What Don’t We Know About U.S. Income and Wealth Inequality?

Esther Peterson Consumer Policy Lecture

​

John Sabelhaus

Non-Resident Senior Fellow

The Brookings Institution

​

This presentation was prepared for the American Council on Consumer Interests Virtual Conference, May 18-21, 2021. The analysis and conclusions set forth are those of the author and do not necessarily reflect those of the Brookings Institution or its funders.

Overview

• Focus on “top shares” of income and wealth.
• Top 1% (≈1.2m families) gets most attention.
• Increasingly, top 0.1% (≈120k) or .01% (≈12k).
• Top 10% (≈12m) interesting for some measures.

​

• Key questions, both income and wealth:
• Why do we know more now than 20 years ago?
• What are the key methodological issues?
• Is a consensus emerging?

“Top Shares” Research is Booming Business

• In 2003 Piketty and Saez published a seminal study on U.S. income top shares, using IRS tax data.

​

• Prior to that, most income inequality research was based on public use Current Population Survey (CPS).
• CPS based on “area probability” address sampling.
• Wealthy very good at not answering surveys.

• On the wealth side, in 2016 Saez and Zucman used IRS income tax data to estimate top wealth shares using so-called “capitalization” technique.

​

​

CBO and SCF Also Key Data Sources

• Congressional Budget Office (CBO) was producing periodic income distribution studies focused on tax policy using IRS tax data supplemented with CPS.

​

• Survey of Consumer Finances (SCF) now has a long history (1989+) and widely used to study top shares.
• SCF combines a CPS address-based sample with an IRS-based high wealth oversample.
• SCF has both household income and wealth.
• Your faithful presenter is not a neutral observer.

​

Measuring Income Inequality Using Tax Data

• IRS Statistics of Income (SOI) branch has produced summary statistics on tax filings for many decades.
• TAXSTATS: little or no focus on top income shares.

​

• Tax data ideal for capturing top shares in one sense, because failure to fill out the “survey” not an option.

​

• Tax data less than ideal in other ways.
• Income concept limited to what’s on the form.
• Population limited to those who file.

​

• Piketty/Saez made drawbacks “academic” in 2003.

​

Piketty/Saez: Top 90-95, 95-99, 1% Income Shares

Piketty, Thomas, and Emmanuel Saez. 2003. “Income Inequality in the United States, 1913–1998.” Quarterly Journal of Economics 118, no. 1: 1–39.

Top 1% income share began exploding in the 1980s, and by 1998 had nearly doubled.

​

​

Meanwhile, shares of other groups within top 10% grew, but slowly.

​

​

Piketty/Saez: Top .01% Income Share

Top .01% income share rose by a factor of five!

​

CBO Top Income Growth Rates Same Pattern…

Congressional Budget Office, 2017. The Distribution of Household Income, 2017. (December)

​

Top .01% income growth well above the next .09%, which was well above the next .9%, which was well above next 19%, etc…

…but CBO Top Income Shares Are Lower

CBO published top 1% incomes shares lower and rising much more slowly than Piketty and Saez.

​

SCF Top 1% “Preferred” Income Close to CBO

Bricker, Jesse, Alice Henriques, Jacob Krimmel, and John Sabelhaus. 2016. “Measuring Income and Wealth at the Top Using Administrative and Survey Data,” Brookings Papers on Economic Activity, 1:2016, p. 261-321.

Income tax data is from Piketty and Saez

Unadjusted SCF income measure directly from survey, close to tax concept

Preferred measure is SCF survey + imputations for unmeasured incomes that CBO includes

SCF: Top .1% Income Shares

Bricker, Jesse, Alice Henriques, Jacob Krimmel, and John Sabelhaus. 2016. “Measuring Income and Wealth at the Top Using Administrative and Survey Data,” Brookings Papers on Economic Activity, 1:2016, p. 261-321.

Gap between SCF and Piketty and Saez for top .1% income share similarly increasing

Why Do Piketty/Saez and CBO/SCF Disagree?

• Population coverage not the problem.
• Tax data is the universe, adding non-filers little impact.
• SCF vetted sample, add Forbes 400 => whole population.
• Comparability of CBO and SCF shows not the problem.

​

• Unit of observation part of the problem.
• Top 1% of tax units is about 1.6 million.
• Top 1% of families is about 1.2 million.
• Piketty/Saez top 1 percent in effect top 1.33 percent.

​

• Income concepts/measurement most of problem.
• CBO/SCF include components like employee benefits.
• Hard-to-measure incomes captured to different degrees.

​

​

So Where Are We On Top Income Shares?

So Where Are We On Top Income Shares?

Problem: Comparing Apples to Oranges (and Lemons and Limes)

Solution: Distribute National Income

National Income is wages and salaries, employee benefits, proprietors’ income, rental income, corporate profits, net interest

Retirement income is shifted: accumulation phase in National Income, pay out phase in top share tables

Distributional National Accounts

Auten, Gerald, and David Splinter, 2019. “Income Inequality in the United States: Using Tax Data to Measure Long-term Trends,” manuscript. PSZ series is from Piketty, Thomas, Emmanuel Saez, and Gabriel Zucman, 2018. “Distributional National Accounts: Methods and Estimates for the United States,” Quarterly Journal of Economics 133, no. 2: 553-609.

Using same underlying tax data and same total income, Auten and Splinter get top shares well below Piketty/Saez/Zucman

Under DNA approach, compute income shares of some agreed-upon income total, in this case, pre-tax National Income (with retirement income re-timed)

Methodological Details

• Who gets credited for undistributed profits?
• In 2020, about $2.1 trillion in corporate profits.
• Assign to stock owners using dividends? Stocks owned by pension funds, tax preferred retirement accounts, etc.

​

• Who is assigned “missing” pass-through income?
• BEA estimates $1.7 trillion in pass-through (proprietor’s) income, but less than half of that reported in tax data.
• Auten-Splinter assign missing $s to lower income families, PSZ adjust proportionally, so most goes to the top.
• Aside: SCF captures more proprietor’s income (think respondents vs their accountants) closer to PSZ.

​

How do Taxes and Transfers Change Answers?

Personal Income same as National Income for income components like labor income but excludes corporate retained earnings and adds back in transfers

Disposable Personal Income is what households keep on net, after taxes and transfers (pension income still retimed)

Taxes and Transfers Lower Top Shares

Auten, Gerald, and David Splinter, 2019. “Income Inequality in the United States: Using Tax Data to Measure Long-term Trends,” manuscript. PSZ series is from Piketty, Thomas, Emmanuel Saez, and Gabriel Zucman, 2018. “Distributional National Accounts: Methods and Estimates for the United States,” Quarterly Journal of Economics 133, no. 2: 553-609.

Different assumptions about “incidence” of taxes and transfers amplifies gap in pre-tax

Adjusting for taxes and transfers lowers top 1% share in all years for both sets of authors

So Where Are We On Top Income Shares?

So Where Are We On Top Income Shares?

“Income Inequality is Worse and Rising Faster”

Sabelhaus, John, and Somin Park. 2020. “U.S. income inequality is worse and rising faster than policymakers probably realize.” Washington Center for Equitable Growth. (May).

Haig-Simons income includes all capital gains, realized or not

Adds another 3 ppts to growth of top income shares since late 1980s, puts SCF closer to PSZ (but for a different reason)

Distribute aggregate capital gains using SCF wealth holdings

Top Wealth Shares

• Top wealth share literature also a booming business.
• As with incomes, impetus is availability of tax data.

​

• But we don’t tax wealth (for living people).
• Using income tax data to estimate wealth involves “capitalizing” income flows to estimate wealth stocks.
• Using estate tax data => reverse engineering mortality, but only for a small (and declining) subset of decedents.

​

• SCF is once again a key player. SCF uses tax data to select the sample, but then survey collects wealth.

​

• The various approaches all agree, right?

​

​

Top 0.1% Wealth Shares

Saez, Emmanuel and Gabriel Zucman. 2016. “Wealth Inequality in the United States Since 1913: Evidence from Capitalized Income Tax Data,” Quarterly Journal of Economics, 131(2):519–578.

Saez and Zucman show methods agreed up to late 1980s, diverge after that

SCF Top .1% Wealth Shares

Bricker, Jesse, Alice Henriques, Jacob Krimmel, and John Sabelhaus. 2016. “Measuring Income and Wealth at the Top Using Administrative and Survey Data,” Brookings Papers on Economic Activity, 1:2016, p. 261-321.

Gap between SCF and “capitalized” wealth shares rising faster than Saez and Zucman report

Differences like tax units versus families, DB pension imputation details amplify wedge

Capitalization: How it Works

Capitalization: What Can Go Wrong

Saez and Zucman UPDATED Top 0.1% Wealth Shares

Saez, Emmanuel, and Gabriel Zucman. 2019. “Progressive Wealth Taxation,” Brookings Papers on Economic Activity, pp. 437-511. (Fall)

Revised capitalization, same units, SCF DB pension imputations eliminates most of gap

This chart on tax unit basis, so in effect top ≈.133% wealth share

Smith, Zidar, Zwick (SZZ) Top .1% Wealth Shares

Smith, Matthew, Owen Zidar, and Eric Zwick. 2019. “Top Wealth in the United States: New Estimates and Implications for Taxing the Rich,” U.S. Treasury Department. Available at: https://scholar.princeton.edu/sites/default/files/zidar/files/szz-wealth.pdf

Capitalization and other details still open questions

Top two (average capitalization) and bottom (estate tax)

basically straw men

Lines in middle are emerging consensus

So Where are We on Top Wealth Shares?

• Top 0.1 percent wealth share is up from ≈10-12% in late 1980s to ≈15-17% in mid 2010s.

​

• This does not mean all issues are resolved…
• SCF wealth concentration more dependent on value of privately held (pass through) businesses than SZZ.
• Capitalization requires “known” aggregates, but no market data (think government statisticians vs SCF respondents).
• Some researchers* criticize SCF because pass-through income is too high AND business values too high.

​

• SCF house values once too high, then aggregates changed!

*Bhandari, Anmol, Serdar Birinci, Ellen R. McGrattan, and Kurt See. 2020. “What Do Survey Data Tell Us About US Businesses?” AER Insights, 2(4):443-58.

Expanded Wealth Concept: DB Pensions

• Direct wealth inequality measurement/capitalization limited to “marketable” household wealth.
• DB pensions less important now, but still ≈ size of DC.
• Top wealth shares shown above all impute DB.

​

• Defined Benefit (DB) pension wealth is the present value of a future benefit stream.
• Total DB liabilities owed to households from BEA/FRB.
• Outstanding question of funded/unfunded benefits.
• Sabelhaus/Volz (2019)* imputations now widely used.

*Sabelhaus, John, and Alice Henriques Volz. 2019. “Are Disappearing Employer Pensions Contributing to Rising Wealth Inequality?” Federal Reserve Board of Governors, FEDS Note. (February 1)

Expanded Wealth: Social Security

• Social Security wealth (SSW) dominates all other forms of wealth for most families.

​

• SSW is present discounted value of benefits - taxes.
• Aggregate SSW is zero (except for Trust Fund).
• Individual SSW rises (then falls) over lifecycle.
• 1970s policy changes greatly increased taxes and benefits.
• Progressivity => SSW more important for bottom 99%.

​

• Does growth of SSW offset/explain rising inequality in other components of household wealth?

​

DB, SSW Lower Top Shares, Trends Same

Sabelhaus, John, and Alice Volz. 2021. “Social Security Wealth, Inequality, and Lifecycle Saving,” Forthcoming in Measuring and Understanding the Distribution and Intra/Inter-Generational Mobility of Income and Wealth, National Bureau of Economic Research, Studies in Income and Wealth. Chicago: University of Chicago Press.

DB, SSW Lower Top Shares, Trends Same

Sabelhaus, John, and Alice Volz. 2021. “Social Security Wealth, Inequality, and Lifecycle Saving,” Forthcoming in Measuring and Understanding the Distribution and Intra/Inter-Generational Mobility of Income and Wealth, National Bureau of Economic Research, Studies in Income and Wealth. Chicago: University of Chicago Press.

Top Shares Takeaways

• Top income shares:
• More concentrated than consensus/widely cited published estimates (i.e., CBO) because of “missing” incomes.
• Top shares rising faster than widely believed because those same missing incomes growing in importance.

​

• Top wealth shares:
• Less concentrated than consensus/widely cited published estimates (i.e., SCF, SZZ) because SSW not included.
• Top shares rising faster (at least in a proportional sense).

​

• SCF pass-through business incomes/assets versus IRS/published aggregates remain contentious.

Policy Takeaways

• Should we raise income taxes?
• Focus on tax base first, finding “missing” pass-through business incomes, both non-compliance and legal evasion.
• All income should be treated equally, equalizing capital gains rates one step in right direction.

​

• Should we tax wealth?
• Rather than taxing wealth >$x million, focus on taxing wealth accumulated because original income was untaxed.
• Estate tax used to be income tax backstop, restore?*
• Clarification on “taxing unrealized gains at death.”

*Gale, William G., Christopher Pulliam, John Sabelhaus, and Isabel V. Sawhill. 2020. “Taxing Wealth Transfers Through an Expanded Estate Tax,” Brookings Institution Report. (August 4)

​

Thanks!

​

jsabelhaus@gmail.com

jsabelhaus@brookings.edu

www.johnsabelhaus.com

​

​

​

Extra Slide: Two Great Papers on Wealth by Race

Bhutta, Neil, Andrew C. Chang, Lisa J. Dettling, and Joanne W. Hsu. 2020. "Disparities in Wealth by Race and Ethnicity in the 2019 Survey of Consumer Finances," FEDS Notes. Washington: Board of Governors of the Federal Reserve System, September 28, 2020, https://doi.org/10.17016/2380-7172.2797.

​

Thompson, Jeffrey P., and Gustavo A. Suarez. 2019. “Accounting for Racial Wealth Disparities in the United States.” Federal Reserve Bank of Boston Research Department Working Papers No. 19–13. https://doi.org/10.29412/res.wp.2019.13

​

SCF 2019 Income and Wealth Thresholds