Philosophy of science for economists, scientists, and decision-makers.
Economic models shape your world. They design the markets you trade in, match students to schools, and set the level of your taxes. Used with discipline, they make institutions fairer and more effective. Used as if nothing had been left out, they can do real harm.
By the end of your visit, you'll be able to look at any economic model — in your coursework, in a policy debate, in the brief on your desk — and ask the questions that separate responsible use from careless use. Does what this model shows hold up where you mean to apply it, and in enough detail to matter? And what did it leave out to stay simple — whose interests, whose experience — that no better model could ever bring back? Philosophy of science has built tools for putting such questions to work. The museum hands you the few that matter most, and shows you through real cases what it looks like to use models well — and badly.
You'll leave with a rare and useful literacy: the ability to engage with models on their merits, neither trusting them blindly nor dismissing them out of hand. That is responsible economics.
Some trust economic models too easily. Others not at all. This museum is for those who suspect the right answer lies in between — and who want the tools to judge when a model deserves trust, and under what conditions.
Hello, I'm Filikon
Welcome to the Museum of Economic Models!
Two rooms are open today — manipulable models and causality — along with the Entry Hall, where every visit begins. Room III and nine more are taking shape behind their doors. I can show you the map and controls in under a minute, or you can dive straight in.
Drag to pan · Scroll to zoom · ← → to rotate · Drag Filikon into a room to step inside · Tab + Enter to visit by keyboard · Click tags to read
Play the Monopoly game!↓
Reading Activity!
Have you seen the early models exhibited on the back side of the room? Incredible! Some are mathematical, some are tables, and one is a material object! The funniest bit, really, is to find out in which sense these were “manipulable”!
Bring me into a room!
Guided Tour
Welcome! Click Next to begin.
SW
Keyboard controls
Museum map
Tab move between rooms · Enter visit the focused room ←→ rotate view · Shift+arrows pan · +− zoom · T guided tour
Inside a room
←→ turn · ↑↓ walk Tab face the next exhibit or door (rightward) · Shift+Tab leftward Enter open what you face · Esc back to the map
Dialogs
Tab move between controls · Esc close
Anywhere
Ctrl/⌘+K search & jump to any exhibit · ? this help
🏠 The Village Monopolist!
You own all 50 houses in the village! Given the demand you face, the more you rent out, the lower the price you can charge.
How many houses would be rented out in a perfectly competitive economy (short run)? How many will you rent out?
1 (few)25 (half)50 (all)
✓ Competitive
Price
—
Rented
—
Profit
—
⚡ Your Monopoly
Price
—
Rented
—
Empty
—
Profit
—
Based on Hal Varian's housing market model (Intermediate Microeconomics, 8th ed.).
The demand curve is logarithmic — P = −ln(Q/103) — making the profit-maximising
quantity non-obvious. Experiment with the slider: the model adjusts itself when you push its knobs.
The shaded green area is monopoly profit; the red triangle is deadweight loss relative to the competitive benchmark.
Market Machine · Room III podium
A Double-Auction Laboratory
Six buyers, six sellers, one unit each. Run a market day and watch trade prices against the competitive equilibrium — then change who is trading, and see what the result actually depends on. (Smith 1962; Gode & Sunder 1993.)
Before you trade, one notice: this podium isn’t a display about an experiment. It is one — and in the third mode you are its subject, one more trader in the data. Keep one step for yourself, because the machine can’t take it for you: whatever holds in this laboratory holds of this laboratory. Carrying it to a real market is an inference, and the gallery ahead is about what that inference costs.
Before anything runs, make your call. These robots bid and ask at random: no strategy, no learning, no memory. With the budget rule on (ZI-C), how much of the possible gains from trade will the market extract?
You are B1A unit is worth — to you · profit so far 0
Your profit on a trade = your value − the price you pay. Bid below your value to profit — but too low and no seller will ever accept. (Bid above it if you like; nobody will stop you.) Submitting a bid opens the market day if it isn’t running.
100
best bid — · best ask —
Dots are trades (rust = traded at a loss, possible only in ZI-U); the dashed line is the competitive equilibrium computed from the step supply and demand curves. Efficiency = realised gains from trade ÷ maximum possible gains.
DN-LAB · ROOM II PODIUM
The Auditor's Game
You are the lab auditor. Three case files probe Hempel's deductive-nomological model. Earn findings; assemble a verdict.
Audit Log0 / 3 findings
⚐Symmetry⚐Sketches⚐Cp-Hydra
1 / 4
CASE FILE #1
The Symmetry Trap
DN says: derive the explanandum from laws + initial conditions and you have explained it. Both candidate arguments below are valid DN derivations. Are both genuine explanations?
👇 Click both directions to compare the two derivations.
DN Derivation
Law of rectilinear propagation of light — light travels in straight lines.
Initial condition — the pole is 80 m tall; the sun's elevation is 36°.
∴ The shadow is tan(54°) × 80 ≈ 110 m long.
✅ Feels right. The pole causes the shadow — of course.
DN Derivation
Law of rectilinear propagation of light — light travels in straight lines.
Initial condition — the shadow is 110 m long; the sun's elevation is 36°.
∴ The pole is 110 / tan(54°) ≈ 80 m tall.
🤔 Also valid — by DN standards. But did the shadow explain the pole's height?
Both derivations are logically valid by DN standards. Same law, same structure — different initial conditions. Yet only one tracks the actual causal order: the pole casts the shadow, not the other way round.
⚑ FINDING: Symmetry Flag
Genuine explanation is asymmetric: the pole explains the shadow but not vice versa. DN treats both derivations as equivalent and so loses the causal direction that explanation requires. The first commitment of DN — that subsumption-under-law is explanation — does not hold without further machinery.
CASE FILE #2
Sort the Sketches
Five candidate “explanations” reach the lab bench. Sort each into one of three boxes. The auditor cares: is the missing detail findable in principle, or has the claim hidden its counterexamples in a ceteris paribus clause?
✓ Genuine DN⊙ Sketch (completable)✗ Cp-trap
“Water at sea level reached 100 °C; water at sea level boils at 100 °C; therefore this water boils.”
“The 2008 housing collapse was driven by a build-up of subprime exposure, a sharp tightening of short-term funding, and a feedback loop of forced asset sales.”
“Wages rise when unemployment falls — ceteris paribus.”
“Smith resigned because he felt his proposal was being systematically ignored by the committee.”
“Markets are efficient — except when they are inefficient.”
Sort all five to receive the audit finding.
⚑ FINDING: Sketches & Cp-traps
Hempel allowed for explanation sketches — DN derivations not yet completed, but completable in principle. They are legitimate provisional scientific products. The cp-clause is something else entirely: it absorbs counterexamples by definition. The next case file tests whether you can ever pin one down.
CASE FILE #3
Demolish the Hydra
A textbook claim arrives on the bench. Your job: enumerate the cp-conditions and turn it into a genuine DN law. Click each candidate condition to add it to the cp-clause.
“Wages rise when unemployment falls, ceteris paribus.”
Each enumeration will be tested against the historical record.
Conditions enumerated: 0Counterexamples reborn: 0
⚑ FINDING: Cp-Hydra
Every cp-condition you enumerated, a fresh counterexample appeared. The cp-clause is not a placeholder for a finite list of conditions — it is a structural blocker for completion. A “law” of this form cannot be cashed out as a DN explanation. The second commitment of DN — that explanations rest on well-confirmed laws — loses its supply line in much of social science.
AUDITOR'S REPORT
What Survives of the DN Model
Three findings collected. The verdict:
⚑Symmetry Flag — DN cannot distinguish cause from effect in symmetric derivations.
⚑Sketches & Cp-traps — sketches are legitimate placeholders; cp-traps are not.
⚑Cp-Hydra — many social-science generalisations resist DN completion structurally.
What survives
In closed systems — the laboratory bench, isolated celestial mechanics, perhaps the simple-substitution case of the law of demand — mechanisms produce the stable regularities DN needs, and explanation can take the deductive form Hempel described. The radiator on a freezing night is a clean DN case.
In open systems — the social world — those regularities are scarce. The DN schema cannot run. What survives as explanation is the identification of mechanisms and the analysis of the conditions under which they fire.
→
Step out of the lab and walk the gallery. Cartwright on the right wall, Elster on the front, and Mäki across to the back develop three independent critiques of what you just diagnosed — each followed by a worked economic example, and pulled together in a wrap-up before you cross to the island.
Curator’s Desk
Leave a note for Filikon
The museum is in early preview. Your impressions, confusions, and suggestions shape what the next rooms look like.
Thank you.
Your note has reached the curator’s desk. Filikon will read it personally.
Stay in the loop
Museum Newsletter
New rooms, new exhibits, new ideas — 4–5 messages per year. No noise, no marketing.
About the Museum
The Museum of Economic Models
Economic models are among the most consequential intellectual tools in modern public life. They shape markets, inform policy, and structure the institutions that govern everyday decisions. Whether they do so well or badly depends less on their technical sophistication than on how they are used — what they are asked to represent, and what is allowed to remain outside the frame. This museum is built around that question. Its purpose is to give visitors the conceptual vocabulary and practical tools needed to engage with economic models critically, neither deferring to authority nor dismissing what formal reasoning genuinely achieves. The museum takes no side in the disputes between schools of economic thought and pursues no policy agenda: its commitment is to the quality of reasoning, not to any particular conclusion.
What you will learn
Reading models critically
An economic model is always a selective representation of reality. To read one critically is to hold two questions simultaneously: what does this model make visible, and with what precision? A practiced reader does not ask whether the model is "right" or "wrong" but what it is capable of showing — and what it is not. The museum builds this habit through close engagement with real models across different domains and contexts.
Recognising what a model leaves out
Every model operates by abstracting away from complexity. Some abstractions are legitimate simplifications; others suppress phenomena that are central to the question being asked. The museum introduces a small set of tools — drawn from the methodology of science — for identifying which kind of exclusion is which. Visitors leave knowing how to ask: is what this model ignores genuinely irrelevant, or is it doing silent, consequential work?
Distinguishing responsible from careless model use
A model can be technically valid and still be misused — applied beyond its domain, stripped of its caveats, or presented as decisive when it is only suggestive. The museum illustrates both patterns: cases where models were deployed with appropriate care and explicit acknowledgement of limits, and cases where they were not. Visitors learn to recognise the difference, and to demand it from others.
How the museum is built
The museum draws on the practice of philosophy of science, which has developed a rich body of tools for analysing the structure, scope, and limits of scientific models. Rather than attempting a comprehensive survey of that literature, the museum works with a curated subset — the tools that travel most usefully across the range of economic modelling and remain accessible to non-specialists. Exhibits feature both exemplary cases, where models were used with discipline and acknowledged scope, and cautionary cases, where their limits were ignored or obscured.
Who is behind the museum
The museum was conceived and built by an academic researcher working at the intersection of economic theory and the philosophy of science. The curator holds a PhD in economics and has spent several years teaching economic methodology at university level. The museum reflects an ongoing commitment to making rigorous methodological thinking genuinely accessible outside specialised academic settings.
The museum is a living project. If you have questions, spot an error, or would like to contribute an exhibit, a case study, or a teaching resource, the curator welcomes the conversation — use the feedback form below or write directly via the Impressum.
Contact & legal
For questions, corrections, or collaboration enquiries, use the
.
Legal information and contact details are in the
Impressum.
Philosophy of science for economists, scientists, and decision-makers.
Floor Plan
Museum of Economic Models
Scroll or pinch to zoom · drag to pan · double-click to reset. Room titles will be updated as the exhibition develops.
