1

Who welcomes them? The spatial distribution of refugees in Italy between attitude and opportunity

Ugo Fratesi Politecnico di Milano ugo.fratesi@polimi.it

Marco Percoco1

Università Bocconi marco.percoco@unibocconi.it

Paola Proietti

Gran Sasso Science Institute paola.proietti@gssi.infn.it

Abstract

This paper investigates the institutional and socio-economic determinants of the location of asylum seekers in the case of Italy where, to face the pressure of arrivals, a complex multi-level system of hosting has been set up. In this system, asylum seekers are allocated to local communities through periodic calls (i.e. with a bottom-up procedure where communities bid for them). This makes it an interesting case, in which local attitudes and economic opportunities are both at play. The econometric analysis explores the economic, social and political drivers of such redistribution findings that, counterintuitively, social capital is negatively related to willingness to host asylum seekers, probably due to the desire to maintain cohesive communities.

1 Corresponding author

2

1. Introduction

Civil wars, political instability, fear and uncertainty are affecting the lives of millions of people and

families in Africa and the Middle East. To escape the fury of these tragic events, an unprecedented

mass of refugees, migrants and asylum seekers are fleeing their countries of origin and moving

towards Europe.

The UNHCR (2016) estimates that more than 1 million people arrived in Southern Europe by boat

during 2015, most of them from Afghanistan, Iraq and Syria. The majority of these arrived through

the Aegean Sea from Turkey to Greece, while another important proportion came to southern Italy

from Libya. This route is affected by the intervention of Italian and European ships under Operation

Triton, managed by Frontex (the European Union's border security agency). While southern

European countries are generally not the final destination of refugees, they are the front-door of

Europe and, under current European border legislation, have to identify and manage the process by

which the eligibility for asylum is determined. This should in principle last 30 days but de facto

spans several months and quite often over one year. In fact, the Dublin Regulation (Regulation No.

604/2013) aims at preventing asylum seekers from submitting applications in different countries

and as such stipulates that the responsible Member State will be the one through which the asylum

seeker first entered the EU. Mechanisms of re-allocation are also implemented. For instance, the

European Council of September 22, 2015 decided to relocate 120,000 refugees in two years, but these

mechanisms are still de facto only concerning a limited number of people.

In a period of significant economic and political uncertainty, even within the borders of the

European Union, regions and cities of southern European countries are therefore currently facing

the challenge of hosting a large number of people. This is combined with the perspective of receiving

many more if the political problems in North Africa and the Middle East are not solved in a

reasonable amount of time.

In this paper, we study the determinants of the spatial distribution of refugees in Italy, a country

that, because of its baricentric position in the Mediterranean Sea, is among those hosting more

refugees and asylum seekers in Europe. Most of the literature deals with the impact of mass

migration on local economies. However, comparatively less attention has been devoted to the

determinants of the location. This second aspect is even more important in the case of Italy, as the

distribution of refugees across locations is not the outcome of a free choice of individuals or by the

State. Instead, it is the result of a bargaining process between the central government, local

authorities and operators, mainly not-for-profit firms and associations. As a consequence, asylum

seekers in the ordinary hosting system are not allocated with a top-down procedure but are instead

3

allocated to those communities which successfully bid for them. This makes for an interesting case

study, in which local attitudes and economic opportunities are both at play.

The location of asylum seekers does not, therefore, depend on their own preferences2, but on the

attitudes of Italian local communities towards their hospitality. This is expected to be driven by

social, institutional and political settings. However, economic reasons are also likely to be relevant,

as the central government pays a daily fee to those who host asylum seekers and this might represent

a significant opportunity for places facing economic distress.

The institutional framework governing the distribution of refugees is therefore influenced by

political, economic and even cultural factors. Social capital in particular may be thought to play a

fundamental role in such processes. Banfield (1954) has in fact argued for a positive correlation

between pro-social or cooperative behavior and generalized trust or even the willingness to provide

help and assistance to others.

Figure 1 depicts a simple correlation between the number of refugees per 1,000 inhabitants in Italian

regions (at NUTS2 level) and a commonly used measure of social capital: blood donations. The

negative implication of the correlation is striking, as it indicates that regions with a higher stock of

social capital are also the ones hosting a lower number of refugees. Of course, this correlation needs

to pass the test of controlling for other factors. However, the Figure clearly points to the non-

triviality of the analysis of the spatial distribution of refugees and asylum seekers.

Our econometric analysis, conducted at a finer spatial scale (NUTS3), confirms that social capital is

negatively correlated with the decision of a territory to accept refugees, although we could not find

similar results in explaining the size of the spatial distribution of such populations.

Taken together, these results cast some doubts about the genuineness of “generalized trust”, as the

extent of such feelings and propensity is perhaps limited in space and may even impact negatively

on some specific pro-social norms.

The paper starts with a review of the literature regarding the evidence available on the

environmental, political and economic determinants of mass migration flows, while also considering

the impact of refugees on innovation, wage structure and local development in general. The paper

then proceeds with an empirical analysis of the spatial distribution of asylum seekers and refugees

in Italy, in order to investigate to what extent the different factors are at play in local attitudes

towards refugee hosting.

2 Only when their asylum request is accepted do refugees become free to choose their favourite domicile.

4

2. Related literature

International migration is a global phenomenon which has expanded in recent years. The

international migration report (UN, 2013) estimates that the number of international migrants has

been constantly growing over the period 1990-2013. Europe is the main destination for migrants,

being the residence of about one third of the total, and is closely followed by Asia. Even more

interesting is the fact that international migration flows have specific characteristics in terms of

origin-destination. North-south flows have remained stable and very low, while north-north flows

have grown slowly and steadily. The most remarkable trend, however, is the very sizeable growth

of south-south and south-north flows, which now account for the very large majority of migrants

(UN, 2013).

Global migration flows have been significantly studied in the literature, at both the national the

regional level. Particular focuses of academic research have been the reasons behind choices to

migrate, the determinants of the location of migrants (e.g. Pedersen et al., 2008; Wang et al., 2016)

and the impact of migration on the places of origin and of destination of migrants (e.g. Kanbur and

Rapoport, 2005; De Haas, 2010; Di Maria and Stryszowski 2009; Dustmann et al., 2011; Borjas, 2015;

The World Bank, 2006)

Refugees constitute a relatively small part of global migration, since they are defined by the 1951

Refugee Convention as a person who “owing to a well-founded fear of being persecuted for reasons

of race, religion, nationality, membership of a particular social group or political opinion, is outside

the country of his nationality, and is unable to, or owing to such fear, is unwilling to avail himself

of the protection of that country”.

However, refugees also constitute a rapidly growing and highly visible part of international

migration. The UNHCR (2016) estimates that more than 2 million asylum applications were received

in 38 European countries in 2015. This is a substantial increase from the slightly more than 700,000

lodged in 2014.

A large (and now growing) body of literature has long considered the effect of mass migration or

refugees on local economies. An early study was conducted by Card (1990) on the effect of the Miami

boatlift in the early 1980s on the Florida labour market, finding limited evidence of an effect on

wages and employment.

Hunt (1992) and Carrington and De Lima (1996) have considered the case of repatriates from the

African colonies of France and Portugal, reporting a negative effect on employment and wages in

locations where repatriates settled.

5

These early papers on the impact of “return” migrants are interesting, as they consider a shock

occurring in other countries as an influence on the decision to migrate, although it was sometimes

not a properly free decision. However, they are not immune from critique of their identification

strategy, given the highly-aggregated data they made use of and the endogeneity of the location

decision. To circumvent the issue of self-selection, Glitz (2012) exploits an exogenous variation in

German migration law which occurred in 1989. This made the migration of family members of

immigrated workers easier. In this case, a small and negative effect of wages and internal migration

were found.

In general, however, the main effect of large migration inflows, if any, could be found in the re-

distribution of income. If migrants are on average less skilled than residents, then unskilled workers

will face tougher competition for jobs and a consequent reduction in wages, which may be not

apparent if only average wages are considered.

The case of refugees and asylum seekers is different from the general case of migrants, as in most

cases they are not allowed to work in hosting countries, at least in the short term. Subsequently, in

the medium-term they have lower employment rates than other immigrants and finally have higher

benefit use rates (European Parliament, 2016; Martìn, I. et. Al, 2016; Ruist, J. 2015;). These features

imply that their impact on local economies is usually worse than that of economic migrants.

Besides Europe in these last few years, Africa is certainly the continent that has witnessed the largest

displacements of people because of wars, political unrest and natural disasters. Therefore, such cases

have attracted some scholars aiming to shed new light on the relationship between forced migration

and development in receiving countries. The most interesting case in recent years has certainly been

the refugee crisis in Tanzania, due to masses of individuals fleeing from Burundi (1993) and Rwanda

(1994).

In a first attempt to evaluate the effect of such phenomena, Baez (2011) has argued in favour of a

negative effect, since the socio-economic stress imposed by refugee influence is too strong to be

absorbed by weak local labour markets. Interestingly, Maystadt and Duranton (2014) consider the

same case and reach opposite conclusions. In a rigorous econometric analysis, they consider a panel

of Tanzanian households over the period 1991-2010 and find that the presence of refugees from

Burundi and Rwanda has had a positive effect on welfare. They also argue that such positive change

was due to public investment in infrastructure, decreasing transportation costs and leading to a

higher mobility of individuals. This result is interesting and important, since it implies that a

temporary population shock may have permanent positive effects thanks to efficient and

appropriate public policies.

The Syrian conflict is currently generating large refugee inflows in neighbouring countries,

6

especially Lebanon and Turkey, deeply affecting local economies. Akgunduz et al. (2015) exploit the

geographical distribution of refugee camps in Turkey to analyse the impact of the inflows of Syrians

in terms of housing and food prices, as well as market outcomes. They find a slight increase in the

level of prices and no significant changes in the employment rate of natives.

As it is also clear from the work of Maystadt and Duranton (2014), considering a long period of time

is of paramount importance to disentangling the effect of massive population shock. An interesting

case is the Great Migration to the USA from Europe, which occurred between the second half of the

nineteenth century and the beginning of the twentieth. Those migration flows were driven by both

economic reasons and persecutions, especially in the case of Jews in Eastern Europe facing the

worrying diffusion of pogroms. Rodriguez-Pose and von Berlepsch (2014) find that countries that

had attracted more migrants between 1880 and 1910 had better economic indicators in 2005, even

after controlling for self-selection of migrants into locations.

Finally, it should be noted that one of the most important drivers of development in modern

economies is innovation capacity. Moser et al. (2014) find that high-skilled refugees may boost

innovation in destination countries through scientific spillovers. In particular, they consider the

impact on the number of patents in the USA of Jewish scientists who were expelled from German

universities and who fled to the USA.

As should be evident at this point of the review, the economic literature on refugees has primarily

focused on the impact of displaced individuals while comparatively less attention has been devoted

to the analysis of the location of refugees. This is affected by a variety of factors, some of which are

eminently political and cultural.

In particular, there is little evidence on the factors that, at a sub-national level, influence the attitude

towards hosting refugees. The International Organization for Migration (IOM) (2015) provided a

survey of public attitudes towards immigration worldwide, finding that in most of the world the

attitude towards immigrants is rather positive. Europe is an important exception in which the

majority of the population believes immigration levels should be decreased. This is the outcome of

two different models, one prevailing in northern Europe in which migration is welcome, and

another prevailing in southern Europe in which migration is considered to be excessive. This might

also be related to the fact that southern Europe is not only the front-door of migration into Europe

and the hub of the refugee crisis, but is also the part of the continent which has been affected most

by the economic crisis. Therefore, the economies of these countries are structurally weaker and have

important issues in terms of public finances (Moro and Becker, 2016; Fonseca and Fratesi, 2016).

This is consistent with the fact, also found in IOM (2015), that the attitude towards migrants is more

7

negative among those people who are more likely to compete with migrants for job opportunities

(i.e. the lower educated and unemployed).

Yet, those same data are not available at a sub-national level. However, Italian institutional settings

allow for testing revealed attitudes to migration, at least towards asylum seekers.

3. Institutions and the spatial distribution of refugees

As argued in the previous section, the literature has primarily focused on the impact of refugees on

labour markets. However, little is known about the allocation of asylum seekers and refugees across

space, especially at a sub-national level.

However, obtaining more precise knowledge of national redistribution schemes is becoming

essential, since the bargaining among European countries to share the refugees’ hosting has also

become tighter.

One of the few studies trying to propose corrections for the refugees’ redistribution program (which

is stated in the European Agenda on Migration (European Commission, 2015a) and still not

functioning), is Rapoport and Fernandex-Hertas (2015).

In particular, they consider a market for tradable quotas of refugees (with correction on the basis of

refugees’ and states’ preferences), advancing the idea that a competitive market may help to reach

an efficient allocation.

In this “market”, countries would trade quotas previously assigned according to the allocation

scheme proposed by the European Agenda on Migration and equilibrium would be reached, while

also taking into consideration a combination of refugees’ and states’ preferences.

Thielemann et al. (2010) present a review of the allocation mechanisms across Europe. According to

this work, refugees in the UK were assigned to local authorities on the basis of indicators such as

number of refugees per capita. In Germany, the spatial distribution of asylum seekers was decided

by the federal government, and local authorities did not play a significant and explicit role. In

Sweden and France, refugees chose where to locate almost freely and public funds were allocated

consequently.

In reality, national pictures were already more complex in 2010 and are even more complex now

than those emerging from the previously mentioned and other comparative works.

For example, regarding the situation in the UK, there are many rigorous papers dealing with asylum

seekers’ dispersal policies which are exclusionary and always more privatized (Darling, J. 2016;

Hynes, P. 2011; Phillips, D. 2006 and Robinson, V. 2003;).

However, there is also evidence from other countries such as Germany and France, where the

allocation of refugees involves both national and local administrative units, although with different

8

criteria.

For example, in Germany, the methodology utilized to distribute asylum seekers among Länder is

the “Königstein Formula”. This is calculated each year according to the tax receipts and population

numbers. However, after this, “two or more Länder may agree that persons requesting asylum who

are to be admitted by a specific Land, in line with its admission quota, are admitted by other Lands.”

(Federal Law Gazette).

Interestingly, the mechanisms at work at a subnational level in France and Italy also mimic a market

mechanism; local authorities participate in hosting calls issued at the national level.

In Italy, basically, there are three phases associated with the arrival and stay of refugees:

1. The arrival, at points mainly in the south of Italy and in Sicily, with first hospitality

essentially set up under national government management.

2. The geographical distribution of refugees and asylum seekers from the moment they are

waiting for the verdict on their refugee status until six months after a positive answer to it or

the end of their first recourse. If they got a negative answer and decide to appeal, there is a

bidding process.

3. After successful dismissal from the hosting system, refugees can settle according to their

preferences.

In this paper, we consider the allocation mechanism between the asylum request and the asylum

seekers’ or refugees’ dismissal from the hosting system, since this period involves several public and

private entities and may shed light on the attitudes of territories to hosting refugees.

Furthermore, we consider only the ordinary hosting through SPRAR Sistema di Protezione per

Richiedenti Asilo e Rifugiati – Protection System for Asylum Seekers and Refugees) and not the

extraordinary one, mainly provided by CAS (Centri di accoglienza straordinaria – Centres for the

extraordinary hosting).

The SPRAR is organized in the form of a multi-level governance structure, in which local authorities

and not-for-profit organizations form a coalition, called progetto territoriale (territorial project), to

host a given number of refugees.3

It should be noted that refugees not only are granted accommodation and food, but also a series of

ancillary services with the aim of improving the conditions of their integration in order to achieve

an independent and satisfactory post-hosting life.

Every two or three years until 2016, the Ministry of Interior has issued a call for proposals to allocate

funds to host refugees. Local authorities apply by proposing drafts of their progetti territoriali, which

3 For an analysis of multi-level governance in Italy, see Percoco (2016) and Percoco and Giove (2009).

9

are then selected on the basis of quality indicators for services and hosting capacity.

As can be seen in Figure 2, there has been a dramatic increase in the number of refugees hosted

through the system of progetti territoriali, with a threefold increase between 2012 and 2013 and a

doubling between 2013 and 2014. Figure 2 also shows that, in 2015, provinces involved in the hosting

through SPRAR were 92 with the highest relative number of refugees hosted in Crotone (0.27%),

Agrigento (0.26%) and Rieti (0.15%). This illustrates little overlapping with the provinces of arrival.

In 2015, there were 430 progetti territoriali, involving 376 local authorities (municipalities, provinces,

metropolitan areas, mountain communities and a union of municipalities).

Interestingly enough, the SPRAR system is an indicator of revealed preferences of territories to host

refugees, because of the involvement of several levels of government and the public-private

partnership nature of progetti territoriali. Furthermore, the sharp increase between 2012 and 2015

shown in Figure 3 makes the spatial distribution of refugees an interesting quasi-natural experiment

to study the territorial attitudes of provinces.

In this paper, we test the role of social capital in the decision of territories to host refugees. In general,

social capital is considered to play a crucial role in the determination of institutional collective

actions (Percoco, 2016), in this case in the form of progetti territoriali. However, the long-lasting

economic crisis might have changed the mechanism for the allocation of resources. Communities

with strong social ties may prefer to allocate more resources to its members, as opposed to

communities with weak ties among its members. Although we cannot directly test this hypothesis,

we will test for the negative sign of the correlation between the number of refugees in Italian

provinces and the level of social capital.

4. The determinants of the spatial distribution of refugees: an empirical analysis

In this section, we analyse the determinants of the spatial distribution of refugees by means of a

regression analysis.

Our baseline regression, in particular, considers the probability of hosting refugees in the province

as a function of economic, cultural and political factors:

Pr(𝑅𝑅𝑅𝑅𝑅𝑅𝑅𝑅𝑅𝑅𝑅𝑅𝑅𝑅𝑖𝑖 = 1) = 𝛼𝛼𝐸𝐸𝐸𝐸𝐸𝐸𝐸𝐸𝐸𝐸𝐸𝐸𝐸𝐸𝑖𝑖 + 𝛽𝛽𝐶𝐶𝑅𝑅𝐶𝐶𝐶𝐶𝑅𝑅𝐶𝐶𝑅𝑅𝑖𝑖 + 𝛾𝛾𝑃𝑃𝐸𝐸𝐶𝐶𝑃𝑃𝐶𝐶𝑃𝑃𝐸𝐸𝑃𝑃𝑖𝑖 + 𝜀𝜀𝑖𝑖

Where the dependent variable takes the value of 1 if Province i hosts refugees in SPRAR in 2015 and

0 if it does not.

Among economic variables, we include the unemployment rate in 2013 and the value added per

10

capita. In both cases, the source for these data is ISTAT.

As is standard in the economic literature, we use the number of not-for-profit firms per capita as a

proxy for social capital. In this case, the source of data is the Censimento del Nonprofit conducted by

ISTAT in 2011.

As for politics, we consider a dummy variable indicating the political party ruling in province i. This

variable assumes three possible values: right, left, and extraordinary administration. Data is taken

from the Ministry of the Interior website.

Furthermore, we control for whether the regional chief-town is located in the province or not, as

well as whether the province is a point of direct arrival for asylum seekers.

In Table 2, we report results (odds ratios) of logistic regressions. In all models, we control for a

dummy indicating whether the province is a point of arrival and whether the regional chief-town is

located in its territory. Subsequently, the economic, cultural and political variables are added,

first separately and then together.

The results are consistent among the different specifications, with only a few coefficients become

slightly more significant in some specifications. However, all of them hold the same sign and the

same magnitude.

Interestingly, it emerges that points of arrival provinces are generally less likely to host refugees in

SPRAR (with a sort of specialization of provinces, either in the first or in the secondary hosting).

Moreover, this decision is not significantly correlated to local economic conditions.

Concerning the economic variables, it appears that they do not exert a very important role. The

regional value added per capita is generally negative and insignificant, meaning that this is not a

clear determinant of the willingness of communities to host asylum seekers. In line with the results

displayed in the survey by IOM (2015), the unemployment rate is negatively related to openness to

refugees as they are probably seen as competitors in a tight labour market. Despite this, this also

tends to be of little significance, especially when inserted into the same regression with regional

value added per capita (which is correlated with the unemployment rate).

When the indicator of social capital is added, we find a negative and strongly significant correlation.

As we proxy social capital by means of the size of the not-for-profit sector (and this sector is directly

involved in the process of distribution of refugees over spaces), it is important to disentangle the

cultural traits embedded in the sector and the service supply. To this end, we also consider a variable

11

indicating the number of not-for-profit entities explicitly dedicated to immigrants (per capita), and

the negative relation between social capita and willingness to acquire refugees holds.

We then control for politics, which does not appear to play an important role. Provinces

administered by the centre-right (which nationally tends to be more against migration) are slightly

less likely to host refugees, while provinces administered by the centre-left (which nationally tends

to be more pro-migration) are slightly more likely to host them. However, no coefficient is significant

and as such there is no statistical difference between them and those provinces (around 1/3 of the

total) where the administration is either a special administration or one which cannot be classified

under the national label.

A test has also been introduced for the past presence of immigrants, intending to determine whether

the attitude is influenced by past behaviours and the spontaneous location of foreign immigrants.

This coefficient appears to be positive but not significant either.

Since spatial effects can be at play, Table 3 reports the same logit regressions as Table 2, but in this

case with spatially corrected spatial errors (using the routines developed by Tim Conley). All the

significances are the same, with the exception of the share of foreign citizens. This becomes

significant if inserted alone, but remains insignificant when inserted in the general regression model.

In the introduction of this article, we presented a negative correlation between the number of

refugees and social capital. We have also estimated regressions similar to the ones in Table 2 with a

continuous dependent variable measuring the number of refugees. In such cases, we could not find

any significant results, indicating that social capital shapes the decision of whether or not to host

refugees. However, it does not appear to influence the number of places which are made available

to refugees.

Why can social capital be negatively correlated with willingness to host asylum seekers? At first,

appears to be a counter-intuitive result but is much less so when the essence of social capital is

considered. In fact, the literature on this subject distinguishes between two essential categories of

social capital: bonding and bridging (Putnam, R. D., 2000). The first should help individuals by

keeping stable links among people that feel part of the same community. These people share values,

institutions and often the territory in which they are located. Bonding social capital, therefore, is not

expected to imply openness to people arriving from distant places with very different cultural

backgrounds.

12

Bridging social capital, on the other hand, should help individuals by allowing them to make

connections and safely deal with other people they do not know. Bridging social capital is therefore

more related to openness, as it makes it easier to see the opportunity of meeting and relating to

different people. The presence of voluntary associations (our social capital variable) is generally

linked with bridging social capital in the literature. However, the spatial scale of bridging social

capital is yet to be defined. Does it go so far as to link people from different continents, languages,

religions, education and culture? Does it extend to substantial users of public benefits?

Our results appear to support that this is not the case, and that communities with more social capital

are less likely to harbour asylum refugees.

The negative sign can even be expected, given the social structure of Italian provinces. Those with

more social capital, in fact, also tend to be more internally cohesive, with communities sharing

values and (especially) an identity. These tend to be small-city provinces, often in northern Italy,

where the arrival of groups of refugees in the middle of the town can be seen as possibly disrupting

the social tissue. Since the communities have to apply for a bid in order to host refugees, persons,

associations and administrations which bid can be seen as those who disturb a way of living

characterized by internal cohesion and peaceful quality of life.

Provinces with less social capital, on the other hand, are less cohesive. In this case, with less fear of

reprisal4, it is easier for altruistic organizations willing to hospitality to asylum seekers to bid, as

well as for unconcerned entrepreneurs looking to grasp the economic opportunity to host refugees

in otherwise run-down touristic structures.

There is a literature observing that, historically, the diffusion of divisive ideas in politics is made

easier by stronger social capital (Riley, 2010; Satyanath et al.; 2013), and that lower acceptance of

refugees can be seen as a clear sign of closure. However, it is also interesting to note that, at the time

of our analysis in 2014, economic conditions were precarious because of long-lasting crises and that

our results are an indication that local communities with strong internal ties tend to increase their

4 Recently, newspapers recorded protests against hotels hosting refugees in a number of Italian small towns, such as Collio (BS) (http://brescia.corriere.it/notizie/cronaca/15_agosto_29/collio-non-vuole-profughi-assediato-l-hotel-che-li-ospita-brescia-f6247b00-4e25-11e5-a97c-e6365b575f76.shtml) , San Genesio (PV) (http://laprovinciapavese.gelocal.it/pavia/cronaca/2014/03/21/news/rifugiati-condotti-a-pavia-all-alba-lega-li-accoglie-con-presidio-di-protesta-1.8893852) San Zeno (VR) (http://www.veronasera.it/politica/prada-manifestazione-lega-nord-contro-rifugiati-lago-garda-15-novembre-2015.html), Cosio Valtellino (SO) (http://www.rainews.it/dl/rainews/articoli/Valtellina-nuove-minacce-a-albergatore-che-ospita-i-profughi-Diamo-fuoco-a-hotel-948de8ff-8fff-4fc2-ae47-7efeb52ecf76.html) and, more recently the working-class Gorino (FE) (http://www.lastampa.it/2016/10/26/italia/cronache/tra-le-barricate-di-gorino-alla-fine-del-po-non-razzismo-abbiamo-paura-tYoV6YB7tFpFZuQgaGmYBN/pagina.html) and the leftist but elitist Capalbio (http://www.corriere.it/cronache/16_agosto_13/migranti-cinquanta-profughi-le-ville-vip-anche-capalbio-fa-barricate-efedd9fa-6118-11e6-8e62-f8650827a70c.shtml).

13

closeness in periods of economic downturn. During these times, the allocation of financial and non-

financial resources may strongly privilege members of the community over outsiders.

5. Conclusions

In this paper, we analysed the attitudes determining the willingness of local communities to host

asylum seekers and early refugees. In particular, some determinants are expected to act in these

processes: the economic situation of places, including the unemployment rate; the political attitudes

of communities, since some political parties are pro-migration and others are against it; and finally,

the social capital of places.

In particular, to test these assumptions, we were able to analyse a very interesting case: Italy. This

country is interesting because there are three phases refugees must go through during their

application process. The first phase is the hosting in the first days and weeks after arrival, in which

it is the government which sets up structures to serve as host points for newcomers. During the

second phase, refugees need to wait for the outcome of their application to be officially considered.

The mechanism set up for this phase is peculiar, since in SPRAR (the scalar fix of the hosting in Italy)

it involves bids to provide places for asylum seekers, where participants are local administrations

involving entities from civic society. In this way, the distribution of asylum seekers in this phase is

not due to centralized planning but to a spontaneous process from the local communities. As a

consequence, actual distribution can reveal local entities’ attitudes towards refugees.

During the third phase, the demand of a refugee is approved and the refugee is dismissed by the

hosting system, leaving them free to choose where to locate.

The second phase is, therefore, the most interesting and innovative with respect to the existing

literature, since it becomes the result of bottom-up collective action.

To test empirically what determines the refugees’ location in this second phase, we analyse data

concerning the Italian provinces which do and do not host asylum seekers using a logic regression

model. The results are very interesting insofar as they have a stronger and apparently

counterintuitive conclusion.

On the one hand, economic variables play a role which is not very significant but is in line with the

expectations, with provinces with higher unemployment being more willing to host refugees.

Furthermore, the political variables play an almost insignificant role, with provinces administered

by the left hosting more but not significantly so.

On the other hand, it is social capital that is the most significant factor in explaining why some

14

provinces are hosting more refugees than others. However, the sign for social capital is negative,

implying that regions with more social capital are less likely to host the asylum seekers. This is

apparently counterintuitive, as social capital has been defined as “the ‘glue’ generating excess

cooperation” (Paldman, 2000, p. 629). Yet, we interpret this as evidence that, firstly, bridging social

capital cannot go as far as to connect people from different continents and cultural backgrounds

with a high dependence ratio. Secondly, the arrival of asylum seekers may be seen by people as

damaging to the social tissue of a community, especially in those places where it is particularly

cohesive. Consequently, it is easier for private entities and administrations in places with less

cohesive communities to bid to host refugees, as social ties are looser and self-identification is lower.

The literature already shows that there is a correspondence between places attracting more migrants

and lower social cohesion (e.g. Huggins and Thompson, 2015, in the case of the UK). However, what

is commonly investigated in the literature is the outcome of the spontaneous location of migrants,

who find better economic opportunities in some places, whereas in the case of refugees it is the local

communities which decide whether they are interested in welcoming them or not. This evidence

also integrates the literature on the ‘dark side of social capital’ (Bowles and Gintis, 2002; Fukuyama

2001; Portes, 1998; Putnam, 2000,) which observes that group solidarity in human communities is

often purchased at the price of hostility toward out-group members.

Part of this is also related to the fact that local communities do not normally see asylum seekers as

providing an economic spark, but more as a group which has to be sustained using the already

strained finances of the Italian state. These migrants are very different from scientists and other high

qualified persons who can boost regional growth by establishing knowledge links with other regions

abroad (Trippl, 2013). Even if the qualifications of asylum seekers are not high, Levie (2007) noticed

that immigrants are normally more likely to become entrepreneurs than lifelong residents. However,

he also noticed that ethnic minorities are less likely to be as such once their younger age is

considered. A very significant part of those seeking asylum in Italy come from African countries,

where the perception of social values toward entrepreneurship is higher than world average (Singer

et al., 2014). In addition, Italy is a country which, with respect to the rest of the European Union,

holds entrepreneurs in higher regard (ibid.). However, it is a country which has suffered a greater

public finance crisis than most parts of Europe (Moro and Becker, 2016) and newly arriving asylum

seekers, due to the length of bureaucratic procedures and the normal delay in acquiring linguistic

and institutional competences, are not expected to significantly contribute to the local economy for

some time.

15

The evidence presented here comes from a single country, Italy. Therefore, even though the fact that

it and Greece represent the southern doors of Europe makes it very important in terms of migration,

results cannot necessarily be generalized to other countries. For this reason, we see scope for further

research to be pursued in two directions. The first of these is comparison with different potentially

more internally homogenous European countries that have a slightly different cultural background

or a better economic situation. Doing this would allow researchers to see whether this this negative

correlation between social capital and openness to migration is peculiar to Italy or not. The second

direction is to test whether these attitudes also influence the location of refugees at the next stage,

once their asylum demands are accepted and they become free to choose their location and seek

employment.

16

References

Akgunduz Y.E., M. van den Berg and W. Hassink (2015), The impact of refugee crises on host labor markets: the case of the Syrian refugee crisis in Turkey, IZA Discussion Paper n. 8841.

Baez, J. E. (2011), Civil wars beyond their borders: The human capital and health consequences of hosting refugees, Journal of Development Economics, 96(2), 391-408.

Banfield, E. (1954), The moral basis of a backward society,

Borjas, G. J. (2015) Immigration and Globalization: A Review Essay, Journal of Economic Literature 53: 961–74.

Bowles, S., & Gintis, H. (2002) Social capital and community governance. The Economic Journal, 112: F419-F436.

Card, D. (1990), The impact of the Mariel Boatlift on the Miami labor market, Industrial and Labor Relations Review, 43(2), 245-257.

Carrington, W. J., and P. J. De Lima (1996), The impact of 1970s repatriates from Africa on the Portuguese labor market, Industrial and Labor Relations Review, pp. 330-347.

De Haas, H. (2010) Migration and Development: A Theoretical Perspective. International Migration Review 44 (1): 227–64.

Di Maria, C., and Stryszowski, P. (2009) Migration, Human Capital Accumulation and Economic Development, Journal of Development Economics 90 (2).: 306–13.

Dustmann, C., Fadlon, I., and Weiss, Y. (2011) Return Migration, Human Capital Accumulation and the Brain Drain, Journal of Development Economics 95 (1). 58–67.

European Commission (2015) A European Agenda on Migration. Communication from the Commission to the European Parliament, the Council, the European Economic and Social Committee and the Committee of the Regions. Brussels.

European Parliament, directorate general for internal policies. Policy department A: Economic and Scientific Policy (2016) Labour Market Integration of Refugees: Strategies and good practices.

Fonseca, M. and Fratesi, U. (eds.) (2016). Regional Upgrading in Southern Europe - Spatial Disparities and Human Capital. Berlin: Springer.

Glitz, A. (2012), The Labor Market Impact of Immigration: A Quasi-Experiment Exploiting Immigrant Location Rules in Germany, Journal of Labor Economics, 30(1):175-213.

Huggins, R. and Thompson, P. (2015). Culture and Place-Based Development: A Socio-Economic Analysis. Regional Studies, 49(1), 130–159.

Hunt, J. (1992), “The impact of the 1962 repatriates from Algeria on the French labor market, Industrial and Labor Relations Review, pp. 556-572.

International Organization for Migration (IOM) (2015) How the World Views Migration, Geneva.

Kanbur, R., and Rapoport, H (2005) Migration Selectivity and the Evolution of Spatial Inequality. Journal of Economic Geography 5 (1): 43–57.

Levie, J. (2007). Immigration, In-Migration, Ethnicity and Entrepreneurship in the United Kingdom. Small Business Economics, 28(2–3), 143–169.

Martìn, I, et al. (2016) From Refugees to Workers Mapping Labour-Market Integration Support Measures for Asylum Seekers and Refugees in EU Member States. Gu�tersloh: Bertelsmann Stiftung.

Maystadt J.-F. and G. Duranton (2014), The development push of refugees: evidence from Tanzania,

17

University of Pennsylvania, mimeo.

McKenzie, D., and H. Rapoport (2010), Self-selection patterns in Mexico-US migration: the role of migration networks, The Review of Economics and Statistics, 92(4), 811-821.

Moro, B., & Beker, V. A. (2016). Modern Financial Crises: Argentina, United States and Europe. Berlin: Springer.

Moser, P., A. Voena and F. Waldinger (2014), German-Jewish Émigrés and U.S. Invention, American Economic Review, 104(10):3222-3255.

Paldman, M. (2000) Social Capital: One or Many? Definition and Measurement, Journal of Economic Surveys, 14(5), 629-653.

Pedersen, P.J., Pytlikova, M. and Smith, N. (2008) Selection and Network Effects-Migration Flows into OECD Countries 1990-2000. European Economic Review 52 (7): 1160–86.

Percoco, M. (2016), Strategic Planning and Institutional Collective Action in Italian Cities, Public Management Review, 18(1): 139-158.

Percoco, M. and M. Giove (2009), Multilevel Governance at Work: Evidence from Structural Funds Management in Lombardia, Italy, Environment and Planning C: Government and Policy, 27(3):381-398.

Portes, A. (1998). Social capital: Its origins and applications in modern sociology. Annual Review of Sociology, 24: 1-24.

Putnam, R. D. (2000), Bowling alone: the collapse and revival of American community, New York: Simon & Schuster.

Rapoport H. and J. Fernandez-Huertas Moraga (2014), “Tradable Refugee-Admission Quotas (TRAQs), the Syrian Crisis and the New European Agenda on Migration”, IZA Journal of European Labor Studies, 2015: 4-23.

Rodriguez-Pose A. and V. von Berlepsch (2012), When migrants rule: the legacy of mass migration on economic development in the United States, Annals of the Association of American Geographers, 103(3):628-651.

Ruist, J. (2015), The Fiscal Cost of Refugee Immigration: The Example of Sweden. Population and development review, 41(4):567-581.

Ryan, D. (2010), The Civic Foundations of Fascism in Europe: Italy, Spain, and Romania, 1870–1945. Baltimore: Johns Hopkins University Press.

Satyanath, S., N. Voigtländer, H.-J. Voth (2013), Bowling for Fascism, Journal of Political Economy, forthcoming

Singer, S., Amoros, J. E. and Moska, D. (2014). Global Entrepreneurship Monitor, 2014 Global Report. Global Entrepreneurship Research Association (GERA).

Thielemann, E.; R. Williams; C. Boswell; and Matrix Insight Ltd. (2010), What system of burden-sharing between Member States for the reception of asylum seekers?. Study. Directorate General for Internal Policies, Policy Department C: Citizens' Rights and Constitutional Affairs, Civil Liberties, Justice and Home Affairs, European Parliament, Brussels.

Trippl, M. (2013). Scientific Mobility and Knowledge Transfer at the Interregional and Intraregional Level. Regional Studies, 47(10), 1653–1667.

United Nations High Commissioner for Refugees (UNHCR) (2015) Global report, Geneva.

United Nations, Department of Economic and Social Affairs, Population Division (2013). International Migration Report 2013, New York.

18

Wang, Z., De Graaff, T. and Nijkamp, P. (2016) Cultural Diversity and Cultural Distance as Choice Determinants of Migration Destination, Spatial Economic Analysis, 11:2, 176-200.

World Bank (2006), Global Economic Prospects 2006 Economic implications of remittances and migration, The International Bank for Reconstruction and Development / The World Bank , Washington, DC.

19

Figure 1: Social capital and refugees in Italian NUTS2 regions

Abruzzo

Basilicata

Calabria

CampaniaEmilia Romagna

Friuli VG

Lazio

Liguria

Lombardia

Marche

Molise

Piemonte

Puglia

sardegna

Sicilia

ToscanaTrentino AA

Umbria

Valle d'Aosta

Veneto

66.

57

7.5

8R

efug

ees

(per

'000

resi

dent

s; in

logs

)

3.2 3.4 3.6 3.8 4Social capital (blood donations per '000 residents; in logs)

20

Figure 2 Total number of refugees over the period 2003-2015.

13652237 2199 2428 2411

4388 3694 31433979 3979

10381

2075221613

0

5000

10000

15000

20000

25000

2003 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014 2015

21

Figure 3. Number of attendees in SPRAR on population across Italian NUTS3 regions, year

2015.

Source: elaboration of the authors on ISTAT and Servizio Centrale data

22

Table 1. Spatial criteria for hosting structures

2014-2016 call 2016-2017 call

Which entities are eligible to host asylum seekers and refugees?

Associations, foundations, entities of the church, public entities, social-private entities.

Experience required It is highly recommended to already have in place some projects for asylum seeker and refugees.

Which are the structures involved in SPRAR? How many places does each structure have?

• Apartments; • Small centres hosting maximum 15 people; • Medium centres hosting maximum 30 people; • Big centres hosting more than 30 people.

How many places in collective centres must be devoted to asylum seekers and refugees hosting?

70% minimum.

Additional places Structures are obliged to activate additional places if required by the Ministry.

No

Localization The structure should be localized in inhabited places, easily reachable by public transportation.

Source: elaboration of the authors

23

Table 2. Logit regression results Province of arrival of refugees -0.445 -1.856 -2.745 -3.023 -2.599 -2.455 -0.670 -0.491 0.589 -3.497 (0.542) (0.960) (1.192) (1.253) (0.861) (0.869) (0.567) (0.618) (0.855) (1.723) * ** ** *** *** ** Province with regional chief-town 1.464 2.252 1.331 1.739 2.525 2.942 2.370 1.457 1.547 3.789 (1.062) (1.240) (1.071) (1.223) (1.357) (1.656) (1.574) (1.063) (1.073) (2.158) * * * * Value added per capita -0.000146 -7.42e-05 -0.000213 (7.70e-05) (9.19e-05) (0.000171) * Unemployment rate 22.50 18.04 9.429 (10.36) (11.60) (12.83) ** Presence of no profit organizations -1.003 -0.917 -0.797 (0.263) (0.284) (0.316) *** *** ** -7.045 -20.49 -6.767 (9.502) (9.136) (10.73) ** Province administered by "centre-right" parties -0.271 -0.437 (0.671) (0.839) Province administered by "centre-left" parties -0.0242 0.676 (0.751) (1.009) Share of foreign residents 17.61 25.61 (11.80) (19.32) Constant 1.609 5.189 -0.421 1.815 8.364 7.978 2.255 1.732 0.0337 9.054 (0.337) (1.964) (0.955) (2.928) (1.885) (1.943) (0.448) (0.625) (1.067) (4.842) *** *** *** *** *** *** * Observations 110 110 110 110 110 110 110 110 110 110

Standard errors in parentheses (*** p<0.01, ** p<0.05, * p<0.1)

24

Table 3. Logit regression results with standard errors corrected for spatial effects Province of arrival of refugees -0.445 -1.856 -2.745 -3.023 -2.599 -2.455 -0.670 -0.491 0.589 -3.497 0.537 0.888 1.263 1.251 1.095 1.119 0.552 0.656 0.749 1.643 ** ** ** ** ** ** Province with regional chief-town 1.464 2.252 1.331 1.739 2.525 2.942 2.370 1.457 1.547 3.789 1.067 1.199 1.075 1.131 1.120 1.682 2.240 1.067 1.070 1.540 * ** * ** Value added per capita -0.000146 -7.42e-05 -0.000213 0.000 0.000 0.000 ** Unemployment rate 22.50 18.04 9.429 11.925 13.217 12.223 * Presence of no profit organizations -1.003 -0.917 -0.797 0.342 0.383 0.399 *** ** ** -7.045 -20.49 -6.767 9.506 10.118 9.740 ** Province administered by "centre-right" parties -0.271 -0.437 0.660 0.776 Province administered by "centre-left" parties -0.0242 0.676 0.810 0.864 Share of foreign residents 17.61 25.61 9.808 17.823 * Constant 1.609 5.189 -0.421 1.815 8.364 7.978 2.255 1.732 0.0337 9.054 0.331 1.664 1.083 2.785 2.557 2.688 0.457 0.657 0.908 4.808 *** *** *** *** *** *** * Observations 110 110 110 110 110 110 110 110 110 110

Spatially corrected standard errors in parentheses (*** p<0.01, ** p<0.05, * p<0.1)