An economy of sheaves

Applying the mathematical sophistication of Gérard Laumon— specifically his genius for geometrization and sheaf theory — to economic theory provides a lens for reimagining how we model human systems.

Traditional mathematical economics relies heavily on classical calculus, linear algebra, and static equilibrium models, such as general equilibrium or expected utility. These frameworks often assume an economy is a rigid machine composed of isolated, rational actors.

If we examine economic theory through the abstractions Laumon used on the Langlands program, economics can move from a study of flat equations to a study of dynamic, topological spaces.

In traditional macroeconomics, data is treated as isolated numbers or vectors: GDP, inflation, or an interest rate at a point in time. Laumon’s work relied on sheaf theory, which attaches algebraic structure to the open sets of a space so local data can be tracked across scales.

An economy is rarely uniform. A policy or a supply shock behaves differently in a rural market than in a global hub. Replacing simple scalars with economic sheaves binds local context — regional habits, supply-chain proximity, regulatory nuance — to the global movement. Rather than averaging the data away, the sheaf keeps the local geometry as it scales up.

Classical economics searches for one stable general equilibrium, where supply meets demand. In algebraic geometry a moduli space is a space whose points are entire systems.

An economy is not a machine seeking a final rest. It is a shifting landscape of possible states, behaviors, and rules. Seen as a moduli space, the question is no longer “where will it settle?” It is “what is the topology of all possible market states, and how do we pass from one to another?”

The core of Laumon’s work is the Langlands program: a dictionary between arithmetic and geometry that turns a hard problem in one field into a solvable problem in another.

Economics is split the same way. Microeconomics is the person. Macroeconomics is the aggregate. The bridge is the friction. An economic Langlands duality would be a strict map from the category of agent networks to the category of systemic movement, so a pattern of persons can be read on the larger book, and the larger book can be read back onto the persons.

Standard models reduce a person to a utility function, one number said to maximize happiness or profit. Categorification replaces numbers with spaces, and equations with maps between them.

Utility is not one real number. It is a network of values, non-linear relations, and constraints. That structure holds a human act more faithfully than an optimization curve.

Edward Frenkel’s warning is not that mathematics is dangerous. It is that bad mathematics, kept in a few hands, is a way to spend other people’s lives. In Love and Math he writes that where there is no mathematics, there is no freedom, and that people need the tools to protect themselves from arbitrary decisions by the powerful few in a math-driven world. He has said the powers that be exploit ignorance, and manipulate us more when we are less able to read the formula. The account of his argument after 2008 is plain: inadequate models, used by bankers and traders, were allowed to stand in for the market. A formula for the cost of living can be revised the same way, and a pension moves while the public is still looking at the fence.

That is the economic ramification. A bad model does not stay in the paper. It prices a thin room off a thick one. It folds a charge into a discount so a privilege looks like a rate. It books a flattering profit, then sends the tail to someone who never signed it. The crisis was that pattern at full size. The firm kept the model. The state arrived later to socialize what the model would not hold. The person was not in the room.

The third seat is that room. The state may tax. The firm may issue. The person sits beside them and may refuse a line that cannot be shown. This book is the order. lgme.app is the small desk where the refusal is practiced. It is not a plan for the whole economy. It is the means by which a broader economic life stops taking a formula on faith.

It can do this, and it must. A model that cannot be refused will be written by whoever is paid to make it flattering. The desk has already watched that happen in miniature. A hub price of 97 offered itself in place of a thin room’s 92. A swapped charge offered itself as a friend rate. Both had to write nothing. That discipline is not a demo to leave behind. It is the step.