Mata Kuliah Kebijakan Publik

Prodi Doktor IPDN

Prof. Dr. Nurliah Nurdin, MA

Implementation Theory
The Top-down/Bottom-up Debate
The Discovery of the ‘Missing Link’

The assumption made by the new students of implementation that their subject had been
neglected can be challenged. To some extent it was merely the case that political scientists
began to use a new concept ‘implementation’ in policy analysis and administrative studies.

Furthermore the neglect of the examination of administrative processes can be partly

attributed to the difficulties involved in looking into the ‘black box’, after parliamentary
processes, particularly in a secretive administrative culture like the British one.

In formal statements of constitutional law, implementers, like civil servants, have often
been entirely ignored.

While the topic of how to distinguish implementation from policy creation

influenced the top-down/bottom-up argument, it was only one component of a
larger difficulty of identifying the characteristics of a very complicated process that
occurred over time and space and involved various players. It will be observed that
implementation authors respond to complexity in a variety of ways.
While the top-down/bottom-up debate was heavily influenced by the
question of how to separate implementation from policy formation,
that was only part of a wider problem about how to identify the
features of a very complex process, occurring across time and space,
and involving multiple actors.

We have in mind here two considerations of this kind, which we will

find many of the implementation theorists struggling with:

1. Variations between policy issues, or types of policy issues; and

2. Variations between institutional contexts, which may include
questions about the extent to which generalizations apply outside
specific political systems or national contexts.
The Classical Top-Down Writers
Jeffrey Pressman and Aaron Wildavsky: The Founding Fathers

This particular formulation has been seen as responsible for a pessimistic tone since it
suggests that purposive action will be very difficult to achieve wherever there are multiple
actors.

It is perhaps more appropriate to use game theory rather than probability theory to analyse
them.

They argue that if action depends upon a number of links in an implementation chain, then
the degree of co-operation between agencies required to make those links has to be very close
to a 100% if a situation is not to occur in which a number of small deficits cumulatively create
a large shortfall.

Pressman and Wildavsky’s original work takes very much a ‘rational model’ approach: policy
sets goals; implementation research is concerned with considering what then makes the
achievement of those goals difficult.

It is of more than biographical interest to note that Wildavsky’s new collaborator, with whom
he wrote a new last chapter called ‘Implementation as Evolution’, was an Italian,
Giandomenico Majone.
Donald Van Meter and Carl Van Horn: System Building

In developing their theoretical framework Van Meter and Van Horn describe
themselves as having been ‘guided by three bodies of literature’ (1975: 453).

• Organization theory, and particularly work on organizational change–here they

recognize the importance of the concerns about organizational control in
sociological work influenced by Max Weber (1964);
• Studies of the impact of public policy and particularly of the impact of judicial
decisions, such as Dolbeare and Hammond’s study of the factors that influenced
responses to US Supreme Court rulings on school prayers (1971);
• Some studies of inter-governmental relations, in particular the work of Derthick
(1970, 1972) and, of course, Pressman and Wildavsky.
The six variables (surely they are in fact clusters of variables) identified in Figure 3.1 are:
 Policy standards and objectives
 The resources and incentives
 The quality of inter-organizational relationships
 The characteristics of the implementation agencies
 The economic, social and political environment;
 The ‘disposition’ or ‘response’ of the implementers
Eugene Bardach: Fixing the Game

Bardach suggests that implementation processes need to be

perceived as involving ‘games’, and he outlines the wide variety of
the games that may be played.

One of these concerns the need for great care in the ‘scenario
writing’ process, so as to structure the games in the right way to
achieve desired outcomes.

The other prescription from Bardach is that attention needs to be

given to ‘fixing the game’. This involves two related usages in the
notion of ‘fixing’, in colloquial American: as mending and as
something rather close to cheating.

We see in Bardach’s work the very clear exposition of a view that

implementation is a ‘political’ process, and that ‘successful’ imple-
mentation from a ‘top-down’ perspective must involve a very full
‘follow- through’.
Paul Sabatier and Daniel Mazmanian: Process Modelling

We see here a very clear distinction being made between policy formation
and policy implementation, but at the same time a recognition of a feed-
back process.
The factors impacting upon the implementation process are then seen as
falling under three headings (Sabatier and Mazmanian, 1980: 544):

• Factors affecting the ‘tractability of the problem’;

• ‘Nonstatutory variables affecting implementation’; and
• The ‘ability of the statute to structure implementation’.

Involving identifying Recommendations to

factors that will cause the ‘top’ about the steps
Both a methodology:
difficulties and factors to be taken to try to
that may be controlled control implementation

They have a great deal in common with the list

produced by the next theorists we will consider.
Brian Hogwood and Lewis Gunn: Recommendations for Policy Makers

It has already been noted that Hogwood and Gunn, like Sabatier and Mazmanian, offer
propositions that can be read as recommendations to policy makers. These are that
policy makers should ensure:

• Circumstances external to the implementing agency;

• Adequate time and sufficient resources are available;
• There no constraints in terms of overall resources;
• Policy to be implemented is based upon a valid theory of cause and effect;
• The relationship between cause and effect is direct;
• Single implementing agency need not depend upon other agencies for success;
• A complete understanding of, and agreement upon, the objectives to be achieved
• In moving towards agreed objectives it is possible to specify, in complete detail;
• Perfect communication among the various elements involved in the programme;
• Authority can demand and obtain perfect obedience
The Bottom-Up Challenge

Michael Lipsky: Street-Level Bureaucracy

Michael Lipsky’s analysis of the behaviour of front-line staff in policy

delivery agencies, whom he calls ‘street-level bureaucrats’, has had an
important influence upon implementation studies.

Lipsky argues that, therefore, to cope with the pressures upon them,
street-level bureaucrats develop methods of processing people in a
relatively routine and stereotyped way.

Such workers see themselves as cogs in a system, as oppressed by the

bureaucracy within which they work. Yet they often seem to have a
great deal of discretionary freedom and autonomy.
Benny Hjern: Implementation Structures

Hence Hjern and his colleagues saw activities as within ‘implementation

structures’ formed from ‘within pools of organizations’ and ‘formed through
processes of consensual self-selection

They used a methodology that, whilst starting from an identified pool of

relevant organizations, ‘snowballed’ to collect a sample of respondents who
were working together.

This is more than an argument about methodology. Hjern and Hull go on to

argue that:

“Once we are clear about who participates how and with

what effect in policy processes, then we can begin to think
about how politics and administration could and should be
(re-)combined in the policy process.”
Susan Barrett and Colin Fudge: Policy and Action

They argue that there is a tendency in the top-down

implementation literature to depoliticize the policy–action
relationship. Their alternative view emphasizes the continuing
political processes occurring throughout implementation.

In effect this suggests that it is very difficult to separate

implementation from policy formation.

However, if it is not possible to separate policy formation

from implementation, there is a difficulty in setting the
limits for an implementation study.

And even more seriously: How can effectiveness be

assessed in ‘getting something done’ or a compromise be
judged as achieving something as opposed to throwing
away an objective, without reference to at least someone’s
policy goals?
 Conclusions

As mentioned in the introduction to this chapter, the debate between the

top-down and bottom-up perspectives moved on to efforts to synthesize
the approaches, picking out key ideas from each. The methodological
elements in the debate were not, in themselves, particularly contentious.
The same was true of those elements in the debate that concerned the
most realistic way to perceive implementation processes.

The normative debate could not so easily be resolved, embodying as it did

alternative stances on democratic accountability; here synthesis
depended upon recognizing the legitimacy of complex formulations of
this topic. These issues are explored through the examination of writers
whom we loosely describe in the next chapter as ‘the synthesizers’.
THANK YOU