15
Processing Methodology of Tax Processing Methodology of Tax Data at Statistics Canada Data at Statistics Canada Authors: François Brisebois, Richard Laroche and Rossana Manriquez (Statistics Canada) Presenter: Danielle Laroche (Statistics Canada) May 2011

Processing Methodology of Tax Data at Statistics Canada Authors: François Brisebois, Richard Laroche and Rossana Manriquez (Statistics Canada) Presenter:

Embed Size (px)

DESCRIPTION

13/02/2016 Statistics Canada Statistique Canada 3 Tax Data at Statistics Canada  StatCan has access to all tax data through a signed agreement with the Canada Revenue Agency (CRA)  Used in several surveys Primary use: as a direct replacement for collected units (or through a model) Secondary use: auxiliary information used for the purpose of survey design, E&I, data confrontation, etc.

Citation preview

Page 1: Processing Methodology of Tax Data at Statistics Canada Authors: François Brisebois, Richard Laroche and Rossana Manriquez (Statistics Canada) Presenter:

Processing Methodology of Tax Processing Methodology of Tax Data at Statistics CanadaData at Statistics Canada

Authors:François Brisebois, Richard Laroche and Rossana Manriquez

(Statistics Canada)

Presenter:Danielle Laroche (Statistics Canada)

May 2011

Page 2: Processing Methodology of Tax Data at Statistics Canada Authors: François Brisebois, Richard Laroche and Rossana Manriquez (Statistics Canada) Presenter:

23-05-04Statistics Canada • Statistique Canada2

Outline Tax data at Statistics Canada Processing of tax data

• Define each source presented• Overview of processing steps• Current challenges

Page 3: Processing Methodology of Tax Data at Statistics Canada Authors: François Brisebois, Richard Laroche and Rossana Manriquez (Statistics Canada) Presenter:

23-05-04Statistics Canada • Statistique Canada3

Tax Data at Statistics Canada StatCan has access to all tax data through a

signed agreement with the Canada Revenue Agency (CRA)

Used in several surveys• Primary use: as a direct replacement for

collected units (or through a model)• Secondary use: auxiliary information used for

the purpose of survey design, E&I, data confrontation, etc.

Page 4: Processing Methodology of Tax Data at Statistics Canada Authors: François Brisebois, Richard Laroche and Rossana Manriquez (Statistics Canada) Presenter:

23-05-04Statistics Canada • Statistique Canada4

Tax Data at Statistics Canada Three sources discussed:

• Business tax Two sources: T1 form, T2 form Annual data

• Individual consumption tax Goods and Services Tax (GST) Monthly data

Focus of the paper is on the processing of these three sources

Page 5: Processing Methodology of Tax Data at Statistics Canada Authors: François Brisebois, Richard Laroche and Rossana Manriquez (Statistics Canada) Presenter:

23-05-04Statistics Canada • Statistique Canada5

Processing of Tax Data At the CRA

• Only minimal editing is performed for the context of statistical use

At Statistics Canada• Centralized processing (in the Tax Data Division)• Thorough process• Ultimate products: tax data files made available to

survey programs

Page 6: Processing Methodology of Tax Data at Statistics Canada Authors: François Brisebois, Richard Laroche and Rossana Manriquez (Statistics Canada) Presenter:

23-05-04Statistics Canada • Statistique Canada6

T1 – Definition The T1 form is the main form filled out annually

by all Canadians to report their income tax Business surveys are interested in income

earned from self-employment (unincorporated businesses)• Farming, Fishing, Rental, Professional, Commission,

Business• 85% are filled out electronically (available); the

remaining 15% are reported on paper (not available)

Page 7: Processing Methodology of Tax Data at Statistics Canada Authors: François Brisebois, Richard Laroche and Rossana Manriquez (Statistics Canada) Presenter:

23-05-04Statistics Canada • Statistique Canada7

T1 – Processing Two files

• Assessed Record File: two variables for everyone Used as a starting point for the imputation of detailed

variables not available for non-electronic filers• “T1 data”: detailed variables for electronic reports only

Identification of partnerships Balancing edits Imputation of missing records Derived variables (using standardized financial

statements)

Page 8: Processing Methodology of Tax Data at Statistics Canada Authors: François Brisebois, Richard Laroche and Rossana Manriquez (Statistics Canada) Presenter:

23-05-04Statistics Canada • Statistique Canada8

T2 – Definition Corporate tax returns

• Filled out annually• Flexible reporting period• Only 8 fields are mandatory, but many others are

of interest to StatCan survey programs

Page 9: Processing Methodology of Tax Data at Statistics Canada Authors: François Brisebois, Richard Laroche and Rossana Manriquez (Statistics Canada) Presenter:

23-05-04Statistics Canada • Statistique Canada9

T2 – Processing Edits: balancing, invalid or inconsistent, outliers Imputation: deterministic, missing tax data Generic to detail allocation

• Block composed of one generic field and detail fields• Goal is to allocate reported generic amounts to detail fields

when it has not been done• Groups of “detail respondents” are formed; their distribution

is used to impute “generic respondents” Derived variables (standardized financial statements)

Page 10: Processing Methodology of Tax Data at Statistics Canada Authors: François Brisebois, Richard Laroche and Rossana Manriquez (Statistics Canada) Presenter:

23-05-04Statistics Canada • Statistique Canada10

T1 & T2 – Challenges Explore other sources to improve processing Reporting on data quality

• Variance due to imputation Harmonise the processing with the few

remaining programs having their own system

Page 11: Processing Methodology of Tax Data at Statistics Canada Authors: François Brisebois, Richard Laroche and Rossana Manriquez (Statistics Canada) Presenter:

23-05-04Statistics Canada • Statistique Canada11

GST – Definition Tax levied on final consumption of products

and services (VAT) Collected and remitted by businesses with

revenues over $30K• Remitting frequency depends on business size

Two main variables: • Sales• Tax collected

Page 12: Processing Methodology of Tax Data at Statistics Canada Authors: François Brisebois, Richard Laroche and Rossana Manriquez (Statistics Canada) Presenter:

23-05-04Statistics Canada • Statistique Canada12

GST – Processing Monthly file sent from CRA, 7 weeks after

reference month Complex steps:

• E&I Missing, inconsistencies, outliers

• Calendarization Break non-monthly transactions into monthly

values, taking into account seasonal patterns• Allocation

From high-level reporting units to detailed level units (from enterprise to establishment levels)

Page 13: Processing Methodology of Tax Data at Statistics Canada Authors: François Brisebois, Richard Laroche and Rossana Manriquez (Statistics Canada) Presenter:

23-05-04Statistics Canada • Statistique Canada13

GST – Challenges Complexities from the fact that (recent) historical

data are reprocessed every month• Only 70% of the expected data are received on time• Annual and quarterly remitters are “extrapolated” until

reported data are received Fine-tuning the methodology

Page 14: Processing Methodology of Tax Data at Statistics Canada Authors: François Brisebois, Richard Laroche and Rossana Manriquez (Statistics Canada) Presenter:

23-05-04Statistics Canada • Statistique Canada14

Conclusion Centralizing and enhancing the processing

methodology has increased tax data quality• Which in turn has encouraged a greater use

Increased use by survey programs, with strong support by management

Possibility of using more tax data?• Options are being examined in the context of

a project to integrate business surveys

Page 15: Processing Methodology of Tax Data at Statistics Canada Authors: François Brisebois, Richard Laroche and Rossana Manriquez (Statistics Canada) Presenter:

23-05-04Statistics Canada • Statistique Canada15

Thank you

For more information, Pour plus d’information,please contact the authors: veuillez contacter les auteurs:

François Brisebois, [email protected] Laroche, [email protected]

Rossana Manriquez, [email protected]