When the Data Runs Out

Why Science Needs Judgment

Over the past century, our understanding of nutrition has evolved dramatically. Ideas once considered settled have been revised, replaced, or abandoned as new evidence emerged. After more than forty years in the food industry, I have had a front-row seat to that evolution.

The clearest example is fat. When I was growing up, fat was the enemy, full stop. Everyone, including scientists and policymakers, believed it. That belief sent the entire food industry on a mission to remove fat from food. The problem was that food without fat is rarely very appealing, so manufacturers replaced it with sugar and refined carbohydrates.

The result helped fuel an epidemic of obesity.

Thankfully, today the conversation has changed, though not entirely. Many people still believe that eating fat makes you fat.

Then looking back, the fear of eggs almost feels like comic relief. For decades, eggs were treated as a cholesterol bomb. People were warned to limit them or avoid them altogether. And suddenly, almost overnight, the guidance changed. Eggs were quietly removed from the list of nutrients of concern. No apology. No asterisk. Not even, “Never mind. We were wrong.”

The most important example, however, may be the one I watched unfold throughout my entire career: the promotion of ultra-processed food.

For years we obsessed over nutrients: fat grams, calories, vitamins. We measured everything measurable but missed the thing that turned out to matter most. The real issue was not simply what was in the food. It was what had been done to it.

My grandmother understood something simple: food should eventually rot. Most people do not like things that rot, so we learned to prefer food that could sit on a shelf for months (or even years), held together by ingredients with names most of us cannot pronounce. Today, roughly seventy percent of what Americans eat comes from packaged and ultra-processed foods.

None of this means scientists acted in bad faith. They were trying to solve real problems. Food became more abundant, more affordable, and more convenient. Those were meaningful achievements. The point is not that science failed. The point is that reality proved more complicated than the story we told about it.

There is an old saying that for every complex problem there is a simple solution, and it is always wrong. We forgot that lesson. We focused on one variable and confused a useful insight with the whole truth.

What should these reversals have taught us?

Not that science is useless. In many ways, it did exactly what it is supposed to do. It corrected itself.

The lesson is humility.

The Empty Room

We hear it constantly now: “The science says...” It has become the final word, a way to end a conversation rather than inform one. The more interesting question is why.

Part of the answer is that we have dismantled many of the things that once guided us. For most of human history, people had religion, tradition, elders, and rituals. They were imperfect, but they offered some sense of what you should do and what you ought to do. There was comfort in that. We are still looking for that kind of comfort. In some ways, “the science says” has begun to play the same psychological role that “the scripture says” once did. We are searching for an authority that relieves us of the burden of judgment.

We also live increasingly apart from one another. As former U.S. Senator Ben Sasse has observed, we cluster by age. Many people rarely see their parents, let alone their elders. So when we face a hard question, there is often no one older and wiser in the room to say, “I’ve seen this before. Take it easy. Eat an apple.”

Instead, we turn to our data, our metrics, and our smartwatches. The pull is understandable. Ambiguity is exhausting, certainty is comforting, and our information ecosystem rewards certainty. A headline that says, “It’s complicated,” does not travel.

Science walked into an empty room. We left it empty, and it filled the space.

Science and Scientism

None of this is a critique of science. Quite the opposite. At its best, science is one of the most remarkable things our species has ever created.

Science is a method. It asks questions, tests assumptions, gathers evidence, and remains open to correction. Real science knows it might be wrong tomorrow. That humility is the engine.

Scientism is something different. It is the belief that science can answer questions it was never designed to answer, questions of meaning, value, and how we ought to live. It mistakes facts for commandments, confusing how the world works with what we should do about it.

That confusion has consequences. We increasingly hear “the science says” not as an invitation to think but as a way to end the conversation. Wrap a position in the authority of science, and anyone who disagrees is no longer seen as holding different priorities or values. They become anti-science. We were never meant to use a lab coat as a weapon.

The irony is that this posture is profoundly unscientific. Science advances by continual revision. Treating today’s findings as the final word gets the spirit of science exactly backward.

Most of our hardest disagreements are not scientific anyway. They are about what we value, what trade-offs we are willing to make, and what kind of life we want to build. Science can inform those decisions, but it cannot make them for us. It can tell you the caloric content of a meal, but not whether the meal was worth sharing. It can tell you how to extend a life, but not what a life is for.

Science is indispensable. But beyond the edge of measurement, judgment has to carry the rest.

What the Studies Can’t Tell You

Aristotle drew a distinction worth recovering. On one side was technē, technical knowledge. How to perform surgery, build a bridge, or run the regression. On the other was phronesis, practical wisdom. Knowing what to do, for whom, when, and to what end.

We are very good at training the first. We are much less intentional about training the second.

Consider the best doctor you have ever known. The techniques she possesses could be taught to ten thousand other doctors. What makes her exceptional is judgment. She knows the patient, the family, and the moment. She knows when to treat aggressively and when to leave well enough alone. No study can tell her that.

The same is true in business. Consultants can analyze the spreadsheets, but the real decision lives somewhere the numbers cannot reach. Judgment is knowing the right timing, the right person, and how to move people toward the hard thing they would rather avoid.

Skill and judgment are not opposites. The carpenter’s wisdom is built on knowing how to work the wood. Real judgment rests on real competence.

So take expertise as far as it honestly goes. Then understand that at the edge of what can be known, the data runs out and the human part begins.

That edge is where judgment lives.

The Well-Informed Spectator

So what happens when we start handing our judgment to the model, the expert, or the algorithm? It atrophies. Judgment is a muscle. Use it and it grows stronger. Neglect it and it weakens. Every time you let something else make a decision you should have made yourself, you surrender a little of your capacity to decide.

At the leadership level the danger is worse, because data can disguise itself as rigor you do not actually have. I have watched leaders become so surrounded by dashboards and consultants that they can no longer decide anything. They can optimize everything and resolve nothing.

A leader who outsources judgment is not really a leader anymore. He is a very well-informed spectator, sitting at home analyzing a game he is supposed to be playing.

And there is a moral cost, which is the real one. When you outsource judgment, you outsource responsibility. “I was only following the science.” “The model told us to.” Nobody owns the decision, so nobody is accountable. For a person, and for a society, that is corrosive.

Trust Without Deference

The hardest question in all of this is how we rebuild trust in expertise without rebuilding blind deference to it. We need experts. The goal is not to tear them down. It is to restore a healthier relationship with them, and that requires something from both sides.

Experts need humility. The scientist who says, “Here is what we know, here is what we don’t, and here is the precise point where my expertise ends and a value judgment begins,” is someone I will trust more, not less. For a while we got this backwards. We confused confidence with credibility. They are not the same thing.

COVID exposed this distinction clearly. Scientists were exactly the right people to answer scientific questions such as how the virus spread. But once the conversation turned to whether schools should close, or how to weigh transmission risk against childhood development, mental health, education, and livelihoods, we had moved beyond science into judgment. Those were questions about competing goods and difficult trade-offs. Expertise could inform those decisions. It could not make them for us.

The public has responsibilities too. I can trust my doctor deeply while still asking hard questions and owning the final decision about my own body. Trust paired with agency, not obedience.

The inconvenient truth is that some of the lost faith in science is self-inflicted. For years we treated experts as substitutes for our own judgment rather than partners in it. Doing the work does not mean distrusting science. It means learning how to think about evidence. Distrust any single study and trust the broader body of literature on the topic instead. A single study is one witness. A meta-analysis is the whole courtroom. And remember, when someone tells you to trust the data, what they are really asking you to trust is their interpretation of the data. We all carry interpretive filters, noticing the evidence that confirms what we already suspect and overlooking the rest.

Some questions, of course, require genuine expertise. Most of us cannot independently master climate science or vaccine research. I want to live in a society with high trust in genuine expertise. But trust is healthiest when it is paired with humility on both sides: experts who recognize the limits of their knowledge, and citizens who remain engaged in the decisions that belong to them.

Judgment Is Built, Not Born

So here are a few practices that actually train judgment.

Keep a decision journal. Before you know how things turn out, write down why you are making a decision. Then go back later and ask what actually happened. Nothing trains judgment quite like watching your own reasoning meet reality. Before an important decision, do the opposite as well. Imagine it has failed. What went wrong? Looking backward after a decision and forward before one trains the same muscle: better judgment.

Learn to make the strongest version of the argument you disagree with. When a disagreement starts heating toward a fight, try to explain the other person’s position well enough that they say, “Yes, that’s exactly what I mean.” Only then should you argue your own case. It is surprisingly difficult because it forces you to listen rather than prepare your rebuttal. I used this constantly with my kids. I also ask AI to do the same by challenging my conclusions rather than confirming them. You do not have to agree with another view, but you should understand it well enough to explain why thoughtful people hold it.

Another practice is to read old books. History is an inoculation against the certainty of the present moment. Every age was sure of things we now find absurd, and remembering that should make us appropriately suspicious of our own certainties.

Finally, learn to sit in ambiguity. Resist the urge to resolve every question immediately. As Tyler Cowen puts it, hold ideas unsettled. The ability to live with that discomfort is much of what judgment actually is.

Because in the end, that is the whole argument. Science can tell us a great deal about the world. It cannot tell us how to live in it. That part was never going to be outsourced. It is ours to carry, and the carrying is the work of a lifetime.

Your Turn

What is one question in your life that cannot be answered by more data alone?

And where are you waiting for certainty when the situation actually calls for judgment?

I also created a song called That's Where Judgment Begins to accompany this essay. It explores the same idea, just with a little more groove: that science can illuminate the path, but judgment is what carries us the rest of the way.