Family Law Institutions and Intimate-Partner Violence Ana Tur-Prats* Preliminary draft February 2013 Abstract This paper analyses the relationship between intimate-partner violence (IPV) and historical family law systems in Spain. Survey data shows that those regions that have had historically a different family law system in which women had a more prominent role in the family have nowadays a lower IPV rate. I test the causality of this relationship by using the land tenure structure that arose during the Christian ‘reconquest’ of the Iberian Peninsula (712-1492) as an instrument for the different family law institutions. Both OLS and IV estimates are consistent and show a negative and significant relationship. Keywords: Intimate-partner Violence, Legal Institutions, Family Organization, Long-Term Persistence. * PhD Student, Universitat Autònoma de Barcelona. Affiliate Research Student, University College London. Contact: Economics Department, University College London. Gower Street. London WC1E 6BT, United Kingdom. E-mail: ana.tur-prats.09@ucl.ac.uk. I thank Marcos Vera-Hernández for his guidance and supervision. For comments and discussions, I am also thankful to Prashant Bharadwaj, Gordon Dahl, Mercedes Fernández-Martorell, Isabel Fernández-Mateo, Amy Finkelstein, Christian Fons-Rosen, Enrique García-Bernal, Rafael Lalive, Albert Lamarca, Nathan Nunn, Javier Ortega, Luigi Pascale, Jaume Puig-Junoy, Pablo Salvador Coderch, Michele Tertilt and Felipe Valencia. I gratefully acknowledge financial support from Fundación Ramón Areces. 1 1. Introduction Intimate-partner violence (IPV) is a social problem of great concern that has serious consequences for women’s health. It is also widely spread across countries, with reported lifetime prevalence that varies from 15% to 71% (Garcia-Moreno et al., 2006). Understanding the factors that cause this kind of abuse is a key question in social sciences, especially since IPV is a complex and multi-triggered issue. From the field of economics, in the last decade several studies have tried to tackle this question. Thanks to these papers we know that emotional cues due to the unexpected defeat of the football team that the man supports lead to an increase of IPV (Card and Dahl, 2011). We also know that husbands can use violence to signal their dissatisfaction with the marriage, and this way extract higher transfers from their wives’ parents (Bolch and Rao, 2002, in rural Indian context). Other studies have identified factors that contribute to the decrease of domestic violence: the introduction of unilateral divorce (Stenvenson and Wolfers, 2006), or the gender wage gap reduction (Aizer, 2010). Regarding to specific policies, e.g. the cash transfers to women in developing countries, authors have found that the effect depends on the level of education of the women and their partners (Hidrobo and Fernald, 2013), and that the total effect on abuse is ambiguous (Bobonis et al., 2013). This paper contributes to the analysis of the IPV causes by studying the relationship between intimate-partner violence (IPV) and historical family law systems in Spain. Data shows that those regions that have had historically a different family law system (Basque Country, Navarre, Aragon, Catalonia and Balearic Islands, also known as ‘foral’ territories) have nowadays a lower IPV rate. In these territories, women had a more prominent role in the family: wives had greater management of marital goods, widows could appoint the heir, and they also held the life interest in the property in ‘foral’ territories whereas in the rest of Spain widows had no legal protection. Map A1 and A2 in the appendix show IPV rates by provinces and ‘foral’ provinces respectively. My hypothesis is that these different family law institutions shaped a different gender identity and that this has had a long-term and persistent impact that is important for explaining violence against women today. Both the historical origins of countries’ laws (La Porta, 2 López-de-Silanes and Shleifer 2008) and the origins of gender roles (Alesina, Giuliano and Nunn, 2013) have received recent attention in the economics literature. I then test the causality of this relationship by using the land tenure structure that arose during the ‘reconquest’ as an instrument for the different family law institutions. The ‘reconquest’ is a centuries-long period (712-1492) in which several Christian kingdoms repopulated and took control of the Iberian Peninsula from the Islamic rulers. The uniqueness of this event in the history of Europe provides an excellent source of exogenous variation to study the consequences of different family institutions and household organizations. 2. Historical background The so-called ‘foral’ territories –i.e. Vizcaya (in Basque Country), Navarre, Aragon (formed by 3 provinces: Huesca, Zaragoza and Teruel), Catalonia (formed by 4 provinces: Lleida, Girona, Barcelona and Tarragona) and Balearic Islands-, all have in common the age-old existence of a house; namely a stable peasant broad family together with the farm that supports their living. The small size of the farm would only assure the family and future generations’ survival if it was undivided. This kind of peasant family remained very stable throughout the centuries until the industrialization. Their family and inheritance legal systems’ rationale was thus to guarantee the conservation and continuity of the family heritage. This is shown in specific institutions that were distinct from the ones established in the rest of Spain, where Castilian law was in force. In this regard, the most paradigmatic family institution of the ‘foral’ regions is the single heir/heiress as opposed to the equal division of bequest between offspring that was promoted under Castilian law. For its economic feasibility the farm in ‘foral’ territories required all family members’ participation and cooperation. This family structure fostered greater organization to reinforce cohesion. Moreover, women had a more prominent role in the family: wives had greater management of marital goods, widows could appoint the heir/heiress, and they also held the life interest in the property in ‘foral’ territories whereas in the rest of Spain widows had no legal protection. Since societies that were initially more pro-women could had selected into this kind of household structure and family law system we need an instrument to try to establish causal 3 effects. Therefore, I use the land tenure structure that arose during the ‘reconquest’ as an instrument for the different family law institutions. The reconquest is a centuries-long period (712-1492) in which several Christian kingdoms repopulated and took control of the Iberian Peninsula from the Islamic rulers. In the north, where the reconquest started, the resettlement was more spontaneous and had a less official nature. It favoured small and private land ownership by free and independent peasants. This is known as “presura” in the northwest and “aprisio” in the northeast. As time went by, towards the south, state structure was developing and clergy and nobility participation in the reconquest were rewarded with vast extensions of land. Therefore, the way resettlement took place gave rise to a different land tenure structure. In the north there were more small holdings owned by free peasants. These small farms were mainly subsistence-oriented and they needed to be undivided in order to sustain the family. In the south, however, more large states owned by nobles or church were found. These large plots are known as latifundia, and they hired landless peasants or day labourers to work the land. 3. Data IPV data comes from the four cross-sectional surveys on violence against women in Spain “macroencuesta sobre la violencia contra la mujer en Espana”. The first three surveys were conducted by phone in 1999, 2002 and 2005 (sample sizes 20552, 20652 and 28423, respectively). The last survey was conducted through a personal interview in 2011 (n=7898). The advantage of these surveys is that they contain a broad sample of women living in Spain (n=77,525) and different IPV measures: self-reported and objective. In this paper I use the socalled objective measure since self-reported measures tend to underestimate domestic violence and the degree of bias can be non-random. In the questionnaire there are 26 questions specifically designed to detect this kind of violence; 13 of them are considered as an indicator of domestic violence. I then construct an IPV variable that takes the value 1 if the woman answers “often” or “sometimes” to at least one of these 13 questions, and 0 otherwise. These surveys also include information at the individual level on the woman’s and partner’s level of educational, woman’s job status, household’s reference person, marital status and 4 religious and political beliefs. We also know in which province they live. Regional data on provinces’ per capita GDP and unemployment rate from the National Institute of Statistics are added. To quantify the instrument, I use the time in which each province was reconquered. Based on the map of the Spanish reconquest by Lomax (1978), I assign to each province a date from a set of 7 categories: 914, 1080, 1130, 1210, 1250, 1480 and 1492. Map A3 in the appendix shows the different stages of the reconquest. 4. Results OLS estimates show that having the ‘foral’ family system is associated with 2 percentage points less intimate-partner violence, which amounts to 20% of the national average rate. This effect persists after controlling for socio-demographic variables, regional economic development, religious and political beliefs, current institutions and share of immigrants. It remains stable through the different specifications and statistically significant as shown in table 1. Table 1: OLS estimates. ‘Foral’ family law system Economic development Religious and political beliefs Current institutions ImmigrantĖ N (1) (2) -0.017*** (0.0037) No No No No 54,332 -0.017*** (0.0040) Yes No No No 54,332 (3) (4) -0.019*** -0.021*** (0.0050) (0.0064) Yes Yes Yes Yes No Yes No No 32,533 32,533 (5) -0.016** (0.0071) Yes Yes Yes Yes 15,394 Model (1) includes age, children, woman’s and partner’s level of education, woman’s job status, household’s reference person, marital status, habitat size and year when survey was conducted. Model (2) adds provinces’ per capita GDP and unemployment rate. Model (3) adds a categorical variable indicating if the woman is a catholic and a measure (1-5) of political ideology (1: left wing, 5: right wing). Model (4) adds a dummy variable for provinces with separation of property marriage regime. Model (5) adds information on women's nationality. Standard errors, in brackets, computed applying a cluster structure by province. *** Statistically significant at 1%, ** 5%. Ė Information on immigrant status is only available in 2005 and 2011 survey. I then test the causality of this relationship by using the year in which each province was reconquered (seven categories) as an instrument for family law institutions. I also use an alternative instrument based on the percentage of the province surface with large plots 5 (greater than 500 hectares), and finally I report estimates combining both instruments. Table 2 shows the IV estimates; these are consistent with OLS estimates. I find a negative and statistically significant effect of having the ‘foral’ family law system on IPV of approximately 3 percentage points. Table 2: IV estimates (a) ‘Foral' family law system Time in which province was reconquered % of surface with large plots N -0.030** (0.0131) Yes No 54,332 (b) (c) -0.037*** (0.0116) No Yes 52,367 -0.029*** (0.0084) Yes Yes 52,367 Model (a) uses the time in which the province was resettled as an instrument for having a different family law system. Model (b) uses the % of the province’s surface occupied by plots larger than 500 hectares as an instrument. Model (c) combines both instruments. Control variables are age, children, woman’s and partner’s level of education, woman’s job status, household’s reference person, marital status, habitat size, provinces’ per capita GDP and unemployment rate, and year when survey was conducted. Standard errors, in brackets, computed applying a cluster structure by province. *** Statistically significant at 1%, ** 5%. 5. Conclusions This paper analyses the relationship between intimate-partner violence (IPV) in Spain and historical family law systems. My hypothesis is that different family law institutions shaped a different gender identity and that this has had a long-term and persistent impact that is important for explaining violence against women today. Survey data shows that those regions that have had historically a different family law system (Basque Country, Navarre, Aragon, Catalonia and Balearic Islands, also known as ‘foral’ territories) have nowadays 2 percentage points less intimate-partner violence. This effect persists after controlling for socio-demographic variables, regional economic development, religious and political beliefs, share of immigrants and current institutions. I then test the causality of this relationship by using the land tenure structure that arose during the Christian reconquest of the Iberian Peninsula (712-1492) as an instrument for the different family law institutions. Both OLS and IV estimates are consistent and show a negative and significant relationship. 6 This study contributes to the understanding of how history can help to explain family organization and how legal institutions can have a persistent and long-term impact on interpersonal relations. 7 References Aizer, Anna. 2010. “The Gender Wage Gap and Domestic Violence”. The American Economic Review, 100(4): 1847-59. Alesina, Alberto, Paola Giuliano, and Nathan Nunn. 2013. “On the Origins of Gender Roles: Women and the Plough”. Forthcoming in The Quarterly Journal of Economics, 128(2). Bobonis, Gustavo, Melissa González-Brenes, and Roberto Castro. 2013. “Public Transfers and Domestic Violence: The Roles of Private Information and Spousal Control”. American Economic Journal: Economic Policy, 5(1): 179-205. Bolch, Francis, and Vijayendra Rao. 2002. “Terror as a Bargaining Instrument: A Case Study of Dowry Violence in Rural India”. The American Economic Review, 92(4): 1029-43. Card, David, and Gordon B. Dahl. 2011. “Family Violence and Football: The Effect of Unexpected Emotional Cues on Violent Behavior”. The Quarterly Journal of Economics, 126: 1-41. Castán Tobeñas, José. Derecho civil español, común y foral. T I, Vol 1. Reus S.A. Madrid.1988. De Moxó y Ortiz de Villajos, Salvador. Repoblación y sociedad en la España cristiana medieval. Ediciones Rialp. Madrid. 1979. Garcia-Moreno, Claudia, Henrica AFM Jansen, Mary Ellsberg, Lori Heise, Charlotte H Watts, on behalf of the WHO Multi-country Study on Women’s Health and Domestic Violence agains Women Study Team. 2006. “Prevalence of intimate partner violence: findings from the WHO multi-country study on women’s health and domestic violence”. The Lancet, 368(7): 1260-1269. Hidrobo, Melissa, and Lia Fernald. 2013. “Cash transfers and domestic violence”. Journal of Health Economics, 32: 304-319. La Porta, Rafael, Florencio López-de-Silanes, and Andrei Shleifer. 2008. “The Economic Consequences of Legal Origins”. Journal of Economic Literature, 46(2): 285-332. Lomax, Derek W. The Reconquest of Spain. Longman, London and New York. 1978. Mikelarena Peña, Fernando. 1992. “Las estructuras familiares en la España tradicional: geografía y análisis a partir del censo de 1860”. Boletín de la Asociación de Demografía Histórica, X(3): 15-61. Moret y Prendesgast, Segismundo y Luis Sevilla. La familia foral y la familia castellana. Academia Matritense de Jurisprudencia y Legislación. Madrid. 1863. 8 Segalen, Martine. Love and power in the peasant family: rural France in the 19th century. University of Chicago Press. 1983. Stevenson, Betsy, and Justin Wolfers. 2006. “Bargaining in the Shadow of the Law: Divorce Laws and Family Distress”. The Quarterly Journal of Economics, 121(1): 267-88. 9 Appendix Map A1. IPV rate, by provinces Source: Own elaboration from survey data on violence against women in Spain (1999, 2002, 2005 and 2011) Map A2. ‘Foral’ provinces (in dark grey) Source: As listed in Castán Tobeñas (1988), core group formed by 10 provinces: Vizcaya, Navarre, Huesca, Zaragoza, Teruel, Lleida, Girona, Barcelona, Tarragona and Balearic Islands. 10 Map A3. Stages of the reconquest. Source: Lomax (1978), p.ix. 11