The Risk of Becoming Data-Blind in the Name of Becoming Colorblind
What happens when a country stops collecting the data it needs to understand itself?
The Trump administration’s proposal to remove race and ethnicity questions from the 2030 Census raises an issue that goes far beyond a few questions on a government form.
It raises a much bigger question:
Can we understand inequality, opportunity, representation, health, economic conditions and demographic change if we stop measuring the differences that help reveal them?
There is an understandable argument for becoming a more colorblind society.
But being colorblind and being data-blind are not the same thing.
And this is not simply an issue affecting minority populations.
Reliable race and ethnicity data can also help us identify important issues affecting White Americans — including differences in opioid deaths, suicide, economic decline, aging communities, education, employment, health and other areas.
Without the data, those problems do not disappear.
They simply become harder to see.
Our new Preference Analytics white paper explores:
• What would actually change if race and ethnicity were removed from the decennial Census
• What would still be available through the American Community Survey and other data sources
• Implications for voting rights, civil rights, healthcare, education and public policy
• Potential effects on businesses, market research and demographic targeting
• Why the loss of these data can affect White Americans as well as minority populations
• What administrative records and AI might — and might not — be able to replace
• Seven possible scenarios for what could happen between now and the 2030 Census
The broader point is simple:
The goal should not be to categorize people more.
The goal should be to understand people better.
You cannot determine whether differences are disappearing if you stop measuring the differences.
And you cannot solve problems you can no longer see.
A note on how this paper was created
The idea, framing, perspective and overall direction of the paper came from Preference Analytics. We used OpenAI’s GPT-5.6 Sol with High reasoning to support the deep research, synthesis and writing.
The final interpretation, conclusions and perspective are ours.
The Risk of Becoming Data-Blind in the Name of Becoming Colorblind
Comments