Admin Data Eyed as Substitute for Labour Income Survey

We are seeking feedback on whether admin data could be used to create viable replacements for Household Labour Force Survey (HLFS) income measures. Read our evaluation and share your feedback on Proposed changes to Household Labour Force Survey income data.

Evaluating the use of admin data as a substitute for Household Labour Force Survey income variables explores whether admin data could be used, alone or linked to the Household Labour Force Survey (HLFS), to create HLFS income survey measures.

Summary of key points

Stats NZ's strategic priority of becoming an organisation that uses administrative (admin) data first relies on evaluation of current measures against admin alternatives so potential changes are well understood. Even where the underlying admin data quality is high, as is the case for income data, comparisons need to be created and shared to determine whether newly derived measures based on admin data would be timely, high-quality, and comprehensive enough to meet customer needs.

Our comparison of HLFS income measures with those reproduced using a linked HLFS and admin income dataset, and a full-coverage admin income dataset, indicate that:

  • current HLFS income measures are high quality and highly comparable with admin data results where substitution is viable
  • linked HLFS and admin data for wage and salary, government transfer, and ACC income appear to be viable substitutes for HLFS income measures
  • unlinked admin data is not a viable substitute for HLFS hourly income measures
  • self-employment income records are not timely enough for replacement using linked or standalone admin data.

We recommend the following two future state options are considered further:

  • Option 1: Continue with the current state of collecting HLFS income measures through survey questions in June quarters.
  • Option 2a: Shift to a linked HLFS and admin data model in which wage and salary, government transfer, and ACC income are provided through admin data, while self-employment continues to be collected in the HLFS.

Option 1 will remain the default position if there is no further work on option 2a.

Work towards option 2a will need to include:

  • customer consultation (to confirm whether linked admin income data would be a viable alternative)
  • a project to build and test how linking these sources would work in a production environment, along with further testing of results
  • adjustments to HLFS income survey content.

We recommend one year minimum in which both options are run simultaneously before introducing any changes that would affect the collection of HLFS income survey questions.

ISBN 978-1-991431-13-4

/Stats NZ Public Release. This material from the originating organization/author(s) might be of the point-in-time nature, and edited for clarity, style and length. Mirage.News does not take institutional positions or sides, and all views, positions, and conclusions expressed herein are solely those of the author(s).View in full here.