Brookings' Hamilton Project Issues Paper on Return-Free Proposal

The Project is named after Alexander Hamilton, the nation's first treasury secretary, who laid the foundation for the modern American economy. Consistent with the guiding principles of the Project, Hamilton stood for sound fiscal policy, believed that broad- based opportunity for advancement would drive American economic growth, and recognized that "prudent aids and encouragements on the part of government" are necessary to enhance and guide market forces.

The costs of complying with federal income tax requirements are large, and are particularly unwelcome to many middle-class taxpayers whose tax situations are quite simple. Indeed, the data show that the costs of complying are regressive -- lower, as a share of income, for wealthier taxpayers.

This paper proposes a program known as the "Simple Return," which would make it much easier for the millions of taxpayers with a relatively simple tax status to file their taxes. The Simple Return might apply to as many as 40 percent of Americans, for whom it could save up to 225 million hours of time and more than $2 billion a year in tax preparation fees. Converting the time savings into a monetary value by multiplying the hours saved by the wage rates of typical taxpayers, the Simple Return system would be the equivalent of reducing the tax burden for this group by about $44 billion over ten years. A Government Accountability Office report estimated in 1996 that a plan similar to the one proposed here could save the IRS close to $36 million per year by reducing the number of errors in tax filings and the subsequent need for investigations.

Around two-thirds of taxpayers take only the standard deduction and do not itemize. Frequently, all of their income is solely from wages from one employer and interest income from one bank. For almost all of these people, the IRS already receives information about each of their sources of income directly from their employers and banks. The IRS then asks these same people to spend time gathering documents and filling out tax forms, or to spend money paying tax preparers to do it. In essence, these taxpayers are just copying into a tax return information that the IRS already receives independently. The Simple Return would have the IRS take the information about income directly from the employers and banks and, if the person's tax status were simple enough, send that taxpayer a return prefilled with the information. The program would be voluntary. Anyone who preferred to fill out his own tax form, or to pay a tax preparer to do it, would just throw the Simple Return away and file his taxes the way he does now. For the millions of taxpayers who could use the Simple Return, however, filing a tax return would entail nothing more than checking the numbers, signing the return, and then either sending a check or getting a refund.

 Contents

The burden of an income tax can be divided into two parts, one visible and the other hidden. The first burden is obvious -- it is the actual tax payment. The second burden is less obvious. Part of this burden arises when taxes alter the economic decisions that people make. Another major part of that unseen burden, however, comes from the cost of complying with the system: the hours spent preparing forms, gathering documents, and reading instructions; or the money a tax filer pays someone to do the tax preparation for him.

This paper proposes a program called the "Simple Return," which would make it much easier for the millions of taxpayers with a relatively simple tax status to file their taxes. As many as 40 percent of all U.S. taxpayers would be eligible to use the Simple Return, collectively saving up to 225 million hours of time and more than $2 billion a year of tax preparation fees. Converting the time savings into a monetary value by multiplying the hours saved by the wage rates typically paid to middle- and low-income workers, the Simple Return could reduce tax compliance costs by about $44 billion over ten years. This benefit would come at a very small cost to the government. Indeed, a Government Accountability Office report (GAO 1996) estimated that a plan similar to the one proposed here could save the Internal Revenue Service (IRS) almost $36 million a year by reducing the number of errors in tax filings and the subsequent need for investigations.

It is important to clarify what the proposal would not address. There is no doubt that some taxpayers face high compliance costs because of the way that their economic situations interact with complex tax code provisions. These taxpayers might have income from a variety of sources -- including an employer, self-employment, investments, and capital gains. They also might have large tax deductions, such as mortgage interest and other itemized deductions. They have the most complex tax positions in the country and the highest costs of compliance. They also have the highest incomes in the country. These taxpayers file the 1040 tax form, along with pages of additional information and schedules. The Simple Return would not affect or apply to them.

Most Americans, however, have a relatively straightforward tax status. About two-thirds of taxpayers take only the standard deduction and do not itemize. Frequently, all of their income comes from wages from one employer and interest income from one bank. For many people, taxes are determined by their earned income and their family status. The IRS already receives information about the sources of income for almost all of these people directly from employers, banks, and so on. The IRS then asks these people to spend time gathering documents and filling out tax forms (or paying tax preparers to do so) in order to provide the IRS with information that it already already receives from other sources. Indeed, if these taxpayers do not fill out their tax forms correctly, the IRS eventually will contact them and tell them exactly how much they owe (or how much they are owed as a refund).

With the Simple Return, the IRS would take the information about income directly from the employers and banks and, if the person's tax status were simple enough, send that taxpayer a return that the IRS had prefilled with the necessary tax information. The Simple Return program would be strictly voluntary. Any taxpayer who preferred to fill out his own tax form or who preferred to pay a tax preparer to fill out the form could do just that. Of course, the IRS would not be able to gather complete information on everyone, and some people's family or tax situation might change during the year, making the taxpayer ineligible for this program. But for the millions of taxpayers who could use the Simple Return -- potentially as many as 40 percent of all U.S. taxpayers -- filing a tax return would entail nothing more than checking the numbers, signing the return, and then either sending a check or, more typically, receiving a refund.

The first response of many who contemplate the issue of reducing compliance costs for middle- and low-income Americans is that the benefits cannot be very large. After all, many people with a relatively simple tax status do not need to file a standard 1040 form. Instead, they file the 1040EZ, a one-page tax form for filers with less than $100,000 in income who do not need to adjust that income. Others can file the 1040A tax form, which allows a limited number of adjustments, without any need to file a sheaf of accompanying tax schedules. Yet all the forms entail gathering information, computing, and reading somewhat technical documents that include tables, worksheets, and instructions. The instruction booklet for the 1040EZ, for example, is thirty-six pages long! Doing taxes takes time and energy, and is clearly something the typical American would prefer to avoid. Anyone, especially those without a college degree or with rusty math skills, can find the forms downright maddening. This is one reason more than 30 percent of 1040EZ filers and 56 percent of 1040A filers used a paid tax preparer in 2001.¹

For the typical U.S. taxpayer, the costs of tax compliance do not arise from a massive tax code interacting with a complex economic situation. Instead, the taxpayer's costs of compliance arise because the government is not using the information it already receives independently to calculate the taxpayer's taxes. This method of having the government tell taxpayers what it knows about their incomes and sending out a prepared tax form -- known as a Tax Agency Reconciliation (TAR) system, because the tax agency does the computation -- has been tested at the state and international levels. For 2004, the state of California carried out a pilot TAR program with the state income taxes of more than fifty thousand taxpayers; this program was very similar to the Simple Return proposed here. The system was well received. It provides a useful basis for considering the benefits and costs of a national program. TAR systems also have been used with success in several European countries (see "The International Experience" below).

If designed judiciously, the program would have little cost and might even save the IRS money by reducing the error rate of filers. Whatever costs might apply would be modest compared with the reductions in the costs of tax compliance. In effect, the Simple Return program proposes that the government take advantage of the extraordinary gains made in information technology in the past two decades to lighten the burden of tax compliance on American families. The benefit will accrue mainly to those with low and middle levels of income because they are the people who file relatively simple tax returns.

II. The Simple Return Program

The Simple Return program would begin early in the calendar year with taxpayers receiving a pre-filled return and a letter explaining the program (see exhibit 1, a copy of the letter that went out with the California ReadyReturn pilot program). As in the California pilot program, anyone receiving a Simple Return would be free to throw it away if he was not interested in participating. Alternatively, the taxpayer could use it as a starting point for doing his own tax return. If he chose to use the enclosed Simple Return, he would confirm that the statements about his income and family status were correct, and then either file the form electronically, at the IRS website, or mail it to the IRS. If the taxpayer were due a refund, the IRS could quickly process it.

The Simple Return system would phase in with three distinct waves representing different sets of taxpayers; a possible fourth wave is also discussed below.

The First Wave

The first wave would begin with the easiest cases. It would consist of automatic mailings of the Simple Return to single filers with the following situation: They have no dependent children, they are not dependents themselves, they had only wage income on their last return, they have no other credits, and they did not itemize in previous tax filings. The system could start with federal workers where the matching of incomes from the employer is particularly easy. About nine million taxpayers might be eligible for the first wave. As table 1 shows, most of these taxpayers already file the 1040EZ or 1040A forms.

The IRS would prepare a Simple Return for each of these people. The IRS would then mail that individualized Simple Return and a note explaining that the recipient is eligible to participate in the Simple Return program to the taxpayer, using the address from the taxpayer's current W-2 form.² To participate, the taxpayer need only verify the information on the form, including income and family status. To make it easier, the taxpayer still could file the Simple Return as long as he had earned less than a threshold amount of interest income -- say, $100. If there were a minor problem with the return such as the wrong address, the taxpayer could make the changes online or on an enclosed form and file the adjusted return. If something major had changed, or if the filer had other income such as capital gains that made the Simple Return invalid, he would just do his taxes the conventional way.

The first wave also would include a second group of taxpayers -- married filers, people with child credits, and people who are dependents, who meet the restrictions on income and filing status noted above.³ The IRS could put the wages of both spouses on the same Simple Return or compute the child credit. Of course, the Simple Return would be valid only if the taxpayer's personal situation had not changed in the intervening period. If a child had been born to or adopted by the filer during the year, for example, the filer would need to file his taxes the conventional way to receive the full child credit.

With this second group included, the number of taxpayers in the first wave who would be eligible for the Simple Return could be as high as 17.5 million. Most of these taxpayers currently fill out the 1040EZ or 1040A forms, as shown in table 1.

Filers in this group would not receive a Simple Return automatically, but could request participation in one of two ways: by including required information on a W-4 form, or by checking a box on the previous year's conventional tax return.

 

Exhibit 1. Sample Letter for California ReadyReturn Pilot

 

Program

 

 

 

The taxpayer could request participation when completing the standard W-4 form, which is filled out by every employee when starting a job or reporting a change in family status that might affect the amount of taxes withheld. Anyone filling out the W-4 form would be given the opportunity to include additional information about himself and his family status (such as child credits, spouse's Social Security number, and so on) and to request a Simple Return. Because many people only file a W-4 immediately after being hired and other people might not want their employer to have detailed personal information about them or their family status, employees could also fill out a postcard version of the information and mail it directly to the IRS, without their employer seeing it.

"Simple Return" Box on Conventional Tax Return Another way taxpayers could request participation would be for the IRS to add a few lines to conventional tax returns informing taxpayers of the program. Taxpayers could check a box if they wanted the IRS to use their current year information (sources of income, whether they have children, marriage status, and so on) to predict whether they would be eligible for a Simple Return the following year. Taxpayers also could check a box indicating that they did not wish to be considered for the Simple Return. Thus, any tax filer would be free either to request consideration for Simple Return status or to opt out of the Simple Return program. At every stage, the IRS would make clear that the Simple Return gives taxpayers more options without restricting existing options in any way. Checking the Simple Return box would not obligate the filer to do anything. At any time, a taxpayer who is eligible for a Simple Return could choose to file his taxes the conventional way.

The Second Wave

In the second wave of the program, the Simple Return system would expand to include people with income from what are called "with-holdable sources." For our purposes, this means the kinds of income the IRS would receive information on from 1099s and other sources besides W-2s. This would include interest, dividends (though not capital gains income), pensions, Social Security benefits, unemployment insurance, and individual retirement accounts. As shown in table 1, the estimated number of taxpayers potentially eligible for a Simple Return under these conditions was 38.6 million in 2003. In principle, the second wave also could include capital gains distributions from mutual funds, although it could not include net capital gains from the sale of assets. In practice, however, this additional step of including capital gains distributions may not be worth the effort: It would only raise the number of taxpayers eligible for the Simple Return by about four hundred thousand (to approximately thirty-nine million people), as shown in table 1.

 Table 1. Estimating the Number of Filers Eligible for a Simple

The Third Wave

The third wave of the Simple Return program could occur simultaneously with the second wave if the ongoing IRS modernization goes smoothly. It would use the taxpayer's tax return information from the previous year to prepare a Simple Return that would be sent out to anyone who did not itemize deductions and was otherwise eligible for a Simple Return based on the previous year's information. This would include a Simple Return for recipients of the earned income tax credit (EITC).⁴ Adding the EITC to the mix of tax provisions that was covered under the Simple Return would increase the number of people who were potentially eligible to more than fifty-two million.

It is important to note that, in most cases, the taxes calculated on the Simple Return would be the same as if the taxpayer did his own tax return or paid a tax preparer to do so. For people with as simple a situation as those cited here, there simply are not many preferences or deductions available that a person would be able to take advantage of if he filed his taxes the conventional way. However, at higher incomes there will be some borderline cases. Perhaps such filers will be close to the cutoff of whether to justify itemizing their deductions, rather than using the standard deduction. Some of these people would receive a Simple Return, but might have a lower tax bill by filling out the conventional tax form and itemizing, perhaps with the help of a paid tax preparer, or by changing their behavior to take advantage of particular tax incentives. The Simple Return instructions would need to state clearly that the calculations are done assuming the simplest situation, using the information the IRS has available. The Simple Return is neither a tax planning service nor an attempt by the government to become involved in a line of business. Rather, it is a way for government to serve citizens better by helping those taxpayers who have simple tax situations avoid the need to fill out a return. Many of these people would likely still go to a paid tax preparer to see how much additional money they could save. They would be free to do so under the Simple Return system.

III. Real-World Experience with Return-Free Filing

The California Pilot Program

In 2004, the Franchise Tax Board of California conducted a pilot program for its state income tax along the lines of the Simple Return system. Bankman (2005) provides an excellent overview of the program. The California ReadyReturn went out to over fifty thousand single taxpayers who did not have any dependents and who had only wage income; the ReadyReturn was accompanied in each case by a cover letter (exhibit 1). The returns used wage and withholding information that employers had directly reported to the state. Overall, more than eleven thousand people used the ReadyReturn -- more than one-fifth of those who received the ReadyReturn and letter. However, interest in using ReadyReturn was even higher: Twenty-two percent of the people who declined participation cited as their reason that they had already filed their tax return by the time that they received the mailing. Participation among those who had not already filed was approximately 27 percent. There was no advertising of the program ahead of time -- people simply received the form in the mail.

ReadyReturn users were pleased with the program. More than 90 percent of ReadyReturn users said that they saved time using the system and that it was more convenient than the conventional filing (Franchise Tax Board 2006). California's Franchise Tax Board also created a control group of people who would have been eligible for the ReadyReturn, but who were not invited to participate. The median filing time for ReadyReturn users who filed electronically was nearly 80 percent lower than the median filing time for the control group. The ReadyReturn filers also said that they saved money using the program. The median ReadyReturn filer paid nothing to complete his state return; the median control group filer paid $30. More than 98 percent of ReadyReturn users said they were "satisfied" or "very satisfied" with the program, and more than 97 percent said that they would use it again next year. Once people experience the convenience of receiving a completed tax return, they do not want to give it up. Of the few complaints, almost all were of a technical nature such as, "What is a PIN?" rather than about the program as a whole (Bankman 2005).

Of those who did not choose to participate, about 25 percent said they prefer to use a tax preparer and 10 percent said they were uncomfortable receiving a prefilled return. This evidence suggests that about one-third of those who are eligible for a Simple Return might just discard the Simple Return form and the letter, at least for the first few years after program introduction.

California's Franchise Tax Board also observed that the ReadyReturn filers were significantly less likely to have errors than the control group's returns. Three percent of the control group received an error notice from the Franchise Tax Board, compared with 0.3 percent of filers from the ReadyReturn group. The Franchise Tax Board also noted that the rate of electronic filing rose dramatically among the ReadyReturn filers, which further reduced the Board's processing costs. Indeed, the Franchise Tax Board was so confident of the cost savings that it asked the California legislature to reduce its long-term budget to reflect the savings from the program (Franchise Tax Board 2005). There is every reason to believe that these same benefits of taxpayer satisfaction and cost savings would hold for a Simple Return applied at the federal level.

The International Experience

TAR systems such as the Simple Return have been used in Europe; in some cases, they have been used more extensively than the Simple Return would be even at its maximal use in the United States. The U.S. Department of the Treasury (U.S. Treasury 2003) observed that, in 1999, about 87 percent of tax filers in Denmark and 74 percent of filers in Sweden had their returns prepared through a TAR system. Finland and Norway also have experimented with TAR systems. One reason that Denmark and Sweden have been able to generate such high participation rates is that their underlying tax systems are simpler than the U.S. tax system, so more people qualify. Not surprisingly, the compliance costs in these countries are substantially lower than they are in the United States. In Sweden, for example, compliance costs have been estimated at about 1 percent of revenue (U.S. Treasury 2003). Slemrod (2004) estimates the compliance costs for the individual income tax in the United States at more than 10 percent of revenue.

IV. Questions and Concerns

What Share of Those Eligible for the Program Would Participate?

It is not clear how many people would accept the Simple Return offer in the near term, but the number of participants is likely to be substantial. A survey conducted by the U.S. Treasury (2003) suggested that 39 percent of people were interested in using such a system, and another 25 percent might be interested.

In the California ReadyReturn system, with no advertising and with only one mailing, about 27 percent of the people who received the return and had not already filed their taxes chose to participate. This is substantially higher than the reports in the U.S. Treasury (2003) survey, where only 17 percent of people said they would try a return-free system immediately. There is every reason to expect a substantial increase in usage over time. More than 97 percent of the people who used the California ReadyReturn said they planned to participate again, and the U.S. Treasury survey evidence suggests that initial success would likely indicate rapid future expansion. Twenty-eight percent of people said they would be comfortable trying such a system after its first year of existence, and another 27 percent said they would be comfortable trying it after its first few years. In other words, the survey evidence suggests interest could double or even triple within a few years.

With proper explanations of the program, with advance media exposure, and with a system enabling people to opt into the system by filling out a postcard, the takeup rate among eligible participants might be as high as 50 percent in the short run and even higher in the long run. Regardless, the Simple Return does not need to achieve anywhere near complete participation to be a success. If it could reduce compliance costs for millions of people, it would obviously be a boon to U.S. taxpayers.

How Much Can Costs of Compliance Be Reduced for U.S. Taxpayers?

To estimate compliance costs, it is helpful to separate those who do not itemize into two groups: those who pay someone to prepare their tax filing and those who prepare their own tax filing.

There has been little systematic evidence on how much low- and middle-income taxpayers pay tax preparers to complete the taxpayers' 1040EZ or 1040A forms. The average price for paying a tax preparer to fill out a tax form that included the EITC, the provision aimed at the working poor, was around $200, although this typically included the cost of a loan made in advance of an anticipated tax refund, and a charge for e-filing and state income tax preparation (Berube et al. 2002). In the instructions for tax forms for calendar year 2005, however, the IRS gave an estimated cost for paying a preparer to fill out the given form. Interestingly, the IRS also estimates that paying a preparer still requires significant time from the tax filer but, to be conservative, the estimates in this paper only include the cost of paying the preparer. The IRS estimates the average fee for a 1040EZ form to be $81, for a 1040A form to be $122 and, a bit incongruously, for the full 1040 without itemizing, without self-employment income, and without capital gains income to be $121. For 1040 itemizers, the fee on an individual basis can be predicted using the most recent IRS Public Use Data File (U.S. Treasury 2001). That file reports all the information on the tax return of a nationally representative sample of more than one hundred forty thousand randomly sampled taxpayers. Because tax preparation fees in the previous year are a reported expense, the actual fees serve as predictors of the fee, depending on the situation of the itemizer.⁵

 Table 2.  Compliance Times Estimated

By design, the Simple Return will tend to be used by those with low and middle incomes. Because they tend to have fewer years of schooling, those taxpayers are estimated to take longer than average to fill out tax forms. In the performance scores of people in the National Longitudinal Survey of Youth, the response time performance for numerical calculations and for coding speed -- skills used to fill out a tax return -- for people at the median income level were 30-35 percent better than for people at the 25th percentile of income.⁶ The median income for people filing the 1040 in 2001 was $43,850, versus $19,120 for the 1040A and $11,290 for the 1040EZ, based on the IRS Public Use Data File (U.S. Treasury 2001). Since these income differences would suggest that filing times for people with lower incomes might be 20-40 percent longer than estimated, the estimates in this paper increase the IRS-estimated time for these 1040A and 1040EZ tax forms by 20 percent when computing the compliance times.

The conventional economic approach for turning these times into an equivalent monetary cost of compliance is to place a reasonable monetary value on the person's time and multiply that by the hours saved. To calculate the value of time for each person, the taxpayer's yearly wages are divided by total hours worked, assuming the filer worked full time (two thousand hours a year); for anyone with an imputed value of time below the minimum wage, the estimates use $5.15 an hour.⁷ It is common practice for policy analysts who work with tax data to truncate the extreme ends of the income distribution. This practice effectively avoids the extreme and often unrepresentative situations of people who earned at either end of the spectrum (very high or very low). Few people who earn at the high end are likely to be eligible for a Simple Return, anyway. In that spirit, the sample is restricted to people with adjusted gross income above $5,000 and below $125,000 a year -- around the 10th and 95th percentiles, respectively, of filers in the sample of 2001 taxpayers.

Table 3 summarizes these calculations across income groups. Each column presents the mean adjusted gross income and tax liability for people in that group, followed by the average compliance costs in either preparer fees (for paying filers) or monetary value of time (for self-filers). In the last row, the table divides the compliance cost for each person by that person's adjusted gross income and gives the median share in the group.

One insight from these calculations is that the costs of compliance amount to 5-10 percent of the entire tax revenue paid by many middle- and low-income taxpayers. Many taxpayers at the bottom and lower middle of the income distribution owe little in federal income taxes (they typically owe much more in the payroll taxes that are collected to finance Social Security and Medicare).

                                  $5,000-     $25,000-     $45,000-

In the bottom row, the evidence also makes clear that across the wider span of the income distribution, compliance costs are generally regressive. They are highest at the bottom and generally fall with income. They are lowest for the top income bracket. This is why something that reduces compliance costs is likely to be progressive. If that something, such as the Simple Return, is geared toward the middle- and low-income filers, the progressivity of the reform will be even greater.

Based on the evidence from California's ReadyReturn pilot project, the Simple Return may reduce filing and compliance times by about 80 percent for qualifying taxpayers. This would save 3.7 hours for 1040EZ filers, 11.2 hours for 1040A filers, and 13 hours for (nonitemizing) 1040 filers who do their taxes themselves. The California evidence showed that ReadyReturn filers typically no longer needed a paid tax preparer, thus presumably saving the entire fee. The top rows of table 4 illustrate the average savings for paying filers and self-filers by form (as well as the share of people filing that form who use a paid preparer).

The bottom rows of table 4 then sum these gains across all the people in the economy (valuing the self-filers' time at their wage rate). If everyone who was eligible to use the Simple Return actually did so, the total savings in the first wave would amount to $2 billion a year. About half of that would be reduced fees paid to preparers; the other half would be the value of saved time. After Waves 2 and 3, the savings could be as high as $4.4 billion a year. Over a ten-year budget window, the value of time saved by the Simple Return would reach an amount as large as a $44 billion tax cut for middle- and lower-income people. Even if participation rates were only 50-75 percent of those eligible, savings from the Simple Return would be substantial.

Can the IRS Speed Up Its Processing Time to Issue Refunds More Quickly?

Currently, most businesses must submit information returns to the IRS by the end of February. These businesses include banks and other financial institutions (who submit 1099 information on financial income) and employers (who submit W-2 wage information). The W-2 information goes first to the Social Security Administration, which processes it and then turns over a master file to the IRS later in the year. The delays in processing and transfer of the files in the current system would need to be improved in order to avoid delays in refunds. (About 80 percent of taxpayers receive refunds, with the probability being about equal for itemizers and nonitemizers. See U.S. Treasury 2001).

 Table 4.  Maximum Compliance Cost Savings from the Simple Return

Furthermore, the state unemployment information on employees is filed quarterly in most states, so wage information, at least, would potentially be available early if the federal government were to collaborate with individual states. This information was what the state of California used for its ReadyReturn pilot, and California was able to generate the returns well before the filing deadline.

If needed, the IRS could accelerate the filing deadline for large employers who already file their information electronically and who already report their information to the states. Currently, these employers file the information with the Social Security Administration by the end of February. If they submitted the information by the end of January, it would allow the IRS to incorporate interest and dividend income into the Simple Return and to expand significantly the share of Americans who would be eligible. If the program did move up the filing date, it would certainly exempt any organization that does not already file electronically or that is below a threshold size, and would probably give a small tax credit to offset the minor inconvenience to the employers.

Even with these changes by the IRS, some small employers will turn in their information returns late, some can only do paper-based returns, and some will have other such problems. Obviously, their employees would not be eligible for a Simple Return. This poses no conceptual problem with the Simple Return, though, and does not need to delay its implementation. If it proves difficult to match information on some category of employer or some type of income in the early years of the program, the affected groups simply would not participate. Indeed, if people appreciate the Simple Return, it is entirely possible that they will pressure their employers and financial institutions to send the information in to the government earlier to make them eligible. Given the problems that the IRS has had trying to incorporate small employers and the marginal economy into their processing, these pressures would directly benefit the government by expanding coverage.

What Are the Costs to the Government of Implementing a Simple Return Program?

Previous discussions of TAR and other return-free systems (U.S. Treasury 1987, GAO 1996, U.S. Treasury 2003) have grappled with how feasible it is for the IRS to use the information returns it already has to prepare tax forms in time for the April 15 deadline. Indeed, a Simple Return form would need to be sent out well before April 15, so that those who wanted to fill out their own form would have reasonable time to do so. There is no doubt that the Simple Return program would require some modernization of the processing capabilities of the IRS and its ability to match to the Social Security Administration records. The IRS already has a modernization program under way. If the IRS accelerated this modernization and implemented the Simple Return at an earlier date, it would thus offset the resultant additional costs, partially or wholly, by savings to the IRS from fewer errors in tax returns and increased electronic filing.

The entire California ReadyReturn pilot program cost less than $300,000, including the processing and mailing of the returns to fifty thousand people. The California Franchise Tax Board believed it could have increased usage of ReadyReturn up to tenfold with no increase in cost had the Web service been used more intensively.

Past estimates of the costs of a TAR type have varied widely. An IRS (U.S. Treasury 1987) study from two decades ago estimated that such a system would increase IRS costs by $284 million. A GAO (1996) study from one decade ago estimated that such a system would reduce IRS costs by $36 million. The GAO report pointed out that a TAR system requires far less document matching. Processing the standard returns that taxpayers send in is more expensive and time consuming than processing the Simple Return. The GAO noted that the IRS could reduce its number of "under-reporter" cases, whereby the IRS investigates discrepancies between the return filed by an individual and the information returns filed by employers and other organizations declaring sources of income for the individual. In 1996, these under-reporter cases cost an average of $17.61 for each taxpayer.

Much of the difference between the IRS estimate and the GAO estimate can be accounted for by the sharp decline in the prices of information technology and the corresponding increase in processing capabilities between the 1980s and the 1990s. Those cost trends have continued in the intervening decade. The GAO (1996) report of ten years ago may have been overly optimistic on the immediate cost savings at that time, but even being more realistic and assuming smaller initial cost savings, a Simple Return program implemented with modern technology would reduce the burden of tax compliance at little expense to the government.

As noted above, the modernization needed to enact the Simple Return is consistent with the IRS Business Vision Goal of reducing the amount of paper documents it needs to process, and with the ongoing modernization program at the IRS. To date, the IRS has focused most of its attention on increasing the amount of electronic filing to reduce paper forms. However, the Simple Return system also would entail a great deal less paper from the filers submitting their tax returns the conventional way. Raising the priority of modernization investments needed to accelerate the handoff of information from the Social Security Administration to the IRS will speed the date at which the Simple Return can be fully implemented and the associated cost savings realized.

What about Extending the Federal Tax Filing Deadline for the Simple Return?

The GAO (1996) report argued that the IRS could have enacted a TAR return-free system even a decade ago, if it extended the deadline for Simple Return filers past April 15, but that a hurdle to extending it was that many states would retain their April 15 deadlines. About two-thirds of all U.S. taxpayers live in a state where the state income tax requires some data that comes from the federal tax return, such as federal adjusted gross income or federal taxable income. These people would be unlikely to take part in a delayed Simple Return system because they would need to have completed their federal tax by April 15 in order to file their state tax returns.

Extending the tax deadline seems an inferior option, though it is possible if things do not go well in the development stages of the program. Clearly, it would be preferable to process the information sooner than to delay the filing deadline. At first glance, extending the deadline might be well received by taxpayers, but because about 80 percent of the people would be receiving refunds, most would prefer to receive their money as soon as possible. If the number of eligible people exceeds the capacities of the IRS to carry out a Simple Return in a timely manner, the program could be phased in. The IRS could first send Simple Returns to people in states that have no state income tax, or states in which the state income code does not rely in any direct way on the federal tax form.

How Would the Simple Return Deal with Mistakes and Corrections?

A Simple Return program would ultimately send out tens of millions of prefilled tax forms. Even if 99.99 percent of the returns were accurate, there would still be thousands of people receiving returns with mistakes. Some people might feel that they were being pressured, or that they were compelled to accept the government's numbers in the Simple Return; others might worry that they would not find a mistake on the return, or not be able to correct the mistake after finding it.

While it is impossible to avoid all mistakes in a Simple Return system, it is worth noting that the current system is not foolproof, either. Some employers misstate income on taxpayers' W-2 forms, or send the forms to the wrong address. People make mistakes on their own tax forms, leading to entanglement in an IRS process. Paid tax preparers are not a foolproof option either. The GAO director of tax issues testified before the Senate Finance Committee that 5 percent of the seventy-one million users of tax preparers had no confidence that they had not overpaid their taxes, and cited a nonrandom survey by Consumer Reports of twenty-six thousand of its readers, in which 6 percent had discovered an error made by their tax preparers (White 2003). In a different study, the GAO claimed that in 1998 as many as one million people using a paid tax preparer (that is, up to 1 percent of total taxpayers) had overpaid their taxes (GAO 2002).

There is, therefore, little evidence that an automated system such as the Simple Return would increase the number of mistakes in the tax system. Indeed, the Simple Return would allow some mistakes to be corrected more quickly. For example, if the government has the wrong amount of income on the W-2 form from a taxpayer's employer, the taxpayer will see the error when the IRS sends him the Simple Return, before he files his taxes. This would allow the taxpayer to mail in a copy of the correct W-2 with the return and avoid any further problems. Under the current tax code, the taxpayer would not even know about this mistake until after filing the tax return, having it go through IRS processing, and then receiving a letter of inquiry from the IRS months later.

The Simple Return could significantly reduce the time lag in resolving disputes and accelerate the time to receive a refund. The Simple Return, of course, would not remove ultimate responsibility from the taxpayer. The government would turn over the information it had on income and family status, but if that information were wrong -- for example, if the taxpayer actually had a large amount of capital gains income, disqualifying him from using the Simple Return; or if the taxpayer had married during the year, requiring a different tax calculation -- the taxpayer still would be responsible for reporting it. Failure to do so, under current law, would be just like failing to report income or misreporting one's marital status on a conventional tax return.

There is no viable alternative to the taxpayer being responsible for filing correct tax returns. If the government were legally to absolve the taxpayer from his obligation of checking the Simple Return, it would open up the possibility of fraud and give people an incentive to convey misleading information to the government in an effort to induce the IRS to send them a faulty Simple Return that they could then sign and make official. To ease fears that a government mistake that people did not find would end up costing them a great deal of money, the IRS could set up a safe harbor level. With such a system, if the taxpayer failed to check the government's return and simply signed it and returned it and it later became clear the taxpayer had failed to report some kind of income or change in family status, as long as the tax owed was below a certain amount (for instance, $250) there would be no additional penalty beyond the tax itself and interest. This is similar to the way a taxpayer can request an extension beyond the April 15 deadline in the current system.

A related issue to the one about mistakes is the issue of whether a taxpayer would be intimidated into paying more than he should, or into agreeing with something incorrect on the return because the document comes from the IRS. It is important to remember that the program is voluntary. If any taxpayer feels that he is being intimidated or cheated, or even feels annoyed by the Simple Return, he can simply discard the Simple Return and file his taxes the conventional way. The evidence from the California ReadyReturn shows most people have no problem rejecting an offer from the tax authority and doing their taxes on their own. Furthermore, most of these taxpayers do not have access to extensive deductions and credits, and so would not risk losing them by filing a Simple Return.

The point of the Simple Return is to make life easier for U.S. taxpayers, not to increase their stress level, however. For that reason, it would probably be better for the Taxpayer Advocate System to send the letter offering the Simple Return.⁸ A letter coming from the IRS directly might make people think they were being ordered to pay a specific amount or believe that this was, in fact, the beginning of an audit.

Does the Simple Return Raise Privacy Concerns?

Some opponents of return-free filing find it invasive or inappropriate that the government would print up and mail out forms listing income and taxes. Indeed, some opponents suggest that a Simple Return would require people to divulge additional personal information to the government. Just to be clear, however, the Simple Return does not require employees to give their employers or the government any more information about themselves than they give now, nor does the Simple Return entail the IRS receiving any more information than it receives now about wages or family status. Indeed, some privacy advocates strongly supported the California ReadyReturn pilot program because the system inherently requires the government to turn over all the information that it has on each individual -- to lay all its cards on the table, in a sense.

Obviously, when tens of millions of these forms are mailed out, a small fraction may go to the wrong address. This is no different from the current system that mails W-2 information, or the Social Security Administration's mailing that lists year-by-year earnings over a lifetime. The risk that some of this information will fall into the wrong hands is not new.

Would the Simple Return Unfairly Infringe on Private Enterprise?

In California, opponents of the ReadyReturn argued that return- free filing constituted an inappropriate government intrusion on private enterprise. Bankman (2005) has pointed out that this argument implies making the tax system more complex and more painful is desirable because doing so would increase the employment of tax preparers. In practice, the government already seeks to reduce the compliance burden of taxation in various ways. For example, the government provides people with a printed tax table indicating the tax burden for a given level of taxable income so that people do not need to calculate using the formula. Few would advocate removing the tax tables because they undermine the market for paid tax preparers. Likewise, a taxpayer who does not have enough withholding at tax time and owes a penalty can ask the IRS to calculate it for him, because the rules are a bit complicated. Again, few people complain that such a service undermines free enterprise. For the government to release the information it already has and to give taxpayers an idea of what their tax situation is (which the government already does now if the taxpayer makes a mistake when he first files) should hardly be considered entering into competition with a private sector business.

Would the Simple Return Raise Taxes?

Antitax groups and some in the U.S. Congress (see Americans for Tax Reform 2005) publicly oppose return-free filing. For example, Grover Norquist, president of Americans for Tax Reform, testified before the President's Tax Reform Commission against any kind of automatic filing (Norquist 2005). At first, such opposition seems ironic, because antitax groups have long been the most vocal critics of the compliance costs of the tax system. However, these groups seem to believe that, if compliance with the tax code were to be less painful, people would be less adverse to higher tax rates. These critics typically ask rhetorically, "Do you trust the government to do your taxes for you?" And they argue that return-free filing is just a way for the government to raise taxes that people will not notice and a way to expand the power of the IRS over people's lives.

But the Simple Return is completely voluntary. No one needs to trust the government, share any additional information with the IRS, or pay a higher tax rate than they would without the Simple Return. Every taxpayer has the right to set aside the Simple Return and file the conventional way. In addition, the government is only doing this for people with extremely simple tax positions so there is little room for the government to cheat people out of their deductions or to induce them to pay higher taxes. People are not sharing any additional information with the IRS. Everything on the Simple Return comes from information the IRS already has about the taxpayer and information that he gives already to the IRS every time he files his tax return. Indeed, if the taxpayer were accidentally to leave this information off his tax return, the IRS would contact him and tell him that he forgot to report the information.

V. The Interaction of the Simple Return with Other Tax Policies

The Simple Return is consistent not only with current tax law, but also with the major tax reform proposals that have been the subject of recent debate. In contrast, alternative proposals to achieve return-free filing (such as exact withholding) would be possible only if major changes were made to the U.S. tax code.

The Simple Return and Federal Tax Reform

The tax code is continually evolving. The Simple Return could be adapted easily to handle almost any changes in basic tax rates and standard deductions, and even could be adapted to handle many of the more sweeping tax reforms that are sometimes contemplated.

One kind of tax reform would broaden the base and lower the rates, meaning reduce or eliminate a number of tax deductions and credits -- such as the tax deductions for mortgage interest or state and local taxes -- and then take the money that is saved and use it to reduce tax rates. The Tax Reform Act of 1986 proceeded along these lines, and the recent President's Advisory Panel on Federal Tax Reform (the Advisory Panel) produced one proposal similar to this. A second kind of proposal for tax reform would focus on switching from the current income tax to a consumption tax. Some proposals along these lines would exempt from income tax all returns from financial investment such as interest earned, dividend payments, and capital gains income. Yet another substantial reform discussed is to abolish or sharply limit the alternative minimum tax (AMT). The Advisory Panel was particularly concerned with this issue. The AMT was originally passed in the late 1960s in an attempt to ensure that all high-income taxpayers would pay at least some taxes, no matter what. Over time, though, an increasingly large share of taxpayers has been affected by the AMT, and it is slated to become an issue for even middle-income people in the near future.

The Simple Return would work well with any of these major tax reforms and, indeed, would likely be available to even more people if such reforms were enacted. The Simple Return system is not dependent on the current income tax system to function; the benefits of the policy could be sustained no matter how the income tax changed. If a tax reform eliminated a number of deductions and, as a result, reduced the number of taxpayers who itemize deductions, then the number of taxpayers who could use a Simple Return would rise. One of the major factors preventing more people from qualifying for the Simple Return is that capital gains income cannot be dealt with in the return-free setting because capital gains are not reported to the government by any third party (unlike, say, the interest earned from a financial institution). A consumption tax that exempted capital gains income would thus entitle more people to use a Simple Return.

The Fourth Wave

A reform that limited the sweep of the AMT might even allow the launch of a fourth wave of Simple Returns that would include what have been called simple itemizers. People in this group itemize deductions, but only a few of them -- mortgage interest, state and local taxes, and charitable contributions. Gale and Holtzblatt (1997) estimated that this group could include almost five million additional filers, or about 10 percent of those who currently itemize deductions (see table 1). In a fourth wave, taxpayers would be sent both a basic Simple Return and an itemized 1040 form with all deductions filled out as zero except three blank lines for mortgage interest, state taxes, and charitable contributions. The filer would complete these three lines, add up the numbers, and compute his tax. If the Simple Return with itemized deductions yielded a smaller tax bill than the basic Simple Return, the filer could just mail in the itemized return (and keep his documentation of the itemizations).

This variation of the Simple Return to cover simple itemizers currently is impossible because of the presence of the AMT. Every taxpayer owes either the taxes calculated on his version of the 1040 form or the taxes calculated on the AMT form, whichever is greater. If a taxpayer received a Simple Return that allowed for a few itemized deductions, that taxpayer might fill in values for charitable contributions, for example, that were large enough for him to reduce the taxes he owed to a level that would oblige him to pay AMT. There would be no way for him to know about this possibility, though, if he filled out the Simple Return. Such a taxpayer would, then, most certainly underpay his taxes and have to pay penalties. If the AMT were limited to those with very high incomes, this problem would disappear and the fourth wave of the Simple Return could take place.

The Simple Return and State Income Tax Changes

The Simple Return would likely put pressure on states to offer their own versions of a Simple Return. People clearly would not like having to do a tax return for their state income tax when the federal government has sharply reduced the cost of compliance. The California ReadyReturn case, however, suggests that many states would potentially be able to do this. In fact, state employment offices are thought to be better equipped than the Social Security Administration to convey wage information in a timely manner because the state unemployment insurance numbers often are updated throughout the year. There certainly would be a great interest among many states in a joint program with the federal government whereby states could jointly send out a state form with the Simple Return. This is especially relevant for the majority of state income taxpayers who live in states where the federal definition of gross income or taxable income is the starting point of the state income tax. Of course, in the seven states that have no state income tax -- including populous states such as Texas and Florida -- there would be no need for this cooperation.

Why Not Exact Withholding?

A frequently mentioned alternative to a TAR system such as the Simple Return is an exact withholding system, in which government adjusts the amount that is automatically withheld from paychecks so that the total amount withheld at the end of the year is the exact amount of taxes owed. In this way, exact withholding eliminates the need for filing a tax return. This system is used in a few countries, including Japan and the United Kingdom. Adopting this system in the United States would require some significant changes to the tax system. Indeed, this is one of the main conclusions of the report prepared by the U.S. Treasury (2003) entitled "Return-Free Tax Systems: Tax Simplification is a Prerequisite."

In countries with exact withholding, there tends to be either no taxation of capital gains and interest income, or else such income is withheld at the source, similar to wage income. Furthermore, the tax structure tends to treat all filers as separate individuals filing alone, rather than as married; there is only a limited role for allowing special deductions or credits, because everything must be clarified ahead of time in order for the firm's withholding to match exactly the tax bill. In a tax code such as the U.S. tax code -- that taxes capital gains and interest income, does not withhold at the source on such income, changes the tax rate according to marital status and children, and has an EITC -- it would be difficult to exactly calculate withholding. Gale and Holtzblatt (1997) describe in more detail the kinds of changes one could make to the U.S. system to make it easier to enact an exact withholding system, but several hurdles seem insurmountable. First, the government would need to make significant changes to the tax code (such as not taxing capital gains and treating all taxpayers only as individuals) before it could enact exact withholding. Second, any time the government changed the tax system it would have to ensure that all employers changed their withholding rates to make sure they stay exact. Third, exact withholding would require employers to gather significantly more information about employees than they currently do, in matters such as number of children, house payments, charitable deductions, and so on. This entails both some privacy concerns for the employees who may not want their employers to have the additional information, and some cost concerns for the employers who will have to do the exact withholding calculations. Fourth, in a recent survey, the public preferred a TAR system such as the Simple Return over exact withholding by a more than three to one margin -- 65 percent to 19 percent, with 17 percent having no preference or not answering (U.S. Treasury 2003).

The Simple Return -- whereby the tax authority simply takes the information it already receives from employers and financial institutions, computes the tax, and mails out the completed form to qualifying taxpayers -- requires almost no change to the tax law before implementation, and is preferable.

VI. Conclusion

The Simple Return is a straightforward idea. It requires that the tax authority take the information it already receives from employers and banks on the income and tax situation of taxpayers and use it, wherever possible, to send out a return that can spare the taxpayer the hassle of filling out a tax return or hiring a preparer. Under the current tax code, the Simple Return could eventually encompass nearly 40 percent of U.S. taxpayers. It could reduce the burden of tax compliance on Americans by about $4.4 billion a year. Most of these benefits would accrue to taxpayers with middle and low incomes. In recent decades, the tax authorities have made a substantial push toward collecting information from employers, financial institutions, and taxpayers in electronic form. As the costs of information technology and communications continue to plummet, it is time to take the next step and enact the Simple Return.

References

I wish to thank Jen Paniza for research assistance and Peter Orszag, Timothy Taylor, Joshua Bendor, and the participants at The Hamilton Project authors' conference for helpful comments.

Author

AUSTAN GOOLSBEE

Professor of Economics, University of Chicago Graduate School of Business

Austan Goolsbee is the Robert P. Gwinn Professor of Economics at the University of Chicago Graduate School of Business. He is also a senior research fellow at the American Bar Foundation and a research associate at the National Bureau of Economic Research. His areas of expertise include tax policy, budget and fiscal policies, public and antitrust law, and the information economy. He is a Sloan Research Fellow and a Fulbright Scholar and has been named one of the 100 Global Leaders for Tomorrow by the World Economic Forum in Switzerland. Previously, Goolsbee served as a special consultant for Internet Policy, Department of Justice, Antitrust Division, 2000-1; was lead editor for the Journal of Law and Economics, 2001-4; and was associate editor for Law and Social Inquiry, 1997-2001.

² The W-2 form is the form that an employer must send each year to employees, the IRS, and the Social Security Administration reporting each employee's wages and the amount of taxes that the employer withheld during the previous year.

³ The child credit reduces a family's taxes according to a schedule that is based on family income and the number of dependent children.

⁴ The EITC is the tax credit that is available for low- income families and is adjusted according to income earned and the number of dependent children.

⁵ This predicted fee comes from regressing preparer fee (for people who hired a tax preparer and reported expenses greater than $0) on AGI and dummies for each schedule A to F, and for the EITC. The average value of the predicted fee for itemizers was around $190.

⁶ The National Longitudinal Survey of Youth has followed a group of people, aged 14 to 21 in 1979, throughout their lives.

⁷ Assuming some part-time rather than full-time work typically raises the implied hourly wage of the lower-income workers (and thus the implied compliance costs among the low-income workers). Using only single taxpayers to avoid any issues associated with two taxpayers both working full time gave similar results, so the listed calculation is quite robust.

⁸ The Taxpayer Advocacy System is an independent organization within the IRS set up to protect the rights of taxpayers and advocate their interests in the event of disputes with the IRS.