Transcription

JESSE BRICKERJACOB KRIMMELFederal Reserve BoardUniversity of Pennsylvania1ALICE HENRIQUESFederal Reserve BoardJOHN SABELHAUSFederal Reserve BoardMeasuring Income and Wealth at the TopUsing Administrative and Survey DataABSTRACT    Most available estimates of U.S. wealth and income concentration indicate that the top shares are high and have been rising in recent decades,but there is some disagreement about specific levels and trends. Household surveys are the traditional data source used to measure the top shares, but recentstudies using administrative tax records suggest that these survey-based topshare estimates may not be capturing all of the increasing concentration. In thispaper, we reconcile the divergent top share estimates, showing how the choicesof data sets and methodological decisions affect levels and trends. Relative tothe new and most widely cited top share estimates based on administrative taxdata alone, our preferred estimates for both wealth and income concentrationare lower and have been rising less rapidly in recent years.Understanding the determinants and effects of wealth and incomeinequality are mainstays of political economy. Within the generaltopic of inequality, the study of the top wealth and income shares garnersparticular interest. Measuring and explaining wealth and income concentration has challenged economists at least since Vilfredo Pareto (1896) andSimon Kuznets (1953), and the high-quality, micro-level administrativetax data that have recently been made available are generating renewedinterest in the shares of resources controlled by the top wealth and incomegroups. Indeed, the striking trends in top U.S. wealth and income sharesreported in the most widely cited studies based on these newly availableadministrative data sets are now accepted as facts to be embraced andpotentially addressed by policymakers. These observations about levels1. This paper was written while Jacob Krimmel was a research assistant at the FederalReserve Board.261

262Brookings Papers on Economic Activity, Spring 2016and trends in top wealth and income shares have begun to transcend academic debates, entering the mainstream political arena through best sellers such as those by Raghuram Rajan (2010), Joseph Stiglitz (2012), andThomas Piketty (2014), and through political movements such as OccupyWall Street.Despite the political controversies generated by the estimated top wealthand income shares, relatively little attention has been paid to these estimates’ sensitivity to data and methodology.2 For example, using administrative income tax data, Emmanuel Saez and Gabriel Zucman (2016)estimate that the top 1 percent (by wealth) had a wealth share of 42 percentin 2013, up from 29 percent in 1992. However, the Survey of ConsumerFinances (SCF), which combines administrative and survey data, showsless than half the increase in the top 1 percent’s wealth share, rising from30 percent in 1992 to 36 percent in 2013 (figure 1).3 Similarly, Piketty andSaez (2003)4 show that the top 1 percent (by income) had a 23 percentincome share in 2012, an increase of 10 percentage points since 1992. TheSCF shows a 20 percent income share for the top 1 percent in 2012, anincrease of 8 percentage points since 1991 (figure 2).5 Differences in levelsand trends in the top wealth and income shares at higher fractiles, such asthe top 0.1 percent, are even more striking.6The goals of this paper are to investigate why the various types of dataand approaches are giving different answers about top wealth and incomeshares, and to provide preferred estimates that reflect what can best begleaned from all the available data, including macro data. The two mainsources of micro data used here are administrative tax records and the SCFhousehold survey. These data sources rely on different wealth and incomeconcepts as well as different measurements of wealth and income. In this2. Notable exceptions include, for the top income shares, Congressional Budget Office(2014); Burkhauser, Larrimore, and Simon (2012); Burkhauser and others (2012); andSmeeding and Thompson (2011). For the top wealth shares, notable exceptions includeKopczuk (2015b).3. Bricker and others (2014) describe the results from the latest SCF, conducted in 2013.A slow rise in the top wealth shares is also consistent with estimates derived from administrative estate tax data (Kopczuk and Saez 2004).4. Piketty and Saez regularly update the tables and statistics from their 2003 paper.The most recent version, updated to 2014, is available at http://eml.berkeley.edu/ saez/TabFig2014prel.xls. We refer to these updated data throughout this paper.5. SCF income values are for the year preceding the survey.6. These issues are not unique to the United States. See, for example, Atkinson,Piketty, and Saez (2011), who provide a multinational and longer-run view of rising incomeinequality.

263BRICKER, HENRIQUES, KRIMMEL, and SABELHAUSFigure 1. The Top Wealth Shares, 1989–2013Top 1 percent wealth sharesPercentIncome tax data4035Survey 200420072010Top 0.1 percent wealth sharesPercentIncome tax data2015PreferredaSurvey 10Sources: Survey of Consumer Finances; Saez and Zucman (2016).a. Our preferred wealth measure is the Survey of Consumer Finances measure, plus defined-benefit pensionwealth, plus the wealth of the members of the Forbes 400. See the text and the online appendix for details.

264Brookings Papers on Economic Activity, Spring 2016Figure 2. The Top Income Shares, 1988–2012Top 1 percent income sharesPercentIncome tax dataSurvey ofConsumerFinances b20Preferred a151991199419972000Year200320062009Top 0.1 percent income sharesPercentIncome tax data10Preferred a5199119941997Survey ofConsumerFinances b2000Year200320062009Sources: Survey of Consumer Finances; Piketty and Saez (2003).a. Our preferred income measure is the Survey of Consumer Finances measure, plus the value of employerprovided health insurance and government health care programs, plus the value of in-kind government transfers,plus the imputed incomes of the members of the Forbes 400. See the text and the online appendix for details.b. The Survey of Consumer Finances collects income data for the calendar year preceding each triennial survey.

BRICKER, HENRIQUES, KRIMMEL, and SABELHAUS265paper we document that resolving these conceptual and measurement differences also resolves most of the difference in wealth and income concentration estimates from the two data sources.In the case of wealth, concentration measures derived from administrative income tax records can yield improbable results and are sensitive tomodel assumptions. There are no administrative wealth data in the UnitedStates, so “administrative” estimates of wealth must infer wealth by capitalizing taxable income through a common rate of return on asset types.Wealth inferred in this way is heavily dependent on model parameters, andwealth share estimates can be sensitive to small deviations in assumed ratesof return. For instance, the return on fixed-income assets of the wealthyassumed by Saez and Zucman (2016) implies as much as four times morewealth than does a market rate of return, and two times more wealth thanrates of return estimated from estate tax filings. When wealth concentrationis reestimated, changing only the return on fixed-income assets to either ofthese alternate rates of return, the trend and level of wealth concentrationover the past 10 years are identical to SCF estimates that are constrainedto use administrative data wealth concepts and units of measurement.Essentially, the entire difference in wealth concentration estimates is dueto assumptions about measurement and data construction.Adjusting income concepts and the unit of measurement generally alsobrings estimated income shares in the administrative tax data (Piketty andSaez 2003) and the SCF into agreement. However, neither data set is ableto provide a full accounting of total personal income in the United States.The central goal of this paper, then, is to go beyond reconciliation andprovide preferred top share estimates of wealth and income. These preferred estimates marry the concepts from macro data to micro data andcover the full target population, which is all U.S. families. We provideevidence that augmenting the SCF gets us close to this ideal. Overall,the top share estimates derived in this paper show much lower and lessrapidly increasing top shares than the widely cited values from the Saezand Zucman (2016) and Piketty and Saez (2003) studies mentioned above(figures 1 and 2).7To produce new and improved estimates of wealth and income concentration, we begin by considering the preferred concept of wealth and7. The top share estimates from Piketty and Saez (2003) and Saez and Zucman (2016)are regularly updated and published in the World Wealth and Income Database, which ismaintained by Facundo Alvaredo and Tony Atkinson, along with Thomas Piketty, EmmanuelSaez, and Gabriel Zucman. This database is accessible at www.wid.world.

266Brookings Papers on Economic Activity, Spring 2016income from an economic point of view. The preferred concept of wealthincludes all assets over which a family has a legal claim that can be used tofinance its present and future consumption. This concept mirrors the household wealth concept used in the Financial Accounts of the United States(FA) because it includes a family’s liabilities and both its financial andnonfinancial assets, as well as its rights to defined-benefit (DB) pensions.8The preferred income concept includes all income received by a family,whether or not it is fully taxed, partially taxed, or untaxed. This conceptmirrors the personal income category in the National Income and ProductAccounts (NIPA). Both the FA and NIPA are aggregate data, however, andmicro data sets are needed for distributional analysis.Several challenges must be confronted when estimating wealth andincome distributions with micro data, such as the SCF and the administrative tax data. The first is that micro data sets do not include every FAwealth concept or every NIPA income concept. Untaxed income, such asthe value of employer-provided health insurance and some governmenttransfer income, is never collected in the income tax data and is only sometimes collected in a survey. The SCF wealth estimate typically does notinclude DB pensions, while most forms of consumer debt cannot be estimated when wealth is inferred from income tax data.A second estimation challenge concerns differences in population coverage and measurement between these micro data sets. Household surveysare generally thought to reliably cover the full income and wealth distribution, save perhaps the very top. Administrative tax data can reliably coverthe top, but coverage suffers at the bottom of the distribution because manyfamilies are not required to file tax returns.Differences in measurement also arise in the units of analysis, which aretax units in the income tax data and the family in a household survey. Thereare many more tax units (161 million) than families (122 million). Familiesin the bottom 99 percent are often split into multiple tax units, but a tax unitin the top 1 percent is almost always a family. Counting the top 1 percent(1.61 million) of tax units, then, effectively includes more families thancounting the top 1 percent (1.22 million) of families in a survey.In addition to the conceptual, coverage, and unit-of-analysis difficulties that plague efforts to measure either income or wealth concentration,estimating top wealth shares using administrative tax data introduces yetanother potential source of errors. Wealth can only be measured indirectly8. The Financial Accounts of the United States (Statistical Release Z.1) are availablefrom the Federal Reserve Board (http://www.federalreserve.gov/releases/z1).

BRICKER, HENRIQUES, KRIMMEL, and SABELHAUS267in income tax data—meaning that wealth is inferred mainly by “capitalizing” income flows—which is at the heart of the approach taken by Saez andZucman (2016).9 In a survey like the SCF, wealth is measured directly byasking families about their balance sheets. Accounting for these measurement differences by constraining the SCF to match administrative tax dataconcepts resolves the discrepancies between the various top wealth shareestimates. In particular, the evidence given here and by Wojciech Kopczuk(2015b) shows the sensitivity of wealth inferred from income tax data.By marrying the concepts from the macro data to the micro data, wecan provide preferred top share estimates that cover the full target population: all U.S. families. We provide evidence that augmenting the SCF getsus close to this ideal. We first demonstrate that the SCF represents the fullfamily income and wealth distribution, save for the Forbes 400. By augmenting the SCF household survey along these lines, and by aligning thepreferred wealth and income concepts and measurement laid out above, wederive preferred top share estimates.Our preferred estimates for wealth shares at the top are lower and growing more slowly than in the widely cited capitalized administrative taxdata from Saez and Zucman (2016), but this is mostly for methodologicalreasons, especially the specific capitalization factors used to estimate certain types of wealth (cited above). Indeed, our preferred top wealth shareestimates are quite similar to the published SCF values—because oneadjustment, adding the Forbes 400, pulls up the SCF top wealth shares;and another adjustment, distributing DB pension wealth, pushes top sharesdown by a similar amount (figure 1).Our preferred estimates for top income shares are also lower and rising less rapidly than the recent and widely cited estimates from Pikettyand Saez (2003), which were derived from administrative tax data (figure 2). However, those administrative tax data income