alternatives public policy

1 PPA 670 Public Policy Analysis Basic Policy Terms and Concepts Essential Definitions POLICY Lasswell & Kaplan: Policy is a projected program of goals, values and practices.

Thomas Dye: Whatever government chooses to do or not do.

Charles Jones: Functional analytic category -- a course of action rather than specific decisions. Essential Definitions PUBLIC POLICY Definition: "Purposive action by actors acting in public institutions to produce direction in government.

Key terms:

Purposive action Acting in public institutions Direction in government 2 Approaches to Policy Analysis The Eight-fold path:

1. Define the problem 2. Assemble some evidence 3. Construct the alternatives 4. Select the criteria 5. Project the outcomes 6. Confront the trade-offs 7. Decide 8. Tell your story The Process Model I.D. Recognize Structure Agendize Prob. State.

Alt. Det. Alt. Test Decision Implement Monitor Evaluate Verification Feedback Iteration Public Policy Environment Key Institutions:

Chief Executive Bureaucracy Legislature Courts Interest Groups & Lobbyists Media 3 Public Policy Environment Key Individuals:

President, governors, city managers, mayors Senior policy-makers (eg.

Department heads) Key legislators (eg. Speaker) Key lobbyists (eg. Nader) Media stars Public Policy Environment Key feature: Policy is a product of public institutions, must be legitimized.

Policy without legitimization is just rhetoric. Ethical Issues:

Roles of ideology and objectivity of key individuals. Recognizing and serving the “Public Good” 4 Why PUBLIC Policy?

Political reasons Moral or Ethical reasons Economic and Market Failures PROBLEM RECOGNITION Key to the start of the analytical process. If the wrong problem is identified, the quality of analysis is moot.

Public policy problems arise as a result of changeor pressure for change. I.D. Recognize Structure Agendize Prob. State. Alt. Det. Alt. Test Decision Implement Monitor Evaluate Problem Recognition Phase 5 Recognizing Public Policy Problems Public vs. private problems Policy vs. management problems Solvable vs. Unsolvable problems If we fail to properly recognize a public policy issue, we cannot hope to solve it:

G.I.G.O.

Problem Recognition Criteria Asking proper questions is critical to establishing the correct criteria:

Four key questions:Where did the problem come from?How do you know about the problem?What are the dimensions of the problem?Who is involved and why? Where did the problem come from?

What is the history?

Have we seen this problem before?

What do we know now?

What do we need to know to solve the problem? 6 How do you know about the problem?

How did this problem come to public awareness?

What are your sources of information?

FactsOpinionsPrimary and secondary dataWho do you trust? Why? How much information is enough?

What are the dimensions of the problem?

Purpose: to set a context and limit the inquiry.

The six critical dimensions of a policy problem:

TypeScaleLocationIntensityExtensivenessTime-line Dimension: Problem Type Political Social Economic Technical Virtually all public issues combine multiple problem types.

Defining the type of problem leads to types of appropriate policy solutions. 7 Dimension: Problem Scale Macro (societal) Micro (organizational/individual) Dimension: Problem Location Bounded by:

Physical location Cultural/Social characteristics Political venue Domestic or International (Note:

international policy problems are not bounded by U.S. laws or culture – separate field entirely.) Dimension: Intensity How important is it to solve this problem now?

How strongly to people feel about the problem?

How is the problem being portrayed in the media?

How is the problem being portrayed politically? 8 Dimension: Extensiveness How many stakeholders are involved?

How large is the target population?

How extensive are the geographic boundaries?

Dimension: Time-line Two basic time-lines for problems:

1. Longitudinal – for these problems, the passage of time causes changes in the solution to the problem.

Problems are time-dependent. 2. Cross-sectional – for these problems, the passage of time does not cause the solution to change.

Problems are time-independent. Who is involved in the problem?

Stakeholder identification is critical to successfully defining a problem.

Elements to stakeholder ID:

1. Who has a stake in the policy solution? – both institutions/ groups and individuals. 2. What are the positions on the issue for each stakeholder? 3. How much influence does the stakeholder have on the policy solution? 4. What will the policy consequences be if the stakeholder preferred solution is selected? 9 Implying Causality in Policy Must be careful of implying causation when examining change and interaction among variables.

Correlation -- the variables show a consistent relationship Causation -- must show both necessaryand sufficient reasoning (extremely difficult in policy) Often look for proximate (doable) rather than ultimatecauses for policy solutions PROBLEM STRUCTURING Problems exist on a structural continuum:Low structure (seek political solutions)Moderate structure (seek mixed solutions)High structure (seek economic/technical solutions) Understanding the structure of a problem aids defining a solution set Problem Structuring Approaches For low/moderate structure problems:

“Back-of-the-Envelope” calculations.

Provides:

Estimation of boundaries of the problemIndication of direction of movement of problemRough idea of the magnitude of the problem Use known and available indicators to develop a rough estimate of the dimensions of the problem. This is speculative. 10 Problem Structuring Approaches For moderate/high structure problems:

Systematic Analysis:

Relies on forecasting. (Projecting existing situations to a point in time when policy action will be taken.) Long-term, very expensive.

Forecasting Techniques:

Modeling -- simplified version of reality. Trend extrapolation -- time series analysis Monitoring -- continual review of critical info. Tools to Aid Structuring “Borrowing” problem definitions:Avoiding re-invention of the wheelBest-practices researchBest-of-Breed recognition Reasoning by analogy: SynecticsResearching similar problemsLearning from the mistakes of othersDistinguishing the relevant from the irrelevant. Others (see readings)Classification analysisHierarchy analysisAssumptional analysis I.D. Recognize Structure Agendize Prob. State. Alt. Det. Alt. Test Decision Implement Monitor Evaluate Policy Planning Phase 11 Agenda Setting:

Limited resources require choice of issues to continue analysis.

Basis (Cobb & Elder):

Distribution of access and influence has inherent biases Range of issues which can be considered is limited The system's inertia make it difficult to change the prevailing biases Agenda Building:

Two agendas Systemic -- discussion only Institutional -- set of items up for active and seriousconsideration Only items on the Institutional agenda are acted on by policy institutions. Division of the Institutional Agenda:

Problem definition agenda. "Active and serious" research Proposal Agenda. Shift to finding a solution Bargaining Agenda. Support for proposals is developed Continuing agenda. Items keep resurfacing 12 Agenda-setting options:

Let it happen. Pluralistic, gov't is passive Encourage it to happen. Gov't assists in definition and articulation of views Make it happen. Gov't takes an active role Non-decision. Agenda-setting forces:

Interest groups Media Politicians Events Notables Violence Formal Problem Statement Once a problem is structured and on the institutional agenda.

Statement of problem to be solved is “locked” for final analysis.

Must be defined prior to alternative solution search. 13 DEVELOPING ALTERNATIVES What is an alternative?

A potentialsolution to the policy problem under consideration. To be effective, a selected policy must be the product of a comprehensive alternative search.

Factors Influencing Alternative Searches:

Theory of political action of potential decision makers. (Eg. “Law of large solutions.) Systemic inertia -- do nothing.

Limits on resources: Time, money, expertise, etc.

Anchoringand parochialismwhich tend to foreclose options before they are explored.

Alternative Sources Literature Reviews Existing policy proposals Best-Practices/Best-of-Breed Generic solutions ExperiencesPilot studiesBrainstorming Custom-made solutions –tailored to a unique situation (most expensive and difficult) 14 Information to be provided on each alternative: Narrative Description Costs (monetary and not) Spillovers (externalities) Estimated effectiveness Other considerations:

Morality Risk Political feasibility Validating alternatives:

Does the alternative meet basic criteria and objectives?

Has new information impacted on previously developed alternatives?

Does it have an appropriate time horizon?Does it foster both internal and external policy consistency?

Is it workable& are resources available?Haveriskand uncertaintybeen accounted for? I.D. Recognize Structure Agendize Prob. State. Alt. Det. Alt. Test Decision Implement Monitor Evaluate Policy Analysis Phase 15 Selecting Criteria for Evaluating Policy Alternatives What is the value of the policy solution? Establish value prior to alternative test.

Informed by:

Definition of the problem (dimensions/boundaries) How the problem was structured Alternative test methods should be chosen to validate the selected criteria.

Alternative Evaluation Criteria Policy effectiveness (quality of outcome) Policy efficiency (cost/reward) Equity (fairness for target group) Liberty/Freedom (enhancement of rights) Political Feasibility (politically acceptable solution) Social acceptability (public support) Administrative feasibility (ease of implementation) Technical feasibility (technology is viable and appropriate) Testing Alternatives Purpose: To critically evaluate each proposed alternative.

Values:

Assessment of strengths/weaknesses of each alternative.

Provides a method of directly comparing alternatives Provides information to decision makers on final selection. 16 Alternative Test All alternatives considered must be tested using the sametest methods.

Tests may be qualitative(more political) or quantitative(more economic) in nature.

Multiple methods should always be used.

Policy Models Purpose: To analyze alternative proposed solutions.

Why? Most alternative are too complex for complete analysis, so we use models.

Definition:

Simplified version of reality.

Levels of models:

Descriptive Conceptual (more qualitative) Predictive (more quantitative) Policy analysis tends to focus on conceptual and predictive models.

Selecting the correct models of analysis is critical to successful policy selection. 17 Sample Policy Models Models used for more qualitative analysis:InstitutionalGroupPublic Choice (aka Rational Choice) Models used for more quantitative analysis:Game TheoryDecision TheoryEconomic Rationalism TheoryCost-benefit AnalysisCost-effectiveness Analysis The Institutional Model:

Policy as institutional output.

Key focus is on the structural elements that produce policy. How the relationship between institutional arrangements and the environment affect the content of policy. Institutional Model Environment Actors Institutions Public Policy 18 Institutional Model Policy Examples Competing agencies – auto safety & EPA Distribution of resources among levels of government – Education support The Institutional model tends to be appropriate for more structuredissues The Group Model:

Policy as group equilibrium.

Basic assumption: Interaction among groups is the basis for politics and therefore policy.

How well does the policy alternative manage group conflict? How:

Establish rules of the game.

Arrange compromises and balance interests.

Enact compromise as policy.

Enforce the compromise.

How is equilibrium maintained:"Latent group" basically supports the status quo.Overlapping group memberships prevent too extreme positions.

Checks and balances from group competition. 19 Group Theory Model Policy Examples Ideology – Abortion, Religion in schools Competing Interests – Environmental protection vs economic development Public Choice Model:

Public choice theorists focus on the question of what government policies are likelyto be implemented in a given political setting, rather than what policies wouldproduce a desirable outcome if they wereimplemented.

Key Aspects:

Individual “self-interest” drives policy.“Government Failure” akin to market failure produces bad policy.

Local government is best choice to act.

Public Choice Model Examples Alaska “Bridge to Nowhere” – Sen. Ted Stevens.

Funded political junkets – Sen. Tom Delay Any “pork-barrel” project (eg. military bases, highways, parks) keyed to constituent self-interest, not public good. 20 Game Theory Models:

Rational choice under competitive circumstances.

“Zero-sum”games Pay-off matrices Risk preference analysis: Maxi-max, Maxi-min, Mini-max Classic Game Theory:

Prisoner’s Dilemma Two prisoners (A and B) taken into custody Charged with same offense Unknown to each other, held separately, cannot communicate Both A and B are offered the same “deal” The Deal A. If you confess the crime was commited by both of you, and the other prisoner denies it or remains silent, you go free and the other goes to prison for 5 years. B. If both deny the crime, there is enough circumstantial evidence to jail for 3 years each. C. If you both confess, you each get one year. 21 Prisoner’s Dilemma Pay-off Matrix Pay-off Matrix Action Pay-off ABAB Cooperate Cooperate -3( R ) -3( R ) Cooperate Defect -5 (S) 0 (T) Defect Cooperate 0(T) -5(S) Defect Defect -1(P) -1(P) The “Payoff Matrix” Involves:

Strategies Alternative conditions.

Probabilities Calculated payoffs from all probabilities Payoff Matrix Example Should I carry an umbrella today? Rain No Rain Carry Umbrella 10 0 No Umbrella -30 5 22 Payoff Matrix Example Add in probabilities:

30% chance of rain Rain (30%) No Rain (70%) Total Value Carry Umbrella 10(.3)=3 0(.7)=0 3 No Umbrella -30(.3)=-9 5(.7)=3.5 -5.5 Payoff Matrix Example Add in probabilities:

5% chance of rain Rain (5%) No Rain (95%) Total Value Carry Umbrella 10(.05)=.5 0(.95)=0 .5 No Umbrella -30(.05)= -1.5 5(.95)=4.75 3.25 Accounting for Risk Preference Used when multiple players have different risk- preferences.

Requires subjective probability estimates.

Perceived pay-off values.

Approaches:

Maximax: maximizes payoff (optimist)Maximin: Avoid lowest pay off (pessimist)Minimax: Minimizer of regret 23 Airport Case Basic pay-off matrix S1-Sell N1 –Decrease services S2 – Status Quo N2-Status Quo S3-Increase funding N3-Increase services N1 N2 N3 S1 20 30 40 S2 0 50 60 S3 -50 10 100 Aiport Case -- Maximax Select maximum payoffS1=40 S2=60 S3=100 Coefficient of Optimism: p=.6 Maximu m p=.6Minimum p=.4To t a l S1 40 20 40(.6)+20(.4)=32 S2 60 0 60(.6)+0(.4)=36 S3 100 -50 100(.6)- 50(.4)=40 Airport Case -- Maximin Minimum Payoff S120 S2 0 S3 -50 24 Airport Case -- Minimax Regret Table: N1 N2 N3 S1 0 2060 S2 20 040 S37040 0 Decision Theory Decision Tree models Used to determine the utility or value of possible outcomes Procedure:

Use the value (estimated) of each outcome.

Use the probability (subjective) of each outcome. Decision Trees Steps:

Determine all components:Initial stateDecision pointsOutcome pathsValueProbability Diagram the system of decision points and outcome paths (tree structure) 25 Decision Trees Calculate conditional probabilities for each outcome path.

Sum up conditional probabilities for each type of result.

Determine which result has the greatest utility. Decision Tree Examples D-Tree Contracting Emergency Services A resort town wants to know how much it should spend on a contract to provide emergency services for accidents on weekends. They determine the initial conditions to be:

 Each weekend lasts three working days.  The estimated probability of an accident occurring on any weekend day is 20%  Since the service is contracted on a day-basis. It does not matter whether there is one accident or 10 on that day. The service provision cost remains fixed:  Cost of a one accident-day weekend = $10,000  Cost of a two accident-day weekend = $25,000  Cost of a three accident-day weekend = $50,000 26 D-Tree Structure D-tree Analysis Accident possibilities:

0 accident-days: (NNN) = .512 1 accident-days: (NNA) .128 + (NAN) .128 + (ANN) .128 =.384 2 accident-days: (NAA) .032 + (ANA) .032 + (AAN) .032 = .096 3 accident-days: (AAA) =.008 Value (utility) of a weekend is the average cost for service:

.512(0) + .384(10,000) + .096(25,000) + .008(50,000) 0 + 3,840 + 2,400 + 400 = $6,640 This means that the town should only accept a bid at $6,640 per week or less for services. Economic Models Basic philosophy: Maximizingsocial benefit.

Derived fromPareto Optimality: No one canbe made better off without making someone else worse off.

Kaldor-Hicks Criteria: A Pareto situationcan be improved to the extent that the losers can be sufficiently compensated for their losses.

Major difficulty: Each individual has their own priority ordering. 27 COST-BENEFIT ANALYSIS (CBA) CBA is a life cycle analysis of both the costsand benefits associated with a project.

CBA is usually associated withcapital projects Basic information needed:

!Measurement of all costs and benefits.

!Determination of theNet Present ValueofallC&B.

!Assessment of the relationship between the costs & benefits. Cost-Benefit Process 1. Specify objective.

2. Identify alternatives 3. Collect & analyze info.

4. Specification of target groups 5. ID. All C&B 6. Discount all C&B 7. Estimate risk & uncertainty 8. Specify criteria for selection Cost-Benefit Process Steps 1 - 4 are the same as in the basic policy analysis process.

Identification of all costs and benefits is a very complex undertaking.

Use of classification schemes can assist in the identification process. 28 Identifying Costs/Benefits Use three questions:

1. Is the C/B Internal (inside) or External (outside) to a target group?

2. Is the C/B directly measurable (tangible)or indirectly measurable (intangible)result of the program?

3. Do the combined C & B create a real (net efficiency) increase in utility or a redistributionalshift?

C/B Tree Structure: Ty p e o f C/BInside OutsideDirectly Measurable Directly Measurable Indirectly Measurable Indirectly MeasurablePrimary Secondary Primary Secondary Primary Secondary Primary SecondaryNet Efficiency Redistributional Net Efficiency Redistributional Net Efficiency Redistributional Net Efficiency Redistributional Net Efficiency Redistributional Net Efficiency Redistributional Net Efficiency Redistributional Net Efficiency Redistributional Variety of costs must be considered:

1. Capital costs 2. Maintenance & operating costs 3. Opportunity costs 4. Social costs–externalities (eg. Pollution) Major problem: many public-sector C&B are not directly measurable. Must use shadow pricing 29 Shadow Pricing for CBA Uses a surrogatemeasure because the Cost or Benefit cannot be measured directly.

Examples:

Cost of a human life (actuarials) Cost of pollution (clean-up) Benefit of quality of life ( property value) CLEAR Program Example Tabel IV.2. Monetized Benefits (Costs per Victimization) Cate gory Tan gible Intan gible Total Cost Assault $1,753 $8,822 $10,575 Arson $22,055 $20,358 $42,413 Ra pe $5,768 $92,063 $97,832 Robber y $2,601 $6,447 $9,048 Car jackin g $5,485 $6,447 $11,932 Murder $1,164,930 $2,160,210 $3,325,140 Kidna pping $7,804 $28,275 $36,079 Making Costs & Benefits Comparable: Discounting Allows comparison of C&B incurred over a period of time. Uses theNET PRESENT VALUEto allow direct comparisons.

Since the value of money declines with time, we must find a way of equalizing it:The Discount Factor 30 The Discount Factor Formula:

i = discount rate (inflation) t = Time period in years from initiation ti DF ) 1 (1   ti DF ) 1 (1   Discount Factor Example Value of $10,000 in 5 years at a discount rate of 5%:

.7835 x 10,000 = $7,835 7835 .

) 2762 . 1 (1 ) 05 . 1 (1 5     DF 7835 .

) 2762 . 1 (1 ) 05 . 1 (1 5     DF DISCOUNT TABLE To simplify matters, use a spreadsheet or Present Value table.

Tables represent discount factors for a variety of discount rates Take care to use 4 decimal-digits in calculations 31 Using the NPV Excel Function Excel contains built-in financial formulas Formula for Net Present Value:

=npv(rate, value) Be sure costs are represented by negative values: (50000) or -50000 Choosing a Discount Rate D.R. should accurately reflect the declining value of money over time -- usually keyed to expected inflation.

D.R. should also account for the risk preferenceof the decision maker.

Risk Philosophies 1.High Risk philosophy (projects are inherently very risky). Marginal productivity of capital in private investment. Results in conservative (low) investment rate. (12% -- 20%) 2. Moderate philosophy. Social Rate of Time Preference. Compensation necessary to induce consumers to refrain from spending. (1.5% -- 4%) 32 Risk Philosophies 3.LowRisk Philosophy (we do government projects forthe public good, not pay-back).Government borrowing rate without regard for time preference.

Rate at which federal government is willing to borrow money. Usually keyed to 6 mos. T-Bills. (0.5 –2.0%) 4.Internal Rate of Return. Rate at which C&B (Net Present Value) are equal. –This is the best way to determine the value of a project without philosophical compromises . Federal Government Discount Rates Real Discount Rates (in percent)3-Year 0.1 % 5-Year 0.4 % 7-Year 0.7 % 10-Year 0.9% 20-Year 1.2 % 30-Year 1.4 % Source:

http://www.whitehouse.gov/omb/circulars/a094/a94_appx-c.html Revised: Dec. 2014.

Strengths of Cost-Benefit Both costs and benefits are measured in dollars as a common unit of value.

Allows us to go beyond the confines of a singlepolicyorprogramandlinkC&Bto society as a whole.

Allows direct comparison of programs in widely differing areas. 33 Limitations of Cost-Benefit Exclusive emphasis on economic efficiency excludes equitycriteria.

Monetary value is an inadequate measure of responsiveness (e.g. $1,000 to a rich person and a poor person) The frequent use of shadow prices may be arbitrary and unjustifiable.

The bottom line is subject to intentional manipulation to produce a pre-determined outcome. COST-EFFECTIVENESS Bases:

Maximize value received for a given expenditure.

Minimize expenditure for a given value. This means not all benefits must be defined or measured -- only key value.

Cost-Effectiveness Procedure Steps 1 -- 4 are the same as CBA Step 5 -- Choose a comparison value criterion (what is the value to be received, lives saved, etc.) Step 6 -- Calculate the value for each alternative proposed (to NPV).

Step 7 -- Rank alternatives in terms of highest cost-effectiveness C.E. = Value/Costs Step 8 -- Recommendation 34 C.E. Example C.E. Example Recommended strategy: New Drug Expected cost: $1093 Expected effectiveness: 0.812 (an expected prob. of cure) Cost-Effectiveness Advantages Less work than CBA More applicable to "soft" programs -- can be used on non-capital projects Flexible Faster and requires far fewer resources than CBA 35 Cost-Effectiveness Disadvantages Selection of the value criterion is arbitrary Quantification problems Tends to undervalue externalities POLICY SELECTION (DECISION) Two options:

Non-selection Selection of alternative Circumstances that lead to non-selection:

!Acknowledge S.Q. and reaffirm it !The decision is so embedded in the estimation process all options are closed out !Make a non-decision !Decision maker(s) do not want a clear decision. 36 Factors involved in selection:

!Context -- How the nature of the issue affects the decision !Leverage -- political feasibility !Importance -- relative political significance of the issue !Information -- What information is available to the decision maker?

!Personality -- Whowill be involved in the decision Selection Realities There is no guarantee a policy analysis will be used in policy selection.

Rationality is not a requirement.

Politics will usually prevail. I.D. Recognize Structure Agendize Prob. State.

Alt. Det. Alt. Test Decision Implement Monitor Evaluate Policy Assessment 37 Implementation Definition: The translation of legislative or executive decisions into operational regulations, programs and actions, including the delivery of the policy consequences to the target population.

Variables involved in the implementation process: 1. Tractability of the problem (is the problem easy to deal with?) 2. Nonstatutory Variables (external factors present in society) 3. Ability of Statute to Structure implementation (effectiveness of implementation infrastructure) Stages in the process: 1. Policy outputs of implementing agencies leads to ...

2. Compliance with policy outputs by target groups leads to ...

3. Actual impacts of policy outputs leads to ...

4. Perceived impacts of policy outputs leads to ...

5. Major revision in statute. 38 Factors influencing Implementation:

Source of the policy Clarity of the policy (legislative intent) Support for the policy Complexity of administration of the policy Incentives for implementation Resource allocations For Successful Implementation: It must work in both the  Political and Programmatic arenas.

EVALUATION:

Purposes:

Assure correct alternative is implementedAssure the alternative does not haphazardly change during implementation Determine if desired impacts are occurringDetermine future course for policy --ContinuationModificationTermination 39 The Policy Evaluation Continuum:

Ex-ante policy analysis-- pre-program policy analysis procedures (policy cycle including implementation) Policy Maintenance-- Analysis of the policy or program as implemented to insure it was implemented as designed. Ex-ante Analysis M aintenance M onitoring Ex-post Evaluation The Policy Evaluation Continuum:

Policy Monitoring-- Recording of changes after the policy or program is implemented.

Ex-post policy evaluation-- Quantitative and qualitative analysis of whether the policy objectives were achieved.

Ex-ante Analysis M aintenance Monitoring Ex-post Evaluation Ex-Post Evaluation Traditional methods:

Hearings and Discussions Site visits Program measures (output) Comparison with professional standards Evaluation of citizen complaints All of these are problematic, impressionistic. 40 Contemporary Approaches: 1. Before vs. After program comparison.

2. Time trend projection of preprogram data vs. actual post-program data.

3. Comparisons with jurisdictions or population segments not served by the program.

4. Controlled experimentation (control group use).

5. Comparisons of planned vs. actual performance. Key Evaluation Components: Definition of the end (goals/objectives) to be achieved.

Specification of the policy, program or actions intended to achieve the end.

A method for observing and measuring the change or outcome.

A method for comparing the outcome with the desired end.

A way to modify the policy as a result of the evaluation.