Content uploaded by Yukio Adachi
Author content
All content in this area was uploaded by Yukio Adachi on Dec 22, 2020
Content may be subject to copyright.
政策創造研究 第 5 号（2011年11月） 61
The Process of Policy Design
Yukio ADACH
This paper examines what policy design is and what elements or phases comprise the
activity. Design activity, whatever is being designed, entails three common elements:
(1) accurately understanding the context, (2) achieving target values, and (3) fl eshing
out the concept. Public policy design has the added element of requiring designers to
keep the public viewpoint in mind throughout the design activity, as the client
comprises all members of society. More specifi cally, the task consists of three cyclical
processes. The fi rst process is analyzing and identifying the problem. The second is
defi ning the policy objective. The third is conceptualizing and selecting specifi c
prescriptions for achieving the policy objective. It is vital for policy designers to be
aware that problems, objectives, and prescriptions are closely interrelated and that,
therefore, systemic thinking is the main requirement.
1. What is policy design?
To design something, whatever it may be, entails a number of common elements, as I emphasized
in my book, The Modes of Thinking in the Policy Processes: Towards a General Theory of Public
Policies (Adachi 2005:62-66).
The fi rst element is obtaining an accurate understanding of the context in which the design activity
is being conducted. Designers are not given carte blanche to create whatever design they like. In fact,
their freedom is restricted to a surprising degree. For example, an architectural designer is greatly
constrained by the preferences and budget of the client, by the natural and physical conditions of the
building site, and by construction laws and regulations, including ancillary rules, administrative
guidance, local area agreements, etc. Designers are required to have an accurate understanding of the
design context and the ability to clearly differentiate between those aspects that must be accepted as
given preconditions, or constraints, of the design activity and those aspects that can be changed by their
own talent and efforts. The arrogant attitude of refusing to accept any constraint on the design activity
and enforcing one’s own preferences and values/world view on a client without sensitivity to the
design context is alien to design thinking. Such an attitude exists at the opposite end of the spectrum
from design thinking.
政策創造研究 第 5 号（2011年11月）62
The second common element is achieving target values. Making value judgments is an integral part
of design activity, regardless of what is being designed. Even designing an automobile is no exception.
A glance at an automobile catalogue shows that there is a highly diverse range of concepts (ideas)
represented by cars on the market today. Difference in concept boils down to how much importance is
placed on which values—safety, economy, roominess, comfort, eco-friendliness, appearance, etc.—
consumers may seek in a car, the thing being designed. One of the most important aspects of design
activity is to establish the concept, either by choosing from among existing concepts or newly creating
an original concept.
The third common element in design activity is the process of fl eshing out the concept or, in the
words of Stephen H. Linder and B. Guy Peters, the process of “fashioning an instrument that will work
in a desired manner” (Linder and Peters 1985:253). This process requires extensive examination of
means versus objectives—in other words, tenacious efforts to fi nd the best method, or prescription, for
achieving the stated objective. The task is not simply a matter of the means being automatically
determined once the objective has been established. In fact, the process of fl eshing out the concept is
creative and instinctive and, moreover, relies heavily on a high degree of tacit and experiential
knowledge.
While the three elements discussed above are common to design activity in general, be it
architectural, industrial, or any other type, public policy design entails an additional element unique to
the fi eld. This is the fact that the stakeholders are all members of society, not just a certain segment.
Architectural designers only have to focus on designing and completing the best possible structure
that maximizes the client’s satisfaction under various constraints. Likewise, industrial designers usually
only have to consider whether the products they design will be accepted by a particular consumer
segment. If they are designing an economy car, for example, the preferences of sports car or luxury car
buyers do not need to be considered.
But this is not the case with public policy designers. A public policy is intended to deal with social
and public issues, which are not the same as private or special-interest issues. The range of
applicability includes all members of society or all those members of society who fulfi ll a given
condition. That is why public policies are, and should be, evaluated and judged from the standpoint of
the interests of society as a whole (how society in general stands to benefi t) rather than private or special
interests (how specifi c individuals or organizations stand to benefi t). Public policy designers must
always keep this in mind when carrying out their tasks.
The fact that all members of society or all those members of society fulfi lling a given condition,
rather than particular customers or consumers, are the client is an aspect unique to public policy
design. Owing to this aspect—coupled with the “wicked” nature of the problems that public policies
The Process of Policy Design
（ADACHI）63
address and the high degree of uncertainty that unavoidably inhibits the application of rational analysis
methods—, the tasks of accurately understanding the context, achieving target values, and fl eshing out
the concept are made particularly diffi cult. In a sense, public policy design requires a much higher
level of capabilities than architectural or industrial design.
A physician, through patient interview, palpation, and various tests (such as blood tests, blood
pressure reading, urinalysis, fecal occult blood test, and ultrasound), obtains an accurate understanding
of a patient’s symptoms, identifi es the illness causing these symptoms, decides what to prescribe, and
embarks upon a certain course of treatment. A symptom or a set of symptoms can be caused by
numerous factors, and the same illness can cause different symptoms in different patients. Medication
can act differently depending on the individual, and the presence, absence, and degree of side effects
can vary greatly from patient to patient. Therefore, in order to provide a defi nite diagnosis or to
identify what is causing the symptoms, a physician must have a solid store of expert knowledge
pertaining to pathology, pharmacology, and other branches of medicine, as well being a thoroughly
experienced clinician. Identifying the cause of disease and determining a course of treatment is a
riddle-solving process consisting of ruling out the possibilities of other diseases one by one through a
systematic trial and error approach to diagnosis and treatment. Causal reasoning plays a central role in
this process, where a set of signs (phenomena) are reviewed to arrive at an estimation of the factor
underlying or causing the symptoms.
The work of a policy designer is similar to that of a physician in that both try to discover the
causes of a problem or symptom and search for an effective treatment to solve the problem. Both are
required to have advanced theoretical and analytical knowledge, as well as practical and experiential
knowledge. However, there is one enormous difference between the two that overwhelms all
similarities.
As is well known, public policy design is no longer the sole province of governments, which have
the responsibility of formulating and preparing policy drafts and implementing and managing policies
approved by parliament. Today, political parties, think tanks, industrial organizations, citizen groups,
non-profi t and non-government organizations, journalists, policy scholars, and diverse other actors are
also involved in public policy design and serve as policy advocates. Nevertheless, we must not
overlook the fact that no matter what person or organization is leading or undertaking public policy
design, there are several stages or phases common to the activity and the reasoning that guides the
activity. These stages or phases are discussed below, including observations on their similarities and
differences vis-à-vis physicians’ diagnostic (determining a treatment regimen) and treatment practices.
政策創造研究 第 5 号（2011年11月）64
2. Analyzing the problem
Public policy design starts with the recognition of a problem to be addressed through policy. If
there is no problem to be addressed, or there is a problem but nothing can be done about it (no
solutions seem to be in the offi ng and it is necessary to adopt a wait-and-see attitude for the
foreseeable future), there is no need for policy. If a seemingly serious problem exists but is likely to be
resolved with the passage of time, acting in haste may cause more harm than good. It may be more
prudent to wait it out. On the other hand, if there is not just a single but multiple problems to be
addressed, public policy designers must make a choice which one to start with. They have no choice
but to give greater priority to more serious and urgent problems. If the resolution of a problem is
theoretically possible but extremely diffi cult from a practical standpoint owing to political or economic
factors, the people’s lifestyle or mindset, or the values or worldview underlying such lifestyle or
mindset, anyone attempting to resolve such a problem would have to be deemed unwise (if perhaps
courageous).
As is partly evident from the above discussion, the process of discovering a problem to be
addressed through policy is not a process of discovering a problem that exists as an objective reality,
independently of the will or interpretation of the observer or analyst, and of defi ning and
formulating—namely identifying—the problem through external analysis. Rather, it is a highly
subjective process where a discrepancy between the existing condition and a desirable (to be aimed at)
future condition is recognized as a problem. How the existing condition is perceived and what kind of
desirable future condition is envisioned may differ depending on the observer or analyst, and there are
no objective, universal criteria or standards in this regard. The problem is whatever the individual
observer or analyst believes is the problem. Most problems subject to public policy are given life only
upon their formulation by an observer or analyst and do not pre-exist as an objective reality. In short,
a problem is a product of subjectivity.
In view of the foregoing, it can be said that the core aspect of problem identifi cation—whether the
persons concerned are aware of it or not—is to strongly impress upon the public that the existing
condition is deplorable and intolerable and that the desirable future condition that a policy is designed
to bring about is not only worth pursuing but is actually realizable if the necessary efforts are made.
The next task in policy design is to investigate the causes of the problem identifi ed as one of great
urgency—analyzing the problem from diverse angles to determine why and how it arose.
It is not always easy to pinpoint the causes of a problem. Only in very rare cases is a phenomenon
or problem caused by a single factor. There are many phenomena and problems whose emergence or
The Process of Policy Design
（ADACHI）65
unique features can only be explained on the basis of multi-causation: (1) several concurrent factors,
including random events, produce a given effect (α + β + γ → A), (2) a chain of factors produces a
given effect (α → β → γ → A), (3) several factors produce the same effect (α or β or γ → A), or (4)
the same factor produces several effects (α → A or B or C). Or rather, a one-on-one relation between
cause and effect is the exception. It can be easily understood that this is so by considering how
diffi cult it is for a physician to arrive at a defi nitive diagnosis and fi nd an effective treatment and
patients are fortunate if they can fi nd a capable physician.
How deeply must one probe the causes of a problem? How much resources, including time, should
be allocated to the task? If a careful analysis of the history and background of the problem is not made
and action is taken in haste—with only the symptoms treated or measures taken against something that
is mistaken to be the root cause but is in fact simply the trigger, the situation is unlikely to improve
and may even be aggravated, making it more diffi cult to solve the problem. That said, if too much time
is given to analysis and investigation, and nothing is decided or carried out until a fail-safe solution
has been found, a prime opportunity to improve may be lost. Eventually, the problem itself may get
worse or change, making all the effort already expended a futile endeavor. The correct answer as to
when to stop looking for causes lies at a point midway between these two extremes. Unfortunately,
there is no defi nitive guideline for determining where that midpoint should be. It must be determined
separately for each individual case.
The most important task in the problem analysis process is to examine what government behaviors
contribute to or are involved in problem creation, intensifi cation or prolongation and to what degree,
and based on that examination to identify the relevant government behaviors—behaviors that are at
least partly responsible for creating, aggravating or prolonging the problem in question. This task is a
process of critical importance, one that enables an analysis of the existing condition to be translated
into change or improvement of the existing condition.
The somewhat vague term government “behavior” is used here for a reason. A government’s
failure to act (not making the effort to explore, identify, and/or execute effective actions to address a
problem, or not taking the initiative to seek a solution) can create or exacerbate a problem or contribute
to such creation or exacerbation, just as government action (the act of preparing and implementing
policies) can have similar effects. Failing to act, therefore, is just as much a behavior as committing an
act. Accordingly, the calm observation of a situation without proactively taking any action can be
regarded as a form of behavior.
In cases where a government’s action contributes to the generation, aggravation or prolongation of
a problem, two possible scenarios can apply. One is where the problem is caused, primarily or partly,
by government action (policy) to begin with. The other is where the problem—although not caused by
政策創造研究 第 5 号（2011年11月）66
government action—shows no improvement or even worsens or is prolonged owing to the
government’s inappropriate action to address the problem. A representative example of the latter is
situations of “government failure”—used actively by political economists of the public choice school
as a framework or tool kit for analyzing policies and institutions—where inept intervention by the
government at times of market failure, i.e., when the market fails to deliver an effi cient allocation of
resources, results in a more ineffi cient allocation of resources. Diverse types of government behavior
not directly aimed at achieving more effi cient allocation of resources can also produce similar results.
For example, reforms introduced to prevent the national pension program from going bankrupt may
hasten the bankruptcy, or a program intended to improve the quality of civil service employees may
lead to further deterioration.
If a government takes no initiative to prevent or address a problem in a situation where its prompt
intervention could have prevented a problem from arising or worsening to any appreciable extent, or if
it loses the opportunity to intervene by postponing its decision due to not knowing what action to take,
the government cannot evade responsibility for creating, worsening or prolonging the problem or, at the
very least, for failing to act—just as it can be held responsible when its activity (its committing of an
act) was the main cause of the emergence, aggravation or prolongation of a problem.
Once the preliminary analytical process of problem identifi cation, investigation of causes, and
identifi cation of relevant government behaviors has been completed, the so-called monozukuri process
of policy design (formulating government action aimed at breaking a deadlock or improving a situation)
begins. Policy design, like design in general, consists of the process of conceiving and creating
“something that has form” and the preceding process of preliminary analysis and review. The quality
of the something-that-has-form that is the design output—namely the public policy—is critically
dependent on how well the initial, background process of analysis and review is carried out.
If government action is the main cause or one of the causes of a problem, it is necessary to
remove the cause itself. The government should remove itself promptly from the fi eld in question, or
shift its policy to one that is benign, or at least less harmful, to the situation. If government action
itself is not the main cause of a problem but a government policy implemented to address the problem
is inappropriate and unlikely to produce the intended effect, the government should search for a more
effective policy. If government action worsens a problem rather than solving it, the best option for the
government is to discover and implement a policy capable of ameliorating or resolving the problem. If
the chance of formulating such a policy is virtually nonexistent, the government has no recourse but to
shift to a policy that will not worsen the situation or completely refrain from implementing any policy
concerning the problem. If the government’s failure to act leads to the generation, exacerbation or
prolongation of a problem, the government may be required to explore and formulate effective and
The Process of Policy Design
（ADACHI）67
appropriate actions to deal with the problem.
3. Defi ning the policy objective
The series of intellectual activities leading from problem analysis to policy design includes, in
addition to cause and effect reasoning, thought processes that require strong conceptual and
imaginative abilities to envision a desirable future condition to be attained through public policies and
to design a roadmap for realizing that condition. Also included is communicative ability capable of
vividly describing through persuasive discourse the ‘charm’ of the desirable future condition and the
attainability of such a condition given the right efforts.
Once the “where” and ”what” of a problem has been identifi ed and an overall direction established
toward which government activity is to be re-directed, the remaining task is to fl esh out the direction
by determining the specifi cs. This process consists of two consecutive phases: (1) establish the policy
objective, i.e., conceive and formulate the desirable condition to be achieved through policy
implementation and (2) select and present specifi c means (prescriptions) for achieving that policy
objective.
There are four levels of government activity (governmental policies) that form a closely linked
hierarchy: “policy” in its narrow sense, plans, programs, and projects. Plans are implemented to
achieve the objective of a policy; programs are carried out to achieve the objective of each plan; and
projects are executed to achieve the objective of each program.
The objectives of public (governmental)policies at the lower end of the scale are more specifi c and
controllable. Conversely, the objectives of policies at the higher end are more abstract and generic.
Objectives at the top of the policy objectives system—i.e., the most abstract and generic—are those
pertaining to the realization or enhancement of a public value1. Public values are those values that, if
achieved or enhanced through or with the help of a public policy, mean that the greater part of society
will regard the policy as desirable according to the degree of perceived contribution. Examples include
freedom, equality, economic growth, social welfare, environmental protection, and public safety and
security.
As with policies and policy objectives, public values have different levels of abstraction and
generality. For example, the top (the most abstract and generic) level may include environmental
protection, and the level below it (second level) may include CO2 emissions reduction, water
purifi cation, and biodiversity conservation. In the realm of biodiversity conservation, public values to
be achieved or enhanced through policy can be formularized in a variety of different ways depending
on the target region and (endangered) species to be protected.
政策創造研究 第 5 号（2011年11月）68
In general, various public values at the second level or below in the public values hierarchy are
subject to public policy. For a public value to be fully functional as a guideline for government
activity and a yardstick for evaluating government activity, it must have suffi cient specifi city and
controllability (operationalizability) to allow judgment as to whether the public value in question was
attained and to what degree.2) For example, it is not enough simply to establish CO2 reduction as the
objective. It is necessary to specify how much CO2 is to be reduced and by when3).
Thus, in essence, the core task of establishing the objective of a public policy is to defi ne what
kind of public value is to be achieved to what degree by when. The defi nition is not easy, however,
primarily because there are multiple public values to be considered when designing, determining,
implementing, and evaluating public policies. Furthermore, different public values that are in
themselves socially justifi able can lead to policy decisions that are in confl ict with each other.
The following is an example of the foregoing: Over the past few decades Japan has been working
on criminal law reform, and one of the biggest points of contention yet to be resolved pertains to
restriction orders4). A restriction order system may be warranted from the standpoint of ensuring public
safety and security. However, if in fact the treatment of mental patients and alcohol and drug abusers
would have a better chance of success under an open system—where the patient is allowed a certain
level of contact with the outside world—than a closed system, locking up mentally disordered or
alcohol- or drug-addicted people who have committed serious crimes in a high-security treatment
center under the control of the Ministry of Justice and keeping them there for many years simply due
to the risk of repeat offense may be in confl ict with a number of public values, including personal
liberty and well-being.
If surveillance cameras are placed all over town as in London or if people are encouraged or
required to report to the authorities if they notice any suspicious characters as in Nazi Germany or
Japan immediately before and during World War II, it may be possible to reduce ‘crime’ and increase
public safety and security. However, such actions come with a high price in the form of privacy
invasion and a weakening of basic trust within a community. Other public values that are typically in
confl ict with each other include effi ciency versus equity, and environmental protection versus economic
growth.
As the foregoing examples indicate, the desirability from a public standpoint of a policy aimed at
contributing to the realization or enhancement of a given public value can only be guaranteed
conditionally. In other words, assuming that desirability can be guaranteed, it can only be guaranteed
“ceteris paribus”—only so long as other conditions are the same, that is to say, unless there are some
stronger reasons to justify a different evaluation. However, the reality is that other conditions are
hardly ever the same.
The Process of Policy Design
（ADACHI）69
This view concerning public value more or less corresponds to Ronald Dworkin’s understanding of
“principle.” Dworkin states, “A principle like ‘No man may profi t from his own wrong’ does not even
purport to set out conditions that make its application necessary. Rather, it states a reason that argues
in one direction, but does not necessitate a particular decision.……There may be other principles or
policies arguing in the other direction.……If so, our principle may not prevail, but that does not mean
that it is not a principle of our legal system, because in the next case, when these contravening
considerations are absent or less weighty, the principle may be decisive.” (Dworkin 1977: 26).
Thus, confl ict with other public values increases as the level of the public value targeted by a
policy rises or the strength of commitment toward achieving or enhancing a given value increases. For
example, once the level of public safety and security passes a threshold value, serious infringement of
personal liberty and privacy can result. If the targeted level of safety and security is higher, the
negative effects will be greater. Conversely, if freedom and privacy are considered inviolable under any
circumstances, it can have the highly detrimental effect of undermining public safety and security.
The same can be said about timeframe. A shorter timeframe for achieving the policy objective (i.e.,
the targeted level of a given public value) can make satisfying the demands for other public values
more diffi cult. Conversely, a longer timeframe will make it easier to achieve adjustment and
coordination among values.
There are no universal guidelines for determining how high to set a policy objective or what
timetable to establish for achieving that objective. Judgments must be made on a case-by-case basis in
line with conditions specifi c to the situation and responsibility for the consequences must be accepted.
There may be times when designers are forced to set the policy objective at a much lower level than
theoretically or technically possible or establish an unnecessarily long timeframe owing to an
overriding need to achieve social integration and gain consensus. Conversely, there may be times when
immediate efforts to realize the highest possible level of a given public value are allowed or
demanded—such as where radical reforms of a revolutionary nature are being courageously
implemented. A different set of considerations is required when faced with two opposing public values
that cannot be easily reconciled. I will discuss this issue on a different occasion.
4. Conceptualizing and selecting specifi c prescriptions
Designing a policy, whether public or not, is not a systematic two-step process consisting of fi rst
establishing the policy objective based on a careful analysis of the problem to be addressed and then,
and only then, fi nding the best means of achieving that objective. While the exploration of means
cannot be performed without some idea of the objective to work toward, in general it is not easy to
政策創造研究 第 5 号（2011年11月）70
discover and establish an appropriate policy objective at the early stages of policy design. In reality,
the designer follows a zigzagging process consisting of an initial cycle of problem analysis and
identifi cation, establishment of policy objective, and exploration of specifi c measures, followed by
additional cycles (as many as necessary) of problem re-analysis, adjustment of policy objective, and
adjustment of specifi c measures. Finally, at the end of all these tasks, the designer is able to determine
the policy objective and the means for achieving that objective—both at the same time. Such a process
is not rare. Rather, it can be said to be the norm. That is how strongly policy objectives and means are
linked together.
The activity of formulating a specifi c prescription, therefore, is not a unilinear, instrumentally-
rational (zweckrational) process consisting of exploring and fi nding the most effective means for
achieving a clearly-defi ned policy objective. Rather, it is a process that requires strong powers to
conceptualize and judge—consisting as it does of thinking up as many objective-means combinations
as possible, predicting what social consequences, i.e., benefi ts and damage, are likely to arise to what
degree of probability as a result of effecting individual combinations, and selecting the combination
that is expected to produce the best, or most preferable, consequence, for incorporation in the policy
process.
As mentioned earlier, policy design, like architectural, industrial, and other types of design,
commonly entails the task of fl eshing out the concept, along with having an accurate understanding of
the design context and selecting and determining the concept (target values). The task of fl eshing out a
concept, or giving shape to an abstract and generic idea, is never easy and just as diffi cult as the other
two tasks. The ability to do so is not achieved overnight. It requires years of practical experience
working with policies, and only those who have gained penetrating insight and keen instincts can
become a master of the prescription-conceptualization process. Unfortunately, the expertise of such
masters cannot be described in logical steps or compiled into a how-to manual.
Problems that may be addressed by public policies are enormously diverse, and no two problems
are entirely the same in real life. Because of this, ultimately, the task of conceptualizing a specifi c
prescription that is the means to achieve a given objective is essentially a unique, one-time endeavor,
one that requires fl exible thinking adapted to the situation at hand, as well as the capacity to make
judgments.
This does not mean, however, that there are no tools to assist this diffi cult task. It is possible to
fi nd several standardized forms, or stock strategies, for public policies that have been proven, through
past policy implementation, to have a certain level of effectiveness as prescriptions against similar
problems occurring under similar conditions. Examples of stock strategies relating to the reduction of
air pollutants such as CO2 and SO2 include emissions restriction, consumption restriction, pollution tax,
The Process of Policy Design
（ADACHI）71
surcharges, and emissions trading. When conceptualizing a prescription, it would be highly useful to
consider if one of the stock strategies —which vary greatly in the degree and form of government
involvement—can be utilized or if it is possible to increase effectiveness by combining some of the
stock strategies.
Of course, there may be times when none of the stock strategies appears useful and therefore
original thinking is called for. It is also important to avoid the trap of relying excessively on stock
strategies to the extent of losing the ability to make necessary changes or adjustments. Provided this
point is kept in mind, it makes good sense to start the design process with one or more stock
strategies. We must not dismiss the wisdom (practical and experiential knowledge) of our predecessors,
as there is more to be learned from them than is usually assumed.5)
Once the basic form has been determined, the remaining task is to decide the specifi cs. This task
also entails enormous diffi culties. What guidelines should be followed to determine the specifi cs, or
fl esh out the concept? What criteria should be used to compare and rank the multitude of similar
prescriptions that differ in the details? While answers to these questions will be considered at a future
opportunity, it should be noted here that whatever criteria are used to compare and rank alternative
prescriptions, it is fi rst of all necessary to gather and organize two types of basic information needed to
carry out the comparison and ranking.
One type of basic information required is the cost incurred in implementing each prescription.
There are two types of cost to consider. The fi rst is the cost of resources, i.e., the monetary value of
resources likely to be used to implement the policy. The second is the cost of any undesirable or
adverse effects likely to be generated by implementing the prescription Another type of basic
information required for designing and selecting prescriptions is the direct and indirect impacts of
implementing each prescription—what impacts are expected to what degree.
5. Policy Design Context
A keen sensitivity to context is a prerequisite of good design. Those who are insensitive to the
context of design activity and incapable of differentiating between those aspects that can be changed
by their own efforts and talent and those aspects that must, or should, be accepted as given
preconditions or constraints of the design activity cannot hope to design well. A designer’s freedom to
design is restricted to a surprising degree, and they must exercise their creativity within the scope of
those restrictions to produce the best possible design. The success of a design hinges on how well the
designer understands and is conscious of the context, i.e. the design constraints. This applies as much
to public policy design as to any other type of design.
政策創造研究 第 5 号（2011年11月）72
The most important constraint in public policy design is the inescapable fact that government
policy is ultimately determined and implemented through social interaction, i.e. politics in the broad
sense, and is not a product of purely intellectual, rational analysis. 6)
As is well known, the recognition that policy decisions and their product, government policies, are
the result of social interaction in the political arena—i.e., the political process—arose and was refi ned
as part of the development of U.S. political science from the early twentieth century onward. Aside
from minor differences in the phase of the political process and the political actors that they focus their
discourses on, Arthur F. Bentley, David B. Truman, Graham T. Allison, and other political process
scholars having a positivist orientation all discuss and interpret government policy as an output of the
political game, which includes complicated negotiations, bargaining, and compromises that can take
place among political (policy) actors and other emotional and non-rational elements.
The same can be said about policy implementation. One may well imagine that policies, once
determined, are faithfully implemented by the relevant government agency without any changes.
However, this is hardly the case in reality. “Implementation is evolution. Since it takes place in a world
we never made, we are usually right in the middle of the process. ……At each point we must cope
with new circumstances that allow us to actualize different potentials in whatever policy ideas we are
implementing. When we act to implement a policy, we change it. ……As we learn from experience
what is feasible or preferable, we correct errors. To the degree that these corrections make a difference
at all, they change our policy ideas as well as the policy outcomes, because the idea is embodies in the
action.” (Pressman and Wildavsky 1979:190-191).
Those who seek to deal effectively with public issues through public (government) policy and create
a better society and political environments must think and act always keeping in mind the fact that
government policy is ultimately determined and implemented through social interaction in the political
arena.
Policy design is not a monological activity, rather, if properly conducted it is, and should be, a
dialogue. The six logically but not necessarily chronologically consecutive processes comprising public
policy design as discussed here—problem identifi cation, investigation of causes, identifi cation of
relevant government behaviors, establishment of overall direction for changing the existing condition,
formularization of objectives, and exploration and determination of specifi c prescriptions—are in fact
descriptions of the process of interaction (dispute, negotiation, bargaining, compromise, etc.) among
policy actors as they move toward a social choice, or determination of government policy, at the same
time that they are prescriptions that individual policy actors should follow in their thought processes
when designing a public policy.
The recognition that something is a problem to be addressed by public policy is not limited to the
The Process of Policy Design
（ADACHI）73
perception of individuals. Society as a whole can arrive at such recognition as well. More accurately, a
common understanding concerning a problem to be addressed by public policy spreads gradually
through a population via interaction among diverse actors identifying the problem via diverse methods,
eventually leading to a widespread consensus. The same can be said about the other fi ve processes.
To repeat, a public policy sees the light of day only when it safely weathers the above-described
process of social interaction and is formally adopted as a government policy. This is the reality,
whether one likes it or not. Public policy design and any standards or guidelines to be followed by all
actors aspiring to public policy design must be in line with this incontrovertible reality.
What standards or guidelines should we follow to design public policies? What public value to
what degree of realization should be established as the policy objective? What yardstick or criteria
should be used to select the best prescription from among the numerous alternatives available? These
basic questions concerning public policy design will be discussed in the next issue of this journal.
Notes
1）
In my work, Seisaku to kachi—gendai no seiji tetsugaku (Policies and Values: Contemporary Political Philosophies ,
1991), this type of value that provides the ultimate rationale for public policies is termed “policy rationale.” Over the
past few years, however, I have used the term “public value” instead, in view of the fact that rationale is a form of
value and policy rationales serve as guidelines for designing and evaluating public policies. The same term is used in
this paper.
2）
It is frequently stated that there are two types of policy purpose. One is “objective,” which corresponds to purpose in
the narrow sense, as used here—that is to say, a purpose that has been made specifi c and controllable enough to enable
judgment as to whether that purpose has been achieved and to what degree. The other is “goal,” which corresponds to
a more general (less specifi c and controllable) purpose to be achieved through policy. It is, however, diffi cult to draw a
clear line between the two. Accordingly, in this paper I have intentionally avoided any discussion about the
differentiation of objectives from goals and instead mentioned that there are diverse levels of specifi city and
controllability in policy objectives.
3）
Quantifi cation is not necessarily required to achieve controllability of a policy objective. While quantifi cation offers great
advantages, it can turn into the be-all and end-all of policy implementation if not handled carefully. First, the
numerical target must, above all, be realizable, with a certain amount of leeway built into the target. An inappropriate
and unrealistic fi gure is liable to discourage rather than motivate policy executors. Second, it is also essential to avoid
the trap of focusing only on the quantifi able portion of a policy objective and losing sight of the non-quantifi able
portion.
4）
When a person who has committed a crime punishable by imprisonment or death (such as murder or arson) is found not
guilty or given a light sentence by reason of insanity or diminished capacity (alcohol or drug abuse) and deemed by the
court to be highly likely to reoffend unless treated for the mental disorder or addiction, that person may be placed in a
(newly created) high-security treatment center, different from a normal psychiatric hospital, under the jurisdiction of the
Justice Ministry. Restriction orders consist of treatment orders applicable to those with mental disorders and
政策創造研究 第 5 号（2011年11月）74
confi nement orders applicable to alcohol or drug abusers.
5）
What types of standardized forms are there in public/government policy design? How are they effective, and what, if
any, are their drawbacks (side effects, limits, etc.)? Research on such points is, frankly speaking, still at the primitive
stage of development. As well, criteria for classifying standardized or similar forms of policy measures differ from
person to person. There are only a few examples of notable research to date, including those by Weimer and Vining
(1989), Carabresi and Bobbitt (1978), Hood (1983), and Lowi (1979).
6）
Other constraints that must always be kept in mind when designing a public policy include (1) hard-to-mediate confl ict
in values between different segments of society and (2) the existence of a high degree of uncertainties. These constraints
are not discussed here owing to space limitations.
References
Adachi, Yukio (2005) The Modes of Thinking in the Policy Processes: Towards a General Theory of Public Policies (in
Japanese:
『政策学的思考とは何か
―
公共政策学原論の試み』
), Keiso-shobo.
Adachi, Yukio (2009) What is Public Policy Studies? (in Japanese:
『公共政策学とは何か』
), Minerva-shobo.
Dworkin, Ronald (1977) Taking Rights Seriously, Duckworth.
Linder, Stephen H. and B. Guy Peters (1985) “From Social Theory to Policy Design”, Journal of Public Policy 4-3.
Pressman, Jeffrey L. and Aaron Wildavsky (1979) Implementation, University of California Press.
