[{"data":1,"prerenderedAt":-1},["ShallowReactive",2],{"detail-sidebar-cat-1-en-105":3,"doc-seo-239515-105":53,"doc-detail-239515-en":126},{"code":4,"msg":5,"data":6},0,"success",[7,14,19,24,29,34,39,44,49],{"id":8,"doc_module":9,"doc_module_name":10,"category_name":11,"show_sort_weight":12,"slug":13},11,1,"Template","Presentations",90,"presentations",{"id":15,"doc_module":9,"doc_module_name":10,"category_name":16,"show_sort_weight":17,"slug":18},12,"Resumes",80,"resumes",{"id":20,"doc_module":9,"doc_module_name":10,"category_name":21,"show_sort_weight":22,"slug":23},14,"Invoices",70,"invoices",{"id":25,"doc_module":9,"doc_module_name":10,"category_name":26,"show_sort_weight":27,"slug":28},15,"Posters",60,"posters",{"id":30,"doc_module":9,"doc_module_name":10,"category_name":31,"show_sort_weight":32,"slug":33},16,"Social Media",50,"social-media",{"id":35,"doc_module":9,"doc_module_name":10,"category_name":36,"show_sort_weight":37,"slug":38},17,"Forms",40,"forms",{"id":40,"doc_module":9,"doc_module_name":10,"category_name":41,"show_sort_weight":42,"slug":43},18,"Letters",30,"letters",{"id":45,"doc_module":9,"doc_module_name":10,"category_name":46,"show_sort_weight":47,"slug":48},21,"Paper Templates",5,"papers-templates",{"id":50,"doc_module":9,"doc_module_name":10,"category_name":51,"show_sort_weight":4,"slug":52},158,"General","general-158",{"code":4,"msg":54,"data":55},"ok",{"site_id":56,"language":57,"slug":58,"title":59,"keywords":60,"description":61,"schema_data":62,"social_meta":119,"head_meta":121,"extra_data":123,"updated_unix":125},105,"en","practical-session-tax-gap-estimation-toolkit-part-2","Practical session: Tax Gap estimation - Toolkit Part 2","","Training slides for a practical session on tax gap estimation using the ATI Tax Gap Estimation Toolkit (Toolkit Part 2). The session recap focuses on the BE approach, data cleaning, preparing and importing data into Stata, handling duplicates, and merging audit and return data. Participants practice VAT gap estimation with machine learning (Random Forest), including setting up VAT variables, defining RF iterations and variable replacements, generating audit/evasion indicators, and running the RF model. Exercises also cover disaggregated results and calculating VAT gaps by return year.",{"@graph":63,"@context":118},[64,80,101],{"@type":65,"itemListElement":66},"BreadcrumbList",[67,71,74,77],{"item":68,"name":69,"@type":70,"position":9},"https://docshare.wps.com","Home","ListItem",{"item":72,"name":10,"@type":70,"position":73},"https://docshare.wps.com/template/",2,{"item":75,"name":11,"@type":70,"position":76},"https://docshare.wps.com/template/presentations/",3,{"item":78,"name":59,"@type":70,"position":79},"https://docshare.wps.com/template/practical-session-tax-gap-estimation-toolkit-part-2/239515/",4,{"url":78,"name":59,"@type":81,"image":82,"author":87,"headline":59,"publisher":90,"fileFormat":93,"inLanguage":57,"description":61,"dateModified":94,"datePublished":95,"encodingFormat":93,"isAccessibleForFree":96,"interactionStatistic":97},"DigitalDocument",{"url":83,"@type":84,"width":85,"height":86},"https://docshare.wps.com/thumbnails/practical-session-tax-gap-estimation-toolkit-part-2/239515.png","ImageObject",442,249,{"name":88,"@type":89},"8796093062539","Person",{"url":68,"name":91,"@type":92},"DocShare","Organization","application/pdf","2026-11-04","2026-09-11",true,{"@type":98,"interactionType":99,"userInteractionCount":25},"InteractionCounter",{"@type":100},"ViewAction",{"@type":102,"mainEntity":103},"FAQPage",[104,110,114],{"name":105,"@type":106,"acceptedAnswer":107},"What is the main objective of the session on tax gap estimation?","Question",{"text":108,"@type":109},"The session demonstrates the steps toward tax gap estimation using the toolkit, including how to investigate disaggregated results and the effects of aggregate tax gaps.","Answer",{"name":111,"@type":106,"acceptedAnswer":112},"Which software and machine learning approach are used for estimating the VAT gaps?",{"text":113,"@type":109},"The exercises focus on preparing a Stata environment and using machine learning code with a Random Forest (RF) model for VAT gap estimation.",{"name":115,"@type":106,"acceptedAnswer":116},"How is the VAT gap by return year calculated in Exercise 2?",{"text":117,"@type":109},"Exercise 2 collapses assessment and VAT payment by return year, computes potential VAT (v01) as the sum of assessment and VAT payment, then calculates the annual percentage VAT gap (vat_gap01) from assessment divided by v01 and multiplied by 100.","https://schema.org",{"og:url":78,"og:type":120,"og:title":59,"og:site_name":91,"og:description":61},"article",{"robots":122,"canonical":78},"index,follow",{"doc_id":124,"site_id":56},239515,1791480813,{"code":4,"msg":5,"data":127},{"doc_id":124,"user_id":128,"nickname":88,"user_avatar":60,"doc_module":9,"category_id":8,"category_name":11,"doc_title":59,"doc_description":61,"doc_content":129,"file_id":130,"file_url":131,"file_type":132,"file_size":133,"view_count":25,"is_deleted":4,"is_public":9,"is_downloadable":9,"audit_status":9,"page_count":134,"language":135,"language_code":57,"site_id":56,"html_lang":57,"table_of_contents":136,"faqs":137,"seo_title":138,"seo_description":61,"update_tm":139,"read_time":140},8796093062539,"United Nations University  \nWorld Institute for Development Economics Research  \n09/10/2024  \nPractical session: Tax Gap estimation  \nBreak out groups:  \nMaria Jouste + Gerald Agaba Amina Ebrahim + Mostafa Bahbah Kwabena Adu-Ababio + Hilja-Maria  \n2024 ATI Tax Gap Estimation Toolkit Training | October 7-8 2024  \nRecap of estimation goal  \nRecap of steps yesterday  \n• What did we achieve yesterday?  \n• 1. We covered the concepts of BU approach  \n• 2. Theory of Data cleaning  \n• 3. Collecting preparing data  \n• Importing data to Stata  \n• Dealing with Duplicates  \n• Merging Audit and return data.  \nAim of this session  \n1. Show/Demonstrate/Teach the steps towards the estimation – Toolkit Part 2.  \n2. Show how to investigate disaggregated results (by sector) .  \n3. Show how to investigate disaggregated results (by return year) .  \n4. Show how to investigate the effects/consequence of aggregate tax gaps.  \nWe are not teaching how to use Stata merely guiding the steps to follow. Own implementation requires contextualizing.  \nPreamble and Preparatory Steps  \nIn the following exercises we will use the Machine Learning code to estimate tax gaps for VAT CIT and PAYE. For practical reasons we focus on only one tax type. However we encourage you to practice with the other tax types and use other variables to fully understand the usefulness of the toolkit. Before engaging in the preparatory steps run the code lines 18 to 30 to prepare the Stata environment.  \nSTEP 1 : Declare VAT variables needed for Random Forest (RF) Model  \n• VAT variables: rtn year tax  sales vat tax  sales zero  rate  sales exempt  sales import  sales vat  import  sales export  sales capital  sales nondeduc  sales taxunregistered  sales vat taxunregistered  sales total  sales total vat  sales tax  purchases vat tax  purchases zero  rate  purchases exempt  purchases import  purchases vat  import  purchases export  purchases capital  purchases nondeduc  purchases taxunregistered  purchases vat taxunregistered  purchases total  purchases  \ntotal vat  purchases vat  payment city region isic sector tax  region  \nSTEP 2: Declare iterations needed and number of variables to replace for RF Model  \n• VAT selection: 400 iterations (insert figure in syntax “iter(400)”) and 32 variable replacements (insert figure in syntax “numvar(32)”)  \nSet path in line 36 or use similar path as done in Data Prep exercise. In line 47 indicate the data to use as “vat  audit”  \nExercise 1: Set up and utilities  \nAim: How to start and prepare for the gap estimation  \n1. To set up audit and evasion data for easy identification perform the following cleaning steps:  \na) Create a dummy variable “reg_audit” for all audits greater than zero.  \nb) Create a new variable “evaded” from the “vat_evasion” variable.  \nc) Run lines 67-69 to create and replace new variables (indicator and mimic_tax) .  \nd) Run the RF model using VAT variables and numbers from STEPs 1 and 2.  \nExercise 1 – Solutions  \na) Create a dummy variable “reg_audit” for all audits greater than zero.  \nb) Create a new variable “evaded” from the “vat_evasion” variable.  \nc) Run lines 67-69 to create and replace new variables (indicator and mimic_tax)  \nd) Run the RF model using VAT variables and numbers from STEPs 1 and 2 .  \nExercise 1 – Solutions continued  \nd) Run the RF model using VAT variables and numbers from STEPs 1 and 2.  \nExercise 2 – Estimating VAT gaps by return year  \nAim: Run gap estimation by return year.  \n2. Estimate VAT gaps by return year by editing lines 92-104 based on the following questions:  \na) Collapse the total assessment (assessment) and vat payment (vat_payment) variables using the return year (rtn_year) variable.  \nb) Generate the potential VAT (v01) by summing the total assessment (assessment) and VAT payment (vat_payment) variables.  \nc) Generate the percentage annual VAT gap (vat_gap01) by dividing the assessment (assessment) by the potential VAT (v01) variable and multiply the result by 100  \nd) Plot the ","cbCainqyatQTp1yp","https://ap.wps.com/l/cbCainqyatQTp1yp","pdf",1483501,23,"English","# Recap of estimation goal\n# Aim of this session\n# Preamble and preparatory steps\n## STEP 1: Declare VAT variables for Random Forest model\n## STEP 2: Declare iterations and variable replacements\n# Exercise 1: Set up and utilities\n## Exercise 1 – Solutions\n# Exercise 2: Estimating VAT gaps by return year\n## Exercise 2 – Solution","[{\"question\":\"What is the main objective of the session on tax gap estimation?\",\"answer\":\"The session demonstrates the steps toward tax gap estimation using the toolkit, including how to investigate disaggregated results and the effects of aggregate tax gaps.\"},{\"question\":\"Which software and machine learning approach are used for estimating the VAT gaps?\",\"answer\":\"The exercises focus on preparing a Stata environment and using machine learning code with a Random Forest (RF) model for VAT gap estimation.\"},{\"question\":\"How is the VAT gap by return year calculated in Exercise 2?\",\"answer\":\"Exercise 2 collapses assessment and VAT payment by return year, computes potential VAT (v01) as the sum of assessment and VAT payment, then calculates the annual percentage VAT gap (vat_gap01) from assessment divided by v01 and multiplied by 100.\"}]","Practical session: Tax Gap estimation - Toolkit Part 2 | PDF",1789149955,8]