Our academic experts are ready and waiting to assist with any writing project you may have. From simple essay plans, through to full dissertations, you can guarantee we have a service perfectly matched to your needs.
GET A 40% DISCOUNT ON YOU FIRST ORDER
Your first group paper should define the problem by providing background information necessary for your analysis. You should place the problem in context, identify stakeholders, and provide appropriate theories or data to allow the client to understand the scope and severity of the issue. For this project no extensive data gathering or analysis is necessary, but I recommend that you include some information to substantiate and support the points you make in your paper.
As I mentioned, policy analysis is not policy ADVOCACY. It is the process of assessing and deciding among feasible alternatives to resolve a problem, so a group has to define what the problem is in the first place. Each group has to pay extra attention to clarify a problem statement as possible as you can. Please identify a narrow, specific issue or problem that you guys want to analyze. Do not select a big or complicated issue. Also, a specific law will not be a good/feasible problem for policy analysis unless you intend to replace the law with another. As the syllabus noted, your group has to clarify
– “What is the problem? How important is it? Why do you say that? What are its dimensions? Use evidence as appropriate to outline the extent of the problem and the nature of that problem.”
*** You have to follow the common process of policy analysis described in the supplementary document about Problem Definition and Criteria-alternatives Matrix.
Introduction to Public Policy Analysis
Problem Definition and Criteria-alternatives Matrix
Table of Contents
1. DEFINING PUBLIC POLICY PROBLEMS AND GOALS …………………………… 1
2. RELEVANCY OF PROBLEMS DEFINITION/STRUCTURING FOR CRITERIAALTERNATIVES
MATRIX …………………………………………………………………………. 1
3. THE CRITERIA-ALTERNATIVES MATRIX ……………………………………………… 2
4. DECISION CRITERIA ………………………………………………………………………………… 3
5. POTENTIAL SOURCES OF CRITERIA ……………………………………………………… 5
6. POTENTIAL SOURCES OF POLICY ALTERNATIVES …………………………….. 6
7. RULES OF THUMB FOR FORMULATING POLICY ALTERNATIVES …….. 7
8. CRITERIA–ALTERNATIVES MATRIX PROCEDURE ………………………………. 9
9. A CRITERIA-ALTERNATIVES MATRIX EXAMPLE: TRANSPORTATION
ALTERNATIVES IN METROPOLITAN ATLANTA …………………………………. 12
DEFINING PUBLIC POLICY PROBLEMS AND GOALS
Please remember that how you define a policy problem and the goal you set should determine
the alternative policy solutions you consider and the alternative you recommend. When you
formulate policy problem and a goal:
The problem (and/or policy goals) should be stated clearly.
Problem statements are usually negative in nature (e.g., “The problem of traffic congestion…”).
Goal statements are about corrections of such negative situations (e.g., “The goal of this study
was to reduce traffic congestion…”).
Preferably both problem and goal statements should be included in a policy paper. This may be
redundant. If that is so, only a goal statement should be included.
Think about the best ways to express the problem and goal for your specific topic of study. Be
clear and specific in your goal statement.
The problem and goal definitions should not include (or suggest) particular policy alternatives.
They should be neutral in relation to the alternatives.
Problems and goals should be defined for:
– A particular jurisdiction: geographic area or specific group of people (State of Alabama,
County of Montgomery. City of Harrisburg; Children with disabilities in the County of
– A particular decision-maker who has the power to take an action on the problem within that
jurisdiction (President, Congress, Governor, State Legislature, Mayor, City Council, etc.).
The problem addressed in this paper is the failure of the thirteen- county in
metropolitan Atlanta area to meet the Environmental Protection Agency’s
(EPA) air quality standards. The goal then is for the Georgia Regional
Transportation Authority to improve the air quality in these counties to the level
to meet the EPAs standards by the year 2010.
RELEVANCY OF PROBLEMS DEFINITION/STRUCTURING FOR CRITERIAALTERNATIVES
The criteria and alternatives to be used in a criteria-alternatives matrix may be gathered from
various sources. One of these sources is problems definition. Ideally, a policy analyst should
define a policy problem, in collaboration with stakeholders, and then extract the criteria and
alternatives from this definition. In real life this does not happen very often; usually policy
analysts are instructed to use criteria and alternatives that are given to them by their bosses or
In either case, the policy alternatives considered should be potential solutions to the problem
as it is defined. They should help the decision-makers to reach the goal as it is defined.
THE CRITERIA-ALTERNATIVES MATRIX
The criteria-alternatives matrix is a useful tool many policy analysts, but not all, use. The matrix
is known by other names as well. In your future readings, you may encounter names like
decision grid, selection matrix, problem matrix, problem selection matrix, solution matrix,
criteria-based matrix, etc.
The following is a hypothetical example of the criteria-alternatives matrix. This matrix was
generated to answer the question: How could educational opportunities be improved for children
in Pennsylvania? In this example four alternative policy actions are compared on five decision
criteria. The alternatives are presented in the columns. They are “No action (status quo),”
“Increase property taxes,” “Increase sales taxes,” and “Use gambling money.” The criteria are
presented in the rows. They are “Increase access to education (the goal),” “Cost,” “Political
feasibility,” “Equity,” and “Legality.” Please note that with the exception of the goal, these are
generic criteria. Each of these generic criteria can be specified, as you will see in examples in
this lesson. I will discuss why these generic criteria should be at least considered to use in
criteria-alternatives matrices in the next section.
Table 4.1: A hypothetical example of the criteria-alternatives matrix
As you can see in the table, the four alternatives are rank-ordered on each criterion, from 1 to 4.
(These numbers can be considered a mathematical matrix; hence the name “criteria-alternatives
matrix.”) “Increase property taxes” is ranked the best alternative on the criterion “Increase access
to education (Goal).” The second-best alternative is “Increase sales taxes” on this criterion. You
can see how all the alternatives are ranked on all the decision criteria. You can also see the total
scores for the alternative at the bottom. The winner is “Use gambling money”; it is the
alternative with the lowest score. Therefore, the policy analyst should recommend this as the best
alternative to policymakers.
Increase access to
education (Goal) 4 1 2 3
Cost 1 4 2 3
Political feasibility 2 4 3 1
Equity 3 2 4 1
Legality Legal Legal Legal Legal
Total Score 10 11 11 8 (Winner!)
Decision criteria are the grounds for choosing among alternatives paths of action. They
represent decision makers’ values and priorities. They are closely related to how the problem is
defined in a policy analytical process. If you consider Figure 4.1 again, the goal in this matrix is
a statement of the desire to remedy and undesirable situation: to increase access to education.
Problems may be defined as undesirable situations (deficit or excess) or as opportunities to
improve current conditions (positive problems). Then the goal is defined as the statement of a
desirable, positive situation.
Obviously, what is a desirable or positive situation in public affairs depends on policymakers’
and/or stakeholders’ values and preferences. There is no universal or scientific way of
determining what should be desirable. Consequently, not one, but multiple goals may be used in
a criteria-alternatives matrix. An analyst should be careful in sorting out the values and
preferences of the stakeholders selecting the goals to be included in a criteria-alternatives matrix.
There is an important issue we need to consider: Why should analysts use multiple criteria to
select the best alternative? Why not only one criterion: the goal (e.g., increasing access to
education)? The first reason is that in a policy analytical situation there may be more than one
goal expressed by the policymakers and/or stakeholders and the analyst may be obligated to
consider these multiple goals. You will see an example of this in transportation planning later in
this lesson. The second reason is that in any policy analysis process, the analysts, policymakers,
and stakeholders have to consider other aspects of taking a policy action as well. These other
aspects include the cost of taking that action, as well as its political feasibility and ethical and
legal implications. These aspects are listed as criteria in Figure 4.1.
Cost should be a decision criterion for the obvious reason that every policy action costs some
amount to someone. Either the governmental entity taking the action has to allocate money to
reach the policy goals, and/or the action costs to the community or some segments of it. If the
municipal government of a city decides to install cameras at street corners to deter crime, this
will obviously cost the government. If the government does not take this action, it may cost the
home owners in the community because not only that they would be victims of crime, but they
may end up installing security systems in their homes, which would cost them some amount of
money. Cost must be a decision criterion in any criteria-alternatives matrix.
How it is defined and calculated are important considerations. It should be clear in a criteriaalternatives
matrix analysis for whom the costs are calculated (for a governmental entity,
community, segments of a community, etc.). Also, the calculation methods like total cost, or
cost-effectiveness, etc. should be stated clearly. The distribution of costs of a policy action in a
community and how they should be calculated may be highly controversial.
The details of these controversies cannot be discussed in this lesson, but I want to refer you to
the elaborate discussions by MacRae and Whittington (Expert Advice for Policy Choice,
Georgetown University Press, 1997, chapter 2). Briefly, the authors bring up issues like whose
costs should be considered and whether the commonly held assumptions about costs should be
taken for granted. They ask, for example, why should only the costs to humans considered? How
the cost of building a chemical plant to the animals and plants in the neighborhood? Should only
financial costs be included in the analyses, or should the intangible costs to the lifestyles,
aesthetics, or well-being of people be considered as well? We will discuss some of these issues
in more detail in Lesson 9 in which we will cover cost-benefit analysis.
The political feasibility, legality, and ethical implications of policy actions also should be
considered in any policy analysis, particularly in the applications of the criteria-alternatives
matrix. I must stress that you probably will not see these criteria listed in all applications of the
matrix. However, I urge you consider including them in your analyses. (You are required to use
them in the policy issue paper you will write at the end of this semester.)
You should consider political feasibility because without taking the preferences of political
actors for policy options into account, you cannot conduct an analysis that would be really
relevant. Even the best policy options may not be feasible to implement if there is no political
The importance of the legality of policy actions is obvious. If an action is not permissible by law
or the constitution of the country, it should not be taken at all. The government cannot confiscate
the properties of citizens without a legitimate reason, for example. In some cases, the legality of
a policy action may not be very clear; it may be questionable. Then it is advisable that the analyst
ask for the opinions of legal experts. The legality of a policy action may not be clear even after
the analyst receives the opinions of legal experts. You will remember from many examples that
the policy actions taken by governments may be challenged by interest groups at courts, after the
laws the actions were based on were passed. For example, the Patient Protection and Affordable
Care Act of 2010 (otherwise known as “Obamacare”) was challenged at the US Supreme Court
after the legislation was passed. The Court ruled that the law was constitutional in June 2012.
The legality of a policy option may not be clear when the analyst conducting her analysis, but it
is still important to consider it in the analyses. It is still better be prepared legally, than facing
unexpected legal challenges later. In Figure 4.1 all the options are considered legal. Therefore,
there is no need to rank them in terms of their legality. In such a case the analyst can simply say
all options are legal and may leave the legality criterion out of the criteria-alternatives matrix.
(This is what you can do in your policy issue paper for this course.)
The ethical implications of policy actions are often ignored in analyses, but I urge you to
consider them seriously. The equity criterion in Figure 4.1 is about these ethical implications. Is
the policy action equitable for the stakeholders? As you can imagine, assessing the equity of a
policy action can be tricky. What is an equitable distribution of resources is debatable. There are
different ways of defining equity and we will discuss them briefly in Lesson 8.
You will see more detailed examples of criteria-alternatives matrices in Lesson 8. For now,
please note that the criteria listed in Figure 4.1 (goal, cost, political feasibility, legality, and
equity) are generic and they should be specified in analyses. Also note that you will specify
(define) these criteria in your policy issue paper.
POTENTIAL SOURCES OF CRITERIA
Where do the decision criteria come from? You can infer answers this question based on the
discussions in the previous section. A brief summary here would be helpful.
In real-life analyses, the most likely scenario is that the “boss” of an analyst (the president,
governor, or the members of a parliamentary committee) provides him with the criteria to be
used in an analysis. If the boss has clear values and preferences, he/she may ask the analyst to
include them as goals in the analysis.
Also common is that laws or regulations may mandate that certain criteria should be used in
analyses. You will this in the example of the analysis of transportation alternatives in
Metropolitan Atlanta later in this lesson and in Lesson 8. In that case some of the criteria analysts
used were mandated by the regulations of the US Federal Transit Administration.
Less common, the analyst may have some say in selecting or formulating the criteria. Rarely the
analyst may be allowed inject her own values directly into the analysis. More often, the analyst is
allowed to formulate criteria based on others’ values and preferences. For example, the analyst
may be asked to synthesize the values of the political community or some commonly held public
values to formulate decision criteria. She may examine the community’s values, such as freedom
and equality, and generally accepted procures, such as the rule of law, in formulating some of the
decision criteria to be used in analyses. For example, the goal stated in Figure 4.1 (increase
access to education) is an expression of a socially acceptable value: access to education. The
analyst may formulate this general value in a specific goal: increase access to education for all
school-age children by 50% by the year 2020.
How would analyst know what values and preferences are dominant among the public at the time
of an analysis? The general answer is that she can consult with experts (academics, legal experts,
or theologians) to better understand these values. She may also refer to the opinion polls that are
conducted on various issues. An example of such polls is the General Social Survey, which is
conducted by the National Opinion Research Center (NORC) annually to gather date about the
values and attitudes of the American people (http://www3.norc.org/GSS+Website/). The NORC
also collaborates with its international partners to conduct cross-national surveys (see the
In some cases, the analyst may have the opportunity to conduct a thorough problem structuring
exercise, using Dunn’s model or some variation of it. At the end of this lesson, I will describe a
method of problem structuring, which is based on Dunn’s model.
Policy alternatives are the options an analyst and policymakers consider to address the policy
problem they defined and/or the goal(s) they want to reach. Alternatives are listed in the columns
in Figure 4.1. Policy analysts conduct analyses to evaluate the alternatives on the criteria listed in
criteria-alternatives matrices; in the end they will select the best alternative (the best method of
solving the problem). In the hypothetical example Figure 4.1, the best alternative (the winner)
was to “use gambling money.”
In both Figure 4.1 the alternatives were expressed as simple phrases. This does not always have
to be the case. It is not always preferable to express alternative sin simple phrases either.
Alternatives may be expressed in full sentences, paragraphs, or pictures/maps. There is no
universal rule for describing policy alternatives, but the rule of thumb is that the analyst should
be as clear and detailed as necessary in describing alternatives. You will see examples of
using maps in describing alternatives in the Metropolitan Atlanta example later in this lesson. I
expect that the policy in your policy issue paper you describe each of your alternatives in one
Alternatives can be formulated and reformulated in successive approximations (from more
general to more concrete and specific; from protoalternatives to alternatives). An analyst may
first define “protoalternatives” (rough sketches of descriptions) and the, define detailed
descriptions, which are called “alternatives.”
Alternatives may also be defined in discrete or continuous terms. The examples in Figure 4.1
are discrete alternatives: status quo, increase property taxes, increase sales taxes, and use
gambling money are distinctly different from each other. Continuous alternatives provide
analysts with flexibility and allow compromises be made in policy debates. If the idea of levying
carbon tax is agreed on, then a common ground may be found between the $20 and $50 options.
It is important to be flexible in defining and revising policy alternatives in real-life analyses. An
analyst should be prepared to formulate and reformulate alternatives during the implementation
POTENTIAL SOURCES OF POLICY ALTERNATIVES
Potential sources of policy alternatives are similar to those of decision criteria. In general either
the “boss” provides the analyst with the alternatives or the analyst may have some say in
identifying or formulating alternatives. In many cases, policy analysts have more flexibility in
identifying and formulating alternatives than in defining or formulating decision criteria.
Policymakers and stakeholders usually have their values and preferences set and ask for certain
criteria to be used in analyses, but they may not have clear ideas about alternative paths of
action. Then the analyst would have more flexibility.
In some cases the boss provides the analysts with her pet ideas. Then the analyst can formulate
specific alternatives using these pet ideas. Formulate, institutional inventories, and proposals that
are awaiting a window of opportunity may be the sources of this kind of alternatives.
When the analyst has the choice to formulate alternatives, he may use one or more of the
1. Idea generation. Creativity plays a role in idea generation. Idea generation may be done
individually, but preferably in groups. There are several methods of idea generation, such
as brainstorming, nominal group technique, cognitive mapping, and analogies (synectics)
for this purpose.
2. Using scientific theories and models. An analyst may employ econometric models,
multiple regression models, etc. to investigate a problem area and extract ideas of viable
policy solutions from them. Some of these models are discussed in PADM 550. Briefly
here, the analyst may identify the manipulable variables (or “intervention points”) in
policy models and reformulate them as policy alternatives.
3. Consulting with experts. This is similar to what an analyst can do in defining decision
criteria. He may consult with academics, lawyers, theologian, practitioners in the area of
study, and others to gather policy ideas.
4. Alternatives suggested in public debates and expressed values. The analyst may identify
the ideas expressed in the public discussions in the media and formulate alternatives from
A rich source of ideas for policy alternatives is policy think tanks. In the United States
alone, there are hundreds, even possibly thousands, of think thanks that generate ideas
and take policy positions. Most of these organizations take ideological, even partisan,
positions; therefore their policy ideas are usually slanted. Still a neutral analyst can find
useful ideas in their policy issue papers and other publications and adapt and refine them
for the purposes of his analysis.
5. Parallel cases or analogies. The analyst may adopt policy alternatives that were used in
other geographic locations or other areas of policy.
6. Generic sources of alternatives. Here are some generic ideas for policy alternatives:
• Subsidies and grants
• Service provision
• Information dissemination and educating the public
• Structuring private rights and framing economic activity
• Contracting out
RULES OF THUMB FOR FORMULATING POLICY ALTERNATIVES
The following rules of formulating policy alternatives are not universal or definitive. Each policy
analysis situation has its own conditions that may force the analyst to formulate alternatives
differently. The following are useful rules of thumb and they have been culled from various
1. The analyst should always include the status quo (no action) as an alternative. This is
about what happens if no action is taken. Even if there is a consensus among the
boss(es) of the policy analyst and stakeholders on not maintaining the status quo and
seeking an alternative path of action for the policy problem, it is useful to keep it as an
alternative in the analysis.
Including the status as an alternative will provide the analyst with a “base line” against
which all the other alternatives will be compared. Thus, the analyst can measure how
much improvement can be made by taking an alternative path, compared to the current
situation. This information can help policymakers to judge whether it is worth taking
any policy action at all. For example, if launching a remedial education program for the
youth in a community is going to improve their educational attainment only marginally,
then policymakers may decide not to take any action at all.
2. When formulating policy alternatives, the analyst may begin with a large number of
alternatives, but in the end the number of alternatives should not be more than 3 or
4. This is particularly important if the analyst does not have a long time period and a
large budget to conduct his analyses. Also the client of the analyst or stakeholders may
not have enough attention span to absorb the information about a large number of
When you formulate your alternatives for the policy issue paper you will write for
this course, abide by the rule of maximum 3-4 alternatives. Consider this: Including
too many alternatives may will increase the both the money and time the analyst will
end up spending exponentially. (You will see in Lesson 8 that each cell in a criteriaalternatives
matrix requires a separate set of analyses. This means that every time you
add another alternative to the matrix, you will multiply the number of analyses you have
to conduct by the time of the number of criteria you have in the matrix.)
3. An analyst may combine alternatives if reasonable or necessary. Not all alternatives
are mutually exclusive. Alternatives may be combined as a result of negotiations and rethinking
of the policy problem. In the example of how to improve educational
opportunities for children in Pennsylvania (Figure 4.1), policymakers and stakeholders
may end up combining the options of using sales tax and gambling money, for example.
4. An analyst should follow budgetary cycles, electoral cycles, and changes in the
economy (e.g., business cycles, unemployment trends, demographic changes) to
identify feasible alternatives. Some great policy ideas may not be feasible at certain
times. For example, it is not likely that policymakers in the United States would accept
increasing taxes as a viable alternative in an election year because of the aversion of the
American electorate to taxes.
5. In some cases the basic alternative (strategy) and its variants can be separated. If that
is the case, the analyst should clearly state the basic strategy and its variants. For
example, in Bardach’s outcome matrix, using “financial mechanisms” (the last group of
alternatives) is the basic strategy and the “$20 carbon tax” and “$50 dollar carbon tax”
are the variants.
CRITERIA–ALTERNATIVES MATRIX PROCEDURE
For numerical assessments, you will need to either rank-order alternatives on criteria or rate
alternatives on criteria. Ranking makes intuitive sense, but it would complicate the calculations
(rankings should be converted to ratings), as you will see in the illustrations below. The
following steps include rank-ordering alternatives, but you may skip this first step.
Step 1: Rank-order alternatives on criteria
Rank the alternatives on each of the decision criteria. These rankings should reflect your
assessments. These assessments may be based on some quantitative calculations, qualitative
judgments, or both. (As an example, see Bardach’a table for the summaries of his quantitative
assessments of each of the alternatives.) Whatever method assessment you use, in the end the
alternatives should be ranked.
Table 8.7: Initial Rankings of Alternatives on Criteria
Status Quo Alternative #1 Alternative #2
Goal 1 2 3
Cost 2 3 1
Political Feasibility 2 1 3
Equity/Fairness 2 3 1
Legality OK OK OK
In this example the alternatives are not ranked on the criterion of legality. Instead, the
analyst determined that they all are legal. Legality can be used as a screening criterion. If
any of the alternatives is not legal, it should not be considered at all; it should be removed
from the matrix. If there are any alternatives whose legality is questionable, then that
should be noted and the analyst should explore further before making a recommendation
about the alternatives.
How would you determine if an alternative would be legal, or constitutional? Ask legal
experts. Even they may not be able to give you a clear answer to the question whether an
alternative is legal. As you know, the legality of many new laws is questioned. Their
legality is determined by courts, eventually high courts, like the US Supreme Court.
For example, as you will remember, the constitutionality of the Affordable Care Act of
2010, otherwise known as “Obamacare,” was challenged by its opponents. The US
Supreme Court ruled in 2012 that the law was constitutional. The law’s most
controversial component was the “individual mandate,” which requires all Americans to
purchase health insurance or pay a “shared responsibility payment” to the government. In
its ruling, the court held that the law could be upheld on the grounds that the individual
mandate was a tax, and therefore it was within Congress’ taxation power.
Because the rankings in Table 8.5 are hypothetical, I cannot provide justification for
them. Let’s assume that the analyst investigates all these alternatives and the rankings in
the table are based on good analyses and judgments.
Note that there is an implicit assumption in Table 8.5. It is that all the criteria are equally
important. This may be problematic analytically. Usually in decision-making not all
criteria are equally important. Some may be more important than the others. For example,
the costs of the alternatives may the most important concern because the policymakers
(e.g., a city council) may be bound by budgetary restrictions. Then how can the different
levels of importance of the criteria be differentiated? They can be assigned weights. But
this would require another step.
Step 2: Convert rankings to ratings (or use ratings from the beginning)
Weights cannot be assigned to rankings. (If you are curious why, try to assign weights to
them and do the following calculations with rankings. Then you will see that the
calculations will not work.) To assign different weights to criteria, they should be
converted to ratings first. This could be done as follows.
Table 8.8: Converting Rankings to Ratings
In this example, I assumed that a rating scale of 1, 2, and 3 can be used. On this scale the
best alternative should be rated 3, the second best 2, and the worst alternative 1. See the
conversions of the rankings in Table 8.6.
Table 8.9: Conversions from Rankings to Ratings
Status Quo Alternative #1 Alternative #2
Goal 1 to 3 2 to 2 3 to 1
Cost 2 to 2 3 to 1 1 to 3
Political Feasibility 2 to 2 1 to 3 3 to 1
Equity/Fairness 2 to 2 3 to 1 1 to 3
Legality OK OK OK
Instead of ranking alternatives first, you may use a rating scale from the beginning. Then,
you would skip these conversions and rate the alternatives. Then the matrix would like
the one in Table 8.7.
Table 8.10: Direct Ratings
Status Quo Alternative #1 Alternative #2
Goal 3 2 1
Cost 2 1 3
Political Feasibility 2 3 1
Equity/Fairness 2 1 3
Legality OK OK OK
Step 3: Weighting criteria
As I mentioned earlier, the criteria use in a matrix are usually are not equally important.
Then they should be assigned different weights. This is easy to do arithmetically, but you
should be careful in assigning these weights. You should think about why you assign a
certain weight to each criterion. These assigned numbers will have consequences; they
will affect the decision in the end.
In the hypothetical example in Table 8.8 the cost is assigned 2.5, while the goal is
assigned 2. Because this is a hypothetical example, I cannot explain why these weights
are assigned to them, but I can tell you that cost will affect the decision (see the total
scores) more than goal will. If the cost were assigned 2 and goal 1, then cost would affect
the decision twice as much. In your policy issue papers you will need to justify your
selections of weights.
You will see that legality is not weighted. This is because if an alternative is either is
legal or not.
Table 8.11: Weighted Scores
Status Quo Alternative #1 Alternative #2
Goal (2) 3*2=6 2*2=4 1*2=2
Cost (2.5) 2*2.5=5 1*2.5=2.5 3*2.5=7.5
Political Feasibility (3) 2*3=6 3*3=9 1*3=3
Equity/Fairness (1.5) 2*1.5=3 1*1.5=1.5 3*1.5=4.5
Legality OK OK OK
Total Score 20 (Winner) 17 17
After you assign the weights, you should multiply each weight with the corresponding
rating of each alternative on each criterion (see Table 8.8). Then you should add up the
weighted ratings (bolded scores in Table 8.8) to calculate the total score for each
alternative. The alternative with the highest score will be the “winner” (the best
alternative). In this example the winner is the status quo.
A CRITERIA-ALTERNATIVES MATRIX EXAMPLE: TRANSPORTATION
ALTERNATIVES IN METROPOLITAN ATLANTA
The following is a real-life example of using criteria-alternatives matrices for policymaking and
The Atlanta metropolitan region (Metro Atlanta) has grown rapidly since the 1970s. The Georgia
state government and the local governments in the region built highways and byways to catch up
with this growth and attract even more economic and population growth. This rapid growth
created traffic congestion and air pollution over time. In the 1990s the US Environmental
Protection Agency found the counties in Metro Atlanta out of compliance with the National
Ambient Air Quality Standards (http://www.epa.gov/air/criteria.html).
Consequently the leaders of the state and local governments in the region decided to take action
to reduce the traffic congestion and air pollution in Metro Atlanta. The ensuing political and
policy process has been quite complex. If you are interested in the history of what happened, a
partial account can be found in the following:
Morçöl, G., Zimmermann, U., & Stich, B. (2003). The Georgia Regional Transportation
Authority: A smart growth machine? Politics and Policy, 31(3), 488-511.
Zimmermann, U., Morçöl, G., & Stich, B. (2003). From sprawl to smart growth: The case
of Atlanta. In M. J. Lindstrom, & H. Bartling (Eds.), Suburban sprawl: Culture, theory,
and politics. Lanham, MD: Rowman & Littlefield.
There are there are three regional state agencies that are directly or indirectly involved in
developing transportation alternatives and analyzing them in Metro Atlanta: The Georgia
Regional Transportation Authority (GRTA), Atlanta Regional Commission (ARC), and
Metropolitan Atlanta Rapid Transit Authority (MARTA). The GRTA and MARTA developed
the “transportation alternatives” and analyzed them, with the technical support provided by ARC
in the 1990s and 2000s. MARTA refined the studies and developed multiple transportation
projects (see http://www.itsmarta.com/Planning.aspx). I will focus only on the “Westline
Corridor Project” as an example (http://www.itsmarta.com/west-line-corr.aspx).
If you are not familiar with Metro Atlanta, the following map of the rail system and major
highways will help you locate these projects visually. (See the original map and other maps at
http://www.itsmarta.com/rail-map.aspx.) The aim of the “Westline Corridor Project” was to
extend the mass transit options beyond where the East –West rail line (blue line) ends on the
The “Westline Corridor Project” general area map is as follows.
For this project, MARTA developed first general transportation concepts and then refined them
to propose specific alternatives. The general transportation concepts were:
Light rail transit (LRT in the following table)
Heavy rail transit (HRT in the following table)
Bus rapid transit (BRT in the following table)
The specific transportation alternatives MARTA developed can be seen in the following criteriaalternatives
matrix. You can see that the alternatives are grouped under the categories “North
Corridor,” “South Corridor,” ‘central Corridor,” and “TSM” (an alternative bus operation plan).
You can see the document at this link for a detailed discussion of the alternatives:
At this link you can find detailed descriptions of each of the alternatives. The following concept
maps of the alternative corridors for the light and heavy rail options and the bus rapid transit
options, which are presented at this link, illustrate how maps can be used to describe alternatives.
Light Rail Corridors:
Bus Rapid Transit Corridors:
Heavy Rail Corridors:
In the same document, you can also see the following table, which summarizes the alternatives.
For more details, please read the document itself. We will come back to this Westline Corridor
Project example in Lesson 8, when we discuss how criteria-analyses can be made with criteriaalternatives
Table 8.14: Descriptions of Calculations in the Westline Project
Step 1 – For each goal a series of evaluation criteria were developed that reflected the
objectives associated with that goal.
Step 2 – A set of performance measures was established for each evaluation criterion
that would provide a good basis for assessing the alternatives. The number of performance
measures per evaluation criterion varied from 2 to 6.
Step 3 – Performance measure values were calculated for each alternative and then
rated (Very Desirable = 3, Desirable = 1, Less Desirable = -1) based on a quantitative and
qualitative comparison of the range of values calculated across the alternatives. The
process for assigning a performance measure ratings is described in more detail in the
Chapters 4-7, which describes the detailed evaluations comprehensively.
Step 4 – The ratings generated across performance measures were summed to create
a single composite score for each alternative for each Evaluation Criterion. The
scores were then rated (Very Desirable = 3, Desirable = 1, Less Desirable = -1) based on
range of scores across the alternatives.
Step 5 – An overall score for each goal was then calculated by summing the composite
scores for the evaluation criteria. Each goal was then rated again based on the range of
the overall scores across alternatives.
Step 6 – Finally, the ratings for overall score for each goal were summed into an
alternative final score for each alternative. The alternative final score was the basis for
comparing alternatives holistically, to assess their ability to address the project need and
purpose. This method removes any bias across goals and evaluation criteria, meaning that
each goal and criterion will affect the alternative final score equally
See the separate tables for each group of decision criteria on pages 9-5 through 9-8 at the
website. Note the “values,” “ratings,” and “composite scores” in each of these tables.
Tables 8.12, 8.13, 8.14, 8.15 display the ratings for each of the four goals of the project.
Particularly note that in Table 8.15 both costs and cost-effectiveness calculations are done
and they named as one of the four goals. (We will discuss cost-effectiveness analysis
briefly in the next lesson.)
Table 8.15 : Goal 1 – Improving Mobility and Accessibility Evaluation Results
Table 8.16 : Goal 2 – Preserve & Enhance the Environment Evaluation Results
Table 8.17 : Goal 3 – Economic Development & Transit Supportive Land Use Evaluation
Table 8.18 : Goal 4 – Cost and Cost Effectiveness Evaluation Results
The results from these four tables (8.12 through 8.15) are combined in table 8.16 below. (This is
Table 9.7: Final Alternative Scores Evaluation Results on page 9-8 in the report.) You can see
that the “winner” is the combined alternative. This is the alternative the MARTA analysts
Table 8.19: Evaluation Results