96198 Migrants’ Remittances from Italy International remittances and access to financial services for migrants in Turin, Italy A Greenback 2.0 Report MAY 2014 i Migrants’ Remittances from Italy International remittances and access to financial services for migrants in Turin, Italy A GREENBACK 2.0 REPORT A survey by MAY 2014 iii CONTENTS FOREWORD AND ACKNOWLEDGMENTS . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . v LIST OF TABLES AND FIGURES . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . vii EXECUTIVE SUMMARY . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . xi INTRODUCTION . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1 INTERNATIONAL REMITTANCES AND THE ECONOMIC CRISIS: THE CONTEXT OF THE SURVEY . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3 The world scenario: large and growing remittance flows. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3 Italy and Piedmont: a context of severe and protracted economic crisis . . . . . . . . . . . . . 4 THE GREENBACK 2.0 SURVEY: RESEARCH OBJECTIVES AND DATA COLLECTION . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7 Work plan and survey design . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7 Research objectives . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7 Sample structure definition . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7 Sampling method and surveying technique . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8 EMPIRICAL FINDINGS . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9 The demographic and economic profiles . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9 Job type, occupation, and sectors . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 11 Individual income . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 15 Savings & financial inclusions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 16 Savings . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 16 Banking products and services . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 16 Access to credit . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 18 Microcredit . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 21 Remittance behaviors: flows and recipients, operator and costs . . . . . . . . . . . . . . . . . . . . . . 21 Main recipients and motivations. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 21 Channels and Remittance Service Providers . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 23 Volume per transaction and per year . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 27 Transaction costs: a distorted perception . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 27 Trends in remittances. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 30 Received remittances . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 31 CONCLUSIONS . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 35 REFERENCES . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 37 iv GREENBACK 2.0 REPORT 2014 ANNEX I—FROM THE SURVEY DESIGN TO THE FIELDWORK: A METHODOLOGICAL NOTE . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 39 The Questionnaire . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 39 Training of interviewers and pilots . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 40 The fieldwork: sampling strategy and entry points . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 40 Feedbacks from the fieldwork. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 42 Data entry, cleaning and analysis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 43 ANNEX II—DETAILED STATISTICS . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 45 A. Italy’s remittance and foreign resident population data . . . . . . . . . . . . . . . . . . . . . . . . . . . . 45 B. Survey data: remittances . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 47 v FOREWORD AND ACKNOWLEDGMENTS W ith over $400 billion officially recorded annual flows to developing countries, remittances are a critical source of national and families’ income. However, high transaction costs and other inefficiencies in the process often adversely impact migrant workers and their families. The G8 and G20 have, therefore, set among their objectives the reduction of the average cost of remittances by five percentage points in five years (5x5 objective). Achieving the objective would save $16 billion per year: these funds would simply remain to migrants and their families and could contribute significantly to improving the living conditions of the migrants themselves as well as reducing poverty in their countries of origin. Project Greenback 2.0 aims at increasing efficiency in the market for remittances through an innovative approach: promote change inspired by the real needs of the ulti- mate beneficiaries of international money transfers: the migrants and their families. In Project Greenback 2.0, Remittances Champion Cities are selected. The World Bank work to implement initiatives aiming at increasing transparency and efficiency in the market for remittance services. The focus is on migrants and their needs. Cooperation between migrants, remittance service providers, and public authorities is key for the achievement of the Project’s objectives. Turin, Italy is the first “Remittances Champion City.” In this context, the World Bank commissioned this study, which is the result of the joint efforts of FIERI (Forum of International and European Research on Immigration) and Labor (Laboratorio R. Revelli). The team included Ferruccio Pastore (Director, FIERI), Eleonora Castagnone (Senior Researcher, FIERI), Claudia Villosio (Senior Researcher, Labor), and Laura Bartolini (Junior Researcher, Labor), whom we all wish to thank. The fieldwork was coordinated by Laura Bartolini and Marta Pinto (Junior Researcher, FIERI). Data were processed and analyzed by Laura Bartolini and Claudia Villosio. The team wish to thank FIERI’s inter- viewers for their great work: Serena Anastasi, Dario Basile, Elena Evangelisti, Agnese Migliardi, Emanuela Roman. Project Greenback 2.0 is coordinated by Marco Nicolì (Financial Sector Specialist, World Bank). In Turin, the team is coordinated by Viviana Premazzi and includes Romina Car- puci, Leyden Durand, and Khaled Elsadat. I would like to express our gratitude to all of the participating communities, groups, and associations, the City of Turin and other local authorities and, most importantly, the interviewed migrants who gave us their time and their attention. Massimo Cirasino Manager, Financial Infrastructure and Remittances Payment Systems Development Group The World Bank LIST OF TABLES AND FIGURES vii List of Tables and Figures Table 1: Remittance outflows, 2007–2012: first 10 Italian provinces (€ thousand). . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5 Table 2: First three countries of destination in 2012, first 10 Italian provinces . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6 Table 3: Main characteristics of the sample, by country of origin . . . . . . . . . . . . . . . . . 9 Table 4: Documents at the time of first entry in Italy . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 10 Table 5: Average number of relatives, by country of residence . . . . . . . . . . . . . . . . . . . 11 Table 6: Job distribution by type of contract for those arrived before and after 2008 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 13 Table 7: Distribution by type of occupation,* by gender . . . . . . . . . . . . . . . . . . . . . . . . . . . 14 Table 8: Individual income (mean, se mean*) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 15 Table 9: To whom do you send money? Remittance recipients, percent of total flows . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 21 Table 10: Average amount sent per year, by recipient . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 22 Table 11: Number and share of flows per channel, by country . . . . . . . . . . . . . . . . . . . . . 24 Table 12: Average amount sent per flows, by type of channel used . . . . . . . . . . . . . . 27 Table 13: Frequencies of transaction per flow, percent of total flows . . . . . . . . . . . . 28 Table 14: Average amount sent per transaction and per year . . . . . . . . . . . . . . . . . . . . . 28 Table 15: Average cost per transaction as percent of the amount sent, by type of channel used . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 29 Table 16: Comparison between actual average costs and perceived average cost . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 29 Table 17: Cost composition (percent), by type of channel used . . . . . . . . . . . . . . . . . . . 29 Table 18: Intercepted migrants excluded from the survey . . . . . . . . . . . . . . . . . . . . . . . . . 43 Table 19: Remittance outflows by Italian Regions, 2012 . . . . . . . . . . . . . . . . . . . . . . . . . . . . 45 Table 20: Remittance outflows by destination country, 2012 . . . . . . . . . . . . . . . . . . . . . 46 Table 21: Remittance outflows from the Province of Turin, first 20 destination countries 2005–2012 (€ million) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 46 Table 22: Foreign resident population in the Province of Turin, first 20 origin countries, 2003–2011 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 47 Table 23: Flows’ recipients, by main channel . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 47 Table 24: Average amount sent per year (classes), by country . . . . . . . . . . . . . . . . . . . 48 Table 25: Average amount sent per year, by recipient . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 48 Table 26: Average amount sent per year, by type of channel used . . . . . . . . . . . . . . . 48 Figure 1: International financial flows, 1990–2016. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3 Figure 2: Foreign resident population and remittances in Italy, 2005–2012 . . . . . . 5 Figure 3: Remittances from the Province of Turin, 2012, first three destination countries (€ million) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6 viii GREENBACK 2.0 REPORT 2014 Figure 4: The three phases of the research . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7 Figure 5: Municipality of Turin: districts . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8 Figure 6: Transitions in legal status, between the initial period of residence in Italy and 2013 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 11 Figure 7: Distribution by number of relatives residing in the country of origin . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 12 Figure 8: Distribution by number of children left in the country of origin (percent of those with children) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 12 Figure 9: Hours worked per week (percent) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 13 Figure 10: Sectors of occupation,* percent of total sample . . . . . . . . . . . . . . . . . . . . . . . 14 Figure 11: Individual income class, by country of origin (percent) . . . . . . . . . . . . . . . . . 15 Figure 12: Individual income: trend since the arrival in Italy (percent of total subsample) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 16 Figure 13: Could you save since the beginning of 2013? If yes, how much? . . . . . . . 17 Figure 14: Type of savings (percent of those who save) . . . . . . . . . . . . . . . . . . . . . . . . . . . . 17 Figure 15: Bank account ownership . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 18 Figure 16: Types of bank account (percent of total sample). . . . . . . . . . . . . . . . . . . . . . . 18 Figure 17: Number of bank accounts in the HH (percent). . . . . . . . . . . . . . . . . . . . . . . . . . 19 Figure 18: Services and products used (percent of total sample, more than one option allowed) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 19 Figure 19: Type of cards owned (percent of total sample, more than one option allowed) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 19 Figure 20: Loans’ distribution, classes of € (percentage of migrants with a loan) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 19 Figure 21: To whom did you ask for money? (percent). . . . . . . . . . . . . . . . . . . . . . . . . . . . . 20 Figure 22: In case you will need a loan, to whom would you ask? . . . . . . . . . . . . . . . . 20 Figure 23: Have you ever heard about microcredit? If yes, did you ever address a microcredit agency? . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 21 Figure 24: What is the money sent used for (percent more than one answer allowed) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 23 Figure 25: Types of channel used to send remittances (percent of total flows) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 24 Figure 26: Operator used, by type (percent of total flows). . . . . . . . . . . . . . . . . . . . . . . . 25 Figure 27: Bank operators (percent of total bank operators used) . . . . . . . . . . . . . . . 25 Figure 28: Where the transaction took place? (percent of total flows) . . . . . . . . . . 26 Figure 29: How long does it take for the money to be available at destination? (percent by channel) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 26 Figure 30: How did you choose your main channel? (more than one answer allowed) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 27 Figure 31: Pros and cons of main channels, percent . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 30 Figure 32: Did you ever have a problem? If yes, which kind of problem? . . . . . . . . 30 Figure 33: In case you had problems, did you look for help? . . . . . . . . . . . . . . . . . . . . . . . 31 LIST OF TABLES AND FIGURES ix Figure 34: Since your arrival in Italy, is your main channel changed? (percentage of total sample) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 31 Figure 35: What kind of channel did you use at that time? (percentage of those that have changed) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 31 Figure 36: Trend in total amount sent since the arrival in Italy, by country of origin . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 31 Figure 37: Reasons for a decrease in remittance amount (percent) . . . . . . . . . . . . . . 32 Figure 38: Received remittances, by country and person of origin. . . . . . . . . . . . . . . 32 Figure 39: Questionnaire structure. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 39 Figure 40: Map of Turin and data collection points . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 42 xi EXECUTIVE SUMMARY “G reenback 2.0–Torino” is a project led by the World Bank, in partnership with the City of Turin, aimed at fostering the development of an efficient and transparent market for remittances. Within this framework, this survey, conducted by FIERI and Laboratorio R. Revelli between July and September 2013, investigated the demand side of the remittance market in the city of Turin, collecting 480 in-depth interviews of migrants from the Moroccan, Peruvian and Romanian communities. These are the three most numerous communities in the city and the largest in terms of remittances sent to the country of origin. The survey is aimed at describing the economic and financial profile of migrants. Specifically, it investigates the interconnected dimensions of financial behaviors—patterns of consumption and savings, remittances and investments, and financial needs and aspirations. Strong emphasis is given to remittance services and access to different types of financial products and services. Interviewed migrants differ by country of origin, gender, length of stay in Italy, type of occupation and related economic stability. Moroccans and Peruvians tend to have larger families than Romanians, considering both relatives in the origin country and in Italy. Moroccan migrants show on average a longer presence in Italy and hence a higher rate of family reunification than both Peruvians and Romanians, which instead often report to have children left in the country of origin. As for economic conditions and level of inte- gration into the Italian labor market, differences arise across the subsamples and by gen- der. Romanians report the highest average monthly income, while Moroccans have the lowest earnings. This differentiation is consistent with the job and sector specialization of the three subsamples: Moroccan migrants, mostly men, are employed in manufactur- ing and construction, while many Peruvians and Romanians, especially among women, are employed in the care and domestic sector and seem to have more stable and higher earnings. Moreover, the relatively better economic condition among Romanians can also be attributed to their acquisition of the EU citizenship in 2007, which may facilitate their economic integration and stability. Collected data suggests, on average a good level of economic and financial integration of interviewed migrants, who, on average, declared high usage of financial products and instruments connected to a bank account, while microcredit is not popular and used very marginally. As for the remittance service providers and channels for sending remittances, 83 per- cent of all recorded remittance flows are sent in cash through a money transfer opera- tor, and only a very small proportion through a bank account, a prepaid card, or twin cards. Although transaction costs are often deemed to be the crucial element in decid- ing if and how often to send remittances, remittance behaviors also seem to respond to other important characteristics of each channel and service provider, such as the xii GREENBACK 2.0 REPORT 2014 availability of agencies at origin and at destination, the easy access for recipients to withdraw the money, the relative transaction speed, etc. The analysis depicts a widespread misperception of transaction costs and overall service conditions by migrants. Remitters have low awareness of costs’ composition and often do not consider the presence of exchange rate’s margins and fixed costs at destination, irrespective of the preferred remittance service provider. Moreover, highly frequent flows are associated with low-value transactions, which imply an overall higher weight of fixed costs on the total amount sent per year. These characteristics suggest there is room for improvement in terms of competitiveness and transparency of all remittance service providers and the need for greater awareness and financial literacy and management skills among remitters. The survey also provides a picture of trends and changes in migrants’ behaviors since arrival in Italy and demonstrates the impact of the current economic crisis on income and remittance flows. Declining economic conditions in the migrant communities depends, among other things, on different job specializations within the Italian labor market. Nevertheless, our empirical data show on average a severe situation in which the protracted economic instability is threatening the overall level of economic integra- tion of migrants and their capacity to keep remittance flows constant over time. In this regard, it is worth noting the role of reverse remittances, which represent the support from the origin household to migrants’ income to overcome the period of crisis. Never- theless, comparing trends in income and remittances, for the time being migrants’ remit- tances seem to be more resilient than income to the worsening of economic conditions in Italy in the last five years. 1 1 INTRODUCTION M igrants’ remittances represent a substantial amount of money both in terms of contribu- tion to family income in the origin country and countries is a crucial issue in the current aca- demic and political debate. Trends and impacts of the economic crisis reflect themselves in micro- the aggregated financial inflows for receiving behaviors and affect also the transnational activi- economies. ties of migrants. Chapter 3 presents the objectives of the research and the related survey design and In 2013, the World Bank launched “Project Green- sampling techniques for the fieldwork, while a back 2.0—Remittances Champion Cities” and specific methodological note is added at the end selected Turin, which is located in the Northern of the Report on the fieldwork phase. Chapter 4 part of Italy, as the first “Remittance Champion focuses on empirical findings from the analysis of 1 City.” The Project aims at investigating the local the 480 interviews collected. After a brief descrip- market for international remittances and at favor- tion of demographic and economic conditions of ing changes both on the supply and demand side, interviewed migrants, the core of this section is in order to boost transparency and efficiency devoted to the analysis of migrants’ attitudes and and respond to the actual needs expressed behaviors in sending remittances, in engaging in by remittance senders and beneficiaries—the investments and savings, and to explore changes migrants, their families, and their origin countries across time and according to their occupational in general. status at destination. In light of the global com- mitment to the reduction of remittance costs2 as As a part of the Project, FIERI (Forum of Interna- a part of a more comprehensive strategy to foster tional and European Research on Immigration) the migration and development nexus, a specific and Labor (Laboratorio R. Revelli) conducted a section deals with remittance service provid- survey on migrants residing in Turin to explore ers and remittance costs revealing still room for their level of financial inclusion and their preva- improvement for the development of a competi- lent practices in sending remittances to their ori- tive and transparent remittance market. gin countries, with special attention to changing trends in a time of deep economic crisis. The conclusive chapter highlights the main empir- ical findings and suggests further lines of analysis This Report starts by presenting the international on migratory models, migrant economic inte- and Italian context of international remittance gration and transnationalism from the collected flows in the past decade, as a general framework empirical evidence. of the study (Chapter 2). The connection between the labor market integration of migrants with the 2 A global effort is in place for the reduction of remittance prices. The so- more general economic trends in their destination called “5x5 objective” was adopted by the G8 at the 2009 L’Aquila Summit where the commitment was made “to achieve in particular the objective of a reduction of the global average costs of transferring remittances from the present 10 percent to 5 percent in five years.” In 2010, the G20 committed to 1 For more information on Project Greenback 2.0 visit http:// a “significant reduction in the cost of remittances” and established a Devel- remittanceprices.worldbank.org opment Action for Remittances. 3 INTERNATIONAL REMITTANCES 2 AND THE ECONOMIC CRISIS: THE CONTEXT OF THE SURVEY cycle. Indeed, remittances remained remarkably The world scenario: large and stable in the wake of the recent financial crisis, growing remittance flows compared to other types of international financial Since the late 1990s, officially recorded monetary flows (World Bank, 2011). Moreover, compared remittances sent by international migrants to their to official development assistance, which repre- origin countries have exceeded official develop- sents all financial flows provided by national and ment assistance (ODA) and portfolio investments. multilateral official agencies in order to promote Remittance growth rate in real terms has been economic development and welfare of develop- impressive, especially in the decade preceding ing countries, remittances seem to have a more the 2008 financial crisis (1999–2008). Re-gaining direct impact in fighting poverty and promoting momentum after the negative shock experienced social development at the household level (see for during 2008–2010 as a consequence of the global example de Haas, 2007). economic and financial crisis, remittance flows to The possible risks of imbalances in the exchange developing countries reached an estimated $401 rate, due to a non-negligible inflow of foreign billion in 2012 (+5.3 percent compared with 2011: currencies (US dollars and euro, above all) in World Bank, 2013a) and are expected to increase receiving countries, have proved to be serious by 6.3 percent to reach $414 billion in 2013 (World only in very few cases (El Salvador and Moldavia, Bank, 2013b). see Ratha, 2007). Still receiving countries should Compared to foreign direct investments (FDI), avoid becoming excessively dependent upon remittances proved to be less volatile and more remittances since significant annual variations in resilient to idiosyncratic shocks of the economic inflows are fairly common at the country level. FIGURE 1: International financial flows, 1990–2016 700 US$ billion 600 FDI Remittances 500 400 300 Pvt debt & portfolio equity 200 ODA 100 0 1990 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013e 2014f 2015f 2016f Source: World Bank, 2013b:2 4 GREENBACK 2.0 REPORT 2014 Many important issues are related to the study interconnections between informal systems and of remittance flows and of migrants’ behavior in international criminal activities (money launder- sending money to their households and countries ing, terrorism, etc.; World Bank, 2013b). Types of of origin. International development institutions, RSP and channels differ a lot in terms of transfers’ academics and policy makers have progressively speed, distribution of agencies in both sending integrated migration and remittances into the and receiving countries, administrative burdens development discourses, policies and programs. and costs. The issue of sending costs is one of Motivations behind remittances have been deeply the most debated at the international level (Yang, investigated (see for example Rapoport and Doc- 2011; Gibson et al, 2006) and one where the mul- quier, 2005) as well as the extent to which remit- tilateral action taken by the World Bank and other tances affect recipient households’ wellbeing, organizations with private actors is more effec- investment and consumption levels in the receiv- tive. Field studies have shown that terms and con- ing countries all play an insurance role against ditions of remittances services are too often not external shocks. Also the study of remittances fully and easily available to senders, who might may shed some lights on intra-household resource not be able to perceive all the components of the allocation, disentangling preferences and behav- final cost and to then choose the more appropri- iors of migrants and individual household mem- ate RSP for their needs (World Bank, 2007). The bers that receive the money (Yang, 2011). structure and accessibility of the remittance mar- ket is also strictly connected to the wider issue of Both receiving and sending countries play a migrants’ financial inclusion. role in building an appropriate and transparent framework to manage remittance flows in the most efficient and productive way at the macro Italy and Piedmont: a context as well as at the micro (household) level. In this of severe and protracted respect, immigration countries can act on regula- tion of remittances service providers (RSP) and economic crisis channels in order to facilitate access to formal Our empirical investigation has been conducted and transparent services for remitters, to dimin- in a context of severe and protracted economic ish costs and inefficiencies and to lower the crisis in Italy, which since 2008 has dramatically FIGURE 2: Foreign resident population and remittances in Italy, 2005–2012 5.0 8,000 7,500 4.5 7,000 6,500 4.0 6,000 3.5 5,500 5,000 3.0 4,500 4,000 2.5 3,500 2.0 3,000 2005 2006 2007 2008 2009 2010 2011 2012 Foreign resident population,* millions Remittance outflows, millions of € (right axis) Source: Istat and Bank of Italy. *Break in series in 2011: from 2012 data are adjusted for new Census data. INTERNATIONAL REMITTANCES AND THE ECONOMIC CRISIS: THE CONTEXT OF THE SURVEY 5 affected both the native and the immigrant popu- striking if one considers that the total migrant lation. Immigrant workers are often deemed to population residing in the country kept growing, suffer more in terms of employment and income although at a slower pace, also in the last years of reduction than Italian workers because of their economic downturn. relatively weaker contractual position, their lower education level, their concentration in highly pro- Almost half of the entire remittances flow is sent cyclical economic sectors such as manufacturing from only two regions, Lazio and Lombardy. and the construction industry (among the more Piedmont represents 4.2 percent of remittances recent contributions on this issue, see Fondazione outflows and the Province of Turin is the seventh Leone Moressa, 2013; Pastore et al, 2013; Bonifazi for volumes after those of Rome, Milan, Naples, and Marini, 2013; Ministero del Lavoro, 2013; but Prato, Florence, and Catania. While China alone is also Ricucci, 2011). the destination of almost 40 percent of all remit- tances sent from Italy, differences at the province Data on remittance outflows confirm the nega- level reflect the different composition of migrant tive economic conditions of migrants residing population across Italy (see the Annex I-A for in Italy. Differing from the global trend, remit- more detailed data). tances from Italy showed a sensible decline in 2011–2012, with the overall amount sent in 2012 In line with national and local official data, our close to the level of 2009. Official data provided study confirms the negative trends in income and by the Bank of Italy on MTOs, banks, Poste Ital- economic stability for migrants residing in Turin iane, and other financial institutions (Fondazione (see Fig. 3). Although there are non-negligible Moressa, 2013:111–112), show that the decrease in differences among migrant communities, which the total remittance volume is due more to the are mainly driven by their different economic decrease in the number of remitting migrants integration and labor market specializations, the than to the decrease in the amount sent by each average economic conditions are worsening since of them (on average, 1673 € per year in 2012). the outbreak of the economic crisis in 2008 with The decrease in the number of remitting migrants a sensible impact on remittance behaviors and and in the total outflows from Italy is even more prospects (Fullin & Reyneri, 2013). TABLE 1: Remittance outflows, 2007–2012: first 10 Italian provinces (€ thousand)   2007 2008 2009 2010 2011 2012 2011–2012   a.v. a.v. a.v. a.v. a.v. a.v. % % var. Rome 1,500,353 1,697,718 1,784,703 1,786,274 2,040,017 1,938,168 28.40 –5.00 Milan 824,860 862,825 897,412 941,826 1,031,305 965,969 14.10 –6.30 Naples 170,810 183,887 240,856 225,751 305,707 295,600 4.30 –3.30 Prato 449,739 415,823 485,559 191,699 249,102 208,458 3.10 –16.30 Florence 244,295 254,110 253,728 207,345 233,604 197,194 2.90 –15.60 Catania 50,754 57,736 77,992 93,402 156,095 178,292 2.60 14.20 Turin 180,411 180,361 180,262 180,538 193,321 164,577 2.40 –14.90 Brescia 127,297 132,627 131,617 132,094 152,763 134,645 2.00 –11.90 Genoa 95,313 109,471 116,682 119,319 122,450 110,734 1.60 –9.60 Boulogne 126,135 138,722 130,773 130,700 131,858 108,989 1.60 –17.30 Total 6,039,255 6,376,949 6,747,818 6,572,224 7,394,398 6,833,116 100 –7.60 Source: Own calculations based on Bank of Italy dataset on remittances (last update October 2013). 6 GREENBACK 2.0 REPORT 2014 TABLE 2: First three countries of destination in 2012, first 10 Italian provinces   1° Row % 2° Row % 3° Row % Tot, % Rome China 72.6 Philippines 6.0 Romania 5.4 84.0 Milan China 46.1 Philippines 12.5 Peru 7.0 65.6 Naples China 53.9 Ukraine 7.1 Romania 2.7 63.7 Prato China 90.0 Romania 1.6 Morocco 1.5 93.1 Florence China 36.5 Philippines 10.6 Peru 9.5 56.6 Catania China 76.5 Romania 7.9 Sri Lanka 2.2 86.6 Turin Romania 28.2 Peru 10.8 Morocco 8.2 47.2 Brescia China 14.7 India 13.5 Romania 10.6 38.8 Genoa Ecuador 28.1 Romania 9.1 Senegal 6.7 43.9 Boulogne Philippines 15.0 Romania 13.8 China 13.2 42.0 Total China 39.1 Romania 11.9 Philippines 5.4 56.4 Source: Own calculations based on Bank of Italy dataset on remittances (last update October 2013). FIGURE 3: Remittances from the Province of Turin, 2012, first three destination countries (€ million) 60 50 40 30 20 10 0 2005 2006 2007 2008 2009 2010 2011 2012 Romania Peru Morocco Source: Own calculations based on Bank of Italy dataset on remittances (last update October 2013). 7 THE GREENBACK 2.0 SURVEY: 3 RESEARCH OBJECTIVES AND DATA COLLECTION migratory experience in Italy, also highlighting the Work plan and survey design impact of the economic crisis on remittance flows. As illustrated in Figure 4, the research was com- In particular, the study aimed at investigating two posed by three interconnected phases: 1) the interconnected dimensions: research design and the definition of the sample • Financial behaviors: remittance sending, structure, 2) the data collection, registration, and consumption and savings, investments in analysis, and 3) the preparation of the final report. the origin country, insurance and other financial services; This section presents the main characteristics of the research and survey design, while a more • Financial needs and aspirations in terms of detailed description of the fieldwork, with notes costs, availability, and accessibility to different on questionnaire’s definition process, data collec- types of financial products and services, with tion and registration is presented in the Method- stronger attention to remittance services. ological Note (Annex I). SAMPLE STRUCTURE DEFINITION RESEARCH OBJECTIVES In 2012, the number of foreign citizens residing in The main objective of the Study was to explore the Municipality of Turin was 142,157, represent- financial needs and behaviors of migrants resid- ing 16 percent of the total population (Comune di ing in the city of Turin, with a focus on patterns Torino, 20133). According to this data, Romania, of remittances, financial inclusion, and access to Morocco and Peru are the first three countries of banking services in Italy. The survey investigates origin, representing almost 60 percent of total migrants’ relation with banks, their savings allo- migrant population. The largest represented immi- cation in Italy and their origin countries, and the grant community is the Romanian community possible changes since the beginning of their (56,438 individuals, 40 percent of the total immi- grant population), which mostly resides in the 5° District Borgata Vittoria, Le Vallette, Madonna Di Campagna (ibid.). The second largest community FIGURE 4: The three phases of the research is the Moroccan community (20,577 individuals, 14 percent of total migrant population), which is mainly concentrated in the 6° District Barriera di Milano, Falchera, Regio Parco. Peruvians repre- sent around 7 percent of total migrant population (9,569 individuals) and are mostly found in the 3° District Cenisia, Pozzostrada, San Paolo (see Fig. 5). 3 Available figures at the municipality level are not yet adjusted for the new data from the last general Census of 2011. Istat (Italy’s National Institute of Statistics) is currently updating and adjusting population historical data since 2001. See Annex II-A for historical trends in foreign resident population in the Province of Turin by country of origin. 8 GREENBACK 2.0 REPORT 2014 FIGURE 5: Municipality of Turin: districts an interesting picture of migrant population within each community. To be included in the sample, the interviewed migrants had to comply with four pre-requisites: 1. to have resided in Italy for at least one year (with or without a regular residence status); 2. to live in the city of Turin (or in one of the sur- rounding municipalities of the metropolitan area); 3. to have an income (through any type of job or occupation); and 4. to have sent remittances to his/her country of origin at least once since the beginning of 2013. Sampling method and surveying technique Romania, Peru and Morocco are also the first Although administrative data exist on the dis- three destination countries of remittances from tribution of migrant citizens in the city of Turin the Province of Turin, representing respectively (e.g. Anagrafe, the Municipal Residence Regis- 28.2 percent, 10.8 percent, and 8.2 percent of try), up-to-date data on their structure in terms all outflows (Fondazione Moressa, 2013; Banca of employment status and on the presence of d’Italia, 2013). At the same time, these three coun- irregular residents are scarcely available. Hence, tries differ in their geographical position, socio- since the reference migrant population is not economic condition, and modes of migration completely known in advance, a “center sampling to Italy. technique” has been applied in order to design a balanced sample (Blangiardo, 2004; Blangiardo For all these reasons, the research focuses on et al, 2011). these three nationalities and the sample is con- structed of three equally large subsamples As a preliminary step, a significant and repre- according to the citizenship at birth of the sentative number of “centers” in Turin has been interviewees: identified: migration association and organiza- tions, places of work, entertainment, health-care • Subsample 1—Short-range EU: Romania and meeting centers, social-informal services etc. • Subsample 2—Short-range non-EU: Morocco From that list of aggregation centers, the sample • Subsample 3—Long-range: Peru was defined by applying two basic criteria of the “snowball sampling” (see Annex I for more details The total sample size is 480 individuals. The sam- on this method): ple is composed of foreign-born individuals resid- ing in the city of Turin at the time of the interview, • no more than 10 contacts/interviews from the including naturalized immigrants, from the three same entry point or key informant (reference selected countries, between 18 and 64 years old. person); • no more than five contacts/interviews from an Each subsample is stratified according to the individual already included in the sample. length of stay in Italy, with around 20 percent of interviewed migrants who have a ‘short’ migra- The survey has been conducted by a team of six tory experience (of five years or less) and arrived interviewers through paper and pencil interviews after 2008. Subsamples are not equally stratified (PAPI), using printed questionnaires administered by gender, which has been assumed to be less face-to-face and lasting about 20–30 minutes significant with regard to the survey’s objectives. each. The questionnaire is structured with closed- Nevertheless, the final gender composition offers ended questions, skips and connected sections. 9 4 EMPIRICAL FINDINGS Nevertheless at the subsample level, gender dif- The demographic and ferences are more evident: of the migrants inter- economic profiles viewed women represented 61 percent of the The average profile of the final sample is reported Romanians, 54 percent of the Peruvians, and only in Table 3, which includes statistics on gender, 14 percent of the Moroccans. Indeed, interviewers age, length of stay in Italy, and education level of found it particularly difficult to intercept Moroc- the 480 interviewed individuals. can women complying with the survey criteria, mainly because of the low activity rate of female The overall sample is relatively gender-balanced Moroccans compared to the one of their Peruvian (43 percent of interviewed migrants are women). and Romanian counterparts. TABLE 3: Main characteristics of the sample, by country of origin Morocco Peru Romania Total Freq. Col % Freq. Col % Freq. Col % Freq. Col % Sex Male 136 85.53 74 45.96 62 38.75 272 56.67 Female 23 14.47 87 54.04 98 61.25 208 43.33 Arrival in Before 2008 128 80.5 117 72.67 121 75.63 366 76.25 Italy After 2008 31 19.5 44 27.33 39 24.38 114 23.75 Age class 18–24 17 10.69 2 1.24 19 11.88 38 7.92 25–29 18 11.32 8 4.97 23 14.38 49 10.21 30–39 69 43.40 63 39.13 56 35.00 188 39.17 40–49 44 27.67 52 32.30 38 23.75 134 27.92 50–59 9 5.66 30 18.63 22 13.75 61 12.71 60+ 2 1.26 6 3.73 2 1.25 10 2.08 Education None 6 3.77 0 0.00 0 0.00 6 1.25 level Elementary 7 4.40 0 0.00 2 1.25 9 1.88 Lower second 63 39.62 26 16.15 21 13.13 110 22.92 Upper second 58 36.48 95 59.01 94 58.75 247 51.46 Post-sec, non ter. 0 0.00 1 0.62 27 16.88 28 5.83 Short-cycle ter. 1 0.63 1 0.62 0 0.00 2 0.42 Bachelor 11 6.92 10 6.21 9 5.63 30 6.25 Master 13 8.18 28 17.39 7 4.38 48 10.00 Marital Married 83 52.2 78 48.45 74 46.25 235 48.96 status Separated 8 5.03 14 8.7 6 3.75 28 5.83 Divorced 7 4.4 13 8.07 19 11.88 39 8.13 Widow 1 0.62 2 1.25 3 0.63 Single 56 35.22 34 21.12 32 20.0 122 25.42 Cohabiting 5 3.14 21 13.04 27 16.88 53 11.04 10 GREENBACK 2.0 REPORT 2014 TABLE 3: Main characteristics of the sample, by country of origin (continued) Morocco Peru Romania Total Freq. Col % Freq. Col % Freq. Col % Freq. Col % Italian citizenship acquisition 15 9.43 10 6.21 3 1.88 28 5.83 By gender M F M F M F M F Age, avg. 36.6 37.1 42.1 41.1 35.9 38.1 37.9 39.3 Length of stay in Italy, avg. 10.9 10.6 8.8 10.7 9.4 8.6 10.0 9.7 Years of schooling, avg. 10.5 12.4 13.1 13.4 12.5 13.0 11.7 13.1 Total 159 100 161 100 160 100 480 100 * Education levels follow the ISCED international classification (Unesco). The post-secondary non tertiary class includes all vocational or university courses after the high school diploma and which last no more than one year. With regard to formal qualifications and compe- achieved their highest education level in Italy and tences, the number of migrants with an elemen- that 35.6 percent of them have attended some tary level of education (or lower) is low for each sort of professional and/or vocational training of the three subsamples, and the average level of courses in Italy, with high variations in terms of education of the sample is relatively high (10 per- length (from one month to two years) and subject cent hold a master’s degree, 7 percent a bachelor of the course (from carpentry to health care, from degree, 51 percent a high school level degree). computer science to cooking). Here too, however, differences among the three sub-groups are evident. Peruvians show the Almost half of the interviewed migrants are mar- highest level of education attained (59 percent ried. Moroccans have the higher level of single with a high school diploma and 17 percent with individuals (35 percent), while among Romanians a master’s degree), closely followed by Roma- and Peruvians there is a higher incidence of sepa- nians (59 percent with a high school diploma, rated or divorced individuals (respectively 18 per- 17 percent with a non-university qualification after cent and 16 percent). the high school diploma). Almost 40 percent of At the time of first entry in Italy, one third of the Moroccans attained a lower secondary school sample had a tourist visa (33.3 percent), one diploma, while 37 percent an upper secondary quarter was totally undocumented (25.4 percent), diploma. In only 25 percent of cases migrants are and only a 15.6 percent of the sample had a visa sure that their level of education is recognized for work reasons (see Table 4). in Italy, regardless to where it was attained. It is worth noticing that 7 percent of migrants have TABLE 4: Documents at the time of first entry in Italy Morocco Peru Romania Total Freq. % Freq. % Freq. % Freq. % Visa—work 35 22.01 30 18.63 10 6.25 75 15.63 Visa—student 9 5.66 1 0.62 2 1.25 12 2.50 Visa—family reasons 19 11.95 19 11.80 2 1.25 40 8.33 Visa—tourism 26 16.35 65 40.37 68 42.50 159 33.13 Visa—not needed 8 5.03 1 0.62 53 33.13 62 12.92 Visa—other* 1 0.63 2 1.24 3 1.88 6 1.25 Undocumented 59 37.11 43 26.71 20 12.50 122 25.42 Do not answer 2 1.26 0 0.00 2 1.25 4 0.83 Total 159 100 161 100 160 100 480 100 * Other type of visas includes: religious, professional exchange, and visa released by another EU Member State. EMPIRICAL FINDINGS 11 As for the legal status, 25 percent of the inter- TABLE 5: Average number of relatives, viewees declared they were undocumented at by country of residence the time of their arrival in Italy, while in 2013 only 7.5 percent of them (all among Moroccans) are In the without a regular permit of stay. These figures Origin In Other are in line with the migration policy and regula-   Total Country In Italy Countries tion prevalent in Italy in the last two decades Morocco 7.00 4.72 2.16 0.27 (Salis, 2012; Reyneri, 2007), which have progres- Peru 6.92 4.24 2.45 0.21 sively restricted regular possibilities of entering Romania 4.58 2.38 2.06 0.16 the country to work and have at the same time allowed more than 1.6 million foreign individuals to regularize their legal status through succes- Moroccan and Peruvian families are larger, on sive amnesties between 1986 and 2009 (IDOS, average, than Romanian ones. Among those with 2013: 119). children, the vast majority reside with them: the number of children left in the country of origin is Probably because of their longer length of quite low among Moroccan migrants (24.3 per- stay in Italy, the highest share of EU long-term cent) and slightly higher for Peruvians (39.8 per- residence permits is found among Moroccans cent) and Romanians (31.7 percent). (27 percent of them), who also present the high- est rate of citizenship acquisition (9 percent of JOB TYPE, OCCUPATION, AND SECTORS the subsample). The vast majority of Peruvians hold a renewable residence permit linked to their All migrants included in the sample and inter- job position (75 percent of them), while Roma- viewed were employed, in order to ensure that nians no longer need a permit to stay or reside they had control over economic resources in rela- in Italy after the accession to the EU in 2007. tion to remittance decisions. This section briefly Romanians also seem to be less interested in the presents the job position, the prevalent sector naturalization processes. of occupation and the distribution of earnings of the sample. Table 5 and subsequent figures present the com- position of migrants’ families. Interviewees were Migrants were asked to describe their main job asked to list all their first grade relatives (parents, or occupation, the one they consider the most siblings, partners and children), residing in Italy, in important in terms of hours, earnings or stabil- their origin country or elsewhere. ity, without prejudice for informal occupations, FIGURE 6: Transitions in legal status, between the initial period of residence in Italy and 2013 The figure represents transition in legal status from “initial period of residence” (spell 1) to 2013 (spell 2), in absolute values by country of origin. All possible conditions—types of permit to stay, documents or reasons for the lack of them—were considered. The category “Not needed” includes migrants entered when Italy’s immigration law did not require a permit to stay, and European citizens (mostly Romanians after 2007). The “999” indicates the very low share of migrants which refused to answer. 12 GREENBACK 2.0 REPORT 2014 FIGURE 7: Distribution by number of relatives residing in the country of origin Morocco Peru Romania 39.6 40 35 34.6 32.3 32.3 30 25.5 25 23.4 20.5 20 17.7 15.8 16.4 15 10.6 10 5.7 5.0 5.0 6.3 5 3.2 2.5 1.9 1.2 0.6 0.0 0 0 1–2 3–4 5–6 7–8 9–10 11+ 0 1–2 3–4 5–6 7–8 9–10 11+ 0 1–2 3–4 5–6 7–8 9–10 11+ FIGURE 8: Distribution by number of children left in the country of origin (percent of those with children) Morocco Peru Romania 80 75.7 70 68.3 60.2 60 50 40 30 20 17.6 15.7 13.9 13.9 9.5 9.5 10 5.4 6.5 4.0 0 0 1 2 3+ 0 1 2 3+ 0 1 2 3+ non-regular jobs etc. Table 6 presents the dis- workers (27.5 percent and 58.3 percent respec- tribution of job types among migrants, who tively among those arrived before and after specified if the arrangement included a formal, 2008), although they represent less than a quar- written contract or not. Among dependent work- ter of the total sample. In particular, small retailers ers, 12 percent of those who arrived before 2008, and peddlers in outdoor markets often state to be and 21 percent of those that arrived after 2008, non-regular. declared to work without a regular contract. The majority of them declared to work as domestic Interviewees tended to declare the official num- workers (both health care and domestic ser- ber of hours worked per week, as stated in the vices) or in the food service (restaurants, small contract or in the agreement with the employer. ethnic food shops). The lack of regular contrac- However, while many of them reported to work tual arrangements is higher among autonomous part-time (less than 40 hours per week), they EMPIRICAL FINDINGS 13 TABLE 6: Job distribution by type of contract for those arrived before and after 2008 Before 2008 After 2008   Contract Contract   No Yes Total Tot % No Yes Total Tot % Dependent 46 232 278 76.0 24 76 100 87.7 Co.co.co/project worker 6 6 1.6 Occasional 4 5 9 2.5 1 1 0.9 Autonomous—employer 6 6 1.6 2 2 1.8 Autonomous—P.IVA* 2 16 18 4.9 Autonomous—own 13 18 31 8.5 6 2 8 7.0 Autonomous—family firm 1 1 0.3 1 1 0.9 Autonomous—coop. member 2 2 0.5 Coop. member & employee 15 15 4.1 2 2 1.8 Total 66 300 366 100 32 82 114 100 Total percent row 18.0 82.0 100   28.1 71.9 100   *Partita IVA is the Italian VAT identification number, assigned to firms and autonomous workers for value added tax purposes. also admitted to working more than the official specialization seems to apply to each of the three number of hours, where the difference is paid subsamples. Indeed, roughly half of the inter- outside the contract, with no fiscal contributions viewed women tend to be classified as qualified or taxes paid for them. Moreover, 19 percent of the workers in trade and services, a category which sample stated to have one or more second jobs comprises also qualified health workers. Men in addition to the one described in detail. Hence, seem to be more represented in the productive the total amount of hours worked as presented in sectors as industrial workers and farmers (66 per- Figure 9 and the overall level of irregular work is cent of Romanian males, 30 percent of Moroccan surely underestimated. males). High percentages of Moroccans and Peru- vians, both males and females, are employed as Following the official classification of profes- non-qualified workers in occupations within the sions provided by ISTAT, which distinguishes service and the productive sector, which do not jobs in terms of their content (tasks, responsibili- require any kind of specific skills or qualifications ties) and their level of qualification, a “gendered” (around 30 percent and 50 percent respectively). FIGURE 9: Hours worked per week (percent) Hours Per Week Worked (Percent) 45 39.38 40 35 30 25 20 16.88 14.79 15 9.79 10 7.50 5.83 5.83 5 0 0 0 0 0 0 0 0 1 –2 –3 –4 –5 –6 0 0– –1 11 21 31 41 51 61 14 GREENBACK 2.0 REPORT 2014 TABLE 7: Distribution by type of occupation*, by gender Morocco Peru Romania Total M F M F M F M F Freq % Freq % Freq % Freq % Freq % Freq % Freq % Freq % High professionals** 8 6.0 3 13.0 2 2.7 6 6.9 6 9.7 9 9.2 16 5.9 18 8.7 Qualified w. in trade & services*** 33 24.6 12 52.2 22 29.7 36 41.4 4 6.5 50 51.0 59 21.9 98 47.1 Workers & farmers 40 29.9 0 0.0 11 14.9 1 1.1 41 66.1 2 2.0 92 34.1 3 1.4 Drivers of machines & transp. 14 10.4 1 4.3 3 4.1 0 0.0 3 4.8 0 0.0 20 7.4 1 0.5 Non-qualified workers 39 29.1 7 30.4 36 48.6 44 50.6 8 12.9 37 37.8 83 30.7 88 42.3 Total 134 100 23 100 74 100 87 100 62 100 98 100 270 100 208 100 * Occupation categories follow international classifications and ISTAT codes, in the table above at the lowest level of disaggregation (1 digit). ** High professionals include Lawyers & managers, High technicians, High professionals, Executive officers. *** Qualified workers in trade and services include all qualified workers in trade services (shop keepers, retailers), in catering and accommodation services (cooks, barmen), in health care and social services (nurses, care workers), in cultural and personal services. FIGURE 10: Sectors of occupation,* percent of total sample 6% Manufacturing 11% Construction 12% Trade 12% Transport 4% Hospitality & catering Health care 11% Other services Family as employer** 28% 7% Other*** 9% * Sectors of occupation follow the international classification ATECO 2007–ISTAT. ** It includes all workers whose employer is a family (not a private or public firm). In this category are comprised housekeepers and all domestic workers in general. *** Residual, it includes: agriculture, electricity prod., science, rental, entertainment, communication and financial services. The distribution of our sample by sector of families are the direct employers (12 percent: employment is consistent with the latest available housekeeping, elderly and children domestic care data on the migrant workforce in Italy (Ministero etc.), and hospitality and catering (9 percent: del Lavoro, 2013). The service industry employs restaurants, hotels, ethnic food shops etc.). Manu- more than half of all interviewed migrants, with facturing and construction represent, respectively, a strong predominance in health care services 11 percent and 12 percent of total employment in (28 percent), followed by services for which the sample. Although not shown here in detail, EMPIRICAL FINDINGS 15 there are significant differences in the gender dis- for women), while the lowest individual monthly tribution among specific sectors of employment. income is declared by Moroccans (932 euro for Male migrants are concentrated in the construc- men, 815 euro for women). Gender differentials tion, manufacturing, and transportation industries, persist even within each of the three subsamples while women are overrepresented in the health and the total average income gap between men care and domestic sector. The composition of the and women is of 168 euro per month. hospitality and catering sector is more gender balanced. The overall and gender distribution On average, individual incomes are higher for of sampled migrants among sectors is also con- those who arrived before 2008 than for those sistent with their length of stay and the histori- who arrived after that year, but among older cal pattern of quotas and amnesties in the last migrants there is a more widespread distribu- 20 years, which also had an impact in determining tion of income. A longer migratory experience in the specialization of migrant workforce in the Ital- Italy is associated with a better integration into ian segmented labor market (Ministero del Lavoro, the labor market, more mobility and more diver- 2013; Salis, 2012; Reyneri, 2007). sification of income careers, while newly arrived migrants had less time and fewer opportunities INDIVIDUAL INCOME to improve their earnings and are still clustered around the income class €501–€1000 per month. Table 8 shows individual income averages for the Newly arrived Moroccans show the highest per- three subsamples disaggregated by gender. On centage of individuals earning less than €500 per average, Romanians declare the highest individual month. Moreover, the composition of sampled monthly income (1,295 euro for men, 898 euro migrants among types of occupation and sectors of employment helps explain the differentiated TABLE 8: Individual income (mean, se mean*) impact of the economic crisis across the three subsamples and by genders. Men Women In the majority of cases long-term residents (i.e. Morocco € 931.79 € 815.00 those that arrived before 2008) show a decrease 35.66 54.49 in their income since their arrival. Among those Peru € 1,020.00 € 851.78 who arrived in the past five years (after 2008), 35.60 37.21 trends in monthly income differ across the three Romania € 1,295.08 € 898.47 communities: Moroccan migrants, mostly men and 54.04 43.73 concentrated in construction and manufactur- Total € 1,037.65 € 869.98 ing, declare in 45 percent of cases a decrease in income since their arrival, while income is stable   25.20 26.56 for more than half of the Peruvians (52 percent), * The standard error of the mean (SE) is the standard deviation of the who are more represented in the health care and sample-mean’s estimate of a population mean. It shows how close to the population mean the sample mean is likely to be. SE is used to calculate the domestic sector. Within the Romanian subsample, confidence interval of an estimate. FIGURE 11: Individual income class, by country of origin (percent) Entered before 2008 Entered after 2008 Total Romania <500 501–1000 Peru 1001–1500 Morocco >1501 80 60 40 20 0 0 20 40 60 80 16 GREENBACK 2.0 REPORT 2014 FIGURE 12: Individual income: trend since the arrival in Italy (percent of total subsample) Arrived before 2008 Arrived after 2008 which benefits from the EU citizenship since money as a form of saved money, independently 2007, the prevalent economic trend seems more from the final use of it. positive with the majority of them declaring an increase in income (56 percent). BANKING PRODUCTS AND SERVICES About 80 percent of the sample declares to have Savings & financial inclusions at least one bank account. However, differences among the three communities are significant; SAVINGS the Peruvian subsample seems to be the most An entire section of the survey questionnaire “banked” with 69 percent interviewed migrants was devoted to explore features and practices having one bank account and the highest per- of migrants with regard to their saving behavior centage of migrants holding two or more bank since the beginning of 2013 and to their access to accounts (18 percent) (see Fig. 16). and use of financial and banking services. A strik- ing majority of the migrants declare that it was Figure 16 shows the distribution of migrants by not possible for them to save money during 2013 the different types of bank accounts they have due to hard economic and working conditions and their length of stay (distinguishing between (Fig. 13). Since the whole sample is composed of those who arrived before and after 2008). In definitions by remitting migrants, this reveals that the majority of cases bank accounts are held the majority of them do not consider remittance at an Italian bank (57 percent and 64 percent EMPIRICAL FINDINGS 17 FIGURE 13: Could you save since the beginning of 2013? If yes, how much? 18.75 81.25 If yes, how much? Average per month, in euro. 35 30 25 20 15 10 5 0 0–100 101–250 251–500 >500 FIGURE 14: Type of savings (percent of those who save) How do you save? 16.5 CC Italy 1.0 1.0 CC Morocco 2.1 CC Romania CC Romania + Italy 12.4 Savings bonds, Italy Savings bonds, Romania 62.9 Treasury bonds, Italy 2.1 1.0 1.0 Savings fund Other (at home) 18 GREENBACK 2.0 REPORT 2014 FIGURE 15: Bank account ownership one bank account is even more widespread, although 9.8 percent of all migrants live in a 100% 11 15 household which is totally non-banked. 18 90% 80% Among products and services offered by bank institutions, migrants frequently use the possibility 70% of crediting their salary or wage on the account 60% 111 107 112 (52.7 percent), followed by bank transfers (35 per- 50% cent), automatic bill payments (28.8 percent), and 40% online services (18.5 percent)4. About 70 percent 30% of the sample has a debit card, 17.3 percent has a Two or more pre-paid card and 10 percent a credit card. Over 20% One one-fifth of the sample (22 percent), independent 10% 37 31 38 of having a bank account or not, does not possess None 0% any kind of card; 28.8 percent use some type of Morocco Peru Romania financial product and services. respectively), but there are also a small number of ACCESS TO CREDIT migrants with two or more accounts (8.7 percent Figure 20 presents the distribution of loans by and 10.5 percent), one in Italy and one attached classes (in Euro) for the 27 percent of the sample to the country of origin (foreign bank with the which currently has a loan. More than a half of agency in Turin or in the country of origin). loans (56.5 percent) are below 10,000 euro, Among those with a longer period of stay in Italy but interestingly 9.3 percent of them are above (more than five years) there is a higher propor- 100,000 euro, possibly representing mortgage tion (24.6 percent against 14 percent) of migrants loans for purchasing an apartment. In more than without any kind of bank account. 4 The questionnaire used a broad definition of on-line services, including Considering all adults living with the interviewed any type of activity made via internet banking, from checking the move- migrants as a household unit, access to at least ments’ list to ordering payments and transfers. FIGURE 16: Types of bank account (percent of total sample) 64.0 60 57.1 50 40 30 24.6 20 14.0 7.9 10.5 10 7.9 8.7 3.5 1.6 0 None Bank, in Italy BancoPosta Bank, abroad Two or more, Italy & abroad None Bank, in Italy BancoPosta Bank, abroad Two or more, Italy & abroad Before 2008 After 2008 EMPIRICAL FINDINGS 19 FIGURE 17: Number of bank accounts in the HH (percent) 50 45 44.2 44.0 43.2 42.6 42.1 40 34.6 34.2 35 29.3 30 25 20 18.4 15 11.9 9.8 10 7.5 8.2 8.5 7.1 5 4.5 2.6 0.7 1.3 1.5 0 0 1 2 3 4+ 0 1 2 3 4+ 0 1 2 3 4+ 0 1 2 3 4+ Morocco Peru Romania Total FIGURE 18: Services and products used (percent FIGURE 20: Loans’ distribution, classes of total sample, more than one option allowed) of € (percentage of migrants with a loan) 60 52.7 9.3 12.4 50 19.4 < 1000 40 35.0 1001–5000 28.8 28.8 22.5 30 5001–10000 10001–20000 20 14.7 10.8 20001–10000 10 7.7 7.7 21.7 10000 + 5.8 0.8 2.1 1.9 0 r t e B o it su n d s s C n-li rod it ct ser s se s es e en fe ed Sa um ce d t ss vice ag on ne uc e ic ns m n cr cr tg N rv ra tra ay or ry vi er 70 percent of the cases loans have been taken out ’p p M la nk In s lls Sa ng Ba le s Bi on with banks, while only in a few cases, the money O ta C on has been borrowed from co-nationals (15 percent for Moroccans, 7 percent of Romanians, and none FIGURE 19: Type of cards owned (percent of among Peruvians). total sample, more than one option allowed) Although data on actual loans demonstrate a 80 clear preference for formal financial institutions 70.63 70 among migrants who already had access to credit, 60 however, when asked to whom they would ask for money in case of need, the total sample is more 50 diversified (Figs. 21–22). The number of migrants 40 who would ideally opt for a bank loan is sensi- 30 bly lower, especially for Moroccans, whereas a 22.08 higher percentage of migrants in each community 20 17.29 10.0 would ideally contact a co-national or an Italian 10 friend instead. Moreover, many of the interviewed 0.42 2.92 0 affirmed they would avoid asking for money in d rd rd rd t e un ar on ca ca ca co tc N ng it d ac di eb ai re vi d- ep D ol C ar Pr ev C R 20 GREENBACK 2.0 REPORT 2014 FIGURE 21: To whom did you ask for money? (percent) 0 10 20 30 40 50 60 70 80 MA Bank PE RO Financial MA Other Inst. PE RO MA Banco Posta PE RO Consumer MA Credit PE RO MA Credit Micro PE RO MA Nationals Co- PE RO MA Italians PE RO FIGURE 22: In case you will need a loan, to whom would you ask? 0 10 20 30 40 50 MA Nationals Nobody To PE RO MA Co- PE RO MA Friends Italian PE RO MA Bank PE RO MA Banco Posta PE RO Financial MA Consumer Other Inst. PE RO MA Credit PE RO MA Credit Micro PE RO Employer MA PE RO MA Know Don’t PE RO EMPIRICAL FINDINGS 21 any case because they are uncomfortable with the of collaterals) and the impossibility of getting a idea of being indebted (31 percent of Moroccans, bank loan. 15 percent of Peruvians and Romanians) and they fear this possibility. Remittance behaviors: flows and MICROCREDIT recipients, operator and costs To complete the picture on migrants’ financial MAIN RECIPIENTS AND MOTIVATIONS inclusion and access to credit, it is interesting that Sending money to the origin country is a transna- interviewed migrants do not seem to consider tional activity, which typically characterizes the microfinance and microcredit services5 as valu- migration experience. The analysis of remittance able options for accessing financial resources with transfers helps understand the level and depth of requirements and criteria different from those of migrants’ labor market integration at destination banks, even though official data for Italy and for and of their connection with the origin house- the Province of Turin show an increase of micro- holds, which directly influence the frequency, credit service providers. Only slightly more than amount, and regularity of flows. three percent of the total sample has ever con- tacted a microcredit agency (Fig. 23), preferring Interviewed migrants have been asked to describe instead the more favorable conditions (no need all flows of remittances they send to recipients, each with its specific characteristics in terms of 5 amount, frequency, channels, and cost. For this Microcredit is defined as a financial instrument designed for individuals who are not eligible for traditional banks’ loans since they lack real collater- reason, the number of recorded flows is higher als or because of their negative historic records within the banking system. than the number of interviews. The European Commission defines as “micro” a credit that is below the threshold of €10,000 for individual and family reasons, €25,000 for entre- preneurial activities. See: http://ec.europa.eu/enterprise/policies/finance/ Recipients are defined as the individuals who borrowing/microcredit/ materially receive the money. As such they do not necessarily coincide with beneficiaries, as the FIGURE 23: Have you ever heard recipient may use the money to benefit a third about microcredit? If yes, did you ever person. This is the case, for example, of children address a microcredit agency? of minor age left in the country of origin, who are not directly receiving the money, but can benefit Yes from remittances sent to other relatives. 10% Yes 34% For each of the three communities the majority No No 90% of remittance flows is addressed towards parents: 66% 60 percent of flows sent by Moroccan migrants, almost 57 percent of those sent by Peruvians, TABLE 9: To whom do you send money? Remittance recipients, percent of total flows Morocco Peru Romania Total Freq. % Freq. % Freq. % Freq. % Parents 138 60.26 105 56.76 95 46.34 338 54.60 Siblings 62 27.07 29 15.68 39 19.02 130 21.00 Partner 10 4.37 14 7.57 12 5.85 36 5.82 Children 2 0.87 19 10.27 19 9.27 40 6.46 Other relatives 10 4.37 13 7.03 18 8.78 41 6.62 Partner’s relatives 3 1.31 5 2.7 17 8.29 25 4.04 Friends 4 1.75 0 0 5 2.44 9 1.45 Total 229 100 185 100 205 100 619 100 22 GREENBACK 2.0 REPORT 2014 and 46 percent of those sent by Romanians are least for nuclear family members (children and directed to one of the parents or both. Also sib- partner, often a female one). The relatively high lings receive a high share of total flows, especially share of flows directed to children left behind in in the case of Moroccans (27 percent from respec- Peru and Romania can be explained by a shorter tively Moroccans, 16 percent from Peruvians, and migration experience in Italy of the two com- 19 percent from Romanians). munities and also by the prevalence in the two subsamples of female migrants, who are often More divergence among the three communities employed in the domestic service and care sector is found on flows addressed to children: while and for which it might be more difficult to fulfil 10 percent of total flows sent by Peruvians and bureaucratic requisites for a reunification with 9 percent of flows sent by Romanians is directed children. As for Romania moreover, the free move- to their children, this type of flows is almost non- ment within the European Union borders makes existent among Moroccans (0.9 percent). Roma- contacts with the family left in the origin country nians show the highest share of flows addressed relatively easier, diminishing the frequency of to their partner’s relatives (8 percent): this reunifications at destination. could be explained by the higher proportion of women among Romanians, which could have The analysis of recipients is strictly connected a higher propensity to send money to their with the motivations behind the decision to remit. parents-in-law than men. Remittances directed Different motivations can be clustered around to friends and acquaintances mainly represent some distinct categories according to the actual flows for the repayment of a debt incurred with use of money and to the nature of relationships someone outside the family. Although parents that remittances contribute to maintain (Rapoport receive flows more often in all the three sub- and Docquier, 2005). Altruism, compensation for samples, total annual flows directed towards past services, insurance, intra-family loan repay- members of the nuclear family (partner, chil- ment, strategic investment, and savings are all dren) are higher on average in comparison with possible motivations and can operate simultane- flows directed towards other relatives (parents, ously. Quantifying the relative importance of dif- siblings, others). ferent reasons and purposes has proved to be a challenging empirical exercise (Yang, 2011) and Data on remittance recipients and on the size of many social scientists only focused on the study flows by recipient seem coherent with the aver- of remittance impact on recipients both at the age family size in the three origin countries (with household and at a country level (de Haas, 2007; extended families more frequent in Morocco Castagnone et al, 2013). and Peru than in Romania) and with the average length of stay of the three selected subsamples. Our study can only analyze senders’ motivations Moroccans have a longer migration history in Italy behind remittances without exploring the impor- and specifically in Turin, to which is associated a tant issue of the control of money by the receivers more complete process of family reunification at (Ashraf et al, 2011), since we do not have parallel TABLE 10: Average amount sent per year, by recipient Mean Std. Dev. N. flows Min Max Partner € 2,711.11 2057.67 36 100 7200 Children € 2,356.88 1916.45 40 100 8400 Partner’s relatives € 1,452.83 1513.86 23 100 5640 Parents € 1,445.93 1593.59 333 50 14400 Siblings € 960.16 1525.31 127 50 12000 Other relative € 806.25 802.58 40 50 3600 Friends € 611.11 716.67 9 100 2400 Total € 1,425.12 1648.98 608 50 14400 EMPIRICAL FINDINGS 23 FIGURE 24: What is the money sent used for (percent more than one answer allowed) 90 80.4 80 70 60 49.4 50 40 30 24.0 21.5 20.2 20 10.4 10 6.9 3.5 2.7 2.3 0 p. d lth n s g ts gs rt t en ie o io n o m en ea n Fo si sp nc at m su vi ou tm H uc ay an Sa ge on H s Ed ep Tr er ve C Em tr In al eb er D en G information from recipients in origin countries. In offering a very widespread distribution of agen- the vast majority of cases, the money serves for cies and operators also in rural areas of origin increasing the level of consumption of receivers— countries and highly competitive services in terms food expenditures, health care services especially of transfer speed. for the elderly, unspecified general expenditures also connected with specific lifetime events— The prevalence of MTOs on all other RSPs is con- and for supporting the education of children. In firmed by our empirical data (Fig. 25), where bank 20 percent of cases flows were directed to sup- transfers, card transactions, and hand-carried port relatives in case of idiosyncratic (individual) cash transfers represent together less than one emergencies and covariate (aggregated) shocks fifth of total recorded flows. Transaction via in the origin area. Housing expenditures (improve- twin cards, prepaid cards, and account to cash6 ment and maintenance of a property) and invest- represent a very small proportion of total flows ments in new assets seem to be important for (2 percent). 10.4 percent and 6.9 percent of flows respectively. Peruvian migrants use almost exclusively (99 per- cent of flows) a money transfer operator, while a CHANNELS AND REMITTANCE more diversified choice is made by Moroccans and SERVICE PROVIDERS Romanians, who also use hand-carried cash trans- Italian regulation of remittance market is based fers (10.5 percent and 18.5 percent respectively) upon the EU Directive 2007/64/CE, which is and bank transfers (4 percent and 12.7 percent). aimed at protecting clients and regulating opera- tors through more competition and transparency Among private non-financial operators for cash (Fondazione Moressa, 2013: 109). With the Law transfers by hand, migrants are more likely Decree 27 January 2010, payment and financial to choose relatives or friends than transport institutions are allowed to operate as sending 6 Twin cards are two cards connected, one used in Italy to send money channels through new instruments (bank trans- and the other at the destination to receive it. Prepaid cards are cards that fers, debt and credit cards, etc.). Among newly can be used at destination and recharged from a bank or post office in Italy. regulated remittances service providers, money The account to cash allows sending money from a bank account in Italy to a partner agency/bank at destination. The recipient will be able to get his transfer operators (MTOs) are those with the high- money in cash. Some MTOs offer a service called “mobile account to cash”, est share of the Italian market (Bank of Italy, 2013) through which customers with a bank account can send money through their mobile phone to an MTO agency where recipients receive money as well as in the majority of Western countries, in cash. 24 GREENBACK 2.0 REPORT 2014 FIGURE 25: Types of channel used to send remittances (percent of total flows) 0.7 0.2 1.1 5.6 9.8 Twin cards Count to cash Prepaid card Bank transfer 82.7 Hand-carried cash to cash Money Transfer TABLE 11: Number and share of flows per channel, by country Morocco Peru Romania Total Freq. % Freq. % Freq. % Freq. % Money transfer 190 82.97 184 99.46 135 65.85 509 82.23 Hand-carried cash to cash 24 10.48 1 0.54 38 18.54 63 10.18 Bank transfer 9 3.93 0 0.00 26 12.68 35 5.65 Prepaid card 2 0.87 0 0.00 5 2.44 7 1.13 Count to cash 4 1.75 0 0.00 0 0.00 4 0.65 Twin cards 0 0.00 0 0.00 1 0.49 1 0.16 Total 229 100 185 100 205 100 619 100 agencies. Among banks, Unicredit is the most fre- beginning of their migratory experience in Italy. quently used (43.5 percent of flows via bank oper- Among those who changed, 42 percent used to ators), but many other Italian and foreign banks, use a MTO, 20 percent sent money via bank trans- often linked to the origin countries of migrants, fer and 7 percent used the post office (almost are used in a large number of cases (Banque the only available option until about 20 years Chaabi du Maroc, Cec Bank, Attijariwafabank). ago). About 14 percent of them state to have always used a MTO, but to have changed opera- As for MTOs, Western Union is the most frequent tor according to costs and services offered. In the choice (53 percent of flows via MTO), followed majority of cases, migrants moved from West- by Ria (16 percent), and MoneyGram (9 per- ern Union (historically, the first MTO available) cent). Nevertheless, there is a high variability and to another operator more specialized in sending migrants often use operators that have a clear money to their country of origin. regional or local connotation. Chavin, a money transfer operator for Peru, represents alone 7 per- As reported in Figure 28, the vast majority of cent of flows via MTO, followed by Smith&Smith transactions, regardless of the specific channel (Romania), Flouss Express (Morocco), and others. used, take place at a dedicated agency. Among those who use a non-financial intermediary Interestingly, one-third of the interviewees have for a cash to cash transfer, the money is more changed type of channels or operators since the EMPIRICAL FINDINGS 25 FIGURE 26: Operator used, by type (percent of total flows) 45 43.0 40 35 30 25 20 15 12.8 10 7.6 8.0 5.8 4.0 5.5 5.8 5 3.0 0.8 1.3 1.5 0.7 0.3 0 Friend Intermediary Trasport agencies Italian bank, local Italian bank, national* Foreign bank WesternUnion WU+Others** Ria MoneyGram MoneyTransfer Peru's MTO Morocco's MTO Romania's MTO Private operator Bank operator Money transfer operator * Include Poste Italiane. ** In almost five percent of cases, migrants declared to use—for the same flow (same amount, same frequency, same recipient)—Western Union or another MTO specialized for a specific destination country (Flouss, Valuetransh, Smith&Smith). The choice depends from time to time on where the MTO agencies are located in the city and on the presence of temporary offers and promotions. FIGURE 27: Bank operators (percent of total bank operators used) 45 40 39.2 35 30 25 20 15 9.8 9.8 9.8 10 7.8 5.9 5 2.0 3.9 2.0 3.9 2.0 2.0 2.0 0 PostelItliane Bancs d’ Alba Banca Pop. di Novara Unicredit Banca Intesa Barkley’s ING Banque Chaabi (MA) Cec Bank (RO) Attijariwafabank (MA) Raiffeisen bank (RO) BMCE Bank (MA) BNL Italian European Origin Countries frequently put in the hands of a relative or a friend The choice of the channel and service provider rather than using a private transport firm or a depends on many different variables. Both the generic private intermediary.7 sender’s context (existence and accessibility of agencies, opening hours, convenience and ease 7 Some travel agencies offer to transport not only passengers and their of transactions, etc.) and the circumstances of luggage but also packages alone within which migrants usually send mate- rial objects and money. “Private intermediaries” is used here to generically recipients and their area of residence (availability refer to individuals who physically put the money directly into the hands of of agencies, opening hours, easiness to access for the receivers at the destination. 26 GREENBACK 2.0 REPORT 2014 FIGURE 28: Where the transaction took place? (percent of total flows) 80 77.7 70 60 50 40 30 20 10 4.4 6.5 3.6 3.1 3.7 1.1 0 ) ce p t r ry nd ne TO te o ia ffi sh ie en r ed ,M te o /fr lc co In st m ive nk al Po ac r te at C a b (b in l re To an cy a en a a Vi Vi Ag recipients, additional costs, etc.) may influence while bank transfers are generally slower and the the final choice in terms of RSP, amount and fre- hand-carried cash transfer often implies no pre- quency of the flows. determined delivery time. Faster transactions—through which the money The word-of-mouth among friends and acquain- can be available at the destination within the tances appears to influence the final decision hour—are those made via MTO or pre-paid card, more than any other factor; local or national FIGURE 29: How long does it take for the money to be available at destination? (percent by channel) 0 10 20 30 40 50 60 70 <1h Bank transfer same day 1day 2days 3–5days 6days+ <1h same day transfer Money 1day 2days 6days+ Prepaid <1h card same day 1day <1h cash to cash Hand-carried same day 1day 2days 3–5days not preset EMPIRICAL FINDINGS 27 FIGURE 30: How did you choose your main channel? (more than one answer allowed) 350 300 250 200 150 100 50 0 th ta ta gn gn y t in g g n e ss tin tin io ig ou i ,i rn i ei s, re st ba or ia ee ee te or m et ge fo ed In Em at m m ,f re of g s, m ia su st ily p us et d v— ed u v— or m io re ro ’s m fa W ig Ad st G cy Ad v— el v— s/ y en M R As Ad Ag Ad advertisements, and translation in Italian and In the vast majority of cases, transactions to the other languages, seem to be less taken into con- same recipient are sent at very high frequencies: sideration. As for bank transfers, it is worth high- transactions occur every month for more than lighting that the possibility of using this service half of Peruvian (53 percent), for 27 percent of as a way of remitting money has been often sug- Moroccan and for 20 percent of Romanian flows. gested directly by the bank, where the migrant Combining the magnitude and frequency of each had his/her account (Fig. 30). flow, we estimate the average annual amount sent to the same recipient (Table 14). The Peruvian VOLUME PER TRANSACTION subsample, with the highest share of monthly AND PER YEAR transactions, shows the largest total amount per year (€1848), followed by Romanians (€1380) and Remittances are usually sent at relatively high Moroccans (€1117). Average annual values are in frequencies and in smaller amounts as opposed line with the most recent data provided by the to other types of international financial flows to Bank of Italy at national level (Bank of Italy, 2013). developing countries (Yang, 2011). Data from our sample show that the average amount of money More detailed data on flows magnitude by type of sent per transaction is of €236. Romanians show RSP and by recipient are in the Annex II-B. a relatively higher average amount (€254) than Peruvians (€230) and Moroccans (€213). As pre- TRANSACTION COSTS: sented in Table 12, transactions through a bank A DISTORTED PERCEPTION transfer are on average larger (€356) than those made via MTOs (€224) or via hand-carried cash to Transaction fees typically include a non-negligible cash transfers by hand (€225). fixed cost per transaction, for which the classical TABLE 12: Average amount sent per flows, by type of channel used Mean Std. Dev. N. flows Min Max Money Transfer € 224.19 204.51 504 50 2500 Hand-carried cash to cash € 225.00 136.93 60 50 500 Bank transfer € 355.88 522.63 34 100 3000 Others* € 218.75 . 12 100 600 Total € 236.03 254.75 46 50 3000 * It includes Account to cash, Prepaid card and Twin cards. 28 GREENBACK 2.0 REPORT 2014 TABLE 13: Frequencies of transaction per flow, percent of total flows Morocco Peru Romania Total Freq. % Freq. % Freq. % Freq. % <1/y 11 4.85 3 1.62 10 4.88 24 3.89 1/y 24 10.57 19 10.27 32 15.61 75 12.16 2–3/y 74 32.60 28 15.14 81 39.51 183 29.66 3+/y 57 25.11 37 20 42 20.49 136 22.04 1/m 61 26.87 98 52.97 40 19.51 199 32.25 Total 227 100 185 100 205 100 617 100 TABLE 14: Average amount sent per transaction and per year   Per transaction Per year     Mean Std. Dev. Mean Std. Dev. N. flows Morocco 212.88 223.03 1117.68 1288.26 226 Peru 230.05 166.57 1848.32 1600.06 184 Romania 253.875 278.98 1380.13 1950.87 200 economic theory would predict relatively rare Data from our sample on perceived costs per and larger amounts to minimize fees paid per transaction show the expected features: the aver- transaction and would only admit the need of age cost declared by migrants, expressed as a higher frequencies in case of unpredicted shocks share of the amount sent, is higher on average or emergencies affecting the recipient household for transactions via MTO than via bank transfer or and/or the origin country as whole (Yang, 2011). hand-carry transfer (Table 15). Nevertheless, very often a pattern of small individ- ual transactions at high frequencies characterizes Even more importantly, migrants seem to perceive remittance flows. Data from our sample confirm a lower cost per transaction than the actual ones. this economically counter-intuitive characteristic. Table 16 compares the actual cost of sending €140 from Italy to Morocco and Romania as registered One possible explanation of the high-frequency quarterly by the Remittance Prices Worldwide associated with low-value transactions is that they (RPW) database with the average cost perceived reduce the probability of losses in case of adverse by the interviewed migrants.8 events either for the remitting migrant or for the recipient. Sending small amounts of money at Not only is the cost often underestimated, but high frequency could also possibly be due to a many of the interviewed migrants are not fully self-control problem, again either for the sender, aware of cost composition and do not take into who avoids the temptation to spend the money, consideration the role played by commissions or the recipient, who is forced to smooth expen- applied by operators in the sending and receiving ditures over time (Yang, 2011: 144). Some studies agency of the RSP and the role of margins on the also tried to estimate the elasticity of remittances official exchange rate, as Table 17 shows. to changes in transaction costs (see for example When asked about the advantages and disadvan- Gibson et al, 2006). Surely, a careful analysis of tages of the various channels, many interviewees remittance costs and their components depicts felt that bank transfers are cheap (74 percent a more comprehensive picture of the available remittance market and can provide useful insights 8 Remittance corridors from Italy are also scrutinized monthly by the for both users and service providers (World online database Mandasoldiacasa, managed by CeSPI and certified by the Bank, 2013c). World Bank. See http://www.mandasoldiacasa.it/ EMPIRICAL FINDINGS 29 TABLE 15: Average cost per transaction as percent of the amount sent, by type of channel used Amount Sent, Classes   Bank Transfer Money Transfer Prepaid Card Cash to Cash Total 0–100 Avg cost, % 1.5 5.4 6.0 2.7 5.0 St. dev. 1.69 4.42 . 4.45 4.45 N. flows 8 162 1 18 190 100–200 Avg cost, % 1.8 4.6 2.3 1.8 4.2 St. dev. 3.99 2.93 3.19 2.50 3.12 N. flows 13 174 5 14 206 200–500 Avg cost, % 2.0 3.9 1.6 3.5 St. dev. 1.54 2.54 2.05 2.57 N. flows 7 128 22 158 500+ Avg cost, % 2.4 5.4 2.0 3.0 4.7 St. dev. 2.33 7.29 . 5.20 6.60 N. flows 5 28 1 3 37 Total Avg cost, % 1.8 4.7 2.8 2.1 4.3 St. dev. 2.79 3.80 2.96 3.20 3.81 N. flows 33 492 7 57 589 * Account to cash and twin cards excluded, too few data. TABLE 16: Comparison between actual average TABLE 17: Cost composition (percent), by type costs and perceived average cost of channel used Actual Average Cost of Hand-c. Sending €140 Expressed in %, Bank Money Cash to RPW 3Q2013*   Transfer Transfer Cash Bank MTO Total Commissions 79.4 56.4 49.2 Average Average Average Exchange rate 0.0 0.4 0.0 From Italy to Morocco 6.7 7.9 7.4 Others for receivers 0.0 0.8 0.0 From Italy to Romania 3.4 7.1 5.7 Comm. & exchange rate 11.8 2.2 0.0 Perceived Average Cost of Sending €100–200 Expressed All of the above 0.0 36.1 1.6 in %, Survey Results No cost 8.8 1.0 45.9 Bank MTO Total Don’t know/Don’t answer 0.0 3.2 3.3 average average Average 100 100 100 From Turin to Morocco, Peru 1.8 4.6 4.2 and Romania *Data on the Italy-Peru corridor are not available on RPW. Source: Remittance Prices Worldwide database. the channel used, the vast majority of interviewed of cases) and fast (17 percent), while MTOs are migrants state that the selected channel has no perceived as fast in the majority of cases (35 per- cons at all (74 percent in case of bank transfer, cent) but also cheap (23 percent), safe (15 per- 73 percent in case of MT and 57 percent for cash cent) and easy to use (10 percent). A person to to cash). The most frequently cited problem of person cash transaction is described not only bank transfers is that they are slow (11 percent), as cheap (47 percent of cases), but also as safe while MTOs are expensive in 19 percent of cases (18 percent) and easy for the recipients (11 per- and cash to cash transactions are slow (13 per- cent). As for weaknesses and disadvantages of cent) and unsafe (7 percent). 30 GREENBACK 2.0 REPORT 2014 FIGURE 31: Pros and cons of main channels, percent 75 60 Difficult for recipients Far from my home 45 No weaknesses Difficult to use 30 Expensive Unsafe Hours 15 Slow 0 Cheap Fast Easy to use Safe Close to my home Easy for recipients Hours No docs Other goods -15 -30 -45 -60 -75 Pros Cons Bank transfer Money Transfer Hand-carried cash to cash As reported in Fig. 32, in very few cases (8 per- flows and of the migrant, data show a decreas- cent) migrants have had some problem with the ing trend for more renowned MTOs paralleled chosen operator, often because of an error in reg- by an increase in competition due to new opera- istering the credential of the recipient or because tors specialized in some regions of the world of a delay in the delivery. and that in some cases have specific agreements with Central Banks and/or commercial banks of Nevertheless, although costs and conditions of receiving countries. flows and transactions described are not always the best available in the market, interviewed TRENDS IN REMITTANCES migrants show a generally good level of satisfac- tion with the service providers they choose. Migrants were asked if the amount of remittances sent to their origin countries has changed— Although a MTO is used in the vast majority of increased, decreased or been stable—during their cases, regardless of the characteristics of the presence in Italy, since their arrival until 2013. FIGURE 32: Did you ever have a problem? If yes, which kind of problem? 8% YES 18% 8% 2% Money lost Delay Other costs for the recipients NO 72% Wrong credentials 92% EMPIRICAL FINDINGS 31 FIGURE 33: In case you had problems, with the prevalent trends among the foreign resi- did you look for help? dent population in the city of Turin, family reuni- fication processes have diminished the number 2% of remittance recipients in the origin countries 6% and at the same time have increased needs and expenditures in Italy (Fig. 37). No 22% Agency director In 65 percent of cases nevertheless, the reduc- 50% MT director tion in the remittance amount has to be ascribed Agency at destination 20% to the worsening economic conditions during Friend the crisis, with an increase in instability and a decrease in earnings, which prevent migrants from maintaining their precedent levels of con- tributions to their relatives abroad. Nevertheless, Remittances increased for a very small propor- while comparing income and remittance trends, tion of migrants within each community (Fig. 36). it seems that remittances decreased slightly less Among Peruvians the total amount has remained often as they were more resilient than income to stable across time in the majority of cases the worsening economic conditions in Italy during (49.7 percent), while for 56.1 percent of Moroc- the last five years (see Fig. 12 for a comparison). cans and 55.3 percent of Romanians the money sent has diminished. RECEIVED REMITTANCES The reasons behind a reduction in the total amount sent are mainly of two types: either a The literature on the migration and development decrease in the demand for remittances or a defi- nexus has recently enlightened the existence cit in migrant’s supply. In many cases, consistently of reverse remittances as flows from the home FIGURE 34: Since your arrival in FIGURE 35: What kind of channel did Italy, is your main channel changed? you use at that time? (percentage (percentage of total sample) of those that have changed) Bank transfer 7% 20% Yes Count to cash 33% 14% 1% MT 7% Hand-carried cash No to cash intermediary 67% 9% 42% Hand-carried cash to cash friend/relative Other MT* Postal order FIGURE 36: Trend in total amount sent since the arrival in Italy, by country of origin 32 GREENBACK 2.0 REPORT 2014 FIGURE 37: Reasons for a decrease in remittance amount (percent) 0 20 40 60 80 Reduction in income Reduction in demand from recipients Reduction in n. of recipients Increase in expenditures in Italy Increase in family members in Italy More insecure income Less savings Other relatives contribute Less convenient change with FIGURE 38: Received remittances, by country and person of origin Peru From where? From whom? 3% Yes 20% 25% Morocco Partner 3% 38% 36% No 2% Italy Children 75% 75% Egypt Parents Other Other relatives 20% Europe* Friends 3% Peru From where? From whom? Yes 2% 25% 25% No Peru Partner 98% 50% USA Parents 75% 25% Other relatives Romania From where? From whom? Yes 9% 7% 7% Morocco* Partner No 14% Romania 43% 43% Children 91% Italy Other relatives 72% France 14% * An interviewed Romanian migrant is married with a Moroccan migrant and receives money from her partner’s relative in Morocco. EMPIRICAL FINDINGS 33 communities to migrants, which may take differ- out of four Moroccan migrants receives money ent forms (services, goods, but also money) and from their relatives, while percentages are lower contribute to maintain reciprocal relationships for Romanians (9 percent) and Peruvians (2 per- between migrants and their household of origins cent). The money usually comes from origin (Mazzuccato, 2010). countries, but also from some European countries, where other family members live. These reverse Although our study only focuses on monetary monetary flows can be seen as counter-cyclical remittance flows and does not investigate on in- flows and help understanding how deeply the kind and social transfers, survey data on reverse economic crisis impacts the migrants’ economic remittances reveal that a non-negligible share integration and their wellbeing in Italy (Fullin & of migrants receive money from co-nationals Reyneri, 2013). (either a relative or a friend) outside Italy.9 One 9 It is worth noticing that, among those not included in the sample because of their non-compliance with the conditions regarding employ- ment and remittances since the beginning of 2013, between one-sixth and one-fifth of the migrants have declared that their economic conditions have deteriorated since they started to receive remittances from their relatives. For more information of migrants excluded from the sample see the Meth- odological Note. 35 5 CONCLUSIONS The survey conducted within Project “Greenback year. These characteristics suggest there is room 2.0” aimed at describing the economic and finan- for improvement in terms of competitiveness and cial profile of migrants residing in the city of Turin, transparency of all remittance service providers in particular, investigating the interconnected and of greater awareness and management ability dimensions of financial behaviors and of finan- of remitters in order to minimize transaction costs cial needs in terms of accessibility to different and make more efficient decisions. types of financial products and services. The sur- vey focused mainly on the monetary remittance Secondly, our survey dealt with remittances behaviors and services as remittances are perhaps intended as international transfers of money from the most widely understood migrant’s transna- migrants to their origin countries. Nevertheless tional activity and play a crucial role in many the economic contribution of migrants to home migrant sending countries, both in terms of GDP countries takes many other forms, from punctual contribution at the macro-level and as an addi- contributions in correspondence with specific tional income source for origin households at the events (birthdays and funerals, religious and micro-level. The final sample is composed of three national holydays to name a few) to long-term equally large subsamples of Moroccan, Peruvian, commitment for productive and social invest- and Romanian migrants, which represent the first ments in the origin area. Decisions on the type three foreign communities in Turin in terms of and level of economic transnational engagement residents and of remittances sent. vary with the type and length of the experience abroad as well as with the level of integration of The analysis of empirical data presented in the migrants at different stages of their migratory previous sections of this Report puts into light process. A further step of analysis would be to four main strands of results, which unfold further connect remittance levels with the micro charac- opportunities of in-depth investigation. teristics of migrants’ in terms of gender, education level and employment condition, with the meso- First, in spite of the differences in frequency level, such as the household composition at origin and amount per transactions among the three and at destination, but also with country-level selected subsamples, remittances are usually variables, which can either foster or prevent other sent at relatively high frequencies and in small types of transnational engagement. This could amounts and registered average annual amounts also help us explaining the different prevalent are in line with the average level for migrants in behavior among surveyed short-range (Moroccan, Italy. The analysis reveals a widespread misper- Romanian) and long-range (Peruvian) migrant. ception of transaction costs and overall service conditions by migrants: remitters have low aware- A third important issue to be further explored is ness of costs’ composition and often do not con- the role of migratory models in influencing remit- sider the presence of exchange rate’s margins and tance practices and needs. Interviewed migrants fixed costs at destination, irrespectively of the differ by country of origin, gender, length of stay preferred remittance service provider. Moreover, in Italy, type of occupation and related economic highly frequent flows are associated with low- stability. Moroccans and Peruvians typically have value transactions, which imply an overall higher larger families than Romanians, considering weight of fixed costs on the total amount sent per both relatives in the origin country and in Italy. 36 GREENBACK 2.0 REPORT 2014 Moroccan migrants show on average a longer Finally, our empirical Study suggests a non-linear presence in Italy and hence a higher rate of family relation between integration and transnational reunification than both Peruvians and Romanians, engagement. The length of stay abroad is usu- who instead often report to have children left in ally taken as an (imperfect) measure of migrant’s the country of origin. Collected data suggests on integration at destination: the level of economic, average a good level of economic and financial social and political integration is expected to grow integration of interviewed migrants, although with the length of migration, which can be associ- differences arise across subsamples and by gen- ated with a more stable job, an improvement in der. Hence, results suggest that different remit- the type and length of a residence permit, family ting practices, i.e. frequency and amount, type of formation, social engagement and political partic- receivers and purpose of the money pertain to ipation etc. Our data indicate that migrants’ trans- distinct migratory models in terms of familiar net- national economic engagement does not always works and reunification strategies for spouse and diminish with time, as we found migrants who still children, distance from the home country, length remit after twenty or thirty years of migration. of migration, and type of labor market integration Hence, a better integration at destination may not abroad. Migrants who consider their migratory prevent migrants from engaging in transnational experience as permanent and who progressively activities or keeping a connection with the origin proceed with family reunification processes at country. Moreover, these trends are also found destination behave differently from migrants who during periods of severe economic crisis which, consider the possibility of return or of circularity on average, has deteriorated migrants’ economic between the origin and the destination country. conditions, putting at stake the overall level of Moreover, our selected subsamples provide evi- economic integration of migrants and their capac- dence of different gender models, which deserve ity to keep remittance flows constant over time. further inquiry. Patterns of transnational economic activities are consistent with data on the partici- Since remittances seem to be resistant to income pation of men and women in the labor market at drops, this could suggest that under certain con- destination. Although women remit less than men ditions integration and transnationalism are not on average, does the remittance-income ratio dif- mutually exclusive. A deeper analysis of their fer between men and women? Do women remit occupational and economic trajectories since their for different purposes and to different types of arrival in Italy could provide a better understand- recipients compared to their male co-nationals? ing of the reciprocal connections between their These questions will be driving further analyses integration processes at destination and their on the empirical evidence collected through the transnational economic activities. Greenback 2.0 survey. 37 REFERENCES Ashraf, N. Aycinena, D., Martinez, A. C., & Yang, D. Cingolani, P. & Ricucci, R (2013), Transmediterranei— (2011), “Remittances and the Problem of Con- Le collettività di origine nordafricana in trol: A Field Experiment among Migrants from Piemonte tra continuità e cambiamento, Rap- El Salvador.” http://www-personal.umich porto di Ricerca, FIERI, Torino. .edu/~deanyang/papers/aamy_ Comune di Torino, (2013), Stranieri residenti a remittancecontrol.pdf Torino distribuiti per provenienza e circo- Bank of Italy (2013), Le rimesse verso l’estero degli scrizione, http://www.comune.torino.it/ immigrati in Italia http://www.bancaditalia.it/ informacasa/studi_statistiche/processi_ statistiche/rapp_estero/rimesse. demografici/stranieri_torino.shtml Blangiardo, G. C. (2004), “Campionamento de Haas, H. (2007), “Remittances, migration and per centri nelle indagini sulla presenza stra- social development,” in Social policy and devel- niera in Lombardia: una nota metodologica,” opment, n.34, October, www.imi.ox.ac.uk/pdfs/ in Aa.Vv., Studi in ricordo di Marco Martini, unrisd-remittances-mig-dev. Giuffrè, Milano. Fondazione Leone Moressa (2013), Rapporto Blangiardo, G. C., Baio, G. & Blangiardo, M. (2011), annuale sull’economia dell’immigrazione— “Center sampling technique in foreign migra- Edizione 2013, Tra percorsi migratori e compor- tion surveys: a methodological note,” Jour- tamento economico, Bologna, Il Mulino. nal of Official Statistics, Vol. 27, No. 3, 2011, Fullin, G. and Reyneri, E. (2013), “Introduzione. pp. 451–465. Gli immigrati in un mercato del lavoro in crisi: Bonifazi, C. & Marini, C. (2013), “The impact of il caso italiano in prospettiva comparata,” in the economic crisis on foreigners in the Ital- Mondi Migranti 1.2013 Immigrazione e crisi eco- ian labor market,” Journal of Ethnic and nomica, pp. 21–35, Franco Angeli. Migration Studies. Galossi Emanuele, (Settembre 2013), L’impatto Castagnone, E., Mezger, C., Schoumaker, B., Nazio della crisi sulle condizioni di vita e di lavoro T., Rokotonarivo N. (2013), “Understanding degli immigrati: un’indagine dell’Associazione Afro-European labor trajectories: integration of Bruno Trentin, Ass. Bruno Trentin e IRES. migrants into the European labor market, trans- Gibson, J., McKenzie, D. & Rohorua, H. (2006), national economic participation and economic “How cost elastic are remittances? Evidence reintegration into the country of origin. A com- from Tongan migrants in New Zealand,” Pacific parative Study,” MAFE WP 26, January: http:// Economic Bulletin, 21:1:112–28. www.ined.fr/fichier/t_telechargement/57781/ International Organization for Migration (2013), telechargement_fichier_en_wp26_ Questionnaire for the project “Developing understandingafro.eulabortrajectories.pdf a platform for development partnerships Centro Studi e Ricerche IDOS (2013), Dossier between diaspora and entrepreneurs in Egypt, Statistico Immigrazione—Rapporto UNAR Dalle Morocco, Tunisia and Lebanon through dona- discriminazioni ai diritti, Roma. tions, loans and services,” coordinated by IOM CeSPI (2012), Primo Report—Osservatorio Nazio- Regional Office in Egypt, Cairo. nale sull’Inclusione Finanziaria dei Migranti Mazzuccato, V. (2010), “Reverse Remittances in in Italia, CeSPI—Centro Studi di Politica the Migration–Development Nexus: Two-Way Internazionale. Flows between Ghana and the Netherlands,” CeSPI (2013), Secondo Rapporto— Population Space Place 17, 454–468. Osservatorio Nazionale sull’Inclusione Ministero del Lavoro e delle Politiche Sociali 2013, Finanziaria dei Migranti in Italia, CeSPI— Terzo Rapporto Annuale—Gli Immigrati nel Centro Studi di Politica Internazionale. mercato del lavoro in Italia, Roma. http://www 38 GREENBACK 2.0 REPORT 2014 .lavoro.gov.it/PrimoPiano/Documents/Terzo Report, MAFE Methodological Note 2: http:// percent20Rapporto percent20Annuale. www.ined.fr/fichier/t_telechargement/58222/ percent20Gli percent20immigrati percent20nel telechargement_fichier_fr_note.2_ percent20mercato percent20del schoumaker_diagne_2011.pdf percent20lavoro percent20in percent20Italia Strozza, S. (2004). “Estimates of the Illegal For- percent202013.pdf eigners in Italy: A Review of the Literature,” Nicolì, M. & Corazza, C. (2010), The Market for IMR-Center for Migration Studies of New York , Remittance Services in the Czech Republic: 38:1:309–331. Outcomes of a Survey among Migrants. World The World Bank and Bank for International Settle- Bank: https://openknowledge.worldbank.org/ ment 2007, General principles for international handle/10986/2521 remittance services, January http:// Pastore F., Salis E., Villosio C. (2013) L’Italia e siteresources.worldbank.org/ l’immigrazione low cost: fine di un ciclo?, in INTPAYMENTREMMITTANCE/Resources/ Mondi Migranti 1.2013 Immigrazione e crisi New_Remittance_Report.pdf economica, pp. 151–172, Franco Angeli. The World Bank 2011, Migration and Remittances Ratha, D., (2007), Leveraging remittances Factbook—Second Edition, Washington D.C. for development, Policy Brief, Migration The World Bank 2012, Guidance Report for the Policy Institute, http://siteresources Implementation of the CPSS-World Bank— .worldbank.org/INTPROSPECTS/ General Principles for International Remittances Resources/334934-1100792545130/ Services, October, http://web.worldbank.org/ LeveragingRemittancesforDev.pdf . WBSITE/EXTERNAL/TOPICS/EXTFINANCIAL Reyneri, E. (2007), “Immigration in Italy: trends SECTOR/0,,contentMDK:23289589~pagePK: and perspectives,” ARGO—IOM, http://www 210058~piPK:210062~theSitePK:282885,00.html .portalecnel.it/portale/indlavdocumenti.nsf/0/ The World Bank (2012), Migration and Develop- 466486C57FF3FF42C125737F0050A9EC/ ment Brief 19, Migration and Remittances $FILE/Reyneri-percent20Immigration Team, Development Prospects Group, percent20in percent20Italy.pdf Washington DC. Ricucci, R. (2011), Le famiglie straniere di fronte The World Bank, (2013a), Migration and Devel- alla crisi. Istantanee piemontesi, Rapporto di opment Brief 20, Migration and Remittances Ricerca, FIERI, Torino: http://fieri.it/2011/02/11/ Team, Development Prospects Group, April 19, le-famiglie-straniere-di-fronte-alla-crisi- Washington DC. istantanee-piemontesi/ The World Bank (2013b), Migration and Develop- Rapoport, H. And Docquier, F. (2005), “The Eco- ment Brief 21, Migration and Remittances Team, nomics of Migrants’ Remittances,” IZA Discus- Development Prospects Group, October 2, sion Paper Series, DP No. 1531, March 2005. Washington DC. Salis, E. (2012), “Labour migration governance in The Word Bank (2013c), Remittance Prices World- contemporary Europe. The case of Italy,” FIERI wide, Issue No. 8—December 2013, https:// WP, LAB-MIG-GOV Project, April: http://www remittanceprices.worldbank.org/sites/default/ .labmiggov.eu/wp-content/uploads/2012/04/ files/RPW_Report_Dec2013.pdf LABMIGOV_WP1_Italy_Final-report.pdf Yang, D. (2011), “Migrant Remittances,” Jour- Schoumaker, B., & Diagne, A. (2010), Migrations nal of Economic Perspectives, Volume 25:3, between Africa and Europe: Data Collection pp. 129–152. 39 AN N E X 1 FROM THE SURVEY DESIGN TO THE FIELDWORK: A METHODOLOGICAL NOTE (2013), INED (Schoumaker & Diagne, 2010), and The Questionnaire CeSPI-ABI (CeSPI, 2012, 2013). Moreover, every After the research objectives and the survey time it was possible, the content, wording and sample structure were defined together with label of questions were formulated as close as the World Bank, the research team within FIERI possible to those included in the Italian Labor and Labor (Laboratorio R. Revelli worked at the Force Survey and other national and international preparation of the Questionnaire and of the field surveys (Bank of Italy, Eurostat) in order to allow work. The general framework and the questions a comparison of survey results with existing offi- about remittances are inspired by previous stud- cial statistics. ies on migration and remittances. In particular, the research team grounded the questionnaire struc- The questionnaire was intended for face-to-face ture on a previous empirical research run by the interviews lasting around 20–30 minutes each and World Bank in Czech Republic (Nicolì and Carro- administered by trained interviewers in Italian, but zza, 2010) as well as on other surveys conducted with the possibility of translating questions and by the International Organization for Migration concepts also in Spanish, French and English. FIGURE 39: Questionnaire structure Intro & privacy statement Survey on migrants’ financial needs and Section A: Filter questions behavior in Turin –To be in Italy since at least 1 year – –To reside in Turin (and the surrounding area) Questionnaire structure –To have an income (any type of source) –To rend remittances at least one since the beginning of 2013) Section B: Personal information (sex, age, birth and current Section C: Info on family components and HH structure citizenship, marital status, religion, education and training) (sex, age, birth and current citizenship, marital status). Section D: Individual and HH income. Section E: Financial inclusion BCA, savings and other bank products, access to credit and microcredit. Section F: Remittances Intensity, freq., operators & costs, direction & use. Received remittances. Section G: Occupation Type of job, productive sector, contract and hours. Only for those Section H: Legal status in Italy since Section I: Entry documents and permit to stay, at the beginning and now. 2008 and Crisis’s impact before Section L: Training/Information on financial services. Follow up: Availability to be recontracted for training and follow-up. For the interviewer: Post-interview comments. 40 GREENBACK 2.0 REPORT 2014 The structure of the questionnaire (Fig. 39) was magazines, children toys and stationary) to intended to first understand whether or not be offered to the interviewed migrants as an include the migrant in the sample through four incentive/reward for their participation. filter questions and then to address each specific Before starting the fieldwork phase, a pilot of section—demography, income and earnings, finan- around a dozen interviews was conducted in cial inclusion, remittances, job and occupation, order to test the length, structure and consistency legal status and availability to further training—in of the questionnaire. The test was very important a logical sequence for the interviewed. for interviewers to verify the interviewing tech- niques and to test the difficulty of single ques- Training of interviewers tions, their exact wording, and the correct way and pilots to register received answers. Especially for the section on economic and financial situation of The fieldwork team was composed by six quali- migrants and their families, the pilot interviews fied interviewers, one of which was the coordina- provided with important feedbacks on the sensi- tor and the person in charge of keeping contacts tiveness and clearness of each specific question. with the research team and the interviewers at the Finally, interviewers were trained on the most same time. effective way to convince migrants to complete the whole questionnaire, how to offer an incentive As a preliminary phase to the fieldwork, the inter- (voucher) and how to obtain consent to the use viewers attended a specific training on the project of sensible data (privacy statement). objectives and on the overall structure and logic of the study, on the specific content of each ques- Thanks to this pilot, it was noted that some spe- tion of the questionnaire and on the strategies for cific wordings and concepts had to be further the sample definition and the interview process. clarified to interviewed migrants. For example, the During the training session, the questionnaire’s word “remittance” was defined as “the money you draft was given to the interviewers along with the send to your origin country,” which all the inter- related documentation for the fieldwork: viewees understood without any difficulty. While, • leaflet on the project to be distributed to inter- in general, the comprehension was good enough, viewed migrants in some cases migrants needed more explanations • a preliminary list of entry points and key infor- on concepts and content of the question from the mants of the three communities interviewer. Moreover, during the test some ques- • a guide (20 pages) to support interviewers in tions were modified and some response options the fieldwork phase, with practical insights on added (e.g. “member of a cooperative” was added interviewing and data collection techniques among types of occupation). The test revealed and with a detailed description of each section that on average each interview would last around of the questionnaire, to lower problems of twenty minutes. interpretation and wording Feedbacks from the pilot interviews were dis- • privacy statements to be signed by migrants cussed together with interviewers and integrated to guarantee the anonymity of the interview in the final version of the questionnaire. • follow-up forms through which interview- ers collected contacts information (name, telephone number and/or email addresses) The fieldwork: sampling strategy of migrants interested in taking part in edu- and entry points cational and non-educational meetings in a The fieldwork timespan was from 01 July to 05 second phase of the Greenback 2.0 project September 2013. Pilot interviews were collected • personal access to the online platform between 01 and 05 July 2013. Survey Monkey, for the registration of questionnaire data Interviewers were sub-divided into three mini- teams made up of two people, each one focused • vouchers in the value of 10 euros, lasting one on one specific community (80 interviews per year, for purchases in a book shop (books, FROM THE SURVEY DESIGN TO THE FIELDWORK: A METHODOLOGICAL NOTE 41 researcher, 160 interviews per community, 480 basic in order to lower possible bias and distor- interviews in total). This subdivision gave them tions in the final sample definition and structure: the opportunity to improve and optimize com- • no more than ten contacts/interviews from munication and coordination between them- one entry point; selves on the entry points, centers and key • no more than five contacts/interviews from informants specific for each of the three com- an individual included in the sample (already munities, to avoid overlapping and repetitions. interviewed). This community-focused approach helped the insertion of the interviewers in the three commu- The first criterion is grounded on the center sam- nities: analyzing just one community each, they pling technique presented by Blangiardo (2004) were able to obtain a good understanding of the and based upon the recognition of some refer- assigned community. ence points for a specific community (a commu- nity leader, the representative or president of an Starting from an initial list, interviewers constantly association, a renowned cultural mediator etc.), updated a comprehensive dataset of key infor- who collaborated with the interviewers to facili- mants and of collective entry points, with opening tate their approach and to get better understand- hours and contacts of reference individuals. This ing a community. The second criterion applies preliminary phase of mapping the city of Turin in the snowball sampling, a special non-probability terms of places and reference points for each of method used when a desired sample is rare or the three communities was even more fundamen- not fully known in advance. It relies on referrals tal as the survey took place during the summer from initial subjects to generate additional waves period, when many public offices, local associa- of respondents from among their acquaintances, tions and organizations are closed. The final list so that the sampled group appears to grow like is made of centers of interests of many different a rolling snowball10. In our case, snowballing types—places of work, recreational clubs, public allowed interviewers to receive suggestions from offices and local authorities, religious centers and interviewed migrants on some other co-nationals commercial companies—some specific to each with the required characteristics to be included of the three communities and some more cross- in the survey and possibly interested in partici- cultural, multi-lingual and general. For example, pating. A positive aspect of using this recalling researchers concentrated in the areas surrounding technique is that it makes it easier to earn trust the three Consulates for the three communities, in and to engage them in the survey, as the first some parks and green areas renowned to be hang contact is made via the presentation of a friend or out by Peruvian migrants (Pellerina and Colletta co-national. Park), in a market area in Corso Racconigi and in a specific part of the Valentino Park where there are Nevertheless, a large number of total interviews some kiosks and stalls run by Romanians. More were collected through an anonymous approach cross-cultural entry points were represented by of the interviewers in public/open places (pub- public offices such as migrant offices within the lic gardens, streets etc.). The constant presence police headquarters (Questura) and the munici- of the researchers in the aforementioned areas pality of Turin, as well as charities, NGOs and allowed them to be recognized and to be wel- local-based associations for migrant integration comed by the different communities. In some (Gruppo Abele, Alma Mater, Asai, Fratia, etc.). cases, from a preliminary contact in a street or in These places were precious for the data collection a public space, the interviewers were invited to as their characteristics allowed for greater acces- access private dwellings in order to be more com- sibility to migrants with different socio-economic fortable during the interview and to take part into profiles, age structure, origins and length of stay, hence helping to lower the selection bias in build- 10 Proposed by Goodman (1961), this selection process is clearly non- ing the sample. random and it lowers search costs at the expense of introducing bias, because the technique itself reduces the likelihood that the sample will rep- Although the timing of the survey made the map- resent a good cross section from the population. Nevertheless, a number of variants have been tested to overcome limits and bias that non-randomness ping of viable entry points more difficult, inter- implies. As regards to Italy, Strozza (2004) reports on various examples viewers were still able to comply with the two where different types of snowball sampling have been tested. 42 GREENBACK 2.0 REPORT 2014 FIGURE 40: Map of Turin and data collection points community events, presenting an opportunity for a final sample of 480 interviews it was neces- interviewers to earn the trust of the entire com- sary to interview more than twice that amount munity or single individuals. of migrants. Finally, the consistency and representativeness of Moreover, a descriptive analysis of main reasons the final sample should be attributed to a good for exclusion revealed some interesting and wor- composition of timing and hours of fieldwork. rying characteristics of intercepted migrants. Interviewers were able to conduct interviews at Around 60 percent of the excluded did not meet any time during the day and on any day of the one or both criteria regarding employment and week. In this way, interviewers were able to inter- remittances. Many migrants (especially Moroc- cept migrants with different preferences accord- cans) claimed to be unemployed or became ing to their working hours (full- and part-time, on unemployed because of the economic crisis, or call and domestic workers etc.) and their individ- did not sent money to their origin country since ual and religious habits. the beginning of 2013 because of the instability of or reduction in their earnings. Among those used to send remittances but who have interrupted this Feedbacks from the fieldwork activity, some migrants explained they can no lon- A first, somewhat underestimated issue in the ger afford to send money home because they are planning phase was that of entry requirements not able to save enough, while for some others for including a migrant in the sample. Interview- remittances are no longer needed since they have ers excluded a high number of potential sub- completed the family reunification process. jects because of their lack of one or more of the filter criteria (residing in Turin, being in Italy The second issue to be taken into consideration at least since one year, employed, and send- during the fieldwork was the composition of sam- ing remittances). The fieldwork team collected ple. Although the preliminary phase of prepara- basic information on the migrants surveyed, but tion of the fieldwork was very accurate, constant not included in the sample, recording the main coordination and collaboration among interview- reasons of their exclusion. Even though these ers were fundamental to monitor the progress and data are incomplete and underestimate the total to make any necessary adjustment to the data number of people interviewed during the field- collection strategy in order to have a balanced work, not including people who refused any kind sample structure in terms of origin country, age, of approach, it is worth noticing that to obtain length of stay in Italy and gender. As for gender FROM THE SURVEY DESIGN TO THE FIELDWORK: A METHODOLOGICAL NOTE 43 TABLE 18: Intercepted migrants excluded from Finally, although many questions included in the survey the questionnaire regarded sensitive issues like earnings, use of money, relations with banks and Reason for Exclusion Freq. Percent financial operators, legal status etc., almost all Not available 143 26.0 interviewed migrants responded in a collaborative Lives out of Turin 44 8.0 and complete way. Very few interviews were not fully completed if migrants refused to answer to Not in employment 111 20.2 one or more questions because they felt uncom- Does not send remittances 107 19.5 fortable with the subject or were not able to give Does not send remittances anymore 115 20.9 details. No interview has been excluded from the Student 15 2.7 analysis since the overall completion rate was in Does not speak Italian 15 2.7 all cases very satisfactory (more than 97 percent Total 550 100.0 of questions). Connected to this, interesting insights came from migrants’ reactions to the incentive they balance within each subsample, interviewers received (a €10 voucher to be spent in a book- found it difficult to receive feedbacks from Moroc- shop). In many cases, interviewers found it easier can women. Among those contacted to take to give the voucher after the interview, as a form part into the study, many were ineligible because of reward and gratitude to the time the migrant they did not meet one or more of the selection dedicated to the survey instead as incentive to criteria regarding employment and/or control convince a migrant to participate. Moreover, and over earnings. this is maybe due to the specific type of voucher, interviewers report that many interviewed The survey period represented a further issue for migrants demonstrated scarce interest in the book the fieldwork team. Not only some local offices voucher as they were not interested in bookshop and migrants’ association are closed during the products and would have preferred a different summer period, but also migrants tend to return type of benefit (e.g. voucher/discount card for a to their origin countries during July and August supermarket). Nevertheless, they didn’t refuse to to visit their families. Moreover, a special effort participate because of the lack of an appropriate was made to interview Moroccan migrants as incentive and therefore demonstrated on average the survey period coincided with Ramadan, a positive attitude and availability to be included the Islamic holy month of fasting, during which in the study. observant individuals are naturally less willing to be contacted. Data entry, cleaning and analysis Among all public events and celebrations where In order to minimize, as much as possible, data interviewers have been invited, two were particu- entry errors and gaps, interviewers were asked to larly meaningful for the relevant community: the proceed with the data entry on a rolling basis, in Peruvian Fiestas Patrias (National Independence parallel with the fieldwork. Day) celebrated on the 28th July at the Pala Ruffini and the Eid El-Fitr (a religious holiday sig- To collect data in an efficient, easy and timely naling the end of Ramadan) held at Parco Dora on manner, the recorded data was uploaded to an the 8th of August. The participation of interview- online survey tool. The online platform was acces- ers to these two events produced different results: sible only to interviewers to upload data and to while during the celebration of Fiestas Patrias the research team to continuously monitor the the recreational atmosphere facilitated the inter- formation of the sample and to adjust its structure views of many Peruvian migrants, the religious if needed. At the end of the surveying period, the celebration of Eid El-Fitr produce a delicate and online survey platform was closed and the over- intimate environment which prevent interviewers all dataset was exported and processed by the to even try to interview Moroccan migrants that research team, using STATA as statistical software were attending. to produce descriptive statistics and data analysis. 45 AN N E X 2 DETAILED STATISTICS A. Italy’s remittance and foreign resident population data TABLE 19: Remittance outflows by Italian Regions, 2012. € Region € (thousands) % (per capita) 2011–2012 var. % Abruzzi 71,613 1.0 725 –9.2 Basilicata 17,158 0.3 1299 –3.7 Calabria 91,663 1.3 1369 –7.7 Campania 403,896 5.9 2658 –5.0 Emilia Romagna 422,954 6.2 929 –13.0 Friuli V.G. 98,072 1.4 1007 34.7 Lazio 2,022,701 29.6 4725 –5.1 Liguria 177,061 2.6 1581 –9.4 Lombardy 1,451,377 21.2 1524 –7.9 Marche 106,375 1.6 793 –5.1 Piedmont 286,898 0.1 795 –12.1 Apulia 163,636 4.2 1956 9.8 Sardinia 59,802 2.4 1922 –7.7 Sicily 329,015 0.9 2595 3.1 Tuscany 599,240 4.8 1856 –13.7 Trentino Alto Adige 55,311 8.8 646 –12.9 Umbria 65,813 0.8 747 –11.4 Aosta Valley 7,908 1.0 933 –20.0 Veneto 423276 0.1 922 –15.2 Molise 9,346 6.2 1147 –13.1 Total 6,833,116 100.0 1673 –7.6 46 GREENBACK 2.0 REPORT 2014 TABLE 20: Remittance outflows by destination country, 2012. Destination Thousand € % 2011–2012 var. % China 2,674,453 39.1 5.4 Romania 810,950 11.9 –9.4 Philippines 366,807 5.4 –39 Morocco 242,510 3.5 –19.1 Bangladesh 228,178 3.3 –21.4 Senegal 216,264 3.2 –11.9 India 198,060 2.9 –3.7 Peru 187,651 2.7 –3.3 Ukraine 152,705 2.2 –8.3 Ecuador 137,385 2 –11.6 Total 6,833,116 100 –7.6 TABLE 21: Remittance outflows from the Province of Turin, first 20 destination countries 2005–2012 (€ million). Destination 2005 2006 2007 2008 2009 2010 2011 2012 Col %, 2012 Romania 33.980 44.040 44.608 43.384 46.547 48.197 50.916 46.366 28.2 Peru 2.963 7.291 8.655 10.984 13.968 16.252 18.858 17.780 10.8 Morocco 12.402 17.272 21.272 19.860 17.735 16.833 19.649 13.532 8.2 Senegal 16.670 23.824 26.010 24.087 19.490 18.426 16.406 11.448 7.0 China 2.629 2.280 2.823 1.255 1.070 4.894 4.983 9.706 5.9 Brazil 3.928 5.813 5.923 6.308 8.880 5.909 10.923 8.227 5.0 Philippines 9.882 12.064 11.861 15.364 14.776 14.162 13.933 5.541 3.4 Albania 5.200 6.239 6.037 5.685 5.442 5.461 5.210 4.973 3.0 Moldavia 2.459 3.155 3.284 2.863 2.927 4.053 4.638 4.299 2.6 Bangladesh 0.421 0.638 0.973 1.775 2.851 2.980 4.421 3.339 2.0 Ecuador 1.399 3.141 3.070 2.812 2.978 2.866 3.414 3.071 1.9 Spain 3.042 3.131 3.108 3.654 3.509 3.530 3.280 2.677 1.6 Dominican Rep. 0.961 1.315 1.079 1.395 1.737 2.056 2.161 2.168 1.3 France 1.840 2.551 3.559 2.909 2.471 2.312 2.245 1.927 1.2 Colombia 1.382 1.876 2.195 2.262 2.292 2.414 2.505 1.914 1.2 India 0.410 0.706 0.703 1.117 1.280 1.336 1.624 1.558 0.9 Ukraine 0.798 1.111 1.143 1.272 1.328 1.410 1.667 1.503 0.9 Egypt 0.595 0.741 0.746 0.883 1.018 1.234 1.427 1.482 0.9 Nigeria 1.545 1.726 2.813 2.647 2.559 2.432 2.233 1.434 0.9 Ivory Coast 2.419 3.351 3.735 2.710 1.353 0.965 0.760 1.137 0.7 Total 121.758 164.027 180.411 180.361 180.262 180.538 193.321 164.577 100 DETAILED STATISTICS 47 TABLE 22: Foreign resident population in the Province of Turin, first 20 origin countries, 2003–2011. Country of Origin 2003 2004 2005 2006 2007 2008 2009 2010 2011 Romania 10787 24818 33628 39570 44158 73557 85817 91786 95258 Morocco 13540 15636 19532 20830 22511 23895 26247 27999 28693 Albania 5557 6850 7930 8522 9165 9713 10493 10861 11136 Peru 3779 5173 5922 6470 7044 7500 8388 9769 10481 China 2840 3389 4357 4869 5483 5829 6221 6832 7553 Moldova 357 1594 2396 2957 3417 3808 4756 5641 6685 Egypt 1453 1786 2332 2602 2913 3242 3601 3960 4344 Nigeria 1491 1818 2194 2349 2595 2807 3064 3337 3638 Philippines 1758 2014 2278 2429 2562 2748 2995 3227 3494 Brazil 1380 1683 1828 2049 2297 2524 2774 2953 3040 Tunisia 1225 1361 1536 1675 1813 1948 2095 2136 2121 France 1724 1763 1851 1892 1966 2056 2111 2072 2111 Ecuador 271 809 1061 1252 1423 1534 1654 1783 1907 Senegal 808 954 1297 1420 1480 1519 1599 1650 1688 Ukraine 222 501 695 811 915 1011 1123 1279 1388 Poland 595 681 794 916 1049 1209 1270 1313 1347 Spain 868 873 913 950 988 1055 1091 1090 1112 Bosnia Herz. 751 838 808 856 913 931 968 992 1045 United Kingdom 865 897 1001 1002 1028 1072 1087 1060 1031 Germany 781 778 797 854 899 950 984 972 958 Province of Turin, tot 62084 86728 106276 118284 129533 164592 185073 198249 207488 B. Survey data: remittances TABLE 23: Flows’ recipients, by main channel. Bank Transfer Money Transfer H.-c. Cash to Cash   Freq. % Freq. % Freq. % Parents 27 77.14 271 53.24 35 55.56 Children 1 2.86 34 6.68 4 6.35 Siblings 4 11.43 112 22 11 17.46 Partner 30 5.89 5 7.94 Partner’s relatives 2 5.71 21 4.13 1 1.59 Other relatives 1 2.86 35 6.88 5 7.94 Friends 6 1.18 2 3.17 Total 35 100 509 100 63 100 48 GREENBACK 2.0 REPORT 2014 TABLE 24: Average amount sent per year (classes), by country. Morocco Peru Romania   Freq. % Freq. % Freq. % <200 28 12.23 16 8.65 33 16.1 201–500 41 17.9 21 11.35 49 23.9 501–1000 77 33.62 32 17.3 44 21.46 1001–2500 62 27.07 68 36.76 43 20.98 2501–5000 12 5.24 40 21.62 20 9.76 5000+ 9 3.93 8 4.32 16 7.8 Total 229 100 185 100 205 100 TABLE 25: Average amount sent per year, by type of channel used. Mean Std. Dev. N. flows Min Max Money transfer € 1,442.5 1652.9 503 50 14400 Cash to cash € 1,075.0 1249.1 60 100 6000 Bank transfer € 1,677.3 2027.2 33 100 9000 Others* € 1,752.1 . 12 100 7200 Total € 1,425.1 1649.0 608 50 14400 * Includes Count to cash, Prepaid card and Twin cards Migrants’ Remittances from Italy International remittances and access to financial servicesfor migrants in Turin, Italy A survey by Cover photo: courtesy of Michele D’Ottavio