Market Economics Lab: Supply and Demand in the Real World

Overview

Supply and demand is the most famous idea in economics, and most people who can recite the two crossing lines have never once watched the law operate in front of them. In this experiment you are going to. You will choose a real local market — a farmers market, a flea market, a swap meet — and treat it as a laboratory. You will form a real hypothesis, collect real data over more than one visit, and find out whether the textbook holds up when the variables are people, weather, and a vendor who wants to go home.

The Question

You will pick one of these questions, or write your own in the same shape. Each one is testable with data you can actually collect.

• Does the price of the same good fall as the market is about to close? (The supply is perishable and the seller's willingness to hold it drops by the hour.)
• Do stalls with more competitors selling the same thing charge less than stalls selling something rare? (Competition versus scarcity.)
• Does a good that is abundant this week cost less than the same good when it is scarce? (Seasonal supply.)
• Do prices vary for the same item across vendors, and what explains the spread? (Information, quality, and bargaining.)

A good question has a measurable cause (time, number of competitors, abundance) and a measurable effect (price). If you cannot count both sides, sharpen the question until you can.

Background

You need three ideas, and no more, to start.

Demand is how much of something buyers want at a given price. As price rises, people generally buy less. Supply is how much sellers will offer at a given price. As price rises, sellers generally offer more. Price is where those two pressures meet — the number at which the amount people will buy equals the amount sellers will sell.

In a textbook this is two clean lines crossing. In a real market it is messy, because the lines move constantly. A rainstorm thins the crowd and demand drops. A second tomato vendor sets up next door and supply jumps. The market is closing in twenty minutes and a baker would rather sell a loaf at half price than carry it home. Every one of these is a line shifting in real time, and you can watch it happen. Your experiment isolates one of these shifts and measures it.

One warning that separates real economics from a guess: a price difference is not proof of a cause. The closing-time loaf might be cheaper because it is closing time — or because that baker is new, or because the loaf is a day old. Your job is to design the observation so that the cause you are testing is the one most likely to explain what you see.

Hypothesis

Write this before your first visit. The discipline of predicting in advance is the whole point — it stops you from "discovering" whatever the data happened to show.

Before you begin, write down what you think will happen and why:

"I think _______ because _______."

Be specific and quantitative if you can. Not "prices will drop at closing time" but "I predict that the price of perishable goods (produce, baked goods, cut flowers) will be at least 20% lower in the final hour than at opening, while non-perishable goods (crafts, tools) will not change."

Designing a Fair Test

The difference between an experiment and a stroll through a market is control of variables. In a chemistry lab you change one thing and hold everything else fixed. You cannot do that perfectly in a real market — you cannot freeze the weather or clone a vendor — but you can do it well enough that your conclusion means something. Three rules make your test fair:

1. Hold the good constant. "Tomatoes" is not a measurement; "medium vine tomatoes, sold by the pound" is. A higher price on heirloom tomatoes than on Romas is not the law of supply and demand — it is two different products. Define your goods so tightly that you are genuinely comparing like to like across time and across vendors.

2. Change one thing on purpose. Your hypothesis names one variable — time of day, number of competitors, abundance. That is the thing you deliberately vary. Everything else, you try to keep the same. If you are testing time of day, return to the same vendors selling the same goods, so the only thing that moved is the clock.

3. Watch the variables you did not choose. The weather, the crowd size, a vendor running out, a new stall appearing — these are confounding variables, and they will try to fool you. You cannot stop them, so you record them. When you analyze your data, a noted confound ("it rained during the midday round") is the difference between an honest conclusion and a wrong one. The student who writes "prices dropped at closing" without noticing it also poured rain has not run an experiment; they have told a story.

Materials

• A data sheet prepared in advance — columns for: item, vendor, price, time, quantity available (your estimate), number of competing vendors, and notes
• A notebook or spreadsheet
• A camera for documenting price tags (ask before photographing a vendor's stall)
• A watch for accurate timestamps
• Cash, only if a test purchase is part of your design

Procedure

Setup

1. Choose your market and confirm its hours and which days it runs. You need a market you can visit more than once, ideally at different times of day or in different weeks.
2. Choose your single question and write your hypothesis.
3. Build your data sheet. Decide exactly what you will record at each observation so you are not improvising in the field.
4. Decide your sampling plan. Example: "I will record the price of three goods at five different vendors at opening, midday, and the final hour, on two different market days." Write it down. A plan you wrote in advance is harder to bend to fit a conclusion you wanted.

Experiment

1. Visit one — baseline. Walk the whole market once before recording anything, to learn the layout and the typical goods. Then collect your first full round of data exactly as your plan specifies. Record the time of every single observation.
2. Talk to vendors. This is data too. Ask a few sellers neutral questions: "How do you decide what to charge?" "Does the price change during the day?" "What sells out first?" Most vendors will tell you, and their answers are a window into the supply side you cannot get from prices alone. Record what they say verbatim where you can.
3. Visit two (and three). Return at a different time or on a different day and collect the same data the same way. Consistency is everything — measure the same goods, the same vendors, the same way, so the only thing that changed is the variable you are testing. The more rounds you collect, the more confident your conclusion can be: a single price drop could be a fluke, but the same drop seen across three Saturdays is a pattern. Aim for at least two full visits, three if your schedule allows it.
4. Watch for the natural experiment. If a downpour empties the market, a vendor sells out, or a new competitor sets up, note it. Real markets hand you experiments you did not design. Use them.

Record

Capture every observation in the same format so you can compare cleanly.

Analysis

Now turn the numbers into an answer. Resist the urge to skip to the conclusion you wanted.

• What happened? Calculate the averages. What was the average closing-hour price versus the opening price? What was the price spread across vendors? Put the numbers next to each other.
• Did it match your hypothesis? Be honest. If you predicted a 20% drop and found 4%, your hypothesis is mostly wrong, and that is a real and useful result. State the actual number.
• What might explain the results? This is where you separate cause from coincidence. If closing-time prices dropped, was it the time — or did the cheap vendor just happen to be the one still open? Look at your competitor counts, your weather notes, your vendor interviews. Argue for the most likely explanation and against the others.
• What would you change if you did it again? Maybe you needed more vendors, more visits, or a tighter definition of "the same good." Naming the weakness in your own design is the mark of someone who actually understands the experiment.

The Explanation

Here is what is almost certainly going on underneath your data.

Prices in a real market are set by the constant negotiation between what buyers will pay and what sellers will accept, and both of those numbers move with conditions. Perishability is the cleanest force you can observe: a tomato or a loaf of bread loses value to the seller by the minute as closing approaches, because an unsold perishable is a total loss. So the seller's lowest acceptable price falls through the day — supply behavior shifts — and prices soften. A craft or a tool has no such clock, so its price holds. If your data showed perishables dropping and non-perishables steady, you watched supply economics, not a coincidence.

Competition compresses prices. When five vendors sell the same tomatoes within thirty feet of each other, no one can charge much more than the others without losing the sale, because the buyer can simply step to the next stall. Where a good is rare — one vendor with the only local honey — the seller has room to charge more. This is why you found a wider price spread for unique goods than for common ones, if you did.

Information and bargaining explain the rest. A buyer who knows the going rate, who arrives late, or who buys in volume can often negotiate, because the seller's true floor is lower than the tag. The price tag is an opening offer, not a law of nature. Markets are conversations.

What you will not have found is the clean crossing of two straight lines. That diagram is a model — useful, but a simplification. What you measured is the real thing the model is trying to describe: thousands of small decisions, made under pressure, that add up to a price.

Why This Is the Engine of American Dynamism

Step back from the tomatoes for a moment, because what you watched at one weekend market is the same mechanism that runs the entire American economy. A price is not set by a government office or a committee; it emerges from millions of independent buyers and sellers each pursuing their own interest, and that emergent price quietly coordinates the whole system. When a good becomes scarce, its rising price signals producers to make more of it and consumers to use less — without anyone in charge issuing an order. When a good becomes abundant, the falling price tells producers to stop and consumers to buy. The vendor lowering the price of unsold bread at closing time and the global market reallocating steel after a shortage are running the identical logic at different scales. This decentralized, self-correcting coordination — the reason American shelves are stocked without a central planner deciding what gets made — is one of the most powerful and least understood ideas in all of economics. You just watched it operate at human scale, close enough to touch. Most people only ever encounter it as a diagram in a textbook. You measured it in the wild, and that changes how you will read every price for the rest of your life.

Extensions

• Change one variable: Run the same design at a very different market — a grocery store with fixed prices, where bargaining and closing-time discounts mostly do not happen. Why does the farmers market behave differently from the supermarket? (Hint: who sets the price, and how perishable is the inventory in each?)
• Real-world connection: Apply your findings to your own venture if you have one (see the Start a Micro-Business unit). When should you discount? When can you hold your price? The market just taught you.
• Further reading: Naked Economics by Charles Wheelan, for supply and demand explained without jargon. Then read about a real market collapse or shortage in the news and use your framework to explain the prices you see reported.

From Observation to Prediction

The real test of whether you understand a market is not whether you can explain yesterday's prices — anyone can explain the past. It is whether you can predict tomorrow's. After your second visit, before your third, write a prediction: "Next Saturday, by the final hour, the price of [your good] at [your vendor] will be approximately ___, because ___." Then go and check. A prediction that comes true is far stronger evidence that you have found a real pattern than any after-the-fact explanation. A prediction that fails is even more useful — it tells you the pattern you thought you found was partly luck, and it sends you back to look for the variable you missed. Economists, weather forecasters, and engineers all live by this rule: an explanation that cannot make a prediction is just a story, and a prediction that can be checked is the beginning of knowledge.

A Note on Honest Numbers

There is a quiet temptation in every experiment, and it is worth naming so you can resist it. When the data does not match the hypothesis, the easy move is to drop the inconvenient observation, round in the helpful direction, or quietly redefine "the final hour" until the numbers cooperate. Do not. The entire value of running a real experiment instead of reading a textbook is that reality gets a vote, and reality only teaches you something if you record it faithfully — especially when it contradicts you. A surprising, honest result is worth ten tidy ones you nudged into shape. The habit of writing down what actually happened, not what you wanted to happen, is the single most important thing this experiment can teach you, and it transfers to every serious thing you will ever measure.

Safety Notes

This is a green-level investigation conducted in a public space. The risks are ordinary, not chemical.

Chemical/Material Hazards

• None. You are observing and recording. If you make test purchases of food, normal food-handling and allergy awareness applies.

Disposal

• No materials to dispose of. Any food you buy, you eat or share.

Protective Equipment

• None required. The relevant precautions are social: stay in the public market area, keep your phone and cash secure in a crowd, agree on a meeting point and check-in time with the adult who brought you, and approach vendors and strangers politely and only in the open, busy part of the market. Tell the adult advising you your sampling plan before you go so they know where you will be and for how long.