Chapter 14 attached , 1 page only

CHAPTER 14

The Art of Structuring and Writing a Health Porcy Analysis

LEARNING OBJECTIVES

By the end of this chapter you will be able to:

■ ·understand the concept of policy analysis ■ Analyze a health policy issue ■ Write a health policy analysis ■ Develop descriptive and analytic side-by-side tables

~ Introduction

Imagine you work for the governor of a state that recently received a large sum of federal money for antibioterrorism efforts and your boss asks you how it should be spent. How will you respond? Or, assume you are an assistant to the director of a state nutrition program for low-income children and the program's budget was slashed by the legislature. The program's director needs to reduce costs and turns to you for help. How will you approach this problem? Finally, pretend you work in the White House as a domestic policy advisor and the president is consid­ ering revising the administration's methane emissions guidelines. What guidance will you offer?

While the substance of health policy and law has been discussed in detail up to this point, this chap­ ter teaches you what policy analysis is and the skill of addressing complex health policy questions through a written policy analysis.

/1

'

We use the following definition to describe a policy analysis: an analysis that provides informed advice to a client that relates to a public policy deci­ sion, includes a recommended course of action/inac­ tion, and is framed by the client's powers and values. We briefly review each element of this definition in this chapter.

► Policy Analysis Overview In this section we define policy analysis and review the purposes for developing one. In the following section, we provide a step-by-step process detailing how to create a written health policy analysis.

Client-Oriented Advice The client is the particular stakeholder that requests the policy analysis, and the analysis must be developed to suit the client's needs. (The client, for example, could

317

r 7 318 CHAPTER 14 The Art of Structuring and Writing a Health Policy Analysis

be a policymaker who hires you, a fictional policy­ maker in an exercise developed by your professor, or an employer who asks you to analyze a problem.) In general, a stakeholder refers to an individual or a group that has an interest in the issue at hand. There may be many stakeholders related to a particular policy issue. Of course, the client requesting an analysis is also a stakeholder because that person or entity has an inter­ est in the issue. However, to avoid confusion, we refer to the person or group that requests the analysis as the client, and the other interested parties as stakeholders.

Informed Advice Providing informed advice means the analysis is based on thorough and well-rounded information. The information included in the analysis must convey all sides of an issue, not just the facts and theories sup­ porting a particular perspective. If a decision maker is presented with evidence supporting only one course of action or one side of a debate, it will be impossi­ ble for the client to make a well-informed decision. In addition, to be effective in persuading others to favor the recommended policy, your client must be able to understand and, when necessary, refute alternative solutions to the problem.

Public Policy Decision Policy analyses involve public policy decisions. A pub­ lic policy problem goes beyond the individual sphere and affects the greater community.

Providing Options and aRecommendation A key component of any policy analysis is providing the client with several options to consider, analyzing those options, and settling on one recommendation. In other words, a policy analysis is not simply a back­ ground report that identifies a variety of issues relat­ ing to a particular problem; instead, it gives the client ideas about what steps to take to address the problem and concludes by recommending a specific course of action.

Your Client's Power and Values Finally, the analysis should be framed by the client's power and values. Framing the analysis based on the client's power is fairly straightforward and uncontro­ versial: the options presented and the recommenda­ tion made must be within the power of the client to accomplish. On the other hand, the notion of fram­ ing an analysis according to the client's values is more

controversial. In most conceptualizations of policy analysis, including the one discussed later in this chapter, the process is roughly the same: define the problem and provide information about it, analyze a set of alternatives to solve the problem, and imple­ ment the best solution based on the analysis (Patton & Sawicki, 1993, p. 3). As new information is uncovered or the problem is reformulated, analysts may move back and forth among these steps in an iterative pro­ cess (Stone, 2002, p. 47). However, although there is general agreement that politics and values play a role in policy analysis, there is disagreement over at which stage of the analysis they come into play.

To understand this controversy, it is necessary to discuss two models of policy analysis: the rational model and the political model. The rational model was developed in an attempt to base policy decisions on reason and science rather than the vagaries of politics (Stone, 2002, p. 7). In the traditional rational model, the analyst does not consider politics and values. Instead, he or she should recommend the "rational, logical, and technically desirable policy" (Stone, 2002, p. 51). According to the rational model, the decision maker infuses the analysis with politics and values once the analyst's work is complete.

Professor John Kingdon and others have moved away from the rational model and toward a political model. Kingdon suggests that policy analysis occurs through the development of three streams: problems, policies, and politics (Kingdon, 1995). The problem stream is where problems are defined and noticed by decision makers. The policy stream is where solutions are proposed. These proposals may be solutiC>ns to identified problems, but they are often favored projects of policymakers or advocates that exist separate from specific problems that have garnered attention. Finally, the political stream refers to the ever-changing polit­ ical mood. As a general matter, these streams develop separately, only coming together at critical junctures when the problem reaches the top of the agenda, the solutions to that problem are viable, and the pplitical atmosphere makes the time right for change (Weissert & Weisser, 2002, p. 87).

Kingdon's approach discusses occurrences the rational model does not consider, such as why some problems are addressed and others are not, why some solutions are favored even if they are not tech­ nically the best approach, and why action is ta~en at some junctures but not at others (Kingdon, 1995). In addition, the rational model refers to only one cycle of problems. As Kingdon and others have noted, solu­ tions to one problem often lead to unintended conse­ quences that create other problems to be addressed,

resulting in an ongoing policy analysis cycle instead of an event with a start and a finish (Weissert & Weissert, 2002, p. 260).

Professor Deborah Stone also focuses on the role ofpolitics and values in analysis (Stone, 2002, pp. 1-4). She argues that the idea of the rational policy analy­ sis model misses the point because "analysis itself is a creature of politics" (Stone, 2002, p. 8). According to Stone, every choice, from defining a problem, to selecting analytic criteria, to choosing which options to evaluate, to making a recommendation, is political and value-laden. As she states, "Rational policy anal­ ysis can begin only after the relevant values have been identified and ... these values change over time as a result of the policymaking process" (Stone, 2002, p. 32). She contends that policy analysis should do the very things that the rational model does not permit­ allow for changing objectives, permit contradictory goals, and turn apparent losses into political gains (Stone, 2002, p. 9). The goal of the rational model founders-to divorce analysis from the vagaries of politics-is simply not possible in Stone's view.

Having differentiated these models, we now return to our definition of policy analysis: an analysis that provides informed advice to a client that relates to a public policy decision, includes a recommended course of action/inaction, and is framed by the client's powers and values. ·You can see that this definition follows Stone's political model of policy analysis, requiring the analysis to be developed with a particular client's values in mind. After reviewing the numerous exam­ ples provided in the following section, it will be evi­ dent that client values permeate all aspects of a policy analysis. Only after you take into account yotir client's values, combine it with the information you have gathered, place it in the prevailing political context, and understand your client's powers, can you make an appropriate policy recommendation.

Multiple Purposes The ultimate product of a policy analysis is a recom­ mendation to a specific client about how to address a problem. However, a policy analysis has several other purposes as well. It provides general information nec­ essary to understand the problem at hand and may be an important tool to inform stakeholders about a policy problem. In addition, the analysis may be a vehicle for widespread dissemination of ideas and arguments. Although your analysis is targeted to the client requesting advisement, it may also be used to inform and persuade other supporters, opponents, the media, the general public, and others. Finally, it

Structuring a Policy Analysis 319

will help you, the policy analyst, learn how to think through problems and develop solutions in an orga­ nized, concise, and useful way.

Policy analyses can take many forms-a memo­ randum, an oral briefing, a report, and so on-and, correspondingly, have varying degrees of formality. This chapter explains how to construct a short, writ­ ten analysis because it is a commonly used, highly effective, and often practical way to provide a pol­ icy analysis to your client. In addition, the principles embedded in a policy analysis can then be used in whatever other format(s) your client prefers. Whether you are aiding a governor, the director of a state pro­ gram, the CEO of a private business, or any other decision maker, you often will not have the oppor­ tunity to discuss issues in person or for a significant length of time. Furthermore, given time pressures, the demands on high-level policymakers, the need for rapid decision making, and the variety of issues most policymakers deal with, many clients will not read a lengthy analysis. That is why it is essential for anyone who wants to influence policy to be able to craft a clear and concise written analysis.

► Structuring a Policy Analysis We now turn to a five-step method for writing a thor­ ough yet concise policy analysis. Regardless of the subject matter, you can use this structure to analyze the question your client is considering. As you review each part of the analysis, notice the various disciplines and tools that may be part ofwriting an effective policy analysis. Analysts draw from a variety of disciplines'­ law, economics, political science, sociology, history, and others-and use a number .of quantitative and qualitative tools when explaining issues, analyzing options, and making recommendations.

Although policy analyses come in various formats and use different terminology, they will all contain these essential elements:

:. Problem identification: Defines the problem addressed in the analysis Background: Provides factual information needed to understand the problem Landscape: Reviews the various stakeholders and their concerns

■ Options: Describe and analyze several options to address the problem

■ Recommendation: Offers one option as the best action to pursue

The following sections discuss each of these ele­ ments in detail.

320 CHAPTER 14 The Art of Structuring and Writing a Health Policy Analysis

Problem Identification Define the Problem The first step in writing a policy analysis is to clearly define the problem you are analyzing. A problem iden­ tification should be succinct and written in the form of a question that identifies the problem addressed in the analysis. It usually consists of a single sentence, though it may be two sentences if you are analyzing a particularly complex issue. Although a problem iden­ tification is simple to define, it is often one of the most difficult parts of the analysis to do well. It is also one of the most important.

The problem identification is the key to your anal­ ysis because it frames the problem at hand. Indeed, some policy battles are won or lost simply by how the problem identification is crafted. For example, con­ sider the different questions asked in these problem identifications:

Problem identification I:

What type of tax credit, if any, should the presi­ dent include in the next budget proposal?

Problem identification 2:

What type and size of health insurance tax credit should the president include in the next budget proposal?

The first problem identification asks what type of tax credit, if any, should be considered. One possible answer to that question is that no tax credit ofany kind should be considered. Another answer could involve a tax credit, but not one related to health i'nsurance. The second problem identification suggests that the option of not proposing a health insurance tax credit is unac­ ceptable. Instead, the second problem identification lends itself to an analysis of identifying the pros and cons ofvarious health insurance tax credit options. In other words, one option that may be considered based on the first problem identification (no tax credit) is excluded based on the second problem identification.

Consider another example:

Problem identification I:

Which health issue should be the governor's top priority?

Problem identification 2:

Should the governor's priority of reducing the number of obese residents be accomplished by relying on currently existing programs?

Again, the first problem identification asks an open­ ended question about identifying the governor's top

priority. The answer may be obesity, but it may be any other health issue as well. The second problem identifica­ tion starts with the governor committed to reducing the obesity rate and asks how to best accomplish that goal. Again, these problem identification lead to very different analyses and (most likely) different recommendations.

It is possible that your client's values will be evident from the way the problem identification is phrased. For instance, in the second example, the second pol­ icy statement clearly reflects the governor's desire to reduce the obesity rate. Consider another example. You have been asked to write a policy analysis about the merits of importing low-cost prescription drugs from Canada. How might the problem identifica­ tion differ if your client is a pharmaceutical lobbying group versus if your client is an elder rights associa­ tion? Here are two possible problem identifications.

An acceptable way to identify the problem for the pharmaceutical lobbying firm:

How can this firm help improve medical care quality in the United States by reducing the importation of dangerous prescription drugs from Canada?

An acceptable way to identify the problem for the elder rights association:

How can this association help seniors obtain low­ priced prescription drugs from Canada?

The vast differences in these problem identifi­ cations reflect differing viewpoints regarding the importation of prescription drugs. A pharmaceutical lobbying firm is more likely to be concerned about reduced profits for its drug company clients and therefore would want to deter or restrict importation, which could lead to more competition in the market. One way to accomplish that goal is to phrase the issue as a safety/quality-of-care concern. An elder rights association is more likely to be concerned with high­ priced prescription drugs in the United States and would therefore want to promote drug importation (assuming, of course, there 1s no actual safety concern with the drugs). One way to accomplish that goal is to phrase the issue as one of c,ost reduction.

It is also possible to identify the problem in a neu­ tral manner. From the immediately previous exam­ ple, analysts for both groups could use the following phrasing.

Neutral problem identification:

What action should [ the client] take in response to recent congressional proposals relating to import­ ing prescription drugs from Canada?

Structuring a Policy Analysis 321

Neutral wording is not necessarily better or worse than value-driven phrasing. The value-driven phrase provides additional information about the direction of the policy analysis and clearly limits some of the options that might otherwise be considered. Neutral wording is often broader, leaving more options on the table at the outset. Yet, even if neutral phrasing is used, the options the analyst considers and the recommen­ dation the analyst makes will still be constrained by the client's values and needs.

Because it is possible to identify problems numer­ ous ways for any issue, how do you develop the best one? In general, you want to use the one that provides the best fit for your analysis. If you decide to focus on all safety-related concerns, for example, the pharmaceutical lobbying firm example would be a good fit. If, however, the firm chose to have one safety-related option, one state government-focused option, and one convenience­ centered option, then you would want a broader prob­ lem identification that allows for all types of options. For example, "How can this firm help improve consum­ ers' experience purchasing prescription drugs?" would allow for various types of options. In addition, you want to make sure your problem identification is analytically manageable and does not include your recommendation.

Make the Problem Identification Analytically Manageable Acceptable phrasing can be broad or narrow. One is not better than the other; they suit different purposes. Policy_analyses with broad problem identifications may require more diverse information in terms of background and may consider a wider range of issues in the paper's landscape section. Also, the recommen­ dations may promote "big picture" changes instead of specific and tailored ideas. Narrower wording may require less extensive background and landscape infor­ mation, but they may not capture big picture, systemic concerns relating to the problem under consideration.

Reflect on the following examples. They both may be acceptable problem identifications, depending on the needs of your client.

A broad problem identification:

What action should the U.S. Department of Health and Human Services take to avoid another flu vaccine shortage?

A narrow problem identification:

How can the U.S. Department ofHealth and Human Services create incentives for additional manufac­ turers to supply flu vaccine to the United States?

The first one could result in a variety of rec­ ommendations, such as improving surveillance to lower the incidence of flu in the future, developing new vaccines that have longer-lasting immunity, or finding ways to entice additional manufacturers to provide supplies to the United States. It is a broad problem identification that will lead to an analysis that could recommend a wide variety of actions. The second one focuses on one particular way to decrease a flu vaccine shortage-increasing the number of suppliers. Although the analysis will also provide a number of options, all of the options will address the specific issue of increasing suppliers. Again, there is no single right or wrong way to phrase a problem. Whether it is more useful to have a broad or narrow problem will depend on the needs and concerns of your client.

However,, it is possible to make a problem iden­ tification so vague that it will be impossible to write a sound policy analysis. Unfortunately, there is no easy way to differentiate an acceptably broad problem phrasing from an unmanageably vague one when you begin your analysis. You will know your wording is too vague, however, if you find it impossible to write a complete and concise policy analysis and you can think of too many diverse options to include (if you can think of 5 to 10 possible options your statement is probably workable; if you can think of 15 or more options, your statement is probably too broad). Your paper will require too much information in the back­ ground and landscape sections if you draft an overly vague problem identification. In addition, you will find that you cannot devise a coherent series ofoptions addressing the problem because the problem identifi­ cation is too broadly defined. Instead of a concise and useful policy analysis, you will end up with a lengthy and unfocused paper.

If you believe your problem identification may be so vague that it is analytically unmanageable, ask yourself if you are addressing one specific problem that may be countered with a few specific options. If you are having trouble narrowing your problem iden­ tification, you can try including limitations based on geography (e.g., refer to a particular state or city), time (e.g., focus on the next year or over the next 5 years), or numerical boundaries ( e.g., use a goal of reducing a figure by a certain percentage or a budget by a specific dollar amount). Consider this example:

An unmanageably vague problem identification:

What is the best use of the Centers for Disease Control and Prevention's resources to improve the health status of our citizens?

322 CHAPTER 14 The Art of Structuring and Writing a Health Policy Analysis

This phrasing is not analytically manageable. It is extremely broad and unfocused. Using this problem identification, your analysis could address any health issue, such as access to care problems, the need to improve vaccination rates, racial disparities in health care, or many others. The list is endless and your pol­ icy analysis will be as well.

A manageable problem identification:

What preventive health issue should be the top priority for the Centers for Disease Control and Prevention next year?

This one is analytically manageable. It is focused on preventive measures specifically and is limited to determining the top priority. In addition, the problem is focused on what can be done in the upcoming year. This second problem identification allows for a much more concise and directed policy analysis.

Do Not Include the Recommendation in Your Problem Identification Another pitfall when identifying problems is crafting a problem in a way that suggests a particular solution to the issue. A problem identification should define a specific problem; it should not indicate how that prob­ lem should be solved. If the answer is preordained, why bother with the analysis? When phrasing your problem , ask yourself ifyou can imagine four, five, or six potentially viable options to address the problem. If you cannot, then you have not defined the problem well.

For example, assume you've been asked to address how to reduce medical malpractice insurance premi­ ums. Here is a problem identification that leads the reader to one conclusion:

A leading problem identification:

To what extent should jury awards be limited in malpractice cases in order to reduce malpractice premiums?

This phrasing leads to one very specific solution (limiting jury awards) as a way to counter a broad problem (reducing malpractice premiums). The only question presented by this problem identification is what the award limit should be. It does not provide a range of options for reducing malpractice premiums (one of which may be limiting jury awards) for your client to consider. (Of course, if you were specifically asked to address how to limit jury awards in malprac­ tice cases, this would be an appropriate problem iden - tification.) A better way to word the problem would be, "What action should be taken to stem the rise in malpractice premiums nationwide?"

This problem identification lends itself to an analysis that considers several options. Possible alter­ natives include limiting jury awards, enacting regu­ lations that limit the amount insurance companies can increase premiums each year, and a host of other options. This wording also narrows the focus of the analysis to national solutions.

Once you have written your concise and precise problem identification, you have set the framework for your analysis. Every other section of the analysis should relate directly to the problem you identified. Remember that writing a policy analysis is an itera­ tive process; you must review, revise, and tighten the information and arguments throughout the writing process. As you review the other components ofyour policy analysis, it may become evident that what you thought was the best way to describe the problems can be further improved. There is nothing wrong with revising your wording as you craft your analy­ sis, as long as you remain true to your client's values and power.

The Background Section The first substantive information your analysis pro­ vides is in the background section. The background informs the reader why the particular problem has been chosen for analysis. This section should make clear why the issue is important and needs to be addressed now. In addition to providing general information about the topic, your background and landscape ( discussed next) sections provide the information necessary to assess the options you lay out.

Much ofthe information in the background will be relevant regardless ofwho assigned the analysis. How­ ever, because the background provides information necessary to understand the problem, it is essential to understand the knowledge level of your client when constructing the background. For example, assume you are writing an analysis relating to state prepared­ ness planning for smallpox vaccination in the event of a bioterror attack. Regardless of your client, your back­ ground would likely include information about why a smallpox attack is a threat, including (but not limited to) the following:

Reference to the September 11, 2001, attacks on the World Trade Center and subsequent events The belief that although smallpox has been erad­ icated as a natural disease, it is likely that samples of the virus still exist Reference to any information provided by the fed­ eral government or other sources relating to the possibility of a bioterror attack

Ifyour client does not have knowledge relating to smallpox, you would also include details about smallpox transmission, the effects of the disease, the vaccination procedure, and the risks associated with vaccination.

In addition, your background should include whatever factual information is necessary to fully assess the options discussed. Remember, your client needs a complete picture, not just the information that supports the recommended action or your client's viewpoint. By the time the reader reaches your paper's options section, all of the information necessary to evaluate the options should have been presented in your background or landscape.

For example, assume one of your options for the smallpox vaccination state preparedness analy­ sis is immediate compulsory vaccination of all first responders and establishment of a protocol for vacci­ nating the remaining population if there is a smallpox outbreak. In that case, your background (and possibly the landscape) should provide information regarding who first responders are, where they are located, how many there are, legal issues relating to compulsory vaccination, and so on.

Because the background section is an informational-not analytical-part of the analysis, the material provided in it should be mostly factual. The tone of the background is not partisan or argumenta­ tive. It should simply state the necessary information.

The Landscape Section Together; the background and landscape sections frame the context of the analysis for your client. Whereas the background provides factual informa­ tion to assist the client in understanding what the problem is and why it is being addressed, the land­ scape provides the overall context for the analysis by identifying key stakeholders and the factors that must be considered when analyzing the problem. In the following discussion, you will read about numer­ ous types of people, groups, and issues that might be included in a landscape. These examples are meant to provide suggestions and provoke thought about what should be included in an analysis. It would be impos­ sible to include everything discussed here in any sin­ gle landscape section. It is the job of the policy analyst to choose among these options-to be able to identify whose views and which factors are the most salient ones in creating a complete landscape.

Identifying Key Stakeholders Up to this point, the policy analysis discussion has focused on just one stakeholder: the client who asked for the policy analysis. The landscape brings in

Structuring a Policy Analysis 323

other stakeholders who have an interest in the issue. Although it is often impossible to include every pos­ sible stakeholder in a single analysis, it is necessary to identify the key stakeholders whose positions and concerns must be understood before a well-informed decision can be made.

How do you identify the key stakeholders particular to your issue? Unfortunately, there is no magic formula. The best approach is through research and thinking. Also, bear in mind that the stakeholders and issues discussed in the landscape must relate to your overall policy analysis. Your options must address the problem identified initially, and all of the information necessary to assess the options must be presented in the back­ ground and landscape. As you learn about the problem to be analyzed and think about options for addressing the problem, it should become apparent which stake­ holders have a significant interest in the issue.

For example, assume your analysis relates to pro­ posed legislation regulating pharmacists and pharma­ ceuticals. Who are possible key stakeholders regarding this issue? They may include several parties:

Democratic and Republican politicians (you might need to distinguish among those in Con­ gress, state legislatures, and governors) Pharmaceutical industry Health insurance industry AARP and other elder rights groups Advocacy groups for the disabled Pharmacists' lobby Foreign pharmaceutical companies Internet-based pharmaceutical companies

Can you think of others? Deciding which stakeholders must be included

in an analysis depends on the exact problem being analyzed and which options are included in the anal­ ysis. For example, assume you are writing an analy­ sis about requirements relating to how pharmacists inform consumers about their medications. If your problem deals with face-to-face encounters only, Internet-based pharmaceutical companies would not be a key stakeholder. Alternatively, if the prob­ lem definition deals with purchasing limitations over the Internet, Internet-based pharmaceutical companies must be included in the analysis. If one of your options focuses on comprehension among older adults, elder rights groups should be included in the analysis. If your client is someone running for public office in a county where the health insur­ ance industry is a major employer, the views of those companies would be essential to include. In other words, although it is possible to make a generic list of the types of individuals and groups that could be

324 CHAPTER 14 The Art of Structuring and Writing a Health Policy Analysis

included, the specific list for any policy analysis will landscape as the portion of the analysis that provides depend on the client for whom the analysis is being the overall context of the issue. The "overall context" written, the specific problem being addressed, and refers to the mix of factors that are relevant when the options being considered. any decision is being made. These factors are used

to analyze stakeholder positions. Although there is Identifying Key Factors not one comprehensive list of context factors and Once you have settled on a list of key stakeholders, not all factors are relevant to all analyses, the list in it is necessary to analyze their position on the issue BOX 14·1 includes some common factors that could at the center of the analysis. Earlier, we described the be discussed.

BOX 14·1 Possible Factors to Include in aLandscape Section

Political Factors

What is the political salience of the issue? Is this a front-burner issue? Is this a controversial issue? Are your client, legislators, and the general public interested in addressing this issue? Has this or a similar healthcare issue been addressed recently? Do key Constituents, opponents, interest groups, or other stakeholders have an opinion about the issue? Who is likely to

support or oppose change? Is there bipartisan support for the issue? Is there a reason to act now? Is there a reason to delay action?

Social Factors Who is affected by this problem? According to the client who assigned the analysis, are influential or valued people or groups affected by this problem? Is there a fairness concern relating to this issue? Is there a stigma associated with this issue?

Economic Factors What is the economic impact of addressing this problem? Of not addressing it? Are various people or groups impacted differently? Are there competing demands for resources that relate to this issue? What is the economic situation of the state or nation? How does this affect the politics relating to this issue? How will addressing this issue affect healthcare costs/healthcare spending?

Practical Factors Is it realistic to try to solve this problem? Do others need to be involved to be able to solve this problem? Is the technology available to solve this problem? Would it be more practical to solve this problem later? Are other people in a better position to solve this problem? What do we know about solutions that do or do not work? If this problem cannot be solved, is it still necessary (politically, socially) to act in some way to address the problem? Is evidence available to support potential solutions?

Legal Factors Are there legal restrictions affecting this problem? Is there a need to balance public health concerns and individual legal rights? Are there legal requirements that impact the analysis? Is new legislative authority necessary to solve the problem? Is there legal uncertainty relating to this problem? Is future litigation a concern if action is taken?

Structuring a Policy Analysis 325

Quality-of-Care Factors Does this problem address quality-of-care issues? Do some solutions focus on quality of care more than others? Do quality-of-care concerns vary based on which provider is involved? Is evidence available about the best ways to improve quality of care? Has the client already taken any actions to improve quality of care?

The list of factors in Box 14-1 is by no means exhaustive, but rather is intended to provide a sense of the types of questions that are often relevant to under­ standing the context of a problem. Just as it would be impossible to discuss every stakeholder who might be connected to the problem, it would be impossible to include in a concise analysis of all the issues raised by the five factors listed in Box 14-1. It is the policy ana­ lyst's job to identify not only which key stakeholders must be included, but also which key factors must be discussed.

Consider again the example relating to limiting prescription drug purchases made over the Internet. The following key fa<;:tors would likely be included:

Political factor: Who supports or opposes the limitation and how influential are they? Economic factor: Would it be costly to imple­ ment this limitation or costly to key constituents who may need more prescription drugs than are allowed by the limit? Would it provide an eco­ nomic benefit to some stakeholders? Practical factor: Is it possible to implement and enforce a restricted purchasing system over the Internet? Legal factor: Are there legal barriers to limiting Internet prescription drug purchases?

For each factor, the analysis should discuss rele­ vant views of the stakeholders you included in your analysis. For example, when discussing economic fac­ tors, the analysis might explain that Internet-based pharmaceutical companies would experience a loss of revenue, some consumers would pay more for pre­ scription drugs if they needed more than the limited amount, and storefront pharmaceutical companies would make money because consumers probably would have to fill more of their orders in person. Depending on the political situation, the analysis might explain that the client's most influential con­ stituents are older adults who use many prescrip­ tion drugs and are likely to oppose any restriction, or that storefront pharmaceutical companies are large campaign contributors and would support a restriction.

Structuring the Landscape Section The landscape may be organized by stakeholder or by factor. In other words, it is acceptable to identify stakeholder 1 and then describe that stakeholder's views based on various concerns, and then identify stakeholder 2 and describe that stakeholder's views, and so on. Alternatively, the landscape could be struc­ tured based on factors described in Box 14-1. Within that structure, the landscape would address one set of factors (e.g., economic), then another set (e.g., social), and so on. Some stakeholders may not have relevant views for all of the factors, but each stakeholder must be addressed as often a·s necessary to convey their policy position. For example, the economic factors section would identify various stakeholders and their views based on economic concerns. It would include information such as "Canadian companies would oppose regulating Internet pharmacies because it would increase the cost of doing business with Amer­ icans, while U.S. storefront pharmacists would sup­ port regulating Internet pharmacies because it may increase prices for their competitors' products:'

When you discuss a particular view, it is import­ ant to identify which stakeholders hold that view. At the same time, do not insert opinions unattached to individuals or groups. For example, it would not be helpful to write, "Some oppose legislation regulating Internet pharmacies:' Such a statement does not iden­ tify who opposes the legislation or why they oppose the legislation. Ifyour client will be able to assess the polit­ ical, practical, and ethical feasibility of taking a partic­ ular action, the client needs to understand where the various groups stand and why they hold those views. Thus, a more helpful sentence would be: "RxData, a large software company that assists Internet pharma­ cies and is a major employer in your congressional district, opposes legislation regulating Internet phar­ macies because it will add costs to its business and may result in future employee layoffs:' That sentence tells your client who is opposed to the legislation, why they are opposed to it, and why they are a key stakeholder.

The tone of the landscape should be neutral and objective. The landscape is not an argumentative

326 CHAPTER 14 The Art of Structuring and Writing a Health Policy Analysis

section to persuade the reader that one view is better than another. Its purpose is to identify all of the key stakeholders and their concerns so the decision maker is well-informed before assessing options.

The Options Section Once your client finishes reading the background and landscape sections, he or she should have a thorough understanding of the policy problem and the overall context in which a decision will be made. The time has come for him to consider what action to take. This is where your paper's options section comes in, providing three to five alternatives for your client to consider. This section is more than a recitation ofvar­ ious choices; it provides an analysis of each option by stating the positive and negative aspects of pursuing each path.

Identifying Options The first step to writing your options section is to iden­ tify the various options you could analyze. Although you may develop a new option not previously consid­ ered by others, it is likely that you will find numerous possibilities that others have already suggested. Some places t<_? find already-considered options include the following:

r: Media Scholarly articles Interest group recommendations Think tanks/experts in the field Congressional testimony Legislation (passed or proposed) Agency reports

Another approach to developing options is to consider several major actions that policymakers often take to deal with a policy problem. For exam - ple, depending on the circumstances, the following actions could be added, eliminated, or altered by a policymaker in response to a policy problem:

Taxes Subsidies Grants Laws Regulations Programs Government organizations Information Outreach

Once you have compiled a list of options, how do you choose the best three to five options to include in the analysis? As always, the guiding principle

underlying your decision is to base your decision on your client:s values and po~ers. You should not sug­ gest an option that clearly v10lates your client's values because your client will not seriously consider that option. For example, ifyour client has rejected spend­ ing new state money to solve a problem, you should not include an option that would cost the state money (or if you do, you would need to be clear about how these funds would be offset). Also, it is important to remember the extent ofyour client's power. For exam­ ple, on one hand, an analyst might recommend that a member of Congress introduce or sponsor a bill, or that an interest group submit a comment as part of the public notice and comment period before a reg­ ulation is finalized. On the other hand, an analyst should not recommend that a member of Congress force a state to pass a particular law, because that action is not within Congress's power (the responsi­ bility of passing state laws falls to state legislatures). Similarly, an interest group cannot issue an executive order or force an administrative agency to craft a reg­ ulation in a particular way. In other words, any option you include must be within the ability of your client to undertake.

In addition, you will probably want to include in your analysis any major proposals that are currently being considered. You may also choose to include pro­ posals backed by key allies and constituents. Even if your client does not act on these proposals, it is import­ ant to be able to explain why "mainstream'' options are being rejected. Also, it may be appropriate to con­ sider the status quo (the "do-nothing" approach) as an option. The status quo option works best when there is a specific reason for waiting to act; otherwise, it may appear to be an option because the analyst ran out of ideas. For example, if there is an evaluation under way ofa key program, a valid status quo option would be to wait until the evaluation is complete and then recon­ sider the issue based on the findings. Simply suggest­ ing to do nothing without a rationale is not acceptable. Even when considering the do-nothing tactic, it is necessary to evaluate the pros and cons of this option. In addition, your options should be different enough from one another to give your client a real choice. For example, if all of your options involved creating a big, new federal program and your client does not pref~r that strategy, then the analysis is not useful. There ~s one final rule to follow: whichever alternative you ulti­ mately recommend must be analyzed in the options section ofyour analysis.

All of the options included in the analysis must directly address the problem identified in your prob­ lem identification. For example, if your problem identification involved ways to reduce the number

Structuring a Policy Analysis 327

of uninsured individuals, an option that promotes increasing access to care for insured residents would be inappropriate because it does not directly address the problem. Indeed, everything in your analysis must flow from your problem identification, from what is included in the background and the issues discussed in your landscape, to the alternatives suggested in the options section, and (as discussed later) your final recommendation.

Assessing Your Options Your options analysis must include a discussion of the pros (what is useful) and the cons (what is prob­ lematic) of each option. Although policymakers often seek a "silver bullet" that solves a problem without any negative effects, it is highly unlikely that such an option exists. As a result, an analysis must include the positive and negative aspects of an option. Also, all options must be analyzed equally. Do not provide more analysis for the option you decide to recommend even though it is your preferred option. Your client can make a fully informed decision only if you analyze what is both good and bad about each proposal.

To draft your pros and cons, you must · identify the appropriate criteria for the analysis of the pros and cons and apply each criterion to all of the options. Many possible criteria could be used, so you must pick the ones that best fit your client's values and that address the key issues relating to the policy problem. Sometimes your client will provide you with criteria; other times you must deduce the correct criteria based on what you know about your client and the problem. Generally you should choose between three and five criteria for your evaluation of the pros and cons in this type of analysis. Typically, fewer than three will not allow for a full analysis, and more than five are difficult to assess in arelatively short analysis.

Your choice of criteria should reflect the concerns discussed in the landscape, where you identified the key factors relating to the problem. For example, if economic burden was an important aspect of the issue being considered, costs should be one of the crite­ ria used to analyze the options. If the need for quick action was an important aspect of the issue, timeliness and administrative ease might be important criteria to consider. In general, cost and political feasibility are usually important criteria to include in your analysis. However, the specific criteria that are best for your analysis will depend on your client, the specific prob­ lem identification being addressed, and the landscape ofthe issue. BOX 14-2 provides some criteria that could be considered when analyzing options. Can you think of others?

BOX 14·2 Sample Options Criteria

Cost: How much does this option cost? (You may have to break this down: how much does it cost the federal government, state government, individuals, etc.?)

Cost-benefit: How much "bang for the buck" does this option provide? (This is a difficult criterion to assess in shorter, less complicated policy analyses because you often do not have the information necessary to make this determination.)

Political feasibility: Is this option politically viable? Is it likely to become law? Even if it is not likely to become law, is it likely to help your client politically?

Legality: Is this option legal? If so, are there any restrictions?

Administrative ease: Does this option have steep implementation hurdles?

Fairness: Are people affected by this option treated fairly/equally?

Timeliness: Can this option be implemented in an appropriate or useful amount of time?

Targeted impact: Does this option target the population/issue involved?

Effectiveness: How well does this option accomplish the goal set out in the problem identification?

Healthcare-specific criteria: Does this option improve quality of care, lower healthcare spending, reduce the incidence of a particular disease, etc.?

Structuring the Options Section When structuring your options section, you may find it helpful to use headings or bullets delineating each option. Your headings should clearly label each option. Also, it is generally useful to separate out the options by paragraph instead of having a continuous page or two assessing all of the options. The first sen - tence or two of each options paragraph should be a clear description of that option so the client under­ stands what is being proposed and exactly what action your client would be taking; then assess the pros and cons of the option based on the criteria you identified. You should explicitly state the criteria prior to your options description.

For example, assume you are writing a policy analysis for the Secretary of the Department of Health and Human Services about how to increase access to care for immigrants. After drafting the necessary background and landscape information, you decide to assess the following options: (1) creating a public edu­ cation campaign, (2) providing grants to states, and (3) increasing funding to community health centers.

•

328 CHAPTER 14 The Art of Structuring and Writing a Health Policy Analysis

You decide the best criteria to use are cost to the fed­ care, or address concerns in the immigrant eral government, political feasibility, and degree of community about being deported if they par­ increase in the number of immigrants who seek care. ticipate in public programs. In your options section, you would have three differ­

You would write a similar analysis for each of theent parts-one for each option-assessing the pros options under consideration. and cons based on the criteria you chose. Although

the structure of the analysis may vary, you might have a paragraph looking something like this: Side-by-Side Tables

Option 1: Create a Public Education Cam­ You may choose to assess your options with the help paign. The Secretary of the Department of of a side-by-side table. This table may be descriptive or Health and Human Services should order analytic. A descriptive table would provide a descrip­ the Centers for Disease Control and Preven­ tion of each option but not provide any analysis. An tion to create a public education campaign to analytic table would assess the options based on the alert immigrants to the various ways they can criteria chosen. In either case, the table should be access health care in their area. The campaign appropriately labeled and easy to read. As a general will be delivered in English and the most com­ matter, a side-by-side table should supplement, not mon immigrant languages in the area and will replace, your textual analysis because it is difficult to occur on television, over the radio, and in provide sufficient information within the space pro­ public (e.g., billboards). In these difficult eco­ vided by a table. However, a table is a useful visual aid nomic times, this is a relatively low-cost option that may enhance the reader's understanding of your (although if immigrants access more care, the options and overall analysis. TABLES 14·1 and 14·2 pro­ cost of providing care will increase) . In addi­ vide examples of side-by-side tables. tion, the political feasibility is somewhat high As you can see from both tables, additional infor­ because this option does not call for any pol­ mation is necessary for a complete analysis. In the icy changes, but rather merely alerting indi­ text of your options section, you would need to pro­ viduals to policies that already exist. On the vide a more complete explanation of each option. It other hand, the political feasibility is tempered is difficult (but not impossible) to provide sufficient because the option focuses on helping immi­ depth of analysis in a chart without creating a busy, grants, a group that has been under attack difficult-to-read table. politically in recent times. The most signifi­ Although the second table uses the terms low, cant disadvantage of this option is that it may medium, and high to assess the options, other terms not be very effective in increasing the number or symbols (e.g.,+, - ) may be used. In addition, you of immigrants who access care because it does may choose to include phrases instead of single words not increase access points, reduce th~ cost of or symbols. Whatever choice you make, be sure that it

TABLE 14·1 Descriptive Side-by-Side Table

General Campaign on rad io, Federal government provides funds Federal government provides description television, and public areas to states to increase access to care funding to health centers for

for immigrants services

Populations All will hear, focus on Only immigrants in the state All health tenter patients,

t affected immigrants in community including immigrants Length of 1 year 5 years, subject to annual 2 years option appropriations

Payer Federal Federal Federal

Structuring a Policy Analysis 329

; TABLE 14·2 Analytic Side-by-Side Table

Cost to federal government Low Medium High

Political feasibility Medium Low High

Increase in access for immigrants Low High Medium

BOX 14·3 Checklist for Writing aPolicy Analysis

1. Problem Identification Is my problem written as one sentence in the form of a question? Can I identify the focus of my problem ? Can I identify several options (but not too many) for solving the problem?

2. Background Does my background include all necessary factual information? Have I eliminated information that is not directly relevant to the analysis? Is the tone of my background appropriate?

3. Landscape Does the landscape identify all of the key stakeholders? Are the stakeholders'views described clearly and accurately? Is the structure of the landscape consistent and easy to follow? Is the tone of the landscape appropriate? Does the reader have all the information necessary to assess the options?

4. Options Do my options directly address the issue identified in the problem identification? Did I assess the pros and cons of each option? Did I apply all of the criteria to each option's assessment? Are the options sufficiently different from each other to give the client a real choice? Are all of the options within the power of my client?

S. Recommendation Is my recommendation one of the options assessed? Did I recommend only one of my options? Did I explain why this recommendation is the best option, despite its flaws?

is clear which assessment is positive and which is nega­ that balance. An attempt to use quantitative measures tive. When necessary, provide a legend explaining your may lead readers to assume a simple summation will terms. Also, some people prefer to use numerical assess­ suffice-just add up the columns and choose the best ments (e.g., 1 = low, 2 =medium, 3 =high). Although one. Overall, reducing the analysis to numerical labels debates exist regarding the value of this type of quan­ hides the value judgments that are part of every analy­ titative assessment in a side-by-side chart, we strongly sis, fails to address whether certain criteria should be discourage it. Policy analysis is both an art and a sci­ weighted more than others, and makes the optimal ence, and using a numerically based table can obscure solution appear more certain than it probably is.

330 CHAPTER 14 The Art of Structuring and Writing a Health Policy Analysis

The Recommendation Section The time has finally arrived in your policy analysis to make a recommendation. You should choose one of your options as your recommendation. Although it is possible to make more than one recommendation or a hybrid recommendation of multiple options, we discourage those approaches. Making multiple recom­ mendations might make it necessary for your client to conduct further analysis to choose among the options, and making a hybrid recommendation (a combina­ tion of two or more options) is not appropriate unless the hybrid option was analyzed separately as a single option. In general, it is the analyst's job to organize and clarify the issues and place them in the context of the client's views and power. Ultimately, this approach should result in one path that you believe best suits your client.

The recommendation section should begin by clearly identifying which option is favored and why this option is preferred over the other ones. As

References Kingdon, J. (1995). Agendas, alternatives, and public policies (2nd

ed.). New York, NY: Addison-Wesley Educational Publishers. Patton C. V., & Sawicki, D. S. (1993). Basic methods of policy

analysis and planning (2nd ed.). Englewood Cliffs, NJ: Prentice Hall.

mentioned in the options section, every alternative will have pros and cons. The recommendation section does not simply repeat the analysis in the options sec­ tion. Instead, this portion ofthe analysis must explain why, despite the drawbacks, this is the best action to take based on your client's values and power. In addi­ tion, the recommendation section identifies what, if any, actions may be taken to mitigate or overcome the negative aspects of your recommendation.

► Conclusion You now have a basic understanding of what policy analysis is and an introduction to the tools necessary to analyze a policy problem. Additionally, thinking through the process described in this chapter should train you to evaluate your options when you become the decision maker. The checklist in BOX 14·3 provides examples of what should be included in each section of a written policy analysis.

Stone, D. (2002). Policy paradox: The art of political decision making. New York, NY: W. W. Norton.

Weissert, C. S., & Weissert, W. G. (2002). Governing health: The politics of health policy (2nd ed.). Baltimore, MD: Johns Hopkins University Press.