[JPRT, 105th Congress]
[From the U.S. Government Printing Office]

                        [JOINT COMMITTEE PRINT]

                              TAX AMNESTY


                         Prepared by the Staff

                                 of the

                      JOINT COMMITTEE ON TAXATION


                            JANUARY 30, 1998


                     U.S. GOVERNMENT PRINING OFFICE
 46-030                    WASHINGTON : 1998                        JCS-2-98

                For sale by the U.S. Government Printing Office
   Superintendent of Documents, Mail Stop: SSOP, Washington, DC 20402-9328

                      JOINT COMMITTEE ON TAXATION

                      105th Congress, 2nd Session
               SENATE                               HOUSE
WILLIAM V. ROTH, Jr., Delaware,      BILL ARCHER, Texas,
  Chairman                             Vice Chairman
JOHN H. CHAFEE, Rhode Island         PHILIP M. CRANE, Illinois
CHARLES GRASSLEY, Iowa               WILLIAM M. THOMAS, California
MAX BAUCUS, Montana                  FORTNEY PETE STARK, California

                    Kenneth J. Kies, Chief of Staff
              Mary M. Schmitt, Deputy Chief of Staff (Law)
      Bernard A. Schmitt, Deputy Chief of Staff (Revenue Analysis)

                            C O N T E N T S

Introduction.....................................................     1

 I. Summary of Findings...............................................2

II. Overview and Background...........................................4

        A. Overview of Tax Amnesty...............................     4

        B. Historical Background on the Federal Government and 
            Tax Amnesty..........................................     5

        C. Elements of Present Law That Are Similar to Tax 
            Amnesty..............................................     6

III.Economic Perspectives on Tax Amnesties............................9

        A. Economic Theory of Tax Evasion and Tax Amnesties......     9

        B. Evidence on Efficacy of Tax Amnesties.................    15

IV. State and Foreign Experience With Tax Amnesty....................18

        A. State Experience With Tax Amnesty.....................    18

        B. Foreign Experience With Tax Amnesty...................    31

 V. Design Parameters of a Possible Federal Tax Amnesty..............35

VI. Conclusions Concerning Possible Use of a Federal Tax Amnesty.....38

Appendix. Estimated Budget Effects of Possible Federal Tax 
  Amnesty Proposals..............................................    39


    This pamphlet,\1\ prepared by the staff of the Joint 
Committee on Taxation, discusses tax amnesty. It was prepared 
in response to a request dated April 15, 1997, from Congressman 
Bill Archer, Chairman of the Joint Committee on Taxation during 
the First Session of the 105th Congress.
    \1\ This pamphlet may be cited as follows: Joint Committee on 
Taxation, Tax Amnesty (JCS-2-98), January 30, 1998.
    Part I of the pamphlet provides a summary of findings; Part 
II provides an overview and background; Part III provides 
economic perspectives on tax amnesties; Part IV discusses State 
and foreign experience with tax amnesty; Part V discusses 
design parameters of a possible Federal tax amnesty; and Part 
VI presents conclusions concerning possible use of a Federal 
tax amnesty. The Appendix presents the estimated budget effects 
of possible Federal tax amnesty proposals.

                         I. SUMMARY OF FINDINGS

           The staff of the Joint Committee on Taxation 
        estimates that a Federal tax amnesty would result in a 
        net revenue loss to the Federal Government, as 
        presented in the Appendix (the revenue table). This net 
        revenue loss occurs primarily because a Federal tax 
        amnesty is estimated to have the long-run effect of 
        reducing overall taxpayer compliance with Federal tax 
        laws. However, a form of amnesty could be designed to 
        contain the amount of the revenue loss to relatively 
        modest levels.
           Present law contains certain provisions that 
        are similar to elements of tax amnesty programs. These 
        provisions include the installment agreement provisions 
        and offers in compromise. In contrast to a tax amnesty 
        program, however, the present-law provisions are 
        tailored to the situations of specific taxpayers, 
        rather than providing broad relief from past taxes for 
        noncompliant taxpayers. The Internal Revenue Service 
        has substantially increased the use of these provisions 
        in recent years.
           A taxpayer's decision whether to evade taxes 
        depends in part on the resources and efforts that the 
        Government expends on the enforcement of existing laws. 
        While an amnesty may be one way to obtain taxes that 
        have not been paid and to bring new taxpayers into the 
        tax system, increasing resources devoted to enforcement 
        and increasing the penalties for failure to comply with 
        the tax laws are other ways to achieve the same 
           Economists generally believe that taxpayers 
        choose a level of compliance with the tax laws by 
        weighing the tradeoff between compliance and evasion 
        and choosing the level of compliance that will lead to 
        the highest expected level of net benefits. Among the 
        factors that will influence a taxpayer's decision 
        between compliance and evasion are the probability of 
        the evasion being detected through audit, the back 
        taxes and civil and criminal penalties that will be 
        imposed if evasion is detected, the taxpayer's ethics 
        or degree of honesty, damage to the reputation of the 
        taxpayer if the evasion is detected, the taxpayer's 
        level of ``risk aversion,'' and the perceived benefits 
        derived from a successful evasion of taxes.
           In the standard economic model of tax 
        evasion, the creation of an amnesty program alone would 
        have no effect on taxpayer behavior because the 
        taxpayer has chosen a level of compliance based on a 
        rational weighing of the benefits and costs of evasion. 
        If these costs and benefits are not changed, the 
        taxpayer's behavior will not change. If, however, the 
        penalties for evasion are increased or the likelihood 
        of detection through audit increases, then the rational 
        taxpayer may choose to take advantage of the amnesty 
        since the expected costs of previous evasion have been 
           State experience with tax amnesty will not 
        necessarily parallel the experience that could be 
        expected at the Federal level. First, many sources of 
        State tax revenues do not have Federal counterparts. 
        Second, pre-amnesty State tax enforcement efforts have 
        generally been less rigorous than efforts at the 
        Federal level. Thus, State amnesties that are coupled 
        with increased enforcement efforts might be expected to 
        be relatively more successful than a Federal amnesty 
        because past evasion is likely to have been more common 
        with respect to State taxes and taxpayers will view the 
        increased enforcement efforts of the States as 
        increasing the costs of State tax evasion. Further, 
        many State amnesties included accounts receivable in 
        their amnesties, which could be viewed as revenues that 
        the State was likely to collect in any event.
           The experience of foreign governments with 
        amnesty programs cannot generally be compared to the 
        experience that might be expected in the United States. 
        In most foreign countries, a larger portion of the 
        national economy escapes the tax system than in the 
        United States. Most foreign countries have larger 
        underground economies. In addition, many countries 
        exempt transactions occurring outside the country from 
        taxation, creating an incentive to move untaxed profits 
        outside the country. Few foreign countries have the 
        level of tax enforcement that the United States has at 
        the Federal level, especially with regard to the use of 
        computer technology and requirements of withholding and 
        information reporting.

                      II. OVERVIEW AND BACKGROUND

                       A. Overview of Tax Amnesty

Types of amnesty
    There are theoretically several types of tax amnesty 
programs. The narrowest form of amnesty would require taxpayers 
to pay all taxes, interest, and civil penalties, but would 
forgive criminal penalties. The goal of this form of amnesty 
(as well as the variants of it described below) is both to 
collect taxes owing from prior years and to place on the tax 
rolls those who had previously escaped taxation. A broader form 
of amnesty would require taxpayers to pay all taxes and 
interest due, but would forgive all civil and criminal 
    Another form of amnesty would require taxpayers to pay all 
(or a portion of) past taxes, but would forgive all (or a 
portion of) the interest \2\ due on those taxes. In addition, 
all civil and criminal penalties would be forgiven.
    \2\ This form of amnesty highlights two differing views of the role 
of interest in the tax system. Some view charging interest on past due 
taxes as a penalty. Others view interest provisions as reflecting the 
time value of money. Accordingly, absent the requirement to pay 
interest, a taxpayer would prefer to delay paying taxes for as long as 
possible so as to retain the use of the money for as long as possible.
    The broadest form of amnesty would forgive all past taxes, 
interest, and civil and criminal penalties. The goal of this 
type of amnesty is not to collect taxes owing from prior years, 
but to place on the tax rolls for the future those who 
previously had escaped taxation. This type of amnesty has not 
been attempted in the United States.
    Most of the amnesty programs that have been operated by the 
States forgive both civil and criminal penalties; these 
programs have differed as to whether all or a portion of the 
interest due was forgiven. All have required payment of at 
least a portion of the past taxes.
Eligibility for amnesty
    Amnesty programs can differ as to whether only nonfilers 
may participate, or whether individuals and entities that filed 
returns, but also either underreported income or overstated 
deductions or credits, may participate.
    Amnesty programs can also differ as to the extent to which 
known tax evaders can participate in the amnesty. Individuals 
or entities under active criminal investigation or prosecution 
are generally not permitted by the States to participate in 
amnesty. Amnesty programs differ as to whether individuals or 
entities under audit or administrative investigation are 
permitted to participate. Amnesty programs also differ as to 
whether persons in accounts receivable status can participate 
in the amnesty. Persons in accounts receivable status are those 
for whom the audit and administrative processes have been 
completed and it has been determined that the taxpayer owes 
additional money, but the taxpayer has not yet paid the 
additional amounts determined to be owing.

   B. Historical Background on the Federal Government and Tax Amnesty

    The Federal Government has never legislatively instituted a 
program that provided amnesty from interest and civil and 
criminal penalties for taxpayers who both voluntarily disclosed 
that they had underpaid their taxes and then paid those 
    \3\ The Current Tax Payment Act of 1943 initiated the still-
utilized system of income tax withholding by employers from employees' 
wages. Prior to this Act, taxpayers made lump sum payments of taxes. 
During the first year of wage withholding, taxpayers would be saving 
amounts to pay the previous year's taxes as well as becoming subject to 
wage withholding. To provide for first-year transition to the wage 
withholding system, the 1943 Act also provided for the cancellation of 
75 percent of the lower of either 1942 or 1943 tax liability. The 
remaining 25 percent was to be paid in two installments, on March 15, 
1944, and March 15, 1945. See Tax Notes, September 1, 1997, pp. 1241-
1244. This was not, however, a tax amnesty, but was instead a rate 
reduction applicable to all taxpayers. A distinguishing characteristic 
of an amnesty is that it treats similarly situated taxpayers 
differently, depending on whether they had previously voluntarily 
complied with the tax laws and fully paid their taxes, or not.
    The Internal Revenue Service (IRS) had an administrative 
policy, officially discontinued in 1952, that in effect 
provided amnesty from criminal prosecution (but not from civil 
penalties or interest) for taxpayers who voluntarily disclosed 
that they had underpaid their taxes. This program officially 
ended (though it may have continued informally) for a variety 
of reasons: charges of abuse, a desire to prevent use by 
organized crime figures, uneven administration of the tax law, 
and failure to pay the taxes once amnesty had been granted. In 
1961, the IRS issued a news release suggesting to taxpayers 
that, since the IRS was then installing new data processing 
equipment, it might be a propitious time for taxpayers to 
disclose voluntarily any underpayments of tax. The news release 
also noted that the likelihood of criminal prosecution was not 
high in instances of voluntary disclosure, although the news 
release offered no assurances that amnesty from criminal 
prosecution would be granted. A current policy statement of the 
IRS includes voluntary disclosure of tax underpayments as one 
criterion to be considered in determining whether a case 
warrants criminal prosecution.\4\
    \4\ IRM (31)330, ``Criminal Tax Policies and Procedures: Voluntary 
    In his January 25, 1984, State of the Union address, 
President Ronald Reagan instructed the Treasury Department to 
issue a Report on Fundamental Tax Simplification and Reform. 
The Treasury Department submitted its three-volume report to 
President Reagan on November 27, 1984 (``1984 Treasury 
Report''). The 1984 Treasury Report stated that ``the Treasury 
Department rejects'' amnesty.\5\ Treasury rejected any form of 
tax amnesty, whether it is restricted only to amnesty from 
criminal penalties or whether amnesty also extends to civil 
penalties, interest, and past taxes due. The 1984 Treasury 
Report rejected amnesty because of its negative effect on 
taxpayer morale: ``To include tax, civil penalties, and 
interest in an amnesty would further undermine taxpayer morale 
by sending a clear signal to the American public concerning 
non-compliance and tax fraud: `Don't bother to pay now. We may 
forget you owe anything. Even if you have to pay tax, we won't 
charge interest.' '' \6\ The 1984 Treasury Report also stated 
that amnesty ``can only reinforce the growing impression that 
the tax system is unfair and encourage taxpayer noncompliance. 
After reviewing State and foreign experience with amnesties, 
the Treasury Department rejects their use by the Federal 
Government.'' \7\
    \5\ 1984 Treasury Report, Vol. 1, p. 91.
    \6\ Ibid.
    \7\ Ibid.
    On July 25, 1990, Michael J. Graetz, then Deputy Assistant 
Secretary (Tax Policy) of the Department of the Treasury,\8\ 
testified on tax amnesty before the Subcommittee on Commerce, 
Consumer and Monetary Affairs of the House Committee on 
Government Operations. His testimony described the State 
experiences, ``outlined important differences in the State and 
Federal systems that make it difficult to translate the States' 
experiences to the Federal level, . . . and reviewed the 
revenue implications of a Federal amnesty program and explained 
why we [Treasury] believe substantial revenue increases would 
be unlikely.'' \9\ He concluded his testimony by indicating 
``our [Treasury's] lack of support for a general Federal tax 
amnesty. . . .'' \10\ At that same hearing, Fred T. Goldberg, 
Jr., then Commissioner of Internal Revenue,\11\ stated ``[o]n 
balance, we believe that a general Federal tax amnesty would be 
ill-advised and counter-productive.'' \12\
    \8\ Now Justus S. Hotchkiss Professor of Law, Yale University.
    \9\ Page 2.
    \10\ Page 8.
    \11\ Now a partner at the law firm of Skadden, Arps, Slate, Meagher 
& Flom.
    \12\ Page 11.

       C. Elements of Present Law That Are Similar to Tax Amnesty

    Although there has never been a formal Federal tax amnesty, 
there are several aspects of the administration of present law 
by the IRS that bear similarities to elements of tax amnesty. 
These are described below.

Installment agreements

    Section 6159 of the Code authorizes the IRS to enter into 
written agreements with any taxpayer under which the taxpayer 
is allowed to pay taxes owed, as well as interest and 
penalties, in installment payments if the IRS determines that 
doing so will facilitate collection of the amounts owed. An 
installment agreement does not reduce the amount of taxes, 
interest, or penalties owed; it does, however, provide for a 
longer period during which payments may be made during which 
other IRS enforcement actions (such a levies or seizures) are 
held in abeyance. Many taxpayers can request an installment 
agreement by filing Form 9465. This form is relatively simple 
and does not require the submission of detailed financial 
statements. The IRS in most instances readily approves these 
requests if the amounts involved are not large (in general, 
below $10,000) and if the taxpayer has filed tax returns on 
time in the past. Some taxpayers are required to submit 
background information to the IRS substantiating their 
application. If the request for an installment agreement is 
approved by the IRS, a user fee of $43 is charged.\13\ This 
user fee is in addition to the tax, interest, and penalties 
that are owed.
    \13\ This user fee is imposed pursuant to 31 U.S.C. 9701. See T.D. 
8589 (February 14, 1995).
    The IRS enters into a very significant number of 
installment agreements each year. For each of the last three 
fiscal years, the IRS has entered into on average 2.6 million 
new installment agreements. At any given point during the last 
three fiscal years, on average 4.4 million installment 
agreements were in effect (many are in effect for longer than 
one year). The IRS collected $4.7 billion through installment 
agreements in FY 94.
    Installment agreements may in part resemble tax amnesties 
in the effects that they have on certain taxpayers. Taxpayers 
who are unable to pay their full tax liability on a timely 
basis can obtain (1) partial relief from the IRS via a 
lengthier period over which payment is permitted and (2) 
forbearance by the IRS from using more disruptive collection 

Offers in compromise

    Section 7122 of the Code permits the IRS to compromise a 
taxpayer's tax liability. In general, this occurs when a 
taxpayer submits an offer in compromise to the IRS. An offer in 
compromise is a proposal to settle unpaid tax accounts for less 
than the full amount of the balance due. They may be submitted 
for all types of taxes, as well as interest and penalties, 
arising under the Internal Revenue Code.
    Taxpayers submit an offer in compromise on Form 656. There 
are two bases on which an offer can be made. The first is doubt 
as to the liability for the amount owed. The second is doubt as 
to the taxpayer's ability fully to pay the amount owed. An 
application can be made on either or both of these grounds. 
Taxpayers are required to submit background information to the 
IRS substantiating their application. If they are applying on 
the basis of doubt as to the taxpayer's ability fully to pay 
the amount owed, the taxpayer must complete a financial 
disclosure form enumerating assets and liabilities.
    As part of an offer in compromise made on the basis of 
doubt as to ability fully to pay, taxpayers must agree to 
comply with all provisions of the Internal Revenue Code 
relating to filing returns and paying taxes for five years from 
the date IRS accepts the offer. Failure to observe this 
requirement permits the IRS to begin immediate collection 
actions for the original amount of the liability.
    Starting in 1992, the IRS has both liberalized the terms 
under which it will accept an offer in compromise and increased 
the publicity of the availability of offers in compromise. The 
following table shows the numbers of offers in compromise 
received by the IRS.

        Table 1.--Offers In Compromise Received By IRS, 1983-1996       
                       Fiscal year                          (rounded to 
1983....................................................           4,000
1984....................................................           5,000
1985....................................................           5,000
1986....................................................           6,000
1987....................................................           7,000
1988....................................................           7,000
1989....................................................           8,000
1990....................................................           9,000
1991....................................................           9,000
1992....................................................          18,000
1993....................................................          50,000
1994....................................................          50,000
1995....................................................          55,000
1996....................................................          57,000
Source: IRS Annual Reports.                                             

    The percentage of offers in compromise accepted has also 
risen significantly. In fiscal year (FY) 1991, 25 percent of 
offers in compromise that were submitted to the IRS were 
accepted. In FY 1992 the percentage accepted rose to 45 
percent.\14\ For FY 1994, 1995, and 1996, the percentage 
accepted averaged 49 percent. In FY 1996, the IRS accepted $287 
million through the offers in compromise program, and 
eliminated a total of $2.2 billion of taxpayers' liability for 
aggregated taxes, interest, and penalties. Accordingly, in FY 
1996, the IRS accepted on average 12 cents on the dollar 
through the offers in compromise program.
    \14\ 1992 IRS Annual Report, p. 8.
    The IRS is currently in the process of issuing a new 
publication that will highlight the availability of offers in 
compromise and installment agreements.
    Offers in compromise strongly resemble tax amnesties in 
that taxpayers can obtain forgiveness of a portion of the 
taxes, interest, and penalties that they owe. The principal 
difference is that offers in compromise are generally tailored 
to the financial circumstances of each taxpayer; \15\ their 
assets as well as their present and future earnings are 
considered by the IRS in determining whether to accept a 
taxpayer's offer. For example, an offer may be rejected if the 
taxpayer does not agree to liquidate certain assets to pay the 
amount owed. Tax amnesties generally are not tailored to the 
circumstances of individual taxpayers, but instead offer 
identical terms to everyone.
    \15\ It should be noted that installment agreements (described 
above), unlike offers-in-compromise, are generally available, 
regardless of financial circumstances, to any taxpayer owing less than 


          A. Economic Theory of Tax Evasion and Tax Amnesties

    A complete discussion of tax amnesties must also address 
the related issue of tax law enforcement. Tax noncompliance is 
a necessary precursor to an amnesty, and the level of 
noncompliance depends in part on the resources and efforts that 
the government expends on the enforcement of existing laws. 
Thus, the current and past status of the government's 
enforcement efforts will in part determine the amount of back 
taxes that potentially can be raised in an amnesty. 
Furthermore, while an amnesty may be one way to obtain those 
back taxes and to bring new taxpayers into the tax system, 
increasing the resources devoted to enforcement of the tax laws 
is another way to achieve the same objectives. For reasons 
described below, both of these measures may be pursued 
simultaneously. In fact, it generally has been the practice of 
the States that have offered amnesty programs to also increase 
enforcement efforts. Because of this, it is important, while 
interpreting the results of State amnesty programs, to bear in 
mind that the State programs generally have been amnesty/
enforcement programs, and not simply stand-alone amnesty 
    The following is a discussion of the economic 
considerations that in part determine taxpayers' decisions with 
regard to tax evasion or compliance, and whether to take part 
in any amnesty program that is offered. The government's 
decisions with regard to the setting of enforcement parameters 
and the potential offering of amnesty are also examined, as is 
the academic literature that has examined the evidence on 
Taxpayer's decision
    Economists typically view the taxpayer's decision to comply 
fully or not to comply fully with the tax laws as a rational 
choice. In this view, the taxpayer chooses a level of 
compliance by weighing the tradeoff between compliance and 
evasion and choosing the level of compliance that will lead to 
the highest expected level of net benefits. The taxpayer's 
choice will depend on the various factors that affect the 
benefits and costs of tax evasion relative to complying with 
the tax laws. Among the factors that will influence the 
decision are the probability of the evasion being detected 
through audit, the back taxes and civil and criminal penalties 
that will be imposed if evasion is detected, the taxpayer's 
ethics or degree of honesty (alternatively, the expected level 
of guilt that would arise from evasion of taxes), damage to the 
reputation of the taxpayer if the evasion is detected, the 
taxpayer's level of ``risk aversion,'' and the perceived 
benefits derived from a successful evasion of taxes.\16\
    \16\ In theory, the net benefits should include consideration of 
the impact of one's tax evasion on the provision of public goods from 
which the taxpayer benefits. In a large community of taxpayers, the 
marginal personal benefit directly derived solely from one's own taxes 
is likely to be insignificant due to the diffuse benefits of the public 
good. Thus, in large communities the potential tax evader can dismiss 
the effect that his evasion may have on his benefits from the provision 
of public goods and services.
    On the cost side, the audit probabilities and the penalties 
imposed are the principal determinants of the expected costs of 
tax evasion. Anything that would raise these expected costs, 
such as increased penalties for evasion, both civil and 
criminal, or an increased likelihood of evasion being detected 
due to increased audit rates, would decrease the amount of 
evasion undertaken. Another cost factor is the guilt that one 
might feel after evading taxes. While guilt, or more broadly 
one's feelings about the morality of paying one's taxes, is not 
often modeled in the economic analysis of tax evasion, it 
clearly is a factor that influences human behavior. If one 
expects to feel guilty from tax evasion, less tax evasion will 
be engaged in.
    The benefit from tax evasion is primarily the addition to 
after-tax income that results from such cheating, but also may 
include the avoidance of the costs of complying with the tax 
laws. It should be noted that the monetary benefit of tax 
evasion in the form of under-reporting income or overstating 
deductions will depend on the level of tax rates. For example, 
for each dollar of income that is not reported, the taxpayer 
will benefit by the amount of the taxes that would have been 
paid on that dollar of income, which is equal to the taxpayer's 
marginal tax rate multiplied by that dollar. Thus, if other 
factors are held constant, tax evasion would be expected to 
increase the higher the tax rate that is imposed on the tax 
base. While higher tax rates may provide an incentive for 
greater cheating, this incentive could well be offset by higher 
audit probabilities for persons in higher marginal tax 
brackets. If audit rates are correlated with higher tax rates, 
then it will not necessarily be the case that persons in higher 
tax brackets will have greater propensity to cheat. Again, one 
must weigh both the benefit side and the cost side of cheating 
before rationally determining a compliance level that maximizes 
one's expected net benefits.
    Further, there is the taxpayer's ``degree of risk 
aversion,'' or the extent to which the taxpayer likes or 
dislikes risk taking. Taxpayers will vary in their attitudes 
towards risk taking, and thus some will choose to cheat, and 
others to comply, even though they might agree as to the 
potential costs and potential benefits of cheating. The less 
risk-averse taxpayer is more likely to risk the downside of 
cheating--getting caught--in order to reap the benefits of 
cheating--the addition to after-tax income. Essentially, the 
taxpayer's attitude towards risk is the final link in weighing 
the cost and benefits to determine whether to evade taxes. 
While the government can affect the benefits and costs of 
cheating, it basically has no control over the attitudes toward 
risk of a given taxpayer.
    In the standard economic model of tax evasion, the 
declaration of an amnesty standing alone would have no effect 
on the behavior of taxpayers. That is, no one would come 
forward to take advantage of the amnesty. The reason for this 
is that in the standard model, the taxpayer has made a choice 
based on a rational weighing of the benefits and costs of 
evasion, and because these benefits and costs have not changed, 
the taxpayer's chosen level of compliance remains the same. If, 
however, the penalties for evasion also are increased, or the 
relevant authorities increase the likelihood of detection 
through audits, then the rational taxpayer will reevaluate past 
choices and may choose to take advantage of the amnesty since 
the expected costs of previous evasion have been increased. It 
is for this reason that State amnesties have been offered in 
conjunction with increased enforcement efforts.
    Clearly, this theoretical model of tax evasion does not 
represent the behavior of all taxpayers, and the real world 
certainly would produce some taxpayers who would take advantage 
of an amnesty even without changes in the relevant enforcement 
parameters. For example, a tax evader might experience more 
guilt from tax evasion than expected, or might find that living 
with the prospect of the tax evasion being detected is more 
unappealing than expected, and wish to reverse the earlier 
decision to evade taxes.\17\ However, a possible constraint on 
a taxpayer's willingness to take part in an amnesty is that to 
do so labels oneself a tax evader and may thus increase the 
probability that the IRS would audit that taxpayer's returns in 
future years. Even if one intended to comply with the tax laws 
in the future, one presumably would prefer to avoid an audit 
whenever possible. A possible additional constraint on tax 
evaders considering an application for amnesty is the potential 
damage to the taxpayer's reputation should such information 
become public.
    \17\ See Arun Malik and Robert Schwab ``The Economics of Tax 
Amnesties,'' Journal of Public Economics 46, 1991, pp. 29-49, for a 
model of taxpayer behavior and amnesties that considers these issues 
and allows for adaptive behavior on the part of the taxpayer.
    An important complication in the economics of any amnesty 
proposal is the degree to which the amnesty will affect future 
tax compliance. Economists typically have noted that the 
granting of amnesty will forever change taxpayers' perceptions 
of the ``rules of the game'' between the taxing authority and 
the taxpayer. Once an amnesty is granted, taxpayers may 
perceive that the likelihood of future amnesties has increased, 
and thus an incentive may be created to evade current taxes in 
anticipation of the future amnesty.\18\ It should also be noted 
that serious discussions of amnesty proposals could have the 
same effect. While an actual amnesty is likely to increase 
one's expectation of a future amnesty, mere discussions of 
amnesties in advance of an amnesty will also likely raise 
expectations that an amnesty will occur, thus potentially 
inducing greater noncompliance in advance of an amnesty.
    \18\ Several States have already offered second amnesties (see 
Table 2).
    Any amnesty offer, no matter how designed, could 
potentially have an impact on future levels of compliance. Any 
amnesty could affect future compliance if it served to indicate 
to taxpayers that enforcement of existing tax laws was lax, 
leading taxpayers to change their assumptions as to how much 
evasion they might reasonably expect to get away with. Contrary 
to the result intended, an amnesty that is successful at 
increasing tax revenues in the short term might be a signal to 
some that enforcement of tax laws is lax, and could induce 
these taxpayers to begin to cheat on their taxes, or to 
increase existing levels of noncompliance.
    The likely effect on future noncompliance from a current 
amnesty can be mitigated. The incentive for future tax evasion 
stemming from a current amnesty will depend critically on the 
design of the amnesty. An amnesty that forgives only criminal 
penalties, and requires that all back taxes, interest, and 
civil penalties be paid, is unlikely to have a significant 
effect on future compliance. That is, the specific expectation 
of a future amnesty of the same or similar design would be 
unlikely to affect future compliance, though there might still 
be an effect on future compliance from an amnesty per se, as 
discussed previously. Under the amnesty design just described, 
it is unlikely that a taxpayer would simultaneously evade 
current taxes and plan to come forward in a future amnesty to 
pay the back taxes, interest, and penalties. The rational 
taxpayer considering this evasion/amnesty combination would 
recognize it as effectively a very expensive loan from the 
government under the best of circumstances, or worse, an 
expensive loan potentially coupled with a jail term if the 
evasion is detected prior to the next amnesty. However, if 
civil penalties and some or all interest charges are waived, 
then the evasion/amnesty combination begins to resemble an 
interest-subsidized or interest-free loan from the government, 
and the taxpayer willing to risk the prospect of being caught 
prior to the next amnesty may be induced to cheat.\19\ It is 
because of this potential dynamic effect that amnesties are 
usually advertised as one-time-only events, in conjunction with 
a switch to a tougher enforcement regime. However, the 
government's credibility in declaring the amnesty a one-time-
only event may be suspect. If the reasons exist to offer an 
amnesty once, it is likely that the same reasons will exist at 
some future point, and that the same decision to offer amnesty 
will be made.\20\
    \19\ To the extent that this is true, second offerings of amnesty 
under the same terms might be expected to bring in significant revenues 
(though the net revenue effect over the long-run would not necessarily 
be positive). Connecticut recently offered a second amnesty, and over 
10 percent of the revenues came from taxpayers who had participated in 
the first amnesty. New York also recently offered a second general 
amnesty, although taxpayers could not participate in it with respect to 
a particular type of tax if they had taken advantage of the first 
general amnesty for the same type of tax.
    \20\ Georgia provided statutorily that its first amnesty would be 
its only amnesty. It is unclear whether such a law should be viewed 
with any more credibility than the mere promise not to offer a second 
amnesty, as the law could easily be overturned in the legislation 
offering the second amnesty. To date Georgia has not offered a second 
amnesty, although only five years have passed since its amnesty.

Government's decision

    Most of the above discussion has focused on the individual 
taxpayer's rational, benefit-maximizing choices with respect to 
compliance with the tax laws and the use of any tax amnesties 
that are offered. The government objective is to design a tax 
system that raises the desired amount of revenues in an 
equitable and efficient manner, taking into consideration the 
likely response of the public to the policies it adopts.\21\ In 
this system, the government policy options include setting the 
enforcement parameters (civil and criminal penalties and the 
resources devoted to audits) and making the decisions as to the 
granting of amnesties. Like the rational taxpayer, the rational 
government should choose from these options with due 
consideration of the costs and benefits that result from the 
choices it makes.
    \21\ The general design of this system, in terms of the magnitude 
of the revenues raised and from which sources, is beyond the scope of 
the current discussion.
    With respect to the enforcement parameters, the government 
benefits from more stringent enforcement in several ways. 
First, more stringent enforcement in the form of higher audit 
rates will bring in increased revenues by identifying more tax 
delinquents and collecting back taxes and penalties. Similarly, 
more stringent enforcement in the form of increased monetary 
penalties from evasion will result in more revenues for each 
tax delinquent uncovered. Second, increasing the enforcement 
parameters will lead to greater voluntary compliance with the 
tax laws since this will raise the expected costs of evasion to 
the taxpayer, and such greater voluntary compliance results in 
greater tax revenues.\22\ Furthermore, it seems reasonable to 
expect that a given individual is more likely to comply with 
the tax laws if it is believed that others are generally in 
compliance as well. Such a taxpayer may be more motivated by a 
sense of collective responsibility to pay taxes rather than the 
strict individualistic cost/benefit analysis previously 
outlined, and might willingly pay their tax obligation provided 
that others do so as well. However, in the face of widespread 
noncompliance, they may come to feel that the tax system is 
unjust to those who do pay the tax. If the government cannot 
achieve a reasonable level of compliance, a taxpayer's moral 
resolve to continue to pay the tax may begin to slip. For these 
taxpayers, increased enforcement may not directly keep them in 
compliance out of fear of the consequences of cheating, but 
rather indirectly keep them in compliance by virtue of keeping 
other taxpayers in compliance whose motivations may differ.
    \22\ For discussion and analysis of the effects of enforcement on 
compliance, see Frank Malanga, ``The Relationship Between IRS 
Enforcement and Tax Yield,'' National Tax Journal, Vol. XXXIX, No. 3, 
September 1986; Ann Witte and Diane Woodbury, ``The Effect of Tax Laws 
and Tax Administration on Tax Compliance: The Case of the U.S. 
Individual Income Tax,'' National Tax Journal, Vol. XXXVIII, No. 1, 
March 1985; and Jeffrey Dubin and Louis Wilde, ``An Empirical Analysis 
of Federal Income Tax Auditing and Compliance,'' National Tax Journal, 
Vol. XLI, No. 1, March 1988.
    There are costs to increased enforcement efforts as well. 
The most direct of these are the necessary resources devoted to 
audits. These resources clearly include the auditor's time, but 
also include the time of the law abiding taxpayers that are 
subject to audit. Additionally, excessive enforcement efforts 
could undermine compliance if such enforcement leads to an 
unnecessarily adversarial relationship between the taxpayer and 
government, leading taxpayers to ``get back'' at the government 
by evading taxes. The government needs to balance these costs 
and benefits in setting its enforcement policies.
    With respect to the amnesty decision, the potential benefit 
of an amnesty program, from the standpoint of the government, 
is an increase in tax revenues. However, as the above 
discussions have noted, there are potential costs to amnesties 
as well, because they could have a negative impact on future 
compliance. Thus, in the long-run, there may not be net 
increases in revenues from an amnesty. The proponents of 
amnesties argue that the long-run effect of an amnesty on 
compliance and revenues is positive, since an amnesty would 
bring into the system non-filers who would continue in their 
non-filing status but for the chance to come clean during the 
amnesty. This argument presumes that, in the absence of 
amnesty, tax evaders will not come into compliance for the 
current and future years for fear that doing so would expose 
the new taxpayer to audits for the previous years of non-
filing. Thus, the argument goes, even though the taxpayer would 
like to begin complying with the tax laws, he will not do so 
without first having been granted absolution for previous 
    \23\ These long-run effects on revenues would also presumably apply 
to underreporting of income, though to a lesser degree if an increase 
in reported income due to previously unreported income (as opposed to 
from filing de novo) were seen as less likely to trigger an audit for 
past returns. To the extent that such a taxpayer can currently begin to 
report income accurately at any time without triggering audits for past 
behavior, there would be no long-run positive revenue effects from an 
amnesty for such taxpayers.
    The government needs to consider other potential costs of 
amnesties prior to embarking on an amnesty program. Among these 
potential costs are administrative costs (including any 
advertising costs), the production of any new tax forms, and 
the manpower to run the program. Diversion of resources from 
other compliance efforts is likely to have negative 
repercussions that should also be considered. An accurate 
accounting of the costs of an amnesty should also consider the 
lost penalties and interest from taxpayers who take advantage 
of the amnesty but who already have been caught in the usual 
enforcement efforts, or who would be so caught in the future. 
Benefits from forgone administrative expenses of pursuing the 
delinquent taxpayers would offset these costs somewhat.
    Perhaps above all, the government objective should be to 
strive to maintain a tax system that is broadly viewed as fair 
to all. The high level of voluntary compliance with the tax 
laws, as compared to numerous other countries, is one of the 
greatest assets of the Federal tax system, and such voluntary 
compliance will no doubt be aided by fostering fairness in the 
tax code.\24\ Views differ as to the fairness of a general tax 
amnesty. Opponents believe that an amnesty is fundamentally 
unfair to taxpayers who have voluntarily and honestly paid the 
appropriate amount of taxes over the years. They see amnesties 
as letting dishonest taxpayers off the hook, and rewarding them 
for their dishonesty. Proponents argue that amnesties are fair 
to the loyal taxpayer because they bring in tax revenues not 
otherwise collectable--revenues that can finance additional 
public services without raising taxes on the law abiding, or 
that can be used to reduce taxes for the law-abiding. Clearly, 
the proponent's view is dependent on the long-run revenues from 
an amnesty being positive.\25\
    \24\ See testimony on tax amnesties by Michael J. Graetz, former 
Deputy Assistant Secretary, Treasury Department (Tax Policy), before 
the Subcommittee on Commerce, Consumer, and Monetary Affairs of the 
House Committee on Government Operations, July 25, 1990, reprinted in 
Tax Notes, April 11, 1996, p. 32 (herein referred to as ``Graetz 
    \25\ It should also be noted that if the net long-run revenues from 
amnesties are positive, the overall tax structure will be more 
efficient because the tax base is effectively broadened, thus allowing 
for potentially lower tax rates.
    In making its amnesty decision a government must weigh all 
of the pros and cons outlined above. Specifically, a government 
must decide whether it thinks that any long-run compliance 
problems from the amnesty are outweighed by the combination of 
immediate revenue increases from the amnesty and any future 
increases that could possibly stem from adding some new 
taxpayers to the rolls. In a system with a very large number of 
taxpayers, it should be noted that small changes in voluntary 
compliance could have large revenue consequences.

                B. Evidence on Efficacy of Tax Amnesties

    The experience of the States, 34 of which, plus the 
District of Columbia, have had at least one amnesty, might be 
expected to provide some data with which to assess the likely 
effects of a Federal tax amnesty, at least in the short run. 
However, care should be taken in drawing the parallels to the 
potential effects of a Federal tax amnesty.
    An important consideration is that many sources of tax 
revenues for the States do not have Federal counterparts. The 
principal example of this is the common State retail sales tax, 
which only has a modest counterpart in the Federal excise taxes 
on certain products or services such as alcohol and tobacco or 
airline tickets. Many States also rely on personal property 
taxes as a source of revenue. Despite these differences in 
State tax structures compared to the Federal structure, the 
income tax was still a major source of revenue in many State 
amnesties, and is a major component of state taxes. Income 
taxes are also the principal source of Federal revenues. 
Amnesty proponents suggest that differences in the tax bases 
between the States and the Federal Government are not 
significant enough to dismiss the State experiences as a 
predictor of the revenue results from a potential Federal 
    Another consideration in drawing parallels to the States' 
experiences with amnesties is that State tax enforcement 
efforts have typically been lax relative to the Federal 
Government, and thus taxpayers are probably more likely to have 
evaded State taxes than Federal taxes.\26\ Indeed, the evidence 
indicates that most non-filers that came forward in State tax 
amnesties had filed their Federal tax returns.\27\ Hence, State 
amnesties, especially if coupled with increased enforcement, 
might be expected to be relatively more successful than a 
Federal amnesty because past evasion is likely to have been 
more common. Contributing to this likely greater success of 
State programs is the lower cost of coming into compliance with 
State taxes, at least for those who have evaded Federal and 
State taxes, as State taxes and compliance costs are typically 
lower than Federal taxes and compliance costs. Also 
contributing to the likely greater success of State amnesties 
are the additional enforcement levers potentially at a State's 
disposal. For example, States can take actions against evaders 
such as revocation of business licenses necessary to practice 
in a State. Both Illinois and Massachusetts were known to 
threaten such actions as part of their amnesty programs in 
order to bring forward evaders.\28\ The Federal Government does 
not have similar enforcement levers.
    \26\ See Graetz testimony, pp 27-35. See also Treasury Study of Tax 
Amnesty Programs, Department of the Treasury, Internal Revenue Service, 
August 1987, pp. 11-12.
    \27\ Ibid.
    \28\ See Treasury Study of Tax Amnesty Programs.
    One has to consider carefully the conclusions some have 
drawn from the observation noted above that the typical State 
amnesty participant was in compliance with Federal tax laws. 
Some opponents of amnesties have cited it as evidence that a 
Federal amnesty is unlikely to be as successful as State 
amnesties have been, since the data would seem to suggest that 
the typical State tax evader has been in compliance with 
Federal laws. However, this argument ignores the fact that the 
incentives inherent in the State amnesties will likely cause 
only certain types of tax evaders to step forward to apply for 
amnesty. Those who come forward in the State amnesties are thus 
not a random sample of tax evaders but rather a self-selected 
sub-group of all State tax evaders. This self-selection could 
influence the observed characteristics of the group that comes 
forward to participate. Specifically, one would expect that 
evaders of both Federal and State taxes will not choose to 
identify themselves to a State government amnesty program for 
fear that they will then be discovered by the Federal 
Government. Hence, only those State tax evaders who are in 
compliance with Federal laws are likely to identify themselves 
in State amnesties.
    Proponents of amnesties cite this same evidence (and the 
self-selection argument) as likely to indicate that a Federal 
amnesty, especially if coordinated with State amnesties, will 
be more successful than States' amnesties have been, on the 
grounds that the past State amnesties have been hampered by a 
lack of coordination with a parallel Federal effort. However, 
without coordination with the State amnesties, which presumably 
would be difficult to achieve, the same motivations would 
provide a disincentive for those not in compliance with State 
taxes to come forward in a Federal tax amnesty. Furthermore, if 
State tax evasion is more common than Federal tax evasion, 
there may be relatively few people in the group most likely to 
come forward in a Federal amnesty that is not coordinated with 
State amnesties--that is, those who have evaded Federal taxes 
but not State taxes. Thus, it seems likely that a Federal 
amnesty that is not coordinated with State amnesties will be 
less successful than State amnesties not coordinated with a 
Federal amnesty.
    One needs also to examine with some skepticism the claims 
of success of the State tax amnesties even with respect to the 
immediate revenues generated. For example, many of these 
programs included accounts receivable in their amnesties, and 
thus the amnesty revenues include taxes that would have been 
collected in the absence of an amnesty.\29\ With respect to 
these taxes, the amnesties may have merely accelerated the 
payment of taxes rather than produced revenues that would not 
have been collected without an amnesty. Additionally, some who 
took part in an amnesty might have been caught in the future 
enforcement efforts of the States and also been liable for 
greater civil penalties and interest. The reported State 
revenue figures should in theory account for these costs, 
resulting in downward estimates of the net revenue intake. As 
previously discussed, the administrative costs of an amnesty 
should also be taken into account, including any advertising 
costs, the production of any new forms, and the manpower 
necessary to run the program.
    \29\ Ibid.
    To date, the evidence on the long-run impact of State tax 
amnesties on State tax revenues, inclusive of their effect on 
future compliance, has been spotty. One academic study, which 
used experimental methods to model the tax system, found that 
the subjects decreased their compliance after an amnesty.\30\ 
The results of Connecticut's second amnesty, discussed in Part 
IV of this document, could, in part, be interpreted as 
consistent with this experimental result. In the same academic 
study, compliance was found to rise when the amnesty was 
coupled with increased enforcement. Another academic study 
examined the 1985 Colorado tax amnesty and concluded that the 
amnesty, coupled with the increased enforcement efforts, had no 
long-run effect on either the level of tax revenues, or the 
trend growth of tax revenues.\31\ However, this study could not 
distinguish between the effect of the amnesty from the effect 
of the simultaneous increase in enforcement efforts. The study 
thus concluded that Colorado's post-amnesty revenues might well 
have fallen had the amnesty not been coupled with increased 
enforcement efforts. A study of the Michigan amnesty found that 
most non-filers were out of compliance for only a single year, 
perhaps indicating that chronic non-filers do not come forward 
in amnesties.\32\ If true, this would undermine the claims of 
amnesty proponents that an amnesty would induce chronic non-
filers to begin to participate in the tax system and produce 
long-run revenue gains.
    \30\ See James Alm, Michael McKee and William Beck, ``Amazing 
Grace: Tax Amnesties and Compliance,'' National Tax Journal, Vol. 
XLIII, No. 1, March 1990, pp. 23-37.
    \31\ See James Alm and William Beck, ``Tax Amnesties and Compliance 
in the Long Run: A Time Series Analysis,'' National Tax Journal, Vol. 
XLVI, No.1, March 1993, pp. 53-60.
    \32\ See Ronald Fisher, John Goddeeris and James Young, 
``Participation in Tax Amnesties: The Individual Income Tax,'' National 
Tax Journal, Vol. XLII, No. 1, March 1989, pp. 15-27.
    Ultimately, the experience of the States provides an 
incomplete guide to the likely effect of a Federal tax amnesty. 
Federal enforcement practices vis-a-vis the States are a 
crucial distinction limiting the comparison between the State 
experiences and a possible Federal amnesty. An important 
element in the State amnesties has been the simultaneous switch 
to more stringent law enforcement. However, Congress has not 
proved very willing to appropriate funds for increased 
enforcement in the past, and Federal audit rates have been 
declining over time. Because State amnesties have taken place 
in conjunction with greater enforcement, it is not apparent 
that their experiences (even if observers could agree on how to 
interpret them) would be relevant for a Federal amnesty that 
occurred during stable or decreasing enforcement efforts. An 
amnesty under circumstances of decreasing enforcement efforts 
could potentially have deleterious revenue consequences.


                  A. State Experience with Tax Amnesty

In general
    Since 1983, a total of 44 general tax amnesty programs have 
been conducted by 34 U.S. States and the District of Columbia. 
A general tax amnesty is a program involving all or a major 
portion of the taxes levied by the jurisdiction. In addition to 
these general amnesty programs, several States have undertaken 
more narrow amnesties covering a single tax or a few types of 
taxes. While the terms of each program were different, under 
all programs penalties, both civil and criminal, for tax 
liabilities owed with respect to certain periods were abated if 
the tax liability was paid to the State during the amnesty 
period. Almost all of these general amnesties also required 
payment of interest on the tax liability owed, although several 
amnesties permitted interest payments at less than the normal 
interest rate. No State amnesty during the period considered 
has allowed reduction of actual tax liabilities, as opposed to 
penalties and interest. Six States--Connecticut, Florida, 
Louisiana, New Jersey, New York, and Rhode Island--and the 
District of Columbia have each conducted two general amnesty 
programs since 1983. Arkansas has had three general amnesties 
during this period. Many state amnesties have been combined 
with announcements of increased State tax enforcement measures 
to take effect after the amnesty period. Several local 
jurisdictions have also conducted tax amnesty programs, some in 
conjunction with their States, others independently.
    Significant aspects of the general State and District of 
Columbia amnesties since 1983 are set forth in Table 2. As 
shown on that table, most of these amnesties applied to all 
major taxes levied by the jurisdiction, although several 
contained exceptions for certain types of taxes. There was wide 
variance among amnesty programs as to which taxpayers already 
in the audit process could participate in the amnesty. All of 
the general amnesties listed in Table 2 excluded taxpayers who 
were under criminal tax investigation or prosecution. Almost 
all excluded taxpayers who were in civil litigation with the 
State with regard to tax matters.\33\ There was an 
approximately equal division among the States as to whether 
taxpayers were allowed to participate if their audits had 
progressed to the point where the State booked an account 
receivable for an amount of tax liability owed.\34\ There also 
has been an approximately equal division among States as to 
whether taxpayers could benefit from the amnesty if they paid 
their tax liabilities in installments after the amnesty period.
    \33\ Both New York amnesties and New Jersey's second amnesty 
allowed taxpayers who were in civil tax litigation with the State to 
participate in the amnesty.
    \34\ This would be the point in each State's tax procedure that is 
analogous to ``assessment'' under Federal tax procedure.
    There was marked variation in the amounts collected in the 
general State amnesties. The 1985-1986 New York amnesty 
collected the most in absolute terms, generating $401.3 
million. The 1996 New Jersey amnesty collected $350 million, 
which was the most relative to State population. Table 2 
reports both the gross collections resulting from the amnesty 
and collections as a percentage of State revenue for the prior 
fiscal year from the taxes covered by the amnesty. The 
percentage calculation facilitates cross-State comparisons 
since total tax receipts reflect a State's population and 
effective tax rates. The percentage figures also improve cross-
year comparisons because the gross collection figures are not 
adjusted for inflation and, thus, gross collections in the 
1980s are not directly comparable to collections in the 1990s. 
The State amnesties with the largest collections as a 
percentage of prior-year revenue were the 1996 New Jersey 
amnesty, at 2.6 percent, followed by the 1987 New Jersey 
amnesty and the 1984 Illinois amnesty, both at 2.2 percent, and 
the 1985-1986 New York amnesty, at 2.1 percent.
    In addition to the general State amnesties listed in Table 
2, several States have conducted amnesties with respect to a 
single tax or a few taxes. For example, in 1993, Nevada 
conducted an amnesty limited to the use tax on personal 
property brought into the State from other jurisdictions. In 
1994, Vermont conducted an amnesty limited to the State's meals 
and rooms tax for take-out sales. In 1994, New York conducted 
an amnesty covering three types of taxes: (1) income taxes owed 
by non-resident individuals, trusts and estates; (2) corporate 
franchise taxes owed by out-of-State businesses; and (3) use 
taxes on personal property.
    Since 1983, several local jurisdictions have also conducted 
tax amnesty programs. Some of these have been coordinated with 
the State amnesty programs set out in Table 2. For example, 
during the period of both of New York State's general 
amnesties, New York City conducted an amnesty for most city 
taxes on the same terms as the State amnesty and with a 
coordinated publicity and advertising campaign. Some local 
jurisdictions also have conducted amnesties independent of 
State amnesties, such as the 1985 amnesty for Chicago city 
taxes on all computer leases and other leases of property 
involving related parties.

                                                     Table 2.--State Tax Amnesty Programs, 1983-1997                                                    
                                                                                                        Amnesty      Collections                        
                                                                                     Accounts         collections   as percentage       Installments    
             State                  Amnesty period      Major  taxes  covered       receivable            ($         of relevant        arrangements    
                                                                                     included        millions)\1\    revenue \2\       permitted \3\    
Alabama........................    01/20/84-04/01/84   All                     No                         3.2       0.1            No                   
    First......................    11/22/82-01/20/83   All                     No                         6.0       0.2            Yes                  
    Second.....................    09/01/87-11/30/87   All                     No                         1.7       0.09           Yes                  
    Third......................    09/01/97-11/30/97   All                     No                     \4\ 3.0        .05(1994)     Yes                  
California.....................    12/10/84-03/15/85   Individual income       Yes                      154.0       1.7            Yes                  
                                                       Sales & Use             No                        43.0       0.5            Yes                  
Colorado.......................    09/16/85-11/15/85   All                     No                         6.4       0.3            Yes                  
    First......................    09/01/90-11/30/90   All                     Yes                       54.0       1.1            Yes                  
    Second.....................    09/01/95-11/30/95   All                     Yes                       40.9       0.6            Yes                  
    First......................    01/01/88-06/30/88   All                     No                     \5\ 8.4       0.09           No                   
    Second.....................    10/01/92-12/03/92   Most \6\                No                        14.0       0.1            No                   
Georgia........................    10/01/92-12/05/92   Most \7\                Yes                       51.3       0.7            No                   
Idaho..........................    05/20/83-08/30/83   All                     No                         0.3       0.02(1981)     No                   
Illinois.......................    10/01/84-11/30/84   All                     Yes                      160.5       2.2            No                   
Iowa...........................    09/02/86-10/31/86   All                     Yes                       35.1       1.6(1984)      N.A.                 
Kansas.........................    07/01/84-09/30/84   All                     No                         0.6       0.04           No                   
Kentucky.......................    09/15/88-09/30/88   All                     No                        61.1       1.9            No                   
    First......................    10/01/85-12/31/85   All                     No                         1.2       0.04           Yes \8\              
    Second.....................    10/01/87-12/15/87   All                     No                         0.3       0.008          Yes                  
Maine..........................    11/01/90-12/31/90   All                     Yes                       29.0       1.8            Yes                  
Maryland.......................    09/01/87-11/02/87   All                     Yes                   \9\ 34.6       0.7            No                   
Massachusetts..................    10/17/83-01/17/84   All                     Yes                       86.5       1.7            Yes \10\             
Michigan.......................    05/12/86-06/30/86   All                     Yes                      109.8       1.3(1984)      No                   
Minnesota......................    08/01/84-10/31/84   All                     Yes                       12.1       0.3            No                   
Mississippi....................    09/01/86-11/30/86   All                     No                         1.0       0.06(1984)     No                   
Missouri.......................    09/01/83-10/31/83   All                     No                         0.9       0.02(1981)     No                   
New Jersey:                                                                                                                                             
    First......................    09/10/87-12/08/87   All                     Yes                      186.5       2.2            Yes                  
    Second.....................    03/15/96-06/01/96   All                     Yes                      350.0       2.6(1994)      No                   
New Mexico.....................    08/15/85-11/13/85   Most \11\               No                        13.6       1.0            Yes                  
New York:                                                                                                                                               
    First......................    11/01/85-01/31/86   Most \12\               Yes                      401.3       2.1            Yes                  
    Second.....................    11/01/96-01/31/97   Most                    Yes                  \13\ 277.5      0.9            Yes                  
North Carolina.................    09/01/89-12/01/89   Most \14\               Yes                       37.6       0.5            No                   
North Dakota...................    09/01/83-11/30/83   All                     No                         0.2       0.02           Yes                  
Oklahoma.......................    07/01/84-12/31/84   Income, Sales           Yes                       13.9       0.9            No\15\               
Pennsylvania...................    10/13/95-01/10/96   All                     Yes                       93.0       0.6            No                   
Rhode Island:                                                                                                                                           
    First......................    10/15/86-01/12/87   All                     No                         0.7       0.08           Yes                  
    Second.....................    04/15/96-06/28/96   All \16\                Yes                        7.9       0.6(1994)      Yes                  
South Carolina.................    09/01/85-11/30/85   All                     Yes                        7.1       0.3            Yes                  
Texas..........................    02/01/84-02/29/84   All \17\                Yes                        0.5       0.006          No                   
Vermont........................    05/15/90-06/25/90   All                     Yes                        1.0       0.2            No                   
Virginia.......................    02/01/90-03/31/90   All                     Yes                       32.2       0.5            No                   
West Virginia..................    10/01/86-12/31/85   All                     Yes                       15.9       0.9            Yes                  
Wisconsin......................    09/15/85-11/22/85   All                     Yes \18\                  27.3       0.5            Yes                  
District of Columbia:                                                                                                                                   
    First......................    07/01/87-09/30/87   All                     Yes                       24.3       N.A.           Yes                  
    Second.....................    07/10/95-08/31/95   All                     Yes                       19.5       N.A.           Yes                  
\1\ Where applicable, the amount indicates local portions of certain taxes collected under the State tax amnesty program. Adjustments have not been made
  for inflation.                                                                                                                                        
\2\ The figure expresses collections in the amnesty as a percentage of the State's revenue from the taxes covered in the last full fiscal year prior to 
  the amnesty. E.g., where the amnesty covers only income taxes, the figure represents amnesty collections as a percentage of prior fiscal year income  
  tax revenue. Where revenue amounts for the prior fiscal year are unavailable, the fiscal year utilized is indicated in parentheses.                   
\3\ ``No'' indicates requirement of full payment by the expiration of the amnesty period. ``Yes'' indicates allowance of full payment after the         
  expiration of the amnesty period.                                                                                                                     
\4\ Estimated as of January 7, 1998.                                                                                                                    
\5\ Did not include intangibles tax and drug taxes. Gross collections totaled $22.1 million, with $13.7 million in penalties withdrawn.                 
\6\ Did not include personal property taxes.                                                                                                            
\7\ Did not include alcohol and tobacco excise taxes and property and intangibles taxes.                                                                
\8\ Amnesty taxpayers were billed for the interest owed with payment due within 30 days of notification.                                                
\9\ Figure includes $1.1 million for the separate program conducted by the Department of Natural Resources for the boat excise tax.                     
\10\ The amnesty statute was construed to extend the amnesty to those who applied to the tax department before the end of the amnesty period, and       
  permitted them to file overdue returns and pay back taxes and interest at a later date.                                                               
\11\ The severance taxes (including the six oil and gas severance taxes), the resources excise tax, the corporate franchise tax, and the special fuels  
  tax were not subject to amnesty.                                                                                                                      
\12\ In both New York amnesties, availability of amnesty for the corporation tax, the oil company taxes, the transportation and transmissions companies 
  tax, the gross receipts oil tax and the unincorporated business tax was restricted to entities with 500 or fewer employees in the United States on the
  date of application. In addition, a taxpayer principally engaged in a aviation, or a utility subject to the supervisor of the State Department of     
  Public Service was also ineligible.                                                                                                                   
\13\ As of May 1, 1997.                                                                                                                                 
\14\ Did not include real property taxes.                                                                                                               
\15\ Full payment of tax liability required before the end of the amnesty period to avoid civil penalties.                                              
\16\ Employment security taxes were not included.                                                                                                       
\17\ Texas does not impose a corporate or individual income tax. In practical effect, the amnesty was limited to the sales tax and other excises.       
\18\ Waiver terms varied depending upon the date the tax liability was accessed.                                                                        
Source: The Federation of Tax Administrators (Updated by Staff of the Joint Committee on Taxation).                                                     

Detailed discussion of several States' experiences

            New York
    New York's amnesty experience is notable for the number of 
amnesty programs it has undertaken, the high amounts they have 
collected, and the coordination of similar amnesties for New 
York City taxes. New York's first general amnesty program in 
1985 and 1986 yielded $401.3 million, the highest total 
collected in any State amnesty thus far. The 1985-1986 New York 
amnesty ranks fourth among the amnesties set out in Table 2 on 
the basis of collections as a percentage of prior-year revenue. 
This amnesty was coordinated with an amnesty for New York City 
taxes available during the same period. During the early 1990s, 
New York State and New York City conducted limited amnesties 
with respect to specific types of taxes. Then, in 1996 and 
1997, New York undertook a second general amnesty program on 
terms similar to the 1985-1986 program, but which resulted in 
fewer collections. Collections for the 1996-1997 amnesty have 
been estimated at $277.5 million. \35\
    \35\ New York Daily News, May 5, 1997, p. 43.
    New York's first general tax amnesty from November 1, 1985, 
through January 31, 1986, yielded collections of $401.3 
million. This total came from diverse sources. Over 148,000 
taxpayers participated in the amnesty, with payments ranging 
from a few cents to $2.5 million from a single individual, and 
an average of about $3,000 per taxpayer. Payments came from the 
following sources: 44 percent from personal income taxes; 40 
percent from sales taxes; and 14 percent from corporate taxes. 
One-half of the total revenues and 43 percent of total 
applications were received or postmarked on the last day of the 
amnesty period. \36\ Nearly 30 percent of the revenue generated 
by the amnesty was from tax liabilities for which the New York 
Department of Taxation and Finance did not have information, 
with the result that at least 1,500 additional businesses were 
added to the New York tax rolls in 1986. \37\
    \36\ ``Tax Amnesty: the New York State Experience,'' New York State 
Department of Taxation and Finance, p. 6 (February, 1988).
    \37\ ``Federal Tax Amnesty: Reflecting on the State's 
Experiences,'' 40 Tax Lawyer 145, p. 176 (Fall, 1986).
    The New York amnesty provided for a waiver of civil 
penalties for taxes owed for periods before January 1, 1985, if 
taxpayers made payments of the tax liability plus interest on 
or prior to January 31, 1986. The amnesty covered most New York 
State taxes, including all personal income, corporate 
franchise, sales and use, withholding, motor fuels and estate 
and gift taxes. Excluded were excise taxes on alcoholic 
beverages, cigarettes and other tobacco, real estate transfer 
taxes, real property gains taxes, and corporate income taxes on 
banks, insurance companies, public utilities and other 
corporations with more than 500 employees.
    The 1985-1986 New York amnesty was reported to have 
originated as a political compromise between then Governor 
Mario Cuomo and Republicans in the State legislature. \38\ The 
Cuomo administration initially opposed Republican proposals for 
an amnesty. Eventually, the Cuomo administration agreed to an 
amnesty in exchange for passage by the State legislature of the 
Omnibus Tax Equity and Enforcement Act (``OTEEA'') of 1985, a 
comprehensive package of tax enforcement measures that took 
effect after the amnesty. \39\ Under the OTEEA, seven tax 
offenses were increased to felonies, including failure to file 
returns, whether or not willful, for three or more years, and 
failure to pay withholding taxes in excess of $250. The maximum 
fine for a felony was raised to $50,000 for individuals and 
$250,000 for corporations. Civil penalties were also increased, 
and broader enforcement and investigating powers were given to 
the Department of Taxation and Finance. Finally, $68 million 
was appropriated for a new State tax computer system under the 
supervision of a new Revenue Opportunity Division in the 
Department. This division was directed to undertake broad-
ranging projects to match tax data with other data available to 
the State. For example, a State publication about the amnesty 
announced a ``doctors project'' for matching tax returns of 
medical doctors with the payments they received from Blue 
Cross, Blue Shield and Medicaid. \40\
    \38\ ``Tax Amnesty Enters Final Week: New York May Get $200 
Million,'' New York Times, January 26, 1986, p. 22.
    \39\ Ibid.
    \40\ ``Amnesty Briefing,'' New York State Department of Taxation 
and Finance, p. 3.
    New York State made much of the increased enforcement 
measures of the OTEEA in its announcements of the amnesty 
program, which attempted to convince taxpayers that ``the rules 
of the tax evasion game'' had changed. \41\ An advertising 
agency was retained to develop a campaign using bus and subway 
posters and some television and radio commercials centered on 
the slogan, ``It would be a crime to miss out on Amnesty.'' 
Officials of the New York Department of Taxation and Finance 
also gave over 550 interviews, speeches and presentations on 
the program. \42\
    \41\ ``Tax Amnesty: the New York State Experience,'' supra, p. 9.
    \42\ Ibid., p. 10.
    Coordinated with the New York State amnesty program was an 
amnesty conducted by New York City on largely the same terms. 
The amnesty applied to several New York City taxes, including 
the personal income tax, the general corporation tax and the 
unincorporated business tax. The City of Yonkers also conducted 
an amnesty for city taxes in conjunction with the State 
    Both the New York State and New York City amnesties applied 
to all taxpayers other than those who were in criminal 
litigation, or who had previously been convicted, or were 
underinvestigation, for a State tax criminal charge. Thus, taxpayers 
who were under audit, including those for whom accounts receivable had 
been established, were entitled to participate. Taxpayers in civil 
litigation could participate if they withdrew from litigation. 
Taxpayers who could demonstrate severe financial need could qualify for 
the amnesty by committing to make payments in installments; a down 
payment of 50 percent was required for this option. The costs of 
administering the amnesty program were approximately $2.8 million.\43\ 
At the height of the program, 200 employees of the New York Department 
of Taxation and Finance were assigned to a separate State building to 
administer the amnesty.\44\
    \43\ 40 Tax Lawyer, supra, p. 176.
    \44\ ``Tax Amnesty; the New York State Experience,'' supra, p. 6.
    The measures of the OTEEA that were used as the incentive 
for the amnesty program reportedly led to a significant 
increase in New York State enforcement activities.\45\ For 
example, in the first six months of 1986, New York handed out 
more jail terms for tax offenses than in the previous 20 
    \45\ ``Your Taxes: A Guide to Preparing 1986 Returns,'' New York 
Times, February 27, 1987, p. D1.
    \46\ ``No More Mr. Nice Guy,'' Wall Street Journal, July 9, 1986, 
    Based on the success of the 1985-1986 general amnesty, New 
York State conducted two limited amnesties in the early 1990s. 
From April 1, 1992, until February 28, 1993, New York offered 
an amnesty for the State's real property gains tax. In 1994, 
the State of New York granted amnesty for three types of taxes 
normally paid by out-of-State residents. From September 1, 
1994, until November 30, 1994, amnesty was given for penalties 
for income taxes owed by individuals, trusts and estates that 
were not residents of the State and corporate franchise and 
other business taxes owed by out-of-State businesses doing 
business in the State. Also included were use taxes owed by 
individuals and certain small businesses on property brought 
into the State. The terms of the amnesty were largely the same 
as for the earlier general amnesty, except that taxpayers who 
had participated in the first amnesty or who had been contacted 
for audit could not participate. In coordination with the 
State's 1994 program, New York City granted an amnesty for 
payments of four specific taxes: the Commercial Rent Tax, the 
Utility Tax, the Real Property Transfer Tax and the Hotel Room 
Occupancy Tax.
    Most recently, from November 1, 1996, to January 31, 1997, 
New York State offered a second general amnesty involving most 
major taxes. The amnesty applied to penalties on taxes owed for 
periods before January 1, 1995. Once again, New York City and 
Yonkers offered amnesties for city taxes on a coordinated 
basis. To encourage taxpayers to come forward, the State tax 
penalties were increased by 5 percent for periods after the 
amnesty for taxpayers who would have been entitled to amnesty 
under the program. With few exceptions, the taxes and taxpayers 
covered and the requirements for the amnesty were the same as 
for the 1985-1986 general amnesty.\47\ Taxpayers who received 
amnesty under the 1985-1986 program were not entitled to 
amnesty under the 1996-1997 program for the same type of tax; 
for example, a taxpayer who received amnesty for personal 
income taxes in 1985-1986 could receive amnesty for withholding 
tax liabilities in the later amnesty, but not for personal 
income tax liabilities.
    \47\ The New York Beverage Container Tax and the Special Tax on 
Passenger Car Rentals were subject to the amnesty although they were 
not enacted until after the 1985-1986 amnesty.
    Despite a large-scale publicity effort and commitment of 
resources by the New York Department of Taxation and Finance, 
payments under the second amnesty fell short of those under the 
first general amnesty, indicating that the second amnesty may 
have reduced taxpayer expectations of the State's future tax 
enforcement.\48\ The State had projected $450 million in 
receipts from the second program.\49\ Collections as of May 1, 
1997, were estimated at $277.5 million;\50\ $73 million of the 
total was from tax liabilities for which the audit system did 
not have information. Based on these results, State Comptroller 
H. Carl McCall characterized the second amnesty as ``ill 
advised'' and a ``disappointment.'' \51\
    \48\ See discussion at pp. 11-12.
    \49\ New York Times, March 15, 1997, supra.
    \50\ New York Daily News, May 5, 1997, supra.
    \51\ New York Times, March 15, 1997, supra.
    California conducted a tax amnesty program from December 
10, 1984, until March 15, 1985, which covered personal income 
taxes and sales and use taxes. Corporate franchise taxes were 
not covered. Collections under the amnesty program totaled $197 
million, the fifth highest total for State amnesty programs 
thus far, although it ranks somewhat lower as a percentage of 
prior-year revenue. These collections were obtained without a 
large expenditure on advertising, which is a dissimilarity 
between the California program and most of the other high-
yielding amnesty programs.
    The terms of the California amnesty were a waiver of civil 
penalties on personal income and sales and use taxes owed for 
periods prior to January 1, 1984, if the taxpayer paid the tax 
liability during the amnesty period. For both the personal 
income tax and sales and use tax, taxpayers did not qualify for 
the amnesty if they were involved in a proceeding, or were 
subject to an investigation, with respect to a State tax 
criminal charge. Taxpayers with accounts receivable were not 
entitled to participate in the sales and use tax amnesty, 
although other taxpayers under audit could participate. 
Taxpayers under audit, including those with accounts 
receivable, were entitled to participate in the personal income 
tax amnesty. Where full paymentof the tax liability during the 
amnesty period would cause undue hardship, taxpayers could qualify for 
the amnesty by committing to make payments in installments.
    The legislation that approved the amnesty also adopted a 
new set of enforcement measures that took effect after the 
amnesty period. Penalties for State tax offenses were raised to 
a minimum of 5 percent of the liability and a maximum of 
$20,000 and three years imprisonment. Special penalties for 
failure to report cash payments were introduced, which included 
loss of State business and professional licenses. Authorization 
was granted to utilize private collection agencies to collect 
tax deficiencies, to use California State police to serve 
warrants for criminal tax charges, and to institute continuous 
levies against non-wage payments. Requirements of information 
returns for real property sales and registration of tax 
shelters were introduced. Finally, the amnesty legislation also 
provided for a new State tax computer system having the ability 
to cross-reference with respect to a single taxpayer 
information on all State taxes, as well as State and local 
business records.
    In the months leading up to the amnesty, the State 
attempted to project a tough and high-profile image on tax 
enforcement, with announcements of tax criminal prosecutions 
and arrests, public seizures of boats and luxury automobiles 
and auctions of unusual property seized.\52\ Although the State 
conducted an advertising and public relations campaign to 
increase public awareness of the amnesty, its budget for this 
campaign was $550,000, which is one of the smallest for the 
State amnesty programs considered, especially when compared 
with California's size. The advertising campaign included 
brochures, billboards and public-service announcements on 
television and radio with the slogan ``Get to us before we get 
to you.'' \53\ The amnesty also benefited from widespread 
coverage on television news, including dramatic footage by a 
local Fresno station of amnesty applicants running out of a 
State tax office because they did not want to appear on 
    \52\ E. Dronenburg, Jr, ``Amnesty, a Tool for Closing the 
California Tax Gap,'' Tax Notes Today, July 26, 1990, p. 155.
    \53\ Ibid.
    \54\ ``Amnesty Brings Tax Bonanza; Rush to Beat Cutoff Nets $100 
Million,'' Los Angeles Times, March 16, 1985, p. 1.
    The $197 million collected by the California amnesty was 
from diverse sources. About 160,000 taxpayers applied for the 
amnesty,\55\ with an average payment of about $1,200 per 
taxpayer. For the personal income tax, the receipts break down 
as follows: 55 percent from taxpayers who had failed to file 
returns; 40 percent from taxpayers who had failed to pay the 
tax on previously filed returns; and 5 percent from liabilities 
not disclosed on returns as filed.\56\ Payments under the 
amnesty ranged from a few cents to a single payment of $1.7 
million from a corporate taxpayer.
    \55\ 40 Tax Lawyer, supra, p. 163.
    \56\ ``Study of Tax Amnesty Programs,'' Research Division; 
Assistant Commissioner (Planning, Finance and Research), Internal 
Revenue Service, Appendix 1, p. 3 (August, 1987).
    After the amnesty, the State undertook large-scale 
enforcement actions using the new mechanisms available under 
the amnesty legislation. Special emphasis was placed on tax 
protesters.\57\ Based on its favorable experience with the 
general amnesty, California conducted an amnesty for the motor 
vehicles tax from January 1, 1986, to March 31, 1986.
    \57\ ``Rise in Tax Protest Cases Because of Crackdown,'' Los 
Angeles Times, June 30, 1985, p. 6.
    Georgia conducted a general tax amnesty program from 
October 1 through December 5, 1992. Collections under the 
program totaled $51.3 million, which ranks eleventh among the 
general amnesty programs in the period since 1983. Noteworthy 
about the Georgia amnesty is that the State amnesty legislation 
provided that there would be no other general amnesty in the 
    The amnesty applied to almost all Georgia State taxes for 
periods before December 30, 1990, including individual and 
corporate income taxes, withholding, motor fuel and sales 
taxes. Excluded were tobacco and alcohol excise taxes and 
property and intangibles taxes. The amnesty applied to all 
taxpayers other than those who were in litigation with the 
State on criminal tax charges or taxpayers who had received 
notice of a criminal investigation on such charges. Taxpayers 
under audit, including those for whom accounts receivable had 
been established, were eligible for the amnesty.
    The terms of the amnesty were a waiver of civil penalties 
if payment of the taxes due, plus accrued interest, was paid to 
the State no later than December 5, 1992. For taxpayers for 
whom full payment during the amnesty period would be a severe 
hardship, the Georgia Department of Revenue was authorized to 
accept a commitment to make payments in installments.
    Both the legislation approving the amnesty and the State's 
announcements of it stated that the amnesty was a one-time 
offer which would not be repeated,\58\ and no subsequent 
amnesty has been conducted by the State. The legislation 
authorizing the amnesty also adopted stiffer penalties and tax 
enforcement mechanisms that took effect after the amnesty 
period. The legislation instituted a new ``cost of collection'' 
fee of 20 percent of tax deficiencies collected after the 
amnesty, which could be raised to 50 percent by regulations of 
the Department of Revenue. After the amnesty, willful failure 
to file a return, pay taxes, or willful filing of a false 
return was made a felony, punishable by a fine of up to $5,000 
or imprisonment for up to three years. The Department of 
Revenue was authorized to hire additional auditors and agents 
and to utilize private collection agencies, and appropriations 
were provided for a new computer system that would allow the 
Department of Revenue to retrieve information on a taxpayer-by-
taxpayer, rather than a tax-by-tax, basis.
    \58\ Georgia H.B. 1405, which was substantially unchanged from the 
form reprinted in State Tax Notes, April 2, 1992, p. 67.
    Public awareness of the Georgia amnesty program was 
heightened by a media campaign directed by a private 
advertising agency and costing over $2 million.\59\ The 
advertising campaign included images of barking dogs, 
guillotines and open sharks' jaws intended to symbolize the 
consequences of non-compliance with the amnesty program. 
Despite these efforts, the success of the Georgia program 
largely occurred in its last few days. As of December 3, 1992, 
two days before the filing deadline, only $20.4 million of the 
final total of $51.3 million had been received.\60\
    \59\ ``Georgia Cashes in on Tax Amnesty Spots,'' The Wall Street 
Journal, December 17, 1992, p. B9.
    \60\ ``Georgia Tax Amnesty Ends on Successful Note,'' State Tax 
Notes, December 14, 1992, p. 240.
    After the amnesty, the Georgia Department of Revenue 
increased enforcement against tax evaders, especially criminal 
actions using the new mechanisms provided by the amnesty 
legislation.\61\ Based on the success of the State amnesty, 
Georgia authorized a property tax amnesty program in 1994, to 
be administered by local tax authorities, that was modeled on 
the 1992 State program.
    \61\ The Tax Enforcer (newsletter of the Federation of Tax 
Administrators), vol. 2, no. 1 (1994).
    Pennsylvania conducted a general tax amnesty program from 
October 13, 1995, to January 17, 1996. The total collected was 
$93 million, the eighth highest for State programs since 1983, 
although the Pennsylvania total ranks somewhat lower as a 
percentage of prior-year revenue. A distinctive aspect of the 
Pennsylvania program was the requirement that amnesty 
participants file all required State returns and pay all 
required State taxes due within two years after the last day of 
the amnesty period.
    Proposals for a tax amnesty had circulated in the 
Pennsylvania State legislature for a number of years. One 
general amnesty bill was struck down by gubernatorial veto. The 
concept of a general tax amnesty received renewed support in 
1994, when gubernatorial candidate Tom Ridge pledged to seek an 
amnesty program combined with stepped-up enforcement against 
tax violators after the amnesty.\62\ In the final legislation, 
the general amnesty was combined with the enactment of a 15-
percent ``non-participation'' penalty for tax liabilities 
collected after the amnesty. In addition, the legislation made 
appropriations for the ``Keystone Integrated Tax System,'' a 
new computer technology system designed to enhance the State's 
ability to identify delinquent taxpayers.
    \62\ ``Pennsylvania DOR Issues Report Regarding Amnesty,'' State 
Tax Notes, July 24, 1996, p. 143.
    The terms of the Pennsylvania amnesty were abatement of 
civil penalties for tax liabilities for periods before January 
1, 1994, if payment of the tax liability plus interest was made 
during the amnesty period. The amnesty was originally announced 
to be from October 15, 1995, until January 10, 1996, although 
the deadline was extended until January 17, 1996, due to a 
major blizzard on the East coast. Eighteen State taxes were 
covered by the amnesty, including personal and business income 
taxes, sales and use taxes, employer withholding, inheritance 
taxes and the motor fuels tax. As noted above, there was an 
additional requirement for amnesty that the taxpayer file all 
required State returns and pay all required State taxes due 
within two years after the end of the amnesty period. If this 
condition was not met, the State could retroactively assess the 
penalties abated in the amnesty. Taxpayers who could 
substantiate a severe financial hardship could make payments in 
installments, but such payments did not qualify for the 
abatement of penalties under the amnesty. They were, however, 
not subject to the new 15-percent ``non-participation'' 
penalty. All taxpayers were entitled to participate, other than 
those who were in litigation with the State on a criminal tax 
charge or who had received notice from the State of a criminal 
tax investigation. Other taxpayers under audit, including those 
for whom accounts receivable had been established, were fully 
entitled to participate.
    In the weeks leading up to the amnesty, the Pennsylvania 
Department of Revenue mailed 600,000 notices of the amnesty to 
known delinquent taxpayers, which emphasized the higher 
penalties and enforcement effort planned for the post-amnesty 
period.\63\ Public awareness of the amnesty also was increased 
by a $2 million advertising campaign run by a private 
advertising agency with the slogan that the State would ``look 
the other way'' about tax liabilities only for the amnesty 
period. The principal television commercial featured 
conversations with somber men cast as revenue agents who 
proceeded to look away from the camera.\64\ Seventy-five 
employees of the Department of Revenue worked on the amnesty 
full-time. The total administrative costs of the program 
(including advertising) were $10 million.\65\
    \63\ ``Special Edition of `Pennsylvania Tax Update' Announces First 
Ever Tax Amnesty Program,'' State Tax Notes, December 7, 1995, p. 235.
    \64\ ``Tax Amnesty Ads are Heavy-Duty,'' Pittsburgh Post-Gazette, 
November 8, 1995, p. C11.
    \65\ ``Pennsylvania DOR Issues Report Regarding Amnesty,'' supra.
    The $93 million collected under the Pennsylvania amnesty 
was from widely diverse sources. More than 63,000 taxpayers 
applied for amnesty--6,000 of whom were unknown to the 
Department of Revenue prior to the amnesty. Payments ranged 
from a few dollars to a single payment of $1.4 million.\66\ The 
total amount collected came from the following sources: 22 
percent from sales and use taxes; 15 percent from personal 
income taxes; 14.4 percent from employer withholding; and 9.5 
percent from inheritance taxes. Motor taxes, realty transfer 
taxes, fuel taxes and the special oil company franchise tax 
accounted for most of the remaining 39.1 percent.\67\
    \66\ Ibid.
    \67\ ``Pennsylvania Tax Amnesty Yields $93 Million,'' State Tax 
Notes, July 22, 1996, p. 141.
    After the amnesty, the Pennsylvania Department of Revenue 
undertook a large-scale enforcement effort against taxpayers 
who had not applied for amnesty. An additional 174 criminal tax 
prosecutions were filed. Collections due to enforcement actions 
of the Department of Revenue in 1995 (not including the 
amnesty) totaled $488.1 million, just $18.9 million less than 
the record set in previous year.\68\
    \68\ ``Pennsylvania DOR Issues Report Regarding Amnesty,'' supra.
    Connecticut is one of several States that have offered 
amnesty to tax evaders more than once. Connecticut's first 
general amnesty was from September 1, 1990, to November 30, 
1990, and covered all major taxes. This amnesty raised $54 
million. The second amnesty occurred over the same dates in 
1995, and also covered all major taxes. The primary rationale 
for offering the second amnesty was that a new tax, the 
personal income tax, was introduced since the last amnesty, and 
thus there was potentially an entire new class of tax evaders. 
Connecticut's literature discussing the results of the second 
amnesty also states that the Connecticut General Assembly and 
Governor mandated that the Department of Revenue Services (DRS) 
offer the second tax amnesty program, and indicates that the 
DRS was ``given a goal to collect $31 million in additional and 
unanticipated revenues'' and ``to collect $31 million in back 
taxes to meet expected State budget requirements.'' \69\
    \69\ ``1995 Connecticut Tax Amnesty Final Report,'' pp. ii and 1.
    This second Connecticut amnesty generated $40.9 million as 
of June 30, 1996. The principal sources of revenues in this 
amnesty were the sales and use taxes (44 percent of 
collections), and the personal income tax and corporation taxes 
(both at 26 percent).\70\ The revenue raised from the amnesty 
with respect to the personal income tax accounted for less than 
0.5 percent of the annual revenue collections resulting from 
the personal income tax.
    \70\ Ibid., p. 4.
    The data that Connecticut has released concerning the 
second amnesty are worth highlighting. First, more than half of 
the revenues (51 percent) came from accounts receivable, and 
thus, as previously discussed, these revenues may merely 
represent an acceleration of revenues that would have been 
eventually collected. Second, a subset of the data shows the 
amnesty revenues raised from participants in the second amnesty 
who had also participated in the first amnesty. These are the 
only available data on taxpayers who have taken advantage of 
more than one amnesty.
    The 1995 Connecticut amnesty had 219 participants, out of a 
total of 14,929 applications accepted for amnesty, who had also 
participated in the 1990 amnesty. In the 1990 amnesty, these 
219 taxpayers accounted for $2.4 million of the amnesty 
revenues, or about 4.5 percent of total amnesty collections. In 
the 1995 program, these same taxpayers accounted for 
collections of $4.2 million or about 10.3 percent of total 
collections.\71\ In inflation-adjusted dollars, these figures 
represent approximately a two-thirds increase in the dollar 
amount of non-compliance by these taxpayers.
    \71\ Ibid., p. 7.
    Definitive conclusions cannot be drawn from these data 
about the general impact of amnesties on future taxpayer 
compliance. However, these data do show that at least some 
taxpayers who participate in amnesties continue in their non-
compliance, and perhaps even increase the dollar magnitude of 
their non-compliance. For at least these taxpayers, this 
contradicts the notion that amnesty participants ``come clean'' 
and henceforth are compliant taxpayers.

                 B. Foreign Experience with Tax Amnesty

    The history of tax amnesties in foreign countries is far 
longer than in the United States, going back to ancient Rome. 
Many foreign jurisdictions have had tax amnesties, both general 
amnesties and amnesties covering only certain taxes. Some 
foreign amnesties have yielded larger collections than those of 
U.S. States. For example, the general Argentine amnesty of 1995 
yielded about $3.9 billion.\72\ The Irish amnesty of 1988 
yielded more than $700 million.\73\ Some foreign countries have 
made repeated use of amnesties, whereas no U.S. State has had 
more than three general amnesties. Ireland offered five 
amnesties in six years, and Italy has had more than a dozen 
    \72\ ``Argentine Tax Amnesty Promises Big Yield,'' Tax Notes 
International, December 14, 1995, p. 240.
    \73\ ``Irish Run Out of Luck over Tax Amnesty,'' The Independent, 
July 4, 1993, p. 6.
    There are several aspects of foreign amnesties that limit 
their comparability with the United States. In most foreign 
countries, a larger portion of the national economy escapes the 
tax system than probably occurs in the United States.\74\ For 
example, many estimates put the portion of the Italian economy 
that escapes taxation in the 20 percent range.\75\ Most foreign 
countries have larger underground economies than the United 
States, which means that they have greater reliance on cash and 
barter as forms of payment.\76\ Many foreign systems also 
exempt transactions occurring outside the country from 
taxation, giving an incentive to move untaxed profits outside 
the country. Few foreign countries have reached the U.S. 
Federal level of development of enforcement mechanisms, 
especially with regard to use of computer technology and 
requirements of withholding and information reporting. In fact, 
some foreign amnesties have been used as an accompaniment to 
introducing enforcement measures that incurred popular 
resistance, but that are already utilized in the United 
    \74\ L. Talley and W. Morrison, Tax Amnesty: State and European 
Experience,'' Congressional Research Service, p. 13 (1984).
    \75\ The Taxing Problem Italy Faces,'' Financial Times, August 7, 
1989, p. I-15.
    \76\ Talley and Morrison, supra, pp. 13-14.
    \77\ Ibid.
    Some foreign countries that have recently adopted 
sophisticated tax systems, such as those in Eastern Europe, 
have relied on amnesties to deal with honest confusion as to 
what tax liabilities were owed.\78\ Some foreign amnesties, 
including several of the Argentine tax amnesties, have 
coincided with a change in government, and thus suggest a 
repudiation of the previous government's policies.\79\ Another 
factor in the comparability of the foreign experience is that 
many foreign jurisdictions, including those of Western Europe, 
raise a high proportion of revenue from value-added taxes and 
other direct taxes, as opposed to personal income taxes. 
Because value-added taxes are collected by businesses, amnesty 
programs in these countries can generate large totals as a 
result of a few large payments by delinquent businesses.
    \78\ ``Russia Announces Tax Amnesty,'' Tax Notes International, 
November 16, 1993, p. 220.
    \79\ ``Study of Tax Amnesty Programs,'' Internal Revenue Service, 
supra, p. 13.
    The terms offered to taxpayers in many foreign amnesties 
also have been different from U.S. state practices. Many 
foreign amnesties have not only abated penalties but also 
interest and even liabilities for tax. In the 1996 Venezuelan 
amnesty, tax liabilities of participating taxpayers were 
reduced by 75 percent and, in the 1974 Panamanian amnesty, by 
80 percent. Most foreign amnesties have allowed taxpayers with 
accounts receivable or in civil tax litigation to participate. 
Some, such as the 1995 Argentine amnesty, have even allowed 
participation of taxpayers involved in criminal tax 
proceedings. Many foreign systems allow the national tax 
administration to waive penalties and interest as a matter of 
administrative discretion, which can be used for ``standing'' 
offers of amnesty.
    As a result of the level of their economic development and 
the sophistication of their tax systems, the experiences of 
Western Europe and the other countries in the Organization for 
Economic Cooperation and Development are probably most relevant 
to the United States. In 1982, France undertook both a general 
tax amnesty and a special program to encourage repatriation of 
untaxed assets from abroad. The general amnesty applied to all 
income and value-added taxes, and offered an abatement of both 
interest and penalties for participating taxpayers. Collections 
were relatively small compared to U.S. State amnesties, 
amounting to about $19 million from 2,786 taxpayers.\80\ Under 
the repatriation program, French residents who brought back 
capital from abroad that represented undeclared income or that 
was illegally exported were taxed at a flat rate of 25 percent 
(regardless of the original rate owed, which was in most cases 
higher). The repatriation program had 276 participants, from 
whom a total of about $22 million was collected.\81\ In 1986, 
France undertook a second special amnesty for assets held 
abroad with a tax rate of 10 percent on the assets repatriated.
    \80\ Talley and Morrison, supra, p. 14.
    \81\ Ibid., p. 16.
    Ireland conducted a total of five amnesty programs in six 
years, with general amnesties in 1988 and 1993, both of which 
received considerable publicity. The 1988 general amnesty 
yielded more than $700 million in collections.\82\ This amnesty 
offered participants a waiver of all penalties and interest and 
was publicized as an opportunity to pay tax liabilities before 
increased penalties, interest and enforcement measures were 
adopted as part of an overall tax reform. Although the 1988 
amnesty was publicized as a one-time opportunity, Ireland 
undertook a second general amnesty in 1993. Faced with a budget 
deficit, the Irish government announced a special amnesty for 
repatriation of undeclared income from abroad.\83\ In addition 
to waiver of all penalties and interest and promises of 
confidentiality, the repatriated funds were subject to a 
special low rate of 15 percent, in contrast to normal Irish tax 
rates which rose above 50 percent. The amnesty was criticized 
by the parliamentary opposition and trade unions as a 
concession to wealthy taxpayers.\84\ In response, the Irish 
government also adopted a general tax amnesty.\85\ The general 
amnesty offered abatement of both penalties and interest, 
although no liabilities were reduced. Collections for the 1993 
amnesty were widely reported to be significantly lower than for 
the 1988 amnesty,\86\ which is consistent with the view that 
repeated tax amnesties decrease taxpayer expectations of 
enforcement.\87\ Moreover, the 1993 amnesty has continued to 
receive negative publicity, especially due to the revelation 
that the chief suspect in Ireland's biggest robbery benefited 
from the amnesty for a large tax deficiency.\88\
    \82\ ``Irish Run Out of Luck over Tax Amnesty,'' supra.
    \83\ Ibid.
    \84\ ``Irish Count on Windfall from Tax Amnesty to Ease Budget 
Shortfall,'' The Guardian, November 27, 1993, p. 37.
    \85\ Ibid.
    \86\ ``Irish Run Out of Luck over Tax Amnesty,'' supra.
    \87\ See discussion at pp. 11-12.
    \88\ ``Someone's Got to Do It,'' The Guardian, December 4, 1995, p. 
    Italy has conducted more than a dozen tax amnesties, an 
average recently of about one every two years. Collections in 
recent Italian amnesties have not been large, supporting the 
view that repeated use of amnesties reduces their 
effectiveness.\89\ For example, for its general amnesty in 
1982, the Italian government predicted collections of $4.6 
billion, but actual collections totaled less than $700,000.\90\ 
Tax amnesties have occurred so regularly in Italy that the 
expectation of future amnesties has been cited as a factor in 
the low national level of tax compliance.\91\
    \89\ See discussion at pp. 11-12.
    \90\ Talley and Morrison, supra, p. 17.
    \91\ ``The Taxing Problem Italy Faces,'' supra.
    In Canada, a flexible program of tax ``compassion'' was 
begun in 1993 by the Ministry of National Revenue, utilizing 
the Ministry's discretionary authority to abate penalties and 
interest.\92\ Non-filers who voluntarily came forward were 
offered a waiver of all penalties. Others with outstanding tax 
liabilities were promised the opportunity to pay in 
installments and offered the possibility of penalty waivers. 
The tax administrations of several countries, including 
Germany, the Netherlands, Sweden, Norway and Denmark, have in 
recent years had ``standing'' offers of amnesty, under which 
the national tax administration has committed to use its 
authority to abate all or a portion of penalties or interest 
for taxpayers who voluntarily pay their tax liabilities.\93\
    \92\ ``Hard-hit Taxpayers May Get Amnesty from Fines,'' Toronto 
Star, August 17, 1993, p. A1.
    \93\ ``Study of Tax Amnesty Programs,'' Internal Revenue Service, 
supra, pp. 13-17.
    There has been no tax amnesty in the United Kingdom in 
recent years.
    Although not a foreign country, the amnesties conducted by 
Puerto Rico are also relevant. Puerto Rico undertook general 
tax amnesties in both 1988 and 1991 on very similar terms. 
Participating taxpayers were offered an abatement of all 
interest and penalties, plus a special flat rate of 20 percent 
for payment of liabilities (compared with a top individual rate 
of 41 percent in 1988). Both taxpayers under audit and those 
involved in civil tax litigation were entitled to participate.


    The following are the principal parameters employed by the 
staff of the Joint Committee on Taxation in estimating the 
individual income tax amnesty proposals contained on the 
revenue table in the Appendix:
    1. The IRS would be given a lead time of approximately six 
to nine months between the date of enactment and the 
commencement of amnesty. A shorter period would not be feasible 
because it would not give the IRS sufficient time to redeploy 
and train staff resources, to develop and print appropriate 
forms, applications, and instructions, and to publicize the 
coming of tax amnesty. A significantly longer period would not 
be desirable because it would increase the open window during 
which taxpayers may alter their current behavior in 
anticipation of the amnesty.
    2. The determination of the exact starting and ending dates 
of the amnesty must be done carefully to provide the optimal 
scheduling of the amnesty. For example, an amnesty that 
overlapped the April 15 filing date could seriously overburden 
the administrative systems of the IRS and lead to serious 
difficulties with or the failure of the filing season or of the 
amnesty (or of both). Alternatively, an amnesty that occurred 
solely during the summer months or during the Thanksgiving 
through New Year's Day holiday period might not achieve optimal 
results because many individuals' attention would be focussed 
elsewhere. Most State amnesties have occurred in the fall; that 
is likely to be the optimal time for a Federal amnesty as well. 
One alternative would be to permit the IRS to select the exact 
starting and ending dates within legislatively established 
    3. The amnesty would be approximately 90 days in length. A 
shorter period may not allow sufficient time for the amnesty 
publicity to affect taxpayers' behavior. Too lengthy a period 
may cause taxpayers to wait until later in the amnesty period 
to come forward; some of those who delay may never come 
    4. Participants would be permitted to pay through 
installment agreements. Although this is not a universal 
feature of State amnesties, it is a necessary feature of a 
Federal amnesty seeking to maximize receipts. This is because 
the average size of the amounts owed to the Federal Government 
is significantly larger than the average size of amounts owed 
to State governments. Absent installment agreements, some 
taxpayers who wish to avail themselves of amnesty would be 
unable to do so because they could not afford to make one lump 
payment. The permissible period for installment payments cannot 
be too lengthy, however, because increasing its length also 
increases the risk of loss attributable to taxpayers being 
unable to meet their installment agreement obligations. A 
taxpayer would continue to owe interest (as under present law) 
during the period the taxpayer was making installment payments.
    5. The amnesty would apply to all open tax years (except as 
noted in the next item). Taxpayers who, prior to the 
commencement of the amnesty, resolved a dispute with the IRS 
and paid the amount owed would not be eligible for a refund of 
amounts paid (although if they had waited for the amnesty to 
commence, they would have paid a lesser amount).
    6. Amnesty would not be available with respect to 
liabilities incurred in the year in which amnesty is announced 
nor in the year of the amnesty.
    7. Amnesty would not be available to individuals currently 
under criminal investigation.
    8. Participants in the amnesty generally would not be 
subject to criminal penalties.
    9. An amnesty would require widespread publicity. Most 
States provided significant funding for amnesty publicity; 
significant additional funding would also be necessary so the 
IRS can publicize the Federal tax amnesty.
    10. The amnesty would explicitly state that no future 
amnesty would be offered. There are, however, several reasons 
why not all taxpayers will necessarily believe that that will 
be true. First, several States have had second amnesties 
despite pronouncements that the first amnesty would be the only 
amnesty. Second, it is not legally possible to prevent future 
Congresses from enacting a subsequent amnesty.
    In addition to these design parameters, the staff of the 
Joint Committee on Taxation made the following assumptions for 
the purposes of preparing the revenue estimates in the 
    1. There would be a reduction in receipts attributable to 
IRS staff redeployment regardless of how staffing is provided 
for the amnesty. If no new employees are provided to the IRS, 
but rather existing IRS employees are redeployed, there would 
be a reduction in receipts due to reallocation of IRS 
resources. This would occur because IRS employees who would 
otherwise perform audits or bring in collection receipts would 
instead process requests to be included in the amnesty program. 
Even if the IRS is provided with additional personnel to handle 
the work created by a Federal tax amnesty, this effect would 
not be eliminated because there are practical limitations on 
how rapidly the IRS can hire and train new employees; 
consequently, some redeployment (and consequent revenue loss) 
would be inevitable. Providing the IRS with more lead time for 
hiring and training can minimize their redeployment losses, but 
providing more time would also increase the period during which 
some taxpayers would reduce their current compliance levels in 
anticipation of the amnesty (see next item), which would 
further reduce receipts.
    2. There would be a reduction in receipts upon announcement 
of amnesty even if amnesty is not enacted. There are two ways 
in which this would occur. First, some taxpayers currently 
involved in disputes with the IRS would cease working to 
resolve those disputes in anticipation of a possibly better 
deal under amnesty. Consequently, voluntary enforcement 
collections would be substantially reduced. Second, some 
taxpayers might alter their current behavior and reduce their 
compliance in the current year in the erroneous anticipation of 
amnesty applying to them. A small number might do so by ceasing 
to file tax returns; those individuals are, however, likely to 
be caught. Most of those who reduce their compliance are likely 
to do so by altering slightly reporting positions taken on tax 
returns that they will still file with the IRS. The IRS may be 
unable to either detect or respond to numerous relatively small 
changes on tax returns.
    3. Two factors would result in a reduction in receipts in 
years following the amnesty. First, some taxpayers may alter 
their behavior to reduce their compliance in anticipation of a 
future amnesty (whether or not that is explicitly ruled out as 
part of the first amnesty). Second, some taxpayers may believe 
that amnesty was unfair, in that taxpayers who cheated and then 
took advantage of the amnesty receive a ``better deal'' from 
the Government than those who voluntarily complied with the 
laws (or who did not but were caught by the IRS prior to the 
amnesty). To the extent that some of these taxpayers respond to 
these perceptions of unfairness by reducing their compliance in 
the future, there would be a reduction in receipts in years 
following the amnesty.
    4. Tax amnesties generally are not tailored to the 
circumstances of individual taxpayers, but instead offer 
identical terms to everyone. For example, the Minnesota amnesty 
forgave 20 percent of the total interest and penalties owed for 
taxpayers who had filed returns (up to a $2,000 limit). Amnesty 
would consequently have two opposing effects. On the one hand, 
it would accelerate collections due to more prompt payment of 
amounts owed than would otherwise occur through the normal 
collection process. On the other hand, however, amnesty would 
reduce collections in that some taxpayers will avail themselves 
of amnesty (and the consequent reduction in the total amount 
owed) who may otherwise have the ability to pay a larger 
portion of the total owed and who would have paid it in the 
course of the normal collection process. This means that there 
would be an initial acceleration into the first year of the 
amnesty amounts that would otherwise be collected (absent the 
amnesty) in later years, but that there also would be a loss in 
the later years because some taxpayers would pay less under 
amnesty than they would have paid in the course of the normal 
collection process.


    There are several factors that may influence the decision 
as to whether to employ a Federal tax amnesty. One is the 
revenue consequences of a Federal tax amnesty. The staff of the 
Joint Committee on Taxation estimates that a Federal tax 
amnesty would result in a net revenue loss to the Federal 
Government. This net revenue loss occurs primarily because a 
Federal tax amnesty will have the long-run effect of modestly 
reducing overall taxpayer compliance with Federal tax laws.
    Other factors in addition to revenue effects may 
significantly influence the decision as to whether to employ a 
Federal tax amnesty. One factor might be whether the Internal 
Revenue Code is being significantly reformed or replaced. For 
example, in the context of a complete restructuring of the 
Internal Revenue Code, consideration might be given to 
implementing a Federal tax amnesty to ``wipe the slate clean'' 
for prior noncompliance.
    Another context in which a Federal tax amnesty might be 
considered is whether it might be an appropriate element of 
legislation to restore taxpayers' confidence in the fairness of 
the Internal Revenue Service. Some might argue that a Federal 
tax amnesty would provide an appropriate opportunity to resolve 
prior disagreements with the IRS, which could help restore 
confidence. Others might argue that amnesty would not be an 
appropriate mechanism to restore confidence in the fairness of 
the tax system, since some taxpayers who fulfilled their tax 
obligations could view it as unfair that others received a 
``better deal'' under amnesty.

                                      APPENDIX--ESTIMATED BUDGET EFFECTS OF POSSIBLE FEDERAL TAX AMNESTY PROPOSALS                                      
                                                                 Fiscal Years 1998-2007                                                                 
                                                                  [Billions of Dollars]                                                                 
           Proposal                  Effective       1998    1999    2000    2001    2002    2003    2004    2005    2006    2007   1998-2002  1998-2007
1. Individual taxpayers only:   aco 10/1/98         ......     4.2    -1.7    -1.7    -1.8    -1.8    -1.8    -1.9    -0.7    -0.7      -1.0       -8.0 
 waiver of penalties and 50%                                                                                                                            
 interest, include accounts                                                                                                                             
 receivable, for taxpayer                                                                                                                               
 years 1996 and prior; amnesty                                                                                                                          
 period for 90 days.                                                                                                                                    
2. Individual taxpayers who     aco 10/1/98         ......   (\1\)   (\2\)   (\2\)   (\2\)   (\2\)   (\2\)   (\2\)   (\2\)   (\2\)     (\2\)       -0.2 
 have not filed returns with                                                                                                                            
 the IRS only: waiver of                                                                                                                                
 penalties and 50% interest                                                                                                                             
 for taxable years 1996 and                                                                                                                             
 prior; amnesty period for 90                                                                                                                           
\1\ Gain of less than $10 million.                                                                                                                      
\2\ Loss of less than $50 million.                                                                                                                      
Legend for ``Effective'' column: aco=amnesty commencing on                                                                                              
Note. Details may not add to totals due to rounding. Estimates do not include outlay effects for promoting and administering a Federal income tax       
Source: Joint Committee on Taxation.