Is it Possible to Make ‘Easy’ Cuts to Human Services Spending by Attacking Waste, Fraud, and Abuse?

Is it Possible to Make ‘Easy’ Cuts to Human Services Spending by Attacking Waste, Fraud, and Abuse? Elizabeth T. Powers, Associate Professor of Economics, University of Illinois at Urbana-Champaign University of Illinois Institute of Government and Public Affairs February 18, 2014. The Health and Social Service area of the state of Illinois’ budget has comprised roughly 40 percent of state expenditures in recent years. When leaders decide to cut the state budget, this area is inevitably on the chopping block. This area includes Medicaid, child-care subsidies, Temporary Assistance to Needy Families (TANF), services for persons with developmental disabilities, and mental health services, to name a few. Spending took an enormous bump down in FY 2012 (just over 15 percent) and a hefty 5.6 percent cut for good measure in FY 2013 before an uptick in 2014. When viewed in a longer historical context, these cuts fit into a long-term downward trend. Indeed, the state has cut human services in the past decade by nearly one-quarter in inflation-adjusted, per capita terms. In the context of an already-squeezed Human Services budget, policymakers may find strategies of acrossthe-board or program cuts decreasingly palatable. Nearly all the programs left in the human services system today have survived multiple rounds of tough programmatic cuts. Examples of such cuts are eliminating specific services, withdrawing services from the slightly better off, and imposing or increasing co-payments. As programs have been whittled down, individuals who remain eligible tend to have a more acute need for them. Policymakers eventually reach a point where wringing further savings from human services programs through cuts threatens to distort the original programs beyond recognition. Thus, an appealing alternative to program cuts is to get tougher on waste, fraud, and abuse.

In the context of an already-squeezed Human Services budget, policymakers may find strategies of acrossthe-board or program cuts decreasingly palatable. Nearly all the programs left in the human services system today have survived multiple rounds of tough programmatic cuts. Examples of such cuts are eliminating specific services, withdrawing services from the slightly better-off, and imposing or increasing co-payments. As programs have been whittled down, individuals who remain eligible tend to have a more acute need for them. Policymakers eventually reach a point where wringing further savings from human services programs through cuts threatens to distort the original programs beyond recognition. Thus, an appealing alternative to program cuts is to get tougher on waste, fraud, and abuse.

The logic of auditing our way to savings
The public has long taken a dim view of "welfare" programs, including suspicions of criminal activity on the part of recipients, and Illinois is no exception.
In fact, the original "Welfare Queen," a convicted criminal who captured the imagination of the national media and ultimately helped Ronald Reagan to victory, committed her crimes in Chicago.
So it is not surprising that the public and many policymakers consider "Waste, Fraud, and Abuse" a rampant problem. Waste can be present in program administration, although, given Illinois's low government staffing resources, one could well argue that inefficiency in the system is the result of too little investment in Human Services administration rather than too much. Documented cases of fraud involve individuals in programs, higher-ranking individuals in state government, and program vendors. 3 Programs are "abused" when people become overly-dependent on them, or participate in them in the absence of real 3 For an example regarding state employees see Hinz, Greg. (November 7, 2013). No degree, no address, nine namesand one great state job. Crain's Chicago Business. Available at http://www.chicagobusiness.com/article/20131107/ BLOGS02/131109867/no-degree-no-address-nine-names-and-onegreat-state-job. (Footnote continued in next column.) For an example regarding program vendors see Chicago Tribune search, keyword "Medicaid fraud." Available at http://articles. chicagotribune.com/keyword/medicare-fraud. need. This short paper focuses on the question of fraudulent receipt of benefits by program recipients, a focus of recent high-profile efforts to reduce Medicaid spending in the state. 4 As a money-saving strategy, attacking "waste, fraud, and abuse" on the part of recipients relies on three implicit assumptions: • A significant number of clearly undeserving individuals are enrolled in a program. • These undeserving enrollees account for nontrivial state expenditures. • Undeserving enrollees can be identified and investigated without inordinate cost.
If any one of these assumptions does not hold, then this strategy for cutting spending may be ineffective.
Undeserving Enrollees. Defining an undeserving enrollee can be more difficult than appears at first blush. To define the population of eligibles, Human Service programs typically rely on income and asset limits, family structure, disability status, and age. Some programs also rely on further inputs to eligibility determination such as medical bills. Because the initial eligibility process is fairly rigorous, people with characteristics far from the eligibility standards are unlikely to become enrolled in a program except by deception. Because program cut-offs are sharply defined and because a good number of household units appear throughout the lower part of the income distribution, many households live in circumstances that place them near to, but not exactly at, the eligibility cutoff. Households with characteristics very close to eligibility limits are "marginally eligible" or "marginally ineligible" for the program. Because the inputs to the eligibility calculation are always changing, status transitions from being just ineligible to just eligible and back again are frequent.
What are the implications of the realities of program eligibility for the first assumption, namely, that a substantial number of enrollees are nondeserving? First, with the exception of fraudulent deception, egregious cases of ineligibility can be weeded out with competent intake processes. Second, the frequency with which a person's eligibility is re-evaluated 4 A note on terms: I use "eligibles" to describe the group that would be able to enroll in a program due to their circumstances. Individuals in this group may or may not be actually enrolled, however. I use "enrollee," "participant," and "recipient" interchangeably. The state often refers to recipients as "consumers." In years past, recipients were also referred to as "clients." (called "redetermination") may be longer than a day, a week, a month, or even a quarter. Under a highfrequency redetermination strategy, the state would identify (mostly) marginal ineligibles for elimination from the program, but it would face the costs of readministering program re-entry for many of them before too long. In addition, even if marginally ineligible recipients are technically ineligible most of the time, they belong to the range of low-resource households that the program overall seeks to help. Being a few dollars outside an eligibility threshold does not mean the unit is "undeserving." Rather it is a byproduct of the fact that in order to limit a program to a certain size, a definitive cutoff is necessary.
The Cost of Undeserving Enrollees. The bulk of marginal enrollees who are found to be ineligible and dropped from the program may not generate much savings to the state. Many programs provide benefits that are pro-rated in some way, such as the child care program, which imposes co-payments for recipients that rise with their income. When marginal ineligibles are removed from programs with sliding benefits, the state's costs are not reduced by much, because the awarded benefits to these people are small. In contrast, it may seem reasonable to suppose that in the case of an "all or nothing" benefit, such as Medicaid, even kicking out marginally ineligible recipients from the program could save a lot of money. As discussed below, however, recent experience indicates that even this supposition is doubtful.
What of fraudulent enrollees? How numerous and how costly are they? In light of the fact that most program benefits tend to be small, one might question the sanity of individuals who would put their liberty at risk for such a modest 'reward.' States' experiences with overpayments in the Supplemental Nutrition Assistance Program (SNAP) support this intuition. Available statistics suggest that net overpayments attributable to recipient "fault" amounted to less than 1 percent of all payments made by the program in 2011. 5 The Cost-Effectiveness of Redetermination. The third assumption is that identification and removal of undeserving enrollees in a program is a cost-effective proposition. As noted, if a blanket investigation of 5 all enrollees is conducted, most of those identified as problem cases are likely to be in the marginally ineligible group, a group that does not contain many "undeserving" and costly recipients. Some evidence from real-world policy supports this hypothesis. For instance, the auditor of the state of California found that identifying and prosecuting fraudulent activities of people already in their main welfare assistance program was quite expensive, whereas applying additional scrutiny to new applicants who appeared to have high potential for fraud was cost-effective. 6 These findings support a strategy of investing most recipient-targeted audit resources at the initial eligibility determination.
Looking for waste, fraud, and abuse in the state's Medicaid program A substantial driver of human services funding is the Medicaid program. In 2013, the state began making specific Medicaid cuts. By ending prescription drug programs and dental care for adults and seniors, the state reduced program costs by $250 million. These were very painful cuts to vital services, including the elimination of adult dental care coverage. Half again as much ($120 million) was proposed by the state to be saved through eliminating enrollees who would be found to be ineligible for Medicaid. 7 In September 2013, the state engaged a for-profit firm, Maximus, at a cost of $35 million per year, to scrutinize the eligibility rolls. Since state workers make the final determination of eligibility, the firm recommended to them which clients be continued, changed in status, or canceled from Medicaid. IDHS estimates that an additional $21 million was spent on staff support for this redetermination project. 8 ReportSeptember17_%202013.pdf changed in status, and that 230,000 be continued as before. The fact that the contractor recommended nearly half of reviewed cases for cancellation seems to suggest ample "waste, fraud, and abuse" upon which to capitalize. Of the cases processed through that date, state workers approved 70 percent of the recommended cancellations (alternatively, one-third of all cases reviewed). 9 However, of those actually cancelled, as many as 20 percent were shortly thereafter re-enrolled in the program. A high reenrollment rate is not surprising, considering that shorter-run fluctuations in circumstances often drive eligibility. More re-enrollment will likely be seen from this group going forward.
Nevertheless, around 140,000 cases were cancelled without near-term re-enrollment, representing around 238,000 persons. That sounds like a big number of individuals, so what were the cost savings? IDHS found that many of the people cancelled from Medicaid had not claimed services for a long while (6 months or more). While we don't have a breakdown of the reasons why services were not claimed, it is likely that many of the people in this group simply stopped using the program because they no longer had urgent medical needs, no longer faced large medical bills, found other sources of health care resources, or knew themselves to be ineligible (e.g., due to a move). The per-month savings per cancelled client was estimated at just $58. Extrapolating from that experience, the total cost savings from the Maximus contract will be at most $105 million. Given payments to Maximus and additional administrative costs, the state will net at most $50 million dollars. Note also that "large" cost reductions are generated only once, as the large bulk of existing cases is reviewed for the first time. Savings going forward from increased redetermination effort will be a trickle, not a flood. 10 While $50 million is a lot of money, it is far short of the hoped-for $120M total savings in Medicaid. The limited success from this approach suggests that policymakers consider applying other "smart" approaches to cost savings. 9 Illinois Department of Healthcare and Family Services. (February 3, 2014). IMRP to date summary report. Available at http://www2.illinois.gov/hfs/SiteCollectionDocuments/ IMRPReport.pdf 10 With a switch to managed care, IDHS notes that the state will pay a capitated fee. Thus, removing ineligible enrollees will become a more pressing matter. This observation highlights the fact that Medicaid fee-for-service is somewhat self-enforcing with regard to eligibility. That is, the state does not receive bills for enrollees who don't need Medicaid services. This aspect should properly be incorporated when discussing the true savings from a managed care system.

Policies for smart administration of programs
In most programs, even the maximum benefit attainable by the individual is fairly small. And, as spelled out in this paper, the type of recipient likely to be targeted for an audit or discovered to be ineligible at redetermination typically collects smaller benefits still. Because of these realities, the really big money in defrauding most human services programs is mostly available to those in a position to "aggregate" program benefits, not individual recipients. This insight is the reason why provider fraud in Medicaid (and Medicare) is such an important concern at the federal level and also the reason that a great deal of enforcement effort in the SNAP program is directed at stores. 11 While these cases can be complex and expensive to investigate and prosecute, the cost savings-including the value of deterring others who might engage in this behavior-are potentially large.
Evidence from some past redetermination efforts, as well as Illinois' efforts in Medicaid, suggest that we may be able to improve the cost savings per administrative dollar spent by taking a more targeted approach to eligibility evaluation. The California experience, although just one example, suggests that focusing resources on initial eligibility determination and screening may be a good investment. Keeping people with intent to commit fraud out of the system in the first place turns out to be much cheaper than discovering such individuals and removing them from the system once they have laid claim to a benefit. That suggests focusing enforcement efforts on fair but accurate and thorough scrutiny of new applicants, rather than doing a blanket sweep of all recipients, regardless of their use of the system. The Illinois Budget Policy Toolbox is a project by the University of Illinois Institute of Government and Public Affairs. IGPA is a public policy research organization striving to improve public policy and government performance by: producing and distributing cutting-edge research and analysis, engaging the public in dialogue and education, and providing practical assistance in decision making to government and policymakers.
Learn more at igpa.uillinois.edu while at the same time exposing and deterring fraud. In addition, individuals and families can be better helped by timely and accurate service and referral to appropriate programs and other resources that may be available for them. Unfortunately, the state has not made much progress in this area. Human Services workers continue to rely on systems that are programspecific and usually at least 25 years old. 12 Finally, the state has a long history of undisciplined contracting with vendors in many human service areas. The system is short on accountability and transparency and has been rife with special pleading, non-merit-based awards, and even outright fraud. A systemic reform of the state's human services contracting system that emphasizes capacity, merit, and performance is long overdue. 13

Conclusion
It is an undisputed fact that undeserving recipients have sometimes defrauded welfare programs. The purpose of this paper has not been to argue that limiting waste, fraud, and abuse in human services programs is not important, or that the state should take a "know-nothing" attitude toward eligibility. Experience, however, suggests several approaches that might be more productive.
Illinois should be vigilant in guarding against largescale crimes, which originate with entities that are able to aggregate benefits. Theory and evidence also suggest that scouring an entire program for fraud may not be the most cost-effective approach. Instead, it may be better to scrutinize new applicants and selectively target cases for more frequent redetermination that have characteristics associated with fraudulent outcomes. Improving administrative systems overall, especially by creating linked records, would have the benefit of not only discouraging waste and fraud but helping the system run more efficiently for deserving recipients as well. Finally, reforms to state contracting standards and procedures have the potential to improve overall effectiveness of human services programs while also discouraging waste, fraud, and abuse. •