Chapter 1
Understanding Policy Decisions
Abstract The point of departure of the analysis is the recognition that major
policy change is extremely difﬁcult. However from time to time this actually
happens and therefore it seems necessary to understand when and how this is
possible. After a short excursus on the special characteristics of contemporary
public policies, four different theoretical models of how decisions are taken are
presented and discussed. The conclusion is the deﬁnition of a conceptual framework
specifying which are the variables that inﬂuence policy outcomes and in
particular the possibility to adopt non-marginal policy change.
Keywords Decision Á Public policy Á Incrementalism
1.1 The Decisional Problem
We live in difﬁcult times. ‘‘Making the necessary reforms’’, as any government in
the world knows all too well, is easier said than done. Taking decisions able to
signiﬁcantly alter the way collective problems are tackled, is particularly difﬁcult
in all democratic regimes.
The arguments used to explain this difﬁculty in governing and innovating
usually mention the lack of agreement among the societal actors about the goals,
the increasing level of institutional fragmentation, the difﬁcult relationship
between politics and administration, the fact that the coordination between the
different public bodies does not work well or that the implementing procedures can
put at risk what was established in the decisional phase, both as regards timing and
costs. This list could of course continue.
What current explanations often have in common is the fact they resort to
arguments at the level of the system. If it is so difﬁcult to make decisions—especially
important ones—there must be something wrong at a general level, making it
necessary to change some institutional or political features that hinder what is
needed. This is the so-called macro-negative approach that searches for and ﬁnds
B. Dente, Understanding Policy Decisions, PoliMI SpringerBriefs,
DOI: 10.1007/978-3-319-02520-9_1, Ó The Author(s) 2014
1
the general causes of the malfunctions of the political-administrative system
(Dunn 1981, pp. 53–54). As Robert Putnam taught us when explaining the efﬁciency
differentials of institutions between Northern and Southern Italy (Putnam 1993),
such features can be generated by a very long history.
This general, systemic nature of the reasons for failure has various effects. In
the ﬁrst place, because it refers to features that everyone knows, it seems plausible
and persuasive: we all tend to prefer short explanations of phenomena rather than
ﬁnd the speciﬁc cause for every single one. Secondly, it generates a feeling of
dismay: if changing what should be changed is so difﬁcult, maybe it’s worth giving
up. Political apathy is also due to this feeling of impotence. Finally, if the reasons
that make it so hard are at a systemic level, it follows that the innovators, or
presumed as such, are free from any responsibility for failure: we do in fact tend to
forgive the leaders we like for not keeping their electoral promises, blaming the
general features of the system for the negative outcomes.
However, there are two objections that cannot be easily overlooked.
The ﬁrst is that many of the features under suspicion are actually constituent,
even fundamental, characteristics of modern democracies, of the political systems
that try to lead highly differentiated societies.
Just to give an example, institutional fragmentation is the other face of
decentralization and federalism. It would be totally pointless to expect that the rise
in resources and legal powers of the sub-national governments has no consequences
at the level of the decisional system. But this does not mean that federal
states cannot make reforms, as some superﬁcial observers are tempted to say, or
even that to do so is more difﬁcult than in unitary centralised states: who would
claim that decisional efﬁciency is lower in Switzerland—one of the most decentralised
countries of the planet—than in Italy? And in any case the trend towards
greater decentralisation is widespread in all western democracies, and not only
there (Marks and Hooghe 2003).
So, the decentralization of responsibilities, the separation of powers, citizens’
guarantees, the opportunity to use direct democracy and the resort to the judicial
system against a decision of the public authorities, represent some of the basic
elements of modern democracies. Of course these features do have a price, at least
at the decisional level, but they also deﬁnitely have beneﬁts that are even more
important, like the impetuous socio-economic development of western democracies
in the past century proves. This is especially true if compared to the much
more modest performance of authoritarian and centralized regimes, even when
they resorted to planning and coordination tools that on paper should have been a
lot more efﬁcient.
As Charles Lindblom teaches us, societies and political systems based on
preferences and interactions tend to work better than societies based on the reason:
they recognizes the dangers of human failure, do not expect to know what people’s
real needs are simply going by some theory, but are based on the need to supply
suitable answers to citizens’ preferences even when they are contradictory or
irrational (Lindblom 1977, Chap. 19, pp. 247–260).
2 1 Understanding Policy Decisions
Or, to quote Luigi Bobbio, ‘‘democracy doesn’t live in Gordium’’ in the sense
that if deciding also means ‘‘to cut’’, to discard some of the alternatives, ‘‘this
doesn’t mean that the most effective tool for this operation is a sword, or its
modern equivalents, whether they be the will of the majority or the rationality of a
leader-manager’’. Actually, says the author, to follow the example of Alexander
the Great who cuts the knot rather than undo it, is not necessarily a good thing:
those who behave like this ‘‘are not in a better position to conquer Asia, like the
young prince of Macedonia, but simply risk to go nowhere’’ (Bobbio 1996, p. 8).
The second problem we have to face is that, in spite of their plausibility,
‘‘macro-negative’’ explanations don’t always work. Important reforms do take
place also in systems with similar features to the ones considered the cause of
immobilism. Sometimes important decisions are made in a short amount of time
and with acceptable costs. The macro explanations of decisional blockages
resemble the well-known myth according to which a bumblebee should never rise
from the ﬂoor. To recall Dunn, we need a micro-positive approach to discover the
speciﬁc circumstances under which important decisions are taken, also in order to
check their transferability to other contexts.
The fact is that if the systemic features considered the cause of decisional
inefﬁciency are often the other side of democracy, and if they still do not fully
explain what actually happens (since important decisions are made and implemented),
it follows that even the therapies identiﬁed to cure decisional block
syndromes, delays and costs of non-decisions are not suitable at all. These therapies
are based on a mix between the attempt to increase coordination in the
decisional process, an element that belongs to the rational decision making model
that we will present in Sect. 1.4.1, and the advocacy of inclusive strategies open
to the participation of all those interested that we will discuss on Chap. 6.
Consistently with their structure, current explanations of failures search for general
therapies, only sometimes distinguishing the different types of policy decisions.
This could be a pointless struggle since it is legitimate to suspect that it is probably
impossible to ﬁnd a single institutional solution for a whole set of decisions.
On the contrary, it is necessary to go further in-depth in the analysis of the
decisional processes to identify the basic elements that could explain why certain
results are achieved.
This is the path we will follow in this book.
To do this, we will unwaveringly take one point of view that will guide us
throughout the whole book. We will put ourselves in the innovator’s shoes, namely
the person who intends to change the content of a public policy in a non-marginal
way. Chapter 6 will in fact introduce the different strategies available to the
innovator to achieve his/her goals. In other words, even if it is obvious that the
features of the system within which each single decisional process takes place
inﬂuence the outcomes of the process itself, they do not completely determine the
decisional ﬁeld. In most cases, general conditions being equal, it is the combination
of the different elements of the process that explain the success or failure of
the attempt to reform.
1.1 The Decisional Problem 3
This book aims to propose a conceptual framework that allows understanding
decisions in the public sphere. In this sense, the book has an explanatory function,
meaning it deﬁnes the variables that can inﬂuence the processes and therefore
(contribute to) determine the outcomes. Such a framework can be used to explain
events of the past but can also be used in a prescriptive way in order to deﬁne the
courses of action an innovator must adopt to accomplish his/her project. This
means to try and answer the question whether the project itself is feasible from a
decisional point of view and check if there are elements that can hinder the
adoption of the decision. From a prescriptive point of view, it means to understand
what it is necessary to do in order to make the decision possible.
We must point out two further aspects.
First of all, the conceptual framework cannot provide any information about the
quality of the proposal. It cannot tell us if it is a good idea to tackle that speciﬁc
policy problem with that speciﬁc solution. The solution itself may not be suitable
to achieve the goal, or it may not be worthwhile (or may even be self-defeating)
trying to solve that problem at all. Understanding the dynamics of the decisional
process does not mean that one is immune from making mistakes.
However, and it is the second point, a good deal of innovative projects in public
policies fail, in the sense that they do not reach the implementation phase, because
the process has been badly managed. In other words, the correct application of the
analytical framework suggested in the following pages tries to avoid decisional
failure but does not guarantee a substantial success.
This does not mean that we want to introduce a ‘‘science’’ of decisions,
meaning that we have identiﬁed laws able to foresee what will happen. As we will
mention again, quoting Aaron Wildavsky, one of the founders of public policy
analysis, the discipline is more ‘‘art and craft’’ than science, strictly speaking
(Wildavsky 1979). The ability to lead a public policy decision through the many
difﬁculties it will come across requires a lot of bricolage and can be based on very
few certainties. However, identifying the main elements that contribute to the
outcomes, doing a bit of conceptual cleaning up and mostly, warning against the
sweeping generalisations that abound in this ﬁeld, can help in avoiding major
mistakes and identifying possible alternative courses of action.
1.2 What is a Policy Decision
The focus of the book is the concept of ‘‘policy decision’’, so it is from here that we
need to start. The problem would be quite simple if we were to only consider the
etymology of the word decision. To decide—from Latin de-coedere—means to cut,
to eliminate the available options till only one remains. To decide, therefore, would
be synonym of to choose, and decision a synonym of choice. Every day, we all make
a lot of decisions, sometimes consciously and more often unconsciously (or, better
said, automatically). This implies that there are some alternatives for many of the
actions we carry out. As soon as we wake up in the morning we can choose whether
4 1 Understanding Policy Decisions
to drink coffee, tea or a glass of whisky but this means that we have all three drinks
available at home, or that there is a café nearby. However, it is absolutely clear that
we cannot decide to wake up: waking up is the result of a natural process that
happens independently from our will. Just like we can decide to commit suicide by
jumping out of a window or, less tragically, to take a dive from a springboard, but we
cannot change our mind half way through: gravity does not allow us to.
These examples highlight some essential elements: decision implies an act of
will and the existence of alternatives. When one or the other lacks, there is no
decision.
The most obvious example of the importance of the act of deciding is probably
the wedding ceremony, where the questions ‘‘do you Carlo take Lucia to be your
lawful wedded wife’’ and ‘‘do you Lucia take Carlo to be your lawful wedded
husband’’ are explicitly asked. The idea is that in front of an action full of consequences
like choosing the person you will probably spend the rest of your life
with, it is necessary to clearly express the will to take such a big step, offering the
chance to respond negatively.
This example demonstrates what Bobbio wrote (1996, p. 11): ‘‘The act of
deciding, taken in itself … has little analytical interest and is often obscure at an
empirical level’’. Answering to why Carlo married Lucia, thus explaining his
decision, by stating that he did so because he positively answered the question the
celebrant asked him, would be tautological and stupid. Actually, the decision was
certainly made a long time before and its reasons sometimes are not clear to the
two participants either.
This is more true for decisions made in ﬁelds other than private life. To make an
example taken from 2010 Italian political news, the moment politicians decided to
approve a law reforming the way Italian Universities are managed, does not coincide
with the ﬁnal vote of the Senate of the Republic on the text approved by the Chamber
of Deputies. The decision was surely made long time before and the reform’s content
progressively changed also due to the intervention of a great number of actors.
So, together with the fact that a decision implies an act of will and the existence
of alternatives, a further fundamental element is represented by the process
through which the ﬁnal choice came to light, the sequence of elementary actions
and decisions that determined its content.
Studying a decision means to study the decisional processes, the mechanisms
through which we ‘‘decide to decide’’ and analyse or exclude possible alternatives
before reaching the ﬁnal result that can also be to ‘‘decide not to decide’’, leaving
things exactly as they are.
This last consideration highlights a fourth element. A decision must imply a
potential transformation of the world: if we decide to drink coffee at home, this
means that our coffee supplies will decrease. In other words a decision implies a
content.
At a ﬁrst level, the decision can regard the selection of the means available to
reach a goal. We can choose to drive our car to work, or to use public transport or,
if we have enough time, to walk. We can choose whether to spend our holidays at
the seaside or in the mountains, whether to send our children to one school or
1.2 What is a Policy Decision 5
another, and so on. In these three examples, the objectives are clear: we have to go
to work, we want to have a nice holiday and we wish to give our children the best
possible education.
Nevertheless, things aren’t always so simple: to really understand the reasons of
a decision, sometimes we have to go further back, given the fact that the real
decision regards what goals we want to pursue and not the means to choose in
order to do so. An example can be a career decision: the problem to solve is not
about the available alternatives, but about personal goals. The decision to become
a doctor can be driven by the will to be useful to others, by the desire to follow
family traditions, by the attractiveness of employment in an intellectually stimulating
sector, or just by the idea that this is a profession where it is possible to earn
good money. It is quite certain that individual choices to enter the medical school
involve a mix of all these different elements but then the real choice, the essential
decision, it to deﬁne what are the values to pursue that will probably lead to
consider completely different alternatives, such as the missionary, shopkeeper,
physicist or private banker.
The problem whether the ‘‘real’’ choice regards the means or the goals (that at
another level can certainly also be considered means: I want to be rich or I want to
help others because that is what my happiness depends on) is clearly crucial in
political decisions strictly speaking (MacKenzie 1982, pp. 16–17).
Before moving on to public decisions, it is good to add at least a further aspect
that refers to decisions made in private life. Up to now, in fact, we imagined that
choices have consequences only for the individual who makes them. But often
enough this is not true. Even though individual behaviour is always fundamental in
the concept of decision, there are many cases in which the results of the process
depend on the actions of different people. Going back to the example regarding
marriage, Carlo’s wish to marry Lucia would remain such (and probably be a
reason of unhappiness) if Lucia didn’t agree. This is where the complex problem
of how to combine individual preferences comes into consideration.
This problem is crucial for decisions made in the public sphere, that have
consequences—direct or indirect, real or potential—on a whole community, may it
be the population of a village, of a region, of a country or of the whole planet. At
least since the end of absolute kingdoms (but actually even before) and in any case
in democratic governments, these are collective decisions, meaning that the
interaction of a plurality of individuals is necessary. As we will see in the following
paragraph, this has important consequences.
Among the decisions made in the public sphere, the ones that regard public
policies are particularly relevant (policy decisions). It is important to underline that
they are not the only ones to be made in the public sphere: for example,
appointments (through elections or other kinds of designation) to positions with
legal authority are equally important.
Policy decisions, however, are certainly the most visible and interesting part of
the governing activity for citizens, so it is worth trying to explain what a public
policy is in the most accurate way possible.
6 1 Understanding Policy Decisions
One widespread deﬁnition is the one proposed by Dye (1987, p. 1) who states that
a public policy is ‘‘whatever governments choose to do or not to do’’. With the same
spirit, (Mény and Thoenig 1989, p. 129) propose the following deﬁnition: ‘‘a public
policy is the product of the activity of an authority invested with public power
[puissance, in the original version, author’s note] and governmental legitimacy’’.
More detailed is the deﬁnition contained in a textbook translated into various
languages according to which a public policy is the ‘‘connection of intentionally
consistent decisions and activities taken from different public actors, and sometimes
private ones,….in order to solve in a targeted way a problem which, politically,
is deﬁned as collective’’ (Knoepfel et al. 2001, p. 29).
It is clear that, while Dye, Mény and Thoenig state that public policies are just
the activities carried out by public institutions, Knoepfel and his co-authors, even
taking their participation for granted, admit that actors can also be private,
introducing two further qualiﬁcations: the ﬁrst is that the actions must be consistent,
at least in the intentions, and the second that they need to refer to the
existence of a collective problem. At a higher level of abstraction, we can deﬁne a
public policy as a set of decisions and activities that are linked to the solution of a
collective problem, meaning ‘‘an unrealized need, value or opportunity
which…may be attained through public action’’ (Dunn 1981, p. 60).
Within this deﬁnition there are no limitations regarding the consistency of the
actions (those who oppose to the problem’s solution must also be considered actors
of the process), nor the necessary presence of public institutions (even if they
probably will be present). Instead, what is crucial is the existence of a collective
problem, the outlines of which are inevitably subject to an interpretative activity
by the analyst, even if he/she is obviously obliged to take into consideration what
the actors think and how they deﬁne the problem. Nor are there limitations as to
the nature of the problems: in history and in different places, it is totally obvious
that there have been important variations of what was considered a collective
problem, or better said, a problem requiring public intervention.
The reason why we prefer to choose a wider deﬁnition is mainly connected to
the ambition of this book, anticipated in the previous pages and that will be further
detailed in Chap. 6. As we said, we will unwaveringly assume a point of view
putting ourselves in the innovator’s shoes. Innovators are those who want to
substantially modify the content of a public policy and are not necessarily public
actors, even though they almost always try to inﬂuence the behaviour of institutions.
Assuming a necessary link between government activity and the existence of
public policies can hide the fact that behind many policy reforms there are social
actors, experts, interest groups, private individuals.
We could actually even go further, asserting that a policy, for being public, only
needs the presence of a collective problem: is the fact that many cultural policies
are promoted and ﬁnancially supported by private foundations or companies, so
decisive to think that they are not public policies? And also, the invention of
microcredit as a tool for development policies in the poorest countries has generally
been attributed to the Grameen Bank and to its founder Mohammed Yunus.
It is an activity that is mostly carried out by private subjects, but surely the
1.2 What is a Policy Decision 7
problem that it is trying to solve is collective, and in fact it receives funding in free
grants or soft loans by public institutions and non-proﬁt organizations. This is not
the right place to analyse this matter in depth, and in any case, as we will see, the
fact that an activity is carried out in the political sphere has important consequences
on how decisions are made.
However, the deﬁnition we gave allows us to look at the decisions we are
interested in, without assuming that we only have to look at the governments’
activities.
As a consequence, and remembering that a decision involves an act of will that
can regard both the goals and the means and will probably involve many actors, we
can deﬁne a policy decision as the process of choice between alternative ways to
solve a collective problem.
1.3 Typical Features of Contemporary Public Policies
As we just mentioned, the emergence of collective problems, or better said, the
recognition that a problem is collective, has been object of modiﬁcations in time
and in space. What is considered as clearly public in one country, is not in another.
For example, the need to save up money to ‘‘send children to college’’ is something
natural in the USA, but not in the United Kingdom, and therefore the rise of
university fees decided by Tony Blair’s Labour government, and later by David
Cameron’s coalition government, generated harsh discussions in the political and
social spheres. On the other hand, in Europe the introduction of a strict legislation
aiming at preventing the creation of monopolies and oligopolies, similar to the
American one, was considered an inappropriate interference of the State in the
functioning of private economy for a very long time.
These differences among countries are real and consistent, but over the last
250 years it has still been possible to witness a development of public policies that
allows us to highlight the typical features of the current age. This is clearly an
essential step: only by understanding the nature of collective problems and how we
can solve them nowadays, we can imagine how an innovator might intervene to
introduce a small or big change in public policies.
The starting point is represented by the American and French revolutions and
by the gradual but widespread success of the liberal state model. It is in this phase
that some of the basic features of the modern public administration are born, some
of which are still with us. Max Weber skilfully summarized them in the deﬁnition
of legal-rational bureaucracy. At the basis of this model are the tasks and the tools
of the state. The basic challenge the liberal State has to deal with is to ensure
public order, at the same time leaving citizens as free as possible to pursue their
own interests. For this reason, the tools used are essentially regulatory, aimed at
determining individual and collective behaviours, deﬁning general rules that must
be respected, or making certain activities possible only after the authorization of
the public administrations (permits, licenses, concessions, etc.). The administrative
8 1 Understanding Policy Decisions
activity was to be totally dominated by the law so that the liberal State becomes a
synonym of rule of law, deﬁnition that can basically be summarized as follows:
public authorities can only do what is prescribed, should do everything that is
prescribed and they must act only in the way prescribed by the laws approved
by the bodies that have legislative power.
It is important to underline that the XIX century liberal State also carried out
activities that were not regulatory in nature: it maintained permanent armies, it
built roads, channels and bridges, it organized and sometimes managed public
education, and so on. However, it carried out these activities as if they were
regulatory activities. Tendering procedures for public works, for instance, are, in
theory, aimed at selecting the best offer, but in fact the main preoccupation is to
strictly follow the rules in order to avoid any suspicion of lack of impartiality in
assessing the different offers. From this point of view it is apparent how the delay
in the building of the infrastructure is considered a lesser evil than a minor
infringement of the formal rules in the selection of the contractor.
Things change, sometimes very deeply, with the welfare state, i.e. with the
enlargement of the tasks of public powers. This ‘‘great transformation’’ consisted
of assigning the State the responsibility to solve the so-called market failures in the
production of public goods but, especially, to guarantee economic and social
development, full employment, the reduction of citizens’ uncertainties through
health and social security systems. It is in this phase that administrations grow
strongly, as do the resources they absorb (to give an idea, between 1950 and 1985,
the public expenditure as a ratio of GDP goes from 35.3 to 47.3 % in Great Britain
and from 23.6 to 68.5 % in Sweden). Obviously this transformation entailed a
parallel change in the tools of government.
Not only public interests are to be protected by the increasing use of ﬁnancial
transfers (just think about retirement beneﬁts), by the use of positive and negative
incentives (in policies aimed at favouring economic development, as well as environmental
protection) and through the direct production of goods and services (from
education to health), but what is more important, the legislative tools at the basis of
the public action change their meaning. From being ‘‘conditional programmes’’—
formulated in the form ‘‘if…then’’ (for example: if speciﬁc circumstances occur, the
requesting citizen has the right to have the building permit)—they become ‘‘goal
programmes’’, in which the objectives to be reached, and not only the legal procedures,
are established. This is when planning becomes the fundamental form of
action of public powers, by basically deﬁning the available resources, their distribution
among the different policy ﬁelds and the short and medium-term objectives
that must be achieved by carrying out all the activities required.
The immediate consequence is the need to enlarge the autonomy and the discretion
of bureaucracies, but also to abandon uniformity, that was a distinguishing
feature up to that moment. The expansion of the direct production of goods and
services by public administrations gives a great impulse to the decentralization of
responsibilities to sub-national levels, through the creation of new levels of government
(for example, regions in Italy, France and Spain), through the
1.3 Typical Features of Contemporary Public Policies 9
strengthening of the State ﬁeld ofﬁces, and the increase of local administration
tasks. This gradual organizational differentiation has two main objectives: the ﬁrst
is to adapt the organizational structure to its functions as much as possible and the
second is to multiply the areas in which political control is exerted over the
bureaucracy.
All these changes actually happen over a rather long period and, in many cases,
without the need to face the problem of deﬁning a new general model. This is
partly due to the fact that the State’s new tasks did not completely replace the
previous ones and that the principles of the rule of law continue to be valid also
after the creation of the welfare state. If it is possible to clearly identify the exact
moment of the process of transition from the absolute state to the liberal state (the
American and French revolutions), the same is not true for the transition from the
liberal state to the welfare state. It was certainly favoured by important external
events (the 1929 economic crisis, the New Deal under Roosevelt, and, most of all,
World War II), but it happened subtly through a progressive enlargement of public
expenditure and taxation.
Therefore, the administrative model of the welfare state has less deﬁned features
compared to the liberal state. Not only was there no Max Weber able to
systematize its essential features, but also the overlapping of new and traditional
principles, along with the high differentiation of the functions generated a rather
complex situation.
However, some elements of this model are clearly recognizable.
The most important one is the emphasis on effectiveness, i.e. on the success of
public policies. This represents a distinguishing feature of the welfare state and it
becomes the new legitimation model of public power. Even if in most cases this
functional legitimation does not fully replace the legal-rational legitimation of the
previous phase, it is clearly recognizable and represents the basis of public powers
during the XX century.
During the last phase of the XX century, however, the welfare state model
entered a crisis leaving space for a further change that has probably not come to an
end yet.
The main reason for the change regards the modiﬁcation of collective problems
and therefore of the tasks assigned to administrations. It is the different nature of
the policies of contemporary states that explains the transformation of governmental
structures.
As an example, we can consider the following three points:
1. The development of economic globalization, the increase in the markets’
dimensions and of the ﬁnancial market in particular, has weakened many of the
tools governments could use to manage the economy; this does not mean that
the population’s demand for economic development and full employment has
decreased nor that governments are considered less responsible for the prevention
of and for the exit from economic and/or ﬁnancial crises.
2. A vast improvement in the knowledge in the ﬁeld of natural sciences has shown
the growing interdependencies between industrial development and
10 1 Understanding Policy Decisions
environmental transformations, at a local and at a global level; the ‘‘sustainable
development’’ imperative poses very difﬁcult and urgent challenges for
governments.
3. The unbalanced global economic development has fostered migration ﬂows as
never before, that deeply changed western societies; an open question regards
how to guarantee public order, broadly speaking, in a multi-ethnic society
where large minorities do not share anymore a lot of the values that were in the
past considered fundamental.
These three examples identify the type of policy problems that governments
face nowadays. Obviously it is impossible to forecast if and when the present
phase will come to an end. However we can try to point out some fundamental
features of contemporary policy processes.
The ﬁrst of these features is the increase in decisional complexity. Today we
see a growth of the decisional networks, both on the vertical axis (different geographical
scales) as well as on the horizontal one (especially relations between
public and private actors).
Starting from this last dimension, new types of actors enter the decisional
processes. Just think about the creation of independent administrative authorities—
bureaucratic bodies that do not respond to elected political representatives—with
the task of regulating a series of key sectors, from monetary policy to competition
protection, from consumer protection to privacy protection. Or think about the core
role of Non-Governmental Organizations (NGOs), sometimes composed by professionals
as well as by volunteers, in the implementation and sometimes also in
the formulation of several public policies, from social services to development aid
to the poorest countries. Again, let’s consider the expansion of the so-called Civil
Society Organisations (CSOs) that act as ‘‘watchdogs’’ over the government’s
activity, often in relation with the independent authorities we mentioned before.
All these new actors join in with the traditional ones and the result is a pluralisation
of the points of view inside the decisional processes, increasing the gap
between the actual ways in which public decisions are made and what is foreseen
by constitutional law.
On the vertical axis, the sum of the two trends towards globalization and
territorial decentralization led to the creation of the term multi-level governance,
needed to indicate how in almost all policy sectors, ﬁnal results depend on the
actions and the decisions made by different subjects operating at different territorial
levels: global fora, continental bodies (just think of the European Union),
national states, regional authorities, local communities, etc. This means that it is
often impossible to identify the authors of the decisions, with sometimes devastating
consequences on policy accountability.
A second feature of public policies in the contemporary age regards the increase
of uncertainty, To say it in a few words, governments often do not know if their
1.3 Typical Features of Contemporary Public Policies 11
decisions will solve the collective problem they are facing, or if they will make it
worse. What is under discussion today is whether the preferred alternative risks
generating negative effects (negative externalities, in economic terms) in other
ﬁelds or even if it is negative for the problem itself. Hence, a series of dilemmas
apparently without a solution. Is nuclear energy an effective and efﬁcient answer to
development needs, or does it expose to unacceptable risks and/or load future
generations with unbearable costs? Is building infrastructures for vehicle mobility
(roads, parking spaces, etc.) a way to improve mobility or is it an incentive to the
use of private vehicles increasing congestion?
Various factors determine this growing uncertainty.
First of all, it is connected to the increasing decisional complexity that we
discussed previously. If the effectiveness of a local policy also depends on what
will be established in an international treaty, it is clear that the actors do not
control a key element of the issue they have to face and, therefore, their forecasts
on the effectiveness of the choice are groundless.
Secondly, the acceleration of the changes caused by globalization processes
increases the chances of exogenous shocks making it impossible to foresee if and
when the trends will peak and change direction. Financial crises and technological
breakthroughs represent examples of the shocks that inﬂuence policy effectiveness.
But, at a different level and with certainly more serious consequences, who would
have said, 50 years ago, that we would have experienced a growth of religious
fundamentalisms in the transition from the second to the third millennium?
Thirdly, the development of knowledge and scientiﬁc progress widely increased
our capacity to identify possible relations between different phenomena without a
correspondent increase in our capacity to tackle them. Using a medical metaphor,
the development of diagnostic capacities, the ability to identify the different factors
at work, has not been matched by a corresponding progress in ﬁnding the
appropriate therapies.
This means that we have to honestly admit that there are problems we just do
not know how to solve, meaning that we are just ignorant about them. The existence
of economic theories that are ferociously competitive about the factors that
can determine the economic development of a territory, shows that we are just not
certain which behaviours are able to reach a widely shared goal.
Finally, and this is the third of the main features of contemporary public policies,
we are witnessing an increase in conﬂicts among social groups, among
political actors and between citizens and public authorities. Some of these conﬂicts
may be cyclical, meaning that a period in which choices are highly shared is
followed by another in which contrasts seem to be more intense. However as
regards the relationship between citizens and public authorities, it is reasonable to
assume that it is a structural phenomenon. Evidence of this are the opinion polls
that in all developed countries witness a signiﬁcant loss of trust in government.
Further evidence is the increasing role, quite novel in the sector of public policies,
12 1 Understanding Policy Decisions
played by the courts, that are more and more often called to solve conﬂicts where
social groups criticize the decisions of public bodies. The use of direct democracy
is also growing in many countries with successful referendum initiatives. In
conclusion, the method to transmit political demand and the ability of the elected
representatives to respond seem to have got stuck. This can be the effect of a
political system that evolved in a self-referential manner, losing contact with its
electors. Or, and more likely, the reasons of the transformation lie in the modiﬁcation
of the values shared by the citizens of developed countries: the loss of trust
in an indeﬁnite social and economic progress corresponds to the emergency of new
values (the so-called post-materialist values—Inglehart 1977) and in any case, the
intolerance of the need for sacriﬁce (which in some cases means giving up small or
big privileges) for a future that we no longer believe in. Cause and effect of these
changes is also the progressive de-ideologisation, at least along the lines of the
political cultures that developed in the XIX and XX century, and the birth of new
identities and new feelings of belonging that are often very ancient and refer to
religion, ethnic groups, territorial and linguistic afﬁliation, and many other
dimensions. However that may be, this social fragmentation seems to deeply
characterize the world of public policies too and it is certainly at the basis of the
conﬂicts that often take policy makers by surprise.
The typical features of the policy making processes (complexity, uncertainty
and conﬂict) do not only regard large problems with long range consequences. We
actually ﬁnd exactly the same features if it comes to approving a European
directive, a national law or the construction of a parking space in an urban area.
The NIMBY syndrome (Not In My Backyard), that is the populations’ refusal of
the localization of a wide range of infrastructures, demonstrates how it is not the
size of the intervention that generates conﬂicts and untreatable problems, but it is
the conﬂuence of a series of factors that we will see later.
1.4 Decisional Models
The previous considerations give us very confused image of policy decisions,
meaning the processes through which the solutions to collective problems are
selected. Hence the need to adopt a theoretical model, i.e. to identify the variables
that can be used to investigate the whole class of phenomena that we decided to
call ‘‘policy decisions’’ and to make explicit assumptions about the behaviours that
inﬂuence their outcomes.
As Allison (1971) taught us in his study of the Cuban missiles crisis, the models
are ‘‘conceptual lenses’’ we cannot do without and that, implicitly or explicitly,
consciously or unconsciously, we use in order to describe and explain the reality.
In the ﬁrst page of Allison book we ﬁnd the following quotation by Alexis de
Tocqueville:
1.3 Typical Features of Contemporary Public Policies 13
I have come across men of letters who have written history books without taking part in
public affairs, and politicians who have concerned themselves with producing events
without thinking about them. I have observed that the ﬁrst are always inclined to ﬁnd
general causes, while the second … are prone to imagine that everything is attributable to
particular incidents, and that the wires they pull are the same as those that move the world.
It is to be presumed that both are equally deceived.
The fact that in order to ﬁnd a meaning of the world that surrounds us it is
necessary to use theory, and not only empirical investigation, can seem counterintuitive,
but it is not less true. In fact we need conceptual models so much, that we
even use them without knowing it. As J.M. Keynes said ‘‘Practical men, who
believe themselves to be quite exempt from any intellectual inﬂuence, are usually
the slaves of some defunct economist’’.
In the following pages, we will introduce four different decisional models. With
this term, we intend ‘‘an analytical construct that identiﬁes the essential elements’’ of
the decisional process and therefore ‘‘the decision maker, his cognitive features, the
activities of research for a solution, the modalities and the criteria of the choice and
most of all… the relations among these different elements’’ (Bobbio 1996, p. 13).
These models are often assigned a descriptive value even before an explanatory
one, either implicitly or explicitly; but often enough they are also assigned a prescriptive
value, meaning that they are used to indicate how decisions should be made
if we want to maximize effectiveness and efﬁciency in the solution of the problem.
In the ﬁnal paragraph of this chapter, we will present a model that is widely
based on one of those proposed in literature, but that includes some elements from
the other models and seems to be able to realistically take into account how policy
decisions actually work in the real world. The elements that are part of this model
will be analysed and operationalized in the following chapters.
1.4.1 The Rational Model
It is traditional to start the presentation of the conceptual models able to represent
the decisional process by the so-called rational model.
From Aristotle’s logic to the economic planning experimented by communist
regimes, passing through most of economic and management literature, the main
idea seems to be that the choices made by individuals should be—and usually
are—connected to a series of operations.
Basically, according to the rational model, the individual who has to make a
decision should:
1. put his/her values and goals in order of priority;
2. know all the possible means to reach the goals;
3. evaluate the consequences of each alternative;
4. calculate the costs associated to the choice of all the available alternatives;
5. choose the alternative, able to maximise the beneﬁts and to minimise the costs.
14 1 Understanding Policy Decisions
It is clear that, ﬁrst and foremost, this is a prescriptive model, as it assumes that
the best decision for the solution of the problem is the one that will be made at the
end of a process as similar as possible to the one above. As a matter of fact, public
and private management manuals contain a large number of prescriptions that refer
exactly to this model. Some of them have been mostly adopted for budget and
investment decisions, as for example the Planning, Programming, Budgeting
System in the USA in the 60s or the Rationalisation des Choix Budgetaires in
France. But the use of the cost-beneﬁt analysis is also explicitly foreseen by
national and European regulations in the case of Regulatory Impact Analysis,
considered the best way to minimize the risk of placing unjustiﬁed burdens on
citizens and ﬁrms.
Why this model has been and still is so successful?
First of all, this depends on the fact that, as noted by Bobbio (1996, p. 17), it
promises to ‘‘create a space taken away from politics’’, in which important choices,
not the ones on big options and general goals, but the ones representing the heart of
public policies, can be made exclusively at a technical level without requiring long
and hard debates. Making the right decision is just a matter of technically deﬁning
the form and the parameters of an equation and inserting data in a computer.
But a second reason for this model’s success is that it seems to explain
everything that happens in the real world of policy decisions and in particular their
failures.
In fact, the model assumptions are that (Bobbio 1996, p. 18):
1. a clear separation between aims and means is possible, and the former are fully
determined before the latter are chosen;
2. the decisional process can be carried out by a single decision maker or, in any
case, by a body able to express organized and not contradictory preferences;
3. the analysis of alternatives and relevant consequences is able dispel the main
uncertainty sources;
4. there are enough resources for the analysis and in particular there is enough
time to investigate the consequences of all possible options, and even before, to
gain the necessary knowledge on the existence of these options.
It is quite clear that the contemporary presence of all these conditions is almost
impossible, and therefore we always have a ready ‘‘explanation’’ for any unsatisfying
result. There were not enough resources, information or coordination,
communication was not appropriate, and so on. Most of the prescriptions that
management consultants formulate basically suggest that we have to try and make
decisional processes more rational, closer to the ideal model previously described.
And this is where the problem lies.
In fact, if the model were only used from a prescriptive point of view it would
probably be useless and quite expensive, but not too dangerous.
1.4 Decisional Models 15
The problem is that we all tend to use it from a descriptive and explicative point
of view, as an easy way to understand someone’s objectives by observing their
behaviour. The syllogism is therefore the following:
MAJOR PREMISE: All people wanting A choose X
MINOR PREMISE: Mario choose X
CONCLUSION: Mario wanted A
It is quite clear that the syllogism is based on a condition the (ontological) truth
of which we can never be sure about, but cast the ﬁrst stone if you never thought
like this. In everyday life we often think this way, and many times correctly.
Adapting an example proposed by Regonini (2001, p. 94), if we see a friend of
ours running down the road in a suit and tie, we imagine, and this is totally
plausible, that he is late for an appointment.
However, especially in the ﬁeld of policy decisions, due to decisional complexity,
uncertainty of the results and latent or open conﬂicts, basing the interpretation
of what happens on a similar model can lead to a misunderstanding of
what happened and, most of all, of why it happened. The search for who is
responsible for the 2008 ﬁnancial crisis shows a continuous ﬂuctuation around the
rational model: bankers (or governments) are guilty as they acted irrationally on
the basis of incomplete information and inconsistent conditions, or they are guilty
as they actually made totally rational choices from their point of view and pursued
objectives that were not in the common interest. The idea that the results arose for
totally different reasons from the intentions and from the knowledge of a bunch of
people is completely absent in the many conspiracy theories that we ﬁnd in the
newspapers or on the internet all too often.
It is obvious that the problem lies in the conditions given by the model and in
particular in the assumption that there is one decision maker acting in a unitary
logic, perfectly informed about the objectives, the available alternatives and the
consequences in terms of costs and beneﬁts of each alternative.
Do we therefore get rid of the good along with the bad and renounce to any
explanation of decisional processes based on logic and on the assumption that
individuals pursue their own interests?
1.4.2 The Bounded Rationality Model
Not necessarily, according to Herbert Simon (Nobel Prize in Economic Sciences)
in Administrative Behaviour, published right after the end of the Second World
War (Simon 1947).
The rationality of an actor, he says, lies not the fact that he is omniscient, that
he/she knows all the objectives, all the alternatives, all the consequences of each
alternative, but the fact that his/her behaviour is at least potentially purposive,
16 1 Understanding Policy Decisions
aimed at reaching a goal, even if this is not completely deﬁned at the beginning of
the process. The decision maker will try to behave consistently, but will inevitably
suffer from a series of intellectual limits:
1. a limited knowledge of the alternatives available;
2. a limited intellectual capacity (there is a limit to the number of issues or aspects
of the same issue that can be contemporarily considered);
3. a limited memory;
4. a limited attention span (the amount of time spent in searching for the solution
of the decisional problem cannot go beyond a certain limit).
In such a situation, the rational decision maker looks for satisfying courses of
actions, or actions that are ‘‘good enough’’ on the basis of the information he has,
avoiding any pretence of optimization, i.e. of maximisation of the effectiveness of
the solution.
The bounded rationality model is, in its essence, the following: the acceptance
of the cognitive limits and the explicit adoption of a less strict decisional criterion
compared to the one implicit in the rational model. It has a prescriptive value,
meaning that it suggests to accept the ﬁrst alternative that appears satisfactory
without searching any further, and a descriptive and explicative value, meaning
that it assumes that the decision maker’s choice not only needs not to be the one
best way to solve the problem, but can also be based on incomplete or even wrong
analyses.
This position is even more understandable if one takes into account that Simon
especially talks about decisions that are not made by one decision maker, but are
created inside complex organizational and institutional contexts and therefore
require coalitions. Moreover, Simon highlights how his model can essentially be
applied to new and big decisions, while routine decisions are usually made
according to standard operating procedures enforced by organizations in order to
minimize the chances of making mistakes.
Actually, this kind of mechanism explains a great deal of our individual
decisions much better than the previous one, from choosing a new car to choosing
the location of our summer holidays. If we had to follow the indications of the
rational model, we would risk keeping our old car or staying at home for the whole
summer holidays.
This last comment shows how the limited rationality model has got a key
element in common with the previous one, meaning the idea that the decision can
be referred to an individual able to put his/her preferences in a transitive priority
order (according to which if choice A is preferred to choice B and choice B to
choice C, then choice A will be preferred to choice C). As shown in the following
box, this is not always possible if the decision is made by a collective actor.
This is Condorcet’s paradox, later expanded by Kenneth Arrow in the so-called
impossibility theorem, according to which it is impossible to create a single social
1.4 Decisional Models 17
welfare function through the aggregation of individual preferences under
democracy conditions (for further information, see Dunn (1981, pp. 227–230)).
In other words we have to face the challenge to ﬁnd a model that appropriately
represents collective choices.
Condorcet’s Paradox
A commission is appointed in order to decide the most appropriate technology
for the electric power generation, choosing among solar energy (S),
nuclear energy (N) or fossil fuels like oil (O). The members of this commission
are one environmentalist, who we will call Green, one technologist,
who we will call White and an economist, who we will call Red. They are
three rational people, who are fully aware that their opinion may not necessarily
prevail and therefore they should have a second option in order to
possibly avoid the most negative one of the three. In a logic of sustainable
development, Green prefers renewable energies, therefore solar energy,
rather than fossil fuels, but he still prefers oil to nuclear energy which he
believes implies very serious risks. On the other hand and in a logic of
technological and industrial development, White prefers nuclear to solar
energy but is however quite in favour of the latter due to its potential compared
to the use of oil that is still based on old technologies. Finally, Red,
who only considers the cost of the KW/h produced, believes that in the
current situation the use of oil is more efﬁcient than nuclear energy, that
implies extremely high investments, but that the latter is much better than
solar energy, deeply inefﬁcient from an economic point of view. The following
matrix summarizes their positions.
Green White Red
First choice (S) better than (O) (N) better than (S) (O) better than (N)
Second choice (O) better than (N) (S) better than (O) (N) better than (S)
Consequence (S) better than (N) (N) better than (O) (O) better than (S)
What happens with a voting? Solar energy has two preferences (Green and
White) against oil that only has one (Red). Oil can however count on two
votes (Green and Red) against nuclear energy that only has one (White).
From the transitive property, it should follow that solar energy is preferred to
nuclear energy, but the situation is actually the opposite as we can see from
the matrix, with White and Red prevailing on Green.
18 1 Understanding Policy Decisions
1.4.3 The Incremental Model
The starting point of Charles Lindblom, the scholar who proposed the incremental
model, in an article dated 1959 signiﬁcantly titled ‘‘The science of muddling
through’’, is that the close observation of policy making processes highlights the
following recurring features [Lindblom (1959) cit. in Parsons (1995, p. 285)]:
• the values, objectives and empirical analysis of the action to be carried out are
not distinct, but closely linked,
• since aims and means are not distinct, the analysis of the appropriateness of the
means required to reach the aims is often limited,
• the test a ‘‘good’’ policy must pass is typically the fact that different analysts
agree on its adoption, without necessarily agreeing that it is the most appropriate
mean for a shared objective,
• the analysis is drastically limited in the sense that (a) important and possible
consequences are neglected, (b) important alternative policies are neglected and
(c) important values are neglected,
• a series of subsequent comparisons reduces or eliminates any trust in the theory.
The combined effect of these features is that the result of a policy making
process actually able to reach a conclusion, is usually a decision departs but a little
from the status quo, that is incremental. Lindblom wonders if everything only
depends on the total irrationality of decision makers, or if it corresponds to some
speciﬁc and not necessarily negative feature of contemporary political systems.
The answer to this question will progressively come to light during the following
40 years and essentially develops at two levels.
First of all, and not differently from Herbert Simon, Lindblom states that the
research for absolute rationality, that he calls synoptic, is meaningless as it is
impossible to reach and potentially damaging because it can lead to serious and
sometimes irreparable mistakes. Actors’ cognitive limits, the constraints deriving
from the political and institutional context and the impossibility to foresee the
insurgence of exogenous shocks are all factors that make the prescriptive usefulness
of the rational model minimal and its explanatory value non existing.
Nothing new up to here.
But Lindblom makes a further huge step forward by emphasizing the fact that
the choice emerging from the decisional process is almost never an act of will of
the single decision maker, but the product of social interaction. All public policy
decisions are co-produced by many actors, with different values, objectives and
action logics. Even when the choice can be formally referred to an individual
decision maker, in selecting the alternatives he/she will have to take into account
the preferences and the resources of other actors, in order to avoid the exercise
veto powers or sabotages during the implementation of the decision.
More precisely, Lindblom believes that the actors of policy making processes
are usually in a situation of partisan interdependence, meaning that they have
objectives and interests that are in a structural conﬂict although they need each
1.4 Decisional Models 19
other. A typical example could be the absolutely normal case of a decision that has
to be jointly made by various public bodies: the representatives of each administration
will do their best to maximise the advantages for their administration, but
they also know that, in the end, they will have to ﬁnd an agreement. This conﬁguration
explains the decisional criteria and the most likely results of decisional
processes: we choose what we can agree upon, therefore we will choose the closest
alternative to the status quo, the incremental decision.
Lindblom’s model immediately appeared to be highly realistic and it inﬂuenced
a large number of analyses and empirical researches.
Lindblom actually claimed that not only the analytically superiority of the
model, as it was able to better explain what happens in reality, but also its
appropriateness from a prescriptive point of view, as it gives useful information to
improve the type of analysis necessary to prepare the decisions.
In particular, there have been different versions of the incrementalism of the
analysis. The initial choice was to suggest to decision-makers the opportunity to
operate by ‘‘successive limited comparisons’’, in order to simplify knowledge
needs. This means to proceed sequentially, only assessing the options that depart
minimally from the status quo at the same time not taking into consideration the
larger consequences of said options. After this, Lindblom proposed a more complex
version of the same approach, called ‘‘disjointed incrementalism’’ and
‘‘strategic analysis’’. The most important point, however is the fact that the
incremental analysis is, in any case, better that the synoptic rational analysis,
because it is aware of its own incompleteness, while the synoptic analysis is
equally incomplete, but tends to hide it.
Furthermore, the incremental analysis works also because it is based on the
existence of a plurality of actors and the knowledge produced derives from their
interaction. This is clear in a speciﬁc type of interaction called negotiation: ‘‘when
a government decides to control salaries in order to ﬁght inﬂation, the challenge of
deﬁning the right level of salaries can go beyond anyone’s capacity…. In this case,
a form of interaction called negotiation among enterprise representatives, workers
and government in a tripartite commission’’ can deﬁne the acceptable salary
increases (Lindblom 1980, p. 27). But more in general, directly assigning to the
actors the task of generating the analysis that has to inform and inﬂuence the
decisional process, ensures that the knowledge produced will certainly be relevant
for some of them. On the contrary, to expect that the professionalization of the
decisional analysis is able to produce more appropriate information and models is
often deceitful [for a discussion of this issue, see Lindblom and Cohen (1979)].
In any case, the real importance of interaction among actors is the fact that it
determines better results.
First of all, it is not true that the incremental model ignores the need, and
sometimes the urgency, of deep policy changes: ‘‘a fast-moving sequence of small
changes can more speedily accomplish a dramatic alteration of the status quo than
can an only infrequent major policy change’’ (Lindblom 1979, p. 520).
20 1 Understanding Policy Decisions
Secondly Lindblom approach goes further beyond the proposal of a methodology
for decision making. He explicitly states that a society based on preferences
and on interactions tends to work better than one based on reason (Lindblom 1977,
Chap. 19, pp. 247–260). In the book The Intelligence of Democracy, leveraging on
the double meaning contained in the title, he states that understanding democracy
means to recognize its intelligence, implicit in the fact that in democratic systems
it is possible to have more points of view at the same time, and therefore to surpass
the cognitive limits of any unitary actor. Surely, the process resulting from this
could seem confused and contradictory, but a brief comparison of performances in
terms of innovation, economic development and social equity is sufﬁcient to
realize that authoritarian systems have nothing to teach to democratic ones.
Two ﬁnal remarks.
The ﬁrst is that, as seems clear, the unit of analysis in this model becomes the
decisional process, i.e. the set of interactions through which the ﬁnal decision is
generated. As stated by Bobbio (1996, p. 32) it is not just a matter of observing
‘‘the solitary path made by a single decisional centre’’. This transformation is
particularly important because, as we often said and will keep repeating, the point
of view assumed in this volume is the innovator’s, meaning the actor who tries to
introduce a non-incremental change in how a collective problem is dealt with:
understanding the context within which his action will be carried out is much more
important than prescribing how he should search for the best solution to the
problem.
The second remark is that the incremental model is deeply grounded in the
political analysis, meaning that it explicitly considers the fact that an important
part of the interactions have to deal with the ‘‘ﬁght for power’’, that is to say with
the competition to gain an authoritative position and with the need to inﬂuence
other people’s behaviour to achieve one’s own goals and interests. All this, in a
context where the distribution of resources among actors is always unequal and
often very unbalanced.
Even if, the higher the decisional complexity gets, the more decisional processes
that take place in the private sphere will tend to be similar to the ones
carried out in the public sphere, there is still a strong political aspect in public
policy decisions that Lindblom’s model clearly highlights when stating that the
essential decision making criterion is agreement among actors, meaning the
consensus at the basis of authority’s legitimation mechanisms in contemporary
political systems.
1.4.4 The ‘‘Garbage Can’’ Model
With the fourth, and last, decisional model that we will focus once more on decisions
in general, meaningin any organizational context. In brief, this model states that in all
decisional situations where there is ambiguity at the level of objectives (badly
deﬁned by the actors), of the usable technology (that isn’t very clear), and of actors’
1.4 Decisional Models 21
participation (that tends to vary in time) the decision derives from the accidental
meeting—mediated by the intervention of contingent factors—of problems, solutions,
participants and opportunities of choice. This is the core of the model proposed
by James March and Johann Olsen in 1979 that they called ‘‘garbage-can model’’.
The authors explain the metaphor as follows: ‘‘Suppose we view a choice opportunity
as a garbage can into which various problems and solutions are dumped by participants.
The mix of garbage in a single can depends partly on the labels attached to the
alternative cans; but it also depends on the what garbage it is produced at the moment,
on the mix of cans available and on the speed with which garbage is collected and
removed from the scene’’ (March and Olsen 1979, p. 26).
It is interesting to see that the starting point of this idea is represented by the
consideration that in a private company there are various objectives partly in
contrast with one another: to increase sales, to improve the market share, to
increase production and proﬁts. Nowadays, we would also add: to decrease debts
and to ensure liquidity, to produce dividends for shareholders and especially, to
support shareholder value in the market. In other words, even in organizations that
the economic theory deﬁnes monolithic, like ﬁrms, the decisional process actually
sees different groups of participants who negotiate to reach an acceptable compromise
as regards what decision it is better to make. Of course, this is even more
true in political and administrative contexts where public policies are decided.
This plurality of possible goals is translated in a decisional model, widely based
on randomness, essentially for two closely linked reasons.
The ﬁrst is that the actors’ preferences are not exogenous but are formed during
the process. For example, they depend on the preferences expressed by the other
actors: if X, who is my enemy, pushes for solution A, I will oppose it and will
bring forward alternative proposals. But they can also depend on a transformation
of the context, on the fact that I lost interest in participating, or that other and more
serious problems gained importance.
The second is that decisional processes are often carried out over very long
periods. It is believed that Keynes answered a critic by saying ‘‘When the facts
change, I change my mind. What do you do, Sir?’’, which means that the longer a
decisional process takes, the more likely it is that the same course of action can
have different consequences and meanings at different times, and that actors will
modify their behaviour. But time also determines another very important effect:
actors simply change, some enter and some leave the decisional arena.
Therefore it is the conceptualization of the decisional process, intended as the
actions aimed at ﬁnding the solution to a problem, that becomes critical: besides
problems in search of a solution, we also have solutions in search of problems and
participants in search of problems to solve, satisfaction for interests or alliances to
create.
Temporal alignment becomes thus the key element in order to interpret results,
meaning the combination of a problem and a solution in one of the garbage cans, at
one speciﬁc moment. This tends to happen accidentally and chance becomes the
main cause for the ‘‘happening’’ (as we cannot talk about choice and awareness) of
decisions and their outcomes.
22 1 Understanding Policy Decisions
The main assumptions of this model are at the basis of the conceptual framework
created by John Kingdon, called multiple stream approach, to explain how
policy issues are included in the political agenda (Kingdon 1984). Basically there
are three different streams: the stream of problems, made by those issues someone
believes need to be solved; the stream of public policies, a kind of primordial stew
where ideas and solutions sometimes emerge, sometimes disappear; and the
political stream that determines the importance and urgency of an issue on the
political agenda. Sometimes, but not always predictably, the three streams merge,
making it possible to deal with that particular issue using that particular idea.
It is quite clear that the garbage can model is essentially descriptive and
interpretative, and it is able to explain and justify a wide number of situations,
especially in the light of the typical features of the policy decisions we mentioned
in Sect. 1.3.
However, it would be wrong to think that it does not also contain prescriptive
elements.
At a ﬁrst level and in the event of extremely chaotic situations, it justiﬁes and
suggests the idea that, instead of trying to simplify the problem to make it more
manageable, it is better to make a random choice, hoping to stimulate a chain
reaction that sooner or later will lead to a simpler decisional structure, that we are
not able to foresee at present (Grandori 1984).
More in general, the representation of decisional processes as garbage cans
stimulates focus on the ways to ‘‘create and support identities, preferences and
resources that make a political community possible’’ (March and Olsen 1995,
p. 28). Not by chance, the same scholars who elaborated the garbage-can model
were central 20 years later in rediscovering institutions as means to provide some
order in a world made chaotic by rising complexity.
1.5 A Realistic Model of the Decisional Process
The following chart deliberately simpliﬁes what we can consider the essential
features of the four models discussed in Fig. 1.1.
Each one of these models contains important clues to understand how decisions
are made. So, the synoptic rational model attracts our attention to the fact that
individuals usually try to choose the best solution for the satisfaction of their
interests. The bounded rationality model reminds us that our knowledge is
imperfect and that we are often satisﬁed with the ﬁrst ‘‘rather good’’ solution that
we ﬁnd. The incremental model highlights how most of the decisions, especially in
public policies, are the result of compromises among actors with contrasting
interests. And, ﬁnally, the garbage can model shows how the passing of time is not
irrelevant, since it can make the connection between problem and solution either
possible or impossible.
We can imagine that, as the complexity of the decision grows, especially
intended as plurality of the participants’ points of view, the best way to understand
1.4 Decisional Models 23
what happened is to move towards the bottom of the chart. In the light of the
features of contemporary public policies this means that we suppose that in most
cases neither the rational/synoptic model nor the one proposed by Herbert Simon
are able to explain what really happened.
On the contrary, the prescriptive value, in the meaning of giving clear indications
on how it is possible to reach the best decision from a speciﬁc actor’s point
of view, tends to decrease, going from the top to the bottom. The comprehensive
rational model requires decision makers to make important analytical efforts that
are only apparently facilitated by the discovery of techniques like linear programming
or costs/beneﬁts analysis and that require the ability to make long-term
forecasts. On the other hand, the garbage can model does not contain real indications
about how to behave, apart from what we will shortly mention about
decisional timing.
However, as often repeated, both the analyst and the decision maker need a
model to refer to, otherwise they risk not being aware of their own assumptions
and therefore suggesting explanations or creating strategies that are inconsistent or
contradictory.
For this reason we will make a clear choice towards the incremental model,
meaning that we will assume that most policy decisions are made by actors with
contrasting interests who need to reach an agreement to achieve their own goals.
As we will see, this agreement does not mean that everybody will be happy with
the outcome of the process. The basic reason that leads us to prefer the model
proposed by Charles Lindblom is that it provides the best representation of the
conditions that usually take place in political/administrative contexts. It is not a
coincidence that politics has been deﬁned as ‘‘the art of compromise’’ and that not
only the creation of alliances within the power élite, but also the citizens’ consensus
are considered a key element for the stability and effectiveness of political
systems. Decisional processes characterized by complexity and uncertainty are
basically political, which means they deal with power.
This does not mean of course that Herbert Simon’s ideas on one side and March
and Olsen’s on the other, should not be taken into consideration and cannot enrich
the understanding of the phenomena we are interested in.
Decisional model Decision maker Cognitive conditions Decisional criterion
Synoptic rational Unitary Certainty Optimization
Bounded rationality Unitary /coalition Uncertainty Satisfaction
Incremental Partisan
interdependence
Partiality Mutual adjustment
Garbage can Changeable Ambiguity Chance
Fig. 1.1 Decisional models
24 1 Understanding Policy Decisions
In particular, Simon not only taught us that each actor has cognitive limits that
we have to take into account while interpreting or forecasting behaviour, but he
also pointed out that a rational actor does not need to be omniscient, he just has to
be purposive, trying to reach a goal even if it is not perfectly deﬁned since the
beginning of the process. This is why in the following pages we will consider the
contributions of the rational political theory and in particular the ones coming from
the application of the game theory: the actors of the incremental model also behave
rationally.
March and Olsen’s work shows how an actor interested in modifying the real
world should consider the context the decision develops in and in particular the
importance of the moment in which the choice between the available alternatives
is made. The ﬂow of events makes the speciﬁc connection between problems and
solutions possible at time t, while this would not happen at time t - 1 nor t ? 1.
The main indication of the model is to pay attention to the timing of the decision a
point that we will expand somewhat when considering the strategies available for
policy innovators.
The incremental model, however, has a fundamental advantage. It explains why
decisions made in the public sphere usually do not depart much from the status
quo. Therefore, it leaves an open door on the issue of how it is possible to
introduce non marginal changes, and this is the exact problem this volume tries to
give an answer to. It indicates the type of analysis that those interested in introducing
important changes in the ways of solving collective problems have to carry
out, since the main features of the model—partisan interdependence among
decision makers, mutual agreement as a decisional criterion, limitations to the
analysis—are valid regardless of the nature of the problem or of the radicalism of
the solutions proposed.
But if the incremental model tells us how the decisional process will develop, it
does not tell us, nor it could tell us, what are the decisions that will be made, nor,
in speciﬁc and general terms, what kind of decisions it is possible to make.
Therefore we have to go beyond Charles Lindblom’s conceptual framework and
to specify the variables that contribute to determine the possibility of an intentional
and non-incremental change of a public policy, of the way to deal with a collective
problem.
The model proposed in this volume and that will be explained in the following
chapters, can be summarized as follows:
The outcomes of a public policy decisional process depend on the interaction of different
types of actors with different goals and roles who, within a network that can have
different characteristics, exchange resources using different patterns of interaction, to
obtain a stake, within a given decisional context.
In the Chap. 2 we will concentrate on the actors, on the decisional networks and on
the resources, while in the subsequent chapters we will focus on the analysis of the
content of the decision, on the patterns of interaction and on the decisional context.
The challenge, or the general objective of this volume, is to understand what
speciﬁc combinations of these variables (let’s repeat them: actors, resources, stake,
1.5 A Realistic Model of the Decisional Process 25
patterns of interaction) make non incremental policy decisions possible, within a
speciﬁc decisional context and from the innovator’s point of view, considering that
in administrative and political systems actors have important cognitive limits and
are bound by the need to reach an agreement.
In an explanatory key, the model is needed to cast light on the elements that
explain how it was possible for a speciﬁc and important transformation of the ways
used to deal with a collective problem to take place (and, after all, this is a
common experience, although these events are quite rare). In a prescriptive key,
the model suggests to the reformers the type of analysis they have to make to
evaluate the decisional feasibility of their proposals.
A couple more warnings.
1. The conceptual framework proposed does not aim to deﬁne the features that an
institutional system has to assume to secure the decisional feasibility of
changes. Clearly, some of the acquisitions deriving from the application of the
model implicitly contain proposals of useful institutional transformations in
speciﬁc contexts or sectors. However, the validity of the model is relevant at the
level of the individual decision and the individual decisional process (and
therefore also to the decision to transform the institutional system), not at the
level of the political system in which they occur. Any inference from the microlevel
(where our analysis is placed) to the macro-level is completely undue.
2. As it should be clear, the model does not say anything about the substantial
quality of the innovation proposal, meaning that it is not able to predict if it will
be able to effectively solve the problem or be the correct solution to face that
speciﬁc problem. The model assumes that this analysis is carried out by
whoever proposes the change before and during the decisional process itself.
Certainly, the idea at the basis of Lindblom’s model is that knowledge generated
during social interaction is able to signiﬁcantly enrich the ways in which to
deal with collective problems. As we will see in Chap. 6, moreover, there is a
whole family of decisional strategies, the ‘‘inclusive’’ ones, that claim to be able to
improve the content of decisions. However, this result cannot be assured from the
beginning due to the cognitive limits we recalled many times and it is not necessarily
true that the solution mostly agreed on is also the most correct. What is
probably valid for the system—a society based on preferences and on interactions
is generally more likely to solve collective problems, compared to one based on
intellect—does not apply to the single decision. In any case, the ontological
uncertainty that accompanies many public policies, guarantees that very often
effective solutions to several problem simply do not exist.
Our model rather realistically assumes that the innovator proposes the solution
to the collective problem even if important modiﬁcations may occur between the
initial proposal and the ﬁnal decision. From the innovator’s point of view, the
decisional problem can be conceptualised as the existence of an unsatisﬁed
26 1 Understanding Policy Decisions
opportunity: the X alternative is available and able to solve problem Y. The
analysis of the decisional process is needed to identify the elements that made a
speciﬁc choice possible (i.e. decisional success).
This decisional success is possibly translated in a substantial failure, meaning
that not only the collective problem was not solved, but it even got worse due to
the decision made. The model is not able to say anything to this end, as it only
explains and assesses the feasibility of the choice. And however, if it is true that
sometimes it would have been better not to make any decision at all, it is also true
that ﬁghting for a politically (broadly speaking) impossible course of action is
certainly not a better alternative.
References
Allison, G.: Essence of Decision: Explaining the Cuban Missile Crisis. Harper and Collins, New
York (1971)
Bobbio, L.: La democrazia non abita a Gordio. Milano, Franco Angeli (1996)
Dunn, W.N.: Public Policy Analysis: An Introduction. Prentice Hall, Englewood Cliffs (1981)
Dye, T.R.: Understanding Public Policy, 3rd edn. Englewood Cliffs, Prentice Hall (1987)
Grandori, A.: A prescriptive contingency view of organizational decision making. Adm. Sci. Q.
29, 192–209 (1984)
Kingdon, J.W.: Agendas, Alternatives and Public Policies. Little Brown, Boston (1984)
Knoepfel, P., Larrue, C., Varone, F.: Analyse et pilotage des politiques publiques. Helbing &
Lichtenhahn, Basel (2001)
Lindblom, C.E.: The science of muddling through. Public Adm. Rev. 19, 78–88 (1959)
Lindblom, C.E.: Politics and Markets: The World’s Economic System. Basic Books, New York
(1977)
Lindblom, C.E.: Still muddling, not yet through, Public Adm. Rev. 39, 517–552 (1979)
Lindblom, C.E.: The policy making process, 2nd edn., Englewood Cliffs, Prentice Hall (1980)
Lindblom, C.E., Cohen, D.K.: Usable Knowledge: Social Science and Social Problem Solving.
Yale University Press, New Haven (1979)
MacKenzie, W.J.M.: Choice and Decision. In: McGrew, A.G., Wilson, M.J. (eds.) Decision
Making: Approaches and Analysis, pp. 16–18. Manchester University Press, Manchester
(1982)
March, J.C., Olsen, J.P.: Ambiguity and Choice in Organizations. Universitetsforlaget, Oslo
(1979)
March, J.C., Olsen, J.P.: Democratic Governance, Free Press, New York (1995)
Marks, G., Hooghe, L.: Unraveling the central state, but how? Types of multi-level governance.
Am. Polit. Sci. Rev. 97(2), 233–243 (2003)
Mény, Y., Thoenig, J.P.: Politiques Publiques. PUF, Paris (1989)
Parsons, W.: Public Policy: An Introduction to the Theory and Practice of Policy Analysis.
Edward Elgar, Cheltenham-Northampton (1995)
Putnam, R.D., Leonardi, R., Nanetti, R.: Making Democracy Work: Civic Traditions in Modern
Italy. Princeton University Press, Princeton (1993)
Regonini, G.: Capire le politiche pubbliche. il Mulino, Bologna (2001)
Simon, H.A.: Administrative Behavior: A Study of Decision-making Processes in Administrative
Organizations. Macmillan, New York (1947)
Wildavsky, A.: Speaking Truth to Power: The Art and Craft of Policy Analysis. Little Brown,
Boston (1979)
1.5 A Realistic Model of the Decisional Process 27