Empirical Analysis of Intersectionality

Dubrow, Joshua Kjerulf.  2008.  “How Can We Account for Intersectionality in Quantitative Analysis of Survey Data?  Empirical Illustration of Central and Eastern Europe.”  ASK: Society, Research, Methods 17: 85-102.

Data come from European Social Survey, ESS, Round Three (2006). ESS is a cross-sectional, cross-national dataset with individuals as the units of analysis.  I focus on disadvantaged groups of new democracies in Central and Eastern Europe: Bulgaria, Estonia, Hungary, Poland, Slovakia, Slovenia and Ukraine3. Dubrow et al (2008) explored soft political protest among the democracies of Europe and found that disadvantaged groups in new democracies engage less in soft political protest than those in old democracies.  The effect of disadvantaged intersection is most likely to be found and have the greatest magnitude in the new democracies of Europe.

The dependent variable is engagement in soft political protest.  Soft political protest is similar to “conventional” protest, which includes legal demonstration and signing petitions (Jenkins and Form 2005). Added here is contacting a politician, government or local government official because soft political protest may be exercised in alternative forms in various countries. In particular, in one country signing a petition can be treated as the functional equivalent of contacting a politician or official in another country. This seems to be especially relevant in countries with weak petition-signing culture, as is the case in new democracies (see Inglehart and Catterberg 2002). 

I classify a case as representing soft political protest if the respondent pro¬vided a positive answer to at least one of the items: contact official, sign petition, and attend lawful public demonstration.  Idiosyncratic national patterns suggest that the three indicators should be treated as alternative expressions of soft political protest rather than cumulative ones—that is, those measuring the intensity of the underlying common phenomenon.  Thus, for each country included in the ESS wave I created a dichotomy, dividing all respondents into those who engaged in any of three forms (denoted 1) and the rest (denoted 0). 

As independent variables, I use dichotomous variables with disadvantaged groups as the focal category, and non-disadvantaged groups as the reference category.  Specifically:

Gender is coded with woman = 1, man = 0.

Ethnicity is constructed from the combination of (a) respondents answering “yes” to the question, Do you belong to a minority ethnic group in [respondent’s country]?  and (b) positive answers to the question, On what grounds is your group discriminated against? in terms of at least one of the following:  the color or race, nationality, language and ethnic group. Thus, ethnicity is coded with self-report of minority status and/or discrimination based on ethnicity = 1, otherwise = 0. 

I constructed disadvantaged class by dividing the lower end of the EGP class schema (Erikson and Goldthorpe 1992) from the rest, where unskilled workers, agricultural laborers, and self-employed farmers = 1, otherwise = 0 (including those without an occupational code).  In new democracies of Central and Eastern Europe, as elsewhere, members of these class categories are marked as disadvantaged as they typically have the least access to and amount of valued resources, such as income, education and status (Heyns 2005).  These class members tend not to do participate in political actions (Gallego 2008).