Social assistance, social cohesion and stigma. The case of Asignaciones Familiares-Plan de Equidad in Uruguay
Vigorito, Andrea (2016). 'Social assistance, social cohesion and stigma. The case of Asignaciones Familiares-Plan de Equidad in Uruguay' Paper presented at the annual conference of the HDCA, Tokyo 2016.
I. Introduction and main purposes of the study
The aim of this paper is to assess the effect of social assistance on social cohesion and stigma on program beneficiaries, based on data from a large cash transfer program launched in Uruguay in 2008. The relevance of promoting social cohesion as a constitutive aspect of well-being and, hence as key dimension of the development process has been advocated not only by academics such as Sen (1999) but also by multilateral actors such as, for example, World Bank, European Union, OECD or ECLAC.
Although there is a large literature on the impact of cash transfer programs on a varied set of outcomes (Boullion and Tejerina, 2007; Fiszbein et al, 2009), there are scarce quantitative studies assessing their effect on social cohesion. However, from the capabilities perspective this topic is extremely relevant, as social programs can be promoting the alleviation of short run income poverty, but might erode long run social outcomes. Social cohesion is an issue of outmost importance in any effort to correct the traditional dualization of social protection in developing countries (Martínez-Franzoni and Sánchez-Ancochea, 2012).
In the context of developing countries, the scarce existing evidence provides contradictory results. For instance, Attanasio et al (2008) report positive effects of CCTs on social capital measured as willingness to invest in collective projects and social participation. Meanwhile, Chong et al (2009) report that welfare destroys social ties via stigmatization. Vera Soares et al (2010) find no effects on social participation in the Paraguayan transfer program Tekoporá, while Kardan et al (2010) evaluate an emergency transfer plan in rural Zimbabwe and find that the program had negative consequences in social relations due to the targeting process. King et al (2010) carried out a systematic review assessing the impact of community driven development interventions (CDD) promoting social cohesion in sub-Saharan Africa and found a positive effect on trust in community members and a negative one in the case of inter group relations. These results alert on the complexities of within and between group social cohesion. In order to contribute to this ongoing debate, this paper studies the impact of a large Uruguayan cash transfer program—Asignaciones Familiares-Plan de Equidad (AFAM-PE)—on social cohesion related outcomes, based on survey data.
Different definitions of social cohesion available put different emphasis on shared values, inequalities and civic engagement (Colleta, ; Green et al; Jantti, ; Portes and Vikstrom, 2015). In all cases, social solidarity/cohesion entail cooperative behavior towards other individuals and collectives, a sense of value-based commitment (Wilde, 2007; Thome, 1999) and mutual esteem (Honneth, 2007; Wilde, 2007). Solidarity and social cohesion have different dimensions and levels (Whelan and Maitre, 2005; Vergolini, 2011; King et al, 2010), that include objective and subjective manifestations and processes that unfold at the micro (group level), meso (relation with other groups) and macro (sense of belonging to society and trust). Cooperative links between individuals and groups can have a negative as well as a positive effect on desirable outcomes and can also provoke conflict, as it has been highlighted by the feminist and Marxian literatures (Jamaat et al, 2015). In this study, we assess social cohesion by considering three main groups of outcomes: household and the family, social relations and institutions. The first group of outcomes focuses on women empowerment and family ties. The second group is centered in trust in other people, stigma (as the reverse feeling), tolerance and social participation. Finally, in the third group, confidence in a wide range of public and social institutions is explored.
II. The programme
AFAM-PE was launched in 2008. It is a large intervention targeted to vulnerable households with children, carried out by the Uruguayan Ministry of Social Development (MIDES). At present it reaches 45% of the Uruguayan children aged 0 to 17. It provides a monthly stipend to households with children in order to alleviate poverty and foster schooling. The transfer is conditional on school attendance for children aged 6 to 17 and it is mainly delivered to women. Previous studies show that AFAM-PE had a positive effect on school attendance for teen-agers (Bérgolo et al, 2014).
III. Methodology and data
The evaluation strategy exploits the discontinuous nature of program assignment using regression discontinuity techniques (Lee and Card, 2008). Program entrance has been strictly based in a score that predicts poverty status combining a set of household variables. On this basis, different regressions are run for each outcome, in order to explore the robustness of the results to different parametric specifications and the inclusion of baseline control variables. The RDD identification assumption requires outcome variables to be a monotonic function of the score other than for the effect of treatment (Imbens and Lemieux, 2008). The potential discontinuity in the outcome variables around the discontinuity point can, hence, be interpreted as the effect of the program. Standard tests for the validity of the identification assumption are also run (Lee and Lemieux, 2010) and absence of discontinuity in the running variable (McCrary, 2008). One drawback of the RDD is that it provides local average treatment effects in the neighborhood to the discontinuity point that – in the presence of heterogeneous effects - cannot be necessarily generalized to the beneficiaries of the program as a whole.
Data come from the AFAM-PE follow up survey carried out in November 2011-March 2012 and the administrative records of program applicants, which contains information on successful and unsuccessful applicants.
The AFAM-PE follow-up survey contains information on housing, household composition, durables, work, income and schooling, the survey collected information on health, economic expectations, knowledge of political and civil rights, participation in social groups, information on peers for teen agers, a wide range of opinions and social attitudes. The survey sample contains data on approximately 3000 households, including both eligible and ineligible applicants, in a 4% neighborhood of the program eligibility threshold score.