Setting boundaries around the use of data – be that ensuring that data is disaggregated or safeguarding measures are in place for the collection of data see the section on identifying data gaps – ensures that the data can be fully used for society’s benefit. Below you will find information on the risks of data use, some best practices in data ethics and an introduction into anti-discrimination in data.

Ensure that data does not exclude or discriminate

The overriding human rights principle “do no harm” should always be respected when collecting, storing and using data. Historically, there have been cases of misuse of data collected by NSOs and others, with detrimental human rights impacts. For this reason, data collection exercises, whether through census, specialized population surveys or administrative records, must not create or reinforce existing discrimination, bias or stereotypes exercised against population groups. As a policymaker, ensuring that the data you are using does not discriminate against already marginalized groups is a critical element of data ethics.

Follow a human rights-based approach to data

The United Nations Human Rights Commission (OHCHR) published a guidance note on the use of data and statistics consistent with international human rights norms and principles. It sets out a Human Rights-Based Approach to Data (HRBAD) with a focus on data collection and data disaggregation that are formulated under six headings.

‍

Protecting human rights under HRBAD means it is necessary for policymakers to have safeguarding measures in place when collecting data. In national statistics systems, there are often gaps in availability of data on vulnerable populations and there may also be unintended discrimination. For example, women and girls and their living conditions are either underrepresented or not prioritized in data sets, leading to policy designs that do not address their challenges and need to be addressed.

A rights-based approach to data goes beyond the ethical and representative collection of data. It also considers the ethical use of data. The principle of “do no harm” means that policymakers should not endorse or encourage any initiative or practice that risks exposing groups to serious human rights violations. The rights of ethnic minorities, people living in rural areas and non-citizens are not always considered by policymakers. At times, it may be appropriate to involve civil society organizations and human rights institutions to represent groups at risk. Policymakers thus need to take reasonable steps to ensure data is used for the benefit of identified populations and a human rights-based approach to data can help facilitate this.

SPOTLIGHT: Gender Data – How to make sure everyone is counted

The systematic integration of gender perspectives in regular statistical programmes is still missing in many countries. Making sure that gender data is considered in your work is highly important to ensure that all people, no matter their gender, are counted in data. Below are a few resources to ensure any statistics or data you are using are as anti-discriminatory as possible:

Using Gender Statistics from UNECE is a toolkit that trains data users on understanding and proper usage of gender statistics. The toolkit includes short user-friendly descriptions of concepts with practical examples and exercises for use in training sessions.

Integrating Gender Perspectives into Statistics is a manual from the UN Statistics Division that aims to support achieving comprehensive coverage of gender issues in data production activities, incorporating a gender perspective into the design of surveys or censuses and improving data analysis and data presentation to deliver gender statistics in a format that is easy to use by policymakers and planners.  

European Institute for Gender Equality Gender Statistics Database is a guide explaining how to integrate gender into data collection, analysis, and dissemination. It includes recommendations for collecting data on a range of gender-related issues, such as unpaid care work, women's participation in decision-making and gender-based violence.

World Health Organization Gender and Health Equity Assessment Tool is a guide developed by the WHO with a special focus on the health sector and provides guidance on how to assess the gender (and health) equity impacts of policies and programs. It includes recommendations on how to collect and analyze gender-disaggregated data to identify gender disparities in health outcomes. This tool can be used as a reference framework and adapted accordingly to different sectors.

Best practices in data ethics

This section details two practical frameworks to help you and your organization ensure that data ethics are being respected. The first framework from the OECD can be helpful to understand the various interlinkages in data ethics. The second framework from the ODI is from a larger tool, which consists of the canvas and a user manual, which are offered online for free, and an accompanying live training. 

OECD’s Good Practices in Data Ethics

The first framework is the OECD’s Good Practice Principles for Data Ethics, introducing ten guidelines to facilitate your data ethics processes in the public sector:

Open Data Institute’s (ODI) Data Ethics Canvas

The second framework is the Open Data Institute’s (ODI) Data Ethics Canvas. The Canvas is a tool for anyone who collects, shares, or uses data. It helps identify and manage ethical issues during the design, development, delivery and review of the project or product based on data.

In this tool, users can review each bubble to reflect on specific questions, access further guidance and resources, and write down both the current situation and potential next steps for each area. This tool actively supports you in developing ethical guidance for your context. Additionally, ODI provides a helpful tool for organizations to benchmark their maturity in relation to data ethics and develop action plans to improve their practices.

Investing the time and resources in ensuring that you and your team are informed about the risks of data and understands data ethics and that your processes consider safeguards at every stage is critical for protecting the rights of data subjects.