Quick Start for:

Letter from Comptroller Strayhorn


October 25, 2006

The Honorable Eliot Shapleigh
The Honorable Carlos I. Uresti
The Honorable Carter Casteel

Dear Senator Shapleigh, Chairman Uresti and Representative Casteel:

In May 2006, you requested my assistance in researching the Integrated Eligibility and Enrollment contract between the Health and Human Services Commission (HHSC) and Accenture LLP. I have completed this research.

At this writing, the project is behind schedule and $100 million over budget, without a revised plan to get the project back on course. Accenture has not met its performance requirements and has not been held accountable for its failure. Clients are still reporting delays and inaccuracies in processing their applications. HHSC has proven it cannot manage Accenture and the contract.

My conclusion is that this project has failed the state and the citizens it was designed to serve. The contract with Accenture must be ended. I recommend the Legislature pass emergency legislation that removes HHSC's direct management of the project and places the responsibility with a turnaround team composed of experts who can effectively manage state resources and stop the drain on tax dollars. And, most importantly, make sure children receive the health insurance for which they are eligible. In addition, the Legislature should review this administration's 27 major policy changes that have resulted in even more children losing health insurance.

Since Accenture began operations on December 1, 2005, CHIP enrollment has plunged by 8.5 percent or 27,567 children though August 2006. It seems unlikely that many of these children dropped from CHIP were transferred to Medicaid since enrollment in children's Medicaid also fell during this timeframe by 2.9 percent or 53,937 children. Evidence shows that some children were inaccurately denied benefits but were in fact eligible. In addition, I found that rather than saving money in this biennium, this contract will cost the state almost $100 million more than budgeted while fewer children and families receive the needed benefits.

The state's automated eligibility determination system, the Texas Integrated Eligibility Redesign System (TIERS), has cost taxpayers $279 million to date, and despite this expenditure, is being used in just four state offices in Travis and Hays counties after three years. Both HHSC and Accenture have hired hundreds of additional personnel to address a myriad of costly problems with TIERS.

As of August 31, 2006, HHSC has paid Accenture more than $123 million to process eligibility for a fraction of Texas' applicants, and the project that was intended to save the state's budget will end up costing the budget $100 million more. This project is over budget and under performing.

To understand the full range of HHSC's mismanagement and Accenture's substandard performance, one must address the complex, 6,000-plus page contract between HHSC and Accenture. Despite its length, the contract is vague and misdirected, leaving taxpayers to pay millions of dollars for inadequate work.

This is certainly not the impression received from the public statements of HHSC and Accenture. Taxpayers and legislators have been told that Accenture is subject to strong performance standards; that the state pays the company only when it performs well; and that the contract guarantees Texas will be reimbursed for Accenture's failures to date. These statements are unsupported by the evidence.

As you may recall, my office recommended in January 2003 to reduce HHSC's administrative costs for determining eligibility by implementing a call-center for children's Medicaid only. Unlike HHSC's plan, my recommendation did not include massive state worker layoffs, and it would have provided a proving ground for testing the new technologies and approaches that HHSC dived into with inadequate planning and little testing.

Contrary to my recommendation, HHSC expanded the scope of the project to include the Children's Health Insurance Program (CHIP) in the integration, maintenance of the state's eligibility technology system and HMO enrollment brokerage services. This major increase in scope created a project that could not be executed within HHSC's proposed timeline and budget.

Between November 2005 and January 2006, HHSC outsourced the eligibility determination process to a private vendor, Accenture LLP. HHSC estimated that this move would save the state $646 million in all funds by 2010, primarily through state staff reductions and office closures.

Since January 2006, however, the outsourcing experiment has come under intense criticism from parents, advocacy groups, legislators and the press, due to steep declines in the number of Texas children eligible for CHIP and Medicaid.

Furthermore, in case after case, eligible children are not getting enrolled in CHIP despite their parents' best attempts to navigate Accenture's call center and application process. Many cases show eligible children are losing health care coverage through no fault of their own, but due to mistakes and errors made by Accenture and its subcontractors.

This project has been mismanaged at virtually every turn. HHSC's lack of effective planning and contract management has drained scarce taxpayer resources. This controversy led 60 state representatives to ask HHSC to cancel the contract with Accenture.

To date, HHSC has not reported the number of applications Accenture has received or processed for CHIP, Medicaid, TANF and food stamps, nor has HHSC reported the percentage of these cases determined eligible for benefits. Because of this, the Legislature has no way to determine the cost of eligibility determination per applicant.

In the course of our study, we found that:

  • the contract's performance standards are weak, and HHSC is not even monitoring contractor performance in some crucial areas.
  • the contract forces the state to pay Accenture whether it performs well or not.
  • the state most likely will never be fully compensated for Accenture's failures, given the contract's limitations on liability and HHSC's poor contract management.

Some of the key findings that my staff uncovered in their exhaustive review of the program include the following:

1. HHSC estimated the cost of Accenture's contract to be $899 million in all funds over five years. This figure is understated and misleading.

In addition to the publicly released cost estimate:

  • HHSC has removed $94.7 million of expenses from Accenture's contract and will pay these on Accenture's behalf.
  • HHSC has added $5.9 million in amendments for TIERS modifications to be made by Accenture and Deloitte Consulting, the previous TIERS vendor.
  • HHSC hired Deloitte for an additional $2.3 million in 2005 to help transfer TIERS to Accenture.
  • In April 2006, HHSC notified the federal Food and Nutrition Services (FNS) and the Centers for Medicare and Medicaid Services (CMS) that it may hire Deloitte for another year to work on TIERS, for an additional $39.1 million.

The General Appropriations Act assumed the state would save $140.9 million in all funds, including $65.1 million in general revenue, in fiscal 2006 and 2007. Instead, HHSC will spend that $65.1 million and an additional $34.8 million in general revenue on the project in those years. Therefore, the state will spend $99.9 million more than anticipated. Given its track record to date, it seems unlikely that HHSC will save the expected $646 million in all funds by 2010.

2. The contract provides Accenture with perverse incentives to process applications inefficiently.

  • Accenture's payments are based on a complex combination of more than 70 prices for transactions processed per application. This payment structure gives Accenture the incentive to process as many paper-based "touches" to the client as possible, when the intention of the call center model is to make the process simpler, more customer-friendly and cost-effective.
  • Accenture is paid when applications are completed and ready for the state's final determination. Accenture also is paid, however, when applications "time out" because clients have not submitted sufficient information for processing. These applications are sent to the state for denial, and Accenture is paid the same rate as for completed applications. This payment structure does not provide Accenture with any incentive to seek necessary information from clients before their applications time out.
  • The contract specifies that Accenture is to be paid only for completed and "appropriate" transactions, but HHSC has not established any effective mechanism to determine whether transactions are appropriate before paying Accenture. It can only recoup inappropriate payments after the fact, not prevent them.

3. HHSC has failed to hold Accenture to the contract's limits on profit.

  • From July 5 through December 2005, Accenture earned $23.5 million in profit on $75.8 million in revenues spent for transitioning from the old to the new eligibility programs. This greatly exceeded the contractual limit of $5.6 million for this service.
  • Accenture earned $5.7 million in profit for the first four months of operations (January through April 2006), exceeding the contractual limit of $3.3 million.
  • HHSC allows Accenture to retain a portion of the state's payments for future, unplanned capital expenditures. If Accenture does not spend the estimated $31.4 million it will accrue by the end of the contract, it will keep half of this amount, or $15.7 million, plus any interest accrued, as pure profit.
  • Although the contract sets "limits" for Accenture's profit, it also allows the vendor to keep any of the excess profits it actually earns. Thus, these limits are pointless and ineffective. The contract allows HHSC to reduce Accenture's fees after two years if its actual profits are 20 percent higher than the contractual limits, but the new prices then counter-intuitively also increase Accenture's allowable profit limit by 20 percent.
  • In addition, Accenture overbilled HHSC at least $327,000 for labor at rates above the agreed maximum rate for additional services.

4. The contract is a "cost reimbursement" model that attempts to limit Accenture's overhead expenses and profit through various complex requirements. In the end, though, these complexities will have little effect on the state's payments.

  • HHSC has not monitored or audited Accenture's overhead expenses to determine whether the company has complied with the contract terms.
  • The contract does not clearly establish whether Accenture is required to repay the state if its overhead expenses are higher than estimated or if its expenses fail to comply with the contract terms.
  • Accenture charges profit on overhead expenses, and since Accenture is allowed to simply estimate its overhead, it can maximize its profit by overestimating overhead.

5. The contract pays Accenture for effort rather than performance.

  • Payments are made on a regular monthly basis regardless of whether Accenture has performed well.
  • Accenture has failed to deliver various required reports and services, and has well-known performance problems, but HHSC has not withheld any payments due to poor performance.
  • Accenture has failed to meet some key performance requirements, but HHSC again has not assessed any liquidated damages. Liquidated damages are intended to recover damages or costs HHSC incurs as a result of Accenture's failure, and are specified in the contract.
  • Accenture is being paid an additional $5.9 million to modify TIERS. This work is being paid at hourly labor rates, with no requirements for performance.

6. HHSC's contract relies upon 94 "Key Performance Requirements" or "KPRs" to ensure that Accenture performs well, but the agency has failed to define or enforce most of these requirements.

  • Accenture measures and monitors its own performance on KPRs, with no meaningful HHSC oversight.
  • Almost a year after the contract began, HHSC has made no effort to measure or track 36 percent of the KPRs.
  • The contract's KPRs do not measure performance concerning the most serious problems clients report, such as inaccurate information and lengthy delays.
  • Liquidated damages assigned to KPRs appear to be arbitrary and do not realistically represent the actual monetary damage the state might incur.
  • HHSC allowed Accenture to set its own liquidated damage amounts based on what Accenture would be willing to pay, rather than on the potential damage to the state should Accenture fail to meet its performance requirements.
  • In some cases, liquidated damage amounts are so low that it may be cheaper for Accenture to pay them rather than to fix the underlying problem.
  • Liquidated damages do not address instances of grossly or consistently poor performance.
  • "Earn backs" granted for superior performance allow Accenture to avoid liquidated damages. An "earn back" for an unrelated or lower-priority requirement can be used to offset a liquidated damage assessed for an unrelated but higher-priority requirement.
  • KPRs are significant only when tied to HHSC's ability to assess liquidated damages. Yet HHSC has assessed none to date, despite repeated poor performance.

In addition, HHSC limited Accenture's total contract liability to $250 million, or just 27.8 percent of the total contract value, when standard industry practice sets this limit as high as 100 percent or more.

HHSC's legislative testimony and quotes in the media seem intended to convince the public and Legislature that Accenture is responsible for actual damages the state has incurred from the company's failures during the introduction of integrated eligibility. Nothing could be further from the truth.

The reality is that the contract does not require Accenture to pay actual damages, only arbitrarily set and generally inadequate liquidated damages, which HHSC has not yet assessed against Accenture. To obtain actual damages, HHSC would have to sue.

Interestingly, HHSC has said that it expects to recover only about 30 percent of its actual costs caused by Accenture's failures, and part of this reimbursement could come in the form of additional services as well as payments.

7. HHSC has jeopardized federal funding to Texas.

  • HHSC added food stamps to the IEE system without prior federal approval, losing about $6.9 million in federal funding for fiscal 2006 in consequence.
  • HHSC risks losing $23.7 million annually in enhanced federal funding for food stamps because processing timelines for the integrated system do not meet federal standards.
  • HHSC already has identified approximately $3 million in food stamp and Temporary Assistance to Needy Families (TANF) overpayments made through TIERS.

8. HHSC calls the TIERS-driven Integrated Eligibility and Enrollment system a "pilot," but it is not a pilot in the usual sense, because the agency did not test the system before causing major disruption to its existing programs.

HHSC relied on an unfinished and unproven software system, TIERS, to serve as the technological base of the integrated eligibility program.

  • TIERS had been in development since 1997, and had been used only in a limited manner in four field offices since July 2003.
  • TIERS was not and still is not ready to be deployed statewide. The system continues to be plagued with well-known problems.
  • HHSC underestimated the magnitude of the modifications needed to make TIERS operate in a call center model.

HHSC did not allow enough time for Accenture to make the significant changes necessary for TIERS to operate in a call center model. This forced Accenture to integrate MAXIMUS' MAXe system with TIERS to provide the capabilities TIERS lacked.

These systems did not and still do not work properly, together or separately. Numerous clients report egregious problems which are described in the Problem Overview section in the Background Information attached. Despite HHSC or Accenture's corrective action plans, clients continue to provide consistent evidence that the problems haven't been fixed yet.

9. HHSC's implementation attempted to do too much at once.

  • HHSC added CHIP eligibility processing, TIERS maintenance services and health maintenance organization enrollment brokerage to the RFP, compounding the problems inherent in implementing a complex new process.
  • HHSC changed core contractors during the transition to integrated eligibility, losing valuable expertise at a critical moment.
  • HHSC implemented 27 major policy changes to CHIP at the same time it was changing contractors, systems and business processes, causing many clients to unnecessarily lose their eligibility.

10. HHSC did not allow enough time for the vendor to prepare and introduce the system, jeopardizing CHIP and Medicaid coverage for Texas children.

  • HHSC allowed only five months for the CHIP program to change contractors, systems and procedures during major policy changes.
  • HHSC spent seven months in negotiating the final contract but gave Accenture just six months to assume responsibility for TIERS, modify it, introduce a new call center system and establish call centers with hundreds of employees. Accenture was not ready.

11. HHSC received clear indications from stakeholders that the integrated eligibility rollout should be delayed, but proceeded with it anyway.

  • As early as December 23, 2005, Accenture assessed the risk of system failures as "red," or critical.
  • HHSC ignored Food and Nutrition Service's (FNS) pleas for caution and overlooked its concerns.
  • HHSC also ignored reports of serious problems provided both by its own Independent Verification and Validation vendor and one hired by FNS.

12. HHSC decided to proceed with the integrated eligibility rollout before standard acceptance testing was completed.

In other words, Accenture was not asked to prove it was ready before starting to process eligibility applications.

  • Accenture scaled back its acceptance testing and completed it only three days before the call center rollout―and 17 days after it began accepting children's Medicaid applications―leaving the contractor no time whatsoever to correct any problems it discovered.

HHSC was simply unprepared for the possibility that the IEE implementation wouldn't work, and had no meaningful contingency plan to minimize failure.

13. HHSC refuses to allow food stamp clients to apply for benefits by telephone, despite FNS' repeated urging.

  • All this would require is a waiver submitted to FNS, which has already endorsed the proposal.

14. The rollout was plagued by staffing problems at all levels.

  • HHSC notified its employees that they were being laid off before it established that Accenture was capable of taking on the work; many of these workers left before their positions were eliminated.
  • HHSC thus had significant numbers of key positions vacant at critical times in the project.
  • Accenture's call centers were understaffed and its agents were unprepared to answer policy and processing questions.

15. HHSC has not established effective accounting and auditing controls of the contract.

  • For example, as of the end of August, HHSC had approved invoices for $22 million yet the invoices had not been sent to the Comptroller's office for payment.

16. HHSC's contract with Accenture was poorly executed and the agency has made limited effort to manage it effectively since its signing.

17. Mismanaged contractors, systems and policy changes made by HHSC took a human toll―from December 2005 to August 2006, 81,504 Texas children lost their CHIP or Medicaid health care coverage because of it.

  • Since the beginning of the contract, CHIP enrollment has plunged by 8.5 percent or 27,567 children.
  • It seems unlikely that many of these children dropped from CHIP were transferred to Medicaid. Enrollment in children's Medicaid also fell by 2.9 percent or 53,937 children.
  • In addition, there is little evidence that these children's families gained enough income to become ineligible for CHIP or Medicaid. An analysis of other similar states' CHIP caseloads does not support the theory that economic conditions are significantly better for families.
  • Sweeping changes to CHIP policies and procedures were implemented, causing many children to lose coverage unnecessarily.
  • About 87 percent of the net CHIP enrollment decline occurred in service areas with more than one HMO plan option, far out of proportion to these areas' 56 percent of total CHIP enrollment in the state. Some evidence suggests that clients were deemed ineligible if they did not reselect their HMO plan at renewal, even though they were led to believe that they would be re-enrolled automatically, without needing to reselect the plan.
  • HHSC has not required Accenture to meet any meaningful quality standards for CHIP application processing.
  • Analysis of CHIP enrollment patterns and declines is extremely difficult because neither HHSC nor Accenture tracks or reports program statistics accurately.

State oversight mechanisms for this project, such as the state's Quality Assurance Team, which oversees major technology projects, were not sufficient to prevent failure or guide HHSC's contract management.

Solutions

Since the integrated eligibility project demands immediate attention and intervention, I recommend the following strategy:

  1. To immediately address the problems, the 80th Legislature should pass emergency legislation to transfer authority and responsibility for the integrated eligibility project and the Accenture contract to a turnaround team led by a special master reporting directly to the Governor and Legislative Budget Board (LBB).

    The integrated eligibility project can be put back on course only with new leadership. To achieve any savings for the state, the project and HHSC's plans to salvage the project must be subjected to an independent technical and financial assessment.

    The integrated eligibility model fundamentally changed the way HHSC contracts with private vendors, but there was no accompanying change in the skills or abilities of the people assigned to manage the project.

    The Legislature should appoint a turnaround team of experts, led by a special master, specializing in contract management, technology, finance, legal affairs and IT project management. This team should be charged with protecting scarce taxpayer dollars.

    HHSC's commissioner and executive managers should report to the special master for this project only; all other health and human services would remain under the current lines of authority.

  2. The cost of the turnaround team should be funded by Accenture's $20.3 million in excess profits.
  3. The first order of business for the turnaround team should be to end the contract and review the Integrated Eligibility program top to bottom. While the state has had to hire back state employees to do Accenture's job, Accenture has made $20.3 million in excess profit.

    HHSC mismanaged the contract from the very beginning by notifying staff in October 2005 that they were going to lose their jobs. This was three months prior to going live with an untested system and long before they knew if Accenture could handle the workload. Now they are having to hire back state employees to solve Accenture's problems. In addition, HHSC did not create an infrastructure of trained staff skilled in outsourcing to manage the project.

    Accenture should be held accountable for its commitments through the transition period.

  4. To mitigate the risk of further wasteful spending and outsourcing failures, the Texas Legislature should create a new state office of Contract Management to establish and manage large contracts for programs and information technology services such as integrated eligibility, Medicaid claims processing and electronic benefits transfer.

    Contracting problems are not new to the Accenture contract or to HHSC. The State Auditor's Office (SAO) audits of HHSC and its departments in recent years show a pattern of poor contracting practices, inadequate contract terms and lax contract management. Indeed, numerous SAO audits of other state agencies have found significant and repeated problems in contract development and planning, procurement, management, and monitoring. Many of these problems have continued at the same agencies, despite SAO recommendations that could have been implemented in subsequent contracts or extensions of existing ones.

    A Contract Management Office (CMO) could provide all Texas state agencies with the specific capabilities they need beyond their own program expertise. The CMO would allow the state to develop and centralize contract management expertise and make it available to state agencies for the life of each contract.

    Expertise at the CMO should include:

    • legal affairs
    • large contract management
    • technology contracting
    • information technology project management
    • finance and auditing

    State agencies with large IT and outsourcing contracts should be required to use the CMO's expertise to:

    • develop requests for proposals;
    • evaluate proposals;
    • negotiate contracts;
    • manage contracts and vendor relations; and
    • manage IT project development and implementation.

In conclusion, this project has failed the state and the citizens it was designed to serve. Change is needed―change is needed now. And the change must be significant if we are to uphold our promise to save tax dollars while improving customer service to our most vulnerable citizens.

Should you have any questions, please contact Ruthie Ford, my Special Assistant for Expenditure Analysis, by e-mail at ruthie.ford@cpa.state.tx.us or by phone at 463-4263, or you can call me directly at 463-4444.

Thanks for all that you do for our great state. Please do not hesitate to call on me if I can assist you in the future.

Sincerely,

[signed]
Carole Keeton Strayhorn
Texas Comptroller

Enclosure

Required Plug-ins