Edurne Zoco, "Legislators' Positions and Party System Competition in Central America: A Comparative Analysis," Party Politics, 12 (March, 2006), 257-280.

First Paragraph:
What political dimensions are relevant in Central America? What are the political differences between the two major parties within Central American countries? This article conducts a spatial analysis of Central American parties while relying on the assumption that the left-right continuum is a valid instrument with which to organize political competition at the elite (i.e. legislator) level in these countries even though the meaning of 'left' and 'right' in Central America might differ from the traditional connotation of these terms in European and North American contexts. Moreover, the article analyzes the dimensions of political competition in Central America at both individual (legislator) and aggregate (political party) levels to show the absence of a straightforward correspondence between the number and substance of political dimensions dividing legislators and the number and substance of dimensions of party competition. This analysis thus adds new insight to the discussion about the role of agency (i.e. parties and political elites) in affecting the way that cleavages occur in the political domain. The article also shows the absence of a clear correspondence between (a) the degree of intensity of political competition (polarization) and (b) the number and substance of the politically relevant dimensions of competition in each country, implying that strongly polarized party systems are not always associated with a large number or with specific types (e.g. military, economic) of political dimensions.

Figures and Tables:
Table 1. Left-right placement of legislators and parties in Central America
Table 2. Reliability measures (all countries)
Table 3. Factor analysis results
Table 4. Discriminant analysis results: Enter method
Table 5. Ideological polarization
Appendix I. Sample and weights
Appendix II. List of Variables

First Paragraph of Conclusion:
The exploratory analysis reported here reveals that legislators and party systems in Central America are ideologically structured around 'left' and 'right' concepts, contradicting an image of non-ideological politicians and parties, an image that formerly served as the basis for denying the validity of spatial analysis as an approach to studying this region. The analysis shows how issues of interest to legislators bundle together, generating different factors in every country, and how these factors cluster together and/or polarize legislators' opinions. Discriminant analyses reveal, however, that only some of these attitudinal factors divide the major parties' legislators in the various national systems, and that military and economic issues are by far the most important and consistent sources of party division in Central America. Thus, it seems that some attitudinal differences evident at the level of individual legislators are over-represented in some party systems - such as military issues in El Salvador - while other issues are under-represented, such as economic issues in Nicaragua and El Salvador. Furthermore, we have seen evidence of uneven levels of polarization in the Central American region, though it is generally greater than is found in the older democracies of western Europe. Finally, results suggest that there is no clear relationship between the substance and number of political dimensions and the ideological intensity of political competition. Future research should focus on these themes.