Best practices for data minimization

Use the questions in each section to:

• Identify your department’s current practices
• Assign roles and responsibilities
• Put data governance policies in place

Set up your data rules

Everyone in your department needs to know about data minimization and privacy principles. Develop and share policies and guidelines across your department. Train people on data minimization and privacy and escalation for data sharing requests.

These best practices make sure you protect data throughout its lifecycle.

• Put robust security measures in place. Update and audit security protocols to protect data from unauthorized access, disclosure, or misuse.
• Simplify and protect sensitive data. For example, split sensitive data from identifying information in storage. The Massive Data Institute has quick improvements for data privacy. Your department may already have rules for this. They are called the Secretary of State approved retention schedule.
• Establish clear, purpose-driven data retention periods for all data types. Regularly review data holdings and set up a monitoring system to find and get rid of data you don’t need.
• Delete, destroy, or anonymize data once you don’t need it and the retention period ends. Explore automated data deletion where you can.

Roles

• Do you have department-wide data minimization policies and guidelines in place?
  • If not, who is responsible for developing and sharing them?
• Do your department policies include clear protocols for escalating sensitive requests?
• How often do you review and update policies to stay current with changing laws and Information Practices Act (IPA) requirements?
• Do you train all staff on:
  • Data minimization?
  • Privacy responsibilities?
  • Complying with the IPA?
• Who is your:
  • Dataset owner(s) (someone in charge of maintaining a dataset)?
  • Executive officer (someone with the authority to approve and publish data published by your department, like a deputy director)?
  • Program manager (someone who works under an executive officer and oversees dataset owners)?
  • Do those roles have clear responsibilities for data minimization?

Data collection

There are 2 big things you can do to minimize the data you collect:

• Use checkboxes and dropdowns over free-text entry when you can. This collects the simplest data. For example, instead of asking for exact household income, ask people to choose from a set of ranges. Massive Data Institute has a list of strategies to proect your data like this.
• Ask as few questions that can directly identify people as possible. Two ways you can do this are anonymization and pseudonymization.

Consider testing your questions before you start. You can partner with community organizations and pilot with small groups to get feedback. Use your research questions to decide when to involve the community. It’s a good idea if you need a deep understanding of:

• People's experiences
• Their motivations

Use plain language to tell the people or groups whose data is being collected about:

• The department
• How you will use their data
• How you will protect their data
• Who else has access to the data
• If it’s mandatory or voluntary

Some people and groups are concerned about the federal government’s access to their personal data. If you collect data that might be shared with the federal government, consider telling people this.

Data retention and disposal

• Do you have an updated inventory of all your datasets?
• When do you review data to find and get rid of data you don’t need?
• Do you only collect the data relevant and necessary for the intended purpose?
• Does the data you collect have clear retention periods?
• How will you securely destroy or anonymize data once its retention period ends?

Security

• What security measures protect data from unauthorized access, disclosure, or misuse?
• Who has access to data?
• How often do you audit and update your security protocols to address new threats?