The Encyclopedia of Economic Principals

Chapter 69

Externalities, Pigouvian Correction, and Common-Pool Resources

Price the harm, limit the entry, and count the duplication.

Four of the chapter's worked examples, made interactive: rent dissipation in an open-access fishery, a corrective tax set right and wrong, a prize contest with too many teams, and a shared basin. Change one value at a time and watch the figure, the numbers and the hand calculation respond.

Every example here is a constructed teaching example: it uses the hypothetical numbers of the chapter's worked examples, plus a few values added for comparison and labelled as such in each panel. Nothing here measures a real market, firm or household.

Demonstration 1 of 4

Free entry burns the fishery rent

Where does free entry leave the fish stock, and what happens to the rent?

Each entrant compares its own catch with its own cost and ignores the stock it depletes for everyone else. Entry continues until revenue per unit of effort equals its cost, so the stock falls until the rent is gone. Limiting access keeps a larger, cheaper-to-fish stock and a rent.

Equation, written in LaTeX: G(X)=0.5X(1-\frac{X}{100}),

Equation, written in LaTeX: 10(0.01)X=5,

Equation, written in LaTeX: E=\frac{9.375}{0.01(75)}=12.5.

Equation, written in LaTeX: X_{OA}=\frac{5}{20(0.01)}=25.

Scroll sideways for the whole equation

X is the fish stock, G(X) its natural growth, which equals the sustainable harvest in a steady state. Harvest is H = 0.01 E X for effort E. p is the dockside price per unit of harvest and c the full cost of a unit of effort. X_OA is the open-access stock where effort earns zero profit.

Predict first. If the price doubles from $10 to $20 under free entry, what happens to rent?

Your prediction

Choose an example

Scroll sideways for the whole figure

Figure: Free entry burns the fishery rent. Left: the logistic growth curve with the open-access stock 50.00 (harvest 12.50) and the controlled stock 75 (harvest 9.375). Right: revenue and effort cost, $125.00 and $125.00 under open access, $93.75 and $62.50 when controlled.
Dockside price ($): $10, Cost per unit of effort ($): $5
Constructed example: the chapter's hypothetical fishery (growth 0.5X(1 - X/100), price $10, effort cost $5, controlled stock 75, and the $20 price case); prices $7.50 and $15 and costs $2.50 and $7.50 are added for comparison.

Calculated values

Open-access stock X_OA
50.00
Sustainable harvest
12.50
Effort E
25.00
Revenue
$125.00
Effort cost
$125.00
Open-access rent
$0.00
Controlled harvest (X = 75)
9.3750
Controlled effort
12.50
Controlled rent
$31.25

Free entry pushes the stock to where a unit of effort just pays: 10 x 0.01 x X = 5, so X_OA = 50.00. Growth there is 12.50, which needs effort 12.50 / (0.01 x 50.00) = 25.00. Revenue $125.00 equals effort cost $125.00, so the rent is $0.00. At the controlled stock of 75, growth is 9.375 and effort 12.5, giving revenue $93.75 against cost $62.50, so holding the stock at 75 earns a rent of $31.25 a year that free entry would burn.

Worked steps

  1. X_OA = 5 / (0.01 x 10) = 50.00
  2. H = G(X_OA) = 0.5 x 50.00 x 0.50 = 12.50
  3. E = 12.50 / (0.01 x 50.00) = 25.00
  4. Revenue = 10 x 12.50 = $125.00; cost = 5 x 25.00 = $125.00; rent = $0.00
  5. Controlled X = 75: revenue = 10 x 9.375 = $93.75; cost = 5 x 12.5 = $62.50; rent = $31.25

Use the idea

To see whether a shared resource is losing its rent, compare total revenue with the full cost of the effort applied; under open access they converge whatever the price.

Where the conclusion applies

Logistic growth, harvest proportional to effort and stock, constant price and cost, identical fishers and a steady state. Real fisheries add uncertainty, adjustment lags and enforcement costs.

Check your understanding: With price $10 and effort cost $7.50, what is the open-access stock and the rent?
X_OA = 7.5 / (10 x 0.01) = 75; G(75) = 0.5 x 75 x 0.25 = 9.375; E = 9.375 / (0.01 x 75) = 12.5; revenue $93.75 equals cost 7.5 x 12.5 = $93.75, so rent is $0.

Chapter 69 source: section "Gordon open-access rent dissipation".

Demonstration 2 of 4

Setting the corrective tax rate

What tax moves a polluting market to the efficient quantity, and what if the rate is miscalibrated?

The market stops where buyers' value equals private cost and ignores the damage, so it overproduces. A tax equal to the marginal damage makes producers face the social cost. A tax above the damage removes units worth more than their full cost; one below leaves harmful units in place.

Equation, written in LaTeX: MB(Q)=100-Q

Equation, written in LaTeX: MPC(Q)=20+Q.

Equation, written in LaTeX: 100-Q=20+Q+12,

Equation, written in LaTeX: MPC_t(Q)=32+Q.

Equation, written in LaTeX: 100-Q=20+Q+5,

Scroll sideways for the whole equation

Q is output in thousands of units. MB is buyers' marginal benefit, MPC the marginal private cost and MSC = MPC plus the external damage per unit. A per-unit tax shifts MPC up by the tax.

Predict first. If damage is really $5 but the $12 tax is kept, is output too high or too low?

Your prediction

Choose an example

Scroll sideways for the whole figure

Figure: Setting the corrective tax rate. Marginal benefit 100 - Q, marginal private cost 20 + Q and social cost with damage 12. The untaxed market produces 40.00 thousand, the efficient quantity is 34.00 thousand and the tax of 12 gives 34.00 thousand; the welfare loss is 0.00 thousand dollars.
External damage per unit ($): $12, Per-unit tax ($): $12
Constructed example: the chapter's hypothetical market (MB = 100 - Q, MPC = 20 + Q, damage $12 and $5, tax $12); damages $0 and $20, tax rates $0 and $5 and the welfare-loss figure are added.

Calculated values

Untaxed output Q^p (000)
40.00
MSC at Q^p ($)
72.00
Efficient Q* (000)
34.00
Taxed output (000)
34.00
Taxed vs efficient
Efficient
Welfare loss, derived ($000)
0.00

With damage $12 per unit, the efficient quantity solves 100 - Q = 20 + Q + 12, so Q* = (80 - 12) / 2 = 34.00 thousand. A $12 tax gives (80 - 12) / 2 = 34.00 thousand, so output is efficient: the tax matches the damage. The shaded triangle between MB and MSC is the derived welfare loss, 0.00 thousand dollars; the book does not print this figure.

Worked steps

  1. Untaxed: 100 - Q = 20 + Q, so Q^p = 80 / 2 = 40.00
  2. MSC at Q^p = 20 + 40 + 12 = 72.00
  3. Efficient: 100 - Q = 20 + Q + 12, so Q* = 68 / 2 = 34.00
  4. Taxed: 100 - Q = 20 + Q + 12, so Q_t = 68 / 2 = 34.00
  5. Loss = 0.5 x 0.00 x (2 x 0.00) = 0.00 thousand dollars (derived)

Use the idea

Before calling a charge Pigouvian, compare its rate with an estimate of marginal damage at the efficient quantity, not with the damage at today's output.

Where the conclusion applies

Linear schedules, constant damage per unit, competitive pricing and no other distortions. The welfare-loss triangle is derived from these schedules; the chapter does not print it.

Check your understanding: With damage $5 and tax $12, how far is output from efficient and what is the welfare loss?
Q* = (100 - 25) / 2 = 37.5 and Q_t = (100 - 32) / 2 = 34, a gap of 3.5 thousand units; loss = 0.5 x 3.5 x (2 x 3.5) = 12.25 thousand dollars (derived).

Chapter 69 source: section "Pigouvian correction".

Demonstration 3 of 4

How many prize teams is too many?

When does adding prize contestants raise expected social value, and when do teams still enter?

More teams raise the chance of success but with diminishing returns, while cost grows in proportion. Each entrant counts its expected share of the prize, not the duplication it adds, so private entry can continue past the number that maximizes social value.

Equation, written in LaTeX: W_D=0.40(1{,}500{,}000)-300{,}000=300{,}000.

Equation, written in LaTeX: q_4=1-0.75^4=0.68359375.

Equation, written in LaTeX: W_4=0.68359375(1{,}500{,}000)-320{,}000=705{,}390.625.

Equation, written in LaTeX: \frac{q_4P}{4}=170{,}898.4375,

Equation, written in LaTeX: q_{12}=1-0.75^{12}\approx0.96832365,

Scroll sideways for the whole equation

V = $1,500,000 is the social value of success, P = $1,000,000 the prize, and each team spends $80,000 and succeeds independently with probability p. q_n is the chance at least one of n teams succeeds and W_n = q_n V - 80,000 n the expected social value. W_D is the direct contract's value.

Predict first. Does moving from 4 to 12 teams raise or lower expected social value, and do teams still want to enter?

Your prediction

Choose an example

Scroll sideways for the whole figure

Figure: How many prize teams is too many? Left: expected social value against the number of teams at success probability 0.25, with 4 teams giving $705,390.625 against the direct contract's $300,000. Right: the expected prize per team, $170,898.44 at 4 teams, against the $80,000 cost.
Number of teams: 4, Each team's success probability: 0.25
Constructed example: the chapter's hypothetical contest (V $1,500,000, prize $1,000,000, cost $80,000, p 0.25, 4 and 12 teams, direct contract 0.40 and $300,000); 1 and 8 teams and success probabilities 0.15 and 0.35 are added. The book's W_12 = 492,485.475 uses a rounded q_12 and is approximate; the exact value 492,485.472 is shown.

Calculated values

Success chance q_n
0.68359375
Total team cost
$320,000
Expected social value W_n
$705,390.625
Expected prize per team
$170,898.44
Direct contract W_D
$300,000
Teams enter
Yes

With 4 teams each succeeding with probability 0.25, the chance that at least one succeeds is 1 - 0.75^4 = 0.68359375, so expected social value is $705,390.625, above the direct contract's $300,000. Each team's expected prize, $170,898.44, covers its $80,000 cost, so teams want to enter.

Worked steps

  1. q = 1 - 0.75^4 = 0.6835937500
  2. W = 0.6835937500 x 1,500,000 - 4 x 80,000 = 1,025,390.625 - 320,000 = 705,390.625
  3. Prize per team = 0.6835937500 x 1,000,000 / 4 = 170,898.44
  4. W_D = 0.40 x 1,500,000 - 300,000 = 300,000

Use the idea

When designing a prize, compare the marginal gain in success probability from one more team with that team's cost, and use entry limits or staged rounds if private entry overshoots.

Where the conclusion applies

Independent success across teams, a prize split at random among winners, no administration cost and risk-neutral teams. Correlated approaches overstate the gain from adding teams.

Check your understanding: With 8 teams at p = 0.25, what are W_8 and each team's expected prize?
q_8 = 1 - 0.75^8 = 0.8998870850; W_8 = 1,349,830.627 - 640,000 = 709,830.627; prize per team = 899,887.08 / 8 = 112,485.89, above $80,000.

Chapter 69 source: section "Prize Incentives, Contest Design, and Subsidized Experimentation".

Demonstration 4 of 4

Four users, one basin

How far does private use of a shared basin overshoot the coordinated level, and what does it cost?

A user who draws one more unit raises everyone's depletion cost but bears only its own share. The private first-order condition therefore ignores most of the marginal cost, and the gap grows with the number of users.

Equation, written in LaTeX: B_i(x)=20x-x^2,

Equation, written in LaTeX: D(X)=0.1X^2

Equation, written in LaTeX: 20-2x-\frac{0.2(4x)}{4}=0.

Equation, written in LaTeX: W(x)=4(20x-x^2)-0.1(4x)^2.

Scroll sideways for the whole equation

Each of N users draws x and gets benefit 20x - x^2. Total use is X = N x and the depletion cost k X^2 (k = 0.1 in the book) is shared equally. x^N is the noncooperative choice, x* the coordinated one, and W(x) total welfare.

Predict first. As users rise from 4 to 8 (same k), does the private choice x^N change?

Your prediction

Choose an example

Scroll sideways for the whole figure

Figure: Four users, one basin. Total welfare against use per user for 4 users and depletion coefficient 0.10. Private use 9.09 gives welfare 264.46; coordinated use 7.14 gives the peak 285.71.
Number of users: 4, Depletion cost coefficient: 0.1
Constructed example: the chapter's hypothetical basin (4 users, B = 20x - x^2, D = 0.1X^2); 2 and 8 users and coefficients 0.05 and 0.2 are added for comparison.

Calculated values

Private use per user x^N
9.09
Total private use X^N
36.36
Coordinated use per user x*
7.14
Total coordinated use X*
28.57
Welfare W^N
264.46
Welfare W*
285.71
Welfare gain
21.25
Withdrawal cut
7.79

Each of the 4 users bears only 1/4 of the depletion cost, so privately each draws x^N = 20 / 2.20 = 9.09, 36.36 in total. Coordination counts the full cost and sets x* = 20 / 2.80 = 7.14, 28.57 in total. Cutting withdrawal by 7.79 raises welfare from 264.46 to 285.71, a gain of 21.25.

Worked steps

  1. Private: 20 - 2x - 2 x 0.10 x = 0, so x^N = 20 / 2.20 = 9.0909
  2. Coordinated: 20 - 2x - 2 x 0.10 x 4 x = 0, so x* = 20 / 2.80 = 7.1429
  3. W^N = 4 x 9.090909 x (20 - 1.40 x 9.090909) = 264.46
  4. W* = 4 x 7.142857 x (20 - 1.40 x 7.142857) = 285.71
  5. Gain = 285.7143 - 264.4628 = 21.25; cut = 36.3636 - 28.5714 = 7.79

Use the idea

Compare the marginal cost each user actually faces with the full marginal depletion cost; rules, quotas or charges that close the gap align private and coordinated use.

Where the conclusion applies

Identical users, quadratic benefits and depletion cost, equal cost sharing and one period. Real basins add heterogeneity, dynamics and monitoring costs.

Check your understanding: With 8 users and k = 0.1, what are x^N, x* and the welfare gain from coordination?
x^N = 20 / 2.2 = 9.09 (unchanged); x* = 20 / 3.6 = 5.56; W^N = 264.46 and W* = 444.44, so the gain is 179.98.

Chapter 69 source: section "Tragedy of the Commons".