Australian Institute of Criminology

Skip to content

Dealing with fraud

Fraud against the Commonwealth can come to light in a number of ways. On occasions, it will be disclosed through regular fraud risk management activities, while on other occasions it may be discovered inadvertently when a staff member leaves their employment or takes leave, allowing accounting inconsistencies to be discovered. Other staff, or external parties, may also report their suspicions, sometime anonymously, or through hotlines available in the workplace, or externally. The present study sought to explore how fraud was detected within entities and how fraud involving members of the public was identified. The aim was to understand how entities dealt with incidents through the processes of investigation, and referral to police and prosecution agencies. Data came not only from the annual census undertaken by the AIC, but also from reports provided to the AIC by the AFP and CDPP, which outlined their fraud-related investigations and prosecutions undertaken each year.

How fraud was detected

Respondents were asked in indicate the primary method by why incidents of internal and external fraud were detected. Nine categories were provided along with options for unknown and other methods of detection. The results for each year are shown in Figure 18 for incidents of internal fraud and Figure 19 for external fraud. Generally, internal controls were responsible for the highest proportion of detections of internal fraud, although the number of detections of this nature declined between 2010-11 and 2012-13. A large number of detections using internal controls were reported by two large entities in 2010-11, mostly due to the use of data analytics and data mining activities. Detected internal fraud by these entities declined in subsequent years. The large number of detections of internal fraud using ‘other’ methods in 2010-11 were due largely to whistleblowers who made reports to one large entity in that year.

Table 13 Methods of detecting internal fraud and incidents, 2010–11 to 2012–13 (N)
2010–11 2011–12 2012–13
Credit card issuer 2 11 2
Media 15 1 2
External audit n.p. 9 27
Offender self-reported 36 17 27
Not recorded/unknown 206 47 3
Notification by police 190 36 36
External whistleblower n.p. 138 144
Internal anon whistleblower 107 181 70
Other 760 38 23
Staff member 609 855 507
Internal controls 1,663 958 862

n.p. - Category not provided

Source: Commonwealth fraud surveys 2010–11, 2011–12 and 2012–13 [AIC computer file]

In the case of external fraud, internal controls and other types of detections accounted for the largest number of incidents. Detections by other staff and internal anonymous whistleblowers, while prevalent in the case of internal fraud, were less so for external fraud. Indeed, there was a dramatic decline in the number of staff member detections of external fraud from 18,196 in 2011-12 to 2,019 in 2012-13 and a large decline in external whistleblower detections of external fraud from 25,299 in 2010-11 to 1,729 in 2012-13. This may be explained, in part, by the large number of responses in ‘other’ methods of detection of external fraud which include detection through the use of data mining and data analytics, and anonymous tip-offs from the public.

Table 14 Methods of detecting external fraud and incidents, 2010–11 to 2012–13 (N)
2010-11 2011-12 2012-13
Credit card issuer 0 0 17
Media 20 5 5
External audit 116 19 7
Offender self-reported 4,934 8,342 2
Not recorded/unknown 3,142 949 36
Notification by police 348 86 60
External whistleblower 25,299 19,266 1,729
Internal anon whistleblower 17 19 27
Other 101,676 65,198 51,803
Staff member 17,304 18,196 2,019
Internal controls/audits 31,792 79,436 34,544

Source: Commonwealth fraud surveys 2010–11, 2011–12 and 2012–13 [AIC computer file]

Detection of fraud was also facilitated through the work of staff in dedicated fraud control sections who undertake fraud risk analysis and who investigate anomalies in accounting and other procedures. Of course, such sections are more likely to be present in the larger entities that experience more incidents of suspected fraud. Table 15 shows that over the three years, entities with a dedicated fraud control section were more likely to detect fraud incidents than those without.

Table 15 Detection of fraud by presence of a fraud control section (number of entities)
Presence of a fraud control section Internal fraud External fraud Any fraud incident
2010–11 2011–12 2012–13 2010–11 2011–12 2012–13 2010–11 2011–12 2012–13
Yes 31 29 28 32 29 29 39 38 35
No 17 15 17 10 17 19 22 29 29

Source: Fraud against the Commonwealth surveys 2010–11 to 2012–13 [AIC computer file]

When asked to describe the measures implemented that have made a difference in the detection of fraud, respondents mentioned increased staff awareness and training as being beneficial in improving fraud detection. In 2010–11 and 2011–12, two entities also referred to the use of CCTV to monitor staff and property as assisting in the detection of fraud. In addition, the use of monitoring and review activities including data mining and data analytics by the largest Commonwealth entities also enabled large numbers of suspected incidents of fraud to be identified, which could then be further investigated (see Box 1).

Box 1 Detection methods used

The use of data mining, data analytics and forensics to detect fraud incidents was evident in all three years of survey data. The Australian National Audit Office (2011) described these techniques such:

Active fraud detection measures are controls or activities that require the assertive involvement of management.

These measures can be broadly categorised as:

monitoring and review activities, focused on employees and customers at risk; and

data mining and/or data matching (ANAO 2011: 56).

In 2010–11, one entity advised that changes within their entity that had led to a difference in the way they detected fraud was due to an ‘increased level of systemic data mining and analysis of key corporate databases, to identify instances of fraud’. In 2011–12, an entity noted they had noted ‘more use of forensic scripts’, while another entity advised that their ‘continued adoption of an intelligence-led approach to detection of fraud, for e.g. full implementation of the post payment monitoring and analysis program’ was a factor that made a difference in 2011–12 to their fraud detection. In 2012–13, an entity noted that they had introduced ‘data mining/data analytics capability over financial data to identify potential results which require further investigation and expansion of the Ethics Hotline to external parties’, which made a difference to their fraud detection in that financial year.

Source: Commonwealth fraud surveys 2010–11, 2011–12 and 2012–13

How fraud was investigated

Internal fraud control staff

In order to assess the resources allocated to fraud control within the Commonwealth, respondents were asked to specify the number of staff dedicated to various fraud control functions. Information was gathered on the number of staff working in fraud prevention, investigation and other fraud control tasks, and whether or not they possessed formal qualifications such as Certificates or Diplomas related to fraud investigation. Results are shown in Table 16.

Table 16 Fraud control staff and their qualifications, 2010–11 to 2012–13
Category Fraud prevention Fraud investigation Other functions Totala
2010–11 2011–12 2012–13 2010–11 2011–12 2012–13 2010–11 2011–12 2012–13 2010–11 2011–12 2012–13
Staff in fraud section (N) 648 641 843 1,165 1,973 2,089 1,284 422 228 3,097 3,036 3,160
Staff with a qualification (N) 108 93 125 945 1,124 1,133 119 62 161 1,172 1,279 1,419
Fraud section staff with a qualification (%) 17 15 15 81 57 54 9 15 71 38 42 45

a: Note that some staff may have been engaged in more than one type of function resulting in totals being in excess of actual staff numbers

Source: Commonwealth fraud surveys 2010–11, 2011–12 and 2012–13 [AIC computer file]

There were substantial changes to fraud control staffing levels and those with qualifications in fraud control over the three year period, although these changes tended to arise in two large entities with substantial fraud control responsibilities. Between 2010-11 and 2012-13 there were large increases in the number of staff employed in both prevention and investigation functions, by 30 percent and by 80 percent, respectively . During the same period, the number of fraud control staff engaged in other duties, such as policy-related functions declined by 82 percent. By way of contrast, however, the percentage of staff with formal qualifications working in both fraud prevention and in investigations declined, while the percentage of staff with formal qualifications working on other functions increased substantially from 9 percent in 2010-11 to 71 percent in 2012-13. Overall, it is apparent that the percentage of staff with formal qualifications increased by 18 percent over the three years, but that the total number of fraud control staff remained relatively constant from year to year. It seems that some large entities reassigned staff to work in investigations, as opposed to prevention and other functions, although further research would be needed to confirm this.

Who conducted investigations

The 2011 Commonwealth Fraud Control Guidelines state that ‘entities are responsible for investigating routine or minor instances of fraud, including investigating disciplinary matters’ (AGD 2011: 16). The present results proved to be the case in respect of both internal fraud and external fraud. Out of the 2,914 internal fraud incidents investigated or reviewed in 2010–11, 2,751 (94%) were undertaken internally by the entity that detected the alleged fraud incident. Similarly, high percentages of suspected internal fraud incidents were investigated or reviewed internally in 2011–12 (97%) and in 2012–13 (93%).

Respondents were also asked to indicate the number of incidents of alleged internal fraud that were investigated or reviewed externally, once an internal investigation or review had been completed. Figure 17 shows the number of incidents investigated primarily by various external organisations. In 2010–11, by far the largest number of external investigations, following internal entity review, were undertaken by the AFP or external investigators. However, in 2011–12 and 2012–13, the AFP were the external investigator employed for only very small numbers of alleged incidents of internal fraud. Since 2010–11, there appears to have been an increase in use by entities of external investigators other than the AFP or state and territory police for incidents of internal fraud. In each year, there were no external investigations undertaken by media organisations into allegations of internal fraud.

Figure 17 Alleged internal fraud by external investigating organisation, 2010–11 to 2012–13 (number of incidents)

mr24_figure17.ai

Note: These figures only relate to incidents investigated externally following an internal review by entities. In 2010–11, an additional 1,589 incidents of alleged internal fraud were reviewed by entities themselves; 2,101 in 2011–12 and 2,751 in 2012–13

Source: Commonwealth fraud surveys 2010–11, 2011–12 and 2012–13 [AIC computer file]

As with internal fraud investigations, the vast majority of external fraud incidents were primarily investigated by entities themselves, accounting for over 99 percent of alleged external fraud incidents in 2010–11 and 2011–12, and 97 percent in 2012–13.

Figure 18 shows the number of external fraud incidents investigated primarily by various external organisations following internal review by the entities themselves. The AFP and state and territory police appear to have undertaken investigations in increasing numbers for external fraud matters over the three years. Other organisations were involved in the investigation of only a few incidents of external fraud each year. In each year, there were no external investigations undertaken by media organisations into allegations of external fraud.

Figure 18 Alleged external fraud by external investigating organisation, 2010–11 to 2012–13 (number of incidents)

mr24_figure18.ai

Note: These figures only relate to incidents investigated externally following an internal review by entities. In 2010–11, an additional 82,751 incidents of alleged external fraud were reviewed by entities themselves; 178,781 in 2011–12 and 37,449 in 2012–13

Source: Commonwealth fraud surveys 2010–11, 2011–12 and 2012–13 [AIC computer file]

Referrals for criminal action

Respondents were asked to indicate how many and what type of incidents were referred directly to various specified organisations for criminal investigation or prosecution, including the AFP, state or territory police, the CDPP and other organisations in Australia or overseas. These were incidents that had not been dealt with by entities themselves and accordingly, represented more serious matters.

Table 17 shows the number of internal fraud incidents referred by entities for criminal action for each fraud focus category. In 2010–11, the option of ‘other organisation’ was not provided. Over the three years, 297 incidents were referred for criminal action, representing only four percent of all internal fraud incidents detected. Incidents relating to financial benefit fraud were most often referred for criminal action.

Table 17 Internal fraud incidents referred for criminal action by focus category, 2010–11 to 2012–13 (N)
Focus AFP State and territory police CDPP Other organisations
2010–11 2011–12 2012–13 2010–11 2011–12 2012–13 2010–11 2011–12 2012–13 2010–11 2011–12 2012–13
Equipment 13 0 3 0 4 0 0 2 2 n.p. 2 0
Entitlements 16 1 28 0 0 0 0 3 3 n.p. 0 0
Information 29 3 3 1 0 0 2 4 2 n.p. 10 7
Financial benefits 4 10 7 16 6 0 35 34 13 n.p. 13 4
Other 2 0 1 0 0 2 0 1 4 n.p. 7 13

n.p. - Category not provided

Source: Commonwealth fraud surveys 2010–11, 2011–12 and 2012–13 [AIC computer file]

In the case of external fraud, Table 18 shows that the largest number of incidents referred for criminal action focused on ‘entitlements’. Over the three years, 5,344 external fraud incidents were referred for criminal action, representing only two percent of all external fraud incidents detected.

Table 18 External fraud incidents referred for criminal action by focus category, 2010–11 to 2012–13 (N)
Focus AFP State and territory police CDPP Other organisations
2010–11 2011–12 2012–13 2010–11 2011–12 2012–13 2010–11 2011–12 2012–13 2010–11 2011–12 2012–13
Equipment 13 4 1 29 18 17 0 0 0 n.p. 15 6
Entitlements 8 3 1 6 0 0 1,328 1,307 1,217 n.p. 3 1
Information 1 18 5 7 2 1 226 192 200 n.p. 6 4
Financial benefits 11 9 9 15 19 92 111 44 248 n.p. 10 24
Other 19 4 0 1 1 0 30 0 36 n.p. 2 20

n.p. - Category not provided

Source: Commonwealth fraud surveys 2010–11, 2011–12 and 2012–13 [AIC computer file]

Australian Federal Police investigations

In addition to the information provided by respondents to the annual data collection, both the AFP and the CDPP provided data to the AIC concerning the investigation and prosecution of fraud against the Commonwealth. These data were provided in addition to information regarding their own experience of internal and external fraud and fraud control arrangements. The data were received in aggregate format and could not be attributed to any individual incidents reported by specific entities. Nor could the type of fraud investigated or prosecuted be categorised as internal fraud incidents or external fraud. Data collection practices within the AFP and CDPP also differed in various respects as described below.

Paragraph 12.5 of the 2011 Guidelines required the AFP to provide the AIC with the following information each year relating to fraud-related matters against the Commonwealth referred to, accepted or declined by the AFP during the previous financial year. Further information requested of the AFP consisted of:

  • the type of offence;
  • estimated financial loss investigated;
  • fulfilment of AFP service standards relating to case handling; and
  • results of the investigation quality assurance review process, with an analysis of best practice and deficiencies.

The Guidelines provide a list of matters that are considered to be of sufficient seriousness to warrant referral to the AFP:

  • significant or potentially significant monetary or property loss to the Commonwealth;
  • damage to the security, standing or integrity of the Commonwealth or an entity;
  • harm to the economy, national security, resources, assets, environment or wellbeing of Australia;
  • a serious breach of trust by a Commonwealth employee or contractor of an entity;
  • the use of sophisticated techniques or technology to avoid detection that require specialised skills and technology for the matter to be successfully investigated;
  • the elements of a criminal conspiracy;
  • bribery, corruption or attempted bribery or corruption of a Commonwealth employee or contractor to an entity;
  • known or suspected criminal activity against more than one entity;
  • activities that could affect wider aspects of Commonwealth law enforcement (eg illegal immigration, money laundering); and
  • politically sensitive matters.

Figure 19 shows the number of fraud matters accepted and rejected by the AFP each year. Between 2010–11 and 2012–13, there was an 18 percent increase in matters referred and a 39 percent decrease in matters declined for investigation. In addition, at the time of reporting to the AIC, a further two matters in 2011–12 and eight matters in 2012–13 were still under evaluation. Over the three year period, 73 investigations resulted in a criminal conviction. Of these, there were 29 in 2010–11, 20 in 2011–12 and 24 in 2012–13. Some of those investigations had been received by the AFP in previous financial years.

Figure 19 Fraud matters accepted and rejected by the AFP, 2010–11 to 2012–13 (N)

mr24_figure19.ai

Source: AFP internal data supplied to the AIC 2010–11, 2011–12 and 2012–13 [AIC computer file]

The AFP estimated that the financial loss attributable to the 61 matters accepted for investigation in 2010-11 was $12,796,207 , which relates to the initial property value as reported by the referring agency. In 2011-12, the estimated loss for the 65 accepted matters was $120,796,393 , and in 2012-13 the estimated loss for the 72 accepted matters was $102,426,346. As at 30 June 2013, the AFP had 166 fraud-related matters still on hand that had an estimated loss value of $1,053,185,683 . These estimates are based on the initial property value as reported by the referring entity, and may be subject to change following investigation.

Fraud prosecutions by the CDPP

Statistics on Commonwealth fraud cases referred to the CDPP for prosecution and the outcomes of those cases, are provided to the AIC each year pursuant to the paragraph 12.6 of the 2011 Commonwealth Fraud Control Guidelines. For each financial year, individual state and territory statistics were provided on:

  • the number of fraud type matters referred to the CDPP;
  • the number of defendants and charges prosecuted;
  • the amount initially charged in each fraud type prosecution;
  • the outcomes of prosecutions including:
    • the number of convictions;
    • the number of acquittals;
    • the number of other outcomes;
    • amounts ordered by courts by way of reparation orders under the Crimes Act 1914 and pecuniary penalty orders under the Proceeds of Crime Act 1987.
  • the number of charges by offence;
  • the number of charges by referring entity; and
  • the number of proven offences by highest sentencing disposition.

Figure 20 shows changes in these statistics for the three financial years. It should be noted that statistics relate to matters referred or dealt with in the financial year in question, some of which may have originated in earlier years. Thus, the convictions do not necessarily relate to the same matters referred that year, as often prosecutions take more than 12 months to be concluded. Generally, over the three years, there has been a decline in the number of fraud matters referred, prosecuted and convictions obtained. This is largely due to the caseload referred to the CDPP changing, as well as the nature of the individual cases involved.

Figure 20 Fraud matters prosecuted by the CDPP, 2010–11 to 2012–13 (N)

mr24_figure20.ai

Source: AFP internal data supplied to the AIC 2010–11, 2011–12 and 2012–13 [AIC computer file]

Further statistics on the CDPP’s fraud caseload for each state and territory are shown in Table 19. The largest number of fraud cases came from the most populous states of New South Wales, Victoria and Queensland, with numbers in all jurisdictions declining generally over the three years. This follows national statistical trends in recorded dishonesty offences, which have also shown a decline in recent years (ABS 2011–13).

Table 19 Prosecution of fraud by jurisdiction, 2010–13 (N)
Year Referrals Defendants prosecuted Convictions Acquittals Charges prosecuted
New South Wales
2010–11 583 1,064 828 5 3,804
2011–12 511 454 375 7 1,770
2012–13 501 443 342 4 2,165
Victoria
2010–11 397 587 528 0 1,369
2011–12 401 290 249 0 821
2012–13 311 323 245 1 692
Queensland
2010–11 341 673 593 3 1,499
2011–12 267 364 275 1 1,108
2012–13 321 296 241 2 671
South Australia
2010–11 118 291 235 0 1,783
2011–12 150 147 107 0 695
2012–13 76 120 103 0 644
Western Australia
2010–11 138 232 191 1 515
2011–12 104 188 156 1 465
2012–13 111 111 66 1 313
Tasmania
2010–11 76 110 92 2 635
2011–12 52 52 46 0 533
2012–13 48 59 42 0 527
Northern Territory
2010–11 42 48 21 0 175
2011–12 10 41 12 0 115
2012–13 22 19 9 0 106
Australian Capital Territory
2010–11 51 147 127 1 314
2011–12 22 63 48 2 97
2012–13 32 24 14 0 115
Total 2010–11 1,746 3,152 2,615 12 10,094
Total 2011–12 1,517 1,599 1,268 11 5,604
Total 2012–13 1,422 1,395 1,062 8 5,233

Source: CDPP internal data provided to AIC in 2011, 2012 and 2013

In 2010–11, the total amount initially charged in fraud-like prosecutions was $77,960,259, decreasing to $29,692,788 in 2011–12 and increasing to $36,651,470 in 2012–13. These changes reflect the number of defendants prosecuted, as well as the nature of charges alleged that can vary considerably over time. Not all Commonwealth fraud offences are prosecuted by the CDPP, as some may be heard in state and territory courts, particularly where state and territory offences are alleged.

Sentencing dispositions

Each year, the CDPP also provides statistics on the highest sentencing disposition given for proven fraud-related offences. Again, the data relate to the years in which defendants were sentenced, rather than the year in which the CDPP received referrals from entities. Each year, the most frequently imposed sentence was a recognisance order, followed by a fully suspended term of imprisonment. There appears to have been a gradual increase in full-time custodial orders over the three years, with 12.5 percent of the highest sentencing disposition for proven offences being awarded in 2012–13.

Figure 21 Highest sentencing disposition for proven fraud offences, 2010–11 to 2012–13 (%)

figure 21

Source: CDPP internal data provided to AIC in 2011, 2012 and 2013 [AIC computer file]