Reducing social inequality: Why the scope of measures is crucial

In modern social research, sociological questions are increasingly being answered with the help of experiments; for example, whether employers discriminate in personnel selection, whether immigrants are treated less well in social situations, or whether counseling programs help to overcome educational inequalities.

However, a new study by Dr. Irena Pietrzyk and Professor Dr. Marita Jacob at the University of Cologne Department of Sociology and Social Psychology shows that findings relating to social reality can be misleading. In addition, the authors have developed an online tool that clearly shows the extent to which social inequalities between groups can, in fact, decrease—depending on how many people actually have access to a measure.

The study “Why Treatment Prevalence Matters: Overcoming a Blind Spot in Experimental Inequality Research ” is published in the special issue titled “Erklärung und Kausalität in der Soziologie (“Explanation and Causality in Sociology”) of KZfSS Kölner Zeitschrift für Soziologie und Sozialpsychologie.

Experiments are well suited to measuring whether a measure or mechanism has a causal effect. However, experiments do not generally record how many people from different social groups are exposed to such a measure in everyday life. The researchers refer to this factor as treatment prevalence.

“Whether a support program reduces inequality depends not only on whether it works, but also on who takes part in it,” says Jacob. “This has hardly been taken into account in experimental studies to date.”

The researchers illustrate their findings using three recent experimental studies. Two of them derive from international research literature, one from their own work.

A laboratory experiment conducted in Italy showed that people who were not from a migrant background did not tend to trust immigrants less than others in a game involving trust. Nevertheless, immigrants might benefit less from pro-social behavior in everyday life at the group level because they might play a less active role within associations and civil society structures, and therefore benefit less from the willingness to help shown to people who engage in voluntary work.

An experimental study with fictitious applicant profiles in Germany and Italy showed that female applicants for professorships were not rated worse than their male colleagues. However, because men in science are more likely to be first authors of publications than women, structural inequality might still arise at the group level—with the result that women would be appointed less frequently as professors.

In their own research, Pietrzyk and Jacob have carried out a large field study in North Rhine-Westphalia, which has shown that intensive guidance counseling increases the enrollment rate among pupils from less privileged families. However, even an effective program could exacerbate inequalities if it also encourages privileged pupils to study, and—in practice—primarily benefits these pupils instead. Conversely, a program could realize its full potential if it specifically benefits those most in need of support when it comes to accessing higher education.

To help researchers and practitioners systematically analyze these relationships, Pietrzyk and Jacob have developed an interactive, freely accessible visualization tool. It allows for a simulation of how different access patterns influence social inequalities.

“The message for practical application is clear,” summarizes Pietrzyk. “Anyone who wants to seriously combat inequalities must not only ask whether a measure works, but also ensure that it is benefiting the right people.”