a difference? An evidence-based analysis of the trade
effects of the euro, London: CEPR.
BALDWIN, R., R. FORSLID, P. MARTIN, G. OTTAVIANO AND R. ROBERT-NICOUD, 2005, “Economic
Geography and Public Policy”, Princeton University Press.
BALDWIN, R., D. Taglioni, 2006, “Gravity for Dummies and Dummies for Gravity Equations” NBER
Working Papers 12516.
BARTEL, A. P., 1989, “Where Do the New U.S. Immigrants Live? ”, Journal of Labor Economics,
BECKER, G.S., 1996, Accounting for Tastes, Harvard University Press.
BECKER, G. S. AND MURPHY, KEVIN M.: Social Economics. Market Behavior in a Social
Environment. The Belknap Press of Harvard University Press 2000.
BOWLES, S., 1998, “Endogenous Preferences: The Cultural Consequences of Markets and other
Economic Institutions”, Journal of Economic Literature, 36: 75-111.
17 These data are aggregated by regions (21 regions) to correspond to trade data
BROCK, W. A. and S. N. DURLAUF, 2001, “Discrete Choice with Social Interactions” Review of
Economic Studies, 68(2): 235-260
BROCK, W. A. and S. N. DURLAUF, 2002, “A Multinomial-Choice Model of Neighborhood Effects”,
American Economic Review, 92(2): 298-303
CASELLA, A., J. RAUCH, 1998, “Overcoming Informational Barriers to International Resource
Allocation: Prices and Group Ties”, NBER Working Paper 6628.
CAVALLI-SFORZA, L. L., M. W. FELDMAN, 1981, “Cultural Transmission and Evolution: A
Quantitative Approach”, New Jersey: Princeton University Press.
CHISWICK, B. R. and P. W. MILLER, 2004, “Where Immigrants Settle in the United States”, Institute
for the Study of Labor (IZA), discussion paper 1231
CHISWICK, B. R., Y. L. LEE and P. W. MILLER, 2002, “The Determinants of the Geographic
Concentration among Immigrants: Application to Australia”, Institute for the Study of Labor
(IZA), discussion paper 462
COMBES, P.P., M. LAFOURCADE AND T. MAYER, 2005, “The Trade-Creating Effects of Business and
Social Networks: Evidence from France”, Journal of International Economics, 66: 1-29.
CROZET M., 2004. “Do Migrants Follow Market Potentials? An Estimation of a New Economic
Geography Model”, Journal of Economic Geography,. 4(4): 439-458
DISDIER, A.C. AND K. HEAD, 2008, “The Puzzling Persistence of the Distance Effect on Bilateral
Trade”, Review of Economics and Statistics, 90(1): 37-41
DOCQUIER, F., O. FAYE AND P. PESTIEAU, 2008, “Is migration a good substitute for education
subsidies?”, Journal of Development Economics, 86: 263-2
DURLAUF, S. and P. YOUNG, 2001 “Social Dynamics”, Washington, DC: Brookings Institution Press
FEENSTRA, R.C., 2004, Advanced International Trade: Theory and Evidence, Princeton: Princeton
FUJITA, M, P. KRUGMAN, A. J. VENABLES, 2000, “The Spatial Economy”, The MIT Press
FUNKHOUSER, E., 2000, “Changes in the Geographic Concentration and Location of Residence of
Immigrants”, International Migration Review, 34(2): 489-510
GIRMA, S. AND Z. YU, 2002, “The Link between Immigration and Trade: Evidence from the United
Kingdom”, Review of World Economics 138(1): 115-130.
GLAESER, E. J. SCHEINKMAN, 2003, “Non-market Interactions” in Advances in Economics and
Econometrics: Theory and Applications, Eighth World Congress, Vol. I (M.Dewatripoint, L.P.
Hansen and S. Turnovsky, eds.), Cambridge University Press: 339-369
GLAESER, E. L., D. CUTLER and J. L. VIGDOR, 1999, “The Rise and Decline of the American
Ghetto”, Journal of Political Economy, 107(3): 455-506
GLAESER, E. L., B. SACERDOTE and J. A. SCHEINKMAN, (1996) “Crime and Social Interactions”, The
Quarterly Journal of Economics, 111(2): 507-548
GONZALEZ, A., 1998, “Mexican Enclaves and the Price of Culture”, Journal of Urban Economics, 43:
GOULD, D.M., 1994, “Immigrant Links to the Home Country: Empirical Implications for U.S.
Bilateral Trade Flows”, Review of Economics and Statistics 76(2): 302-316
GREENE, W.P., 2003, Econometric Analysis - Fifth Edition, Englewood Cliffs: Prentice Hall.
GUIMARÃES, P., FIGUEIRDO, O. AND WOODWARD D., 2003, “A Tractable Approach to the Firm
Location Decision Problem”, Review of Economics and Statistics 85(1): 201-204
HEAD, K., T. MAYER, 2008, “Detection of Local Interaction from the Spatial Pattern of Names in
France”, Journal of Regional Science 48(1): 67-95
HEAD, K. AND J. RIES, 1998, “Immigration and Trade Creation: Econometric Evidence from Canada”,
Canadian Journal of Economics, 31(1): 47-62
HECKMAN, J. J., 1976, ‘‘The Common Structure of Statistical Models of Truncation, Sample
Selection, and Limited Dependent Variables and a Simple Estimator for Such Models,’’ Annals
of Economic and Social Measurement, 5, 475–492.
HELLIWELL, J. F., 1997, “National Borders, Trade and Migration”, Pacific Economic Review, 3: 165-
HERANDER, M. G. AND L. A. SAAVEDRA, 2005, “Exports and the Structure of Immigrant-Based
Networks: The Role of Geographic Proximity”, Review of Economics and Statistics, 87(2): 323–
MANSKI, C. F., 1993, “Identification of Endogenous Social Effects: The Reflection Problem” Review
of Economic Studies, 60(3): 531-542
MANSKI, C. F., 2000, “Economic Analysis of Social Interactions”, Journal of Economic Perspectives,
MASSEY, D. S., Denton, N. A., 1998, “The Dimensions of Residential Segregation” Social Forces,
MAYDA, A. M., 2007, “International Migration: A Panel Data Analysis of the Determinants of
Bilateral Flows", CEPR, Discussion Paper 6289
PONCET, S., 2006, “Provincial migration dynamics in China: Borders, costs and economic
motivations”, Regional Science and Urban Economics 36: 385–398
PORTES, A., BOROCZ, J., 1989, “Contemporary Immigration: Theoretical Perspectives on Its
Determinants and Modes of Incorporation” International Migration Review, 23(3): 606-630.
RAUCH, J., 1999, “Network Versus Markets in International Trade”, Journal of International
Economics, 48: 7-35.
RAUCH, J., 2001, “Business and Social Networks in International Trade”, Journal of Economic
Literature, 39: 1177-1203.
RAUCH, J. AND V. TRINDADE, 2002, “Ethnic Chinese Networks in International Trade”, Review of
Economics and Statistics, 84(1): 116-130
RAUCH, J., 2004, J. Watson, 2004, “Networks Intermediaries in International Trade”, Journal of
Economics and Management Strategy, 13: 69-93.
REDDING, S. AND A.J. VENABLES, 2004, “Economic Geography and International Inequality”
Economic Policy, Journal of International Economics, 66: 53-82.
SANTOS SILVA, J.M.C. AND S. TENREYRO, 2006, “The Log of Gravity”, Review of Economics and
Statistics, 88(4): 641–658.
SCHAIN, M. A., (2004), “The Politics of Multiculturalism in France and the United States,”
Conference on Transatlantic Tensions: From Conflicts of Interests to Conflict of Values?
CERI/Paris, February 2-3, 2004
SCHEINKMAN, J., “Social Interactions” Forthcoming. in The New Palgrave Dictionary of Economics
and Law, Available at http://www.princeton.edu/∼ joses/wp/socialinteractions.pdf.
STARK, O., 2003, “Tales of Migration Without Wage Differentials: Individual, Family, and
Community Contexts”, Center for Development Research ZEF – Discussion Papers on
Development Policy 73
STIGLER, G.J. AND G.S. BECKER, 1977, “De Gustibus Non Est Disputandum”, American Economic
Review, 67: 76-90.
TABUCHI, T., J.-F. THISSE, 2002, “Taste heterogeneity, labor mobility and economic geography”,
Journal of Development Economics, 69: 155– 177
TAI, S. H. T., forthcoming, “Market Structure and the link between Migration and Trade”, Review of
TOPA, G., 2001, “Social Interactions, Local Spillovers and Unemployment” Review of Economic
Studies, 68(2): 261-295
VANDERKAMP, J., 1971, “Migration Flows, Their Determinants and the Effects of Return Migration”
Journal of Political Economy,.79(5): 1012-1031.
WAGNER, D., K. HEAD AND J. RIES, 2002, “Immigration and the Trade of Provinces”, Scottish
Journal of Political Economy, 49(5): 507-525.
WATTS, M., 1998, “Occupational Gender Segregation: Index Measurement and Econometric
Modeling”, Demography, 35(4): 489-496
WAVODNY, M., 1998, “Determinants of Recent Immigrants" Locational Choices”, Federal Reserve
Bank of Atlanta, Working Paper 98-3
WHITE, H., 1980, “A Heteroskedasticity-Consistent Covariance Matrix Estimator and a Direct Test
for Heteroskedasticity”, Econometrica, 48(4): 817-838.
WILLIAMSON, J. G., T. J. HATTON, 2005, “What Fundamentals Drive World Migration? ” in G.
Borjas and J. Crips (eds.), Poverty, International Migration and Asylum, Hampshire, UK:
Palgrave-Macmillan for WIDER: 15-38