SSRN-id2316233 Presumptive Collection - A Prospect Theory Approach To Increasing Small Business Tax Compliance

Presumptive Collection: A Prospect
Theory Approach to Increasing Small

Business Tax Compliance
KATHLEEN DELANEY THOMAS*
INTRODUCTION
The United States is on an unsustainable fiscal course. Although
revenues are projected to rise relative to government spending in the
short-term, expanding health care costs, the retirement of the baby
boomers, and insufficient revenue are expected to threaten our long-
term financial outlook.1 At the same time, Congress has made little
headway in finding ways to raise additional tax dollars.2 But as the
debate over increasing taxes continues, policymakers seeking to re-
duce the deficit should not lose sight of the $450 billion of tax per year
that goes uncollected, known as the “tax gap.”3
Collectively, U.S. taxpayers pay approximately 83% of their taxes
on a voluntary and timely basis.4 Although an 83% compliance rate
suggests that most taxpayers are highly compliant, a closer look at the
* Assistant Professor of Law, University of North Carolina School of Law. I am

especially grateful to Deborah Schenk and Joshua Blank for reading numerous earlier
drafts of this Article and providing insightful comments and feedback. I would also like to
thank Noël Cunningham, Andrew Hayashi, David Kamin, Mitchell Kane, Jason Oh, Leigh
Osofsky, Gregg Polsky, and Courtney Thomas for helpful comments and discussions. Any
errors or omissions are my own.
1 See Cong. Budget Office, Updated Budget Projections: Fiscal Years 2013 to 2023, at 1
(2013), available at http://www.cbo.gov/sites/default/files/cbofiles/attachments/44172-Base-

line2.pdf; see also David Kamin, Are We There Yet? On a Path to Closing America’s
Long-Run Deficit, 137 Tax Notes 53, 53-54, 67 (Oct. 1, 2012).
2 See, e.g., Lindsey McPherson & Michael M. Gleeson, Parties Disagree on Linking Tax
Reform to Debt Ceiling, 139 Tax Notes 705, 705 (May 13, 2013); Michael M. Gleeson &
Lindsey McPherson, Hatch Gives Tax Reform 25 Percent Chance This Year, 139 Tax Notes
385, 385-86 (Apr. 22, 2013); Lindsey McPherson, Party Leaders’ Rhetoric Hampering Pro-
gress on Tax Reform, 138 Tax Notes 1304, 1304-05 (Mar. 18, 2013).
3 The “tax gap” is the difference between the U.S. federal tax that is owed and the tax
that is voluntarily and timely paid. See IRS, Tax Gap for Tax Year 2006, Overview, at 1
(2012), available at http://www.irs.gov/pub/newsroom/overview_tax_gap_2006.pdf [herein-
after Overview]; IRS, Tax Gap “Map” (2011), available at http://www.irs.gov/pub/news-
room/tax_gap_map_2006.pdf [hereinafter Map]. The 2006 tax gap study, released at the
beginning of 2012, is the most recent data on the tax gap. IRS, The Tax Gap, http://
www.irs.gov/uac/IRS-The-Tax-Gap (last updated Mar. 21, 2013).
4 See IRS, Map, note 3.
111
Electronic copy available at: http://ssrn.com/abstract=2316233

112 TAX LAW REVIEW [Vol. 67:
components of the tax gap reveals an enormous disparity in compli-
ance rates among individuals. At one end of the spectrum5 are wage-
earning employees, who demonstrate a near perfect rate of compli-
ance.6 At the other end of the spectrum are self-employed individuals
earning business income, whose overall compliance rate is less than
50%.7 Tax evasion is a particular problem among self-employed indi-
viduals engaged in cash businesses, who are estimated to report a stag-
geringly low 19% of their income.8
On the surface, the explanation for the disparity between wage
earners and the self-employed seems clear: The self-employed are
noncompliant because their income is generally not subject to with-
holding (an employer subtracting taxes from a paycheck) and is sub-
ject to little information reporting (for example, a bank providing a
Form 1099 showing interest income).9 IRS data reveals that compli-
ance is highest-–approximately 99%-–where there is third-party with-
holding, as is the case with wages.10 Even for amounts that are not
subject to withholding, if there is substantial information reporting, as
is the case with dividend and interest income for example, the compli-
ance rate is approximately 92%.11 In contrast, amounts subject to lit-
tle or no information reporting, such as individual business income,
have a compliance rate of only 44%.12
5 For the sake of simplicity, this Article generally juxtaposes wage earners, whose in-
come is subject to withholding, with self-employed individuals, whose income generally is

not subject to withholding. See IRC § 3402. In reality, wages and business income are not
mutually exclusive, and some taxpayers have income that falls into both categories.
6 IRS, Overview, note 3, at 3 chart 1. Wages and salaries have a “net misreporting per-
centage” (defined as the ratio of net misreported income to true income) of 1%. Id.
7 Only 44% of income tax on individual business income is accurately reported, with an
estimated 61% of individuals engaged in under-reporting. See id.; Gov’t Accountability

Office, GAO-07-1014, Tax Gap: A Strategy for Reducing the Gap Should Include Options
for Addressing Sole Proprietor Noncompliance 3 (2007), available at http://www.gao.gov/
assets/270/265399.pdf (based on tax gap data for 2001). The tax gap data does not indicate
what portion of under-reported tax is due to unintentional errors. The possibility that
much of the tax gap might be due to unintentional error is mitigated by the fact that the
data also accounts for unintentionally over-reported tax, resulting in a “net” misreported
amount. See GAO, supra, at 35. This Article focuses solely on intentionally evaded tax.
8 A Closer Look at the Size and Sources of the Tax Gap: Hearing Before the Subcomm.
on Taxation & IRS Oversight, S. Comm. on Fin., 109th Cong. 14 (2006) (statement of J.
Russell George, Treas. Inspector Gen. for Tax Admin.), available at http://
www.treasury.gov/tigta/congress/congress_07262006.pdf.
9 See, e.g., Kent W. Smith & Karyl A. Kinsey, Understanding Taxpayer Behavior: A
Conceptual Framework with Implications for Research, 21 Law & Soc. Rev. 639, 649
(1987).
10 See note 6. Wages are also subject to information reporting because employers report
wages to the IRS. Id.

11 See IRS, Overview, note 3, at 3 chart 1.
12 Id. One IRS study revealed that the majority of independent contractors with income
under $10,000 failed to report any of the income received in that capacity. Charles T.
Electronic copy available at: http://ssrn.com/abstract=2316233

2013] PRESUMPTIVE COLLECTION 113
Although Congress and Treasury have expanded information re-
porting requirements in recent years in an attempt to increase small
business compliance,13 information reporting is generally unable to
capture the cash economy, which is the primary source of evasion
among self-employed individuals.14 Because income generated
through the cash economy usually involves transactions with numer-
ous retail consumers, it generally is not amenable to third party with-
holding or information reporting.15 Further, information reporting
and withholding requirements can be avoided by small businesses
dealing in cash because they can more easily conceal the existence of
transactions from the IRS. Evasion in the cash economy represents
over a quarter of the total federal tax gap, accounting for more than
$100 billion of unpaid tax per year.16 This likely constitutes the major-
ity of the total income tax unpaid by self-employed individuals each
year ($122 billion),17 making the cash economy the predominant
source of small business noncompliance.18 When combined with
under-reported self-employment tax ($57 billion), individual small
business noncompliance accounts for approximately $179 billion, or
40% of the total tax gap.19 This results in a significant revenue loss,
Clotfelter, Tax Evasion and Tax Rates: An Analysis of Individual Returns, 65 Rev. Econ.
& Stat. 363, 364 (1983).
13 See, e.g., GAO, note 7, at 48-51; Michael R. Phillips, Treas. Dep’t, Ref. No. 2010-30-
024, Significant Tax Issues Are Often Not Addressed During Correspondence Audits of
Sole Proprietors (Feb. 24, 2010), available at http://www.treasury.gov/tigta/auditreports/
2010reports/201030024fr.pdf.
14 As used in this Article, the “cash economy” refers to business income that is not
subject to information reporting or withholding, which largely includes income from cash
transactions–-such as a retail sale for cash-–but may also include credit and debit card
transactions as well as payments by check that are not subject to any information report-
ing. Recent legislation requiring information reporting for certain credit and debit card
transactions should remove some of these transactions from the cash economy. See IRC
§ 6050W (effective beginning with the 2011 taxable year, with the first filing required in
2012).
15 See Leandra Lederman, Reducing Information Gaps to Reduce the Tax Gap: When
Is Information Reporting Warranted?, 78 Fordham L. Rev. 1733, 1752-53 (2010) (arguing

that, although cash-based retailers pose a “substantial enforcement problem,” information
reporting by consumers is not a feasible solution).
16 See Nat’l Taxpayer Advocate, 2007 Annual Report to Congress 35 (2007), available at
http://www.irs.gov/pub/tas/arc_2007_vol_1_cover_msps.pdf; see also IRS, Overview, note 3,

at 3 chart 1 ($120 billion of unpaid tax on amounts subject to little or no information
reporting for 2006 compared to a $450 billion tax gap).
17 The $122 billion of unreported individual business income tax represents 56% of total
business income tax due, resulting in a compliance rate of 44%. IRS, Overview, note 3
chart 1; see note 7.
18 See IRS, Overview, note 3, at 2 tbl.1. Although it is likely that some of that $100
billion figure is attributable to other types of taxpayers, most of the income in the cash
economy is earned by small, unincorporated businesses. See, e.g., Nat’l Taxpayer Advo-
cate, note 16, at vi n.1, 36.
19 See IRS, Map, note 3. Self-employed individuals are thus the largest contributors to
the tax gap. By way of comparison, although corporate tax noncompliance (particularly
causes an inequitable shift of the tax burden from noncompliant small
businesses to compliant taxpayers,20 and distorts taxpayer preferences
among different types of businesses.21
The problem of how to collect tax effectively from small business
owners,22 often referred to as the “hard to tax,”23 is not unique to the
United States.24 A number of foreign countries have addressed this
issue through “presumptive taxes,” which impute income to small
businesses based on easily verifiable external factors, rather than rely-
ing on businesses to self-report their income.25 For example, under a
presumptive tax formerly in place in Israel, a laundromat was subject
to income tax based on the amount of water and electricity it used,
information that could be obtained by the taxing authorities from
meter readings.26
through corporate tax shelters) has garnered much attention in recent years, see, e.g.,
Joshua D. Blank, What’s Wrong with Shaming Corporate Tax Abuse, 62 Tax L. Rev. 539,
539 (2009), just 18% of the income that was underreported in 2006 was attributable to
underreporting of the corporate income tax. See IRS, Map, note 3.
20 See Subsection V.B.2.
21 Evasion in the cash economy essentially amounts to a tax preference, which in turn
may lead to an oversupply of small, cash-based businesses.

22 For the sake of simplicity, this Article uses the term “small business” to refer to those
individually-owned businesses that are the largest source of the tax gap and make up the
majority of the cash economy. Although some small businesses operate through partner-
ships, S corporations, or C corporations, most are operated as sole proprietorships (which
include single-member LLCs). See IRS, Small Business/Self-Employed Division At-a-
Glance, http://www.irs.gov/uac/Small-Business-Self-Employed-Division-At-a-Glance (last
updated July 23, 2013). The IRS defines small businesses as those with assets up to $10
million. Id.
23 See, e.g., James Alm, Jorge Martinez-Vazquez & Friedrich Schneider, “Sizing” the
Problem of the Hard-to-Tax, in Taxing the Hard-to-Tax 11, 13-14 (James Alm, Jorge Marti-
nez-Vasquez & Sally Wallace eds., 2004) (describing the hard to tax as including small and
medium-sized businesses, individual proprietors, professionals, and farmers); Leandra Led-
erman, The Interplay Between Norms and Enforcement in Tax Compliance, 64 Ohio St.
L.J. 1453, 1503 (2003).
24 See, e.g., Alm et al., note 23, at 15-23 (estimating the size of the “shadow economy”
(that is unreported income) for over 100 countries).

25 See Arye Lapidoth, The Use of Estimation for the Assessment of Taxable Business
Income 13-15 (1977); Richard M. Bird & Sally Wallace, Is It Really So Hard to Tax the
Hard-to-Tax? The Context and Role of Presumptive Taxes, in Taxing the Hard-to-Tax,
note 23, at 123; Richard M. Bird & Eric M. Zolt, Redistribution Via Taxation: The Lim-
ited Role of the Personal Income Tax in Developing Countries, 52 UCLA L. Rev. 1627,
1685 (2005); Kyle D. Logue & Gustavo G. Vettori, Narrowing the Tax Gap Through Pre-
sumptive Taxation, 2 Colum. J. Tax. L. 100, 121-22 (2011); Joel Slemrod & Shlomo
Yitzhaki, Analyzing the Standard Deduction as a Presumptive Tax, 1 Int’l Tax & Pub. Fin.
25, 25 (1994); Victor Thuronyi, Presumptive Taxation, in 1 Tax Law Design and Drafting
401-33 (Victor Thuronyi ed., 1996), available at http://www.imf.org/external/pubs/nft/1998/
tlaw/eng/ch12.pdf.
26 Harold C. Wilkenfeld, Taxes and People in Israel 148 (1973).
The U.S. approach to targeting small business noncompliance,
which generally focuses on expanding information reporting,27 shares
a central goal with other countries that use presumptive taxes: Both
focus on reducing the opportunity to evade. Presumptive taxes reduce
the opportunity to evade by removing the element of self-assess-
ment.28 The U.S. approach aims to reduce the opportunity to evade
by increasing the probability that the IRS will detect the taxpayer’s
evasion (through information reporting). While both approaches may
increase compliance in certain circumstances, both suffer serious
drawbacks. Because they do not necessarily reflect an individual’s
ability to pay, presumptive taxes have been criticized as unfair.29
Scholars have also cited concerns that such taxes distort taxpayer be-
havior in potentially inefficient ways,30 as was the case with laundro-
mats in Israel that stopped using hot water so as to reduce the amount
of electricity used, which in turn reduced their tax liability.31 And
while expanded information reporting avoids many of the problems of
presumptive taxation, it generally is unable to reach the cash econ-
omy, since businesses that want to avoid reporting requirements can
transact in cash and conceal their income from the IRS.32
There is another key difference between wage earners and small
business owners that offers an additional explanation for the disparity
in compliance between the two groups. Most wage earners receive a
refund when they file their tax return,33 while small business owners
generally do not. This distinction is important because empirical data
27 Although information reporting generally has little impact on the cash economy,
some small business owners are subject to information reporting on certain types of in-
come, such as that earned from credit card transactions. See notes 84-85 and accompany-
ing text.
28 Presumptive taxes do not completely eliminate the opportunity to evade because tax-
payers may alter their behavior to reduce their presumptive tax liability. See Section III.B.
However, in a number of countries replacing self-reporting with presumptive taxation has
been regarded as a successful means of increasing compliance. See notes 51, 57.
29 See, e.g., Bird & Zolt, note 25, at 1687-88; Thuronyi, note 25, at 429.
30 See, e.g., Lapidoth, note 25, at 129-30.
31 Wilkenfeld, note 26, at 148.
32 Cash transactions are particularly difficult for the IRS to detect. See GAO, note 7, at
12. Not only is there no information reporting in most cases, but the lack of a “paper
trail,” such as that created by depositing a check into a bank account, makes it difficult for
the IRS to uncover such transactions during an audit.
33 Close to 80% of all individual taxpayers claim refunds each year. Damon Jones, Iner-
tia and Overwithholding: Explaining the Prevalence of Income Tax Refunds, 4 Am. Econ.
J. Econ. Pol’y 158, 158 (2012)(nearly 80%); Joel Slemrod, Does It Matter Who Writes the
Check to the Government? The Economics of Tax Remittance, 61 Nat’l Tax J. 251, 265
(2008) (at least 75%). The percentage is higher-–approximately 93% for 2009—for taxpay-
ers subject to withholding. IRS, Statistics of Income 2009: Individual Income Tax Returns
122 tbl.3.3 (2011), available at http://www.irs.gov/pub/irs-soi/09inalcr.pdf (112 billion re-
turns with overpayments compared to 120 billion returns for taxpayers subject to
withholding).
shows that the presence of a tax refund (as opposed to a balance due)
at the time of filing is linked to higher tax compliance.34 It also sug-
gests a different approach to targeting small business noncompliance.
The relevance of a taxpayer’s prepayment position-–whether she
faces a refund or a balance due at filing-–stems from a line of deci-
sionmaking theory called “prospect theory.”35 Through extensive em-
pirical observations, psychologists Daniel Kahneman and Amos
Tversky discovered that individuals make decisions based on whether
they perceive the potential outcome to be a gain or a loss relative to a
neutral reference point-–a phenomenon they termed “framing.”36
Kahneman and Tversky observed that individuals facing a gains frame
are generally risk-averse, while individuals facing a loss are generally
risk-seeking.37
A number of economists and psychologists have found that pros-
pect theory is a relevant predictor of taxpayer behavior.38 In the con-
text of framing, the theory predicts that taxpayers claiming a refund
tend to view the outcome as a gain, and thus will demonstrate risk
aversion, making them more likely to comply. Taxpayers facing a bal-
ance due, generally framed as a loss, tend to be risk-seeking, making
34 See, e.g., Paul Corcoro & Peter Adelsheim, A Balance Due Before Remittance: The
Effect on Reporting Compliance, in Recent Research on Tax Administration and Compli-

ance: Selected Papers Given at the 2010 IRS Research Conference 22-25 (June 29-30,
2010), available at http://www.irs.gov/pub/irs-soi/10rescon.pdf. This is true of both small
business owners and wage earners. Henk Elffers & Dick J. Hessing, Influencing the Pros-
pects of Tax Evasion, 18 J. Econ. Psychol. 289, 292-93 (1997).
The presence of a tax refund is likely one of multiple explanations for high compliance
rates among wage earners. For example, the fact that compliance is high for amounts sub-
ject to information reporting but not withholding also suggests that the probability of de-
tection plays an important role in compliance. See note 11 and accompanying text.
35 Daniel Kahneman & Amos Tversky, Prospect Theory: An Analysis of Decision
Under Risk, 47 Econometrica 263 (1972), reprinted in Choices, Values, and Frames 17
(Daniel Kahneman & Amos Tversky eds., 2000) [hereinafter Prospect Theory I].
Kahneman and Tversky updated their original work in 1992. See Advances in Prospect
Theory: Cumulative Representation of Uncertainty, 5 J. Risk & Uncertainty 297 (1992),
reprinted in Choices, Values and Frames, supra, at 44 [hereinafter Prospect Theory II].
36 Daniel Kahneman, Thinking, Fast and Slow 275 (2011); Daniel Kahneman & Amos
Tversky, Choices, Values and Frames, 39 Am. Psychol. 341 (1984), reprinted in Choices,
Values and Frames, note 35, at 1; Kahneman & Tversky, Prospect Theory I, note 35, at 27-
28.
37 Kahneman & Tversky, Prospect Theory I, note 35, at 22-23.
38 See, e.g., John S. Carroll, How Taxpayers Think About Their Taxes: Frames and Val-
ues, in Why People Pay Taxes: Tax Compliance and Enforcement 43, 57-58 (Joel Slemrod
ed., 1992); Otto H. Chang, Donald R. Nichols & Joseph J. Schultz, Taxpayer Attitudes
Toward Tax Audit Risk, 8 J. Econ. Psychol. 299, 304 (1987); Richard Dusenbury, The Ef-
fect of Prepayment Position on Individual Taxpayers’ Preferences for Risky Tax-Filing Op-
tions, 16 J. Am. Tax. Ass’n 1, 2 (1994); A. Schepanski & T. Shearer, A Prospect Theory
Account of the Income Tax Withholding Phenomenon, 63 Org. Behav. & Hum. Decision
Processes 174, 182-84 (1995); A. Schepanski & D. Kelsey, Testing for Framing Effects in
Taxpayer Compliance Decisions, 12 J. Am. Tax. Ass’n 60, 65, 71 (1990).
them more likely to evade.39 This behavior has been confirmed by
numerous empirical studies involving laboratory subjects, and by stud-
ies of IRS compliance data covering hundreds of thousands of actual
taxpayers.40
Thus far, the legal community has made little use of this empirical
evidence.41 To fill this gap, this Article proposes a novel approach to
target small business noncompliance that relies on the demonstrated
connection between tax compliance and the presence of a tax refund.
I argue that the best way to make use of this connection is through a
system under which advanced tax payments from small business own-
ers would be collected on a presumptive basis (what I call “presump-
tive collection”42) throughout the year. In contrast to true
presumptive taxes, however, the regular income tax ultimately would
apply to determine a small business owner’s final tax liability at year
end. Presumptive collection would operate like wage withholding, ex-
cept that the taxpayers themselves would be responsible for remitting
prepayments of tax, rather than third parties. Because individuals are
more likely to report their tax liability accurately if they find them-
selves in a gains frame at the time of filing, the goal of presumptive
collection is to increase the number of small business owners that
would face such a frame at year end.
39 See Sanjit Dhami & Ali al-Nowaihi, Why Do People Pay Taxes? Prospect Theory
Versus Expected Utility Theory, 64 J. Econ. Behav. & Org. 171, 174 (2007). Gideon Yaniv
offers a theoretical model to show that “obligatory advance payments may . . . substitute
for costly detection efforts in enhancing compliance.” Gideon Yaniv, Tax Compliance and
Advance Tax Payments: A Prospect Theory Analysis, 52 Nat’l Tax. J. 753, 755 (1999).
Yaniv argues that if taxpayers hold realistic assessments of the low probability of detection,
advanced tax payments would not completely eliminate the incentive to evade, although
they would still reduce it. Id. at 759. He notes, however, that taxpayers likely do not hold
realistic assessments of the probability of detection because they tend to overweight small
probabilities. Id. at 760. Thus, he concludes, even small overpayments may be sufficient to
“induce honesty” in taxpayers. Id.; see also Section IV.B for a discussion of the treatment
of the probability of detection under prospect theory.
40 See sources cited in note 38.
41 A number of economists and psychologists have examined the link between prospect
theory and tax compliance, see notes 38 and 39, although they generally have not made
policy recommendations. The legal community, on the other hand, has paid relatively little
attention to the relevance of prospect theory to tax compliance. The few legal scholars that
have focused on prospect theory in the tax compliance context have suggested increasing
wage withholding, which does not address small business evasion. See notes 165-68 and
accompanying text.
42 “Presumptive collection” refers to the proposed system under which the government
would require taxpayers to make advanced tax payments throughout the year that are
calculated based upon external factors, rather than income. The government would deter-
mine the amount of these payments, which are referred to as “presumptive prepayments.”
Any amount by which the taxpayer’s presumptive prepayments exceeded his actual income
tax liability would be refunded when he filed his tax return.
Like true presumptive taxes used in developing countries, small
business tax liability initially would be measured indirectly based on
one or more readily observable external characteristics of the tax-
payer’s business, such as location, square feet, or number of employ-
ees. Unlike true presumptive taxes, however, my proposal seeks to
collect tax on a presumptive basis, but not to impose tax on a pre-
sumptive basis. Taxpayers would be required to make interim tax
payments throughout the year based on their predetermined pre-
sumptive liability.43 At the end of the year, an individual would calcu-
late her actual tax liability based on regular income tax principles, and
would claim a refund (or pay a balance due) based on the difference
between her true tax liability and what she prepaid throughout the
year on a presumptive basis.
The goal of presumptive collection would be to overcollect tax in a
significant number of cases, so that those small business owners would
be facing a gains frame at the time they filed their return. Small busi-
ness owners would be treated in this respect like wage earners, who
make advanced tax payments through withholding, and who often
claim refunds at year end. By aligning the frame of small business
owners with the frame of wage earners, presumptive collection should
bring the compliance rates of the two groups more closely in line. Al-
though not every type of small business would be amenable to pre-
sumptive collection, it could be an important additional tool for
reducing tax evasion that could be targeted at specific industries. The
resulting increase in small business compliance would narrow the tax
gap, generating much needed revenue and resulting in a fairer tax sys-
tem overall.
This Article proceeds as follows: Part II offers an overview of pre-
sumptive taxation as utilized by a number of foreign countries and
examines some of the forms of presumption currently employed by
the United States. Part III evaluates the effectiveness of both infor-
mation reporting and presumptive taxation at targeting small business
noncompliance, concludes that neither is wholly satisfactory, and ar-
gues that presumptive collection fills gaps left by both approaches.
Part IV explores the underlying behavioral economics rationale for
presumptive collection. Part V develops the presumptive collection
proposal, including a description of the proposal, a discussion of the
costs, and the fairness implications of presumptive collection, and ad-
dresses several potential objections to the proposal. Part VI
concludes.
43 Presumptive collection would be different from than the current estimated tax system
because the government, rather than the taxpayer, would determine the amount of the
interim payment. See Subsection V.A.2.
II. PRESUMPTIVE TAXATION
The presumptive collection proposal advanced in this Article draws

on aspects of presumptive taxation used in other countries. Thus, it is
useful to understand how true presumptive taxes work. Under such a
regime, the tax base is measured indirectly based on some readily ob-
servable characteristics of the taxpayer or the taxpayer’s business.44
The indirectly measured tax is presumed to be the taxpayer’s legal tax
liability regardless of what her liability would have been under the
normal tax system.45 Presumptive taxes are common in developing
countries, where the tax base is difficult for authorities to monitor
through techniques like information reporting or withholding.46 For
example, such taxes are employed in a number of African countries
where the cash economy represents a substantial portion of gross do-
mestic product, where many small businesses are unregistered and pay
no tax, and where the tax administration is relatively weak.47
A. Presumptive Taxation Outside the United States
A number of countries currently use, or have previously used, pre-

sumptive taxes.48 Common factors on which presumptive income is
based under such regimes include gross receipts, sales, number of em-
ployees, or assets of the taxpayer.49 Regardless of the specific factors
used, the key is that the presumptive tax is based on information that
the government can obtain without relying on taxpayers to assess their
own net income.50 For example, Colombia enacted a minimum pre-
sumptive tax under which individual income was presumed to be at
44 See Logue & Vettori, note 25, at 124; Slemrod & Yitzhaki, note 25, at 25; Thuronyi,
note 25, at 401.

45 Thuronyi, note 25, at 402. “Normal” tax is used in this Article to mean the income tax
currently in effect in the United States (or other relevant country), as opposed to an ideal
or normative income tax.
46 See Lapidoth, note 25, at 13-14; Bird & Wallace, note 25, at 123; Bird & Zolt, note 25,
at 1664-65; Slemrod & Yitzhaki, note 25, at 25; Dagney Faulk, Jorge Martinez-Vazquez &
Sally Wallace, Taxing Potential Income: A Second Look at Presumptive Taxes 1 (Georgia
St. Univ. Andrew Young Sch. of Pol’y Stud., Working Paper No. 06-32, 2006), available at
http://aysps.gsu.edu/isp/files/ispwp0632.pdf.
47 See Günther Taube & Helaway Tadesse, Presumptive Taxation in Sub-Saharan Af-
rica: Experiences and Prospects 1 (IMF, Working Paper No. 96/5, 1996), available at http://
www.imf.org/external/pubs/ft/wp/wp9605.pdf.
48 See, e.g., Kenan Bulutoglu, Presumptive Taxation, in IMF, Fiscal Affairs Dep’t, Tax
Policy Handbook 258, 258 (Parthasarathi Shome ed., 1995).

49 Bird & Zolt, note 25, at 1685.
50 See id.
least 8% of the taxpayer’s net worth,51 and similar regimes have been
enacted in Argentina, Mexico, and Venezuela.52
Israel used a presumptive system called the “tachshiv” (or “stan-
dard assessment guide”53) for more than twenty years to tax small
businesses, with the objective of estimating net profit based on exter-
nal factors.54 The tachshiv presumed net income based on a number
of factors developed by the government for specific industries.55 For
example, a restaurant would be taxed on the basis of its geographic
location, the number of seats, and the average price of the items on
the menu (among other factors).56
Although the tachshiv in its official form is no longer in use in
Israel, the guidelines continue to be used on an informal basis.57
Other countries, such as Korea and Ghana, have similarly adopted
standard assessment guides.58 Greece also uses a presumptive regime
51 See Indiria Rajaraman, Presumptive Direct Taxation, Lessons from Experience in
Developing Countries, 30 Econ. & Pol. Wkly. 1103, 1111 (1995). The net worth tax per-
centage was reduced to 5% by the mid-1990’s and is currently an alternative minimum tax
on 3% of net worth. Rajaraman, supra, at 1111; Deloitte, Colombia Highlights 2013
(2013), available at http://www.deloitte.com/assets/Dcom-Global/Local%20Assets/
Documents/Tax/Taxation%20and%20Investment%20Guides/2013/dttl_tax_highlight_2013
_Colombia.pdf. The implementation of the presumptive system is regarded as a success
that is estimated to have increased tax collections in Colombia by 25%. Rajaraman, supra,
at 1111; see also Charles E. McLure, Jr. et al., Net Wealth and Presumptive Taxation in
Colombia, in Readings on Taxation in Developing Countries 287, 289 (Richard M. Bird &
Oliver Oldman eds., 4th ed. 1990) (Colombia’s presumptive tax has “undoubtedly been a
major success in the sense of maintaining a certain minimal degree of integrity in the
system.”).
52 See Thuronyi, note 25, at 412.
53 Lapidoth, note 25, at 124-25; Wilkenfeld, note 26, at 144 n.7.
54 See Logue & Vettori, note 25, at 136; Rajaraman, note 51, at 1113; Thuronyi, note 25,
at 423-24; Shlomo Yitzhaki, Cost Benefit Analysis of Presumptive Taxation, 63

FinanzArchiv: Pub. Fin. Analysis, 311, 314-15 (2007). Under the tachshiv, if a business
agreed to pay tax according to the given criteria, its payment was accepted without ques-
tion. If it claimed income below the threshold created by the system, then the taxing au-
thorities conducted a full investigation of its books and records. See Yitzhaki, supra, at 9.
55 Thuronyi, note 25, at 423-24.
56 See Yitzhaki, note 54, at 7-8.
57 The continued use of the standard assessment guides on an informal basis is said to be
the result of a significant decline in revenue experienced when the official system was
halted. Rajaraman, note 51, at 1116-17. The tachshiv was officially abandoned in 1975 in
large part because it was suspected that groups of businesses had undue influence over
setting levels of net income for their particular industry. See Lapidoth, note 25, at 132; see
also Yitzhaki, note 54, at 13. Although no longer an official system, there was a subsequent
publication in 1992 recommending reasonable levels of income for various types of busi-
nesses that covered more than 130 industries. Lapidoth, note 25, at 316; Rajaraman, note
51, at 1113; Thuronyi, note 25, at 423-24.
58 Vito Tanzi & Milka Casanegra de Jantscher, Presumptive Income Taxation: Adminis-
trative, Efficiency, and Equity Aspects 13 (IMF, Working Paper WP/87/54, 1987), available
at http://www.imf.org/external/pubs/ft/wp/8754.pdf (describing efforts by Korea and Ghana
to implement standard assessment systems); Rajaraman, note 51, at 1113 n.66 (describing a
similar scheme used in Korea).
that relies on external indicators, which was implemented to combat
systemic evasion among individuals.59 Presumptive income under the
Greek system is determined by individual factors such as rent from a
second home, costs of household staff, and values of assets such as
cars, boats, or planes.60
France uses a contractual system of presumptive taxation known as
the “forfait” (“contract”) for certain small businesses with income be-
low a specified threshold.61 Under the forfait, the taxpayer and the
taxing authority make an advanced agreement as to the taxpayer’s tax
liability based on estimated income in lieu of actual income. The
agreed-upon amount of tax applies for two years. To participate in the
forfait, the taxpayer must supply information to the taxing authorities
regarding its purchases, sales, number of employees, and value of its
inventory.62 Other countries have adopted similar contract-based pre-
sumptive systems.63
B. Forms of Presumption in the United States

The United States also uses forms of presumptive taxation64 and,
thus, the presumptive element of presumptive collection would be
neither unprecedented nor unfamiliar.65 One example that is specifi-
cally targeted at evasion with respect to income that is not subject to
withholding is the rule for reporting tip income earned by restaurant
servers.66 Large food establishments must report tip income on their
employees’ W-2s that is calculated as 8% of the establishment’s gross
receipts.67 The employees can petition the IRS to lower their tax lia-
59 Manos Matsaganis & Maria Flevotomou, Distributional Implications of Tax Evasion
in Greece 11 (Hellenic Observatory, Eur. Inst., Papers on Greece and Southeast Europe
No. 31, 2010), available at http://eprints.lse.ac.uk/26074/1/GreeSE_No_31.pdf.
60 Bird & Wallace, note 25, at 134; S.N. Makedonskiy, Russian-European Ctr. for Econ.
Pol’y, Taxation Mechanisms Based on Simplified and Indirect Evaluation of Tax Liabilities,
27 (2005), available at http://www.recep.ru/files/documents/Taxation_Makedonskiy_eng
.pdf.
61 Thuronyi, note 25, at 422. Thuronyi notes that while the forfait was once widely appli-
cable, its importance has declined in recent years. Id.

62 Id.
63 Id. at 422-23 (describing a similar system in Belgium).
64 Slemrod and Yitzhaki argue that all taxes are somewhat presumptive because the
ideal tax base can never be measured with complete accuracy. Slemrod & Yitzhaki, note
25, at 25; see also Yitzhaki, note 54, at 2.
65 Wage withholding is a form of presumptive collection, as discussed below in Subsec-
tion V.A.1.
66 Yoram Margalioth, The Social Norm of Tipping,Its Correlation with Inequality, and
Differences in Tax Treatment Across Countries, 11 Theoretical Inq. L. 561, 584 (2010).
67 IRC § 6053(c), Reg. § 31.6053-3(d); see also Logue & Vettori, note 25, at 124 (high-
lighting the rebuttable presumption applied to tip income); Yoram Margalioth, Compara-
tive Tax Law and Culture, 11 Theoretical Inq. L. 561, 584 (2010) (explaining various
reporting methods in several countries regarding tips received by waiters).
bility if they can prove that they earned less than the presumed 8%
amount.68
Additional forms of presumption employed by the United States
include the use of the standard deduction as a substitute for itemized
deductions below a threshold amount,69 the use of fixed depreciation
schedules in lieu of actual measurements of decline in asset value,70
and the allocation of 30% of business profits to personal services for
citizens living abroad.71 The IRS also uses a form of presumptive tax-
ation by filing “substitute returns” for taxpayers that fail to file tax
returns.72
States and localities also have employed forms of presumption to
tax their citizens. For example, localities in Pennsylvania employed an
“occupation assessment tax,” under which counties assigned set values
to various occupations and applied a rate that varied by locality.73
Washington State, which does not have an income tax, imposes a
“business and occupation tax” that applies to gross receipts.74 The
rate of tax levied on each business is determined by its classification in
a number of predetermined industry categories, such as manufactur-
ing, retail, or wholesaling.75
68 IRC § 6053(c)(3)(C), Reg. § 31.6053-3(h)(2)(iii)(A), (B).

69 IRC § 63(c)(2); see Slemrod & Yitzhaki, note 25, at 27.
70 IRC § 168; see Slemrod & Yitzhaki, note 25, at 27.
71 IRC § 911 allows qualifying U.S. citizens and residents living abroad to exclude cer-
tain foreign earnings from income. The exclusion applies to “foreign earned income” at-
tributable to services, such as wages. IRC § 911(a)(1), (b)(1)(A), (d)(2)(A). For taxpayers
“engaged in a trade or business in which both personal services and capital are material
income-producing factors,” the amount treated as “earned income” is limited to 30% of
the individual’s share of the net profits of the business. IRC § 911(d)(2)(B).
72 IRC § 6020(b)(1); IRS, What If You Don’t File Voluntarily (Sept. 3, 2013), available
at http://www.irs.gov/Businesses/Small-Business-&-Self-employed/Filing-Past-Due-Tax-
Returns; see also Logue & Vettori, note 25, at 124 (discussing the IRS practice of recon-
structing the income of a taxpayer that cannot substantiate his deductions using data from
sources such as the Bureau of Labor Statistics).
73 See Faulk et al., note 46, at 3. The occupational assessment tax was repealed in 2001.
H.B. 2126, 2002 Gen. Assemb., Reg. Sess. (Pa. 2002); see David W. Darare, Rising
Costs. . . Falling Funding, Pa. School Boards Ass’n Bull., Feb. 2008, at 19, available at http:/
/www.qesd.org/cms/lib04/PA01000005/centricity/Domain/1/rising-costs.pdf.
74 Wash. St. Dep’t of Revenue, Business and Occupation Tax (July 2011), available at
http://dor.wa.gov/content.FindTaxesandRates/BAndoTax/#Whatis. Although the taxpayer

is not allowed to deduct costs of doing business like labor or material, certain deductions
and exemptions do apply, such as a deduction for bad debts. Id.
75 Id. States have employed presumptive taxes in other contexts, as well. For example,
Maine employed a presumptive use tax until 1999. Me. Rev. Stat. Ann. tit. 36, § 1861-A
(1996), amended by Me. Rev. Stat. tit. 36, § 1861-A (2001). If taxpayers did not report any
use tax on their state return, they were presumed to owe .04% of their adjusted gross
income. Id. Even after the abolition of the presumptive tax, Maine continues to have a
higher use tax participation rate than any other state. See, Research Dep’t, Minn. House
of Rep., Policy Brief Use Tax Collection on Income Tax Returns in Other States (Apr.
2012), available at http://www.house.leg.state.mn.us/hrd/pubs/usetax.pdf.
III. POTENTIAL APPROACHES TO REDUCING SMALL BUSINESS
NONCOMPLIANCE IN THE UNITED STATES
A. The Traditional U.S. Approach to Reducing Small Business

Noncompliance
Before addressing the question of whether presumptive taxation
would be an attractive approach to reducing small business noncom-
pliance in the United States, it is useful to examine the current U.S.
approach, which mainly focuses on information reporting. The aim of
information reporting is to deter noncompliance by greatly increasing
the likelihood that small business evasion will be detected by the
IRS.76 The idea that increasing the likelihood of detection will de-
crease tax evasion is grounded in expected utility theory, which was
applied to tax compliance in a widely cited work by Michael Al-
lingham and Agnar Sandmo.77 The Allingham and Sandmo (“A-S”)
model assumes that taxpayers are rational actors who make the deci-
sion whether to comply with their tax obligations by comparing the
expected cost of compliance (that is, the tax) to the expected cost of
evasion.78 The expected cost of evasion depends on the expected pen-
alty if the evasion is detected, which is equal to the nominal penalty
discounted by the probability of detection.79 The model thus dictates
that the cost of evasion can be made higher (meaning that compliance
can be increased) by increasing the probability of detection, increasing
the penalty for evasion, or some combination of the two.
76 Information reporting increases the odds of detection because the IRS matches infor-
mation returns filed by third parties with the information reported by the taxpayer on his
return under its Automated Underreporter Program. See IRS Data Book 35 (2010), avail-
able at http://www.irs.gov/pub/irs-soi/10databk.pdf.
77 Michael G. Allingham & Agnar Sandmo, Income Tax Evasion: A Theoretical Analy-
sis, 1 J. Pub. Econ. 323 (1972). Allingham and Sandmo extended work by Gary Becker,
which applied traditional expected utility theory principles to the legal arena. Gary S.
Becker, Crime and Punishment: An Economic Approach, 76 J. Pol. Econ. 169 (1968). The
A-S model is widely cited by both economists and legal scholars. See, e.g., James Andre-
oni, Brian Erard & Jonathan Feinstein, Tax Compliance, 36 J. Econ. Literature 818, 823
(1998); Terrance Chorvat, Trust in Taxation, in Behavioral Public Finance 206, 208 (Ed-
ward J. McCaffery & Joel Slemrod eds., 2006); Kyle D. Logue, Optimal Tax Compliance
and Penalties When the Law Is Uncertain, 27 Va. Tax Rev. 241, 266 (2007) (citing Becker);
Alex Raskolnikov, Crime and Punishment in Taxation: Deceit, Deterrence, and the Self-
Adjusting Penalty, 106 Colum. L. Rev. 569, 576-77 n.26 (2006); Daniel Shaviro, Disclosure
and Civil Penalty Rules in the U.S. Legal Response to Corporate Tax Shelters, in Tax and
Corporate Governance 229, 239 n.39 (Wolfgang Schön ed. 2008); see also Joel Slemrod,
Cheating Ourselves: The Economics of Tax Evasion, 21 J. Econ. Persp. 25, 35-36 (2007).
78 See Allingham & Sandmo, note 77, at 324.
79 Id. For example, if a taxpayer has a 1% chance of being detected by the IRS, then a
$100 nominal penalty would result in a $1 expected penalty. For a taxpayer deciding
whether to evade $500 in tax who is facing a 1% chance at incurring a $100 penalty, the
expected cost of evasion (1% of the penalty plus the tax, or $6) will be cheaper than the
expected cost of compliance ($500).
One method of increasing the rate of detection would be to increase
the rate at which taxpayers are audited. But increasing audit rates
well beyond current levels would likely be cost prohibitive.80 Another
option would be to increase the nominal level of tax penalties. But
with the current individual audit rate approximating 1%,81 penalties
would have to be set at approximately ninety-nine times the tax
evaded to sufficiently deter noncompliance under the A-S model.82
Enacting such high penalties is likely impracticable for a variety of
reasons, including political infeasibility, high administrative cost, and
the potential “crowding out” of voluntary compliance.83
Given the practical limitations of increasing the audit rate or penal-
ties, Congress and Treasury have focused their recent efforts to curtail
small business noncompliance on increasing information reporting,
which also raises the rate of detection for certain taxpayers.84 For ex-
ample, recent legislation requires credit card companies to report the
gross receipts collected by merchants through credit card transac-
tions.85 But although current information reporting rules86 (including
80 See, e.g., Gov’t Accountability Office, GAO-13-151, Tax Gap: IRS Could Signifi-
cantly Increase Revenues by Better Targeting Enforcement Resources 1 (2012), available

at http://www.gao.gov/assets/660/650521.pdf. And, although a higher audit rate would ide-
ally result in a corresponding increase in the probability of detection, in reality, studies
indicate that at least half of noncompliance goes undetected during an audit. See Ras-
kolnikov, note 77, at 584.
One goal of presumptive collection would be to use the IRS’s limited resources in a
more cost-effective way by applying information about small businesses on a broad, ex
ante basis instead of through piecemeal, ex poste audits, as discussed below in Subsection
V.B.1.
81 The overall individual audit rate for 2012 was 1.03%, and was close to 1% for each of
the five years prior. See IRS, Fiscal Year 2012 Enforcement and Service Results (Sept. 3,
2013), available at http://www.irs.gov/pub/irs-news/FY%202012%20enforcement%20
and%20service%20results-%20Media.pdf.
82 For example, to deter the decision to evade $500 of tax with a 1% chance of detection,
the nominal penalty would have to be at least $49,500 (.01($500 + $49,500) = $500) (not
accounting for interest or other costs). Many expected utility models do, however, assume
some level of risk aversion, which effectively places an upper limit on penalties. See, e.g.,
Shaviro, note 77, at 239.
83 See, e.g., Lederman, note 23, at 1466; Logue, note 77, at 268; Raskolnikov, note 77, at
635-36; Alex Raskolnikov, Revealing Choices: Using Taxpayer Choice to Target Tax En-
forcement, 109 Colum. L. Rev. 689, 704 (2009).
84 See, e.g., GAO, note 7; Phillips, note 13. In its 2007 report to the Senate Finance
Committee, the GAO recommended improving sole proprietor noncompliance by ex-

panding gross receipts reporting on Schedule C and improving and expanding information
reporting by third parties. GAO, note 7, at 48-51. Expanding withholding requirements to
address small business noncompliance has not been the subject of serious debate due to the
practical limitations. See note 168 and accompanying text.
85 IRC § 6050W.
86 Current information reporting rules require reporting on Form 1099-MISC any pay-
ments for services that exceed $600 made in the course of one’s trade or business, provided
that the recipient is not a corporation. See IRC § 6041(a); IRS, 2012 Instructions for Form
1099-MISC, available at http://www.irs.gov/pub/irs-pdf/i1099msc.pdf. Payments not made
the new credit card requirements) should cover some portion of small
business income, the rules inevitably will not capture segments of the
population that deal in cash, which is the primary source of small busi-
ness noncompliance.87 It may also be the case that, because so much
small business evasion involves cash transactions, increased third
party reporting with respect to credit card transactions will have little
effect on compliance because most credit card transactions are already
reported by taxpayers.88
As things currently stand, information reporting does not reach the
majority of small business transactions, as it covers only an estimated
25% of small business income that is reported on Schedule C,89 and
that estimate does not include Schedule C income that goes unre-
ported.90 Even if Congress continues to expand information reporting
requirements, many small businesses will continue to deal in cash and
avoid the requirements.91 In light of the limitations on the ability of
information reporting to deter small business noncompliance, an addi-
tional approach is needed.
in the course of one’s business (for example, by individual retail consumers) are not cov-
ered, nor are payments for goods. See 2012 Instructions, supra.
87 See notes 8 and 16 and accompanying text; see also Joseph Bankman, Eight Truths
About Collecting Taxes from the Cash Economy, 117 Tax Notes 506 (Oct. 29, 2007); Logue
& Vettori, note 25, at 105; Susan Cleary Morse, Stewart Karlinsky & Joseph Bankman,
Cash Business and Tax Evasion, 20 Stan. L. & Pol’y Rev. 37, 39-40 (2009).
88 A field study found that most cash business owners reported credit card receipts as
taxable revenue, as opposed to amounts paid in cash or by check, which were viewed as
easier to hide. Morse et. al, note 87, at 50 (reporting results from 275 field study interviews
with cash business owners and the tax return preparers and bankers that serve cash busi-
ness owners).
89 Schedule C is an attachment to Form 1040 on which sole proprietors report their
business income.
90 GAO, note 7, at 17, 48.
91 If the U.S. economy experiences a significant decline in the overall use of cash in
coming years, then information reporting should be more successful in shrinking the tax
gap. Although advances in technology will presumably facilitate more cashless transac-
tions, small businesses will continue to have a tax incentive to deal in cash, and IRS data on
the tax gap does not reveal any improvement in levels of small business evasion from 2001
to 2006. See IRS News Release IR-2012-4 (Jan. 6, 2012), available at http://www.irs.gov/
uac/IRS-releases-new-tax-gap-estimates;-compliance-rates-statistically-unchanged-from-
previous-study. While a more in-depth discussion of the merits of a cashless society or the
likelihood of its occurrence are beyond the scope of this Article, it is the subject of a
fascinating debate. See, e.g., Binyamin Appelbaum, As Plastic Reigns, Printing of Money
Slows, N.Y. Times, July 7, 2011, at B1; David Wolman, Time for Cash To Cash Out?, Wall
St. J., Feb. 11-12, 2012, at C3; see also Ilan Benshalom, Taxing Cash, 4 Colum. J. Tax L. 65
(2012) (proposing a Pigouvian tax on cash withdrawals to correct for negative externalities
imposed by the cash economy).
B. Using Presumptive Taxation to Reduce Small Business
Noncompliance
Although the United States employs presumptions in limited con-
texts, it has never imposed a true presumptive tax on any group of
taxpayers. The use of true presumptive taxation, such as that em-
ployed by countries like Israel or Greece, is a popular tool in countries
where compliance rates are notoriously low because it does not rely
on self-assessment. In addition to the administration and enforcement
benefits, presumptive taxes also may enhance equity in a tax system
by subjecting taxpayers that are otherwise “off the radar” to tax at
rates that are similar to those paid by wage-earning employees.92 Fur-
ther, it has been argued that presumptive taxes enhance efficiency by
imposing a zero marginal tax rate on above-average earnings.93 The
intuition here is that, for a presumptive tax that is calculated based
upon a presumed average level of earnings for a particular industry,
any actual earnings above that average will not change the taxpayer’s
level of presumed tax. Thus, unlike a traditional income tax with a
progressive rate structure, increasing productivity will not subject the
taxpayer to a higher tax burden.
Despite these potential benefits, presuming net income based upon
easily verifiable external factors may lack accuracy or be perceived as
arbitrary.94 For example, two taxpayers with similar businesses that
are categorized identically under a system like the tachshiv might have
greatly disparate incomes if one performs well below average for its
industry and one performs well above average, yet their tax liabilities
might be the same.95 Progressivity may also be lessened since pre-
sumptive taxes essentially impose a ceiling on the tax.96 Further, tax-
payers may change their behavior in inefficient ways to manipulate
the external factors on which the presumptive tax is based.97 For ex-
ample, a taxpayer might decline to hire an extra employee, even if the
extra hire would have added revenue, if doing so would have in-
creased her level of presumptive tax.98
Perhaps due to these drawbacks, the use of presumptive taxation in
the United States has received relatively little attention from legal
92 See Bird & Wallace, note 25, at 124; Bulutoglu, note 48, at 258; Tanzi & Casanegra de
Jantscher, note 58, at 7.

93 See Bulutoglu, note 48, at 258; Tanzi & Casanegra de Jantscher, note 58, at 6.
94 See Bird & Zolt, note 25, at 1687; Yitzhaki, note 54, at 320-21.
95 See Bird & Zolt, note 25, at 1687-88; Yitzhaki, note 54, at 320.
96 Yitzhaki, note 54, at 321-22.
97 See Lapidoth, note 25, at 129.
98 This discussion of concerns with presumptive tax systems is by no means exhaustive.
For an in-depth analysis of the potential efficiency, fairness, and complexity concerns cre-
ated by a number of different presumptive tax regimes, see Logue & Vettori, note 25, at
125-44.
scholars. One notable exception, however, is an article by Kyle Logue
and Gustavo G. Vettori, which proposes the use of a type of presump-
tive taxation to reduce evasion among small and medium-sized busi-
nesses in the United States.99 Because the authors propose a true
presumptive tax regime,100 it would be subject to the same fairness
and efficiency problems as any presumptive tax. Additionally, as dis-
cussed further below, Logue and Vettori’s proposal would not capture
a significant portion of small business evasion because it would ex-
clude income from sales to individual consumers and income from
cash transactions between both consumers and other businesses.
Logue and Vettori explore, but stop short of endorsing, what they
describe as a “modified gross receipts” presumptive tax to target small
business tax evasion.101 The modified gross receipts tax would rely on
the actual gross receipts reported by taxpayers, but would apply a
“presumed profit ratio” to determine net income, rather than relying
on taxpayers to report their business deductions.102 Under the propo-
sal, Congress or Treasury would set the presumed profit ratios for va-
rious lines of business in a number of different industries based on a
study of historical profit margins in those industries.103 When calcu-
lating the tax liability for a particular business, all business expense
deductions would be disallowed and the relevant presumed profit ra-
tio would instead apply to the gross receipts reported by the busi-
ness.104 Logue and Vettori would apply the modified gross receipts
tax only to what they call “merchant-to-merchant transactions,” which
is essentially a sale of goods or services between two businesses, as
opposed to a sale between a business and an individual consumer.105
They argue that the presumptive tax would be most effective if col-
lected through compulsory withholding on business-to-business
payments.106
99 Logue & Vettori, note 25. The authors include “medium” sized businesses along with
small businesses in their presumptive tax proposal to emphasize the fact that the cutoff
between small and large businesses is inherently arbitrary. Id. at 111.
100 They also discuss the possibility of having a presumptive regime that is optional on
either an ex post or an ex ante basis. The optional regimes would act like an “alternative
maximum tax,” under which the taxpayer would always pay the lesser of the presumptive
tax or the regular tax. Id. at 138-42.
101 Id. at 104. Although Logue and Vettori focus their discussion on a modified gross
receipts tax, their article contains a thorough analysis of a number of other types of pre-
sumptive taxes, including a lump sum business tax, a pure gross receipts tax, an asset value
tax, and an external factor tax (such as the tachshiv). Id. at 125-37.
102 Id. at 104, 129.
103 Id. at 129. The authors propose that the data could come from audits of a sample of
small businesses in each industry. Id.

104 Id.
105 Id. at 105.
106 Id. at 131. Logue and Vettori note that compulsory withholding on business-to-busi-
ness payments might improve compliance on its own, without presumptive taxation. They
Logue and Vettori acknowledge that their modified gross receipts
proposal would not address tax evasion on income earned from sales
by small businesses to consumers.107 Even if information reporting on
credit card transactions could improve compliance with respect to
some individual consumer sales, cash sales to consumers would not be
covered, and there is some evidence that these transactions make up a
significant portion (if not a majority) of small business tax evasion.108
Further, because the modified gross receipts proposal relies on tax-
payers to report their gross receipts accurately, not all business-to-
business transactions would be covered either, as businesses that deal
in cash could understate their gross receipts and thus lower their pre-
sumptive tax liability. Thus, while the modified gross receipts propo-
sal would have a powerful impact on the ability of small businesses to
overstate deductions, the most costly source of small business non-
compliance–-understated receipts109—would not be addressed.110
Although they argue that there are a number of advantages to im-
posing a presumptive regime like the modified gross receipts tax on
business-to-business transactions, Logue and Vettori ultimately “do
not go quite so far as to advocate such a change.”111 Their primary
concern is the potential for a “massive tradeoff of accuracy of income
measurement for lowered costs of tax enforcement,” and they caution
argue, however, that without a regime like the modified gross receipts tax, businesses could
still evade by overstating their deductions even in the presence of mandatory withholding.
Id. at 119.
107 This is because their proposal is limited to merchant-to-merchant transactions, with
retail transactions presumably remaining under the regular tax system. Id. at 132.
108 See, e.g., id. at 109; Morse et al., note 87, at 49 n.51.
109 See, e.g., IRS, Map, note 3 (attributing just $17 billion of the 2006 tax gap to individ-
ual “adjustments, deductions, exemptions,” compared to $122 billion to understated busi-

ness income); GAO, note 7, at 10-11 (estimating that in 2001 sole proprietors understated
approximately $53 billion in gross income, compared to approximately $40 billion in over-
stated deductions by sole proprietors); IRS, Tax Gap Facts and Figures, Additional Materi-
als on the Tax Gap (Mar. 2005), http://www.irs.gov/pub/irs-utl/tax_gap_facts-figures.pdf
(“Understated income, not overstated deductions, accounts for over 80% of individual un-
derreporting [in 2001].”); see also Morse et. al., note 86, at 51-53 (reporting that the sub-
jects of field studies have also indicated that under-reporting of gross income is
significantly more prevalent than overstating deductions.)
110 Logue and Vettori acknowledged this point, but their focus is on reducing evasion
that is accomplished through overstated deductions, which they view as a significant source
of small business noncompliance. Logue & Vettori, note 25, at 109-10; see also GAO, note
7, at 10-11 (estimating that in 2001, sole proprietors understated approximately $53 billion
in gross income and overstated approximately $40 billion in deductions, but noting that
certain types of gross income, such as unreported cash receipts, are difficult to detect).
While the modified gross receipts regime would not prevent a small business from under-
stating its receipts, it would prevent businesses from overstating their deductions by substi-
tuting deductions with presumed profit ratios.
111 Logue & Vettori, note 25, at 145.
that further research would be needed to explore how accurately net
profit ratios could be determined and at what cost.112
Other legal scholars share Logue and Vettori’s skepticism of the
wisdom of imposing true presumptive taxation in the United States.113
Even the modified gross receipts proposal, which Logue and Vettori
suggest might be the most efficient and fair way to impose presump-
tive taxation,114 does not address the cash economy, a gap that is also
left open by information reporting.115 The shortcomings of true pre-
sumptive taxation and information reporting highlight the need for an
additional approach to reducing small business noncompliance. As
discussed below in Part V, presumptive collection would improve
small business compliance in areas that are not affected by informa-
tion reporting while avoiding many of the downsides of true presump-
tive taxation.
IV. PROSPECT THEORY AS AN ALTERNATIVE BASIS FOR PRESUMPTION
A. Background on Prospect Theory

Presumptive collection is grounded in the relationship between the
taxpayer’s frame of reference and tax compliance. The relevance of
an individual’s frame arises out of Kahneman and Tversky’s empirical
work on prospect theory. Prospect theory is a branch of cognitive the-
ory, which attempts to study the way that individuals think and make
decisions. While traditional decisionmaking theories, such as ex-
pected utility theory, are based on a hypothetical rational actor faced
with a decision under uncertainty, prospect theory is intended to de-
scribe how people actually respond when making such decisions.116
112 Id. Logue and Vettori also propose alternatives such as: (1) an optional, dual-track
presumptive regime, under which naturally compliant taxpayers would be subject to pre-
sumptive tax and noncompliant taxpayers would be subject to a normal income tax with
high penalties and high audit rates; and (2) the use of presumption as an audit strategy for
the IRS. Id. at 140-41, 145; see also Raskolnikov, note 83, at 689 (proposing a dual-track
tax enforcement regime).
113 See, e.g., Bird & Zolt, note 25, at 1687 (“[T]hose concerned with the equity, the
efficiency, and the long-term development of sound tax systems should be less sanguine
about presumptive tax regimes.”); see also Lapidoth, note 25, at 130 (discussing ability to
pay as ideal measure for tax); Thuronyi, note 25, at 429 (“Some countries make little use of
presumptive methods of taxation, given that such methods inherently involve unfair-
ness . . . .”). But see Tanzi & Casanegra de Jantscher, note 58, at 15 (arguing that a well-
designed presumptive regime could be “superior” from an efficiency point of view to an
income tax).
114 See Logue & Vettori, note 25, at 131.
115 Id. at 133. Additionally, even if another form of true presumptive taxation could
effectively reduce the opportunity for cash businesses to evade, the fairness and efficiency
implications counsel against such an approach.
116 Kahneman, note 36, at 271.
By posing a series of hypothetical choices, or “prospects,” to numer-
ous subjects, Kahneman and Tversky found that individuals deviate
from expected utility theory in predictable ways.117
1. Framing/Reference Dependence
Under the rational actor model, utility is evaluated, and decisions
are thus made based on an individual’s final state of wealth.118
Kahneman and Tversky suspected that most people do not actually
evaluate choices in terms of final states of wealth, but rather they eval-
uate prospects in terms of gains or losses.119 For example, given the
choice to be Person A, who had $1 million yesterday and has $5 mil-
lion today, or Person B, who had $9 million yesterday and has $5 mil-
lion today, a rational actor should be neutral. Yet common logic
seems to indicate that Person A is in a more desirable position than
Person B.120 Person A has experienced a $4 million gain (relative to a
$1 million reference point) while Person B has experienced a $4 mil-
lion loss (relative to a $9 million reference point).121
Kahneman and Tversky’s empirical work confirmed their intuition
that individuals make decisions by comparing the potential outcomes
to a neutral reference point.122 They termed the phenomenon “fram-
ing,” because they observed that individuals make decisions based on
whether their options are framed as gains or losses relative to the ref-
erence point.123
2. Risk Seeking, Declining Sensitivity, and Loss Aversion

Some versions of expected utility theory have incorporated an ele-
ment of risk aversion, which would explain, for example, an individ-
ual’s preference for a sure gain over a gamble with an equivalent or
slightly higher expected value.124 Although Kahneman and Tversky
observed risk aversion in individuals facing gains, they also observed
that individuals are actually risk-seeking when faced with losses.125
117 Kahneman and Tversky posed their hypothetical questions to subjects in Israel, Swe-
den, and the United States and found the pattern of results “essentially identical” across
the locations. Kahneman & Tversky, Prospect Theory I, note 35, at 19.
118 See Kahneman & Tversky, Prospect Theory II, note 35, at 45-46. For example, a bet
on a $20 coin toss would be represented as an even chance of ending up with current
wealth W + $20 or W - $20. Id.
119 Id.
120 Kahneman, note 36, at 275.
121 Id.
122 Id.; Kahneman & Tversky, Prospect Theory I, note 35, at 27-28.
123 See Kahneman & Tversky, note 36, at 4-6.
124 See note 82.
125 Kahneman & Tversky, Prospect Theory I, note 35, at 22-23.
For example, given a choice between a gamble with an 80% chance of
receiving $4000 and a 20% chance of receiving zero, or a 100% chance
of receiving $3000, most subjects chose the sure $3000, despite the fact
that the expected value of the gamble ($3200) is higher. On the other
hand, when faced with an 80% chance of losing $4000 and a 20%
chance of losing nothing, or losing $3000 with certainty, most subjects
chose to gamble.126
Kahneman and Tversky also found that the marginal value of both
gains and losses decreases with magnitude.127 It is this “diminishing
sensitivity” feature that explains risk aversion in the domain of gains
and risk-seeking in the domain of losses. For example, because the
pain of a $900 loss is likely more than 90% of the pain of a $1000 loss,
an individual would choose to gamble.128 Further, they observed that
individuals exhibit loss aversion, meaning that losses are more painful
than gains are pleasurable.129 For example, a loss of $3000 would cre-
ate greater disutility than the amount of utility created by a gain of
$3000. Kahneman and Tversky’s studies indicate that the median pa-
rameter for loss aversion is 2.25,130 meaning that losses are roughly
twice as painful as gains are pleasurable.131
B. The Relevance of the Taxpayer’s Prepayment Position Under

Prospect Theory
As discussed above, a central tenet of prospect theory is the concept
that value132 is based upon gain or loss relative to a reference point.133
Kahneman and Tversky generally define the reference point as the
status quo or an individual’s “current asset position.”134 Empirical
126 Id.
127 Id. at 278.
128 Kahneman, note 36, at 285. A gamble of a 90% chance at losing $1000 would be
preferable to a sure loss of $900 in this scenario.

129 See Kahneman & Tversky, Prospect Theory I, note 35, at 32-34.
130 Kahneman & Tversky, Prospect Theory II, note 35, at 59.
131 Kahneman and Tversky identified two phases of decisionmaking during which indi-
viduals incorporate the above-described phenomena. During the initial “editing” phase,
individuals code each prospect as a gain or a loss. Next, during the “evaluation” phase, the
prospects are evaluated and the one with the highest value is chosen. Kahneman & Tver-
sky, Prospect Theory I, note 35, at 274. As part of the evaluation process, Kahneman and
Tversky found that individuals have a tendency to overweight small probabilities and un-
derweight large probabilities of potential outcomes. See id. at 281. Thus, while the ra-
tional actor model discounts potential outcomes by their objective probability of
occurrence, prospect theory uses “decision weights,” which reflect subjective probabilities.
See id. at 280.
132 Prospect theory is couched in terms of maximizing “value” when choosing among
prospects, analogous to maximizing “utility” under expected utility theory. See id. at 275.
The goal is the same in either case.
133 Id. at 274.
134 Id.
studies support this approach in the case of taxpayers, who tend to
view their current cash position as their reference point.135 When
starting from a taxpayer’s current cash position, any tax refund pay-
ment would be coded as a “gain” and any tax balance that must be
paid would be coded as a “loss.” For this reason, the taxpayer’s pre-
payment position (the presence of a refund or balance due) is relevant
to the prospect theory model.
The relevance of the taxpayer’s prepayment position can be illus-
trated with the following example. Assume that Taxpayer has
$100,000 of income (which also represents her total wealth W) and has
tax liability of $30,000. She is considering the decision whether to
evade $2000 of tax. If she is caught, she will be fined at a penalty rate
of 75% of the tax evaded,136 and the probability of detection is 3%.137
Taxpayer’s decision under expected utility theory can be represented
as follows:
135 See, e.g., Michele Bernasconi & Alberto Zanardi, Tax Evasion, Tax Rates, and Ref-
erence Dependence, 60 FinanzArchiv 422, 442 (2004) (support for current cash position);
Schepanksi & Shearer, note 38, at 176, 182 (same); Richard A. White, Paul D. Harrison &
Adrian Harrell, The Impact of Income Tax Withholding on Taxpayer Compliance: Further
Empirical Evidence, J. Am. Tax’n Ass’n, Fall 1993, at 63, 68 (same).
An alternative reference point would be the taxpayer’s expected cash position. For ex-
ample, a payment over and above what the taxpayer expected to owe would be a loss and a
refund amount above what was expected would be a gain. Most subjects, however, have
demonstrated a tendency to use their current cash position. See Phyllis V. Copeland &
Andrew D. Cuccia, Multiple Determinants of Framing Referents in Tax Reporting and
Compliance, 88 Organ. Behav. & Hum. Decision Processes 499, 502, 509 (2002) (stating
that the use of the current cash position as a reference point is consistent with prospect
theory’s reliance on the status quo as a reference point, but study results showed relevance
of both cash position and expectations); Erich Kirchler & Boris Maciejovsky, Tax Compli-
ance Within the Context of Gain and Loss Situations, Expected and Current Asset Posi-
tion, and Profession, 22 J. Econ. Psychol. 173, 176, 188 (2001) (concluding that self-
employed taxpayers based their decisions on their current cash position); But see Michael
S. Schadewald, Reference Point Effects in Taxpayer Decision Making, J. Am. Tax’n Ass’n,
Spring 1989, at 68 (finding neither cash position nor expectation relevant in a study of
MBA students, but noting limited range of outcomes used in the study compared to other
studies with contrary results).
136 This is the highest possible rate for a civil penalty (applied to fraud), and a penalty
with a 20% rate is more commonly asserted by the IRS. See, e.g., IRS Data Book, note 76,
at 42. I have chosen the higher rate for purposes of the example, however, because any
nominal penalty rate inevitably underestimates the true “fine” for noncompliance, since it
does not account for interest owed and any other potential costs incurred during audit and
potential litigation.
137 Although the overall individual audit rate is approximately 1%, audit rates vary
based on certain taxpayer characteristics. The audit rate for sole proprietors is closer to
3%. See GAO, note 7, at 23.
Comply: ($100,000 – $30,000) = $70,000
Evade: (.97) ($100,000 – $28,000) + (.03)($100,000 –
$30,000 – (0.75*$2000))138 = $71,895
The decision to evade has the higher expected value and, thus, a
rational actor would evade.
Under prospect theory, Taxpayer’s total wealth W is not relevant,
but her current cash position is. Assume first that no tax has been
withheld and, therefore, any tax payment will be a loss. If Taxpayer
complies, then she will have to make a $30,000 payment. If Taxpayer
evades and is not detected (97% probability), her tax payment de-
creases by $2000 to $28,000. If her evasion is detected (3%
probability), she will owe a penalty of $1500 (75% of $2000) in addi-
tion to the tax of $30,000, increasing her total payment to $31,500.
Thus, Taxpayer’s decision whether to evade $2000 of tax is initially
coded as follows:
Comply: (−$30,000)139
Evade: (−$28,000, .97; −$31,500, .03)
Studies have shown that individuals tend to simplify prospects by

effectively “cancelling” out any elements common to all prospects.140
138 Under the A-S model, the utility of a taxpayer’s choices can be represented as
follows:
Comply: U (W - x)
Evade: (1 - p)U (W - y) + pU(W – x – f(x - y))
where current wealth = W; correct tax liability = x; reported tax liability = y (such that y <
x); probability of detection = p and the fine for evasion = f. If the taxpayer complies, she
will have to pay the full tax x. If she evades, she will pay y if not caught, but if she is
caught, she will have to pay the full tax (x) plus a penalty f(x - y). See Allingham &
Sandmo, note 77; see also Shlomo Yitzhaki, A Note on Income Tax Evasion: A Theoretical
Analysis, 3 J. Pub. Econ. 201 (1974) (updating A-S model by pointing out that the penalty
in most systems is based on the underpayment of tax, not unreported income).
Under some formulations of expected utility theory, the value of each choice is also
weighted by a coefficient of risk aversion, under the assumption that that there is a declin-
ing marginal utility of wealth. See, e.g., Kahneman & Tversky, note 36, at 2. For the sake
of simplicity, I have adopted the common approach of disregarding the coefficient of risk
aversion in arriving at expected outcomes.
139 Under prospect theory, the choice between a gain of x with probability p and a gain
of y with probability q is expressed as (x, p; y, q). A 100% chance of x would be expressed

as (x, 1) or simply (x). See, e.g., Kahneman & Tversky, Prospect Theory I, note 35, at 266.
140 See id. at 274-75. This cancellation process affects the prospect theory calculus be-
cause the positive and negative value curves flatten as absolute values increase (that is,
there is diminishing sensitivity to gains and losses).
Because Taxpayer will make a payment of at least $28,000 in all
events,141 the prospects likely will be simplified as follows:
Comply: (−$2000)
Evade: ($0, .97; −$3500, .03)
Because taxpayers faced with losses tend to be risk-seeking, the tax-

payer likely will choose the gamble, that is, evasion, in this scenario.
Application of the prospect theory value function confirms this.142
When accounting for loss aversion, diminishing sensitivity to losses,
and Taxpayer’s subjective weighting of the 0.03 probability of detec-
tion,143 the subjective value (V) of the gamble is approximately -$239:
V($0, 0.97; −$3500, .03) = p(.03)n(−$3500) = −$239144
Because the subjective value of the loss from the gamble ($239) is less
than the sure loss of $2000, Taxpayer likely will evade.
141 Since Taxpayer is contemplating evading only $2000 of her $30,000 tax liability, she
will pay at least $28,000 in tax whether she complies or evades. Thus, the value of each
prospect can be reduced by $28,000.
142 When evaluating a prospect, a taxpayer assigns a subjective value to the prospect
based upon what he believes to be the likelihood of its occurrence, whether he perceives it
to be a gain or a loss, and the magnitude of the gain or loss. The evaluation process was
formally modeled by Kahneman and Tversky using the value function V, which adjusts for
loss aversion and diminishing sensitivity to gains and losses and weights probabilities by
the decision weight function p. See note 131. Under the prospect theory value function, if
the taxpayer’s evasion decision resulted in a tax payment of a if detected (with probability
p) and a payment of b if not detected (with probability 1 − p), the prospect would be
evaluated as follows:
V(a, p; b,1 − p) = p(p)v(a) + p(1 − p)v(b)
See Kahneman and Tversky Prospect Theory I, note 35, at 276. The function v (which
accounts for loss aversion and diminishing sensitivity to gains and losses) differs depending
on whether a gain or loss is being evaluated. If a value x ≥ 0, then v(x) = cb. If x < 0, then
v(x) = −q(−c)b, where the parameter -q is the parameter of loss aversion, and b represents
the diminishing sensitivity of both gains and losses. Based upon the median values calcu-
lated from their empirical observations, Kahneman and Tversky suggested b = 0.88 and q =
2.25. Kahneman and Tversky Prospect Theory II, note 35, at 57-59.1
The weighting function p is defined as: p(p) = pg / (pg + (1 − p)g)g. In Kahneman and
Tversky’s empirical studies, the median value for y was 0.61 for gains and 0.69 for losses.
Id. at 59. Another frequently cited paper by Colin Camerer and Teck-Hua Ho estimates y
= 0.56. Colin F. Camerer & Teck-Hua Ho, Violations of the Betweenness Axiom and
Nonlinearity in Probability, 8 J. Risk and Uncertainty 167, 188, 190 (1994); Drazen Prelec,
Compound Invariant Weighting Functions in Prospect Theory, in Choices, Values &
Frames, note 35, at 67; George Wu & Richard Gonzalez, Curvature of the Probability
Weighting Function, 42 Mgmt. Sci. 1676, 1678, 1686 (1996) (citing Kahneman and Tver-
sky’s and Camerer and Ho’s findings and making their own estimate of g − 0.71).
143 Individuals tend to overweight small probabilities and underweight large ones. See
note 131.
144 The value function V was applied using the parameters suggested by Kahneman and
Tversky. See note 142. The value of p(.97)v($0) is equal to zero and drops out of the
equation.
Now assume that Taxpayer’s employer withheld $32,000 and Tax-
payer is owed a refund of $2000. Taxpayer’s decision whether to
evade $2000 of tax is initially coded as follows:
Comply: ($2000)
Evade: ($4000, .97; $500, .03)145
Since Taxpayer receives a gain of at least $500 in all events,146 the

prospects can be further reduced to:
Comply: ($1500)
Evade: ($3500, .97; $0, .03)
Because taxpayers tend to be risk-averse when faced with gains, Tax-

payer likely will comply in this scenario. Application of the prospect
theory value function supports this conclusion because the subjective
value (V) of the gamble is $1105, which is less than the sure $1500:
V($3500, 0.97; $0, .03) = p(.97)n($3500) = $1105147
Taxpayer’s prepayment position, therefore, has a determinative im-

pact on her compliance decision under prospect theory. When faced
with a refund (gain) of $2000, Taxpayer will be less likely to risk re-
ducing her refund to $500 (by incurring penalties) for a chance to in-
crease it to $4000 (by evading $2000 of tax). Rather, she likely will
prefer the certainty of the $2000 refund. On the other hand, when
faced with a balance (loss) of $2000, Taxpayer is more likely to risk
increasing her payment to $3500 (the tax plus a $1500 penalty) for a
chance to decrease her balance to zero (by evading $2000 of tax).
145 If Taxpayer evades and is not detected, her refund increases by $2000 to a gain of
$4000. If her evasion is detected, she will owe a penalty of $1500 (75% of $2000) in addi-
tion to the tax, decreasing her refund to $500.
146 If Taxpayer reports $30,000 of tax (complies), she is entitled to a $2000 refund. If she
reports $28,000 in tax (evades) and is not detected, she will receive a $4000 refund. If she
reports $28,000 in tax and her evasion is detected, she will owe a $1500 penalty (75% of
$2000 evaded) plus the $2000 in tax that she evaded, reducing her $4000 refund to $500.
Thus, the minimum refund in all events is $500.
147 The value function V was applied using the parameters suggested by Kahneman and
Tversky. See note 142. The value of p(.03)v($0) is equal to zero and drops out of the
equation.
To intuit how a $3500 gain (with a 97% probability) drops to a subjective value of only
$1105, consider that (1) subjective probability weighting (p) reduces the 97% probability to
approximately 84%; and (2) diminishing sensitivity to gains (v) reduces the value of $3500
to approximately $1315.
C. Empirical Support for Applying Prospect Theory to Tax
Compliance Behavior
The predictions offered by prospect theory square with the compli-
ance behavior observed in the real world. A number of laboratory
studies have demonstrated that, when faced with tax compliance deci-
sions, subjects exhibit the framing behavior predicted by prospect the-
ory. Specifically, subjects are more likely to comply when faced with a
refund and are less likely to comply when faced with making a pay-
ment.148 One frequently cited study of the application of prospect
theory to tax compliance carried out simultaneous experiments with
over 600 subjects in six different countries.149 Half of the subjects
were told they would receive a refund, and half were told they would
owe a tax payment.150 The authors concluded that:
[S]ubjects facing the prospect of an additional tax payment

exhibited significantly more noncompliance and tax cheating
than did subjects confronting the prospect of a refund.
When the data were reanalyzed using subjects’ perceptions
of their gain versus loss status as the independent variable,
these effects on noncompliance and tax cheating were
sustained.151
In another study, when subjects were asked if they would claim an

unsubstantiated deduction worth $700 when they either owed $2000 of
tax or expected a $2000 refund, there were significantly more “de-
duct” decisions in the tax-owed group compared to the refund
group.152
What is particularly compelling about the application of prospect
theory to tax compliance behavior is the availability of real world data
that supports the findings of laboratory studies. The IRS’s Taxpayer
Compliance Management Program (“TCMP”) and National Research
Program (“NRP”)153 provide compliance data on hundreds of
148 See sources cited in note 38. To control for the possibility that noncompliant taxpay-
ers intentionally manipulate their prepayment position to avoid a refund (and vice versa),
Richard Dusenbury observed risk preferences for various prepayment positions within sin-
gle subjects. See Dusenbury, note 38, at 1-2 (concluding that less income is reported when
an individual faces a balance due as compared to when that individual receives a refund).
149 Henry S.J. Robben, Paul Webley, Russell H. Weigel, Karl-Erik Wärneryd, Karyl A.
Kinsey, Dick J. Hessing, Francisco Alvira Martin, Henk Elffers, Richard Wahlund, Luk
Van Langenhove, Susan B. Long & John T. Scholz, Decision Frame and Opportunity as
Determinants of Tax Cheating, 11 J. Econ. Psychol. 341, 341 (1990).
150 Id. at 351.
151 Id. at 360.
152 White et al., note 135, at 69-71.
153 The Taxpayer Compliance Management Program was a series of intensive, random
audits conducted on approximately 50,000 taxpayers in a single year. The audits took place
thousands of taxpayers, which has been studied and analyzed by schol-
ars looking for patterns in compliance behavior. The studies of this
data further support the relevance of prospect theory to explain tax-
payer behavior. The random audits reveal that taxpayers claiming re-
funds consistently report a higher percentage of their income than
taxpayers who owe additional tax at the time of filing.154 The higher
compliance among taxpayers claiming refunds is consistent across va-
rying levels of adjusted gross income and across varying sources of
income.155
One study of TCMP data demonstrated that the percentage of tax
returns needing correction increases consistently as a function of the
amount of additional payment required at the time of filing. For
nonwage business income, compliance was around 95% when there
was a refund of $1000, decreasing to approximately 90% with a zero
balance due, and then dropped sharply as the level of balance due
increased.156 Compliance rates for wage earners also dropped as the
balance due increased, going from approximately 97% with a refund
of $1000, to 95% with a zero balance, and dropping below 90%, with a
balance due of $1000.157
The IRS recently published a study of 1988 TCMP data and 2001
NRP data that specifically examined the link between prepayment po-
sition and underreporting noncompliance.158 The authors of the study
every several years from the 1960’s through the 1980’s. The purpose of the program was to
allow the IRS to collect data on taxpayer compliance rates, to identify red flags that are
correlated with noncompliance, and to guide the direction of future enforcement activities.
A compliance measurement program was reinstituted in 2002 as the National Research
Program. The audits are less invasive but involve roughly the same number of returns
(approximately 50,000) per year. See George Lowenstein, Deborah A. Small & Jeff
Strnad, Statistical, Identifiable, and Iconic Victims, in Behavioral Public Finance, note 77,
at 41; George K. Yin, John Karl Scholz, Jonathan Barry Forman & Mark J. Mazur, Improv-
ing the Delivery of Benefits to the Working Poor: Proposals to Reform the Earned Income
Tax Credit Program, 11 Am. J. Tax Pol’y 225, 240 n.51 (1994); Robben et al., note 149, at
346.
154 See, e.g., Chorvat, note 77, at 239.
155 See, e.g., Otto H. Chang & Joseph J. Schultz, Jr., The Income Tax Withholding Phe-
nomenon: Evidence from TCMP Data, 12 J. Am. Tax’n Ass’n, Fall 1990, at 88, 92; Robben
et al., note 149, at 345 (IRS data from a sample of 50,000 returns showing that compliance
declines as the size of the refund due declines/balance due rises); Schepanski & Shearer,
note 38, at 174.
156 Elffers & Hessing, note 34, at 293 fig.1. At a balance due of $250, compliance was
approximately 85%, and at a balance due of $1000, compliance dropped to 70%. Id. A
small business owner might claim a refund of tax paid on business income if, for example,
she overpaid her estimated taxes because she experienced an unforeseen loss at the end of
the year.
157 Id. The difference between the compliance rates in the study and the 99% compli-
ance rate for wages cited in the IRS’ 2006 tax gap study, Overview, note 3, likely is due to
the presence of other types of income (for example, interest) earned by the taxpayers in
the study.
158 Corcoro & Adelsheim, note 34.
used a regression analysis to control for other causes of under-report-
ing in order to isolate the impact of the taxpayer’s prepayment posi-
tion.159 They found the data from both the 1988 and the 2001 samples
to be consistent and concluded that:
[A]t a statistically significant level, holding a balance-due

prepayment position was consistently shown to result in an
increase in underreporting noncompliance. This result holds
using the full sample and at each income level. . . . All speci-
fications show that a refund due is associated with smaller
absolute changes in reporting compliance. The reaction to
the loss condition is much greater than the reaction to the
gain condition.160
As noted by the authors, the results of the study are consistent with
the risk-seeking and loss aversion aspects of prospect theory.161
It is important to reiterate that compliance improves among all
types of taxpayers when they face a refund as compared to a balance
due. If there were merely a correlation between refunds and higher
tax compliance because wage earners are more likely to receive re-
funds and are also subject to information reporting and withholding,
then we would not expect to see lower compliance among wage earn-
ers who face a balance due as compared to those who claim refunds.
Further, if information reporting and withholding were the sole fac-
tors driving tax compliance, then we should not see a significant im-
provement in compliance among small business owners who claim
refunds as compared to those who owe a balance. And although it is
possible that some taxpayers intentionally overpay in advance because
they are generally more risk-averse individuals, changes in compliance
were also observed among individual taxpayers across multiple years,
based on whether that individual received a refund or owed a balance
in a particular year.162 The above-described data supports the notion
159 Id. at 8. The authors considered factors such as liquidity (measured by the presence
of investment income like interest), return complexity, occupation, income source, and age.
Id. at 16-17, 24 tbl.7. They found that “[w]ith all the interaction terms included (along with
different combinations of interactions) the taxpayers still appear to be more compliant if
they have overpaid and more noncompliant if they have a balance due.” Id. at 25.
160 Id. at 23.
161 Id.
162 Corcoro and Adelsheim observed compliance among taxpayers for which tax return
data was available for the two years prior to the year for which there was NRP data. Id. at
19. They isolated a subset of 7365 taxpayers that had stable withholding in the prior two
years but who then experienced a significant, unexpected change in tax liability in the year
under study, resulting in a balance due (an “exogenous taxpayer prepayment position”).
Id. at 19-20, 23, 26. The subset showed significantly lower compliance when there was a
balance due. See id. at 23.
that framing does, indeed, play an important role in tax compliance
apart from information reporting and withholding.163
V. PRESUMPTIVE COLLECTION: A PROSPECT-THEORY-BASED

APPROACH
A. Reframing Taxes for Small Business Owners

Prospect theory has been applied to various aspects of the law with
increasing frequency.164 And although some scholars have suggested
applying it in the context of tax compliance,165 the literature is largely
lacking in tax policy prescriptions. The few scholars, be they legal aca-
163 Although the importance of framing effects should not be overlooked, framing is
likely one of multiple determinants of tax compliance. See note 34. For example, Corcoro
and Adelsheim’s study indicates that liquidity is another factor that contributes to compli-
ance among taxpayers with a balance due. Corcoro & Adelsheim, note 34, at 23. Id. While
the empirical data cannot establish a causal link between prepayment position and compli-
ance that is free from doubt, it presents a compelling justification for at least testing the
effect of framing on a small scale. See note 222 (suggesting the implementation of pre-
sumptive collection through test industries).
164 See Richard Birke, Reconciling Loss Aversion and Guilty Pleas, 1999 Utah L. Rev.
205, 219-46 (1999) (exploring the role of framing effects, loss aversion, and risk-seeking in
the context of criminal pleas); John M.A. DiPippa, How Prospect Theory Can Improve
Legal Counseling, 24 U. Ark. Little Rock L. Rev. 81, 104-14 (2001) (suggesting lawyers
should be aware of their clients’ cognitive biases); Larry T. Garvin, Adequate Assurance of
Performance: Of Risk, Duress, and Cognition, 69 U. Colo. L. Rev. 71, 154-62 (1998) (dis-
cussing the application of prospect theory to the adequate assurance doctrine of contract
law); Richard L. Hasen, Efficiency Under Informational Asymmetry: The Effect of Fram-
ing on Legal Rules, 38 UCLA L. Rev. 391, 416-24 (1990) (suggesting that the impact of
framing effects on consumers’ responses to product warnings may mean that an ex post
strict liability standard is more efficient than a negligence standard); Russell Korobkin &
Chris Guthrie, Psychological Barriers to Litigation Settlement: An Experimental Ap-
proach, 93 Mich. L. Rev. 107, 130-38 (1994) (discussing framing effects in settlement be-
havior); Howard Latin, “Good” Warnings, Bad Products, and Cognitive Limitations, 41
UCLA L. Rev. 1193, 1239-41 (1994) (discussing the application of prospect theory to the
legal treatment of product hazards and warnings); Richard W. Painter, Lawyers’ Rules,
Auditors’ Rules and the Psychology of Concealment, 84 Minn. L. Rev. 1399, 1422-24
(2000) (discussing the impact of loss aversion and frames of reference on a lawyer’s deci-
sion to conceal a client’s misconduct or their own errors); Jeffrey J. Rachlinski, Gains,
Losses, and the Psychology of Litigation, 70 S. Cal. L. Rev. 113, 150-60 (1996) (discussing
the impact of framing effects on litigants’ willingness to settle). See generally Chris Guth-
rie, Prospect Theory, Risk Preference, and the Law, 97 Nw. U. L. Rev. 1115 (2003) (dis-
cussing the use of prospect theory by legal scholars in various doctrinal areas).
165 See Edward J. McCaffery & Joel Slemrod, Toward an Agenda for Public Finance, in
Behavioral Public Finance, note 77, at 5-7; Guthrie, note 164, at 1142-45.
Edward McCaffery has also argued that cognitive biases can help explain features of our
existing tax system that otherwise are difficult to understand, such as the historic increase
in payroll taxes and the prominence of tax shelters during the 1970’s. See Edward J. Mc-
Caffery, Cognitive Theory and Tax, 41 UCLA Law Rev. 1861, 1882-83, 1916-23 (1994).
Marjorie Kornhauser recommends incorporating prospect theory into a “tax morale”
approach to tax compliance, which would also incorporate other nonrational factors such
as social norms and demographic factors. Marjorie E. Kornhauser, A Tax Morale Ap-
demics or psychologists, who have suggested ways to utilize the pre-
dictions of prospect theory in the realm of tax compliance have
focused on increasing wage withholding to ensure that more taxpayers
expect refunds.166 But because small business owners derive income
from a potentially large number of sources, expanding withholding re-
quirements to cover small business income is not a viable option.167
Requiring individual consumers who make payments to small busi-
nesses to withhold tax would impose such disproportionate costs and
would be so difficult to enforce that it has not been the subject of
serious debate.168 Even a withholding regime that was limited to busi-
ness-to-business transactions169 would fail to capture cash transac-
tions, which constitute a significant portion of unreported small
business income.170 Businesses that wished to avoid the withholding
requirements could simply agree to make payments in cash, not report
the income, and thus conceal the entire transaction from the IRS.171
1. The Presumptive Collection Proposal
Given the near impossibility of imposing a comprehensive third

party withholding regime on payments to small business owners, a
more viable approach is to make the small business owners172 them-
proach to Compliance: Recommendations for the IRS, 8 Fla. Tax Rev. 599, 602-03, 612-17,
620-22 (2007).
166 See, e.g., Elffers & Hessing, note 34, at 292-93; Guthrie, note 164, at 1145; Elizabeth
F. Loftus, To File, Perchance to Cheat, 19 Psychol. Today 34, 38 (1985); White et al., note
135, at 76.
168 See, e.g., id. (too many individual consumers to enforce compulsory withholding).
Foreign countries that require withholding on certain business-to-business transactions

usually exclude transactions with consumers because they generally are not seen as viable
withholding agents. Slemrod, note 33, at 264; see also Piroska Soos, Self-Employed Eva-
sion and Tax Withholding: A Comparative Study and Analysis of the Issues, 24 U.C. Davis
L. Rev. 107, 184-85 (1990).
169 Logue and Vettori posit that imposing mandatory business-to-business withholding
would not significantly reduce noncompliance because small businesses would still be able
to overstate business deductions. Logue & Vettori, note 25, at 119. But there is no reason
that wage earners cannot overstate deductions, as well, even if they have less opportunity
to do so as compared to small business owners who have business deductions. This Article
contends that if small businesses were subject to withholding on a significant portion of
their income, such that they had a zero balance or a refund at the time of filing, compliance
in fact would increase.
170 See notes 16 & 108 and accompanying text.
171 See note 32.
172 For the sake of simplicity, this Article considers only individually owned small busi-
nesses (that is sole proprietorships), not corporations or businesses operated as pass-

through entities such as partnerships or S corporations. Sole proprietorships account for
72% of all business in the United States. See GAO, note 7, at 5; see also note 22.
selves responsible for the compulsory prepayment of tax.173 To
achieve this, they should be subject to a form of presumptive tax pre-
payment (that is, presumptive collection), which would require tax
payments throughout the year that are calculated on a presumptive
basis. The goal of presumptive collection is to collect more than ac-
tual tax liability,174 or to approximate actual tax liability, in most cases
so that most small business owners would be faced with a refund or
zero balance when they file their tax returns. Unlike true presumptive
taxes, the presumptive collection proposal advanced in this Article of-
fers a unique approach that uses presumptions to shift the taxpayer’s
frame of reference, rather than as a substitute for the normal income
tax.
At the end of each quarter,175 a small business owner would make a
payment of an amount that the government predetermined and that
was somewhat more than his actual tax liability (in most cases). The
presumptive regime would be a collection device analogous to with-
holding, as distinguished from a true presumptive tax,176 so that small
business owners’ ultimate tax liability would be determined under the
regular income tax. When the taxpayer filed her annual tax return,
she would calculate her tax liability under current law and then claim
a refund (or remit payment for any additional amounts owed) based
on the difference between her total presumptive prepayments and her
173 Although the current estimated tax regime also requires small business owners to
make advanced tax payments, there are significant differences between presumptive collec-
tion and estimated taxes. See note 175 and Subsection V.A.2.
174 For a discussion of how the government would accomplish over-collecting, see Sub-
section V.A.3.
175 The timing would be the same as, or similar to, the current system of estimated tax
payments.
Although this Article adopts a quarterly approach for administrative convenience, pre-
sumptive payments could be made on any interim basis, such as monthly. And it may be
the case that making smaller, more frequent prepayments is more palatable to taxpayers
than making fewer, larger prepayments. If the prepayments could be directly debited from
the taxpayer’s bank account, then presumably a monthly prepayment would not be viewed
as administratively burdensome and might be preferable from a budgeting perspective.
See note 233.
176 Some true presumptive tax systems are “rebuttable,” meaning that presumptive tax
liability can be rebutted if the taxpayer can prove that his actual tax liability is lower. Tanzi
& Casanegra de Jantscher, note 58, at 9. Most true presumptive taxes, however, are inten-
tionally designed so that presumptive tax liability is lower than “true” tax liability, which
gives rebuttable presumptive taxes the same practical effect as nonrebuttable ones. Bird &
Wallace, note 25, at 145. Although there are some similarities between presumptive collec-
tion and a rebuttable presumptive tax, the goal of presumptive collection (achieving a gains
frame) is different than that of true rebuttable presumptive taxes (eliminating self-assess-
ment in most cases). Additionally, taxpayers who owe more than their presumptive tax
under the normal income tax would have to make an additional payment at the time of
filing under presumptive collection, which generally is not the case with a rebuttable pre-
sumptive tax.
actual tax liability. The taxpayer would be compensated with market
interest on any amount refunded to her.177
An overarching goal of presumptive collection is to emulate wage
withholding in a context where withholding is not feasible. Withhold-
ing itself is a form of presumptive collection. The withholding tables
presume net income is equivalent to wages less a few specified deduc-
tions, and they do not fully account for each individual taxpayer’s per-
sonal circumstances.178 For example, Form W-4 does not account for
many deductible expenses such as IRA contributions, alimony, stu-
dent loan interest, and health savings account contributions.179 The
operation of Form W-4 and the withholding tables results in over-
withholding for most taxpayers,180 which in turn results in most em-
ployees claiming refunds when they file their return.181 The employee
generally does not control the remittance of withheld tax to the gov-
ernment throughout the year and the “self-reporting” aspect of our
voluntary compliance system does not come into play until the tax-
payer files her return at the end of the year.
177 The current interest rates for overpayments of tax, which approximate a market rate,
would apply. See IRC § 6621.

178 Form W-4 allows an employee to choose a number of “allowances” that lower the
amount of tax withheld from their wages, which are based on factors such as number of
dependents, marital status, and additional jobs. See IRS, Form W-4 (2012), available at
http://www.irs.gov/pub/irs-prior/fw4—2012.pdf; see also Jones, note 33, at 161.
179 See Form W-4, note 178. Form W-4 does instruct individuals to estimate a dollar
amount for certain itemized deductions such as medical expenses and state taxes, but it is
unclear if this instruction has much impact, as taxpayers may find it difficult to estimate
these expenses in advance.
180 See Leandra Lederman, Statutory Speed Bumps: The Roles Third Parties Play in
Tax Compliance, 60 Stan. L. Rev. 695, 731 n.205 (2007) (withholding tables designed to
over-withhold); Jones, note 33, at 160 (same).
For example, a single individual who earned a $50,000 salary, who had no dependents,
took the standard deduction of $9500, and made a deductible contribution of 10% of his
salary to an IRA would have had $5006 of federal income tax liability for 2011. See IRS,
Instructions to Form 1040 EZ 33 (Tax Tables), available at http://www.irs.gov/pub/irs-prior/
i1040ez—2011.pdf. Under the 2011 employer withholding tables, that individual would
have had $6292 of federal income tax withheld if she claimed two allowances and was
subject to biweekly pay periods, resulting in a tax refund of $1286. See IRS, Employer’s
Tax Guide 35-37 (2011), available at http://www.irs.gov/pub/irs-prior/p15—2011.pdf; Form
W-4, note 178 (instructing taxpayer to enter “1” allowance for self plus “1” additional
allowance if single with only one job; no instruction for allowances for retirement contribu-
tions). Even if she had made no retirement plan contributions and had no deductions
other than the standard deduction, she would have received a small refund of $36.
181 See note 33 (93% of wage earners subject to withholding claim refunds). The aver-
age individual refund is approximately $3000. IRS, 2011 Tax Statistics, available at http://
www.irs.gov/pub/irs-soi/11taxstatscard.pdf. This is representative of both lower income
taxpayers, who may be eligible for refundable credits, and of higher income taxpayers who
generally are not. See IRS, note 33, at 122 tbl.3.3 (average overpayment was $2710 for
$30,000-$40,000 income range, $2754 for $40,000-$50,000 income range, $3208 for $50,000-
$75,000 income range, and $3835 for $75,000-$100,000 income range).
The goal of presumptive collection is to similarly delay the self-re-
porting aspect of tax compliance for small business owners until the
time of filing.182 Although the small business owners themselves
would be responsible for remitting quarterly payments to the govern-
ment, they would not be required to self-assess their tax liability
throughout the year because the quarterly payments would be deter-
mined by the government on a presumptive basis.
2. Presumptive Collection Versus Estimated Taxes

Although presumptive collection bears some superficial similarity
to the current system of estimated taxes, there is an important differ-
ence. The key distinction between estimated taxes and presumptive
collection is that the taxpayer does not determine the amount of the
payments made under presumptive collection.183 Since she does not
have to make a decision regarding how much tax to report and pay at
the end of each quarter, the taxpayer does not find herself in a loss
frame because the gains/loss frame is not even relevant. By the time
she does have to self-assess her tax liability at the time of filing, she
will have prepaid her taxes, resulting in a gains frame.184 In contrast,
as explained below, estimated taxes require taxpayers to make a deci-
sion regarding how much tax to report at the time they make esti-
mated payments.
Under the current income tax, small business owners who are not
subject to withholding are required to pay estimated taxes on a quar-
terly basis.185 To avoid a penalty, the sum of the four estimated tax
payments generally is required to equal the lesser of: (1) 90% of the
current year’s tax liability (each estimated payment comprising 22.5%
of the total liability); or (2) 100% of the previous year’s tax liability
(each estimated payment comprising 25% of the previous year’s liabil-
182 As discussed below, under the current system, small business owners have to make a
decision regarding how much income they plan to report at the time they make estimated
tax payments. See Subsection V.A.2.
183 For a discussion of how the government would determine the amount of the tax-
payer’s presumptive prepayments, see Subsection V.A.4.

184 Some lapse in time between the presumptive prepayments and the reporting of ac-
tual tax liability is probably necessary for a gains frame to arise. For example, a taxpayer
who writes a $1200 check on April 14 and files a return reporting $1000 on April 15 might
still experience a loss frame when she files. But a taxpayer who wrote four $300 checks in
the prior year may be in a gains frame by the time she files and reports $1000 in April.
Although it is probably difficult (if not impossible) to precisely quantify the requisite
length of time between payment and filing, making prepayments during the tax year fol-
lowed by reporting on or after April of the following year would create a time lapse similar
to that experienced by wage earners, who demonstrate compliance behavior consistent
with a gains frame when claiming a refund.
185 IRC § 6654(a) (penalty for failure to pay sufficient estimated tax). Estimated tax
payments are due on April 15, June 15, September 15, and January 15. IRC § 6654(c)(2).
ity).186 While estimated tax payments provide time-value-of-money
benefits for the government and ensure that taxpayers do not run into
liquidity problems when they file their annual returns, they likely do
little, if anything, to encourage taxpayers to report their tax liability
accurately. Presumably those who intend to under-report when they
file their tax return will under-report on their estimated tax payments
as well.187
Unlike wage withholding, over which employee taxpayers have lit-
tle control, paying estimated taxes requires self-assessment of one’s
tax liability. At the time of that self-assessment, a small business
owner is unlikely to find herself in a gains frame, but rather is likely to
be facing a loss equal to the value of the estimated tax payment.188
Under prospect theory, taxpayers in this situation (a loss frame) likely
will exhibit risk-seeking behavior that makes them less likely to com-
ply. For example, a small business owner who anticipates annual tax
liability of $50,000, but who plans to report only $30,000, could simply
make an estimated tax payment of $6750 (22.5% of $30,000) for each
quarter and satisfy the 90% test. Similarly, a small business owner
with relatively stable income who under-reported her income in the
previous year could use the prior year calculation method and con-
tinue to under-report in the current year. In both cases, the taxpayer
can easily control her estimated tax payments so that she avoids incur-
ring any estimated tax penalties while simultaneously underpaying her
taxes. She will be motivated to do so because she will be in a loss
frame.
Presumptive collection avoids this result by removing the element
of self-assessment from the quarterly payment process. The govern-
ment, not the taxpayer, would determine the amount of her quarterly
payment. Because she would not have to make a decision regarding
how much tax to report and pay on a quarterly basis, the gains/loss
frame would not be invoked. By the time she would have to make a
186 IRC § 6654(d)(1). For the second method, each payment must equal 27.5% of the
current year’s liability (110% of 25%) if net income exceeded $150,000 in the prior year.
Id. An alternative method of calculation exists for taxpayers whose income is earned un-
evenly across the year, such as those engaged in a seasonal business. IRC § 6654(d)(2).
187 IRS data on taxpayer noncompliance does not include information on estimated tax
noncompliance. See generally IRS, Overview, note 3 (noncompliance statistics do not in-
clude information on estimated tax noncompliance); IRS, Map, note 3 (same); IRS, note
181 (same).
188 This assumes that the neutral reference point is the taxpayer’s current cash position.
If the reference point were the taxpayer’s expected cash position, then a payment of esti-
mated tax would not necessarily be framed as a loss. For example, a taxpayer that ex-
pected to pay $2000 would find herself in a neutral frame if she made a $2000 payment or
even a gains frame if she made a payment of less than $2000. See note 135 and accompany-
ing text.
decision regarding how much income to report (when she files her
return), she will have prepaid her taxes, resulting in a gains frame.
3. Choosing a Presumptive Collection Method

Presumptive collection would avoid many of the drawbacks of a
true presumptive tax, although, as discussed below, there are effi-
ciency, fairness, and complexity implications that would need to be
weighed against its advantages. The outcome of that analysis depends
on the method of presumptive collection that is chosen. Several op-
tions are surveyed briefly below.189 In light of the drawbacks inherent
in the other approaches, this Article recommends using extrinsic fac-
tors as the basis for presumptive collection. This method would strike
the best balance between accuracy and preventing taxpayer
manipulation.
a. Guiding Principles in Designing a Presumptive Collection

Regime
As discussed above, the goal of presumptive collection is for the
taxpayer to have already satisfied her tax liability through prepay-
ments by the time she is required to decide whether to comply or
evade. Thus, it is important to choose a presumptive collection
method that minimizes the potential that the taxpayer must self-assess
her potential liability at the time of the presumptive prepayment. If
the taxpayer views the presumptive prepayment as a compliance deci-
sion that can be controlled and manipulated, then her frame (gain or
loss) becomes relevant. In that case, the payment likely will be
viewed as a loss, the taxpayer will be more likely to cheat, and no
progress will be made from a compliance perspective. The most effec-
tive presumptive collection regime, then, would be one that minimizes
the control that the taxpayer has over determining the amount of the
presumptive prepayments.
Take, for example, a presumptive collection system where the pay-
ments are calculated as a percentage of the taxpayer’s gross re-
ceipts.190 Because the reporting of gross receipts is largely under the
189 For a discussion of different types of presumptive taxes used in other countries, see
generally Amotz Morag, Some Economic Aspects of Two Administrative Methods of Esti-
mating Taxable Income, 10 Nat’l Tax J. 176 (1957); Rajaraman, note 51; Thuronyi, note 25;
see also Logue & Vettori, note 25, at 125-44 (surveying different possibilities for presump-
tive tax regimes in the United States); Richard A. Musgrave, Income Taxation of the Hard-
to-Tax Groups, in Taxation in Developing Countries, note 51, at 299-309 (providing rec-
ommendations for designing a presumptive tax regime in Bolivia).
190 Presumptive taxes based on gross receipts have been employed in other countries.
See Thuronyi, note 25, at 10.

taxpayer’s control, such a presumptive system would require self-as-
sessment at the time of each presumptive prepayment that is analo-
gous to year-end reporting.191 The reporting of gross receipts is,
essentially, as much of a voluntary compliance framework as the cur-
rent tax system, and it is unlikely much would be gained from this
approach.
The goal should be to rely as much as possible on information that
can be obtained from sources other than the taxpayer and that cannot
be easily manipulated by the taxpayer.192 Take, as another example, a
presumptive collection method that relies upon the size and location
of the taxpayer’s store to determine her presumptive tax prepayment.
Assume further that the IRS determines this information by asking
the taxpayer to provide her most recent property tax statement.
While asking a taxpayer to report her gross receipts is essentially ask-
ing her to make a compliance decision (whether or not to report the
full amount), asking a taxpayer to provide her property tax records is
likely to be viewed differently.193 Because the taxpayer cannot make
a decision regarding how much or how little to report in this case at
the time of the presumptive prepayment, the gains/loss frame would
not be relevant.
In addition to choosing a presumptive collection mechanism that
minimizes the amount of taxpayer control, the amount of the pre-
sumptive prepayments applied to each taxpayer or group of taxpayers
would also have to be determined with great care. While many pre-
sumptive tax systems “aim low” due to concerns of over-taxing those
with incomes below the average for their group,194 presumptive collec-
tion would instead aim a bit high, since the ultimate goal is to put
taxpayers in a gains frame at the time of filing. The government
should strive to strike a balance between not collecting so much tax as
to burden small business owners with liquidity concerns, and collect-
ing enough such that a majority of (or a significant portion of) the
group would have satisfied its tax liability by the end of the year. The
current wage withholding tables achieve this.195
Presumptive collection would achieve over-collection196 in a man-
ner similar to the current withholding regime. A small business’ net
191 See, e.g., id. at 11.

192 This is the goal of any presumptive tax, not just presumptive collection.
193 The taxpayer could refuse to turn over the information, in the same way a taxpayer
could refuse to file a tax return. The IRS, however, could obtain the taxpayer’s property
tax records directly from the locality, a fact of which the taxpayer is likely to be aware.
194 See, e.g., id., at 21.
195 See notes 178-81.
196 Even collecting the right amount of tax (or close to it) should have the same effect.
Like taxpayers expecting a refund, taxpayers with a zero balance have been found to be
more compliant than taxpayers with a balance due. Even taxpayers with just a small bal-
business income—gross receipts less deductible business expenses—
would first be estimated as accurately as possible. For example, net
business income could be estimated based on extrinsic factors,197
which would serve as a starting point analogous to employee wages.
The presumptive prepayments then would be adjusted for factors such
as marital status, dependents, and child care expenses, which are simi-
larly reflected on Form W-4 and taken into account to reduce with-
holding payments. Like the current withholding regime, common
deductible expenses, such as retirement account contributions, would
not be factored into the calculation of presumptive prepayments.198
As a result, most small business owners should be entitled to a refund
at year-end, as is the case for most wage earners.199
b. Extrinsic Factors Approach

Presumptive collection could be based on a standard assessment
guide, similar to the Israeli tachshiv, which would rely on extrinsic
factors. Under such a regime, businesses would be classified by a
number of identifiable factors such as industry, size, geographic loca-
tion, and price of products. For example, under the tachshiv, fruit and
vegetable stores were classified into nine categories according to their
location and products, and each category was assigned an estimate of
gross revenue.200 The stores were further divided into those that
bought products from organized suppliers and those that bought prod-
ucts from nonorganized suppliers, and a higher profit margin was as-
signed to the latter group.201 Based on these determinations, a set
amount of tax was imputed to each group.202
A presumptive collection regime using a standard assessment guide
would categorize small businesses in the same manner, and charge a
quarterly tax payment based on the presumed tax assigned to each
group. In designing the standard assessment guide, the goal of
preventing taxpayer manipulation of the presumptive prepayments
would be best achieved by relying on information that either can be
ance due, likely caused by some error or miscalculation in withholding, tend to display the
same compliance behavior as those expecting a refund; whereas those taxpayers with large
balances due demonstrate riskier behavior. See Corcoro & Adelsheim, note 34, at 11, 22-
23.
197 The extrinsic factors method is discussed further in the next Subsection.
198 See notes 178-81. The design of Form W-4 and the wage withholding tables could
serve as a guide for determining which expenses should be factored into presumptive
prepayments.
199 See note 33.
200 Yitzhaki, note 54, at 317.
201 Id.
202 The tachshiv was not strictly mandatory, as a business could submit its books and
records for review in order to a claim a lower amount. See note 54.
provided by third parties or easily verified by the government. The
external factors should also bear a reasonable relationship to the busi-
ness income. For example, the color of the paint on a restaurant’s wall
would be easily verifiable by the government but would not be a sensi-
ble extrinsic factor on which to base presumed income. On the other
hand, one might expect a larger restaurant that seats more customers
to have higher receipts than a smaller restaurant (all other things be-
ing equal).
Examples of extrinsic factors might include the address and square
footage of the business establishment, the amount of casualty insur-
ance, information on a loan application, and average service price or
product price.203 To the extent that it is absolutely necessary to rely
on the taxpayer to provide information for some of the factors, care
should be taken to choose factors that minimize the potential for tax-
payers to distort their behavior in an attempt to influence the calcula-
tion of the presumptive prepayments. For example, it is possible that
the use of “employees” as a factor is inefficient if it drives taxpayers to
hire fewer employees than they otherwise.204
Some industries would be better candidates for standard assessment
guides than others. For example, those small businesses that are part
of the “underground economy,” meaning the entire existence of the
business is concealed from the government, would not be viable candi-
dates because the IRS would not be able to obtain information re-
garding external factors.205 Examples of such businesses include
babysitting, house cleaning, eBay sellers, at-home car repair, and day
laborers.206 The use of standard assessment guides as a method of
calculating presumptive prepayments, therefore, could be limited to
those industries for which the method was feasible and advantageous,
not unlike the way that withholding and information reporting re-
quirements are limited to certain contexts where those mechanisms
are feasible. There are numerous businesses that tend to deal in cash
but do not operate underground because they require licenses and/or
real property space to operate, such as beauty salons and spas, car
203 One of the best sources of third party records is a transaction for which the taxpayer
has no incentive to under-report her assets or receipts. For example, a taxpayer is less
likely to downplay the value of her assets when applying for financing from a bank or for
casualty insurance.
204 There should be less of a behavioral response to presumptive collection, however, as
compared to true presumptive taxation, as discussed in Subsection V.B.1.

205 The existence of these businesses is concealed through methods such as dealing in
cash and operating out of individual residences. Businesses operating in the underground
economy would not be viable candidates for any method of presumptive taxation.
206 See IRS, Underground Economy in Cash Intensive Business Audit Techniques,
Guide - ch. 8 (updated Nov. 7, 2013), available at http://www.irs.gov/pub/irs-utl/cash

chapter8_210737.pdf.
washes, casinos, convenience stores, and taxi cabs.207 If standard as-
sessment guides could be made applicable to identifiable businesses
that typically rely on cash208 and are thus largely unaffected by infor-
mation reporting and withholding requirements, then significant com-
pliance advantages could be gained among a new subset of small
businesses that are not captured under current compliance
initiatives.209
c. Other Possible Approaches
Another method on which presumptive collection could be based

would be to calculate presumptive tax prepayments according to a
percentage of the value of the taxpayer’s business assets.210 A pre-
sumptive tax calculated on this basis is intended to duplicate the aver-
age rate of return that investors would expect on their assets.211 The
desirability of this approach depends largely on whether the IRS
could obtain reliable information regarding the existence of and the
value of the taxpayer’s assets.212 While real property values could be
obtained from property tax records, other assets may be harder to ver-
ify through third party records.213 Additionally, because intangible as-
207 See IRS, Beauty and Barber Shops (updated Nov. 6, 2013), http://www.irs.gov/pub/
irs-utl/cashchapter10_210745.pdf; IRS, Car Wash (updated Nov. 6, 2013), http://www.irs.

gov/pub/irs-utl/cashchapter11_211031.pdf; IRS, Coin Operated Amusements (updated
Nov. 6, 2013), http://www.irs.gov/pub/irs-utl/cashchapter13_211006.pdf; IRS, Convenience
Stores, Mini-Marts and Bodegas (updated Nov. 6, 2013), http://www.irs.gov/pub/irs-utl/
cashchapter14_211003.pdf; IRS, Taxicabs (updated Nov. 6, 2013), http://www.irs.gov/pub/
irs-utl/cashchapter17_248965.pdf.
208 For example, the IRS recommends estimating gross income for taxi cab drivers, in
part, by multiplying the average number of customers per day by the number of days
worked per year, and multiplying this amount by the driver’s entry rate, noting that
“[m]any times this data can be obtained from the transportation regulatory agency.” IRS,
Taxicabs, note 207. For car washes, the IRS recommends consulting the local water depart-
ment and local electric company for information regarding water and electricity usage.
IRS, Car Wash, note 207.
209 Those industries that are not viable candidates would remain under the regular in-
come tax.
210 Several countries have adopted presumptive taxes that are calculated on this basis,
including Argentina, Colombia, Mexico, and Venezuela. Thuronyi, note 25, at 12; see
Rajaraman, note 51, at 1111-12 (discussing asset-based presumptive taxes in various
countries).
211 Thuronyi, note 25, at 12.
212 It is possible that taxpayers would be less likely to conceal assets for this purpose, as
compared to gross receipts, because the IRS can verify the existence of tangible assets
relatively easily.
213 The reliability of taxpayer reporting of asset value would depend, in part, on whether
fair market value or cost basis was the relevant value. For example, we might expect tax-
payers to reliably report the cost basis of assets for which they plan to claim depreciation.
sets pose particular valuation problems,214 some industries (those with
fewer intangible assets) would be better candidates for this method
than others.
The presumptive collection method could also piggyback on the
current information reporting regime, by using the information re-
ported by third parties to estimate cash receipts. The IRS could then
combine the reported income and estimated cash income to arrive at
total estimated income, on which the presumptive tax payments would
be based. For example, Logue and Vettori suggest that “credit-card
reporting may be a useful method of estimating cash income.”215 If
the IRS could determine an “average ratio of cash to card receipts”216
for particular industries, it could calculate total receipts on that ba-
sis.217 It then could apply presumed profit ratios for particular indus-
tries to arrive at net income estimates based on the gross receipts
estimates. As is the case for the other presumptive collection meth-
ods, some industries may be better candidates than others for this ap-
proach, particularly if they demonstrate a fairly consistent ratio of
cash-to-credit card receipts across multiple businesses.
Another option would be to create a presumptive formula for ex-
isting small businesses based on each individual’s previous tax liabil-
ity. For example, in a given year, the sum of an individual’s
presumptive prepayments may total 105% of his prior year’s tax liabil-
ity. Alternatively, the presumptive prepayments may be some multi-
ple of the individual’s average tax liability over a period of several
years (for example, three years), or a multiple of the highest amount
of tax reported in the several preceding years. New businesses would
either be exempted from the presumptive system altogether, or be
subject to another method for computing presumptive prepayments.
Basing the presumptive prepayments on the taxpayer’s previously re-
ported tax liability would be considerably less complex than the other
methods, but the potential for abuse is higher. For taxpayers who
have consistently under-reported their income by a large margin, even
collecting more than 100% of previously reported tax may still result
in the taxpayer prepaying significantly less than her full liability for
214 Thuronyi, note 25, at 20; see also Logue & Vettori, note 25, at 136 (arguing that the
biggest drawback to an asset-based presumptive tax is that intangibles are difficult to

value).
215 Logue & Vettori, note 25, at 133. They do not advocate estimating cash from credit
card receipts for purposes of calculating a presumptive tax, but rather suggest that an audit
be triggered for taxpayers who do not report adequate cash receipts in proportion to their
credit card receipts. Id.
216 Id.
217 Presumably the IRS would have access only to credit card information reports from
the prior year, so it would have to rely on the previous year’s credit card receipts to esti-
mate cash receipts, and to calculate total receipts, for the current year.
the current year.218 To the extent that taxpayers did not claim refunds
at the end of the year, however, the IRS would collect increasing
amounts of tax from such taxpayers over time.
Given that a prior-year method creates the most potential for abuse
of the options surveyed here, this Article does not endorse such a
method for presumptive collection. The inherent difficulty in valuing
and verifying the existence of many small business assets also makes a
net asset method an unattractive approach for presumptive collection.
The most effective presumptive collection method would be one that
minimizes taxpayer control over determining the amount of the pre-
payments while maintaining a reasonable degree of accuracy. Unless
it could be established that cash receipts could be reliably estimated
based on information reporting of credit card receipts, the method
that best achieves these goals is an extrinsic factors approach.219 Such
an approach would have to be limited to those industries for which the
government could obtain reliable sources of third party information
regarding the external factors.
4. Example: Presumption Collection Based on Extrinsic Factors
Because it would not be feasible to design a standard assessment

guide that covered all types of small businesses, the IRS220 might start
by identifying those industries221 that both lend themselves to external
factors and that have particularly high rates of noncompliance.222 For
example, one study of noncompliance among thirty-four different oc-
cupations ranked “vehicle sales” as highest “both in terms of esti-
mated average level of non-compliance ($6406 [average understated
218 Additionally, it is possible that taxpayers would under-report to a greater degree in
the future in order to game the presumptive collection system going forward.
219 Unlike an approach that estimates total receipts based upon credit card receipts, an
extrinsic factors approach has been successfully implemented in other countries.

220 As suggested by Logue and Vettori, Congress could be responsible for designing a
presumptive regime or it might delegate authority to Treasury. See Logue & Vettori, note
25, at 104. For the sake of simplicity, this Article assumes the relevant governmental body
is the IRS.
221 The IRS already employs several hundred different “business or professional activity
codes” to classify small businesses on Schedule C, which might be a useful starting point.
See IRS, 2012 Instructions for Schedule C, C-9, available at http://www.irs.gov/pub/irs-pdf/
i1040sc.pdf.
222 It could be helpful to designate one or two test industries, for which presumptive
collection initially would apply for several years before imposing it more broadly. This
would be no different than singling out restaurant employees who earn tip income for
special presumptive reporting rules under the current regime. See notes 67-68 and accom-
panying text. In the case of the Israeli tachshiv, the government introduced standard as-
sessment guides for only a few new industries per year. Wilkenfeld, note 26, at 147.
tax]) and estimated share of overall taxes not paid (51.1%).”223 As a
simple example, then, assume that vehicle sales comprise a distinct
category of business that is subject to presumptive collection. Further
assume that Taxpayer is an individual that owns, as a sole proprietor,
a car dealership in Fairfax, Virginia that sells used nonluxury sedans.
A hypothetical presumptive collection system could operate as
follows.
In addition to any state law registration requirements for Tax-
payer’s business, presumptive collection could require Taxpayer to
register his business with the IRS through an online registration sys-
tem. Taxpayer would provide the IRS with information such as the
category of industry, address of his business, hours of operation, the
size of the lot, the type of cars on his lot, and the number of cars on his
lot. The IRS would match this registration with information it ob-
tained from the state with respect to Taxpayer’s registration with the
Virginia Motor Vehicle Dealer Board for an auto dealer license,224
similar to Form 1099 matching. Application for an auto dealer license
in Virginia requires, among other things, the address of the business,
proof of local zoning approval, and liability insurance for each license
plate issued.225 From these and other similar sources (such as prop-
erty records for the lot), the IRS should be able to verify the size of
Taxpayer’s business, including the average number of cars for sale.226
In addition to gathering information about Taxpayer’s individual
business, the IRS would have to design a standard assessment guide
for vehicle sales. Using data from sources such as the Department of
Commerce,227 Consumer Reports,228 state agencies,229 and private
223 Brian Erard & Chin-Chin Ho, Mapping the U.S. Tax Compliance Continuum, in Tax-
ing the Hard-to-Tax, note 23, at 179. Other top evaders include tip earners (49.8% of total
taxes not paid), farmers (33%), social and religious workers (23.4%), loggers/fishers/
hunters (23.1%), and construction workers (22.3%). Id.
224 See Motor Vehicle Dealer Bd., Commonwealth of Virginia, New Dealer Require-
ments Checklist, available at http://www.mvdb.virginia.gov/forms/files/MVDB-33.pdf.

225 Dealers are also required to own or lease the property on which the business is lo-
cated, operate at least twenty hours per week, and maintain at least 250 square feet of
enclosed office space. Id.
226 The IRS could estimate the average number of cars from the size of the lot and the
amount of liability insurance, which is required for each plate issued.

227 See, e.g., U.S. Dep’t of Commerce, Advance Report on Durable Goods Manufactur-
ers’ Shipments, Inventories, and Orders, (Mar. 2012), available at http://www.esa.doc.gov/

sites/default/files/ei/documents/2012/April/advancereportondurablegoods28march2012
29.pdf; see also U.S. Census Bureau, Wholesale & Retail Trade: Motor Vehicle Sales,
http://www.census.gov/compendia/statab/cats/wholesale_retail_trade/motor_vehicle_
sales.html (last visited Nov. 26, 2013).
228 See, e.g., Consumer Reports, Sedan Buying Guide, http://www.consumerreports.org/
cro/cars/sedans/buying-guide.htm (last updated May 2013).

229 For example, the IRS might obtain relevant sales information from records of motor
vehicle sales taxes, which are hard to evade because proof of payment is required for vehi-
companies that publish business statistics,230 the IRS could compile
information regarding the average selling price of used nonluxury se-
dans, the average number of sales per year by auto dealers, and how
that varies per region. The initial goal would be for the IRS to deter-
mine a presumed amount of gross income for a used auto dealer in
Fairfax, Virginia based on the number of cars on his lot. The IRS
would then determine a presumed amount of net income by applying
a fixed profit ratio to presumed gross income. The profit ratios would
vary by industry and would be calculated from sources such as IRS
audit data from the relevant industry.231 Finally, the IRS would apply
a presumed tax rate and determine Taxpayer’s annual presumed tax
liability.232 Taxpayer would receive a bill for this amount and be re-
quired to pay one-quarter of it each April, July, October, and
January.233
Assume that Taxpayer had net business income of $200,000 in Year
1 and was subject to a 30% effective tax rate under the regular income
tax, with a resulting tax liability of $60,000. Further assume that under
the above-described system, Taxpayer’s presumptive tax liability was
determined to be $64,000. Under presumptive collection, Taxpayer
would owe $16,000 (one-quarter of $64,000) at the end of each quarter
in Year 1. In Year 2, Taxpayer would declare $60,000 of tax liability
on his Year 1 tax return, would report $64,000 of presumptive prepay-
ments on his return, and would claim a refund of $4000. Taxpayer
would receive a market rate of interest on his refund and thus would
not be disadvantaged economically. Although Taxpayer would still
cle registration. See, e.g., Virginia Dep’t of Motor Vehicles, Motor Vehicles Sales and Use
Tax, available at http://www.dmv.virginia.gov/vehicles/#sut.html (last visited Nov. 26, 2013).
230 For example, in its audit guide for cash businesses, the IRS recommends that audi-
tors consult with www.bizstats.com, a company that provides industry financial benchmark
reports for 157 different types of sole proprietors. See IRS, note 206, at ch. 2.
231 See Musgrave, note 189, at 300 (discussing assigning unique profit ratios to different
industries in designing a presumptive tax for Bolivia); Logue & Vettori, note 25, at 129-30
(suggesting using TCMP-type audit data to design industry-specific presumed profit ratios
for modified gross receipts tax). In the preparation of the tachshiv, a staff of economists
from the Israeli taxing authority similarly would review a sampling of files from taxpayers
within a particular business. Wilkenfeld, note 26, at 148-49. Within a single industry, there
might be multiple profit ratios. See Logue & Vettori, note 25, at 130. For example, dealers
of luxury cars might have a different profit ratio than dealers of nonluxury cars, and deal-
ers in Manhattan might have a different profit ratio than dealers in Fairfax, Virginia. In-
dustry profit ratios could also be obtained from private companies that compile such data.
See note 230.
232 The amount of presumed tax could be determined from the same sources (for exam-
ple, TCMP-like audit data) as the net profit ratios. As discussed above in Subsection
V.A.3.a, the calculation of presumed tax would take into account some, but not all, per-
sonal exemptions and nonbusiness deductions, similar to withholding.
233 One way to further emulate wage withholding would be to provide taxpayers with an
option to have their presumptive prepayments automatically debited from their checking
account.
have the opportunity to evade by understating his actual income
(thereby increasing his refund), the empirical studies cited above
demonstrate that, because he would be in a risk-averse gains frame, he
is much less likely to do so.
B. Evaluating the Proposal

If presumptive collection could bring the compliance levels of small
business owners (currently near 40%) anywhere near the compliance
levels of wage earners (currently near 99%), billions of additional tax
dollars could be collected. But even if a large revenue increase gener-
ated by enhanced tax collections from small businesses would more
than pay for the administrative costs of developing and enforcing a
presumptive collection system, that is not the end of the story. The
other costs of presumptive collection—potential behavioral distor-
tions and added complexity—must be weighed against the benefits.
Additionally, even if the benefits of presumptive collection outweigh
its costs, the fairness implications of imposing this system on small
businesses must be considered.
1. Potential Costs of Presumptive Collection
a. Efficiency Costs
Although increasing small business compliance undoubtedly would
generate revenue, the gain could be offset if taxpayers deviated from
economically efficient behavior in an attempt to lessen their presump-
tive prepayment liability. But because the proposed presumptive re-
gime would be a tax collection device that had no impact on final tax
liability, the behavioral response should be smaller than the response
to a true presumptive tax. Presumptive collection would affect a small
business owner’s cash position in the short term (as is the case with
withholding), but would not have income effects that extend beyond
the tax year for which the presumptive prepayments were made.
Thus, to the extent that taxpayers understood that presumptive collec-
tion would not affect their year-end tax liability and did not distort
their behavior in response to their short-term cash position, the be-
havioral response should be minimal.234
234 A taxpayer might care about his short-term cash position if he thought he could have
earned a higher rate of return than the market rate of interest that would be paid on his tax
refund. Although such a belief could lead to behavioral distortions, this concern would be
likely to arise for only a small subset of taxpayers (who were confident they could beat the
market rate). See note 263. Further, this concern could be mitigated by relying on extrin-
sic factors that could not be easily manipulated by the taxpayer.
Consider, for example, a method that incorporates average product
prices into its calculation of the presumptive tax prepayment. In the
case of a true presumptive tax, a taxpayer might lower her prices in
order to lower her presumptive tax payment, because she might deter-
mine that the revenue lost by lowering prices would be less than the
revenue lost through increased taxes. It is significantly less plausible,
however, that a small business owner would be willing to lower her
prices to manipulate her presumptive tax prepayment if she was aware
of the fact that it would have no impact on her tax liability at the end
of the year. The end result in that case would simply be less revenue
than if she had not adjusted her prices (assuming she had already set
her prices at a revenue-maximizing level). With proper taxpayer edu-
cation regarding presumptive collection, the possibility that small bus-
iness owners would distort their behavior in inefficient ways could be
minimized.
It is possible that small business owners who understand that pre-
sumptive collection would have no impact on final tax liability would
still distort their behavior. For example, some might do so out of con-
cern about their current cash position. For taxpayers with legitimate
liquidity concerns, the presumptive collection regime would address
this problem by allowing them to appeal to the IRS for immediate
relief in exchange for a higher rate of audit.235 Other taxpayers might
distort their behavior because they believed that lower presumptive
tax prepayments would make it easier to evade their actual tax liabil-
ity.236 To the extent that taxpayers without legitimate liquidity con-
cerns distorted their behavior in response to presumptive collection,
the efficiency costs could be minimized by relying as much as possible
on factors that were either out of the taxpayer’s control, or that tax-
payers were less likely to respond to, in calculating presumptive
prepayments.237
235 This should be the case only for a small subset of taxpayers. See Subsection V.B.2.
236 Consider, for example, a taxpayer with $20,000 of actual tax liability who planned to
report only $10,000 on her tax return. She might prefer presumptive prepayments totaling
$12,000 to presumptive prepayments totaling $23,000, because she might believe that
claiming a substantially larger refund ($13,000 versus $2000) would raise a red flag with the
IRS.
237 The IRS could also minimize behavioral distortions by keeping its exact method of
calculating presumptive tax prepayments secret, akin to its Discriminant Function System
(or “DIF”). The DIF is a computer program employed by the IRS in selecting returns for
audit, which assigns each return a score that rates its potential for an underpayment based
on statistical data collected by the IRS. See IRS Fact Sheet No. FS-2006-10 (Jan. 2006),
available at http://www.irs.gov/uac/The-Examination-(Audit)-Process. The formula on
which the DIF is based is not disclosed to taxpayers and is closely guarded by the IRS.
IRC § 6103(b)(2); see also 5 U.S.C. 552(b)(3) (exemption from Freedom of Information
Act); Gillin v. IRS, 980 F.2d 819, 822 (1st Cir. 1992). In the case of presumptive collection,
b. Administration and Enforcement Costs
Developing a presumptive collection system may entail significant
costs on the part of the government.238 For example, while a prior-
year method would involve relatively minimal expense as compared to
the current regime,239 developing standard assessment guides for a
number of industries might require immense amounts of data and spe-
cialized knowledge of various industries. Further, presumptive collec-
tion would not necessarily lower enforcement costs to the same
degree as a true presumptive tax.240 While a true presumptive tax re-
gime would mean that authorities no longer had to monitor the accu-
racy of taxpayers’ self-reported income,241 in the case of presumptive
collection, the IRS would still need to police the regular income tax in
addition to the taxpayer’s presumptive prepayments.242
The level of additional administrative cost imposed by presumptive
collection would depend on the amount of information already at the
IRS’s disposal. For example, the IRS might already have a significant
amount of information needed to design a presumptive collection re-
gime based on extrinsic factors in its DIF database.243 The IRS cur-
rently uses this and other information to reconstruct income when
auditing certain individual taxpayers.244 Presumptive collection would
merely apply such techniques on a broad ex ante basis rather than a
piecemeal ex post basis. The IRS also already reconstructs taxpayer
income when filing a substitute return, using data from various
sources such as the Bureau of Labor Statistics.245 To the extent that
the IRS and other government agencies have already compiled rele-
vant information, the task of calculating presumptive tax prepayments
would be significantly less burdensome.
The cost of presumptive collection also would depend on the extent
to which administrative and enforcement costs imposed by the current
regime could be reduced. For example, the cost of presumptive col-
however, taxpayers would be aware of the extrinsic factors that they reported, which may
allow them to at least partly reconstruct the IRS formulae.
238 See, e.g., Tanzi & Casanegra de Jantscher, note 58, at iii.
239 A prior-year method of calculating presumptive tax has much in common with the
current estimated tax regime, and thus would add little cost.
241 A true presumptive tax would not necessarily eliminate the need to monitor all self-
reporting by taxpayers, since calculation of the presumptive tax might still rely on informa-
tion reported by the taxpayer.
242 Because the regular income tax would still be in place, the same accuracy and delin-
quency penalties could continue to apply. As part of its enforcement of the presumptive
collection system, there should be separate penalties for failure to comply, analogous to
the penalty under § 6654 for failure to pay estimated tax.
243 See note 237.
244 See note 208.
245 See note 72.
lection might be partly offset by reduced audit and litigation costs if
the system achieved a substantial increase in small business compli-
ance. Although auditors might still have to verify the accuracy of fac-
tors (for example, product prices or property value) that went into
calculating presumptive tax prepayments, this should involve less time
and expense than having to determine if small business owners were
hiding gross receipts.246
Ultimately, enforcement costs can be tolerated to the extent that
they generate significant revenue. If presumptive collection resulted
in a sizable increase in small business compliance, the resulting reve-
nue gain should more than pay for the extra administration costs of
presumptive collection. Indeed, the premise of any presumptive tax,
and the reason that they are so popular among countries where com-
pliance is notoriously low, is that any additional administrative and
enforcement costs imposed by the system are outweighed by the com-
pliance gains.
C. Complexity and Compliance Costs
One benefit of true presumptive taxes is that compliance costs for

taxpayers are lowered because they no longer need to maintain the
same level of records required by self-reporting.247 Presumptive col-
lection would not provide this upside because taxpayers would still
need to maintain the same level of records as under the current re-
gime to calculate their actual tax liability at year-end. It is also un-
likely, however, that presumptive collection would add significant new
compliance costs for taxpayers, nor should it add significant complex-
ity to the current regime from the taxpayer’s perspective. Because
year-end tax reporting would be the same both with and without pre-
sumptive collection, the issue would be whether the presumptive pre-
payment system was more complex or more costly than our current
estimated tax regime. There is no reason to think, however, that pre-
sumptive collection would create an undue burden for taxpayers as
compared to estimated taxes. Although there would be complexities
and costs involved in calculating presumptive tax prepayments, that
burden would fall largely on the government, as is the case with any
presumptive tax. Because taxpayers’ presumptive liability would not
affect their ultimate tax liability, they would not be required to wade
through the mechanics of how the government arrived at the amount
246My thanks to Leandra Lederman for this suggestion.

247Yitzhaki, note 54, at 321 (“[T]he major advantage of presumptive taxation is in re-
ducing the bureaucratic burden on the taxpayer.”).
of presumptive prepayments.248 And while calculating presumptive
tax prepayments may require some information reporting on the part
of the taxpayer (property values, for example), this should not be re-
quired more than once annually, and the effort should be minimal.
None of the work required on the taxpayer’s part should significantly
exceed the effort required in calculating estimated tax and making
quarterly remittances.249 In the case of presumptive collection, the
number of remittances (four) would be the same, and under an extrin-
sic factors method the government would do the calculation for the
taxpayer.250
d. Additional Compliance Benefits

Presumptive collection might generate additional compliance bene-
fits that are unrelated to framing effects. For example, collecting ex-
trinsic factors from small businesses might have a deterrent effect.
Small business owners might be less likely to evade because they be-
lieve that the IRS is monitoring them more closely or has more infor-
mation about their business.251 Taxpayers might also believe that
reporting significantly less than their presumptive tax liability creates
a “red flag” for the IRS, and the IRS might be able to use this as an
audit strategy.
In addition to improving small business compliance overall, another
potential upside of presumptive collection is that it might encourage
small businesses to keep better records than they do under the current
system. For dishonest small business owners, there is an incentive
under the current system to not keep records, or to hide or destroy
records that provide evidence of receipts that they do not intend to
report. This incentive would be mitigated by presumptive collection
in two ways. First, to the extent that the framing effects imposed by
248 Taxpayers with legitimate liquidity concerns would be able to request immediate re-
lief from presumptive prepayments. Those without liquidity concerns who still wanted to
challenge the amount of their presumptive tax could apply for an adjustment by demon-
strating that their presumptive tax exceeded their actual tax liability by more than a certain
amount (for example, 10%) over the past several years.
249 If a taxpayer elected to have his presumptive prepayment directly debited from his
account each quarter, the remittance process would be easier.

250 There are a number of ways the government could accomplish this from a logistical
perspective. For example, the government might solicit information from the taxpayer on
an annual (or less than annual basis) and submit a bill to the taxpayer for each year show-
ing her presumptive liability. Given the increasing prevalence of online recordkeeping and
return filing, the system might be implemented in a more efficient and less costly fashion
via the Internet. For example, taxpayers could be asked to enter information relating to
external factors on a specified website, which could then automatically calculate the tax-
payer’s presumptive prepayment liability.
251 I am grateful to Leigh Osofsky for this suggestion.
presumptive collection improved small business compliance, the in-
centive to conceal records would be lessened because small business
owners would be less likely to attempt to hide income. Second, tax-
payers that want to claim a refund likely would be motivated to keep
better records than they did previously so that they could justify the
difference between the amount collected through presumptive pre-
payments and the amount they claimed on their return. Better re-
cordkeeping might also create positive spillover effects that increase
compliance among even more businesses, since dishonest small busi-
ness owners might fear that records kept by their business associates
would create a paper trail that could be traced back to them.252
It is possible that presumptive collection could improve compliance
among tax professionals that prepare returns for small businesses as
well. If return preparers feel pressure to reduce their clients’ balance
due, or to help them get a refund, they may have incentives to take
overly aggressive positions on their clients’ returns or resort to out-
right cheating. To the extent that presumptive collection would al-
ready ensure a refund for those clients without resorting to dishonest
or overly aggressive return positions, return preparers might feel less
pressure to take such positions.
Finally, the increase in small business compliance resulting from
presumptive collection could also have a positive impact on non-
small-business compliance. Relying on what is known as “reciprocity
theory,” scholars have argued that tax noncompliance can create a
“chump effect,” under which otherwise compliant taxpayers reduce
their level of compliance because they perceive that others are not
complying.253 In the current context, this means that the pervasive-
ness of small business noncompliance likely impacts compliance
among other types of taxpayers. The flipside of reciprocity theory is
that taxpayers will be motivated to comply if they believe that others
are doing so. If presumptive collection can change the perception that
small business owners notoriously under-report their income, then this
should have a positive effect on actual compliance among small busi-
ness owners and wage earners alike.
2. Fairness Concerns
Even if the benefits of presumptive collection outweigh its costs in
purely economic terms, the proposal is only desirable insofar as it is an
252 For a discussion of similar spillover effects in the context of a VAT, see Dina Pomer-
anz, No Taxation Without Information: Deterrence and Self-Enforcement in the Value

Added Tax (Nat’l Bureau Econ. Research, Working Paper No. 19199, 2013), available at
http://www.nber.org/papers/w19199.
253 Blank, note 19, at 575-76.
equitable method for increasing compliance.254 Likewise, the admin-
istrative and efficiency costs of presumptive collection may be more
tolerable if presumptive collection would result in a fairer tax system.
Commentators have argued that there is an inherent unfairness in
true presumptive tax regimes because taxpayers with little or no in-
come may still end up having to pay tax.255 Similarly, horizontal eq-
uity may be compromised if taxpayers with comparable incomes end
up bearing disparate tax burdens because of differences in external
factors used to calculate the tax. Others have argued that presump-
tive taxes also violate vertical equity by placing a ceiling on the tax,
imposing a lower effective rate on those with above-average in-
comes.256 True presumptive taxes, however, can also enhance fair-
ness, to the extent that noncompliance under the current tax system
results in unequal tax burdens among similarly situated taxpayers.257
Presumptive collection escapes many of the equity issues imposed
by true presumptive taxation because the prepayments would be re-
fundable and the taxpayer’s ultimate liability would be based on her
income. Thus there would be no ceiling on the tax258 and compliant
taxpayers with similar incomes should bear similar tax burdens. Pre-
sumptive collection not only avoids these problems but should make
the current tax system more equitable. The enormous gap in small
business compliance coupled with the exceptionally high rate of com-
pliance among wage earners indicates a significant horizontal equity
problem. If Taxpayer A earns a $100,000 salary, he likely will pay tax
on the majority of this income through withholding. But if Taxpayer
B is a small business owner who earns $100,000 in net cash income, he
may pay tax on only a fraction of this income, choosing to conceal the
rest from the IRS. The differences in enforcement between these two
groups results in wage earners bearing more of the tax burden as a
group, and all compliant taxpayers likely paying tax at a higher rate as
a result.259 One of the most important benefits of presumptive collec-
254 Cf. Noël B. Cunningham and Deborah Schenk, The Case for a Capital Gains Prefer-
ence, 48 Tax Law Rev. 319, 353 (1993) (“[R]evenue should be raised in as fair and efficient
a manner as possible. Therefore, even if a preference for capital gains raised revenue, we
would evaluate the preference on efficiency and equity grounds.”).
255 See, e.g., Thuronyi, note 25, at 429; see also Bird & Zolt, note 25, at 1687.
256 See Yitzhaki, note 54, at 321.
257 See Thuronyi, note 25, at 429.
258 See Subsection V.B.3 for a discussion of whether presumptive collection places an
effective ceiling on small business taxes.

259 It is possible that noncompliant small business owners are no better off financially
than taxpayers in other sectors, if their lower effective tax rate is simply incorporated into
lower prices and a lower rate of return. If that is the case, then those harmed the most are
compliant small business owners, who presumably have trouble competing in such an envi-
ronment. For a discussion of the economics of evasion in the cash economy, see generally
Bankman, note 87, at 506-08.
tion would be that, in addition to improving small business noncompli-
ance, it would reduce this inequity between small business owners and
those subject to withholding. By essentially subjecting small business
owners to a system of prepayments that operates much like wage
withholding, the tax burdens of the two groups should fall more
closely in line.
Even if presumptive collection makes the tax system more equita-
ble, some might argue that it is unfair to impose a mandatory system
like presumptive collection on small business owners, particularly
when the aim is to over-collect tax. Such critics may argue that pre-
sumptive collection amounts to a forced loan to the government. But
withholding on wages also can be analogized to a loan to the govern-
ment260 to the extent that wage earners incur significant deductible
expenses (for example, medical expenses) that are not reflected in
withholding allowances.261 In either case, the “loan” is only the
amount that was over-collected, which will be a fraction of total tax
liability.262 Further, although tax would be over-collected in many
cases under presumptive collection, small business owners would be
fully compensated through a market rate of return on their refunds.
Thus, at the very least, most would be made whole,263 and to the ex-
tent they would not otherwise have earned a return on the refunded
money, certain taxpayers would be better off. Further, even though
260 See, e.g., Richard L. Doernberg, The Case Against Withholding, 61 Tex. L. Rev. 595,
623 (1982).
261 A wage earner who incurs significant deductible expenses may end up with a tax
refund even if he attempts to adjust his withholding. First, exact matching on Form W-4 is
difficult. Second, some deductible expenses may be unexpected. Although taxpayers can
adjust their withholding by filing a new Form W-4 mid-year, changes to withholding are
prospective only and previously over-withheld amounts cannot be recovered before the
end of the year. See IRC § 3402(f)(3)(B); Jeremy R. Polk, Compensation for the Fruit of
the Fund’s Use: The Takings Clause and Tax Refunds, 98 Nw. U. L. Rev. 657, 668-69
(2004).
262 Consider the hypothetical car dealer from Subsection V.A.4. who had $60,000 of
actual tax liability and $64,000 of presumptive tax liability. Although $64,000 was collected
through presumptive collection, only $4000 of this amount could be considered a loan be-
cause the remaining $60,000 represents tax for which he is liable. He would receive inter-
est on the $4000 overpayment.
263 The taxpayer who would not made whole would be one who would have earned an
above-market return on the amount that was refunded to him had it not been collected on
a presumptive basis. But in light of the popularity among taxpayers of receiving tax re-
funds, see note 273, it seems unlikely that a significant portion of taxpayers would other-
wise invest those funds at a market rate of return. Adding that to the difficulty of earning
an above-market rate of return, it is likely that only a very small subset of taxpayers would
not be made whole with a market rate of interest. Even in the case of more sophisticated
investors, those with a well-balanced portfolio would likely own some amount of short-
term government debt, and they could simply rebalance their portfolio to account for this
short-term loan. See Gregg D. Polsky, Reconstructing the Individual Mandate as an Es-
crow Account, 109 Mich. L. Rev. First Impressions 73, 74 (2011).
employees have more control over the amount of withholding through
elections on Form W-4 than small business owners would have under
presumptive collection, employees generally do not receive interest on
tax refunds.264 In this sense, small businesses would be better off than
their wage-earning counterparts.265
Finally, as discussed above, a presumptive collection regime should
also have exemptions for small business owners who would experience
liquidity problems from making presumptive prepayments. These
concerns are not unique to presumptive collection: A small business
that found itself facing unexpected losses and unable to afford its pre-
sumptive prepayments is in the same situation as a wage earner who
faces unexpected deductible expenses during the year (for example,
medical expenses) and finds that his paychecks are no longer suffi-
cient due to overwithholding.266 In any event, liquidity concerns
should affect only a small subset of businesses—those who earned in-
come that was significantly lower than their presumed income and
who also could not accommodate the additional tax prepayments even
though they would be refunded.267 To alleviate liquidity issues, tax-
payers below a certain income threshold should be exempted from
presumptive collection altogether. For those not below the threshold,
a mechanism would be put in place under which small business owners
could apply to the IRS for mid-year relief from presumptive prepay-
ments (analogous to amending one’s Form W-4). To dissuade abuse,
the act of appealing for immediate relief from presumptive prepay-
ments would raise the taxpayer’s chance of being audited by the IRS.
3. Addressing Potential Objections to the Proposal

Even if presumptive collection is an equitable and efficient method
to increase small business compliance, critics may object to the theo-
264 IRC § 6611(e) (no interest on refund issued within forty-five days of filing of return).
265 Although wage earners do not receive interest, see id., they have some control over
the amount of tax they prepay through their elections on Form W-4. See note 178. Small
businesses subject to presumptive collection should be treated differently with respect to
interest because they would not be able to elect to reduce their presumptive prepayments.
266 A wage earner can adjust his withholding mid-year, but cannot obtain a refund of
any previously overwithheld tax until the end of the year. See note 261. Consider, for
example, a taxpayer that earns a $30,000 salary and suffers $30,000 of deductible expenses
mid-year. See, e.g., Doernberg, note 260, at 623.
267 Consider again the hypothetical car dealer in who had $60,000 of actual tax liability
and $64,000 of presumptive tax liability, payable in four installments of $16,000. The car
dealer would be responsible for an average of $15,000 of tax per quarter under the normal
income tax. For there to be a legitimate liquidity concern caused by presumptive collec-
tion, it would have to be the case that the car dealer’s expenses and income were so closely
aligned that the additional $1000 per quarter of presumptive tax would mean that he could
no longer manage his affairs. This should be an unlikely outcome for most taxpayers.
retical underpinnings of the proposal. Additionally, they may argue
that the proposal has practical limitations that cannot be overcome.
Some of these potential concerns are addressed in this Subsection.
One possible objection to presumptive collection is that any system
that exploits cognitive errors or biases is undesirable.268 Prospect the-
ory predicts that individuals do not act rationally when making tax
compliance decisions, as there is no economic reason that the pres-
ence of a refund or a balance due should have an impact on taxpayer
behavior. Critics may argue that presumptive collection seeks to ex-
ploit this irrationality by manipulating the frame of reference that
small business owners will face at the time they make a tax compli-
ance decision, and that there is something undemocratic or paternalis-
tic about that exploitation.269
It is not immediately clear why designing a mechanism that seeks to
increase tax compliance by exploiting cognitive error is inherently bad
policy. First, governments already employ this practice to encourage
compliance with the law in other contexts.270 And even if there is
something undesirable about exploiting cognitive error, governments
routinely make trade-offs in accomplishing legitimate policy objec-
tives if the benefits are found to outweigh the downsides.271 It is also
possible that reliance upon framing effects to design a tax collection
mechanism is not inherently undesirable, since individuals may actu-
ally prefer policies that are geared towards their cognitive prefer-
ences.272 For example, even when individuals are aware that it would
be economically irrational to overpay their taxes through withholding,
studies reveal that most prefer to receive a refund when they file.273 It
is possible that small business owners would similarly prefer to receive
268 See, e.g., McCaffery, note 165, at 1943.

269 See, e.g., Deborah H. Schenk, Exploiting the Salience Bias in Designing Taxes, 28
Yale J. Reg. 253, 287-89 (2011) (addressing the argument that the intentional use of low-
salience taxes is undemocratic); Joshua D. Blank, In Defense of Individual Tax Privacy, 61
Emory L.J. 265, 338-39 (2011) (addressing the argument that exploiting individuals’ poten-
tially inaccurate perceptions of tax enforcement is paternalistic).
270 Blank, note 269, at 339 (describing examples of governments exploiting cognitive
biases through signs posted in public parks and on highways); see also Richard H. Thaler &
Cass R. Sunstein, Nudge: Improving Decisions About Health, Wealth, and Happiness 108-
09 (2d ed. 2009) (describing how automatic enrollment in retirement plans has improved
savings).
271 Schenk, note 269, at 287; see McCaffery, note 165, at 1943.
272 Schenk, note 269, at 289.
273 See, e.g., Benjamin C. Ayers, Steven J. Kachelmeier & John R. Robinson, Why Do
People Give Interest Free Loans to the Government? An Experimental Study of Interim
Tax Payments, J. Am. Tax’n. Ass’n, Fall 1999, at 55; Donna D. Bobek, Richard C. Hatfield
& Kristin Wentzel, An Investigation of Why Taxpayers Prefer Refunds: A Theory of
Planned Behavior Approach, J. Am. Tax’n Ass’n, Spring 2007, at 93, 109 (empirical study
finding that the majority of subjects enjoyed receiving a tax refund and would not adjust
withholding even when given opportunity to do so); David Gamage & Darien Shanske,
tax refunds. And unlike the preference for refunds among wage earn-
ers, such a preference would not be irrational for small business own-
ers because refunds would bear market interest,274 making the
argument that presumptive collection exploits taxpayer irrationality
less compelling.
Further, any compliance initiative enacted by Congress through a
legitimate legislative process cannot be said to be undemocratic.275
Although presumptive collection is based on framing effects that indi-
viduals may not be aware of, the general operation of presumptive
collection would not be hidden from voters, particularly since small
business owners would be informed at the beginning of each year of
their presumptive liability.276 Nor would presumptive collection re-
strict the choices of small business owners in a meaningful way. Al-
though taxpayers would be bound to make presumptive prepayments,
they would continue to have control over how much taxable income
they reported at year-end, effectively leaving the ultimate decision
whether to comply up to the taxpayer.277 And even if presumptive
collection did exert significant influence over taxpayers’ decisions—
causing them to comply when they otherwise might have evaded—
their freedom will not have been compromised because individuals
are not entitled to break the law.278
Three Essays on Tax Salience: Market Salience and Political Salience, 65 Tax L. Rev. 19,
42 (2011).
274 A rational actor should be neutral between making a tax payment or receiving a
refund with interest (putting aside liquidity concerns and assuming a constant rate of re-
turn). Thus although there would be no rational reason to prefer a refund, it would not be
economically irrational to affirmatively opt for one, as many have argued is the case with
wage-earners. See sources cited in note 273.
275 See Schenk, note 269, at 285-86 (discussing legislative transparency with respect to
low-salience taxes).
276 Nor would the theory behind presumptive collection—the link between framing and
compliance—be hidden from voters. The legislative history could make Congress’s intent
publicly accessible to anyone that chose to understand it, including the intent to over-
collect tax. Further, keeping the precise calculation of presumptive income secret—akin to
the scoring in the IRS’s DIF database—would not make the entire system undemocratic.
The government has a legitimate interest in protecting its privacy to aid certain aspects of
law enforcement (as evidenced by the exemption for the DIF under the Freedom of Infor-
mation Act), and in this case the compliance advantages of keeping the calculations secret
(preventing taxpayer manipulation of the extrinsic factors involved) would outweigh the
costs. See sources cited in note 237. In any event, taxpayers would gain a general sense of
the relevant extrinsic factors involved when they had to supply information to the govern-
ment regarding those factors.
277 The approach of presumptive collection, to influence (but not control) taxpayers’
decisions whether to comply with their tax obligations, is supported by Thaler and Sun-
stein’s “libertarian paternalism” theory, which posits that it is legitimate for policymakers
to influence—but not restrict—behavior to improve individual welfare. See Thaler & Sun-
stein, note 270, at 5; see also Blank, note 269, at 339.
278 See Blank, note 269, at 339.
It might also be argued that forcing small business owners to make
presumptive prepayments is indicative of the government’s mistrust in
taxpayers, which could have a negative effect on taxpayer morale and
voluntary compliance. It could be argued that this disrespect of tax-
payers is further exacerbated by aiming to collect more than their ac-
tual tax liability. This objection, however, would be relevant to almost
all aspects of our voluntary compliance system. The government al-
ready imposes mechanisms, such as withholding and information re-
porting, that demonstrate our mistrust that individuals will report
their tax liability accurately.279 Compelling taxpayers to comply with
the law is an integral part of administering our tax system. If we were
hesitant to design compliance mechanisms that made taxpayers feel
like they could not be trusted, our voluntary compliance rate would
likely be significantly lower than what we have today. Overall, with-
holding has had a positive impact on compliance among wage earn-
ers,280 and the downside of any mistrust created through collecting tax
in this manner is greatly outweighed by its benefits. Presumptive col-
lection seeks the same result for small businesses.
Another potential objection is that, regardless of Congress’ intent in
instituting presumptive collection, in practice the presumptive prepay-
ments would amount to a ceiling on the tax that is collected from
small businesses. The idea is that, to the extent that a small business
owner’s true tax liability was greater than the amount of tax she pre-
paid through presumptive collection, she would fail to report any ad-
ditional tax. It is possible, perhaps even likely, that this is the result in
other contexts in which presumptive-type taxes are employed, such as
the rules requiring large restaurants to report 8% of their gross re-
ceipts as tip income for their employees.281
In the current context, there would inevitably be some individuals
who do not satisfy their entire tax liability through presumptive pre-
payments. To the extent the presumptive prepayments are calculated
for groups of businesses, the above-average earners would fall into
this category.282 In these cases, dishonest individuals who otherwise
279 Like presumptive collection, withholding also results in over-collecting tax for the
majority of taxpayers. See note 33. Because taxpayers have more control over preventing
refunds from withholding than they would with presumptive collection, taxpayers might
view presumptive collection as a significantly greater imposition than withholding.
280 See note 10 and accompanying text.
281 See note 68 and accompanying text. In informal conversations with acquaintances in
the restaurant service industry, I have been told that “no one” reports tip income beyond
the presumptive 8% reported by restaurant employers. If this has become a norm among
restaurant servers, it is possible that many do not even realize they have an obligation to
report actual tip income in excess of the amount reported by their employers.
282 This would also be the case for taxpayers whose income was not adequately captured
by the external factors, even if they were not above-average earners.

would under-report on their tax returns likely would continue to do so
under the proposed system if they found themselves facing a loss
frame at the time of filing.283
But otherwise honest taxpayers would not necessarily fail to report
the balance that they owe just because they prepaid some portion of
their taxes on a presumptive basis. These taxpayers can be distin-
guished from restaurant servers, who might report only the amount of
tip income that is reflected on the Form W-2 that they receive from
their employer and fail to report any excess. In most other contexts,
Form W-2 does provide an employee’s full amount of taxable employ-
ment income. Thus in the unique case of restaurant servers, there is
potential for confusion and for failure to keep adequate records of
additional income. Small business owners, on the other hand, are not
able to rely on an employer to report income for them. Their obliga-
tions to record and report business income would remain unchanged
under the presumptive collection system, and through proper publicity
and education, the IRS could ensure that taxpayers were aware of
their obligation to continue to report their full tax liability.
This still leaves the issue of the dishonest taxpayers who might fail
to report additional tax owed after making presumptive prepayments.
If the presumptive collection system collects sufficient tax from a large
portion of taxpayers, and if a significant number of honest taxpayers
continue to be honest, then this problem should exist for only a minor-
ity of small business owners. For those taxpayers, we essentially
would have shifted to a true presumptive tax system. And if the pre-
sumptive prepayments were higher than the amount of tax that other-
wise would have been reported by dishonest small business owners,
then the system would still raise additional revenue, the amount of
which could be significant.284 Given the current state of affairs in the
context of small business compliance, this is a fairly benign worst-case
scenario.
Another potential objection to presumptive collection is that for
some types of small businesses in which individuals engage, it would
283 It is possible, however, that taxpayers in this situation would report some tax liability
at year-end, even if it were less than the full amount due, to avoid raising a red flag. Given
the imprecision inherent in any presumptive method of calculating tax liability, taxpayers
who repeatedly claim a zero balance on their return likely would raise suspicion in the eyes
of the IRS.
Additionally, if presumptive collection reduces the balance that would otherwise be due
(that is, the taxpayer faces a smaller loss), compliance might improve in some cases be-
cause the diminishing sensitivity to losses predicted by prospect theory indicates that larger
losses lead to riskier behavior. See notes 127-28 and accompanying text.
284 If it is true that restaurant servers generally do not report tip income in excess of the
presumptive 8%, then we have shifted to a true presumptive tax in that context, as well, a
result that presumably Congress anticipated.
be infeasible to calculate a presumptive tax. As discussed above, it is
certainly true that some industries would be better candidates for pre-
sumptive collection than others. For instance, certain businesses may
have few tangible assets or other verifiable qualities on which the IRS
could base estimates of gross income. An example might be an actor,
who alternates between successful and unsuccessful years, and whose
primary business asset is his acting talent. In these situations, pre-
sumptive collection is not likely to be a feasible option. Presumptive
collection also would not be able to capture small businesses that were
able to shield their entire existence from the government because, for
example, they operate out of a residence and are not properly
licensed.285
But presumptive collection does not have to be a universal solution
to the small business noncompliance problem. Presumptive collection
instead should be viewed as an additional method to combat small
business noncompliance that could and should coexist with other en-
forcement mechanisms like information reporting.286 If presumptive
collection could be applied to those identifiable cash-based businesses
that are viable candidates, it could fill significant gaps in small busi-
ness tax compliance.
VI. CONCLUSION
The lack of progress over the past decade in reducing the tax gap,
particularly the failure to reduce the rate of noncompliance among the
self-employed, demonstrates the need for innovative approaches. Re-
cent increases in information reporting likely will have an impact in
some areas, but the elephant in the room is the cash economy. Given
the limitations of standard deterrence techniques, prospect theory has
much to offer in improving compliance among small business owners
for whom information reporting or withholding is not feasible.
The application of prospect theory to taxpayer behavior— particu-
larly the existence of framing effects in taxpayer compliance deci-
sions—is widely supported by a number of empirical studies of
laboratory subjects and by studies of IRS data on hundreds of
thousands of actual taxpayers. Given that taxpayers behave predict-
ably, displaying higher compliance when faced with a gains frame and
lower compliance when faced with a loss frame, policymakers should
utilize this knowledge to their advantage to target small business non-
compliance. Not only would applying this understanding of taxpayer
285 See note 205 and accompanying text.
286 To the extent there is overlap, there could be some instances where presumptive
collection was more efficient than information reporting, in which case information report-
ing could be scaled back, and vice versa.
168 TAX LAW REVIEW
behavior make significant headway in reducing the tax gap, but it

would reduce the current inequity caused by the disparity in compli-
ance between wage earners and small business owners.
Although a number of economists and psychologists have argued
that prospect theory is a relevant predictor of taxpayer behavior, this
Article offers the first concrete proposal for using prospect theory to
increase tax compliance where information reporting and withholding
are not viable. Presumptive collection achieves this by reframing the
tax compliance decision for small business owners so that, like wage
earners, they will be in a gains frame when they decide how much tax
to report. By relying on external factors to collect tax throughout the
year, rather than to calculate year-end tax liability, presumptive col-
lection would increase compliance among small business owners in a
manner that is fairer and more efficient than true presumptive tax re-
gimes. While presumptive collection is not a universal solution, it can
be another important tool aimed at reducing the tax gap, which would
complement current compliance initiatives. If presumptive collection
could be applied to groups of identifiable cash businesses that are cur-
rently able to avoid information reporting, then policymakers could
make significant gains in reducing small business tax noncompliance.

SSRN-id2316233 Presumptive Collection - A Prospect Theory Approach To Increasing Small Business Tax Compliance

Uploaded by

Copyright:

Available Formats

SSRN-id2316233 Presumptive Collection - A Prospect Theory Approach To Increasing Small Business Tax Compliance

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

SSRN-id2316233 Presumptive Collection - A Prospect Theory Approach To Increasing Small Business Tax Compliance

Uploaded by

Copyright:

Available Formats

Presumptive Collection: A Prospect

Theory Approach to Increasing Small

* Assistant Professor of Law, University of North Carolina School of Law. I am

(2013), available at http://www.cbo.gov/sites/default/files/cbofiles/attachments/44172-Base-

Electronic copy available at: http://ssrn.com/abstract=2316233

come is subject to withholding, with self-employed individuals, whose income generally is

estimated 61% of individuals engaged in under-reporting. See id.; Gov’t Accountability

wages to the IRS. Id.

Electronic copy available at: http://ssrn.com/abstract=2316233

Is Information Reporting Warranted?, 78 Fordham L. Rev. 1733, 1752-53 (2010) (arguing

http://www.irs.gov/pub/tas/arc_2007_vol_1_cover_msps.pdf; see also IRS, Overview, note 3,

may lead to an oversupply of small, cash-based businesses.

(that is unreported income) for over 100 countries).

Effect on Reporting Compliance, in Recent Research on Tax Administration and Compli-

The presumptive collection proposal advanced in this Article draws

A. Presumptive Taxation Outside the United States

A number of countries currently use, or have previously used, pre-

note 25, at 401.

Policy Handbook 258, 258 (Parthasarathi Shome ed., 1995).

51 See Indiria Rajaraman, Presumptive Direct Taxation, Lessons from Experience in

at 423-24; Shlomo Yitzhaki, Cost Benefit Analysis of Presumptive Taxation, 63

B. Forms of Presumption in the United States

59 Manos Matsaganis & Maria Flevotomou, Distributional Implications of Tax Evasion

cable, its importance has declined in recent years. Id.

68 IRC § 6053(c)(3)(C), Reg. § 31.6053-3(h)(2)(iii)(A), (B).

http://dor.wa.gov/content.FindTaxesandRates/BAndoTax/#Whatis. Although the taxpayer

A. The Traditional U.S. Approach to Reducing Small Business

cantly Increase Revenues by Better Targeting Enforcement Resources 1 (2012), available

Committee, the GAO recommended improving sole proprietor noncompliance by ex-

Jantscher, note 58, at 7.

small businesses in each industry. Id.

ual “adjustments, deductions, exemptions,” compared to $122 billion to understated busi-

IV. PROSPECT THEORY AS AN ALTERNATIVE BASIS FOR PRESUMPTION

A. Background on Prospect Theory

2. Risk Seeking, Declining Sensitivity, and Loss Aversion

B. The Relevance of the Taxpayer’s Prepayment Position Under

preferable to a sure loss of $900 in this scenario.

Studies have shown that individuals tend to simplify prospects by

of y with probability q is expressed as (x, p; y, q). A 100% chance of x would be expressed

Because taxpayers faced with losses tend to be risk-seeking, the tax-

V($0, 0.97; −$3500, .03) = p(.03)n(−$3500) = −$239144

Since Taxpayer receives a gain of at least $500 in all events,146 the

Because taxpayers tend to be risk-averse when faced with gains, Tax-

V($3500, 0.97; $0, .03) = p(.97)n($3500) = $1105147

Taxpayer’s prepayment position, therefore, has a determinative im-

[S]ubjects facing the prospect of an additional tax payment

In another study, when subjects were asked if they would claim an

[A]t a statistically significant level, holding a balance-due

V. PRESUMPTIVE COLLECTION: A PROSPECT-THEORY-BASED

A. Reframing Taxes for Small Business Owners

1. The Presumptive Collection Proposal

Given the near impossibility of imposing a comprehensive third

Foreign countries that require withholding on certain business-to-business transactions

nesses (that is sole proprietorships), not corporations or businesses operated as pass-

would apply. See IRC § 6621.

2. Presumptive Collection Versus Estimated Taxes

payer’s presumptive prepayments, see Subsection V.A.4.

3. Choosing a Presumptive Collection Method

a. Guiding Principles in Designing a Presumptive Collection