MSD Review and evaluation failures


The May 2021 report by the Auditor-General on the wage subsidy stated:

4.77 As part of our 2019/20 annual audit of the Ministry of Social Development, our Appointed Auditor recommended that the Ministry write to recipients of subsidy payments to seek re-confirmation that they complied with the requirements and obligations. This is consistent with one of Deloitte’s recommendations to increase communications with recipients. We have also included it as part of Recommendation 4.

4.78 Using this type of correspondence is one of the anti-fraud techniques identified as being good practice for emergency situations.

Recommendation 4:

Seeking written confirmation from applicants (which could be targeted towards larger or risk-indicated applicants) of compliance with the eligibility criteria and the obligations of receiving the subsidy.

COMMENT:  The MSD ignored the two 2020 recommendations to write to all recipients. It also did not properly comply with the 2021 recommendation which clearly was about writing and correspondence but did not specify how many recipients to contact.

1000 recipients were asked to confirm their compliance and eligibility.  An OIA request was made in September 2021 for a copy of the letter sent to recipients but it has not been supplied.

OIA Response 19 May 2022

In regard to the sample of 1,000 early Wage Subsidy recipients written to, as at May 2022, 717 replies had been received, of which 713 employers confirmed they met the eligibility criteria and complied with their obligations.

Two reminder emails to non-respondents were delayed due to the establishment of the August 2021 Wage Subsidy scheme, and were sent on 18 February 2022 and 8 April 2022. Where a recipient does not respond to the reminder emails, MSD intends to proactively contact them to confirm their response.

COMMENT: No verifying documents were requested so replying was easy. However, 29% did not confirm that they had met the eligibility criteria and   complied with post payment obligations. This indicates that if an appropriate email had been sent to all recipient, a huge number of repayments could have been requested and received.


The MSD ignored the following paragraphs in the May 2021 Auditor-General report and told staff to contact recipients by phone.

4.5 Most of the post-payment reviews involved phone conversations with people who received the subsidy payment. How accurately those conversations reflect the recipient’s actual situation – for example, as recorded in their business accounts – is unknown. In our view, the reliability of those reviews should be tested against additional evidence.

4.64 This is because the work that was carried out did not routinely involve substantiation of information against a secondary source, such as requesting and reviewing documents to verify information provided verbally.

There were several other references to the need for verifying documents to be provided.  The following information comes from a desk file for staff carrying out the review and and from Appendix3.

19.  It was recommended that the Ministry of Social Development test a sample of these reviews against documentary evidence. This guidance is intended to respond to that recommendation.   

20. It was also recommended that the Ministry of Social Development seek written confirmation from applicants of their compliance with the eligibility criteria and the obligations of receiving the subsidy. Responding to this recommendation is part of a separate workstream.

COMMENT:  It should have been a simple matter to send out a standard email letter requesting the following most important documents that could have been requested in the signed Declaration or in a post payment email to all applicants:

(a)  They had a 30% or 40% drop in revenue.

(b)  They had no cash reserves and could not get assistance from their bank.

(c)  They gave each employee details of the wage subsidy being requested for that employee and got their consent in writing.

(d)  They informed the Ministry of Social Development within 5 working days if anything changed that affected their eligibility or                           entitlement to the subsidy.  (including a post lockdown boom with above normal revenue in the subsidy period)

(e)  They repaid all or part of the subsidy if they were not entitled to retain it.    

22. The process for the post-payment integrity check is as follows:

A number of steps are outlined. Staff are required to first get an understanding of the business, check the wage subsidy history of the business and to check all IRD, Government and publicly available information.

The staff member is then told to “Engage with the applicant to understand their business, the impacts of the Alert Level restrictions during COVID-19 on their business, and how they have made an assessment that they fit the eligibility criteria. Question their recollection of the original application.”

The person spoken to, based on separate conversation guidelines, is to be asked to provide evidence that payment was received for the correct number of staff. Also requested was “2.2 Information evidencing a 30% and/or a 40% decline in revenue (if they applied under the Wage Subsidy Extension Scheme) related to COVID-19. This may include financial or bank statements.”

“What evidence could you show us that your business has met the revenue decline test.”

“Request calculations of revenue decline by email.”

COMMENT: The documents mentioned were optional. A revenue statement could be easily made up so bank statements or better still, GST Returns (as suggested by the Auditor-General) should have been a basic requirement.

It should have been stated that the month/s before and after the revenue drop were required in order to show that the drop was genuine.  The applicant was to be asked in conversation whether they had engaged with their bank, drawn on reserves etc but no documents were required.

Under Appendix 3 the following was stated:

“Documentary Evidence Below are some examples of the types of information that may be sought as verification from the applicant (this is not an exhaustive or complete list but some examples). Requests for verification should be on case-by-case basis and relevant to your integrity check decision making.

COMMENT:  The list included “Copies of IR returns”  but it is doubtful whether most staff would read Appendix3 which contained a lot of duplicate information, or would have the business experience to know what documents to request.

If at any point during the conversation fraud was suspected, the staff member had to issue a caution and read the person their rights and say that they did not have to answer any further questions.

COMMENT: A  standard letter should have been sent requesting documents that had been specified by expert advisors. The caution could have been a way of stopping evidence being provided.

The MSD asks beneficiaries to complete forms and include specified documents without staff having a conversation and giving a caution. In the case of the wage subsidy review of 339 applications, a standard letter stating the documents required should have been sent. Having a conversation was  a very time consuming, subjective and unnecessary process that was unlikely to result in any usable evidence being provided. If the person spoken to had been involved in applying for the wage subsidy up to 18 months earlier, they might have difficulty recalling the details of what happened and could later claim a misunderstanding or that they made a mistake.

It was very easy for applicants to hold back processing of invoices for a couple of weeks to show the required decline in revenue but this would then be followed by a big increase in revenue in the following month or two. MSD would have known about this from the start or certainly by June 2020 so copies of GST returns should have been requested as suggested in the OAG report.

After the conversation, an email, based on a template letter, was to be sent requesting any documents the writer wanted to see. The recipient was told that they could ignore the email if they thought that providing documents could be incriminating and were told that if that happened MSD might then try getting the information elsewhere.

If three attempts to have a conversation using the contact details supplied failed, the staff member had to write requesting any documents that they thought that they might want to see but no instructions were given about what documents to request or must be supplied.

Under Appendix 3, the possible outcomes of discussions were listed including No Further Action. That would have meant that no documents were requested.

COMMENT: Requiring staff, who lacked the knowledge and experience to check business documents, to decide on what documents they wanted to see was a policy designed to fail.

The MSD had a time consuming conversation with 339 recipients and then an email was sent requesting any information or documents a staff member wanted copies of.  Staff were largely inexperienced in checking business documents but were given no guidance as to the most important documents to request and what to check when the documents were received.

The MSD advised the following in response to a request for copies of standard letters:

“The Ministry’s response to the Auditor General’s recommendation to test the reliability of the post-payment assurance work carried out does not include sending a standard letter to the 339 recipient sample, and as such your request is refused.”


MSD asked a sample of 339 recipients of payments in the first iteration of the Scheme for documentary evidence to confirm their entitlement to the Wage Subsidy. As at 27 October 2021, 186 of the 339 recipients had provided documentary evidence.

Of those 186, MSD determined that:

      • 3% (168 cases) required no further action;

      • 5% (14 cases) were required to provide a partial or full refund; and

      • 1% (2 cases) had been referred for further investigation based on the evidence or lack of evidence provided.

    Our assessment of the progress made

    Although these may appear to be small numbers, it is important this information is carefully considered to determine what further work needs to happen to protect the integrity of the whole Scheme, and the significant public funds involved.

    MSD does not yet know whether the findings from its follow-up work are indicative of wider non-compliance with the requirements of the Scheme. If they are, this could involve a substantial amount of public funds. This reinforces the importance of MSD knowing more about the extent of non-compliance across all iterations of the Scheme and taking action to recover funds paid to applicants who did not meet the criteria or their obligations.

    COMMENTS:  It is disturbing that 45% of recipients did not reply within 3 months, even though many may have been asked by staff for documentary evidence that was likely to be of little value. If the sample revealed a significant extent of non compliance by all recipients , the MSD was expected to take action to recover the substantial amount of public money involved but it has not done so.

    An OIA request was made for details about the outcome of the review but we were advised that: “The information you have requested is held in notes on individual application files. In order to provide you with this information Ministry staff would have to manually review hundreds of files. As such I refuse your request under section 18(f) of the Act.”  This indicates that any results from the poorly designed review process will not be of much value.

    The following response to an OIA request regarding the sample of 339 was received in May 2022.

    Documentary verification was requested from these recipients to support their entitlement, including that they met the required revenue decline. This documentary verification included:

    a. Verification that they are a real business

    b. Information evidencing the required revenue reduction, for example, financial or bank statements.

    c. A reconciliation of their employee numbers (if they are an employer) to confirm they employed all of the employees named in the application.

    d. Payroll information showing they continued to retain and pay their employees for the duration of the subsidy. 

    Staff found that 8% of recipients did not have the required revenue drop for a month and 6% did not retain some of the employees they claimed for.  

    Of the final results, in 89% of cases, entitlement for the Wage Subsidy received was confirmed. As noted in MSD’s response to the first part of your request, the sample relates only to those subject to a random post-payment integrity check and it is not appropriate to extrapolate findings to recipients of the schemes in general given the nature of the sample.

    Findings from this work however will feed into broader work as recommended by Audit NZ to analyse the results of all integrity and assurance work undertaken to date to inform a risk-based assessment of next steps, and what further integrity work needs to be completed, if any, to strengthen integrity of the scheme.

    No separate report or findings have been prepared in relation to any results from the assurance work described in this response.

    Staff did not check whether businesses had had the required drop in revenue for all of the 12 weeks of the original wage subsidy they received.

    COMMENT:  There is no record of how many of the 339  were sent a letter and what was requested. Recipients were not even asked verbally about some basic Declaration obligations. Five out of 339 did not respond at all. If it is not possible to extrapolate anything from the results, then there was no point in doing the sampling.  This result was not what the Auditor-General recommended.

    29% of 1,000 did not provide an easy response so, if 339 had all been sent a standard letter requesting them to verify important Declaration requirements, a much higher percentage would not have responded.

    OIA response 18 May 2022

    Furthermore, in our response dated 25 March 2022, we advised that 89 percent of responses received from this sample group confirmed entitlement and require no further action. We also advised that nine percent are required to make a partial or full refund, and that four recipients have been referred for investigation. These figures have not changed.

    The Finance and Expenditure Committee also expected the reliability of earlier checks to be tested as the following quotes from their report state:

    “Most of the post-payment reviews involved phone conversations with people who received the subsidy payment. The OAG identified that those conversations may not accurately reflect the recipient’s actual situation—for example, as recorded in business accounts. In our view, the reliability of those reviews should be tested against additional evidence.

    MSD reported to us that it has made the following improvements to its post payment assurance processes. (The OAG’s other recommendations are not relevant to post-payment assurance work, and are discussed further below.)

    OAG recommendation #3: test the reliability of a sample of the post-payment assurance work it carried out against documentary evidence held by applicants

    MSD is testing the reliability of its post-payment assurance work by comparing a sample of its post-payment assurance work against documentary evidence. To do this, MSD will conduct 3,751 integrity checks from a representative sample of 339 applicants.

    This work began in July 2021 and is expected to take up to three months.

    OAG recommendation #4, part A: seek written confirmation from applicants of compliance with the scheme

    MSD has written to a sample of 1,000 applicants of the Wage Subsidy Scheme asking them to confirm their compliance with the eligibility criteria and their post-payment obligations. The sample is weighted towards larger employers, and represents 10 percent of the scheme’s expenditure.

    This work also began in July 2021.

    The following OIA request was refused because the MSD does not have data on complaints. Those making a complaint are not notified of the outcome so complaints can be easily ignored.

    “Please advise how many complaints have been re-opened for investigation since May 2021 and whether all complaints will be properly investigated, including those resolved, based on the requirements of the Auditor-General.”

    Thank you! Your subscription has been confirmed. You'll hear from us soon.
    Subscribe to our newsletter.