The 2008 AmericasBarometer surveys, carried out by the Latin American Public Opinion Project (LAPOP), involved interviews conducted in 24 nations in Latin America and the Caribbean with a total of 40,567 probabilistically selected respondents, interviewed face-to-face (with the exception of web surveys in the U.S. and Canada).2 In the particular case of support for free trade, a total of 29,982 respondents from 19 countries were asked the following question, to which they could respond on a 7-point scale AmericasBarometer Insights: 2010 (No.40 ) where 1 means not at all and 7 means a lot: The Influence of B48. To what extent do you believe that free trade agreements will help to improve the economy? Religion on Support Figure 1. for Free Trade in Average Support for Free Trade in Latin America and the Caribbean, 2008 1 Latin America Uruguay 63.5 Belize 62.4 Chile 61.0 Alejandro Díaz‐Domínguez Dominican Rep. 58.3 Jamaica 57.4 Vanderbilt University Costa Rica 56.8 Venezuela 53.9 alejandro.diaz‐[email protected] Colombia 52.9 Mexico 50.8 Peru 50.4 hy do individuals in Latin America Panama 50.2 Brazil 49.4 and the Caribbean support free trade Argentina 48.4 policies? Some scholars have argued Nicaragua 47.4 W Ecuador 46.4 that such policies can produce negative Honduras 42.9 Paraguay 42.1 consequences and economic insecurities Guatemala 41.7 (Goldberg and Pavcnik 2004; Merolla, El Salvador 41.4 Stephenson, Wilson and Zechmeister 2005). Yet, 0 20 40 60 support for free trade exists among many in the Support for Free Trade region. In this Insights report, I discuss the 95% C.I. (Design-Effects Based) importance of one particular influence on Source: AmericasBarometer by LAPOP 2008; 19 countries; 29,982 respondents support for free trade policies: religious factors. Specifically, I argue that membership in religious communities provides mental and monetary support, which offsets the costs of, Figure 1 shows national averages for the 19 3 and therefore reduces resistance to, free trade countries surveyed. In order to facilitate policies. comparisons, these responses were recoded on a 2 Funding for the 2008 AmericasBarometer round was The Insights Series is co‐edited by Professors Mitchell A. mainly provided by the United States Agency for Seligson and Elizabeth Zechmeister with administrative, International Development (USAID). Other important technical, and intellectual support from the LAPOP group at sources of support were the Inter‐American Development Vanderbilt University. Bank (IADB), the United Nations Development Programme 1 Prior issues in the Insights series can be found at: (UNDP), the Center for the Americas (CFA), and Vanderbilt http://www.vanderbilt.edu/lapop/studiesandpublications University. The data on which they are based can be found at 3 Across these countries an average of 13.5% of respondents http://www.vanderbilt.edu/lapop/datasets did not answer the free trade question. © 2010, Latin American Public Opinion Project, “Insights” series www.AmericasBarometer.org Page 1 of 6 0‐100 scale. Uruguay is the country with the attendance and affiliations can provide a type of highest support for free trade with an average of social insurance. 63.5 points. A total of eleven countries exceed the 50 unit mid‐point indicating average More specifically, the theory is that adverse life events such as unemployment, illness, or support, whereas only four nations have workplace accidents generate monetary costs, average support levels lower than 43 points. In but also impose important psychological costs, sum, Figure 1 indicates that trade is supported such as stress and loss of self-esteem (Scheve by many in the region, despite the fact that and Stasavage 2006a: 137). Religious attendance popular benefits are debatable (Goldberg and and affiliations potentially reduce the Pavcnik 2004). psychological and financial costs of those adverse life events because Churches can offer Theoretical Perspectives comfort and support during difficult times (Mainwaring 1986; Scheve and Stasavage 2006a; Prior scholarship suggests that there is a link 2006b; Daniels and von der Ruhr 2005; between international economic integration Hagopian 2008). Therefore, if religious associated with free trade and workers’ affiliations do provide such “insurance”, we insecurities due to the elasticity of labor should find all else equal that those who demand, i.e. the ability of firms to easily participate in religious life have higher levels of substitute workers (Scheve and Slaughter 2004; support for free trade than those who do not.5 Merolla, Stephenson, Wilson and Zechmeister 2005: 584). Thus, conventional theories explain Modeling Support for Free Trade support for free trade using variables that reflect individuals’ positions in the economy, by factor In order to test the relationship between (e.g., level of education) and/or sector (e.g., religious variables and support for free trade in agriculture). At the same time, there is some the 19 countries surveyed, I use a survey linear evidence that the importance of those model (a least squares regression for survey determinants is declining in comparison to the design). The dependent variable is respondents’ initial stage of trade reforms in Latin America level of support for free trade, measured using (Baker 2003; 2009).4 the variable described in Figure 1. The key independent variables are religious Assuming there are nonetheless still some costs denominations and church attendance. to free trade, I focus on a different theoretical perspective, which relates religious factors to In addition, I control for a number of potential support for free trade. My argument states that predictors of support for free trade. These certain aspects of religion reduce psychological include sociotropic and pocketbook evaluations and monetary costs associated with free trade. of the economy (Seligson 1999; Merolla, Stephenson, Wilson and Zechmeister 2005); left Put differently, assuming that free trade and right in politics (Milner and Judkins 2004; imposes, at a minimum, some psychological Magaloni and Romero 2008); interpersonal trust and, at a maximum, some economic costs as as measure of potential social networks; interest well, it is possible to import and adapt a in politics; and, both left-right placements and theoretical perspective that suggests the preferences over the role of the state with existence of a religious coping effect. At the core of this perspective is the notion that religious 4 I ran different models using factor and/or sector variables 5 The distribution of religious people in the 19 countries related to the Stolper Samuelson theorem (Rogowski 1989); analyzed is 67.7% of Catholics (20,123 cases); 6.9% of the Hecksher-Ohlin model (Ohlin 1967); and the Ricardo- Protestants (2,043 cases); 11.5% of Evangelicals (3,429 cases); Viner theorem (Alt, Frieden, Gilligan, Rodrik and Rogowski 2.0% of LDS and Jehovah’s Witness (596 cases); 2.1% of 1996), but neither factor nor sector variables reached Eastern and Traditional religions (616 cases); and 9.9% of conventional levels of statistical significance, excepting people who do not profess any faith (2,922 cases). The last levels of education. category will be the reference category in the model. © 2010, Latin American Public Opinion Project, “Insights” series www.AmericasBarometer.org Page 2 of 6 respect to its social role6 and its role as owner of that does not overlap the vertical “0” line (at .05 critical industries (Baker 2003; Magaloni and or better). When the dot, which represents the Romero 2008; Zechmeister and Corral 2010). predicted impact of that variable, falls to the right of the vertical “0” line, it implies a positive Other control variables in the model are relationship whereas when it falls to the left, it demographic and socio-economic measures: indicates a negative contribution.8 levels of education (Merolla, Stephenson, Wilson and Zechmeister 2005; Hainmueller and Hiscox My expectations regarding the coping effect of 2006), wealth measured by ownership of assets religion on support for free trade are supported (Scheve and Slaughter 2001), size of residence, by the model. In particular, Catholics, gender (Seligson 1999), age, and indigenous Protestants, and Evangelicals tend to support identity (Magaloni and Romero 2008). free trade more in comparison to people who do not profess any religion, which is the reference In addition, some scholars argue that a category. In addition, Church attendance has a relationship between religious factors and free positive effect on support for free trade. In other trade would simply suggest that religious words, as expected, those who attend Church people tend to be more conservative (De la O more frequently hold more favorable opinions and Rodden 2008: 439), and others suggest that regarding free trade agreements. Christian (in particular non-Catholic) affiliations tend to be more oriented toward Figure 2. Support for Free Trade in Latin America internationalism, in particular toward the U.S. and the Caribbean, 2008 (Rodriguez 1982). Controlling for socially conservative attitudes and international ties Catholic allows me to assert with more confidence that Protestant the connection between religious factors and Evangelical Eastern and Trad Rel support for free trade is due to a coping effect, LDS and Witness Church Attendance and not one of these other plausible National Eco Situation State Socially Responsible explanations. I therefore include a proxy Left to Right Politics variable for moral traditionalism by means of Interpersonal Trust Personal Eco Situation opinions on homosexuals’ rights to run for Political Interest State Should Own Industries office, and a very rough proxy of international Relatives in the US linkages by means of whether the respondent Education Wealth 7 has contact with relatives in the U.S. Size of Residence Reject Homosexuals' Rights Female Country fixed effects Age and intercept included but not shown here The Religious Coping Effect Indigenous -.05 0 .05 .1 .15 The significance of the variables in the model is 95% C.I.