Caught in the Middle has been added to your bookmarks.
The U.S. government now uses sophisticated analytics to identify and investigate improper payments. By adopting similar tools and processes, companies doing business with the government can reduce the risk of improper payments caused by the misdeeds of subcontractors or employees. Moreover, these tools can help any company to identify improper or erroneous payments in its own value chain.
Improper payments by U.S. government agencies and departments happen for a variety of reasons, including inadvertent errors, inadequate controls and, in some cases, willful acts of contractor fraud.
Whatever their cause, improper payments—that is, overpayments, underpayments, payments made without substantiation, or payments otherwise involving fraud, waste, abuse or errors—are in the government’s crosshairs as never before.
The growing call from politicians and the public to identify, mitigate or at least minimize the effects of waste, fraud, abuse and errors is reflected in stricter laws and regulations emanating from the federal stimulus program, healthcare reform, strengthening of the False Claims Act and other initiatives. Overpayments specifically were targeted in the summer of 2010, when Congress passed and President Obama signed the Improper Payments Elimination and Recovery Act (IPERA).
Most U.S. business leaders, especially prime contractors to the government, recognize that these mandates, along with the clamor for greater accountability, will likely lead to an increase in investigations with the purpose of recovering improper payments. But these leaders may not understand the extent to which, and by what means, the government is adding depth and rigor to those probes. And, rather than relying primarily on whistleblowers and tips to unearth problems, the government is now using sophisticated analytics to identify and investigate improper payments that have already been made. Some agencies are beginning to use predictive analytics and near-real-time transaction monitoring tools to catch errors before funds are distributed to contractors.
These changes have implications for both prime contractors and subcontractors. Prime contractors are, in general, subject to greater scrutiny and accountability – including investigation into their use of subcontractors. Prime contractors also can be held responsible for improper payments made to their subs, even if they are not aware of or involved in the precipitating mistake or fraudulent action.
How can companies doing business with the government lessen the risk of becoming a national news story due to improper payments caused by misdeeds of a subcontractor or even one of their own employees? And, from an internal operational efficiency perspective, how can they curb their own revenue leakage attributable to accidental or fraudulent payments.
The answer may lie in adopting strategies, processes and tools similar to those that the government is beginning to use to proactively root out waste, fraud, abuse and errors. By doing so, companies can potentially preempt government inquiries and expedite corrective action. Advanced tools wielded by analysts experienced in antifraud programs can help contractors—and other businesses, even if they don’t do business with governments—to identify and recoup improper or otherwise erroneous payments.
Such initiatives are neither simple nor cheap, and some senior executives may balk at making such a commitment solely for compliance or even risk management purposes. However, along with addressing regulatory requirements, companies can use these new approaches and technologies to achieve other business objectives, such as identifying and resolving areas of revenue leakage across their value chain and refining operations across the enterprise, including strategic planning, operations, marketing, and human resources.
Improper payments by the U.S. government totaled nearly $110 billion in 2009, a more than 50 percent increase from the previous year (Figure 1).1 This total includes “payments made in error or because of fraudulent claims by contractors and organizations as well as benefits sent to individuals who are dead or in jail.”2 The government has attributed the dramatic rise in identified improper payments to both improved detection efforts by agencies and the significant increase in outlays associated with the economic downturn.
Sources: http://www.paymentaccuracy.gov, Office of Management and Budget
Figure 1: Total federal improper payments, FY2004-2009 (billions)
No federal agency is immune to the problem of improper payments. Social programs, given their responsiveness to a massive population of payees, are particularly susceptible. In 2009, the Department of Health and Human Services reported approximately 60 percent of the $110 billion of improper payments alone. Similarly, the Unemployment Insurance program administered by the Department of Labor ($12 billion), the Earned Income Tax Credit program monitored by the Department of Treasury ($12 billion), and other programs under the responsibility of the Social Security Administration ($8 billion) further added to the collective issue (Figure 2).
Sources: http://www.paymentaccuracy.gov, Office of Management and Budget
Figure 2: Improper payments by program FY2009 (billions)
The Obama administration has taken a number of steps to address improper payments. In late 2009, the president issued an executive order laying out a strategy to reduce improper payments. This was followed by a presidential memorandum in March 2010 directing federal agencies and departments to expand and intensify their use of payment recapture audits.
The passage of the IPERA in July 2010 is viewed as a key step toward realizing the president’s goal of reducing improper payments by $50 billion between mid-2010 and 2012. The act outlines a number of activities aimed at boosting the campaign against improper payments. Specifically, the bill will improve agency efforts to reduce and recover improper payments in several ways, including:
Evidence of the government’s intensified focus on improper payments is also apparent in day-to-day government operations. For example, Task Force 2010 is an initiative of the U.S. military to ferret out corruption among providers of security, supplies and reconstruction work to the Afghan war effort.
Figure 3. Recapture of payment errors to government contractors − cumulative reporting, FY2004-2009 (millions)
As a harbinger of what is perhaps ahead for the entire federal contracting community, Gen. David Petraeus, commander of U.S. forces in the Middle East and Central Asia, has said investigations will go well beyond their historical purview of prime contractors to embrace “not only the subcontractors, but the subcontractors to the subcontractors – literally, where is the money going, and is it all above-board?”3
The arsenal available to governments, contractors and other businesses to identify improper and otherwise inappropriate payments consists of three broad categories of activity:
To address the increasing risks associated with improper payments, government contractors and other businesses can implement their own high-tech analytical and monitoring program focused on payments they receive from the government, as well as payments they make to subcontractors, whether or not as part of a government contract. A clearer understanding of the tools involved – and the types of expertise needed to use them effectively – can help contractors and other businesses decide which might be most helpful in their operations.
Generally speaking, forensics involves combing through data on past activities, events and transactions to identify issues and anomalies. Forensic analysis is superior to traditional auditing techniques because it is capable of testing virtually 100 percent of a population instead of sampling only a percentage. The process includes anomaly detection tests, or rule sets, run against normalized data sets and analysis of the results by forensic accountants, also known as forensic auditors. The number, type and complexity of anomaly detection tests depend on the type of improper payments being targeted. Examples of anomaly detection tests that can be employed include:
In a process the government has idiomatically labeled “pay and chase,” information uncovered in this manner is then used as the basis to recover improper payments. U.S. government agencies make extensive use of data analytics and forensic analysis, also referred to as data mining or data matching. An example of the success of this approach involved the Department of Housing and Urban Development (HUD). HUD achieved a 70 percent reduction in the level of improper payments from 2002 to 2009, from $3.430 billion to $1.022 billion, which the agency attributed to the use of sophisticated data-matching solutions that helped confirm recipient eligibility.4
The Department of Defense (DoD) has deployed continuous monitoring technology to prevent improper payments. The Business Activity Monitoring (BAM) tool has assisted the DoD in preventing more than $700 million in improper payments over the past two years.
Perhaps the best defense in mitigating losses from improper payments is simply preventing such disbursements from being paid out in the first place. A strategy built around prevention is many times more effective than one founded on pay and chase – even if that pay and chase strategy is powered by the advanced analytics capabilities described above.
Continuous monitoring is a relatively nascent technology that provides access to this type of protection. The technology equips an organization with the ability to detect and prevent improper or other inappropriate payments in real time. This technology works by screening each transaction against a predefined list of characteristics, or “rules,” and making automated decisions about that transaction based on the result of this real-time analysis. Depending on the needs of the organization, a transaction can be automatically denied or passed along to other workstream functions for additional review. This type of technology is just beginning to penetrate both the federal and commercial marketplace.
Predictive analytics is a methodology utilizing machine learning and statistics to analyze historical and current data to predict future actions and events. Using advanced analytics and algorithms, such as econometric models, neural networks, decision trees and self-organizing maps, data can be mined for trends, patterns and behavior that will provide, for example, indicators of anomalous activity that go beyond rule-based detection. In short, one variable, or a number of variables acting in concert, are analyzed to assess whether a relationship exists with other variables of interest.
Potential benefits of predictive analytics include:
Application of predictive analytics is varied. Studies might involve views into fraud detection, price optimization, product demand forecasting, customer segmentation and loyalty or the effects of geospatial distribution.
It is possible to combine the best attributes of data analytics, continuous monitoring and predictive analytics into an even more effective approach. Essentially, the elements of past, present and future can be viewed and analyzed in one hybrid fraud detection framework. This affords an organization even deeper insight into the nature of its payments and the effectiveness of its operations.
In this type of multidimensional technology platform, incoming payment data is streamed down two parallel paths, one that operates in a “live” environment and the other in an “offline” setting. At a high level, the continuous monitoring technology is deployed in a real–time or live stage, with predictive analytics and forensic data analytics being completed in an offline stage.
As discussed previously, the transaction data streamed in the live environment is screened with rules and runtime models to make real-time decisions on disposition – for example, a pay or no pay review. With the hybrid approach, the results of these dispositions are sent to the offline analytic setting for further assessment.
There it is possible to query the data to make the monitoring process more effective. Was flagging a particular transaction effective? Should the rules be modified or optimized? If the rules or models are changed, how many more or fewer transactions will be flagged? Are there emerging trends? Clearly, these are the types of questions appropriate for an offline platform—and predictive analytics in particular—rather than slowing down the monitoring function that is applied to live payment data.
The benefit of this architecture is that, in effect, it acts as a giant feedback loop. This is important because perpetrators of fraud continually modify their approach as they learn about new thresholds, whether imposed by governments through new laws or by companies that are simply more vigilant. Companies should continually evolve and refine their capabilities to keep pace with potential fraudsters.
However, the benefit extends beyond fraud. Are the same types of payment processing errors or waste occurring? Can error patterns be identified by frequency or source? If so, this type of insight can guide an organization about how and where processes can be enhanced.
Lastly, given that this hybrid approach captures and analyzes historical data in the offline setting, an organization can incrementally add a forensic querying interface that allows data to be viewed through a different lens. With the mechanism for collecting and storing payment data already in place via the monitoring functionality, this additional capability can be included efficiently. In fact, it is entirely possible that insight gained through this forensic interface can be used to further inform the rules and models used to screen live payment data. This effectively gives the organization aspects of all analytic worlds in one package.
Private sector businesses, whether government prime contractors or not, can and should leverage the power of analytics and monitoring tools. These tools provide data on transactions that can be used by organizations to mitigate and help with recovery of improper and other otherwise erroneous payments, as well as provide value to strategy planning and operational areas across the enterprise.
Specifically, in an era where transparency and accountability are emphasized increasingly, analytics and monitoring tools help government contractors and commercial enterprises:
Benefits derived from advanced analytics and monitoring tools do not end with compliance and recovery. Organizations can gain traction with other diverse and critical business imperatives, including:
To capitalize on these advancements, some companies may need to elevate certain aspects of their existing organization, approach, and capabilities, including:
Culture and tone at the top. Implementing analytics and monitoring can take perhaps six months in a smaller company and up to several years in a large, global enterprise. Whatever the duration, executive-level support is essential.
The board of directors and senior executives need to be willing to undertake an ambitious transformation, investing well beyond the traditional internal controls companies have relied upon. The board and senior management also need to set a strong tone of compliance, honesty and ethics – a tone that should be spread throughout the organization through consistent messages, training programs and reinforcement of appropriate actions and behavior. Finally management commitment needs to be actualized through the dedication of resources to the initiative and management oversight through each stage of deployment.
Collaboration between internal audit and information technology groups. Should a company choose to pursue development of advanced analytics and monitoring capabilities, the chief compliance officer may well own the project. However, much of the heavy lifting in using these tools will likely fall to the internal audit team, which will need to bake the system into the company’s internal audit program, and the information technology (IT) department, which will have the analytics capabilities and be responsible for deploying new technologies and integrating them with the company’s existing IT systems.
Existing processes will need to be revisited from the perspective of how payments are made today – when are they authorized, who authorizes them, and the nature of the entire payment cycle. The internal audit and IT teams will need to prepare for the task by developing new skill sets, including rules development, fraud and anomaly detection, analytics, functional testing skills and alignment with the tone at the top. There also will need to be a new level of cooperation and collaboration between the two groups as they establish the baseline analytics and monitoring capabilities and then iteratively build on those capabilities over time.
Companies doing business with federal departments and agencies, both prime contractors and subcontractors, can look forward to ever more intense scrutiny of payments they receive from the government. Government officials have the authority and tools to identify and rectify improper and otherwise inappropriate payments, and they are motivated to do so.
Fortunately for contractors, the technologies and techniques now being used by multiple government agencies to attack waste, fraud, abuse and errors are also available to them. By understanding and deploying predictive analytics and continuous monitoring solutions, companies can avoid potentially costly problems, stay in good stead with their government clients, and leverage their technology and process investments to improve overall operations and performance.