← Back to the lecture: Gains from Exchange
T2 Lecture Recap · Part 1
Why moving existing goods around creates wealth, what a middleman gets paid for, where a price has to lie for both sides to gain, and when trading stops.
Session 6 · September 10, 2026 · 10-minute review · Final 34 minutes of class · Topic continues Tuesday, September 15
Sam hands Joe some granola bars. Joe hands Sam some bottles of water. No bar was made and no bottle was filled, yet both say they are better off. If trading only moves existing things around, where did the gain come from? From the fact that the two men value the same goods differently. Wealth is not stuff. The same goods are worth more in the hands that value them most, and trade is how they get there.
How to use this page
This recap covers the last 34 minutes of the September 10 class, which opened the gains-from-exchange topic after finishing elasticity (the first 37 minutes have their own recap and transcript, labeled Elasticity Part 3). Class ran from the cold open through the camp tables, the jacket, and the two-panel figure, and stopped just before the crossing; the topic continues Tuesday. Read the four-step diagnostic first. Then use the lecture sections to reconnect each idea to an example from class. Finish with the application checks. The separate complete transcript is available when you want the full explanation, including the candy-bag trading game and the long exchange about middlemen.
Use these steps whenever a question asks whether a trade creates value, who gains from it and how much, whether a middleman is worth his fee, or when two people will stop trading.
What you should be able to do
Opening · The cold open
Up to now the course has been about one person’s demand. This topic asks why people exchange, how much exchange there will be, and when it ends. The opening scenario: Sam and Joe swap granola bars for water, no bar is produced and no bottle is filled, and both say they are better off. Where did the gain come from? From different valuations. Give two people identical bundles and, if they value the goods differently, they will trade, and the trade raises the value of the fixed pool of stuff.
The instructor’s classroom trading game makes the point with candy. Everyone gets a bag of five random pieces and rates each from 1 (meh) to 5 (favorite), so a bag is worth at most 25. Then people trade, first only within their side of the room, and re-rate. The value of the candy on the left goes up, on the right goes up, in the center goes up: the wealth in the room rose with not one more piece of candy. Remove the barrier between the sides and it rises again, because there were gains from trade across the room that the rule had blocked. Nobody with a bag of all fives would trade, and nobody has to; anyone below 25 can do better with someone whose tastes differ.
The TopHat check asked why Sam and Joe can both gain from exchanging existing goods and when their exchange would stop. Each wrong option teaches something. “Only the person who receives the more useful good benefits”: there is no universally useful good, and nobody trades voluntarily to make himself worse off. “Exchange creates new goods and stops when prices are equal”: nothing is produced here, and no grocery store sells everything at one price. “Both benefit only when each receives more goods”: it is the worth of what you get, not the count, and a man crossing the Sahara would give all his bars for one bottle. The right answer: both gain when each values what he receives more than what he gives, and exchange stops when no mutually beneficial trade remains. Trade happens because people are different and therefore disagree about what things are worth; identical people with identical bundles would never trade.
Movement 01 · Trade creates wealth
Sam and Joe each start with 20 bottles and 20 bars. A newcomer arrives with nothing, and for now he can rearrange who owns what at zero cost. He notices that Sam would give up bars for water and Joe would give up water for bars, and he brings the two sides together. That is what middlemen do, and it prompted the day’s longest digression. “Cut out the middleman” (Michael Scott gets accused of being one on bring-your- daughter-to-work day) only makes sense if you look at the dollar price and not the full price. Go straight to the farmer and you take on the inspecting, the contracting, and the driving to a hundred farms that the grocery store does for you; Aldi’s cheaper pears come with the time you spend digging through the bin. Grocery stores are middlemen between you and wholesalers, wholesalers are middlemen between grocery stores and farmers, and every link adds value. The newcomer, a student suggested, has a comparative advantage in seeing what Sam and Joe do not.
The deals: the newcomer takes 7 bars from Sam for 10 bottles, and gives Joe 5 bars for those 10 bottles, keeping 2. Sam would have paid up to 8, so he gains 1 bar; Joe would have accepted as few as 4, so he gains 1 bar. The concept to recall is surplus. On the tables, Sam ends with 30 bottles and 13 bars, Joe with 10 bottles and 25 bars, and the newcomer with 2 bars. Everyone is better off, and the totals are still 40 and 40. The objection that this must be zero-sum because the totals never changed is counting stuff and not value. Middlemen, the instructor said, are the unsung heroes of the economy; the newcomer created wealth and kept a piece of it, and without that piece he would have had no reason to act.
The second TopHat check: the newcomer changes the payment terms while the transfer stays feasible and both men still agree. The gains do not disappear (they agreed, so value was created); the newcomer is not the only one who can gain (an opportunity cost, if we allowed one, would cut into his share); and no higher payment creates more goods (there are only 40 of each). The total gain stays 4 bars and only its distribution changes: the newcomer might keep a bar and a half, or one, or half, and the rest stays with Sam or Joe. One thing to watch in class: the instructor at one point said the total is “still two bars”; the tables and the answer he was confirming both say four.
Movement 02 · Personal worth and the stopping point
Gains from exchange rest on marginal personal worth, which is why the course taught demand first. Assume costless exchange for now: the newcomer had no cost of negotiating, though real middlemen do. Sam and Joe trade as long as each expects to be better off, which means the price, in bars per bottle or its inverse in bottles per bar, has to lie between their two marginal worths. Below the seller’s worth he would not sell; above the buyer’s worth he would not buy. A personal worth is a reservation value, the most you would give to acquire a good or the least you would accept to give it up, which is exactly what eBay asks for when you set a reserve on a listing or a maximum on a bid. The jacket: you would pay up to $40, the seller would take as little as $25, it sells for $32. The buyer gains $8, the seller $7, and any price between $25 and $40 would have left both better off.
Now draw it. Both men start with 20 bottles. Sam’s worth of the 20th bottle is a full granola bar; Joe’s is two-tenths of a bar. So there are gains from trade, and Joe, who values the marginal bottle less, becomes the seller of water even though he bottles nothing. The cost to Joe of drinking his 20th bottle is what he could have sold it to Sam for. Armen Alchian, one of the authors behind this course, used to say “demand: it’s all there is”; supply is demand looked at from the other side, and Joe’s worth curve is his demand curve read as a seller. As Joe gives up bottles his marginal worth of a bottle rises (he moves up his curve), and as Sam collects them his falls. They keep trading until the marginal worth of a bottle is the same to both, and “breaking even” means exactly that. Past that point a further swap would leave one of them, in fact both of them, worse off. Class stopped there.
What continues Tuesday
Class stopped just before the crossing: the slide that puts the two worth curves on one axis and shows where trading stops, how many bottles change hands, and what happens if the men overshoot. Still to come after that: what Pareto-optimal does and does not mean; what happens to the value of a good when the trade is forbidden; who does the middleman’s work when you cut him out; how lower transaction costs make more trades worthwhile; a fee ceiling and a rival dealer; and licensing. Tuesday’s class gets its own recap and transcript page.
Do not memorize an example as a story. Use it to recover the economic principle.
| Example from class | Economic lesson |
|---|---|
| Bars for water, and both say they are better off | Wealth is not stuff; the same goods are worth more once they sit with the people who value them most. |
| Bags of candy rated 1 to 5, traded, and re-rated | Reallocating a fixed pool of goods raises total value; a bag of all fives has no reason to trade, and nobody has to. |
| Left side trades with left side only; then the barrier comes down | Every trade barrier blocks some gain; removing it lets value rise again with no new goods. |
| Crossing the Sahara with a hundred bars and one bottle | What makes a good valuable is its marginal worth to you, not how many units you receive. |
| 7 bars for 10 bottles; 5 bars for 10 bottles; 2 bars kept | Each gain is worth minus what was paid or received (1, 1, and 2); the total is 4 bars and no bar was made. |
| “It’s zero-sum: still 40 bottles and 40 bars” | That counts stuff, not value; exchange is positive-sum because it moves goods to higher valuations. |
| The newcomer who would not bother without his two bars | A middleman is paid for seeing a gain the parties missed; his fee is the price of the wealth he creates. |
| Michael Scott, Aldi’s pears, Publix and the wholesaler | “Cut out the middleman” compares dollar prices and ignores the full price: inspection, search, contracting, and gathering everything in one place. |
| The newcomer keeps a bar and a half instead of two | Changing the terms changes the division of the gain, not its total, so long as both sides still agree. |
| eBay’s reserve price and maximum bid | A personal worth is a reservation value: the least a seller will take or the most a buyer will give. |
| The $40 buyer, the $25 seller, the $32 jacket | Both gain when the price lies between the two worths ($8 and $7 here); the price decides the split of a $15 gain. |
| Sam’s 20th bottle is worth a bar; Joe’s, two-tenths | The man with the lower marginal worth becomes the seller; supply is demand read from the other side. |
| Joe’s worth rises as he sells; Sam’s falls as he buys | Trading continues until the marginal worths are equal; past that point a swap hurts. |
Answer before you open each one. Every question uses only material from class.
Counting stuff instead of value. Sam gave 7 bars for water he valued at 8 bars, and Joe took 5 bars for water he valued at 4. Each is a bar better off by his own valuation, and the two bars that left their hands paid the newcomer for a gain neither had found on his own. Wealth is worth, not inventory.
Sam gains 2 bars (8 minus 6), Joe gains 1 (5 minus 4), the newcomer keeps 1 (6 minus 5). Total: 4 bars, unchanged. Terms move the split; they cannot move the total, because the total is fixed by the gap between the two men’s worths.
No trade. Sam’s worth of the 10 bottles is 8 bars, so a price of 9 leaves him worse off and he refuses. The price has to lie between the two reservation values (4 and 8 bars) for both to gain, and 9 is outside that range. The newcomer, who would have earned 4 bars, gets nothing.
You gain $10 (60 minus 50); the reseller gains $15 (50 minus 35). The whole gain is the $25 gap between the two reservation values, and every price from $35 to $60 produces that same $25; the price only decides who gets how much of it. The jacket, with different numbers.
The dollar price at the farm gate may be lower, but the full price includes the store’s work: inspecting quality, contracting with hundreds of growers and ranchers, and putting everything in one place so you make one trip instead of ten. Cut out the middleman and you do that work yourself, at Aldi’s bin-digging pace or worse. Middlemen create value; their fee is payment for it.
No. Trade needs two people who value the same thing differently. With the same bundles and the same worths, no price lies strictly between a buyer’s and a seller’s reservation value, so no swap can make both better off. Identical bundles with different tastes, as with Sam and Joe, is a different story.
Because his marginal worth of a bottle (0.2 bars) is below Sam’s (1 bar), so he is the one who parts with bottles. Supply is demand read from the other side: the cost to Joe of drinking his 20th bottle is what he could have sold it for. Alchian’s line, “demand: it’s all there is.”
Because the worths move as they trade. Each bottle Sam adds is worth less to him than the last, and each bottle Joe gives up makes the next one worth more to him. Once the two marginal worths are equal there is no price that leaves both better off, and any further swap makes one of them (in fact both) worse off. Trading stops where the worths meet; Tuesday draws that crossing.
Bottom line
Trade creates wealth without creating goods, because the same goods are worth more once they sit with the people who value them most. Each side’s gain is its worth minus what it paid or received; a middleman is paid for finding a gain the parties missed, and changing his terms redistributes the gain without changing it. Both gain only when the price lies between their reservation values, and they keep trading until their marginal worths are equal. Tuesday puts the two worth curves on one axis and finds that stopping point.