Derive comparative advantage and examine the adjustment that a simple model leaves out.
By the end: Calculate opportunity costs while distinguishing potential aggregate gains from compensation of losers.
Opportunity cost, not just absolute productivity
Country A can produce 10 units of grain or 5 of cloth per working day. Country B can produce 4 grain or 4 cloth. A is more productive in both activities, but a unit of cloth costs A two grain and B only one grain. B has comparative advantage in cloth; A has it in grain. Absolute advantage compares productivity in one activity; comparative advantage compares alternatives forgone.
At a trading price of one cloth for 1.5 grain, A can obtain cloth for less grain than it gives up making cloth itself. B receives more grain per cloth than it could produce by switching its own labor. This creates scope for mutual gains under the model. It does not require A to be worse at making cloth, and it does not prove that complete specialization is always desirable.
Build a feasible exchange from production possibilities
Comparative advantage compares costs within each producer before comparing producers. The relevant cost of cloth is how much grain the same resources could have produced instead. In the stated example, a day spent making five cloth in A prevents production of ten grain. Dividing ten by five gives two grain forgone per cloth. B gives up four grain to make four cloth, so its cost is one grain per cloth. These are physical opportunity costs, not money prices or hourly wages. A's higher output in both activities establishes absolute advantage but does not make its opportunity cost lower in both.
To construct a complete paper allocation, give each country two working days, identical within each country. Assume output is divisible, productivity is constant, labor can switch activities without cost, and neither country carries inventories from elsewhere. If each country initially devotes one day to each good, A produces ten grain and five cloth. B produces four grain and four cloth. Together they have fourteen grain and nine cloth. With no trade, each consumes its own production. This initial allocation is feasible, but it is not assumed to be a competitive equilibrium or the unique preferred consumption bundle.
Now move A's cloth-producing day into grain. A produces twenty grain and no cloth. In B, move half of its grain-producing day into cloth. B now spends half a day on grain and one and a half days on cloth, producing two grain and six cloth. Combined production becomes twenty-two grain and six cloth. The production change raises grain by eight and reduces cloth by three. Neither total rising nor specialization alone establishes a welfare gain. To compare consumption fairly, one must show what each country can actually obtain through exchange and whether its bundle is preferred.
More grain alone does not establish mutual gains
With two working days in each country, the initial allocation produces 14 grain; the changed allocation produces 22. These bars deliberately isolate grain, not a combined output index. Cloth simultaneously falls from nine to six. Without preferences and feasible exchange quantities, the extra eight grain cannot establish that either country's complete consumption bundle improves.
Suppose A imports five cloth and exports seven and a half grain at the stated trading price. A consumes twelve and a half grain and five cloth, compared with ten grain and five cloth initially. B consumes nine and a half grain and one cloth, compared with four grain and four cloth initially. A clearly has more grain with unchanged cloth, but B has traded away three cloth. Without a preference assumption, calling this particular allocation better for both is unjustified. A favorable price per unit does not ensure that every possible trade quantity makes each country's entire consumption bundle better.
The failed comparison separates a marginal opportunity from a complete allocation. B gains half a grain relative to its opportunity cost for each cloth exported, but that does not establish that it wants to export five cloth. Its consumption choices also depend on the cloth it keeps. To predict actual exchange, one would need preferences and a mechanism determining the price and quantity, with each country's purchases funded by sales or specified borrowing. The worked exploration below constructs a different allocation that leaves both countries more of both goods. It demonstrates feasibility without claiming to solve for a market equilibrium.
B's export quantity leaves less cloth to consume
In the failed paper exchange, B produces six cloth but exports five, retaining only one versus its initial consumption of four. Extra grain does not establish that B prefers this bundle. The favorable unit price leaves room for a different quantity; preferences are still needed to judge a trade that sacrifices cloth rather than increasing both goods.
Production is spread across firms and borders
Actual production combines design, components, assembly, logistics, and services across many firms. A disruption at one supplier can constrain output elsewhere, even when those other firms have workers and finance available. Inventories, spare capacity, alternative suppliers, and long-term relationships trade higher ordinary costs against resilience. The cheapest observed input price need not minimize the expected cost of disruptions.
Gross trade values can count imported content multiple times as goods cross borders. Domestic value added distinguishes production contributed locally from the value of imported inputs. Exchange rates, transport costs, trade rules, sanctions, and bargaining among firms affect where value is created and captured. Ownership of brands or intellectual property can matter as much as the location of final assembly.
Count supply-chain costs across ordinary and disrupted states
A supply chain is a set of production dependencies, not merely a list of sellers offering prices. If final assembly needs one of each of two components, surplus units of one cannot replace a missing unit of the other. This complementarity makes a small supplier capable of limiting a much larger production process. Money balances or unused assembly workers do not solve the immediate physical mismatch. Substitution may require a redesigned component, new certification, or time to qualify a supplier. Those costs explain why a second nominal source need not be an immediately usable alternative.
Consider a separate paper assembler planning 100 finished units per period. Each needs one specialized component. The regular supplier charges 4 currency units per component, and other assembly costs are 12 per unit. A finished unit sells for 20. With reliable supply, component spending is 400, other costs are 1,200, and sales are 2,000. The resulting operating contribution is 400 before fixed costs. A component supplier's lower price therefore matters, but it is only one part of the margin. Assume that without components the assembler avoids all these variable costs and makes no sales.
Now suppose the regular supplier completely fails with probability one quarter during the period, and succeeds otherwise. These are invented probabilities, not estimates. Expected contribution from relying solely on that supplier is three quarters of 400, or 300. An alternative supplier can guarantee all 100 components at a price of 5 each. Under that alternative, component spending is 500 and contribution is 300 in every state. The two sourcing plans have equal expected contribution under the assumptions. One is nevertheless more uncertain. A firm with debt due or workers requiring stable hours may evaluate that uncertainty differently from an owner concerned only with expected receipts.
Equal means conceal different delivery risks
Currency units per period; probabilities labeled separately
Delivery state
Risky supplier
Guaranteed supply
Success: probability 0.75
400
300
Failure: probability 0.25
0
300
Expected contribution
300
300
The risky plan gives 400 when delivery succeeds and zero when it fails. Its expected contribution is 0.75 times 400 plus 0.25 times zero, or 300. Guaranteed supply contributes 300 in either state. Equal expectations therefore conceal different possible outcomes. Variable production costs are avoided if output cannot be made; fixed costs and risk preferences remain outside this comparison.
Suppose instead the reliable supplier charges 4.8. Component spending becomes 480 and contribution is 320. It now exceeds the risky plan's expected 300 by 20. The reliable option costs more in the ordinary state but avoids enough expected lost production to pay for itself. Conversely, if the disruption probability were only one tenth, the cheap supplier's expected contribution would be 360. The reliable option at 4.8 would then sacrifice 40 in expected contribution. These comparisons depend on the probability, recoverable costs, selling opportunities, and whether the alleged backup survives the same disruption.
Inventory is another option with its own budget and timing. Components held in advance use working capital and may require storage or become obsolete. Their benefit depends on whether they arrive before the disruption and remain usable. A supplier in another country may diversify one local hazard while sharing dependence on the same port or raw material. Neither foreign nor domestic sourcing automatically establishes resilience. The numerical case identifies a mechanism for paying more to reduce risk. A real causal claim about sourcing policy requires evidence on these dependencies and on the production that would otherwise have occurred.
Work through the mechanism
Per working day
Grain capacity
Cloth capacity
Grain forgone per cloth
A
10
5
2
B
4
4
1
Labor is the only input in this constructed example. Comparative advantage depends on relative opportunity costs, not a wage comparison.
Use the static table to check gains and their limits
The worked model is a static production table, with no trade slider, country selector, or playback. Its capacities are per working day, not actual national output observations. Explore it with paper labor allocations while retaining the two-day endowment for each country. Return to the original no-trade bundles of ten grain and five cloth in A, and four grain and four cloth in B. A useful test of a gain is to find an exchange leaving each country at least as much of both goods, with some increase. That avoids relying on an unstated willingness to sacrifice one good for another.
Let A allocate one and a half days to grain and half a day to cloth. It produces fifteen grain and two and a half cloth. Let B devote both days to cloth, producing eight cloth and no grain. Their combined output is fifteen grain and ten and a half cloth. Compared with the original fourteen grain and nine cloth, there is now one additional grain and one and a half additional cloth. This production change is feasible because B can replace A's lost cloth at a smaller grain cost. Full specialization by both countries is unnecessary for this construction.
Exchange three cloth from B for four and a half grain from A. A then consumes ten and a half grain and five and a half cloth. B consumes four and a half grain and five cloth. A gains half a grain and half a cloth relative to its initial bundle. B gains half a grain and one cloth. Imports and exports match physically: the same three cloth leave B and enter A, and the same four and a half grain leave A and enter B. The value of each shipment is equal at the stated price of one cloth for one and a half grain.
Exchange can increase cloth consumption in both places
After the revised production allocation and exchange of three cloth for four and a half grain, both countries consume more cloth than initially. Grain also rises by half a unit in each country, as calculated in the prose. This is a feasible improvement with costless switching and shipping, not an equilibrium prediction; transport costs can undermine the comparison.
The calculation establishes the possibility of gains under its assumptions, not a unique price or allocation. Preferences, bargaining, and market organization determine which feasible trades actually occur. Transport also consumes resources. If shipping the three cloth uses one additional grain in total, it exactly exhausts the constructed aggregate grain gain, leaving the additional cloth. Who pays that shipping cost matters for whether both retain gains. If A alone pays the whole grain cost out of its final bundle, its grain falls to nine and a half. The previous comparison of more of both goods no longer holds for A.
A price outside the opportunity-cost interval fails for a different reason. If A must pay two and a half grain per cloth, that exceeds the two grain it gives up making cloth itself. Producing two cloth at home costs A four grain, while buying them at that price costs five before shipping. A can reject that import opportunity under the model. If B receives only half a grain per cloth, producing grain at home gives it more grain for the same labor. These contrasting cases show why low absolute productivity does not rule out gains, but exchange terms and resource costs still constrain them. The model contains no mechanism that forces a mutually beneficial agreement.
Importing costs more than producing at home
At the later hypothetical price of 2.5 grain per cloth, A must give up five grain to import two cloth, versus four grain to make them domestically. Both bars price the same cloth quantity before shipping. Rejecting imports is feasible under the fixed-productivity model; transport costs would worsen this offer, while changed production possibilities would require a new comparison.
Potential compensation is not actual compensation
Trade can lower consumer prices and expand markets while harming particular workers, firms, and places. Moving between industries takes time and may require new skills or relocation. Aggregate gains do not guarantee that affected people receive transfers, new jobs, or a share of those gains. A policy case must specify the distribution and adjustment process rather than claiming that hypothetical compensation has already happened.
The comparative-advantage example fixes technologies and available resources and omits unemployment, learning, bargaining power, environmental damage, and strategic dependence. Industrial policy may aim to develop future capabilities rather than accept today's pattern, but can also fail through poor information or capture. Evaluate concrete alternatives and evidence, not a blanket claim that all openness or all protection is beneficial.
Trace gains, losses, and compensation through actual budgets
Countries in the production example are aggregates. They conceal workers, owners, consumers, and public budgets. More consumption feasible for a country does not show that every resident obtains more. Owners of export firms may receive additional profits while workers in an import-competing industry lose employment. Consumers gain only to the extent that lower border costs reach retail prices. Bargaining power and market concentration affect that transmission. These mechanisms require a distributional account within each country, rather than treating a national production possibility as a payment to every individual.
Consider an invented annual accounting case with all amounts in thousands of currency units. Consumers gain purchasing power valued at 120 because imported goods become cheaper. Exporting firms receive an additional 50 after their extra production costs. Workers and owners in the contracting industry lose 90. Assume these amounts are nonoverlapping, measured on a comparable basis, and cover all effects for the moment. Adding gains of 170 and subtracting losses of 90 leaves a net gain of 80. This total establishes room for compensation in the constructed accounts. It does not create a transfer institution or identify the people authorized to collect those gains.
Suppose a feasible compensation program collects 100 from the gaining groups, pays benefits of 90 to the losing group, and uses 10 in administrative resources. The losers are restored to their initial measured resources. Gainers retain 70 of their original 170 gain, while administration absorbs 10. The remaining net benefit is therefore 70 after administration. This conclusion assumes that collection does not alter the measured gains or production choices. If only 40 can actually be collected and paid, the losing group still bears 50 of its loss. Describing potential compensation as if it had occurred would hide precisely that difference.
Potential gains become compensation only if paid
The invented accounts initially give gaining groups 170 and the contracting group a loss of 90. Collecting 100, paying 90, and using 10 for administration leaves gainers with 70 and restores losers to zero. This assumes collection changes no production choices. Without the feasible transfer, the positive aggregate balance would leave the original loss uncompensated.
Even a payment of 90 may not restore every lost benefit. A displaced worker can lose occupation-specific skills, predictable hours, or access to nearby family care. Another may find a better job. Converting those consequences into a single money measure requires evidence and explicit valuation assumptions. Geographic concentration matters because falling local spending can affect businesses and public revenue outside the original industry. Such effects cannot simply be added to the initial loss if they count the same lost income again. A careful study defines the boundary and avoids treating every linked receipt as an independent social cost.
Protection also needs a counterfactual budget. A tariff is a charge on imports, usually remitted by an importer under the applicable rules. That filing obligation does not establish whether consumers, foreign suppliers, domestic firms, or workers ultimately bear it. Substitution, pricing power, and exchange-rate changes can affect incidence. Tariff revenue is a government receipt, not automatically an aggregate resource gain, and higher prices may help one domestic producer while raising another's input costs. Learning or security benefits can be legitimate objects of analysis, but they require specified mechanisms and evidence. Neither a favorable aggregate trade sum nor a claim of strategic importance removes the need to compare feasible alternatives, identify who pays, and state the normative criterion used to evaluate the resulting distribution.
Exercises
Find mutually acceptable terms
Using the example, would one cloth exchanged for 1.5 grain offer both countries a potential gain relative to producing the imported item themselves?
Hint
Compare the trading price with opportunity costs of one and two grain.
Worked answer
Country
Own opportunity cost
Trade terms
Potential gain per cloth
A imports cloth
2 grain
Pays 1.5 grain
Saves 0.5 grain
B exports cloth
1 grain
Receives 1.5 grain
Gains 0.5 grain
The calculation does not identify who within each country receives the gain.
Yes. A gives up 1.5 grain rather than 2 to obtain cloth. B receives 1.5 grain for cloth that costs it 1 grain to produce. Actual gains also depend on feasible quantities and preferences; these terms identify a possible mutually beneficial price.
Gross exports are not domestic contribution
A firm imports components for 70, adds domestic processing, and exports the finished product for 100. What is its domestic value added, ignoring other intermediate inputs?
Hint
Apply the production accounting boundary.
Worked answer
Gross export
Imported inputs
Domestic value added
100
70
30
A large gross export flow can coexist with a smaller domestic production contribution.
Domestic value added is 30, not 100. The gross export value includes imported production of 70. This does not establish how the 30 is split among wages, operating surplus, and other domestic claims.
Sources and further reading
The lessons and answers stand alone. These optional references identify underlying models, evidence, and perspectives. All numerical examples here are invented unless explicitly attributed; none is a forecast or a personal investment recommendation.
An institutional argument for trade, not a neutral estimate of every policy effect. The lesson derives comparative advantage and separately examines distribution, adjustment costs, and strategic constraints.
Definitions of capitalism, production, historical change, technology, and institutions. The course is a mechanism-based introduction, not a claim that one institution alone explains growth.