1
Introduction

BACKGROUND AND OVERVIEW

Gross domestic product is the nation’s most influential measure of overall economic activity and growth. The national income and product accounts (NIPAs) that underlie gross domestic product, together with such other key economic data as price and employment statistics, are widely used as indicators of how well the nation is doing.

Since their earliest construction for the United States by Simon Kuznets in the 1930s (Kuznets, 1934), concerns have been voiced that the national income and product accounts are incomplete. The venerable NIPAs meet rigorous standards and enjoy broad acceptance among data users interested in tracking economic activity. These accounts are, however, primarily market-based and, by design, shed little light on production in the home or in other nonmarket situations. Furthermore, even where activity is organized in markets, important aspects of that activity may be omitted from the NIPAs. In some cases, unpaid time inputs and associated outputs are integral to production processes but, because no market transaction is associated with their provision, they are not reflected in the accounts. One illustration is provided by estimates (LaPlante et al., 2002) suggesting that the value of in-home long-term care services provided by family and friends is greater than the value of similar market-provided services.

In other areas, the output resulting from market-based production may be incorrectly characterized or valued. There is wide agreement, for example, that the output of the education sector properly should be considered investment



The National Academies | 500 Fifth St. N.W. | Washington, D.C. 20001
Copyright © National Academy of Sciences. All rights reserved.
Terms of Use and Privacy Statement



Below are the first 10 and last 10 pages of uncorrected machine-read text (when available) of this chapter, followed by the top 30 algorithmically extracted key phrases from the chapter as a whole.
Intended to provide our own search engines and external engines with highly rich, chapter-representative searchable text on the opening pages of each chapter. Because it is UNCORRECTED material, please consider the following text as a useful but insufficient proxy for the authoritative book pages.

Do not use for reproduction, copying, pasting, or reading; exclusively for search engines.

OCR for page 9
Beyond the Market: Designing Nonmarket Accounts for the United States 1 Introduction BACKGROUND AND OVERVIEW Gross domestic product is the nation’s most influential measure of overall economic activity and growth. The national income and product accounts (NIPAs) that underlie gross domestic product, together with such other key economic data as price and employment statistics, are widely used as indicators of how well the nation is doing. Since their earliest construction for the United States by Simon Kuznets in the 1930s (Kuznets, 1934), concerns have been voiced that the national income and product accounts are incomplete. The venerable NIPAs meet rigorous standards and enjoy broad acceptance among data users interested in tracking economic activity. These accounts are, however, primarily market-based and, by design, shed little light on production in the home or in other nonmarket situations. Furthermore, even where activity is organized in markets, important aspects of that activity may be omitted from the NIPAs. In some cases, unpaid time inputs and associated outputs are integral to production processes but, because no market transaction is associated with their provision, they are not reflected in the accounts. One illustration is provided by estimates (LaPlante et al., 2002) suggesting that the value of in-home long-term care services provided by family and friends is greater than the value of similar market-provided services. In other areas, the output resulting from market-based production may be incorrectly characterized or valued. There is wide agreement, for example, that the output of the education sector properly should be considered investment

OCR for page 9
Beyond the Market: Designing Nonmarket Accounts for the United States rather than consumption, and that its value should be assessed in terms of the returns on that investment rather than the cost of the inputs used in its production. The conventional accounts do not include changes in the asset value of human capital production associated with education, health care, and other personal investment activities. Available estimates are rough, but suggest that the value of the human capital stock may be as large as that of the physical capital stock (see Kendrick, 1967, and, for a more recent discussion in the context of analyzing economic growth, Mankiw et al., 1992). Although the importance of nonmarket—but productive—endeavors has long been recognized, few attempts have been made to provide systematic information about even the most quantitatively significant of them. Economic accounting need not, and should not, extend to all nonmarket activities, but there are certain areas in which nonmarket accounts, designed to supplement the NIPAs, could make particularly important contributions. We stress the potential value of new methods of accounting for home and volunteer production efforts, education, health, and environmental improvement or pollution. Given that limited coverage of the NIPAs has long been recognized, why is this report needed at this time? The existence of economically valued nonmarket inputs and outputs is already widely recognized. Examples are easy to find: the higher value of a house sold after improvements are made by the homeowner; the meals that a nonprofit soup kitchen serves to the homeless; or the increase in the productive capacity of the economy attributable to extensions of working life resulting from modern diabetes treatments. But information about these productive activities is not systematically compiled and routinely updated as are the NIPAs and other market-based accounts. The state of nonmarket accounts today resembles the situation for market-based accounting in the 1920s and 1930s before the creation of the NIPAs—new data were becoming available, but they were not organized and published in a systematic accounting framework. Advances in data collection and in economic analysis of nonmarket activities merit the review that appears in this volume. One event alone—the development and recent publication of the American Time Use Survey—justifies a new round of thinking about nonmarket accounting issues. In this report we hope to encourage social scientists to pursue the analysis of nonmarket activities and the development of corresponding data collection and accounting systems. We also point out new ideas and new data sources that have improved the prospects for progress. The Panel to Study the Design of Nonmarket Accounts was charged with evaluating current approaches, determining priorities for areas of coverage, examining data requirements, and suggesting further research on nonmarket accounting. The panel’s charge includes four specific tasks: to review efforts to develop nonmarket accounts by government agencies, as well as by private organizations and scholars, including theoretical as

OCR for page 9
Beyond the Market: Designing Nonmarket Accounts for the United States well as empirical studies and actual implementation of nonmarket accounting frameworks; to make specific recommendations on the framework and sectors for developing nonmarket accounts and determine a set of priorities with respect to developing or phasing nonmarket accounts; to examine and make recommendations with respect to key data that are needed to develop nonmarket accounts, considering such efforts as the Bureau of Labor Statistics’s (BLS) time-use survey; and to investigate and make recommendations for research related to nonmarket accounts in the areas of statistics, economics, psychology, survey research, and other disciplines. The panel worked over a period of 2 1/2 years to fulfill this charge. Our recommendations aim to provide practical guidance for constructing nonmarket accounts. Over the course of its deliberations, the panel found that the conceptually ideal approach often does not align with what can realistically be expected in practice. The panel has tried to be clear about when its recommendations reflect compromise required by practical constraints, such as data availability or measurement feasibility. The rest of this chapter discusses the motivation for this study, priorities for nonmarket accounting, and issues relating to account scope and conceptual framework. Chapter 2 identifies and describes accounting and data issues—particularly issues related to the measurement of time use—that are key to both market and nonmarket accounts. Chapters 3 through 8 provide detailed analyses of the individual areas for which the panel has evaluated development of nonmarket satellite accounts. The chapters on specific satellite accounts each begin with a discussion of the area, considering its importance and the role that nonmarket activity plays in it. As noted above, in many of the areas the panel has chosen to explore, market activity coexists with nonmarket activity. Students pay tuition to attend universities (a market transaction), but there is no market transaction that captures directly the value of the time they devote to this endeavor. Likewise, nonprofit organizations may hire employees (a market transaction), but also may use unpaid volunteers, the value of whose time and output produced is nowhere reflected. Central to each of the chapters is a discussion of the conceptual issues related to the measurement of nonmarket activity in the area it covers. These issues lead to consideration of how inputs—both market and nonmarket—and resulting nonmarket outputs might be quantified. The transformation of inputs into outputs presupposes a production technology and, in a number of the chapters, this technology is also a subject of discussion. Another matter that figures prominently throughout the report is how to attach values to specified quantities of nonmarket inputs and outputs. The panel’s approach, in essence, is to seek prices that can be assigned to the identified

OCR for page 9
Beyond the Market: Designing Nonmarket Accounts for the United States quantities. In some cases, satisfactory price measures can be found; in other cases, available information may be unsatisfactory, either because it is conceptually deficient or because it is insufficiently precise to be useful in the implementation of an accounting scheme. Each chapter devotes substantial attention to the question of whether and how satisfactory price measures might be developed. In many cases, researchers have already developed input or output valuations of the sort the panel contemplates. On the basis of these topic-by-topic evaluations, the panel makes recommendations regarding areas to be viewed as high priorities for the development of nonmarket accounts and areas to be viewed as lower priorities. Finally, for those areas in which the panel believes further work holds promise, the individual chapters include recommendations about ideal data for the intended purpose and steps that might be taken toward the production of such ideal—or at least better—data for the United States. MOTIVATION Extending the nation’s accounting systems to better incorporate nonmarket production promises substantial benefits to policy makers and researchers. One objective of improved nonmarket accounting is to support alternative aggregate measures of economic performance. Nonmarket accounts would enhance the ability of researchers (and statistical agencies) to produce augmented gross domestic product (GDP) measures that reflect a broader array of outputs. Data from nonmarket accounts could provide the information required to construct appropriate price deflators needed for augmented real GDP calculations and productivity measurement. The inherent limitation of the NIPAs—that they fail to consider the full array of the economy’s productive inputs and valued outputs—might not be especially important if market and nonmarket activities trended similarly, but this cannot be assumed. To take one often-cited example, failing to account for the output produced within households may lead to misleading comparisons of economy-wide production, as conventionally measured. According to BLS figures, the female labor force participation rate in the United States has grown substantially, from about 34 percent in 1950 to almost 60 percent in 2000. To the extent that the entry of women into paid employment has reduced effort devoted to household production, the long-term trend in output as measured by GDP may exaggerate the true growth in national output (Landefeld and McCulla, 2000). Similarly, the relatively smaller portion of total output attributable to home production in the United States as compared to many developing countries may exaggerate its national output relative to theirs. Perhaps less well recognized are potential problems with the measurement of national output over the business cycle. If people who lose their jobs during cyclical downturns take advantage of their absence from paid employment to

OCR for page 9
Beyond the Market: Designing Nonmarket Accounts for the United States increase the effort they devote to home production, the short-term decline in national output may be dampened relative to that measured by GDP (Greenwood et al., 1995). Knowing more about the level and distribution of nonmarket activity could be important for other purposes. Such information could, for example, change perceptions of the extent of economic inequality among U.S. households and how that has changed over time. This, in turn, could affect where welfare and poverty lines are drawn (Michael, 1996). Accounting for nonmarket activity also would contribute to our understanding of the sources of economic growth. Researchers studying this topic have long found it necessary to supplement data from the national accounts with external estimates of the contributions of research and development, investments in human capital, and the services of the natural environment (Mankiw et al., 1992). Economists and historians have shown that, over the last few centuries, factors such as technical change, scientific inventions, and discoveries in medicine—many of which have a nonmarket dimension—have accounted for a very large portion of the growth in living standards (Fogel, 1986; Nordhaus, 2003; Cutler, 2004). Historical trends reveal the reality that neither economic production nor contributions to social welfare take place exclusively within the market’s border, but extend to many nonmarket activities. Nonmarket accounting also would illuminate the processes whereby inputs are transformed into outputs in particular sectors. Consider, for example the production of health. Currently constructed health expenditure accounts track market payments but do not identify the outputs in a way that is very useful for measuring price change or productivity. In contrast, a health account would relate health improvements—the real “good” that is produced—to medical treatments, as well as to a wide range of other inputs, including diet, the environment, exercise, and research and development. By most measures, improvements in health have outpaced the increase in spending on medical care. Since medical care interacts with this broad range of interrelated factors, however, we do not know with any certainty the productivity of resources directed into health care (Cutler and Richardson, 1997; Cutler, 2004). Optimally, expenditures and outcomes would be tracked so that changes in people’s health could be linked to different actions; in turn, this information could support better management of expenditures (both private and public) to achieve desired outcomes. To take another example, education accounts might be designed to relate improvements in skill capital—the output—to the various inputs of the educational process. As in the health case, schooling is characterized by a mix of market and nonmarket inputs and outputs. In education, the value of time students spend in school is likely to be of the same order of magnitude as expenditures on marketed inputs. The 2003 Statistical Abstract shows that, in 2000, school expenditures on primary and secondary education amounted to approximately $400 billion and that just over 47 million students were enrolled in primary and secondary schools. Assuming 180 days at 6 hours a day, plus 1 hour of commuting

OCR for page 9
Beyond the Market: Designing Nonmarket Accounts for the United States time and 2 hours of homework per student, students in these grades devoted more than 75 billion hours to their education. If students’ time were valued at $5.35 per hour (purely for illustrative purposes), the value of unpaid student time would be approximately as large as the expenditures measured in the conventional accounts. This report identifies and discusses many of the key and sometimes controversial issues, such as those hinted at above, that underlie nonmarket accounting. One goal of the report is simply to remind readers of the major omissions built into our system of economic measurement. In so doing, the panel hopes to encourage contributions by social scientists to improve the measurement of nonmarket activity and to point out new ideas and new data sources that have improved the prospects for progress. Time is the dominant input to nonmarket production, and the lack of good measures of how people spend their time has seriously handicapped work in this area. The panel is optimistic that the newly developed American Time Use Survey, produced by the BLS, will spur new work to develop informative nonmarket accounts. PRIORITIES, SCOPE, AND OBJECTIVES Scope of Coverage in the NIPAs Traditional national income and product accounts include primarily the output of marketed goods and services—that is, goods and services that are bought and sold in market transactions. An international standard, the System of National Accounts (SNA) (Commission of the European Communities et al., 1994), provides a set of guidelines, accounting rules, classifications, and definitions designed to, among other things, facilitate international comparison of market-based measures. The boundaries to national economic accounting employed in the U.S. NIPAs largely conform to the standards outlined by the SNA.1 At the time of their development, there were good reasons for the U.S. national income and product accounts to focus on market production. In assessing the disruptions attributable to the World Wars and the Great Depression, policy makers understandably had a primary interest in how these events had affected market activity. In addition, as recognized by Kuznets and others, reliable measurement of nonmarket production would have been extremely difficult. Nonetheless, there have long been influential economists who argued that economic accounting should extend beyond the market. One of the great founders of modern economics, A.C. Pigou, wrote that national accounts should include elements that reflect economic welfare and that can “be brought directly or indi- 1   Additional background information on the structure and scope of the NIPAs is provided in Chapter 2.

OCR for page 9
Beyond the Market: Designing Nonmarket Accounts for the United States rectly into relation with the measuring rod of money” (Pigou, 1920, p.11). He emphasized that the word “can” might mean anything from “can easily” to “can with mild straining” to “can with violent straining.” National accounting practices in most countries lean far more toward those elements that “can easily” be measured in money terms than those that can be measured only with “violent straining.” Though the national accounts produced by the Bureau of Economic Analysis (BEA) generally exclude activities that do not involve a market transaction or produce a marketed output, there are exceptions—most notably the imputation for the rental value of owner-occupied housing. This imputation is based on assumptions that are approximately as crude as those for, say, valuing the time spent cleaning a house at the price a cleaning service would charge. One reason for making an imputation for the value of owner-occupied housing is to ensure that the accounts are invariant to trends in home ownership (which, incidentally, has increased significantly in the past half-century; see U.S. Census Bureau, 2004). Other imputations for nonpriced, nonmarketed items in the NIPAs include those for wages and salaries paid in kind, food and fuel consumed on farms, and the services provided by banks, insurance companies, and other financial intermediaries that are not reflected in explicit service charges. The imputations for banking services are somewhat unique. In banking, there are observable market transactions that provide an estimate of the nominal value of banking output. Imputations are necessary, however, to allocate the nominal value of unpriced services between borrowers and depositors (see Fixler et al., 2003, for a more complete discussion). We are not aware of any body of analysis that would set the outer bounds of market accounting at precisely the point chosen by the SNA or by the BEA. One key characteristic of the nonmarket items that are covered in conventional accounting systems is that their consumption is closely related to the sales and purchases of marketed goods and services, making estimation reasonably straightforward if not always precise. For example, the rental value of owner-occupied housing is imputed from observed rents for similar housing. Similar, straightforward imputations could serve as a basis for price estimation for some nonmarket items excluded from coverage in the national accounts but, for many others, such a close comparison will not be possible. While the national accounts exclude the output resulting from many areas of nonmarket activity, information relevant to these activities often is included. In many cases, purchases of inputs that contribute to nonmarket outputs are treated as expenditures for final demand. Spending on food, cleaning supplies, and laundry detergent is counted as part of personal consumption. In an accounting system that considered meals, house cleaning, and laundry services as elements of consumption, these kinds of expenditures would appear as intermediate inputs to the production of consumption goods. Similarly, government expenditures on education are included in the accounts. These would become intermediate prod-

OCR for page 9
Beyond the Market: Designing Nonmarket Accounts for the United States ucts if the human capital produced through educational activities were treated as the relevant output. And many of the inputs to medical care are included, but the accounts contain relatively little separate information about the value of the services provided or the health capital formed. For the cases we have elected to explore, the conventional accounts generally do not reflect the full range of inputs used in the production of the output of interest. And in no case is the value of the resulting output, whether goods and services produced for current consumption or the creation of a productive asset, measured fully and independently of the value of the inputs used in its production. Independent measurement of output entails estimating prices and quantities of that output, as opposed to building up valuations from cost side factors. Satellite Accounts While acknowledging the limitations in their coverage, the panel is not proposing that the scope of the existing NIPAs be expanded to encompass currently omitted nonmarket activities. The existing core accounts have the important virtues of consistency over time, hard-won comparability across countries, and, aside from a limited number of agreed-upon imputations deemed necessary for certain nonmarket activities, solid grounding in observed market transactions; we would argue strongly that they should be preserved. Likewise, the statistical agencies should not develop a comprehensive measure of total economic output as a substitute for the existing GDP measure. Even at a conceptual level, there is no consensus about what an appropriate replacement for the existing measure of GDP might be. We expect that data from the satellite accounts we advocate will allow researchers to develop a variety of expanded GDP measures. It would be premature, however, to endorse a specific measure—indeed, it seems unlikely that a consensus around any one measure ever will develop. Instead, we recommend the development of satellite accounts to report on selected activities not included in the conventional accounts. Satellite accounts can link to the existing economic accounts as appropriate, but also expand into areas that the NIPAs do not cover. Furthermore, satellite accounts can be developed even where standards of accuracy and data quality are not up to the level of the NIPAs, without compromising the conceptual basis or technical integrity of the conventional accounts. Similarly, where no consensus yet exists regarding the best way to measure a particular area of nonmarket activity, satellite accounts allow for the flexibility to experiment with alternative methodologies that might go against convention. The goal is to extend the accounting of the nation’s productive inputs and outputs, thereby providing a framework for examining the production functions of some difficult-to-measure nonmarket activities. The idea of satellite accounts is not a new one. The BEA has long conducted research on topics beyond the scope of the conventional accounts. A representa-

OCR for page 9
Beyond the Market: Designing Nonmarket Accounts for the United States tive BEA definition of satellite accounts is as follows (Bureau of Economic Analysis, 1994a, p. 41): [S]atellite accounts are frameworks designed to expand the analytical capacity of the economic accounts without overburdening them with detail or interfering with their general purpose orientation. Satellite accounts, which are meant to supplement, rather than replace, the existing accounts, organize information in an internally consistent way that suits the particular analytical focus at hand, while maintaining links to the existing accounts. In their most flexible application, they may use definitions and classifications that differ from those in the existing accounts…. In addition, satellite accounts typically add detail or other information, including nonmonetary information, about a particular aspect of the economy. The System of National Accounts offers a similar description (Commission of the European Communities et al., 1994, pp. 45, 489): Satellite accounts provide a framework linked to the central accounts and which enables attention to be focused on a certain field or aspect of economic and social life in the context of national accounts; common examples are satellite accounts for the environment, or tourism, or unpaid household work…. Satellite accounts or systems generally stress the need to expand the analytical capacity of national accounting for selected areas of social concern in a flexible manner, without overburdening or disrupting the central system. The panel views the accounting frameworks described in this report as harmonious with these definitions and concludes that, for a number of industries and sectors with significant nonmarket components, satellite accounts can be developed that will generate meaningful and useful data to inform policy and to advance research. Measurement Objectives Although they are widely used in comparisons intended to shed light on trends and relative levels of national well being, the national accounts are not designed to measure the welfare of the population or the contributions to welfare associated with particular sectors. Rather, the NIPAs provide a measure of the nation’s market output (and income). Output obviously correlates to some extent with consumers’ welfare—indeed, there is a well-developed literature in cost-benefit analysis showing that, under certain conditions, changes in the former may approximate changes in the latter2—but national income is not the same as national welfare. 2   Dreze and Stern (1987) show that, under certain conditions, the net social benefits of a “project” (that is, some change in output) are equal to the change in GDP measured at market prices.

OCR for page 9
Beyond the Market: Designing Nonmarket Accounts for the United States For some potential policy or research applications, an account based on a traditional (price times quantity) valuation may be most useful; for others, an account oriented toward “social welfare valuation” might be conceptually preferable. If resources were unlimited, multiple accounts would be developed to serve all of the various purposes that arise. But this is not the case, and development of separate accounts to meet all possible needs is not a realistic option. Certain nonmarket activities—much household work, for example—result in the production of goods and services very similar to those sold in markets. These goods and services presumably are welfare enhancing as well. A logical first step in developing a set of satellite accounts is to quantify and value these “near-market” activities. If coverage of the nonmarket accounts were limited to items for which there are close market substitutes, however, many factors that affect well-being would be excluded. Broader accounting approaches also can be envisioned. One option would be to include nonmarket activity if it is possible to conceive of a procedure that would reveal the monetary value that people place on it. This approach would lead to a more expansive set of nonmarket accounts. Under this definition, an activity such as leisure might be considered in scope. Even such a seemingly noneconomic outcome as “family disintegration” could enter the accounting structure if a method could be developed that revealed what society would be willing to pay to prevent its occurrence. Our ability to conceive of such procedures has increased in the years since Becker’s seminal work on marriage and the family (e.g., Becker 1973, 1974), and the boundary between what is and is not a proper candidate for inclusion, as determined by the rule that an entry must receive a monetary valuation, undoubtedly will change over time. As noted above, one of the important potential uses of satellite accounts for nonmarket activities is to support broader aggregate measures of national output. A leading example of an alternative aggregate output measure is James Tobin and William Nordhaus’s measure of economic welfare (Nordhaus and Tobin, 1972). Although various alternatives exist, there is no professional consensus around any single expanded measure, and we do not believe that such a measurement objective provides the organizational principles for a system of satellite accounts. A more realistic approach to developing augmented measures of output would begin with a set of largely independent accounts covering a wide range of nonmarket activities. While acknowledging that different users require different kinds of data and that new methods may be developed for valuing outcomes previously considered noneconomic, it is the panel’s view that resources should initially be directed toward developing a more complete accounting of output rather than attempting to measure happiness or well-being. This position can be defended on practical measurement, as well as conceptual, grounds. Accounting for nonmarket activities and results that parallel those already represented in the national accounts would enable the developers of the new accounts to build more directly on past

OCR for page 9
Beyond the Market: Designing Nonmarket Accounts for the United States experience. Such things as household production of cleaning services or participation in educational activities have market analogues and are (relative to, say, leisure) more closely aligned with what is currently viewed as output in the accounts, meaning that the hurdles to clear in devising sensible measurements are lower. Because improving output (and corresponding input) measures is a prerequisite to any vision for an expanded set of accounts, this is where the panel focused its energies. Even if the accounting objective is limited primarily to measurement of output, there is still a formidable amount of work to be done. Nonmarket Accounting Priorities In defining the boundary of what should be counted when developing measures of nonmarket output, some have advocated the application of Margaret Reid’s (1934) third-party criterion: Is the output something that a person could have hired someone else to produce? This criterion seems appropriate for certain areas—for example, a household production account could be designed to include such things as meals, clothing services, shelter services, and a basic component of child care, but to exclude fertility, studying, and exercise. For other areas of nonmarket activity, such as education and health, the third-party criterion is clearly inappropriate: there is no replacement for self involvement in the activities required to enhance a person’s cognitive skills or improve his or her health—activities that nonetheless produce valuable but nonmarketable capital outputs. In considering how broadly nonmarket accounting work should be cast, one must ask what standards of accuracy and reliability should be applied to measures of inputs and outputs. We do not have a full answer to this question. Traditionally, the statistical agencies responsible for economic accounting—the Bureau of Economic Analysis of the Department of Commerce and the Bureau of Labor Statistics of the Department of Labor—set high standards of accuracy. At least initially, nonmarket satellite accounts presumably will have to be constructed under a more forgiving standard, but there undoubtedly will be debate about how imprecise a number can be and still remain useful. In market accounting, outside researchers often produce accounts from ingredients supplied by the statistical agencies—for example, a consulting firm publishes monthly GDP from data supplied by the BEA, though the BEA does not consider the data to be sufficiently reliable in all sectors to produce an official monthly measure. We anticipate similar private accounting efforts with nonmarket data. As detailed in this volume, potentially valuable areas of nonmarket accounting are at different levels of development with respect to measurement concepts and available data. For that reason, the panel favors a staged approach. Work should begin in areas where potential gains are high and costs (mainly in terms of new data collection needs) are low. For example, we believe that accounting for certain types of home production should begin soon, based on data from the American Time Use Survey. In contrast, neither the conceptual understanding

OCR for page 9
Beyond the Market: Designing Nonmarket Accounts for the United States in the pattern of time use reflect changes in population mix or some other cause. The demographic data to support such an effort are, for the most part, already available, largely from the Census Bureau but in some cases from the Bureau of Labor Statistics, the National Center for Health Statistics, and other agencies. A determined researcher could compile these data from existing sources. But it would be very helpful if the information were assembled in a single place, adjusted to be consistent over time. The demographic database would not itself be a satellite to the existing economic accounts, but it would assist in the development and use of those accounts. The ease with which the quantity of nonmarket outputs can be measured varies widely. Relatively good data are available, for example, on the educational attainment of the working-age population. These data provide a starting point for quantifying the output of the educational sector. Changes in mortality and morbidity are similarly well documented and could provide a basis for quantifying changes in the health status of the population, particularly if combined with demographic data tracking changes in population mix. In other cases, considerable creativity may be required to measure the quantities of nonmarket outputs, and doing an adequate job ultimately may require the collection of new data. Tracking air quality would require better measures of the pollutants to which the public is exposed and of the costs they impose. Tracking the output of the household sector would require data on such things as meals prepared or loads of laundry washed and dried. But, at least in principle, it is possible to see how this task might be approached. To elaborate on the laundry example, the accounts would, on the input side, tally the number of hours devoted to laundry and the wage of a domestic employee or the opportunity cost or predicted market wage of the person doing the laundry (these methods are discussed below). The remaining inputs would be the capital services of the household’s washing machine and dryer, together with the necessary materials, electricity, water, and detergent. These inputs would be reported in quantity and price terms. On the output side, the accounts would report the amount of laundry done and its price, estimated on the basis of what it would have cost to have the laundry done commercially. More thought needs to be given to what productivity measures mean when they are based on market substitute valuations. In the absence of direct measures of the output of nonmarket activities, one might impute them from observed market activities but, in such cases, productivity measures for nonmarket activities may simply recover the imputation scheme. Assigning Prices Anyone contemplating the development of nonmarket accounts must decide how best to value inputs and outputs in the various accounts, given the absence of prices. Valuation typically involves finding market analogues for the nonmarket

OCR for page 9
Beyond the Market: Designing Nonmarket Accounts for the United States inputs or outputs in question. Given the distance from the market of some utility-generating activities, however, this approach is not always feasible. How to measure the value of unpaid time devoted to nonmarket production is the central input valuation issue. One approach has been to value this time at market substitute prices—the wage that would be paid to a person hired to perform the task in question. Another approach is to value time at the opportunity cost of the person performing the nonmarket activity. The two approaches may give quite different answers if higher-wage individuals devote time to tasks for which the market wage is relatively low. At first blush, it may seem puzzling why anyone would choose to perform activities that compensate—in the form either of wages paid or value of nonmarket output produced—at a rate below the wage that could be earned in market employment. Further reflection makes clear that such decisions may be entirely rational. Economic theory conceives of people making marginal choices about their allocation of time to different activities. At the point of maximum satisfaction, the marginal earnings from an hour spent working for pay should be equated to the combined value of the output plus the value of the extra enjoyment intrinsic to an hour spent in nonmarket activity in comparison with market work. A key point in this theory is the following: Even at the same moment, an individual may require different amounts of compensation, in the form either of dollars or the value of output produced, for engaging in different activities because different activities may yield different amounts of intrinsic satisfaction. A lawyer who commands $200 per hour from corporate clients may do work at $50 per hour for a charity. In this case, providing the work to the charity has an offsetting personal benefit (enjoyment) absent from working for a corporation. By the same principle, highly paid individuals may choose to prepare meals at home that could have been purchased in the market at a cost far below the wages the individual could have earned by working for pay instead of cooking. The recreation component of cooking means that the marginal value of the cooking performed is lower than the wage, if there is no similar recreational value in the person’s job. In both cases—the lawyer performing work for a charity or the highly compensated person cooking meals at home—use of an unadjusted opportunity cost wage to value the time spent in activities the individual finds enjoyable would overstate the cost of inputs to nonmarket activities and understate their productivity. We turn to economic theory for guidance in attaching an appropriate replacement cost value to time spent in nonmarket activities that someone else could have been hired to perform. The simplest case is that in which market and nonmarket production employ the same underlying technology. Consider a production function that relates productive inputs—labor (L) and capital (K)—to an output (Q): (1.1) Q = f(biL, K).

OCR for page 9
Beyond the Market: Designing Nonmarket Accounts for the United States Quantitatively, people’s time (L) is the most important unmeasured input in nonmarket production. In the nonmarket context, an unpaid labor input must often be compared to a market replacement. People performing nonmarket tasks may be less skilled and work less hard on the task, on average, than people doing similar work in the market for pay. In the production function for nonmarket output, bi is a measure of the relative efficiency of individual i in the nonmarket task as compared to the efficiency of market labor. If our speculation is correct, bi will be a number between zero and one, though the precise value likely will vary across individuals. An appropriate procedure for cases in which a family member performs work at home that could have been performed by someone hired in the market is to count the family member’s hours as measured and to value those hours at a rate equal to the efficiency factor, bi, multiplied by the market wage for someone performing the type of work in question. Thus, if a homeowner chooses to reroof the house and, using the same materials and tools, takes twice as long to do so as it would have taken a professional roofer making $30 per hour, we would record the time the homeowner spent on the task and value that time at $15 per hour. Further, we would use the same $15-per-hour valuation whether the homeowner earns $100 or $10 in his or her own market job. In the case of the $100-per-hour person, we are implicitly assigning the roofing task an amenity value of $85 per hour, while in the case of the $10-per-hour person, we implicitly would be assigning it a disamenity value of $5 per hour. With respect to a task that cannot be given to another person—such as studying or exercising—there is little option but to use an opportunity cost valuation of the time expended. For people who work in the market, the market wage rate commonly is used as a measure of opportunity cost; for those who do not, a market wage must be imputed. Having said this, it should be recognized that people perform activities at home at different times of the day, week, and year than those they perform in the market. The marginal wage rate that a person might obtain in the market at the times many household activities are performed is unlikely to equal the average wage rate the person receives for his market activities. For example, the relevant hours for home production may be the 2-3 hours at home after putting in a full workday at a market job. It is not clear that the market value of those hours is the same as the individual’s marginal or average hourly rate. Furthermore, many people in the labor market are unable to sell additional hours at their wage rate. Wage workers have little control over their hours; salaried workers, by virtue of the compensation structure of their jobs, are typically not paid at all for extra hours of work (Nordhaus, 2004, p. 18). For these reasons, the market wage rate may be a poor measure of the opportunity cost of an hour of the person’s time. Recognizing this problem is easy; doing anything about it is difficult. Certainly, more research into the relationship between opportunity cost and market wages would be helpful here. More fundamentally, even beyond the difficulty of measuring the opportunity cost associated with time devoted to nonmarket production, one would want

OCR for page 9
Beyond the Market: Designing Nonmarket Accounts for the United States in principle to adjust that opportunity cost to reflect differences between the amenity of work and the nonmarket activity in question to estimate the true cost of time devoted to the latter. Making this adjustment would require being able to quantify the value of enjoyment that individuals derive from different activities. Provided that individuals are able to equalize the marginal value of time across all of the activities in which they engage, the adjusted opportunity cost measure we advocate for valuing tasks that someone else cannot be hired to perform is conceptually equivalent to the adjusted replacement cost measure that seems most appropriate for valuing time devoted to nonmarket activities satisfying Reid’s third-party criterion. To see why this is so, return to the roofing example discussed above. Let WS equal the hourly wage of a roofing specialist (replacement wage) and WOi the foregone hourly market earnings (opportunity wage) of a homeowner taking on a roofing task. The productivity-adjusted replacement cost of time the homeowner devotes to roofing would be (1.2) where, as before, bi is a parameter between zero and one that reflects the relative productivity of the homeowner as compared to the professional in the activity under consideration. The amenity-adjusted opportunity cost of time the homeowner devotes to the activity—in this case, roofing—rather than to paid employment would be (1.3) where ei represents the value of the extra enjoyment that individual i receives from spending an hour in the nonmarket activity rather than in paid employment. But for any given activity, an optimizing individual will seek to equate the value of the marginal product associated with time devoted to an activity (biWS) to the net opportunity cost of that time (WOi − ei). This discussion implies that there are two alternative but potentially equivalent routes to arriving at a valuation for time devoted to nonmarket activity. If we know the relative efficiency of the family member, the effective wage can be calculated as the professional’s wage multiplied by that relative efficiency. We could estimate the amenity value the person attaches to the nonmarket work by subtracting the effective wage from the individual’s market value of time. Or, if we know the amenity value of the activity to the individual, we can subtract it from the family member’s market value of time to estimate the effective wage. As a byproduct, at least in the case of activities that could have been performed by a third party so that an observable wage exists for market providers, we could estimate the person’s efficiency relative to a professional in the same line of work

OCR for page 9
Beyond the Market: Designing Nonmarket Accounts for the United States as the ratio of the effective wage to the professional’s wage. Where it is an option, it seems likely that adjusted replacement cost will be easier to estimate than adjusted opportunity cost, but conceptually the two approaches should lead to much the same result. Additional factors, such as taxes, complicate the relationship between opportunity and replacement costs. All else the same, income taxes reduce the opportunity cost of a person’s own time devoted to nonmarket production, but not the replacement cost of hiring someone else to perform the task. This suggests the use of pre-tax wage rates in replacement cost calculations but after-tax wage rates as the basis for opportunity cost calculations. Still, the basic concepts to be applied are reasonably straightforward. Transaction costs may be an important consideration in certain cases. A homeowner who hires a roofer does not entirely escape the need to spend time on the project—time must still be devoted to finding and negotiating with the roofer and to overseeing the roofer’s work. Incurring these costs may be worthwhile for a large job but not for a small job. Market service providers also may incur transactions costs that make it unattractive for them to take on small jobs. It would be very difficult, for example, to find anyone willing to come to your house for 10 minutes each morning just to prepare a pot of coffee! We should also acknowledge that things are more complicated when different technology is employed for market and nonmarket production. Continuing with the example above, a professional roofer might find it worthwhile to purchase specialized tools and equipment that a homeowner doing the same job would not possess. Even if skill levels and intensity of work effort were similar, this might be another reason to expect that a professional roofer could complete a job more quickly than the average homeowner. The homeowner’s decision between hiring a roofer and doing the work personally will depend on the full relative costs, not just on the time requirements, associated with the two options. In many if not most cases, however, market and nonmarket producers perform their tasks in similar ways—doing yard work, ironing shirts, and making a salad for dinner come to mind as examples. In such cases, the only significant difference between market and nonmarket production is the source of the labor supplied and, even in cases where other factors come into play, it is likely to be the most important difference. Recommendation 1.4: A replacement cost measure, adjusted for differences in skill and effort between nonmarket and market providers, should be adopted for valuing time inputs devoted to nonmarket production in cases in which someone else could have been hired to perform the work in question. Where this is not the case, an opportunity-cost-based measure, ideally adjusted to account for the intrinsic enjoyment associated with the activity in question, should be used for valuing time devoted to nonmarket production activities.

OCR for page 9
Beyond the Market: Designing Nonmarket Accounts for the United States We believe that the valuation of nonmarket outputs should, where possible, follow the principle of treating nonmarket goods and services as if they were produced and consumed in markets. Under this approach, the prices of nonmarket goods and services are imputed from a market counterpart. Many youth sports organizations, for example, are operated largely by volunteers. Although a fee may be charged for participation in the activity, that fee cannot be viewed as a market price. But there are also private firms that offer opportunities for children to participate in similar recreational activities that do charge a market-determined price. Given information on the relevant output quantities—for example, the numbers of children participating in the various recreational programs of a nonprofit youth sports organization—the price charged for participating in similar activities offered by private firms could be used in valuing the nonprofit organization’s output. In some cases, there may be differences in quality between home-produced outputs and market outputs, just as there may be between home and market production inputs. In principle, the valuation of nonmarket outputs should take into account any differences in the quality of those outputs as compared to similar market outputs, much as we proposed for the valuation of nonmarket as compared to market labor inputs. Even in the case of near-market goods, market and nonmarket outputs may be imperfect substitutes, complicating comparisons of their value. More difficult yet are the cases in which a nonmarket good is an asset that has no direct market counterpart and is never sold. A possible approach in these cases may be to use market prices to value the stream of output produced by the asset over time and then to treat the present value of the returns as a measure of the asset’s value. This approach has a clear grounding in the standard theory that underlies the valuation of marketable capital assets and is the approach taken, for example, by Jorgenson and Fraumeni (1989, 1992) in their work on the valuation of investments in human capital. They begin by calculating the increments to earnings associated with successive increments to education. The present value of the earnings increments, cumulated over a person’s productive lifetime (and assuming that education enhances the value of market and nonmarket time equally), then is used as a measure of the value of the incremental investment in human capital. Investments in health also yield a flow of nonmarket services over time. Improved health increases not only expected years (and quality) of labor market activity, and thus labor market earnings, but also the expected number of years available in which to enjoy all that makes life rewarding. Developing a market-based measure of the marginal value of additional years of life that may flow from health care investments is controversial, though labor market data have proven useful for this purpose. Specifically, the fact that different occupations are associated both with different risks of fatal injury and different relative wage rates has been exploited to derive estimates of the value of an additional year of life. Such measures, while far from perfect, have the advantage of being based on

OCR for page 9
Beyond the Market: Designing Nonmarket Accounts for the United States real-world decisions that yield observable market outcomes and, for that reason, they have appeal. Different approaches may be necessary for the case of nonmarket outputs that are public in nature, such as crime rates and air quality. Again, however, it may be possible to develop measures of the value of these outputs on the basis of market transactions. The levels of many, if not all, of these nonmarket outputs are likely to differ across localities. People presumably will be willing to pay more to live in communities with low crime rates and good air quality than in communities that lack these attributes. The value of such positive attributes are reflected in house prices. At least in principle, one could derive an estimate of the value of lower crime rates, better schools, or higher air quality from a hedonic model that relates house prices to these (and other) community characteristics.6 There are a number of areas for which market valuation, or even imputations based on nonmarket analogues, are simply unavailable and impossible to obtain. Examples of these might include some aspects of social capital, such as degree of social cohesion or cooperation; the effect of terrorism on the population’s sense of well-being; or the “existence” and “legacy” values of national monuments, such as the Grand Canyon. In these cases, any attempted valuation would have to rely on more indirect evidence. As indicated above, we would argue strongly that attention should be directed first to those categories of nonmarket output for which the most defensible, market-based approaches to valuation are possible. Counting and Valuation Issues The national accounts have a consistent structure for reporting prices and corresponding quantities. The two have an intimate connection because prices form the basis for aggregating the quantities of different products. The national accounts have adopted the approach long advocated by index-number theorists—chain-weighted quantity indexes of groups of products are computed by weighting the percent change of the quantity of each product by its share in the dollar value of all the products. As a result, the accounts support productivity calculations directly. Productivity growth for any group of products—including all the products in GDP—is the percent growth of the aggregate quantity less the corresponding weighted growth of the inputs. In the market economy, monetary aggregates generally are the most accessible measures of the level of activity—dollar values of sales, dollars paid as wages and salaries, and so on—and measuring quantities often is more difficult. By definition, however, nonmarket activity does not involve monetary transactions. Consequently, data on monetary aggregates that form the building blocks 6   See Heckman and Vytlacil (1999) and Heckman et al. (2003) for an indication of advances in the theory of hedonic price measurement in this context.

OCR for page 9
Beyond the Market: Designing Nonmarket Accounts for the United States for traditional national income accounting are simply not available. Instead, available data may consist of physical or other quantity indicators of the level of activity, such as hours of time devoted to home production, student-years of education provided, or ambient concentrations of various air pollutants. On one side are those who argue that no nonarbitrary way exists for assigning monetary values to a heterogeneous set of nonmarket inputs or outputs, and that any such assignment unavoidably will reflect value judgments that are inappropriate for a statistical agency (see, e.g., van de Ven et al., 1999, p. 8). The counterposition holds that, without an attempt to assign monetary values to the quantity indicators that are the basic unit of measurement for nonmarket outputs, it will be difficult for policy makers to digest and use the information (see, e.g., National Research Council, 1999, p. 123). This may mean that nonmarket outputs end up being ignored, which implicitly assigns them a value of zero. Alternatively, policy makers may assign a value to the nonmarket output using subjective methods that are less defensible than the methods that would be employed by a statistical agency. In either case, there is a good argument for measurement specialists to provide estimates based on the best possible methods, even if these are highly imperfect, rather than leaving a statistical void. Another argument for attempting to assign monetary values to quantity indicators is that the effort filters out indicators that may be of minor economic importance. One problem with purely physical accounting systems is that, useful as they may be for some research topics, they tend to be encyclopedic and difficult to comprehend. Economics can minimize biased value judgments by providing scientific guidelines and accounting conventions for approximating prices in many cases. And with a monetary metric, the aggregation of detailed measures of output to larger, useful indexes is possible. For these reasons: Recommendation 1.5: Wherever feasible, nonmarket inputs and outputs should be valued in monetary, not just physical unit, terms. The usefulness of a monetary valuation approach depends on the extent and accuracy with which monetary values ultimately can be assigned to the inputs and outputs in question. In order that such assignments be as objective as possible, we favor basing these valuations wherever possible on information derived from the terms of observable market transactions or their analogs. Even when it is difficult to base valuations on market transactions, it is important that calculations be, in principle, reproducible by independent observers. In certain instances, assigning prices to outputs (or inputs) may be so controversial that publishing physical quantity accounts may be the best available option. Given that both price and quantity data are needed to calculate values for the conventional monetized accounts, however, it is reasonable to produce the best price and monetary estimates available (along with the quantity data), as long as sets of assumptions are clearly stated. Limiting an account to physical quantity reporting should be the exception, not the rule. We also again emphasize the desirability of giving priority

OCR for page 9
Beyond the Market: Designing Nonmarket Accounts for the United States in the near term to work in those areas of nonmarket accounting for which valuation can draw from market comparisons. Marginal and Total Valuation Economic valuation methods fall into two broad categories: the first, which tracks the framework of the national accounts, relies on prices, which reflect marginal benefits; the second considers the full amount consumers would be willing to pay for a good or service, which includes a consumer surplus to the extent that amount is greater than the price. Thus, the two approaches differ in the way benefits are measured. In the case of a product or service sold in a competitive market, the price is set at a value that equates the cost of producing and the value of consuming the marginal unit of output. In a standard supply-and-demand diagram, shown in Figure 1-1, a value of P* is assigned to each of the Q* units of output, for a total market valuation represented by the shaded rectangle. The total value to the consuming public of these Q* units of output, however, is the larger, total shaded area. The difference between these two areas, the shaded triangle, is the consumer surplus associated with the consumption of Q* units of the product or service in question at a price P*. FIGURE 1-1 Consumer surplus.

OCR for page 9
Beyond the Market: Designing Nonmarket Accounts for the United States In many cases, knowing consumers’ willingness to pay for first and subsequent units of a good or service does not matter for any decision. Although the public enjoys a large consumer surplus from the production of ice cream—that is, enjoyment exceeding in value the total price paid for the ice cream—there is no policy or accounting issue relating to that surplus. Productivity and other types of measurement use the marginal values revealed by the market price. The same principle applies to many of the nonmarket goods and services that would be included in satellite accounts. In the context of national accounting, the potential usefulness of total valuation seems to us to be restricted to the case of products new to the market where it may be important for accurate comparisons of the value of output in the current period as compared to that in prior periods. It has been argued that the value consumers place on new products should be reflected in properly constructed price indexes as a decline in the price level (see, for example, Hausman, 1996). While there is not yet a consensus on this issue in the price index literature, we would note that deflating nominal expenditures with a price index that accounted for the value realized by the purchasers of new goods would yield an estimate of real output that included consumer surplus associated with the introduction of these goods. It is meaningless, in a national income accounting context, to estimate total value for existing products. Sometimes total value data will be needed for a cost-benefit analysis, and this is fine; cost-benefit analysis and national accounts rest on different conceptual ideals and objectives. Recommendation 1.6: As a general rule, nonmarket accounts should measure the marginal valuations associated with covered goods and services.

OCR for page 9
Beyond the Market: Designing Nonmarket Accounts for the United States This page intentionally left blank.