Back Matter
Author:
Elizabeth Gavin
Search for other papers by Elizabeth Gavin in
Current site
Google Scholar
PubMed
Close

Annex 1. Examples of Differences between Administrative Data and Survey Data

This note emphasizes that users of administrative data and survey data need to understand differences between data from different sources. Some examples of typical differences that may be encountered in working with tax administration data and official statistics are as follows:

  • The entities contained in registers may differ. Taxpayer registers will contain records of legal and natural persons that are required to register for tax. In the case of large, complex businesses, it may be possible to have more than one registration for the same tax type (for example, where different payrolls are run within the business, each may be associated with a registration for pay-as-you-earn). SBRs may record distinct establishments, where components of an enterprise differ in economic activity and physical locations. As the economic activity of each entity in a taxpayer register or SBR should be recorded, this can, for example, lead to differences in sectoral aggregates.

  • Unless there is close collaboration between the RA and NSO on industry classification, there are many reasons why disparate sectoral aggregates may not be associated with the same set of businesses. One of the reasons for this follows from the point above: the economic activity of each entity in a taxpayer register or SBR should be recorded. Thus, an RA may be recording all the economic activity associated with a large business as one type of activity, while this may be broken down into several economic activities in the SBR.

  • Typically there is no exact correspondence between economic aggregates available from the compilation of National Accounts and any particular tax base. For this reason, adjustments are made to economic statistical aggregates to estimate tax gaps when applying RA-GAP methodologies (see, for example, Ueda (2018) where the relationship between the tax base for corporate income tax and gross operating surplus is set out).

  • Trends in tax revenue aggregates do not mirror trends in measures of economic activity. In addition to the point made previously, tax policy changes and taxpayer behavior (including compliance) also affect tax revenue aggregates over time.

  • There may be differences in the reference periods for statistical collections and the accounting periods for tax. Adjustments may be necessary to make meaningful comparisons of tax administration and economic statistics aggregates over time.

  • There may be timing differences in the recording of taxable transactions and economic activity in national accounts. The System of National Accounts requires that transactions be recorded on an accrual basis. Most RAs record transactions on a cash basis.

References

This paper was reviewed by Michael Keen and benefitted from comments by Katherine Baer, Joseph Crowley, Claudia Dziobek, Andrew Masters, Andrew Okello, Lisbeth Rivas, and Mick Thackray.

1

See IMF Technical Manuals and Notes TMN/17/04 (Hutton, 2017), TMN/17/05 (Thackray, ,2017) and TNMEA2018002 (Ueda, 2018) for detail on estimating the tax gap for VAT, Excise and CIT respectively.

2

See IMF Working Paper “Using Administrative Data to Enhance Policymaking in Developing Countries: Tax Data and the National Accounts” (Rivas 2018).

3

An example of how VAT fraud impacted trade and balance of payment statistics in the United Kingdom is discussed in the report “VAT Missing Trader Intra-Community Fraud: the Effects on Trade Statistics” (OECD 2004).

4

The application of revenue administration data to the evaluation and formulation of tax policy has been covered in another how-to-note (Grote 2017).

5

Rivas (2018) covers country examples showing the ways in which tax data have been used in compiling national accounts.

6

For simplicity, the term “taxpayers” will be used to refer to businesses, individuals, households, and other organizations in exercising either their obligations as taxpayers or as respondents to official statistical surveys.

7

TADAT describes the desired outcome of taxpayer registration as: “All businesses, individuals, and other entities that are required to register are included in a taxpayer registration database. Information held in the database is complete, accurate, and up-to-date.” (TADAT 2019)

8

Fifty-six and 58 percent, respectively, of tax administrations indicated that they had formal programs to improve register quality in 2015 and 2016, respectively (ISORA 2016). This figure rose to 73 percent in 2016 and 2017 (ISORA 2018).

9

See, for example, http://www.wiesbaden2018.bfs.admin.ch/ agenda-papers/ where country reports to the Wiesbaden Group identify inaccuracies in address information and economic classification in taxpayer registers as challenges.

10

Cooperation with data providers is necessary to understand the concepts, ensure the continued supply of data, and enable linkages to be made between various sources (United Nations 2020).

11

See, for example, the discussion on compliance risk management in TADAT (2019).

12

See, for example, the country reports to the Wiesbaden Group on Business Registers http://www.wiesbaden2018.bfs.admin.ch/agenda-papers/.

13

For a more detailed discussion, see Keen (2015).

14

TADAT (2019) requires a measure of “the extent of initiatives to detect businesses and individuals who are required to register but fail to do so.”

15

All tax administrations indicated that they had powers to gather information in ISORA 2016 (Crandall 2019).

16

Sixty-one percent of RAs undertake taxpayer surveys, and of these, half outsource surveys (ISORA 2016).

17

Many RAs publish statistics on revenue collection (for example, Rwanda [RRA 2020], South Africa [National Treasury 2020], Uganda [URA, 2020]). The publication of tax statistics can serve to improve public understanding of revenue collections, offer transparency on tax incidence, and promote accountability.

  • Collapse
  • Expand
How to Collaborate Effectively to Improve Data Quality and Use in Revenue Administration and Official Statistics
Author:
Elizabeth Gavin