Note: Where available, the PDF/Word icon below is provided to view the complete and fully formatted document
Joint Committee of Public Accounts and Audit - 17/07/2014 - Auditor-General's reports Nos 10 to 31 (2013-14)

BOX, Mr Darren, General Manager, Debt, Appeals and Health Compliance, Department of Human Services

BURTON, Ms Donna, Executive Director, Performance Audit Services Group, Australian National Audit Office

CHAPMAN, Mr Steve, Deputy Auditor-General, Australian National Audit Office

IOANNOU, Dr Anastasios (Tom), Group Executive Director, Performance Audit Services Group, Australian National Audit Office

JACKSON, Ms Kylie, Senior Director, Performance Audit Services Group, Australian National Audit Office

JONES, Mrs Alice, General Manager, Health Programmes, Department of Human Services

MCLEAN, Mr Colin, National Manager, Enterprise Architecture; and ICT Security Information Officer, Department of Human Services

MCCLUSKY, Ms Melissa, General Manager, Health and Government to Business Systems, Department of Human Services

MCPHEE, Mr Ian, Auditor-General, Australian National Audit Office

SANDISON, Mr Barry, Deputy Secretary, Health and Information, Department of Human Services

SCHIAVO, Ms Emilia, Director, Performance Audit Services Group, Australian National Audit Office

STERRENBERG, Mr Gary, Chief Information Officer, CIO Group, Department of Human Services

Committee met at 10:43 .

CHAIR ( Dr Southcott ): Welcome. I declare open this public hearing of the Joint Committee on Public Accounts and Audit for its review of Auditor-General’s reports Nos 26 and 27. The first part of the hearing is focused on audit report No. 26 2013-14, Performance audit: Medicare compliance audits. The second part of the hearing will focus on audit report No. 27 2013-14, Performance audit: integrity of Medicare customer data. The committee resolved to review these audit reports on 5 June 2014.

On audit report No. 26, Medicare compliance audits, I now call representatives from the Australian National Audit Office and the Department of Human Services. Although the committee does not require you to speak under oath, you should understand that these hearings are formal proceedings of the Commonwealth parliament and that the giving of false or misleading evidence is a serious matter and may be regarded as a contempt of parliament. I remind you that the hearing is public and is being recorded by Hansard. The hearing is also being broadcast live.

I acknowledge that the Hon. John Dawkins MLC, member of the South Australian parliament, is viewing the proceedings in the public section of the gallery.

I now invite you now to make a short introductory statement. The committee will then proceed to questions. Time is limited, so I will ask that you limit your opening comments to a few minutes only.

Mr McPhee : Chairman, I do have an opening statement. It summarises both audit reports. In the interests of time, I am happy to table it if that is the wish of the committee.

CHAIR: Thank you. We will accept that as tabled and we will authorise its publication. Mr Sandison?

Mr Sandison : Very briefly, again in the interests of time, I just want to note that the Department of Human Services acknowledged the report and accepted the recommendations. The key point we would make is that there has been significant activity during the course of the audit as it raised a number of issues that we could work on straightaway as we worked with the ANAO about the findings. Since the report has been tabled we have continued those activities. If there are questions in particular that you want to ask about activities that we may have undertaken, Mr Box or I can answer those.

CHAIR: Thank you very much. I might go back to the 2008 budget and the target that was committed to in that budget. It was planned that there would be $147.2 million in savings through this initiative. How was that figure calculated?

Mr Box : The savings were calculated based on best efforts at the time based on the history of what we were getting as far as debts and what we perceived to be behaviour change within the community from the audits and the compliance activity that was already underway, and then transposed to the increased compliance activity which the budget measure funded.

CHAIR: As I understand it, prior to that you would do 500 compliance audits in a year?

Mr Box : Correct.

CHAIR: And the proposal was for that to go up to 2½ thousand compliance audits in a year, so a fivefold increase. Was the 147.2 a multiplication by five? How did you come up with that?

Mr Box : Without getting into the costings, it was probably fairly simple: 'This is what we are finding now. If we have the new powers and we do more compliance activity, do more audits, we could expect to get an increase in results.' So it was effectively a multiplication. Whether it was a straight five-factor multiplication, I would have to look at the costing. But there was a factor increase based on more audits, increased powers, more results.

CHAIR: Is it true that when you expand the number of compliance audits you see, firstly, diminishing returns in terms of picking up activity which is outside the law, outside what it is allowed to be and, secondly, diminishing returns in terms of what you collect?

Mr Box : I think it is hard to say that the more you do the more you will see diminishing returns. It goes back to how well you are targeting your compliance activity et cetera. It depends upon what is happening in the community and the claiming patterns. The other thing to point out is that the savings were also based on the expectation that we would be able to measure impact and behaviour change. It is not always about a debt. It is not always about getting money back. It is about education and getting compliance activity in the community through education. Then, because we have done an audit activity, the community understands it is something they need to be aware of, and behaviour change results. Unfortunately, we are still working on that to be able to measure the behaviour change. The savings were not a direct debt-based saving, if that makes sense. Back to your question about the more you do the less return you get, I do not think that is necessarily the case.

Mr Sandison : Chair, I think one of the points that is worth noting is that it is not just depth of audits. You might do more of the same audits, but one of the issues is the nature of the audits and the breadth. So you can go deep, and that might be sometimes more limiting returns, but if you broaden out, with your understanding of what you have been doing previously, you actually explore other areas.

That is one of the points made in the recommendations from the ANAO about the risk framework and better targeting combined with what Mr Box was talking about in relation to the assessment of the results, and the results are behavioural as well as straight-out debts and what you might recover. So, with those combinations, that increase from 500 to 2½-thousand gives us scope to actually broaden how we approach the nature of audits and look at different targeting within the overall health system, not just do more of the same.

CHAIR: In terms of the debt that you raise and what you recover, what do you expect to recover of that debt?

Mr Box : If we raise a debt we would expect that we obtain the repayment of that debt.

CHAIR: Have you ever done that?

Mr Box : There are obviously providers out there who completely repay their debt, but if you look at the total debt raised versus recovery you will end up with less because for some people their financial circumstances and other reasons mean they are unable to settle their debt.

CHAIR: Okay. Through this initiative you raised $49.2 million in debt and you recovered $18.9 million. Is that consistent with previous compliance activities?

Mr Box : I think that is fairly consistent with what we have found in the health space, from the data the audit was looking at. I think the reasons why that is happening would be fairly multifactored. It can be for various reasons. Sometimes people go bankrupt; there are a whole range of reasons.

CHAIR: Sure. In terms of the value of the debt you raise and what you recover, it is about a third?

Mr Box : Yes.

CHAIR: Was that factored into your advice to government in 2008?

Mr Box : I would have to take it on notice as to how we calculated it, whether we discounted it by such a factor. I am not sure, to be honest.

Mr Sandison : We can check on that. The other thing as well is the time frame in which you do the collection. You identify the debt at a point in time. While we have a collection amount, sometimes, depending on the time frames for the actual management of the collection, that is not going to be an instant amount. If you identify $10 million of debt you are not going to get $10 million of repayment happen within necessarily the same financial year or even over a number of financial years, depending on how the repayments are organised. So as well as the issue of some organisations either becoming bankrupt or having lesser ability to pay, some of the arrangements are actually made over a number of years as well.

CHAIR: Are you saying that we might do better than $18.9 million out of $49.2 million?

Mr Sandison : Correct. With a lot of our debt collection as an organisation, be it on the health side or the welfare side, there is a tail, if you like, that goes through; it is not an instant 'here is the full payment in one go.'

CHAIR: All right.

Mr CONROY: Can I go to the other part of the expected savings to the budget, which is behavioural change. I have a couple of questions. First, are you seeing any behavioural change? Second, how do you measure that and hence how do you return that to the budget as a savings?

Mr Box : As in the audit that made the recommendation which we agreed with, we do need to finalise our work in being able to measure the impact from the compliance activity. Outside of straight debts, that is largely behavioural change, so seeing a change in claiming patterns. As to how we will measure that, we are still looking at other jurisdictions, both in Australia and internationally, and how they are measuring their behaviour change. At a really blunt level, the way you would do it is look at claiming patterns for a cohort of providers before the audit activity and then look at the claiming patterns after the audit activity to see a change, potentially.

It would probably be too blunt to say that you could attributed 100 per cent of change to your compliance activity, because there are a range of reasons why claiming patterns might change, but it is an indication of an impact from the compliance activity. That is reasonably blunt and we want to get more sophisticated than that. There is the area of, say, voluntary acknowledgements. An increase in voluntary acknowledgements from providers in the area where we have done a compliance activity might show an impact. So there are a range of things and we have not settled 100 per cent on exactly how we are going to measure going forward, but it is something we are working on and we hope to have resolved.

Mr Sandison : If you are interested in that area, you step into the education element of compliance as well. Sometimes if you look at compliance in the nature of audits, the education side of simply writing to people informing them about the ways in which they should comply with the legislation in undertaking their activities can make quite a difference.

What we are trying to do—again, the audit discussed the use of letters—is make sure that we are really clear about what is an audit in the formal sense versus what is education and information that we provide. Both of them have the same issue about behavioural impacts. We are getting better in the use of data, whether it is by geography or specialisation or claim, as to what happens if you do a 'before and after'. One of the issues that was measured here and commented on by the ANAO was the extent to which we made a baseline beforehand and understood what was happening and then were able to measure the impact and effect, be it an audit or an education process. Our measurement is not only something that is of interest to the ANAO but, needless to say, our colleagues in the Department of Finance have a strong interest about whether—this goes to the second part of your question—that is real money returned to budget and if it was not spent in the first place how can it be a 'save'. That has always been a fraught area in compliance, in health or welfare, about how that gets recognised in various initiatives that are put forward.

Mr CONROY: How much of the $147 million in expected gross savings is expected to be this area? You might need to take that on notice.

Mr Sandison : I think we would. It started in 2008 so it would have been before that time when the assumptions were being made. The world of behavioural science and behavioural economics has expanded somewhat over the last four or five years and there is a greater discussion and, dare I say, acceptance. At the time we were probably more talking about what would be the nature of compliance and audit activity. We can take on notice if there was some element around that in the paperwork.

Mr CONROY: Presumably, Finance is not a department that takes claimed savings lightly, so if any of that snuck in it would have looked at it reasonably closely.

Mr Sandison : There is a robust discussion about those sorts of claims, yes.

CHAIR: I would like to ask you about the savings. One of the things that came out of the performance audit was that the department was not monitoring the savings, there were no systems or processes enabling it to report against the savings target and there was no follow-up by Human Services to the ERC's request that the minister report back to the government three years later. Why did none of those things happen?

Mr Sandison : We have acknowledged those findings in our response to the audit and some of our explanation is in our response. I think the primary issue was we moved from being an independent organisation, Medicare, and went into Human Services as a larger organisation with quite a different governance regime that has improved, I would say, out of sight in terms of where we were, and there were some issues where we accept we omitted to do the responses to some of the elements that were required of us. That is exactly what the ANAO found.

CHAIR: Mr McPhee, do you want to comment on that?

Mr McPhee : I think it is fairly clear. Government expects agencies to monitor these new measures and their implementation and, of course, ministers in particular would want to know if things were not going according to plan. It is a governance issue and, as Mr Sandison has mentioned, I think it is true that agencies' governance processes have improved considerably from the period we are talking about until this day. So let's hope that it is a point in time issue and that we would not see it again. But it is a fundamental issue for all agencies and an expectation that they will monitor, particularly in the early years, new policy measures to see whether they are achieving the results that government expects and within the parameters that government expects.

CHAIR: Why did the department not inform their minister of the proposal to expand the types of compliance activities that it would conduct under that budget measure?

Mr Box : You are talking about the inclusion of letters being counted in that target. We have accepted that was something that we should have done. What we would say is that the inclusion of those letters is a valid compliance activity, and I think the ANAO accepts that, and we had had conversations with the policy department, being the Department of Health, through a formal committee process. Notwithstanding that, we do accept that going back to ministers, because of the relationship to the budget measure, was a requirement we should have done and one we have agreed to in the audit. So there were some mitigations in that we thought that we were consulting with the Department of Health as the policy owner, but, notwithstanding that, we still had an obligation to talk to the minister.

CHAIR: As I understand it, when the initiative was announced, the 2,500 would not have been including letters.

Mr Box : Correct.

CHAIR: That is correct?

Mr Box : The initiative in the budget measure talks about audits.

CHAIR: But during the period you changed the performance measure or the reporting measure?

Mr Box : Because the budget measure talked about audits, we should have been counting—the funding was provided to conduct audits. Whilst letters are a valid compliance activity, that should be on top of the audits.


Mr McPhee : Can't argue with that.

Mr LAUNDY: Mr Box, going back to something you said earlier, about education in the community, my understanding is that this is an audit process conducted on providers.

Mr Box : Correct.

Mr LAUNDY: So education of them is what you really mean.

Mr Box : Yes, I was talking about the community of providers.

Mr LAUNDY: Okay. With that as a backdrop, they provide 340-odd million services a year. Am I right to say that, prior to moving it to 2½ thousand a year through this initiative, we checked 500 providers?

Mr Box : We were doing 500 audits before the budget initiative.

Mr Sandison : I think we have to be clear that, depending on what the scale of the audit is, an audit might cover a practice with 20 doctors or it might cover a health network or a hospital. So it is not the case that 500 doctors with 500 claims were the source of the audit process. It really depends, and that is why—

Mr LAUNDY: Because that does not appear on numbers to be robust.

Mr Sandison : No, and in part that is why there was an expansion.

Mr LAUNDY: How many health providers are there in the country? I found the figure of 344 million services.

Mr Box : I would have to take that on notice.

Mr Sandison : It is tens of thousands, depending again on—

Mr LAUNDY: I would have suspected that. It did not seem like a robust sample, even at 2½ thousand, although you might argue that the more we do the more we lose. But it just did not seem very robust. Do you have an opinion on that? Is it robust?

Mr Box : The 500 or the 2,000?

Mr LAUNDY: Either. Was 500 too light? Is 2½ thousand enough? Should it be more?

Mr Box : With the position we are in at the moment with the funding to allow us to do the 2½ thousand audits, as Mr Sandison said, it is not 2½ thousand providers; it is more than that. Then there are the education and the letters on top of that, notwithstanding the comments from the audit. I would say that we have a fair amount of money as an organisation to spend on compliance activity. If we implement the recommendations, as we are, from the audit, I think we will improve our compliance approach even more. Again, there are with the controls within the Medicare system. Funds are always limited, but I would say that we have the ability to provide a robust compliance approach for the funding that we are provided, notwithstanding that we do have to make sure that we are as efficient as we can be. Education achieves a lot, and the first recommendation from the audit talks about targeting and analysing our risks. If we get that right, with the education as well, I think that we can deliver a robust compliance approach.

Mr LAUNDY: Thank you. Mr McPhee, given that we have not gone close to what we were supposed to do by ramping up the audit and finding the savings that we were meant to under the change in the numbers, in your appraisal of this as Auditor-General did you at any stage look at the cost-benefit analysis of the increased resources on top of the missing of the funds—how far we missed the target by?

Mr McPhee : I would need to check with my colleagues, but I suspect we would need to be checking the outcome of the audits themselves to see what the returning was like and see whether there was a scope for revising the audit program based on the findings and the risk assessments coming through. So that is quite a detailed exercise. I look to my colleagues to see if we can add anything further to that.

Dr Ioannou : I will provide part of an answer. The department has just observed that resources are always constrained, so we have to be respectful of the parameters established by government. What we were looking to add in the context of the audit was, within that framework, to ask whether the department did what it was asked to do. We found that only once, as I recall, did it manage to reach the target. In terms of cost benefit, in resource constrained environments I think the real trick in all of this is to target your resources as well as you can; hence, the focus on a targeted risk based approach. And, of course, we have a recommendation along those lines. Did we go and review each of the assessments? I do not think we did. We did not look at the substantive audits. That would have been quite a different exercise for us.

Mr McPhee : I think the other thing that is interesting is that the tax office uses a similar approach with their focus on compliance. There is education, correspondence, reviews—which are a light approach to an audit—and then full, in-depth audits. So you have this graduated approach, which is what the department's approach is. I think probably the tax office has particularly refined its approach across the years. I appreciate the department saying, 'looking internationally in the sphere here in terms of Medicare sort of arrangements', but the tax office is also a very useful source of advice in this area, including the whole approach to risk assessments, where they get the returns from their audits. In fact, we have done some audits in the tax office space as well and highlighted a few issues for them to pay greater attention to. But it is becoming more sophisticated and all directed to putting your resources in the best areas. It is not just a case of audits or this or that; it is this graduated approach. When you have clear indicators of issues based on your risk assessments, or in the tax office's case that is strong data analysis to show transactions or situations which are pointing to red flags. That is where you put your audit focus. It is quite a graduated approach. The good thing is agencies are on this wavelength, improving the sophistication of their methods and targeting to get more cost-effective compliance approaches, which goes to your question.

Mr LAUNDY: I notice that recommendation. Following on from that, with computerised systems that produce red flags, working smarter into the future and conducting these audits, do you feel that the system that we operate Medicare under gives you the flexibility to best execute your task? I know through personal experience with the ATO, pre-politics, that it is that unusual behaviour, something that will throw a red flag, that comes with a light approach. I agree with the recommendations made in this report. Do you have the IT support in the system to allow you to best execute your job?

Mr Sandison : I think a debate on IT systems in our organisation could fill in more time than you have this morning. We have an IT system that enables us to do a reasonable amount of the targeting. There are always new ways to look at IT and how you build datasets and so on to put new flags into place because you did not realise 10 years ago you might need them. That is always going to be the case. Through the data that we have just now, we can target a range of different things that give us the benefit of looking at particular risk based approaches. One of the most common examples we have is writing to doctors who are approaching what is called an 80-20 rule, where if they see up to 80 patients in a day 20 times in the course of the year, that is seen as a barrier not to go past. So we would write to them. So the data will show when that is getting closer and we would write out or engage with those doctors. It is just a small example.

We can also see the before and after claiming patterns of particular claims because it is in the datasets about how many times a doctor might be charging against a particular claim and whether that is anomalous compared to other doctors of a particular specialisation or something like that. If a range of specialists of category X claim 10 times a month and one is claiming 40 times a month then in the risk based approach that is where we would target. We would not go and look at all 100. That is something with the recommendations from the ANAO. And to be fair to some of our predecessors, people were working towards how to use the data within the systems. It is always open for 'Yes, there can be improvements.'

Mr TAYLOR: I was interested in the original object of this program and the idea that we should spend $77 million to get $147 million back. That seems to me to be a fairly high-risk proposition and it clearly turned out to be so. When I look at the ATO or any tax office type metrics, I do not think they would ever consider that to be a good payback. Have you benchmarked this program and the objectives of the program against what tax offices do particularly because I think that is the most comparable benchmark you could use?

Mr Sandison : It is comparable in some regard, but I do not think any of the various compliance related systems—welfare, tax, health—are directly comparable because you are looking sometimes for different things. While money is one measure and obviously the ATO might look at a two-for-one return and say, 'That's not worth it,' if they were going after to big business or the general taxpayer. Part of the compliance regimes that we run both on the welfare side and on the health side are about deterrence.

Mr TAYLOR: That is true for the tax office as well.

Mr Sandison : True, absolutely. If we are looking at providers, even though there are tens of thousands of them, as was mentioned, we are dealing with a smaller target group and you have to consider whether there is an assurance to the taxpayer about the way in which we are out there looking at the health system. If you look at GPs, specialists and diagnostic imaging, you have pathology and the range of different areas. So the scale of the work—and I think it was part of the discussion back in 2008—would simply be, yes, there might be a return in terms of what is targeted with a compliance activity, but also back to the breadth and depth of what we do to make sure people are aware we are out there. When it was 500, it was probably small-scale and the chances of touching certain areas of the health system with 500 audits would have been low. So money is one part, but it is a broader issue.

CHAIR: How does a 13 per cent return compare with your historical benchmarks? Leaving aside ATO, with Medicare and the Department of Human Services, how does a 13 per cent return compare with what you would normally do?

Mr Box : We would have to take that on notice.

Mr Sandison : We can provide that to the committee.

CHAIR: You cannot give us anything? If you set out to recover a certain amount of money, surely you have an idea of what you would expect to receive.

Mr Box : I think the point to make is that—and the second recommendation in the audit talks about measuring our impact—the return based on the numbers, as correctly quoted in the audit, would probably be consistent with what has been there historically. That is my instinct, but we would have to take it on notice.

CHAIR: So 13 per cent is about half?

Mr Box : I would have to check. The problem is that the bit that we are missing is that we have not yet measured the behaviour change in a way which was anticipated in this budget measure. That is work which needs to be finalised.

CHAIR: How does a 13 per cent return compare with other compliance activities?

Mr McPhee : I do not have the answer to that one, but for me the issue is how strong and how good was the business case in the first place to take to government, suggesting that for the investment of 76.9 the return would be 147. How strong was that case?

It seems it was a bit thin on all accounts but perhaps in the department's defence, finance runs its eyes over these business cases before they go to government and generally would raise a red flag if it saw a particular issue. So the position was accepted at the time. We did not go into the business case. It was not the focus of our audit but for me, if I wanted to learn from this particular matter, it would be around how come we have such a big variation going here and see what we can learn from that experience.

Mr Sandison : If we were preparing a business case now along similar lines, there would be a different discussion on a range of areas about how indeed we would have to argue the case. It would be expected by the Department of Finance, watched carefully by ANAO, but it would be a debate about how we target risk approaches, what kind of return, is there a set number, but the expectations around behavioural change as well. Normally we would still see in there an element of to what degree government want us out there anyway regardless. You could say, 'Don't do any compliance.' If you went purely for financial return, there is a return about being active and engaged which you sometimes never see in relation to compliance activity.

Mr TAYLOR: That is true for any revenue collection organisation, including commercial organisations. I would just say that that ratio of 77 to 147, having looked at many of these things over the years, is not a ratio I would have found particularly interesting. I am amazed that it got through the screens.

CHAIR: Are there any other questions on this?

Dr HENDY: To cover some ground you have already talked about, when you were initially answering questions from the chair about the costings of the $147.2 million saving, he asked you whether the actual result you got was in accord with what might have been in practice and you said it was fairly consistent with practice— that was your answer. If it is fairly consistent with practice that that was the outcome, then your costing was completely false—is it not?

Mr Box : The way I answered the question was to say that the result we are getting as far as pure debt results, raising of debts, the costing I think would have been based on what we were getting historically. The costing then, is my understanding, included an element of what we would likely see as far as savings are concerned from behaviour change because of compliance activity which is not debt related. That has not been measured historically and we still need to get to that point, which is the second recommendation of the audit which we have agreed to. So the pure debt raising percentage of it, while we would always try to target our resources in the way to get a best impact, is only one element of what we had anticipated to get from that costing as it was developed, putting aside the validity of the business case as we have just previously finished talking about. I hope that makes sense.

Dr HENDY: It makes sense but it does not answer my question because you have produced an estimate, which is 147 million. The behavioural effect is a relevant point but it does not go to your actual costing. If you are saying that the behavioural effect would be part of your costing and you have not even got to the stage now a number of years later of checking that, your costings were false. That is my point.

Mr Sandison : I would acknowledge that there are perhaps some inaccuracies in the assumptions that were made when we were preparing it

Dr HENDY: Okay. How do you prepare these costings? I am interested in who in the department or the agency does the costing initially and then how is it checked up the ranks to become a decision of the department that that is the costing?

Mr Sandison : Whenever we are doing a costing for any of our activities, it is normally driven by the line area which has responsibility. It would be prepared normally by some of the experts in the area. They will link to service delivery people, to our chief finance officer area and to our IT area because the discussion is: to what extent can you get the data or do the IT systems need to be changed to achieve it?

There is a range of different things, including: do we have front-line staff or call centre staff involved? We have to get across a range of areas.

In relation to compliance, we would make some judgements in relation to past history but would also start to look ahead to what we are trying to change, because often with a new initiative there will be elements that would be brought in. In any of those costings, there is a range of assumptions that have to be made and they will be debated internally. We would then work with the policy departments. There would be a joint approach with the policy department and sometimes they are just as vigorous as the Department of Finance. If we get to an agreed figure, we would then get sign-off. The current practice is right up through to the secretary, depending on the scale of the proposal, and then it goes to Finance. In Finance, they will go through it again and look at it to try and judge whether it is an appropriate cost framework.

Dr HENDY: So the secretary is the person responsible for the costing?

Mr Sandison : Under the current processes. That was not necessarily the process—

Dr HENDY: No-one more junior is responsible for the costing? It is not a division head or a deputy secretary?

Mr Sandison : As I said, the responsibility goes through, but we have various thresholds. If it is a significant costing with a major cost to government, then it will be signed off by the relevant deputy secretary, the chief finance officer and then it will go to the secretary.

Dr HENDY: I had the same sort of questioning with other agencies. In this particular example there has been a failure in the outcome. I think you have acknowledged that. There is a process in which a costing is produced. It goes up the ranks. I understand—I worked in the Public Service in different parts of my career. It is touched by a lot of different people—any calculation—and there is diffuse responsibility. Who can you pin the mistake on? That is what I think is important to try and do. More to the point is: through that range of maybe six levels of different officers who have been involved in the calculation and given that this has been a failure, in my view, who has been counselled? Whose performance review on an annual basis has been touched by the fact that this has been a bad result? What processes have you had in place in terms of dealing with this area of your departmental or organisational structure to feed back that there is a failure in performance? Have people's performance appraisals been affected by this poor result?

Mr Sandison : I cannot give you a specific answer. Mr Box has been looking after health compliance for 12 or 13 months and I came in about seven months ago. In terms of taking it forward, that is the role that we have. All I can make is an observation that there has been significant change in some of the areas that had responsibility for delivery of the measure. Some people are no longer in the Public Service and some are in different positions—they have been moved around. The extent to which there were discussions, I am not aware. It would not have been part of my role or responsibility at the time. As it stands, in terms of how we respond to it, we have our targets of responding to the ANAO and how we conduct the audits, and that is a very clear direction. It sits in the performance agreements of individuals, where some of the things that I think were deficiencies in terms of things that should have been picked up that ANAO made comment on should have been very clear—'Your role and responsibility is to deliver on this and report on this at a regular stage.' That is now built into how we manage the workflows.

Dr HENDY: The performance assessments of individual officers—it is built into—

Mr Sandison : The delivery of this measure is a key part of both my broad role and also a more specific one with Mr Box.

Dr HENDY: Acknowledging that you are both reasonably recent to your positions, I do not think I can get many more answers. But it is a relevant point.

CHAIR: I would like to ask Mr McPhee to respond to that.

Mr McPhee : Ideally, the governance arrangements of an agency would monitor these new measures. If it was a 2008-09 budget measure, certainly by 2009-10 you would be starting to ask, 'How's it going? What's it looking like?' and by 2010-11 you would have a feel for things. Most agencies would say, 'The actual outcome is not lining up with our expectations. What can we do?' and manage it that way. So you manage it before it becomes a serious problem and if your estimates happen to be way out and are most unlikely to be able to be recovered, you would let the minister know fairly quickly and perhaps reset expectations. I think we said in the report that ERC had asked for a report 12 months after. So even the ERC were saying ,'Keep us posted on how the measurement was going.' That did not happen for the reasons mentioned earlier. For me at the highest level it is a failure of governance arrangements over the measure at the time and it goes back quite some years.

Mr LAUNDY: Mr McPhee stole a little bit of my thunder because I was going to say they were 100 per cent out within 12 months of what they expected, to what it cost. I direct your attention to page 75. Following on from Mr Taylor, am I right in understanding that the 77 million in table 4.1 is additional resources put in place or is that the whole departments funding?

Mr McPhee : That is the new money.

CHAIR: The next paragraph breaks it down.

Mr LAUNDY: Am I right, Mr McPhee, to conclude that we spent 77 million to try to get 147? We got 49.2 in theory but only got 18.9 million of that?

Mr McPhee : That is correct.

Mr LAUNDY: So it is not a 13 per cent return. We spent 77 million. By doing this, we have lost 50 million bucks.

Mr McPhee : The only proviso I need to say to pick up Mr Box's point is that it may have had any effect on behaviour, the fact that more audits were down. Certainly in the terms of direct savings that were expected by government, your analysis is correct.

Mr LAUNDY: Is spot on.

CHAIR: To clarify, the $147.2 million is what you were expecting to receive.

Mr Sandison : The expectation.

CHAIR: That was not counting in any behavioural change at all.

Mr Sandison : Yes.

Mr Box : I need to be just a bit careful. I need to check the costing. My belief here—and I am happy to stand corrected on such costings— is that that costing, as far as expectations of returns are concerned, had an element of behaviour change and reduction of the outlays as well as the direct recovery of debts.

CHAIR: Can you provide for the committee the breakdown of that $147.2 million, what you were expecting to receive in the hand and what you were expecting was going to be some behavioural change on the NDS.

Mr LAUNDY: The 147.2 is a budget by-line.

CHAIR: As I understand it, it could be two things. It could be that you received the money or it is reduced expenditure on the NDS.

Dr HENDY: The Audit Office's answer is that they did not raise the—

CHAIR: I follow that. I am going back to the 147, the original costings for the 2008 budget.

Mr Sandison : And what were the assumptions in the 147?

CHAIR: If you could provide to the committee on notice the assumptions on how much of that money was actually expected to be recovered and how much was due to behavioural change.

Mr CONROY: I want to follow up on Dr Hendy's line of questioning about accountability, whether the Public Service learns from this and holds people accountable. You commented that ultimately for a program of this size the secretary to the department is responsible for the costing. Who was the secretary to the department at that time?

Mr Sandison : At that time it would have been the chief executive of Medicare. I would have to check who that was back in the preparation. If it commenced in 2008, it would probably have started work in 2007 to go through the budget process in time to commence in 2008.

Senator CONROY: With all this talk about recovery of the money and behavioural change, one thing seems not to have been really discussed in depth. All the costings were based on doing 2½ thousand audits a year. You only reach 2½ thousand audits once. In what appears to be reasonably generous supplemental funding to this area, why were you unable to hit the 2½ thousand target in those other three years?

Mr Box : We talked about the letters. The letters are a valid compliance approach but we had been counting letters in that target. We accept that they should be on top of the conduct of the audits. This year we will exceed the 2½ thousand audits as well as having the letters on top of that, so we will achieve it this year. The reason we did in only one of those other years is an issue of making sure the resources are focused in the correct areas. We have now, and we will hope to—

Mr Sandison : To be fair, I think you could read into it that in the two recommendations from the ANAO we did not do those things; therefore, that limited our ability in governance to make sure we were achieving the 2½ thousand audits, with any letters being in addition.

Senator CONROY: The ANAO detected a nine per cent data inaccuracy rate in the subset of compliance audits. This meant that a number of compliance claims were incorrectly recorded as non-compliant. What is the story there?

Mr Box : That is not to say we got the wrong result, that we found a provider non-compliant when they had been compliant. The audit result was that we got the right result. If you ran the report, the right result was given to a provider. If they were non-compliant they were told they were non-compliant. If they were compliant they were told they were compliant. They were not disadvantaged or harmed in that sense. But when they then ticked the box, in the data, they picked the wrong reason. That is a quality issue around our data. We have now educated and provided guidance to people so that the data is accurate.

Senator CONROY: So one in 10 data-recording incidents was incorrectly—

Mr Box : We hope we have rectified that now.

Senator CONROY: That is a fairly significant incidence of ticking the wrong box. That is not just one that slipped; that is a fair number, is it not?

Mr Box : They were picking the wrong reason codes, in the outcome of the audit. I can only say that we have accepted the finding of the audit. We have provided training to people and we think we have resolved that issue. We will keep monitoring it, going forward.

CHAIR: Is a nine per cent inaccuracy rate an acceptable level, looking at other areas of program administration, and what are the implications of a nine per cent inaccuracy rate, extrapolated to the program as a whole?

Mr McPhee : Generally you would say nine per cent is getting too high. We auditors say less than five per cent is immaterial, but once you get above five per cent you are starting to get into areas of significance. As a guide, we would say it is pushing a material inaccuracy in the information. It means there is either a training issue or form-design issue, fundamentally.

CHAIR: The committee has taken a lot of interest in implementation of audit recommendations. When we get to the next report on the integrity of customer data what we find is a concern that going back to 2004 and 2005 there were a number of recommendations, which your department accepted and totally agreed with, and they were not implemented. You accept the report and you agree with the recommendations. What steps are you taking to implement those recommendations?

Mr Sandison : In our governance arrangements in the department since we were created three years ago, the audit side within the department, firstly, works very closely with the ANAO and, secondly, runs a standard report around how many audits there are, internal and external, what recommendations there are and how many of those are outstanding. The secretary gets those regular reports. Each of the relevant deputies of the executive is accountable for making sure that we respond to the findings. We are held to account to respond to them so that they do not get left sitting. At particular times, if there is a time frame that we have identified when we said we would respond to certain things, that is what our internal audit people would also hold us to account on—'It's now the end of December. You said you'd review the policy settings as a result of an ANAO report. Have you concluded them yet?' There might sometimes be valid reasons for delay such as complexity, but you have to report. Rather than just leave it to the line area and, dare I say, hope that the right things are done, there is now full accountability for all audits as a central area.

Dr Ioannou : To provide a little additional comfort to the committee, we tabled Audit report No. 34 as part of the last program and we looked at DHS's implementation of ANAO performance audit recommendations. This was tabled on 14 May. We observed in that report:

DHS has well developed monitoring and reporting processes which provide oversight of the implementation of audit recommendations …

So I think some of what is being said by the department was borne out by that audit.

CHAIR: Firstly, I require a resolution that we accept an opening statement, and I ask: is it the wish of the committee that the opening statement tabled by the Auditor-General be accepted as evidence for the review of Auditor-General's reports nos 26 and 27 and be authorised for publication? Moved by Mr Laundy and seconded by Mr Conroy. It is carried. I thank the witnesses for making time to appear before the committee today and for their evidence. The secretariat will be in contact if the committee has any further questions. I now call representatives from the same agencies, the Australian National Audit Office and the Department of Human Services.