POLICY ANALYSIS IN PUBLIC ADMINISTRATION
Policy paradox: the art of political decision making – Deborah Stone
Introduction
Politicians have two goals: 1) Policy goals: the program/proposal they would like to see
accomplished/ defeated. 2) Political goals: politicians want to preserve/gain power, to accomplish
their policy goals.
Three aims of the book: 1) The rationality project misses the point of politics. The dark, self-
interested side of political conflict should be recognized. At the same time, politics is a valuable
creative process for social harmony. 2) The rationality project worships objectivity and seeks modes
of analysis that lead to the objectively best results for society. 3) The field of public policy is
dominated by economics and its model of society as a market (as opposed to a model of community).
The project of making public policy rational rests on three pillars. 1) A model of reasoning: rational
decision making: a) Identify objectives. b) Identify alternative courses of action for achieving objecti-
ves. c) Predict the possible consequences of each alternative. d) Evaluate the possible consequences
of each alternatives. e) Select the alternative that maximizes the attainment of objectives. 2) A model
of society: the market model. 3) A model of policy making: a production model, where policy is
created in an orderly sequence of stages.
Chapter 1 – The market and the polis
Market model: a social system in which individuals pursue their own welfare by exchanging things
with others whenever trades are mutually beneficial (» Robinson Crusoe society). Participants
compete with each other for scarce resources. Individuals act only to maximize their own self-
interest. Ultimately, competition raises the levels of economic well-being of society as a whole.
Polis: a model of political society. 1) It is a community (perhaps multiple) with ideas, images, will, and
effort quite apart from individual goals and behaviour. Unlike the market, a model of the polis must
assume collective will and collective effort. A community must have members and some way of defi-
ning them, which sometimes is the primary political issue (example: immigrants). A political
community is a group of people who live under the same political rules and structure of governance.
A cultural community is a group of people who share a culture and draw their identities from shared
language, history, and traditions. Membership in a community defines social, economic and political
rights. Mutual aid among members transforms a collection of individuals into a community. It’s also a
kind of social insurance. In the market model, insurance is a financial product that firms sell in order
to make a profit and buyers buy in order to create economic security for themselves. In the polis,
mutual aid is a good that people create collectively in order to protect each other and their
community. 2) Its mem-bers are motivated by both altruism and self-interest. Altruism: acting in
order to benefit others rather than oneself. In the rationality paradigm, altruism is almost invisible.
There is a paradox of altruism: when people act to benefit others, they feel satisfaction. The strict
self-interest paradigm, therefore, makes altruism impossible by definition. 3) It has a public interest,
whose meaning people fight about and act upon. Citizens have two sides: a private, rather self-
interested side and a more public-spirited side. We might think of public interest as what the public-
spirited side desires. Public interest could also mean those goals on which there’s a consensus. It’s
what most people want at the moment, and so it changes over time. Finally, public interest could
mean things that are good for a community. All communities have a general interest in having some
governing processes and some means for resol-ving disputes without violence. The members of a
community almost always have an interest in its survival. In market theory, public interest is the net
1
,result of all individuals pursuing their self-interest. In polis, even if we only want to understand how
people pursue their self-interest, we need to under-stand how conceptions of the public interest
shape and constrain people’s strategies for pursuing their own interests. 4) Most of its policy
problems are common problems: situations where self-interest and public interest work against each
other (example: dumping waste in a river). In market theory, com-mons problems are thought to be
the exception, most actions don’t have social consequences. In the polis these problems are
frequent, because actions have side effects and long term or ripple effects. 5) Influence is pervasive,
and the boundary between influence and coercion is always contested. Our ideas about what we
want and the choices we make are shaped by education, persuasion, and socia-lization. Often our
choices are conditional. 6) Cooperation is as important as competition. a) Politics involves seeking
allies and cooperating with them in order to compete with opponents. When there are only two
players, there’s no possibility for strategic coalitions and shifting alliances, or for joint effort,
leadership, and coordination. b) Cooperation is essential to power; it’s often a more effective form of
subordination than coercion. 7) Loyalty is the norm. There’s no ‘glue’ in buyer-seller relations. In the
polis there’s a presumption of loyalty: people expect that others will normally stick by their friends
and allies, and that it would take a major event to get them to switch their loyalties. 8) Groups and
organizations form the building blocks. a) People belong to institutions and organizations, even when
they aren’t formal members. Their opinions and interests are shaped by, affected by and depend on
organizations. b) Policy making is also about how groups are formed, split, and re-formed to achieve
public purposes. Groups coalesce and divide over policy proposals. c) Decisions of the polis are
collec-tive. Even when officials have ‘sole authority’, they’re influenced by outside opinion and
pressure. 9) In the ideal market, information is ‘perfect’, meaning it’s accurate, complete, and
available to everyone at no cost. In the polis, information is ambiguous, incomplete, often
strategically shaded, and some-times deliberately withheld. Interpretations are more powerful than
facts. What we believe about information depends on who tells us (the source) and how it’s
presented (the medium, the language, the context, the timing). Political actors strive to control
interpretations. Information is never fully and equally available. Political actors very often
deliberately keep information secret. 10) It’s governed by the laws of passion as well as the laws of
matter. Laws of matter: resources are finite, scarce, and used up when they are used. People can only
do one thing at a time, and material can only be one thing at a time. Laws of passion: a) Like passion,
political resources are often enlarged or enhanced through use, rather than diminished. b) The whole
is greater than the sum of its parts (example: a demonstra-tion). c) Things can mean and therefore be
more than one thing at once (» ambiguity, symbolism). 11) Power: derives from all other elements
and can’t be defined without reference to them. It is a pheno-menon of communities (1). Its purpose
is always to subordinate individual self-interest (2) to other interests – sometimes to individual or
group interest (8), sometimes to public interest (3). It operates through influence (5), cooperation
(6), and loyalty (7), and through strategic control of information (9). Finally, power is a resource that
obeys the laws of passion (10).
Change: in the market model, change is driven by exchange. In the polis, change occurs through the
interaction of mutually defining ideas and alliances.
Chapter 2 – Equity
Distributions are at the heart of policy controversies. In a distributive conflict, all sides seek equality;
the conflict comes over how the sides envision a fair distribution. There’s a paradox in distributive
pro-blems: equality often means inequality; the same distribution may look equal or unequal,
depending on where you focus. Equality: sameness, it signifies a part of a distribution that contains
uniformity. Equity: distributions regarded as fair, even though they contain both equalities and
2
,inequalities. Every distribution has three important dimensions: a) The recipients (who gets
something; number 1, 2, 3 and 4 below). b) The item (what is being distributed; number 5 and 6
below). c) The process (how is the distribution carried out (number 7, 8 and 9 below). Process is a key
dimension of equity because in the polis, distributions are policy decisions carried out by real people,
not by invisible hands. The processes of distribution can create or destroy things of value such as
loyalty, community spirit, or jobs, apart from the things they explicitly distribute. Process is
important too, because our notion of fairness includes not only a fair end result but also a fair
decision-making process.
1) Dilemma: Equal slices but unequal invitations.
Example: Students who came to class are getting cake, the ones who didn’t show up don’t,
even though they would’ve shown up if they had known that there would be cake.
Issue: This challenge questions the definition and boundaries of membership in a
communi- ty: who is a member of the group of recipients?
2) Dilemma: Unequal slices of unequal merit but equal slices for equal merit.
Example: Students have to do a test, those who score the highest get the most cake,
students who score lower get less cake.
Issue: Represents the ideal reward for individual accomplishment (merit-based
distribution). All modern liberal societies prize individual achievement as the standard of reward,
and aspire to minimize the role of race, gender, ethnicity, and other immutable personal
charac- teristics in determining citizen’s fortunes.
Critique: a) Some question how individual merit is measured, how we identify achievement
and aptitude (example: a test could be irrelevant to really test someone’s abilities). b)
Many social scientists question how much credit an individual ought to get for their accomplish-
ment. Nevertheless, the idea of reward for individual achievement provides the major
justi- fication for income inequality. c) Government policies exert much more influence on how
income gets distributed than individual effort and skill.
3) Dilemma: Unequal slices for unequal ranks but equal slices for equal ranks.
Example: Divide the cake according to the university department’s hierarchy.
Issue: Distributions is based on ranks (internal subdivisions of society). In economics, the
concept of equity based on internal ranks is called horizontal and vertical, with horizontal
equity meaning equal treatment of people in the same rank and vertical equity meaning
unequal treatment of people in different ranks. Rank-based distribution is widespread.
Hierarchical organization justify their rank-based differences in power, money and prestige
by merit.
Critique: One can ask whether the different ranks indeed represent different skills,
knowled- ge, or other relevant factors.
4) Dilemma: Unequal slices but equal social blocks.
Example: Gender roles and gender divisions in society.
Issue: Group-based distribution can lead to major internal cleavages of society. It holds
that some major divisions in society are relevant to distributive equity, and that membership in
a group based on these divisions should sometimes outweigh individual characteristics in
determining distribution (example: military veterans). In societies with liberal individualist
ideologies, group-based distribution is usually proposed as a remedy for previous violations
of merit- or rank-based distribution. The analogy in contemporary politics is affirmative
action: a policy of giving preference to members of groups that have been the victims of
historical discrimination. This is not the same as quotas! Quotas are a means of reserving a
certain portion of an item for members of a particular group.
3
, Critique: a) It can be questioned whether the definition of relevant groups reflect some
mea- ningful social reality. b) Ascriptive identify characteristics such as race, gender and natio-
nality, don’t really correspond to the actual experience of disadvantage or discrimination,
yet the primary rationale for group-based distribution is to compensate people for past
disadvantage. In this view, demographic groupings are too ‘rough’. They make
unwarranted presumptions about individual cases and give preferences to people who never
suffered any disadvantage. Group-based distribution should be used only as a tool to correct
deviations from merit-based selections. c) Race and gender are always illegitimate criteria for
distribu- tion, even when used in compensatory fashion. The main argument for affirmative
action starts from the main feature of the polis: community. This view sees diversity as a
character- istic of cohesive, strong, and vibrant communities.
- Rank-based distributions assign people to groups according to more or less fine-tuned
indivi- dual measurements. The justification for assignment to a rank usually has something to do
with the individual’s past performance or achievement.
- Group-based distributions assign people to groups based on immutable traits (things a per-
son can’t change) that have nothing to do with individual qualifications/performance (race).
5) Dilemma: Unequal slices but equal meals.
Example: Students who had lunch get less cake, students who haven’t get more cake.
XXXXX
Issue: To make something part of a larger entity is to expand the boundaries of what’s
being distributed. Expanding the definitional boundaries of the item is always redistributive,
because it calls for using the more narrowly defined item to compensate for inequalities in
a larger sphere. Challenges to the definition of an item are generally not either/or choices,
but choices about how expansively to define the item along a continuum.
6) Dilemma: Unequal slices but equal value to recipients.
Example: Some students don’t want chocolate cake, they ask for their portions to be
reallo- cated to those who do want cake.
Issue: The item is valued to the individual » switch from standardized value (e.g. the weight
of a cake slice) to customized value (e.g. how much nutrition someone derives from the
cake). Conflicts over the value dimension of equality are especially intense in social policy.
Education and medical care are ‘delivered’ through relationships and derive much of their
value from the relationships’ quality and the ability of the provider to tailer the serve to an
individual’s needs.
7) Dilemma: Unequal slices but fair competition with equal starting resources.
Example: Give everyone a fork and let them have a go at the cake.
Issue: Competition (opportunity as starting resources).
8) Dilemma: Unequal slices but equal statistical chances of winning.
Example: Put everyone’s name in a hat, draw a ticket, and give the whole cake to the
winner. Issue: Lottery (opportunity as a statistical chance).
9) Dilemma: Unequal slices but equal votes.
Example: Hold an election and let students vote on who gets to eat cake.
Issue: Voting (opportunity as political participation).
The point is there’s no right way. Decisions are hard; people sometimes resort to the two solutions
that free them from making decisions: equal slices and lotteries. Equal slices is intuitively powerful,
because the outcome ‘looks and feels’ equal. Lotteries have the political virtue of symbolizing
absolute fairness and they dampen citizen’s anger, losers are more likely to blame fate than
politicians. A distri-butive problem is analysed by figuring out the answers to three questions: 1) Who
4
