Policy Sciences 4 (1973), pp.

197-209
© Elsevier Scientific Publishing Company, Amsterdam--Printed in Scotland

The Policy Implementation

Process
THOMAS B. S M I T H
School of Political Science and Public Administration, Victoria U~iversity of Wellington, Wellington,
New Zealand

ABSTRACT
There is an implicit assumption in most policy studies that once a policy has been formulated the
policy will be implemented. This assumption is invalid for policies formulated in many Third
World nations and for types of policies in Western societies. Third World governments tend to
formulate broad, sweeping policies, and governmental bureaucracies often lack the capacity for
implementation. Interest groups, opposition parties, and affected individuals and groups often
attempt to influence tile implementation of policy rather than the formulation of policy.
A model of the policy implementation process is presented. Policy implementation is seen as a
tension generating force in society. Tensions are generated between and within four components of
the implementing process : idealized policy, implementing organization, target group, and
environmental factors. The tensions result in transaction patterns which may or may not match
the expectations of outcome of the policy formulators. The transaction patterns may become
crystallized into institutions. Both the transaction patterns and the institutions may generate
tensions which, by feedback to the policymakers and implementors, may support or reject further
implementation of the policy.
By application of the model, policymakers can attempt to minimize disruptive tensions which
can result in the failure of policy outcomes to match policy expectations.

A general a s s u m p t i o n is often m a d e by scholars who analyze policies a n d b u i l d m o d e l s

o f the policy processes t h a t once a policy has been m a d e the policy will be implemented.
This a s s u m p t i o n , in large part, accounts for the neglect o f the policy i m p l e m e n t a t i o n
process in the models constructed to explain p o l i c y m a k i n g o r to instruct p o t i c y m a k e r s
in m e t h o d s o f better policy making.
In this p a p e r the implications o f this a s s u m p t i o n will be e v a l u a t e d - - e s p e c i a l l y in
terms o f policy processes in T h i r d W o r l d nations. I will also construct a m o d e l o f the
policy i m p l e m e n t a t i o n process which m a y be useful in the analysis o f p o l i c y m a k i n g
a n d in the formulation o f policies.

Policymaking and Policy Implementation

There are a n u m b e r o f reasons for the neglect o f the i m p l e m e n t a t i o n o f p o l i c y in the
c u r r e n t models. As m e n t i o n e d before, there is an implicit a s s u m p t i o n in m o s t m o d e l s

197
that once a policy has been "made" by a government, the policy will be implemented
and the desired results of the policy will be near those expected by the policymakers#
This assumption rests upon certain political and organizational conditions. However,
these conditions may not be found in all societies. Indeed, the societies of the Third
World often are devoid of all or some of these conditions.
Recent interpretations of politics in Asian and African nations have indicated that,
unlike in many Western nations, the governmental policies seldom are the result of
demands and pressures by interested parties. As James C. Scott has written, "The open
clash of organized interests is often conspicuously absent during the formation of
legislation in these nations."z Indeed, the channels for such influence (political parties
and interest groups) may be undeveloped or suppressed by official governmental
action. Governments will initiate policies often without consultation with the
interested or affected individuals or groups. According to Scott,
To conclude from this, however, that the public has little or no effect on the eventual
"output" of government would be completely unwarranted. Between the passage of
legislation and its actual implementation lies an entirely different political arena t h a t . . .
has a great effect on the execution of policy. 3
The real role of interest groups and other interested groups, including political
parties, comes when the policies are implemented by the government. It is at the
implementation stage of the policy process when the policy may be abandoned by the
government, implemented or modified to meet the demands of the interested parties.
This process represents a radical departure from that found in many Western nations.
Interests and opposition in Western societies focus upon the policymakers. If a policy
is made (or not made) the various interests know that this decision will be implemented.
And often there is little they can do about it at that stage. Research and theory building
also has focused on the "inputs" or upon the processes by which policies are made.
However, in Third World nations the emphasis by interested parties is upon the policy
itself or upon the bureaucrats who must implement the policy. 4
Even if a government of a Third World nation is committed to the implementation
of a particular policy, the bureaucracy that must implement the policy often is lacking
in the capacity to implement.5 This situation usually is not found in Western societies.

1 Citizens also make this assumption. According to Dye, "We assume that when Congress adopts
a policy and appropriates money for it, and when the executive branch organizes a program, hires
people, spends money, and carries out activities designed to implement the policy, the effects of the
policy wilt be felt by society and the effects will be those intended by the policy." Thomas Ro Dye,
Understanding Public Policy (Englewood Cliffs, N.J.: Prentice-Hall, 1972), p. 291.
2 James C. Scott, "Corruption, Machine Politics, and Political Change," American Political Science
Review 63 (December 1969), p. 1142,
3 Ibid., p. 1142.
4 Scott writes, "Influence at the enforcement stage often takes the form of 'corruption' and has
seldom been treated as the alternative means of interest articulation which in fact it is." Ibid., p. 1142.
For three cases of interest concentration upon policy implementation in India, see Robert W. Stern,
The Process o f Opposition in India (Chicago: University of Chicago Press, 1970).
5 On administrative capacity in Third World nations, see Ralph Briabmati, ed., Political and
Administrative Development (Durham: Duke University Press, t969); Fred W. Riggs,Administration
in Developing Countries (Boston: Houghton Mifflin, 1964); and Edward W. Weidner, ed., Development
Administration in Asia (Durham: Duke University Press, 1970).

198
Generally, the Western bureaucracies are relatively efficient and effective in policy
implementation. For the Third World nations a myriad of factors can contribute to
the emasculation of any government policy: lack of qualified personnel, insufficient
direction and control from political leaders, opposition to the policy itself, corruption,
etc. The process by which the government made the policy may have been the most
rational, elaborate, "modern" method. But this may not be meaningful if the
administrators cannot implement it.
One reason the bureaucracies in the Third World nations find it difficult to im-
plement governmental policies is the nature of the policies which these political
systems must formulate. Policies in Western nations tend to be incremental in nature.6
Normally, the policies deviate little from past policies in the same area of concern.
The fact is that the Third World nations have not been able to afford the luxury of
incremental policymaking. Policies tend to be ambitious, sweeping programs designed
to bring about development and social reform.7 This was especially true in the years
immediately following independence or the overthrow of traditional rulers. Budgets
soared, massive new programs were formulated, and foreign assistance money,
advisors, and packaged programs penetrated the policymaking systems. I think it is
accepted knowledge that incremental policies are much easier to implement than are
ambitious non-incremental ones. Therefore, many policies in the non-Western nations
have not met with success in attaining their objectives.
The problems of policy implementation may be more widespread than commonly
acknowledged. Modern Western nations such as the United States may have difficulty
implementing certain kinds of programs, especially if the programs are of a new, non-
incremental nature. The experience of America's War on Poverty programs bears this
out.8 In another policy area, Schneider, in his case study of urban mass transport,
notes that "There may be legal or technical difficulties in implementation."9 He points
out that,
Too often the public reads that the most modern transportation system in the world has
been approved, but a year later finds that the new system is little different from the un-
pleasant existing lines. In these situations, it is not surprising that the feedback to decision
makers results in unfavorable attitudes toward future mass-transportation projects.10
Models of policymaking have not taken seriously enough the importance of policy
implementation in the analysis of public policy. For the policy process in the Third
World nations, the implementation of policies is an acute phase in the process. I
contend that we must re-evaluate the importance of policy implementation if we are
to explain policy processes or to seek to inform policymakers on how they should go
about making policy. The model of the policy implementation process outlined in this

6 Dror points out that policies in Western nations normally are politically and economically
feasible. Y. Dror, Public Policy-making Reexamined (San Francisco: Chandler, 1968), p. 89.
7 Ibid., pp. 109-110.
8 See Thomas Dye's evaluation of the innovative social programs in his Understanding Public Policy,
op. cir. Also see, Daniel P. Moynihan, ;~aximum Feasible Misunderstanding (New York: Free Press,
1969).
9 Lewis M. Schneider, "Urban Mass-Transportation: A Survey of tlhe Decision-Making Process,"
in Raymond A. Bauer and Kenneth J. Gergen, eds., The Study of Policy Formation (New York: Free
Press, t968), p. 277. lo Ibid., pp. 277-278.

199
paper is a step in that direction. I will outline a model of the implementation process
and identify the key variables in that process.

Change and Governmental Policy

In order to develop a model of the policy implementation process, it is essential to
view the policy process from the point of view of social and political change. Govern-
mental policies are designed to induce changes in society J1 By the implementation of
governmental policies, old patterns of interaction and institutions are abolished or
modified and new patterns of action and institutions are created. Governmental
policies may be designed to result in sweeping changes in life styles, or they may be
incremental in nature and deviate little from past policies. But each policy is an
attempt by a government to induce changes in patterns of interaction between or
within individuals, groups or institutions. With this perspective as the basis of a model
of policy implementation, the process of societal change can be examined for concepts
and approaches that will provide insights into the policy implementation process.
It is useful to view society and policy processes in systemic terms. However, one of
the general problems with the use of systems theory in the social sciences has been the
static quality of the systems models. 12 The analytic systems developed were not geared
to the explanation of change processes except as related to the maintenance of
equilibrium within the system. In a recent, lucid book Walter Buckley concentrates
upon the "other side" of systems analysis--the change processes within social
systems. 13 His interpretation is immensely useful in the analysis of policy processes.
Buckley defines two aspects of system regulation:
M o r p h o s t a s i s - - r e f e r s to those processes in complex systems-environment changes
that tend to preserve or maintain the system's given form, organization, or state. I4
M o r p h o g e n e s i s - - r e f e r s to those processes which tend to elaborate or change a
system's given form, structure or stage.15
Buckley then develops a model of the morphogenic process. In this model the root
cause of changes in society is identified:
The internal source of dynamics for the ongoing process is the continuous generation of
varying degrees of tension, "stress," or "strain" within and between the interacting
components; such tension is now being recognized as an inherent and essential char-
acteristic of such systems. 16
There are a number of other social scientists who identify the major initiating factor
of societal change to be the tensions within and between components of a society.
Robert Chin, in a brief essay, writes of the sources of tensions in society:
We find built-in differences, gaps of ignorance, misperceptions, or differential perspectives,

11For comprehensive definitions of public policy, see Austin Ranney, "The Study of Policy
Content," in Austin Ranney, ed., Political Science andPublic Policy (Chicago: Markham, 1968), p. 7.
Also see Dror, op. cit., pp. 12-17.
12See A. P. Hall and R. E. Eagen, "Definition of a System," General Systems, 1 (t956), 18-28.
Also see Oran R. Young, Systems of Political Science (Englewood Cliffs, N.J. : Prentice-Halt, 1968).
13Walter Buckley, 5beiology and ?vIodern Systems Analysis (Englewood Cliffs, N.J. : Prentice-Hall,
1967). 14Ibid., p. 58. i5 Ibid., p. 58. i6 Ibid., p. 129.

200
 internal changes in a component, reactive adjustments and defenses, and the requirement
of system survival generating tensionsS
The importance of identifying societal tensions is noted by Chin:
The identification of and analysis of how tensions operate in a system are by all odds the
major utility of systems analysis for practitioners of change. The dynamics of a living
system are exposed for observation through utilizing the concept of tension, stress and
strain, and conflict. ~8
George K. Zollschan provides a more comprehensive model of how societal tensions
result in changes in society. Zollschan and his associates prefer to use the term
"exigency" in place of tensions. An exigency is defined as "a discrepancy (for a person)
between a consciously or unconsciously desired or expected state of affairs and an
actual situation.'q9 Zollschan identifies three main types of discrepancies which may,
alone or in combination, make up an exigency :z0
1. A discrepancy between a legitimate pattern or arrangement and an actual situation.
2. A discrepancy between a prediction (or explanation or expectation) and an
observation.
3. A discrepancy between a desired objective and what is actually achieved.
Exigencies (tensions) are linked directly with changes in society. Zotlschan postulates
that tensions "must trigger a series of phase processes."zl The phase process is outlined
in Figure 1.
I....... I
T~n~iori~ill ~ Articulation~ Action~ Institutions
SocietyI

Fig. 1. The Zollschan institutionalization model.

Zollschan defines the phase articulation as "the r e c o g n i t i o n . . , of the existence of

the exigency and the postulation of goals for its removal, prevention, or ameliora-
tion."22 After tensions have been articulated and recognized, the tensions may then
be referred to as "needs." Action is defined by Zollschan as "locomotion toward
postulated or unconscious goals."23 Whether or not action will or will not take place
depends upon the nature of the tensions and the social and political system in which
the tension and articulation is taking place. The final state, "institutionalization," is
conceptualized by Zollschan as the change in old, stable, crystallized patterns of
interaction and/or the substitution of newly crystallized patterns for old patterns.Z4
Zollschan's phase model has many merits; however, several modifications must be
made before it will have utility in explaining the implementation of policy. That
17 Robert Chin, "The Utility of System Models and Development Models," in Jason L. Finkle and
Richard W. Gable, eds., Political Development and Social Change (New York: John Wiley and Sons,
t966), p. 12. 18Ibid., p. 14.
t9 George K. Zollschan, "Working Papers on the Theory of Institutionalization," in George K.
Zollschan and Walter Hirsch, eds., Explorations in Social Change (Boston: Houghton Mifflin, 1967),
p. 89. 2o Zollschan regards exigencies as occurring only at the individual level.
21 1bid., p. 89. 221bid., p. 90. 231bid., p. 91.
24 For other, related definitions of the concept of institutionalization, see Bertram Gross, The
Management of Organizations (New York: Free Press, 1964); Philip Selznick, Leadership in Administra-
tion (Evanstan, Ill. : Row, Peterson and Co., 1957); and Buckley, op. cit.

201
tensions are the chief causal factors in the change process is accepted. But the chain
process of articulation to action should be examined more closely. Perhaps the phase
between tensions and institutions should be bridged by a more general phase which
can be called the transaction phase. Within this phase, the tensions concerned could
be articulated and could manifest themselves in new behavioral patterns and new
relationships. These patterns of action might or might not become permanent,
institutionalized patterns. Most societal tensions probably do not end in the institu-
tionalization of new patterns and relationships, but result only in the creation of
uncrystallized action patterns which I will call transactions.
Tension~ irt
Transactions ~ Instit utkm~
Society [ 1
¢ ~ - - - Feedbalk Phase ~t.. . . . . .

Fig. 2. A general model of the change process.

Another consideration in the process of change concerns the never-ending process

of tension generation. Newly created transaction patterns or institutions may be the
catalyst for new and powerful tensions which may be supportive or disruptive to the
new or modified patterns of relationships. With this in mind a feedback phase must
be added to the model of the societal change process. The modified model is dia-
grammed in Figure 2. In this model societal tensions can result in transaction patterns
which, in turn, may or may not result in their institutionalization. The transaction
patterns and institutions may generate tensions which may cause further modifications
in their structure or processes.

The Policy Implementation Process

Governmental policies have been defined as deliberate action by a government to
establish new transaction patterns or institutions or to change established patterns
within old institutions. Policy formulated by a government, then, serves as a tension-
generating force in socieO:. While policies are implemented, tensions, strains and
conflicts are experienced by those who are implementing the policy and by those
affected by the policy. The tensions generated by the implementation of policies may
cause transaction patterns and, in some instances, the establishment of institutions
required for the realization of policy goals. Also, the process of policy implementation
may generate tensions that trigger changes in other related institutionsY It is im-
portant to note that institutions need not emerge from every government policy that
is implemented. Transaction patterns may be the only result.
Once public policies are regarded as tension generating forces in society, it is a
necessary step to consider the context of the implementation of policies--the policy
relevant components which form the tension generating matrix. I wish to identify four
such components which are important in the policy implementation process: (1) the
idealized poIicy, (2) the implementing organization, (3) the target group, and (4) environ-
mental factors. Figure 3 represents the model of the policy implementation process.
25 See~Dye,op. cir., pp. 291-299.

202
 The model presented in Figure 3 represents the central core of the policy implementa-
tion process. As policy is implemented, interaction within and between the components
of the policy implementation system result in discrepancies and tensions. The tensions
result in transaction patterns--non-permanent patterns related to the aims and goals
of the policy. The transaction patterns may or may not result in institutionalization.
Feedback in the form of relieved tensions or increased tension is introduced back into
the tension generation matrix from transaction patterns and institutions.

--1
POLtCYMAKING
PROCESS ~ POLICY ~
¢/
Poli
I
]
TENSIONS

TRA SACIIONS

" FEEDBACK ~l INSTITUTIONS

Fig. 3. A model of the policy implementation process.

In the remainder of the paper I will add to the policy implementation model by
identifying the major variables in each of the components of the policy matrix.
Secondly, I will attempt to further define the concept of tensions and make it especially
relevant to policy implementation. Lastly, I will elaborate upon the transaction,
institutions and feedback phases of the model.

The Idealized Policy

The idealized policy is defined as the idealized patterns of interaction that the
policymakers are attempting to induce. There are four relevant categories of variables
of the idealized policy.
The Formal Policy--The formal decisional statement, law, or program that the
government is attempting to implement. In short, what form does the policy take ?
The Type of Policy--The type of policy can be of three categories:
(a) Policies may be complex or they may be simple in nature. This distinction is
similar to broad, non-incremental policies and small-scale incremental ones.
(b) Policies may be categorized as organizational or non-organizational. An
organizational policy requires the modification (or the establishment) of a formal
organization.Z6 Non-organizational policies, on the other hand, call for the establish-
ment of patterns of interaction outside the formal organizational context.
(c) Policies also may be classified as distributive, re-distributive, regulatory, self-
regulatory, or emotive-symbolic.27

26To use an Indian example, the implementation of the Panchayati Raj program required the
establishment of formal organizations such as village, block and district panchayats. However, a
block panchayat may introduce policies of taxation that would not require the establishment of a new
formal organization.
27 Regarding the utility of a policy-arena type classification, see Theodore J. Lowi, "American
Business, Public Policy, Case Studies and Political Theory," WormPolitics 16 (July 1964), 677-715;
Robert H. Salisbury, "The Analysis of Public Policies," in Austin Ranney, ed., PoliticalScience and

203
15
 The Program--There are three aspects o f the p r o g r a m o f the policy:
(a) Intensity o f support. The degree to which the government is committed to the
implementation of the policy.28
(b) The source o f the policy. Is the policy required to satisfy needs and demands
in the society or was it formulated with little demand or support ?29
(c) Scope. Is the policy b r o a d in nature and universal in scope or is it a con-
centration u p o n a small geographic or subject area ?30
Images of the Policy--It is essential to consider the images the policy invokes in the
society. The images o f those affected by the policy" and those who implement the policy
are most important.3a

The Target Group

The target group is defined as those who are required to adapt new patterns of
interaction by the policy.32 They are the people within organizations or groups most
affected by the policy. It is they who must change to meet the demands o f the policy.
Several factors are relevant here:
1. The degree of organization or institutionalization of the target group.33
2. The leadership of the target group. Leadership can be aligned for or against the
policy. Or the leadership may be indifferent to the policy. The nature of the leadership
is also important.
3. The prior policy experience of the target group. Has the group been affected by
governmental policies in the past ? W h a t has been their experience and response to
governmental policy. Are they compliant, rebellious or indifferent ?34

The ImplementingOrganization
The implementing organization is responsible for the implementation o f the policy.
I n most instances, the organization is a unit o f the governmental bureaucracy.3S There
are three key variables to consider in the implementation o f policy.

Public Policy (Chicago: Markham, 1968), pp. 151-175; and T. Alexander Smith, "Towards a Com-
parative Theory of the Policy Processes," ComparativePolitics 1 (July 1969), 498-515.
2s Fred Riggs has noted the distinction between goals and policy. He contends that much of what
passes for "policy" in developing nations should be called "goals," Goals are not meant to be seriously
implemented. See Riggs, op. eit., pp. 332-334.
29 One aspect of this variable refers to foreign aid programs. See Eugene B. Mihaly's interpretation
of the impact of massive foreign assistance on Nepal. Eugene B. Mihaly, Foreign Aid and Politics in
Nepal (New York: Oxford University Press, 1965).
30 In this regard, see Lewis A. Froman, "An Analysis of Public Policies in Cities," Journal of
Politics 29 (February 1967), 94-108.
31 In regard to the images of the American poverty program, see Aaron Wildavsky, "The Empty-
Headed Blues: Black Rebellion and White Reaction," The Public' Interest 4 (Spring 1968).
32 Dye, op. eit., p. 292.
33 A number of studies have indicated that the more highly institutionalized a group or organization,
the more difficult it is to initiate change.
34 One wonders how receptive the American low-income, black, ghetto dweller will be to future
government programs. Surely, their current attitudes toward future policy have been influenced by
past policies.
35 1 do not wish to revive the old policy-administration dichotomy. An implementing organization
may have been in the role of policymaker. The organization can implement its own policy.

204
 1. The structure andpersonnel--The stability of the structure and the qualifications
of the personnel who must implement the policy are important to understanding
implementation. An unstable administrative organization and unqualified personnel
may reduce the capacity to implement.
2. The leadership of the administrative organization--This variable, like the leader-
ship of the target group, refers to the style and nature of the leadership.
3. The implementing program and capacity--The program and capacity of the
implementing organization refers to the intensity and care taken to organize for the
implementation and to the general capacity of the organization to meet the objectives
o f program implementation.

Environmental Factors
The environmental factors are those factors which can influence or be influenced
by the policy implementation.36 Environmental factors can be thought of as a sort of
constraining corridor through which the implementation of policy must be forced.
For differing kinds of policy, differing cultural, social, political, and economic
conditions may prevail. For example, in policies related to local self-government in
Third World nations, the basic cultural and social life-styles at the village level may
be an environmental constraint of great magnitude.

Tensions
After identifying the major component categories within the policy implementation
processes, it is in order to explore further the nature of tensions. Tensions may occur
both within and between the four component categories of idealized policy, implement-
ing organization, target group, and environmental factors. An example of a within-
component tension would be the discrepancy in the implementing organization when
an administrative unit is instructed to implement a policy for which the unit has
inadequate personnel in numbers or skills for policy execution. A between-component
tension may occur between the target group and the ideal policy component when the
policymakers expect a policy to be welcomed by the target group and the actual
reception is a hostile one.
It is necessary to examine at what societal level tensions may be prevalent. Zollschan,
in his model, conceives of tensions only at the individual level. His argument is that all
systemic tensions can be reduced to those felt by individuals. This restriction is unduly
harsh and need not be followed in studying policy implementation. In addition to
tensions of an individual nature, there also can be tensions on the group level. Although,
in analysis, group tensions may be reduced to individuals within the group, it may be
more meaningful to speak of group tensions. For example, the clerks in a government
department may feel that they are underpaid and overworked. But their supervisor
may be convinced that they are overpaid and underworked. In cases such as this it is
proper to speak of the clerks and supervisors as groups rather than as individuals.
Tensions can be identified also at the formal structural level. For instance any office,
in order to carry out a policy, must contain enough qualified personnel to do the task.
36See Riggs, op. cit., pp. 426-429.

205
If the office is understaffed, the assigned tasks cannot be completed. A definite tension
is prevalent and can be operationalized if the level of structure is taken into considera-
tion. Tensions, then, can be conceptualized at the individual, group or structural
levels. But, as Zollschan rightly points out, the tensions are all reduceable to the
individual level.
Zollschan, in his model of social change discusses three types of discrepancies.
However, these categories must be expanded for applicability in policy implementa-
tion. The tension matrix presented in Table 1 is more comprehensive and more
applicable than the one proposed by Zollschan. Within the tension framework there
are ten types of discrepancies which can occur at three different levels (individual,
group, and structure).
TABLE 1
Policy Implementation Tension Matrix

Ideal Actual Perceived Expected

Ideal 7 1 2 3
Actual 8 4 5
Perceived 9 6
Expected 10

Type 1. A discrepancy between what is an ideal state and an actual situation. Example:
In an ideal situation in a government office there would be a full administrative staff
for that office. The actual situation may be that key posts are vacant. Therefore a
tension of a structural level regarding the discrepancy between the ideal and the actual
exists,
Type 2. A discrepancy between an ideal state and a perceived situation. Example: At
the individual level a district administrator may perceive a village leader as ignorant
and resistant to change. The administrator may wish for the ideal situation in which
village leaders would be well informed and welcome change in their villages.
Type 3. A discrepancy between an ideal situation and an expected situation. Example:
The ideal situation in village level planning is for village governing bodies to make
plans that will be implemented. The district officer in charge of planning may not
expect village governments to be able to implement any of their plans.
Type 4. A discrepancy between actual and perceived situations, Example: A district
political leader may perceive of the police in his district as unfairly engaging in
political repression. The actual situation may be that the police are dispensing with
their duties in an impartial and restrained manner.
Type 5. A discrepancy between an actual situation and an expected situation. Example:
A researcher from the central government might expect to find records kept at a
district office to be orderly and up-to-date. The actual record-keeping by a district
office may fall short of this.
Type 6. A discrepancy between an expected situation and a perceived situation,
Example: An official sent by a central government to help a lower level governmental

206
unit formulate a plan of action may expect the personnel of the lower unit to be
appreciative of his help and implement his suggestions. The official may receive
shabby treatment because his suggestions are not good ones. But he may perceive
that the poor treatment results from his caste or class background.
Tensions can occur within the ideal, actual, perceived or expected categories. If
individuals, groups or structures within the system have varying images regarding the
variables in the category, discrepancies can occur. Discrepancies of this type have been
numbered types 7, 8, 9 and 10 in the discrepancy matrix in Table 1.
The following would be an example of a discrepancy of this type. Within the
"ideal" category a national planning commission may have in mind an ideal state of
regional planning in which all key decisions regarding the style and rate of regional
growth would be determined by the national planning commission. The regional
leaders, however, may have in mind an ideal regional planning system whereby the
regional governments are autonomous and determine the style and rate of development
in their areas.
The ten types of discrepancies, then, can be prevalent at the individual, group and
structural levels. And the discrepancies may occur within each of the four implementa-
tion contextual components or between the components. By operationalizing the
concept of tension in this way, the tensions within the policy implementation frame-
work, though varied and complex in nature, can be identified and measured in
intensity.

Transaction Patterns
Transaction patterns have been identified as uncrystallized patterns of interaction.
These patterns are responses to the tensions, stresses and strains within and between
the component parts of the policy implementation context. The patterns, at this stage,
do not have the permanence of institutions.
The transaction phase of the policy implementation process is important. Many
times governmental policy is intended to result only in transaction patterns and not
in the establishment of permanent institutions. And even though institution building
may be one of the prime goals of a governmental policy, institutions may not be
formed. Particular attention must be paid to whether or not the policy is being
implemented. In Third World nations, as well as in many of the developed
Western ones, there are many "idealized" policies written into law or placed in the
constitution.37 Programs supported in a half-hearted way often are never implemented
with vigor. In these cases, the transaction behavior seldom results in the attainment
of the desired objectives of the policy.
In exploring the properties of the transaction phase of the policy implementation
process, the researcher should attempt to identify the actions that emerge when the
policy is being implemented. In examining a policy which requires the establishment
of a popularly elected local governmental unit, one might examine manifestations of
response to tensions in the way the members of the unit are chosen, the unit's relation-
ship with the dominant administrative sector in the area, and the unit's ability to
37Ibid., pp. 15-19. Riggs calls this "formalism."

207
satisfy prescribed functions. Also to be taken into account would be the initial
relationships between the local unit and the population it is supposed to represent.

Institutions
Since the policy implementation process is an ongoing process, it is difficult to
determine when (if ever) interaction patterns become firmly crystallized into institu-
tions. The exact degree of institutionalization is difficult to determine, but several
measures can serve as guidelines. These are :38
1. The institution's ability to survive in its environment.
2. The extent to which the institution is viewed by surrounding societal components
to have value (both autonomy and influence).
3. Whether the relationship patterns become normative for other social components.
It should be stressed that it is a simple process for governments to establish adminis-
trative or political organizations for the purpose of policy implementation. The
structural arrangement functions with the support of the government that established
it. Whether the organization eventually turns into an institution is another matter. A
test of an established organization's institutionality is whether demands for the
activity would be sufficient to force its continuance.39
In the policy implementation process a methodology for determining whether a
policy meets its objectives might be to examine the newly established patterns or
institutions in comparison with the operation of the "idealized" policy goals of its
creators. In order to do this, the idealized policy's functions and relationships can be
itemized and then compared with the actual outcome of the implementation process.
It is possible that in some instances the patterns of the idealized policy have never
materialized or that the outcome of the change process has been to crystallize patterns
of resistance to the idealized policy's program and goals. 40
In considering the degree of institutionalization, the time factor must be taken into
account. Complex, non-incremental policies may take more time for patterns to
institutionalize than would simple incremental policies. Although it may be impossible
to predict how long it takes for certain processes to become institutionalized, enough
time should be allowed for definite patterns to emerge before conclusions are made
regarding the institutional outcomes of policy implementation. It should be considered
that the policy implementation process as indicated by this model contends that the
policy process is a continuous one that may or may not have a final definite end.

Feedback
Once transaction patterns emerge or institutionalization occurs, they can serve as
tension generating entities. The tensions can--though not always--relate back to the
implementation matrix of the idealized policy, target group, implementing organiza-

38 From Milton J. Esman and Hans C. Blaise, "Institution Building Research: Some Guiding
Concepts," National Institute o f Public Administration JOurnal (Karachi) 5 (June 1966), 88-98.
39 Agencies or departments established may soon appear to gain a life of their own. See F. E. Rourke,
Bureaucracy, Politics and Public Policy (Boston: Little, Brown, 1969).
40 See Dye, op. c#., pp. 291-299.

208
tion and environmental factors. 41 By the creation of new patterns and new institutions,
tensions may be generated which could lead to further changes within the system. In
this regard, there is no assurance that the former state of the system will come into
existence again. There is also no guarantee that there will be a distinct end to any
implementation process. The new tensions may be supportive or disruptive to the new
patterns and institutions brought about by the policy implementation. An example of
disruptive tensions might be the case of the establishment of a viable system of
autonomous local government able to process all problems of its area. An efficient
and effective functioning of this local government system may evoke tension from
some ruling leaders at the national level who felt that local units should be tightly
controlled by the central government. In this case, the new tensions created might lead
to the reshaping of governmental policy and the issuance of instructions to adminis-
trative units to exercise more control over local government.
With the feedback process built into the model, it becomes difficult to speak of the
"end product" of a policy implementation process. The feedback phase is an im-
portant part of the model because it clearly indicates that the policy process is an
ongoing, continuous process that may never have a final and definite end.

Conclusions
Given the importance of understanding the total policy process, it is essential that the
implementation process be explored and evaluated. The model presented in this paper
--while tentative in nature--is one method of explaining the process. Surely more
variables (and relationships between variables) can be added to the model as more
research and thinking is devoted to the implementation process.
An implication of the model for the making and execution of policy is that policy-
makers (especially those in Third World nations) should devote more effort to under-
standing and mastering implementation strategies. Cost-benefit analysis may not be
enough to ensure that policy outcomes match policy expectations. Policymakers, by
utilizing this model, can attempt to minimize discrepancies in the implementation
process and, with this goal in mind, can formulate and apply more effective policies.
41For some examples of American policy cycles, see Charles O. Jones, An Introduction to the Study
of Public Policy (Belmont, Calif.: Wadsworth, 1970), pp. 107-134.

209