Example:

A contractor bidding for a contract with the National Roads Agency (NRA) appoints an agent to assist it win the contract.  It is agreed that if the contractor is awarded the contract, the contractor will pay the agent a fee of $1 million (5% of the contract price).  The agent is appointed under a formal agency agreement which states that the agent will carry out specified services.  However, the fee being paid to the agent is grossly in excess of the market value of the legitimate services which the agent actually provides.  Although the contractor does not actually know that the agent will use the fee for corrupt purposes, the contractor thinks it likely that this is the case due to a high level of corruption in the market in which the project is located, and the significant disparity between the value of the legitimate services to be carried out by the agent and the amount of the fee.  The agent agrees to pay the Chief Executive of the NRA $750,000 if the contractor is awarded the contract.  The contractor is awarded the contract.  The contractor pays the agent the $ 1 million fee.  The agent pays $750,000 to the Chief Executive.  The cost of the fee (and therefore of the bribe) is included by the contractor in the contract price.  The NRA therefore pays more than it would have done had there not been a bribe, and the contract is not awarded to the best evaluated contractor, but to one which paid a bribe. 

Explanation:

The above example illustrates what has historically been one of the most common methods of paying a bribe in relation to major projects.  This method avoids direct contact or payment between the bribe payer and bribe recipient.  As in most cases agents perform a legitimate and valuable commercial service to organisations, it can accordingly be difficult to distinguish corrupt agents from legitimate agents, and to identify whether the agent intends to act, or has acted, corruptly.  This difficulty is compounded by the fact that many agents work on a success fee basis (i.e. they are only paid a fee if a contract is awarded in consequence of their services).  A legitimate agent may need to undertake extensive work in order to help an organisation be awarded a contract.  It may need to widely market the organisation’s products, participate in tenders and contract negotiations, and support the delivery of the organisation’s goods.  And if the contract is not awarded to the company appointing the agent, the agent will have performed these services without compensation.  An illegitimate agent on the other hand may need only to have direct contact with corrupt personnel working for the relevant government or project owner, and may not need to undertake any services whatsoever other than to agree and process the bribe.  In some cases, an agent may perform a combination of legitimate and illegitimate services.  The success fee basis makes it difficult to relate the value of the fee to the value of legitimate services carried out.

If it is established that the agent has actually paid a bribe to a project owner representative or government official in order to ensure that an organisation wins a contract, then the organisation which appointed the agent and won the contract will normally be liable for the corrupt actions of the agent if it can be shown that the organisation appointed the agent with the purpose of it paying a bribe, or knew or suspected that the agent would pay a bribe but did not take reasonable steps to prevent this from occurring. 

Judges in a prosecution may infer knowledge and therefore guilt. They may determine that the circumstances of the appointment are such that the organisation must have known that a bribe would be paid.  For example, the judge may examine whether the agent’s fee was reasonable and proportionate to the legitimate services and risk undertaken by the agent.  If the fee is very high, and the level of services and risk is very low, and a bribe was actually paid by the agent, then the judge may infer that the organisation must have known – why else would it pay such a high fee for so little work and risk?  While a success fee based on a percentage value of the contract can make this inference based on value difficult, it is not impossible.  For example, knowledge and guilt may be inferred if a $1 million success fee is agreed in circumstances where it is obvious that the agent will be providing very few legitimate services in return, and it is therefore obvious that the fee is grossly excessive.

Using an agent as a conduit for a bribe is not the only method.  Bribes can be concealed in, and paid out of, the contract prices of sub-contractors, suppliers, consultants and joint venture partners.