Week 5 · Dr. Boaz Zik

General Equilibrium Production Economy & Externalities

General Equilibrium Production Economy & Externalities

Part of: Micro-Economics II | Key concepts: Production Economy, Efficiency Locus, PPF, RPT, Second Welfare Theorem, Externalities, Pigouvian Tax, Coase Theorem, Market Failure


What This Presentation Is About

This note picks up from Lec_04-Theory of the Firm, Partial & General Equilibrium, which established the First Welfare Theorem in an exchange economy. Here we:

  1. Complete the welfare theorems — the Second Welfare Theorem and why lump-sum taxes dominate distortionary taxes.
  2. Extend general equilibrium to production — two firms produce two goods using two inputs. We find what "efficiency" means, how the Production Possibility Frontier is derived, and why competitive markets attain the efficient outcome (MRS = RPT).
  3. Analyse externalities — one of the main reasons markets fail to be efficient. We cover consumption externalities, production externalities, the Pigouvian tax, and the Coase theorem.

Part 1 — Completing the Welfare Theorems

Recap of the First Welfare Theorem

In a competitive exchange economy, the equilibrium allocation is Pareto efficient. This is remarkable: no government coordination is required — free agents pursuing their own interests, guided only by prices, achieve a socially efficient outcome.

The Second Welfare Theorem

Second Welfare Theorem

In a competitive free market, any Pareto efficient allocation can be achieved as a competitive equilibrium, provided the right initial endowments are chosen.

What this tells us about redistribution: If society wants a particular "fair" allocation — not just any efficient one — it can achieve it by redistributing endowments and then letting markets work. This is better, in principle, than direct intervention in prices (subsidies, price controls), because:

  • Distortionary taxes/price controls alter agents' incentive margins, potentially pushing the economy off the Pareto frontier.
  • A lump-sum tax (a tax independent of the agent's choices) does not distort margins — the agent still faces the same relative prices and will trade to an efficient outcome.
The practical caveat

The Second Welfare Theorem is partly hypothetical. To implement the "right" endowment, the government would need complete information about preferences and production functions — computationally impossible. In practice, taxes always create some distortion, so there is an unavoidable equity–efficiency tradeoff: redistributing income for fairness costs efficiency.


Part 2 — General Equilibrium: Production Economy

Setting Up the Economy

We now consider an economy with production:

  • Two goods: XX and YY.
  • Two inputs: labour LL and capital KK (fixed quantities Lˉ\bar{L} and Kˉ\bar{K}).
  • Two firms: Firm 1 produces XX with f1(k1,l1)f_1(k_1, l_1); Firm 2 produces YY with f2(k2,l2)f_2(k_2, l_2).
  • A representative consumer with utility U(x,y)U(x, y) who owns all inputs and firm profits.

The feasibility constraint is k1+k2=Kˉk_1 + k_2 = \bar{K} and l1+l2=Lˉl_1 + l_2 = \bar{L}.


The Edgeworth Box for Production

Just as in the exchange economy we used an Edgeworth box for consumption bundles, we use one here for input allocation:

  • Horizontal axis = total labour Lˉ\bar{L}.
  • Vertical axis = total capital Kˉ\bar{K}.
  • Origin O1O_1 (bottom-left): Firm 1's input allocation.
  • Origin O2O_2 (top-right, flipped): Firm 2's input allocation.
  • Each point in the box is a feasible (k1,l1,k2,l2)(k_1, l_1, k_2, l_2) allocation.

Firm 1's isoquants run from O1O_1 in the standard orientation. Firm 2's isoquants are flipped around O2O_2 and overlaid on the box.

Efficiency in Production

An input allocation is Pareto efficient if there is no reallocation that strictly increases one firm's output without reducing the other's. Graphically, this occurs precisely where the two firms' isoquants are tangent.

At a tangency, the firms have the same Rate of Technical Substitution:

RTS1(l1,k1)=RTS2(l2,k2)\boxed{RTS_1(l_1, k_1) = RTS_2(l_2, k_2)}

The ==efficiency locus== (also called the Pareto set) is the set of all Pareto efficient input allocations — all points where the isoquants are tangent.


From Efficiency Locus to the PPF

Each point on the efficiency locus corresponds to a specific (x,y)(x, y) pair (the outputs of the two firms). Tracing out all these pairs gives the ==Production Possibility Frontier (PPF)==.

Production Possibility Frontier (PPF)

The PPF shows all combinations of (x,y)(x, y) that can be produced when the economy uses its inputs efficiently. Points inside the frontier are feasible but inefficient; points outside are infeasible.

Shape: We assume the PPF is concave (bowed out), because of diminishing marginal productivity — transferring inputs from YY to XX yields diminishing returns as XX grows.


Rate of Product Transformation (RPT)

Rate of Product Transformation (RPT)

The RPT at a point (x,y)(x, y) on the PPF is the number of units of YY that must be sacrificed to produce one more unit of XX while remaining on the frontier: RPT(x,y)=−dydx∣PPFRPT(x, y) = -\frac{dy}{dx}\bigg|_{\text{PPF}} We assume RPT is increasing in xx (the PPF gets steeper as more XX is produced).

Why RPT increases

As the economy shifts production toward XX, the inputs employed in XX have lower marginal products (diminishing returns), while the inputs released from YY are the least productive ones first. The opportunity cost of an extra unit of XX therefore rises — RPT increases.


Optimal Production: MRS = RPT

The representative consumer maximises U(x,y)U(x, y) subject to the feasibility constraint that (x,y)(x, y) lies on (or below) the PPF. Since UU is increasing, the optimum is on the PPF, and:

Efficiency condition for a production economy

The socially optimal allocation (x∗,y∗)(x^*, y^*) is the point on the PPF where: MRS(x∗,y∗)=RPT(x∗,y∗)\boxed{MRS(x^*, y^*) = RPT(x^*, y^*)}

The MRS (consumer side) equals the RPT (production side).

Intuition: MRS tells us how much YY the consumer is willing to give up for one more unit of XX (without losing utility). RPT tells us how much YY must be sacrificed in production to get one more XX. When MRS > RPT, shifting production toward XX makes the consumer better off (they value the extra XX more than it costs). When MRS < RPT, shift back. At MRS = RPT, no reallocation can improve welfare.

Scenario Implication Direction to move
MRS>RPTMRS > RPT Consumer values XX more than it costs to produce Increase XX
MRS<RPTMRS < RPT Producing XX costs more than consumer values it Decrease XX
MRS=RPTMRS = RPT Optimal — cannot improve welfare Stay here

Equilibrium in a Production Economy

In a competitive equilibrium:

  1. Firms minimise costs → each firm's RTS equals the input price ratio pl/pkp_l/p_k. Since both firms face the same input prices, both have the same RTS → they operate on the efficiency locus → production is on the PPF. ✓

  2. Key result — RPT equals cost ratio: Along the PPF, total cost C=plLˉ+pkKˉC = p_l \bar{L} + p_k \bar{K} is fixed. At any point (x,y)(x, y), this means C(x)+C(y)=constC(x) + C(y) = \text{const}. Differentiating, the slope of the iso-cost locus is:

RPT(x,y)=C′(x)C′(y)RPT(x,y) = \frac{C'(x)}{C'(y)}
  1. Firms price at marginal cost (perfect competition) → px∗=C′(x∗)p_x^* = C'(x^*) and py∗=C′(y∗)p_y^* = C'(y^*). Therefore:
RPT(x∗,y∗)=px∗py∗\boxed{RPT(x^*, y^*) = \frac{p_x^*}{p_y^*}}
  1. Consumer maximises utility → sets MRS(x∗,y∗)=px∗/py∗MRS(x^*, y^*) = p_x^*/p_y^*.

Combining: MRS=px∗/py∗=RPTMRS = p_x^*/p_y^* = RPT — the competitive equilibrium achieves the socially optimal outcome. This is the Third Welfare Theorem result for a production economy (an extension of Theorem 1).

How equilibrium prices are found in practice

To find the full equilibrium:

  1. Derive the efficiency locus (set RTS1=RTS2RTS_1 = RTS_2 using resource constraints).
  2. Derive the PPF.
  3. Find (x∗,y∗)(x^*, y^*) by solving MRS=RPTMRS = RPT.
  4. Normalise one output price (e.g., py=1p_y = 1), set px=MRS(x∗,y∗)p_x = MRS(x^*, y^*).
  5. Use the cost function and px∗=C′(x∗)p_x^* = C'(x^*) to find input prices.
  6. Verify the budget constraint holds: pxx∗+pyy∗=plLˉ+pkKˉ+π1∗+π2∗p_x x^* + p_y y^* = p_l \bar{L} + p_k \bar{K} + \pi_1^* + \pi_2^*.

Worked Examples

Example 1 — Linear PPF

Economy:

f1(k1,l1)=l11/4k13/4,f2(k2,l2)=2l21/4k23/4f_1(k_1,l_1) = l_1^{1/4} k_1^{3/4}, \quad f_2(k_2,l_2) = 2 l_2^{1/4} k_2^{3/4}
Kˉ=Lˉ=50,U(x,y)=x1/2y1/2\bar{K} = \bar{L} = 50, \quad U(x,y) = x^{1/2} y^{1/2}
Step-by-step solution

Efficiency locus: RTS1=k1/(3l1)RTS_1 = k_1/(3l_1), RTS2=k2/(3l2)RTS_2 = k_2/(3l_2). Setting equal with k2=50−k1k_2 = 50-k_1, l2=50−l1l_2=50-l_1 gives k1=l1k_1 = l_1.

PPF: Substituting k1=l1k_1=l_1 into f1f_1: x=l1x = l_1. For YY: y=100−2l1=100−2xy = 100 - 2l_1 = 100 - 2x. So the PPF is 2x+y=1002x + y = 100 — a straight line, with RPT=2RPT = 2.

Optimal point: MRS=y/xMRS = y/x. Set y/x=2y/x = 2 so y=2xy = 2x. Substituting into PPF: 4x=100⇒x∗=25,y∗=504x = 100 \Rightarrow x^* = 25, y^* = 50.

Example 2 — Quadratic PPF

Economy:

f1(k1,l1)=l1k1,f2(k2,l2)=l2k2f_1(k_1,l_1) = l_1 k_1, \quad f_2(k_2,l_2) = l_2 k_2
Kˉ=Lˉ=50,U(x,y)=x1/4y3/4\bar{K} = \bar{L} = 50, \quad U(x,y) = x^{1/4} y^{3/4}
Step-by-step solution

Efficiency locus: RTS1=k1/l1=k2/l2=RTS2⇒k1=l1RTS_1 = k_1/l_1 = k_2/l_2 = RTS_2 \Rightarrow k_1 = l_1.

PPF: x=l12x = l_1^2, so l1=xl_1 = \sqrt{x}. Then y=(50−l1)2=(50−x)2y = (50 - l_1)^2 = (50-\sqrt{x})^2. PPF: y+100x−x=2500y + 100\sqrt{x} - x = 2500.

RPT: RPT=50x−1/2−1RPT = 50x^{-1/2} - 1.

Optimal point: MRS=y/(3x)MRS = y/(3x). Setting MRS=RPTMRS = RPT: y=150x−3xy = 150\sqrt{x} - 3x. Substituting into PPF and solving: x∗=625/4x^* = 625/4, y∗=5625/4y^* = 5625/4.

Example 3 — Symmetric PPF

Economy:

f1(k1,l1)=12(l1+k1),f2(k2,l2)=12(l2+k2)f_1(k_1,l_1) = \tfrac{1}{2}(\sqrt{l_1}+\sqrt{k_1}), \quad f_2(k_2,l_2) = \tfrac{1}{2}(\sqrt{l_2}+\sqrt{k_2})
Kˉ=Lˉ=50,U(x,y)=x1/2y1/2\bar{K} = \bar{L} = 50, \quad U(x,y) = x^{1/2} y^{1/2}
Step-by-step solution

Efficiency locus: RTS1=k1/l1=k2/l2⇒k1=l1RTS_1 = \sqrt{k_1}/\sqrt{l_1} = \sqrt{k_2}/\sqrt{l_2} \Rightarrow k_1 = l_1.

PPF: x=l1x = \sqrt{l_1}, y=50−l1=50−x2y = \sqrt{50 - l_1} = \sqrt{50 - x^2}. PPF: x2+y2=50x^2 + y^2 = 50.

RPT: RPT=x/50−x2RPT = x/\sqrt{50-x^2}.

Optimal point: MRS=y/xMRS = y/x. Set y/x=x/50−x2⇒y=x2/50−x2y/x = x/\sqrt{50-x^2} \Rightarrow y = x^2/\sqrt{50-x^2}. Substituting into PPF: x2=50−x2⇒x∗=y∗=5x^2 = 50 - x^2 \Rightarrow x^* = y^* = 5.


Part 3 — Externalities and Market Failure

Why Markets Can Fail

The First Welfare Theorem assumed:

  • Competitive markets (price-taking firms and consumers)
  • Perfect information (no hidden quality)
  • No externalities (agents' decisions affect only themselves)

A violation of any of these creates a market failure — the competitive equilibrium is no longer Pareto efficient. This section analyses externalities.

Externality (definition)

An externality exists when the choices of one agent directly enter the utility function of a consumer or the cost function of a firm — without being reflected in market prices.

Note: If A's purchases raise the price of a good and this hurts B, that is not an externality — it is transmitted via the price mechanism. An externality bypasses the price mechanism entirely.

Examples: Second-hand smoke (negative), vaccination (positive), R&D spillovers (positive), river pollution (negative).


Externalities in Consumption

Why the Market Equilibrium Fails

Consider Homer and Barney, each consuming cigars cc and whiskey ww:

UH(cH,wH,cB)=v(cH,wH)−ϕ(cB),UB(cB,wB,cH)=v(cB,wB)−ϕ(cH)U^H(c_H, w_H, c_B) = v(c_H, w_H) - \phi(c_B), \quad U^B(c_B, w_B, c_H) = v(c_B, w_B) - \phi(c_H)

Key observation: Each agent's MRS (between their own cigars and whiskey) is independent of the other's consumption. So each agent's demand function is the same as if there were no externality — the equilibrium allocation is computed as if ϕ\phi doesn't exist.

But if ∂v/∂ci<∂ϕ/∂ci\partial v/\partial c_i < \partial \phi / \partial c_i at the equilibrium (marginal private benefit of smoking < marginal harm imposed on the other), then both agents would be better off smoking less. The equilibrium is not Pareto efficient.

Worked Example

UH=cH1/2wH1/2−23cBU^H = c_H^{1/2} w_H^{1/2} - \frac{2}{3} c_B, UB=cB1/2wB1/2−23cHU^B = c_B^{1/2} w_B^{1/2} - \frac{2}{3} c_H. Endowments (10,10)(10,10) each.

Equilibrium: Each agent's demand (ignoring the external term) gives c∗=5+5/pcc^* = 5 + 5/p_c. Market clearing: cH+cB=20⇒pc∗=1⇒cH∗=cB∗=10c_H + c_B = 20 \Rightarrow p_c^* = 1 \Rightarrow c_H^* = c_B^* = 10, wH∗=wB∗=10w_H^* = w_B^* = 10.

Each agent's utility = 101/2⋅101/2−23⋅10=10−6.67=3.3310^{1/2} \cdot 10^{1/2} - \frac{2}{3} \cdot 10 = 10 - 6.67 = 3.33.

Compare the allocation cH=cB=9c_H = c_B = 9, wH=wB=10w_H = w_B = 10: Utility = 91/2⋅101/2−23⋅9=9.49−6=3.49>3.339^{1/2} \cdot 10^{1/2} - \frac{2}{3} \cdot 9 = 9.49 - 6 = 3.49 > 3.33.

Both agents are better off at (9,10)(9,10) than at the equilibrium (10,10)(10,10). The market overproduces cigars.

The Coase Theorem — Tradeable Externalities

Why does the market fail? Because the harm from smoking is non-tradeable — Homer cannot pay Barney to smoke less through a market transaction.

The fix: Make the externality tradeable. Suppose we create a market for "smoking rights". Barney is assigned a maximum quota Cˉ\bar{C} of cigars; Homer can buy reductions from Barney. Now:

  • Barney's utility over (wB,cB)(w_B, c_B) is unchanged.
  • Homer's utility is UH(wH,Cˉ−cB)U^H(w_H, \bar{C} - c_B), where (Cˉ−cB)(\bar{C} - c_B) is the amount Barney doesn't smoke — which Homer buys.

This is structurally identical to the exchange economy — two agents trading two goods. By the First Welfare Theorem, the competitive equilibrium is Pareto efficient.

Coase Theorem (informal)

If property rights over externalities are clearly defined and transaction costs are zero, then market participants can negotiate to an efficient outcome, regardless of the initial assignment of rights.

Coase Theorem limitations

The Coase theorem requires: (1) well-defined property rights, (2) low transaction costs, and (3) no wealth effects. In practice, many externalities involve millions of parties (e.g., global pollution), making negotiation infeasible.


Externalities in Production

Setup — Steel Mill and Fishery

  • Firm S (steel) produces output ss and pollution xx, with cost CS(s,x)C_S(s, x).
  • Firm F (fishery) produces fish ff, but is harmed by pollution: cost CF(f,x)C_F(f, x) with ∂CF/∂x>0\partial C_F / \partial x > 0.

Key properties of CSC_S:

  • ∂CS/∂s>0\partial C_S/\partial s > 0 (producing more steel costs more).
  • CSC_S is convex in xx: it decreases in pollution up to a level xˉ(s)\bar{x}(s), then increases. So firm S benefits from some pollution (it reduces abatement costs) up to the point ∂CS/∂x=0\partial C_S/\partial x = 0.

The Competitive Equilibrium Produces Too Much Pollution

Firm S's problem (ignoring F):

max⁡s,x  ps⋅s−CS(s,x)\max_{s,x} \; p_s \cdot s - C_S(s, x)

FOCs: ps=∂CS/∂sp_s = \partial C_S/\partial s and ∂CS/∂x=0\partial C_S/\partial x = 0.

Firm F's problem (treating xx as given):

max⁡f  pF⋅f−CF(f,x)\max_f \; p_F \cdot f - C_F(f, x)

FOC: pF=∂CF/∂fp_F = \partial C_F/\partial f.

Social optimum (maximise combined profits):

max⁡f,s,x  pss−CS(s,x)+pFf−CF(f,x)\max_{f,s,x} \; p_s s - C_S(s,x) + p_F f - C_F(f,x)

FOCs:

ps=∂CS∂s,pF=∂CF∂f,−∂CS∂x⏟abatement benefit=∂CF∂x⏟harm to fisheryp_s = \frac{\partial C_S}{\partial s}, \quad p_F = \frac{\partial C_F}{\partial f}, \quad \underbrace{-\frac{\partial C_S}{\partial x}}_{\text{abatement benefit}} = \underbrace{\frac{\partial C_F}{\partial x}}_{\text{harm to fishery}}

The last condition says that at the optimum, the marginal saving to S from emitting one more unit of pollution must equal the marginal harm to F. Since ∂CF/∂x>0\partial C_F / \partial x > 0, the right-hand side is positive, so ∂CS/∂x<0\partial C_S/\partial x < 0 at the optimum — meaning S should be producing to the left of where ∂CS/∂x=0\partial C_S/\partial x = 0.

x∗<xe\boxed{x^* < x^e}

Conclusion: Firm S overproduces pollution in the competitive equilibrium because it does not internalise the harm it imposes on F.

Remedy 1 — Pigouvian Tax

Pigouvian Tax

A Pigouvian tax is a per-unit charge tt on pollution imposed on the polluter. Set at: t∗=∂CF∂x(f∗,x∗)t^* = \frac{\partial C_F}{\partial x}(f^*, x^*) This makes S face the full social cost of its pollution.

How it works: Firm S's problem becomes max⁡s,xpss−CS(s,x)−t⋅x\max_{s,x} p_s s - C_S(s,x) - t \cdot x. The FOC for xx becomes:

t=−∂CS∂xt = -\frac{\partial C_S}{\partial x}

Setting t=∂CF/∂xt = \partial C_F/\partial x restores the social optimum condition.

Practical difficulty

Implementing the Pigouvian tax requires the government to know the firms' technology CSC_S and CFC_F — which is private information. In practice, regulators use approximations.

Remedy 2 — Creating a Pollution Market (Coase)

Assign F the right to clean water, which it can sell to S. Let qq be the price per unit of pollution and xx the quantity of pollution S is allowed to buy:

  • Firm S: max⁡s,xpss−q⋅x−CS(s,x)\max_{s,x} p_s s - q \cdot x - C_S(s,x) → FOC for xx: q=−∂CS/∂xq = -\partial C_S/\partial x.
  • Firm F: max⁡f,xpFf+q⋅x−CF(f,x)\max_{f,x} p_F f + q \cdot x - C_F(f,x) → FOC for xx: q=∂CF/∂xq = \partial C_F/\partial x.

Market clearing (qq same for both): ∂CF/∂x=−∂CS/∂x\partial C_F/\partial x = -\partial C_S/\partial x ✓ — the social optimum is achieved.

The efficient outcome is the same regardless of whether F or S holds the property right (Coase Theorem) — only the distribution of profits differs.

Approach Mechanism Requires
Pigouvian tax Government sets t∗=∂CF/∂xt^* = \partial C_F/\partial x Full knowledge of CFC_F, CSC_S
Pollution market Create tradeable rights, let markets clear Well-defined property rights; low transaction costs
Command & control Regulation caps xx at x∗x^* Knowledge of social optimum x∗x^*

Worked Example

Setup: CS(s,x)=s2−12(s−x)2C_S(s,x) = s^2 - \frac{1}{2}(s-x)^2, CF(f,x)=f2+fxC_F(f,x) = f^2 + fx. Prices: pS=65p_S = 65, pF=39p_F = 39.

Competitive equilibrium

FOCs for S: pS=∂CS/∂s=2s−(s−x)=s+x⇒s+x=65p_S = \partial C_S/\partial s = 2s - (s-x) = s+x \Rightarrow s+x=65 ... and ∂CS/∂x=s−x=0⇒se=xe=32.5\partial C_S/\partial x = s-x = 0 \Rightarrow s^e = x^e = 32.5.

FOC for F: pF=2f+x⇒2f+32.5=39⇒fe=3.25p_F = 2f + x \Rightarrow 2f + 32.5 = 39 \Rightarrow f^e = 3.25.

Profits: πSe=65×32.5−32.52=1056.25\pi_S^e = 65 \times 32.5 - 32.5^2 = 1056.25, πFe=39×3.25−3.252−3.25×32.5=10.56\pi_F^e = 39 \times 3.25 - 3.25^2 - 3.25 \times 32.5 = 10.56. Total: 1066.81

Social optimum

FOCs: pF=2f+x⇒f=(39−x)/2p_F = 2f + x \Rightarrow f = (39-x)/2; pS=s+x⇒s=(65−x)/2p_S = s + x \Rightarrow s = (65-x)/2; −∂CS/∂x=∂CF/∂x⇒s−x=f-\partial C_S/\partial x = \partial C_F/\partial x \Rightarrow s-x = f.

Substituting: (65−x)/2−x=(39−x)/2⇒65−3x=39⇒x∗=13(65-x)/2 - x = (39-x)/2 \Rightarrow 65-3x = 39 \Rightarrow x^* = \mathbf{13}; s∗=26s^* = 26; f∗=13f^* = 13.

Total profit: 1098.5 — higher than the competitive case (1066.81). The market failure destroyed 31.69 in total surplus.

Pigouvian tax for this example

t∗=∂CF/∂x∣f∗=13=f∗=13t^* = \partial C_F/\partial x\big|_{f^*=13} = f^* = 13. The tax on pollution should be set at 13 per unit of xx.


Summary

Concept Key formula / idea
Second Welfare Theorem Any Pareto efficient allocation is achievable; redistribute via lump-sum taxes to avoid distortions
Efficiency locus (production) RTS1=RTS2RTS_1 = RTS_2 — both firms have same rate of technical substitution
PPF Maps efficiency locus into output space; concave due to diminishing marginal returns
RPT Opportunity cost of XX in terms of YY along the PPF; RPT=C′(x)/C′(y)RPT = C'(x)/C'(y) in equilibrium
Optimal production MRS=RPTMRS = RPT — consumer's willingness to trade must equal production's opportunity cost
Competitive equilibrium Firms price at MC → px/py=RPT=MRSp_x/p_y = RPT = MRS → First Best achieved
Externality Effect of one agent's choices on another's utility/cost, not priced by the market
Consumption externality Market overproduces the externality-generating good; marginal private benefit < social cost
Coase theorem Tradeable rights restore efficiency if transaction costs are low
Production externality Polluter sets ∂CS/∂x=0\partial C_S/\partial x = 0 but optimum requires −∂CS/∂x=∂CF/∂x-\partial C_S/\partial x = \partial C_F/\partial x
Pigouvian tax $t^* = \partial C_F/\partial x\big