"TAXCOMPLEXITY,TAXSALIENCE,ANDTAX POLITICS"

December 2014

ABSTRACT

Thispaperconsiders the implications of thetax salience literature for the UK. First, the different categories,and definitions, of tax salience that have developed inthe literature are reviewed, and someof the prescriptive implications of these terms are introduced. Tax salience refers, essentially, to the capacity of taxpayers to understand legislation. Thus, the potential reasons behind tax complexity, and thepotential beneficiaries of it, also are addressed. The paper considers thecontributionthat thetax salience literature may maketo existing analyses in the UK, particularly 1) in the debate surrounding principles based legislation, and 2) in the context of newly formed government offices devoted, to some extent, to considering the role and value of salience. It is suggested that it is worth heeding the admonitions of Schenk, Gamage and Shanske, and others, that “salience” is a layered term, that should be employed with specific reference to political,economic and legal contexts. It is also acknowledged thatthe conclusions of someportions of the tax salience literature that legislative clarityshouldbe includedonly within alist of priorities for legislativedrafting, and not necessarily at the top of that list, is evocative of existing analyses of “principles based drafting” in the UK, New Zealand and elsewhere. Neithersalience, nor principlesbased drafting, have the potential, on their own, to “fix” the problem of tax complexity. Thepaper, however, considers theimplications of eitherraising, or lowering,salience on the list of desirablefactors in drafting, in the particular context of fostering trust betweentaxpayers and the state.

Tax Complexity, Tax Salience, and Tax Politics

Should it matterif taxpayerscannotunderstandtaxlaw? It is perhapsinstinctive that,yes, itshould matter. Indeed, itshould not be too difficult to makethe caseagainst complexityinfiscallegislation, in that perhaps themost significantproblem with overlycomplextaxlegislation(accordingtotraditionalarguments)isthat it has a negative impact upon theabilityof taxpayers to complywith theirobligations. (CucciaCarnes, 2001, p. 113)Additionally,complexlegislation maycauseaperception that the law is unfair.(ibid) It issomewhat surprisingthusto learnthatthesearguments, howeverintuitively appealing,arenotsupportedbyadeveloped empiricalliterature; and, moreover, scientific investigations intowhethercomplexitysignificantlyaffects compliancelargelyhas produced “mixed” results.(id.)

Theabsence of developed researchin the empirical effects of complex tax legislation hasnot prevented theemergence of asophisticated bodyof theoreticalwork. This paper addresses one aspect ofthis literature,whichmaybe referred toas the “taxsalience,” interdisciplinarybranch of legal theoryandbehaviouraleconomics. Although this literaturehas been developed primarilyin theUS todate, its theoretical basesseem to be constructed in a waythat is deliberatelyuniversal inpotential application.

Theterm “taxsalience” refers to the gapbetweena taxpayer’sperception ofataxobligation,and the amount that is actuallyowed. Two leading authors inthis field,Gamage and Shanske, haveexplained that, ultimately,“[t]axsalience refers to howthe presentation of taxpricesaffectstaxbehaviour.” (2011, p. 19)Theyprovide awidereview of thedifferingdefinitions of salience,(ibid, fn. 14-17)but,generally, explain thathigh taxsalience indicates that taxpayers have anaccurate understandingof thelaw;whereasthe description‘low taxsalience’refers to a significantgapbetweenataxpayer’s impressionof legal obligations andthose that actuallyexist, whetheror not thetaxpayer is awareof thisgap.This literaturechallenges thegeneralassumption ofpoliticaldiscourse that ataxwith low salience must beinneed of reform. The argument is notthat low taxsalience is infact a positive aspect of legislation; but, rather, that itis one ofanumberofpotentialfactorsto be balancedwhen legislation isbeingdrafted, and, as such, itshould not necessarilyoccupyatop position on this list of factors.

This paperwillconsider the implications of thisliterature fortheUK, in threeparts.First, thedifferentcategories,anddefinitions, of taxsalience that have developed inthe literaturearereviewed,and someof theprescriptive implications of these terms areintroduced.Thesecondsectionaddressesthe question oftaxcomplexityspecifically,andaims to considerthepotential reasons forit, thepotential beneficiariesof it, and the contributionthat thetaxsalience literaturemaymaketo existinganalyses. Thefinalsection will consider salience inthe context of taxpayerconfidence in the system,andperceptions of fairness. It will considerthepotentialof therelativelynew Office ofTaxSimplification(The Officeof TaxSimplification, 2013) forconfidence,fairness,andsalience,and the interactionbetweenpoliticsand taxlaw.Thispaperwillexploreathesis that decisionsabout the importance oftaxsalience involve valuejudgments, and,as such, widerprescriptions about the relativerankingofsalience on alist of taxlegislative desirables should be avoided.

What is Tax Salience?An introduction to the literature

Thestartingpoint for this paper isGamage and Shanske’s suggestion that the problematic oftaxcomplexityshould beconsideredwithin thecontext ofthe interdisciplinarytaxsalienceliterature, in part becausethis literature"measureshow taxpayer [behaviour]departsfromkeyassumptions of neoclassicaleconomic theory."(2011, pp. 19, 33)In this sense,theychallengesan existingorthodoxy. Theyadvance the basic precept that thecomplexityof taxlegislationis inextricablylinkedwithtaxpayers’perceptions of it, and,thus, it is from the taxpayers’perspective thatcallsto simplifyorreform legislation must beconsidered. Theissueis thus entirelyoneof subjectiveperception, of whichtherearetwocategories: what taxlegislation actuallyrequires,and whattaxpayers think that thelaw requires.(ibid) In traditional, politicaldiscourse, thegapbetween the two should beassmallaspossible. (id.)

GamageandShanske suggest, however, that there aresome problems with the traditionalapproach to tax complexity.(p. 33) First, focusingon narrowingthegap,and on makingtaxlegislationasclearandas simple aspossible, necessarilyplaces simplicityat thetop ofalist of priorities.Thisprioritisation mayoccurwithout sufficientconsideration,forexample,ofthe role oftaxpayererrors. (ibid)Perhapslegislation is perceived tobe complexbytaxpayersbecause theyhavemisunderstood it, or because theyhaveapplied it incorrectlyto theircircumstances. Thechallenge is that it is not necessarilyclear ifeveryerror should bedismissed as the product ofpoor drafting. Secondly,the prioritygiven to taxsalienceas agoal– or, inotherwords, theassumption that it is important and valuable that taxpayers should be abletounderstand thelaw–maybe exaggerated,whencompared with othergoals of taxlegislative drafting. (id.)GamageandShanske donot suggestthat saliencedoes not belongonalist of priorities fortaxlegislationdrafting; but, rather, theargumentis that salience does not necessarilybelongat thetop of that list.

Thereareidentifiable consequencesforboth highand low saliencesystems. Inataxsystemwith complete salience,taxpayersareawareof“actual” taxes whentheymakeeconomicdecisions, or politicalchoices See, inter alia, (Finkelstein, 2009, p. 972; Auerbach,et al.,2010; Congdon,et al., 2009)In a system with lowersalience,taxpayers do not realise the fullcosts of taxes, but theydo possess a “perception”of them. (ibid)Further,oneof thepurported consequencesof decreasing taxsalience isgrowth in thesizeof government, aconcern that canbefound in the writings of Mill. (ibid, at p.970, citing Mill, 1848; andGamageShanske, 2011-2012, p.19).Indeed,Buchanan, inparticular,argued that “taxmisperception” is “exploited”bygovernments in order to justifyexpansionof thestate. (id.,citingBuchanan 1967) Within these arguments,governments traditionallyare perceivedasbeneficiaries oflow taxsalience, or, simply, of taxcomplexity.Discernible within theseassumptions is alingeringsuspicionthat,perhaps,governments deliberatelyconspiretorendertaxlegislationascomplexaspossible,such that taxpayerswill not be able to tell when theyareoverpaying.

Asmentionedinfra, there is not a great deal of empiricalresearchconsideringwhethertaxpayers struggleeitherto complywith, or to protectthemselves adequatelyfrom,complexlegislation. Perhaps the reason forthis is that the negative consequences of taxcomplexityareperceivedto be self-evident. Yet one interesting,recent studydid attempt toput these theories to thetest(and,as aconsequence,hasbeenwidelycited).(Finkelstein,2009,citedby, inter alia,Chetty,etal., 2007; and,Auerbach,et al., 2010)Finkelsteinexaminedwhetheranylink couldbe discernedbetween theintroduction of electronictoll passesin theUS, and the rate of thetoll. (ibid.)Electronic toll passesgenerallyworkbyaffixingaplasticboxto the windscreen on atollpayer’s car, so thatthe drivermaydrivequicklythrough tollstops, without the need forpersonalinteraction with atoll attendant.Given the seamlessexperience of payment enabled bythe electronic toll pass,Finkelsteinqueried,aredriversless aware of the amount of thetoll that theyarepaying? Additionally, is this lack of

awareness exploited bygovernments, who thenproceed to raise therates?Finkelstein’sstudyconcluded that thereis a “robust” link between levels of electronic toll pass usage,and

higherrates of tolls, so this is aclear possibility(although sheemphasises that the “normativeimplications of these findings are ambiguous”).(id., p.1009)

Thereason forthis ambiguity,Finkelstein suggests, is that evidence has not beencollected on“what is done”with theextrarevenue that electronic toll passesenable to becollected. (id.)She alsoemphasises thata taxpayer’s response to lowersalience for acomparativelysmalltoll paymentmaysignificantlydifferfrom theinterest taken in higher valuetaxes.[1] (id.)Thehypothesis, however,is thattaxsalience can haveparticularresonance in consideration oftopics such as the potentialintroduction of valueadded taxes (ashas been considered in theUS), the reduction of budgetdeficits,the desirabilityof taxexpenditures, and electronic taxreturnfiling,amongst others. (GamageShanske, 2011, p. 19)

Although this paper has discussed salience as if the term were synonymous with complexity, broadly understood, it is worth emphasizing that taxsalience is a term with several connotations of meaning. As Gamage and Shanskeexplain, there aretwocategories of “taxsalience”within the literature: “economic salience,”(ascoinedbySchenk)which indicates the impactof thepresentation oftaxupon economicdecisions;and,“politicalsalience,”whichrefers to the impact of taxpresentation on thepoliticalprocessand election outcomes.(id., at p.20, citing Schenk, 2011, pp.272-273)GamageandShanske emphasise that thesetwo categoriesmustbe distinguished,becauseincreasedpolitical saliencecan result inreducedeconomic salience(and, indeed, theconverse). (ibid) Additionally, whilst theydo not suggestthat either economic or politicalsalience should be“ignor[ed],”theyalsocaution as to the importance of rememberingthatclaimsaboutsalience in“real-worldpolicy”arenot supported byempiricalevidence. (id.)

Salience, as an interdisciplinary term, exists, as a concept, at the borderline of behavioural economics and political discourse. What is its potential value in legal analyses? Onepotential benefit of introducingterms from behaviouraleconomics into analysesof taxlaw andpolicyis thepossibilityof aclearer,public understandingofthe impact ofcomplextaxlegislation. Indeed, inanearlierpiece,Schenkemphasised the importance ofclarifyingtermssuchas“transparency”,“complexity” and“salience”,andwarned that theyareoften confused,especiallyin thecontext of thepublic discourse surrounding “hidden taxes”. (2011,p. 255)Especiallywhensalience is presented asaremedyforpurported hidden taxes,Schenkemphasises, it is important to clarifywhat,exactly,transparency, complexityandsalience mean. (ibid) Transparency, traditionallylaudedasaverypositivegoal, is valuedboth as avirtue,and for its own sake, but also as atool with instrumentalistpotential forimprovingthe politicalprocess. (id., at p.257-259) The meansbywhich transparencymightbe achieved,however, areless clear. (id.,at 259)

Schenkwrites that oneofthe precepts behind thepublicdiscussion ofcomplexityis“[t]hegeneral idea… that thegovernmentshould operate in the openand not in secretand shouldstrive to makeinformation availableto thepublic.”(id.)This idea, or value, is evident insome aspects of the political process(forexample,whenlegislatures publiclydebateproposedlegislation),but not in all (i.e.,whentaxpayers areunableto contributeto thehearings or to the process). (id.)It is against this background that the impact of taxcomplexityshould beassessed.

Schenk defines the term complexity itself as, unsurprisingly, a complicated term, but one which essentially indicates a taxpayer’s capacity both to understand, and to comply, with legislation. (id., at p.261) The complications emanate from the different categories of complexity, including compliance, “transactional” (or behavioural responses to law, undertaken in thehopeofpayingless tax), and “rule complexity” (simply, difficultyunderstandingthe law). (ibid, p.261)These categories of reactions both “overlap”andobscureeach other.(id.) Schenk,however, utilises the term to refer to “rule complexity”, forthe purposes of her analyses of taxsalience.

As, in the context of tax, salience analyses involve discussions about rules, it follows that specificity of expression is important in this area, not least because differenttypes ofcomplexityhavedifferenttypesofimpact. Additionally, some forms of impact are more important than others. Economic, or market,salience is not considered to be a significant problem, becausetaxpayers havethecapacityto “learn fromexperience”. (GamageShanske, 2011, p. 21) Infact,attemptingto curb low market saliencecould cause moreproblemsthan aresolved.An increase in politicalsalience, conversely,could be construedas essentiallyprovidingtaxpayerswith“false orarbitraryinformation abouttaxcosts”.(ibid.) The reason forthis is that “[n]eitherthe fields of philosophynor ofvoterpsychologyaresufficientlydeveloped toguide usas towhatinformationvotersought to focus on when assessingreal-world fiscalpolicies.”(id.)Given this, “[d]emocratic values thus provideno indicationas to whetherpolitical salienceshould bemadehigher or lower.”(id.)

GamageandShanske’sambivalenceover thevalueof politicalsalience isechoedbyWeisbach, who has suggested that thedesirabilityof increasingknowledgeof taxlawdepends upon threequestions. (Weisbach, 2013) These include“beliefs” about taxlegislation“in theabsenceof knowledge”; “thequalityof thetaxlaw”(i.e.,how well is itdesigned?);and“thetypeof tax”.(ibid, p.188) On thethird point,Weisbach explains thatit is most likelydesirablethat taxpayers should understandthe intention behind and designofso calledPigouvian taxes, or taxes that aredesigned to correct negative externalities.(id.)As Pigouvian taxes aredesigned tocorrectinefficiencies in themarket, they,simply,aremeant to impact upon taxpayer behaviour.[2] Theydiffer in this wayfrom taxes that aredesignedonlyto raise revenue, and for which, Weisbach argues, thereis aless compellingargument in favour ofincreased taxpayer understanding. (id.)

Theclearimplication from Weisbach’sargumentsis that,sometimes, thecapacityof ataxpayer to understandcertaintypes of taxes is not a veryimportant point to consider inassessment ofits overallvalue. Indeed,Weisbachemphasises that it is crucial that thespecific context of each taxis clarified,beforeassumptionsaremade about thedesirabilityof increasingtaxpayer understanding.(id., p.189) Forexample,thresholdsandmarginalratesfor taxcredits in theUS are designed inawaythat makes them difficult to understand.(id.) Should they be simplified, and rendered easier to understand? Perhaps, but it is perhaps worth recalling that some forms of general anti-avoidance legislationappearto be predicated to someextent on thebeliefthat it is sometimesundesirable for taxpayersto understandlegislationwell; or, at theleast,wellenough so asto be able to spot the loopholes(id.)

Weisbachconcludesthat“…unfortunately,…whether knowledgeof thetaxlaw improves theefficiencyof thesystemiscontingent on a numberof hard to observe factors”,and thus,“clean” answers to the questions surroundingthe desirabilityof increasingtaxpayerunderstandingmaynot bepossible. (id., p.210)Further,giventhat sometaxlaw is designedwith the objective of influencingtaxpayerbehaviouras little aspossible, it then mayfollowthat it would be useful to aim to ensure that taxlegislation is difficult to understand. (id.,p.187) It should be emphasisedthough that thiswould bequiteaspecific type of tax, and thissuggestion should not beappliedto all forms of taxlegislation.

What is taxcomplexity? Consideration ofthetarget of the tax salience literature

Despite Weisbach’swarning against ‘clean’ answers to the question of whether it is desirable for taxpayers to understand legislation,anddespite the veryspecificnatureof thetypesof taxesto which the desirability of comprehensibility may/not apply, the radical suggestion that taxcomplexityin some instances, indeed,maybepositivedoesprovideauseful transition to the followingquestions. When is taxcomplexityagoodthing?Ishigh taxsalience,generally,ofdubious, or at least relativevalue? One wayof addressingthese questions is throughseparate contexts.First, the valueoftaxsaliencecanperhapsbeunderstood in the area ofconsumerprotection. Studieshave shown that if,as is common inthe US, taxes arenot included in the displaypriceof aproduct, thereis a potential thatconsumerswill underestimatethe ultimate cost ofa product;and, thus, over-consume.(Nussim, 2010, p. 222)Secondly, the issue of taxpayersclaimingback taxpresents anexampleof an issuefor which it is appropriate toask, forwhom does“value” haverelevance? Deductions forpensions,and charitable giving, are claimed back in theUK atnotablylow rates, presumablybecause taxpayersdo not understandthat theyareentitled tomakethe claim.(Greenwood, 2010;Hills, 2006)Yetwhentaxpayersarepresentedwith theoption of automatic assessment– of notreceiving aletterthrough the post from HMRC– on

the whole, the discomfortwith which manytaxpayers view HMRC interaction mayleadthemto assentreadily, without awarenessthat this maylead to payingmoretaxthan is required.[3] (See generally, albeit in an Australian context,Braithwaite,2013).

In this sense, perhaps, low taxsalience has value, butperhaps forthetaxcollector,andnotfortaxpayers. Yet taxcomplexitymayarisenot simplybecause it benefits thetaxauthority,butalso as a consequence of pursuingworthwhile and necessaryobjectives in theinterest of the taxpayer. For example,complexrules maybenecessaryto measureincomeaccurately, or to ensurethat benefits are received bythe taxpayerswho need them. (GAO(U.S.Government AccountabilityOffice), 28June2011) Additionally, complexrules maybe aneffective responsetononcompliance,whichhas the potential to be viewedas positiveactiontaken on behalf oftaxpayerswho docomply. (ibid)

A third possible source of complexity(in addition to the pursuit of goals which benefiteitherthe taxauthority,or thetaxpayer)is that,perhaps,taxpayersask too much oftaxlegislation,or place unreasonable demands upon thecapacityof writtenlegislation to communicate,effectively,what at firstglance might appear to besimple concepts - for example, the tax terms of asset, orgain. Itmaybethe case that the rules themselves have been designed with unrealisticexpectations, orwithout the benefit of a clearlyarticulatedpolicy. Hart, in his accountof the “opentexture”of language,famouslyreferred to therelative ignoranceoffact, and the relative indeterminacyofthe aim of thelegislator requiringreassessmentat thepoint of application. (Hart, 1961, p. ch.7;Shaw, 2013)This applies clearlyto much of thetaxlexicon.

On the point of policy, Sawyer and others have suggested that clear policy, in fact, is the sine qua non of clear tax legislation. Indeed, this argument is that for tax legislation to be considered successful, or “effective,” the policy should be clear from the statute itself. (Sawyer, 2013, p.321, citing Oratore 2010) Similarly, if the policy underpinning the legislation is unclear,then it is difficult,Sawyerargues,forthe statute tobe effective.(p. 322)

He explains that thetaxlegislative process in New Zealand is based upontheseprinciples,whicharehighlightedasprioritieswithin a systemof extensive collaborationbetweentaxacademics, thegovernment, and thetaxauthority,but comparativelylimited input frompoliticians. Sawyersuggeststhat theheavyinvolvement oftaxacademicsin the designandreview of taxlegislation is both anadvantage, in terms ofthe lack of politicisation,yet alsoadisadvantage, in terms ofa struggle forlegitimacy. (p. 329)Turninghis attentionfromNewZealand to the UK, Sawyer proposes that the needto "fasttrack”legislationthroughParliamenthasmeant that effortstowardsclarity,sometimes, are simplylost.(pp. 330-331)Remindingus that the UK’s taxlegislation isthe longest in the world (p. 334), Sawyer writesthat hewelcomed theestablishment of the Officeof TaxSimplificationasastep in thedirection of addressingthe simplification oftaxpolicy,in its purestsense,and not justanodto improvingthe form oftaxlegislation. (p. 342)

Salience,complexity,andthemechanicsof(tax)lawproduction

In this sense, Sawyer isevocative of much of the literaturesurroundingthe benefits of“principlesbaseddrafting”.(Cooper,2010;Freedman, 2010) Principles based draftinghasbeenat theheart ofdiscussions in the UKabout the reform of complexlegislation, at leastsinceAveryJones’ famed call to place it at theheart of reformefforts,(AveryJones, 1996)andFreedman’sclarifyingreturn to the question in 2010.(Freedman, 2010) Principlesbaseddraftingis a subject offrequentdiscussion, but lessfrequentefforts towards definition– and,indeed, Cooper explicitlydeclines to define“principlesbaseddrafting,”although hedescribes it as similarto“anoperative rule”. (Cooper, 2010, p. 341)That is to say, “[i]t isnot a statement of the object or purposethat is to be accomplished byanotherrule, noris it anaid to interpretingthe operative provision.” (ibid) Whilstthe principlesbased draftingliteraturehasa comparativelylonghistoryin theUK,(AveryJones, 1996) Freedmanhascautioned that, in addition to acknowledgingits promiseforshorterlegislation and fewer

anti-avoidance provisions, it is not a“…solution toall ills”. (Freedman, 2010, pp. 717, 735)Onereason for caution is thevarietyof statutes themselves, whichrange inimportance.(Katzmann, 2012, p. 645)A furtherreason to view proposed“solutions to all ills” withsuspicion is the varietyofcultural, and indeed, structuraldifferences between differentformsof legislation. Braithwaite also has advisedcaution,andsuggests that generalprinciplesshould be balancedwith specificrules. (Picciotto, 2007, p. 15,citing Braithwaite, 2002,2003,2005)

Theaim of this section is not to revisit the principlesbased draftingdebate,whichhas beendeveloped amplybyleadingauthorities in the UK,Australia,andbeyond, but to step behindthe question of “how are draftspersons startingtheprocess of design– is itwith a clearprinciple in place?”,[4]to a more basic issue surroundingthenatureof thedraftspersonsthemselves. Is the capacityof alegislatureto articulatepolicywithin legislation in anywayinfluencedbysuperficialfactors,suchas size? Putdifferently-andremainingmindful ofFreedman’s warningthatprinciplesbased drafting is not an all-purpose panacea – aresometypes of legislatures more capable ofproducing principles basedlegislation(in all of itscomplexityof meaning) than others?If so, do these legislaturesprioritisehigh levels oftaxpayercomprehension, or salience, in the legislative process? Ultimately,is there aconnectionbetween taxsalience,andprinciplesbaseddrafting?

Principlesbased drafting is verymuch about the mechanics of lawproduction.Theconsideration of principlesbased draftingfortheUKlargelyhas proceeded over theyearswith comparativeglanceselsewhere. This is not unusual, and, indeed, it has been common, perhapsatleastsince(in the US) Surrey, tofindthat comparativeanalyses oftaxsystemsare underpinnedbyreferencetomechanicaldifference. Although these differences are highlighted as the point of the comparative exercise, the pragmatic value or normative underpinning of the exercise itself is less clear. An assumption of commonality amongst differing common law tax collection systems is discernible. InSurrey’scase, he suggested thatcoherenttaxlawmaybeeasierto produce inthe UK,than in theUS, becauseof thepresenceinthe UK of the Parliamentarywhip system.(1956, pp. 1154-1157)Similarly, as mentioned,infra, Sawyersuggests that a significant,comparative(withNewZealand)featureof UK taxlawmakingis the“fast tracking”procedure. (2013, pp. 230-231)