Australian Institute of Criminology

Skip to content


Comparing internal and external frauds

One of the clear findings from the present survey concerned the differences that existed between agencies’ experience of internal as opposed to external fraud. As in previous years, agencies were asked to report their findings separately for incidents of fraud allegedly committed by staff and contractors (internal fraud) and by parties external to the agency (external fraud). The following discussion compares the results obtained from the survey on this variable to determine if similarities or differences were present in the incidence and cost of fraud experienced as well as in the nature of the incidents reported.

Fraud types

Overall in 2009–10, external fraud affected slightly more Australian Government agencies (n=51 agencies, 34%) than internal fraud (n=47 agencies, 31%). Consistent with the greater number of agencies affected by external fraud, agencies also reported a greater number of incidents related to external fraud (n=702,941) than internal fraud (n=3,001). Yet while external fraud affected more agencies generally, the fraud types that resulted in the largest number of incidents tended to affect only a small number of agencies. The two external fraud types that produced the largest number of incidents—‘fraud relating to social security’ and ‘fraud relating to visas and citizenship’—were reported by only two agencies. For internal fraud the most frequent incident type—‘obtaining or using personal information without authorisation’—affected only eight agencies.

The method of ‘obtaining or using personal information’ accounted for the largest number of incidents of internal fraud (n=869, 29% of incidents), while only 0.1 percent of incidents of external fraud involved this method. When looking at fraud types that affected most agencies, rather than those that resulted in the largest number of incidents, fraud involving ‘financial benefits’ was the most prevalent type for agencies in both internal and external fraud, affecting 20 percent and 21 percent respectively (Table 28). This is consistent with the findings reported in 2008–09.

Table 28: Agency experience of fraud by focus of incident—summary (number and % of agencies)a
Type of incident Agencies that experienced an internal fraud incident (n) Total agencies (%) Agencies that experienced an external fraud incident (n) Total agencies (%)
Equipment 27 18 24 16
Entitlements 25 16 19 13
Information 22 14 22 14
Financial benefits 30 20 32 21
Other frauds 5 3 11 7

a: percentages have been rounded to whole numbers

Source: Commonwealth fraud survey 2009–10 data (AIC computer file)

Fraud methods

The most common methods used to carry out fraud were consistent whether the alleged offender was an internal employee within the agency or external to the agency. ‘Misuse of documents’ was reported as being committed internally in 16 percent of agencies (n=24) and 17 percent externally (n=26). ‘Misuse of IT’ and ‘corruption’ were the next most commonly committed methods for internal fraud (20 agencies in both cases), while ‘misuse of identity’ and ‘other’ fraud methods were the next most common for external fraud (17 and 16 agencies respectively) (Table 29). These findings mostly remained consistent with the 2008–09 findings.

Table 29: Agency experience of fraud by method of commission—summary (number and % of agencies)
Type of incident Agencies that reported method for internal fraud (n) Total agencies (%) Agencies that reported method for external fraud (n) Total agencies (%)
Misuse of information technologies 20 13 10 7
Misuse of identity 7 5 17 11
Misuse of documents 24 16 26 17
Corruption 20 13 10 7
Other methods 13 9 16 11

Source: Commonwealth fraud survey 2009–10 data (AIC computer file)

Some methods used to commit fraud, however, showed marked differences depending on whether the accused was internal or external to the agency in question. The principal areas of difference were as follows.

  • The specific method used in the largest proportion of incidents involving external fraud was ‘failing to provide documents when required to do so’ (n=7,573), an increase from 2008–09 (n=794), while for internal fraud it was ‘accessing information or programs via a computer without authorisation’ (n=1,011), showing a decrease from 2008–09 (n=1,816).
  • There was a large discrepancy between internal and external fraud in the number of incidents that involved the ‘misuse of IT’ and ‘accessing information or programs via a computer without authorisation’ more specifically. This was the most frequent method of internal fraud incidents; however, it was reported in 460 incidents of external fraud by four agencies, compared with only three incidents by three agencies in 2008–09.
  • By way of contrast, ‘unauthorised use of another person’s Tax File Number or Australian Business Number’ was one of the most common external fraud methods used in 2009–10, although it was only reported by two agencies. In the case of internal fraud this method was reportedly not experienced by any agencies, representing only a reduction of one incident reported by one agency in 2008–09. In 2008–09, agencies reported no incidents of external fraud committed through the use of ‘unauthorised use of another person’s Tax File Number or Australian Business Number’. Further inquiries by the AIC revealed that changes in one agency’s interpretation of the definition of fraud and consequential changes to counting rules were responsible for the large increase in incidents involving this method in 2009–10.
  • In 2008–09 no agencies reported any external fraud involving ‘unauthorised use of another person’s Tax File Number or Australian Business Number’ as the primary method. It is likely that the reporting or business activities in the two agencies that reported this method have changed, accounting for the increase in fraud using this method.

It should also be noted that across both internal and external fraud, there were fraud incidents for which agencies could not categorise the method. Most significantly, seven agencies reported 23,123 fraud incidents in the ‘misuse of documents’ method and were unable to further categorise these. A further 8,711 fraud incidents fell into the ‘misuse of identity’ category; however, they also could not be further defined. As discussed below, 40 agencies reported the extremely large number of 608,704 external fraud incidents that were simply classified as ‘other’ for method of commission.

Losses and recoveries

Of all agencies that experienced at least one incident of fraud (n=61), there were 41 agencies (42%) that specified a financial loss arising from either internal or external fraud. This finding was consistent with the number of agencies that suffered a loss in 2008–09. For internal fraud, 25 agencies identified a loss; for external fraud 33 agencies identified losses, compared with 29 agencies for both internal and external fraud in 2008–09.

External fraud resulted in substantially higher losses to Australian Government agencies than internal fraud. The majority of external fraud losses were a result of fraud targeted at entitlements, with an estimated loss of $487,585,211 across eight agencies. The 2009–10 loss represents a slight decrease from the loss due to entitlements fraud experienced in 2008–09, which was estimated as a $488,588,745 loss. Similarly, for internal fraud the loss from entitlements was also the most expensive category, resulting in an estimated loss of $1,282,076, which had increased from $860,862 in 2008–09. Overall, more than $497m (n=$497,573,820) was lost to fraud in 2009–10.

For both internal and external fraud incidents, the fraud type that resulted in a loss for most agencies was fraud targeted at ‘financial benefits’. This affected 17 agencies that reported internal fraud and 18 agencies that reported external fraud. Compared with the 2008–09 reporting period, fewer Australian Government agencies reported financial losses, with 23 and 21 agencies reporting financial loss from internal and external fraud respectively.

Agencies were able to recover money from external frauds (n=22) and internal frauds (n=15) in 2009–10. While rates of recovery differed, the methods used to recover the largest amounts of money were administrative methods for both internal and external fraud. Money recovered from losses relating to ‘entitlements’ was most common for both internal fraud and external fraud recoveries.


Suspects were slightly more likely to be identified by agencies for internal fraud than for external fraud. In total, 94 percent of agencies with an internal fraud identified at least one suspect, compared with 88 percent for external fraud. The high rate of identified suspects is not surprising given that internal fraud is committed by employees who are known to the agency. The high rate of identified suspects for external fraud is more surprising, since external fraud can be committed by any member of the public, although this rate has dropped slightly for external fraud since 2008–09 (96%). Consistent with findings from the 2010 KPMG survey, non-management employees were the most likely group to commit fraud, at least with respect to the number of incidents, although not with respect to losses sustained (KPMG 2010: 6).

Looking at fraud more generally, of the 61 agencies that reported any incident involving either internal or external fraud, 74 percent (n=45) identified at least one suspect for one fraud type. These results showed a reduction in the suspects identified compared to 2008–09, with 97 percent of suspects identified as having committed at least one fraud.

However, while agencies were more likely to identify individuals suspected of internal fraud, the number of individuals suspected of external fraud was much larger. The fraud type that had the highest number of external suspects, ‘entitlement fraud’, involved 691,111 suspects and 13 agencies, while ‘information fraud’ had the highest number of internal fraud suspects at 961. Compared with 2008–09, the number of suspects fell from nearly 790,000 for external fraud and decreased from around 1,800 for internal fraud out of 3,001 internal fraud incidents. In 2009–10 the highest number of suspects identified from one agency was 613,498 relating to external ‘entitlement fraud’ and the highest number of suspects identified from one agency for internal fraud was reportedly 548 for fraud relating to ‘corruption ’.

Detection and investigation

The most prevalent methods for detecting frauds were the same for both internal and external fraud, namely ‘internal controls/internal audit’ which was used to detect 1,275 incidents of internal fraud and 596,094 incidents of external fraud. ‘Internal controls/internal audit’ was also the method of detection found to be most prevalent in the 2008–09 reporting period, accounting for a similar number of internal fraud incidents as in 2009–10. In the case of external fraud, this method of detection decreased from 730,145 in 2008–09 to 596,094 in 2009–10. Consistent for both internal and external fraud, the greatest number of agencies detected fraud through ‘internal controls/internal audit’, with 42 agencies using this method for internal fraud detection and 33 agencies using this method for external fraud. Once incidents were detected, both internal and external frauds were most likely to be investigated by the agency alone.

In addition to detection and investigation within agencies, financial crime investigations have also been undertaken by taskforces that involve law enforcement and intelligence agencies as well as the government agency in question. Project Wickenby is an example of a multi-agency collaborative taskforce implemented to investigate organised tax fraud and associated offshore money-laundering activities. It involved a number of investigative and intelligence agencies, including the Australian Taxation Office (ATO), AFP, Australian Crime Commission (ACC), Australian Securities and Investments Commission (ASIC), and was supported by the Australian Transaction Reports and Analysis Centre (AUSTRAC), the Attorney-General’s Department and the Australian Government Solicitor (AGS) and international partners when necessary. The CDPP played an important role in prosecuting offenders and recovering proceeds from these crimes under the Proceeds of Crime Act 2002. Project Wickenby was funded between February 2006 and June 2010, and as at 30 September 2010 the CDPP had prosecuted 60 defendants in various jurisdictions investigated by the AFP and ACC. In addition, the CDPP has recouped a total of $513.2m in outstanding tax collections, compliance dividends and other monies as a result of a number of Project Wickenby matters (ATO 2010b). Box 4 shows some of the successful outcomes arising from Project Wickenby.

Box 4: Australian Taxation Office

The ATO is one of Australia’s largest public sector agencies, which in 2009–10 collected net cash revenue of $253.2b. As such, it is an attractive target for those seeking to defraud the Commonwealth.

Since 2006, Project Wickenby, a multi-agency taskforce, has worked across Australia to deter, identify, audit, investigate and prosecute promoters, intermediaries and participants involved in abusive tax haven schemes. Since 2006, 7,243 cases have been completed under Project Wickenby and raised liabilities totalling more than $1.7b.

As a result of Project Wickenby, the ATO has collected $174m from active compliance adjustments and a further $302m has been collected due to improved compliance activities from those previously subject to Wickenby action. As such, Project Wickenby is proving an effective deterrent to tax evaders. The ATO has uncovered that there has been a reduction since 2007–08 in the flow of finances from Australia to tax havens such as Switzerland, Vanuatu and Lichtenstein, areas of focus for Project Wickenby, with a reduction of between 18% and 45% in cash flow between 2007–08 and 2008–09 and from 12% and 40% between 2008–09 and 2009–10. The ATO has put in place a number of publications and educational measures to deter would-be tax evaders and tax practitioners from offending.

One individual was detected and subsequently prosecuted for tax evasion under Project Wickenby. The accused, a solicitor, was a director of a company at the time the alleged offences were committed. Over a period from January to March 2003, the defendant concocted two separate situations where he was evading tax through the use of his overseas banking account based in Switzerland. This led to the defendant being charged with two counts of defrauding the Commonwealth, under the Crimes Act and one count of conspiring to dishonestly cause a risk of loss to a Commonwealth entity, under the Criminal Code. On 15 April 2010, the defendant was convicted by the Supreme Court of Victoria and sentenced to 12 months imprisonment with a condition that he be of good behaviour for 12 months thereafter.

Source: ATO 2010a ; CDPP 2010b

Changes to agency fraud procedures

Agency-specific changes may have an impact on fraud detection and investigation, including advances in staff resourcing and training. In 2009–10, the Australian Customs and Border Protection Service delivered training in fraud control and ethics to 2,473 officers, a substantial increase from the 57 officers trained in 2008–09. The training was provided throughout Australia in accordance with the Fraud Control Framework, which seeks to encourage preparedness, prevention, detection and resolution of fraud-related activity. This level of enhanced training may have contributed to the observed reduction in fraud incidents reported by the Australian Customs and Border Protection Service during that period. In support of prevention activities, the Australian Customs and Border Protection Service implemented the Compliance Action Plan in 2009–10. This is a risk-targeted program of treatments to ensure compliance with Customs and Border Protection requirements. This led to 244 audits of imports and exports and 259 other post-transaction activities to investigate revenue and regulated goods risks; visits to 62 percent of licensed depots and warehouses; engagement with cargo terminal operators; and pre-clearance intervention on approximately 161,000 transactions to test compliance and risk.

In April 2010, the Australian Government announced new anti-smoking measures, which included raising the tobacco excise by 25 percent. As such, an increase in tobacco smuggling resulted and led to 60 convictions, $513,616 in fines and penalties and nine custodial sentences ranging from one to 10.5 years (Australian Customs and Border Protection Service 2010).

Box 5 highlights some of the Australian Customs and Border Protection Service activities relating to fraud detection and investigation for 2009–10.

Box 5: Australian Customs and Border Protection Service

The Australian Customs and Border Protection Service investigates serious, complex and sensitive breaches of a range of border controls in accordance with legislation covering customs and the Australian border control. Investigations comply with the Commonwealth Fraud Control Guidelines and the Australian Government Investigation Standards in collaboration with other agencies such as federal, state and territory police.

In 2009–10, the Australian Customs and Border Protection Service inspected (via X-ray) 101,822 sea cargo containers; 1,492,762 air cargo containers; and 62,209,682 items of mail and examined (by actually opening the item) 14,175 sea cargo containers; 66,821 air cargo containers; and 202,858 items of mail.

The activities of the Australian Customs and Border Protection Service led to the prosecution of an excise fraudster during the reporting period. The defendant, one of two directors of a company that was predominantly involved in the importation of goods, (the mailing of clothing into Australia), arranged false statements from six Chinese companies on 57 occasions. These invoices were then used to falsely declare the goods to the Australian Customs and Border Protection Service. By understating the value of the goods, the defendant avoided paying excise duty of $195,524.01. Both company directors were charged with 57 counts of obtaining a financial advantage by deception, under the Criminal Code; the defendant was also charged with an additional 45 counts of aiding, abetting, counselling or procuring the company to obtain a financial advantage by deception. Both company directors pleaded guilty in the Downing Centre Local Court in Sydney on 16 April 2010. The company was fined $195,524.01 and the defendant was sentenced to 16 months imprisonment and released forthwith on condition that he be of good behaviour for 26 months.

Source: Australian Customs and Border Protection Service 2010; CDPP 2010

Similarly, Medicare Australia’s National Compliance Program is risk-based and identifies a mixture of activities to support and manage compliance. Factors identified that may lead to non-compliance include incorrect claiming and inappropriate practice by health care providers. Medicare Australia focuses on further refining and extending support services and enhancing the audit system to overcome non-compliance (Medicare Australian 2010). In addition, Medicare Australia implemented a new product to identify and analyse compliance risk in 2009–10. This new product involved examining the nature of the risk, relevant legislation, any factors that cause the risk to occur and any other issues that could affect how Medicare Australia responds (Medicare Australia 2010).

In 2009–10, Medicare Australia received 2,876 tip-offs from health care professionals and members of the public that were further assessed to determine whether non-compliance had occurred. Medicare Australia has seen a significant reduction in frauds reported, which could be attributable to these practices (Medicare Australia 2009). Box 6 highlights some of Medicare Australia’s achievements for 2009–10.

Box 6: Medicare Australia

As part of its responsibilities to protect the public interest, Medicare Australia has a fraud control program that complies with the Commonwealth Fraud Control Guidelines. Under this program Medicare Australia:

  • prepares fraud risk assessments and fraud control plans in accordance with the Guidelines;
  • puts in place appropriate fraud prevention, detection, investigation and reporting procedures and processes; and
  • collects and reports annual fraud data in line with the Guidelines.

The Australian Government services fraud tip-off line helps Medicare Australia conduct further assessments into fraudulent claims. In 2009–10, Medicare Australia received 1,864 tip-offs from health professionals and members of the public, and a further 1,011 through other avenues which were further assessed to determine whether non-compliance had occurred.

Medicare Australia continued to build on the new compliance framework developed in 2008–09, which expands compliance responses to include provision of targeted information, audits to verify compliance, reviews of practitioners and investigations. During 2009–10, Medicare Australia finalised 3,594 compliance claims, of which eight individuals were referred to the CDPP for criminal prosecution. These eight referrals and four prior referrals led to the prosecution of 12 members of the public, four medical practitioners and one pharmacy/pharmacist and the recovery through repayment orders of almost $418,942 in 2009–10. In addition, Medicare Australia recouped almost $10.3m in the recovery of benefits incorrectly paid through administrative means.

Source: Medicare Australia 20010

Fraud trends in 2009–10

Australian Government agencies are subject to a diverse range of fraud risks. Agencies can be targeted internally, as well as from external entities, using diverse methods of offending, and fraud attempts can focus on the various benefits or services an agency may provide. The threat posed to governments by external fraud is often different to threats from internal fraud. This year’s survey demonstrated that the types of fraud incidents and methods used for internal and external fraud were often different. Further, the experience of external fraud has been quite different across agencies, and the types of fraud incidents that affect each agency are often unique to the service provided by that body.

While the risks relating to fraud involving social security or visas may appear high due to the large number of incidents reported, these fraud types will only affect the few agencies that are involved with delivering those services. These results highlight the ongoing need for individual agencies to be aware of the unique risks they face and to continually update fraud prevention strategies and control plans so they are tailored to the needs of their agency.

While the results of the survey showed that internal and external fraud are often targeted at different benefits and use different methods, it appeared that they were not entirely separate phenomena. Agencies that experienced either internal or external fraud were significantly more likely to experience the other type as well. This relationship worked in both directions, indicating that while there were differences in the types of incidents between internal and external fraud, agencies were not necessarily more likely to be a victim of one type over the other.

Another consideration to arise from the survey results was that when attempting to measure fraud levels and the risk of victimisation, it is important to analyse data looking at both the number of agencies affected as well as the number of incidents recorded. It has been shown consistently since 2007–08 that a large number of fraud incidents affect only a small number of agencies. Although some fraud types resulted in a substantial loss to the Australian Government because of the large number of incidents or the amount of money lost, these often tended to only affect specific agencies. Fraud types which resulted in a smaller financial loss could affect a larger proportion of Australian Government agencies.

The size of agencies was also shown to be an important factor in fraud victimisation. Larger agencies in the 2009–10 survey were more likely to have experienced fraud than smaller agencies. This supported previous findings in the private and not-for-profit sectors that larger agencies are more vulnerable to fraud (BDO Chartered Accountants and Advisers 2008; KPMG 2009). The 2010 ANAO survey found that smaller agencies with fewer than 249 employees had less rigorous fraud reporting and fraud control procedures in place (ANAO 2010). This could, arguably, lead to fewer fraud incidents being detected and reported, thus contributing to the overall lower incidence of reported fraud among smaller agencies. Further research is needed to explore this aspect in more depth.

A positive relationship was also found in the results of the present survey between the agency having a dedicated fraud section and detection of fraud. There are a number of possible explanations for this. It could be that having staff dedicated to fraud control leads to enhanced vigilance in detecting fraud within the agency. Alternatively, it could be argued that agencies with enhanced fraud risks, or previous experience of fraud, have found it necessary to increase fraud control resources and staffing to deal with the problem. Again, further research is needed to explore this.

The present report did not attempt to calculate the cost of responding to fraud, although the burden that responding to fraud places on agencies through time, resources and financial losses should not be underestimated. Nor has this report attempted to calculate the cost of fraud prevented through anti-fraud awareness campaigns.

This report has, however, provided policy-relevant information about the types of fraud that affected Australian Government agencies and the methods used to commit them. As the survey used to collect the data for the report is substantially the same as in 2008–09, it was possible to make certain comparisons between the results for the two years. In future years, the results from the annual survey will be able to provide trend data over a greater period of time, which will benefit agencies in preparing fraud control policies based on knowledge of changing trends in fraud and its prevention. The present report also shows the need for more consistent data-recording practices and measurement of fraud, particularly concerning the question of the extent to which regulatory non-compliance ought to be included within the scope of the Guidelines and survey.

Future data collection

Scope of the Guidelines and definitions of fraud

In gathering information for the current report, some agencies alerted the AIC to a number of difficulties they had encountered in responding to the annual survey and in interpreting the provisions of the current Fraud Control Guidelines. Difficulties arose because of the wide diversity in agencies subject to the Guidelines both for fraud risk as well as the size and nature of their operations.

One of the major limitations for agencies in complying with the Guidelines when reporting fraud to the AIC lies in the inability of the Guidelines to ensure consistent data is collected from across agencies and within agencies from year to year. The competing interests of adhering to the Guidelines and developing policy, reporting ‘fraud’ and managing daily, ongoing compliance concerns create complex issues for larger Australian Government agencies. Over the last three years, considerable differences have emerged in the way in which large agencies have defined fraud for reporting purposes, resulting in large variations from year to year. This has led to a lack of clarity in measuring overall levels of fraud experienced.

Although the Guidelines provide a clear statement of what agencies should report each year, there remains a need for greater consistency by agencies in how they interpret the Guidelines regarding annual fraud reporting to the AIC. This could be achieved through publishing explanatory notes and illustrative examples for agencies to use when deciding which matters to report each year.

Training within Australian Government agencies

As agencies are responsible for investigating the bulk of fraud allegations themselves, rather than referring them to the AFP, it is imperative that the quality of training provided to the fraud personnel is adequate. In 2009-10, agencies indicated that 99 percent of external fraud incidents were investigated by agencies themselves (659,899 of 665,846 incidents where information on investigation was provided), and 97 percent of internal fraud incidents % (2,553 of 2,633 incidents where information on investigation was provided. While these differences may be attributed to the differences between the two fraud types, it does demonstrate the importance of the investigation process. The large number of fraud investigations carried out by agencies increases the importance of agencies having a clear understanding of the nature and extent of fraud affecting Australian Government agencies. These agencies need to be funded adequately and provided with the resources necessary to prevent and control fraud as it affects them. As noted above, 34 agencies completed the open-ended questions pertaining to training for fraud personnel in the 2009–10 survey, and 32% (n=11) expressed a desire for fraud personnel to have enhanced training opportunities, such as tertiary qualifications in fraud control or investigations, as long as it was uniform across the APS agencies involved. Currently, the Certificate IV in Government (Investigations) is recommended for personnel in fraud sections. However, the quality of the information within the course differs between training institutions, which can lead to non-uniform practices between agencies.

Implications for future monitoring of Commonwealth fraud trends

To address these developments and improve the Commonwealth’s fraud control arrangements, the Australian Government Attorney-General’s Department undertook a review of the Guidelines. Following the review, completed in 2011, a number of improvements have been achieved and greater consistency in reporting should be apparent under the new regime. In May 2010 the ANAO also released a performance audit of the fraud control arrangements of Australian Government agencies (ANAO 2010). Both these activities have provided an opportunity for refining and improving federal fraud control arrangements.

The ANAO found that 97 percent of respondent agencies reported that they use the definition of fraud as specified in the Guidelines (ANAO 2010). This represents a marked improvement since 2002, when the ANAO reported that only 50 percent of agencies were defining fraud as specified in the Guidelines (ANAO 2010).

However, recalling the ANAO findings presented earlier in this paper, the ANAO echoed AIC findings that an inherent issue in collecting comparable data relating to fraud relates to the inconsistent use of fraud definitions—‘The integrity of such trend information is contingent upon common definitions for fraud’ (ANAO 2010:14).

As was evident in the previous year’s fraud report, the results for 2009–10 provide an increased level of policy-relevant information and more analysis than was available in previous years. Including specific questions on fraud types and methods has allowed for a better understanding of the types of fraud affecting agencies and resulting in loss of revenue for the government. However, this survey has been, and will continue to be, limited owing to the manner in which the Guidelines are expressed and the definition of fraud adopted. Although the ANAO report (2010) found that agencies generally defined fraud in accordance with the definition provided in the Guidelines, there is no available evidence to indicate whether agencies adopt the same interpretation of fraud as each other. The way agencies define fraud and interpret the guidelines has a direct impact on what they then report in the survey and how that information can then be used for preventive action and policy reform. Further qualitative investigations are needed to explore this aspect in greater depth, to determine how uniform the definition of fraud is across agencies and how differences may affect the nature and extent of fraud reported.

Future research could further explore the problem of external fraud, which accounts for by far the largest proportion of fraud detected by agencies, particularly the largest agencies. It would be useful to explore why these large agencies have apparently good levels of protection against internal fraud and yet continue to remain vulnerable to external fraud risks.

Other areas of external fraud risks that future surveys could address include the need for further categorisation of methods of fraud. The 2009–10 report demonstrated that agencies tend to overuse the ‘other’ categories when given the option. For example, when examining the methods used in external fraud incidents, 608,704 incidents, reported by 40 agencies, were classified as involving ‘other’ methods of commission. Ways of enabling agencies to provide more specific information to deal with this problem will be investigated for use in future surveys. It is important to note, however, that it is only with greater understanding of how fraud incidents occur that preventive measures can be devised and implemented.

A further area of research for future reports includes the determination of the profile of fraudsters offending against the Commonwealth. The revised 2010–11 survey instrument will include a new section asking respondents to provide information about the ‘most serious fraud’ case of internal fraud that they experienced during the preceding financial year. It is hoped that this information can lead to an enhanced understanding of the types of fraud affecting agencies, how employees are involved in fraud and what preventive actions can be taken to reduce the risk of fraud incidents taking place in Australian Government agencies.

Fraud remains a problem for all public sector bodies and, because the Australian Government deals in such large amounts of revenue collection and payment of benefits and refunds, opportunities for dishonesty will always be present. Knowing with certainty where fraud takes place and how individuals perpetrate acts will provide an evidence base that can be used by all agencies to design appropriate risk-reduction measures. If successful, these could lead to considerable savings for the Australian Government, arguably far in excess of the amounts required to implement the control measures in question. It is through the regular collection of data on agencies’ experience of fraud that the effectiveness of fraud-reduction activities can be measured and adjusted to maximise loss reduction across the sector as a whole.