Benefit-cost Analysis

A Political Economy Approach

A. Allan Schmid

• Publisher: Taylor & Francis
• Year: 2019
• Print ISBN: 9780367003166
• eBook ISBN: 9780429718601
• Edition: 1
• Topic: Politics & International Relations
• Subtopic: Politics
• Index: Politics & International Relations

Choice is the name of the game. Government sets the size of the public budget and decides which public projects it will invest in and which transfers and regulations it will implement. To do this systematically the government must have a procedure that displays the consequences of the alternatives. This book is an exposition of benefit-cost analysis (BCA), an analytic framework for organizing thoughts, listing the pros and cons of alternatives, and determining values for all relevant factors so that the alternatives can be ranked. A major question illuminated by this text is whether the results of such an analysis can instruct government--in the sense of telling it what it must do to avoid being labelled stupid, corrupt, irrational, and/or inefficient. How and when, we will ask, can the benefit-cost analyst label a particular governmental investment, policy, or regulation as political (in the pejorative sense) as opposed to economic (in the laudatory sense of being economically justified)? This book will argue that BCA is much like a consumer information system. Consumer information neither tells consumers what to do nor tells them what they should want. However, it does tell them which products will perform in selected ways and at what costs. And this information, together with the independently arrived at wants, helps the consumer make intelligent choices.

1
Introduction

Choice is the name of the game. Government sets the size of the public budget and decides which public projects it will invest in and which transfers and regulations it will implement. To do this systematically the government must have a procedure that displays the consequences of the alternatives. This book is an exposition of benefit-cost analysis (BCA), an analytic framework for organizing thoughts, listing the pros and cons of alternatives, and determining values for all relevant factors so that the alternatives can be ranked. A major question illuminated by this text is whether the results of such an analysis can instruct government--in the sense of telling it what it must do to avoid being labeled stupid, corrupt, irrational, and/or inefficient. How and when, we will ask, can the benefit-cost analyst label a particular governmental investment, policy, or regulation as political (in the pejorative sense) as opposed to economic (in the laudatory sense of being economically justified)?

This book will argue that BCA is much like a consumer information system. Consumer information neither tells consumers what to do nor tells them what they should want. However, it does tell them which products will perform in selected ways and at what costs. And this information, together with the independently arrived at wants, helps the consumer make intelligent choices.

Similarly, informed benefit-cost analysts cannot tell government what public projects to fund. They can provide an information system that illuminates and informs the decision-making process and that implements politically chosen objectives and facilitates public monitoring for consistency with objectives, even occasionally encouraging these objectives to be rethought. The adoption of this stance involves a clear understanding of the division of labor between analyst and politician. In this book, we differentiate, in particular, between questions involving political matters, for which the analyst must obtain answers from the political system, and questions that involve technical matters to be solved within the discipline of economics.

The distinction between efficiency and distribution (sometimes referred to as the allocation and distribution branches of economics ) is central to most treatises on public investment analysis.¹ The government can implement the demands of sovereign consumers given the distribution of income and other property rights, or it can redistribute that income via projects which constitute transfers of goods in kind (or in cash grants). Much of the literature of public investment analysis is devoted to separating these effects, but this text will develop a different point of view. Instead of taking income and rights as independently given and to be implemented by BCA in a Pareto-better fashion, we will see income, rights, and the analytic rules of BCA emerge as interacting parts of the same structure. In practical terms, after all, a person's wealth is made up not only of titles to land and stock (factor ownership) but also of rules governing tax incidence and government spending (rules for project evaluation).

According to the potential Pareto-better criterion, a project should be implemented if the gainers can compensate the losers, or, put in terms of property rights, if the buyers of the goods can compensate the owners of the opportunities (inputs) used to produce the goods. One of the ingredients in the buyers' willingness to pay is their ownership of wealth. Essential to the equation is the legitimacy of ownership (rights) that produces the income behind the willingness to pay and the legitimacy of ownership (rights) of the inputs necessary to produce the good in question (Kennedy 1981).

If the present distribution of wealth is accepted, then taxes and public spending only serve to implement the allocative preferences of those with income. But if it is rejected (or continually being renegotiated), the spending of tax receipts is more like a stock dividend by which some of the owners get their share of the nation's net income. In this setting, a tax implements a social dividend to a stockholder to be named later (when the money is spent by the government, which acts as both rights giver and agent for the owner). And to push this account a step further, the appropriation process is a resolution of the conflict over who the stockholders are; that is, who owns the Treasury?

The process becomes hard to follow when the way in which the money is to be spent is part of the political agreement to implement a given group's ownership. Analysis would be simpler if basic ownership never changed or, if changed, was always done prior to analysis by changes in private property or lump-sum tax and direct income transfers (Musgrave 1969). But this is in fact not done, and any counsel to do so would be part of a power play to protect the status quo.

Welfare economics is concerned with evaluating whether any proposed change in government spending or regulation will improve total welfare. But as Hutchison (1964, 165) points out, "excessive claims for welfare economics have fostered the illusion of policies without politics, or that significant policy recommendations can be made without controversial value judgements." Welfare economics can produce policy conclusions only under the following conditions: Citizens must agree on the initial rights of the players. Second, government must be able to achieve any distribution of income and welfare it desires by using costless lump-sum taxation (Tresch 1981, Ch.4). Third, people must have identical preferences and income elasticities that equal unity (i.e., the elasticities are homothetic). These conditions are known as the first-best assumptions.

However, propositions and decision rules appropriate for a first-best world are often incorrect for the real world. This book works out the practical implications of second-best theory, which accepts the fact that government cannot achieve desired distribution with lump-sum taxes and, indeed, that public spending is one instrument used to achieve a desired distribution. A second-best analysis accounts for the following facts: taxes alter resource allocation and consumption; government uses budget constraints; government commandeers production inputs and distributes output free of charge or at prices below opportunity cost; monopoly power exists in the private sector, so some prices do not equal marginal opportunity cost. If the difference between first-best assumptions and second-best reality is not recognized, analysts will inevitably substitute their own implicit distributional judgments for those of elected participants in the political process.

Analysis is seen here as a means for systematically implementing politically chosen objectives on the size and content of production and its distribution. Analysis should show decision makers the specific choices that follow from more generally stated preferences chosen from outside the analytic system. The generality of preferences is, of course, a matter for political choice. The objectives can be stated so precisely that they imply specific projects and leave no role for further analysis. But complete systems analysis is Utopian. Systematic analysis allows observers to determine the objectives being pursued and whether these objectives are pursued consistently or piecemeal over some chosen scope of application. The analysis should allow different groups with different interests to see how the government's proposed budget choices will affect them. The material to follow is, then, built on the premise that good analysis will facilitate widespread, informed public participation in decision making.

Government directs the allocation of resources not only via public spending projects but also by regulation. For example, a redirection of resources is accomplished by a rule requiring industry to install pollution control equipment, or by government paying for the equipment, or by a hospital treating those affected by the pollution. Systems analysis broadens the consideration of alternatives, although there are practical limits to placing all of policy analysis into the BCA framework. For example, a public firm might be created to compete with a private monopoly, and in that sense public projects would be a substitute for antitrust regulation. But it may not be useful to force all of antitrust analysis into the BCA format. So, even though most of the references in this book are to public spending projects, we show that the same principles apply to public regulations and tax spending (e.g., when reduced taxes are used to affect resource use and distribution).

1.1. Demand for Systematic Analysis: Implication for Second Best

In the United States, government BCA reports are common in the areas of water and forest resources, regulatory action (especially relating to health and safety), and some human services. But the demand for systematic analysis should not be overestimated. Although many projects are studied, few are actually chosen according to how they fare under BCA. For example, water resource projects in the United States have been required to undergo a BCA since 1936, but Congress does not necessarily fund those with the highest returns. The analysis is used more to qualify a set of eligible projects rather than to establish the priority of members of the set. Similarly, the World Bank requires a BCA before it will make a loan, but its allocation of funds among countries has no systematic relation to relative rates of return.

There are several possible rationales for this behavior. The least charitable is that politicians do not wish to make the reasons for their choices explicit and subject to examination and public debate. The public itself may opt for self-delusion when faced with tough choices, such as health or safety for one group versus another. Another explanation is that there are many objectives that are not quantifiable and must be more flexibly and qualitatively assessed. Related to this are the information requirements of any analysis. The results of BCA must meet the test of practicality; that is, analysis and information must themselves be subject to BCA. Systems analysis makes political decisions more open and explicit, but some second-best compromises are necessary to deal realistically with information and management costs as well as political pressures.

1.2. Steps in Analysis

The preparation of a report of the costs and benefits of public investments or regulations follows a series of logical steps, and this logic is reflected in the chapter titles of this book.

Theories of Public Investment: The first question is why a proposed investment should be made in the public rather than the private sector. The analyst working for a particular agency with an assigned task may not have to ask this question. Nevertheless, the rationale for public investment supplies the background for any subsequent analysis of particular projects. In Chapter 2 we analyze some of the more commonly expressed justifications for public investment and/or regulation, which frequently are based on particular characteristics of the project or service in question.

Program Information Structure: What does the project produce, and what are the physical units for measurement of its quantity? This question is often difficult to answer and is related to the goals and objectives of the public program. Its answer will facilitate or hinder comparison with other projects and regulatory programs across agencies. These issues, explored in Chapter 3, also provide a convenient context for a discussion of the contemporary practice of program budgeting.

Estimating Project Effects: The next step is to determine if the project's inputs caused a change in the outputs identified above in the program structure. The question is, did the project make the difference? Economists call this step estimating the physical production function, or more commonly known as program evaluation, which is a bit of a misnomer because it has little to do with values or prices. This is a specialized topic, and Chapter 4 will merely indicate the major threats to causal validity and familiarize the reader with alternative quasiexperimental designs.

Valuation of Direct Effects: After we know the product to be produced and the role of the project in its production, we must consider the valuation of any other products produced by the project and any inputs required. Chapters 5, 6, and 7 discuss side effects--what economists usually call "externalities." We argue that there is much confusion in the literature on this concept and that clarity can be obtained by careful attention to distributive and property rights issues. The resolution of these issues requires an explicit political choice among conflicting values of different individuals. An especially important problem, that of assigning values to effects when there is no market available to set them, is addressed in Chapter 5.

Opportunity Cost Adjustments: Even when prices exist, they might not reflect opportunity costs. Adjustments are often necessary because of taxation, monopoly, subsidies, foreign exchange, and labor policies. Projects not only produce goods and services directly used by consumers but they also indirectly affect employment and income of input suppliers and the subsequent processors of the direct output. These are often referred to as multiplier effects. Chapter 6 distinguishes regional gains from gains to the total economy. The use of unemployed resources is given special treatment.

Nonmarginal Projects: Particular attention will be given in Chapter 7 to the effect of project size on product prices. The discussion will be sensitive to how value (political) judgments enter into calculations of what is often asserted to be "net social gain."

Distribution Effects: The point has been made that the analysis and design of projects affect income distribution in many ways, some of which are often masked as net social gains. In Chapter 8, different ways of incorporating an explicit redistribution objective into project selection will be discussed. Systematic treatment of redistribution objectives has not previously been accomplished in the literature or practice. Topics discussed in this connection include tax incidence, pricing, and allocation of overhead costs in multiple-product projects.

Valuation over Time: Projects differ in the pattern and length of flow of benefits and costs. Alternative investment criteria and the role of the discount rate are explored in Chapter 9. Particular attention is given to the problems raised by capital rationing, and which interest groups' opportunity costs and reinvestment possibilities are given preference under real-world situations of disequilibrium.

Uncertainty: Projects also differ with respect to the predictability of outcomes. People do not always agree on the trade-off between different-sized benefits and their variance. One approach to irreducible uncertainty (discussed in Chapter 10) is to design, implement, and manage projects so that adjustments can be made as more information becomes available.

Political Economy of Budgeting: In the final two chapters, we return to the question of the demand for systematic analysis. Information is a resource to political bargainers, and no good study is neutral in its effect on conflict resolution. Resources for analysis have been growing, but use of the results in actual budget formulation is still limited. Reasons for this limitation and implications for the role of the analyst vis-a-vis the politician are explored. A final topic is the relationship of project analysis to budget formulation and national planning.

1.3. Skills of the Applied Economist

Public investment analysis places heavy demands on the skills of the applied economist. No subfield of economics is irrelevent. It is necessary to draw on production economics and theory of the firm to estimate benefits and size effects, industrial organization to adjust for effects of monopoly on prices, location and regional economics to estimate indirect effects on employment, macroeconomics and monetary policy to deal with discount rates, on international trade to take account of tariffs and foreign exchange policies, and so forth. Nor should we ignore the techniques of measurement from statistics for experimental design and contributions from other disciplines to facilitate implementation and management. No single book can address the necessary skills in detail. But it is possible to indicate which concepts and experience from these fields are particularly relevant for public investment analysis and to provide a framework for interpretation of information produced by other specialists when it is useful to apply such information to the choice of public investments.

This book discusses what BCA is and what it can do, provides guidelines for its proper use and interpretation. Some will see it as a critique of BCA, but that is a misunderstanding of our purposes. The book critiques not BCA itself, but only its presentations and practices that fail to be systematic, that do not recognize the costs of analysis, that ignore second-best reality, that misconceive or misapply important concepts, or that allow the analyst to usurp the job of politicians. It is a call to do BCA better than we have been doing it.

Notes

1. Treatises on BCA that provide background for this book but are not elsewhere cited are Aboucher (1985), Austin (1981), Cohn (1972), DeGarmo (1967), Dorfman (1965), Gittinger (1982), Howe (1971), Hufschmidt et al...

Table of contents

• Cover
• Half Title
• Title
• Copyright
• Dedication
• Contents
• List of Tables and Figures
• Foreword
• Acknowledgments
• 1 Introduction
• 2 Public Investment: Rationales and Objectives
• 3 Program Information Structure
• 4 Estimating Project Effects
• 5 Valuation of Direct Effects
• 6 Opportunity Cost Adjustments
• 7 Valuation of Nonmarginal Projects
• 8 Distribution Effects
• 9 Valuation over Time and Selection Criteria
• 10 Uncertainty
• 11 The Political Economy of Budgeting
• 12 Conclusion: Systematic Analysis in Perspective
• References and Bibliography
• Author Index
• Subject Index