Economic Analysis: Cost-Benefit Analysis, and Other Applications
Political Science 320B 2013
Finish other lecture
Lectures up to now will be posted this week.
Cost-Benefit Analysis (CBA)
Existence of market failure shows that the market outcome (no
govt intervention) is not optimal; there is some potential benefit
of govt action. But govt action has costs. Taxes, or burdens of
regulation. The question: are the benefits worth the costs?
CBA—a set of techniques for putting monetary values on the
benefits and costs of public policies— to decide which ones are
warranted. Thus you can compare: dollars (benefits) to dollars
(costs)—which are greater?
Often fairly easy. Major elements: actual spending by the
government; financial costs to citizens or businesses of complying
Some costs are harder to measure, much like some of the benefits
(e.g, a welfare program, supplementing income of low-income
families, may discourage employment).
The easiest case: government produces some benefit that is already
bought and sold in a market, with indentifiable prices. E.g.
government builds a dam, which will produce electric power.
Benefits of electric power is simply the price that people are already
paying for the amount of power. OR govt has training program for low-income people. Benefits are equal to the increase in their wages
(from before and after comparison.
What if there is no market for the benefit? E.g., cleaner air; more
equality; safer streets: there is no place to buy these things!
Two major approaches, if no direct market valuation
Use a survey to measure “willingness to pay.”
ask some people how much they would pay. Then multiply by the number of
individuals who would be in the same situation.
This produces exaggerated values—people want everything; they ignore budget
constraints. But you can discount the amounts. Thus this approach is of some
Find market prices that reflect the benefit indirectly.
Pollution. A real example: What is it worth to reduce pollution by 50%. Use
evidence from real estate prices in places with high and low pollution. (Use
controls for other factors).
What do you think? (Any other suggestions?) A major, controversial example: how much is it worth to save one
life? Relevant for policies on safety or health, designed to save lives.
E.g. how much to spend on organ transplants?
Is this a good idea? Why? Why not? DISCUSS.
Why: No one should like discussing this. But if policies don’t set
consistent values, then they waste lives!!!! Suppose $10 million for a
highway improvement saves 2 lives (before highway is rebuilt) in
highway safety; but the same amount would saves 100 lives in
immunizations, not being performed. Govt should move the money to
immunizations; save 98 more lives. Govt should use a fairly consistent
Viscusi study (forthcoming article)—Do quickly.
Basics of his study: law requires companies or EPA to clean up hazardous
wastes (e.g. a dump, not properly sealed; sometimes owner out of business).
Study combines three kinds of estimates: 1) at the site—how many people would
be exposed to a certain substance, at what levels, if no cleanup? 2)
Epidemiological or animal studies—how many cancer deaths from a given
amount of exposure (if 10,000 people exposed, how many die?) 3) Cost of
Combine all this: you get how much money is spent for each life saved. His result: cleanup program costs $100 million per life saved!! Very bad
payoff, compared with many other expenditures. (E.g. better to just keep people
away from the hazardous area.)
How to set a value on life? (Note. Viscusi just estimates cost per life
saved in a given program. Doesn’t address, what value to use in
One old approach: the value of a life is the persons’ expected lifetime earnings.
This measures economic contribution. It doesn’t make sense—that’s not why
we value a life!. Also: obnoxious results: disabled person—no earnings—does
that mean their life is worth less?
More recent approach to setting value for life: price paid for dangerous jobs.
How much extra do people demand in wages if job involves risk of loss of life?
Two jobs, equally skilled and difficult: one job has 1/250 greater chance of
premature death during entire career. Suppose the dangerous job pays $270 per
year more, apparently because of this risk. Then worker values his life at 250 X
$270 X 30 yrs = $2,025,000. Indicates: govt should value a life at about $2
million. WHAT DO YOU THINK? Is this right???
What to expect from this exercise. Not to get the “right” value. But get a
reasonable number. Otherwise how can you say, $30,000 or $500 million? And
it’s beneficial to set a value for any given government and use somewhat
consistently. Other forms of policy research
Intro: Most policy research is not based on economic theory. It
just seeks to find out the effects of policies on conditions in
society. Or sometimes tries to find the causes of social
problems—so that policy can try to remove the causes.
The effects of capital punishment: an example.
state policies provide almost an experiment on the effects of capital
punishment (the death penalty) in deterring crime.
In the late 1960s, the Supreme Court ruled that all existing state and
federal laws authorizing capital punishment were unconstitutionally
vague—i.e. they didn’t say clearly who should be executed—and
therefore subject to racial and other biases in implementation. In
1976, stated criteria for permissible statutes.
Over a period of two decades, many of the states passed new laws
and reinstituted capital punishment. Of those, some used it fairly
frequently, especially Texas; some did not. The result was a lot of
change over a fairly short period of time, and a lot of differences
between states. Most states had significant periods without capital
punishment; then many adopted, at different times; and with different
numbers of executions. Perfect for research. (