problem is a mass enforcement problem and probably cannot be licked just by applying more manpower and going out and seeing the taxpayer. It indicates we must utilize machines and these processing division techniques and low-cost methods of getting to the final

answers.

We must somehow strive to discover, either how to perfect old ways of doing old things, or finding new ways of doing old things. So much for the size of the job.

MONEY VALUE OF TAX ADJUSTMENT DUE TO INCORRECTLY COMPUTED

RETURNS

Of course, you are interested in the money aspect of it. How much money is involved in all of this work? That is more tenuous than these figures I have given you. These figures are very preliminary, but we may wind up with 14,000,000 instead of 13,600,000, or we may wind up with 13,200,000.

It is a pretty good figure anyway.

These next figures I am going to give you are the best that we have at the moment from this project. About a month or so ago, we designed a little test in this audit control program so that we could see what kind of bugs we had in our tabulations so that we would clean up the problems before we came to making our estimates.

We picked three States because they were rather small and we had a considerable higher proportion of returns than in some other areas. We had the fewest problems there we thought. These were the States of Georgia, North Carolina, and Indiana.

We did not pick them because they were average or representative. We just took them because those were the States that happened to be handy.

That sample was incomplete, but we did get some amounts, some averages. Using those averages and applying them to, roughly, 13,000,000 returns which were then estimated to be in error, we estimate that for 1948 returns alone that on the basis of these preliminary averages that the total tax adjustment is about one and a third billion dollars. That figure will change and we do not want you to hold us too closely to it. We are doing the best we can as we go along. That includes both deficiencies and overassessments.

In other words, there is a certain proportion of this one and onethird billion dollars which we have to give back, but our responsibility is to determine the correct tax. If the man has paid too much, heaven knows we ought to give it back to him.

We also attempted, because it is so rough, to do another type of estimate. The errors of taxpayers who had changes, amounted to about ten and two- or three-tenths percent of the total tax all taxpayers reported.

In other words, when we found a taxpayer with an error, both the small and the very large, it averaged ten and two- or three-tenths percent of the tax all taxpayers said they owed. We took that figure and applied it to an estimated $17,000,000,000 liability for the year 1948, and that gives us about 134 billion dollars.

So, our estimate at this time is that there is outstanding, or let me put it this way: For the year 1948 there is a tax adjustment in

volved of somewhere between one and a third and 134 billion dollars in the individual income-tax field alone for this 1 year.

There are additional adjustments involved for other years.

Mr. CANFIELD. You use the word adjustment. I assume you mean just that?

Mr. FARIOLETTI. I mean tax adjustment, one way or the other. We do have an indication that all but somewhere around 100 million of this 113 to 134 billion dollars is owing the Government, but that is more tenuous because it is limited to a few returns.

It is difficult to make a reasonable estimate when you only have a few returns. When we get the final estimates, we will get a much more stable figure. It may go up. Right now it looks a little low.

We tried to figure out what that meant. How much are we missing? We figured something like this: We said that in fiscal 1949 we actually collected an additional 800 million dollars from individual income taxpayers. That involved mostly 1946 and 1947 returns, and it also includes interest and penalties.

So, we took this 113 and 134 and we increased it by 10 percent on this theory: That if we did examine all of these returns within 3 years, on the average, there would be a 9 percent interest adjustment, or 112 years at the rate of 6 percent.

Then, we added 1 percent for penalty, which is, I think, a little low, but anyway, we got 10 percent. That increased the additional assessments estimates up 1.4 billion dollars to 1.8 billion dollars.

Then we said his $800,000,000, that involves returns for years other than 1948.

Beginning in 1948 we had the Revenue Act of 1948 which increased exemptions and reduced rates and gave the income-splitting features. We estimated at the time when that was done that that would reduce tax liability on the average about 20 percent. It varies.

We took this $800,000,000 and we reduced it by 20 percent, and that gave us $640,000,000 dollars.

When we subtracted the 640 million dollars from the 1.4 billion dollars, and 1.8 billion dollars and we said, if we examine the 1948 returns with the same procedures and the same manpower that we used in examining returns in fiscal 1949, we will not catch from, roughly, three-quarters to 1.2 billion dollars. That would remain outstanding and unenforced. That is how we got to this estimate that you heard of, say, roughly three-quarters of a billion dollars.

The next thing I think that I would like to point out to you is just what does this mass enforcement mean. Let us take this one and one-third billion dollars. Roughly 80 percent of that one and onethird billion dollars, and let us call it $1,000,000,000 falls into two areas that is the collectors' 1040 area, and the agents' under $25,000 where we showed in our estimates that there were about 33,000,000 returns and about 11,000,000 tax errors.

That just shows you that even if you take a cycle approach to this thing, after you got your cycle of 3 years, you have caught up and you are back where you were before.

Somehow, either through educational programs or through paper shuffling, through these machines, et cetera, we have got to find ways of getting to better leads than we have, so let us look a little bit at the kind of mistakes that we found; the major errors we found these people making.

TYPES OF ERRORS REFLECTED ON RETURNS

In the Form 1040-A area there are 11.6 percent of the taxpayers who made mistakes. There are 2.2 million of them, let us say. A third of those 2.2 million, roughly, made the major error in income. They either put in too much, or they put in too little income. Mostly they put in too little, and while they are subject to withholding, there are many people who work for more than one employer. Every once in a while they forget to put in more than one withholding return from their employers.

The Processing Division must match their original withholding returns which are attached to their income tax forms and that should help us pick that up.

The other two-thirds make mistakes in claiming exemptions. If we can discover patterns of error, what kind of people, what kind of taxpayers, or what kind of exemptions are most frequently in error, then we can devise means for sending out additional questionnaires to that kind of a taxpayer saying thaat we would like more information about the exemption in question.

Where we have attempted to do that already, a high proportion of them write back and say that somebody told them to claim it. They say that they made a mistake. They ask us to please bill them. In other cases you have to go out and see them.

That, of course, involves manpower which costs more money. In the collector's 1040 area 50 percent of the errors are in income. That is because we are already getting into a different kind of taxpayer. We are getting into the small farmers and small-business men, rents, interests, and dividends, you see. About a fourth of them had errors in deductions. That is, they made mistakes in itemizing deductions. They claimed too much medical deductions, too much contributions. or too little, or they picked the wrong standard deduction. Instead of taking the tax table, they took the 10 percent, which is incorrect.

A fifth of them claimed the incorrect exemption. So we have a big exemption problem in those two groups because we have over 3,000,000 taxpayers involved. Then the others had mathematical errors or the major error was a mathematical error, and the machines about which Mr. Williams has been talking should pick that kind of thing up.

As we move up into the higher income tax returns what we find is that the exemption problem disappears. It drops down and virtually disappears, but the problem of reporting correct income becomes a very troublesome issue.

Sometimes they do not report the receipts correctly, and sometimes they do not report the deductions correctly. For example, a restauranteur or a retail grocer or butcher will take things off the shelf and use them for his family. That goes into his cost of goods sold whereas actually it is for his own consumption. It should not come out of his business income which he understated by blowing up his cost of goods sold. That is a much more difficult error to find and cannot be discovered by paper shuffling and matching documents.

There you have to get out and actually contact the taxpayer and examine his books and records.

In these higher income areas also we have the problems of personal deductions and curiously enough the errors in personal deductions become relatively more important as the income increases. As we pointed out, the mathematical errors can be important.

To get into the details which we do not even have a rough idea about now, let us take the business returns. We plan to take all the business returns that are in this sample and classify them by type of business.

METHODS OF CATCHING ERRORS ON RETURNS

Then we plan to show the changes and the source of changes and the size of changes, the importance of that size relative to what was originally disclosed. So that we can see whether doctors or lawyers or liquor dealers or restauranteurs or retail food outlets are in these areas that appear to be more out of line than others and should have a priority of selection. Those are the sort of things we are attempting to do.

We do not know what the final results are going to be because this is another experiment, but we think it will pan out.

We are going to analyze it not only in Washington by statisticians, but we are going out in the field with these things, and we are going to say, "Look, this is what you have found. We have put it together and made the estimates. What do you think now?

Tell us what

you think about it, and we will try to get as many views as we can on how to improve our attempts to get to the areas which need. enforcement before we get to areas that do not need enforcement as much."

It is the old priority problem because you cannot do all of them. You cannot attempt or look forward to auditing fifty-some-million individual returns.

PLAN OF ENFORCEMENT OPERATION FOR THIS YEAR

For 1949 we are in the unfortunate position that we have to plan ahead, even though the 1948 results are not all in. We have to get started now so that we can work in our 1949 program on time. We plan at this time to continue in the individual area, but reducing the size of the individual income tax sample drastically, except in the business area where it looks like most of the big problems or many of the big problems are.

We are also going into small corporations and we are also going to attempt to do something like this: When a deputy collector or an agent goes out to see you because your audit-sample return has been selected for examination, if you are an employer, that deputy or that enforcement officer will, in addition to doing your income tax, also do your withholding tax, your social-security taxes, and if you are subject to excise taxes if you are a manufacturer or a retailer who must collect excise taxes, he will also examine that aspect.

So that one man going out to see the taxpayer one time does the whole job. That is quite an undertaking because it involves training and bringing in a lot of different parts of the multifarious laws, but we think it can be done. We are going to try it.

As a result of that, we will then be able to see whether there were any big changes in 1949 as compared to 1948 in the individual area. We can then do some rather detailed analyses of the United States small corporations, that is, the 80 percent of the corporations that have assets of under $250,000.

We can get an insight to begin thinking about how are we going to improve our enforcement methods in the, perhaps, even more difficult areas of excise-tax enforcement. That is the nature of the auditcontrol program.

Mr. GARY. We are very glad to have had that statement from you, Mr. Farioletti. I think it will help to work out some of your problems. Are there any questions?

Mr. CANFIELD. I have no questions, except that I want to say, Mr. Chairman, that I have listened very attentively and I have gotten much out of the statement of the witness. I wish that all the members of the subcommittee of this committee that thought about these things 2 or 3 years ago and the Joint Revenue Taxation Committee that made their studies and criticism of the Bureau could be here today to hear the progress that is being made.

We are certainly marching forward, Mr. Commissioner, and I am glad to hear of this work.

Mr. SCHOENEMAN. Thank you.

Mr. FARIOLETTI. Pardon me, Mr. Commissioner, but I would like to say, Mr. Chairman, that as soon as our preliminary estimates are available, we will present them to the members of the committee and for the record, and also send them to the joint committee.

As soon as our final estimates are available, we will do the same thing. So that these two committees who have a responsibility for the Bureau of Internal Revenue will have as much information as we have in evaluating and helping us determine what avenue we should pursue more intensively than in the past in this very difficult enforceinent problem.

Mr. COUDERT. Mr. Chairman, I would like to add a word of comment. Your analysis of the program that your Bureau has put into effect, Mr. Commissioner, is very interesting. I trust that the administration and Congress will draw the moral from the apparently large proportion of error in these returns, to wit the hopeless difficulty and the great burden that our complex and almost incomprehensible system of income-tax law imposes upon the unfortunate taxpayer of the United States.

Mr. CANFIELD. For which the Commissioner and the Bureau are not responsible.

ESTIMATED INCREASE IN 1951 IN DIRECT TAXES ON INDIVIDUALS OVER 1950 AND DECREASE IN DIRECT TAXES ON CORPORATIONS

Mr. GARY. Mr. Commissioner, I notice in the budget for fiscal 1951 an estimated increase in direct taxes on individuals over 1950 and a decrease in direct taxes on corporations.

Can you tell us why that is?

Mr. SCHOENEMAN. I think Dr. Atkeson can make an estimate, or a guess on that, only. He may be able to throw some light on it.