Corrupt procurement: rethinking the roles of principals and agents

Abstract Public sector procurement is a multi billion dollar activity and requires the highest level of integrity. Even in countries that are relatively free of corruption, there exist many opportunities for corrupt behavior, and there are always individuals who seek out these opportunities. This paper uses a principal/agent analysis to note that when agents behave corruptly in procurement activities there is usually a clear remedy, but when principals behave corruptly the path forward is not clear at all, and leaves government quite vulnerable to loss and mistrust. This analysis of corrupt procurement through the lens of principals and agents examines some Australian cases but the themes are applicable across many countries.


Introduction
Corruption in procurement occurs in both rich countries and poor countries but the manifestations vary in prevalence and subtlety. In rich countries we generally do not see direct payments to politicians or officials to do things or not do things. We do not see kleptocrats looting the treasury and stashing their gains in offshore bank accounts. We do not see citizens denied a service that is their right, unless they pay a corrupt civil servant a consideration. We do not see officials buy a place in a bureaucracy, and then use that to both extort citizens and repay superiors. This paper argues that where corruption is a transgression that is outside the everyday expectations of citizens, then building cultures of integrity, making improvements in bureaucratic processes and compliance and enforcement do have a positive impact on lessening the risk of corruption. Critics argue whether developing better processes really has an effect. Some (Persson, Rothstein, and Teorell 2019) argue that this has no impact on corruption control and that collective action is the only way forward. This is a key difference between rich country corruption and poor country corruption.
In OECD countries public procurement comprises about 12 per cent of GDP and 29 per cent of general government expenditure. Of government procurement in OECD countries on average one third of procurement is by central government, two thirds by sub-national government (OECD 2016, 5). In 2014 The European Commission estimated that in Europe about e120 billion is lost to procurement corruption each year (European Commission 2014). Transparency International estimates that of the US$9.5 trillion spent globally on government procurement about $2 trillion "disappears" (Transparency International 2014, 4 and 8).
Writing about corruption in procurement more than a decade ago Ware et al. (2011) observed that the African Union estimates that approximately one quarter (or US$148 billion) of Africa's gross domestic product is 'lost to corruption each year'. In Latin America, the Inter-American Development Bank (IADB) estimates that about 10% of GDP is on average lost to corruption annually; and the Asian Development Bank (ADB) reports that in one Asian country, approximately US$48B may have been lost to corruption over a twenty year period.
We are dealing with a widespread global phenomenon, though measuring corruption is notoriously difficult (Heinrich and Hodess 2011). It does not appear in crime statistics, survey work does not measure incidence or prevalence of corruption, and most often when it occurs, it is a consensual activity between the parties, neither of whom is willing to disclose it.

Theoretical issues
The corruption literature is replete with definitions which will not be rehearsed here, but there is a theoretical issue in placing corruption in scope. Is it a transgression by an official, an official who fails to implement policy and seeks personal benefit through the transgression? Or is it something that is more rotten to the core where the policies are corrupt and favor powerful individuals, or their cronies or supporters, or their political parties?
The debate becomes one of "administration" versus "politics," and the principal/ agent relationship involves the rational choice of politicians and officials. Is corruption caused by poor systems and exploitable loopholes, built upon low civil service pay and/ or sheer greed (Bauhr 2017). Or is it something that is more about the nature of political systems, and is the contestation over things like democratic processes, transparent transactions, and genuine representation of human rights (Mungiu-Pippidi 2020). If there are transgressions, is this a breach of the principal/agent relationship, or is it something more politically systemic?
In the simplest terms the principal, the politician or minister, delegates to the agent the task of implementing the principal's policies, and the agent delivers as required. As Marquette and Peiffer observe (2015) there is a problem where the agent and principal have divergent interests or if the agent has more information than the principal. This is compounded when the public is unable to monitor or hold public officials or politicians accountable. Klitgaard (1988) sees problems when there are imbalances in accountability, discretion and monopoly of services.
In rich countries, if there is a corrupt exchange the assumption generally is that the principal is not corrupt, but the agent may be. This then becomes a matter for public administration or the criminal justice system. However there are occasions when the principal is corrupt, or if not corrupt, insufficiently accountable, and makes decisions outside the realm of evidence and transparency. There is also another element to the debate, and that is that in a democratic society, the 'real' agent is the politician, and the principal is the electorate; and in this sense, the elected politician is there to implement the notional will of the electorate. Philosophers and political scientists have been debating this point at length.
Another way of looking at the theory is to conceptualize it as a supply side/demand side approach to corruption. Anti-corruption initiatives to limit the supply of corruption (or opportunities for corruption) focus on reforms of processes and better administration as well as greater transparency, reductions in discretion and harsher penalties for offenders. Demand side initiatives, on the other hand, are more politically oriented and aim to empower citizens and civil society to scrutinize government and make public decision-making more transparent and accountable (Peiffer and Alvarez 2014). In short, supply side initiatives focus on agents, demand side on principals. Paul Heywood's analysis teases out the arguments between those who see corruption as a structural, principal-agent problem and those who see it as a political system problem or democratization problem (Heywood 2017, 22-25). One way forward is to identify red flags that alert observers to corruption in procurement. Red flags are discussed in some detail below, and it is noted that they are more apparent when dealing with agents than when dealing with principals.

Procurement context
Despite Australia being a rich country that ranks well in the Transparency International Corruption Perception Index, corruption in procurement does exist. Australia is ranked 18 th (least corrupt) out of 180, an enviable position, but of concern is that it has slipped from 7 th in 2012 (Transparency International 2022). One possible explanation for the fall is that corruption is not being called out in the political arena.
In Australia the Federal government spent about $300 billion on procurement between 2014 and 2019. If expenditure is similar to OECD patterns, sub national governments would have spent an additional $600 billion in procurement.
Corruption always follows opportunity. To minimize harm and to put in place preventive measures it is important to understand the concept, and understand how corruption plays out in different settings (Graycar 2015). We need to understand how corruption varies in different settings, the component parts that make up a corrupt event or transaction, and we need to understand how to prevent corruption through minimizing opportunities that may arise (Graycar and Prenzler 2013).
Procurement provides many opportunities for corruption. It ranges from minor padding of invoices to manipulation of major contracts. High levels of discretion by decision makers can provide opportunities for kickback brokers, bid rigging or the use of front or shelf companies. The opportunity structure for corruption plays out differently when exploited by agents (Graycar and Sidebottom 2012) or principals (Praino and Graycar 2018).
In Australia there are strict standards and strict guidelines about public procurement, and yet there are examples of corruption that come to light. When it occurs it is more an exception than something systemic, yet there are enough examples to cause concern. Two things come together, opportunities exist, and there is somebody on the make. The challenge is to build and maintain cultures of integrity without having a horrendous amount of red tape that ties up every procurement transaction.
There is a grey area when examining conflict of interest, or cutting corners to speed things up, or going to known and reliable suppliers who might be friends. However there is no ambiguity about taking secret commissions or bribery. This is plainly criminal. But corruption is bigger than bribery.
Different approaches to corruption in procurement come into play if we focus on a rich country/poor country spectrum, or on a principal/agent divide. Rich country corruption is very different to poor country corruption. Different syndromes are in play (Johnston 2005) and the differences have been explored in the literature. (Andersson 2008;Graycar and Monaghan 2015).
In Australia it is reasonable to assume that corruption in procurement is not the norm, and when cases come to the surface these are usually exceptions. The transgressions are likely to be episodic and opportunistic rather than systemic. There are legal and managerial mechanisms to deal with these. When the principal (here, the political leadership) is involved the situation is more difficult because they usually fall beyond the scope of legal and managerial rectification (Ng 2021).
Corruption in procurement can be classified in terms of what it is that is being bought. In general, procurement can be classified into five types. These are not expanded here, but description and analysis can be found elsewhere (Graycar 2019). They are listed here because each is built on and creates different opportunities for corruption in which principals and agents can play a role.

Standard Customized Intangible Complex Unique
There is one form of corruption in procurement where neither principal nor agent is involved. This is in cartel behavior where, rather than compete with each other, businesses collude to" fix" a procurement process. Various techniques are used. Some of these include: price fixing where collusion occurs to set a higher price for all bids and perhaps share excess profits, rigging bids, where suppliers decide on who will submit the lowest bid and share future opportunities, or segmenting or sharing markets where competition is limited and collaborators do not stray onto each other's turf.
In Australia the Australian Competition and Consumer Commission investigates cartel behavior (ACCC, n.d.). Generally with cartel behavior no public official is involved. However if either a principal or an agent colludes with a cartel or knowingly looks the other way, this is clearly a corrupt transaction and would come within the scope of an anti-corruption agency. Cartel behavior is not discussed further in this paper.
Australia is a federal system, and each state has an anti-corruption agency. They vary in their scope and powers, but corruption in procurement is certainly on their radar. There is a federal anti-corruption body whose scope is narrowly limited to law enforcement agencies. But the federal government does not have a general anti-corruption agency (ACA), and the conservative coalition government which was in power until May 2022 had resisted pressure to establish one (Brown 2021); however, the new Labor government elected in 2022 has promised an ACA by the end of the year.
In Australia we have witnessed some examples in recent times of both principals and agents engaging in dubious procurement activity, activity that has almost universally been labeled corrupt, but which in many instances has not breached any specific laws because the activities were being steered by government ministers. Some leading examples of both agent and principal mis-behavior are described below.

Agents behaving badly
Exploring corruption in procurement a research paper (Graycar 2019) examined all of the procurement cases which went to a public hearing over a 30 year period in one state ACA in Australia (the New South Wales Independent Commission against Corruption) . It explored how these can be identified and classified, and how interventions might be structured. What is notable is their "ordinariness." Some of these cases are described below. The full reports of each case can be found at https://www.icac. nsw.gov.au and then typing in the name of the operation.

Operation siren
Employees of the Sydney Water Corporation (SWC) sought and/or received corrupt payments or rewards from numerous accredited contractors in relation to the performance of their SWC duties. They violated rules in overseeing the work of contractors providing services to SWC and authorizing invoices for services (totaling $293,000) without the appropriate authority, knowing certain invoices contained false information.

Operation barcoo
A Project Manager with the NSW Department of Education and Training (DET) received corrupt benefits valued at about $437,000 by improperly causing DET officers to appoint persons employed by or otherwise connected with his own private company from which purchases were made. He took the profits.

Operation kanda
A cleaning and site manager of the Sydney University Campus Infrastructure Services (CIS) received 267 payments (totaling over $350,000) by manipulating procurement processes that favorably granted cleaning contracts to a company jointly owned by her and her husband, and by concealing from the University as her employer, her relationship with this private company.

Operation aztec
An officer in public housing was responsible for the disposal of properties. The process involved realtors selling the properties on behalf of the department, and getting the best price for them. The officer chose two realtors who were friends to sell five properties. They were sold at greatly reduced prices, to his relatives or to companies owned by the official, and then were re-sold at true market value (about 40% more). The profits were split. He also received a number of non-cash benefits. The official concealed conflicts of interest, numerous frauds and the receipt of secret profits.

Operation torrens
A council officer favored a contractor by forging quotes and by arranging for another contractor in the air-conditioning industry to submit dummy quotes to falsely demonstrate that he had followed council procurement procedures.

Operation monto
Within the state railways, the Commission investigated allegations of fraud, bribery, improper allocation of contracts, submitting false and misleading invoices for goods and services not delivered unauthorized secondary employment, failure to declare conflicts of interest, falsification of time sheets, and the cover-up of a safety breach. In financial terms RailCorp employees were found to have improperly allocated contracts totaling almost $19 million to companies owned by themselves, their friends or their families, in return for corrupt payments totaling over $2.5 million.

Operation vika
A NSW Rural Fire Service (RFS) contracts officer received corrupt payments totaling $408,382 over a period of three years in return of favors to the owner of catering companies dealing with RFS business. This included concealing the undersupply of snack packs, facilitating the payment of false invoices, and raising invoices and paying for RFS purchase orders for goods that were never delivered to RFS.
Operation yancey A Department of Justice assistant director (responsible for managing of capital works) improperly and corruptly exercised his official functions in order to financially benefit his private contractor friends in awarding contracts to refurbish NSW courthouses. This assistant director was assisted by another government employee who was to financially benefit from obtaining a higher paying position within the department based on false information, constructed by both of these officials.

Operation crusader
A campus services manager at the University of New England (UNE) corruptly solicited and received benefits for himself and others (e.g. contributing to preferred fundraising functions, and receiving free hospitality, gifts and cash) in return for improperly favoring private companies in the awarding of UNE contracts and for approving payment of false invoices submitted by those companies.
From another jurisdiction: Operation dunham (IBAC) (see IBAC 2017 in references) A senior official of a state Department of Education proposed a new learning system initially budgeted for $60.5 million. With dogged determination, he ensured that two companies with whom he had a long-standing relationship succeeded at all costs. He manipulated departmental procurement processes to create the corrupt tendering process that led to the two companies being awarded the contract to develop the system. The budget blew out substantially. He was accused of abusing the standards of conduct expected of a public servant, manipulating departmental procurement processes, and deceiving and bullying anyone who stood in his way.
In three quarters of the cases studied the initiator of the corrupt activity was a civil servant; and in the remaining cases, somebody external to the departmenta company or a private person. Three quarters of the civil service perpetrators were acting alone, as a "lone wolf," and one quarter in conjunction with other colleagues in the department. Three quarters (76%) of the cases involved a senior public servant only, one sixth (16%) had senior and junior public servants involved together, while the remainder was unknown. Two thirds of the cases fell within three sectors: Transport, 26%; Education, 19%; and Local government 17%. One quarter of the cases involved monetary losses of between $1 m and $5 m; 40% between $100,000 and $1million, while 16% involved less than $100,000. Almost one third of the cases involved no clear bribe or kickback to the public official. These involved steering contracts and work to the benefit of friends, associates or families. Some people became romantically involved, and misused their position to impress or benefit their prospective partner or their partner's family (Graycar 2019).
These cases, and many more, are examples of the agent acting corruptly in procurement. They are transgressions and there is no contest about their being inappropriate behavior. But they are not always necessarily illegal. In Australia every government department has rules, processes and procedures about procurement which if breached have sanctions for the perpetrators, if they are caught.
These processes go through many steps, some of which include Is procurement required? Is there a business case Where will funding come from What approvals are needed? Acquisition plan/purchase recommendation/contract preparation/contract signature What policies or legislation are relevant What risks are there Are there probity or conflict of interest issues Who writes the specification How is a supplier identified and found Evaluation methodology/supplier record Value for money/timeliness/supplier service record/human resource management/ safety record/past performance/financial viability/technical competence/quality assurance Contract management Document management Deliverables specified.
When standards are breached, and if an agent makes discretionary decisions to benefit a supplier, and if the exchange is hidden, and if the agent receives a payment, it is clearly a corrupt exchange.

Principals behaving badly
A common manifestation in many political systems is that of "pork barrelling." This involves the allocation of funds or purchase of things to gain political advantage. Is it corruption? Is it corruption in procurement? Or is it politics as usual with a bit of sleazy behavior on the side?
In some of the examples below there is a mixture of malfeasance, sleaze and incompetence. While they all demonstrate inappropriate behavior, can they all be considered as corruption in procurement?
Where grants are being allocated, or projects funded there are usually guidelines and criteria for the allocation of those funds, or for the assessment of need for the procurement. There have been several cases in Australia in recent times where there have been allegations of corruption where expert recommendations have not been followed or where things have been bought in dubious circumstances allegedly to gain political advantage. Are grants a form of procurement? They involve buying a service or contributing to a desirable state of affairsor could be examples of cynically currying favor and seeking to buy votes.
Sports grants. Governments in Australia operate many grant programs. One such national program is Sport Australia's Community Sport Infrastructure Program, which in 2019 allocated $102.5 million for local sports infrastructure. The Australian National Audit Office (ANAO, the federal government's independent auditor) published a report documenting irregularities in the grants process (Australia National Audit Office 2020a). This led to accusations that the Minister for Sport, Senator Bridget McKenzie had awarded grants to projects which would otherwise not have been successful, but for their being in marginal electorates. As decisions were not based on merit she was accused of abusing ministerial discretion in order to "pork barrel" in the run up to the 2019 Australian federal election.
The administering authority, Sports Australia, used its expertise to assess all applications and make recommendations to the Minister who in normal circumstances would sign off on the experts' recommendations. Interestingly Sports Australia, a government body which disperses public funds and procures infrastructure is not subject to the Commonwealth Grant Rules and Guidelines as it is a corporate Commonwealth entity and not subject to the same administrative controls as is a Department.
Three assessment processes took place in December 2018, February 2019, and April 2019. The 15 January 2020 ANAO report found there was a color coded spreadsheet being used by the Minister which highlighted electorates. This spreadsheet was analyzed in depth by Leigh and McAllister (2021). In the final round, 73% of the grants recommended by Sports Australia were not approved, and numerous non-recommended projects were funded.
The ANAO concluded that grant funding was "not informed by an appropriate assessment process and sound advice" and that successful applications "were not those that had been assessed in terms of the publishing program guidelines." Sports Australia was found broadly to have followed the guidelines. The Minister was found to have used a different assessment process, "using other considerations." 412 of the grants approved by the Minister did not meet Sports Australia's cut for merit, and the ANAO also found evidence of distribution bias.
There was a huge outcry in the media and severe disappointment from community groups who missed out. The Minister resigned, but not ostensibly because of the pork barreling, but rather because she had failed to declare her membership of a sports club which had successfully received a grant.
The government's response to the criticism was to have an internal review conducted by the Head of the Prime Minister's Department. This review found that political considerations did not play a part in the decision-making process by the Minister however, the report is confidential and has never been released.
The ANAO report came shortly after the government had won the 2019 election by the narrowest of margins. An academic study into the first item below concluded that there was no consistent evidence that electorates (districts) which received more sports grant funding tended to swing toward the government (Leigh and McAllister 2021).
Car parks. In 2021 the Australian National Audit Office (ANAO) released a report entitled Administration of Commuter Car Park Projects within the Urban Congestion Fund (Australia National Audit Office 2021c). Reporting on a procurement exercise of $650 million, it found that there was "no competitive selection process, no formal guidelines and no formal call for submissions." The car park fund was part of the Urban Congestion Fund, a 10 year strategy for urban improvement funded largely by the federal government, but done in conjunction with state governments. The ANAO found severe deficiencies in fund design; project identification and selection; project funding approval; and project delivery.
On 19 July 2021 a special Senate estimates committee questioned officials about the car parks (Australian Parliament 2021). The committee heard that the then Infrastructure minister had prepared spreadsheets identifying state by state the top twenty marginal seats in the country. The Deputy auditor Brian Boyd informed the Senate committee that the minister and his staff had then canvased their party's MPs and candidates in those marginal seats, looking for projects to fund.
Twenty seven of these were allocated the day before the election was called, which by convention meant that the government went into caretaker mode and was obligated to announce no new initiatives. They then became "election commitments" which means they are not subject to the same due diligence requirements.
The Australian public service is legally bound to ensure that public moneys are spent properly, according to the Public Governance, Performance and Accountability Act 2013, as well as the public servants' obligations under the Public Service Act 1999. The cornerstone of the PGPA Act is that it mandates the "proper" use of public money, which is itself defined when used in relation to use or management of public resources as efficient, effective, economical and ethical. Two years later only two car parks had been completed, and these at very high cost (Burgess 2021).
Selection criteria were inappropriate. There was no design to make the process open and transparent. There was no engagement with the relevant stakeholders (state governments and councils) which increased the risk that the projects would not deliver the desired outcomes. It was not demonstrated that the projects were selected on merit and the department did not provide an assessment of how the investment met policy objectives. "The distribution of selected projects reflected the geographic and political profile of those given the opportunity by the government to identify candidates for funding consideration" (Burgess 2021). The Senate passed a motion that these spreadsheets be tabled, but the department and the minister declined to do so.
Relationship problems. A third example involves a principal who was in an intimate personal relationship with a member of parliament who was seeking $5.5 million for a construction project for a local sporting club. The Minister (principal) did not disclose the relationship. The bureaucratic advice was that the project not be funded because it did not meet the criteria for funding, and on a score sheet it fared badly. Over a period which involved several submissions the proposal consistently failed to establish a business case. The bureaucratic advice was overruled by the senior politician and the funding went ahead (ABC News 2021) . The relationship ended, the senior politician resigned, and is currently under investigation for failing to declare a conflict of interest.
Land purchase near airport. A new airport is being built on the outskirts of Sydney, Australia. A second runway is not due to be developed until 2050. In anticipation, in 2018 the Commonwealth Government Department of Infrastructure, Transport, and Regional Development purchased for $32.8 million a 12.26 hectare parcel of land known as the Leppington Triangle. The land was valued at $3 million. This attracted the attention of the Auditor-General (ANAO) which conducted an audit (Australia National Audit Office 2021b).
This created considerable political controversy because the land was owned by two men who are high-profile donors to political parties, in particular the government party. They also have a history of litigation with the government over land purchases and would have resisted attempts at compulsory acquisition. The issue raised concerns about integrity and probity and the suspicion was that the government had been defrauded.
The ANAO reviewed and reported on the case, and to test whether the Commonwealth might have been defrauded, in an unprecedented move, referred the matter to the Federal Police before the audit was complete. The ANAO audit found that processes for managing conflicts of interests were not good. One official involved in the acquisition had declared a conflict with the seller of the land, but was allowed to continue working on a key project related to the land acquisition. Officials in the department appeared to be hiding information of interest. The possible cover up attracted discussion in the Australian Senate.
Two Inquiries were called, one into whether a staff member had failed to properly manage conflicts of interest during the Leppington deal, the other into whether an official had breached the public service code of conduct over the sale. These reports have not been released.
The Australian Federal Police found no evidence of a crime, specifically, no evidence of bribery, conspiracy to defraud, or abuse of public office. One journalist alleged a cover-up by the government (Elmas 2020). The purchase was in line with the Lands Acquisitions Act. One journalist (Keane 2021) has described this as not surprising, "given the scandal always looked far more like a case of spectacular incompetence and industry capture rather than overt corruption." Procuring offshore refugee processing and security services. In 2017 a contract for $420 million was awarded to the Paladin Group. The Department of Home Affairs was seeking somebody to run garrison support and welfare services for its offshore refugee processing sites on Nauru and Papua New Guinea from 2017 to 2019 and to run security services on Manus Island. The Department approached four companies for this work.
The key concerns raised were that Paladin did not demonstrate value for money in PNG; was a thinly capitalized, little known company with no corporate structure; and that the contract was awarded without going to open tender, thus excluding other companies which were interested in bidding. The company did not have sufficient funds to start its work and $10 million had to be advanced by the government (ABC News 2019).
Paladin Group was registered to a beach shack on an offshore island and a post office box in Singapore. The Australian Financial Review reported (18 February, 2019) on links between its two key shareholders and PNG politicians. There were inferences of political influence or lack of probity or pressure coming from the relationship with these politicians.
The ANAO (Australia National Audit Office 2020b) found that the management of the half billion dollar procurement was "largely appropriate" and "largely in accordance with the Commonwealth Procurement Rules." The department used exemptions based on human health and security grounds to exempt the tender process from the normal open tender requirements. The department did not document the reasons for seeking quotes from only the four companies that achieved the contracts, and found the department did have the time to secure other offers from other companies also interested in bidding. The process seems to have been done by the department "on the run" via letters of intent rather than properly organized contractsthese came later.
The ANAO found value for money was not achieved with the PNG contracts and there were questions around the probity framework, adequate monitoring, recordkeeping, risk assessment. It made a number of recommendations to improve transparency and probity.
Water procurement. In 2017, the Commonwealth Department of Agriculture Water and the Environment paid $80 million to buy back water licenses from two Queensland properties owned by Eastern Australia Agriculture (EAA). The advice from the department had been that this was not to go ahead, but the sale was signed off by the then Minister who was also Deputy Prime Minister.
There was no open tender and the water in question was allegedly sold for many times its worth (Tee 2020). This prompted an investigation by the ANAO which found that the procurement arrangements were not fully effective. The department did not assess the procurement adequately and did not apply policy consistently. It did not develop a framework to maximize value for money and did not properly assess and manage procurement risks (Australia National Audit Office 2021a).
A senior government minister had been a director of the company but at the time of sale had no direct or indirect financial interest in the company, having resigned on becoming a member of parliament. He stated that he did not benefit from the water licenses sale. Although an Australian company, the profit from the water licenses went overseas to a tax haven in the Cayman Islands, and thus no tax on the windfall was paid in Australia.
There was a question about what the government was actually buying. It was not water that existed, but rather rights to overland flows, which are only available in floods. In a sense this is a non-existent asset (insofar as it is highly unreliable) and therefore represented very poor value for money as the purchases were not securing appropriate environmental outcomes. In summary, an exorbitant price was paid for a non-existent asset (hence, providing little to no value for money), and the company had close links with a senior federal minister.

Red flags for agents
In regard to cases of transgression by an agent, there are red flags that can be a warning. A corrupt offender might sometimes be a government official or sometimes be a supplier, or sometimes they could be in collusion. If they are not in collusion, officials and suppliers are all well placed to identify red flags during the procurement process. An Australian anti-corruption agency the Independent Broad-based Anti-corruption Commission in Victoria has prepared a set of red flags to alert people to corruption risk in procurement (IBAC 2015). The process is structured into five stages of vulnerability The bidding process Choosing the preferred supplier Paying for goods and services Delivery of goods and services Contract management (variations, extension, blowouts) For each, it lists red flags for potentially corrupt employees, and potentially corrupt suppliers.
For example, in the bidding process red flags for employees might include writing tender specifications in a way that favors a particular supplier; frequently using exemptions to circumvent competitive procurement; accepting late or suspicious bids; not declaring connections with bidder (e.g. family member or friend), accepting offers of gifts, benefits or hospitality; releasing sensitive information (e.g. tender specifications) to a particular bidder.
For suppliers of services, red flags might include: submitting suspicious bids (e.g. fake companies, subsidiaries, shell companies or affiliates) to give the appearance of competition; not declaring connections with another bidder (e.g. same names, contact details); submitting bids that vary significantly from others; or offering gifts, benefits or hospitality. As controls on such behavior, the document suggests a range of due diligence measures, regular and random audits, policies for probity and conflict of interest; as well as regular training and staff development (IBAC 2015). The document lists many red flags for each of the steps in the procurement process. Readers may also find useful the many red flags outlined in international studies by Ware et al. (2011). Examples of both system weakness and system failure have also been identified in Australia (Graycar 2019). In of these Australian cases there were rules in place which were being evaded or manipulated. Having rules, robust and concrete as they may be, is insufficient to address and prevent procurement corruption. Simple monitoring of processes in place could have either prevented, or identified much earlier, corrupt conduct in many of the cases examined. Some cases involved bypassing certain rules for the sake of getting things done to meet deadlines and other external pressures, all of which were taken advantage of by corrupt officials. Almost half of these cases pointed to the failure of people around the corruptors (especially their supervisors) to meet their organizational responsibility to apply existing procurement rules and procedures. Sometimes this is linked to incompetence, sometimes negligence, and sometimes due to pressure from above, or the need to meet targets. There were also system failures. For example, in half of the cases there were no concrete procurement rules and processes in place. This provided opportunities for misconduct and for officials to make up processes as they went along. There was an absence of procurement checks and balances, and few clear and practical guidelines, policies and procedures.

Red flags for principals
In three of the (principals) examples listed above the procurement involved using grant money to further a politician's political advantage. This is currently regarded as neither illegal nor criminal. As the Auditor General found, processes were not followed, but to make a ruling on criminality or corruption is beyond the jurisdiction of the Auditor. In two other cases the Auditor found that the officials were more likely to have behaved incompetently, leading to very bad procurement outcomes, but again no criminality. There seem to be no red flags other than post-event reports or media exposure.

Responding
In most cases where the agent behaves badly there is either a legal remedy or the anticorruption agency recommends that the department improve its processes and procedures and report back on its process of rectification. Following Operation Dunham above (IBAC 2017) concerning the Education Department learning system, two major investigations found the facilitating factors for the corruption in procurement were very much cultural and managerial. They identified: a culture of noncompliance, bullying and entitlement a lack of accountability, and lack of a culture of integrity in senior leadership inadequate controls around procurement inadequate governance of risk and major projects poor controls for financial management in the department and in schools incompatible financial systems inadequate auditing. A 'willingness of some senior leaders in the Department to deceive' The Department responded (Victoria Department of Education and Training 2017) with a strategy for huge cultural change, documented in its response to the anticorruption agency which had investigated the procurement lapses. It proposed three lines of defence and stated these in aspirational terms A culture of integrity underpins everything we do Smart systems and controls to oversee and guide our actions Independent and robust assurance In the first line they outlined how they would embed values, improve capacity, and ensure people feel safe to raise concerns. In the second line they outlined how they would develop smart systems and corporate governance to address delegation of authority, lack of clarity about accountability, lack of oversight and unclear roles and responsibilities. In the third line they outlined how they would enhance internal audit plans, financial services data analytics and an audit and risk committee. The reform process guided and often consumed the department for two years after the corrupt procurement issue was first identified. This example highlights the responses available for dealing with agents who transgress as compared to principals who do so. There is no similarly structured response for principals who behave badly.

General observations
In rich countries which observe the rule of law, the principal/agent theory works almost exclusively on the basis that the principal is principled, and the agent may be a potential transgressor. More affluent countries seem less corrupt in part because established models either miss a great deal of what is going on or routinize it, as exemplified above. Outright bribery is easily modeled, but there are more subtle exchanges that require close examination. Public administration in theory and practice has active debates over the merit of rules based versus values based corruption controls.
Generally there are three approachesmore rules, better enforcement, "better" culture. The agent generally faces real time monitoring, supervision, detection, and if appropriate, prosecution. The principal however is subject only to ex-post detection. This is often in the form of a report which can be acted upon, discredited, or ignored, and whatever sanctions might be available are likely to operate in the political domain.
An important part of responding to corruption is careful diagnosis of the problem. In rich countries, as Michael Johnston (2005) has pointed out corruption is characterized by trading in influence, making connections for a fee, using wealth to gain access and multiply that wealth and sharing the spoils (contracts, and favorable laws). It all works within the system, and may be lawful, but as many have observed "lawful but awful" (Lessig 2011;Lessig 2013;Thompson 2018).
When subject only to ex-post detection the principal may be either immune from consequences or able to re-orient the playing field. As noted by Persson et al. (2013), defining the problem is a key part of the analysis of systemically corrupt systems, and when corruption is seen through a principal/agent lens it can easily miss a significant dimension of the problem. In essence it puts a band aid on the infection, but does not solve the underlying problem in circumstances where the principal is not principled. Currently, the main mechanisms for dealing with misuse of public funds is through better enforcement of the rules and penalties for transgressors. This is insufficient. By developing frameworks that include a focus on "lawful but awful" principals, we can better define the situation and better respond.

Conclusion
When an agent does not meet standards of integrity in procurement there are clear responses available. Investigations by an anti-corruption agency can lead to the tightening of procedures, the bringing of criminal charges, or the re-orienting of culture in a department and different methods of guidance, management and control.
However, when it is the governmental principal who is delinquent there is little remedy. The common answer offered by politicians is first of all to deny that there is any wrong-doing, second to say that this is the way politics has always been played, third to say this is a politically motivated smear by political opponents, and finally, if people don't like it "they can vote me out at the next election." The Australian examples above of an agent flouting procurement processes are clear cut and actionable. The examples of the principal however are conceptually different. They show a principal using public funds for procurement to buy both votes and infrastructure, and to do so for the benefit of the decision-maker rather than the benefit of the community. This analysis is applicable across many countries.
The way forward is to have a clear theoretical understanding of the distinction between administrative corruption and political corruption, or between supply side corruption and demand side corruption, both of which involve an analysis of the behavior of principals and agents. By understanding the differences and the overlaps, appropriate responses and remedies can be developed.

Disclosure statement
No potential conflict of interest was reported by the author(s).