Tax Policy: Summary of Estimates of the Costs of the Federal Tax 
System (26-AUG-05, GAO-05-878). 				 
                                                                 
In 2005, Americans will pay about $2.1 trillion in combined	 
federal taxes, including income, payroll, and excise taxes, or	 
about 16.8 percent of Gross Domestic Product (GDP). However, the 
amount of taxes paid does not reflect the total cost to taxpayers
of the federal tax system. In addition to taxes paid, taxpayers  
also bear compliance costs and efficiency costs. Understanding	 
the magnitude of these additional costs is important because	 
every dollar spent on compliance and lost due to inefficiency	 
represents a dollar that society could have spent for other	 
purposes. In response to a congressional request for information 
on the magnitude of the compliance and efficiency costs of the	 
current federal tax system, this study describes the nature of	 
these costs, presents the difficulties associated with estimating
them, and summarizes existing estimates of their magnitude. GAO  
did not make independent estimates of compliance or efficiency	 
costs nor did we replicate any of the studies. GAO is not making 
any recommendations in this report.				 
-------------------------Indexing Terms------------------------- 
REPORTNUM:   GAO-05-878 					        
    ACCNO:   A34598						        
  TITLE:     Tax Policy: Summary of Estimates of the Costs of the     
Federal Tax System						 
     DATE:   08/26/2005 
  SUBJECT:   Cost analysis					 
	     Federal taxes					 
	     Financial analysis 				 
	     Tax administration 				 
	     Tax administration systems 			 
	     Tax expenditures					 
	     Taxpayers						 
	     Voluntary compliance				 
	     Cost estimates					 
	     Gross Domestic Product				 

******************************************************************
** This file contains an ASCII representation of the text of a  **
** GAO Product.                                                 **
**                                                              **
** No attempt has been made to display graphic images, although **
** figure captions are reproduced.  Tables are included, but    **
** may not resemble those in the printed version.               **
**                                                              **
** Please see the PDF (Portable Document Format) file, when     **
** available, for a complete electronic file of the printed     **
** document's contents.                                         **
**                                                              **
******************************************************************
GAO-05-878

                 United States Government Accountability Office

GAO

                       Report to Congressional Requesters

August 2005

TAX POLICY

          Summary of Estimates of the Costs of the Federal Tax System

                                       a

GAO-05-878

[IMG]

August 2005

TAX POLICY

Summary of Estimates of the Costs of the Federal Tax System

                                 What GAO Found

Complying with the federal tax system costs taxpayers time and money.
Estimating total compliance costs is difficult because neither the
government nor taxpayers maintain regular accounts of these costs and
federal tax requirements often overlap with recordkeeping and reporting
that taxpayers do for other purposes. Although available estimates are
uncertain, taken together, they suggest that total compliance costs are
large. For example, combining the lowest available estimates for the
personal and corporate income tax yields a total of $107 billion (roughly
1 percent of GDP) per year. As noted, whether this is a definitive lower
bound for compliance costs is uncertain.

The tax system also results in economic efficiency costs because tax rules
cause individuals to change their behavior in ways that ultimately leave
them with lower-valued combinations of consumption and leisure than they
would have obtained if the tax system did not affect their behavior.
Estimating efficiency costs is very challenging because the tax system has
such extensive and diverse effects on behavior. In fact, we found no
comprehensive estimates of the efficiency costs of the current federal tax
system. The two most comprehensive studies we found suggest that these
costs are large-on the order of magnitude of 2 to 5 percent of GDP each
year (as of the mid-1990s). However, the actual efficiency costs of the
current tax system may not fall within this range because of uncertainty
surrounding taxpayers' behavioral responses, changes in the tax code and
the economy since the mid-1990s, and the fact that the two studies did not
cover the full scope of efficiency costs.

The goal of tax policy is not to eliminate compliance and efficiency
costs. The goal of tax policy is to design a tax system that produces the
desired amount of revenue and balances the minimization of these costs
with other objectives, such as equity, transparency, and administrability.
In addition, whether compliance and efficiency costs could be reduced by
redesigning the tax system and, if so, by how much would depend critically
upon the detailed characteristics of the new tax system.

  Components of the Total Cost of a Tax to Taxpayers United States Government
                             Accountability Office

Contents

                                    Letter 1

Results in Brief 3

Background 4

Estimates of Tax Compliance Costs Are Uncertain Because

Taxpayers Generally Do Not Record Them; However, Total Costs

Are Likely to Be Large 6 Efficiency Costs Arising from Tax-Induced Changes
in Behavior Are Likely to Be Large but Can Only Be Modeled with
Considerable Uncertainty 16

Agency Comments and Our Evaluation 25

Bibliography

Tables        Table 1: Features of the Tax System as of July 2005        4 
              Table 2: Available Estimates of Tax Compliance Costs for     
                            Individual Taxpayers (Including Self-employed) 13 
              Table 3: Available Estimates of Tax Compliance Costs for     
                            Corporations and Partnerships                  14 
            Table 4: Estimates of the Efficiency Gains from the Removal of 
                   Selected Distortions in the Current Tax System          20 
Figures    Figure 1: Components of the Total Cost of a Tax to Taxpayers  6 
                 Figure 2: Compliance Burden Is One Cost Taxpayers Face in 
                            Complying with the Tax System                   7 
              Figure 3: Efficiency Costs Are One Cost Taxpayers Face in    
                            Complying with the Tax System                  17 

This is a work of the U.S. government and is not subject to copyright
protection in the United States. It may be reproduced and distributed in
its entirety without further permission from GAO. However, because this
work may contain copyrighted images or other material, permission from the
copyright holder may be necessary if you wish to reproduce this material
separately.

A

United States Government Accountability Office Washington, D.C. 20548

August 26, 2005

The Honorable Thomas DeLay
Majority Leader
House of Representatives

The Honorable John Linder
House of Representatives

In 2005, Americans will pay about $2.1 trillion in combined federal taxes,
including income, payroll, and excise taxes, or about 16.8 percent of
Gross
Domestic Product (GDP). These taxes fund the services provided by
government. As taxpayers, we balance the costs of taxes with the benefits
of government.

However, the amount of tax revenue collected does not reflect the total
cost to taxpayers of the federal tax system. In addition to taxes paid,
taxpayers also bear compliance costs and efficiency costs. Compliance
costs include the value of the taxpayer's time and resources, along with
any
out-of-pocket costs to paid tax preparers and other tax advisors, spent to
ensure compliance with the tax laws. Efficiency costs result when taxes
alter the economic decisions that people make-decisions such as how
much to work, how much to save, what to consume, and where to invest-
in ways that reduce overall well-being. Understanding the magnitude of
these additional costs is important, because every dollar spent on
compliance and lost due to inefficiency represents a dollar that society
could have spent for other purposes. Moreover, these costs could change if
policymakers decide that the projected imbalances between federal
revenues and expenditures can only be addressed by changes to both sides
of the fiscal system.

Despite the existence of these costs and their size, the goal of tax
policy is
not to eliminate compliance and efficiency costs. The goal of tax policy
is
to design a tax system that produces the desired amount of revenue and
balances the minimization of compliance and efficiency costs with other
objectives, such as equity, transparency, and administrability.

In response to your May 17, 2004, letter to the Comptroller General and
follow-up meetings with your staffs, this report summarizes what is known
from the academic literature about the current tax system's compliance
and efficiency costs. For both costs this report describes their nature,

presents the difficulties associated with estimating them, and summarizes
existing estimates of their magnitude.

To describe the nature of compliance costs and the difficulty in
estimating their magnitude, we reviewed the relevant economic literature.
To summarize the relevant cost estimates available from the literature we
reviewed studies that present original, empirically based compliance cost
estimates. We report on all studies that we found that estimate the
compliance costs to individuals or businesses attributable to the current
(post-1986) tax system. We did not make independent estimates of
compliance costs nor did we replicate any of the studies.

To describe the nature of efficiency costs and the difficulty in
estimating their magnitude, we reviewed the relevant economic literature.
To summarize the relevant cost estimates available from the literature, we
reviewed 1,567 abstracts from peer-reviewed journal articles and books
(dating back to January 1986) and nonacademic research organization
publications and economic working papers (dating back to January 1, 2000)
that contained some reference to taxes, costs, and efficiency. On the
basis of reading the abstract or paper, we excluded studies that did not
produce original empirical estimates or that made estimates for the
federal tax system existing prior to the major 1986 tax reform. We found
no study that comprehensively estimated the cost of all behavioral
distortions caused by the tax system. We also consulted with experts
outside of GAO, asking them if they were aware of studies we overlooked.
We report on all studies that we found that estimate the efficiency costs
attributable to selected aspects of the current system. The coauthor of
one of these studies is a key contributor to this report. To ensure the
objectivity and independence of this product, a second economist, within
GAO, verified that the work in this section was presented in an accurate
and unbiased manner. We also report on a set of studies that estimate
efficiency gains that would result from replacing the current federal tax
system with an alternative system that raised the same amount of revenue.
As was the case with compliance costs, we did not make independent
estimates of efficiency costs nor did we replicate any of the studies.

We conducted our work in Washington, D.C., from November 2004 through July
2005 in accordance with generally accepted government auditing standards.
We provided a draft of this report in August 2005 to the Commissioner of
Internal Revenue. We received technical comments via email from the IRS
Office of Research. Where appropriate, we made changes in our report in
response to these comments.

GAO is making no recommendations in this report.

Results in Brief	The federal tax system imposes a wide range of
recordkeeping, computational, and filing requirements upon businesses and
individuals. Complying with these requirements costs taxpayers' time and
money. Neither the government nor taxpayers maintain regular accounts of
these costs and many important elements of the costs are difficult to
measure because, among other things, federal tax requirements often
overlap with recordkeeping and reporting that taxpayers do for other
purposes. Available estimates of aggregate compliance costs vary in terms
of the scope of costs that they include, the tax years that are
represented, assumptions regarding the monetary value of an hour of time
spent on tax compliance, and other methodological factors. Although the
Information Collection Budget for the Department of the Treasury contains
the most comprehensive estimate, analysts both within Treasury and outside
consider this estimate to be very uncertain because it is based on survey
data from the early 1980s that have been updated each year with an overly
simplified methodology. The Internal Revenue Service (IRS) is in the
middle of a long-term research effort to improve its methodology for
estimating compliance burden. Preliminary, partial results from that new
effort and evidence from other researchers indicate that compliance costs
are large, even though the total remains uncertain. For example, combining
the lowest available (and incomplete) estimates of individual and
corporate compliance cost yields a total of $107 billion (roughly 1
percent of GDP) per year; however, other studies estimate costs 1.5 times
as large. Whether compliance costs could be reduced by redesigning the tax
system and, if so, by how much would depend upon the details of the new
tax system, since all tax systems have compliance costs.

The tax system also results in economic efficiency costs. These costs
occur when tax rules cause individuals to change their work, savings,
consumption, and investment behavior in ways that ultimately leave them
with a combination of consumption and leisure (now and in the future) that
they value less than the combination they would have obtained under a tax
system that did not alter their behavior. Some of the incentives to change
behavior are intentionally designed into the tax system, others are
unintended consequences of rules designed to achieve other objectives,
such as equity or increased revenue yields. Estimating efficiency costs is
very challenging because federal taxes have such extensive and diverse
effects on behavior. In fact, we found no comprehensive estimates of the
efficiency costs of the current federal tax system. Nonetheless, we did
find

some studies that estimate the efficiency costs attributable to selected
aspects of that system. Although none of these studies, either
individually or in the aggregate, provide a basis for estimating the total
efficiency cost of the tax system, they do indicate that those total costs
are likely to be large. The two most comprehensive studies we found show
costs on the order of magnitude of 2 to 5 percent of GDP each year (as of
the mid1990s). However, the actual efficiency costs of the current tax
system may not fall within this range because of uncertainty surrounding
taxpayers' behavioral responses; changes in the tax code and the economy
since the mid-1990s; and the fact that the two studies did not cover the
full scope of efficiency costs. The current tax system, as well as all
major alternatives, imposes efficiency costs. As with the case of
compliance costs, the extent to which efficiency costs could be reduced by
tax system redesign would depend upon the details of the new system.

Background	The current tax system in the United States consists primarily
of five types of taxes: (1) personal income taxes; (2) corporate income
taxes; (3) social insurance taxes (employee and employer contributions for
Social Security, Medicare, and unemployment compensation); (4) estate and
gift taxes, and (5) a variety of other taxes such as excise taxes on goods
and services. Table 1 summarizes several selected features of the current
federal tax system.

              Table 1: Features of the Tax System as of July 2005

Type of tax Tax base Tax rates

Personal Income Taxes (PIT) Regular PIT Regular PIT

Personal income, including income from wages, Graduated rate structure:
interest and dividends, capital gains, and small Statutory marginal rates
of 10, 15, 25, 28, 33 percent,
business income. and 35 percent. Deductions and other tax

expenditures, such as refundable tax credits, such as Numerous tax
expenditures exist that reduce the the Earned-Income Tax Credit, create a
group of size of the tax base. taxpayers who have no tax liability or
receive an outlay

from the tax system.

Personal Alternative Minimum Tax (AMT) Personal AMT

Taxable income exceeding certain threshold 26 or 28 percent depending on
taxable income

amounts based on filing status.	subject to the AMT. Eligible for a credit
for a portion of the AMT paid in a prior year.

(Continued From Previous Page)

     Type of tax            Tax base                     Tax rates            
Corporate Income                                                           
        Taxes              Regular CIT                  Regular CIT
        (CIT)       Corporate profits (total         Statutory marginal rates 
                    revenues less total           fluctuate between 15 and 35 
                    expenses). Numerous tax       percent. Due to special     
                    expenditures exist               provisions, some         
                     that reduce the size of  corporations may pay a marginal 
                          the tax base.               tax rate of 39          
                                                         percent.             
                          Corporate AMT                Corporate AMT          
                    Broader definition of the    20 percent for all corporate 
                    tax base (corporate             income subject to the tax 
                    income) than regular CIT,  less the AMT credit for that   
                          less generous                  tax year.            
                        accounting rules.     

Social Insurance      Social Security        Social Security               
Taxes                                        
                         First $90,000 of                6.2 percent employee 
                         employee wages.                        contribution. 
                                                         6.2 percent employer 
                                                                contribution. 
                                                      12.4 percent for        
                                                       self-employed.         
                         Medicare               Medicare                      
                                                        1.45 percent employee 
                         All wages.                             contribution. 
                                                        1.45 percent employer 
                                                                contribution. 
                                                      2.90 percent for        
                                                       self-employed.         

Unified Transfer Tax- Estate Tax Estate Tax
Estate, Gift, and Fair market value of the decedent's cash and Maximum tax
rate of 47 percent in 2005. As a result
Generation Skipping securities, real estate, trusts, annuities, business
of recent tax legislation, estate tax rates will fluctuate
Taxa interests, and other assets less allowable before the estate tax is
eliminated in 2010. However,

deductions in excess of $1.5 million in 2005. the estate tax will be
reinstated in 2011.
There is an unlimited deduction for transfers to a
surviving spouse.

Gift Tax Gift Tax

Tax is imposed on lifetime taxable transfers of Maximum tax rate of 47
percent in 2005. Rates
gifts of property. Applicable exclusion amount of fluctuate in the same
manner as for the estate tax in
$1 million for 2005. In addition, there is an coming years. Gift tax will
be retained following repeal
annual exclusion of $11,000 per donee and an of estate and generation
skipping tax.
unlimited exclusion for tuition and medical
payments.

Generation Skipping Tax (GST) Generation Skipping Tax

Total generation skipping transfers (such as from 47 percent (or highest
statutory marginal tax rate for a grandparent to a grandchild) in excess
of $1.5 the estate tax) in 2005. GST rates decrease until the million in
2005. tax is repealed in 2010. GST is reinstated in 2011.

Excise and Other Taxes	Selected goods, services, and other items (e.g.,
Various rates apply to different goods, services, and gasoline, alcoholic
beverages, tobacco, airline other items. tickets, etc.).

Source: GAO analysis of IRS information.

aThe unified transfer tax is based on the value of property to be included
in a decedent's estate at death and the value of taxable lifetime gifts
made by the decedent. Beginning in 2004, the estate and gift tax
applicable exclusion amounts differ.

The definition of the tax bases, the tax rates, the various tax credits,
and the specific characteristics of each taxpayer determine how much that
taxpayer must pay to the U.S. Treasury. The total cost of taxes from a
taxpayer's point of view is the sum of the tax liability, the costs of
complying with the tax system, and the efficiency costs that the system
imposes, as shown in figure 1. In subsequent sections we discuss the
various factors that cause the latter two types of costs.

Figure 1: Components of the Total Cost of a Tax to Taxpayers

                                  Source: GAO.

  Estimates of Tax Compliance Costs Are Uncertain Because Taxpayers Generally Do
  Not Record Them; However, Total Costs Are Likely to Be Large

The costs of complying with the tax system are uncertain, but likely to be
large-estimates are roughly on the order of about 1 percent of GDP. The
costs include the computational, reporting, planning, and recordkeeping
requirements of the tax system. Estimates of compliance costs are
uncertain because taxpayers generally do not keep relevant records
documenting their time and money spent complying with the tax system and
many important elements of the costs are difficult to measure because,
among other things, federal tax requirements often overlap with
recordkeeping and reporting that taxpayers do for other purposes.

Taxpayers Incur Costs The federal tax system imposes a wide range of
recordkeeping, planning, When Complying with the computational, and filing
requirements upon businesses and individuals. Computational, Reporting,
Complying with these requirements costs taxpayers time and money. Many

of these requirements are complex, reflecting both the complexity of
ourand Recordkeeping modern economy and intent of policymakers to build
progressivity andRequirements of the Tax various incentives into the tax
system. As shown in figure 2, these costs to System taxpayers are above
and beyond what they pay to the government in taxes.

Figure 2: Compliance Burden Is One Cost Taxpayers Face in Complying with
the Tax System

Source: GAO.

The tax compliance requirements for individuals include the documentation
of their incomes and their qualifications for various exemptions,
deductions, and credits available under the federal income tax. Although
the documentation of income is straightforward for a large proportion of
the individual taxpayer population who earn only labor and interest income
and capital income within a retirement account, substantial numbers of
taxpayers receive income from capital gains, rents, self-employment, and
other sources that involve additional calculations and recordkeeping. In
addition, many individuals must familiarize themselves with (or pay
someone to advise them on) the sometimes complex rules for determining
whether they qualify (and, if so, to what extent) for any of the numerous
tax benefits in the federal tax code. Moreover, in cases where multiple
tax expenditures have similar purposes, taxpayers may have to devote
considerable time to learning and planning in order to make optimal use of
these tax benefits.

For example, the IRS publication Tax Benefits for Education1 outlines 12
tax expenditures, including 4 different tax expenditures for educational
saving. Three of these savings tax expenditures-Coverdell Education
Savings Accounts, Qualified Tuition Programs, and U.S. education savings
bonds-differ across more than a dozen dimensions, including the tax

1Department of the Treasury, IRS, Publication 970, Tax Benefits for
Education, 2004.

penalty that occurs when account balances are not used for qualified
higher education expenses, who may be an eligible beneficiary, annual
contribution limits, and other features. Similarly, 3 other tax
expenditures, all of which help students meet current costs-the Hope
credit, Lifetime Learning credit, and the tuition deduction-differ in
terms of eligibility criteria, benefit levels, and income-related
phase-outs. The use of one of these tax expenditures can affect whether
(or how) a taxpayer is allowed to use the other tax expenditures.
Moreover, the use of one of these tax expenditures may affect a student's
eligibility for other forms of federal assistance for higher education,
such as Pell grants and subsidized loans.2

Tax compliance requirements for businesses are even more extensive and
complex than those for individuals. Rules governing the computation of
taxable income, expense deductions, and tax credits of U.S. corporations
that do business in multiple foreign countries are particularly complex.
But even small businesses face multiple levels of tax requirements of
varying difficulty. In addition to computing and documenting their income,
expenses, and qualifications for various tax credits, businesses with
employees are responsible for collecting and remitting (at varying
intervals) several federal taxes on the incomes of those employees.
Moreover, if the businesses choose to offer their employees retirement
plans and other fringe benefits, they can substantially increase the
number of filings they must make. Businesses also have
information-reporting responsibilities-employers send wage statements to
their employees and to IRS; banks and other financial intermediaries send
investment income statements to clients and to IRS.3 Finally, a relatively
small percentage of all businesses (which nevertheless number in the
hundreds of thousands) are required to participate in the collection of
various federal excise taxes levied on fuels, heavy trucks and trailers,
communications, guns, tobacco, and alcohol, among other products.

2 For a fuller discussion of the difficulties that taxpayers face when
trying to make use of the various tax expenditures for education, see GAO,
Student Aid and Postsecondary Tax Preferences: Limited Research Exists on
Effectiveness of Tools to Assist Students and Families through Title IV
Student Aid and Tax Preferences, GAO-05-684 (Washington, D.C.: July 29,
2005).

3 Although this information reporting increases the compliance burden on
businesses, it does enable IRS to enforce tax compliance by wage earners
and investors at lower cost. This reduction in administrative costs, which
are paid out of the federal budget, means that taxes are slightly lower
than they otherwise would have to be.

The tax filings and records of both individuals and businesses are subject
to review by IRS. Attempts at tax evasion, unintentional filing errors, or
differing interpretations of the tax law can all lead IRS to request more
information from taxpayers and, possibly, to litigation, all of which
would add to the taxpayers' costs.

    Estimates of Compliance Costs Are Uncertain Because Relevant Records Are Not
    Kept

It is difficult for researchers to estimate the total compliance costs of
the federal tax system because taxpayers generally do not keep records of
the time and money spent complying with that tax system. Moreover, many
important elements of the costs are difficult to measure because

o 	recordkeeping and reporting to comply with federal tax requirements
often overlap with recordkeeping and reporting that taxpayers do for other
purposes;

o 	some costs that taxpayers incur in dealing with IRS may result from tax
evasion, rather than tax compliance; and

o 	there is no consensus on the appropriate method for determining the
dollar-value of each hour that an individual spends on tax compliance.

Businesses and individual taxpayers have little, if any, motivation to
keep records of the time and money that they spend specifically on
complying with federal tax requirements. Consequently, when researchers
attempt to collect data on compliance costs, they typically have had to
contact samples of taxpayers, through surveys or interviews, and ask them
for their best recollection of the total time and money they spent on
particular compliance activities. Evidence from one past compliance cost
study suggests that respondent recall error may be substantial.4 Moreover,
conveying the appropriate definition of federal tax compliance costs to
mail or telephone survey respondents and getting those respondents to
apply that definition uniformly is difficult, as will be clear from the
discussion below. In fact, most of the studies that we found had
considerable difficulty getting taxpayers, particularly businesses, to

4 As part of the study that forms the basis for IRS's current estimates of
compliance costs, the study's authors used two data collection methods-a
mail survey and a diary study. The cost estimates yielded by the two
methods varied significantly. The average burden of 14.8 hours for the
mail survey respondents was 78 percent higher than the average burden of
8.3 hours reported by diary respondents.

respond at all. A high response rate is important for the accuracy of
estimates, particularly for the population of business taxpayers that is
highly diverse in terms of size, industry, and business organization and,
therefore, is likely to be diverse in terms of its compliance costs.

A major difficulty in measuring compliance costs is disentangling
accounting and recordkeeping costs due to taxes from the costs that would
have been incurred in the absence of the federal tax system. For example,
where the rules regarding the calculation of income for tax purposes
coincide with the rules for determining income for financial statement
purposes, the additional costs of taxation can be minimal. Public
corporations must file financial statements with the Securities and
Exchange Commission; other businesses are required to provide financial
statements to banks and other lenders when they are seeking credit.
Similarly, many individuals need to document their income when applying
for mortgages or financial aid for college. Moreover, most businesses and
individuals must comply with state, and sometimes local, income tax filing
requirements that are often very similar to those imposed by the federal
tax system. Consequently, if one wishes to define the compliance costs
attributable to the federal tax system as only those costs that would not
exist in the absence of that system, then one needs to assume what the
requirements of state and local tax systems, as well as those of financial
accounting systems, would look like in the absence of the current federal
tax system and make sure that none of the costs of complying with those
requirements are included in the estimate of federal compliance cost. The
studies that we reviewed typically exclude normal business accounting
costs from their definition of federal tax compliance costs; however, they
do not exclude the costs of complying with tax requirements that are
imposed by both federal and state governments.

Two additional definitional issues that have been discussed in the
compliance cost literature are how to treat costs associated with either
tax planning or tax evasion. Although tax planning expenses are excluded
from the definitions of compliance costs established by the Paperwork
Reduction Acts of 1980 and 1995, some studies that we reviewed do include
such expenses in their definition of compliance costs.5 None of the
studies we reviewed recommended the inclusion of costs associated with
illegal tax evasion; however, it is difficult for researchers to
completely exclude tax evasion costs from their estimates because, in many
cases, the

5 Pub. L. No. 96-511 (1980) and Pub. L. No. 104-13 (1995).

determination of whether a particular activity constitutes tax evasion or
not is not made until a year or more after the activity is completed (and
recorded in a survey).

There is no consensus among researchers regarding the appropriate monetary
value to be assigned to each hour of time spent on tax compliance
activities. Some of the many issues involved in the choice of the
appropriate value are: whether to use different values for time that is
taken away from the taxpayer's leisure and time taken away from additional
hours at work; whether to use different values for time spent on
activities that the taxpayer could have paid someone to do and time spent
on activities that only the taxpayer could do; whether to relate the value
to each taxpayer's own hourly wage rate or to some uniform wage rate; and
whether to use a before-tax or after-tax wage rate. The choice of the
monetary value of an hour spent on tax compliance obviously will have a
substantial effect on the estimated magnitude of aggregate compliance
costs.

    Existing Estimates of the Compliance Costs of the Federal Tax System, Though
    Incomplete and Imprecise, Suggest Those Costs Are Large

The government's Information Collection Budget (ICB), for the Department
of the Treasury, annually estimates the time that taxpayers spend on
prefiling and filing activities for every form issued by the Treasury. The
vast majority of these forms are tax related. Treasury has estimated that
during fiscal year 2004 individuals, businesses, and exempt organizations
spent a total of 6.4 billion hours on Treasury's forms. Treasury does not
convert this time estimate into a monetary value. If this time burden were
monetized at rates between $15 per hour and $30 per hour (the range used
for individual taxpayers in the studies that we found), the total cost
would amount to between about $100 billion and $200 billion.

Many analysts within Treasury and outside believe that the ICB estimates
are not very accurate. The estimates are based primarily on an IRS model
that was developed with survey data for tax year 1983. The simplicity of
the model's approach for updating the data each year leaves it with a very
limited ability to adjust for changes in compliance burden resulting from

changes in tax policy or tax system administration.6 Indeed, in an effort
to more accurately measure paperwork burden, IRS is currently developing a
revised methodology for estimating the compliance burden individuals incur
when complying with the U.S. tax system. The new model will reflect a
change in the approach for measuring burden: it focuses on the taxpayer
and the taxpayer's characteristics rather than the forms the taxpayer
uses. Key drivers of taxpayer burden in the model are the taxpayers'
activities and the method of return preparation (by a paid preparer, by
the taxpayer with tax preparation computer software, or by the taxpayer
without computer assistance). In addition, IRS is also working on new
methodologies for estimating the compliance burdens of various types of
businesses and tax-exempt entities, and the burdens associated with
postfiling activities.

Most of the other estimates of the compliance burden that we found for
large segments of the taxpayer population were produced by contractors to
IRS or by analysts who used IRS's data. The disparities among the
estimates, their likely margins of error, and their incompleteness (few
include postfiling costs, such as audits, and all are limited to the
income tax) prevent a more definitive conclusion about the total
compliance burden of the U.S tax system.

Studies we found that focus on the compliance costs of individual
taxpayers estimate it to be between $67 billion and a little over $100
billion per year. The variation in these estimates is primarily
attributable to the wide range in monetary values that the different
studies apply to an hour spent on tax compliance activities. These studies
are summarized and compared in table 2.

6 The only factors used in the model are measures of return size (numbers
of forms and attachments), form size (number of words or number of line
items), the number of references to the tax code in forms, and
instructions, and the number of line items requiring records. The model
does not take into account new developments in tax preparation technology,
such a personal computer software and electronic filing, nor does it
differentiate between simple and complex types of line items on tax forms.

Table 2: Available Estimates of Tax Compliance Costs for Individual
Taxpayers (Including Self-employed)

Compliance Study costs (publication (dollars in Important differences in
scope and date/estimate year)a billions) assumptionsb Notable
methodological issues

       IBM/IRS     67-99 o  Covers taxpayers'  o  Based on surveys of two     
                            time, preparer     samples of taxpayers:          
(2003 /TY 2000)       fees, and any other    one (in 2000) for individuals 
                         out-of-pocket                     who earn only wage 
                                               and investment income; a       
                               expenses.       second (in 2001) for           
                         o  Taxpayers' time is   self-employed individuals.   
                           monetized at $15    
                         per hour for the low  
                         estimate and at       
                         $25 per hour for the  
                         high estimate.        

      Moody-Tax     104 o  Covers taxpayers' and     o  Based on data from an 
                               preparers'             IRS survey of taxpayers 
                        time but no out-of-pocket            for TY 1983; the 
      Foundation        expenses.                    methodology for updating 
                                                                         that 
(2002 /CY 2002)c         o  Taxpayers' and     data is simplistic and does 
                           preparers' time is     not account for             
                          monetized at $30 per     changes in tax preparation 
                                  hour.                     and recordkeeping 
                                                          technology.         

       Slemrod      85 o  Covers taxpayers' time,  o  Based on the author's   
                                preparer             informed judgment of     
(2004 /TY 2004)d    fees, and any other            accumulated research on 
                       out-of-pocket                this topic, including his 
                                                     own study of a sample of 
                               expenses.               Minnesota taxpayers in 
                         o  Taxpayers' time is              1989 (Slemrod and 
                            monetized at $20      Blumenthal) and the IBM/IRS 
                               per hour.                    study.            

Source: GAO analysis of papers cited.

aThe abbreviations CY and TY stand for calendar year and tax year
respectively.

bThe scopes of all of these studies are limited to the prefiling and
filing burden associated with taxpayers' own income tax returns; they do
not cover costs of complying with payroll or income tax responsibilities
with respect to any employees they may hire. None of these studies
includes postfiling costs, such as responding to notices or providing
information for audits.

cMoody included the self-employed in his estimate for business burden. He
also provided detailed estimates by specific tax forms. We used this
form-level detail to approximate his estimate for the selfemployed and
shift that amount ($20 billion) from table 3 to this table to improve
comparability to the other studies.

dThis estimate updates a similar exercise the author undertook in 1996.

Studies we found that focus on the compliance costs of businesses estimate
them to be between about $40 billion and $85 billion per year. None of
these estimates include the costs to businesses of collecting and
remitting income and payroll taxes for their employees. The accuracy of
these business compliance cost estimates is uncertain due to the low rates
of response to their data collection surveys. In addition, the range in
estimates across the studies is due, among other things, to differences in
monetary values used (ranging between $25 per hour and $37.26 per hour),
differences in the business populations covered, and differences in the
tax years covered. These studies are summarized and compared in table 3.

Table 3: Available Estimates of Tax Compliance Costs for Corporations and
                                  Partnerships

Study (publication date / estimate year)a

                                       Compliance costs (dollars in billions)

Important differences in scope and
assumptionsb Notable methodological issues

Moody-Tax 85  o  Covers all corporations and  o  Based on data from an IRS
survey of Foundation partnerships. taxpayers for TY 1983 (with a response
rate (2002 /CY 2002)c  o  Covers taxpayers' and preparers' time of less
than 38 percent); the methodology for

but no out-of-pocket expenses; does not updating that data is simplistic
and does not

cover any postfiling costs. account for changes in tax preparation and

o  Taxpayers' and preparers' time is recordkeeping technology. monetized
at $37.26 per hour.

Slemrod and Venkatesh (2002 /TY 2001)

22  o  Excludes the largest 1,350 corporations, all businesses with less
than $5 million in assets, and all partnerships with less than a certain
number of partners.

o  Covers taxpayers' time, preparer fees, and any other out-of-pocket
expenses, including all postfiling costs.

o  Costs are stated in terms of businesses' actual expenses for in-house
or external tax compliance services.

o  The effective response rate for the survey was only 10.25 percent.

o  Large discrepancies existed between the estimates of outside expenses
given by the taxpayers versus those given by tax professionals.

o  Large discrepancies existed between the asset sizes reported by IRS
versus the sizes reported by the taxpayers themselves. (Some respondents
may have reported as part of a consolidated group when they were sampled
to represent only a single corporation.)

Slemrod 40  o  Covers all corporations and  o  Represents author's best
judgment, based on (2004 /TY 2004) partnerships. his other studies cited
in this table and an

o  Covers taxpayers' time, preparer fees, educated guess about the cost to
small and any other out-of-pocket expenses, businesses other than sole
proprietors. but does not appear to cover postfiling costs.

o  Taxpayers' time is monetized at $25 per hour.

Slemrod and 2  o  Covers only the 1,329 largest  o  Response rate for the
1992 survey was 27.5
Blumenthal corporations. percent.
(1993 /TY 1992)  o  Covers taxpayers' time, preparer fees,  o  Slemrod
(1997) did a follow-up survey in

and any other out-of-pocket expenses, 1996 that had a lower number of
Follow-up study by including all postfiling costs. respondents. The
overall cost estimate was Slemrod  o  Costs are stated in terms of
businesses' roughly the same in each study. (1997/TY 1996) actual
personnel costs for in-house tax

compliance activities and all costs for external tax compliance services.

Source: GAO analysis of the papers cited.

aThe abbreviations CY and TY stand for calendar year and tax year
respectively.

bThe scopes of all of these studies are limited to the taxpayers' own
income tax returns; they do not cover costs of complying with payroll or
income tax responsibilities with respect to their employees.

cThis study estimated that nonprofits spent over $5 billion on compliance
costs (above and beyond the $85 billion). None of the other studies
included the costs of nonprofits.

Frequent changes to the tax code over time reduce the relevance of past
estimates of compliance costs to policymakers who are interested in the
costs of the current system. Since the comprehensive tax reform of 1986
there have been changes to the tax code every year; many of these changes
can be characterized as major.7 Two examples of the numerous changes that
have likely affected compliance burden are

o 	the revenue provisions of the Small Business Job Protection Act of
19968 that tightened the pension nondiscrimination rules that businesses
must follow; and

o 	the creation of Hope and Lifetime credits for post-secondary education
(in the Taxpayer Relief Act of 19979); the fact that taxpayers must chose
between the two credits adds to their need for tax planning.

    The Extent to Which Compliance Costs Could be Reduced Depends Upon the
    Details of a Redesigned Tax System

The estimates presented above do not represent the potential cost savings
to be gained by replacing the current federal tax system. Any replacement
tax system will impose significant compliance costs of its own. Moreover,
given that many state and local government income taxes depend upon the
same compliance activities as the federal income tax does, taxpayers would
still bear the costs of those activities unless those other governments
replaced their own taxes to conform to the new federal system. In
addition, if some of the subsidies, such as the earned income tax credit
and the research tax credit, that are provided by the current federal tax
system are replaced by spending programs under a reformed system, tax
compliance costs may be reduced, but only as a result of their being
shifted to those new programs. Similarly, if a replacement tax system no
longer requires businesses and individuals to compute and document their
incomes, those businesses and individuals will still need to document
their incomes for borrowing and other purposes, and government statistical
agencies will incur expenses to replace the data that they currently
obtain from income tax returns.

7 According to The President's Advisory Panel on Federal Tax Reform, Staff
Presentation on Complexity and Instability presented on July 20, 2005,
there have been 14,400 changes to the tax code since 1986.

8 Pub. L. No. 104-188 (1996).

9 Pub. L. No. 105-34 (1997).

  Efficiency Costs Arising from Tax-Induced Changes in Behavior Are Likely to Be
  Large but Can Only Be Modeled with Considerable Uncertainty

Economic efficiency costs occur when tax rules cause individuals to change
their work, savings, consumption, and investment behavior in ways that
ultimately leave them with a combination of consumption and leisure (now
and in the future) that they value less than the combination they would
have obtained under a tax system that did not distort their behavior.
Estimating efficiency costs is very challenging because the tax system has
such extensive and diverse effects on behavior. We found no comprehensive
estimates of the efficiency costs of the current federal tax system.
However, we did find some studies that estimate the efficiency costs
attributable to selected aspects of the current system. Although none of
these studies, either individually or in the aggregate, provide a basis
for estimating the total efficiency cost of the tax system, they do
indicate that those total costs are likely to be large. The more
comprehensive estimates show costs on the order of 2 to 5 percent of GDP
each year. However, the efficiency cost of the current tax system may not
fall within that range because of uncertainty surrounding taxpayer's
behavioral responses to tax rules, changes in the tax code and the economy
since the mid-1990s, and the studies do not cover all of the sources of
efficiency costs.

    Efficiency Costs Arise When Taxpayers Alter Their Behavior in Response to
    Tax Rules

Many aspects of the federal tax system provide incentives or disincentives
for taxpayers to undertake particular activities. Some of these incentives
and disincentives were intentionally designed into the system; others are
unintended consequences of rules designed to achieve other objectives,
such as equity or increased revenue yield. By changing the relative
attractiveness of highly taxed and lightly taxed activities, taxes alter
decisions such as what to consume and how to invest. Households and firms
generally respond to taxes by choosing more of lower taxed items and less
of higher taxed items than they would have otherwise. When taxpayers alter
their behavior in response to tax rules, they often end up with a
combination of consumption (broadly defined) and leisure that they value
less than the combination they could have achieved if they made decisions
free of any tax influences. Economists refer to this reduction in value as
a "welfare loss" or an efficiency cost; they also generally refer to

behavioral changes in response to taxes as "distortions," even though not
all changes have negative consequences. 10

Figure 3: Efficiency Costs Are One Cost Taxpayers Face in Complying with
the Tax System

Source: GAO.

As an example of the efficiency costs of taxes, suppose an investor is
choosing between two potential investments, one that has an expected
return of $1.10 on every dollar invested and a second that has an expected
return of $1.20. If neither investment is taxed, or if the investments are
taxed equally, the investor will choose the second investment with its
higher economic rate of return. However, if the first investment continues
to be untaxed, while the second is subject to a 10 percent tax, the
decision will be based on the investments' after-tax rates of return. In
this case the after-tax return on the first investment continues to be
$1.10 for every dollar invested, while the after-tax return on the second
investment is now $1.08. An investor would choose the first investment
because it has a higher after-tax return. This tax distortion causes
investors to earn $.10 less on every dollar invested, relative to the
no-tax case, even though no tax is paid to the government. This decline in
income ultimately leads to lower total consumption.

10 A tax on pollution is an example of an efficiency-enhancing tax that
causes a beneficial change in behavior. Pollution may be viewed as a
negative product that consumers are involuntarily forced to consume. A tax
on pollution can provide polluters with an incentive to reduce their
emissions, thereby increasing the well-being of consumers.

Efficiency gains or costs are not the same as increases or decreases in
economic output (normally measured in terms of GDP). For example, a
reduction in taxes on wages could encourage some individuals to increase
the number of hours they work, which in turn would increase economic
output and the amount of consumption those workers could achieve. However,
the welfare gain to these individuals may be considerably less than the
increase in their consumption because they would have to forego some
leisure time that they value in order to achieve the gain in welfare. In
fact, some researchers who have examined both the output and efficiency
effects of replacing the current tax system with various alternatives have
reached similar conclusions concerning output effects while reaching
different conclusions concerning the efficiency effects of similar
alternatives.11

The efficiency gain or loss due to a change in a tax system is defined as
the difference between total welfare that is achieved under the existing
tax system and that which would be achieved under the replacement system
that raised the same amount of revenue. The total efficiency cost of the
current federal tax system would have to be estimated by comparing it to a
tax system that raised the same amount of revenue while generating no
efficiency costs at all-in other words, a tax system that had no effect on
taxpayers' behavior at all. Although such distortion-free tax systems can
be designed in theory, none exist as primary sources of revenue in
practice because they are generally viewed as inequitable.12

11 For example, both Jorgenson and Yun, Investment Volume 3: Lifting the
Burden: Tax Reform, the Cost of Capital, and U.S. Economic Growth,
(Cambridge, Massachusetts, MIT Press, 2001) and Alan Auerbach, "Tax
Reform, Capital Allocation, Efficiency and Growth," in Economic Effects of
Fundamental Tax Reform (Washington, D.C., Brookings Institution Press,
1996) examine the output and efficiency effects of changing to a
consumption-based tax. While they both find significant output effects,
only Jorgenson and Yun consistently find large efficiency gains from
switching to a consumption tax.

12 A head tax, which is a tax that collects the same amount of money from
all taxpayers regardless of how much they earn or consume, is an example
of a tax that does not distort behavior. This type of tax does not meet
either of the two commonly recognized criteria for equitable taxes-first,
that tax liabilities should be related to taxpayers' ability to pay and,
second, that tax liabilities should reflect the benefits that taxpayers
receive from the government.

    Estimates of Efficiency Costs Are Highly Uncertain Because the Tax System
    Has Such Extensive and Diverse Effects on Behavior

Estimating the efficiency costs of the federal tax system is an enormous,
complicated, and uncertain task, given the complexity of existing tax
rules, the breadth and diversity of the U.S. economy and population, and
the limited empirical evidence available regarding how individuals and
businesses change their behavior in response to tax rules. In order to
obtain a precise estimate of the efficiency costs researchers would need
to identify all of the significant incentives and disincentives imbedded
in the federal tax system; they would also need to know the extent to
which large, heterogeneous populations of individuals and businesses have
changed their behaviors in response to these incentives.13

In practice, researchers have not been able to obtain and analyze all of
the detailed data they need to produce efficiency cost estimates that are
free from a large degree of uncertainty. The mathematical models that are
used to analyze the important interrelating and cascading effects of the
entire tax system, or even significant components of the system, are quite
complex, even when researchers limit their examination to highly
simplified representations of the actual taxes. It is impractical to
incorporate all of the significant details of tax rules that result in
efficiency costs. Similarly, data constraints and computational
practicality lead researchers to limit the extent to which their models
reflect the variations in behavioral responses to various tax rules across
large and heterogeneous populations of businesses and individuals.
Finally, researchers attempting to estimate efficiency costs often have
little, or conflicting, empirical evidence upon which to base their
assumptions about various behavioral responses to tax distortions because
the underlying research into those behavioral responses is, itself,
subject to considerable uncertainty. For example, one piece of information
that is critical to the estimation of efficiency costs is the extent to
which individuals' taxable incomes change in response to tax changes.
Researchers have had difficulty estimating this responsiveness because of
the difficulty of controlling for all of the other factors that affect
income growth, such as changes in the economic environment, returns to
investments in education, and the changing age distribution of the

13 Moreover, it is not sufficient simply to know how various tax
incentives directly distort an individual's or business' decisions. The
researchers would also need to know how those distorted decisions, in
turn, caused other decisions to be distorted in a cascading effect. For
example, the distortion of savings decisions can affect the amount of
investment in the economy, which in turn can affect the productivity of
labor and, therefore, the wage rate paid to labor (which generally is
directly related to the productivity of labor). The change in the wage
rate then affects individuals' choices between work and leisure.

population. Moreover, this responsiveness is likely to be different for
different subpopulations of taxpayers.14

    While No Comprehensive Estimates of the Tax System's Efficiency Costs Exist,
    Available Partial Estimates Suggest Those Costs Are Large

None of the studies we reviewed provides a comprehensive estimate of the
efficiency cost of the U.S. federal tax system since the tax reform of
1986. However, a variety of studies do provide some evidence that the
efficiency costs are large. All of these studies are summarized in table
4. The more comprehensive studies we found show costs on the order of 2 to
5 percent of GDP each year (as of the mid-1990s). The other studies that
we reviewed examined more limited aspects of the tax system's distortions.
The efficiency cost estimates for those selected distortions cannot be
summed directly to an overall estimate; however, they are each significant
enough to support the conclusion that the combined cost of all distortions
is large.

Table 4: Estimates of the Efficiency Gains from the Removal of Selected
Distortions in the Current Tax System

Tax(es) included in
the study Behavior altered

                                Feldstein (1999)

Employs his own estimates of the responsiveness of taxable income to
Personal Income Tax Work versus leisure
tax changes and those of other researchers to estimate the efficiency
Payroll Tax
gains from replacing personal income and payroll taxes with
revenue-Tax-preferred
neutral nondistorting taxes. Estimated efficiency costs of $137 billion to
consumption (e.g., fringe
$363 billion in 1994. The author recognizes that the responsiveness of
benefits) versus other
taxable income to changes in taxes is a key parameter, subject to
consumption
ongoing research and that a range of estimates exist. This cost recurs
each year.

Gravelle (2004, 1989) Gravelle and Kotlikoff (1993)

Simulates how taxes affect the allocation of capital investment across
Personal Income Tax Corporate versus
the corporate and noncorporate sectors. Estimated efficiency costs of
noncorporate
nearly $13 billion in 1988, based upon one set of assumptions. This
Corporate Income Tax investment
cost recurs each year. In 2004, the author notes that the subsequent
reduction in taxes on dividends has contributed to a decrease in the
efficiency cost from the corporate/noncorporate distortion.

14 Seth Giertz, Recent Literature on Taxable Income Elasticities,
Congressional Budget Office Technical Paper Series, number 2004-16,
December 2004.

(Continued From Previous Page)

Tax(es) included in
the study Behavior altered

Cai and Gokhale (1997)

Estimates how taxes on capital income distort decisions about when
Selected Aspects of the Current consumption
and what to consume. Estimated efficiency costs of $45 billion in 1995,
Personal Income Tax versus savings
in the authors' base-case. The authors note that the distortion can be
substantial for the consumption of durable goods, such as housing,
Selected Aspects of the Durable consumption
which have relatively low rates of depreciation. This cost recurs each
Corporate Income Tax versus nondurable
year. consumption

                        Lui and Rettenmaier (2002, 2004)

Estimates the excess burden of the Social Security payroll taxes using
Social Security Payroll Work versus leisure
two estimates of labor supply responsiveness. Estimated efficiency Tax
costs range from $49 billion to $82 billion in 2001, depending on the
degree of labor supply responsiveness assumed. The authors note that
their estimates do not include impacts of payroll taxes on savings
decisions or tax-preferred consumption. This cost recurs each year. In
the second paper the authors suggest an alternative derivation of the
excess burden, which produces estimates that are 10 to 50 percent of
the original.

                     Holtz-Eakin and Marples (2001a, 2001b)

Estimates the efficiency cost associated with the distortion of wealth
Estate Tax Wealth accumulation
accumulation decisions by the estate tax relative to a uniform tax on and
allocation to
capital income. Estimated efficiency cost of $38.4 billion in 1999, using
bequests versus lifetime
the authors' preferred set of assumptions. The authors note that their
consumption and
data exclude the "super-rich" who are most affected by the tax and that
other lifetime uses of
the literature on the percentage of bequests that are intentional wealth
presents a large range of estimates. This cost recurs each year.

                            Jorgenson and Yun (2001)

Estimates the efficiency cost of replacing the personal and corporate
Personal and Corporate Income
income tax with a nondistortionary revenue neutral tax. Estimated Taxes at
the Federal, State,
efficiency costs at the local, state, and federal level of 19.5 percent of
and Local level
combined collections. If applied to only federal collections, this would
equal about $200 billion in 1997.

Behaviors altered by the personal income tax, such as savings, work
effort, and housing choices; and behaviors altered by the corporate income
tax, such as debt versus equity finance, organizational form, and dividend
decisions.

                     Source: GAO analysis of papers cited.

The two studies with the broadest scopes among those that we reviewed were
by Jorgenson and Yun and by Feldstein.15 The first set of authors
estimated that the efficiency cost of federal taxes on capital and labor
income in 1997 was equal to about 19.5 percent of the revenues collected
from those taxes.16 Applying this percentage to federal corporate and
personal income tax collections in 1997 would yield efficiency costs of
about $200 billion or, roughly, 2.5 percent of GDP in that year. Feldstein
examined the effects of several distortions caused by the federal personal
income tax and payroll taxes, including those related to decisions about
how much to work and what to consume. He estimated that these distortions
resulted in efficiency costs of between $137 billion and $363 billion in
1994 (depending on his assumptions regarding the size of taxes effects on
various decisions). Those estimates were roughly equivalent to between 2
and 5 percent of GDP in 1994.

The other studies that we found focused on more limited aspects of the tax
distortions caused by the federal tax system. For example, Cai and Gokhale
examined how selected aspects of the federal personal and corporate income
taxes distorted the choices between savings and consumption and found that
these distortions generated $45 billion per year in efficiency costs as of
1995. As another example, Holtz-Eakin and Marples found that the
distorting effect of the estate tax on choices among consumption, leisure,
and wealth accumulation resulted in efficiency costs of over $38 billion
in 1999.

The estimates from the various studies shown in table 4 cannot be combined
to obtain a comprehensive estimate of the current tax system's efficiency
costs for several reasons. First, no combination of the estimates covers
all of the tax system's distortions (e.g., none of the estimates cover the
effects of payroll taxes on savings). Second, an estimate of the costs
arising from all of the tax system's distortions can only be made by

15 Jorgenson and Yun 2001 and Martin Feldstein, "Tax Avoidance and the
Deadweight Loss of the Income Tax," The Review of Economics and
Statistics, 1999.

16 The authors actually estimate the cost of these taxes at the federal,
state, and local level; however, one of the authors told us that the costs
as a percentage of revenue should be approximately the same if the state
and local taxes were excluded from the analysis.

examining the removal of all of the distortions simultaneously.17 Finally,
the various estimates are made for different years and, therefore, reflect
different tax systems, each of which is somewhat different than today's
system. As noted earlier, there have been frequent and significant changes
to the tax system since 1986. Even as recently as 2001 and 2003 there have
been changes that are likely to have affected the efficiency costs of
federal income taxes. For example, the Economic Growth and Tax Relief
Reconciliation Act of 2001 (EGTRRA)18 decreased the marginal tax rates on
individual income and estates and increased the exemption from estate
taxes. These changes could have decreased the efficiency cost for
individuals. Also, Gravelle (2004) notes that the temporary reduction in
taxes on dividends contained in the Jobs and Growth Tax Relief
Reconciliation Act of 200319 contributed to a reduction of the
corporate/noncorporate distortion and its efficiency costs. Nevertheless,
the estimated efficiency costs associated with the selected distortions
are significant enough individually to suggest that the total efficiency
cost of the tax system is large.

We also reviewed a number of other studies that estimated the difference
in efficiency costs between existing federal income taxes and potential
substitutes for those taxes.20 These estimates do not represent the
absolute costs of the existing taxes because the substitute taxes all
generate efficiency costs of their own; however, they would (to the extent
that they are accurate) represent lower bounds for those costs. The
findings of these studies varied considerably, depending on the nature of
the taxes that were presumed to replace the federal income taxes. For
example, Lim Rodgers

17 For example, Douglas Holtz-Eakin and Donald Marples, "Distortion Costs
of Taxing Wealth Accumulation: Income Versus Estate Tax," National Bureau
of Economic Research Working Paper 8261, 2001, estimated the efficiency
costs of the estate tax in the presence of all of the distortions
associated with the federal income taxes. Their estimate could have been
significantly different if they had assumed that a nondistortionary tax
had been in effect at that time, instead of the income taxes. For this
reason, it is inappropriate to assume that eliminating the distortions of
the estate tax would have yielded the same $38.4 billion in 1999 on top of
the gains that Jorgenson and Yun 2001 estimate could have been realized by
removing all income tax distortions.

18 Pub. L. No. 107-16 (2001).

19 Pub. L. No. 180-27 (2003).

20 In addition, a number of studies have estimated the effects of
fundamental tax reform on economic output. However, as we noted
previously, output effects are not the same as efficiency effects, so
those output estimates are not relevant to the questions addressed in this
report.

and Jorgenson and Wilcoxen found that there would be little, if any,
efficiency gain from replacing the existing income taxes with a
consumption-based flat tax.21 In contrast, Jorgenson and Yun estimated
that the efficiency gains of replacing income taxes with a pure flat-rate
income and/or sales tax would yield gains of about $210 billion per year
(measured in 1997 dollars).22

As noted earlier, considerable uncertainty surrounds all of the estimates
we have cited. The estimation of efficiency costs involves complicated
modeling based on numerous assumptions about the behavioral responses of
individuals and businesses to changes in taxes and other factors. Results
are often quite sensitive to the assumed magnitude of key responses and
those assumptions are often based on empirical research that continues to
evolve over time or, in other cases, has yet to be undertaken. For
example, the consensus of recent research is that individuals are less
responsive to changes in taxes than Feldstein assumed them to be when he
made his estimates. As another example, Holtz-Eakin and Marples noted that
there was significant disagreement in the empirical research over one of
the factors that was key to their estimate-the extent to which actual
bequests differed from intended bequests.

    The Extent to Which Efficiency Costs Can be Reduced by Tax System Redesign
    Is Uncertain

As some of the results presented in the preceding section demonstrate, the
extent to which efficiency gains could be realized by switching to an
alternative tax system depends critically on the detailed characteristics
of the alternative. All of the alternative tax system proposals that have
received serious consideration in recent decades would have imposed
significant efficiency costs. Moreover, in assessing the potential
efficiency gains from any tax reform proposal it is also important to
consider compensating changes that may be made on the spending side of the
federal budget. For example, if any tax expenditures in the current
federal income taxes are replaced by grants, spending programs,
regulations, or

21 The characteristics of this replacement tax were specified by the Joint
Committee on Taxation as part of an exercise for a symposium that the
committee organized to test the feasibility of incorporating macroeconomic
effects in revenue estimates. See JCS-21-97, "Joint Committee on Taxation
Tax Modeling Project and 1997 Symposium Papers," November 20, 1997, for a
complete description.

22 These replacement taxes in Jorgenson and Yun 2001 are "pure" in the
sense that they have no exemptions (except for investment goods in the
case of the sales tax), deductions, credits, or special rates.

other forms of nontax subsidies, those subsidies can result in efficiency
costs similar in magnitude to those associated with the tax expenditures
they replaced.

  Agency Comments and Our Evaluation

We provided a draft of this report in August 2005 to the Commissioner of
Internal Revenue. We received technical comments via e-mail from the IRS
Office of Research. Where appropriate, we made changes in our report in
response to these comments.

As agreed, unless you announce the contents of this report earlier, we
plan
no further distribution until 30 days from the date of this report. At
that
time, we will make copies available to others on request. In addition, the
report will be available at no charge on the GAO Web site at
http://www.gao.gov.

If you or your staff have any questions on matters discussed in this
report
or would like additional information, please contact me at (202) 512-5594
or [email protected]. Contact points for our Offices of Congressional
Relations and Public Affairs may be found on the last page of this report.
Key contributors to this report were James Wozny and Donald Marples.

James R. White
Director, Strategic Issues

Bibliography

Auerbach, Alan J. "Tax Reform, Capital Allocation, Efficiency, and
Growth." in Economic Effects of Fundamental Tax Reform. Edited by Henry J.
Aaron and William G. Gale. Washington, D.C.: Brookings Institution Press,
1996.

Blumenthal, Marsha. "Burden Reduction Research and Analysis." Requisition
Number 053BPPX0039. Washington, D.C.: Internal Revenue Service, 1996.

Blumenthal, Marsha and Joel Slemrod. "The Compliance Cost of the U.S.
Individual Income Tax System: A Second Look After Tax Reform." National
Tax Journal 45, no. 2 (1992).

Cai, Jinyong and Jagadeesh Gokhale. "The Welfare Loss From a Capital
Income Tax." Federal Reserve Bank of Cleveland Economic Review 33, no.

1. (1997).

Feldstein, Martin. "Tax Avoidance and the Deadweight Loss of the Income
Tax." The Review of Economics and Statistics. 81, no. 4. (1999).

GAO. Tax Administration: IRS Is Working to Improve its Estimates of
Compliance Burden. GAO/GGD-00-11. Washington, D.C.: May 2000.

GAO. Student Aid and Postsecondary Tax Preferences: Limited Research
Exists on the Effectiveness of Tools to Assist Students and Families
through Title IV Student Aid and Tax Preferences. GAO-05-684. Washington,
D.C.: July 2005.

Giertz, Seth. Recent Literature on Taxable Income Elasticities. Technical
Paper Series no. 2004-16. Washington, D.C.: Congressional Budget Office,
2004.

Gravelle, Jane. "The Corporate Tax: Where Has It Been and Where Is It
Going?" National Tax Journal 57, no. 4 (2004).

-----"Differential Taxation of Capital Income: Another Look at the 1986
Tax Reform Act." National Tax Journal 42, no. 4 (1989).

-----and Laurence Kotlikoff. "Corporate Tax Incidence and Inefficiency
When Corporate and Noncorporate Goods Are Close Substitutes." Economic
Inquiry 31 (1993).

Bibliography

Guyton, John, John O'Hare, Michael Stavrianos, and Eric Toder. "Estimating
the Compliance Cost of the U.S. Individual Income Tax." National Tax
Journal 56, no. 3 (2003).

Holtz-Eakin, Douglas, and Donald Marples. "Distortion Costs of Taxing
Wealth Accumulation: Income Versus Estate Taxes." National Bureau of
Economic Research Working Paper 8261, National Bureau of Economic
Research, 2001a.

Holtz-Eakin, Douglas, and Donald Marples. "Estate Taxes, Labor Supply, And
Economic Efficiency." Center for Policy Research Special Report.
Washington, D.C.: American Council for Capital Formation, 2001b.

Joint Committee on Taxation. The Joint Committee on Taxation Tax Modeling
Project and 1997 Tax Symposium Papers. JCS-21-97. Washington, D.C.: Joint
Committee on Taxation, 1997.

Jorgenson, Dale, and Kun-Young Yun. Investment Volume 3: Lifting the
Burden: Tax Reform, the Cost of Capital, and U.S. Economic Growth.
Cambridge, Massachusetts: The MIT Press, 2001.

Jorgenson, Dale, and Peter Wilcoxen. "The Effects of Fundamental Tax
Reform and the Feasibility of Dynamic Revenue Estimation." In The Joint
Committee on Taxation Tax Modeling Project and 1997 Tax Symposium Papers.
JCS-21-97. Washington, D.C.: Joint Committee on Taxation, 1997.

Liu, Liqun, and Andrew Rettenmaier. "The Excess Burden of the Social
Security Payroll Tax." Public Finance Review 32, no. 6 (2004).

Liu, Liqun, and Andrew Rettenmaier. "The Economic Cost of the Social
Security Payroll Tax." National Center for Policy Analysis Policy Report
No. 252. Dallas, Texas: National Center for Policy Analysis, 2002.

Moody, J. Scott. "The Cost of Complying with the Federal Income Tax."
Special Report No. 112. Washington, D.C.: Tax Foundation, 2002.

Rogers, Diane Lim. "Assessing the Effects of Fundamental Tax Reform With
The Fullerton-Rogers General Equilibrium Model." In The Joint Committee on
Taxation Tax Modeling Project and 1997 Tax Symposium Papers. JCS-21-97.
Washington, D.C.: Joint Committee on Taxation, 1997.

Bibliography

Slemrod, Joel. Written Testimony Submitted to the Committee on Ways and
Means. Subcommittee on Oversight. Hearing on Tax Simplification. June 15,
2004.

-----"Measuring Taxpayer Burden And Attitudes For Large Corporations: 1996
and 1992 Survey Results." Report to the IRS Coordinated Examination
Program. Washington, D.C.: Internal Revenue Service, 1997.

-----"Which is the Simplest Tax System of Them All." In Economic Effects
of Fundamental Tax Reform. Edited by Henry J. Aaron and William G. Gale.
Washington, D.C.: Brookings Institution Press, 1996.

-----and Jon Bakija. Taxing Ourselves: A Citizen's Guide to the Great
Debate over Tax Reform. Cambridge, Massachusetts: The MIT Press, 2004.

-----and Marsha Blumenthal. The Income Tax Compliance Cost of Big
Business." Washington, D.C.: Tax Foundation, 1993.

-----and Varsha Venkatesh. "The Income Tax Compliance Cost of Large and
Mid-Size Business." Report to the IRS LMSB Division. Washington, D.C.:
Internal Revenue Service, 2002.

Steuerle, C. Eugene. Contemporary U.S. Tax Policy. Washington, D.C.: The
Urban Institute Press, 2004.

GAO's Mission	The Government Accountability Office, the audit, evaluation
and investigative arm of Congress, exists to support Congress in meeting
its constitutional responsibilities and to help improve the performance
and accountability of the federal government for the American people. GAO
examines the use of public funds; evaluates federal programs and policies;
and provides analyses, recommendations, and other assistance to help
Congress make informed oversight, policy, and funding decisions. GAO's
commitment to good government is reflected in its core values of
accountability, integrity, and reliability.

Obtaining Copies of The fastest and easiest way to obtain copies of GAO
documents at no cost

is through GAO's Web site (www.gao.gov). Each weekday, GAO postsGAO
Reports and newly released reports, testimony, and correspondence on its
Web site. To Testimony have GAO e-mail you a list of newly posted products
every afternoon, go to

www.gao.gov and select "Subscribe to Updates."

Order by Mail or Phone	The first copy of each printed report is free.
Additional copies are $2 each. A check or money order should be made out
to the Superintendent of Documents. GAO also accepts VISA and Mastercard.
Orders for 100 or more copies mailed to a single address are discounted 25
percent. Orders should be sent to:

U.S. Government Accountability Office 441 G Street NW, Room LM Washington,
D.C. 20548

To order by Phone: 	Voice: (202) 512-6000 TDD: (202) 512-2537 Fax: (202)
512-6061

  To Report Fraud, Contact:
  Waste, and Abuse in Web site: www.gao.gov/fraudnet/fraudnet.htm

E-mail: [email protected] Programs Automated answering system: (800)
424-5454 or (202) 512-7470

Congressional 	Gloria Jarmon, Managing Director, [email protected] (202)
512-4400 U.S. Government Accountability Office, 441 G Street NW, Room 7125

Relations Washington, D.C. 20548

Public Affairs	Paul Anderson, Managing Director, [email protected] (202)
512-4800 U.S. Government Accountability Office, 441 G Street NW, Room 7149
Washington, D.C. 20548
*** End of document. ***