equality data
Equality data
Definition of equality data
Equality data are the quantification of discrimination and inequality in society through structural processes of data collection and analysis. The aim behind equality data is combatting exclusion and maltreatment of people by clearly documenting their situation as targets of racism, sexism or religious discrimination. Data on racism, discrimination, sexism and ethnic minorities on a (supra)national level are invaluable bases for the design of adequate policies and measures as well as for the raising of societal consciousness concerning the concrete reality of discrimination. Examples of equality data are the collection of how many people of the LGBTIQ+-community experience discrimination, how often and in what form. The relation with the concept of discrimination is a quite concrete one, because equality data are factual representations of measures of inequalities that obtain through discrimination of society. A related but different concept is that of equity data. Rather than focusing on creating equal situations, equity data has to do with the enforcement of justice with regard to the data access of marginalized communities and their empowerment by use of data against domination.
- #comment/Anderson: This is a rather subtle distinction; could be elaborated: equity data
Implications of commitment to equality data
⇒ Pursuing equality data means being serious about mapping discrimination in society on a structural and professional basis to expose how bad the inequalities of society actually are. Following the sad fact that people’s testimonies of lived experience are not enough for a collective impetus towards tougher anti-discrimination policies or more vocal antiracism, equality data are posed as unreputable leverage for changing society for the better. The consequence of equality data for AI is that the data that serve as input to machine learning processes or other types of algorithms need to be handled by curators, proving it to be equality data, i.e. either conscious of the discrimination of society or with discriminating aspects purged from it. Harshly constraining what one inputs into algorithms is a different way to debias algorithmic outcomes and decisions. There is a concern with the collection of equality data, however. It is difficult to collect data about ethnicity without enabling actors to use them for racial profiling. Part of the challenge is collecting the data necessary for proving the harsh reality of discrimination without treading outside of necessary data protection safeguards and ensuring they are used for the benefit of discriminated groups.
⇒ debiasing
Societal transformations required for addressing concerns raised by equality data
What cultural, educational, institutional, or societal changes are needed to address concerns related to this concept?
To follow through on the promise of equality data, collective change on two fronts I required: a legal front and a cultural front. On the one hand, the national and supranational legislations need to be changed to accommodate for the ethical collection of the sensitive information categories that constitute equality data. On the other, a sphere of trust will have to grow in society that local, national and European governmental organisations will safeguard the data and use it for just causes. Currently, the distrust of governmental organisations and fellow citizens’ intentions is abundant and present, while minorities do have the insight that equality data would help them. Compliance with the collection of sensitive equality data thus faces cultural and legal obstacles.
Sources:
- European Network Against Racism. “Data Equality.” https://www.enar-eu.org/about/equality-data/
- Joyce Lee Ibarra. “Data Equity: What Is It, and Why Does It Matter?” https://www.jliconsultinghawaii.com/blog/2020/7/10/data-equity-what-is-it-and-why-does-it-matter.)