Policy Analysis: Frameworks and Models

1
Policy Analysis Frameworks and Models

• Stokey and Zeckhauser Ch1-3
• Jenkins-Smith Ch1-3
• Weimer and Vining Ch1-2

2
Where we left off with Weimer Vining

• Executive Summary
• Introduction
• Statement of the Problem
• Current State
• Statement of the Policy Goals
• New Ways of Organizing
• Comparing the Alternatives with the Policy Goals
• Summary Table
• Evaluation and Recommendation

3
Stokey and Zeckhauser Framework for Analysis

• What do you do when a complicated policy issue
  lands on your desk?
• Establish the context
• Lay out the alternatives
• Predicting the consequences of each alternative
  including likelihood
• Valuing the outcomes
• Making a choice
• These are craftsmans tools you must learn to
  wield them with skill.

4
Models

• Model a simplified representation of some aspect
  of the real world
• Types
• Diagrammatic flow charts, decision trees they
  identify distinct stages in a process
• Conceptual models taking complex ideas and
  concepts and simplifying them
• Long division large numbers are sometimes
  confusing
• Tragedy of the Commons (Garrett Harding)
  describes the incentives associated with the
  common grazing ground of a medieval English
  village. Where individuals ignore the cost their
  use has on others with the inevitable result of
  overgrazing that is costly to all.

5
More on Types

• Simple formal mathematical models that describe
  explicitly the quantitative changes in a
  particular variable in response to stimuli.
• Example in a savings account/interest
• Descriptive describing the way the world
  operates
• Prescriptive provide rules for making an optimal
  choice prescribing courses of action
  (normative/optimizing models)
• First, construct a descriptive model that
  encompasses all choices open to the decision
  maker and predicts the outcome of each
• A set of procedures for choosing among
  alternative actions given the decision makers
  preferences among the outcomes.

6
A couple more

• Deterministic the outcome is assumed to be
  certain. Natural Laws Emc2
• Probabilistic Where the outcome of a particular
  action is not unique, instead there is a range or
  a number of possible outcomes.

7
Judging a model

• How do we judge a model?
• We judge a model on how well it works or how
  accurately it predicts.
• Taking great care with our assumptions and model
  specification (Are we including the right
  explanatory variables?)
• A useful model is streamlined and relatively
  simple (How do we decide on which variables to
  omit?)

8
Why use a model?

• The discipline helps us get our thinking
  straight it forces us to think about fundamental
  principles
• The possibility of experimenting with the model
  rather than the system itself
• Facilitates communication among those concerned
  why did you make those assumptions? Why did you
  leave out those important variables?

9
The Model of Choice

• Allocating scarce resources with a focus on the
  public sector

There are two primary elements of any act of
choice

1. The alternatives available to the decision maker
2. The decision makers preferences among the
   alternatives

• Trade-offs are the essence of difficult decisions
• This model assumes little significant uncertainty

10
Alternatives available to the decision maker

• Two definitions
• Efficiency a combination of attributes is said
  to be efficient if, given the available
  alternatives, it is impossible to increase one
  output without giving up some of at least one
  other.
• Domination When comparing two alternatives (A
  and B), A dominates B when A is better in every
  respect. Dominated points can never be efficient.

p.24 of SZ
11
More

• The possibility frontier the set of efficient
  alternatives. The frontier may be straight or
  curved, continuous or discrete, or a few isolated
  points.

This curve tells us the maximum achievable output
of water for every possible output of
electricity. Point F indicates that with an
electrical output of 8 thousand kwh, the maximum
water output is 32 million gallons, and
conversely.
Electricity (thousands of kwh/day)
20
E
i
10
F
i
W
i
0
30
10
20
40
Water (millions of gallons/day)
12
The decision makers preferences

• The 2nd element of the fundamental choice model
  describes the decision makers preferences
  Indifference Curves

Electricity (thousands of kwh/day)
i
Each of these points are found to be equally
satisfactory and all of those along the curve
they lie on the same indifference curve because
they are equally satisfactory.
S
20
i
R
10
P
i
I1
i
Q
0
30
10
20
40
Water (millions of gallons/day)
13
A family of Indifference Curves

• Similarly, we could draw other indifference
  curves depicting lower and higher levels of
  satisfaction.

Remember, we dont assign specific values to
these levels of satisfaction, we say I1 is better
than I0. Also, there is no implication that
movements of equal distance across the graph are
equally valuable. Things get better as we move
north and east. Such a family of curves is
called an indifference map.
Electricity (thousands of kwh/day)
20
10
I3
I2
I1
I0
0
30
10
20
40
Water (millions of gallons/day)
14
The Best Choice

• Here we combine the continuous possibility
  frontier and the indifference map.

The best choice for the planner is the
combination of electricity and water represented
by point __. This is because only at that point
can the planner reach the highest possible
indifference curve.
Electricity (thousands of kwh/day)
20
E
i
G
i
T
i
10
F
i
W
i
0
30
10
20
40
Water (millions of gallons/day)
15
MRT
At any point on the curve, the rate at which one
output can be transformed into another is given
by the slope at that point on the possibility
frontier. We refer to the rate at which one
output can be transformed into the other at a
particular point as the marginal rate of
transformation (MRT).
Electricity (thousands of kwh/day)
20
E
i
G
i
10
F
i
W
i
0
30
10
20
40
Water (millions of gallons/day)
16
MRS
The slope of an indifference curve may be
interpreted in a similar way. It represents the
way in which the decision maker is willing to
trade electricity for water while still remaining
on the same indifference curve. The steepness of
the indifference curve indicates the rate at
which the planner is willing to trade off between
the two outputs. This trade off rate is called
the marginal rate of substitution (MRS).
Electricity (thousands of kwh/day)
i
M
20
10
N
i
I1
0
30
10
20
40
Water (millions of gallons/day)