Role of Intelligence in Modern Tax Administration Dumre

A changing role of Economic Intelligence Unit (By Pramod Dumre, Deputy Commissioner, EIU) All over the world tax departments are adopting onept o! voluntary tax omplian ompliane e whih is a trust "ased "ased system# Under modern modern tax system, only limited limited num"er o! ases an "e seleted !or audit or srutiny "eause o! sare resoures#  $hus the trust "ased ompliane system re%uires regulation through a entrali&ed monitoring system to ath hold o! those taxpayers who are misusing the system !or under'reporting o! tax o"ligation or who are involving in tax !rauds or who have intelligently devised their omplianes !or tax avoidane# $his allows appropriate prioriti&ation o! audit and taxpayer servie worload and ena"les the alloation o!  resoures to high'ris groups#  $he main !untions o! modern tax administration is to manage tax ompliane (a) "y deteting and preventing delin%uent "ehavior and (") "y providing taxpayer servie and eduation in order to help taxpayers to disharge their tax o"ligations with ease# $hus a onept o! entrali&ed Compliane is *anagement (C*) is "eing adopted "y various tax departments# C* C* is an impo import rtan antt elem elemen entt o! e+e e+eti tive ve and and ei eien entt manag managem emen entt o!  ompliane# -ust lie a private "usiness alloates its resoures to areas that have the most potential !or generating pro.ts, a modern tax administration uses methods that !ous on high'ris taxpayers to selet ases !or audits# $his targeted approah is liely to raise higher revenue and provide a stronger deterrene !or non'ompliane# In the last /0 years, inreased use o! external in!ormation and automation have allowed tax administrations to generate pre.lled tax returns 1!or example, in the 2ord 2ordi i ountr ountries ies (Denmar Denmar,, 3inland 3inland,, Ieland Ieland,, 2orway 2orway and  and 4weden 4weden)) , 4pain, 4pain, and Chile5, whih are !ollowed up with online notie o! audits# $he availa"ility o! validated data should "e taen into onsideration when .ne'tuning tax audit strategies and ris'"ased approahes# A on!erene was held in Istan"ul, $urey, $urey, in Deem"er /006, !or tax admin administ istrat rators ors in the Europe European an and Centra Centrall Asian Asian regio region n to intera interat t and disuss disuss the tehnial tehnial hallenges hallenges involved involved in implement implementing ing ris'"ased ris'"ased audits with their regional ounterparts and international experts# $he 7orld 7orld Ban supervised this pro8et, in whih many ountries made presentations# $his oasion was attended "y parti partiipa ipants nts !rom !rom /0 ountr ountries ies throug throughou houtt Europe Europe and Asia Asia to share share their their own own experienes, learn !rom eah other, and esta"lish a !oundation !or !uture ooperation and and in!o in!orm rmat atio ion n shar sharin ing g on the the vari variou ous s syst system ems s "eing "eing impl implem ement ented ed in thei theirr ountries# Partiipants represented Armenia, A&er"ai8an, Belarus, Bulgaria, Canada, 3rane, 9eorgia, 9eorgia, 9ermany, 9ermany, India, India, :a&ahsta :a&ahstan, n, the :yrgy& yrgy& epu"l epu"li, i, *ongolia, *ongolia, the 2etherlands, 4weden, $a8iistan, $urey, Uraine, the United :ingdom, the United 4tates, 4tates, and U&"eista U&"eistan# n# Parti Partiipant ipants s shared shared experien experienes es on how ris'"ased ris'"ased audit systems improve the eay o! revenue olletion and use priniples that an "e applie applied d to all segmen segments ts o! taxpa taxpayers yers## $hese $hese systems systems help help tax admini administra strator tors s to alloate resoures and manage per!ormane#  $he e+etive and eient implementation o! is "ased audits and C* re%uires re%uires use o! Business Intelligen Intelligene e tools# tools# Business Business intellige intelligene ne (BI) is an um"rella um"rella onep oneptt that that overs overs arhi arhite tetur tures, es, appli appliat ation ions, s, data"a data"ases ses,, analyt analytia iall tools tools and methodologies used !or deision support# Analyses o! historial and urrent data, situations and per!ormanes give deision maers valua"le insight and ena"le them to mae more in!ormed and "etter deisions# 3rom a proess point o! view BI an "e seen seen as a tran trans! s!or orma mati tion on wher where e data data are are .rst .rst re.ne re.ned d to in!o in!orm rmat atio ion, n, then then to deisions, and .nally to ations#  $he arhiteture o! BI has !our main omponents; (<) data warehouse, with its soure data= (/) "usiness analytis > a olletion o! tools !or manipulating, mining

and analy&ing the data in the data warehouse= (?) "usiness per!ormane management > methodology and appliations !or monitoring and analysing organi&ational per!ormane= and (@) user inter!ae, suh as a dash"oard#  $he Eonomi Intelligene Unit o! *4$D has used a"ove nowledge "ase e+etively !or managing tax ompliane "y the use o! automated systems to (<) gather in!ormation !rom external agenies and math it with dealer data using PA2 as a uni%ue taxpayer identi.ation num"er, (/) undertae seletive audits "ased on ris analysis and (?) provide timely ade%uate in!ormation to support management deision maing and tax poliy !ormulation#  $he *aharashtra 4ales $ax Department (*4$D) has innovatively used Business Intelligene (BI) 4ystem not only !or implementing onept o! Computer assisted audit tehni%ue (CAA$) in its internal proess o! auditing taxpayers details "ut also !or taing poliy deisions "ased on reports o! BI system# $he reent ativities "eing done in EIU are as under# <# 9eneration o! ases !or omprehensive audits where multiple veri.a"le ris parameters are applia"le# /# Identi.ation o! Issue Based Audit ases where ris parameters are diretly ationa"le and results in reovery o! taxes# $he ompliane an "e through diret we" site ompliane and losure o! audit issues in omplied ases are without human intervention# $his ompliane system is alled Computeri&ed Des Audit (CDA)# ?# $he generation o! Business Intelligene *odels in Data *ining tool suh as predition o! non'genuine dealers, ase seletion !or audit et# @# $he prioriti&ation o! non'omplianes suh as raning o! return non'.lers !or !ollow'up in the order o! tax redits passed on to "uyers# # $he supply hain Analysis identi!ying the laims o! "ogus re!und laimants in irular trading and "u+er suppliers# # $he star rating o! dealers !or pre!erential treatment in provision o! servies# # $he Deision 4upport 4ystem (D44) !or tax poliy deisions suh as F7hat I!G analysis !or showing the impat o! hange in tax rate on revenue o!  government# H# $he !oreast analysis !or revenue !oreasting "ased on historial trends as well as "ased on maro'eonomi !ators# 6# $he utlier Analysis !or identi.ation o! outlier taxpayers not dislosing tax o"ligation ommensurate with the prevailing trends in their setor# <0#EIU is now integrating all ris analyses into a ?0 degree ris pro.le report !or eah dealer# Challenges

I! we go through all a"ove ativities we o"serve that the strategy o! *4$D has "een reative# 7e are .nding out the disrepany a!ter we get transations data in 3orm e'0@ a!ter a"out 6 months !rom the end o! the year# $he dealers as well as oers o! *4$D are engaged in resolving issues o! past periods where there might "e inadvertent errors in uploading o! data or the dealers were unaware a"out the de!aulting status o! their suppliers# 7e an proatively prevent suh ourrenes o!  tax evasion either ' <# By providing validations at the time o! upload o! transation data along with eturns,

/# By providing real time in!ormation to dealer a"out status o! their suppliers J "uyers, and ?# By auditing the dealers on real time on the "asis o! Data *ining modeling and Preditive Analytis#  $wo partiularly prominent uses !or data mining are identi.ed !or any tax administration; (<) Risk Rating: $ax administrations an "uild up truly ris'"ased worKows !or proessing the registration, .ling, reporting and payment transations that the taxpayers mae or should mae# A omprehensive ris rating, "ased on data mining modeling, an "e applied to eah transation so that all availa"le relevant data are utilised# As a result, high'ris transations an "e Kagged !or ase' spei. treatment while low'ris transations an move on to automated routine proessing# (/) Compliance Rating (Star Rating); $ax administrations an segment the taxpayers and identi!y segment spei. ompliane pro.les in terms o! diverse a"ilities and propensities to omply# $his helps tax administrations "etter design and target their servies and ompliane ations# *4$D has already started apturing party'wise sales and purhase transations data along with returns and also the mismathes J un'mathes are "eing shown to the dealers on *ahavat portal a!ter eah return period# 7e have also planned to implement CDA !or diret we"'site ompliane !or weeding out pending audits up to the periods /0
<# 7e started apturing annexures along with returns during /0<@'0<, "ut we are not validating 4upplier $I2 status while upload o! data# *4$D is maing availa"le on portal some parameters o! non'admissi"le I$C a!ter eah return period, "ut there is no ompliane mehanism !or the dealer# 4uh ompliane mehanism will only "e availa"le when we implement 4AP Automation system within one year# 7e have purhased 4AP $ax and evenue *anagement ($*) whih is a C$4 produt along with their other supporting modules lie C* (Customer elationship *anagement), IC* (Integrated Cases *anagement), et# I! we use the apa"ility o! 4AP tools then while aepting returns along with sales and purhase transations we an ensure proper allowane o! input tax redits and monitor the redit Kow e+etively "y showing the Computeri&ed Des Audit results !or every return period#

/# $he issue o! invoie mathing an "e resolved "y implementing mandatory e' invoiing# $he ountries where there was mathing o! sellersG and "uyersG invoies !or allowing tax redits are now shi!ting to the onept o! mandatory e'Invoiing# $he ountries lie 4outh A!ria, Bra&il, $aiwan, Long :ong, 4ingapore, Australia, and 2ew Mealand are some o! the ountries where tax redits are validated with sellerGs invoies "ut they are now shi!ting to e' invoiing to get real'time validation o! I$C# $he ountry lie China has implemented 9$P > 9olden $ax Pro8et using I$ system to ur" NA$ evasion "y mathing tax redits# 4AP has now a%uired Ari"a, a ompany whih provides onneted solutions that ena"le unparalleled olla"oration and "usiness per!ormane within and "etween enterprises inluding generation o! eletroni invoies# I! we implement mandatory e'invoiing using 4AP tools lie that !rom Ari"a !or B/B transations then mathing o! transations will not "e re%uired !or on.rming I$C# ?# At present we are analy&ing data a!ter a"out a year# 2ow as we are getting transations data along with return we an generate is ating !or all the !untions on monthly "asis# $hus real time ris ratings an "e generated on 4AP Business Intelligene and Business "8ets (BIB) $ool !or egistration ris in ase o! !raudulent dealer, eturn non .ler ris where "uyers are laiming I$C or Bogus re!und ris due "u+er transations et# as now all the data o!  Portal and $ransations are availa"le on 4AP BI server# @# $he EIU Power Users will "e given speial privileged aesses on 4AP BIB !or analy&ing data on real time "asis# $here will "e a 2ew EIU *odule on 4AP through whih there will "e exhange o! data "etween 4A4 and 4AP servers#  $hrough Integrated Case *anagement tool EIU will monitor the progress o!  ationa"le ases and also !eed"a o! eah is parameter will "e automatially aptured where there is any ation taen "y any !untional oer# $he ases will "e distri"uted through 4AP IC* and a!ter ation through C* *odule, !eed"a will "e aptured# # $he ?0 Degree is Pro.le o! Dealers along with 4tar rating !or in'time omplianes will "e availa"le to eah !untional oer with respet to his !untion# # Under the new 4AP 4ystem we will "e apturing L42 odes along with an annual return and 3orm e'0@# $hus it will now "e possi"le to e+etively arry out ommodity analysis partiularly $ax ate mislassi.ation analysis# # I! we use the Data *ining tools properly then the Pattern Analysis and preditive analysis is possi"le !or identi!ying year'on'year trends as under' a# $he dealers who are ha"itual short .lers an "e predited to "e short .lers "# $he dealers who are ha"itual non'.lers an "e predited to "e non'.lers # $he dealers who are ha"itual suppressors an "e predited d# $he pattern o! tax reeipts loation'wise an also "e studied# e# $he gap analysis in terms o! setor'wise revenue an "e studied on historial trends# Data mining an provide novel and e+etive ways !or tax administrations to develop their in!ormation systems and there"y redue the in!ormation asymmetry vis'O'vis taxpayers# $his is mainly due to "eing a"le to om"ine data !rom various

soures and onsolidate plenti!ul varia"les into easy'to'interpret, ationa"le lassi.ations and preditions# Being "etter in!ormed, tax administrations an improve their ontrol measures and the hit rate o! deteting !raud and errors# IrelandGs initiative o! developing audit target seletion, !or instane, provides support to this assumption In addition to the diret "ene.t o! the smarter ontrols in atually athing the !raudsters, there is also the potential indiret "ene.t o! less attempted !raud, owing to raising pu"li awareness o! the tax administrationGs "etter ontrols# $axpayers are liely to align their "ehaviour in line with the applia"le tax laws and regulations a!ter reali&ing that the tax administration annot "e deeived#  $he U4AGs initiative to use ard payment data to improve small "usinessesG transpareny is an example where Data mining, with its lustering tehni%ues, an help tax administrations segment their heterogeneous taxpayer "ases into suiently uni!orm su"populations whose needs, "ehaviour and attitudes an "e disovered, understood and responded to# AustraliaGs and DenmarGs initiatives to determine the olletion ations with the use o! analytis, as well as AustraliaGs notion o! developing the overall servie orientation give support to this assumption# Data mining, most nota"ly with its soial networ analysis tehni%ues, an help tax administrations get hold o! the ever growing omplexity o! "usiness relationships, ownership strutures and other types o! linages and dependenies among and aross taxpayers# $he in Nisuali&ation Analysis or 2etwor diagram is used to .nd the relationships in a luster o! taxpayers# EIU has used this tehni%ue to identi!y Cirular trading e!und !rauds# $he examples progressing towards this diretion are in at least the Australian, Irish and 4wedish tax administrations# $hus in !uture the use Data mining and preditive analytis would inrease in EIU Analysis#