52.01.03 · economics / microeconomics

The Edgeworth box, Pareto efficiency, and the two welfare theorems

shipped3 tiersLean: none

Anchor (Master): Edgeworth 1881 Mathematical Psychics (Kegan Paul); Pareto 1906 Manual of Political Economy; Arrow-Debreu 1954 Existence of an equilibrium for a competitive economy (Econometrica 22); Debreu 1954 Valuation equilibrium and Pareto optimum (PNAS 38); McKenzie 1959 On the existence of general equilibrium for a competitive market (Econometrica 27)

Intuition Beginner

Two people, Alice and Bob, are stranded on an island. Alice has 100 apples and 0 oranges; Bob has 0 apples and 100 oranges. Neither can eat only one fruit. They will trade. The question is: at what price, and with what final allocation?

Francis Ysidro Edgeworth's 1881 answer draws a box. The width of the box is the total number of apples (100); the height is the total number of oranges (100). Alice's origin is the bottom-left corner; Bob's origin is the top-right. Any point inside the box represents a full allocation of the two goods between them, and the two consumers' preferences are represented by their indifference curves drawn through the box — Alice's curves bow one way, Bob's the opposite.

Trade moves the allocation point inside the box. The set of allocations where neither can be made better off without the other being made worse off is the contract curve — the locus where Alice's and Bob's indifference curves are tangent. The competitive equilibrium, where prices clear both markets, lands on this curve. So does every Pareto-efficient allocation. The deep result, the First Welfare Theorem, runs the other way: every competitive equilibrium is Pareto-efficient. The market, under the right conditions, wastes nothing. The Second Welfare Theorem adds that every Pareto-efficient allocation can be achieved by some competitive equilibrium, provided initial endowments are suitably redistributed — so efficiency and equity are separable.

Visual Beginner

The picture shows the Edgeworth box: a rectangle 100 wide and 100 tall. Alice's origin is bottom-left, Bob's origin is top-right. Alice's indifference curves bow up-and-right from her origin; Bob's bow down-and-left from his. The endowment point $e$ is at the top-left corner (Alice has all the apples, Bob has all the oranges). The contract curve runs from Alice's origin to Bob's origin through the points where the two families of curves are tangent. The competitive equilibrium is marked at a point where a price line through the endowment is tangent to both indifference curves simultaneously.

The tangency of indifference curves at the competitive equilibrium is the geometric content of Pareto efficiency: at the margin, Alice and Bob value the two goods in the same proportion, which is the market price.

Worked example Beginner

Compute the competitive equilibrium for two Cobb-Douglas consumers in an Edgeworth-box economy.

Step 1. Set up. Two goods, apples ( $a$ ) and oranges ( $o$ ). Alice has endowment $ω^{A} = (100, 0)$ and utility $u^{A} (a, o) = a^{1/2} o^{1/2}$ (she weights the two goods equally). Bob has endowment $ω^{B} = (0, 100)$ and the same utility $u^{B} (a, o) = a^{1/2} o^{1/2}$ . Normalize the price of apples to 1, so $p_{a} = 1$ , and let $p_{o} = p$ be the price of oranges.

Step 2. Compute Alice's demand. Alice's income is $p \cdot ω^{A} = 1 \times 100 + p \times 0 = 100$ . Cobb-Douglas demand with equal weights spends half the income on each good: $a^{A} = 50$ and $o^{A} = 50/ p$ .

Step 3. Compute Bob's demand. Bob's income is $1 \times 0 + p \times 100 = 100 p$ . So $a^{B} = 50 p$ and $o^{B} = 50$ .

Step 4. Clear the apple market. Total apple supply is 100; total demand is $a^{A} + a^{B} = 50 + 50 p$ . Setting demand equal to supply: $50 + 50 p = 100$ , giving $p = 1$ . The competitive-equilibrium price of oranges is 1 apple.

Step 5. Verify the orange market. Demand is $50/1 + 50 = 100$ , matching supply. Alice ends with $(50, 50)$ ; Bob ends with $(50, 50)$ . Each is at the symmetric point of equal consumption.

What this tells us: with identical Cobb-Douglas preferences and equal-but-opposite endowments, the competitive equilibrium splits the goods equally at a price of 1. The endowment inequality washes out through trade, and the final allocation is Pareto-efficient.

Check your understanding Beginner

Exercise (easy, multiple choice).

The contract curve in the Edgeworth box is the set of allocations where:

A. Alice and Bob have equal utility B. The two consumers' indifference curves are tangent C. Total output is maximised D. Prices are zero

Hint

At a tangency, the marginal rates of substitution are equal, so no mutually beneficial trade remains — the definition of Pareto efficiency in the box.

Answer

B. The contract curve is the set of allocations in the Edgeworth box at which Alice's and Bob's indifference curves are tangent, equivalently at which their marginal rates of substitution between the two goods are equal. At any tangency, no Pareto-improving trade exists (the indifference curves through the point do not enclose a "lens" of mutual improvement), so the contract curve is exactly the set of Pareto-efficient allocations in the box.

Formal definition Intermediate+

A pure-exchange economy is a measure space of consumers with endowments and preferences, together with a price vector at which all markets clear. Its core notions are the allocation, the Pareto order, the Walrasian equilibrium, and the two welfare theorems.

Definition (exchange economy). A pure-exchange economy is a tuple $(I, (ω^{i}, u^{i})_{i \in I})$ where $I$ is a finite set of consumers, $ω^{i} \in R_{+}^{n}$ is consumer $i$ 's endowment of the $n$ goods, and $u^{i} : R_{+}^{n} \to R$ is consumer $i$ 's utility function. An allocation is a list $x = (x^{i})_{i \in I}$ with $x^{i} \in R_{+}^{n}$ and $\sum_{i} x^{i} = \sum_{i} ω^{i}$ (the allocation is feasible).

Definition (Pareto efficiency). An allocation $x$ is Pareto-efficient if there is no other feasible allocation $y$ with $u^{i} (y^{i}) \geq u^{i} (x^{i})$ for all $i$ and $u^{j} (y^{j}) > u^{j} (x^{j})$ for some $j$ . The set of Pareto-efficient allocations is the Pareto set.

Definition (Walrasian equilibrium). A Walrasian (competitive) equilibrium of the exchange economy is a pair $(p^{*}, x^{*})$ with $p^{*} \in R_{+}^{n} ∖ {0}$ and $x^{*}$ a feasible allocation such that each $x^{* i}$ maximises $u^{i}$ on the budget set ${x : p^{*} \cdot x \leq p^{*} \cdot ω^{i}}$ .

Counterexamples to common slips Intermediate+

Conflating Pareto efficiency with equality. Pareto efficiency is silent on distribution. An allocation giving all goods to one consumer can be Pareto-efficient if the other consumer's utility cannot be raised without lowering the first's. Equality is a separate normative criterion.
Treating the contract curve as a single point. The contract curve is a continuum of allocations, not a single competitive outcome. Which point on the curve the market reaches depends on initial endowments and the resulting equilibrium prices.
Assuming the Second Welfare Theorem requires no redistribution. The Second Welfare Theorem says every Pareto-efficient allocation can be supported as a competitive equilibrium after a suitable lump-sum redistribution of endowments. Without the redistribution, the equilibrium generally lands elsewhere on the contract curve.

Economic theory Intermediate+

Theorem (First Welfare Theorem). If preferences are locally non-satiated, then every Walrasian-equilibrium allocation is Pareto-efficient.

Proof. Let $(p^{*}, x^{*})$ be a Walrasian equilibrium and suppose for contradiction that some feasible allocation $y$ Pareto-dominates $x^{*}$ : $u^{i} (y^{i}) \geq u^{i} (x^{* i})$ for all $i$ and $u^{j} (y^{j}) > u^{j} (x^{* j})$ for some $j$ . By local non-satiation at each $x^{* i}$ , there is a nearby bundle $z^{i}$ with $u^{i} (z^{i}) > u^{i} (x^{* i})$ and $p^{*} \cdot z^{i}$ arbitrarily close to $p^{*} \cdot x^{* i}$ . Since $x^{* i}$ was chosen at budget $p^{*} \cdot ω^{i}$ , any $z^{i}$ with $u^{i} (z^{i}) > u^{i} (x^{* i})$ must have $p^{*} \cdot z^{i} > p^{*} \cdot ω^{i}$ (otherwise $x^{* i}$ was not utility-maximising). So $p^{*} \cdot y^{i} \geq p^{*} \cdot ω^{i}$ for all $i$ and $p^{*} \cdot y^{j} > p^{*} \cdot ω^{j}$ for the $j$ where utility strictly rose. Summing: $p^{*} \cdot \sum_{i} y^{i} > p^{*} \cdot \sum_{i} ω^{i}$ . But feasibility of $y$ requires $\sum_{i} y^{i} = \sum_{i} ω^{i}$ , contradicting the strict inequality. So no Pareto improvement exists. $□$

Theorem (Second Welfare Theorem). If preferences are convex and continuous, then every Pareto-efficient allocation $x^{*}$ can be supported as a Walrasian equilibrium after a suitable lump-sum redistribution of endowments. That is, there exists a price vector $p^{*} \neq = 0$ such that each $x^{* i}$ minimises expenditure $p^{*} \cdot x$ over the set ${x : u^{i} (x) \geq u^{i} (x^{* i})}$ .

Reconstruction. The proof uses the separating hyperplane theorem. The set of aggregate net trades that would improve everyone's utility is convex (by convexity of preferences) and does not intersect the non-positive orthant (by Pareto efficiency). The separating hyperplane between these two convex sets is the price vector $p^{*}$ ; the supporting-hyperplane property gives the expenditure-minimisation statement. The redistribution of endowments moves the budget hyperplane through $x^{* i}$ for each $i$ .

Bridge. The two welfare theorems build toward 52.04.01 game theory by establishing the formal sense in which competitive equilibrium is "optimal" — the analogue in strategic settings is the efficiency of Nash equilibrium in specific game classes — and appear again in 52.01.01 microeconomics foundations as the rigorous content of the invisible-hand intuition. The foundational reason the First Welfare Theorem holds is that the budget-set constraint at equilibrium prices rules out Pareto-improving trades: any allocation preferred by everyone would cost more than the aggregate endowment at the equilibrium prices, which is infeasible. This is exactly the structure that identifies a competitive equilibrium with a Pareto optimum via price-mediated separation, and the bridge is from decentralised price-taking behaviour to social efficiency, with the budget hyperplane as the separator. The pattern generalises across the modern general-equilibrium programme — the existence proofs (Arrow-Debreu, McKenzie), the core-convergence theorem (Debreu-Scarf), the characterisation of incomplete markets — each of which treats competitive equilibrium as the decentralised institution that implements Pareto efficiency under prices, and the central insight is that efficiency and equity are separable: the market solves the efficiency problem given any endowment distribution, and redistribution solves the equity problem separately.

Exercises Intermediate+

Exercise 1 (easy, multiple choice).

The First Welfare Theorem requires which key assumption?

A. Identical preferences for all consumers B. Local non-satiation (more is always better nearby) C. Concave utility functions D. Equal endowments

Hint

The proof uses local non-satiation to ensure that a Pareto-preferred bundle would cost strictly more at the equilibrium prices.

Answer

B. The First Welfare Theorem requires local non-satiation: for every bundle, there is a nearby bundle that the consumer strictly prefers. The assumption ensures that any Pareto-improving reallocation would cost strictly more at equilibrium prices than the aggregate endowment, contradicting feasibility. Without local non-satiation (e.g., consumers fully satiated at their equilibrium bundle), Pareto-improving moves might not be available at any price, and the theorem can fail.

Exercise 3 (medium, symbolic).

Show that at any Pareto-efficient interior allocation in the Edgeworth box with smooth preferences, the two consumers' marginal rates of substitution between the two goods are equal. State the converse and identify the additional condition needed.

Hint

At an interior Pareto optimum, no Pareto-improving trade exists. If the MRSs differed, the consumers could trade at an intermediate ratio and both gain.

Answer

At an interior Pareto-efficient allocation $x^{*}$ , suppose for contradiction that Alice's MRS ( $MRS^{A} = (\partial u^{A} / \partial a) / (\partial u^{A} / \partial o)$ ) differs from Bob's. Then Alice values apples more than Bob does at the margin (in one direction), so Alice could give up some oranges for Bob's apples at a ratio strictly between the two MRSs, leaving both strictly better off — a Pareto improvement, contradicting efficiency. So $MRS^{A} = MRS^{B}$ at any interior Pareto optimum. The converse requires convexity of preferences (diminishing MRS): equal MRS guarantees efficiency only when preferences are convex, so that the indifference curves do not re-cross. With non-convex preferences, an MRS-equal allocation can be Pareto-dominated.

Exercise 4 (medium, numeric).

In a two-consumer Cobb-Douglas economy with Alice endowment $(50, 0)$ , Bob endowment $(0, 50)$ , both with utility $u = a^{1/2} o^{1/2}$ , find the equilibrium price ratio $p_{o} / p_{a}$ using the apple-market-clearing condition.

Hint

Cobb-Douglas demand: Alice spends half her income on apples, half on oranges. Alice's income $= 50 p_{a} + 0 p_{o}$ . Apple demand $= 25 p_{a} / p_{a} = 25$ ... wait, recompute carefully. Alice's apple demand is (income / 2) / $p_{a}$ .

Answer

1. Normalize $p_{a} = 1$ . Alice's income $= 50 \cdot 1 + 0 \cdot p_{o} = 50$ . Her apple demand is $(50/2) /1 = 25$ . Bob's income $= 0 \cdot 1 + 50 \cdot p_{o} = 50 p_{o}$ . His apple demand is $(50 p_{o} /2) /1 = 25 p_{o}$ . Apple market clearing: $25 + 25 p_{o} = 50$ (total apples), giving $p_{o} = 1$ . The equilibrium price ratio is $p_{o} / p_{a} = 1$ . Final allocations: Alice $(25, 25)$ , Bob $(25, 25)$ — symmetric to the worked example.

Exercise 5 (medium, multiple choice).

The Second Welfare Theorem differs from the First in that it:

A. Claims competitive equilibrium is efficient B. Claims every efficient allocation can be achieved as a competitive equilibrium after a suitable redistribution of endowments C. Refutes the First Welfare Theorem D. Applies only to single-consumer economies

Hint

The Second Welfare Theorem addresses the reverse direction (efficiency $\Rightarrow$ market-supportable) and adds that redistribution is the lever for equity.

Answer

B. The Second Welfare Theorem states that every Pareto-efficient allocation can be supported as a competitive equilibrium, provided that initial endowments are suitably redistributed by lump-sum transfers before the market operates. The theorem separates efficiency and equity: the market solves the efficiency problem given any endowment distribution, and redistribution solves the equity problem separately. The theorem requires convex preferences and continuity.

Exercise 6 (medium, multiple choice).

The core of an exchange economy is the set of allocations that:

A. Maximize total utility B. No coalition of consumers can block (improve upon for all its members using only its own endowments) C. Are reached by competitive equilibrium D. Are Pareto-efficient and equal-utility

Hint

The core is a cooperative-game concept: an allocation is in the core if no subset (coalition) of consumers could do better on its own, using only its members' endowments. The core refines Pareto efficiency.

Answer

B. The core of an exchange economy is the set of feasible allocations that no coalition of consumers can improve upon — that is, no subset of consumers could reallocate their own endowments among themselves to make every coalition member strictly better off. The core refines Pareto efficiency (which considers only the grand coalition of all consumers). The Debreu-Scarf core-convergence theorem shows that as the economy is replicated (each type of consumer appears in many copies), the core shrinks to the set of competitive equilibria.

Exercise 7 (hard, symbolic).

Prove the core-convergence theorem (Debreu-Scarf) in outline: in an exchange economy with $n$ types of consumer, replicated $k$ times for each type, the core shrinks toward the set of Walrasian-equilibrium allocations as $k \to \infty$ .

Hint

A Walrasian-equilibrium allocation is in the core for every $k$ (easy direction). For the converse, take a non-equilibrium allocation and construct a blocking coalition of "winners" and one "loser" from each type who would deviate.

Answer

Sketch: Walrasian-equilibrium allocations are always in the core (by the First Welfare Theorem applied to any coalition's sub-economy). For the converse, suppose $x^{*}$ is a core allocation in the $k$ -fold replicated economy that is not a Walrasian-equilibrium allocation. Then for some consumer type $i$ , the bundle $x^{* i}$ is not expenditure-minimising at any candidate price. Construct a coalition consisting of one consumer of each type, with the type- $i$ member receiving a slightly better bundle and every other type receiving the same bundle as at $x^{*}$ . By the equal-treatment property (core allocations treat same-type consumers identically), this coalition can be assembled from the $k$ copies and can improve on $x^{*}$ using only its members' endowments, contradicting the core property. The contradiction forces $x^{*}$ to be a Walrasian equilibrium. So the core converges to the set of Walrasian-equilibrium allocations as $k \to \infty$ . $□$

Exercise 8 (hard, multiple choice).

Why does the Second Welfare Theorem fail in economies with non-convex preferences (e.g., fixed costs, increasing returns to scale)?

A. Non-convex preferences make consumers irrational B. With non-convexities, the separating hyperplane that supports a Pareto optimum may not exist, so no set of prices can support the allocation as a competitive equilibrium C. Non-convex preferences violate the First Welfare Theorem D. Non-convexity makes the economy unstable

Hint

The Second Welfare Theorem's proof relies on convexity to apply the separating hyperplane theorem. Non-convex preferences can produce Pareto optima at corners or isolated points where no tangent price line exists.

Answer

B. The Second Welfare Theorem's proof uses the separating hyperplane theorem to construct supporting prices at a Pareto optimum. The theorem requires convexity of preferences (and production sets) to ensure the convex-set separation applies. With non-convexities (fixed costs, increasing returns, indivisibilities), the Pareto optimum may lie at a point where no tangent price line supports the preferred set, so no price vector can decentralise the allocation as a competitive equilibrium. The First Welfare Theorem does not require convexity, but the Second does. Non-convexities are a major reason competitive markets may not implement efficient outcomes (natural monopoly is the canonical example).

Advanced results Master

Result 1 (Edgeworth box, 1881). Francis Ysidro Edgeworth introduced the box diagram as a geometrical representation of a two-consumer, two-good exchange economy, with the contract curve as the set of tangency points of indifference curves. Edgeworth also defined the core of an exchange economy and conjectured its convergence to the competitive equilibrium as the economy grows; the conjecture was proved rigorously by Debreu and Scarf in 1963.

Result 2 (Pareto efficiency, 1906). Vilfredo Pareto defined the efficiency criterion that bears his name: an allocation is Pareto-efficient if no reallocation can make anyone better off without making someone else worse off. The criterion replaced the earlier utilitarian aggregation (which required interpersonally comparable cardinal utility) with an ordinal criterion that is well-defined without comparison across people. The criterion is the standard efficiency benchmark of modern welfare economics.

Result 3 (Arrow-Debreu existence theorem, 1954). Kenneth Arrow and Gerard Debreu proved that under standard conditions (continuous preferences, strictly positive endowments, compact consumption sets), a Walrasian equilibrium exists. The proof uses Kakutani's fixed-point theorem on the aggregate excess-demand correspondence defined over the price simplex. The result closed a half-century of informal argument about whether market-clearing prices always exist.

Result 4 (First Welfare Theorem, Arrow 1951). Every Walrasian-equilibrium allocation is Pareto-efficient, under the assumption of local non-satiation. The theorem formalises Adam Smith's invisible-hand intuition: competitive markets allocate resources without waste.

Result 5 (Second Welfare Theorem, Arrow 1951, Debreu 1954). Under convexity and continuity of preferences, every Pareto-efficient allocation can be supported as a Walrasian equilibrium after a suitable lump-sum redistribution of endowments. The theorem establishes the separability of efficiency and equity: the market solves the efficiency problem given any endowment distribution, and redistribution solves the equity problem.

Result 6 (Debreu-Scarf core convergence, 1963). As an exchange economy is replicated (each type of consumer appearing in many identical copies), the core shrinks monotonically and converges to the set of Walrasian-equilibrium allocations in the limit. The theorem formalises Edgeworth's 1881 conjecture and provides a cooperative-game foundation for the competitive-equilibrium concept: in large economies, no coalition can improve on the competitive outcome.

Synthesis. The Edgeworth-box theory of exchange builds toward 52.04.01 game theory by establishing the formal sense in which competitive equilibrium is "optimal" — the cooperative-game analogue is the core, and the equivalence emerges in large economies — and appears again in 52.01.01 microeconomics foundations as the rigorous content of the invisible-hand intuition. The foundational reason the welfare theorems hold is the price-mediated separation of the budget set from the Pareto-improving set, which uses only local non-satiation (First) and convexity (Second). This is exactly the structure that identifies a competitive equilibrium with a Pareto optimum via the supporting-price hyperplane, and putting these together with the Arrow-Debreu existence proof and the Debreu-Scarf core convergence, the bridge is from decentralised price-taking behaviour to social efficiency, with the price vector as the separator and the lump-sum redistribution as the equity lever. The pattern generalises across the modern general-equilibrium programme — incomplete-markets theory (Radner, Geanakoplos), sunspot equilibria (Cass-Shell), the Modigliani-Miller theorem in production — each of which extends the welfare-theorem logic to settings with uncertainty, time, or production, and the central insight is that the competitive market is the decentralised institution that implements Pareto efficiency under prices, with redistribution as the separate equity instrument.

Full proof set Master

Proposition (First Welfare Theorem). Under local non-satiation, every Walrasian-equilibrium allocation is Pareto-efficient.

Proof. Suppose $(p^{*}, x^{*})$ is a Walrasian equilibrium and $y$ is a feasible allocation that Pareto-dominates $x^{*}$ . For each $i$ with $u^{i} (y^{i}) \geq u^{i} (x^{* i})$ , local non-satiation gives a nearby $z^{i}$ with $u^{i} (z^{i}) > u^{i} (x^{* i})$ and $p^{*} \cdot z^{i}$ arbitrarily close to $p^{*} \cdot y^{i}$ . Since $x^{* i}$ was utility-maximising on the budget set ${x : p^{*} \cdot x \leq p^{*} \cdot ω^{i}}$ , any strictly preferred $z^{i}$ must cost more: $p^{*} \cdot z^{i} > p^{*} \cdot ω^{i}$ . By continuity, $p^{*} \cdot y^{i} \geq p^{*} \cdot ω^{i}$ . For the $j$ with $u^{j} (y^{j}) > u^{j} (x^{* j})$ , the same argument gives $p^{*} \cdot y^{j} > p^{*} \cdot ω^{j}$ . Summing: $p^{*} \cdot \sum_{i} y^{i} > p^{*} \cdot \sum_{i} ω^{i}$ . But feasibility of $y$ gives $\sum_{i} y^{i} = \sum_{i} ω^{i}$ , contradicting the strict inequality. So $x^{*}$ is Pareto-efficient. $□$

Proposition (Second Welfare Theorem). Under convexity and continuity of preferences, every Pareto-efficient allocation can be supported as a Walrasian equilibrium after a suitable lump-sum redistribution.

Proof sketch. Let $x^{*}$ be Pareto-efficient. Define the aggregate preferred set $P = \sum_{i} {y^{i} : u^{i} (y^{i}) > u^{i} (x^{* i})}$ . By convexity of each preferred set, $P$ is convex. By Pareto efficiency, $P$ does not intersect the non-positive orthant ${z : z \leq 0}$ (no aggregate destruction can make everyone strictly better off). The separating hyperplane theorem gives a non-zero $p^{*}$ such that $p^{*} \cdot z \geq p^{*} \cdot 0 = 0$ for all $z \in P$ , with strict inequality on the interior. This $p^{*}$ is the supporting price vector. Each $x^{* i}$ minimises expenditure $p^{*} \cdot x$ over ${x : u^{i} (x) \geq u^{i} (x^{* i})}$ (by the supporting-hyperplane property applied to consumer $i$ 's preferred set). Setting each consumer's endowment to $x^{* i}$ (a lump-sum redistribution) makes the budget hyperplane pass through $x^{* i}$ , so $x^{*}$ is the competitive outcome at prices $p^{*}$ . $□$

Connections Master

Microeconomics — scarcity, choice, and equilibrium 52.01.01. The Edgeworth-box theory of exchange is the rigorous foundation of the supply-and-demand intuition in the foundations unit. The two welfare theorems formalise Adam Smith's invisible-hand intuition and define the conditions under which it holds.
Game theory — Nash equilibrium and strategic interaction 52.04.01. The core of an exchange economy is a cooperative-game concept, and the Debreu-Scarf core-convergence theorem establishes the equivalence of competitive equilibrium and the core in large economies. The bridge between cooperative and non-cooperative game theory runs through the core.
Revealed preference and Afriat's theorem 52.01.02. Both revealed-preference theory and welfare theory extract structural conclusions (rationalisability, efficiency) from observable market data (choices, allocations). The shared methodology is the inference of the unobservable from the consistency of the observable.

Historical & philosophical context Master

Francis Ysidro Edgeworth's Mathematical Psychics ^{[Edgeworth1881]} of 1881 introduced the box diagram (now called the Edgeworth box), the contract curve, and the core of an exchange economy. Edgeworth's central conjecture — that the core shrinks toward the competitive equilibrium as the economy grows — was proved rigorously by Gerard Debreu and Herbert Scarf in 1963. Vilfredo Pareto's Manual of Political Economy ^[Pareto1906] of 1906 defined the efficiency criterion that bears his name and replaced the utilitarian aggregation with an ordinal criterion.

The modern synthesis is due to Kenneth Arrow and Gerard Debreu. Arrow's 1951 paper "An Extension of the Basic Theorems of Classical Welfare Economics" ^[Arrow1951] gave the modern proofs of the two welfare theorems. The Arrow-Debreu 1954 Econometrica paper ^{[ArrowDebreu1954]} proved the existence of Walrasian equilibrium via Kakutani's fixed-point theorem. Lionel McKenzie gave an independent proof in 1959 ^{[McKenzie1959]} via Brouwer's fixed-point theorem. Debreu's 1954 PNAS paper ^[Debreu1954] gave the modern proof of the Second Welfare Theorem via the supporting-hyperplane construction. Arrow and Debreu shared the 1972 Nobel Prize in Economics. The deeper lineage runs through Leon Walras's Elements of Pure Economics (1874), which first formulated general equilibrium as a system of simultaneous equations, and through John Hicks's Value and Capital (1939), which reworked the theory in modern ordinal-utility terms.

Bibliography Master

@book{Edgeworth1881,
  author = {Edgeworth, F. Y.},
  title = {Mathematical Psychics: An Essay on the Application of Mathematics to the Moral Sciences},
  publisher = {C. Kegan Paul \& Co.},
  address = {London},
  year = {1881},
}

@book{Pareto1906,
  author = {Pareto, V.},
  title = {Manual of Political Economy},
  publisher = {Macmillan},
  address = {London},
  year = {1906 (English translation 1971, A. M. Kelley)},
}

@article{ArrowDebreu1954,
  author = {Arrow, K. J. and Debreu, G.},
  title = {Existence of an equilibrium for a competitive economy},
  journal = {Econometrica},
  volume = {22},
  number = {3},
  pages = {265--290},
  year = {1954},
}

@article{Debreu1954,
  author = {Debreu, G.},
  title = {Valuation equilibrium and {Pareto} optimum},
  journal = {Proceedings of the National Academy of Sciences},
  volume = {40},
  number = {7},
  pages = {588--592},
  year = {1954},
}

@article{McKenzie1959,
  author = {McKenzie, L. W.},
  title = {On the existence of general equilibrium for a competitive market},
  journal = {Econometrica},
  volume = {27},
  number = {1},
  pages = {54--71},
  year = {1959},
}

@article{DebreuScarf1963,
  author = {Debreu, G. and Scarf, H.},
  title = {A limit theorem on the core of an economy},
  journal = {International Economic Review},
  volume = {4},
  number = {3},
  pages = {235--246},
  year = {1963},
}

@book{MasColellWhinstonGreen1995,
  author = {Mas-Colell, A. and Whinston, M. D. and Green, J. R.},
  title = {Microeconomic Theory},
  publisher = {Oxford University Press},
  address = {New York},
  year = {1995},
}

Prerequisites

52.01.01
52.01.02

Tier anchors

beginner: Varian — Intermediate Microeconomics (Norton 2014) Ch. 31 on exchange and the Edgeworth box; Nicholson-Snyder — Intermediate Microeconomics (Cengage 2016) Ch. 15
intermediate: Mas-Colell-Whinston-Green — Microeconomic Theory (Oxford 1995) Ch. 15-16 on general equilibrium and welfare; Stiglitz-Walsh — Economics (Norton 2006) Ch. 16
master: Edgeworth 1881 Mathematical Psychics (Kegan Paul); Pareto 1906 Manual of Political Economy; Arrow-Debreu 1954 Existence of an equilibrium for a competitive economy (Econometrica 22); Debreu 1954 Valuation equilibrium and Pareto optimum (PNAS 38); McKenzie 1959 On the existence of general equilibrium for a competitive market (Econometrica 27)

References

Edgeworth — Mathematical Psychics: An Essay on the Application of Mathematics to the Moral Sciences · Kegan Paul, London 1881; originator of the Edgeworth box, the contract curve, and the core convergence theorem
Arrow and Debreu — Existence of an equilibrium for a competitive economy · Econometrica 22, 265-290 (1954); originators of the modern existence proof of Walrasian equilibrium via Kakutani's fixed point theorem
Debreu — Valuation equilibrium and Pareto optimum · Proceedings of the National Academy of Sciences 40, 588-592 (1954); originator of the modern proof of the two welfare theorems
Mas-Colell, Whinston, and Green — Microeconomic Theory · Oxford 1995, Ch. 15 (general equilibrium), Ch. 16 (welfare), Ch. 17 (positive theory)
Pareto — Manual of Political Economy · Macmillan 1906; originator of the Pareto efficiency criterion
McKenzie — On the existence of general equilibrium for a competitive market · Econometrica 27, 54-71 (1959); independent originator of the existence proof via Brouwer's fixed point theorem

Estimated time

beginner: 15m
intermediate: 40m
master: 80m