| Popis: |
In three earlier controlled experiments that elicit ethnic discrimination among individuals who belong to a majority group (which we refer to as NL2020, UK1920 and NL1415), we found that ethnic salience induced them to seemingly act more pro-socially towards interaction partners who belonged to an ethnic minority group. Ethnic salience refers to the salience of issues related to ethnicity and discrimination to the public or to the salience of other people’s ethnicity. NL2020 and UK1920 are pre-registered at OSF (see https://osf.io/yd82n/ and https://osf.io/syp83, respectively) and NL1415 includes a wave reported in Cettolin and Suetens (2019) and a wave ran two months after the initial wave. In NL2020 the variation in ethnic salience was induced by the researchers giving part of the decision-makers additional information related to the ethnic background of the matched partner (i.e. the birth place of the parents). The additional information made decision-makers matched to a minority participant more aware that the experiment had something to do with behavior towards ethnic minorities, immigrants with non-Western origin or discrimination, as was revealed when asked about the purpose of the research in a post-experimental questionnaire. In NL1415 the variation in ethnic salience was in fact a natural variation related to the timing of the experiment: one wave was run right before the terrorist attack on the Charlie Hebdo office in Paris in January 2015 and another wave (with other participants) was run right after, thus giving a low and high ethnic salience condition, respectively. In UK1920 the variation also followed from the timing, with one wave ran at the time that the #Black Lives Matter movement spread globally (high salience condition), and an earlier wave ran about half a year before (low salience condition). We back up our ethnic salience assumptions with a time trend of the intensity of Google searches on topics related to ethnicity and discrimination in the Netherlands and the UK, following the suggestion of Epstein and Segal (2000) to use appearance in the media as a proxy for issue salience. We explored the mechanisms behind the result by merging the experimental data with individual survey data collected in post-experimental or independent surveys and by studying patterns in voting preferences in the Netherlands and the UK. Doing so led us to hypothesize that the results are due to social desirability bias: ethnic salience has made it more likely that decision-makers matched to a minority interaction partner anticipated that the research study was about ethnic discrimination, which induced them to act more pro-socially towards a minority partner than they otherwise would. Social desirability bias is a well-known bias in experiments and surveys; it refers to participants responding or behaving in a more socially acceptable way than they would behave otherwise, for example, because they are concerned about their social image or reputation towards the researchers (Levitt and List, 2007). In the current study we further investigate the social desirability hypothesis by testing whether being aware that the research topic involves ethnic discrimination has an effect on choices of majority decision-makers that affect a minority partner. To do so, we compare choices that affect a minority partner (with a Turkish or Moroccan immigration background) made by majority decision makers in the experiment pre-registered at https://osf.io/s2g6n, in which decision-makers are informed about the research topic with those in a treatment in which decision-makers are informed. In the latter treatment, they are shown the following information on the consent screen: “The study is about discrimination. It is being studied whether people with a Turkish or Moroccan background are treated differently than people with a Dutch background.” Both treatments are run in the same experimental wave. |
