If You Care About What Works, You Believe in Critical Race Theory
Understanding patterns, power, and context isn’t ideology, it’s evidence.
What evaluation teaches us about why programs work for some and not for others.
The ultimate question evaluators are asked is simple: Is it working?
But that question only makes sense if we understand for whom and why.
Consider a school-based nutrition program in Indonesia that provided healthy lunches and nutrition education to children. When researchers looked at the overall results, there was little measurable change in students’ health. Funders began to doubt its value. But when they broke the data down, a different story emerged.
Among a subgroup of underweight and anemic students, the effects were dramatic: anemia rates dropped by half, hemoglobin levels rose, and children’s diets improved. For them, the program worked. But why? Because it addressed the specific needs of children who had the least access to nutritious food and targeted the behaviors most relevant to their reality. In aggregate, the change looked small; in context, it was transformative (Andarwulan et al., 2023).
Now look at an education initiative in rural Afghanistan that built community-based schools for girls. At first glance, the data looked underwhelming. Across all students, there was little measurable improvement. But when evaluators disaggregated the data, they found that the gains were concentrated among girls in remote villages with no existing schools nearby. For them, enrollment and literacy rates rose sharply. For those already within reach of public schools, the change was minimal.
The difference was context. For girls who faced long travel distances and safety concerns, the program wasn’t just effective it was revolutionary. It worked because it was built around local norms, relationships, and constraints (Burde et al., 2021).
Finally, consider school voucher programs in the United States. Many appear successful on paper: families report higher satisfaction, and some students show improved academic outcomes. But when researchers examine the counterfactual—what would have happened without the vouchers—the picture shifts. The benefits are concentrated among students who were already performing well or whose families could navigate the system. For lower-income or marginalized students, outcomes are flat or negative (National Bureau of Economic Research, 2022; Brookings Institution, 2023).
A program can look ineffective overall yet have a major effect on a subgroup. Or look effective overall while deepening inequities beneath the surface. Averages hide both truths.
That’s what happens when we ignore culture and context. One program fails because it didn’t fit how people actually live and decide. Another “succeeds” because it measures the wrong thing and mistakes correlation for impact.
If we are blind to the differences among our populations, we have no chance of answering why a program worked or didn’t work. Disaggregation helps us tell those stories honestly. When we break the data down by gender, age, race, income, or geography we see that outcomes are rarely uniform. The real learning happens in those differences: who benefited, who didn’t, and why.
Disaggregation isn’t about politics; it’s about precision. If we can’t see which groups are being reached and which are being missed, we can’t make programs more effective, or spend money wisely.
Real effectiveness demands that level of honesty. We can’t fix what we can’t see, and averages hide as much as they reveal.
Looking for Patterns
When we start asking why a program worked or didn’t, the answers rarely stay contained within that single program.
You disaggregate the data, and new shapes begin to emerge. You do it again on the next project, and the same shapes appear. The people least reached by one intervention are often the same people least reached by others across sectors, across geographies, across time.
At first, it feels like coincidence. But as the patterns repeat, it becomes impossible to ignore. The gaps aren’t random.
This is where evaluation starts to teach a harder truth: effectiveness isn’t just about doing more, it’s about understanding why some groups consistently benefit less, or not at all.
Recognizing patterns doesn’t require a political stance. It requires curiosity. And curiosity is where all good evaluation begins.
Patterns Have Histories
Once you start seeing patterns, curiosity turns into harder questions. Why do the same groups keep falling through the cracks? Why do some programs flourish in certain communities but struggle in others, even with identical designs?
Those answers rarely live in the data alone. They live in context, in history, trust, geography, and lived experience.
Maybe a community has seen dozens of short-term projects come and go, each promising change but leaving little behind. Maybe a policy once designed to help families now discourages participation through layers of paperwork or stigma. Maybe the language of “support” carries memories of control.
These histories shape how people engage with programs long before we arrive with a survey or a spreadsheet. Ignoring them doesn’t make them go away; it just blinds us to why outcomes differ.
Understanding effectiveness means understanding context. The data tells us what is happening; culture and history tell us why. And if we want programs that truly work, we have to be willing to see both.
What We Can’t Pretend Not to See
When we trace patterns across programs, we start to see that the differences among subpopulations aren’t random, they’re reflections of something deeper.
The success or failure of a program rarely depends only on its design. It depends on the culture it enters, the power structures it interacts with, the biases it carries, the strengths it builds on, and the relationships it either honors or ignores. These aren’t side notes; they’re the conditions that determine whether change takes hold.
That’s why the idea of being “color-blind,” or pretending not to see difference, simply doesn’t hold up. Difference is real. Cultures are real. History is real. Power is real. To claim not to see them is to willfully look away.
Ignoring these realities doesn’t make programs fairer, it makes them ineffective. It’s like trying to understand a forest by studying only the tallest trees. You might think you’re seeing the whole picture, but you’re missing the roots, the shade, the soil, the context that makes the system work or fail.
If we say we care about whether something is working, then we have to care about the conditions that make it possible for some people and not for others. We have to look at how history shapes trust, how hierarchy shapes access, and how bias shapes opportunity.
We can’t pretend we’re evaluating impact if we’re blind to the forces that create uneven impact in the first place.
What the Backlash Misses
In recent years, terms like Critical Race Theory and intersectionality have drawn criticism and confusion. Much of that debate overlooks what these ideas actually represent. At their core, they are frameworks for understanding how history, culture, and power continue to shape present-day outcomes, and they help us understand relationships of power today.
Critical Race Theory examines how laws, institutions, and social norms were built within specific historical contexts and how those legacies persist. It does not assign moral blame; it traces cause and effect. In that sense, it mirrors evaluation itself—both seek to understand how prior decisions produce current results.
Intersectionality extends that lens. It recognizes that people experience the world through multiple, overlapping identities like race, gender, class, age, ability, and that these intersections shape how opportunity and constraint are felt. A policy can appear effective overall yet still exclude or disadvantage certain groups; intersectionality helps make that visible.
These frameworks are not political platforms but analytical tools. They allow us to see difference clearly rather than interpret it as error or deviation. When we understand how culture, history, and power influence outcomes, we can design programs that respond to reality rather than to assumption.
Ignoring these dynamics does not make systems more equal or programs more fair. It simply leaves us blind to why results differ and what would be required to make them better.
Seeing Clearly
Effectiveness isn’t only about effort or intention. It’s about understanding the conditions that make change possible, and the barriers that quietly hold it back.
Culture, history, and power are not distractions from the work of evaluation or policymaking; they are the work. They explain why outcomes differ, why trust is uneven, why programs succeed in some places and stall in others.
If we care about whether something is working, then we have to care about why. And that means being willing to see the patterns that shape results, even when they are uncomfortable to name.
In the end, evaluation and learning both depend on the same discipline: the willingness to look closely at what is real, not just at what we hoped to find. That is how evidence becomes understanding, and how understanding becomes change.
Anthralytic is built on the idea that clarity creates impact. We use evidence, context, and human insight to help organizations understand not just whether something works, but why.
