Tax Administration: Data Needed on Whether to Regulate Filers of Information Returns for Others

Published by the Government Accountability Office on 1990-07-03.

Below is a raw (and likely hideous) rendition of the original report. (PDF)

                                                                                      l_.l   ____.           -_-._-..         .   ..““^..
._-._            --..--.---_---.-.--


                                                       .-..-   I   .._.....   -.l-l                  --.-.              ---
  __-__                                -   .____   I

   .July. l!I!tO
                                                                                                                                                           Data Needed on
                                                                                                                                                           Whether to Regulate
                                                                                                                                                           Filers of Information
                                                                                                                                                           Returns for Others


          ___~                             -__....-_-                                         “-._~___-                                     ___-_...   -
      United States
GAO   General Accounting Office
      Washington, DC. 20548

      General Government    Division


      July 3, 1990

      The Honorable Lloyd Bentsen
      Chairman, Committee on Finance
      United States Senate

      The Honorable Dan Rostenkowski
      Chairman, Committee on Ways and Means
      House of Representatives

      This report responds to section 7716 of the Omnibus Budget Reconcilia-
      tion Act of 1989. Section 7716 requires that we study whether service
      bureaus that transmit information returns or other documents to the
      Internal Revenue Service (IRS) on behalf of other persons (payors)
      should be subject to registration or other regulation. On the basis of dis-
      cussions with the House Ways and Means Committee, we limited our
      review to service bureaus that submit information returns on magnetic
      media for payors.

      In a 1989 study of civil tax penalties, IRS reported that its Martinsburg
      Computing Center (MCC) often has to reject information returns sub-
      mitted on magnetic media by service bureaus because the media cannot
      be processed. Magnetic media include magnetic tape, 8“ disks, and
      floppy disks. According to the study, these media contain errors, such as
      missing data elements or incorrect record formats, that prevent them
      from being processed on IRS’ computers. When incorrect magnetic media
      are returned to service bureaus, payors who are responsible for filing
      these information returns may be assessed late filing penalties. Also, if
      the returns are not corrected, IRS may not be able to use the rejected
      information returns in its document matching programs. However, the
      Internal Revenue Code has no provisions to penalize those service
      bureaus that may be responsible for submitting unacceptable informa-
      tion returns on magnetic media.

      IRS could not identify how many service bureaus file information returns
      on magnetic media and could not provide us with any quantitative anal-
      ysis on problems specifically relating to service bureaus. However, using
      ES databases, we were able to estimate the potential population of ser-
      vice bureaus using magnetic media and compare problems related to ser-
      vice bureau filings with problems of payors filing information returns
      for themselves.

      Page 1                                        GAO/GGD-90-96   Tax Administration

                       does not appear to have more of a problem with service bureaus that
Results in Brief   IRS
                   file magnetic media than it has with those who file magnetic media
                   information returns for themselves. We found that over 96 percent of
                   the tax year 1988 information returns submitted on magnetic media by
                   all filers, including service bureaus, were readily processed, that is,
                   processed the first or second time they were submitted.

                   Our analysis shows that rejection rates for disks and magnetic tape of
                   filers likely to be service bureaus was 12.4 percent for the first time
                   they were submitted and 2.3 percent for the second time for tax year
                   1988. These figures were about the same as the 12.5 percent first-time
                   and 2.2 percent second-time rejection rates for businesses filing their
                   own information returns on magnetic media.

                   While the vast majority of magnetic media submitted for tax year 1988
                   was readily processed, IRS had to use resources to resolve problems with
                   over 23 million information returns that were not acceptable the second
                   time submitted. As many as 6 million of these returns may have been
                   submitted by up to an estimated 203 filers that are likely to be service
                   bureaus. The 203 filers were filing information returns for 8,757 payors.
                   These 203 service bureaus represent about 5 percent of the universe of
                   potential service bureaus. We do not know to what extent payors paid
                   penalties due to problem returns submitted by service bureaus because
                   IRS does not have a database that identifies service bureaus or their

                   The potentially large number of magnetic media information returns and
                   payors affected by these service bureaus, coupled with IRS’ lack of hard
                   data on service bureaus filing information returns on magnetic media,
                   suggests that IRS needs information to better define the problem. IRS
                   should know who the service bureaus are, exactly what kind of mag-
                   netic media processing errors they cause, and how much regulating or
                   controlling service bureaus would cost IRS and the industry. As part of
                   the cost, IRS should consider the impact any regulation or control would
                   have on new small businesses entering into the industry so industry
                   growth or viability would not be unduly stifled.

                   Once IRS has this type of information, it will be better able to consider
                   whether various options for regulating service bureaus are warranted.
                   Imposing the same controls on service bureaus submitting information
                   returns as IRS imposes on filers of other returns is one way to deal with
                   all service bureaus, including the problem ones. IRS does not oversee
                   those that submit information returns for others on magnetic media

                   Page 2                                        GAO/GGD-SO-96   Tax Administration

             with the same level of standards and controls that it imposes on those
             that submit income and employment tax returns for others, either elec-
             tronically or on magnetic media. For these latter filers, IRS sets specific
             media standards and controls for approving, denying, and revoking
             filing privileges.

             However, the overall error rates may be too low to justify new regula-
             tions for the whole industry when the problems appear to be caused by
             relatively few service bureaus. Other options are available that would
             allow IRS to deal more effectively with the relatively few service
             bureaus causing problems without subjecting the entire industry to reg-
             ulation or other controls. These options would include applying penal-
             ties to service bureaus with repeated transmission errors, authorizing
             IRS to revoke transmission privileges, and providing information to
             payors on service bureaus with repeated transmission problems. These
             other options would only be applicable to that small percentage of ser-
             vice bureaus failing to correct errors after reasonable notice and feed-
             back from IRS.

             Service bureaus may file information returns with IRS on payments
Background   made by payors for interest, dividends, and other types of income. Ser-
             vice bureaus filing these returns are the ones the 1989 act intended for
             us to study. Service bureaus may also file other returns such as employ-
             ment tax returns and wage and tax statements (Form W-2).

             Before tax year 1989, information returns sent to IRSgenerally had to be
             submitted on magnetic media if the payor submitted 50 or more of a
             single type of information return. The 1989 Omnibus Budget Reconcilia-
             tion Act raised this threshold to 250. Thus, payors who submit fewer
             than 250 of a single return may now use paper information returns. The
             250-return threshold applies separately to each type of return and sepa-
             rately to both original and corrected returns. Thus, corrections to
             returns originally filed on magnetic media, if less than 250, may be made
             using paper forms.

             Payors must send information returns required to be filed on magnetic
             media directly to MCC, while information returns that are filed on paper
             forms must be sent to the appropriate IRS service center. Most informa-
             tion returns are due on February 28, the major exception being the Indi-
             vidual Retirement Account return (Form 5498), which is due on May 31.
             Payors may be assessed a penalty of up to $50 for each information
             return that is incorrect, not filed on time, or not filed in the proper

             Page 3                                         GAO/GGD-9096 Tax Administration

                        format, depending on when a corrected information return is filed. The
                        objective of the penalty is to encourage payors to file the information
                        returns in a usable format in time for use by IRS. IRS uses the returns to
                        identify taxpayers who may not have filed an income tax return or
                        reported all their income.

                        For tax year 1988, MCC received 670,309,254 information returns in
                        66,793 shipments of magnetic media from over 32,011 filers.’ These
                        numbers do not include Forms W-2 (Wage and Tax Statements) and W-
                        2p (Statements for Recipients of Annuities, Pensions, Retirement Pay, or
                        IRA Payments) because they are filed with the Social Security

                        Pursuant to the Omnibus Budget Reconciliation Act of 1989 and our dis-
Objectives, Scope,and   cussions with the House Ways and Means Committee, our objectives
Methodology             were to (1) determine the extent of any problems IRS had with magnetic
                        media information returns filed by service bureaus and (2) explore the
                        need for any administrative or legislative measures to regulate the ser-
                        vice bureaus submitting these returns.

                        IRS  does not maintain information that readily identifies service bureaus
                        or any problems encountered with the information returns they submit.
                        As a result, IRS does not know how many service bureaus submit infor-
                        mation returns on magnetic media and could not identify problems spe-
                        cifically relating to service bureaus. Therefore, to meet our first
                        objective we first estimated the number of potential service bureaus
                        sending in information returns on magnetic media. To do this, we
                        matched the employer identification number (EIN) of the payor        in IRS’
                        Payor Master File against the EIN of the filer in MCC’S tax year 1988
                        information returns database. This comparison identified 4,102 filers
                        with EINS that did not match those of the payor. These 4,102 filers
                        became our population of potential service bureaus.

                        Thus, we defined a potential service bureau submitting information
                        returns as any organization that filed information returns on magnetic
                        media to MCC for payors identified by a different EIN. This definition
                        could also encompass filers that submitted magnetic media information
                        returns for only one other EIN, financial institutions that filed for trusts,

                        ‘These figures do not include about 7.6 million returns shipped from IRS’ Detroit Data Center that
                        were erroneously included in the transmitter database.

                        Page 4                                                         GAO/GGD-90-96     Tax Administration

                         and parent organizations that filed for subsidiaries and affiliates.2 We
                         had to adopt this broad definition because IRS data do not identify ser-
                         vice bureaus separately. While this definition is imperfect-for
                         instance, it includes related parties, such as parent companies, as poten-
                         tial service bureaus-we believe and IRS agreed it presents the best defi-
                         nition possible given the limited information IRS had available in its

                         We matched these potential service bureaus against those filers whose
                         magnetic media were rejected by MCC to identify potential service
                         bureaus that had problems providing acceptable magnetic media to MCC.

                         To meet our second objective of exploring the need for administrative or
                         legislative measures to regulate service bureaus filing information
                         returns, we interviewed IRS officials; representatives of service bureaus;
                         and an official of a data processing association-an umbrella group rep-
                         resenting service bureaus filing on magnetic media. In addition, we
                         reviewed existing regulatory and statutory requirements concerning the
                         regulation of service bureaus filing other types of tax returns on mag-
                         netic media to determine IRS’ authority to regulate service bureaus that
                         submit information returns. Our review was done between January and
                         June 1990 using generally accepted government auditing standards.

                         The proportion of magnetic media submissions rejected from all filers
The Number of Media      including service bureaus has significantly declined since 1987. How-
Submissions Rejected     ever, over 99 million of the 670 million information returns submitted
Has Declined, but        on magnetic media for tax year 1988 could not be processed the first
                         time submitted, and over 23 million could not be processed the second
Millions of Returns      time submitted. Our analysis of IRS data shows that the returns sub-
Still Cannot Be          mitted on magnetic media by potential service bureaus do not create
                         more problems for IRS than the information returns it receives on mag-
Readily Processed        netic media from other filers,

Problems With Magnetic   MCC  data show substantial improvements for all filers, including service
Media Have Declined      bureaus, in the percentage of all magnetic media information returns
                         received that can be processed. As shown in table 1, MCC reported it
                         rejected 27 percent of magnetic media submissions for tax year 1987, 16

                         “We did not include aa service bureaus organizations that converted paper returns to magnetic media
                         when the media were transmitted directly by the payor. In these cases the payor and transmitter
                         were the same entity, and the payor would be the first to know if the media submitted could not be

                         Page 6                                                       GAO/GGD-90-96     Tax Administration

                                          percent for tax year 1988, and 7 percent for tax year 1989 as of June 1,
                                          1990.3 Because most media submissions contain hundreds of information
                                          returns, the table also shows our calculations of rejection rates for the
                                          number of returns contained on the rejected media, which showed a sim-
                                          ilar decline.

Table 1: Magnetic Media and Information
Returns Received by MCC and Returned                             Media submissions                       Number of information returns
to Transmitters                           Tax war           Received   Returneda Percent                 Received    Returnedb Percent
                                          1987                 108,736          29,213         27      687,627,224    226,405,275              33
                                          1988                  90,769          14,418         16      670,309,254     99,183,103              15
                                          198gc                 82,885           5,670          7      645805,046      80,429,737              12
                                          aNumbers rcturned at any time for resubmission. MCC adds one each time the media replacement is
                                          bNumbers returned the first time for resubmission

                                          %formation returns for tax year 1989 were being processed at the time of our review. The number of
                                          media and documents received and returned are as reported by MCC for the period ending June 1,
                                          1990, and are adjusted to exclude those received but not worked.

                                          MCC  officials stated that an IRS format change for submitting magnetic
                                          media in tax year 1987 probably contributed to the high rejection rate
                                          that year. The lower rejection rate for tax year 1989 suggests that filers
                                          may be improving their submissions as they acquire more experience
                                          with IRS’format requirements.

Lower Error Rates When                    For tax year 1988, IRS did not have written criteria for how many times
Using Criteria for                        magnetic media could be returned to the submitter for correction before
                                          a penalty may be considered. However, it does have written criteria for
Processing Paper                          processing information returns submitted on paper documents. These
Documents                                 criteria for paper state that if the first document submitted is not
                                          acceptable, the problem should be identified and the document returned
                                          to the payor with a request that it be corrected. The payor is allowed a
                                          second opportunity to submit an acceptable paper document before it is
                                          subject to a failure-to-file or late-filer penalty. An IRS official informed
                                          us that MCC was orally told to apply the paper information return cri-
                                          teria to magnetic media information returns submitted for tax year

                                          %ee appendix I of this report for our assessment of the questionable methodology IRS used to calcu-
                                          late its media error rates.

                                          Page t?                                                       GAO/GGD-90-96     Tax Administration

                            Using the IRS criteria for processing paper returns, we found that over
                            96 percent of all 1988 information returns submitted on magnetic media
                            by all filers were acceptable and successfully processed, According to
                            MCC’S database, 85.4 percent of the over 670 million 1988 information
                            returns sent to MCC on all types of magnetic media were successfully
                            processed the first time through. After the rejected media had been sub-
                            mitted a second time, 96.5 percent of all information returns originally
                            submitted on magnetic media were successfully processed.

Service Bureau Problems     While MCC officials described a number of problems they have encoun-
and Rejection Rates Are     tered with specific service bureaus, they said no problems are unique to
                            service bureaus in general. They said that some of these problems
Similar to Those of Other   include incorrect format and unreadable media. Because MCC’S database
Transmitters                of magnetic media filers does not identify which filers are service
                            bureaus, MCC officials could not provide us with any quantitative anal-
                            ysis concerning information return shipments rejected from service
                            bureaus. However, by matching MCC’S database of filers with IRS Payor
                            Master File, which shows the payors that filed information returns, we
                            were able to estimate the potential number of service bureaus that sub-
                            mitted tax year 1988 information returns on magnetic media and the
                            extent to which these service bureaus caused MCC processing problems.
                            We defined a potential service bureau submitting information returns as
                            any organization that submitted information returns on magnetic media
                            to MCC for payors identified by an EIN different from the submitting

                            We estimate that of 3 1,300 filers of magnetic media information returns
                            for tax year 1988,4,102 (13 percent) were potentially service bureaus
                            while the remaining 27,198 were payors that submitted their own infor-
                            mation returns.4

                            We found that the rejection rates for service bureaus were about the
                            same as the rejection rates for payors submitting for themselves,
                            whether measured in the number of shipments, information returns, or
                            media, as shown in table 2.

                            “The file of transmitters we analyzed consisted of 32,011 transmitters, but we could not determine if
                            711 filers filed for themselves or for others because updated information was not included on our
                            copy of the Payor Master File. Table 1 includes the 711 filers but table 2 does not.

                            Page 7                                                         GAO/GGD-90-96     Tax Administration

Table 2: Comparison of Potential Service
Bureaus’ and Other Transmitters’           Potential service bureaus’
Rejection Rates for Tax Year 1988                                                                Information                    Units of
                                                                  Shiamentsb-      Percent
                                                                                    _.__~_            returns      Percent      mediaC Percent
                                           Total                         15,425        100.0     183.856.261         100.0        23.096        100.0
                                           Rejected   once                1,563         10.1      29,466,490           16.0         2,874           12.4
                                           Rejected twice                   235           1.5       5,974,392           3.2           522            2.3
                                           Other payers/ transmitters
                                           Total                         50,492        100.0     484.021,638         100.0        66.438        100.0
                                           Rejected   once                5,791         11.5      68,330,726          14.1         8,300         12.5
                                           Rejected   twice                 966           1.9      17,101,572          3.5         1,483          2.2
                                           ‘See appendixes II and Ill for stratified data on information returns submitted by service bureaus for
                                           payors and rejected for tax year 1988.

                                           bOne shipment can contain one or more pieces of media (8” disk, 5 i/4” or 3 l/2” floppy disk, or
                                           magnetic tape).
                                           C”Units of media” refers to the combined total number of 8” disks, 5 l/4” and 3 l/2” floppy disks, and
                                           magnetic tape.

                                           The 23 million tax year 1988 information returns from both service
                                           bureaus and payors that had to be returned for correction a second time
                                           is a relatively small proportion of the 670 million information returns
                                           that were submitted on magnetic media. Of these 23 million returns,
                                           over 25 percent (about 6 million) were submitted by potential service
                                           bureaus that filed information returns for 8,757 payors. Of the 4,102
                                           potential service bureaus in our universe, 203 (5 percent) were respon-
                                           sible for these erroneous transmissions. Eight of these potential service
                                           bureaus sent in 13 pieces of magnetic media containing 226,105 informa-
                                           tion returns that were rejected four or more times because the media
                                           were unacceptable.

                                           As noted before, IRS headquarters told MCC to apply paper information
                                            return procedures that could result in penalties for payors whose infor-
                                            mation returns filed on magnetic media are not acceptable on the second
                                           processing attempt. Applying these criteria would suggest that less than
                                            4 percent (6 million) of these information returns would have been sub-
                                           ject to penalties. Since service bureaus are not subject to penalties, the
                                            penalties would have been assessed against as many as 8,757 payors
                                           that were clients of the 203 potential service bureaus that submitted
                                            faulty magnetic media.

                                           Page 8                                                              GAO/GGD-90-95    Tax Administration

                            We found that 3,899 (95 percent) of the 4,102 potential service bureaus
IRS Needs Better Data       submitted acceptable information returns on magnetic media. With this
Before Regulation           level of compliance, it may not be necessary to impose standards or con-
I>ecisions   Can   Be Ma&   trols over service bureaus. However, before this determination can be
                            made, IRSneeds to obtain better data on the nature and extent of any
                            problems it may have with these service bureaus.

                            Although we found that large numbers of information returns were
                            being rejected by IRS as unacceptable for processing, we could not defini-
                            tively determine how many of these returns were filed by service
                            bureaus. Our definition of potential service bureaus included some orga-
                            nizations, such as parent corporations filing for their subsidiaries, that
                            were not service bureaus. However, our estimates were the best we
                            could make in a restricted amount of time given the limitations of IRS’

                            IRSdoes not know how many service bureaus submit information
                            returns on magnetic media and cannot identify the extent of the
                            problems specifically relating to service bureaus. It does not record
                            which shipments of information returns come from service bureaus,
                            although MCC has the capability of identifying the shipments by service
                            bureaus from the accompanying forms currently used to post informa-
                            tion to the database. Having such information readily available would
                            enable IRS to analyze and compare the rejection rates for media from
                            service bureaus with those for media from other filers. These data could
                            also provide the information necessary to justify, if warranted, some
                            type of regulation or control of problem service bureaus continually sub-
                            mitting unacceptable information returns for payors.

                            To identify whether or not a filer is acting as a service bureau, IRS could
                            add a one-digit field to its transmitter control code database. With this
                            information, IRS could compile and sort, by type of filer, information
                            already in the database, such as the number of times media were
                            returned for replacement and the reason they were rejected. Such infor-
                            mation could provide a statistical basis for comparing data on media
                            from service bureaus with data on media from other filers. It would
                            allow IRSto determine if it returns media submitted by service bureaus
                            more often than it returns media submitted by others, and if so, what
                            types of problems service bureaus most often have.

                            IRScould also obtain better data on the frequency and types of magnetic
                            media errors made by filers. The structure of the database fields IRS uses

                            Page 9                                         GAO/GGD-So-96   Tax Administration

                        to record media problems is too limited for detailed quantitative anal-
                        ysis. Each of the three “problem code” fields, which record the reasons
                        for returning media for replacement, allows two entries. Some of the
                        media submitted had as many as five types of errors, but MCC personnel
                        can list only two of them in the database for each time the media are

                        In addition, IRS’ current method of calculating the media return rate is
                        flawed. The Magnetic Media Status Reports for tax year 1988 showed a
                        rejection rate of 15.9 percent for all media received from all transmit-
                        ters. Our calculation of the same information shows a rejection rate of
                        13.9 percent. As appendix I describes in greater detail, the difference in
                        IRS’ and GAO’S calculated rejection rates results from including different
                        numbers in the rates’ denominators.

                        We believe and IRS agreed that making the modifications described
                        above to the existing database would enable IRS to identify magnetic
                        media transmitters that are service bureaus. In addition, IRS would be
                        better able to identify the types of errors filers are making so that
                        appropriate corrective action can be taken.

                        However, before it makes a determination of whether to impose stan-
                        dards and controls over service bureaus, IRS also needs to obtain data on
                        the additional cost such regulation would impose on IRS and service
                        bureaus. IRS should also consider the impact any regulation or control
                        may have on the entry of new small businesses into the service bureau
                        industry. If considered regulatory approaches are so onerous that new
                        firms will be reluctant to compete with current service bureaus or take
                        the place of firms that leave the business, these possibilities have to be
                        weighed before any regulation is adopted.

                        Once IRS obtains data identifying service bureaus and the extent and
Regulation of Service   nature of any problems with them, several options for regulation or con-
Bureaus: Options for    trol can be considered. These options range from keeping things as they
Consideration           are now to extending current IRS controls over service bureau filing of
                        other returns to information returns.

                        Several criteria should be kept in mind in considering the need for regu-
                        lation and the various options. These criteria include (1) the prevalence
                        of problems in service bureau transmissions; (2) the impact of such
                        problems on the tax system; (3) the added burden imposed by regula-
                        tory options on IRSand the service bureau industry, including the impact

                        Page 10                                       GAO/GGD-SO-96   Tax Administration

                        of added requirements on the ability of new firms to enter the market;
                        and (4) which option can best address current problems with the least
                        increase in costs and burden to the system.

The Current Oversight   IRSdoes not regulate filers who provide information returns on magnetic
System                  media. Anyone who has a valid business EIN and requests a transmitter
                        control code number is given one. This transmitter code generally identi-
                        fies the filer and the type of magnetic media submitted. Once a trans-
                        mitter code is obtained, filers do not have to reapply each year. Filers
                        need only request a new number if the old transmitter code has not been
                        used for a year.

                        Under current procedures, a filer is not required to submit test files
                        either before or after IRS gives it a transmitter code. For any filer who
                        wishes to submit test files, IRSwill review them for such things as
                        proper format to identify any processing problems before the filing

                        In addition, current procedures do not provide for withdrawing the
                        transmitter code of those service bureaus or payors that continuously
                        file unacceptable files, Media that contain processing problems, such as
                        coding or format errors, are returned to the filer for correction.

                        Service bureaus submitting information returns on magnetic media are
                        not subject to any penalties. However, the payor employing the service
                        bureau may be subject to failure-to-file or late-filer penalties if the ser-
                        vice bureau continues to submit unacceptable data or is late submitting
                        acceptable data to MCC.

                        The current system essentially relies on the market mechanism to weed
                        out service bureaus continually transmitting media that cannot be
                        processed. Payors penalized by a service bureau’s mistakes can use a
                        different service bureau; payors also have recourse to the legal system
                        to sue a service bureau or take other legal action. However, relying on
                        penalized payors to rid the system of service bureaus who file improp-
                        erly may be of limited effectiveness. First, payors may not be penalized
                        and as a result either may not be aware of problems or may lack incen-
                        tive to correct them. Second, even if there are occasional penalties, the
                        probability of a payor being penalized and dismissing a service bureau
                        may be so low that service bureaus lack sufficient incentive to perform
                        properly. Third, even if penalties are assessed, problem service bureaus
                        may move on to new customers without their reputations catching up to

                        Page 11                                         GAO/GGJHO-95   Tax Administration

                                  them. Fourth, payors may have insufficient information on service
                                  bureau performance to know in advance to avoid those bureaus with
                                  continual problems, so the discipline of the market can be exercised only
                                  after the damage has been done.

Another Option: Extending Current IRS regulations applying to service bureaus submitting other
IRS Regulation of Other    types of tax data, such as employment tax returns, could be extended to
                           service bureaus submitting information returns. Such controls include
Activities to Service      IRS preapproval of formats and revocation of filing privileges for service
Bureaus Filing Information bureaus with repeated errors. Such controls could affect all service
Returns                    bureaus submitting information returns.

                                  IRSallows and encourages service bureaus to file other types of returns
                                  besides information returns but puts strict requirements on those that
                                  do. For example, to participate in the individual income tax electronic
                                  filing program, filers, including service bureaus, must agree to follow
                                  certain requirements and specifications.” All electronic filers who func-
                                  tion as transmitters are required to complete acceptance testing and
                                  pass required suitability checks. Some of the factors to be considered in
                                  a suitability check include a clean criminal record, a history of paying
                                  and filing tax returns correctly and on a timely basis, and a background
                                  of never having been suspended from practice before IRS. IRS reserves
                                  the right to revoke the electronic filing privilege of any electronic filer
                                  who varies from these requirements and specifications or who does not
                                  consistently submit error-free income tax returns.

                                  IRS  also sets requirements and conditions under which service bureaus
                                  can furnish the following returns on magnetic tape instead of on paper:
                                  Form 940, Employer’s Federal Unemployment Tax Return;” Form 941,
                                  Quarterly Federal Tax Return; and Form 941E, Quarterly Return of
                                  Withheld Federal Income Tax.7 IRS requires these service bureaus to
                                  submit an application and a test magnetic tape that must be approved
                                  by an IRS service center before permission to file on magnetic media is
                                  granted. The service center approves, denies, and revokes magnetic tape
                                  filing privileges for service bureaus and other filers. Generally, a filer
                                  must submit magnetic tapes that can be read on IRS’ computers and that
                                  “Revenue Procedure for Electronic Filing of Individual Income Tax Returns (Tax Year 1989) (Oct. 11,
                                  1989) IRS Publication 1345.
                                  “Revenue Procedures for Magnetic Tape Reporting of Form 940 (Aug. 8, 1988) IRS mtblication 1314.
                                  7Revenue Procedures for Magnetic Tape Reporting of Form 941, Quarterly Federal Tax Return, and
                                  941E, Quarterly Return of Withheld Federal Income Tax (Aug. 8, 1988) IRS Publication 1264.

                                  Page 12                                                      GAO/ND-90-95     Tax Administration

                       have an error rate on the returns of 5 percent or less in order to have its
                       application for filing and its continuation of filing privileges approved.

                       Extending this kind of regulatory regime to service bureaus filing infor-
                       mation returns may serve to weed out in advance those firms whose
                       transmissions are likely to be the most troublesome. However, several
                       issues should be considered in deciding whether to extend these stan-
                       dards. First, such controls would apply to all service bureaus, regardless
                       of their track record. If subsequent IRSdata prove that 95 percent of
                       service bureaus cause no problems, such a broad regulatory scope may
                       not be cost effective. A second concern is that standards can be set so
                       high that effective competition may be hindered. For example, if stan-
                       dards specify certain media while disallowing other media currently
                       used, some smaller service bureaus could be significantly affected.
                       Third, prior approval and testing of service bureau magnetic media sub-
                       missions could impose significant burdens and costs on both service
                       bureaus and IRS Finally, if revoking filing privileges is considered too
                       severe a response to careless submissions, the regulatory standards
                       might fall into disuse by IRS

Other, More Targeted   Other options could be considered that would target sanctions to those
Options                service bureaus submitting magnetic media information returns with
                       repeated errors in data transmission. These options would include (1)
                       applying penalties to service bureaus, (2) revoking filing privileges for
                       those with repeated problems, and (3) providing information to payors
                       on service bureaus with repeated data filing errors. Unlike the more
                       extensive regulations applied to service bureaus filing other returns,
                       these other options would only be applicable to that small percentage of
                       service bureaus failing to correct errors after reasonable notice and
                       feedback from IRS.

                       As far as penalties go, current IRS regulations focus responsibility on the
                       payor, not the service bureau, for submitting correct information
                       returns on magnetic media. However, payors using service bureaus as
                       filers may not know that the magnetic media submitted are unaccept-
                       able until they receive a notice of a failure-to-file or late-filer penalty.
                       Payors employing problem service bureaus can terminate the relation-
                       ship once they discover the problems, allowing the free market system
                       to weed out service bureaus continually submitting unacceptable mag-
                       netic media.

                       Page 13                                        GAO/ND-90-95   Tax Administration

              However, if IRSwere to have the legal authority to assess penalties
              against service bureaus- something that would require legislation-it
              would not have to rely only on the free market system. If these penalties
              were reasonably certain and of sufficient magnitude, they could provide
              the incentive for proper filing by service bureaus. They could also cause
              enough financial burden that low-quality firms would be forced out of

              Another option IRScould explore would enable it to revoke the filing
              privileges of service bureaus that repeatedly make filing errors. Unlike
              IRS’ current oversight of filers of Form 940 and electronic income tax
              forms, this option would enable IRS to revoke filing privileges without
              first having gone through the process of approving those privileges. Of
              course, IRS would have to develop criteria governing the circumstances
              under which it would use its revocation measures. It would also have to
              be mindful of the pitfalls of revocation that we discussed earlier, such as
              the cost and burden imposed on both service bureaus and IRS.

              A final option that IRS might want to consider is making information
              available to the payor community on service bureaus that have repeated
              filing problems. If public disclosure requirements permit, IRS could pub-
              lish lists of problem service bureaus or tell payors that request informa-
              tion which service bureaus have had a history of problems. Payors
              would then not be as likely to first find out about faulty service bureau
              submissions when they receive penalty notices from IRS. They might also
              then be able to see if paying more for the work of particular service
              bureaus is worth it if they receive higher-quality submissions. Making
              more information available to the consumers of service bureau services
              upfront would make the current market relationship between payor and
              service bureaus work more efficiently.

              Service bureaus generally appear to be submitting data that IRS can pro-
Conclusions   cess. Moreover, their error rates are similar to those of the general
              payor community submitting information returns on magnetic media.
              Nevertheless, although filing errors seem to be confined to about 5 per-
              cent of the potential service bureaus in our universe, they involve
              processing delays affecting nearly 6 million information returns and
              could result in penalties to over 8,700 payors.

              Because the percentage of service bureau transmissions with repeated
              errors is so small, we believe that a broader regulatory scheme applying
              to all service bureaus may not be appropriate at this time. However, the

              Page 14                                       GAO/GGJHO-96   Tax Administration
                      lack of more definitive information on service bureau filings and the
                      large numbers of returns and payors affected by those relatively few
                      filings with repeated errors suggest that IRS needs more information on
                      the scope and extent of the problem.

                      Specifically, IRS needs to modify its records to identify which filers are
                      service bureaus, better pinpoint the types of errors involved, and
                      develop a more accurate way of calculating the error rates of various
                      types of filers. IRS needs information on the costs to IRS and service
                      bureaus of imposing additional regulatory standards and controls,
                      including the potential impacts of regulation on the entry of additional
                      businesses into the industry.

                      With this information, IRS and Congress will have a better basis to
                      decide whether regulation is warranted. At this point, a number of regu-
                      latory tools could be considered. Current IRS regulations applying to ser-
                      vice bureaus submitting other types of tax data, such as employment
                      tax returns, could be extended to service bureaus submitting informa-
                      tion returns. Such controls include procedures for revoking transmission
                      privileges for service bureaus with repeated errors and IRS preapproval
                      of formats. Such controls would affect all service bureaus submitting
                      information return data. Although promising to further reduce the error
                      rate, preapproval requirements could increase the burden on both the
                      service bureaus and IRS.

                      Other options to be considered would target sanctions at those service
                      bureaus with repeated errors in data filing. These options would include
                      applying penalties to service bureaus, authorizing IRS revocation of filing
                      privileges for those with repeated problems, and providing information
                      to payors on service bureaus with repeated data filing errors. Unlike
                      prior approval requirements, these options would be applicable only to
                      that small percentage of service bureaus failing to correct errors after
                      reasonable notice and feedback from IRS. Although further study is
                      needed, these options could offer IRS and the payor community protec-
                      tion from problem service bureaus at less overall cost and burden for
                      everyone involved.

                      We recommend that the Commissioner of Internal Revenue collect infor-
Recommendations       mation on problems caused by service bureaus that submit information
         Y            returns. Specifically, IRSshould

                  l   modify its records to identify which filers are service bureaus,

                      Page 15                                        GAO/GGJMO-96   Tax Administration
. modify its records to more accurately identify magnetic media
  processing errors,
l modify its method of calculating the magnetic media rejection rate,
l obtain data on the cost and benefits to IRS and service bureaus of insti-
  tuting standards and controls, and
. obtain data on the impact any regulation or control may have on the
  entry of new small businesses into the service bureau industry.

    This information should be used by the Commissioner to decide what
    type of regulation or control, if any, is warranted.

    The contents of this report were discussed with IRS officials. They gener-
    ally agreed with the facts presented and with our recommendations, and
    their comments were considered in preparing our final report.

    Major contributors to this report are listed in appendix IV. If you have
    any questions regarding this report, please call me on 272-7904.

    Paul L. Posner
    Associate Director, Tax Policy
      and Administration  Issues

    Page 16                                      GAO/GGDSO-96   Tax Administration
Page 17   GAO/GGD-90-95   Tax Administration

Appendix I
GAO Recalculation of
IRS’ Overstated Media
Error Rates
Appendix II
1988 Service Bureau
Information Returns
Submitted on Magnetic
Appendix III
1988 Service Bureau
Information Returns
Submitted on Magnetic
Media and Rejected a
Appendix IV                                                                                            23
Major Contributors to   General Government Division, Washington, D.C.
                        San Francisco Regional Office
This Report
Tables                  Table 1: Magnetic Media and Information Returns                                 6
                            Received by MCC and Returned to Transmitters
                        Table 2: Comparison of Potential Service Bureaus’ and                           8
                            Other Transmitters’ Rejection Rates for Tax Year


                                  employer identification number

                        IRS       Internal Revenue Service
                        MCC       Martinsburg Computing Center

                        Page 18                                    GAO/GGDsO-95   Tax Administration
Page 18   GAO/GGD90-95   Tax Administration

GAO Recalculation of IRS’ Overstated Media

               In its Magnetic Media Status Report, IRS’ calculation of the percentage of
               media submissions returned overstates the rejection rate. IRS calculates
               its rejection rate by dividing “media returned for replacement” by
               “returns received.” This calculation overstates the rejection rate
               because IRS includes the number of times each piece of media was
               rejected in its calculation of the number returned for replacement
               (numerator) but excludes the receipt of replacements from its calcula-
               tion of the total number of information returns received (denominator).
               The following formula represents the IRS calculation:

                 sum (number returned X number of times returned)
                                                                      x 100
               sum (number received, excluding replacements received)

               IRSwould have provided a more accurate representation of the per-
               centage of media submissions returned by also weighting the total
               number of media received by the number of times each piece was
               received (denominator) as shown below:

               sum (number returned X number of times returned) x IO0
               sum (number received X number of times received)

               Applying each of these methods to the data on the Magnetic Media
               Status Report for 1988 information returns produces the following

               Returned for Replacement       g       x 100 = 15.9%
               Returns Received                   ,

               GA0 calculation
               Returned for Replacement            14,420
                                                              x 100 = 13.9%
               Returns Received plus          90,769 + 12,974
               Replacements Received

               Page 20                                       GAO/GGD-SO-96   Tax Administration


Appendix II

1988 Service Bureau Information Returns
Submitted on Magnetic Media

               Number of        Number of                          Number of              Number of
               payors per          service   Number of   information returns    information returns
               service bureau     bureaus      Davors              submitted      reiected first time
               l-50                  3,820      21,865            77,343,784                t&884,193
               51-100                  128       9,096            29,098,660                6,095,977
               101-150                  65       8,038            23,619,319                1,673,782
               151-200                  26       4,481            17,039.029                7.998.322
               201-250                  12       2,648             1,218,142                  488,914
               251-300                   9       2,467             2,764,992                  147,961
               301-350                   8       2,599             2,851,092                  636,914
               351-400                   4       1,513             1,203.078                  263,910
               401-450                  11       4,668             8,590,230                  922,462
               451-500      -            6       2,896             7,064,072                  760,287
               501-1.000                 6       4.155             7.293.056                1.356.349
               1,001.2K                  2       2,835               303,808                        0
               2,001-3K                  2       5,399               156,302                  113,810
               3,001-4K                  1       3,985               496,128                   30,319
               7,001-8K                  1       7,658             2,147,440                   62,998
               OverlOK                   1      10,823                73,015                   71,864
               Total                4,102       95,126          163,262,147               29,506,062

               Page 21                                          GAO/GGD-So-95    Tax Administration
Appendix   III                                                                             !?J

1988 Service Bureau Information Returns
Submitted onMagnetic Media and Rejecteda
                                        Number of                 Number of information
                 Number of payors per      service   Number of         returns rejected
                 service bureau           bureaus       payors             second time
                 l-50                          168       1.426                   2.215.224
                 51-100                         12         901                   2,431,329
                 101-150                         3         366                     150,985
                 151-200                         4         651                     353,103
                 201-250                         4         895                     185.051
                 251-300                         2         576                       4,749
                 301-350                         3         990                      77,051
                 351-400                         2         721                      51,542
                 401-450                        3         1,284                    445,686
                 451-500                        2           947                     57,063
                 Total                        203        8,757                  5,974,39?

                 Page 22                               GAO/GGDW-96    Tm Administraticin
 p&ndix    IV

&jor Contributors to This Report

                        Natwar M. Gandhi, Assistant Director, Tax Policy and Administration
General Government        Issues
Division, Washington,   Lawrence M. Korb, Assignment Manager
                        Ralph T. Block, Regional Management Representative
San Francisco           George A. Zika, Evaluator-in-Charge
Regional Office         Daniel F. Alspaugh, Evaluator
                        Terry Shenks, Programmer Analyst

(268443)                Page 23                                     GAO/GGD-90-96   Tax Administration
  First-<hss  Mail
Postage & Pws Paid
  Permit, No. G 100