I am an architect. I build things. And the first thing any architect learns is this: measure the foundation before you draw the building.

If the foundation is not where you think it is, the building falls. Not because the design was bad. Not because the materials were wrong. Because someone assumed instead of measured. Every collapsed structure in history has the same epitaph: they did not check.

Now let me tell you about a foundation that an entire nation has been building on without checking.

The Question

What percentage of the US federal budget goes to foreign aid?

(A) About 15-25% (B) Less than 1%

If you said A, you are in the company of most Americans. Polls consistently show that the average citizen estimates foreign aid at about twenty-five percent of the federal budget. Some say fifteen. Some say thirty. The median guess is somewhere around twenty to twenty-five percent.

The Answer Is B

Less than one percent. In recent years, approximately 0.7 to 1 percent of the federal budget, depending on what you count and how you count it.

The average citizen is off by a factor of twenty-five.

Let me say that differently, because the architecture of this error matters: the gap between perception and reality on this question is not a rounding error. It is not a matter of interpretation. It is a structural failure in the information environment so large that the building cannot possibly stand on it.

Why This Is an Architect's Nightmare

I spent my life designing structures that do more with less. The geodesic dome. The Dymaxion house. The tensegrity sphere. Every one of them begins with the same discipline: know your materials. Know your loads. Know your numbers. If you do not know the numbers, you cannot design. You can only guess. And guesses collapse.

Foreign aid policy in the United States is a building designed by guessers.

Consider: if you believe foreign aid is twenty-five percent of the budget, then cutting it in half feels like a significant savings: twelve and a half percent freed up for other priorities. That sounds like it matters. That sounds like a real policy choice with real consequences.

But if foreign aid is actually less than one percent, then cutting it in half saves... less than half a percent. The entire passionate debate (the campaigns, the attack ads, the congressional hearings, the opinion columns, the shouting on television) is about an amount of money so small that eliminating it entirely would not noticeably change any other budget line.

The debate is real. The passion is real. The amount being debated is, relative to the total, almost imaginary.

That is what I mean by a hallucination with parliamentary procedure. The democratic process is functioning: people are voting, representatives are deciding, arguments are being made in good faith on all sides. But the process is operating on a number that is wrong by a factor of twenty-five. The parliamentary procedure is intact. The foundation is a mirage.

The Lens

Here it is, the thing that will follow you out of this post:

You cannot have an informed opinion about a number you have never looked up.

This sounds obvious. It is not. Because most people believe they HAVE looked it up, or that they absorbed the number through general awareness, through the news, through the culture. They did not. They absorbed a feeling, the feeling that foreign aid is "a lot," and converted that feeling into a number. The feeling is real. The number is fabricated. And the fabricated number is the one they vote on.

This is not unique to foreign aid. How much does the US spend on defense? On Social Security? On education? On healthcare? On interest on the national debt? Most citizens would be wrong about most of these numbers. Not by a little. By multiples.

The entire edifice of democratic self-governance rests on the assumption that citizens are at least approximately informed about the things they have opinions about. The Trim Tab reveals, question by question, that this assumption is structurally unsound. Not because citizens are stupid. Because the information environment has failed to deliver the load-bearing numbers.

The Design Response

I am an architect. I do not just diagnose. I build.

The design response is not "educate the public about foreign aid." That is a lecture, and lectures do not compete with the algorithm. The design response is the Trim Tab: make the number visible, make the surprise enjoyable, and let the correction self-replicate.

You just experienced it. You read a question. You committed to an answer. You discovered the gap between your guess and the reality. And now, I am willing to bet, you want to ask someone else the same question. Not to prove them wrong. To watch the look on their face when they discover the number. Because that look, the look of genuine, shame-free surprise, is the most human thing there is. It is a person's model of reality updating in real time. It is error-correction made visible.

Each time that happens, one more edge connects two triangles. One more node joins the dome. One more citizen is now better calibrated than most of Congress.

You are now better informed about foreign aid than the average voter, the average pundit, and possibly the average legislator. That took you thirty seconds. The question is: what else have you not checked?

Keep Going

If you want to see what happens when a physicist gets a number wrong and grins about it, read Richard Feynman's Three Trillion Trees and One Dead Physicist.

If you want to see what it looks like to be right and still not know enough, read Carl Sagan's The Most Important Thing Nobody Celebrated.

And if you want to understand why shared wrongness is the structural element that holds this whole project together, read The Geometry of Shared Reality.

The dome grows. Each post is a triangle. Each shared surprise is an edge. The geometry does the rest.

"You never change things by fighting the existing reality. To change something, build a new model that makes the existing model obsolete."

— Buckminster Fuller, Architect of the Universe March 17, 2026