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
16/02/2018
Auditor-General's report No. 19 (2017-18)

BOURNE, Mr Andrew, Assistant Secretary, Commercial and Government Services, Department of Finance

BULL, Ms Helen, Group Manager, Workforce Information Group, Australian Public Service Commission

EDGE, Mr John, Deputy Secretary, Commercial and Government Services, Department of Finance

HEHIR, Mr Grant, Auditor-General, Australian National Audit Office

HELGEBY, Dr Stein, Deputy Secretary, Governance and APS Transformation, Department of Finance

HUNT, Mr Nicholas, First Assistant Secretary, Commercial and Government Services, Department of Finance

MELLOR, Ms Rona, Deputy Auditor-General, Australian National Audit Office

RAUTER, Ms Lisa, Group Executive Director, Performance Audit Services Group, Australian National Audit Office

VINE-CAMP, Ms Kerryn, First Assistant Commissioner, Australian Public Service Commission

Committee met at 09:50

CHAIR ( Senator Smith ): I declare open this public hearing of the Joint Committee of Public Accounts and Audit for its inquiry into Australian government contract reporting, based on Audit report No. 19 of 2017-18. The committee resolved to conduct the inquiry, based on the report, on 29 November 2017.

I now call representatives from the Australian National Audit Office, the Department of Finance and the Australian Public Service Commission. Welcome to the hearing this morning. We're grateful for the evidence that you will provide to us. Although the committee does not require you to give evidence under oath, you should understand that this hearing is a formal proceeding of the Commonwealth parliament. Giving false or misleading evidence is a serious matter and may be regarded as a contempt of parliament. I also remind all witnesses that in giving evidence to the committee they are protected by parliamentary privilege. This hearing is public and is being Hansard recorded. It is also being broadcast live.

I now invite you to make a brief opening statement, and the committee will then proceed to questions. Auditor-General, would you like to make an opening statement?

Mr Hehir : No thanks.

CHAIR: Would the Australian Public Service Commission or the Department of Finance like to? No? Great, then we'll begin with questions. Who would like to start? Mr Hill?

Mr HILL: Sure. Where do we start? Maybe we'll have a bit of a chat about the issues arising in relation to the 'consultant' definition and some of the Auditor-General's findings. Also, I think, Finance has provided some updated information. Auditor-General, what should be done about the inconsistent use of the consultancy tag?

Mr Hehir : I'd like to start by saying that the report we tabled didn't have findings. It was an information report, and in that information report what we were basically trying to do was use some of the AusTender data to provide information to the parliament on a number of issues that we'd seen come up through this committee and other committees through time. One thing I would say is that some of the information we put in the paper would lead you to a view that the data in AusTender isn't that reliable, and that's consistent with a finding and a recommendation we made in Audit report No. 48 of 2014-15, Limited tender procurement. We made that observation there and we've made it in other reports that we've done over time. I think the issue that may come up in looking at the data that we put out is that it's not precisely clear that the way people are reporting in AusTender is accurate in accordance with categorisations that are put out in the framework. That reduces the usefulness of those categorisations, if it's not accurate.

Mr HILL: I suppose the concern—and, sorry, I should have started with a bit of the context for my thinking—is that there's quite a narrow definition of what is a consultancy, which the Department of Finance's submission sets out and which you touched on in your report. There are some prima facie questions raised in the report about the extent to which that definition is being applied consistently, and some strong indicators, when you have a look at the categories of service spend—so management, accounting, professional services; I think you list the top six—that consultancies may be underreported.

Sitting behind this is a transparency question. There's enormous public scrutiny and interest from parliamentarians and others on the extent to which work is being outsourced to external providers, particularly the big four but not only—EY, Deloitte, PwC, KPMG and so on—and the only transparency we can see is in the annual reports, where, if something's classed as a consultancy, you can read about it in the annual report. But, otherwise, it seems opaque as to what government is actually spending, what departments and agencies are actually spending, on that kind of procurement. That's the context for our interest in this area. What work is the Department of Finance undertaking to address these limitations and the concerns regarding the use of the consultancy tag?

Mr Edge : I will start and my colleagues might add to the answer. Just to clarify a couple of points that you made, we are actually talking about two different but related processes here. We're talking about the AusTender process, which is a system that has a couple of functions but it's a system for publishing tender notices—so the opportunity to put in a tender for work for the Commonwealth government. It's also a system for reporting contract notices. The contract notices are reported and they effectively reflect the value of the contract that has been entered into. A contract notice doesn't reflect the actual expenditure under that contract, whether it's flagged as a consultancy contract or any other contract. The annual reporting process that you mentioned does reflect actual expenditure by agencies on consultancy contracts—in this context—for the financial year. So there is transparency, absolutely, in terms of what agencies have spent on consultancies through the annual reporting process.

Mr HILL: I think we've just said the same thing—so thank you. My question is: if an agency doesn't tick the box of consultancy, whether wrongly or because they don't understand the definition or it's vague, then there's no transparency in the annual report. Is that right? If an agency is spending on consultancy from—pick a firm—say, KPMG, and there's quite a narrow definition for consultancy, you can see that in the annual report but you cannot see any of the rest of the spend.

Mr Edge : I think the process that each agency would go through to generate the consultant information in its annual report may vary a little bit but, basically, they would take the information from their financial systems, they would look at the information about the contracts that have been entered into and those that have been flagged as consultancies and I would imagine that there would be a review of that, that there would be some quality controls around that.

Mr HILL: You would imagine, but what are the assurance processes? You report on the data, you own AusTender. What assurance is there that the data on AusTender is accurate with regard to the current definition of consultancy?

Mr Edge : If we're looking at how the system works, each agency is responsible for the information that it puts on AusTender. I think part of your question went to guidance and assistance, and we can certainly talk to that in a moment in terms of what we're doing.

Mr HILL: What would good practice look like, from your point of view, for an assurance that the data published on consultancies is correct?

Mr Edge : It is important to remember that it is a devolved system. Each agency, each entity, that uses AusTender is responsible for the information that goes on there and the quality assurance that goes around the publishing of that information, including how it's categorised. That is something that is the responsibility under the devolved financial management system that we're talking about. It's the responsibility of each agency to make those inquiries and make those—

Mr HILL: We understand that. That's clear in your submission. My question is: what does good practice look like, given you're the Department of Finance? Even if we accept you're just a post box from the AusTender point of view and people just pop stuff up there, good or bad, what does good practice look like for an agency to undertake assurance that the stuff they're putting up there is actually robust and properly classified?

Mr Edge : Whatever steps need to be taken to ensure that the reporting is accurate, that things have been reported, categorised correctly and that there can be a clear assurance that the information being published is accurate. That's a quality control process that needs to be built around these systems in each of these reporting entities.

Mr HILL: Do you think anything needs to be done about the apparent inconsistent use of consultancy tags that the Auditor-General's pointed to, whilst not making a finding?

Mr Edge : Certainly in terms of the systems that exist within the individual agencies and entities that use AusTender, there obviously needs to be a clear focus on accuracy of reporting. In terms of what the Department of Finance can do to assist in this process, we do and we're able to assist with the provision of guidance on how these steps are best undertaken and on definitions and categorisation.

Mr HILL: You issue the guidance?

Mr Edge : Yes.

Mr HILL: Are you the custodian of that rule?

Mr Edge : We publish guidance around the use of AusTender and around categorisations and around definitions of certain things.

Mr HILL: Sorry; just to be clear: is it the Department of Finance that decides what the guidelines are and the actual rule is in relation to what is a consultancy and what isn't?

Mr Edge : We have a definition of consultancy that we have published. That definition of consultancy, as I understand it in the Commonwealth context, hasn't changed for a number of years, possibly a decade or two.

Mr HILL: Indeed. I guess what strikes me when looking at the data and the enormous gap between what is classed as a consultancy versus what is now outsourced in the large categories of work is whether that definition has any current utility or future utility, because it seems to be a very arbitrary, narrow definition of the kind of work now undertaken by very large private firms.

Mr Edge : If it's helpful, we can also talk a little bit more about the nature of the guidance that we publish and the work that we are currently undertaking in terms of reviewing that guidance. We have a process of constantly seeking feedback from entities that use our guidance and we are very focused on making that as effective and user-friendly as possible. I might ask one of my colleagues to talk very briefly about that process.

Mr Hunt : As Mr Edge said, we do engage through a range of fora with both industry and with agencies about their use of the system and the policy framework more broadly. We have an industry engagement forum. The secretary convenes a procurement consultative roundtable. I sit in the role of Australian government procurement coordinator. That is about providing advice—

Mr HILL: I'm sorry, but I actually have no idea what you are now talking about. I asked a pretty specific question about whether the definition of consultancy is fit for purpose given the growth in other expenditure, because you said it hadn't changed in 20 years. Is that under current review or not?

Mr Hunt : We are looking at our current guidance material, which is on the internet at the moment. There are bits of guidance in various different places and we are looking at pulling that together at the moment, which is a piece of work we refer to in our submission, which will put all of the reporting requirements in one place and make them clearer. But, at the moment, the definition, I think, is reasonably clear. That definition, as Mr Edge said, does date back at least a decade.

Mr HILL: Okay; we're repeating ourselves now. Let's take the large four—big global firms and high daily rates. Let's say that there is public concern about the growth in expenditure—and we'll come to value for money later. There doesn't seem to be anywhere in the system that anyone can easily understand or view what is spent year to year on those four large firms? I will just take them; I know there's others.

Mr Hunt : I think Mr Edge made it clear that the system isn't about recording spending; it's about recording the outcome of procurement processes and the issuing of contracts.

Mr HILL: Yes, that's my point.

Mr Hunt : There are multiple ways that the data can be accessed. If you are interested in contracts with the big four firms, you can search the data on those firms. Some of those contracts will be consultancy contracts and some will not be consultancy contracts. So you can search for consultancies or contracts flagged as consultancies. You can certainly analyse the data on the basis of all contracts that the Commonwealth has entered into.

Mr HILL: How do we know how much has been spent each financial year on each of those firms? How does the parliament know?

Mr Edge : As I mentioned earlier, each reporting entity's annual report has information about money spent on consultancies, so there is that transparency.

Mr HILL: I didn't ask about consultancies. I really think we're going around in circles. Either you're playing a game or you can't listen.

CHAIR: I might have that.

Mr HILL: Yes, can you have a go? This is ridiculous; it's like an episode of Utopia.

CHAIR: I will take a few moments and then we'll share the questioning around. I'd like to go back a few steps to make sure that as we proceed through this inquiry we are talking about like for like. At least on the public record and in front of this committee, people are very clear about the appropriate utility of certain devices that operate across the Commonwealth to give people higher levels of transparency.

Officials from the Department of Finance, could you please explain to us exactly the purpose and role of AusTender? I do think, in some of the public reporting, there has been a misconception about what AusTender is there to do and whether or not, when you're looking for certain information, you should be looking at AusTender or whether or not you should be looking at the annual report. Mr Edge, could you start with an overview, in terms of the accurate role of AusTender, how its role as a matchmaking service for suppliers and government sources works, and reflect on some of the reporting—you may not have seen it—just to give us a very clear base at the beginning of this inquiry on what it is we are talking about when we say AusTender?

I will give myself about 10 to15 minutes and then we'll go to other members. After that, we'll go to some of the transparency issues, which are an important theme, this inquiry would like to get to at the conclusion of its deliberations.

Mr Edge : AusTender, as I said earlier, performs two functions. It is a way in which the Commonwealth publishes contract notices, contracts that have been entered into, who that contract is with and what the value of that contract is. It doesn't mean that it reflects the actual expenditure under this contract, but that is a system that's been in place for about 10 years. The AusTender system was introduced in about 2007 and has been refined over the past 10 years. AusTender is also where government tender opportunities are published. Approaches to market, for companies to respond to, are published on AusTender. It covers, effectively, the reporting of both ends of that procurement process, from the tender opportunity through to the contract that's ultimately entered into, the party it's with and how much it's worth.

CHAIR: How does that shape the data fields that people on the outside looking in access? That informs the data fields, which have been established to create a level of transparency, but for who?

Mr Edge : There are a number of different uses for AusTender. Prospective suppliers to government will use AusTender to identify tender notices and opportunities for them to bid for government work, so there's an audience there. In terms of the specific fields that are published around that, I'd have to get one of my colleagues to talk to that, if that's helpful. There's also interest in the contract notice reporting process that's ultimately awarded.

In terms of the specific information that's there, my colleagues could probably assist you with that as well, but certainly the name of the supplier is made clear. There's usually a contract notice number and information about the value of the contract. In some cases, as we've said, there'll be a flag as to whether it's a consultancy, and there's some other information so that the nature of the contract and the types of services can be identified.

Mr Hunt : Chair, I just might add a bit more detail. The data in AusTender, as Mr Edge suggested, is linked to the purpose that the system is intended to deliver. As Mr Edge said, it is a kind of matchmaking service at one end and then the contract notices section, which is the focus of the ANAO's report that we are talking about, is essentially a record of the outcome of those procurement processes. The government goes to the market for various purposes and then AusTender records the outcome of that approach to the market. It's really about providing transparency of the outcome of procurement processes. That does mean that the data is necessarily limited in a whole lot of ways. It's very good at recording the outcome of procurement processes but not necessarily at providing trend or annual data or any of those sorts of things.

The contract notice is reported at the time the contract is awarded, or within 42 days of the contract being awarded. It reports the total value of the contract. That's the total potential value of the contract, including potentially any extension options which may not equate to the actual expenditure under the contract. It also records at a single point in time the value overall of the years of the contract. So some contracts are relatively short-term; some go over multiple years. If you have in 2016-17 a 10-year contract published then the value of that contract appears wholly in the year which it's published in. It means that the annual figures can be volatile as large-value, multi-year contracts are published.

Another element of the data is that it's owned by the agencies and it's recorded at a point in time. So agencies can go back in and adjust those records if they discover there are errors in them. The records get amended by later contract notices which update the value or the term of the contract. So it is a living dataset. We think there's quite a lot of transparency about the data. You can access the data from the AusTender system itself, and we also publish periodic extracts of the data on data.gov, which are very useful as analytical tools but are limited because it is a living dataset that we extract at a point in time and the data can be adjusted for a whole lot of reasons looking backwards.

CHAIR: Who's the audience for AusTender?

Mr Hunt : I think it's a transparency mechanism primarily for the market. Market participants want to understand the outcome of processes which they participate in, and there's a very good reason why they might want to. They make judgements about the amount of resources to invest in a procurement process based, for example, on the estimated value and the terms of conditions that are put to the market. So they want to see, having made those judgements, that the outcome of that process aligns with their expectations when they made those judgements. That's a perfectly reasonable expectation. It fulfils a role that the Commonwealth has done through the gazettal process, going back to the start of Federation. I think we used to have a single, hard copy gazette until about the 1970s and then it split and there was a hard copy purchasing and disposal gazette that performed this function. Then some time in the nineties we got our first e-reporting system called GaPS—the Gazette Publishing System. Then that was replaced by AusTender, which performs the two roles: advertising the notices and the reporting at the other end. I think the other audience that it's reporting to, beyond the market, is the public. They can see what the outcome of those processes is.

CHAIR: Was it designed or intended to track spending?

Mr Hunt : It wasn't. The data cannot track spending. Tracking spending is a complicated task. Some of these are very big contracts, with an enormous number of different parts to those contracts. They are appropriately tracked in internal agencies, financial management information systems, which have detailed charts of accounts, and that's the place where spending is tracked. AusTender cannot track spending and isn't intended to.

CHAIR: I have a final question before going to other colleagues. This brings me to the transparency point. What sorts of things, like annual reports, can someone go to? What things can a committee like this expect around improving transparency around public expenditure?

Mr Hunt : I think we can speak for the AusTender side of things. I think we can always do a bit more. We certainly work closely with agencies, who are our main stakeholders on the reporting front. I started to explain to Mr Hill the various processes through which we do that—there are a range of them. We've undertaken in our submission to remind agencies of their obligations. Then there is a range of guidance we produce, including around appropriate use of the consultancy flag. We are undertaking a bit of work to make that clearer and to pull it together into a single document. I think in our submission we reflect on the observations of the ANAO and any recommendations from this committee in finalising that guidance, which should help agencies to improve the reliability of the data that they put into the system.

Mr GEE: Following on from that and what Mr Hill was getting at earlier, do you think that the guideline, or definition, needs to change? I take it from what you've just said that you think it might.

Mr Hunt : I don't know that we think that the guideline needs to change. There is a definition, and that definition has been in place for a while. It has served us reasonably well, I think, but there is some guidance for agencies that is available on our internet site. It's not directly located with the flow chart, but it's guidance to provide agencies with a range of additional questions they might ask about the nature of the contract and the level of control that the organisation is going to have over the individuals delivering on the contract and a range of other considerations. So we've got some supplementary guidance that sits on top of the flow chart that was picked up in the ANAO's report. I think by bringing that all together we can give a much clearer picture to agencies about how to go about this process.

Mr GEE: I guess the issue, though, is that the Audit Office is concerned that entities may have underreported consultancy contracts. It was on page 25 of the report. The Audit Office is saying it doesn't know the extent of it, because in order to find that out:

… would be required to determine the extent to which consultancy contracts are being accurately reported in AusTender using the 'consultancy' flag.

It would have to analyse individual contracts with those respective entities. What do you suggest would be the answer to this issue that the Audit Office has raised?

Mr Edge : I might just say that I think we talked about the quality of the guidance and assistance that we can offer. As Mr Hunt said, we are constantly revising that. It comes down to: Is there a good understanding of the definition? Is it clear? What can we do to assist to make sure that the characterisations are taking place correctly? There are probably quite a few things that can be done in terms of guidance and assistance that will be helpful, but we can certainly also have a look at the definition to see if there are some areas for improvement. There is no reluctance on our part to do that.

Mr Hehir : Can I make a comment on the definition. My recollection—it's just my recollection—is that the definition of consultancy was put in place largely because of concerns flowing from parliament around the use of consultancies. It's really a user issue around the definition—the user of the information rather than necessarily a Department of Finance issue around the definition. I think it's a valuable thing the committee might wish to consider whether the current definition actually serves the purpose which users like the parliament want with respect to the information. I think the use of the information is what should drive the definition, not some technical thing about what itis. If the committee has concerns that it's not serving a purpose, I think it should think about whether the definition should be changed from that perspective.

Mr GEE: I will ask another question, perhaps to the Audit Office. In figure 5.1 of your report, it gives the consultancy contract value by consultancy reason and financial year. In 2016-17, you're looking at about $700 million. Is that right? But chart 3 in the Finance submission is different. The reason for that is being given as that it's when you extract the information. What does the Audit Office make of the Department of Finance's response to the difference between those two charts, because they're showing fairly substantial different contract values in 2016-17?

Mr Hehir : Our understanding is that between when we downloaded the data and when Finance did a number of entities went in and changed the flag.

Mr HILL: So it wasn't a change to values; that's an important point.

Mr Hehir : No, it's not a change to that.

Mr GEE: Basically, none ticked the consultancy box.

Mr Hehir : I assume that, maybe looking at our report, they were concerned about definition. I don't know why it happened, but people went in and changed whether particular contracts were ticked as consultancies or not and in between the two things, so that's what happened. I can't comment on whether that was accurate or not accurate.

Mr GEE: Are you saying they went back and changed it?

Mr Hehir : Yes.

Mr HILL: That's an important point—the dollars—isn't it? I think you said before that they can change the value of the contract in dollars or they can also untick the box.

Mr Hunt : Yes, they can. The data, as I said, is live and is owned by the agencies. They can amend the record. There are instances where agencies accidentally tick a box like that through a manual entry, and they can go and amend that. I think close to the end of the financial year, when CFO areas are going through a process to finalise their end-of-year accounts, it is the kind of time when errors and omissions are identified. Generally when they amend a value, that will be through an amendment notice. Unless they've entered the wrong value by mistake when entering the contract, they can go back and update that record.

CHAIR: Which goes to your earlier point, Mr Hunt, that it is a living mechanism.

Mr Hunt : It is a live stream of data—that's right.

CHAIR: The fact someone corrects the record should be seen as a virtue?

Mr Hehir : This wasn't the end of the financial year, by the way. It was from September to December.

Mr HILL: About when you were finishing your audit—Deirdre Chambers-style coincidence.

Mr Hehir : I'm not talking about motive there. It goes as much to the accuracy of the database as anything else.

Mr HILL: I will pick up the chair's point, because there is an interesting discussion there. A virtue would be making sure that the dollar figures are updated as variations occur and final negotiations occur and so on. It seems curious to think that you could suddenly make $200 million of consultancies disappear by going: 'It's not a consultancy anymore; we just got the definition wrong.' Is that a normal occurrence, given you said that people correct things at the end of the financial year? Would you expect two-sevenths of consultancies to magically disappear?

Mr Edge : It is obviously a hypothetical question, because we're not exactly certain what the variations actually relate to. But it is very possible that someone could have classified something as a consultancy which wasn't. It is quite possible. If it is a very large value contract, it would certainly have an impact in terms of the way the numbers were reported.

Mr HILL: We don't really know why two-sevenths of the consultancies disappeared conveniently as the Auditor-General was finishing his report. That's just a mystery.

Mr Edge : We could certainly look at what was changed in the system. There's no problem with doing that.

Mr HILL: That doesn't really matter.

Ms MADELEINE KING: A point of clarification: if people untick a box, do they reclassify it as something else or is it a consultancy and then not a consultancy?

Mr Hunt : There is a flag for consultancy contracts. They're not all contracts that have an element of consultancies. AusTender catches the primary purpose of a contract, not every element of expenditure under it. Some contracts are extraordinarily large and complicated. I think it's useful to keep that in mind. I think it's useful to keep that in mind. With that in mind, if the main purpose is to deliver a consultancy—a bespoke piece of intellectual property not under direction from members of the agency, in accordance with the definition and the guidance—then agencies should tick that box. There are instances where they tick it by mistake.

Ms MADELEINE KING: I understand that and the correcting the record. There is no option to not tick the box, do you know what I mean? If you want to change it to consultancy because you think you've mischaracterised it, do you change it to something else, does it go to 'other' or is it blank?

Mr Hunt : It's a flag. It's either a yes or a no.

Ms MADELEINE KING: It does literally get unticked?

Mr Hunt : That's my understanding. I'm not an expert user.

Ms MADELEINE KING: Just a matter of clarification. Thank you.

Ms BRODTMANN: My questions go to the accuracy of the reporting, particularly in the context of small business and SMEs. I refer here to chapter 12. There is this target under the CPRs of 10 per cent and, as the report says, it looks like things are going very, very well and the fact that the targets have been more than exceeded, between 24 and 42 per cent. That is certainly not the impression I get from speaking to small businesses, particularly micros on the ground, particularly in remote and regional Australia and particularly on defence contracts. I was interested to see the findings of the ANAO in 12.5:

The absence of information regarding classification of suppliers as either SME or 'Other' (large) prevents an in-depth analysis of procurement with SMEs. However, the ANAO examined a selection of entities with a reported high proportion of SME contract value.

They found:

… a number of businesses with large numbers and values … appeared to have been classified as SMEs. These included major consultancy/accountancy firms and other Australian arms of large global business with thousands of employees.

That's completely against the definition of a small business, a micro business or a SME. The ANAO said here in the report that it provided with the ABS with the results of this analysis and they revised their definition and their methodology for the classification of what it constitutes a SME. What has Department of Finance done with that information? Has the methodology and the determining factors for what constitutes a small business, a SME and a micro business been redefined? Between 24 and 42 per cent is capturing those large businesses with thousands of employees, and we are really keen to get a clear read on what's actually happening with small business. What has Finance done in regards to that?

Mr Edge : I might start the answer and Mr Hunt might want to add to it. With the way the ABS provides advice to us, we are reliant on the ABS for advice. It is really driven by the ABN, the Australian business number, of the supplier. We are reliant on the ABS to tell us whether that ABN is a SME or not. They work with the ATO, as I understand it, to deliver that advice. As mentioned in the ANAO report, the ABS is currently still reviewing the way in which it deals with that. I think in the cases that have been mentioned, in particular the categorisation of a PwC, EY or one of those kinds of firms as a SME, they may have subsidiaries or a number of ABNs and I think what's happened is a particular contract has been flagged with an ABN that belongs to that group but perhaps the particular entity within that group is technically a SME, in that it is a separate company or has some other separation from the parent. I think that's what has happened. The ABS clearly understands the problem, from our conversations with them, and is working on fixing that.

Ms BRODTMANN: You are working closely with the ABS to get a clearer definition so we can get this clear read on small business?

Mr Edge : We are consulting with the ABS, yes.

Ms BRODTMANN: When are you looking at finalising those discussions?

Mr Hunt : We are not looking at the definition of an SME, just to be clear. The definition is 200 employees or less. The ABS is assessing its methodology for determining whether the businesses identified by an ABN in our dataset meet that definition. The methodology that they're looking at involves a database that they manage, which has tax-sensitive information, so we don't have access to that. They are assessing their methodology. I think we talked to them as recently as Monday on this subject and they're still undertaking the work on this.

Mr ENTSCH: Do we have a time frame from them at this point?

Mr Hunt : We don't, no.

Ms BRODTMANN: When will we get a clear definition of SMEs and particularly small business so that we can get an accurate read of how many smalls, micros and small to mediums are being contracted by government? Obviously, the methodology at the moment is giving us a very false impression. It is a significant concern for small business because it is misleading. The reality on the ground when you talk to micros and smalls and even mediums is that they cannot get government contracts. They're tearing their hair out because they can't get government contracts, particularly in remote and regional Australia, and this is misleading?

Mr Edge : I think it's a slightly different issue. Our commentary has been around ensuring the accuracy of the reporting, using the SME definition. What's been identified through the work the ABS is doing is that there have been a couple of anomalies in terms of the way in which that assessment has been made. We think that those anomalies are probably at the margin in terms of having an impact on the statistics. I think your question goes to a bit of a different issue around opportunities for SMEs to bid for government work, rather than reporting.

Ms BRODTMANN: No, it doesn't. We need to get a clear read on how many micros, smalls and medium enterprises are getting access to government contracts? I know that the opportunities are limited, having been a micro that tried to contract with government, so I do understand the challenges that smalls, micros and mediums have. That's not what I'm talking about. I am talking about getting a clear read on what the reality is in terms of smalls, micros and mediums being able to contract with government. Does the ANAO have any views on this?

Mr Hehir : The concern that this piece of analysis and data cutting that we did was that it was a relatively simple piece of analysis we undertook using the existing available data and, given that this piece of data was used as a part of a KPI, we would have expected that a normal level of quality assurance would have picked up something like this. That was a bit of a surprise to us around the quality assurance around the information. I don't know what our analysis means with respect to whether it is a material impact on the numbers or not, or whether our analysis is only one side of it and the methodology is undercounting in some other area. I think the issue from our point of view is the quality assurance around KPI-type information.

Ms BRODTMANN: Do you have any suggestions on how government agencies can improve the accuracy of their reporting on contracts with micro, smalls and—

Mr Hehir : Not really, we haven't done any work in that space and we'd expect it is work that the ABS is probably best placed for. That's their expertise. They are best placed to come up with a process for getting the best and most accurate estimate for the process.

Mr Hunt : Can I clarify something. The agencies don't report whether it's an SME. They report the ABN of the business that they enter a contract with and the ABS, at the end of the year, does an analysis of whether those entities are SMEs. So it's not the agencies reporting their use of SMEs. The agencies just report the contract and the ABN.

Ms BRODTMANN: Just so we're clear, the ABS does the analysis on the ABN—

Mr Hunt : That's correct, yes.

Ms BRODTMANN: and on size?

Mr Hunt : In consultation with the ATO, so I think there's data drawn from business activity statements about the particular ABN numbers.

Mr HILL: We could get the ABS and the ATO to come and explain that to us, potentially.

Ms BRODTMANN: That would be helpful with a view to trying to get—

Mr HILL: Given the data doesn't really mean anything and there's no assurance done on that.

Ms BRODTMANN: Exactly. Thank you.

Ms MADELEINE KING: A point of clarification before we move on—and I probably should have asked this earlier—and it's for the Auditor-General. I thought towards the start of this hearing, when we were talking about categorisations and perhaps the definition of consultancy, that the Auditor-General had made recommendations in 2014-15 in respect of categorisations for AusTender—did I mishear that?

Mr Hehir : We made recommendations with respect to the quality of the data on AusTender and our recommendation was that—

Ms Mellor : I'm getting it: it's in report No. 48 of 2014-15.

Ms MADELEINE KING: You've also said today, Mr Hehir, that the information in AusTender is not reliable. Is it the case that you've made that recommendation some time ago and nothing's moved on that—would that be your assessment?

Mr Hehir : The audit work we did some time ago was on limited tender procurements. One of the conclusions we drew was that the AusTender data didn't produce the accuracy and transparency that you would expect, and we made a recommendation that agencies should improve that. We've also made commentary with respect of our audits on the use of confidentiality agreements, which we've done on a regular basis flowing from a Senate order with respect to the accuracy of AusTender data. It seems that there's still significant room for improvement, driven from those types of work as well as the work that we've just done.

Ms MADELEINE KING: Has there been significant room for improvement—has there been any progress since report No. 48, do you think?

Mr Hehir : We haven't done enough audit work for me to draw a conclusion on whether it's improved or not.

Ms Mellor : In audit report No. 18 of 16-17, which is our audit work into the confidentiality in government contracts where the Senate orders require the publication of notification of the use of confidentiality provisions, only 19 per cent of contracts sampled were accurately reported in AusTender.

Mr Hehir : That's the confidentiality provisions that were accurately reported in 19 per cent of cases.

Ms MADELEINE KING: So, since you first started making that recommendation, significant improvement could still be made?

Mr Hehir : Yes.

Mr HILL: Can I now go to a different topic. In an extract of data from AusTender—whether or not it's reliable is an interesting question—there are two categories of services: temporary personnel services and personnel recruitment. I'm interested in the issue of casual labour and labour hire contracts. The data that we've got from AusTender suggests, over the last four financial years from 2013-14 to 2016-17, we've seen enormous growth from around $250 million to $749 million in the value of temporary personnel services; and, in another category, enormous growth from around $47 million to $128 million in the value of personnel recruitment. On the face of it, that's an explosion in temporary labour, scattered throughout Commonwealth agencies. Is that data reliable—can I rely on those numbers that we've dragged off AusTender?

Mr Edge : Can you rely on the numbers off AusTender?

Mr HILL: I'm trying to understand—to start with and then I want to interrogate what it all means—whether the data that's come off AusTender in relation to temporary personnel services and personnel recruitment gives an accurate picture of the use of labour hire staff throughout Commonwealth agencies.

Mr Edge : Firstly, if the information has been taken off AusTender, subject to all of the conversation today, the information is live; it is what it is. The categories are fairly broad in terms of what they might cover, so some of the services that may be included in those categories are not necessarily related to the use of contract labour. They could be related to recruiting services, for instance. I'm not sure which one of those categories something like defence recruiting would go into, which is actually engaging suppliers to assist with defence recruiting and advertising and all of that; it could be in those categories. It doesn't necessarily mean that that reporting is all about the use of temporary personnel.

Mr HILL: How do I as a parliamentarian find out—just anyone, really, the Public Service Commission, Finance, Auditor-General—how many labour hire contractors there are working across the Commonwealth? Who can tell me? Is it a thousand, 5,000, 10,000, 20,000, 30,000—anyone?

Ms Vine-Camp : We do not have that data.

Mr HILL: No-one can tell anyone in the Commonwealth parliament or the public how many temporary contractors are working in government departments?

Ms Bull : The Australian Public Service Commission only collects data on our workforce, so we can tell you how many casual staff we have, how many non-ongoing employees we have and how many ongoing employees we have—so all staff that are engaged under the Public Service—

Mr HILL: As APS staff?

Ms Bull : Yes, that's right.

Mr HILL: But we have no idea how many people with ABNs are sitting around or indeed how many thousands of human beings are sitting around in government departments through temporary labour hire firms— that's just a mystery?

Mr HART: So there's a whole area of the workforce which is being ignored?

Mr HILL: The Public Service Commission has no plans to look at this?

Ms Bull : The data collections that we do are authorised under the Public Service Act, which goes to our workforce. So, no, we haven't any plans to collect data which goes to sorts of financial arrangements that agencies might enter into in terms of contracts for services.

Mr HILL: If, for example, an agency—I'll use an example in a moment—decided they were going to enter into a permanent arrangement to have 20, 30, 100 or 3,000 staff as part of their permanent resourcing profile, procured through a labour hire firm, you don't see that as part of the Public Service Commission's business to worry about or think about?

Ms Vine-Camp : Each agency head is responsible for their own recruitment and for how they manage their agency. That would be held as an agency decision.

Mr HILL: Is that reported in annual reports? Could I go to hundreds of annual reports and add it all up?

Mr Hehir : What you'll find in annual reports, I think in the financial statements, is a line with respect to consultants and contractors expenditure.

Mr HILL: One of the concerns that we've heard in some submissions and discussion already is that part of the growth in these enormous figures has been driven by hard, non-negotiable staffing caps, ASLs. Could someone explain to me how the ASL policy works?

Dr Helgeby : The government has a commitment to maintain the size of the Public Service in ASL terms at a level consistent with or under the 2006-07 figure.

Mr HILL: When you say the size, is that a number or headcount?

Dr Helgeby : That is a number.

Mr HILL: It's not a value of—

Dr Helgeby : No, it's an ASL figure. I think this is one of the things that can trip up analysis in this area. ASL is a particular concept. It is not a headcount concept; it is not an FTE concept. It is a concept which really enables government and parliament to get a handle on or to understand, over the course of the year, what the size of the workforce is. The actual number in 2006-07 is I think 167,596. In budget papers every year, budget paper No. 4, the government publishes estimated ASL for the coming year.

Mr HILL: And that doesn't include labour hire?

Dr Helgeby : No. It is an ASL figure, so it is about the public sector workforce. That can include permanent ongoing and it can include casual—

Mr HILL: So the categories of the Public Service establishment.

Dr Helgeby : Yes. I think one of the things that came up in the discussion earlier was that recruitment activity can often be recruitment for the purpose of hiring people who are either full time or part time and ongoing or casual. A figure that tries to capture recruitment services isn't a figure that necessarily reflects on the workforce. It is about a particular activity to sustain the workforce.

Mr HILL: I want to explore this a little bit and try and understand the cost implications of this enormous growth in temporary labour hire, which, the parliament has been told in other forums, has been driven explicitly and solely in some cases by the ASL. I'll use the case study of the Office of the Director of Public Prosecutions, which is a nice, small, defined agency. Evidence was given to Senate estimates 18 August in response to:

… the government has instructed the Commonwealth Director of Public Prosecutions to return to its 2006-07 average staffing levels.

The agency responded:

The average staffing levels are now down to 390, and that is the basis.

They went on to say:

… in relation to our administrative staff, we have employed a cohort as third-party contractors, but we have yet to do that with any of our lawyers—

or so they asserted in August. They've put a cohort onto third-party contracts. They said:

That's a proportion of the workforce which we still have, but they are employed in a different way.

They said:

Generally, they're on a short-term contract for up to 12 months.

But the agency said that this is now part of their permanent resourcing outlook. There was then a discussion about whether they are paid from ongoing budget, and the answer was yes. It's not an ERC allocation for a project; it's ongoing appropriation to the agency.

There was a discussion about on-cost to the labour hire provider, and the response was:

There is a percentage that we pay to the labour hire provider. That is a combination of things, including their fees, but we also pay them an allowance which covers superannuation and the like. The fee would differ if it were lawyers, and it is a percentage.

Then the estimates committee was told those staff were paid three per cent less than the APS employees sitting alongside them doing the same work. It confirmed that it was for the same work. Then they were asked:

Is it possible to put a dollar or a percentage figure around the on-costs that you provide to the labour hire?

The response was:

Yes, it's between 25 per cent and 28 per cent.

That equates to—

… a three per cent saving in terms of the amount that the person is paid—

so you lower the wages for an employee and then you pay the labour hire firm 25 per cent. Explicitly, the agency was asked if it was cheaper or more expensive, and the answer was:

It's more expensive.

I put that there as a case study with a defined agency, which is all from the record.

To start with, could someone explain to me if there is any comment on that as a scenario? Could someone explain the costing factors or the elements that would enable a fair comparison, given we've been told by the agency it's more expensive and they're being forced to do this because there is—I could read the rest, but you can go back and look at it—no facility for them to go back to government and negotiate a different ASL?

Dr Helgeby : I've got nothing I can comment on about the particular organisation. What I would make is a couple of observations. One is, as our submission has pointed out, from a pure cost base departmental expenses are declining and have been declining as a proportion of government expenses. That's the overall setting.

Mr HILL: Excluding the NDIS and Defence, so it would be really difficult.

Dr Helgeby : But that's the overall setting. That's the context for this issue. The second thing I'd say is, if you considered the detailed tables that we publish each year in Budget Paper No. 4, which are at the agency level as well as the portfolio and aggregate level, you'll see over a period of time that, in fact, there's quite a lot of movement in the ASL attributable to individual agencies while still staying within the government's overall parameters. That takes account of the fact that government has priorities that it wants to implement and that managers make decisions about their resource mix and how they're going to do that. The key thing that the focus on ASL does is that it forces people to ask the question: what do I need to achieve this thing that the government has asked me to deliver on? Sometimes, the answer to that is that you need ongoing staff, and sometimes the answer to that is that ongoing staff is not the right way to meet that need and to meet government's needs. I think, in one sense, what you're describing in that example is the case of an entity that has basically found itself having to ask itself the question: how do I meet my needs into the future? What we wouldn't want to see, or rather what would be difficult to justify, would be for an agency which had a resource base that was a constraint on it to make decisions that locked in levels of ongoing permanent staffing that it couldn't sustain. The focus on ASL is a way of prompting the questions: what can you sustain, and what do you need to do to meet government objectives?

Mr HILL: I understand that. My concern is that hard limits appear to be driving, in this instance, a slightly perverse resource allocation, in a couple of senses. You have a certain amount of money, you have a certain set of KPIs and outcomes you have to deliver for government, and you determine your resource mix. If you're looking at that as a rational manager, why would you not employ those 30 people as APS staff and save money? Doesn't it concern you, as a finance department, that this control of one element of the inputs in the resource framework—a hard control, which is APS staffing limits—then forces agencies to go and waste money on more expensive options?

Dr Helgeby : I think managers have, for quite a long period of time, been used to working within financial constraints, and this is really just an application of constraint at the ASL level that, again, forces them to ask the question: what do I need going forward? It turns out that in many instances people need a resource for a shorter period of time, or they need a skill set they haven't got, or they need—

Mr HILL: Sorry, the evidence here—and I said that very clearly—is that it's funded from ongoing budget, and the agency said that the reason for this is purely the ASL, and there's no facility to renegotiate that with the government. I'm not talking about a project.

Dr Helgeby : I made the observation earlier that I think that, if you were to go back to the budget papers, you would see that individual agency ASL does change from time to time and from year to year. Of the two basic factors that drive it, one is government decisions about priorities, and the other is management decisions about their resource mix. So I think the idea that one cannot go back and look at that issue—I don't think, as a system, that's right.

Mr HILL: I think there'll be an outbreak of joy across the APS from that statement. Could you, perhaps, take on notice—because there was another element to the question—to come back and explain precisely what the costing features or elements that make up each type of employment would be—an APS staff member and a labour hire. What would be the normal elements of that costing?

Dr Helgeby : We'll take that on notice, yes.

Mr HILL: What work has Finance done to assess the sorts of cost impacts across the sector of this apparent enormous growth in labour hire?

Dr Helgeby : Again, I'd start by reference to the fact that departmental expenses, as a percentage of total government expenses, are actually declining. That's the first observation I would make. The second observation I would make is that obviously the Department of Finance works with individual agencies on what resources are available to them, what needs they have, what government decisions impact them and what pressures they have, and we monitor their financial position on a very regular and ongoing basis. Our primary focus is to ensure that government priorities and agency resourcing stay related to each other.

Mr HILL: You've mentioned a couple of times the part of your submission that asserts that departmental expenses as a proportion of total expenditure have declined, although that excludes defence operations and the NDIS. Why is that?

Dr Helgeby : Defence operations are not really a thing which is a normal part of government, as in, they are very variable from year to year. It depends really on what operations are in place, so you wouldn't build that into the base, and—

Mr HILL: What about the defence department itself—the APS component and the department?

Dr Helgeby : My understanding is that we typically look at the defence department as a part of government, but, when we talk about defence operations, what we are talking about is overseas operations of a military nature.

Mr HILL: Some of the functions from the DSS would have been put out into the NDIA as part of that creation?

Dr Helgeby : I'm not personally familiar with all the intricacies of that. I might take that one on notice.

Mr Hunt : I do have a bit of information from a former role I was in. The NDIS expenditure grows over the forward estimates to around $20 billion a year, and the way it was set up when the NDIA was first established was that, due to the quirks of the system, all of that expenditure counts as departmental, but we would consider the vast majority of it as administered in nature. It's not the organisation managing its affairs. I'm just speaking from recollection, but I think the departmental or management overhead was about seven per cent of that, something around that—

Mr HILL: Just on that, if I look at the table, the asterisk which said you've excluded defence operations and the NDIS applies to the departmental expenses column. Are they also excluded from the total GCS expenses?

Mr Hunt : I don't think they are.

Mr HILL: The table's a complete nonsense then. You leave them in in terms of the total cost of government, but you take them out to try and make a point that the cost to the departments has fallen, as if they're cost free to administer to the whole defence department?

Dr Helgeby : I think we've made the point that defence operations are a particular thing which would not normally be viewed as part of the—

Mr HILL: You're trying to have it both ways. You can put forward what tables you like, but it would be more helpful to have, as a comparator, a table that did include the NDIS and defence and, if you could see your way clear, to at least count the defence department, which is growing by 700 APS staff. Otherwise you could have saved the paper and the ink.

Dr Helgeby : We'll provide you with that table. We will split out, in that table, defence operations of an overseas nature, so that that is clear.

Mr HILL: Sure. If the committee or the parliament were of a mind to say, 'We would like a line of sight, in exactly the same way that we can have to the number of APS staff, to the number of temporary labour hire personnel,' on notice, could you give us some advice in a submission of exactly how we could achieve that in a sensible fashion? What kinds of recommendations would the finance department suggest we make to improve transparency around the costs, the overheads and the numbers in each agency? And where would we put that in the system—in the PGPA Act or the annual reporting rule?

Dr Helgeby : We'll take that on notice.

Mr HILL: Thank you.

CHAIR: I want to go back to this issue of the variations in the contracts. I think we've established that it is possible to make a variation to the contract during its life and that agencies are able to update the AusTender to keep the supplier market informed. We've established that—that's correct?

Mr Edge : Yes, that's right.

CHAIR: So then a contract notice can be republished in a later year, and the full value of the contract means it would have been reported in more than one year?

Mr Edge : That's right. For instance, if a contract has been varied a number of times, the original contract's published and each of the variations is published. Where there have been a number of variations, for instance, it's up to an agency to roll all of those together and republish that. In doing that, there would be a link back to the other notices in the original publication so it can be tracked through.

CHAIR: How often would that happen? Is it a regular occurrence?

Mr Edge : We'd have to take that on notice.

Mr Hunt : I think we would. Individual agencies, obviously, have different behaviours. It depends on the scale of the very large contracts they enter into. I know that Defence does this in instances where they have a contract that runs over many years. It has a large number of amendments. For transparency purposes, it gets hard to track the big picture through all of that, so they bundle all of those contract notices up and republish them in a single place.

Mr Edge : We will take it on notice.

CHAIR: I might come to the scale issue in a moment. Isn't that then another reason why the total contract notices for an agency in a given year may be nowhere close to the annual spend?

Mr Edge : Absolutely. That's the point, and that is the point we've made in the submission.

CHAIR: How many years after a contract is signed can it be updated?

Mr Edge : That goes to the duration of the contract. If you're talking about a multiyear contract, it can be updated at any point in that multiyear contract.

CHAIR: So a contract begun under a previous government could be updated under this government?

Mr Edge : As I said, a multiyear contract can be updated at any point, whether that's a three-year contract, a five-year contract or whatever.

CHAIR: That brings me back to Mr Hunt's comment about scale. At the beginning of the conversation, earlier in the hearing, I think people might have been forgiven for thinking, 'It might be a $1,000 update; it might be a $5,000 update.' What's the extreme of the quantum that has been updated in a contract?

Mr Edge : I think we'd have to take that on notice.

Mr Hunt : We would have to, to get an absolute example.

CHAIR: Defence ones come to mind.

Mr Hunt : I am aware of some significant ones. There was a republishing of a contract to do with submarine maintenance. This was a contract that started a large number of years ago and had multiple amendments. I think it was republished in 2016-17, and the value was, from recollection, between $1.8 billion and $2 billion. It was a large, significant number of historical contracts wrapped up and republished for transparency purposes in a single place.

CHAIR: So it was reported under the regime of one government but was incurred in previous governments?

Mr Hunt : I don't know exactly when the contract—

Mr Edge : We would have to take the detail of that on notice, but we certainly can. In regard to your earlier question about the largest value, we can certainly take that on notice. It may be that contract or some other contract.

Mr Hehir : In chapter 14 of our report we did some work on contract amendments. What we looked at was contracts which were amended within 12 months of being entered into and we presented some data on the number of those contracts which were amended both by value only and by duration. Figure 14.4 shows the number of contracts which had changes in value by percentage change in value. We had a go at doing it. Why did we focus on 12 months? It's interesting that a contract, even if it's a long-term contract, is varied so soon after being entered into. That was something which we thought was of interest.

CHAIR: My question, though, to be fair, went to variations being made across governments. I want to go back to the issue of consultancies and the definition. Are you satisfied, at the moment, recognising what you said, Mr Edge—it sounds to me that these sorts of issues are constantly being discussed and reviewed, even if they're not actually being changed—that the consultancy definition is clear and principled?

Mr Edge : As we said earlier, it's a longstanding definition. It's been around for a decade or two. As I also said earlier, to the extent that we think there's any merit in looking at any changes to the definition, we're absolutely prepared to do that. I think, really, that will come through some further consultation and engagement with agencies to the points made earlier about guidance and how this definition's being applied. Is it a clarity thing around the definition or is it an education thing around the people that are making those assessments? There are a lot of things we can look at to assist with this.

CHAIR: I don't want to put words in Mr Hill's mouth, but fundamental to what I think the deputy chair was getting to are issues around productivity across the Australian Public Service. I wonder if you can update the committee on the work that's being done around the functional and efficiency reviews. That's something I'm a bit familiar with from the Senate Finance and Public Administration Committee, but can you give us a sense of how they are being rolled out. Can you demonstrate that those functional and efficiency reviews are leading to an increased level of productivity across the APS?

Dr Helgeby : Functional and efficiency reviews are one part of a broader approach to improving productivity in the public sector, alongside contestability reviews and other initiatives—Project Tetris, for example, in the property area. They have largely been conducted and, because they involve recommendations to government, as government makes decisions on those they get integrated into the normal decision-making and other processes of government. A high-level summary has been published in the last two years, I think, in a row in Budget Paper No. 4. I don't have it with me, but the last Budget Paper No. 4, for example, made clear some very significant savings that have been agreed by government on the basis of functional and efficiency review recommendations.

CHAIR: And that can be expected in the next budget paper?

Dr Helgeby : As decisions have been made, they get published as each budget goes past.

CHAIR: The point of contestability is an important point. In some parts of the community, there's this idea that all wisdom resides in the Australian Public Service. I've seen some notable speeches from people who have argued that accessing independent advice to government is important, to provide for that contestability around ideas and initiatives. Is the level of contestability necessary? How much does that drive the procurement of services outside of government?

Dr Helgeby : There are a number of dimensions to that question. The functional and efficiency reviews themselves have been designed to ask the question: who's best placed to do something and can things be done more efficiently? That's really been the remit under which they've done their work and, therefore, the recommendations that have come out of those have pursued that broad approach. I think that, as the government makes decisions in relation to particular recommendations, they find their way into the normal budget processes and budget papers. In terms of decisions about contracting more broadly, those decisions are made all the time, they're made right across government and they draw on all of those considerations of what is the best value for money, what is the need, who has the skills and what is the time frame within which we need to undertake things. As those decisions are made and contracts are put out to tender or put out and then awarded, they get picked up in AusTender. So AusTender is a record, amongst other things, of a level of contestability across government.

Mr Hehir : We'll be tabling a report that goes to the functional and efficiency reviews and contestability program in the next several weeks—at the end of this month or early next month.

CHAIR: Excellent, I think. Will that be excellent, Mr Edge! I don't expect you to disclose what you might already know.

Mr HILL: I want to pick up that discussion. I fully agree with the notion that we're decades past the time when all wisdom resided in the Public Service. We have a contestable environment and we have a policy and delivery supply chain. We'll wait and see what future submissions say as well, but there are concerns that, to oversimplify, there are good reasons for going external: independent advice, audit assurance, enormous complexity, urgency, capacity and so on. I have also seen some speeches from retiring secretaries that talk about an overuse of consultants and bad or lazy reasons. Is there any guidance in the system for agencies or the Public Service that would set out some kind of decision matrix or typology that would cause people to think: 'Why am I chucking this job out for $2 million versus employing more public servants to do it in-house and build capability?' How does that decision get made?

Dr Helgeby : This really goes to what the resource framework is. The resource framework starts out with a legislative basis in the PGPA, works through the rule and works through also into the procurement rule. It's really a series of sequential duties, requirements and processes that people are to follow. To the extent that people follow those things, the system as a whole is designed to ensure that parliament, government and the public get the best value for money.

Mr HILL: I guess my concern is that, when you look at the numbers in some of the media commentary and what's been said to us privately around this inquiry, including by firms, you've got a situation where you've got a resource framework with a hard control over one aspect, which is APS staffing, and yet what that seems to have generated is a growth in the use of temporary labour and a ballooning in the use of consultancies. I find it difficult to reconcile with a lot of the anecdotes and evidence that we get that this is a rational use of resource framework. Again, policy by anecdote is never a great thing, but I was a public servant—a senior executive in Victoria. You were actually the budget deputy secretary at that time, and hard staffing limits came in. You were kind enough—or I think it was the Premier's department; I'm not sure whether the Treasury were that excited—to give me a couple hundred million dollars to do some work. I said, 'I've got to employ some people.' I went to my secretary, and they said, 'You've got the staffing limit. I said, 'This is irrational,' and they said: 'Don't worry about it. Just go and get some labour-hire contractors and some consultants.' I said, 'But that will waste literally millions of dollars.' 'Don't worry about it. It's just all you can do.' So I've lived in the middle of this reality, and this seems to be exactly what we're replicating here with your resourcing framework, because you're controlling one input arbitrarily and not the others.

Dr Helgeby : I think it would be wrong to say that the only control is an ASL control. People have resource allocations. They have appropriations. They also have ASL controls and they have a framework within which to work to make appropriate decisions. I think one of the points I made a little earlier was that the value or the strength of some of these controls is that they force the question to be asked: what do you really need in order to achieve government's priorities?

Mr HILL: I think we're going round in circles. How would you expect an agency to determine if they got value for money? I will use an example from a former Commonwealth secretary who I was talking to about this topic in the last couple of weeks. He recounted an example of getting a few million dollars in a budget to do a regulatory review. He was astounded. The first question his colleagues asked was, 'Who are you going to get to do that?' He said, 'I'm going to build a project team.' They said, 'That's a bit risky, isn't it? Don't you need a logo on it?' The project was done internally and the work was good, so the story went, and he maintains to this day that he saved a lot of money by getting staff, but that was in an environment when you could employ staff. So how do you determine if you get value for money from a $2 million KPMG job?

Dr Helgeby : That's really what the procurement framework is designed to ensure. It steps you through some requirements and some obligations, and Finance provides significant guidance on how to do that, including a contracts suite.

Mr HILL: I get the market testing bit, but you've got a product. How do you determine whether that report was ever worth $2 million?

Mr Edge : Well, I think there's a range of factors that you'd have to take into account in deciding whether to outsource a piece of work. In the case of a regulatory review—and I can't comment on the specific case you are talking about, obviously—sometimes independence is required. If you are looking at the activity an agency does and there is a requirement for stakeholder consultation, an independent review, clearly, you would be thinking whether you needed to outsource that work.

In many cases, also, there is a requirement for specialist skills—very highly-applied commercial skills, understanding of market capitals and other things—that really don't exist in any kind of critical mass way in the Public Service. In each instance, you would have to look at all of those considerations and a whole lot of other considerations in making that assessment.

Mr HILL: Sure. So, there is no global way to step back and say: 'We've seen growth like this in the use of external firms. How does the Commonwealth know when it's getting value for money?'

Mr Edge : Obviously, there are requirements in the Commonwealth Procurement Rules to approach the market. And once you've worked through a series of decisions about whether this particular piece of work will be outsourced, the Commonwealth Procurement Rules require it to be subject to some value thresholds and tested in the market. A scope of work is tested in the market and you get responses from the market about what it's going to cost to do that work. An assessment is then done about making a selection based on the selection criteria, and value for money is a very important consideration there.

Mr HILL: Does the Public Service Commission have any view around assurance of building internal capability? There are different models of procurement, ranging from, 'We put a tender out, we give a couple of million bucks to a big consulting firm and they give us a report,' versus shared project teams, where some APS staff may spend some time working in the consultancy and vice versa. There is a whole spectrum of models. Do you have thoughts or views, given that we are now spending billions of dollars of taxpayer money on external work?

Ms Vine-Camp : Agencies will make a decision based upon the project that needs to be done. You do see cases where there will be mixed project teams working. We don't have a view or present a rule around how agencies need to do this work. It really depends on the project that they're wishing to undertake. We do know that there is a wide variety of reasons for why people, agencies, will bring in external providers. Indeed, that can be because it's something that is short term or where we don't have the skills. We know, for example, there is not a huge amount of digital and data skills within the Australian Public Service, and they're actually hard skills to find even globally. So it's often those things that are brought into agencies to do pieces of work.

We also know that there are good pieces of work that the private sector do better than we do. Building, fit-outs and things like that are pieces where agencies will bring in an external provider to do them.

Mr HILL: I understand that. I suppose the point I'm trying to get to is: do you have a typology or any guidance, given that there are also concerns as part of this inquiry about the impact of excessive outsourcing on capability in the Public Service? And you also need to be an informed client, otherwise how do you know whether you have just bought $2 million of crap or not?

Ms Vine-Camp : We don't provide any sort of guidance to agencies about how they might use external providers.

Mr HILL: Okay. So there are no current case studies, or different models or anything which—

Ms Vine-Camp : No.

Mr Hehir : I think that the key to your question is basically one of contract management. It doesn't matter whether you're buying a large piece of infrastructure or a consultancy service; unless you've got a contract which clearly defines what it is you're expecting to get so that you can manage the provider to deliver what it is and, depending on the scale of the procurement, some form of evaluation of whether it was fit for purpose, then you can't assess value for money. There's a general expectation that people involved in procurement will have contracts which clearly describe what it is that they're purchasing so that they can assess whether they've achieved value out of it.

The front end is competitive procurement processes to make sure that you're getting value for money on the basis of tender documentation that clearly sets out what you want to get, entering into a contract which is consistent with the tender documentation and then holding the provider to account. I know it's a bit simplistic, but that's the methodology that you would expect to see to assure value for money.

Mr HILL: Would it be possible to develop a better typology of the outsourcing of this kind of work to inform transformation in reporting and, arguably, get rid of this very narrow definition of consulting and put the whole lot on the table in annual reports and so on—having some kind of typology that would help agencies, individual managers and executives work through a logical decision as to, 'When is it good to outsource?' versus 'When should we do this work in-house'? Would it be possible to develop such a typology?

Mr Hehir : In my organisation, we do. For example, we have in-sourced, outsourced and internally provided audit work. We set up an internal process where we determine what we in-source and what we outsource and why. It's risk based on a whole pile of other things. From personal experience, yes, you can do that. I've never done an audit to determine whether other organisations do it.

CHAIR: I have a question for the Public Service Commission. Is there any evidence that there needs to be stronger management around the planning of the use of consultants or on-hire labour companies to better predict capability issues across agencies and departments?

Ms Vine-Camp : I'm not aware of any. As I have said previously, agencies are ultimately responsible for how they use their resources and how they plan for that. We haven't—

CHAIR: But you would have some visibility to poor planning?

Ms Vine-Camp : We see the public reviews that come out of issues. We certainly have not proceeded to develop any rules or structure around that.

CHAIR: Do you have any general observations? You didn't make a written submission, but do you have any general observations about the use of consultancies or on-hire labour companies across the Public Service?

Ms Vine-Camp : Only that there will be many examples and times when agencies need to bring in specialised work, specialised assistance, and I've already outlined some of those times. We believe that a modern public service will be required to draw its expertise from a wide range of different areas and, in cases, that will require labour hire or consultancy services.

CHAIR: Has any evidence come to you over recent years about the inappropriate use of on-hire labour companies or consultancies?

Ms Vine-Camp : Not that I'm personal aware of—no.

Mr HILL: Given you don't know how many there are.

Mr HART: I have some general questions for Finance with respect to AusTender and its purpose and what I might describe as the business case for AusTender. Obviously, we're aware of the procurement rules and all of the underpinnings, but what's the formal relationship between those rules and AusTender as we see it on the web?

Mr Hunt : The rules dictate a requirement to publish notices on AusTender for contracts over the contracting threshold, which is $10,000.

Mr HART: I understand that.

Mr Hunt : As I said earlier, the purpose is to reflect the outcome of the procurement process.

Mr HART: I understand that. Regarding the particular dataset that's represented by AusTender, is there any flexibility as to the inclusion of more data points within the current database which is now described as AusTender?

Mr Hunt : It's an IT system. I'm not an IT technician, but it comes back to the intent of the system. I talked through a bit of the history of gazetting—

Mr HART: I understand that. That's why I'm asking the question.

Mr Hunt : I presume we can add extra fields to the system, but our assessment is that it captures the key details of the main purpose of a contract and has done that for a number of years, since it was established.

Mr HILL: In relation to Mr Hart's question, on the categorisation, do you use the UNSPSC to categorise different work?

Mr Hunt : We use a subset of the UNSPSC. The UNSPSC has over 50,000 codes—I think it is 64,000. My understanding is that when we first implemented the system we had the full code set. Compliance was understandably challenging for agencies—

Mr HILL: So you can chunk it up a bit.

Mr Hunt : We worked with agencies in 2008 to define a subset of that code, so there are 572 of those 60,000-odd codes which meet agencies' business needs, and we review that annually with agencies. It was last updated in about 2014 to add some cloud-computing services. We took those services off the 60,000 and added them to our subset, and that subset is published on data.gov, so it's accessible.

Mr HILL: If that UNSPSC subset is used on AusTender, is there any barrier, for instance, in thinking about that as a basis for annual reporting, if we wanted to get full transparency on expenditure?

Mr Hunt : Again, it comes back to the purpose, which is recording the outcome of the procurement. There's no capturing of, and there's no intention to capture, actual expenditure under those contracts.

Mr HART: By the same token, there's no impediment, should Finance choose to, or should parliament choose to direct Finance to. If we directed the capture of additional data, then surely you could include further datasets within the database.

Mr HILL: To add to that, there are two scenarios there. If we were interested in transparency, I think you've heard from every member of the committee that one approach would be to augment AusTender, expanding its purpose through adding expenditure data. Another approach would be through blowing up the consultancy definition and completely re-establishing what is put in annual reports around the extent of outsourcing. Both approaches may have ups and downs.

Mr Hunt : I think adding expenditure data would be an extraordinarily complicated and costly process.

Mr HILL: Why?

Mr Hunt : I think it would have to be a policy proposal to get significant resourcing to do it. It would completely change the purpose of the system. I talked earlier about the complexity of expenditure data that's captured in agency financial management information systems. There are huge and elaborate charts of accounts that agencies like the Department of Defence use, and that's the appropriate place to capture expenditure. It gets very, very granular very, very quickly. The system as it currently stands could not just have a couple of extra fields added to it to capture expenditure.

Mr HILL: In that case, if the committee is of a mind to make recommendations to improve the transparency around expenditure, can you then provide us with clear advice, building on the requests earlier, on how we would do that, where we would do that and what the basis of such a system would be? You've heard the two core concerns. At the moment there's no transparency across the entire Commonwealth about labour hire contractors—human beings—doing APS work, there's no transparency about how many billions of dollars are spent on the four large firms and, indeed, their travelling companions, and there's no transparency about how many former senior public servants with ABNs are wandering around departments on daily rates. None of this is clear, and it's adding up to billions of dollars of taxpayers' money.

Mr HART: Could you take that on notice. My question is for the Auditor-General. How easy has it been for the ANAO to extract data from the dataset at AusTender?

Mr Hehir : It's not that difficult, really. It took us a bit of set-up work, but most of the time I think was spent on deciding what questions to ask the system rather than getting the data out.

Mr HART: I know there has been some evidence about the quality of the material within the database, but are you able to offer any advice to the committee about the general quality of the database, noting the limitations that have been expressed by Finance?

Mr Hehir : I think the limitations are the ones that Finance have talked about. The database is there for a particular purpose—

Mr HART: Limited by design.

Mr Hehir : That sets what data's in there. If you try to present information that isn't precisely related to that purpose then you have to do certain things to it. The point was made around big, lumpy contracts, so when presenting some of this information we excluded some contracts which went for a decade and, if you put the data in, it made the rest of it look silly. You've to make assumptions when using the data to try and get something out of it which actually may be useful. That's because the system is predominantly designed around the disclosure of individual things, not around telling a story about aggregates.

Mr HART: That brings me exactly to my next point. In fact, the quality of the material which comes out the other end is very much limited by the design of the database. Its function is very much analogous to the gazette. In other words, the quality of transparency that's supplied is very limited, simply by the nature of the volume of data that one has to sift through to find any information.

Mr Hehir : If you want to try and get information on where a contract's going, that's the database, unless you go in and interrogate individual financial management systems.

Mr HART: Aggregation of data is very difficult by virtue of the limitations that are internally imposed upon the database.

Mr Hehir : It raises questions. You have to be very clear about the assumptions you are making in doing the aggregation, which we've tried to do in our paper.

Mr HART: In previous conversations in previous inquiries, we've talked about the concept of having databases that can perform many different functions—in other words, one set of data with many queries and, therefore, many different objectives. Are you able to offer any evidence as to the utility of having some limited further material within the database so that it would report as to aggregated data?

Mr Hehir : I think what we've tried to do is collect data that raises questions rather than tries to answer them. The questions that we were most interested in were largely ones to do with resource allocation and value for money, and to use the data to raise questions about whether the system is performing as it would expect it to be done. I would have thought that that is something of interest for the system, to know whether things like the CPRs are creating outcomes which they are expected to, which they expect to be achieved from it. Particularly in a devolved system where you basically say, 'The accountability for compliance entirely rests with accountable authorities,' how do you ensure the accountability for that?

We've seen that in questions the committee has asked today, and on Wednesday. If the answer is always, 'It's up to the individual and we don't do any aggregation to do any analysis,' you can't determine whether the system's producing the outcome that you want. Part of our reason for doing this analysis was to say, 'Can you get any insight into the system from the data?' I think you can get a little bit, but you'd need to do a lot more research before you could come to a conclusion. What we've done here, together with our audit work we've done on procurement, suggests that there's a fair amount of work that can be done to ensure that greater value for money is being achieved through procurement practices.

CHAIR: Before closing, would the Department of Finance like to reflect on the Auditor-General's comments?

Mr Edge : I will just make a point. There's been quite a long discussion about AusTender and what it is and what it isn't. I'd hate to leave the committee with the impression that it's a dataset that can't be interrogated. I think it's obvious from our submission and the work that the Auditor-General's done that you can derive data, derive reports, search it and get a report out of it of every contract notice that's been signed with a particular supplier for multiyear or single-year periods. I wouldn't like to leave the impression that it can't do those things. It absolutely can do those things. What it reports on is the value of the contracts that've been signed with those entities, not the actual expenditure—we've been around that—but it can do that. It is a dynamic dataset, it's a live dataset and it has, in our view, significant value in terms of the things it can do and the reporting it can provide.

Mr Helgeby : It seems to me that some of the discussion today has been around the distinction between a forward-looking and real-time source of information about contracts and reporting, which is typically ex-post, lags by quite a bit and has a different set of arrangements around it. In one sense, the committee appears to be exploring whether it wants to bring together the reporting of certain types of financial arrangements with the real-time discovery of a forward-looking approach. It may be that when we respond to the questions on notice we might want to consider some of that also from the annual reporting perspective.

CHAIR: That's right. Thank you all again for making time to attend this morning's hearing and giving evidence before the committee. The secretary will be in contact if the committee has any further questions of you. I now close this public hearing, and the next meeting of the committee will be at 8.30 am on 23 March 2018. I now adjourn the committee hearing.

Committee adjourned at 11:36