Testing the Effect of Tax Office and Client Interactions on the Meeting of Employer Obligations

How to cite

Sinning, Mathias; Biddle, Nicholas. 2016. "Testing the Effect of Tax Office and Client Interactions on the Meeting of Employer Obligations." AEA RCT Registry. AEARCTR-0000838. https://doi.org/10.1257/rct.838-2.0

Download: BibTeX · RIS — for reference managers without a browser translator. Zotero and most others pick up this page's citation metadata directly.

Mathias Sinning* , Nicholas Biddle

* Primary investigator

Registry ID
AEARCTR-0000838
DOI
https://doi.org/10.1257/rct.838-2.0
First registered
Last updated
Status
Completed
Trial dates
2016-02-29 – 2016-12-31
Countries
Australia ()
JEL codes
H25, H26, H83
Registry record
http://www.socialscienceregistry.org/trials/838

Abstract

The objective of this trial is to improve internal procedures and client engagement strategies of the Business Area "Employer Obligations" (EO) within the Australian Taxation Office (ATO). Employers in Australia have tax, superannuation, and other government obligations that vary depending on whether workers are employees or contractors. For employees, employers may have obligations for pay-as-you-go (PAYG) withholding, superannuation, and fringe benefits tax. For contractors, employers may have superannuation obligations. In addition to these obligations, employers may have to meet other federal, state and territory obligations (such as occupational health and safety requirements).

This trial consists of two parts. The first part of the trial aims to assess the effect of changing internal procedures (by adopting the procedures of a different ATO Business Area, "Indirect Tax" (ITX)) on the debt collection rate, the duration of case cycles, and the quality of the interactions between EO field auditors and their clients. The analysis is based on de-identified administrative data provided by the ATO and data obtained from a survey of field auditors at the beginning and the end of the trial.

The second part of the trial intends to improve the interaction between EO desk auditors and their clients. The intervention focuses on studying the effect of changing the contents of letters that desk auditors send to their clients and the effect of changing the script that desk auditors use for phone conversations. The outcome measures are (1) the debt collection rate, (2) the propensity of clients to lodge their tax return, (3) the duration of case cycles, and (4) the quality of interaction between EO desk auditors and their clients (based on a survey of desk auditors at the beginning and the end of the trial).

Intervention

The trial consists of two parts:

- Part 1: Change in internal procedures used by field auditors of the ATO. Copies of the procedures are provided in Appendix A of the analysis plan.
- Part 2: (a) Change in the script that desk auditors use for phone conversations and (b) change in the contents of a letter that desk auditors send to their clients. Phone scripts are provided in Appendix B of the analysis plan. Sample letters are included in Appendix C of the analysis plan.

The treatment effect of each intervention will be measured by comparing average outcomes between treatment and control group.

Experimental design

Our analysis is based on four sub-samples: (A) treatment cases of field auditors, (B) control cases of field auditors, (C) treatment cases of desk auditors, and (D) control cases of desk auditors. We will compare the members of sample A to the members of sample B; the members of sample C will be compared to the members of sample D. Moreover, we will compare the members of each treatment group (A or C) to all members of the two control groups (B and D). We will estimate separate regression models for each of the four comparisons: (Case I) A vs. B, (Case II) C vs. D, (Case III) A vs. B+D, and (Case IV) C vs. B+D.

Design details

Randomization method
Randomization is based on a random variable generator in Stata, using a random choice of the underlying seed. A copy of the Stata code that has been written to randomly assign field and desk auditors to treatment and control groups is included in Appendix D of the analysis plan.
Randomization unit
Intervention 1: 5 sites; 30 EO field auditors; up to 600 clients. Intervention 2: 2 sites; 12 desk auditors; up to 240 clients.
Clusters
Intervention 1: 15 EO field auditors were randomly assigned to the Treatment Group; 15 EO field auditors were randomly assigned to the Control Group; it is expected that each team member will complete about 10 and no more than 20 cases. Intervention 2: 6 EO desk auditors were randomly assigned to the Treatment Group; 6 EO desk auditors were randomly assigned to the Control Group; it is expected that each team member will complete about 10 and no more than 20 cases.
Observations
Intervention 1: 30 EO field auditors; about 300 clients. Intervention 2: 12 desk auditors; about 120 clients.
Arms
Intervention 1: Treatment Group: 15 EO field auditors; about 150 clients; Control Group: 15 EO field auditors; about 150 clients. Intervention 2: Treatment Group: 6 EO field auditors; about 60 clients; Control Group: 6 EO field auditors; about 60 clients.
Primary outcomes
We will use the following outcome measures in our analysis: 1. Continuous: - The duration of the case cycle - The typical duration of the case cycle reported by the auditor (based on survey data, see Appendix F of the analysis plan) - The duration until the current business activity statement (BAS) is lodged - The amount owed by the taxpayer at the start of the case - The amount collected during the audit - The amount owed by the taxpayer after the case is closed (updates daily with general interest charge) - The debt estimate raised by ATO during the audit 2. Binary: - A variable indicating whether an activity statement has been lodged by the taxpayer during the audit - A variable indicating the status of the BAS after the case is closed - A variable indicating whether an arrangement has been made with the taxpayer during the audit - A variable indicating whether the ATO lodged an overdue default assessment 3. Ordered: - The payment conversations with my clients are easy to establish and maintain (based on survey data, see Appendix F of the analysis plan) - The payment conversation procedures (e.g. letters sent to clients, telephone scripts, field audit guidelines) are easy for me to understand and implement (based on survey data, see Appendix F of the analysis plan) - The payment conversation procedures (e.g. letters sent to clients, telephone scripts, field audit guidelines) are helpful in the process of obtaining payment of outstanding liabilities (based on survey data, see Appendix F of the analysis plan) - Audit cases are referred to the Debt Business Line at an appropriate point in time, so that the chances of obtaining outstanding liabilities are maximised (based on survey data, see Appendix F of the analysis plan) - It is easy to negotiate payment arrangements for outstanding liabilities (based on survey data, see Appendix F of the analysis plan) - I encounter clients who are willing to pay outstanding liabilities (based on survey data, see Appendix F of the analysis plan) - I feel rushed to finish audit cases in order to meet recommended time limits (based on survey data, see Appendix F plan) - Negotiating payment arrangements with clients is stressful (based on survey data, see Appendix F of the analysis plan) - I am given sufficient autonomy to make arrangements that will likely result in the payment of outstanding liabilities (based on survey data, see Appendix F of the analysis plan) 4. Survival data: - The number of days until the current business activity statement (BAS) is lodged

Papers from this trial

  1. Biddle, Nicholas, Fels, Katja, and Mathias Sinning (2017): Behavioral Insights and Business Taxation: Evidence from Two Randomized Controlled Trials, Tax and Transfer Policy Institute (TTPI) Working Paper No. 2/2017, Australian National University. https://taxpolicy.crawford.anu.edu.au/sites/default/files/publication/taxstudies_crawford_anu_edu_au/2017-06/complete_bi_biddle-fels-sinning.pdf

Cited by

No citation data available for this registration yet.

Keywords