Costs of Environmental Regulation

Econ 2277

Prof. Richard L. Sweeney

(print this presentation)

Overview of next two lectures

  • economic concept of costs

  • incidence

  • next class: employment and innovation


What comes to mind when you think of environmental policy costs?

Economic Concept of Costs

Opportunity Cost: value of whatever must be sacrificed in order to obtain something (i.e. foregone net benefits)

  • Example: costs of graduate school?

  • Opportunity costs typically do not coincide with monetary outlays – accountant’s measure of costs

  • Opportunity costs are usually (but not necessarily) greater than out-of-pocket costs.

Types of costs

Private compliance costs

  • Equipment purchase, installation, operation, and maintenance

    • most frequently calculated
    • example: scrubber
  • Changes in production process and/or inputs

    • more difficult to measure, but increasingly important
    • example: fuel change
    • [draw isocost-isoquant graph]
  • Also, cost of time spent on permit applications, reports, etc.

Government regulatory costs

  • Costs to government of administering, monitoring, and
    enforcing regulation

  • Frequently small relative to private-sector compliance

    • exceptions: SDWA, CERCLA/Superfund

Social Welfare Costs

Definition: Losses in consumer and producer surplus due to
rise in price or decrease in output of good and services
(caused by regulation)

Basic Story: if regulation leads to increase in price of
good, leads consumers to buy less of good and/or switch to
substitutes, leading to a decrease in surplus.

Indirect costs:

  • Changes in market structure

    • if regulatory costs are great, firms can exit industry, causing changes in concentration/competitiveness (hence impacts on both producer and consumer surplus)
  • Reductions in product quality

    • for example, more stringent fuel-efficiency standards result in production of lighter cars
  • Discouraged investment

    • e.g., R&D funds reallocated to compliance

General equilibrium costs:

  • increases in product prices can spill over into other markets

  • for example if regulated good is a large input to production, or industry is a large employer

  • typically we assume regulated sector "small" enough that it doesn't matter

Example: Fracking

Fracking = Multistage hydraulic fracturing + horizontal drilling

Incredible Increase in US Oil Production


Problem: Wastewater


Hydraulic Fracturing Background

  • New process that dramatically altered US ability to extract gas and oil

  • Works by pumping very high pressure water into shale formations to break up the rock and release oil

  • Produces large amounts of wastewater which contains all sorts of bad stuff: solids, organic and inorganic chemicals, metals, and naturally-occurring radioactive materials

  • Water is typically disposed of using injection wells or brought to a treatment facility.

    • Recent evidence suggests that many of these treatment facilities are not equipped to process this water.

Policy Proposal

Consider a regulation which would require firms to either pre-treat or recycle water

  • EPA estimates this will cost $\$$10/bbl of oil
    • variable cost

How can we estimate the total cost of this proposal?


  • Marginal costs for a single well: $MC_i = .5q_i$
  • Current price: $\$$50/bbl

First imagine the regulation applies to a single well


Initial quantity:

$$MC = .5q = 50$$

$$q = 100$$

Initial profits:

$$.5*50*100 = 2500$$

Regulation shifts MC up by $10


New quantity:

$$MC = 10 + .5q' = 50$$

$$q' = 80$$

Policy profits:

$$.5*40*80 = 1600$$

  • lose $\$$10 per unit on 80 barrels still produced
  • lose $\$$5 per unit on 20 barrels no longer produced

So full opportunity cost larger than actual expenditure, but smaller than an "engineering" approach that ignores behavioral response.

Now imagine it applies to all 10K wells in the US


Assume aggregate demand curve:

$$P = 100 - Q/(20,000)$$

To get aggregate supply, note that quantity is now 10,000 times bigger at every MC:

$$10,000 MC = .5Q$$

$$MC = Q/(20,0000)$$

$Q = 1,000,000$

If entire industry faces regulation, both consumers and firms respond


To find the new Q':

$$MC' = 10 + Q'/(20,000) = 100 - Q'/(20,000)$$

Q' = 90
P' = 55

Change in surplus:

  • both sides now lose $\$$5 on 900,000 barrels still consumed
  • $\$$2.50 on 100,000 barrels screened out by the policy

What about general equilibrium effects?

Do you think they might be large?

What are some spillovers in this setting?


Who bears the cost of regulation?

• In the previous example, both consumer and producer surplus
declined even though the statutory burden was imposed solely
on producers

• This is a key point: Incidence of a tax (or regulation) is
independent of the side the tax is imposed on

Example: Gasoline Excise Taxes


What is the economic incidence of these taxes?

  • Imagine that without taxes, the price at the pump would be

  • Consider a 50 cent excise tax levied on retailers

    • For every wholesale gallon a gas station owner buys, it must pay the state $0.50
  • Now consider a rule change which would shift the statutory burden to consumers

    • Now drivers, not station owners, cut a check to the state
  • How would this change the incidence of a 50 cent tax?

    • Should consumers oppose this change?

Statutory burden does not equal economic incidence


Statutory burden does not equal economic incidence


What does determine incidence?

  • In this example, price goes up by $\$$0.3 after a $\$$0.5 tax

  • This tells us that consumers incur 60% of the economic burden

    • And therefore producers incur 40%
  • What determines these proportions?

The true burden of regulation depends on the relative responsiveness of demand and supply

  • In the fracking example, demand and supply had the same slope.

    • So their surpluses declined by the same amount
  • We typically measure responsiveness in terms of price elasticities:

Elasticity of demand:

$$\epsilon_{D}=\frac{\%\Delta Demand}{\%\Delta Price}$$

Elasticity of supply:

$$\epsilon_{S}=\frac{\%\Delta Supply}{\%\Delta Price}$$

Inelastic factors bear taxes


Elastic factors avoid them


Inelastic factors bear taxes; Elastic factors avoid


Formula for tax incidence

Fraction paid by consumers:


Fraction paid by producers:


Consumers bear full burden if:

  • Perfectly inelastic demand: $\epsilon_{D}=0$
  • Perfectly elastic supply $\epsilon_{S}=\infty$

Producers bear full burden if:

  • Perfectly inelastic supply: $\epsilon_{s}=0$
  • Perfectly elastic demand $\epsilon_{D}=-\infty$

What determines these elasticities?


  • Availability of substitutes
  • How “essential is a good”


  • Availability of inputs
  • Capacity constraints

What about the long run?

Are things more or less elastic?

  • In general, more elastic
  • Demand: Divers buy different cars
  • Supply: Can open up new oil fields

What about incidence among subgroups?

  • Most often, want to know if otherwise socially disadvantaged
    consumers pay a disproportionate share of costs

  • Do you think environmental policy is typically progress or
    – Electricity tax
    – Flood plain restrictions

  • Determined by how much of a good each group uses, and their

    • If full response involves expensive capital goods, situation could look worse in the long run

Ultimately, costs are borne by individuals

Either consumers, taxpayers, or shareholders


Employment Effects


(Photo by Win McNamee/Getty Images)

Group discussion

Do you think the employment effects of environmental regulation are likely large or small?

  1. Yes or no (justify your answer)

  2. What factors determine whether the effects are large or small?

Some Context on Employment

Here are some highlight from the most recent BLS monthly labor report.

  • Job openings: 6.9 million
  • Hires: 5.7 million
  • Total separations: 5.5 million
    • includes quits, layoffs, other

Over past 12 months: hires totaled 66.7 million and separations totaled 64.2 million, yielding a net employment gain of 2.5 million

Some Context on Employment


Source: Politifact

Extent of employment effects depends on regulatory asymmetry

  • Are other firms in the same industry similarly regulated?

  • Can firms easily move to other locations

Gross Job Losses Can Be Large

Greenstone (2002) found that the counties unaffected by the Clean Air Act gained 590,000 jobs relative to those affected as a result.

However, we actually care about net jobs.

  • Reason to suspect net effects are much smaller

  • A recent study on a $\$$40 carbon tax found that it would reduce long run employment by 0.3 percentage points, for net job losses of about 480,000.

    • As a contrast, NERA estimated losses of 2.75 million jobs..

When might we actually care about gross numbers too?

Why all the fuss then?

One snarky WaPo article noted that the entire coal industry employs fewer people than Arby's.

However, as a reply noted:

The average coal salary is $\$$15 - 30 per hour, while the average fast food worker makes $\$$8.

So what really matters is how long you're out of work, and the difference in salaries.

Empirical evidence suggests these transition costs can be large

Walker (2016) looked at workers in plants regulated under the 1990 Clean Air Act, and found that

  • Average earnings decline by 5%, and take 5 years to fully recover
  • Net effect is a 20% loss in pre-regulation earnings.

Should we explicitly be counting jobs in BCA?

  • jobs are a cost not a benefit
    • we don't care about jobs per se, but as a means to an end (output)
  • exception is when there's high unemployment
  • absent that
    • regulation could increase or decrease jobs in a sector
    • we care about efficiency
  • broken window fallacy

Summary on jobs

  • Discussion of environmental policy in the US dominated by
    employment concerns

  • But even in regulated sectors, net effects can ambiguous

    • Pro: "green jobs"
    • Anti: "job killing regulations"
  • Economics suggests this gets much more attention than it warrants

    • Overall employment is determined by macro factors like investment, labor supply and technology
    • For context, about 20% of jobs in the US manufacturing sector are gained or lost in a given year
  • Under normal employment conditions, jobs "lost" in one sector are quickly "gained" in another
    • To the extent that's not true, transition costs not trivial

Next week: Jobs are a cost not a benefit!


Policy costs my a be offset by innovation

Hypothesized in a very influential article by Michael Porter

"Weak" version of the Porter Hypothesis

"stringent policies should trigger greater investment in developing new pollution-saving technologies. If these technologies induce input (e.g., energy) savings that would not have occurred without the policy, they may offset part of the compliance costs"

Group discussion:

Do you believe the weak Porter hypothesis?

  • What settings do you think will be most likely to find it?

Induced innovation has long history in economics

Sir John Hicks (1932): "a change in the relative prices of the factors of production is itself a spur to invention, and to invention of a particular
kind-directed to economizing the use of a factor which has become relatively expensive"

Newell, Jaffe, Stavins (1999)

Study spike in energy prices in 1970's


NJS (1999)

Find firms responded by making more efficient appliances


What about the "strong" version of the PH?

"Porter and van der Linde (1995b) go further, arguing that environmental regulations can actually “trigger innovation that may more than fully offset the costs of complying with them,” i.e., lowering overall production
costs and boosting competitiveness"

Group discussion:

Do you believe the strong Porter hypothesis?

  • Again, what factors do you think make this more or less likely?
  • What does this imply about firm's and free lunches? Should the government go around regulating things to boost productivity?

What about the strong Porter hypothesis?

Porter and van der Linde (1995), gave several examples:

At Ciba-Geigy's dyestuff plant in NewJersey, the need to meet new environmental standards caused the firm to reexamine its wastewater streams. Two changes in its production process-replacing iron with a different chemical conversion agent that did not result in the formation of solid iron sludge and process changes that eliminated the release of potentially toxic product into the wastewater stream-not only boosted yield by 40 percent but also eliminated wastes, resulting in annual cost savings of $740,000 (Dorfman, Muir and Miller, 1992).

Similarly, 3M discovered that in producing adhesives in batches that were transferred to storage tanks, one bad batch could spoil the entire contents of a tank. The result was wasted raw materials and high costs of hazardous waste disposal. 3M developed a new technique to run quality tests more rapidly on new batches. The new technique allowed 3M to reduce hazardous wastes by 10 tons per year at almost no cost, yielding an annual savings of more than $200,000 (Sheridan, 1992).

What do we think of these examples?

  • Unclear if its worth the fixed costs
  • Ex post vs ex ante rationalization
  • Anecdotes != data

In order for Porter to shape how we think about regulation, such examples must be quite common

Would imply:

  1. that firms are making large mistakes

    • possible
  2. government intervention triggers attention in a way that corrects these

    • hardly obvious

Takeaway on Porter

  • long run costs likely smaller if firms can innovate

    • in fact, sound economic theory and evidence to suggest they will
    • some evidence that we systematically underestimate this (ie overestimate regulatory costs)
  • possible that innovations are so great that it actually increases profits

  • to be useful, this has to be widespread and expected

  • economists skeptical, little evidence to suggest otherwise


  • Care about opportunity costs, not actual outlays

    • this has a very natural mapping into changes in consumer and producer surplus
  • Although it does not effect efficiency, we are often also interested in economic incidence

    • relatively inelastic factors bear burden
  • In the long run, costs will be...

    • higher if firms exit the market
    • lower if consumers can adapt
    • lower if innovation is directed towards regulatory compliance