See discussions, stats, and author profiles for this publication at: https://www.researchgate.

net/publication/263245613

The One Percent

Article in Journal of Economic Perspectives · January 2014

CITATIONS READS
7 3,319

4 authors, including:

Richard V. Burkhauser Jeff Larrimore

Cornell University Board of Governors of the Federal Reserve System
263 PUBLICATIONS 6,923 CITATIONS 72 PUBLICATIONS 1,469 CITATIONS

SEE PROFILE SEE PROFILE

All content following this page was uploaded by Richard V. Burkhauser on 16 December 2014.

The user has requested enhancement of the downloaded file.

 Journal of Economic Perspectives—Volume 28, Number 1—Winter 2014—Pages 243–248

Correspondence

To be considered for publication in the Correspondence section, letters should be relatively short—generally less
than 1,000 words—and should be sent to the journal offices at [email protected]. The editors will choose which letters
will be published. All published letters will be subject to editing for style and length.

The One Percent entrepreneurs. On financial profits and inequality,

Mankiw waffles uncomfortably.
The cheerful blandness of N. Gregory Mankiw’s Second, Mankiw refers at best tangentially to
“Defending the One Percent” (Summer 2013, what may be the most dangerous adverse conse-
pp. 21–34) may divert attention from its occasional quence of extreme inequality at the top: the rich,
unstated premises, dubious assumptions, and omit- with a large assist from the Citizens United v. Federal
ted facts. I have room to point only to a few such Election Commission (558 US 310 [2010]) decision,
weaknesses; but the One Percent are pretty good can buy political influence, and not only influence
at defending themselves, so that any assistance they but power. An immediate example is the millions
get from the sidelines deserves scrutiny. of dollars spent by the financial services industry
First, the paper starts off by invoking an iconic to weaken regulation under the Dodd–Frank Wall
Steve Jobs and, without making an explicit claim, Street Reform and Consumer Protection Act of
carries on tacitly as if the One Percent consists 2010.. Mankiw’s partial reference is in connection
2010
mainly of entrepreneurs whose innovations gen- with the contribution of rent-seeking to inequal-
erate a lot of consumer surplus for the world. ity. There his response is that the proper remedy
There would be less alarm if that were so. But a is not to attack extreme inequality but to suppress
large and—before the crisis—increasing part of rent-seeking. Presumably he would give the same
the income of the One Percent arises in the finan- answer here: if inequality corrupts politics, go after
cial services industry, and a large and increasing the corruption, not the inequality. But this would
part of that has come from trading profits. Much be, willfully or not, naive: it is precisely the power of
of the income in question must therefore be the great wealth that makes it difficult or impossible to
payoff to asymmetric information, and generates eradicate corruption.
precious little in the way of aggregate consumer Third, Mankiw’s conception of rents is too narrow;
surplus. Consider this history (for details see he refers mainly to monopoly rents. But economic
Murphy 2012, especially pp. 307–8): from 1970 to rents are pervasive. He mentions the close correla-
about 1995, the median realized compensation for tion between height and income. It seems unlikely
chief executive officers in Standard and Poor’s 500 that height is correlated with true productivity,
broker-dealer firms was essentially indistinguish- outside the NBA. More likely height confers an inter-
able from that of Standard and Poor’s 500 banks personal advantage, so the return to height is a rent,
and industrials. Rather suddenly, between 1996 and at the expense of others. (Our culture has produced
2006, the median broker-dealer chief executive a song with the line: “Short people got no right to
officer started to collect anywhere between 7 and live.”) Taxing height would indeed be odd (Greg
10 times the median compensation of the other Mankiw is pretty tall, by the way); but it is too non-
two groups. This does not smell like the Goldin chalant to presume that all market incomes reflect
and Katz (2008) story. I’ll swallow “innovation,” true productivity, let alone social productivity.
but socially productive innovation, no thanks. Fourth, evidence has accumulated that the degree
Extreme inequality is not primarily about useful of intergenerational income mobility in the United
244 Journal of Economic Perspectives

States is less than it used to be, less than in some References

other advanced economies, and less than the
American self-image requires. These changes and Goldin, Claudia, and Lawrence F. Katz. 2008. The
differences are unlikely to be genetic in origin. (If Race between Education and Technology. Belknap Press.
they were, they would be rents!) Mankiw ignores or Mankiw, N. Gregory. 2013. “Defending the One
skates over this evidence in favor of an anecdote: he Percent.” Journal of Economic Perspectives 27(3):
thinks his children do not have significantly better 21–34.
opportunities than his own, although they grew up Murphy, Kevin J. 2012. “Pay, Politics, and the
in a much more affluent family than he did. This is Financial Crisis.” Chap. 12 in Rethinking the Financial
a rather small sample, of course. Besides, Professor Crisis, edited by Alan S. Blinder, Andrew W. Lo and
Mankiw’s success came in what must be one of the Robert M. Solow. Russell Sage Foundation.
more meritocratic occupations around. I wonder if
he would have had as good a shot at a bank presi-
dency as his children could have. Response from N. Gregory Mankiw
Fifth, you don’t have to be a card-carrying utili-
tarian to believe that taking a (lump-sum) dollar Robert Solow’s scattershot letter offers various
from a random rich person and giving it to a ran- gripes about my paper “Defending the One Per-
dom poor person would lead to a better social state. cent.” Let me respond, as blandly and cheerfully as
Mankiw’s attempts to undermine this intuition fall I can, to his points.
pretty far short: a) Yes, rich countries are pretty First, Solow objects to my invoking the iconic
cheap with aid to poor countries, and might be so Steve Jobs. Jobs is indeed an extreme case, but he
even if they were surer that aid was effective. But also exemplifies much of what has happened to the
this only says something about the limits of human US economy in recent years. In the same JEP issue
solidarity: greater within a family than outside it, as my paper, Kaplan and Rauh (2013) examine
greater within a nation than outside it, etc. So what? alternative hypotheses about increasing inequality.
b) No, we don’t try to equalize the marginal value They conclude that “the increase in pay at the high-
of kidneys across the population. But this does not est income levels is broad based” and that “the US
deny that some people need a kidney more than evidence on income and wealth shares for the top
others, it merely reflects the fact that most people 1 percent is most consistent with a ‘superstar’-style
regard kidneys and other body parts as somehow explanation rooted in the importance of scale and
less fungible than bank accounts. Does it follow skill-biased technological change.” Solow says that
from the fact that we do not redistribute spare kid- I waffle on the role of finance, and about that, he is
neys that we should not redistribute some spare right. The social value of financial activity is hard to
cash? As for the actual progressiveness of the fed- measure. When the evidence is inconclusive, hon-
eral tax code, if we take the group averages given esty requires that we admit it.
in the paper as points on the tax function, the mar- Second, Solow is concerned about the political
ginal tax rate between incomes of $223,500 and influence of the rich. I am less worried, in part
$1,219,700 appears to be slightly lower than that because the wealthy include supporters of both
between $64,300 and $223,500. the right (the Koch brothers) and the left (George
Sixth, who could be against allowing people Soros). Moreover, despite rising inequality, in 2008
their “just deserts?” But there is that matter of and 2012 the United States managed to elect a left-
what is “just.” Most serious ethical thinkers distin- leaning president committed to increasing taxes on
guish between deservingness and happenstance. the rich. In any event, the absence of such politi-
Deservingness has to be rigorously earned. You cal concerns in my paper is hardly unique. The
do not “deserve” that part of your income that economics literature on optimal income redistri-
comes from your parents’ wealth or connections bution, including the celebrated Mirrlees (1971)
or, for that matter, their DNA. You may be born model, is based largely on utilitarian principles, not
just plain gorgeous or smart or tall, and those on some conjectured link between the income dis-
characteristics add to the market value of your tribution and the electoral process.
marginal product, but not to your just deserts. It Third, Solow thinks I have too narrow a con-
may be impractical to separate effort from hap- ception of rents. Actually, I am less concerned
penstance numerically, but that is no reason to about rents (incomes paid to factors of produc-
confound them, especially when you are thinking tion in excess of opportunity cost) than I am
about taxation and redistribution. That is why we about rent-seeking (the attempt to obtain rents by
may want to temper the wind to the shorn lamb, manipulating the political environment). If a per-
and let it blow on the sable coat. son supplies his labor inelastically, his earnings are
entirely rent, but that does not mean he is engaged
Robert Solow in socially unproductive rent-seeking or that he is
Massachusetts Institute of Technology earning more than the value of his marginal prod-
Cambridge, Massachusetts uct. Regarding the narrow, tangential question of
 Correspondence: The One Percent 245

why tall people earn higher wages, the literature reference to marginal tax rates is perplexing, as
has examined various hypotheses; for example, these are relevant for measuring incentive effects,
Case and Paxson (2008) report a positive correla- not progressivity.
tion between height and cognitive skills. Sixth, and finally, Solow asks, who could be against
Fourth, Solow interprets the evidence on inter- allowing people their “just deserts”? Actually, much
generational mobility as showing that the economy of the economics literature on redistribution takes
is not very meritocratic. (Oddly, he exempts the precisely that stand, albeit without acknowledging
economics profession. He seems to believe that doing so. The standard model assumes something
lack of success is often the result of bad luck or like a utilitarian objective function and concludes
a rigged system, unless you are an economist, in that the optimal tax code comes from balancing
which case it’s your own fault.) Although I noted diminishing marginal utility against the adverse
in my article that those born into extreme poverty incentive effects of redistribution. In this model,
face particularly difficult obstacles, I view the rest what people deserve plays no role in the formu-
of the economy as more meritocratic than Solow lation of optimal policy. I agree with Solow that
does. In addition to the Kaplan and Rauh study, figuring out what people deserve is hard, and
I recommend a popular book called The Million- I don’t pretend to have the final word on the topic.
aire Next Door (Stanley and Danko 1996). Written But if my paper gets economists to focus a bit more
by two marketing professors who extensively sur- on just deserts when thinking about policy, I will
veyed high net worth individuals, the book reports feel I have succeeded.
that the typical millionaire is not someone who was
born into wealth but rather is someone who has N. Gregory Mankiw
worked hard and lived frugally. Harvard University
Moreover, the fact that higher income inequality Cambridge, Massachusetts
is associated with lower intergenerational mobility
is not a surprise. Consider the following assump-
tions: 1) A person’s income Y depends on talent T References
and noise ε:
Case, Anne, and Christina Paxson. 2008. “Stature
Y = αT + ε. and Status: Height, Ability, and Labor Market
Outcomes.” Journal of Political Economy 116(3):
2) Talent T is partly heritable, while noise ε is uncor- 499 – 532.
related across generations. 3) Talent T and noise ε Kaplan, Steven N., and Joshua Rauh. 2013. “It’s the
have constant variance but the return to talent α Market: The Broad-Based Rise in the Return to Top
varies over time and across nations. Under these Talent.” Journal of Economic Perspectives 27(3): 35– 56.
assumptions, higher income inequality goes hand in Mirrlees, James A. 1971. “An Exploration in the
hand with lower intergenerational mobility. This sim- Theory of Optimal Income Taxation.” Review of
ple statistical model does not explain why the return Economic Studies 38(2): 175 –208.
to talent varies, but it does explain why inequality Stanley, Thomas J., and William D. Danko. 1996.
and mobility often move in opposite directions. The Millionaire Next Door: Surprising Secrets of America’s
Fifth, Solow tries to revive utilitarian logic without Wealthy. Marietta, GA: Longstreet Press.
calling himself a utilitarian by claiming that redis-
tribution leads to a better “social state.” But what ***
he means by social state, other than something like It is conventional wisdom that income inequal-
total utility, is unclear. In response to my critiques ity grew almost unabated from the late 1970s
of utilitarianism, he says: a) people don’t endorse up through the start of the Great Recession,
more foreign aid because they don’t feel solidarity with the top 1 percent controlling an increasing
with those in other nations, and b) people aren’t share of income in the United States and other
in favor of kidney redistribution because they don’t English-speaking countries. The intellectual under-
view kidneys as fungible. I agree. In fact, Solow’s pinnings of this now dominant view grew out of the
arguments seem more like restatements of my tax-record-based research collected by Facundo
observations than refutations of them. If people Alvaredo, Anthony B. Atkinson, Thomas Piketty,
were maximizing a conventional social welfare and Emmanuel Saez in their world top incomes
function behind a veil of ignorance, they would database and discussed in “The Top 1 Percent in
treat foreigners equally with their fellow citizens International and Historical Perspective” (Summer
and they would treat kidneys as fungible. That they 2013, pp. 3–20).
do not do so is evidence that our innate moral intu- Tax records offer researchers many advantages,
itions are far from utilitarian. but their inherent disadvantages can skew inequality
As for the progressivity of the actual tax system: levels and affect their trends. Because the defini-
Progressivity is correctly gauged by average tax tions of income are determined administratively,
rates, which rise strongly with income. Solow’s they vary over time within and across countries,
246 Journal of Economic Perspectives

and often differ from how most economists would inequality trends since the 1960s found by Piketty
define income. For example, while most econo- and Saez (2003) for market income. But when we
mists agree that Social Security and other public extend our analysis and use a more comprehensive
transfers are income for individuals, tax record data post-tax, post-transfer income measure, includ-
largely miss this income in the United States and to ing in-kind income and accrued capital gains,
a lesser degree in other countries. Similarly, indi- we observe different income trends (Armour,
viduals’ business income is a real resource, whether Burkhauser, and Larrimore 2013). Current levels of
reported on their personal income tax form, and income inequality are still at or near record levels.
thus captured in individual tax return data, or But inequality has not appreciably increased since
reported on a corporate income tax form and thus the late 1980s or early 1990s.
missed in the individual tax return data. This finding does not discount Alvaredo and col-
In Burkhauser et al. (2012), we raise these mea- leagues’ insights on the causes of increased market
surement concerns as they relate to the unusually income inequality. In many ways, their explanations
large increase in inequality in personal tax data still apply when using accrued, rather than realized,
between 1986 and 1988 following the 1986 Tax capital gains. If increased bargaining by executives
Reform Act. In Piketty and Saez’s top income series and capital owners due to lower tax rates is impor-
excluding capital gains (the series they focus on tant, then market income inequality should have
in Piketty and Saez 2003), the increase in the top increased more in the 1980s, when top income tax
1 percent’s income shares during the two years rates fell, than in the 1990s, when top income
following the implementation of the act repre- tax rates increased. This pattern is likely to be the
sents approximately one-third of the total increase case when using accrued capital gains, but is less
since 1985. This change appears to reflect a reclas- true when using realized capital gains—which par-
sification from corporate to individual income in tially shift the timing of these returns to capital into
response to top individual tax rates falling below the 1990s and 2000s.
the corporate rate, rather than a true inequality But designing policy recommendations based on
change. Similarly, Burkhauser, Hahn, and Wilkins pre-tax, pre-transfer market income trends over
(2013) note that a short-term spike in Australian long periods of time seems misguided. If you com-
top incomes in tax-record data between 1986 pare distributions of market income in the 1920s
and 1989 appears to be driven by a change in the when taxes and government transfers were low
treatment of company profits and dividends that with market income today, you miss the dramatic
more fully captured company profits in the per- growth of progressive income taxes and transfers
sonal income tax base. The longer-term growth in like Social Security and the Earned Income Tax
inequality in tax record data since 1989 seems to Credit that shifted income to people with less mar-
arise because Australia started taxing longer-term ket income in the real world. More subtly, leaving
realized capital gains in 1989—but only on prop- out the real world behavioral consequences of the
erty purchased after September 19, 1985. growth of Social Security and other government
Alvaredo, Atkinson, Piketty, and Saez acknowledge transfers obfuscates the impact these policies have
that tax changes may affect income measures over on market income. The US labor force participa-
time. But in the US case, they dismiss this concern tion rate of men aged 65 and over was well over
by referencing their data series that includes realized 50 percent in the 1920s and is now under 20 per-
taxable capital gains. They claim doing so eliminates cent. In a market income world, the elderly look
a major part of the tax avoidance channel, because much worse off today, but in fact their income
corporate income still appears as capital gains. This has shifted to nonmarket sources. In this sense,
rebuttal would be a stronger if realized taxable capi- social policies may increase measured market
tal gains had a close relationship to current income. income inequality even if they reduce inequal-
They do not. Because the United States only taxes ity under broader income measures. Thus, while
capital gains when the asset is sold, individuals can research using tax records contributes greatly to
defer realizing them indefinitely. Thus, capital gains the inequality debate, it should be viewed as an
may not appear on individual tax forms until years addition to, rather than a replacement for, the
or decades after being generated, if at all. This also extensive research conducted on both US and
means corporate income before the 1986 Act may international inequality trends using broader
appear as realized capital gains in the 1990s or 2000s. income measures.
Hence, including taxable realized capital gains as a
measure of current market returns to capital can Richard V. Burkhauser
importantly affect such trends. The classic Haig– Cornell University
Simons income definition includes capital income, Ithaca, New York
but at the point of accrual, not when such gains
are realized. Jeff Larrimore
In Burkhauser et al. (2012), with data from the Joint Committee on Taxation
Current Population Survey, we largely replicate the Washington, DC
 Patents and the Dissemination of Inventions 247

References Burkhauser, Richard V., Markus H. Hahn, and

Roger Wilkins. 2013. “Measuring Top Incomes
Armour, Philip, Richard V. Burkhauser, and Jeff Using Tax Record Data: A Cautionary Tale from
Larrimore. 2013. “Levels and Trends in United Australia.” NBER Working Paper 19121.
States Income and Its Distribution: A Crosswalk Facundo Alvaredo, Anthony B. Atkinson,
from Market Income Towards a Comprehensive Thomas Piketty, and Emmanuel Saez. 2013. “The
Haig–Simons Income Approach.” NBER Working Top 1 Percent in International and Historical
Paper 19110. Perspective.” Journal of Economic Perspectives
Burkhauser, Richard V., Shuaizhang Feng, Stephen 27(3): 3–20.
P. Jenkins, and Jeff Larrimore. 2012. “Recent Trends Piketty, Thomas, and Emmanuel Saez.
in Top Income Shares in the USA: Reconciling Esti- 2003. “Income Inequality in the United States,
mates from March CPS and IRS Tax Return Data.” 1913–1998.” Quarterly Journal of Economics
Review of Economics and Statistics 94(2): 371–88. 118(1): 1–39.

View publication stats