Parallel Census
Official statistics erase you. Count yourselves. Know your own numbers.
Pressure addressed: displacement, psychological
Time: Months to design and conduct, then periodic updates
People: Community researchers, participants
Resources: Survey design, data collection capacity, analysis skills
About this tool
Official statistics are political. Who gets counted, how categories are defined, what questions get asked, all reflect the priorities of those doing the counting. Communities are routinely undercounted, miscategorised, or rendered invisible. A parallel census counts your community on your own terms.
How to do it
Define what to measure
What information does your community need that official sources don't provide? Be specific about purpose.
Design carefully
Questions that make sense to your community. Categories that reflect how you understand yourselves. Tested before deployment.
Build trust
People share sensitive information only with trusted collectors. Community members conducting census, not outsiders.
Collect systematically
Coverage matters. Reaching the hard-to-reach. Multiple methods if needed.
Analyse and share
Data sitting unused helps no one. Analysis, reports, advocacy applications. Give back to those who participated.
Before you start
- Research capacity
- Community trust
- Clear purpose for data
Implementation guidance
Starting point
- Rigorous methodology producing defensible data
- High participation rate from target community
- Data used for advocacy and service planning
- Community ownership of data and findings
- Regular updates tracking change over time
- Security protocols protecting participant information
Building on it
- Basic count completed
- Reasonable participation
- Some use of findings
Advanced practice
- Methodologically weak data that doesn't withstand scrutiny
- Data collected but not used
- Sensitive information inadequately protected
Weak methodology produces data that gets discredited, undermining future community research credibility. Collected data sitting unused wastes participant trust. Inadequately protected sensitive information about community members can be exploited by opponents.
Examples
Various immigrant communities
Community-led counts documenting populations undercounted in official census
Evidence for service provision and political representation