The Dangers of ‘Trump‑Style’ Modeling: How to Keep Math Models Honest

Some models appear rational but are merely a façade of technical illusion.

On April 2, 2025, U.S. President Donald Trump signed an “equitable tariff” executive order, a policy driven by his senior trade advisor Peter Navarro.

Trump’s tariff logic likely contradicts economic principles: it is not a deep equilibrium model nor a strategy optimization based on international negotiations, but a simple proportional formula—trade deficit divided by total imports multiplied by a coefficient.

The detailed formula is referenced in an article titled “Trump’s tariff model is exactly the same as AI’s!”.

I call this approach “Trump‑style modeling”, characterized by a lack of domain knowledge, intuition‑driven guesses, and linear rules.

Navarro, long notorious in academia, fabricated an “economist” Ron Vara to support his books, used student reports as fake evidence, and employed emotive language and propaganda videos to create a public‑war model—what I term “Navarro‑style modeling”, a pseudo‑academic populist performance with fake data, citations, and experts.

“Trump‑Style Modeling”: Lack of Domain Knowledge

The purpose of mathematical modeling is to translate real‑world problems into quantifiable, inferable, and verifiable structures. Designers must deeply understand the problem background, variable relationships, logical structure, and constraints. “Trump‑style modeling” jumps over this process with excessive simplification.

Consider the model used in the U.S. “equitable tariff” policy:

In the formula, D_i represents trade deficit, M_i represents total imports, and a 50% adjustment coefficient is applied. The logic seems clear and may work short‑term, but it lacks theoretical grounding: no marginal substitution effects, no global value‑chain structure, no expected impact on domestic consumption prices or employment.

This is the danger of “head‑bang modeling”: using mathematical symbols without domain‑specific logic, applying a model’s skin without its soul.

“Navarro‑Style Modeling”: Fraud in Academic Disguise

Compared to “Trump‑style guesswork”, Navarro’s modeling violates academic ethics and modeling principles fundamentally.

In several best‑selling books, Navarro repeatedly cites a non‑existent economist Ron Vara. The name is an anagram of “Navarro” and was fabricated to lend false credibility.

Peter Navarro

Such fabricated experts, fake citations, and emotional manipulation produce models that are false, manipulative, and emotional, ensuring they will not endure.

Modeling Baseline: Honesty, Logic, and Responsibility

Whether solving everyday problems or competing in international arenas, the true value of mathematical modeling lies not in “pretty formulas” or “exciting conclusions”, but in the modeler’s ability to respect reality, face data honestly, reason logically, and bear responsibility for decisions.

It is theoretically guided problem abstraction , not experience‑driven guesswork.

It is data‑driven mechanism exploration , not imaginative stitching.

It is clear causal reasoning , not result‑oriented rhetoric.

It is transparent result verification , not reliance on authority.

Especially in the AI era, when language models can output seemingly reasonable formulas, we must strengthen training in modeling thinking to avoid becoming new “Trump‑style” or “Navarro‑style” modelers.

We discuss why “mathematical modeling should not be Trump‑like nor Navarro‑like” not to criticize a president or denigrate a scholar, but to emphasize that modeling is both a technical task and a value‑driven practice.

We need a modeling culture that prioritizes solid verification over quick virality, serves problem solving rather than propaganda, and relies on honest logic instead of decorative data.

The future will be a world where human thinking and algorithmic reasoning co‑construct solutions, making us co‑guardians of model governance, demanding greater rigor, self‑discipline, and reverence for truth.