Toxic Pollution: Tragedy of the Commons, or a Surprisingly Common Tragedy

1
Toxic Pollution Tragedy of the Commons, or a
Surprisingly Common Tragedy?

• Berry, Freudenburg and Howell, 2006

2
Economy vs. Environment?
COMPATIBILITY The Wealth of Nations (Smith) Ecolo
gical Modernization (Mol, Spaargaaren) Environment
al Kuznets Curve (Grossman and Krueger
)
ENDURING CONFLICT An Essay on the Principle of
Popn (Malthus) The Treadmill of Production
(Schnaiberg) The Tragedy of the
Commons (Hardin)
To what extent, and under what conditions, does
economic activity harm the environment?
Are all polluters equally responsible for
environmental harm?
3
Testing the Competing Hypotheses H1
Proportionality (polluters equally
responsible) H2 Disproportionality (bad
actors) Is pollution uniform across groups of
producers? Is pollution uniform within groups of
producers? Are there just a few bad
actors? ...if so, are they necessary?
4
Case Study Toxic Releases

Level 1 Across major industry types Level 2
Within major industry Level 3 Facilities
within specific industry
Between 2-digit SICs
Select largest polluter
Within 2-digit SIC
Select largest polluter
Within 3-digit SIC
Data Sources Toxic Release Inventory, 2000
RSEI Version 2.1
5
Expected Distributions
Probability Density
Cumulative Fraction of Polluters
Toxic Releases per unit Production
Toxic Releases per unit Production
Normal Distribution Cumulative Fraction Plot
H1 Proportionality, allowing for many small
differences between firms
6
Expected Distributions
Probability Density
Cumulative Fraction of Polluters
Toxic Releases per unit Production
Toxic Releases per unit Production
Lognormal Distribution Cumulative Fraction Plot
H2 Disproportionality
7
Test Statistic Kolmogorov-Smirnov (K-S test,
with Lilliefors Correction)
D
8
Across Industries
How important is the tail of the distribution?
9
Across Industries
K-S Test Result Normal? (0.000) Log-normal
(0.200)
10
Across Sectors within the Primary Metals Industry
K-S Test Result Normal? (0.002) Log-normal
(0.200)
11
Across Facilities within the Highest Polluting
Sector (Primary Smelting Refining of
Nonferrous Metals)
K-S Test Result Normal? (0.000) Log-normal
(0.031)
12
Is the Pollution Necessary for the Economy?
K-S Test Result Normal? (0.000) Log-normal
(0.028)
13
Is the Pollution Necessary for Jobs?
K-S Test Result Normal? (0.000) Log-normal
(0.058)
14
Is the Pollution Necessary for Jobs?
K-S Test Result Normal? (0.000) Log-normal
(0.083)
15
Tragedy of the Commons?
16
How much does each cow affect the pasture?
Cows
Pasture
17
How much does each cow affect the pasture?
10
Cows
Pasture
10
18
How much does each polluter affect the overall
level of pollution?
Polluters
Toxic Pollution
19
How much does each polluter affect the overall
level of pollution?
10
Polluters
Toxic Pollution
10
Plants within Smelting and Refining Industry
20
How much does each polluter affect the overall
level of pollution?
10
Polluters
Toxic Pollution
Plants within Smelting and Refining Industry
21
How much does each polluter affect the overall
level of pollution?
10
Polluters
97.8
Toxic Pollution
Plants within Smelting and Refining Industry
22
How much does each polluter affect the overall
level of pollution?
Polluters
Top 10 to median
Toxic Pollution
Plants within Smelting and Refining Industry
23
How much does each polluter affect the overall
level of pollution?
Polluters
Top 10 to median
97.4
Toxic Pollution
Plants within Smelting and Refining Industry
24
How much does each polluter affect the overall
level of pollution?
Top 10 to median
Polluters
65
Toxic Pollution
Plants within Smelting and Refining Industry
Omitting top polluter (Rowley, UT)
25
Conclusions

• The proportionality hypothesis is consistently
  rejected.
• The tail of the distribution of pollution
  production accounts for an
• extremely disproportionate amount of the total
  pollution at all levels
• of analysis.

• This disproportionality does not appear to be
  necessary.

• Regulation that targets bad actors can
  substantially lower
• environmental harm with minimal impact to the
  economy.

• More empirical research is needed to substantiate
  and provide
• explanations for the above phenomenon.