IV. >> Illustrating the Framework - Indicators for Some Rights >> Selecting the indicators Box 23 Measuring discrimination Measuring discrimination is not easy. A different treatment or outcome is not necessarily the result of clearly identified acts of discrimination, but the result of complex processes involving multiple and cumulative discrimination, or simply due to other factors. Moreover, victims are sometimes unable to identify the discrimination that they are subjected to. Certain social and cultural practices create high tolerance levels for discrimination among certain population groups, which results in the acts of discrimination being frequently overlooked. Also, they are often unaware of the available legal remedies or unable to use them. Thus, the number of convictions for discrimination in court is not a good indicator for assessing discrimination in a country. Given these limitations in using the events-based information in monitoring discrimination, statistical techniques, as well as direct surveys, are vital for assessing the prevalence of discriminatory practices in a country. Some useful statistical tools in this context are:  ocioeconomic statistics disaggregated by prohibited grounds of discrimination (e.g., life expectancy, S age-specific sex ratios and unemployment rates broken down by ethnic origin) measure disparities and differential outcomes that are often the result of multiple and accumulative discrimination; E conometric models based on multiple regression analysis help in estimating the portion of differences in outcomes attributable to discrimination as opposed to observable variables (e.g., percentage of the wage differential between women and men that cannot be explained by “observable” criteria, such as the number of working hours or socioprofessional characteristics, etc.); Population surveys measuring experiences, perceptions and attitudes regarding discrimination (e.g., percentage of members of ethnic minorities reporting racially motivated victimization and discrimination by public/private personnel);a and  iscrimination or situation-testing surveys to measure directly discrimination in specific instances, such D as those related to access to work, housing, health care, private educational institutions or other public services (see box 24). It may be desirable to use any of these procedures to assess periodically the extent of discrimination in a country, especially where multicultural, racial, religious and linguistic communities are seen to be competing for scarce resources and opportunities. Concrete evidence in support of discriminatory practices in different social spaces of human engagement, including the political space, could facilitate the strengthening of legal and administrative remedial measures in such instances. a. See, for instance, European Union Agency for Fundamental Rights, EU-MIDIS: European Union Minorities and Discrimination Survey (2009). Available from www.fra.europa.eu/fraWebsite/minorities/minorities_en.htm. 82 Human Rights Indicators