Below are the first 10 and last 10 pages of uncorrected machine-read text (when available) of this chapter, followed by the top 30 algorithmically extracted key phrases from the chapter as a whole.
Intended to provide our own search engines and external engines with highly rich, chapter-representative searchable text on the opening pages of each chapter.
Because it is UNCORRECTED material, please consider the following text as a useful but insufficient proxy for the authoritative book pages.
Do not use for reproduction, copying, pasting, or reading; exclusively for search engines.
OCR for page 58
.
A Policy Framework:
Three Levels of Decision Making
Policy making affecting families and children is, as we
anticipated when we initiated the case studies, time-
consuming, polycentric, and complex; this much is obvious
to participants and observers alike. The panel sought to
go beyond these observations to create an analytic account
of the policy formation process that would draw on exist-
ing scientific understanding of policy determination,
resonate with the panel's experience and expertise, and
be capable of yielding prescriptions for practitioners.
The panel's search for explanatory frameworks was
guided by certain predilections. First, policy making is
not only time-consuming but is also sequential, specific
events being significantly shaped by what has occurred
before and virtually never ending. Second, policy making
seems to occur at several distinct hierarchical levels of
government--from the top political leadership to middle-
level program officials--with horizontal interactions
among participants at each level and with policy-making
activity at the different levels taking place concur-
rently. Third, the character of policy making differs
depending on the kind of government action being con-
templated--for example, whether the issue involves the
tax code and the tax-writing committees of Congress or
whether it involves direct expenditure programs and,
consequently, authorizing committees of Congress.
A large, rapidly growing, and richly diverse body of
research exists on public policy formation. We judged it
to be beyond the scope of this report to present a compre-
hensive survey of this literature. We reached the conclu-
sion, however, that no single model or approach presented
in existing literature adequately captures the cumulative
"feel" of the cases and the types of policy making they
represent. Thus we were led to use the information in
58
OCR for page 59
59
the cases to derive an analytic framework, the application
of which would fulfill the criteria outlined at the
beginning of this chapter.
From our examination of the case studies we observed
distinguishable patterns of interactions among components
of the policy formation process.
These patterns reflect
the dynamic relationships among the nature of the policy
issue, the participants who are involved in the policy
debate, and the types of resulting government action. We
describe these relationships according to an analytic
framework involving levels of decision making that are
distinguished by what issues are to be resolved, who is
typically involved in resolving them, and how they are
resolved.
This framework may be usefully expressed in terms of a
game metaphor. Its analytic appeal derives largely from
its capacity to depict complex problem-solving phenomena
involving participants who employ strategies to maximize
their positions in contexts constrained by agreed-on
rules, laws, and conditions.
We believe the framework presented below provides a
useful basis for formulating conjectures about the complex
relationships among observable elements of the policy
formation process. We stress that this presentation is
designed to be suggestive in interpreting complex policy
developments; it is not a definitive statement of testable
or tested propositions. Considerable further work is
needed before such a statement is possible.
THE POLICY FRAMEWORK
We believe that policy making in the federal government
can be described as occurring at different levels, and,
for purposes of analysis, we postulate three levels:
high, middle, and low. They differ along three major
dimensions: the nature of the issues in question, the
participants who are involved, and the types of government
action that can result. The nature of policy issues can
vary from highly value laden--for example, whether govern-
ment should mandate and support preschool education for
all children--to essentially technical--for example, how
to most effectively and efficiently immunize children
against polio once a decision to immunize has been made.
In addition, the degree of consensus among participants
in the decision-making process over a policy issue may
vary. At each level, decision making involves a large
_ ~, _ _ _ ,
OCR for page 60
60
.
number of participants who can be distinguished by their
goals and the resources available to them. These
resources include their position or office, expertise,
information, and time. Moreover, the types of resulting
Some may involve major
alterations In the existing social order--for example,
mandatory employment of all welfare mothers. Others may
involve less dramatic initiatives, such as expanding the
Head Start budget. At each level of decision making the
role of various components of policy formation--contextual
factors, constituency pressure, media presentations, and
research--and their interactions change. The high level
essentially involves deciding whether government action
is warranted and appropriate. The middle level involves
deciding more concretely what the government's role
should be. The low level involves the precise design of
that role and selecting the details of its execution.
These three levels of decision making represent dist-
inct arenas in which certain actors in the process have
greater control and advantage. Depending on the level of
activity, different strategies are more or less appropri-
ate for different actors. In addition, the levels of
decision making presented in this schema do not represent
a hierarchy. Policy making at the high level, although
more visible, is not necessarily more significant or more
essential than policy making at the middle or the low
level. The stakes may be equally great at all levels.
The three levels of decision making suggest a logical
priority for determining the nature of a social problem,
agreeing on a programmatic response, and finally estab-
lishing the legal regulations and guidelines to implement
it.
government action may also vary.
~ . . . . . . .
The High Level
At the high level, decision making involves the definition
of social conditions as social problems, the formulation
of solutions to those problems, and the resolution of
major conflicts in societal values. The high level
represents the contest to make an issue political--that
is, a legitimate object for government action. Policy
making at this level addresses major questions concerning
the nature of social goals. Does society have a problem
that requires government action? What is the nature of
that problem? Is more or less government intervention
warranted? Why? The debate is about philosophies of
OCR for page 61
61
government, the fundamental responsibilities of institu-
tions, and basic principles of social justice. Intense
controversy is likely, often fueled by the actions of
single-issue constituencies or powerful elites. Policy
proposals generally involve significant alterations of
the existing social order--for example, the nature and
extent of government involvement in the lives of American
children and their families. The principal governmental
actors are the President, the congressional leadership,
and the Supreme Court. Powerful private interests are
also significant participants. Among the components of
policy formation, media presentations, influential
leaders, and large coalitions of interests are predomin-
ant. Often, new visions wrought by economic, demographic,
or cultural changes are principal components at the high
level. Research is less significant. Although it may
serve to illuminate high-level issues, it rarely settles
them.
The original School Lunch Act of 1947 represents a
classic example of decision making at the high level.
The major issue was whether the federal government had
responsibility to ensure that children are adequately
nourished. It entailed defining a role for the federal
government in ~ordinary" times--no war and no depression--
that it had never had previously. Large coalitions of
education and farm interests supported the measure. The
President and leaders of Congress made it a legislative
goal. Though television was not yet a major force, other
media presentations examined and editorialized the issue.
Research on the medical etiology of men rejected from
military service illuminated the problem.
Similarly, the hunger crisis of 1967-1968 also involved
decision making at the high level. A mass media expose
on malnourished children and adults ignited widespread
public response and focused presidential attention on the
problem. It helped define a new goal for the federal
government--the elimination of hunger in America.
Research illuminated the issues. Large coalitions of
interest groups and the public at large, led by influ-
ential individuals in the Senate, pressed for congres-
sional action. The decision-making process at the high
level ended when the elimination of hunger became an
accepted national goal and a federal responsibility.
a
OCR for page 62
62
The Middle Level
.
The middle level of decision making involves the choice
of means to achieve high-level goals. It is a contest
over the allocation of authority and resources to attain
an agreed-on objective. Since government must act, what
type of action should it take? How should responsibility
be assigned among levels of government and government
agencies? How much should be spent on what types of
benefits and services? Policy proposals generally
represent programmatic responses to acknowledged social
problems--for example, the establishment of the Child
Abuse and Neglect Program to help deal with the problems
of physical and psychological abuse of children by family
members. The debate is about the results of alternative
government actions--i.e., their effectiveness and
efficiency, their fairness, costs, and distributional
effects, and the administrative competence of alternative
agencies. The principal participants are presidential
appointees, members of Congress, and the designees of
either group. Among policy components, the media are
less prominent than in the high arena. Interests are
more parochial and coalitions smaller. Ideas and visions
are more technocratic and focused; they deal in probabili-
ties, not possibilities. The choices made are among
existing options and structures of values. Research
frequently can help resolve middle-level issues. Although
it is in some senses more political than decision making
at the high level, reasoned compromises are frequently
easier to reach.
The policy debate surrounding the child care tax
deduction/credit centered at the middle level of decision
making and concerned the equitable distribution of the
tax burden. It was a contest among members of Congress
and presidential appointees. Media presentations influ-
enced the decision making infrequently. With the excep-
tion of the issue of women's equality, which touched many
politically sensitive issues besides the deduction, there
was little presidential involvement. Changes among major
contextual factors--for example, changing patterns of
women's labor force participation, particularly among
those with young children--and visions of equal treatment
for women in the marketplace were high-level issues, but
their ramifications for the tax system were not. Consti-
tuency pressure was parochial: unions with women in their
membership, employer organizations with a high percentage
of women workers, and some professional legal associations
OCR for page 63
63
concerned with the integrity of the tax code. Research
was significant in revealing the effects of various
proposals for child care tax benefits and in preventing
the passage of some measures that projected significant
revenue losses.
The Low Level
The low level of decision making involves the design of
means chosen at the middle level to achieve an end deter-
mined at the high level. It is a technical contest over
how best to implement an agreed-on approach to a problem.
Since government must act and the type of action has been
determined, how precisely should that action be imple-
mented? How will eligibility, standards, or exceptions
and exclusions be defined? How will compliance be deter-
mined, monitored, and enforced?
How will vendors be
selected and funds transferred? The answers to these
questions reflect the judgments of specialists and tech
Cans concerning factors such as the feasibility of
. . .
-
administration, legal sufficiency, costs, etc. Decision
making at the low level involves the fine-grained
processes of government and tends to reflect the concerns
of those with fiscal (budgeting, enforcement, auditing)
or programmatic (administration, staffing, efficacy)
responsibilities. The principal participants, therefore,
are the staffs of Congress and the executive branch
agencies. Media presentations and vision are less
significant at the low level.
Constituency pressure is
very specific and targeted--often involving the use of
technical experts in a particular area. Research and
evaluation generally loom large in decision making at
this level as arbiters of disputes over the effectiveness
and efficiency of alternative plans.
Throughout most of the process of regulation writing,
the development of the Federal Interagency Day Care
Requirements involved decision making at the low level.
The issue was technical:
govern federal day care Programs.
how to design requirements to
The principal actors
were federal agency staff members. Media and contextual
factors had no discernible impact. Consitituency
pressures by a few advocacy groups were narrowly
focused. Research was very important, if only in the
form of advice from experts on what was in the best
interests of children.
OCR for page 64
64
.
In many respects the nature of an issue is the key to
determining the level of decision making. If the issue
involves the intrinsically value-laden question of what
the role of government is, decision making must take
place at the high level. If it involves the more tech-
nical question of how the government shall act, decision
making usually takes place at the middle level. If it
involves the question of how a policy or program shall be
designed, decision making is generally centered at the
low level. In short, an issue involving social values is
resolved at the high level, an issue involving equity at
the middle level, and an issue involving efficiency at
the low level. Although the existence of an issue is the
basis of decision making at each level, it does not
necessarily have to involve conflict. Decision making
takes place regardless of whether there is a dispute
concerning appropriate government action. It is not the
magnitude of agreement or disagreement that distinguishes
the level of policy making, but rather the character of
the agreement or disagreement.
Similarly, although many participants take part in
federal policy making, the level is distinguished by the
actor who is capable of resolving the issue. For example,
an OMB budget examiner participates, as a rule, in
decision making at the low level, but he or she might
prepare an analysis of congressionally approved legisla-
tion that significantly influences a presidential veto.
If the legislation involved an issue of social justice,
the budget examiner would in effect have participated at
the high level. The President alone, however, is capable
of resolving the issue by exercising veto power. The
presence of the President in conjunction with other
factors distinguishes the level of decision making as
high.
Certain actors are typical to each decision-making
level. Actors who participate at a level other than
their typical one take greater risks, expend more
resources, cannot participate alone, and often perform
ineffectively. A President, for example, who becomes
involved in regulation writing is participating at the
low level--not the typical level. The President expends
more resources--namely, time--to deal with regulation
writers, who are generally technocrats, and frequently
performs poorly at this level, lacking the technical
expertise and the necessary amount of time to participate
effectively.
OCR for page 65
65
The stakes or potential outcomes of policy making vary
according to the level of decision making because possible
types of governmental action at each level differ. Their
significance and ultimate impact are not necessarily
greater at the high level than at the middle or low level.
Congress and the President agreed, for example, that
handicapped children should have equitable access to
educational opportunities and passed a law mandating the
provision of special facilities and services to all handi-
capped children of public school age (P.L. 94-142). The
crux of the issue and the greatest stakes, however, rested
on the regulations promulgated to effect that high-level
end. Whether federal regulations would mandate states
and localities to spend billions of dollars on special
transportation systems, teachers, and other facilities
for the handicapped was in fact the crucial question and
involved the highest stakes. Decision making at the low
level, therefore, does not necessarily involve low stakes.
As we suggested earlier, no hierarchy is necessarily
implied among the decision-making levels, although there
is a logical priority. A broad social issue is resolved
into a programmatic issue of how to execute the decision
made at the high level; the resolution of a programmatic
issue requires a decision of how best to implement an
agreed-on initiative. Policy making, however, rarely
reflects this smooth linear flow. The policy process is
often characterized by solutions looking for problems,
rather than problems looking for solutions. Policy makers
do not always reach agreement first on broad goals, then
on programs, and finally on details of implementation.
Policy making is a dynamic, convoluted process of conflict
resolution, in which consensus at the high level can
easily be destroyed at the middle or low levels, where
programmatic and implementation issues are addressed. In
addition, because policy making is a fluid and dynamic
process, events and participants can shift the level of
decision making. And a given policy issue may involve
decision making at more than one level simultaneously.
In the WIC case we observed that once the elimination
of hunger became a national goal, the policy debate
shifted from the high level to the middle level. There
the contest involved choosing among available program-
matic means: direct distribution, food stamps, the
provision of supplemental food, and other possible
programs to reach the goal. More parochial interests
surfaced over which vehicle would most benefit particular
constituencies, including farmers, the schools, and the
OCR for page 66
66
.
poor. Members of the cabinet, the subcabinet, and
congressional committees debated the size of appropria-
tions and the expansion Of specific Programs.
Targeting
food assistance to pregnant women, infants, and young
children appeared as an attractive strategy in part
because of the findings of several research studies that
explored the physical and developmental effects of mal-
nutrition on infants and vouno children.
Of means was resolved, the debate shifted again to the
low level, where the major concern was for the precise
design of a targeted supplemental food program. New
issues arose over the purchase price of food stamps, the
contents of a supplemental food package, and the require-
ments governing program participation. Staffs dealt with
these questions; middle-level participants approved their
answers. An expert from a nongovernmental antipoverty
group advised on a regulation. A research study of the
nutritional needs of pregnant women was influential in
determining the contents of the food packages. The
~r~yram was ~mp~emencea ana over time expanded. The
media accorded little or no attention to issues and
decisions at this level.
Decision making at the low level in the development of
the Federal Interagency Day Care Requirements was
disrupted continually by the middle- and even high-level
issues that embroiled it. It is worth noting that when
the requirements were moved into the middle or high level,
the interests, ideas, and actors changed substantially.
Child-staff ratios that agency heads and their staffs
could agree on became one of untenable once threats of
their enforcement created new, unresolved conflicts at
the middle or high level. The issue became one of
whether the federal government should regulate and
standardize care for Preschool children and, if so, by
_ _
After the issue
~ . . . ~
what measure and to what extent. The interests, ideas,
and research that had sustained the requirements at the
low level simply could not carry it at the middle level.
In vetoing the Mondale-Brademus bill, President Nixon
transformed a middle-level issue of program design to a
high-level issue of government interference in the family.
Whether Nixon actually believed that the Comprehensive
Child Development Act threatened the sanctity of family
life and promoted communal approaches to childrearing,
the administration was dissatisfied with the legislation,
particularly its administrative provision--reauthorizing
the Office of Economic Opportunity and establishing a
network of community-based prime sponsors. Philosophical
OCR for page 67
67
issues concerning the child development program were in
fact never raised during the congressional debate over
the legislation. Nixon's use of inflammatory language in
his veto message suggests that the administration realized
its only hope of defeating the legislation was to elevate
it from a middle-level debate over programmatic means to
a high-level debate over social values. The political
costs of trying to override the veto were too great for
most members of Congress to risk, and the bill died.
The case studies provide numerous instances of decision
In each
making at all three levels of the policy schema.
situation the nature of the issue, the goals and resources
of the involved actors, and the objectives of the policy
proposal determined the level of decision making. Yet in
ammo -=cmc__f~r "x~mn1 e. the drafting of the day care
requirements and the negotiations surrounding the Compre-
hensive Child Development Act, we observe that the level
of decision making shifted in response to events in the
policy process and the specific and conscious actions of
key participants to achieve their goals. These actions
and their consequences for policy outcomes suggest that
not only is our analytic framework helpful in explaining
past events but also that it has operational significance
for participants in future federal policy making as well.
OPERATIONAL IMPLICATIONS
The analytic framework provides a means of bringing order
to complex phenomena and of drawing lessons about effec-
tive participation in the policy formation process. At
each level of decision making, a number of conditions
exist that affect the potential roles of the components
of policy making, the options available to different
actors in the process, and the possible policy outputs.
Understanding the nature of these conditions and their
influence on policy making suggests certain strategic
opportunities or levers that are available at each level
to aid participants.
As previously noted, certain actors are typical, even
necessary, to decision making at each level and therefore
have positional advantage. At the high level, issues can
be resolved only by the President, selected members of
the congressional leadership, and the Supreme Court. At
the middle level, decision making requires the participa-
tion of members of Congress, cabinet members, and other
presidential appointees or their designees, but not
OCR for page 68
68
necessarily the President. At the low level the principal
actors are congressional staffs and officials in the
executive branch agencies. Nongovernmental interests can
participate at all levels of decision making, but to be
most effective they must involve strong and visible
coalitions at the high level and more specific political
advantage and technical expertise at the middle and low
levels. Although all of these actors operate to some
extent at each level, they are most effective at their
typical level. For all participants there are costs
associated with participation at levels other than their
typical ones. They must build effective ties, develop
relationships, and often trade favors, all of which
require the expenditure of resources, especially time.
Without such ties and relationships, a participant is
less likely to succeed.
One important aspect of effective participation at any
level of decision making is access to the key participants
who are necessary to the resolution of the issue in
question. The high level is most difficult to penetrate
for the average participant in federal policy making
because he or she must enlist the support of a few,
largely inaccessible key actors. Strategic opportunities
are principally a matter of access: to the most visible,
high-level policy makers in government; to powerful non-
governmental interests; to the mass media, which are
integral tools for mobilizing the public and interest
groups; and to institutional controls, such as high-level
appointments, legislative vetoes, and the authority to
rule existing laws and statutes unconstitutional. Because
there are more actors who typically participate at the
middle and low levels, these arenas are more easily
entered by outsiders.
Another important aspect of effective participation is
recognizing how and when changing social, economic, demo-
graphic, and political factors create conditions that are
favorable to new policy initiatives. The changing socio-
economic status of women and their rapid entry into the
labor force stimulated the establishment of the child
care tax deduction/credit. Similarly, the civil rights
movement contributed to the social awareness that created
a favorable climate for nutrition programs, compensatory
education programs, and other initiatives designed to
overcome poverty. Contextual factors frequently present
windows for change. At each level of decision making,
actors who recognize these opportunities and act on them
can significantly advance their causes.
OCR for page 69
69
Power at the high level, it has been noted, is largely
the power to persuade and to dramatize. It is the power
to shape public opinion and define or redefine social
goals. To achieve success at the high level requires an
understanding of broad currents of public opinion, of the
larger strategic issues of national politics, and of the
large strategic power blocs. It also requires greater
political support or consensus than at other levels.
Therefore, participants seeking access to policy making
at this level usually do so through coalitions and
alliances. Although necessary to effect change, such
alliances generally involve compromises among interests
and often result in the distortion of an idea or policy
initiative from its original form. They frequently
require a significant commitment of time to build and a
great deal of energy to maintain. Because of its
visibility, decision making at the high level is very
seductive, yet it can also be very costly for partici-
pants who lack the essential strategic opportunities and
resources. Many short-term political appointees and
elected officials, for example, find it difficult and
impractical to enter the high level of decision making.
Because of their generally brief tenure of office (just
over two years on the average), they can accomplish more
on behalf of a particular constituency and have greater
influence over federal policy initiatives at the middle
and low levels, which are their typical arenas. Under
normal circumstances, only when especially favorable
opportunities present themselves can such actors effec-
tively participate in decision making at the high level.
At the middle level of decision making, strategic
opportunities are more concrete. They involve access to
individuals and institutions with the power to initiate
and enact legislation affecting the authorization of
programs and agency structures and the appropriation of
funds. Power at the middle level is the power to control
programs, personnel, and budgets. It is not the power to
shape public opinion or define public goals, but rather
to design and initiate programmatic means to achieve them.
To achieve success at the middle level requires a
different type of political support than at the high
level. It requires continuous political interaction with
other participants and frequently creates problems of
divided loyalties. In contrast to the high level,
effective participation at the middle level is a matter
of establishing effective working relationships with
committee and subcommittee chairmen, program officials,
OCR for page 70
70
.
and the executive staffs of interest groups with high
stakes in specific outcomes. Coalitions of interests may
be smaller than at the high level and have more specific
goals--for example, the enactment of a program or the
expansion of an appropriation that does not involve
issues or conflicts concerning fundamental social values.
Decision making at the middle level is less visible and
therefore participants are less concerned with access to
the media than access to researchers and other profes-
sional communities having knowledge and strong interests
in program content and size. Information about program-
matic content, effects, and political feasibility as well
as organizational control over the structure of programs
and administering agencies are far more essential levers
to decision making at this level than visibility. Effec-
tive participation requires a sense of timing, maneuver,
opportunism, and an instinct for identifying trade-offs
and fashioning compromises. Though middle-level decision
making is frequently a prolonged process, it is generally
less time-consuming than decision making at the high
level. For that reason, many elected officials and
political appointees find they can be more productive by
participating at this level. They can frequently wield
significant influence by introducing a bill, adding an
item to the President's budget, or instituting an agency
reorganization.
At the low level, the primary strategic opportunity or
lever is expertise. Power at the low level resides in
regulation writing and project management--the implementa-
tion of policies and programs established at the high and
middle levels. Technical knowledge is essential to
achieving success. Research results, especially those of
evaluation studies, are frequently influential in the
design of delivery systems, the establishment of eligi-
bility requirements, or the drafting of specific regula-
tions and guidelines. Effective participation at the low
level is a matter of communicating with experts and tech-
nicians. It requires a sense of how specialists work, of
how long it takes for them to produce answers to ques-
tions, and of the limitations on their perspectives.
Political support is also significant in decision making
at the low level, though generally it is much more
specifically targeted than at the high or middle levels
"11U uses non usually require One establishment of large
coalitions among interests. Public visibility generally
offers no advantage. In fact, participants at this level
generally believe their position is enhanced by not
OCR for page 71
71
attracting broad attention.
· . . .
/
When attention is drawn to
low-level decision making, unresolved middle- and
high-level issues are likely to surface and shift the
policy-making arena as well as introduce new key actors.
As we have noted, the level of decision making is
generally determined by the nature of the issue in
question, the goals and resources of the participants,
and the objectives of the policy proposal. From time to
time, however, actors may decide to shift the level of
decision making in order to advance their cause, particu-
larly as a means of blocking a particular policy initia-
tive. By shifting the level of decision making a partici-
pant not only introduces new issues into the debate but
opens the process to new participants. Resolution depends
on the strength and support of a new group of key actors.
In the defeat of the Mondale-Brademus bill, this strategy
was successful. The President vetoed the bill, and
congressional proponents could not muster the votes to
override. Nevertheless, without assured support and
control at a higher or lower level, a move to shift the
level of decision making can be risky. In general, it is
a less effective strategy for initiating a new policy
proposal than for blocking one already under consider-
ation.
In summary, a major lesson emerges from our application
of the analytic framework to the concrete instances
presented in the case studies: To enhance one's position
in the policy formation process, a participant must under-
stand the conditions and constraints of decision making
at each level. He or she must recognize the typical
level at which different types of issues can feasibly be
resolved and with what possible policy outcomes.
Similarly, a participant must recognize the level of
decision making at which he or she has the greatest
positional advantage and concentrate his or her energies
there. When there is cause to shift the level of decision
making, a participant must carefully assess the costs and
the risks associated with operating at a different level.
If one can gain access to key actors and can attract the
necessary political support, visibility, administrative
control, and expertise to participate effectively, the
change may produce a desired outcome; if not, it is
likely to result in failure.
In the next chapter we apply these general principles
to participation in federal policy formation affecting
children and families.
Representative terms from entire chapter:
middle level
