Tax & Accounting Software Corp. v. United States

                                                                       F I L E D
                                                                 United States Court of Appeals
                                                                         Tenth Circuit

                                                                        AUG 30 2002
                                    PUBLISH

                   UNITED STATES COURT OF APPEALS                    PATRICK FISHER
                                                                             Clerk
                                 TENTH CIRCUIT



 TAX AND ACCOUNTING
 SOFTWARE CORPORATION; TIM
 E. KLOEHR; and SHERYL KLOEHR,

       Plaintiffs - Appellees,

 v.

 UNITED STATES OF AMERICA,
                                                       No. 00-5196
       Defendant - Appellant.




 COLORADO SOFTWARE AND
 INTERNET ASSOCIATION;
 SOFTWARE FINANCE AND TAX
 EXECUTIVES COUNCIL,

        Amici Curiae.



                 Appeal from the United States District Court
                   for the Northern District of Oklahoma
                            (D.C. No. 98-CV-363)


Curtis C. Pett (Richard Farber with him on the briefs), Attorneys, Tax Division,
Department of Justice, Washington, D.C., for the Defendant-Appellant.

Rebecca M. Fowler (Richard H. Foster with her on the brief), Tulsa, Oklahoma,
for the Plaintiffs-Appellees.
Mark Reinhardt, Denver, Colorado, filed an amicus brief on behalf of Colorado
Software and Internet Association.

Mark E. Nebergall, Washington, D.C., filed an amicus brief on behalf of Software
Finance and Tax Executives Council.


Before LUCERO and McWILLIAMS, Circuit Judges, and STAGG, * District
Judge.


LUCERO, Circuit Judge.



      Plaintiff taxpayers filed a refund suit seeking money allegedly owed them

from a tax credit for research and development expenses under I.R.C. § 41. In a

matter of first impression in this circuit, we interpret the scope of “qualified

research” under I.R.C. § 41, including the requirement in § 41(d)(1) that the

taxpayer must intend to “discover[] information” using a “process of

experimentation.” The district court granted summary judgment to the taxpayers

and the government appealed. Our review of this case was abated from January

18, 2002, until May 24, 2002, pursuant to the government’s request. We have

jurisdiction under 28 U.S.C. § 1291, and we reverse and remand.




      *
        The Honorable Tom Stagg, District Judge for the Western District of
Louisiana, sitting by designation.

                                         -2-
                                            I

        Plaintiffs—Tim Kloehr, his wholly owned Subchapter S corporation, Tax

and Accounting Software Corporation (“TAASC”), and Mr. Kloehr’s wife Sheryl

Kloehr—filed suit seeking a refund of taxes paid by Mr. Kloehr for his 1993 tax

year and by him and his wife for 1994. 1 TAASC is an Oklahoma corporation that

develops and sells software for use by tax and accounting professionals. In 1993

and 1994, TAASC developed four computer software products for sale to its

customers: EasyACCT, EasyMICR, Professional Tax System, and EasyTEL. The

research and development expenses for these four products are at issue in this

suit.

        EasyACCT is an integrated accounting program that collects data for the

preparation of financial statements and allows for the transfer of this information

to TAASC’s tax-preparation software. The parties agree that at the time of its

introduction to the public EasyACCT was unique in the functions that it provided.

        EasyMICR is a software program designed to print magnetic-ink-character

banking transit codes on blank check stock. It was developed in both DOS and

Windows versions. A commercial failure, EasyMICR was subsequently integrated

into EasyACCT.



        1
            The plaintiffs will be collectively referred to as TAASC throughout this
opinion.

                                           -3-
      Professional Tax System integrates a number of tax-preparation software

programs into a single, seamless package and allows for the preparation of tax

returns for both state and federal tax forms using the same data. It was the first

commercial software program to allow for electronic filing with state and federal

governments. The program was also designed to run with minimal memory and

therefore is able to run on a wide range of computers.

      EasyTEL is an automated, multi-task call-processing system that allows

businesses to answer and transfer calls, take messages, provide information over

the phone, convert faxes to e-mails, and distribute faxes automatically. EasyTEL

is unusual because it was designed to run on low-cost computer hardware and

requires little maintenance.

      In 1993, TAASC incurred a total of $1,838,756 in research and

development expenses for the development of these software products; in 1994, it

incurred a total of $2,444,938 in expenses. TAASC claimed a portion of these

expenses as research and development tax credits under I.R.C. § 41, but the

Internal Revenue Service (“IRS”) disallowed those tax credits. As a result, Mr.

Kloehr faced tax deficiencies of $123,764 in 1993 and $192,510 in 1994. Mr.

Kloehr paid the deficiencies and brought a refund suit on May 14, 1998 under

I.R.C. § 7422(a) and 28 U.S.C. § 1346(a)(1). On the parties’ cross-motions for

summary judgment, the district court granted summary judgment for TAASC.


                                         -4-
      We review the district court’s grant of summary judgment de novo. Simms

v. Oklahoma ex rel. Dep’t of Mental Health & Substance Abuse Servs., 165 F.3d

1321, 1326 (10th Cir. 1999). Summary judgment is appropriate “if the pleadings,

depositions, answers to interrogatories, and admissions on file, together with the

affidavits, if any, show that there is no genuine issue as to any material fact and

that the moving party is entitled to a judgment as a matter of law.” Fed. R. Civ.

P. 56(c). When applying this standard, we view the evidence and draw reasonable

inferences in the light most favorable to the non-moving party. Simms, 165 F.3d

at 1326. If there is no genuine issue of material fact in dispute, we determine

whether the district court correctly applied the substantive law. Id.

                                          II

      Central to the parties’ dispute in this case is what constitutes “qualified

research,” which entitles a business to a tax credit under I.R.C. § 41. We begin

our analysis with an overview of the provisions of I.R.C. § 41(d)(1), which

defines “qualified research,” and a brief history of the § 41 tax credit.

      We initially note that the provision in question, I.R.C. § 41, is separate

from the tax code provision that provides for a deduction for any “research or

experimental expenditures” made by a business. I.R.C. § 174(a)(1). The latter

expenditures have been defined by regulation as expenditures for “activities

intended to discover information that would eliminate uncertainty concerning the


                                         -5-
development or improvement of a product.” Treas. Reg. § 1.174-2. Our

interpretation of the proper scope of the tax credit under § 41 has no bearing on

the scope or applicability of the deduction under § 174.

                                            A

       Section 41(d)(1) lays out five separate requirements for “qualified

research”: 2

       (1) The research must qualify as expenses under I.R.C. § 174, the tax code

provision providing for deductions of research expenses. I.R.C. § 41(d)(1)(A).

       (2) The research must be “undertaken for the purpose of discovering

information.” § 41(d)(1)(B).



       2
           The full text of I.R.C. § 41(d)(1) is:

       (d) Qualified research defined.—For purposes of this section—
             (1) In general.—The term “qualified research” means
             research—
                    (A) with respect to which expenditures may be treated as
                    expenses under section 174,
                    (B) which is undertaken for the purpose of discovering
                    information—
                          (i) which is technological in nature, and
                          (ii) the application of which is intended to be
                          useful in the development of a new or improved
                          business component of the taxpayer, and
                    (C) substantially all of the activities of which constitute
                    elements of a process of experimentation for a purpose
                    described in paragraph (3).
             Such term does not include any activity described in paragraph
             (4).

                                            -6-
      (3) The information discovered must be “technological in nature.”

§ 41(d)(1)(B)(i).

      (4) The “application” of the information described above must be

“intended to be useful in the development of a new or improved business

component of the taxpayer.” § 41(d)(1)(B)(ii).

      (5) “Substantially all” of the research must “constitute elements of a

process of experimentation” for a valid purpose under the tax credit. 3

§ 41(d)(1)(C).

      The government concedes that TAASC has met the first and fourth

requirements. The parties also do not dispute that the research undertaken by

TAASC was “technological in nature” as required by I.R.C. § 41(d)(1)(B)(i).

Thus, we consider only the proper interpretation of requirements two (the

“discovering information” requirement) and five (the “process of

experimentation” requirement).

                                          B

      The tax credit embodied in § 41 was originally codified as § 44F of the

Internal Revenue Code and was enacted as part of the Economic Recovery Tax

Act of 1981, Pub. L. No. 97-34, § 221, 95 Stat. 172, 241–47. Like § 41, § 44F


      3
          The research that qualifies for the § 41 tax credit may only be for certain
purposes, listed in I.R.C. § 41(d)(3). The validity of TAASC’s research purposes
is not at issue in this case.

                                         -7-
allowed a credit for “qualified research,” but unlike § 41, § 44F provided no

definition of “qualified research.” Instead, the statute simply referred to

“qualified research” for purposes of the existing tax deduction under § 174. See

Pub. L. No. 97-34, § 221, 95 Stat. 172, 242. In 1981, “qualified research” for

§ 174 had no statutory definition and was defined by Treasury Regulations as

“research and development costs in the experimental or laboratory sense.” H.R.

Rep. No. 97-201, at 110 (1981) (citing Treas. Reg. § 1.174-2(a) (1960) (amended

1994)).

      In 1986, Congress reenacted the research tax credit with significant

changes, creating the current § 41. The provisions in I.R.C. § 41(d)(1) defining

“qualified research” were added, including the “discovering information”

requirement and the “process of experimentation” requirement.

      Throughout the late 1980s and 1990s, the research tax credit was reenacted

without any substantive changes to the relevant provisions. See Credit for

Increasing Research Activities, 63 Fed. Reg. 66,503, 66,504 (Dec. 2, 1998).

      In 1998, Congress reenacted the § 41 tax credit, and the relevant

Conference Report included language regarding the definition of “qualified

research” in § 41(d)(1). See H.R. Conf. Rep. No. 105-825, at 1548 (1998). In

1999, another reenactment of the tax credit prompted more Conference Report




                                         -8-
language regarding the meaning of § 41(d)(1). See H.R. Conf. Rep. No. 106-478,

at 132 (1999).

      Whether the legislative history is relevant to the disposition of this case, of

course, depends on whether the meaning of each statutory requirement is plain.

See United States v. Simmonds, 111 F.3d 737, 742 (10th Cir. 1997) (“[I]f the

statutory language is ambiguous, a court can then resort to legislative history as

an aid to interpretation.”); Edwards v. Valdez, 789 F.2d 1477, 1481-82 (10th Cir.

1986) (“When the meaning of the statute is clear, it is both unnecessary and

improper to resort to legislative history to divine congressional intent.”). We will

later discuss whether the meaning of each contested statutory requirement

(“discovering information” and “process of experimentation”) is sufficiently plain

to preclude consideration of the legislative history.

                                          C

      TAASC’s interpretation of the “discovering information” test is that it only

requires a taxpayer to “discover[] technological information that led it to develop

innovative . . . products.” (Appellees’ Br. at 8.) Alternatively, TAASC argues

that the “discovering information” test in § 41 should have the same meaning as it

has in the definition of research for purposes of § 174 of the tax code. 4 TAASC


      4
        There is no definition of research in the text of § 174. The current
Treasury regulations state that an expenditure represents a research cost if it is
                                                                      (continued...)

                                         -9-
also implies that because its products were new and innovative compared with the

other commercially available products in the field of tax and accounting software,

its research and development expenses for those products therefore must qualify

for the tax credit.

       The government argues that the 1986 amendments, including the addition

of the “discovering information” requirement, were intended to narrow the scope

of the tax credit significantly. As a result, to meet the “discovering information”

requirement, qualified research must “rely upon, and expand or refine, principles

of the physical or biological sciences, engineering or computer science.”

(Appellant’s Br. at 13.) According to the government, TAASC’s research did not

“expand or refine” principles of computer science, a point that TAASC conceded

before the district court. (1 Appellant’s App. at 80–81.)

       The district court, however, agreed with TAASC that § 41(d)(1) does not

require the taxpayer to expand or refine principles of science or engineering in

order to qualify for the tax credit. It added that the “emphasis should be on

whether the information qualifies as being ‘technological in nature’ . . . , not

whether the work could be considered a revolutionary discovery in the scientific

sense.” Tax & Accounting Software Corp. v. United States, 111 F. Supp. 2d



       (...continued)
       4

“intended to discover information that would eliminate uncertainty concerning the
development or improvement of a product.” Treas. Reg. § 1.174-2(a).

                                         -10-
1153, 1158 (N.D. Okla. 2000). Indeed, the court reasoned that the allegation that

there is a separate requirement that the taxpayer “discover” information is

erroneous and a “strained and improper reading without any support in the

legislative history to back it up.” Id. It concluded that as a matter of law TAASC

had met the requirements of “discovering information” for purposes of § 41(d)(1),

id. at 1160, relying in part on the fact that TAASC’s products were “new and

more efficient combination[s] of software that [were] not available to the public.”

Id.

      The parties also disagree whether TAASC’s software development process

qualifies as a “process of experimentation.” TAASC argues that by trying various

programming methods in order to achieve its desired results, it undertook a

process of experimentation. The government responds that because TAASC used

generally known computer programming skills to achieve its final objective, there

was no “experimentation,” even if TAASC was unsure which particular technique

would allow it to achieve its goal.

      The government also contends that TAASC did not use a “process of

experimentation” because TAASC believed at the outset that the final results of

its research—the development of the software—were certain to be technically

feasible. According to the government, the “process of experimentation”

requirement requires the taxpayer to be uncertain as to whether the final result


                                        -11-
can be feasibly achieved. The district court concluded that TAASC’s research

satisfied the “process of experimentation” requirement. Tax & Accounting

Software, 111 F. Supp. 2d at 1160–61.

                                         III

      We first analyze what Congress intended when it required that taxpayers

“discover” information in order to qualify for the tax credit. In doing so, we

apply traditional tools of statutory construction and interpretation, starting with an

overview of the applicable principles.

                                          A

      At the time that this case began, the government’s interpretation of

§ 41(d)(1) had not been incorporated into a final regulation that had gone through

the notice-and-comment rulemaking process. 5 Under Christensen v. Harris

County, 529 U.S. 576 (2000), this means that the government’s position is only

due respect to the extent that it has the “‘power to persuade,’” id. at 587 (quoting

Skidmore v. Swift & Co., 323 U.S. 134, 140 (1944)), instead of the much greater



      5
         The government has issued a new set of proposed regulations interpreting
§ 41(d)(1), superseding regulations that became final in January 2001, but none of
these regulations are at issue in this case. See Credit for Increasing Research
Activities, 66 Fed. Reg. 66,362 (Dec. 26, 2001) (proposed regulations); Credit for
Increasing Research Activities, 66 Fed. Reg. 280 (Jan. 3, 2001) (final
regulations). Although the regulations are not at issue in this case, we
nonetheless consider them as part of our analysis of the respect due the
government’s position under Skidmore v. Swift & Co., 323 U.S. 134, 140 (1944).

                                         -12-
deference due to regulations promulgated through the Administrative Procedure

Act’s notice-and-comment rulemaking process, see Chevron, U.S.A., Inc. v.

Natural Res. Def. Council, 467 U.S. 837 (1984). 6

       The respect that is owed to an agency position under Christensen and

Skidmore is based on the “body of experience and informed judgment” that

agencies have in the interpretation, enforcement, and implementation of the

statutes for which they are responsible. Christensen, 529 U.S. at 587; Skidmore,

323 U.S. at 140. The respect owed to the agency interpretation “will depend upon

the thoroughness evident in its consideration, the validity of its reasoning, [and]

its consistency with earlier and later pronouncements.” Skidmore, 323 U.S. at

140.

       The government’s position regarding the “discovering information”

requirement has been consistent in § 41(d)(1) litigation, see, e.g., Wicor, Inc. v.

United States, 263 F.3d 659, 660 (7th Cir. 2001); United Stationers v. United

States, 163 F.3d 440, 444–46 (7th Cir. 1999); Wicor, Inc. v. United States, 116 F.

Supp. 2d 1028, 1034–36 (E.D. Wis. 2000); United Stationers, Inc. v. United

States, 982 F. Supp. 1279, 1283–86 (N.D. Ill. 1997); Norwest Corp. v. Comm’r,


       6
         The district court applied a Chevron analysis in this case. Tax &
Accounting Software, 111 F. Supp. 2d at 1157. Although its analysis was not
explicit, the district court concluded that the language of § 41(d)(1) was plain and
contradicted the government’s position and that therefore the government’s
position was foreclosed. Id. at 1158–60.

                                        -13-
110 T.C. 454, 492–94 (T.C. 1998), but it has changed significantly as the

government has developed regulations to implement § 41(d)(1). Compare Credit

for Increasing Research Activities, 66 Fed. Reg. 280, 282–84 (Jan. 3, 2001)

(stating that the “discovery requirement” can only be met when information is

discovered that “exceeds, expands, or refines the common knowledge of skilled

professionals in a particular field of science or engineering”), and Credit for

Increasing Research Activities, 63 Fed. Reg. at 66,504–05 (same), with Credit for

Increasing Research Activities, 66 Fed. Reg. 66,362, 66,363 (Dec. 26, 2001)

(stating that the new proposed regulations “eliminate[] . . . the requirement that

qualified research must be undertaken to obtain knowledge that exceeds, expands,

or refines the common knowledge of skilled professionals in a particular field of

science or engineering”). Under Skidmore and Christensen, we conclude that the

government’s position is not entitled to substantial deference.

      An additional factor to consider is that in construing the Internal Revenue

Code, courts are expected to show “respect” and “deference” to the IRS’s

interpretation of the Code. See Jewett v. Comm’r, 455 U.S. 305, 318 (1982);

Callaway v. Comm’r, 231 F.3d 106, 131 (2d Cir. 2000); Dewees v. Comm’r, 870

F.2d 21, 31 (1st Cir. 1989). However, given the IRS’s changing position on this

issue, we believe that deference is inappropriate here.

      Finally, we note that tax credits are “a matter of legislative grace, and


                                         -14-
taxpayers bear the burden of clearly showing that they are entitled to them.”

Schumacher v. United States, 931 F.2d 650, 652 (10th Cir. 1991) (citing New

Colonial Ice Co. v. Helvering, 292 U.S. 435, 440 (1934)). Thus, they are to be

strictly and narrowly construed. See Texasgulf, Inc. v. Comm’r, 172 F.3d 209,

214 (2d Cir. 1999); Inland Steel Co. v. United States, 677 F.2d 72, 79 (Ct. Cl.

1982). Where there is clear provision in the statute for a particular credit,

however, the credit is allowable. See New Colonial Ice, 292 U.S. at 440.

                                          B

      In interpreting a statute we begin with its plain language. United

States v. Morgan, 922 F.2d 1495, 1496 (10th Cir. 1991). “In ascertaining the

plain meaning of the statute, the court must look to the particular statutory

language at issue, as well as the language and design of the statute as a whole.”

K Mart Corp. v. Cartier, Inc., 486 U.S. 281, 291 (1988) (emphasis added).

      The word “information” has a broad meaning. See Webster’s Third New

International Dictionary Unabridged 1160 (1993) (defining “information” as, for

example, “knowledge communicated by others or obtained from investigation,

study, or instruction,” or “knowledge of a particular event or situation”). Among

other definitions, “discover” can mean “to make known (something secret,

hidden, unknown, or previously unnoticed),” “to reveal the identity of,” or “to

disclose to view (something hidden, covered, or previously unseen).” Id. at 647;


                                         -15-
see also id. (stating that the word discover “means to come to know something not

previously known, either by purposive search and investigation or by accident”).

In the present statutory context, i.e., as one of the requirements for “research” that

the government would have an interest in encouraging through a tax credit, we

conclude that the word “discover” requires that what is “discovered”—in this

case, information—be something new or previously unknown. 7

      Moreover, the new information that is discovered must also be information

that is “intended to be useful in the development of a new or improved business

component of the taxpayer.” § 41(d)(1)(B)(ii). The new information

therefore cannot merely be the product itself but must also have independent

value that can be applied in the development of a new product. Cf. Wicor, 263

F.3d at 660 (distinguishing, for purposes of § 41, “the domain of genuine research

from that of implementation of existing research findings”).

      In short, the statute requires that the taxpayer’s research, in order to

qualify, must have developed new information that is applied towards the



      7
         This information must not only be new to the taxpayer, but also be
information that is generally unknown to the public as well. We cannot conclude
that Congress intended to subsidize through tax credits research that might instead
have been achieved through a trip to the library or a review of the relevant
technical and professional literature. TAASC does not appear to dispute this
point. It argues that its programs qualify under the discovery test because
“TAASC was the first to make the discovery” that the programs could be
developed. (Appellee’s Br. at 30.)

                                         -16-
development of a product. This is the “discovery” requirement of § 41. 8 Contrary

to TAASC’s arguments, mere evidence that the taxpayer has developed a new and

useful product in and of itself will not qualify. Contrary to the government’s

argument, the new information need not “expand, or refine, principles of the

physical or biological sciences, engineering, or computer science.” (Appellant’s

Br. at 13.) In other words, each of the positions of the parties is incorrect.

      More importantly, for our purposes, the district court erred when it held

that any “discovery” test was based on a “strained and improper reading without

any support,” Tax & Accounting Software Corp., 111 F. Supp. 2d at 1158, a

holding that essentially read the “discovering information” language out of the

statute. The district court also erred when it rejected the argument that the

“discovering information” language requires “newness and expansion of existing

knowledge.” Id. at 1159. The term “discovery” means that the researcher must


      8
         We reject TAASC’s argument that the § 41 discovery requirement should
be the same as the one for § 174. We find it difficult to understand how
Congress, when it drafted the “discovering information” requirement of § 41 in
1986, intended to adopt a definition of “discovering information” that was only
developed in the Treasury Regulations after § 41 was enacted. See Credit for
Increasing Research Activities, 66 Fed. Reg. at 66,363; Research or Experimental
Expenditures, 58 Fed. Reg. 15,819, 15,820 (Mar. 24, 1993) (proposing the
definition of “discovering information” for § 174). TAASC’s argument that the
“requirement for ‘discovery’ was established in IRC § 174” and that therefore, by
adopting the § 174 definition of research, the original § 44F tax credit also
adopted any definition of “discovering information” in § 174, is historically
inaccurate. There was no “discovery” requirement in the § 174 regulations until
after 1981. See H.R. Rep. No. 97-201, at 110.

                                         -17-
find new information or, in other words, must expand existing knowledge.

Finally, the district court erred when it relied on the mere newness and innovation

of the product itself, without determining that TAASC discovered new

information separate from that product, in concluding that TAASC met the

requirements of § 41. Id. at 1160.

      To summarize, the “discovering information” language of § 41 establishes a

separate requirement that the taxpayer must meet in order to qualify for the tax

credit. Under that requirement, the taxpayer must show that he discovered new

information and that information must be separate from the product that is

actually developed. 9

                                          IV

      The next issue is whether TAASC’s work amounted to a “process of

experimentation.” We analyze both (1) whether the “process of experimentation”

test allows a taxpayer to use methods that are generally known, and (2) whether

the test requires the taxpayer initially to believe that there is uncertainty as to

whether the final result is feasible.




      9
        Because we conclude that the statute’s plain meaning addresses the
question, we need not reach the parties’ legislative history arguments.

                                          -18-
                                          A

       In construing the “process of experimentation” requirement, we again start

with the language of the statute. “Process” has a broad meaning that supports

TAASC’s interpretation. See Webster’s Third New International Dictionary

Unabridged at 1808 (defining process in this context as “a progressive forward

movement from one point to another on the way to completion,” “the action of

passing through continuing development from a beginning to a contemplated

end,” “the action of continuously going along through each of a succession of

acts, events or developmental stages,” or “a particular method or system of doing

something, producing something, or accomplishing a specific result”).

       The definition of “experiment” is “a test or trial,” “a tentative procedure or

policy,” or “an act or operation carried out under conditions determined by the

experimenter (as in a laboratory) in order to discover some unknown principle or

effect or to test, establish, or illustrate some suggested or known truth.” Id. at

800.

       Here, the definition of “experiment” is ambiguous with respect to the

questions at issue. Concerning the first issue, none of the definitions clarifies

whether the methods may be generally known. Regarding the second issue,

whether the taxpayer must initially believe that there is uncertainty as to whether

the final result is feasible, the first two definitions would best fit with TAASC’s


                                         -19-
broad reading of the term. The last definition could support either the

government’s argument that the final results of the experimental process must be

uncertain as to feasibility—the final result being an “unknown principle or

effect”—or TAASC’s position that the final results may be certain as to

feasibility—the final result may also be a “known truth.” Given that the “process

of experimentation” must be used to discover new information (as we discussed

above), however, the government’s position regarding this last definition is

stronger. New information cannot be a “known truth.”

      In light of the conflicting definitions, we conclude that the term “process of

experimentation” is ambiguous as to the points at issue: namely, whether the

final results must be certain as to feasibility and whether the methods to be used

in the “process of experimentation” may be generally known. Because of this

ambiguity, legislative history may properly be invoked in construing the “process

of experimentation” requirement. See Simmonds, 111 F.3d at 742 (1997)

(discussing the circumstances in which resort to legislative history is appropriate).

                                          B

      By all indications, the government has consistently argued in § 41(d)(1)

litigation that there must be uncertainty as to whether the final result can be

achieved for the “process of experimentation” test to be satisfied. See, e.g.,

United Stationers, 163 F.3d at 445–46; Wicor, 116 F. Supp. 2d at 1035–36;


                                         -20-
United Stationers, 982 F. Supp. at 1284–86; Norwest, 110 T.C. at 496. However,

the government has apparently not raised the issue of whether generally known

techniques may be used in the “process of experimentation” in any of these cases.

      In the various versions of the regulations implementing § 41(d)(1), the

government implies that the end result may be certain so long as the method of

achieving that result is uncertain at the outset. Over time, however, the

government has apparently changed its view regarding whether the converse is

true—i.e., whether the method of achieving the result may be certain provided

that the end result itself is uncertain. Compare Credit for Increasing Research

Activities, 63 Fed. Reg. at 66,508 (stating that the result must be one “where the

means of achieving that result are uncertain at the outset” (emphasis added)), with

Credit for Increasing Research Activities, 66 Fed. Reg. at 66,368 (stating that

“the capability or the method of achieving [the] result” of the research, “or the

appropriate design of that result, [must be] uncertain as of the beginning of the

taxpayer’s research activities” (emphasis added)), and Credit for Increasing

Research Activities, 66 Fed. Reg. at 290 (stating that “the capability or method of

achieving [the] result” of the research must be “uncertain at the outset” (emphasis

added)).

      As for the issue of whether generally known techniques may be used in the

“process of experimentation,” the government has not addressed that question at


                                         -21-
all in any of the proposed or final regulations. See Credit for Increasing Research

Activities, 66 Fed. Reg. at 66,364, 66,368–69; Credit for Increasing Research

Activities, 66 Fed. Reg. at 284, 290; Credit for Increasing Research Activities, 63

Fed. Reg. at 66,505, 66,508. Because the government has either taken no position

or has altered its position on both issues, its interpretation of the “process of

experimentation” requirement is entitled to little deference. Skidmore, 323 U.S.

at 140.

                                           C

      We conclude that the term “process of experimentation” can include

research in which the taxpayer tries alternative methods to achieve a result and all

of the methods are already commonly known, but it is uncertain which method

will allow the taxpayer to achieve the result. As noted above, the definition of

“experiment” does not seem to prohibit the use of commonly or generally known

methods. The government has apparently not made the argument that the methods

cannot be commonly or generally known in previous litigation, nor has it

discussed or considered this interpretation of the statute in its evolving

regulations. Nor does the legislative history reflect a contrary legislative intent.

Congress stated that a “process of experimentation” would include efforts to

“develop, test, and choose among viable alternatives.” H.R. Conf. Rep. No. 99-




                                          -22-
841, at II-72 (1986) (emphasis added). Viable alternative methods need not be

new to the taxpayer nor must they be generally unknown to the public.

      Moreover, as the Treasury Department itself noted in discussing the

regulations implementing § 41 early in 2001, “virtually all research utilizes

existing scientific principles and technology.” Credit for Increasing Research

Activities, 66 Fed. Reg. at 283. Most scientific research requires the use of

known methodologies to determine whether or not particular results are

achievable. Although Congress intended to narrow the scope of the research tax

credit in 1986, 10 we conclude that the government’s position in this case

overstates how narrow Congress intended the research credit to be. Congress did

not intend to eliminate the use of already known techniques in the discovery of

new information through its narrowing of the tax credit clause.

                                          D

      The second issue is whether the taxpayer must show that it was uncertain

about whether the end result of the “process of experimentation” was technically

feasible at the time that it commenced the research. Courts that have previously

decided this issue have generally agreed with the government’s position. See,

e.g., United Stationers, 163 F.3d at 446 (stating that the legislative history


      10
          See, e.g., H.R. Rep. No. 99-426, at 178 (1986) (stating that the prior
definition of qualified research had been “applied too broadly in practice” and
that the new definition “targets the extended credit”).

                                         -23-
strongly suggests that “qualifying research must from its outset involve some

technical uncertainty about the possibility of developing the product” and

concluding that “debugging programs” alone can not constitute a “process of

experimentation” because of the lack of uncertainty as to the final result);

Norwest, 110 T.C. at 496 (holding that “at the outset uncertainty [must] exist

about the ability to develop the product in the scientific or laboratory sense”).

      The government substantially relies on the legislative history. The 1986

House Report defines “process of experimentation” as:

      a process of scientific experimentation or engineering activities to
      design a business item where the design of the item as a whole is
      uncertain at the outset, but instead must be determined by developing
      one or more hypotheses for specific design decisions, testing and
      analyzing those hypotheses (through, for example, modeling or
      simulation), and refining or discarding the hypotheses as part of a
      sequential design process to develop the overall item.
             [C]osts of developing a new or improved business item are not
      eligible for the credit if the method of reaching the desired objective
      (the new or improved product characteristics) is readily discernible
      and applicable as of the beginning of the research activities, so that
      true experimentation in the scientific or laboratory sense would not
      have to be undertaken to develop, test, and choose among viable
      alternatives.

H.R. Rep. No. 99-426, at 180–81 (1986) (emphasis added). The 1986 Senate

Report language is substantially identical, see S. Rep. No. 99-313, at 696 (1986),

and the 1986 Conference Report used slightly different language, see H.R. Conf.

Rep. No. 99-841, at II-72 (stating that a process of experimentation “means a

process involving the evaluation of more than one alternative designed to achieve

                                         -24-
a result where the means of achieving that result is uncertain at the outset”). This

legislative history suggests that the credit is inapplicable when the final design is

certain at the outset or the taxpayer knows how to achieve the result at the

beginning of the claimed research.

           TAASC cites subsequent legislative history from 1998 and 1999 to

support its position on the issue. Specifically, it notes that during the 1998

reenactment of the tax credit, the conference committee stated that the process of

experimentation test included research “even if the taxpayer knows at the outset

that it may be technically possible to achieve the result.” H.R. Conf. Rep. No.

105-825, at 1548. However, we cannot consider this subsequent legislative

history.

      In Pierce v. Underwood, 487 U.S. 552 (1988), the Supreme Court

considered whether legislative history included with the 1985 reenactment of the

Equal Access to Justice Act (“EAJA”) would control the interpretation of a

provision of the 1980 version of the EAJA that was reenacted without change in

1985. The legislative history indicated that Congress approved of the

interpretation of that provision of the EAJA by some courts, but not by other

courts. The Court held that Congress could not control the interpretation of a

statute in that way.

      If this language is to be controlling upon us, it must be either (1) an
      authoritative interpretation of what the 1980 statute meant, or (2) an

                                         -25-
      authoritative expression of what the 1985 Congress intended. It
      cannot, of course, be the former, since it is the function of the courts
      and not the Legislature, much less a Committee of one House of the
      Legislature, to say what an enacted statute means. Nor can it
      reasonably be thought to be the latter—because it is not an
      explanation of any language that the 1985 Committee drafted,
      because on its face it accepts the 1980 meaning of the terms as
      subsisting, and because there is no indication whatever in the text or
      even the legislative history of the 1985 reenactment that Congress
      thought it was doing anything insofar as the present issue is
      concerned except reenacting and making permanent the 1980
      legislation.

Id. at 566–67.

      As was the case in Pierce, neither the 1998 nor the 1999 reenactments of

the tax credit substantively changed the definition of “qualified research” in the

text of the statute. The only reference to the definition was in the legislative

history, and that legislative history does not indicate any intent to change the

textual meaning of the “process of experimentation” requirement. See H.R. Conf.

Rep. No. 106-478, at 132 (stating that Congress intended to “reaffirm” the

original meaning of the requirement). Congress cannot retroactively change the

meaning and intent of previously enacted statutory language through the

introduction of legislative history which purports to state what the original

meaning of that statutory language was. Congress must change the wording of the

statute itself if it wishes to change the meaning of the statute. Following Pierce,

we will not consider the subsequent legislative history urged upon us by TAASC.

See also Reno v. Bossier Parish Sch. Bd., 520 U.S. 471, 484 (1997) (refusing to

                                        -26-
consider a footnote in subsequent legislative history to the Voting Rights Act

because the Court “doubt[ed] that Congress” would amend the statute “by

dropping a footnote in a Senate report instead of amending the statute itself”);

Hadden v. Bowen, 851 F.2d 1266, 1268 (10th Cir. 1988) (refusing to consider

subsequent legislative history in connection with the EAJA, following Pierce).

      Looking to the legislative history for § 44F, the predecessor to § 41, it

appears that at least part of the goal of the tax credit is to provide incentives for

companies to invest in research that might not otherwise be undertaken because of

its high risks. See H.R. Rep. No. 97-201, at 111 (suggesting that the § 44F tax

credit was intended to encourage businesses otherwise “reluctant to allocate

scarce investment funds for uncertain research”). 11 Thus, the very uncertainty of

the research is a rationale for the tax credit in the first place. Allowing

experimentation to qualify for the tax credit where the feasibility of the final

result was certain would undermine that rationale, and might encourage

companies to be more conservative in their allocation of resources, concentrating

on problems with a solution that is evident from the outset.

      TAASC’s interpretation of the statute would allow relatively basic

activities, such as debugging software, to be included as “experimentation” within


      11
         The conference report for the § 44F tax credit states that the final
version enacted by Congress followed the House bill. See H.R. Conf. Rep. No.
97-215, at 223–24 (1981).

                                          -27-
the scope of § 41. See United Stationers, 163 F.3d at 445. Debugging software

often requires the utilization of multiple known methods to eliminate the bug

from the software. The software developer often has little or no doubt that one of

the methods will eliminate the bug. Thus, without the government’s limitation on

what “experimentation” can mean, all debugging would qualify for the tax credit.

The legislative history quoted above indicates that Congress did not intend such a

result. We are further persuaded by the maxim that tax credits are to be narrowly

interpreted. See New Colonial Ice Co., 292 U.S. at 440. In enacting the § 41

credit, Congress did not repeal the deduction under § 174, suggesting that less

risky research activities were to remain deductible, not creditable.

                                          V

      We now apply our legal analysis to the record before us. With respect to

the “discovering information” test, much of the record is filled with the

conclusory assertions of both parties that the evidence clearly demonstrates that

TAASC has, or has not, met the formulation of the test advanced by that party.

TAASC’s evidence emphasized that its products were new and innovative, and

therefore must meet the test; the government’s evidence minimized the

innovations in TAASC’s work. But a crucial question, as shown by our analysis

of the plain language of the statute, is whether any new information that TAASC

discovered in the course of its research was independent and separate from the


                                         -28-
new products that it developed. This issue was never developed by either party

below, and thus summary judgment for TAASC was inappropriate on this issue.

      As for the “process of experimentation” requirement, the affidavits

submitted by TAASC under seal indicate to us that TAASC did explore a variety

of alternatives to achieve the result in question. However, TAASC has also

conceded in its briefs that “TAASC believed . . . [its] goal was technically

feasible,” at least with respect to the Professional Tax Software project.

(Appellee’s Br. at 34.) And before the district court, TAASC “acknowledge[d]

that its programmers did not question the technical feasibility of the products” in

question. (1 Appellant’s App. at 77.) TAASC therefore cannot, as a matter of

law, meet this requirement.

                                         VI

      The judgment of the district court is REVERSED, and the case

REMANDED.




                                         -29-