Chapter 9. Public Policy
Last
updated 11-23-2009
Copyright
2009 Robert E. Botsch
My policy is
to have no policy. Abraham Lincoln
OUTLINE
I. Public
Policy: Definition
II. General
Steps in the Policy Process and Problems in Each Step
A. Goal selection—Demands
B. Choosing among
alternatives
1.
Defining proposals
2.
Legitimating the proposal
a. Support
building
b. Enactment
C. Implementation
1.
Steps
a. Assigning
bureaucratic responsibility
b. Changing
legislation into rules and regulations
c. Resource
allocation—providing resources to carry it out
2.
Problem of language
3.
Limits on compliance
a. Calculations of
costs and benefits by those affected
b. Social values
and customs
c. Vague
regulations
d. Administrative
and/or political opposition
e. Cost
D. Policy
evaluation
1.
Types of evaluation
a.
Seat-of-the-pants
b. Middle
level—financial audits
c. Systematic
evaluation
2.
Obstacles
a. Unclear
objectives
b. Direction of
causality
c. Difficulty of
measuring long term impact
d. Self-interest of
evaluators
III. Types of
Policies
A. Classifying by
subject area
B. Classifying by
impact on society
1. Distributive
2. Redistributive
3. Regulatory
IV. Theoretical
Models that Aid in Understanding Public Policy
A. Purpose of
models
B. Overview of the
possibilities
C. Two models in
more detail
1. Rational
model—maximizing benefits and minimizing costs
2. Incrementalism
TEXT
I. Public
Policy: Definition
We could discuss a number of complicated
definitions of public policy, an important area of political science, but they
would only serve to confuse. As Thomas Dye argues in the
seventh edition of his classic text on the subject, Understanding Public
Policy, all of the complicated definitions boil down to a rather simple
definition. Public policy is what governments choose to do and
what they choose not to do. Even I can understand this definition!
You should note that this includes inaction
as well as action. Therefore, having no policy is in effect having a
policy. (See the
II. General
Steps in the Policy Process and Problems in Each Step
The policy process proceeds in stages, or
steps. If you are to understand policies that affect your life and want to play
a role in shaping these policies, you need to understand how policies get made.
All policies go through four general stages
in being made: 1) goal selection, in which demands get made and the agenda is
set to respond to these demands, 2) choosing among alternative proposals, 3)
implementation, or putting the policies in place and enforcing them, and 4)
evaluation, in which the effects of policies are measured, or at least
attempted to be measured. Even policies of non-action may go through these
steps. More likely, however, these policies to not do something are a result of
never getting past the first step: goal selection.
A. Goal
selection—Demands
If you remember back to the theoretical
model of political systems, we talked about inputs and outputs. The goals of a
policy are often defined by the demands that government leaders hear from their
followers. Do something about _________! You can fill in the blank. The demand
could be creating jobs, providing health care, reducing taxes, keeping illegal
aliens out, reducing crime, and so on.
In some cases, the goals are pretty well
agreed upon. We all want jobs,
health care, a clean environment, high wages, well educated children, keeping
taxes as low as possible, and peace. So long as we keep these kinds of goals
rather general and broad, we usually get agreement. These goals create what we
earlier called "valence issues" in political campaigns. The issue is
not whether we want employment or health care, but who can do a better job in
realizing these goals.
In other cases, the goals are themselves
controversial. We do not agree on
the goal of ending abortion, or on the goal of affirmative action, or on the
goal of reducing income inequalities among citizens.
The question of political power arises here.
For a demand to be heard and for leaders to feel that they must respond, those
making the demand have to be important enough or create enough of a disturbance
to gain leaders' attention. Otherwise the demands may get ignored. The
ability to get decision makers to worry about a problem is called agenda
setting. The ability to set the agenda is a very important political power.
In the
The question of goal selection is further
complicated by the fact that even some of the broad goals around which there
is a consensus come into conflict with each other. A clean environment can
cost jobs, at least temporarily. Improved health care can cost jobs and
increase taxes. Well educated children can increase taxes.
Finally, as soon as we take the broad
goals that translate them into policy choices, that is, specific ways to
achieve those goals, they become controversial. Agreement often ends.
American citizens agreed with Ross Perot in 1992 that the deficit was a major
problem. But many, if not a majority, did not at all like his policy proposal
to reduce the deficit through a gasoline tax. Nor did they like George H.W.
Bush's tax increase in 1990 and violated his "read my lips" pledge.
Nor did they like the tax increase proposed and passed by Bill Clinton and the
Democrats in Congress in 1993 that reduced the deficit. Nor did they like
B. Choosing
among alternatives
After those who make policy become aware of
a problem and a goal is chosen for consideration, policy makers must define
and choose among alternative actions or policies to try and reach that
goal. Several things happen here.
1. Defining
proposals
By defining proposals, we mean deciding
what the options are. What are all the possible ways of addressing a
problem? What are the details for each of the possibilities? How much does each
option cost? What places have any of them been tried before? What are the
tradeoffs?
Who develops these possible proposals? In
the
Many ideas for national policies come from
policies that are tried out in the states. Thinking cross-nationally,
nations borrow ideas from each other. In the
Governments have their own policy experts
who develop ideas. Cabinet offices in parliamentary governments have
professional policy analysts whose responsibility it is to study policy
options. In
Often governments hire outside experts
to examine alternative policies and make recommendations. Many of these experts
are housed in "think tanks," whose members make their living
from consulting.
2. Legitimating
the proposal
Choosing a policy to solve a problem is just
the beginning. In a any system, especially a
democratic system, you have to build sufficient support to make it
acceptable enough to get it adopted by the government. This is not easy!
a. Support
building
The first thing policy makers must do is to
build support for the proposal. The first thing that is usually done is to
float "trial balloons." This means to leak the idea and see
how other elites, the media, interest groups, and the public reacts. If the
reaction is favorable, then it is safe to take the proposal to the legislation
stage. If not, you have to either abandon it, modify
it, or build support for it.
Building support is the most difficult of
these alternatives. It takes a long time, and office holders don't have a long
time. Take, for example, the idea of a national health care system in the
b. Legislative enactment
Once support has been built, the policy must
be enacted into law. This involves more than just popular support. It requires
support among the groups and leaders who control the legislative process.
Even if they represent minorities, they may be able to block change if the
occupy key strategic locations. For example, white southern committee chairs of
key committees in the House and Senate were able to block civil rights
legislation for many years after popular majorities were built in support of
this kind of legislation. Republicans and conservative Democrats in Congress
were able to block
In sum, this is a much more complicated
process than most people believe. One cannot just simply choose the best policy
and hold hands with everybody and then "just do it." Even in
parliamentary systems where the executive commands a majority in the
legislature to get bills passed, pushing one’s party
members too far can result in losing the majority. If the majority is not
really a majority, but rather a coalition government that depends on members of
minority parties, the coalition is even more fragile. A leader who pushes too
far will have to stand for reelection. Only a dictator can "just do
it."
C. Implementation
Assuming the policy is passed into law by
some legislative body, it then has to be carried out. You should keep in
mind that we are talking about a wide range of possible legislative bodies. The
body could be anything from a student council through a school board to state
legislatures, regional assemblies, and national legislatures. You could even
think of it outside of government in terms of corporate boards of directors or
governing boards of civic groups. Regardless, someone has to implement the
policy.
1. Steps
There are three steps involved no matter who
passes the policy and no matter who is to implement it. Each is critical to the
success of the policy.
a. Assigning bureaucratic
responsibility
First, someone has to be made responsible for implementation. If no clear responsibility is
assigned, the policy will surely fail. This is as true on the personal level as
it is on the governmental level. Suppose in a family the mother or father says
that it is this family's policy that the trash be removed daily from the
holding bin under the kitchen sink. They then explain the reasons for this
policy in terms of good sanitation and appearance. They end the pronouncement
with that discussion and assign no one the responsibility to actually get it
done. What happens? Most likely nothing at all!
b. Changing
legislation into rules and regulations
The agency or people made responsible must decide
what rules and regulations are necessary to carry out the policy. This may
involve creating forms, setting dates, laying out procedures, and so on.
Even in a simple case of taking out the
trash, some of this happens. If the eldest child is made responsible for trash
removal, she may decide that the best time to do this is before her shower in
the evening. Therefore, she tells everyone that all trash must be deposited in
the receptacle before 9 pm each evening or else it will not be taken out.
c. Resource
allocation--providing resources to carry it out
Passing a policy and assigning someone
responsibility does little good if they do not have sufficient resources
to carry it out. Schools cannot educate if they do not have teachers,
classrooms, books, and operating budgets. Health care providers cannot provide
services if they lack similar kinds of resources. Even in a family, a small
child cannot carry out the trash or mow the yard if they lack sufficient
strength.
2. Problem of
language
Legislatures often pass policies using
quite general language. They don't
translate the policy into details. The details are usually left up to those who
carry out the policies. The ability to fill in the details creates what is
called administrative discretion. This discretion can be used to
actively and strongly implement the policy, or to passively and weakly
implement it.
For example, in 1970 Congress passed the
Occupational Safety and Health Act. It created the Occupational Safety and
Health Administration (OSHA) and gave it the responsibility to set limits on
worker exposure to toxic substances found in the workplace like cotton dust.
The law set no definite timetable and no precise dust limits. All this was left
up to OSHA to decide. As of the 1972 presidential election, no standards had
been set. In fact, Nixon administration campaign officials promised to
set no standards as a way to generate campaign contributions from executives in
the textile industry. They kept their promise. No real move to set the
standards took place until after Nixon left office. Nothing of substance
happened until after the election of Jimmy Carter. Even then, textile interests
created a great deal of pressure to set standards at levels that would provide
no real protection to workers. All of this was possible because of the discretion
allowed those responsible for carrying out the policy.
3. Limits on
compliance
Even if you pass and make a good faith
effort to implement a policy, other factors may prevent it from being
effective. What follows is a list of factors that can limit the effectiveness
of policies.
a. Calculations
of costs and benefits by those affected
Every policy (or any law, for that
matter) can be interpreted as a question: will I abide by it or not? When the textile industry faced temporary cotton
dust standards in the 1970s while the permanent standards were being
considered, many textile plants calculated that they would be better off not
trying to meet the standards and simply paying any fines that were imposed.
They knew that the permanent standards, if ever passed, would require different
and probably more expensive technological measures. Money spent on meeting the
temporary standards would probably be wasted. So why not save money and only
spend it when they find out what would be needed to meet the permanent standards?
To use a much simpler example, when we go
over the speed limit on the highway, we are making a cost/benefit calculation.
We are calculating that the danger of being caught and being fined is so small
that getting there a little bit faster is worth the risk.
b. Social values
and customs
Policies that violate social values and
customs are particularly difficult to enforce. Consider the effort to have a policy that forbade the consumption of
alcohol in the 1920s—prohibition. It violated customs of strong drink that
existed in many American subcultures and also violated the norm of individual
freedom. The consequences were disastrous. Prohibition created a whole new
class of criminals. It was so unpopular that people paid the police to not
enforce it—bribes and corruption. Ultimately, prohibition proved impossible to
enforce.
Today, many people make similar arguments
against current efforts to follow policies of strict prohibition with respect
to drugs and gambling. Perhaps we would be wiser to regulate them, as do many
other nations, rather than outlaw them.
Birth control and family planning policies
in many third world nations run into many of the same problems. The belief in
large families makes enforcement quite difficult.
c. Vague
regulations
Regulations that are vague are difficult
to enforce because efforts to
enforce them may not hold up in the courts. For example,
Alternatively, vague regulations may
allow those who are supposed to comply with the law to escape with minimal
compliance. Until changes took place in 1992 following "Operation Lost
Trust," the vote buying scandal that ultimately placed nearly 10% of the
South Carolina Legislature in jail, legislators felt free to ignore the
requirements that they report sources of campaign contributions. Even when they
did report them, the report was in such a form that it was nearly impossible to
tell exactly who was giving them money. Names were illegibly handwritten and
unintelligible abbreviations were frequently used. Vague requirements in record
keeping combined with weak enforcement allowed legislators to consider the
policy as a joke.
d.
Administrative and/or political opposition
Many policies fail because those charged
with enforcement oppose them. This was the case with the cotton dust
regulations when the Nixon administration was in power. It was the case for
many environmental standards during the Reagan years. A court ordered policy on
banning group-led prayers in public schools may be (and in fact has been)
ineffective in that teachers and school officials often simply ignore the
policy. Of course, practically speaking, they are the ones who must carry out
the policy because the court does not have the means to carry it out.
D. Policy
evaluation
Assuming that a policy choice is made, a
policy is enacted into law, and the policy is enforced to at least a minimal
degree, the next step is evaluation of that policy. Is it doing what it was
supposed to do? That is not as easy a question to answer as one may think.
1. Types of
evaluation
We can divide the kinds of evaluation that are
done into three broad groups: seat-of-the-pants, middle level limited
evaluations, and systematic evaluations.
a. Seat-of-the-pants
This type of evaluation relies on
impressionistic data that is usually not randomly selected. The report may cite
cases of where the policy succeeded giving the impression that these cases are
typical. In fact, the cases chosen may not be at all typical. The evidence may
only be fragmentary, and the cases may have been selected in a biased fashion.
Sadly, for reasons we will discuss below, policy makers quite often rely on
seat-of-the-pants evaluations.
b. Middle
level evaluation--financial audits
Middle level evaluations are more
systematic, but they are of limited help in evaluating whether or not the
policy is actually working. Typically, these middle level approaches involve a
financial audit to make sure that money is being spent in the ways it was
supposed to be spent. It tells the policy maker little about how successful the
policy is in addressing the problem the policy was supposed to address.
c. Systematic
evaluation
Logically, systematic evaluation sounds like
the best kind to perform. In systematic evaluation, the policy maker must
specify the desired measurable outcomes before the policy is implemented. Then,
after the policy is enacted, representative samples must be drawn and the
outcomes are then measured in an objective way. If the measurements do not meet
expected outcomes, then the policy is less than totally successful.
For example, if a job training program is
supposed to get people off welfare and into productive employment, the
evaluator needs to compare groups of otherwise similar people in and out of the
program. She must track them and see if statistically significant differences
persist over a reasonable amount of time. Are more of those in the program
employed? Do they make significantly higher wages? Do they stay employed a year
or two years after leaving the program?
A less than systematic evaluation might
compare people in the program with people who have significantly more
impediments (less education, criminal records, drug problems, poorer previous
job records, etc.) That is not a fair comparison. Those in the program need to
be compared with people with similar backgrounds. Alternatively, a seat-of-the-pants
evaluation may only look at a few success stories in the program and present
them as case histories. These people may have found success whether or not they
were in a program.
2. Obstacles
Despite the obvious advantages to systematic
evaluations, they are not performed very often. A number of reasons
explain this failure.
a. Unclear
objectives
Lack of clear objectives makes it
difficult for evaluators to measure outcomes because they don't know what to measure. Lack of clear objectives
bedevils those in the field of education. Is education supposed to produce high
scores on standardized knowledge tests? Some would say yes. Indeed, we can
teach to maximize knowledge. That is what the Japanese do and they do it quite
well. Unfortunately, that also stifles individual creativity and reduces the
quality of life for those who spend their lives memorizing fact after fact
after fact. Ok, so schools should also encourage creativity and prepare us for
living full well rounded lives. How can you measure creativity? What exactly is
creativity? It's a pretty vague concept. In fact, we may not know if a person
is creative until years after she leaves school. And then we can't be sure that
the school had anything to do with her creative activities. While we are at it,
many argue that schools should promote moral values. But exactly what we mean
by moral values is less clear. Do moral values include patriotism and loyalty
to nation, even if the nation is following otherwise immoral policies? Should
tolerance of those who are different be included as a moral value? Should that
tolerance be extended to unpopular groups like homosexuals? You can see how
controversial this gets.
b. Direction of
causality
Even if one is able to lay out clear
objectives, demonstrating whether the policy brought about the change is
difficult. We may be dealing with a spurious relationship. Some third
external variable may have brought about the change—not the policy.
Again, consider your college education.
Suppose we give you an entrance test and an exit test. Suppose the tests
measure improvements in reading and writing skills. How do we know that the
specific programs in this school led to those changes.
Any program might have led to positive change. Simply having time to read books
might have improved scores. Just maturing as one would do almost anywhere could
have improved scores. The only way you can tell for sure is to use a control
group. But in practice that is not always practical or possible. Can we
withhold education from a group of young people who are perfectly matched with
the entering students at a school? Ethical values often limit the kinds of
assessment research we are able to perform.
c. Difficulty of
measuring long term impact
Even if immediate change is measured, how
can we be sure that the change will be lasting? The only way is to track
those involved in the program a number of years after the program is completed.
This is often done, but long term tracking is very expensive and quite complex.
We could even argue about the importance of
long term impacts. Even if in the long run no real difference exists, can we
conclude that improving a person's life in the short run is not worth the
effort? Exactly how long must the effects last to be considered successful? In
the long run we are all dead!
d. Self-interest
of evaluators
If those doing the evaluation are those
implementing the policy, then they have a built-in self-interest in
demonstrating at least some kind of success. I remember attending a
conference that explored the success of a unique human services program. Those
running the program didn't have much to show for all the money and time they
spent. Nevertheless, they declared it as a success. They said they learned how
NOT to do it! In a very real sense they were correct. Those who designed later
efforts to improve human services did learn a number of things from this
effort.
From the point of view of the evaluator who
reports to the same boss who runs a program, reaching a negative conclusion is
a very risky enterprise. Where you stand depends on where you sit, remember?
If objectivity is a problem, then why not
just hire outside evaluators to come in and take the measurements? Even outside
evaluators have built-in biases. Any outside evaluator knows that they must
find something wrong if they ever expect to be hired again.
In the
e. Cost
Finally, policy analysis, especially
systematic policy analysis, is very expensive. It takes time, money,
expertise, and may disrupt the delivery of critical services in the process.
With budgets already extremely tight, what policy maker is willing to spend
scarce resources on something that, at best, will only prove that what she
thinks to be true is in fact true—the policy is working.
At worst, the evaluation will prove that she has been wasting money and
embarrass her. Going with your gut feelings that the program
is working, providing "evidence" by some economical seat-of-the-pants
evaluation, and spending available money on delivering the services that people
demand seems wiser. I know how much I resent having to spend my scarce
time proving that my teaching is working. I want to spend my time teaching, not
making up tests about what I do.
Evaluation is not really the final step. The
policy process is an ongoing process. Evaluations may uncover problems
and may suggest ways to improve. This naturally leads to discussions about new
policy alternatives or modifications. Modifications need to be legitimated,
enacted, put into the form of rules and regulations, and then enforced and once
again evaluated. Thus, the policy process is much like the scientific process.
Problems lead to efforts to improve existing theory, which in turn leads to new
questions to explore.
III. Types of
Policies
There are a number of ways we could classify
policies. Here we will cover two of the more obvious ways, by subject area and
by the impact they have.
A. Classifying
by subject area
This is the most obvious way to classify
policy. People in the policy field typically specialize in one or two areas:
human services policy, personnel management policy, urban policy, environmental
policy, transportation policy, education policy, criminal justice, defense
policy, tax policy, and so on. If you ever take a policy course, chances are
that your instructor will emphasize one particular policy area that matches
their own interests.
B. Classifying
by impact on society
This second method of classification is a
bit more controversial. It was developed by a political scientist named
Theodore Lowi. He argues that policies can be
classified by their impact into one of three general areas. Policies in each
area create similar patterns of political behavior.
1. Distributive
These policies are the least
controversial. They typically include "pork barrel"
projects that are not critical, but distribute jobs and government money and
services to constituents. They are passed in what political scientists call
a "log-rolling" process in which legislators vote for a
project in someone else's district with the expectation that others will vote
for projects in their own districts.
About the only time these policies become
controversial is when they lead to budget deficits. Because deficits
have been a problem in the
While the line-item veto sounds like a
common sense solution, it may not be any real solution at all. First,
pork-barrel projects only contribute a small percentage to the annual budget
deficit. If you vetoed all of them, there would still be a large deficit.
Second, pork is in the eye of the holder. A project in someone else's district
may look like pork to you, but in the eyes of those who live there it may be a
needed improvement in the region's infrastructure. Finally, many observers feel
that a line-item veto would merely transform pork distribution from Congress to
the President. It would increase the power of an office that many feel is
already too powerful.
2. Redistributive
Redistributive policies take resources
from one group and give them to another. They create more conflict and are
much more difficult to pass. For example, tax policies are redistributive in
nature. A major issue in the 1992 presidential campaign was President Bush's
proposal to decrease the tax rate capital gains. Democrats saw this as another
move to redistribute wealth from the middle and working classes to the upper
classes. The health care policy proposals that ultimately failed in 1994 were
hurt by their redistributive aspects. Insurance companies feared that their
profits would be reduced. Doctors saw limits placed on services they could
provide. Ultimately, middle class consumers feared that their own insurance
benefits would be reduced to provide benefits for those without insurance.
Although few were happy with the status quo, the proposals coming from the
All policies are to some extent
redistributive in nature. So we have
to talk about degrees here. For example, the Social Security system is not
terribly redistributive, though it does transfer money from the younger
generation to the older one. Regardless, people collect Social Security no
matter how much money they make. This makes it a largely distributive policy.
Its distributive nature is one important explanation for its deep-seated
popular support. Efforts to use means tests in giving benefits will move
it toward a redistributive policy. (A means test would mean that only those
below a certain income can get benefits.) The upper middle classes and the
wealthy will no longer feel they have a stake in the system. That could spell
the beginning of the end of this near sacrosanct program.
Because distributive policies are less
controversial and more easily passed, legislators often try to transform
redistributive policies into distributive ones. They try to put a little
something for all important groups into policies. The end result makes
political sense, but it may not make economic or common sense to those who are
not aware of political logic.
3. Regulatory
Regulatory policies are those policies that attempt
to control the behavior of people or institutions, like corporations. The
cotton dust regulations we have been using is a good example of regulatory
policy. Regulatory policies invite the involvement of interest groups,
especially those with an economic stake in the outcomes. Sometimes the parties
who are regulated transform these policies into economic advantages. For
example, regulations that set prices or limit competition are often favored by
businesses that do not want to face competition.
One criticism of this category is that
regulatory policies are really distributive or redistributive in nature. If a cotton dust regulation protects workers and
costs industry money, then it is redistributive. A regulation that limits
competition may result in increased prices for consumers. Acreage allotments
may distribute benefits to farmers. So in fact, regulatory policies may not be
a separate category at all.
IV. Theoretical
Models that Aid in Understanding Public Policy
A. Purpose of
models
If you remember our discussion of theory
back in chapter 2, we defined theory as a set of specified relationships
involving political matters that focus and organize out inquiry in our attempt
to describe, explain, and predict political events. In this case, what we
are attempting to describe, explain, and predict are policies, i.e. actions and
inactions of governments. Therefore, theories, or models, of the policy process
are "middle range" theories, to review another term we studied in
chapter 2.
B. Overview of
the possibilities
Political scientists who specialize in
public policy have used a number of theoretical models to help them understand
how policies are made. The list includes the elite model, the group or
pluralist model, the game theory model, the incremental model, the
institutional model, the process model, the public choice model, the rational
model, and the systems model (inputs and outputs, remember?).
No single model is clearly best in the sense that it explains more about more
policies than any other. You might think of them as different lenses through
which we can view the same phenomenon. Each gives us a slightly different view
of reality.
C. Two models in
more detail
To keep things simple, we will briefly
review two of the theoretical models you have not seen before in this on-line
text.
1. Rational
model—maximizing benefits and minimizing costs
In the abstract ideal world, the rational
model may well be the most desirable model of all. The policies chosen by
this model will result in the greatest benefits for the least cost.
Logically, all the policy maker has to do is
make a number of measurements and calculations. First, the policy maker must
decide what society wants. That is, she must know what the preferences
of all citizens are and how important each of these preferences are. This measure of the importance of each preference is
called its "weight." Second, the policy maker must know what all the policy
options are. Third, she must know what the results of each policy choice
will be in terms of providing the preferences. Fourth, she must calculate
the costs of each policy. Fifth, she makes a simple calculation comparing
the ratio of preference provision to cost. Finally, the most efficient
policy is chosen. That is, the policy with the highest ratio of benefits to
costs is chosen.
This sounds nice and logical. However, in
practice the model is usually not at all practical because the policy
maker can rarely make all the measurements. If the measurements can't be
made, then the model gives no real guidance.
Nevertheless, policy makers frequently
attempt to make cost/benefit calculations. And when they do make them, the
process by which they are made and the conclusions that are drawn from them
almost invariably become matters of political controversy. Whether or not they
are even required to make the calculations can become a point of contention.
For example, when congress passed the
Occupational Safety and Health Act of 1970, it directed the Occupational Safety
and Health Administration (OSHA) to set a standard for cotton dust exposure
that "to the extent feasible, on the basis of best available evidence,
that no employee will suffer material impairment of health or functional
capacity...based upon research, demonstrations, experiments,...the latest
scientific data in the field, the feasibility of the standards, and experience
gained under this and other safety and health laws." The key to this law
is what is meant by "feasible." The textile industry argued that
feasible should include economic feasibility. Those in favor of a strict standard,
which ultimately included OSHA, argued that technologically possible limits
were all that was meant by feasible. Both sides massed over 5,000 pages of
testimony, including hundreds of scientific studies, before OSHA made a ruling
and issued a regulation. All sides had different calculations about what it
would cost for each life saved. While one could calculate benefits in terms of
wages saved, medical costs saved, calculating the benefits of reducing pain and
suffering was harder and even more controversial. Yet to do a real cost/benefit
analysis, all costs and benefits had to be quantified. This was not the end of
it. When the textile industry didn't like the OSHA ruling, they appealed it to
the courts. Ultimately, the U.S. Supreme Court ruled in June of 1980 that
feasible did not include balancing the costs to industry against the lives of
workers. In effect, the Court said that Congress had rejected any cost/benefit
approach in writing the legislation and that if the standard was
technologically feasible, the industry had to abide by the regulation.
In addition to the difficulties of making
the calculations, other obstacles stand in the way of the rational model.
First, every policy has unintended consequences. We are not smart enough
to know what these will be. They can have a great effect on costs and benefits
after policies are put into place. For example, one of the results of the
cotton dust regulations was that it forced many companies to modernize their
plants. Some textile officials privately admit that they were forced to
modernize more quickly than they otherwise might have and that this helped
their competitive position. No one included these benefits in initial
calculations.
Second, political pressures from citizens
demand action while searching for the best policy creates delay. Therefore,
unless policy makers want to lose power, they have to do something, even if it
is not always the best thing.
Third, most policies are built on existing
policies and existing programs. The government and taxpayers have "sunk
costs" in these programs. To replace them with new ones, even though
they may be better ones, creates a great deal of turmoil. It's kind of like
remodeling a home. You make compromises to account for what you already have
built rather than tear it all down and start from scratch.
The end result of all this is that policy
makers may do a little cost/benefit calculation, but they rarely depend on it
entirely in choosing policies. Let us turn to a more realistic model.
2. Incrementalism
Because it would cost so much to look at
every program searching for the ideal replacement, what policy makers do is to
incrementally change them. (To do otherwise leads to the situation where the
search for the "perfect" becomes the enemy of the merely "good,"
or put more succinctly as policy-makers do, "the perfect is the enemy of
the good.") Incrementalism refers to
taking most of an existing program as a given and seeking ways to improve it.
Incremental change is a lot quicker and cheaper on a year to year basis. It is
also more conservative in that it sacrifices potential gains for eliminating
colossal failures that could come with completely revamping policies. It fits
with the common sense cliche' to not "throw out
the baby with the bath water" or "don't fix it if it ain't broken."
Because it is a less controversial approach,
it is easier politically. Dramatic new programs generate opposition. Consider
how much easier it would be to pass a national health care system that
preserved the traditional role for private insurance companies than it would be
to have the government be the sole provider. But as President Clinton (and
maybe President Obama?) learned, even with this compromise, the task proved
impossible.
KEY TERMS AND
IDEAS
public policy
steps in the policy process
agenda setting
think tanks
legitimating the proposal
trial balloons
enactment
implementation (also steps in)
administrative discretion
limits on policy compliance
types of policy evaluation
seat-of-the-pants evaluations
middle level evaluations
systematic evaluations
obstacles to systematic
evaluations
General Accounting Office
Legislative Audit Council
types of policies
distributive policies
"pork
barrel"
"log
rolling"
line-item veto
redistributive
means tests
regulatory policies
rational policy model
unintended consequences
sunk costs
incrementalism