University of Minnesota Law School University of Minnesota Law School Scholarship Repository Scholarship Repository Minnesota Law Review 2020 Eligible Subject Matter at the Patent Office: An Empirical Study of Eligible Subject Matter at the Patent Office: An Empirical Study of the Influence of Alice on Patent Examiners and Patent Applicants the Influence of Alice on Patent Examiners and Patent Applicants Jay P. Kesan Wang, Runhua Follow this and additional works at: https://scholarship.law.umn.edu/mlr Part of the Law Commons Recommended Citation Recommended Citation Kesan, Jay P. Wang, Runhua, "Eligible Subject Matter at the Patent Office: An Empirical Study of the Influence of Alice on Patent Examiners and Patent Applicants" (2020). Minnesota Law Review. 3212. https://scholarship.law.umn.edu/mlr/3212 This Article is brought to you for free and open access by the University of Minnesota Law School. It has been accepted for inclusion in Minnesota Law Review collection by an authorized administrator of the Scholarship Repository. For more information, please contact [email protected].
92
Embed
Eligible Subject Matter at the Patent Office: An Empirical ...
This document is posted to help you gain knowledge. Please leave a comment to let me know what you think about it! Share it to your friends and learn new things together.
Transcript
University of Minnesota Law School University of Minnesota Law School
Scholarship Repository Scholarship Repository
Minnesota Law Review
2020
Eligible Subject Matter at the Patent Office: An Empirical Study of Eligible Subject Matter at the Patent Office: An Empirical Study of
the Influence of Alice on Patent Examiners and Patent Applicants the Influence of Alice on Patent Examiners and Patent Applicants
Jay P. Kesan Wang, Runhua
Follow this and additional works at: https://scholarship.law.umn.edu/mlr
Part of the Law Commons
Recommended Citation Recommended Citation Kesan, Jay P. Wang, Runhua, "Eligible Subject Matter at the Patent Office: An Empirical Study of the Influence of Alice on Patent Examiners and Patent Applicants" (2020). Minnesota Law Review. 3212. https://scholarship.law.umn.edu/mlr/3212
This Article is brought to you for free and open access by the University of Minnesota Law School. It has been accepted for inclusion in Minnesota Law Review collection by an authorized administrator of the Scholarship Repository. For more information, please contact [email protected].
A. TheAliceDecisionRegardingEligibleSubjectMatter.........5371. AbstractIdeaandStatutoryLimits.....................................5372. ImplementationbythePTO...................................................543
B. UncertaintiesinEligibleSubjectMatter....................................5451. InnovationandUncertaintiesCreatedbythe
A. DataandMethodology......................................................................5551. DataSourcesandStudyObjects...........................................5562. DescriptiveAnalyses..................................................................5593. Methodology..................................................................................564
B. RegressionResults..............................................................................5681. CorrelationBetweenAliceRejectionsandOther
A. Increased§101RejectionsbyUSPTOPatentExaminers...............................................................................................5881. Bioinformatics..............................................................................5892. BusinessMethods.......................................................................5903. SoftwareArtUnits......................................................................591
B. TheAbilityofPatentApplicantstoOvercome§101Rejections................................................................................................5931. DifficultiesinOvercoming§101Rejectionsin
lins,Johnson,andStiversdraftedabilltoreform35U.S.C.§101ofthePatentActtoaddressissuesrelatedtopatenteligibility.1SincetheSu-premeCourtrulinginAliceCorp.v.CLSBank,2theindustryhasbeenconfronting uncertainties in the prosecution of patent applicationsandinpatentenforcementasaresultofthelawgoverningpatenteli-gibility,whicharguablyharms innovation.3In thenextyearor two,Congressisonceagainlikelytobeunderpressuretoaddresseligiblesubjectmatterreform,astheU.S.SupremeCourthasrecentlychosennottorevisitthistopicbydenyingcertiorari inaseriesofcasesin-volvingpatentablesubjectmatterinboththesoftwareandbiotech-nologyfields.4
EligiblesubjectmatterinpatentlawisathresholdrequirementofpatentabilityandreferstosubjectmatterthatcanlegitimatelybethesubjectofaU.S.patent.5Patentlawacceptsfourcategoriesofin-ventions—machines,articlesofmanufacture,compositionsofmatter,andprocesses—asproperlybeingthesubjectofaU.S.patent.6Byju-dicial exceptions, however, abstract ideas, natural phenomena, andlawsofnaturearecategoricallyexcludedfrompatentprotection.7Ithas been difficult to definewhat the three categories of exclusionsmeaninpractice,partlybecausethemeaningsoftheseexclusionsareunclear.Asaresult,courtshavestruggledtospecifylegalteststoop-erationalizetheseexclusions.
Since2010, theU.S.SupremeCourthasmadethree forays intodefiningthejudicialexclusionsinBilski,Mayo,andAlice.8Thesecasesmotivatedthedraftingofthe“Coons-Tillis”billtoreform§101ofthePatentAct.9Mostrecently,in2014,theU.S.SupremeCourtaddressedtheabstractideasexceptionandoutlinedatwo-parttestfordetermin-ing thescopeofpatent-eligiblesubjectmatter inAlice.10In the firststep,theCourtaskswhetherthepatentclaimatissueisorincorpo-rates an abstract idea.11 If not, the claim is patent-eligible.12 If theclaiminvolvesanabstractidea,however,thesecondstepapplies,andtheCourtaskswhethertheabstractideahasbeentransformedintoaninventiveconceptbyincludingadditionallimitationstothepatentclaim,therebyrenderingtheclaimeligibleforpatentprotection.13
TheAlicedecisionhasbeenineffectforoverfiveyears.14Thereissignificantscholarlydebateaboutwhetherthecurrentlawaddress-ing eligible subject matter after Alice creates uncertainties andwhetherAlice fails to providemeaningful guidance. Some scholarsworrythattheAliceframeworkharmsinnovation.15Becauseofuncer-taintiesinpatentingstandards,inventorsmayprefertohavetheirin-novationsprotectedundertradesecretlawinsteadofrelyingonthecurrentpatentregime.16Moreover,financiersandventurecapitalists
8. Bilskiv.Kappos,561U.S.593 (2010);MayoCollaborativeServs.v.Prome-theusLab’ys,Inc.,566U.S.66(2012);AliceCorp.Pty.Ltd.v.CLSBankInt’l,134S.Ct.2347(2014). 9. SeeSens.TillisandCoonsandReps.Collins,Johnson,andStiversReleaseDraftBillTexttoReformSection101ofthePatentAct,supranote1(“Noimplicitorotherjudicially created exceptions to subjectmatter eligibility, including ‘abstract ideas,’‘lawsofnature,’or‘naturalphenomena,’shallbeusedtodeterminepatenteligibilityundersection101,andallcasesestablishingorinterpretingthoseexceptionstoeligi-bility are hereby abrogated.”);MichaelBorella,Senate Subcommittee on IntellectualPropertyHoldsHearingsonProposedRevisionsto35U.S.C.§101,PAT.DOCS(June17,2019), https://www.patentdocs.org/2019/06/senate-subcommittee-on-intellectual-property-holds-hearings-on-proposed-revisions-to-35-usc-101.html [https://perma.cc/S4C6-XRE2] (“Themotivationbehind thebill and thesehearingswas thewide-spreadunderstandingthataseriesofSupremeCourtdecisionsinthelastdecade...had‘madeahash’ofpatenteligibility.”). 10. AliceCorp.Pty.Ltd.,134S.Ct.at2355(citingMayo,566U.S.at75–80). 11. Id. 12. Id. 13. Id. 14. Id.at2347. 15. See,e.g.,Quinn,supranote3(interviewingScottAlterwhobelievesthatAliceis“arguablynotagooddecision,”foritharmsinnovationincentivesanddoesnotre-wardandprotectinnovation). 16. SeeJoannaBrougher&KonstantinM.Linnik,PatentsorPatients:WhoLoses?,32NATUREBIOTECHNOLOGY877,880(2014)(suggestingthatsomeinventorsmayhave
2020] IMPLICATIONSOFALICE 531
relyonthepredictabilityofthepatentlawsandtheavailabilityofpa-tentprotectiontoassesstheeconomicviabilityoftheinnovationsinwhichtheymightinvest.17Accordingly,investorspreferreliableandreasonableexpectationsregardingthelawthatgovernspatenteligi-bilitysotheycanminimizeeconomiclossandmaximizeeconomicef-ficiency.18Notall scholars,however,agree thatAliceharms innova-tion.MarkLemley argues that stringent restrictionspreventing thepatentingofabstractinventionswillresultinamorecompetitivere-search and development (R&D) environment. 19 Jason Schultz andBrianLovebelievethatpatentsareusedasadefensiveweaponandthusarelesslikelytobetheresultofadesiretoprotectinvestmentsinR&Dorrewardinnovation.20JoshuaSarnoffisinfavorofthecur-rentjudicialexclusionsthatprotect“thepublicdomainofscience,na-ture,andideas”from“unwarrantedencroachment.”21Thesescholarlydebateswerereflectedinthehearingsonproposedlegislationtore-formpatentlaw.22ItiscurrentlyunclearifabillabrogatingAliceand
theirinventionsprotectedunderthetradesecretmechanisminsteadofpatentunderthiscircumstance). 17. See Ognjen Zivojnovic,Patentable SubjectMatter AfterAlice–DistinguishingNarrow Software Patents from Overly Broad Business Method Patents, 30 BERKELEYTECH.L.J.807,838(2015)(suggestingthatwemustweighthebenefitbroughtbythisinventionagainstthesocialcostofagrantedpatent). 18. SeeBenDugan,MechanizingAlice:AutomatingtheSubjectMatterEligibilityTestofAlicev.CLSBank,2018U.ILL.J.L.TECH.&POL’Y33,41(2018)(arguingthatrea-sonableandreliablepredictionbasedonAlicecansaveasignificantamountoftimeandcost). 19. MarkA.Lemley,MichaelRisch,TedSichelman&R.PolkWagner,LifeAfterBilski,63STAN.L.REV.1315,1331(2011). 20. JasonM.Schultz&BrianJ.Love,BriefofAmiciCuriaeLaw,Business,andEco-nomicsScholarsinSupportofRespondentsinAliceCorp.Pty.Ltd.v.CLSBankInterna-tional,etal.,4N.Y.U.J.INTELL.PROP.&ENT.L.358,366(2015). 21. TheStateofPatentEligibilityinAmerica,PartI:HearingBeforetheSubcomm.onIntell.Prop.oftheS.Comm.ontheJudiciary,116thCong.13–23(2019)(statementof Professor Joshua D. Sarnoff, Professor of Law, DePaul University), https://www.judiciary.senate.gov/imo/media/doc/Sarnoff%20Testimony.pdf [https://perma.cc/R9KQ-C9CX](criticizingthe“Coons-Tillis”billforharminginnovationbyitsoverpro-tectionofpatentsinsteadofpreservingthepublicdomain). 22. SeeBorella,supranote9;seealsoJasonRantanen,GuestPostbyProf.Ghosh:AFitter Statute for the Common Law of Patents, PATENTLY-O (Aug. 1, 2019), https://patentlyo.com/patent/2019/08/fitter-statute-patents.html [https://perma.cc/BBQ6-ZQ8X](arguingthatthe“Coons-Tillis”billwouldlimitjudicialexceptionstothePatentActandwouldlikelybefoundunconstitutional);Briefof19LawProfessorsasAmiciCuriaeinSupportofPetitionforaWritofCertiorari,Sequenom,Inc.v.AriosaDiagnos-tics, Inc.,788F.3d1371(Fed.Cir.2015) (Nos.2014-1139,2014-1144) (noting thatlowercourtshave invalidatedpatents thatare legitimatebecause theyhavemisap-pliedtheAlicetest).
532 MINNESOTALAWREVIEW [105:527
other judicial exceptionswill be enacted, and thewidespread con-cernsaboutAlicepersist.
WhiletheAlicetestforeligiblesubjectmatterismostapplicableto computer-implemented inventions (i.e., computer software), 23lowercourtdecisionspost-AliceshowthatnoneofthepatentclaimsinanytechnologyareaaresparedfromreviewundertheAliceframe-work (e.g., an improved high-performance computer memory sys-tem).24Businessmethodsthataresoftware-implementedandinvolvethe Internet often develop new types of e-commerce.25 Patents onbusinessmethods,asubjectareasimilartothepatentatissueinAlice,maybeeligibleforpatentprotection,unlesstheymerelyinvolveanabstractideaandareinsufficientlytiedtoaparticularreal-worldim-plementation.26 Ognjen Zivojnovic believes that Alice kills all purebusiness methods patents because all business methods patentsmerelyreciteanabstracteconomicpracticeandsimplyemployagen-eral purpose computer to implement the businessmethod.27 PeterMenellagreesthattheU.S.ConstitutionandthePatentActwerenotmeanttoprotectbusinessmethods.28Bycontrast,AlexDejeanarguesthat technologicalapplications, suchasonlineshopping, individual-izedadvertising,andautomatedcustomerservice, ledtoatransfor-mation in the patent-eligibility of computer-implemented subjects,
23. TysverBeckEvans,ApplyingStepOneoftheAlice/MayoTest,BITLAW:GUID-ANCE,https://www.bitlaw.com/guidance/patent/applying-step-one-of-Alice-Mayo-test.html [https://perma.cc/68MY-66ZQ] (indicating “[a]ll three of the identifiedshadowtestsseemmostapplicabletocomputer-implemented”inventions,althoughtheymaybeapplicabletootherareasaswell). 24. HungH.Bui,ACommonSenseApproachtoImplementtheSupremeCourt’sAl-iceTwo-StepFrameworktoProvide“Certainty”and“Predictability,”100J.PAT.&TRADE-MARKOFF.SOC’Y165,230(2018). 25. Nam Kim, Software and Business Method Inventions After Alice, SHEPPARDMULLIN (Sept. 23, 2016), https://www.intellectualpropertylawblog.com/archives/software-and-business-method-inventions-after-alice[https://perma.cc/64TB-2YCC](“Businessmethodsrefertomethodsofdoingbusiness,includingnewtypesofe-com-merce,insurance,banking,etc.,oftenimplementedassoftwareincomputersandin-volvingtheInternet.”). 26. Bilskiv.Kappos,561U.S.593,608–09(2010);seealsoZivojnovic,supranote17,at813(explaininghowcourtsdonotapprovepatenteligibilityforsoftwarethatdoesnotaccompanynewandusefulhardware);MarkA.Lemley,SoftwarePatentsandtheReturnofFunctionalClaiming,2013WIS.L.REV.905,962(2013)(notingthatmostbutnotallFederalCircuitdecisionsafterBilskihavedeniedsoftwarepatentclaims). 27. Zivojnovic,supranote17,at827. 28. PeterS.Menell,FortyYearsofWonderingintheWildernessandNoClosertothePromisedLand:Bilski’sSuperficialTextualismandtheMissedOpportunitytoReturnPatentLawtoItsTechnologyMooring,63STAN.L.REV.1289,1312–13(2011)(“Thereisnoreasontobelievethat‘businessmethods’havebecomeascienceortechnologyfittingthefunctionalpatentmoldduringthecourseofthepasttwocenturies.”).
2020] IMPLICATIONSOFALICE 533
includingbusinessmethods.29Along the same lines,DavidReardonandGeneQuinnurgethatAlicemustbereversedbecausethetrans-formativecharacteristicsofsoftwarearetechnologicalinnature.30
In addition to software and businessmethods,Alice has nega-tivelyimpactedpatenteligibilityinbiotechnology(e.g.,biocomputingandbioinformatics).31HallieWimberlysuggeststhatCongressortheSupremeCourtshouldbroadenthescopeofpatentsubjectmatterel-igibilitybecauseofthehighburdenplacedonbiotechnologicalinven-tionsafterAlice.32TheBiotechnologyIndustryOrganization(BIO)andPharmaceutical Research and Manufacturers of America (PhRMA)botharguethattherestrictionsoneligiblesubjectmatterafterAliceshouldbeloosened.33
TheAlice test impacts theentire lifecycleofapatent, includingpatentapplicationpreparation,patentprosecutionintheU.S.PatentandTrademarkOffice (PTO), andpatent enforcement in the courtsandinpost-issuanceproceedingsinthePTO.34Thiscreatessignificant
29. AlexDejean,ACritiqueoftheSupremeCourtHoldinginAliceCorpv.CLSBankwithNewRhetoric,12COLLOQUY52,59(2016). 30. DavidReardon&GeneQuinn,AliceisDueforReversal:ScienceProvesItsRea-soning Unsound, IPWATCHDOG (Mar. 21, 2019), http://www.ipwatchdog.com/2019/03/21/alice-due-reversal-science-proves-reasoning-unsound[https://perma.cc/5P36-XY9B](claimingthatallactivesoftwareisastransformativeas“DNA[m]anipu-lation,a[f]orge,or[p]harmaceuticals”). 31. EugeneKim,BiotechPatentEligibility:ANewHope,2017COLUM.BUS.L.REV.1157,1160(2017)(“[T]hisframeworkwasusedtodenypatenteligibilityforanon-invasivemethodof accessing fetalDNAusingpreviouslydiscardedcell-free cffDNAand amethod for gene detection by amplifying and analyzing significantly shorter‘non-codingregionsknowntobe linked to thecodingregion’of interest.”);seealsoAriosaDiagnostics,Inc.v.Sequenom,Inc.,788F.3d1371(2015)(holdingthataprena-talfetalDNAtestwasnotpatenteligiblebecauseitwasdirectedatanaturalphenom-enon);GeneticTechs.Ltd.v.MerialL.L.C.,818F.3d1369(2016)(holdingthatapatentfordetectionofgeneticvariationswasinvalidbecauseitwasdirectedatlawofnature);LidiaYamamoto,DanielSchreckling&ThomasMeyer,Self-ReplicatingandSelf-Modi-fyingProgramsinFraglets,2BIO-INSPIREDMODELSNETWORK,INFO.,&COMPUTINGSYS.159(2007)(“Artificialchemicalcomputingmodelsaregainingincreasingprominenceinthe design of bio-inspired software with self-organizing and emergent proper-ties....”). 32. HallieWimberly,Comment,TheChangingLandscapeofPatentSubjectMatterEligibility and Its Impact onBiotechnological Innovation, 54HOUS.L.REV. 995,1025(2017)(“Consideringtheoutspokendissatisfactionwiththestrictstandard,thetimeisaptforeithertheSupremeCourttorevisitthematterandbroadenthescopeofpa-tentsubjectmattereligibilityorforCongresstostepinandreiteratetheideathatpa-tentablesubjectmattershouldbegivenbroadscope.”). 33. Id.at1020. 34. Dugan,supranote18,at41.
534 MINNESOTALAWREVIEW [105:527
uncertaintiesinalloftheseproceedings.35PaulGugliuzzaandLemleyempiricallyreviewed104FederalCircuitdecisionsonpatentablesub-jectmatterafterAlice.36LemleyandSamanthaZyontzalsoempiricallyreviewed808federalcourtdecisionsaboutpatentablesubjectmatterafterAlice.37DataontheimpactofAliceonpatentprosecutionarealsobeing updated on blogs by patent practitioners and commentators(e.g.,BilskiBlog,IPWatchdog,andthelike).38Thatsaid,thereislim-itedempiricalworkfocusingontheuncertaintiesinpatentprosecu-tionsystematically.TherearenostudiesregardingAlice’simpactonpatentexaminersatthePTOinvarioustechnologycenters,noronpa-tent applicants’ responses to theAlice regime or patent applicants’ability toovercomeeligible subjectmatter rejectionsduringpatentprosecution.39
ThisArticlepresentsacausalempiricalstudyofAliceandcare-fullyexploreshowAliceimpactspatentexaminersandpatentappli-cantsinvarioustechnologyareas.Itconsidershowpatentapplicantsemploydifferentstrategies(e.g.,filingamendedpatentclaims,filingnewpatentapplications,orchoosingtoabandonorneverseekpatentprotection)toovercomePTOrejectionsforineligiblesubjectmatterbycomplyingwithAlice.ThestudydeploysdataofallthePTOofficeactions over the five-year period between 2012 and 2016, roughlytwoyearsbeforeandaftertheAlicedecisionin2014.Itcoversatotalof4.48millionpatentofficeactionsandpatenteeresponses.Thestudyalsocontainsdetailsregardingartunitsandthespecificgroundsfortherejectionsgivenbypatentexaminers.Ourmethodologiesincludelogisticregressionsanddifference-in-difference(D-i-D)regressions.Logisticregressionsaredeployedtoexploretheassociationbetween
35. SeeJasperL.Tran,TwoYearsAfterAlicev.CLSBank,98J.PAT.&TRADEMARKOFF.SOC’Y 354,358–59 (2016) (showing statistical evidence that thePTABand theFederalCircuitinvalidatedalargeproportionofpatentsafterAliceunder§101). 36. PaulR.Gugliuzza&MarkA.Lemley,CanaCourtChangetheLawbySayingNothing?,71VAND.L.REV.765,767(2018). 37. MarkA. Lemley& Samantha Zyontz, DoesAlice Target Patent Trolls? (un-publishedmanuscript)(onfilewithauthors). 38. See, e.g., Robert Sachs, Alice: Benevolent Despot or Tyrant? Analyzing FiveYearsofCaseLawSinceAlicev.CLSBank:Part1,IPWATCHDOG(Aug.29,2019),https://www.ipwatchdog.com/2019/08/29/alice-benevolent-despot-or-tyrant-analyzing-five-years-of-case-law-since-alice-v-cls-bank-part-I[https://perma.cc/ZG37-PEDG](findingthattheAlicetesthasresultedinasignificantnumberofrejectionsforpatentineligiblesubjectmatterandabandonedapplications). 39. But seeColleenChien& JiunYingWu,DecodingPatentable SubjectMatter,2018PATENTLY-OPAT.L.J.1(Oct.16,2018),https://patentlyo.com/media/2018/10/Chien.Decoding101.2018.pdf[https://perma.cc/M3JQ-KH6U](presentingastatisticalanalysisofofficeactions).
2020] IMPLICATIONSOFALICE 535
thepatentclaimrejectionsunderAliceandstatutoryrejectionsgivenbyexaminersunder35U.S.C.§§101,102,103,and112(a)–(f).Thisstudyalsoexploresthecorrelationbetween§101rejectionsandtheotherstatutoryrejectionsbecausetheAlicedecisionimplicatesnov-elty,obviousness,writtendescription,enablement,andclaimdefinite-nessunderthePatentAct.ThestudydeploysD-i-Dregressionstoex-plorewhetherAliceresultsinmore§101rejectionsinthesoftware,business methods, and biotechnology areas, and sub-categorieswithin those three areas. Patent applications in themanufacturingsectorareselectedas thecontrolgroupbecausetheyarerarelyre-jectedundertheabstractideasexclusionandarethereforeveryun-likelytobeaffectedbyAlice.Thestudycomparespatentapplicationsinthecontrolgroupwiththepatentapplicationsinthreetechnologyareas—business methods, bioinformatics, and software—and in-cludesbroadandnarrowdefinitionsforthesecategories.
ThestudydemonstratesthattheU.S.SupremeCourt’sdecisioninAliceimpactspatenteligibilityindifferenttechnologyareastodiffer-entdegrees.Moreover,theimplementationofthedecisionbythePTOstrengthenedtheeffectsofthedecision.Inallthreebroadtechnologyareasthatwestudied,applicantsreceivedmoreAlicerejectionsand§101rejectionsafterAlice;theyarepositivelyassociated.Alicerejec-tionsarenotalwayspositivelyassociatedwithothertypesofstatu-toryrejections,however.Moreover,patentapplicantsinallthreetech-nology areas filed fewer patent applications post-Alice, with thegreatest reduction occurring in bioinformatics. Patentees adjustedtheirpatentingstrategiesafterAlice.Somestrategieseffectivelyover-came§101rejections,butothersdidnot.
Aliceplacesthehighestcostofpatentingonbioinformatics.40Ap-plicationsforbioinformaticsreceivedmanymore§101rejectionsbe-causeofAlice,andtheapplicantsalsoexperienceddifficultiesinover-comingtheserejections.Similarly,applicationsforbusinessmethodsreceivedmore§101rejectionsbecauseofAlice.41Nevertheless,aver-age applicants in business methods learned from Alice, receivingfewer§101rejectionswhentheyfiledapplicationspost-Alice.Appli-cantsinthebusinessmethodsofe-commerceandfinance,however,stillfounditdifficulttoovercome§101rejectionsbothpost-Aliceandwhenrespondingtotheexaminers’initialroundofrejectionsunder§101.Alicealsoimposedvaryingdegreesofcostofpatentingfordif-ferenttypesofsoftwareinventions(e.g.,cryptographyandsecurity,
40. SeeinfraPartIII.A.1. 41. SeeinfraPartIII.A.2.
536 MINNESOTALAWREVIEW [105:527
databasesandfilemanagement,GUIanddocumentprocessing,com-puterarchitecture,digitalandopticalcommunication,computernet-works,telecommunications,digitalcameras,recordingandcompres-sion,computergraphicsprocessing, telemetryandcodegeneration,and artificial intelligence (AI)). Some art units related to softwarefacedhighercostsofpatentingafterAlice,42butAlicemightnotbeadirectorsignificantreasonforthe increasinguncertainties inotherartunitsrelatedtosoftware.OurempiricalresultsportrayamurkypictureofhowAliceplaysoutindifferenttechnologysectors—quiteunworkableinseveralsectors,yetprovidingpredictableguidanceinafewareas.
PartIofthisArticleintroducestheeligiblesubjectmattertestun-derAliceandexplainshowithasbeenimplementedbycourtsandthePTO,includingthePatentTrialandAppealsBoard(PTAB)andpatentexaminersinexparteprosecution.ThetestfordeterminingwhetherapatentclaimisabstractunderAlice,therebyfallingwithinajudicialexceptionto35U.S.C.§101,shouldnotbefaciallyconfusedwithotherstatutory patentability requirements, such as novelty in § 102 andnon-obviousnessin§103.Unfortunately,theAlicedecisionitselfcre-atesmany uncertainties in this regard, in addition to uncertaintiessurroundingitsapplicationtodifferenttypesofpatentclaimsanddif-ferent technologicalsubjectmatters.Part IIdiscussesourempiricalstudydesign,includingdataandmethodology,andanalysis.Wepro-videadescriptiveanalysisofthedataandacausalanalysiswiththeregressionresults.PartIIIdiscussestheimplicationsoftheempiricalresults,explainingtheeffectsofAliceonthetechnologiesintheareasofbioinformatics,businessmethods,andsoftware.
42. SeeinfraPartIII.A.3. 43. MemorandumfromAndrewH.Hirshfeld,DeputyComm’rforPat.Examina-tion Pol’y, U.S. Pat. & TrademarkOff. to the Pat. Examining Corps (June 25, 2014),https://www.uspto.gov/sites/default/files/patents/announce/alice_pec_25jun2014.pdf[https://perma.cc/GJ7M-WNS4].SeegenerallyGugliuzza&Lemley,supranote36;AliceCorp.Pty.Ltd.v.CLSBankInt’l,134S.Ct.2347(2014).
A. THEALICEDECISIONREGARDINGELIGIBLESUBJECTMATTERThisSectionexplainstheAlicedecisionandhowithasbeenim-
plementedbythePTO.Itdealswiththeabstractideaexceptiontopa-tenteligibilitywithatwo-steptest,butAliceitselfdoesnotclearlyde-fine what constitutes an abstract idea. 44 The Alice opinion alsoappears to implicate other statutory requirements forpatentabilitysuch as novelty and non-obviousness.45 In implementing Alice, thePTOprovidedsomestepstodefineabstractideasandfurtherspecifytheAlicetest.46
The two-step test inAlice couldarguablybe seenasprovidinggreaterclaritytopatenteligibilitystandardsinharmonywithotherforeignpatentregimes,butinreality,itsapplicationcomeswithmanychallenges. Under Article 52 of the European Patent Convention(EPC), discoveries, scientific theories, mathematical methods, aes-theticcreations,schemes,rulesandmethodsforperformingmentalacts,playinggames,ordoingbusiness,programsforcomputers,andpresentationsof information,shouldnotberegardedas inventions;however,additionaltechnicalfeaturesrecitedintheclaimscanconferpatent eligibility. 50 Although China, Japan, and South Korea treat
computer-related inventions and business methods slightly differ-ently,theytreatallpatentapplicationssimilartoEurope,i.e.,asbeingeligibleif“technicalcharacteristics”arepresent.51
a. “AbstractIdea”asa§101IssueTheU.S.SupremeCourtinAlicedidnotdefinetheprecisescope
of the categories of “abstract ideas” or explain how to determinewhetherthepatentclaimcontainedan“abstract idea.”54Thevague-nessoftheconceptof“abstractidea”maybetracedbacktoArticleI,Section 8, Clause 8 of the U.S. Constitution, which is the basis forproviding exclusive rights for inventions and creations and for§101.55MallaPollackpointsout that thisArticle itselfdoesnotex-plainhowonedecideswhichdiscoveriespromotetheprogressoftheusefularts.56
bythetechnicaldisclosure(i.e.,thepatentspecification),raisingwrit-ten description and enablement concerns under § 112(a), or thatclaimsarenotsufficientlydefiniteunder§112(b).64
TheSupremeCourtinAliceprominentlynotedthepolicyconcernofpreemptionraisedbyabstractpatentclaims.65JosephCraignotesthattheFederalCircuitalsorequirespatentclaimspecificitytoreducepreemptionconcernsintheireligibilityanalysisunderAlice.66Forex-ample,inInternetPatentsCorp.v.ActiveNetworkInc.,theFederalCir-cuitheldthataninventiondirectedatasolutiontothetechnicalprob-lem of data loss in browsing websites failed to claim a technicalsolution in sufficiently concrete terms to limit preemption, therebyrendering itpatent-ineligibleunderAlice.67Moreover,AndrewChinhighlightstheissuesofpreemptionconcernsonthesoftwareindus-try.68ChinarguesthatthepreemptionconcernsinAlicefailtoprovideclearguidanceforpatent-eligibility.69
AlicelimitsanNPE’sabilitytoownbasicandeverydayideasthataremerelyimplementedwithacomputer.73Likewise,JeffreyLefstinar-guesthattheeligibilityofpatentsandotherfundamentalprinciplesofpatentability (i.e., §§ 102, 103, or 112) are historical tools used bycourtstodealwiththepreemptionissue.74Alice,then,wasnotarea-sonforexcludingpatentabilitybecauseofpreemptionconcerns,butaresultoftheambiguousimplicationofthatrationale.75Indeed,inre-centcasessuchasBerkheimerv.HPInc.,76AatrixSoftwareInc.v.GreenShadesSoftwareInc.,77andMyMailv.ooVoo,78theFederalCircuitclar-ifiedthatthedeterminationofpatenteligibilityunder§101mayre-quirepreviousconstructionofcriticalanddisputedclaimterms.79
c. AbstractIdea,InventiveConcept,and§§102and103ManyscholarsnotethattheSupremeCourtinAlicedecidedthe
issueofpatenteligibilityunder§101bybleedingintothenoveltyandnon-obviousness requirements under §§ 102 and 103. John DuffycommentsthatthejudicialexceptionsinAlicewereinterpretedliber-allyandexpansively,soastohavethepotentialto“swallowallofpa-tentlaw.”80Additionally,MariaSinatrasuggeststhatAlice’svagueandambiguous language regarding abstract ideas further blurs and
interjects§§102and103requirements into the§101analysis.81Anegativeconsequenceofblurringtheeligibilityandpriorartdistinc-tionsinpatentlawisincreasedtransactioncostsinbothpatentlitiga-tion andpatent examination, as patent eligibility counterclaims (orthroughmotionstodismissunderRule12(c))andPTO§101rejec-tions play an increasingly important role.82Indeed, afterAlice, it iscommontoseethatcourtsanalyzeclaimtermsforjudgingnoveltyornon-obviousnessunder§§102and103inordertodeterminepatent-eligibilityunder§101.83Manydistrictcourtshaverequireddefend-antstoarguearoundnovelty/anticipationunder§102andobvious-nessunder§103beforeaheavydebateover§101.84
Bycontrast,thePTOdoesnotthinkthattheAlicetestaddresses§§102or103.85Section102addressesnoveltyand§103addressesthe issue of obviousness, and PTO examiners compare the patentclaimstothepriorartunderthoserequirements.86Butthoserequire-mentsaredifferentfromthe“additionalelements”inthetwo-steptestinAlice.87Therefore,regardlessofwhethertheSupremeCourtfurtherdefineswhatconstitutesan“abstract idea” in the future,until then,
81. SeeMariaR.Sinatra,DoAbstractIdeasHavetheNeed,theNeedforSpeed?:AnExamination of Abstract Ideas afterAlice, 84FORDHAML.REV.821, 841, 849 (2015)(showingthatdistrictcourtsusedthetermsof“conventional,”“longprevalent,”“rou-tine,”and“wellknown”intheirdecisionscitingAlice). 82. SeeSchultz&Love,supranote20,at360,366(believingthatthesubstantiveconditionsofpatentabilityin§§102,103,and112havemuchgreaterlitigationcostscomparedtothelitigationcostwithrespectto§101). 83. SeeRobert Sachs, Twenty-TwoWays Congress Can Save Section 101, BILSKIBLOG (Feb. 12, 2015), https://www.bilskiblog.com/2015/02/twenty-two-ways-congress-can-save-section-101 [https://perma.cc/4N6T-NKA6] (introducing how a§101patenteligibilityissuemergedwiththecriteriaofotherpatentstatutes,includ-ing§§102and103);Parkerv.Flook,437U.S.584,600(1978)(Stewart,J.,dissenting)(criticizingthatpatenteligibilityin§101shouldnotimportintoitsinquirythecriteriain§§102and103);Diamondv.Diehr,450U.S.175,211(1981)(Stevens,J.,dissenting)(admittingthefailureofthecourtsinrecognizing“thecriticaldifferencebetweenthe‘discovery’requirementin§101andthe‘novelty’requirementin§102”). 84. SeeMatthewBultman,GilstrapChangesPlayingFieldwithPatentEligibilityRule,LAW360(Aug.14,2019,7:44PM),https://www.law360.com/articles/1188573/gilstrap-changes-playing-field-with-patent-eligibility-rule[https://perma.cc/7MME-NY3L](“Thereareanumberofcourts, theEasternDistrictofTexas included,withlocalrulesrequiringdefendantstopresentinvaliditycontentionsbasedonanticipa-tionandobviousnessearlyoninacase.”). 85. SeeMemorandum from Robert W. Bahr, supra note 65 (“The question ofwhetheradditionalelementsrepresentwell-understood,routine,conventionalactiv-ityisdistinctfrompatentabilityoverthepriorartunder35U.S.C.§§102and103.”). 86. Seeid. 87. Seeid.
theU.S.SupremeCourt,thePTOissuedPreliminaryExaminationIn-structionsinviewofthiscase.89TheinstructionsexplainthattheSu-premeCourtheldthatclaimsinvolving“abstractideas,”particularlycomputer-implemented“abstract ideas,”arepatent-ineligibleunder§101. 90 The instructions provide four examples of abstract ideas,learningfromtheAlicedecision,including(1)fundamentaleconomicpractices, (2) certain methods of organizing human activities, (3)ideasthemselves,and(4)mathematicalrelationshipsorformulas.91ThePTOnotedthatAliceisanextensionofMayobecausetheframe-workoftheAlicedecisionwascurrentlybeingusedbythePTOtoex-amineclaimsinvolvinglawsofnatureafterMayo“buthadnotbeenusedforclaimsinvolvingabstractideas.”92
InDecember2014,sixmonthsafterAlicewasdecided,thePTOformallyimplementedtheAlicedecisionbyissuingInterimEligibilityGuidance(InterimGuidance)tothepatentexaminingcorps.93TheIn-terimGuidancemergesothertestsforpatent-eligibilityissuedbytheSupremeCourtinMyriad,94Mayo,95andBilski96anddevelopsatwo-steptest.97Thefirststep(Step1)istodetermineifthepatentclaimisdirectedtoaprocess,machine,manufacture,orcompositionofmattersoastobepatent-eligibleunder§101.98Ifnot,theclaimisstatutorilynon-eligible and rejected without consideration of those judicial
opinions.99 The second step is a two-part test (Steps 2A and 2B),whichdealswithjudicialexceptions;itapplieswhentheclaimspassStep1.100ItcomparesAlicetothosethreeearlierSupremeCourtdeci-sionsonpatent-eligibilityissues.101
Step2AinthePTOInterimGuidancedetermineswhetherthepa-tent claim is directed to a judicial exception, such as an “abstractidea.”102TheInterimGuidanceexpandsthefourexamplesofabstractideasinthePreliminaryExaminationInstructionstomanyexamplesof abstract ideas.103 Notably, “software is not automatically an ab-stractidea.”104Forsoftwareclaimsthatmayincludeanabstractidea,examinersare instructedto furtheranalyzetheclaimasawholetodeterminepatenteligibility.105
Iftheclaimisdirectedtoanexceptionforabstractideas,natureornaturalphenomena,ornature-basedproducts,Step2Bisappliedtodeterminewhethertheclaimamounts tosignificantlymorethantherelevant judicialexceptions.106InStep2B, ifapatentclaimasawholedoesnotreciteadditionalelementsthatamounttosignificantlymorethantherelevantjudicialexceptions,theclaimisnotpatent-eli-gibleandisrejectedunder§101.107IftheclaimpassesStep2B,itwillbefurtherexaminedunder§§101(utility,inventorship,anddoublepatenting),102(novelty),103(non-obviousness),and112(enable-ment,writtendescription,bestmode,andclaimdefiniteness,amongothers).108Whenaclaimisrejectedbecause it fallswithina judicialexception, the PTO Interim Guidance requires that the examinersidentifythespecificjudicialexceptionintherejection.109
Anysubsequentofficeactionsonthemeritsareusuallyfinalre-jections.110However, ifexaminersrejectapatentclaimunder§101afterStep2B,whichdoesnotrequireapplicantamendments to theclaim,thenewgroundsforrejectionarenon-final.111Meanwhile,the
ardtoevaluatewhetherapatentclaimisapatent-ineligibleabstractidea.122SamuelRegerbelievesthatthismanifestlyevidentstandardcan reduce litigation costswhen courts applyAlice.123Practitionersandscholarshavepresentedempiricalevidencesuggestingtheexist-enceofsignificantuncertaintyregardinghowAliceshouldbeappliedby the Federal Circuit and other federal courts when those courts
implementtheAlicedecision.124Forinstance,JoeMullinobservedthatintheEasternDistrictofTexas,whereNPEshaveahigherwinratecomparedtootherdistricts,Alicedidnotcausethiscourttorulemorefrequently againstNPEsor “patent trolls.”125Further,morepatentssurvivedpost-Alice126eventhoughthenumberofpatentapplicationsfiledbyNPEshasrecentlydropped.127Moreover,afterDaniKassre-viewedthecostofIPlitigationovermanyyears,sheconcludedthatAlicecontributestothedecreaseinpatentlitigationandtheincreaseinthecostofIPlitigationamonglargecompanies.128RobertSachsre-viewedallfederalcourtdecisionsfromthedateoftheAlicedecisiontoJuly1,2015,andfoundthatfederaldistrictcourtsandtheFederalCircuitinvalidated66.1%ofallpatentsand76.7%ofallclaimschal-lengedunder§101.129SachsrecentlyextendedthedatatoJune2019andfoundthat62%ofthecasesregardingpatentablesubjectmatterinfederaldistrictcourtsandtheFederalCircuitinvalidatedthosepa-tents.130This rate is slightly lower than the judicial data from fouryearsago.131
GugliuzzaandLemleyreviewed104casesonpatentablesubjectmatterdecidedbytheFederalCircuitbetweenJune20,2014,andJune19,2017.132Theirdatapresentatoughstoryforpatentees:theFed-eral Circuit is very likely to invalidate claims based on patentable
124. See,e.g.,Gugliuzza&Lemley,supranote36,at780(observingtheuncertain-tiesinthejudicialsystembyempiricallyreviewingthecasedecisionsmadebythefed-eralcircuitsafterAlice). 125. JoeMullin,ManyPatent-HoldersStopLookingtoEastTexasFollowingSupremeCourt Ruling, ARSTECHNICA (Oct. 12, 2017, 2:50 PM), https://arstechnica.com/tech-policy/2017/10/patent-cases-in-east-texas-plunge-more-than-60-percent [https://perma.cc/JXN4-XNJL]. 126. Id. 127. MarkCurriden,PatentFilingsPlummetinEastTexas,CHRON(May22,2018,5:30 AM), https://www.chron.com/business/article/Patent-filings-plummet-in-East-Texas-12932436.php[https://perma.cc/793P-GWNX]. 128. Dani Kass, IP Litigation More Costly, Risky Than Ever Before, MoFo Says,LAW360 (Aug. 8, 2019, 9:25 PM), https://www.law360.com/articles/1186755/ip-litigation-more-costly-risky-than-ever-before-mofo-says[https://perma.cc/V9F2-SPGB](“Theactualnumberofsuitsisdiminishingthough,whichthefirmattributedinpart to theU.S.SupremeCourt’sTCHeartland andAlicedecisions,bothofwhichmadeitharderforplaintiffsinlitigation.”). 129. RobertSachs,#AliceStorminJune:ADeeperDiveintoCourtTrends,andNewData on Alice Inside the USPTO, BILSKI BLOG (June 30, 2015), https://www.bilskiblog.com/2015/06/alicestorm-a-deeper-dive-into-court-trends-and-new-data-on-alice-inside-the-uspto[https://perma.cc/XB5H-E6XY]. 130. Sachs,supranote38. 131. Id.;Sachs,supranote129. 132. Gugliuzza&Lemley,supranote36,at782.
548 MINNESOTALAWREVIEW [105:527
subject matter in cases appealed from the district courts and thePTAB.133Manyof thosedecisionsarenon-precedential.134In thirty-threeprecedentialopinionsoutofthetotal104decisionsreviewed,onlysevenopinions(21.2%)foundthepatentatissuetobevalid.135Ofthe104totaldecisions,patentsinonlyeightofthedecisions(7.7%)wereallowedtosurvivebytheFederalCircuit.136
Thetechnologiesinvolvedinthose104caseswereeitherinfor-mationtechnology(IT)orbiotechnology.137GugliuzzaandLemleyob-served that biotechnology ismore likely to survive eligibility chal-lenges post-Alice compared to IT.138This finding is consistentwithanotherstudyinwhichLemleyandZyontzreviewed808decisionsonpatentablesubjectmatterdeliveredbytheFederalCircuitandthefed-eral district courts. 139 There, federal courts invalidated patents in65.1%ofthe724softwareorITcases,butonlyinvalidated50%oftheseventy-sixbiotechnologyorlifesciencecases.140
Theaboveresults,however,donotmeanthatwhenitcomestobiotechnology,federalcourtsareprovidedclearguidanceunderAlice.AnempiricalstudybyLemleyandZyontzfoundthatbiosciencepa-tentsfaredbetterinthecourtswithrespecttoeligibility.141Lookingintotheprocess(i.e.,theAlice,Myriad,andMayodecisions)forhowcourts determine eligibility in biotechnology,142 Rebecca Eisenbergfoundthatthepolicyimplicationsofrestrictionsonpatenteligibilityareunclear.143EugeneKimargues thatalthough theFederalCircuitdecisioninCellzDirecthelpsbiotechnologypatentsnotdirectedatdi-agnostics,therearesignificantuncertaintiesinthedecisionsregard-ingdiagnosisandthetreatmentofdisease.144
Forartificialintelligenceandbigdata-relatedinnovations,AliceandsubsequentdecisionsbytheFederalCircuithavecastdoubtonwhethergrantedpatentsandnewpatentapplicationscansatisfythestill-evolvingAlice test forpatent eligibility.145PractitionersbelievethatAlicehasdramaticallyreducedthevalueofissuedpatentsinpar-ticulartechnologiesandchangedhowpatentapplicationsaredraftedand prosecuted.146That said, some practitioners are confident thattheuncertainties imposedbyAlicewilleventuallydiminishthroughevolving court decisions or new congressional legislation. 147 EventhoughAIandbigdatainnovationcanstillbeprotectedwithpatents,theuncertaintiessuggestthatpatentsarenotthebestmechanismtoprotecttheseinventions.148
InEnfish,LLCv.MicrosoftCorp.,acasethatispost-Alice,theFed-eralCircuitheldthat“softwarecanmakenon-abstractimprovementstocomputertechnologyjustashardwareimprovementscan.”149JerrySuvabelievesthatEnfishisamanifestationandapplicationofAlice’slegalclaim,whichfurtherclarifiesthatimprovementstoatechnolog-icalprocessortothefunctioningofthecomputeritselfarepatenta-ble.150However, even though the Federal Circuit has found severalcomputerpatentstobeeligibleinpost-AlicedecisionssuchasEnfishandBASCOM,151Kimcautionsthatthesamehasnotyethappenedinbiotechnologycases.152Evenworse, ina recent case,ElectricPowerGroupLLCv.AlstomS.A.,theFederalCircuitexpandedthefirststepoftheAlice test,holding that “collecting information, analyzing it, anddisplayingcertainresultsofthecollectionandanalysis”isanabstractideaoracombinationof“abstract-ideaprocesses.”153
in decisions regarding diagnosis and treatment of disease.”); Rapid Litig. Mgmt. v.CellzDirect,Inc.,827F.3d1042(Fed.Cir.2016). 145. DouglasH.Pearson,OgnianV.Shentov,CarlA.Kukkonen,AndrewWeissJef-fries&PatrickT.Michael,ProtectingArtificial IntelligenceandBigData InnovationsThrough Patents: Subject Matter Eligibility, JONES DAY (Mar. 2018), https://www.jonesday.com/protecting-artificial-intelligence-and-big-data-innovations-through-patents-subject-matter-eligibility-03-12-2018[https://perma.cc/Q7U9-D9R2]. 146. Id. 147. Id. 148. Id. 149. Enfish,LLCv.MicrosoftCorp.,822F.3d1327,1335(Fed.Cir.2016). 150. JerrySuvaII,SlaydenGrubertBeardPLLC,CLEPresentationattheStateBarofTexasAdvancedIntellectualPropertyLawCourse:PatentableSubjectMatterUp-datefromtheFederalCircuit(Feb.23,2017). 151. BASCOMGlob.InternetServs.,Inc.v.AT&TMobilityLLC,827F.3d1341(Fed.Cir.2016). 152. SeeKim,supranote31,at1181. 153. Elec.PowerGrp.LLCv.AlstomS.A.,830F.3d1350,1353–54(Fed.Cir.2016).
550 MINNESOTALAWREVIEW [105:527
The Federal Circuit has found it difficult to consistently deter-minepatent-eligibility.InAthenaDiagnostics,Inc.v.MayoCollabora-tiveServices,LLC,theFederalCircuitreleasedeightseparateopinionsregardingtheissueofsubjectmattereligibility:fourconcurredwiththedenialoftheenbancpetitionandfourdissentedfromthatdeci-sion.154Somedissentingjudgesarguedthatpatentsondiagnostickitsand techniques should be protected for inventiveness.155 However,somejudgesintheirconcurrencesinvitedtheSupremeCourtorCon-gress to fix the law governing patent eligibility.156Those disparateopinionsbroadlysuggestthattheFederalCircuitjudgesagreethatAl-iceandMayocreatedconfusion.157However,theU.S.SupremeCourthas repeatedlydeclined to re-visit the topicofeligible subjectmat-ter.158
3. UncertaintiesatthePTOThePTABhears appeals frompatent applicants engaged in ex
parte prosecution, post-issuance patent validity challenges filed bypetitioners,orthroughthepre-AIAexpartereexaminationsystem.159ThecurrentsystemreliesheavilyontheFederalCircuittoreviewthePTAB’sdecisionsonpatentvaliditychallenges,andtheinteractionbe-tweentheFederalCircuitandthePTABcanbringuncertaintiesfromthecourtstothePTAB.160Post-issuance,eligiblesubjectmatterchal-lengesunder§101canberaisedthroughmechanismssuchaspost-
154. Athena Diagnostics, Inc. v. Mayo Collaborative Servs., LLC, 927 F.3d 1333(Fed.Cir.2019). 155. Id.at1362. 156. Id.at1337. 157. Athenav.Mayo:ASplinteredFederalCircuitInvitesSupremeCourtorCongresstoStepUpon101Chaos,IPWATCHDOG(July8,2019),https://www.ipwatchdog.com/2019/07/08/splintered-federal-circuit-invites-supreme-court-review-athena-v-mayo/[https://perma.cc/4TJE-UYXL]. 158. E.g.,Athena Diagnostics, 915 F.3d 743, cert. denied, 140 S. Ct. 855 (2020);Berkheimer v. HP Inc., 881 F.3d 1360 (Fed. Cir. 2018), cert. denied, 140 S. Ct. 911(2020);VandaPharms.Inc.,v.West-WardPharms.Int’lLtd.887F.3d1117(Fed.Cir.2018),cert.denied,140S.Ct.911(2020);PowerAnalyticsCorp.v.OperationTech.Inc.,748F.App’x334(Fed.Cir.2019),cert.denied,140S.Ct.910(2020);CellspinSoft,Inc.v. Fitbit, Inc., 927 F.3d 1306 (Fed. Cir. 2019), cert. denied, 140 S. Ct. 907; (2020);ChargePoint,Inc.v.SemaConnect,Inc.,920F.3d759(Fed.Cir.2019),cert.denied,140S. Ct. 983 (2020);TradingTechs. Int’l, Inc. v. IBGLLC,767F.App’x1006 (Fed.Cir.2019),cert.denied,140S.Ct.955(2020);TradingTechs.Int’l,Inc.v.IBGLLC,921F.3d1084(Fed.Cir.2019),cert.denied,140S.Ct.954(2020);SRIInt’l,Inc.v.CiscoSys.,773F.App’x1090(Fed.Cir.2019),cert.denied,140S.Ct.1108(2020). 159. 35U.S.C.§6(a)–(b);Gugliuzza&Lemley,supranote36,at783–84. 160. RochelleCooperDreyfuss,GivingtheFederalCircuitaRunforItsMoney:Chal-lengingPatentsinthePTAB,91NOTREDAMEL.REV.235,258(2015).
2020] IMPLICATIONSOFALICE 551
grantreviews(PGRs)andcoveredbusinessmethodreviews(CBMs)atthePTAB.161Interpartesreviews(IPRs),amechanismwithinthePTAB to challengepatent validity, are farmorenumerous than theothermechanisms that challengepatent validity, such asPGRs andCBMs,andfarmoreprevalentthaninitiallypredictedbythePTO,butIPRs cannot be employed to raise subject matter eligibility chal-lenges.162About87%of thePTABpetitionschallengedpatents thatwerebeingenforcedindistrictcourtafterapre-suitinvestigation.163Under these circumstances, the former Federal Circuit Chief JudgePaulMichelbelievesthatAliceimposesmassiveuncertaintyoverthevalidityofcountlessthousandsofpatents,mostofwhichwereissuedlong before Alice or evenMayo.164 Likewise, Federal Circuit JudgeToddHughescontendsthatsuchuncertaintiesharmtheU.S.patentsystemandinnovationecosystem.165
TheseuncertaintiesmaynotbeeliminatedinthedualsystemsofthefederalcourtsandthePTO,includingthePTAB.PaulGugliuzzaex-plainsthatthejudicialsystemandthePTABadaptdifferentstandardsofproofanddifferentrulesofclaimconstruction(atleastforthepe-riodoftimethatisthefocusofthisstudy).166Gugliuzzacriticizedthedual proceedings for increasing litigation costs and incentivizing“wastefulproceduralmaneuvering,”therebyexacerbatingtheuncer-tainties.167
thatprocess.168SamuelHayimandKateGaudrystudied500exparteappealdecisionsforappealsfiledpost-Alicethatwerewithinthetech-nologyareasofTechnologyCenters(TCs)2100,2400,2600,andbusi-nessmethodsandthatwererenderedbythePTABinthetwoyearsafterAlice.169They found that amere 16% of the initial rejectionsbasedonpatenteligibilitywere“fullyreversed”bythePTAB.170
AlthoughitisunclearwhetherthePTABshouldadoptthesamecriteriaascourts,RochelleDreyfussbelievesthatthePTABcanfur-nish a blueprint for clarifying the uncertainties because a narrow,clearscopeofpatentclaimscanreducethelitigationandtransactioncostsimposedbypatenttrolls.171Moreover,basedontheirreviewofalargenumberofFederalCircuitdecisions,GugliuzzaandLemleypre-dictthattherewillbenoincreaseinthepercentageofpatentsbeingupheldinappealsfromthePTABbecauseofthepeculiaritiesoftheadministrativeprocess.172TheybelievethatitisunlikelythattheFed-eralCircuitwillhearmanyappealschallengingthePTABrulingsthatconfirmpatentvalidity.173
4. PreviousEmpiricalStudiesofOfficeActionsbythePTOBesides the summary statistics that are published on various
InChienandWu’sempiricalstudy,theycountedthenumbersofpatentapplications,§101rejectionsimposedbythePTO,andpatentapplication abandonments by applicants. 180 However, directly ob-servingthefluctuationofthesenumberscannotsupportaconclusionthattheabandonmentswerecausedbyAlice.Theyalsodidnotindi-catehowpatentapplicantsadjustedtheirfilingandprosecutionstrat-egiespost-Alice.181OurempiricalresearchdesignwithD-i-DanalysispresentedinthisArticleobserves(1)whetherAlicewasacauseofthe§101rejectionsandthedecreaseinthenumberofpatentapplicationsand(2)whetherpatentapplicantsadjustedtheirfilingandprosecu-tionstrategiespost-Alice.
arsaresplitonwhethertheseguidelinesmaybeeffectiveineliminat-ingtheuncertaintiesarisingfromapplicationsoftheAlicetestinthecourts.182TranispersuadedthatthePTOgrantssoftwarepatents ifmeaningful limitations go beyond generally linking the use of an
176. Seegenerallyid. 177. Seeid.at16–17. 178. Seeid.at17(“101isplayinganincreasinglyimportantroleintheexaminationofsoftwareandmedicaldiagnosticspatents....[T]hevastmajorityofinventionsex-aminedbytheofficearenotsignificantlyimpactedby101.”). 179. Sachs,supranote129. 180. SeeChien&Wu,supranote39,at14(describingtheirmethodologyasade-scriptiveanalysis, rather thanadiff-in-diff regressionanalysis, even though theau-thorstermedtheirworkaD-i-Dstudybecauseadiff-in-diffanalysisrequiresaniden-tificationstrategy,whichtheworklacks). 181. Id. 182. CompareTran,supranote168,withKenyon,supranote88.
554 MINNESOTALAWREVIEW [105:527
abstract idea to a particular technological environment.183 By con-trast,BrooksKenyoncriticizesthePTO’sinternalguidelinesbecausetheyonlymirrortheFederalCircuit’sdecisionsonsoftware,resultinginanalarminglyhighrejectionrate.184Kenyonpredictsthatexamin-erswillhesitatetoissuepatentclaims,andsuchhesitationispremisedontheguidelinesandordersfromtheirsupervisorsintheexaminingcore.185Indeed, since theAlice decision, patent examiners have re-jectedastaggeringnumberofpatentapplicationsindifferenttechnol-ogyareasunder§101.186
Inordertoprovideclearguidancetopatentexaminersforevalu-atingsubjectmattereligibility,thePTOissuedRevisedPatentSubjectMatter Eligibility Guidance (Revised Guidance) in January 2019.187TheRevisedGuidancedoesnotrevisetheearlierInterimGuidanceforStep1,whichpertainstowhetherthepatentclaimfallswithinastat-utorycategory.188ItonlyrevisesSteps2Aand2BandtriestoclarifythejudicialexceptionsrelatedtoAliceandMayo.189
TherevisedStep2Arequiresmorethanaskingwhetherthepa-tentclaimsareabstractideasornotsoastofallwithinajudicialex-ception,asoutlinedintheearlierStep2A.190Itisnowatwo-prongtestthatinvolvesallowingpatentclaimsthatreciteajudicialexceptionifthejudicialexceptionisthenintegratedintoapracticalapplication.191Becauseofthissecondprong,examinersareinstructedtogiveweighttoalladditionalelementsintheclaim,includingwhethertheyarecon-ventionalwhenevaluatingwhether the judicialexceptionsare inte-gratedintoapracticalapplication.192
ThegoalofStep2BisclarifiedintheRevisedGuidancetofocuson evaluating whether the patent claims provide an inventive
183. Tran,supranote168,at537,541–42. 184. Kenyon,supranote88,at4–5. 185. Id.at5. 186. MichaelStein,USPTOUrgedtoReviseInterim§101GuidancetoRequireExam-iners to Present a Proper Prima Facie Case Supported by Factual Evidence, BAKERHOSTETLER: IP INTEL. (Mar. 23, 2015), https://www.ipintelligencereport.com/2015/03/23/uspto-urged-to-revise-interim-%C2%A7101-guidance-to-require-examiners-to-present-a-proper-prima-facie-case-supported-by-factual-evidence[https://perma.cc/R6R7-ZHUV];seealsoSachs,supranote129. 187. 2019RevisedPatentSubjectMatterEligibilityGuidance,84Fed.Reg.50(Jan.7,2019). 188. Id.at54. 189. Id.;seealsoMayoCollaborativeServs.v.PrometheusLab’ys,Inc.,566U.S.66(2012). 190. 2019RevisedPatentSubjectMatterEligibilityGuidance,84Fed.Reg.at54. 191. Id. 192. Id.at55.
2020] IMPLICATIONSOFALICE 555
concept.193Aninventiveconceptmaybepresentedwhentherearead-ditionalelementsthatadd limitationsthatarenotwell-understood,routine,conventionalactivitiesinthefield.194Itisnotenoughtopre-sent an inventive concept if the additional elements only generate“well-understood,routine,conventionalactivitiespreviouslyknowntotheindustry...atahighlevel.”195
Table1 shows thenumberofofficeactionsbetween2012and2016anddisclosesthespecificnumbersofrejectionsunderthedif-ferentstatutoryrequirementsandbasedonAlice.Table2showsthenumber of office actions by technology areas, addressing selectionbias concerns and supporting the robustnessof the empirical anal-yses. Patent applications are reviewed by patent examiners in
Thisstudyconductsanempiricalanalysisforthreedifferenttech-nologysectorsmostdirectlyaffectedbytheAlicedecision—bioinfor-matics,businessmethods,andsoftware.208Italsoincludessub-cate-gorieswithin those technology areas (e.g., specific art unitswithinthese three sectors).209 In the bioinformatics sector, this study ex-plorespatentapplicationsundertwomeasures.Itemploysanarrowdefinitionofbioinformaticsbychoosingspecificartunits1631and1639andabroaddefinitionofbioinformaticsunderthemoregeneralartunit1630.210Thisstudyalsoexplorespatentapplicationsdirectedatbusinessmethods,asdefinedinTC3600.211IttestssomespecificartunitswithinTC3600,includingartunits3620and3680fore-com-merceandartunit3690forfinance.212E-commerceisfurtherdividedintospecificartunitsforhealthcareandcryptography.213Finally,thisstudyexploressoftwarepatentapplications,broadlydefinedasthose
206. Patent Classification, U.S. PAT.& TRADEMARKOFF., https://www.uspto.gov/patents-application-process/patent-search/classification-standards-and-development[https://perma.cc/3BX6-CHTC]. 207. Id. 208. SeeinfraPartsII.A.2,II.B. 209. SeeinfraPartsII.A.2,II.B. 210. See infra Parts II.A.2, II.B; seealsoClassesArrangedbyArtUnit,U.S.PAT.&TRADEMARKOFF.,https://www.uspto.gov/patents-application-process/patent-search/understanding-patent-classifications/patent-classification[https://perma.cc/3BX6-CHTC](outliningartunit1631referstodataprocessingandartunit1639referstocombinatorialchemistrytechnology);TC1600ManagementRoster,U.S.PAT.&TRADEMARKOFF.,https://www.uspto.gov/patent/contact-patents/tc-1600-management-roster[https://perma.cc/VP5P-8KXV](explainingthatartunit1630re-ferstothetechnologysectorsofmolecularbiology,bioinformatics,nucleicacids,re-combinant DNA and RNA, gene regulation, nucleic acid amplification, animals andplants,andcombinatorial/computationalchemistry).SeegenerallyPatentClassifica-tion,U.S.PAT.&TRADEMARKOFF.,https://www.uspto.gov/patents-application-process/patent-search/classification-standards-and-development[https://perma.cc/3BX6-CHTC](explainingthedefinitionofartunitsreliesontheclassificationpro-videdbythePTO). 211. SeeinfraPartsII.A.2,II.B. 212. See infra Parts II.A.2, II.B; seealsoTC3600ManagementRoster,U.S.PAT.&TRADEMARKOFF.,https://www.uspto.gov/patent/contact-patents/tc-3600-management-roster[https://perma.cc/8MN6-P3XX]. 213. See infraParts II.A.2, II.B; seealsoTC3600ManagementRoster, supranote212.
ice over all office actions issued to patent applications in differenttechnologysectorsafterAlice.Figure2belowshowstheproportionof§101rejectionsoverallofficeactionsindifferenttechnologysectors.Thetwofiguresshowthefrequencywithwhichpatentexaminersre-jectedpatentapplicationsforpatentineligibilityunderAliceor§101.Figures1and2showthatpatentapplicationsinbioinformatics(un-derthenarrowdefinition)receivedbothmorerejectionsthatcitedAl-iceasareason(i.e.,Alicerejections)andmore§101rejectionscom-pared to patent applications in business methods or software.However, the narrower categories in businessmethods for financeande-commercereceivedmoreAlicerejectionsand§101rejectionscomparedtoallothertechnologyareas.
Inbusinessmethods,13.87%oftheofficeactionsforapplicationsfiledbeforetheAlicedecision,butexaminedaftertheAlicedecision,citedAlice and imposed a rejection. This rate increased slightly to15.21%forapplicationsfiledbeforeAlicebutexaminedafterthePTOimplementation ofAlice. The rejection rate reduced significantly to
0.00%
10.00%
20.00%
30.00%
40.00%
50.00%
60.00%
Bio-Inf
ormati
cs
Busine
ss M
ethod
s
Busine
ss M
ethod
s of F
inanc
e
Busine
ss M
ethod
s of E
-Com
merce
Software
(in g
enera
l) AI
Data B
ases a
nd File
Man
agem
ent
Cryptog
raphy
and S
ecurit
y
Compu
ter N
etwork
s
Teleco
mmunica
tions
Digital
Cam
eras
Compu
ter G
raphic
Process
ing
Telemetr
y and
Cod
e Gen
eratio
n
Rejections based on Alice as a fraction of all office actions for patent applications filed before Alice
Rejections based on Alice as a fraction of all office actions for patent applications filed after Alice
Softwarereceivedrelativelyfewer§101rejectionsandAlicere-jectionscompared tobothbusinessmethodsandbioinformatics. Inthegeneralsoftwaresector,1.04%oftheofficeactionsincludedrejec-tionsunderAliceforapplicationsfiledbeforeAlicebutexaminedafterAlice.Thisrateincreasedslightlyto1.17%forapplicationsfiledbeforetheAlicedecisionwasimplementedbythePTO.Therejectionratein-creasedto1.7%forsoftwareapplicationsfiledafterAliceanditsim-plementationbythePTO.BeforeAlicewasdecided,10.75%ofallof-fice actions for software applications were § 101 rejections. AfterAlice, 10.93%of all office actions for applications filed beforeAlicewere§101rejections,whichthenincreasedslightlyto12.98%forap-plicationsfiledafterAlice.Moreover,6.47%offinaldecisions(e.g.,al-lowancesorrejections)forsoftwareapplicationsincluded§101re-jectionsbeforetheAlicedecision.Thisratewentupslightlyto9.73%afterAlicewasdecidedforapplicationsfiledbeforeAliceandto9.83%forapplicationsfiledafterAlice.
Patentapplicationsinthevarioussub-categorieswithinsoftwarereceivedAlicerejectionsand§101rejectionstovaryingdegrees.Inthespecificsoftwareartunits, theunit thatreceivedthemostAlicerejectionswascomputernetworks,inwhich3.31%ofallofficeactionsforpatentapplicationsfiledbeforeAlicewasdecided,butexaminedafterAlice,wereAlicerejections.Thisrateincreasedto5.66%forap-plicationsfiledaftertheAlicedecision.EighteenpercentofallofficeactionsforapplicationsincomputernetworksfiledbeforeAlicewasdecided,butexaminedafterAlice,were§101rejections.Thisratein-creased dramatically to 26.5% for applications filed post-Alice. Incryptography and security, 2.55% of office actions for applicationsfiledbeforeAlice,butexaminedafterAlice,wereAlicerejections,andthisincreasedto5.14%forapplicationsfiledpost-Alice.Eighteenper-centofallofficeactionsforapplicationsincryptographyandsecurityfiledbeforeAlicewasdecided,butexaminedafterAlice,were§101
562 MINNESOTALAWREVIEW [105:527
rejections,which then increased to 24.39% for patent applicationsfiledafterAlice.
Moreover, patent applications in computer networks, GUI anddocumentprocessing,databasesandfilemanagement,cryptographyandsecurity,andcomputergraphicsprocessingreceivedahighper-centage (about8–10%)of final rejectionsunder §101beforeAlicewasdecided.Therateincreasedto19.32%post-Aliceforapplicationsin cryptography and security filed before Alice and increased to22.53%forapplicationsfiledpost-Alice.Therateincreasedto18.18%after Alice was decided among applications in computer networksfiledbeforeAlice and increased to23.28% for applications in com-puter graphics processing filed post-Alice. Compared to these in-creasedpercentagesofrejections,patentapplicationsinGUIanddoc-ument processing, computer architecture, telecommunications, andrecordingandcompressiondidnotreceivemorefinalrejectionsun-der§101afterAlice.
Inabroadview,thenumberofpatentapplicationsdecreasedinbroadandnarrowbioinformatics,businessmethods,andsoftwareingeneral,asshowninFigure3above.BeforeAlice,therewere,onaver-age,392patentapplicationsfiledpermonthinbroadbioinformatics,withinwhichninetypatentapplicationswereinnarrowbioinformat-ics. AfterAlice, therewere about 339 patent applications filed permonth in broadbioinformatics, and thepatent applications filed in
0
2000
4000
6000
8000
10000
12000
14000
16000
18000
2012 2013 2014 2015 2016 2017
Time (monthly)Manufacturing Broad Bio-Informatics Narrow Bio-Informatics Business Methods Software (in general)
Decision Date Implementation Date
2020] IMPLICATIONSOFALICE 563
narrowbioinformaticsdecreasedto81.Businessmethods-patentap-plicantsfiled4,206patentapplicationspermonthonaveragebeforeAlicebut filed3,843patentapplicationspermonthafterAlice.Soft-ware-patent applicants, on average, filed 8,780 patent applicationspermonthbeforeAlicebutfiled7,910patentapplicationspermonthafterAlice.Thetotalpatentapplicationsalsodecreasedby10.64%af-terAlice.From2012to2016,about26%ofthetotalpatentapplica-tionswerefiledinsoftware.About12%wereinbusinessmethodsandabout1%wereinbroadbioinformatics.
foreandafterAlicebydifferenttimewindowsbecausethemarketandpatentapplicantsneededtimetoreacttoAliceandadjusttheirpatentstrategies.ComparingtheaveragenumberofpatentapplicationsfiledpermonthoneandahalfyearsafterAliceandoneyearaftertheAliceimplementationbythePTOtotheaveragenumberofpatentapplica-tions filedpermonthbeforeAlice,patentapplications inbroadandnarrow bioinformatics, businessmethods, software in general, andtheoverallindustrydecreasedatahigherlevelcomparedtotheear-liercomparisons.Thetotalpatentapplicationsfiledpermonth,onav-erage,fell29.56%afterAlice.Whilethe20.08%decreaseinpatentap-plications inbroadbioinformaticswas lower than thedrop in total
Even though the language of Alice does not directly address§§102,103,or112,thesestatutesarerelevanttothepatentabilityofan invention andare implicatedby theAlice decision.216Therefore,thePTOrejectionsunderallfourstatutesshouldbeindividuallycon-trolledintheregressionanalysisasindependentvariables.Sincethepresenceorabsenceofeachstatutorycategoryinanofficeactionisbinary(zeroorone),thisstudyfirstdeployslogisticregressionsandincludes statutory rejections as independent variables to estimatewhetherarejectionunderAlicewasissued.
In logisticregressions,weobservetheassociationbetweenthepresenceofAlicerejectionsandanyoneofthefourcategoriesofstat-utoryrejections(i.e.,rejectionsbasedon§§101,102,103,or112)asaninitialorfinalrejectionwhentakingallofficeactionsregardingallfourstatutesintoaccount.Itisimportantforlogisticmodelstocontrolformonthandfortechnologycentersorartunitsasbeingfixed.Weaddedfixedcontrolsbecauseallthesefactorscouldbedirectorindi-rectreasonsforstrengtheningtheassociationbetweentheAlicerejec-tionsandanyonetypeofstatutoryrejection.217
b. Difference-in-DifferenceAnalysesWehaveobservedavariationin§101rejections(whichareei-
therinitialrejectionsorfinalrejections)amongallofficeactionsbe-foreandaftertheAlicedecision.218Specifically,moreinitialandfinalrejections were given by examiners under § 101 and Alice for
The interventionof theAlice decision is consideredunder twodates.Onedateisthemonth(June2014)whentheopinionwasdeliv-eredby theU.S. SupremeCourt,220and theotherdate is themonth(December2014)whenthePTOintroducedtheInterimGuidanceim-plementing the Alice decision.221 The two dates are separately de-ployedintheD-i-Dregressions.Theregressionresultsbetweenthetwodatessuggestadifferenceinhowtheopinionandimplementationeventsaffectedpatentapplicationsandtheofficeactionsissuedbypa-tentexaminers.
Weselectedthepatentapplicationsinmanufacturingdevicesandprocesses,machinetools,andhandtoolsinartunits3722–3727andinartunit3729asthecontrolgroup.Asthecontrolgroup,weexaminewhethertheyarelesslikelytobeaffectedbytheAlicedecision.Spe-cifically,thecontrolgroupofpatentapplicationsinmanufacturingde-vicesconsistentlyreceivedaverysmallnumberof§101rejections,whichwereatmost3.7%ofallofficeactionspermonthand0.06%ofallofficeactionsonaveragepermonthduringtheentireperiodfrom2012to2016.Wecomparedthiscontrolgroupwithourstudyobjectsof patent applications in business methods, bioinformatics, and
software. Patent applications in these three technology areaswerethenconsideredas independent treatedgroups.222All these treatedgroupsreceivedmuchmoreinitialandfinal§101rejectionsbetween2012and2016thanthecontrolgroupandcomparedtoothertech-nologyareasoutsidethescopeofourdataselection.Figure4belowshowsthespecificmonthlyvaryingtrendsintheinitialandfinal§101rejections for the control group of manufacturing and the treatedgroupsofbusinessmethods,bioinformatics, andbusinessmethods.Notonlydidthefourgroupsreceive§101rejectionsintheirinitialandfinalPTOdecisionsatdifferentlevels,butthetreatedgroupsalsoreceivedmanymore§101rejectionsintheirinitialandfinaldecisionsgivenbythePTOafterAlicewasdecided.
Groupdifferenceisabinaryvariable,wherezerorepresentsthecontrolgroupandonerepresentsthetreatedgroup.Timedifferenceisalsoabinaryvariable,whichcontrolsthetimepriortotheAlicede-cisionaszeroandthetimepost-Alice(decisionorimplementation)asone. InD-i-Dregressions, thecoefficientof the interactiontermbe-tweengroupdifferenceandtimedifferencesurrogatesaD-i-Deffect.Inamodeltoestimatetheprobabilityofreceivinga§101rejectionintheofficeactionsgivenbythePTO,apositiveD-i-Deffectorapositivecoefficient for the D-i-D effect with statistical significance suggeststhatAliceinducedagreaternumberof§101rejectionsforthetreatedgroupintheirinitialandfinaldecisionsgivenbythePTO.
This study deploys logistic regressions to explore the correla-tionsbetweentheAliceofficeactionrejectionsandthedifferentstat-utoryrejections.Section101rejectionsforalloftheartunitsforbio-informatics, business methods, and software were positivelycorrelatedtoAlicerejectionsatastatisticallysignificantlevel.Thissta-tisticalsignificancemeansthatpatentapplicationsthatwerefiledin
TheartunitwiththestrongestassociationbetweenAlicerejec-tionsand§101rejectionsamongalltechnologyareaswasdigitalandopticalcommunication.224Digitalandopticalcommunicationwasanart unit where patent applications that received an Alice rejectionweremostlikelytoberejectedunder§101ininitialandfinaldeci-sions,comparedtootherartunitsthatalsoreceivedAlicerejections.Patentapplicationsincomputergraphicsprocessing,databasesandfilemanagement,cryptographyandsecurity,computernetworks,dig-italcameras,telemetryandcodegeneration,softwareingeneral,bio-informatics,andbusinessmethodsalsohadaverystrongassociationbetweenAlicerejectionsand§101rejections.Eventhoughartunitsforspecificcategoriesinbusinessmethodsforfinance,e-commerce,healthcare,andcryptographyalsohadastrongpositiveassociationbetweenAlicerejectionsand§101rejections,theassociationwasnotasstrongastheassociationingeneralbusinessmethodsorotherpre-viouslydiscussedartunitsforthespecificcategoriesinsoftware.
Therecordingandcompressionartunithadtheweakestpositiveassociation betweenAlice rejections and § 101 rejections. In otherwords, patent applications in recording and compressionwere theleastlikelytosimultaneouslyreceivemoreAlicerejectionsandmore§101rejections intheir initialor finaldecisionscomparedtoothertechnologyareas.Amongallofthetestedartunits,theonlyartunitwhereallcategoriesofstatutoryrejectionswerepositivelycorrelatedtoAlicerejectionswasdatabasesandfilemanagement.Thismeansthatpatentapplicationsindatabasesandfilemanagementwereal-waysmorelikelytoberejectedunderAliceregardlessofthetypeofstatutoryrejectionsthatwereissued.
In most of the other art units in the three technology areas,§§102,103,or112rejectionswereeithernegativelycorrelatedtoAl-icerejectionsornotcorrelatedtoAlicerejectionsatastatisticallysig-nificantlevel.Thissuggeststhatpatentapplicationsintheseartunits
Table 4 shows the correlations between § 101 rejections and
otherstatutoryrejectionsthatwerereceivedintheinitialorfinalde-cisionsandaretheindependentvariablesdeployedinthelogisticre-gressions.Notonlycouldthesestatutoryrejectionsshareanunderly-ingrationaleaddressedintheAlicedecisionitself,butthecovariationsamongthestatutoryrejectionshelpuslocatethetrueassociationbe-tweenthemandtheAlicerejections.Whenthecovariationorcolline-arityishigh,differentindependentvariablesmayrepresentthesamestatisticalinformationandneednotbeindependentlyexplored.225Re-garding§112, thisstudyspecifically focuseson§112(a)rejections(i.e., written description and enablement of specification 226 ) and§112(b)(i.e.,definitenessofclaims227)rejections.Inourdataofallof
InalltestedtechnologyareaslistedinTable1otherthanmanu-facturingdevices (the controlgroup), the frequencyof§101 rejec-tionswaspositivelycorrelatedwiththe frequenciesof§§102,103,and112(b)rejections,regardlessofwhethertheofficeactionswereissuedbeforeorafterAlice.Inotherwords,apatentapplicationthatwasrejectedunder§101wasalso likelytoreceiveanother§§102,103, and 112(b) rejection. In bioinformatics, narrowly defined, thecorrelation between § 101 rejections and § 112(b) rejections wasmuch stronger than the correlation between § 101 rejections and§102rejections.Inothertechnologyareas,thelattercorrelationwasmuchstrongerthantheformercorrelation.Patentapplicationsinnar-rowly defined bioinformatics that received a § 101 rejection weremorelikelytoreceiveonemore§112(b)rejectionthantoreceiveonemore§102rejection.Bycontrast,patentapplicationsinothertech-nologyareasofbroadlydefinedbioinformatics,businessmethods,orsoftwarethatreceiveda§101rejectionweremorelikelytoreceiveonemore§102rejectionasopposedtoreceivingonemore§112(b)rejection.
Thecorrelationbetween§101rejectionsand§112(a)rejectionswasweakerthanthecorrelationbetween§101rejectionsandotherstatutoryrejections(i.e.,§§102,103,and112(b)),exceptfortheofficeactionsforbusinessmethodpatentapplicationsafterAlice.Itmeansthateventhoughpatentapplicationsthatreceiveda§101rejectionwere likely to receiveonemore§112(a) rejection, thisprobabilitywas lower thantheprobabilityofsimultaneouslyreceivinganotherstatutoryrejectionotherthan§112(a).Amongtheofficeactionsis-sued after Alice, the correlation between § 101 and § 112(a) wasstrongerthan(1)thecorrelationbetween§101rejectionsand§102rejections and (2) the correlation between § 101 rejections and§112(b)rejections.
Thecorrelationbetween§101rejectionsand§112(a)rejectionsvariedamongtechnologyareas.Insomesub-categoriesoftechnologyareas,includingbusinessmethodsoffinance,AI,andcomputerarchi-tecture, there was no correlation between § 101 rejections and§112(a)rejectionsatastatisticallysignificantlevelamongtheofficeactions that were issued either before or after Alice. In these
228. Id.§112(e)–(f).
2020] IMPLICATIONSOFALICE 573
technologyareas,wedidnot find thatapatentapplicationrejectedunder§101wasalsolikelytosimultaneouslyreceivea§112(a)re-jection. Innarrowlydefinedbioinformatics,digitalandopticalcom-munication,computernetworks,digitalcameras,andtelemetryandcodegeneration,therewerepositiveassociationsbetween§101re-jections and §112(a) rejections at a statistically significant levelamongofficeactionsissuedafterAlice.Indatabasesandfilemanage-mentandcryptographyandsecurity,therewerepositiveassociationsbetween§101rejectionsand§112(a)rejectionsatastatisticallysig-nificantlevelamongtheofficeactionsissuedbeforeAlice,whichweremuchweakercomparedtotheassociationbetween§101rejectionsandotherstatutoryrejections(e.g.,rejectionsunder§§101,102,and112(b)).
applications filed before the Alice decision in order to explore thecausaleffectoftheAlicedecisiononexaminers,thecoefficientsfortheinteractiontermintheD-i-Dregressionsarepositiveandstatisticallysignificantinthemodelstoestimatetheprobabilityofinitiallyandfi-nallyreceivinga§101rejectioninbusinessmethods,bioinformatics,andtheartunit fortelemetryandcodegeneration.Whileaparalleltrendofreceiving§101rejectionsbetweenthecontrolgroup,whichreferstopatentapplicationsinmanufacturingdevicesandprocesses,andthetreatedgroupsbeforeAlicecannotbeprovenasshowninFig-ure4andFigureA1,230becauseofthefluctuationsin§101rejectionsreceivedbythetreatedgroups,thedifferenceinthelevelofreceiving§101rejectionsbetweenthecontrolgroupandthetreatedgroupsarestable,suggestingacounterfactualtrendtocomplementtheparalleltrendassumption.231
§101.Therefore, thosepositivecoefficientssuggestthatAlicemadepatent applications in those technology areasmore likely to be re-jectedunder§101intheinitialorfinaldecisions.Meanwhile,thepos-itivecoefficients,interpretedasapositiveD-i-Deffect,surrogateneg-ativeeffectsofAliceonpatentapplications:patentapplicationsweremorelikelytoberejectedduetoAlice.
ThecoefficientsfortheD-i-DeffectoftheimplementationofAlicehaveasimilardegreeofstatisticalsignificanceandsimilarvalueasthecoefficientsfortheD-i-DeffectoftheAlicedecision.Theformercoef-ficientsare slightly stronger than the latter coefficients,which sug-geststhateffectsoftheAlicedecisionanditsimplementationonpa-tentexaminerswereconsistent,andthePTOimplementationoftheAlicedecisionhadaslightlylargereffectonexaminersthantheAlicedecisionitself.
ThecoefficientsfortheD-i-DeffectoftheAlicedecisionareposi-tiveandstatisticallysignificantinthemodelsthatestimatetheprob-abilityofissuing§101rejectionsforbioinformatics,someartunitsforthesub-categoriesinbusinessmethods,andtwoartunitswithrespecttosoftware(e.g.,computernetworksandtelemetryandcode-genera-tiontelemetry).ThepositivecoefficientssuggestthattheAlicedeci-sioncausedpatentapplicationsfiledpost-Aliceinthesetechnologyar-eas to be more likely to be rejected under § 101. Similar to theregressionswithrespecttoapplicationsfiledbeforetheAlicedecision,butexaminedpost-Alice,thecoefficientsfortheD-i-DeffectofthePTOimplementation ofAlice have a similar degree of statistical signifi-canceandvalueasthecoefficientsfortheD-i-DeffectoftheAlicede-cision.ThissimilaritysuggeststhatAlice’seffectwasconsistentastoits impact on examiners reviewing applications filedpost-Alice.WenowturntotheeffectsofAliceonpatentexaminersissuingofficeac-tionsineachtechnologyareaandhowAliceincreasedthelikelihoodofreceiving§101rejections.
a. BioinformaticsThisSubsectionfirstexploresartunits1631and1639withre-
specttobioinformatics,narrowlydefined.Amongalltypesoftechnol-ogieslistedinTable1,narrowlydefinedbioinformaticshasthehigh-estpositivecoefficientfortheinteractiontermbetweenthedatethatAlicewasdecidedandthetwogroupsofofficeactions,onegroupin-cludingtheofficeactionsgivenbeforetheAlicedecisionandtheothergroupincludingtheofficeactionsissuedaftertheAlicedecision,butonlyforapplicationsfiledbeforeAlice.Theprobabilityofinitiallyandfinally rejecting patent applications in narrowly defined
2020] IMPLICATIONSOFALICE 575
bioinformaticsfiledbeforeAliceunder§101increased83%aftertheAlicedecision. It increasedslightlyto84.3%aftertheAlicedecisionwasimplementedbythePTO.Inotherwords,theAlicedecisionmadepatentapplicationsinnarrowlydefinedbioinformaticsfiledbeforeAl-icebutexaminedafterAliceaboutfourtimesmorelikelytoreceivea§101rejectionthannottoreceivea§101rejection.Thislikelihooddecreasedtotwotimeswhenweemployedthedataofartunit1630forbroadlydefinedbioinformatics.
576 MINNESOTALAWREVIEW [105:527
Table5.D-i-DLogitRegressionson§101Rejectionsfor(Narrow)BioinformaticsThemodelsshownbelowarelogisticregressions.Thedependentvariableisthe§101rejec-tion.Itisbinary,soarejectionrefers1andanallowancerefers0.Intheindependentvariables,AliceisacategoricalvariablecontrollingforthetimeperiodbeforeandaftertheAlicedecisionorthetimeperiodbeforetheAlicedecisionandaftertheimplementationoftheAlicedecisionbytheUSTPO.Technologyisacategoricalvariablecontrollingforthecontrolgroupandthetreatedgroup.TimereferstothedecisiondateorimplementationdateofAlice.ThecoefficientontheinteractiontermsurrogatestheD-i-Deffect.Whethertheofficeactionalsogivesa§102,§103,or§112rejectionisindependentlycontrolledasfixedinthemodel.Time(month)iscontrolledasfixedinthemodel.Technologycenteriscontrolledasfixedinmodel1,3,5,to8.James Stock’s Heteroskedasticity-standard errors are shown in parentheses, *** p<0.01, **p<0.05,*p<0.1.
Table5introducestheD-i-Dmodeldesignandshowstheresultsof theD-i-Dregressions fornarrowlydefinedbioinformatics.232ThecoefficientfortheD-i-DeffectforpatentapplicationsfiledafterAliceisslightlysmallerthanthecoefficientwithrespecttothepatentappli-cationsfiledbeforeAlicebutexaminedpost-Alice.Thespecificcoeffi-cientssuggestthattheprobabilityofpatentapplicationsfiledaftertheAlice decision receiving an initial or final § 101 rejection increased79%aftertheAlicedecision,whichisabout16%lowerthanthein-creasedprobabilityof§101rejectionsforapplicationsfiledbeforeAl-icebutexaminedafterAlice.Inotherwords,whiletheAlicedecisionresultedintheprobabilityofinitiallyandfinallyrejectingapplicationsfiledbeforeAlicebutexaminedafterAliceunder§101tobeaboutfourtimeshigherthantheprobabilityofallowingpatentapplicationsun-der§101,theformerprobabilitydecreasedtoalowerdegreeamongapplicationsfiledandexaminedafterAlice.
AmongnarrowlydefinedbioinformaticspatentapplicationsthatwerefiledbeforetheAlicedecision,examinedafterAlice,andrejectedonce under § 101,Alice caused the probability of them receiving a§101rejectionintheirfinaldecisiontobeincreasedbyabout93%.Inotherwords,theAlicedecisionmadethoseapplicationsabouttwelvetimesmore likelyto fail inovercomingtheir initial§101rejectionscompared to successfully overcoming their initial § 101 rejections.Narrowlydefinedbioinformaticspatentapplicationsfiledandexam-inedafterAliceandapplicationsfiledbeforeAlicebutexaminedafterAlicefacedasimilarchallengeinovercominganinitial§101rejection.Alicemadetheseapplicationsabouttwelvetimesmorelikelytofailinovercomingtheinitial§101rejections.Moreover,thecoefficientfortheinteractionterminthemodelwiththePTOimplementationdateastheeventdateislargerthanthecoefficientinthemodeldeployingtheAlicedecisiondate.ThissuggeststhatthePTOimplementationoftheAlice decisionmade the applications filedbeforeAlice nineteentimesmore likelyto fail inovercomingtheir initial§101rejectionsreceivedfromtheexaminers.
Bycontrast,theAlicedecisiondidnothaveastatisticallysignifi-cant effect on how applications in broadly defined bioinformaticsovercametheirinitial§101rejections.However,thePTOimplemen-tationofAlicehadanegative,statisticallysignificanteffectonhowthebroadlydefinedbioinformaticspatentapplicationsfiledbeforeAliceandexaminedafterAliceovercametheirinitial§101rejections.TheimplementationofAlicecausedpatentapplicationsinbroadlydefinedbioinformaticstobethreetimesmorelikelytofailinovercomingtheirinitial§101rejections—asmallereffectthanthenegativeeffectoftheAlice implementation on applications in narrowly defined bioinfor-matics.
b. BusinessMethodsTheD-i-Dmodels for patent applications in businessmethods
likelytoreceiveaninitialorfinal§101rejection.However,thecoeffi-cientfortheD-i-DeffectoftheAlicedecisionwasnegativeandstatis-ticallysignificantinthemodelswithrespecttobusinessmethodap-plicationsfiledaftertheAlicedecision.ThiscoefficientsuggeststhatAlicedidnotinduceanincreaseininitialandfinal§101rejectionsforbusiness method applications filed after Alice. These applications,whichwerefiledandexaminedafterAlice,werestill55%morelikelyto receive an initial or final § 101 rejection compared to businessmethod patent applications filed and examined before Alice. Thatnumberdecreasedto30%whenweappliedtheinteractiontermwiththeimplementationdateoftheAlicedecisionbythePTO.
Moreover,Aliceaffectedtheprobabilityoffailingtoovercometheinitiallyreceived§101rejectionsforbusinessmethodpatentapplica-tionstoincreasebyabout94%.Inotherwords,Alicemadebusinessmethod applications filed before Alice about fourteen times morelikelytofailinovercomingtheirinitial§101rejectionsreceivedfrompatentexaminers.ThecoefficientfortheinteractionbetweenthePTOimplementationdateandthetechnologytypesislargerthanthecoef-ficientfortheinteractionbetweentheAlicedecisiondateandthetech-nologytypes.ThisdifferencesuggeststhatthePTOimplementationofAlicehadastrongereffectonpatentapplicants’failuretoovercometheir initial § 101 rejections. Specifically, the PTO implementationmadebusinessmethodpatentapplicationsfiledbeforetheAlicedeci-sionbutexaminedafterAliceabouttwenty-onetimesmorelikelytofailinovercomingtheirinitial§101rejections.Althoughpatentappli-cationsfiledafterAlicewerealsolesslikelytoovercometheirinitial§101rejections,thoseapplicationswerelesslikelytofailinovercom-ingtheirinitial§101rejectionscomparedtotheapplicationsfiledbe-foreAliceandexaminedafterAlice.TheAlicedecisionmadetheappli-cations filed after Alice about two times more likely to fail inovercomingtheirinitial§101rejections.
In studying the three sub-categories in business methods, wefoundthattheircoefficientsfortheD-i-Deffectaremuchhighercom-paredtonarrowlydefinedbioinformaticsorbusinessmethodsingen-eral.Moreover,theeffectsoftheAlicedecisiononapplicationsfiledbeforeAlicebutexaminedafterAliceweresimilartotheeffectsonap-plicationsfiledandexaminedafterAlice.Amongthebusinessmethodsinfinance,theprobabilityofissuing§101rejectionstoapplicationsfiledbeforeAliceincreased98%becauseoftheAlicedecision.TheAl-icedecisionmadeapplicationsofbusinessmethodsinfinanceaboutfifty-onetimesmorelikelytoreceivea§101rejection.Theprobabilityofissuing§101rejectionstotheapplicationsinthebusinessmethodsoffinancefiledbeforeAliceincreased97%duetotheAlicedecision.TheAlicedecisionincreasedthislikelihoodofreceivinganinitialorfinal§101rejectionbyafactoroftwenty-eight,butthiseffectofAlice
Amonge-commerceapplicationsfiledbeforeAlice,theprobabil-ityofreceivinga§101rejectioninaninitialorfinaldecisionincreased95%fortheAlicedecision.ThissuggeststhattheAlicedecisioncausedtheseapplicationstobeabouttwentytimesmorelikelytoreceivea§101rejectioncomparedtonotreceivingrejectionsunder§101intheirinitialandfinaldecisions(i.e.,receivinganallowanceoraninitialorfinalrejectionunder§§102,103,or112).Similartoapplicationsofe-commercefiledbeforetheAlicedecisionbutexaminedafterAlice,theprobabilityofreceivinga§101rejectioninaninitialorfinaldeci-sionalsoincreasedbyabout95%becauseoftheAlicedecisionamonge-commerceapplicationsfiledafterAlice.Thee-commercepatentap-plications filedafterAlice alsohadahigherprobabilityof failing inovercomingtheinitial§101rejectionsthanaveragebusinessmethodapplications.TheAlicedecisionmadee-commerceapplicationsfiledbeforetheAlicedecisionbutexaminedafterAliceabouttwenty-onetimesmorelikelytofailinovercomingtheirinitial§101rejections.Furthermore,thePTOimplementationoftheAlicedecisionincreasedthislikelihoodtothirty-eighttimesmorethanthelikelihoodofsuc-cessfullyovercomingtheinitial§101rejections.Additionally,theAl-icedecisionmadee-commerceapplicationsfiledafterAliceaboutthir-teen times more likely to fail in overcoming their initial §101rejections;thisissmallerthanthelikelihoodofthe§101rejectionsasafinalrejectionreceivedbythoseapplicationsinthebusinessmeth-odsine-commercefiledbeforetheAlicedecision.
Forbusinessmethodsinhealthcare,theprobabilityofpatentap-plicationsfiledbeforetheAlicedecisionthatreceivedaninitialorfinal§101rejectionincreased97%becauseoftheAlicedecision.Thissug-geststhattheAlicedecisionmadethehealthcareapplicationsaboutthirty-two times more likely to receive an initial or final §101
2020] IMPLICATIONSOFALICE 583
rejectioncomparedtonotreceivinga§101rejectionintheinitialorfinalPTOdecisions.TheprobabilityofapplicationsfiledandexaminedaftertheAlicedecisionthatreceivedaninitialorfinal§101rejectionincreasedabout97%becauseoftheAlicedecision,suggestingthattheAlicedecisionmadehealthcareapplicationsabout thirty-five timesmorelikelytoreceiveaninitialorfinal§101rejection.Moreover,theAlicedecisionmadehealthcareapplicationsfiledbeforetheAlicede-cisionaboutthirty-twotimesmorelikelytofailinovercomingtheirinitial§101rejections,whichishigherthanthelikelihoodoffailingtoovercome the initial § 101 rejections in general businessmethods.WhilethePTOimplementationofAlicehadastrongernegativeeffectonthelikelihoodthatanapplicationwouldovercomeitsinitial§101rejection,thelikelihoodoffailingtoovercometheirinitial§101rejec-tionsamongthehealthcareapplicationsfiledandexaminedafterAl-icewasslightlylowerthantheapplicationsfiledbeforeAlicebutex-aminedafterAlice.Specifically,theAlicedecisionmadethehealthcareapplications filedandexaminedafterAlice twenty-nine timesmorelikelytofailinovercomingtheirinitial§101rejections.
Inbusinessmethodsofcryptography,theprobabilityofapplica-tionsfiledbeforeAlicereceivinganinitialorfinal§101rejectionin-creased95%because of theAlice decision. This suggests thatAlicemadetheseapplicationsaboutseventeentimesmorelikelytoreceiveaninitialorfinal§101rejectionascomparedtonotreceivinga§101rejectionintheirinitialandfinaldecisionsfromthePTO(e.g.,receiv-inganallowanceoraninitialorfinalrejectionunder§§102,103,or112).Moreover,theprobabilityofapplicationsinbusinessmethodsofcryptographyfiledaftertheAlicedecisionreceivinga§101rejec-tionincreased95%becauseoftheAlicedecision.ThissuggeststhattheAlicedecisionmadetheseapplicationsabouteighteentimesmorelikelytoreceiveaninitialorfinal§101rejectionascomparedtonotreceivinganinitialorfinalrejectionunder§101(e.g.,receivinganal-lowanceoraninitialorfinalrejectionunder§§102,103,or112).TheAlicedecisionalsomadetheapplicationsinbusinessmethodsofcryp-tographyfiledaftertheAlicedecisionforty-twotimesmorelikelytofailinovercomingtheirinitial§101rejections.
Inartunit2686withrespecttotelemetryandcodegeneration,applicationsfiledaftertheAlicedecisionweremorelikelytoreceiveaninitialorfinal§101rejectionascomparedtoapplicationsfiledbe-fore the Alice decision. The probability of applications receiving a§101 rejection in their initial or final decisions increased 70%be-causeoftheAlicedecision,about4%higherthantheincreasedprob-abilityofreceivinginitialorfinal§101rejectionsamongapplicationsfiledbeforetheAlicedecision.However,thecoefficientfortheD-i-DeffectortheparameteroftheAlicedecisionwasnegative,eventhoughtheD-i-DparameterispositiveforapplicationsfiledbeforetheAlicedecisionandsuggestsapositiveeffectfortheAlicedecisiononthese
233. SeeinfraAppendixCTableC1.
2020] IMPLICATIONSOFALICE 585
applications. In otherwords, among patent applications in art unit2686,thepossibilitythatapplicationsfiledandexaminedafterAlicewouldreceiveaninitialorfinal§101rejectionisslightlylowercom-paredto thatsamepossibility forapplications filedbefore theAlicedecisionbutexaminedafterAlice.
Moreover, the implementationofAlicehadastrongereffectonapplications in computernetworks filedbefore andexaminedaftertheAlicedecisionatastatisticallysignificantlevel.Notethatapplica-tionsincomputernetworkswerenotmorelikelytoberejectedunder§101intheirinitialandfinaldecisionsfromthePTOaftertheAlicedecision. Inotherwords,patentapplications incomputernetworksbecamemorelikelytobeinitiallyorfinallyrejectedunder§101afterthePTOimplementationoftheAlicedecision.However,thisincreasedlikelihoodisverysmall—about0.28timesmorelikelytobeinitiallyorfinallyrejectedunder§101thannotreceivinga§101rejectionintheirinitialandfinaldecisionsfromthePTO(e.g.,receivinganallow-anceoraninitialorfinalrejectionunder§§102,103,or112).Ontheotherhand,patentapplicationsincomputernetworksfiledaftertheAlicedecisionweremorelikelytoreceiveaninitialorfinal§101re-jectioncomparedtoapplicationsfiledbeforetheAlicedecisionbutex-aminedafterAlice.TheprobabilityofapplicationsfiledaftertheAlicedecisionreceivinganinitialorfinal§101rejectionincreased61%be-causeoftheAlicedecision;thisisabout6%higherthantheincreasedprobabilityofreceivinginitialorfinal§101rejectionsforapplicationsfiledbeforetheAlicedecision.
Asarobustnesscheck,wealsostudiedtheD-i-DeffectonlywiththedataafterJanuary2013becausesoftwareingeneralhadasharpdecreaseinthepercentagereceiving§101rejectionsoverallofficeactions received in January 2013.We cannot explain the sharp de-creasein2013,whichwasalsoignoredinthediscussionofAlice’sef-fectonsoftwarepatentapplicationsbyotherscholarsorprofession-als.234ItcouldbealaggedeffectoftheAmericaInventsAct(AIA)235orMayo,236oritusedtobehighallalong,perhapsaftertheBilskidecisionin2010,237andthe2013datawereonlyanaberration.
AftertheadjustmentofthetimeperiodbeforetheAlicedecision,softwareingeneralalsodidnothaveastatisticallysignificantD-i-DeffectoftheAlicedecision.Specifically,TC2100withrespecttocom-puterarchitectureandTC2600withrespecttocommunicationsdidnot showastatistically significantD-i-Deffectof theAlicedecision.However, some art units for sub-categories in software, includingcomputer networks, data bases and filemanagement, and cryptog-raphyandsecurity,showedsmallpositiveandstatisticallysignificantcoefficientsfortheD-i-DeffectoftheAlicedecision.Thecoefficientssuggest that the applications filed in those three areas weremorelikelythannottoreceiveaninitialor final§101rejectionaftertheAlicedecision.
Figure5belowpresentsthecoefficientsoftheinteractiontermintheD-i-Dregressionsforthesub-categoriesofsoftwaretechnolo-gies to estimate the probability of receiving final rejections under§101.238ErrorbarsinFigure5refertothestandarderrorofthere-gressionsresults,representingthevariabilityofthedata.Longerrorbars indicate that a coefficient is not statistically significant, so itsvalueordirectionisnotsufficientlyreliabletoexplaintherelationshipintheregression.
Assuggestedbyfinalrejections,applicationsincryptographyandsecuritywerenegativelyaffectedbytheAlicedecisioninovercomingtheirinitial§101rejectionsatastatisticallysignificantlevel.Thisef-fectwassmallerthantheeffectonapplicationsinthebusinessmeth-odsofcryptography.Inotherwords,patentapplicationsinthebusi-nessmethodsofcryptographyfacedgreaterdifficultyinovercomingtheirinitial§101rejectionsthanpatentapplicationsinthesoftwareofcryptographyandsecurity.Moreover,eventhoughapplicationsinthesoftwareofcryptographyandsecurityweremorelikelytoreceiveinitialandfinal§101rejectionsthannottoberejectedunder§101(i.e., receiving an allowance or an initial or final rejection under§§102,103,or112)beforetheAlicedecision,thislikelihooddidnot
reachastatisticallysignificantdegreeaftertheAlicedecision.TheAl-icedecisionmadeapplicationsinthesoftwareofcryptographyandse-curityfiledbeforeAliceabouttwotimesmorelikelytofailinovercom-ing their initial § 101 rejections, and this increased to three timesundertheeffectofthePTOimplementationofAlice.Patentapplica-tionsinthesoftwareofcryptographyandsecurityfiledaftertheAlicedecisionwerenotinfluencedbyeithertheAlicedecisionoritsimple-mentationatastatisticallysignificantlevel.
Figure5.Coefficientsof the InteractionBetweenAliceDecisionandTechnologytoEstimateFinal§101RejectionswithErrorBarsintheLogitD-i-DRegressions
Bycontrast,theimplementationoftheAlicedecisionmadedigitalandopticalcommunicationapplicationsfiledbeforeAlice twotimesmorelikelytofailinovercomingtheirinitial§101rejections,whiletheAlicedecisiondidnothaveastatisticallysignificanteffectonhowapplicationsindigitalandopticalcommunicationovercametheinitial§101 rejections. Similarly, in telemetry and code generation, eventhoughapplicationsfiledbeforetheAlicedecisionbutexaminedafterAlicedidnotexperiencemoredifficultyinovercomingtheinitial§101rejectionsatastatisticallysignificantlevelduetotheAlicedecision,thePTOimplementationofAlicemadethoseapplicationsthreetimesmorelikelytofailinovercomingtheirinitial§101rejections.Further-more, the Alice decision made telemetry and code generation
-1.50
-1.00
-0.50
0.00
0.50
1.00
1.50
2.00
2.50
softw
are (i
n gen
eral) AI
graph
ical u
ser in
terfac
e and
docu
ment p
roces
sing
data
bases
and f
ile m
anag
emen
t
crypto
graph
y and
secu
rity
compu
ter ar
chite
cture
digita
l and
optic
al co
mmunica
tion
compu
ter ne
twork
s
telec
ommun
icatio
ns
digita
l cam
eras
recor
ding a
nd co
mpressi
on
compu
ter gr
aphic
proc
essing
telem
etry a
nd co
de ge
nerat
ion
After Alice for Applications Filed Before Alice After Alice for Applications Filed After Alice
588 MINNESOTALAWREVIEW [105:527
applicationsfiledaftertheAlicedecisionfivetimesmorelikelytofailinovercomingtheirinitial§101rejections,astatisticallysignificantlevel. Besides the applications in the above three sub-categories insoftware(i.e.,cryptographyandsecurity,digitalandopticalcommu-nication,andtelemetryandcodegeneration),theD-i-Dregressionre-sultsinPartIIforothersub-categoriesinsoftwarelistedinTable1donotshowanystatisticallysignificanteffectsoftheAlicedecisionoritsimplementationonapplicationseitherfiledbeforetheAlicedecisionbutexaminedafterAliceorfiledafterthedecision.Inotherwords,ourempiricalevidencedoesnotsuggestthatAliceincreasedthebarrierofpatentabilityforthosetechnologyareasinsoftware.
creaseduncertainties,includinguncertaintiesinpatentapplicational-lowancesandrespondingtopatentofficeactionrejections,imposead-ditional costs on patent applicants during patent prosecution. 239When patent applicants cannot successfully overcome these rejec-tions,thecostsaretransformedintoexpensesinaccountingterms.240Theincreasedcostsorexpensesareadirectresultoftheuncertaintiesand increased transaction costs.241This Part presents an efficiencyanalysisbasedontheempiricalresultsandthenanalyzeswhetherthedesignoftheRevisedGuidance242canimproveefficiencywhenimple-mentingAlice.
A. INCREASED§101REJECTIONSBYUSPTOPATENTEXAMINERSThissectiondiscussesthediscrete§101rejectionsissuedbypa-
tobeinitiallyandfinallyrejectedunder§101comparedtoothertech-nologyareasbeforeAlice.AfterAlice,abouttwo-thirdsofallofficeac-tionsforapplicantsinbioinformaticswereinitialorfinal§101rejec-tions.Theproportionoffinal§101rejectionsinfinalofficeactionsforbioinformaticsalsobecame largerafterAlice.Our findingregardingnarrowlydefinedbioinformaticsisconsistentwiththeobservationbyGaudryandHayim,whoonlytrackedtherejectionsandallowanceforbioinformaticsinartunits1631,whichisnarrowerthanourdatase-lection.243Thetimeperiodfortheirdataislongerthanours—itcoversofficeactionsbetween2013and2019.244Theirdatashowthatbioin-formatics inartunits1631sufferedan increaseof§101rejectionssinceAlice,buttheyareoptimisticregardingpatentapplicationsinbi-oinformaticsbecause theirdata suggest that since2018, the allow-ancerateofpatentapplicationsinartunit1631increasedbacktothedegreebeforeAlice.245However,areversaltothepreviousallowanceratesdoesnotremovetheproblemscreatedbyAlice.
TheAlicedecisionisthereasonfortheincreased§101rejectionsissuedtonarrowlydefinedbioinformaticsapplicationsintheirinitialorfinalexaminationrounds,andthatisprovenbytheD-i-DregressionresultsinPartII.246Theincreasedinitialandfinal§101rejectionsduetoAlicesuggestincreaseduncertaintiesinpatenteligibility.247Appli-cants innarrowlydefinedbioinformaticsneed tospendmoneyandtimetoovercometherejectionsthattheyreceivedfromthePTOintheinitialroundofpatentexamination.Moreover,thePTOimplementa-tionoftheAlicedecisionstrengthenedthiseffectoftheAlicedecisionfurtherandresulted ina larger likelihoodof initiallyandfinallyre-ceiving§101rejectionsforapplicationsinnarrowlydefinedbioinfor-matics. In other words, the PTO further increased the applicationand/or prosecution costs in narrowly defined bioinformaticswhenimplementing the Alice decision. Compared to bioinformatics, nar-rowlydefined,Aliceanditsimplementationhadasmallereffectonthetechnologieswithinthescopeofbroadlydefinedbioinformaticsinart
2. BusinessMethodsIn business methods, the data suggest that the Alice decision
clearlycausedanincreaseinpatentapplicationrejections.Patentap-plicationsinbusinessmethodsreceivedbothmore§101rejectionsandmoreAlicerejectionsintheirinitialandfinaldecisionsfromthePTObecauseoftheAlicedecision.Inotherwords,Aliceinducedanin-creaseinuncertaintiesrelatedtopatenteligibilityinbusinessmeth-ods.Theincreaseduncertaintiesandincreasedprobabilityofreceiv-ing initial or final rejections under § 101 and Alice increasedprosecutioncostswhenpatentapplicantstriedtoovercomethosere-jections.248However,forapplicantswhoaddressedtheirinitiallyre-ceived§101rejectionsorAlicerejectionswhenrespondingtopatentexaminers, an increasedproportionof them failed toovercome the§101rejectionsorAlicerejections.AccordingtotheD-i-DregressionresultsinPartII,theAliceeffectthatpatentapplicationsofbusinessmethodsreceivedmoreinitialorfinal§101rejectionsduetoAlicewasstrengthenedwhenthePTOpubliclydecidedto implement itsdeci-sion.249ThisfindingisconsistentwithwhatChienandWuobservedwith PTOoffice actions during a period of time that is longer thanours.250 Even though Chien andWu did not observe howAlice in-structedapplicantstoadjusttheirapplications,251ourresearchdesignallowsustoprovethatpatenteesweresuccessfulinovercomingAlice-basedrejectionsforapplicationsfiledaftertheAlicedecision,butnotforapplicationsfiledbeforetheAlicedecision.252
WhengeneralbusinessmethodsinTC3600receivedmoreinitialorfinal§101rejectionsbasedontheAlicedecisionanditsimplemen-tation,theeffectsofAliceonthesub-categoriesinthetechnologyar-easwithinbusinessmethodsinTC3600weredifferent.Specifically,businessmethodsinfinanceandbusinessmethodsinthee-commerceof health care or cryptography faced stronger Alice effects than
jectedunder§101intheirinitialorfinaldecisionsreceivedfromthePTObeforeAliceascomparedtobioinformaticsandbusinessmeth-ods,whereinpatentapplicantsreceivedafargreaterpercentageofre-jectionsunderAliceorthePTO’simplementationofAlice.Thepropor-tionofofficeactionsinsoftwarewithinitialandfinal§101rejectionsdidnotincreasemuchaftertheAlicedecisionoritsimplementationbythePTO.D-i-DregressionresultsinPartIIdonotshowthateithertheAlicedecisionorthePTO’simplementationofAliceincreasedtheinitialandfinal§101rejectionsforpatentapplicationsforsoftwareingeneral.Wefindthattherejectionrateunder§101forsoftwarepa-tentapplicationswasrelativelyconsistent(around12%)intheperiodbetween2012and2016,exceptforasmall,inexplicabledropinthe§101rejectionrateduring2013(toaround8%).The§101rejectionrateforsoftwareinventions,however,wasmuchhigherthanthere-jection rate formechanical inventions in the control group (below1%).Thissuggeststhattheincreasein§101rejectionsforsoftwareinventionsmayhaveoccurredatatimepriortotheAlicedecision,per-hapsas a result of theBilski v.Kappos decision in2010253or someotherdevelopmentspriorto2012.Thispointmaybeexploredfurtherinfutureempiricalwork.
Insomesub-categoriesofsoftware,suchastheartunitsforcom-puternetworksandGUI,anincreasedproportionoftheapplicationswereinitiallyorfinallyrejectedunder§101aftertheAlicedecision.In addition to the increased initial and final §101 rejections, somesoftwareartunitsalsoreceivedincreasinglymoreAlicerejectionsintheirinitialandfinaldecisionsfromthePTO.UncertaintiesinpatenteligibilityincreasedafterAlice,andtheapplicantsinthoseareasspentmoretimeandmoneyonovercoming§101rejectionsafterAlice.254
Thissuperficialoverviewofthe§101rejectionsandAlicerejec-tions, however, does not necessarily mean that the Alice decisioncaused the increase in the rejections, even though theirAlice rejec-tionswerecorrelatedtotheir§101rejections.Ontheonehand,theD-i-DregressionresultscanonlyprovethattheAlicedecisiondirectlyresultedinanincreaseininitialandfinal§101rejectionsforapplica-tionsintelemetryandcodegeneration(onlyartunits2686and2688).Ontheotherhand,basedontheD-i-Dregressionresults,wecannotconcludethattheAlicedecisionhadadirecteffectontheinitialandfinalrejections intheareaofcomputernetworks.Wedofind,how-ever,thatthePTO’simplementationoftheAlicedecisiondirectlyre-sultedinmoreinitialandfinal§101rejectionsforpatentapplicationsincomputernetworks.Inotherwords,regardlessofthedirecteffectoftheAlicedecisionfromtheSupremeCourt,thePTO’simplementa-tionofAliceincreasedtheapplicationand/orprosecutioncostsforap-plicationsdirectedatcomputernetworks.255
Inadditiontotheincreasedcostsofpatentprosecutionforsoft-ware inventionsthatreceivedan increasingnumberof§101rejec-tionsandAlicerejectionsintheirinitialandfinaldecisionsreceivedfromthePTO,thesoftwareindustrymayhavelimitedaccesstocapitalfrominvestorsasaresultoftheseeligibilityrejections.InDavidTay-lor’ssurveyof475venturecapitalistsandprivateequityinvestorsbe-tween2009and2017,hefoundthatinvestorsingeneralconsiderpa-tenteligibilitywhenmakinginvestmentdecisions.256Eventhoughtheinability toobtainpatentprotectionmaynotdirectlydrive themtoreduceinvestmentinsoftwareandInternetinventions,257onceinves-torsareawareof casessuchasAlice,258Myriad,259andBilski260thataddress patent eligibility issues, they become overwhelmingly
255. Seesupranote82andaccompanyingtext. 256. DavidO.Taylor,PatentEligibilityand Investment,41CARDOZOL.REV.2019,2027(2019)(“[O]verall,74%oftheinvestorsagreedthatpatenteligibilityisanim-portant consideration in firmdecisionswhether to invest in companiesdevelopingtechnology....”). 257. See id.at2028(“Investorsoverwhelmingly indicated, forexample, that theeliminationofpatentswouldeithernotimpacttheirfirm’s’decisionswhethertoinvestincompaniesoronlyslightlydecreaseinvestmentsincompaniesdevelopingtechnol-ogyinthe...softwareandInternet(80%)...industr[y].”). 258. AliceCorp.Pty.Ltd.v.CLSBankInt’l,134S.Ct.2347(2014). 259. Ass’nforMolecularPathologyv.MyriadGenetics,Inc.,569U.S.576(2013). 260. Bilskiv.Kappos,561U.S.593(2010).
2020] IMPLICATIONSOFALICE 593
negativeabout investing in the industriesof software, the Internet,andbiotechnology.261Inotherwords,Taylor’sfindingthattheuncer-tainties inpatent eligibilityharm investment and innovation in thesoftwareindustry262issupportedbyourgeneralempiricalevidenceshowinghigherrejectionsratesforapplicationsdirectedatsoftwareinventions.
B. THEABILITYOFPATENTAPPLICANTSTOOVERCOME§101REJECTIONSInthissection,wefocusondiscussingtheeffectofAliceonthe
industriesofbioinformaticsandbusinessmethods.Theapplicantsinthesetwo industries facedgreateruncertainties inpatenteligibilityandasaresultboremorecostsorexpensesinpatentprosecutionduetoAlice.263ThiscouldbeaninevitableresultofAliceseekingtopre-ventpeoplefrompatentingsomethingthatraisesanissueofpreemp-tion and thereby to benefit the public interest. 264 Alternatively, itcouldmerelyshowthattheincreasedtransactioncosts,resultingfromtheuncertaintiescreatedbyAlice,265harminvestmentincentivesandinnovationincentivesinthosetechnologyareas.
Patent applicants in bioinformatics, narrowly and broadly de-fined,becamepessimisticaboutfilingmorepatentapplicationsafterAlice.Theaveragenumberofpatentapplicationsfiledinbioinformat-ics, narrowly defined, decreased by 74.21% per month during the
261. See Taylor, supra note 256, at 2082–83 (showing that 63% of eligibilityknowledgeableinvestorsreportednegativeimpactswithinthesoftwareandInternetindustry and86%of eligibility knowledgeable investors reportednegative impactswithinthebiotechnologyindustry). 262. Seeid.at2083–85(suggestinganegativeimpactoftheSupremeCourt’sdeci-sionsoninnovativecompanies’value). 263. Seesupranote82andaccompanyingtext. 264. SeeAliceCorp.Pty.Ltd.V.CLSBankInt’l,134S.Ct.2347,2360(2014). 265. SeesupraPartI.A.1.c.
594 MINNESOTALAWREVIEW [105:527
nineteen months after the Alice decision (June 2014 to December2015)comparedtothesamelengthoftimebeforetheAlicedecision(January2013toMay2014).266Themediannumberofpatentappli-cationfilingspermonthwas144beforetheAlicedecision.Thisnum-berdecreased to thirty-oneafter theAlicedecision.267ThesharpestdecreaseoccurredinJune2014,whentheAlicedecisionwasdeliveredbytheSupremeCourt.ApplicationsfiledinJune2014were48.56%lower than the average patent applications in bioinformatics, nar-rowlydefined,filedinthenineteenmonthsbeforetheAlicedecisionandcontinuedtodecreaseovertime.InJanuary2015,thefirstmonthafterthePTOimplementedtheAlicedecision,thisnumberfurtherde-creasedby63.07%ascomparedtobeforetheAlicedecision.Theap-plicants in narrowly defined bioinformatics made cautious adjust-mentsbymodifyingthedisclosuresandclaimsaftertheAlicedecision,andapplicantsinbroadlydefinedbioinformaticsdidthesame.Thepa-tentapplicantsinbroadlydefinedbioinformaticsalsofiledfewerpa-tent applications after Alice. The number of applications filed permonthonaveragedecreasedby51.06%aftertheAlicedecision.Themedianforthenumberofpatentapplicationswas638beforetheAlicedecision,andthisnumberdecreasedto307afterAlice.
Fromtheexaminers’perspective,Alice causedexaminers tobemorelikelyinitiallyandfinallytorejectapplicationsfiledafterAliceunder§101,eventhoughapplicantsfiledincreasinglyfewerpatentapplicationsfornarrowlydefinedbioinformatics.Inotherwords,forthosedecreasedapplicationsself-selectedbytheapplicants,theAlicedecisionalsoaddedsomeexpensesfortheseapplicants.Ourdataalsoshowthatthedecreasednumberofapplicationsinnarrowlydefinedbioinformaticswereneverthelessmorelikelytobefinallyrejectedbyexaminersunder§101.Examinersgaveincreasinglymorefinalrejec-tionstopatentapplicationsinnarrowlydefinedbioinformaticsunder§101.Forpatentapplicationsthathadinitiallyreceiveda§101rejec-tion,theD-i-DmodeldesigninthisstudyshowsacausaleffectoftheAlice decision; specifically, applicants in narrowly defined bioinfor-maticswerelesslikelytoovercometheirinitial§101rejectionsandmorelikelytoreceiveafinalrejectionunder§101.
Moreover,applicationsinnarrowlydefinedbioinformaticsfiledafter the Alice decision had greater difficulty in overcoming their
initiallyreceived§101rejectionscomparedtoapplicationsfiledbe-foretheAlicedecision.IntheD-i-DregressionresultsshowninTable5,thecoefficientinthemodelforapplicationsfiledaftertheAlicede-cisionislargerthanthecoefficientinthemodelforapplicationsfiledbeforetheAlicedecision.Accordingly,applicantsinnarrowlydefinedbioinformaticswerenotclearlyinstructedbytheSupremeCourtinitsAlicedecision,eventhoughtheyreactedtothelawandadjustedtheirpatentpreparationandfilingstrategies.Theadjustedfilingstrategiesor modified disclosures and claims failed to work, and applicantsmerelyspentmoremoneyonpatentapplicationsthatwouldnotbeapproved.WhatisworseisthatAlicemayhavenotonlyaddedmoreapplicationorprosecutioncostsinthebusinessofbioinformatics,butthecasemayhavealsorestrictedapplicants’accesstoinvestorsandcapitalthroughthemarket.
Bioscience,whichisbroaderthanbioinformatics,however,facesvariouslayersofuncertainties,suchastheuncertaintiesinresearch,infindinganapplicationoftheresearchtohumanhealth,andinpa-tentapplications.PatentsincentivizescientistsandfirmstoengageinbioscienceR&Ddespite theseuncertainties.268Even though innova-tioninbiosciencemayharmorbenefitmankind,theSupremeCourthasnotedthatitdoesnotmeantodeterordisincentivizeinnovationinbioscience,especiallygeneticresearch,byrequiringanarrowscopeofpatentablesubjectmatterinbioscience.269Theutilitybarexcludessomebioscienceresearchfrompatentabilityiftheapplicationoftheresearchisuncertain,regardlessoftheeffortsmadetoobtainthoseresearchfindings.270
Whentheutilityrequirementcannotbesatisfiedduetouncer-taintiesintheresearchresults,inventorsmayattempttobringmorespecificityorclaritybyengaginginfurtherR&D.However,whentheAlicedecisioninducedmore§101rejections,thepreviousutilityre-jectionsunder§101wereexpandedtoincluderejectionsforineligiblesubject matter, also under § 101, further compounding the uncer-taintyinprosecutionoutcomesandincreasingtheoverallcostofpa-tentprosecution.AsDavidTaylor’ssurveyregardinginvestorsshows,notreceivingpatentprotectiondirectlyresultsinlesserinvestmentinbiotechnologybyventurecapitalistsandprivateequityinvestors.271
ThereisacontinuingconcernthattheAlicedecisionobfuscatesthedistinctionbetweeneligibilityunder§101andnon-obviousnessunder§103byfocusingonwhatisgenerallyknownintheart.Thisconcernisalsoaddressedinthetextoftheproposedeligibilitylegis-lationintroducedintheHouse.272Nevertheless,wefindthattheasso-ciationbetween§101rejectionsand§§102,103,or112(b)rejectionsdecreasedafterAliceamongpatentapplicationsinnarrowlydefinedbioinformatics(e.g.,artunits1631and1639).Bycontrast,theassoci-ationbetween§101rejectionsand§112(a)rejectionsforpatentap-plications in narrowly defined bioinformatics increased afterAlice.Thatstrengthenedassociationmaysuggestthatovercomingtheprob-lemsofwrittendescriptionandenablementmayhelpwithovercom-ing§101eligibilityrejectionsinbioinformatics.
2. DiverseReactionsinBusinessMethodsandSoftwarePatent applicants filed fewer patent applications in business
methods afterAlice. In TC 3600 for general businessmethods, thenumberofpatentapplicationfilingspermonthonaveragedecreasedby38.16%inthetwelvemonthsaftertheAlicedecision(June2014–June2015)comparedtothesamelengthoftimebeforetheAlicedeci-sion(May2013–May2014).273Themedianforthepatentfilingnum-berspermonthwas9,018beforetheAlicedecision,anditdecreasedto5,445aftertheAlicedecision.Thedegreeofthedecreasewashigherforpatentapplicationsinthebusinessmethodsoffinanceande-com-merce.Infinance,theaveragenumberofpatentapplicationfilingspermonthdecreasedby57.63%inthethirteenmonthsaftertheAlicede-cision.Ine-commerce,theaveragenumberofpatentapplicationfil-ingspermonthdecreasedby79.11%aftertheAlicedecision.Withine-commerce,aftertheAlicedecision,theaveragenumberofpatentap-plicationfilingsincryptographypermonthdecreasedby67.41%,andthe averagenumber of patent application filings in health care permonthdecreasedby86.41%.
theirfirm’sinvestmentsinthebiotechnology(77%).”). 272. SeeRestoringAmerica’sLeadershipinInnovationAct,H.R.6264,115thCong.§7(2018)(draftingas“[t]heeligibilityofaclaimedinventionundersubsections(a)and(b)shallbedeterminedwithoutregardastotherequirementsorconditionsofsections102,103,and112ofthistitle,ortheclaimedinvention’sinventiveconcept”). 273. Theaverageinitialactionpendencyintechnologycenter1600is18.2months,soourdataderived fromReedTechmaydeduce the filingnumbers in the thirteenmonthsaftertheAlicedecisionforbusinessmethodsandtheartunitswithinbusinessmethods.
2020] IMPLICATIONSOFALICE 597
applicationsinbusinessmethodsstarteddecreasinginJuly2014,onemonthaftertheAlicedecision.Theaveragenumberofpatentapplica-tions decreased 24.03% in July 2014, compared to the thirteenmonthsbeforetheAlicedecision.Thedecreaseinpatentapplicationsforbusinessmethodswasincremental,andasharpdecreaseoccurredwhenthePTOdecidedtoimplementtheAlicedecision.274InJanuary2015, the firstmonthafterAlicewas implementedby thePTO, thisnumber furtherdecreasedby48.33%compared tobefore theAlicedecision.Patentapplicationsinthebusinessmethodsofhealthcare,whichistheartunitthatdecreasedthemostinthesub-categoriesinbusinessmethodsaftertheAlicedecision,decreasedtotwenty-seven,comparedto237,themediannumberforpatentapplicationsinthethirteen-monthperiodbeforetheAlicedecision.
Overall,thisstudyfindsthatpatentapplicantsinbusinessmeth-odsweregraduallyadjustingtheirapplicationstrategies:applicantsfiled fewer applications after theAlice decision, especially those inhealthcarebusinessmethods. Inaddition toapplicants’ reaction tothelaw,theyalsofiledfarfewerpatentapplicationsaftertheimple-mentationofthelawbythePTO.ThehighdegreeofthedecreaseaftertheAlicedecisionanditsimplementationbythePTOmayalsoshowthattheapplicantsarepessimisticabouttheallowanceprospectsfortheirpatentapplications.However,thedecreaseinpatentallowancedoesnotnecessarilymeanthatAliceimpedesinnovationinbusinessmethods.Whileanarrowerscopeofpatenteligibilitymaynotincen-tivizeinnovationinparticulartechnologies,itdoesnotnecessarilyde-ter innovation in those technology areas. 275 Innovative companiespracticingbusinessmethodsinthoseareasmayusetradesecrets,ra-ther than patents, if they are barred from receiving patent protec-tion.276
Moreover,thepatentapplicationsinbusinessmethodsfiledaftertheAlicedecisionwerenotmorelikelytobeinitiallyorfinallyrejectedbypatentexaminersunder§101,assuggestedbythenegativeesti-matoroftheinteractiontermintheD-i-DregressionsinTable6.277TheabilitytoadjusttotheAlicedecisionismostclearlyseeninthebusiness methods of cryptography, wherein Alice did not increase
initialandfinal§101rejectionsfortheapplicationsfiledafterAlicecomparedtoapplicationsfiledbeforeAlice.Bycontrast,thepatentap-plicationsfiledafterAliceinfinanceore-commerceweremorelikelytobeinitiallyorfinallyrejectedunder§101comparedtopatentap-plicationsfiledbeforetheAlicedecision.Thus,Aliceimposedcostsonbothexaminersandapplicants in the fieldsofbusinessmethods inhealthcare,finance,ande-commerce.
Examinersgavemorefinalrejectionstoapplicationsinbusinessmethodsunder§101aftertheAlicedecision.Inotherwords,aftertheAlicedecision,applicantsfaceddifficultiesinsuccessfullyovercoming§101rejections.Forapplicationsthatinitiallyreceiveda§101rejec-tion,Alicemade itmoredifficult for them to overcome their initial§101rejections.TheimplementationofthelawbythePTOincreasedtheuncertaintiesinpatenteligibilityandthedifficultiesinovercom-ingtheseuncertaintiestoahigherdegree.
Alicealsoincreasedtheprosecutioncostsforpatentapplicationsin businessmethods,which could be absorbedby applicantswhentheymodifiedtheirdisclosuresandclaimstorenderthemlesslikelytobefinallyrejectedunder§101(i.e.,receiveanallowanceorafinalrejectionunder§§102,103,or112).However, forthoseapplicantswhofailedtomodifysuccessfullytheirdisclosuresandclaimsintheirpatent applications and overcome these § 101 rejections, Alice in-creasedoverallpatentprosecutioncosts.Inthebusinessmethodsoffinanceande-commerce,applicantswerenotclearlyguidedbytheAl-icedecision,andtheydidnotsuccessfullyadjusttheirpatentingstrat-egies,despitefilingfewerpatentapplications.TheseapplicantsfacedhigherpatentprosecutionexpensesbecauseofAliceandthePTO’sim-plementationof it.The increase in theexpenses forpatentapplica-tionsinthebusinessmethodsoffinancewashigherthantheincreaseintheexpensesforpatentapplicationsinthebusinessmethodsofe-commerce.
The goal of the Supreme Court in Alice was to exclude thoseclaimsthatconstitutethe“buildingblocksofhumaningenuity,”whichcreaterisksofpreemption.278Therefore, increasedrejectionsunder§101 afterAlice might serve as evidence showing that the risk ofpreemptioncreatedbypatentshasalsobeenreducedbypatentexam-iners.Additionalevidenceconcerningthisriskofpreemption is theincreasedassociationbetweenthefrequencyofreceiving§101and§112(a)rejectionsinbothinitialandfinalofficeactionsfromthePTOafterAlice.
Analternativeexplanationfortheincreased§101rejectionsinthe three sub-categories inbusinessmethods is that the law is notclearenoughtoinstructexaminersandpatentapplicantsandmerelycreatescostlyuncertainties.Thisstudyobservedthattheassociationbetween§101and§103rejectionsamongthepatentapplicationsinbusinessmethodswasheavilystrengthenedafterAlice.ThisresultistheclearestsupportforthewidespreadcriticismthattheteachingsinAlicethatfocusonwhatis“generallyknown”bothnecessarilyandin-correctlydrawthepriorart intotheeligibility inquiry.Therelationbetween§§101and103forbusinessmethodpatentapplicationsistheoppositeofwhatwehaveseenearlierinbioinformatics.
WecannotprovethatAlicecausedmore§101rejectionsingen-eral software andmost sub-categories of software technologies.279However,weobservethattheAlicedecisionresultedinfewer§101rejectionsinsomesoftwaretechnologies(e.g.,databasesandfileman-agement,cryptographyandsecurity,GUIanddocumentprocessing,andcomputerarchitecture).Inthesefourtechnologyareas,theasso-ciationbetween§101rejectionsand§112(a)rejectionswasweak-enedafterAlice.Incontrast,theassociationbetween§101and§102or§103rejectionswasstrengthenedafterAliceforsoftwareapplica-tionsinthesetechnologyareas.
C. THEFUTUREOFSHIFTINGTRANSACTIONCOSTSTOTHEPTOThisSectiondiscusseshowthe2019RevisedGuidancefromthe
PTOtriestomitigatetheincreasedcostsofpatentprosecutionfacedbypatentees.TheD-i-DempiricalresultsinPartIIshowtheeffectoftheAlicedecisionand its implementationby thePTOonexaminersandpatentapplicants.280Insometechnologyareas,suchasbioinfor-matics,businessmethods,andsoftwareoftelemetryandcodegener-ation,we find that theAlice decision inducedmore initial and final§101rejectionsissuedbypatentexaminerstoapplications,especiallyforapplicationsfiledbeforetheAlicedecision.Thedirectionoftheef-fectofthePTOimplementationisconsistentwiththeAlicedecision,butthePTO’simplementationhadastrongereffectthanAlice itself.After thePTOpublished itsspecific InterimGuidance to implementtheAlicedecision,applicantsreceivedmore§101rejectionsinthein-itialroundofpatentexamination,anditbecamemoredifficulttoover-cometheserejections.Therefore,thePTOimplementationofAlicefur-therincreasedthecostofpatentprosecutionforpatentees.
InJanuary2019,thePTOissuedRevisedPatentSubjectMatterEligibilityGuidance.281ThePTOaddedmoredetailed instruction inthisRevisedGuidanceinordertodecreasetheuncertaintyofpatent-ing and the transaction costs created byAlice and its implementa-tion.282BesidestheadditionalStep2AandStep2B,comparedtotheInterimGuidance,thereisonemorestepaddedafterthetwo-prongtestforcarefullyexploringtheeligibilityofpatentapplications.283
TheRevisedGuidanceenumeratesthreetypesofabstractideasinaccordancewithseveralprecedentsfromtheSupremeCourtortheFederalCircuit.286Thethreetypesofabstractideasaremathematicalconcepts, 287 certain methods of organizing human activity, 288 andmentalprocesses.289Anyrejectionsforrecitingan“abstractidea”thatisnotenumeratedintheRevisedGuidancemustbeapprovedbytheTechnology CenterDirector andmust provide justification forwhysuchclaimlimitation(s)aretreatedasrecitinganabstractidea.290
Both the additional procedure and the additional indicationabout reciting “abstract ideas” in the justification add transactioncoststothePTO.FouryearsafterthePTOformallyimplementedtheAlicedecisionundertheInterimGuidance,thePTOputforthfurthereffortstoclarifytheboundarybetween“abstractideas”andnon-ab-stract, patent-eligible subjectmatter.291This additional analysis re-sultsinanincreaseinadministrativetransactioncostsatthePTO.292
Thisincreasecouldbearesultofshiftingtheincreasedtransac-tioncostsinthemarketthatarebornebypatentapplicantsandinves-torstothePTO.Forexample,ourempiricalresultsinPartIIshowthatboth theAlicedecisionand its implementationby thePTO increasepatentprosecutioncostsorexpensesonapplicantsinsometechnol-ogyareas,suchasbioinformatics,businessmethodsofe-commerce,businessmethodsoffinance,andsoftwareofcryptographyandsecu-rity. Even though applicants adjusted their application strategies,modifiedthedisclosuresandclaims intheirapplications,andspentmoremoneyonpatentingbecauseofthechangesinthelaw,therewasstillsignificantdifficultyinovercoming§101rejections.TheRevisedGuidancesuggeststhatthePTOdecidedtotakeontheburdentoclar-ifythelawthroughtherevisedStep2andtheaddedsecondreviewprocedurespertainingtotheelementsthatarenot listed intheRe-visedGuidance.293
filed before and examined after the implementation of theRevisedGuidancewerelesslikelytoreceiveafirstofficeactionwitha§101rejectioncomparedtothosefiledandexaminedbeforetheimplemen-tationoftheRevisedGuidance,theprobabilityofreceivingafirstof-ficeactionwitha§101rejectionaftertheimplementationoftheRe-visedGuidancedecreasedlessthanbeforetheAlicedecision.296TooleandPairoleroalsoshowedthatpatentexaminersservingtheTCsaf-fectedbyAlicewerelesslikelytoissueafirstofficeactionwith§101rejectionsaftertheimplementationoftheRevisedGuidance.297How-ever,onceagain, theprobabilityof thoseexaminers issuinga§101rejection decreased to a lower degree than before the Alice deci-sion.298Inaddition,thetwopairsofcomparisonsdonotconsiderthevariationinthetotalnumberofpatentapplicationsfiledbeforeandafterAlice.299Thus,theseresultsmayconfirmthePTO’seffortstode-creasetheuncertainties inpatenteligibility,300buttheycannotesti-mateboth the transactioncostsbornebyapplicantswhomayhaveadjustedtheirapplicationbehaviorsafterAliceandtheimpactonin-novationthroughapplicantsforegoingallpatentprotectionfortheirinventions. Second, the Revised Guidance may not affect how thecourtsdeterminepatenteligibilityorhowtheyapplyAlice301becausethejudicialsystemisalsoacriticalplayerincontinuouslycreatingun-certaintiesinpatenteligibility.302Itishard,however,topredicthowthe PTO’s justification addressing the uncertainties of patentabilitywillbeperceivedonreviewbythejudicialsystem.303
ThePTABprovidesamechanismbywhichwecanexaminethepossible efficiencies gained at thePTO through the judicial system.TheFederalCircuitreviewsthePTO’stestsregardingpatenteligibilityasappliedinPGRsorCBMsatthePTABthatemploytheRevisedGuid-ance.304Therefore,notonlywilltheenumeratedspecifictypesofab-stractideasbepresentedtotheFederalCircuit,butothernewtypesofabstractideasoutsidethescopeofpatentprotectiondynamicallyaddedbytheTCdirectorswillalsobereviewedbytheFederalCircuit.
AftertheRevisedGuidancefor§101cameoutin2019,thePTOrevisedtheGuidancefor§112becauseaclaimdraftedbroadlyisnotanissueunder§101(i.e.,patentablesubjectmatterorutility),butitisan issueunder§112(i.e.,writtendescriptionorenablement).305Thiscouldbemoreimportantinsomeparticularsoftwaretechnolo-gies,suchasdatabasesandfilemanagement,cryptographyandsecu-rity, computer architecture, GUI and document, and computergraphicsprocessing.Thedataanalyzedinthisstudyshowthattheas-sociationbetween§101rejectionsand§112(a)rejections inthosetechnologyareasdecreasedafterAlice.AliceandtheRevisedGuidancefor§101maynotbesufficienttodealwiththepreemptionprobleminthosetechnologyareas.
Alice’stwo-parttestatthePTO,focusingonthreetechnologyareas—software,bioinformaticsandbusinessmethods.OurcausalempiricalstudyoftheAlicedecisionrevealshowthatcaseimpactedbothpatentexaminersandpatentapplicants,increasingthetransactioncostsas-sociatedwithpatentprosecutionandcreatinguncertainoutcomesinpatent allowance. Patent applicants employed different strategies(e.g.,filingamendedpatentclaims,filingnewpatentapplications,orchoosingtoabandonorneverpursuepatentprotection)toattempttocomplywithAlice.
Patentapplicantsinallthreetechnologyareasdecreasedtheirre-liance on the patent system and filed fewer patent applications ascomparedtothetimeperiodbeforeAlice,withthegreatestreductionoccurringinbioinformatics.Patenteesinsometechnologyareas(e.g.,businessmethodsingeneral)weresuccessfulinovercoming§101re-jectionsafterAlice,butpatenteesinotherareas(e.g.,bioinformaticsandfinanceore-commercebusinessmethods)werenotassuccessfulinovercomingAlice-basedrejections.Applications inbioinformaticsreceivedmanymore§101rejectionsbasedonAlice,buttheseappli-cants also faced difficulties in overcoming those rejections. Mean-while,patentapplicationsbasedonbusinessmethodsalsoreceivedmore§101rejectionsbasedonAlice.ButpatentapplicantsinbusinessmethodslearnedfromAliceandreceivedfewer§101rejectionswhenthey filedpatentapplicationsafterAlice.Alicealso imposedvariousdegreesofpatentingcostsfordifferenttypesofsoftwareinnovation.
InadditiontothePTO,otherpatentinstitutions,suchastheFed-eralCircuit,havestruggledtooperationalizeAliceandtherebymiti-gate its uncertain application. Indeed, the Federal Circuit has ex-pressed futility at staying within the Alice framework. Since theSupremeCourthasnotshownadesiretore-visititsAlicedecision,wearenowleftwiththehopethatCongressionallegislationoneligibilitymightbringsomemuch-neededclaritytothisthresholdrequirementinpatentlaw.
and𝑡 ∈ 𝑇 = {201201,… ,201612} .𝐼 represents office actions. n do-natesthetotalnumberofofficeactionsgiventoindividualtechnologyareas.𝑡$denotestheintervention(i.e.,AlicedecisionorthePTOim-plementation).𝜀 denotes an idiosyncratic error term uncorrelatedwithotherindependentvariablesorcontrols.
𝑅𝑒𝑗101indicateswhether examiners gave a § 101 rejection. Itequals0whentheofficeactionwasanallowanceandequals1whenthe office action was a rejection.𝑝!"#"$ denotes the probability of𝑅𝑒𝑗101=1.TheD-i-DmodelsestimatetheaveragedegreeofR&Din-tensity.𝑇𝑒𝑐ℎ# and𝐴𝑙𝑖𝑐𝑒"$ absorb category- and event-fixed effects.Thecoefficientontheinteraction,𝛽&,surrogatesfortheD-i-Deffect.𝛾"denotesthemonth-fixedvariable.Thecontrolvariables forotherstatutoryrejections(e.g.,§§102,103,and112rejections), industrycategory(e.g.,technologycentersorartunits)arealsoincluded.
-2.00
-1.50
-1.00
-0.50
0.00
0.50
1.00
1.50
softw
are (i
n gen
eral) AI
graphic
al us
er inter
face a
nd do
cumen
t pro
cessi
ng
data base
s and
file
manag
emen
t
cryptogra
phy an
d sec
urity
compute
r arch
itectu
re
digital a
nd opti
cal c
ommunica
tion
compute
r netw
orks
telec
ommunic
ation
s
digital c
amera
s
record
ing and
compres
sion
compute
r grap
hic proc
essin
g
telem
etry an
d code
genera
tion
telem
etry an
d code
genera
tion-26
88
telem
etry an
d code
genera
tion-26
86
After Alice for Applications Filed Before Alice After Alice for Applications Filed After Alice
610 MINNESOTALAWREVIEW [105:527
FigureC1.CoefficientsoftheInteractionBetweenAliceDecisionandTechnology toEstimateAll§101RejectionswithErrorBars in theLogitD-i-DRegressions
notestheintervention(i.e.,Alicedecisionortheimplementation).𝑑"$denotesthestaticD-i-Dequation.𝑑"$,( areasequenceoffutureval-ues.𝑑"$,( are a sequenceof exantevalues.𝐸"$ denotesexpectationtakenwithrespecttoatreatmentattheAlicedecision.𝑒"$isanidio-syncraticerrortermuncorrelatedwithotherindependentvariables.
306. AnupMalani&JulianReif,InterpretingPre-TrendsasAnticipation:ImpactonEstimatedTreatmentEffectsfromTortReform,124J.PUB.ECON.1,1–2(2015). 307. AnupMalani&JulianReif,AccountingforAnticipationEffects:AnApplicationtoMedicalMalpracticeTortReform5(JohnM.OlinL.&Econ.WorkingPaperNo.578,2011),https://chicagounbound.uchicago.edu/cgi/viewcontent.cgi?article=1027&context=law_and_economics [https://perma.cc/7CLY-62MN]; see also Gregory C.Chow,RationalVersusAdaptiveExpectationsinPresentValueModels,71REV.ECON.&STAT.376(1989)(examiningrationalexpectationsandadaptiveexpectations).