Free trade agreements and regional alliances: support from Latin American legislators

The creation of international organizations and the signature of free trade agreements have become common in Latin America. However, few studies have tackled the attitudes of political elites. This work aims to analyze the determinants of support by Latin American legislators for free trade agreements with the US and the EU, and for the Pacific Alliance and ALBA. Results show that ideology, pro-state/market positions, and trustworthiness vis-à-vis Chinese and Russian governments are the main predictors. This paper aims to analyze the causes of support in Latin American parliaments for Free Trade Agreements (FTAs) with the US and the EU, as well as for two regional alliances, the Pacific Alliance and the Bolivarian Alliance for the Peoples of Our America (ALBA). To achieve this, we employ the Latin American Elites Database from the University of Salamanca (PELA-USAL), which gathers interviews with Latin American legislators on issues such as democracy, economy, political careers and, in recent years, international relations. This article extends to all Latin American countries where questions on the aforementioned FPIs have been made available: Bolivia, Chile, Colombia, Costa Rica, the Dominican Republic, Guatemala, Honduras, El Salvador, Mexico, Panama, Uruguay, and Venezuela. This research can be seen as appropriate due to its special focus: legislators are themselves the who eventually must approve these FPIs in their respective legislative chambers. Understanding what legislators think and what drives their support can clearly help to identify the causes of success or failure of and regional alliances in Latin America, specifically at the elite legislative level.


Introduction
I n recent decades, several international organisms have been created in Latin America and the Caribbean. At the same time, free trade agreements have become common in the region; not only among Latin American countries but between Latin American countries and other international actors such as the United States (US) or the European Union (EU). Nevertheless, the level of support for these Foreign Policy Instruments (FPIs) among political elites, especially legislators, remains unknown, as do the reasons behind their support or rejection of such instruments. Few studies have approached the attitudes of political elites in Latin America towards regional alliances and free trade agreements. A notable study by Achard et al. (2001) interviews Latin American political leaders and concludes that, although a majority of these elites express belief in Latin American cultural identity, they also defend the existence of regional blocks and favor closer integration with the US.
comes not only from national governments but also from markets (Mattli 1999;Malamud 2011) and from civil society (Puchala 1968;Thakur and Langenhove 2006).
Five types of regionalism can be identified within the literature: regionalization or transfer of state functions to regional organizations; creation of a regional identity to strengthen symbolic and cultural ties; interstate cooperation to approach common conflicts jointly; economic integration to create a common market; and political integration to generate supranational institutions (Hurrell 1995). Economic integration can be said to have four phases: establishment of a free trade area that eliminates customs barriers; a customs union that imposes a single tariff on products from third countries; a common market that implies not only the customs union but the mobility of capital and labor, and finally; an economic union that involves the creation of a common currency and a single monetary policy (Balassa 1965;Malamud 2011).
Regarding Latin American integration, the first salient point is that this region includes territories with highly distinct political and economic realities, although the majority of countries share a similar colonial past (Sánchez Muñoz 2012, 704). Integration processes in this part of the world have until now expressed a state-centric orientation (Vieira 2008, 17), resulting in part from the important role that heads of state have played (Malamud 2008). This situation has been emphasized even more since the early 21st century through the emergence of strong leaders and new models critical of liberalism and FTAs (Hobsbawm 2008;Lowenthal 2018). Unlike in Europe, Latin American integration has not been promoted as a way to end wars, nor has it implied significant changes of sovereignty or generated supra-regional institutions with decision-making capacities (Malamud 2008).
However, Latin American regionalism has been affected by external actors, fundamentally the US and EU, with which certain countries have adopted various political and economic relationships (Sanahuja 2004;Russell and Tokatlian 2008;Vigevani and Magnotta 2016). These foreign entities' attitudes toward Latin America have served as an "important external catalyst of the new Latin American regionalism, because they have accepted, and ultimately encouraged, a regional relationship (beyond country-to-country relations) as regards trade agreements" 1 (Sanahuja 2004, 83). However, during the first decade of the 2000s, regional progressive governments reviewed their relationships with former allies, especially with the US (Monteiro 2014). In fact, Latin American regionalism has oscillated between models of cooperation under the tutelage of the US and the development of intra-Latin American integration schemes (Bernal-Meza 2008, 155). At the same time, China and other emerging countries from Asia, Europe, and the Middle East are taking interest in the region and have gained significant economic and commercial weight over the last decade (Cesarín and Moneta 2005;Quiliconi and Salgado 2017;Lowenthal 2018).
In this sense, we expect that right-leaning Latin American legislators support FTAs with the US due to the importance of pro-US ideology (Amoroso and Silva 2017). Likewise, we expect that legislators with high trust in the Chinese and Russian governments show greater support for ALBA and less support for FTAs with the US and EU, or for the Pacific Alliance, because they may perceive China and Russia as alternative powers to counter the military and commercial hegemony of the US (Öniş 2017). Historical antecedents have also played a fundamental role in the current scenario of Latin American integration, which still is the subject of debate "between the Bolivarian conception of democratic 'continentality' and the nationalist conception of the creole elites" 2 (Vieira 2008, 17-18). In the past, there have been alliances and ruptures among the different states in terms of political integration, such as the Federal Republic of Central America (1824)(1825)(1826)(1827)(1828)(1829)(1830)(1831)(1832)(1833)(1834)(1835)(1836)(1837)(1838) in the first years of independence. Throughout the 19th century, Hispanic-American and Pan-American conferences were periodically held, and these continued into the middle of the 20th century. Taken altogether, these provide the historical background for more recent conferences (Vieira 2008).
All of the great Latin American integration projects first began to take shape following the Second World War, in three differentiated stages. The first, between 1950 and 1980, was promoted by the Economic Commission for Latin America and the Caribbean (ECLAC) and inspired by Developmentalist and Dependence Theory (Cardoso and Faletto 1979). In this period, the Viña del Mar Latin American Consensus was reached, in which twenty countries, advised by the ECLAC, agreed upon certain trade integration proposals ("Consenso Latinoamericano de Viña del Mar." 1969). Elsewhere, numerous other projects for regional integration were fostered, including: the Latin American Free Trade Association (LAFTA), which was the first attempt to implement a free trade area in the region; 3 the Latin American and the Caribbean Economic System (SELA), whose aim was to specify common trade strategies between the various countries; and the Latin American Parliament (Parlatino), which sought to encourage the development and political integration of all Latin American countries against colonialism. Likewise, sub-regional integration organizations were fomented, such as the Organization of Central American States (ODECA), the Andean Community (CAN), and the Caribbean Community (CARICOM). Similarly, the first international blocs in the region were created during this era (Vieira 2008;Sánchez Muñoz 2012).
The second stage, called open regionalism, began in the 1980s and emphasized trade openness, economic liberalization, and negotiation of FTAs (Vieira 2008, 19;Sanahuja 2009, 12). During this period, the Enterprise of the Americas Initiative was established by the US to promote free trade and foreign investment in Latin America. Other initiatives included the Rio Group, to stimulate political agreement and cooperation, 4 the Central American Integration System (SICA), to help achieve peace and development in Central America, and the Southern Common Market (Mercosur), to create a free trade area amongst the countries of the Southern Cone (Vieira 2008;Sanahuja 2009;Sánchez Muñoz 2012;Bernal-Meza 2013). Open regionalism had varied results: on the one hand, Latin American countries managed to eliminate certain customs duties and to increase foreign investment; on the other hand, trade barriers were maintained, common policies were not implemented, stable institutions were not created, and initiatives failed to correct structural asymmetries at the national and inter-country levels (Obaya 2007;Sanahuja 2009).
The third stage, known as post-neoliberal regionalism, was sparked by the failure of the Free Trade Area of Americas (FTAA) conference at Mar de Plata in 2005, and the creation of the Union of South American Nations (USAN) three years later. This period is characterized by reaffirmation of the state's relevance, and by focusing the integration agenda on political and social aims, instead of free trade and foreign investment (Veiga and Ríos 2006;Sanahuja 2009;Bizzozero 2011;Bernal-Meza 2013;Briceño Ruiz and Morales 2017). Thus, post-neoliberalism is not only a set of foreign policies but an ideological project (Marston 2015;Ruckert et al. 2017).
However, unlike open regionalism (which enjoyed consensus by all countries around free trade), the present stage of regional integration is not homogeneous but can be distinguished into three Colombia, Chile, Mexico, Peru, the Dominican Republic, and all the Central American countries except Nicaragua constitute the open integration axis. These countries, with neoliberal or globalist governments, support FTAs in alliance with transnational firms, excluding some strategic sectors of the state (Mercier 2000;Carranza 2017;Laursen, 2018). Such ideological patterns can be found in the main organization of open integration, the North American Free Trade Agreement (NAFTA) (Briceño Ruiz 2013, 15), established between the US, Canada, and Mexico (the latter incorporated in 1994). This agreement is based on general rules and has little institutional structure, and its performance "looks more like a couple of bilateral agreements with a center in the US than a triangular relationship" 5 (Malamud 2011, 236). The US tried to expand this model through the FTAA but, after its failure in 2005, chose to pursue bilateral FTAs with the countries from the open integration axis. In 2007, these countries grouped together into the Forum of the Latin American Pacific Arc, which became the Pacific Alliance in 2011. This is a regional bloc whose aim is the promotion of deep integration and free trade in the region as well as the establishment of a mechanism for common negotiations (Briceño Ruiz 2012Santos et al. 2017;Ortiz 2017;Serrano 2017). This initiative defends economic and trade integration instead of political integration, calls Brazil's regional leadership into question, and turns the page on the Latin American vs. South American discussion by incorporating Mexico (Malamud 2012, 3-4).
Argentina, Brazil, Paraguay, and Uruguay belong to the revisionist axis. All are countries with nationalist governments that promote a productive regionalism model and look for integration compatible with a reaffirmation of national sovereignty (Serbin 2011, 147). These countries share some elements with developmentalist regionalism, such as favoring industrial development, diversifying national economies, and the promotion of participation by local and national companies (Briceño Ruiz 2013; Errejón and Guijarro 2016). Mercosur is the main organization of the revisionist axis. Although this integration process was created based on the mandates of open regionalism, since 2003 it has emphasized its social and productive contents and has tried to consolidate a "new South American regionalism" led by Brazil (Briceño Ruiz 2013, 17). The other great project of revisionist integration − also promoted by Brazil − is USAN, a regional alliance that groups all countries from South America and advocates integration with political, social, environmental, and military aims. It is worth noting that with Dilma Rousseff 's assumption of the presidency, Brazilian foreign policy was oriented to strengthening its relations with US (Caballero 2011, 169). This "turn" was accentuated during the governments of Michel Temer in Brazil and Mauricio Macri in Argentina, who defended a policy of commercial opening, closer to an open integration axis (Frenkel and Azzi 2018).
Finally, Cuba, Bolivia, Ecuador, Nicaragua, Venezuela, and some Caribbean countries are located in the anti-systemic axis. All of these are supporters of a social regionalism model based on the promotion of social rights, redistributive policies, and the fight against poverty and social exclusion, in collaboration with civil society (Yeates 2005; Deacon et al. 2007;Briceño Ruiz 2013). This axis also defends a decolonial approach of regional integration, summarized in the sentence: "our north is the south" (Gomes 2018, 2). The anti-systemic countries are grouped into ALBA, a regional alliance promoted by Hugo Chávez in 2002 and consolidated two years later (Dabène 2009;Briceño Ruiz 2011). Some academics consider that this alliance advocates not only economic integration but also a space of political and economic influence (Buck 2010, 397) that "seems to have a more ideological than commercial face" 6 (Altmann 2008, 5). Meanwhile, other authors describe it as an initiative of anti-capitalist and anti-imperialist integration (Martínez 2006) and hold it to be an instrument of economic integration, because it promotes supranational productive projects and establishes a common economic area (Briceño Ruiz 2011Santos et al. 2017).
Consequently, and in view of prior works (Briceño Ruiz 2013; Marston 2015; Ruckert et al. 2017;Atkins 2018), we expect that the ideological-programmatic positions of Latin American legislators may explain their levels of support for FTAs and regional alliances. Legislators who consider themselves ideologically left-wing and/or favorable to the "state" will manifest higher support for ALBA and lower support for the FTAs and the Pacific Alliance. Conversely, right-leaning legislators and/or those favorable to the "market" will exhibit higher support for FTAs and the Pacific Alliance, and lower support for ALBA.

Methodological approach: political elites and international relations
In this work we analyze the determinants of support for regional alliances and FTAs among political elites, who can be defined as persons that, due to their strategic positioning, affect national political results on a substantial and regular basis (Higley and Gunther 1992). It is important to study these elites, in order to understand their preferences and profiles (Coller et al. 2014). Few studies clearly define who these political elites that rule a country really are, what they think, or what the political implications of their attitudes might be (García and Mateos 2006, 12).
Among the different types of elites, political elites are those with the greatest influence on the strategic decisions that shape the standards of living within a society (Hoffmann-Langue 2007).

Figure 1. Regional integration axes in Latin America
These elites can be composed of sub-groups (Blondel and Müller-Rommel 2007), such as members of a cabinet, mayors, senators, representatives, etc. Among these subgroups, national legislative elites are the most relevant, as they represent the citizens, have some control over the Executive, and influence the political system (Aberbach et al. 1981); moreover, they can be clearly identified through the positional approach (Coller et al. 2016;Hoffmann-Langue 2007). In democratic regimes, parliaments are the place of representation par excellence (Alcántara and Llamazares 1997), and the elected representatives are those who make fundamental decisions (García et al. 2013, 136).
For these reasons, understanding the support for regional alliances and trade agreements is critical, because it helps in characterizing the foreign policy approaches of countries. Hence, to analyze legislators' attitudes toward our four FPIs of interest, we rely on the Latin American Elites Database from the University of Salamanca (PELA-USAL), which compiles interviews with Latin American legislators from 18 countries since 1994 (García et al. 2013). These personal interviews are conducted by members of the PELA-USAL research group following every legislative election. 7 The questionnaires cover a wide range of issues (Barragán 2015), the main topics being democracy, elections, political parties, institutions, the military, the role of the state, public policies, clientelism, corruption, ideology, political careers, social values, and foreign policy.
In recent years, due to the growing importance of the latter issue, four new questions have been introduced on two different international instruments: international integration alliances (ALBA and Pacific Alliance), and two hypothetical FTAs between Latin America and the US and EU. Legislators are asked to answer on a 10-point scale, 1 being very negative and 10 very positive, in order to assess these four FPIs. So far, these four questions have been posed in the following legislatures, each from a different country:  After a descriptive analysis of the four FPIs of interest to this work, we analyze the determinant factors of Latin American legislators' support for free trade agreements with the US and EU, as well as for ALBA and the Pacific Alliance, by way of four linear regressions that consider the following groups of explanatory variables: • Ideological-programmatic positions. A self-placement question that asks legislators to locate themselves on a 10-point left-to-right scale. As this measure is not always a reliable indicator in all Latin American countries, we also include a self-placement question on the role of the state in the economy along a 10-point scale, 1 being state and 10 market. The state/market logic is not necessarily associated with the left/right logic in Latin America (Alcántara and Rivas 2007), since pro-state rightist (Luna and Rovira Kaltwasser 2014) and pro-market leftist (Weyland et al. 2010;Flores-Macías 2010) positions also exist. Therefore, based on the evidence of our data 8 and prior research through PELA-USAL (see Martí and Santiuste 2008;, we have chosen to include both variables as representative of ideological-programmatic positions. As mentioned above, we expect that right-leaning legislators and proximity to "market" preference will imply greater support for both FTAs and the Pacific Alliance, and lower support for ALBA. • Trustworthiness of the Chinese 9 and Russian governments. Legislators are asked if they consider these governments very trustworthy (4), trustworthy (3), untrustworthy (2) or very untrustworthy (1). As mentioned above, because China and Russia are geopolitical rivals to the US (and EU), we expect that the more supportive legislators are of these foreign governments, the less they will support FTAs with the US and EU and the Pacific Alliance (open-integration axis, in tune with US), but the more positively they will assess ALBA (anti-systemic axis).
• Country of origin. In view of the important differences between cases with regard to foreign policy, as will be shown in the next section, we have included dummy variables for every country to control for a possible country effect, beyond ideology or the perceived trustworthiness of foreign governments. We have chosen Mexico as the country of reference because it is the country with the widest sample of interviewees (100). Therefore, following Figure 1 and Table 1, one country can be called revisionist (Uruguay), two are anti-systemic (Bolivia and Venezuela), and nine are in favor of open integration (Chile, Colombia, Costa Rica, Honduras, Panama, El Salvador, Mexico, Guatemala and the Dominican Republic). Thus, given the differences in these three axes (Briceño Ruiz 2013), we expect open integration countries to be more supportive of both FTAs and the Pacific Alliance and less supportive of ALBA. The opposite will be true regarding anti-systemic integration, and for revisionists we expect mixed results, although with higher support for ALBA, -given that it is a clearly Latin American FPI.
• Government/opposition. Given that it would be particularly challenging to include the political party dimension in the explanation (covering 12 countries with very distinct party systems), we opt for an intermediate solution to control for possible political party effects, thus without theoretical expectations. We include a dummy proxy (1 for opposition) to control for whether the legislator belongs to the opposition party or the party in government. Beyond this methodological rationale, there is a further objective. Because convergence between government and opposition might confirm the very existence of consensus (Mújica and Sánchez-Cuenca 2006;Miller and Stecker 2008), here we focus on attitudes of government and opposition deputies to explore the degrees of consensus on that dimension. This approach is not new when analyzing, for instance, market reforms in the region (Armijo and Faucher 2002). Absence of statistical significance means government/opposition consensus, and presence means dissension.
• Socio-demographics. Along with government/opposition, we have included the following standard control variables: gender (1 for female), age, education (6-point scale, 1 being no studies, and 6 being postgraduate studies), and specifically whether a legislator has studied abroad or not (1 for abroad). Figure 2 describes the average support for FTAs with the US and the EU, and with ALBA and the Pacific Alliance, in 12 Latin American parliaments. An FTA with the EU is the international instrument with most support from legislators, followed by the Pacific Alliance and an FTA with the US. The least valued international instrument by the legislators is ALBA. This means that Latin American elites generally prefer a foreign policy leaning toward open integration, rather than anti-systemic positions, following the classification by Briceño Ruiz (2013), given that ALBA is a clear example of the Bolivarian regional project.

Descriptive analysis
Source: own elaboration from PELA-USAL Figures 3 to 6 represent the weighted support for these four FPIs by country. With regard to an FTA with the US (Figure 3), Chilean legislators are the most favorable, followed by legislators from Mexico, the Dominican Republic, El Salvador, Panama, and Honduras. Uruguay is the country whose deputies are the least favorable to an FTA with the US, followed by Bolivia, Costa Rica, Venezuela, Guatemala, and Colombia. Not surprisingly, the open integration axis is clearly in favor of trading with the US, with the exception of Colombia, Guatemala, and Costa Rica. Figure 4 shows the weighted support for an FTA with the EU in Latin American parliaments. As with the US, Chile is the country whose deputies are most favorable to an FTA with the EU, followed by Mexico and the Dominican Republic. However, Uruguayan parliamentarians are not necessarily opposed to an FTA with the EU (as occurs with the US), but instead are above the regional average in terms of support. Once again, Costa Rica, Guatemala, and Colombia (open integration axis) are below the regional average, with Bolivia showing higher support. Finally, the country that least supports an FTA with the EU is Venezuela.
In relation to regional alliances, the results of the weighted support for ALBA among Latin American legislators run counter to the results for both FTAs. As can be seen in Figure 5, Bolivian parliamentarians, sometimes detractors of an FTA with the US, are most favorable to ALBA, and these are followed by Hondurans, Dominicans and Mexicans -all clear supporters of the FTAs. On the other hand, Chile is the country where the legislators are least favorable to ALBA, followed by Guatemala, El Salvador, and Panama; all of which show high support for the FTAs.  1 represents the average support in the region (6.07) Source: own elaboration from PELA-USAL

Figure 3. Weighted support for an FTA with the US throughout Latin American parliaments
Interestingly, Venezuelan legislators are below the regional average; that is, they are (comparatively) slightly contrary to ALBA although their own country promotes this regional alliance. It is worth mentioning that the legislature included in this work (2016-2021) has been composed of an opposition majority, following the 2015 legislative elections. These opposition forces are generally against Bolivarian proclamations and projects, such as ALBA, which are linked to the ruling party, the United Socialist Party of Venezuela (PSUV in Spanish).   Finally, the weighted support for the Pacific Alliance in Latin American parliaments is presented in Figure 6. Mexican parliamentarians are the most supportive of this alliance, followed by Bolivia (without access to the Pacific Ocean), Chile, Colombia, and Uruguay. On the other hand, Venezuela (also without access to the Pacific Ocean) is the country where the deputies are least favorable to this alliance, followed by Costa Rica, Guatemala, Panama, El Salvador, the Dominican Republic, and Honduras -all of whom have access to the Pacific, with the exception of the Dominican Republic. Table 2 shows the four regressions and, consequently, the main findings of the analysis. With regard to control variables, although we do not hold theoretical expectations, findings are noteworthy and deserve some description. Gender has no relevance and age is significant, but with a rather low coefficient and with significance only for the Pacific Alliance, which older legislators dislike. On the contrary, education does appear to be significant: higher levels of education lead to lower levels of support for ALBA and the Pacific Alliance, but to higher support for an FTA with the US. Also, interestingly, the fact of having studied abroad is both strong and significant only when assessing ALBA. Legislators who have studied abroad, usually in the US or Europe (mainly the United Kingdom, France, and Spain) have a more negative opinion of ALBA, the Bolivarian Latin American alliance. This variable is irrelevant for the rest of the FPIs. None of these  profile-related variables show significance for an FTA with the EU, meaning that its determinants are more attitudinal than socio-demographic. The explanatory power of both ideology and degree of statism confirms our hypothesis. Ideological-programmatic bias for all four projects is quite evident in the light of these results. ALBA is preferred by left-wing, pro-state legislators while free trade with the US or EU and the Pacific Alliance (which aims to represent, in a certain way, a free trade area) is preferred by right-wing, pro-market legislators. All coefficients are significant above 0.05, and rather high, particularly for ideology in the case of an FTA with the US. Along with these ideological positions, what legislators think about the Chinese and Russian governments does matter in terms of relations with the US, the EU, and other Latin American countries. The trustworthiness of the Chinese government affects the perception of FTAs with both the US and EU, not negatively but positively, contrary to popular belief, and with positive strong and significant coefficients. Thus, China would not be viewed as a rival or substitute, but as another complementary partner. However, attitudes towards the Chinese government are not influential in the assessment of Latin American alliances.

Determinants of Latin American foreign policy attitudes in parliaments
On the other hand, contrary to China, in terms of the trustworthiness of the Russian government, the effects confirm our hypothesis: the more legislators trust the Russian government, the less they support free trade with both the US and EU, and the more they support ALBA. In this specific case, Russia can be considered a geopolitical rival and substitute for Western powers. Worth noting is the high coefficient for ALBA -one of the strongest at the individual level for all four regressions -behind only ideology in relation to an FTA with the US.
The government/opposition variable is significant only for an FTA with the US and for ALBA, not for an FTA with the EU or the Pacific Alliance. The latter have similar levels of support (or rejection) from both government and opposition parties (government/opposition consensus), the former more dependent on which party holds which offices (government/ opposition dissension).
Hence, ALBA receives more support from legislators whose party or coalition holds the presidency, and free trade with the US receives more support from legislators that have lost elections or that do not belong to the presidential coalition. It is worth mentioning, once again, that half of the included legislatures had left and center-left governments during this period (Bolivia, Chile, Costa Rica, El Salvador, Uruguay, and Venezuela), jointly characterized as the left-turn (Cameron and Hershberg 2010). Consequently, if ideological position has this strong explanatory power, it should not be surprising that the government/opposition variable has the aforementioned effects. Still, two of our FPIs of interest (free trade with the EU and the Pacific Alliance) are outside this winner/loser dynamic, for better or worse.
Finally, in relation to country effects, the results must be interpreted bearing in mind that Mexico is the country of reference. In Figures 3 to 6, Mexico showed clear, above-average support in all the FPIs, which would explain the generally negative values in all four regressions (although with some exceptions). However, differences between the countries do exist, beyond ideology and profile. When compared to Mexicans, Guatemalans and Venezuelans are more reluctant toward free trade with the US or EU and with our two regional alliances of interest. Meanwhile, only Chile and Uruguay are more supportive than Mexico with regard to an FTA with the EU. It is worth mentioning that ALBA has the least significant country effects; Colombia, Costa Rica, the Dominican Republic, El Salvador, Honduras, and Panama show no significant differences when compared to Mexico.

Conclusions
Latin American foreign policy remains far from consensual, or from becoming a significant national issue, -even among political elites. The fact that there are so many significant variables in the analysis proves that clear and identifiable determinants that explains legislators' attitudes towards FTAs and regional alliances do indeed exist. If integration and economic openness enjoyed more neutral and transversal support, then ideology, statism, or the trustworthiness of international actors such as Russia would not condition opinions regarding FTAs, ALBA, or the Pacific Alliance.
Among the socio-demographic variables, only two have been found to be significant. Deputies with high levels of education (Masters or PhD) are more reluctant toward both regional alliances, even though this effect disappears for free trade. More interestingly, the fact of having studied abroad incites higher levels of reluctance only toward ALBA. Thus, ALBA can be said to be (more) rejected by legislators who have been highly educated abroad -a kind of hyper-elite, so to say.
According to legislators' ideological positions, ALBA is a leftist, pro-state project, while the FTAs and Pacific Alliance are rightist, pro-market projects. Also, Russia and China have proved to be relevant, although in different ways. While Russia is (still) considered a rival or competitor to the US and EU, confirming our expectations, China is seen more as a partner to Latin America, which contradicts our expectations. Trust in the Russian government leads to lower support for FTAs, and higher support for ALBA; trust in China's government leads to higher support for FTAs.
What legislators think about these governments, or whether the state should have an active role (or not) in the economy, or where they place themselves on the left/right ideological scale will determine their attitudes towards foreign policy. Moreover, this is coincident with the three phases of regionalism already mentioned (developmentalism, neoliberalism, and post-neoliberalism), with evident ideological rationale. Hence, this paper has found empirical evidence that foreign policy depends in part on what party holds the legislative majority, which eventually will hinder the consolidation of strong and durable Latin American integration projects, or which will maintain stable free trade agreements as medium-or long-term prospects.
If Latin American legislators' attitudes toward foreign policy depend so much on other attitudinal factors, it comes as no surprise that the international position of every country will change (and has changed) depending on what party wins elections. Once again, the opposition vs. government dissension proved to be significant. In fact, until Mauricio Macri and Michel Temer took office, in 2015 and 2016 respectively, Brazil and Argentina (revisionist axis) showed no negative behavior toward ALBA, and less-than favorable attitudes toward FTAs with the US, especially in the case Argentina. One final observation with regard to free trade agreements should be made. In the PELA-USAL questionnaire, the question inquiries about an FTA with US/EU for Latin America, and not for individual countries. Currently, bilateral agreements are more common than multilateral agreements.
In short, this article investigates support for different FPIs in Latin American parliaments with empirical survey data, finding that ideological-programmatic positions, perceived trustworthiness of the Chinese and Russian governments, and country of origin are the key determinants of support for FTAs with the US and EU, the Pacific Alliance and ALBA. These findings constitute a theoretical contribution to the new agenda of international relations in Latin America, since they confirm the ideological bias of these projects and the unmistakable influence of external actors in the region's international arena. This being a first step, future research should focus on the evolution of legislators' opinions, parallel to the inclusion of new countries from the three axes into the analysis, in order to better understand the reasons for the successes and failures of FPIs in Latin America.
Probably, future hypothetical changes in governments and legislative majorities will lead to the disappearance of some of these FPIs, and the birth of new ones, as was the case throughout Latin America during the 20th century. President Donald Trump in the US is not especially rhetorically favorable to free trade. Domestic problems of the EU members and the deep economic crisis in Venezuela, which leads ALBA, are political phenomena that might well affect the field of international relations on the continent.
Be that as it may, as long as ideological positions and the trustworthiness of non-Latin American governments continue to matter in the formulation of foreign policy opinions, there will be no lasting projects for trade and integration. As long as there is no consensus, at least among political elites, there will be no true integration in Latin America.