Realistic Automation
& Fair Process

  The State's efforts to improve child support enforcement have been derailed by the Statewide Automated Child Support System (SACSS) -- a trouble-plagued $300 million computer system.

Automation has the potential to transform child support programs, speeding the enforcement process and making it harder for delinquent parents to hide, but it must be balanced with measures that assure fairness to all parties.

Custodial and non-custodial parents alike are frustrated by the inability to get questions answered or to resolve complaints. As automation moves forward these problems will become even more pressing.







Realistic Automation
& Fair Process

Finding 4: The attempt to automate child support casework statewide has sacrificed current financial support, has failed to put a priority on delivering the easy benefits of automation quickly and reliably and is creating due process concerns for future cases.

On the path to statewide automation the State has traded proficiency today for the promise of efficiency tomorrow. Implementing a Statewide Automated Child Support System has become such a burden that for years some counties have performed below par as resources have been diverted toward SACSS.

Automation provides two central benefits to child support enforcement efforts: It allows thousands of routine cases to be processed quickly and it enables caseworkers to reach into electronic tax, employment, drivers license and other records to find absent parents, seize assets and attach wages. Congress, recognizing the benefits of automation, mandated in 1988 that states develop automated child support systems by October 1995, with the federal government to pay 90 percent of the cost. But nine years later, only a handful of states are automated and California's troubled $300 million system is barely off the ground. Of equal importance, as technology makes the child support dragnet broader and faster, serious questions arise about how to best ensure that the system is both fair and effective.

The SACSS Saga

The Department of Social Services has looked to the Statewide Automated Child Support System (SACSS) -- a massive computer network linking the 58 counties and the State -- to solve most of the deficiencies in California's Child Support Enforcement Program. DSS maintains that SACSS will improve the performance of many counties and resolve the inconsistent statistical reporting that has compromised accountability and hampered policy makers.

While many California counties still process cases by hand, most counties had some degree of computerization even before SACSS. Those with the resources built sophisticated computer systems that track cases, tap into databases, generate forms and initiate enforcement actions.

The Governor's Child Support Court Task Force noted in 1995 that in San Francisco County the number of cases in which paternity and support orders were established increased between 200 and 300 percent in the first year after the family support bureau became automated. The number of enforcement actions in that period increased by nearly 40 percent.

The Massachusetts child support enforcement program increased collections by 45 percent between 1991 and 1996 using a similar automated system. The chief counsel described the computer's role:

With a statewide system, all California counties would gain electronic access to state and federal databases. A linked statewide system will ease some of the conflicts associated with the same case being worked in two different counties at the same time. Moreover SACSS is intended to go beyond a computerized filing system to actually automate cases -- moving them through the process with little, if any, human action.

But with $82 million spent, and a final bill expected to top $300 million, SACSS -- the centerpiece of the State's automation plans -- is mired in seemingly intractable technical problems. Many of the county family support directors doubt that SACSS can ever be made to work. And while the State struggles to get the system up and running, child support collections are paying the price.

The History

California responded to the congressional mandate to automate by contracting in 1992 with Lockheed Martin/IMS to build SACSS under the supervision of the state Department of Social Services. The cost of the project at the time was set at $152 million. The state plan called for 57 counties to be put on SACSS. Because of its huge caseload and historically poor performance, the 58th county, Los Angeles, was required by federal officials to automate in advance of the State. The Los Angeles system -- known as the ACSES Replacement System (LA ARS), also built by Lockheed -- is designed to connect with SACSS.

In 1995, with concern over costs and delays mounting, the supervision of SACSS was moved from DSS to the Health and Welfare Agency Data Center. The data center renegotiated the Lockheed contract and revised the implementation schedule. By that time, the projected cost had ballooned to $260 million -- with a $28 million state share, a $23 million county share and the balance to be paid by the federal government.

Also in 1995, with Montana the only state automated, Congress extended the automation deadline to October 1997. The penalty for missing the deadline is a sanction equal to 5 percent of block grant funding and a drop in federal funding for the automated system from 90 percent to 66 percent. DSS believes the deadline will be extended again.

California began cautiously rolling out SACSS in 1995. The smallest counties were brought on line first so that glitches could be resolved before the large counties were put on line. In January 1996, the Health and Welfare Agency Data Center suspended installation so it could reassess how the program was being implemented, and in particular how it could expand the training necessary for the county workers to become proficient on the complicated system. In November 1996, the first two large counties were added -- Ventura and San Francisco -- which brought the number of counties on SACSS to 23.

Ventura and San Francisco had been selected because they were already automated and officials believed that would ease implementation. Instead, the enforcement efforts in both counties were brought to a halt. The Ventura County family support director -- an early supporter of SACSS who also serves as president of the California Family Support Council -- summarized the problems in a January 1997 letter to the State Office of Child Support:

With counties on the verge of revolt, the State put the project on hold, stopping the roll-out in January 1997 while technicians tried to work out the bugs.

At its February 1997 annual meeting in Palm Springs the California Family Support Council, which is made up of family support directors throughout the state, passed a resolution urging the DSS to explore alternatives to SACSS.

As doubts about SACSS increased, the State contracted with a consultant -- Logicon Inc. -- to determine whether the system could be fixed. In February 1997 Logicon reported its tentative conclusion: that SACSS is salvageable, but only if Lockheed can resolve some 1,400 remaining technical problems.(106)

As of April 1997, the future of the project remains a question mark. Lockheed is working through a corrective action plan that calls for problems to be resolved according to a specified schedule. Logicon recommended that the vendor's progress in meeting that schedule be reassessed in late May 1997.

Meanwhile, some of the system's most important functions -- locating addresses for absent parents, automatically generating forms, linking with automated databases and processing account information -- are not working properly. The locate function has been turned off until problems can be resolved, the forms function is slow and account information is plagued with errors. Counties also say SACSS keeps repeating information already known and antagonizes employers by billing repeatedly for wage assignments.

San Francisco's family support director summarized the county's problems:

Even if SACSS were working properly, counties like San Francisco that already had highly functioning computer systems regard SACSS as a step back into the cybernetic Dark Ages. The counties report that the system is awkward, glacially slow and inconsistent in what it requires of the user. They complain that it has almost 400 difficult-to-read screens and requires cumbersome maneuvers for even the simplest functions.

The problems have counties vying to be the last to connect to SACSS. Counties already using SACSS say the system has brought their collections to a standstill. Those not yet on the system are having to divert resources away from enforcing child support to getting ready for SACSS -- training staff in SACSS procedures, foregoing upgrades to existing systems and spending months recoding cases for transfer onto SACSS. The limbo effect is severe for counties like San Bernardino -- which has separate, incompatible systems for case management and accounting, neither of which can communicate with the county welfare department computer system.

The Department of Social Services has responded to the problem by granting "hold harmless" status to counties that fail performance reviews because of SACSS. In the 1995-96 performance review, 12 counties fell into that category.

Despite the years of controversy, Department of Social Services officials have been unflagging in their optimism about SACSS. In January 1996 the chief of the state Office of Child Support said:

As late as October 1996, the Director of the Health and Welfare Agency Data Center testified that he is confident SACSS will be implemented in time for California to meet the October 1997 federal deadline.(109)

While now it is almost certain that the deadline will be missed, DSS officials report that their federal counterparts have indicated the October 1997 deadline will not be rigorously enforced. The federal law provides 90 percent reimbursement for systems that are certified by the deadline and 66 percent reimbursement for systems that are certified after the deadline. The issue now, however, is how much if any of the project costs the federal government would assume if the State decided the California's children would be best served if SACSS were scrapped and a better system installed.

Among the Problems

Part of the blame for the difficulties with SACSS lies with federal legislation. In a well-meaning attempt to save money, Congress required states to use existing technology to build the automated child support systems -- requiring states to adapt systems already in use. The effect of that mandate was to render SACSS obsolete before the procurement contract was signed.

A second factor rests in California's county-based system of administering child support. Although Congress mandated states to build statewide systems, the federal government bowed to California's desire to preserve county autonomy by allowing the State to build a county-linked system, providing it could show that a linked system would work as well.

In practice, that made implementing SACSS a nightmare. Because every county has its own existing system and its own level of technology, SACSS has had to be shoe-horned to fit each county. The director of the Health and Welfare Agency Data Center testified that the original SACSS bid assumed there would be a high degree of consistency among the county automated systems and that SACSS technicians would have to write fewer than 30 conversion programs. In fact, he said, more than 50 conversion programs will have to be written.(110)

Faced with similar problems in the past, the State has created a mechanism for scrutinizing automation projects. After a $44 million computer system at the Department of Motor Vehicles was scrapped, the Department of Information Technology (DOIT) was established in 1995 by SB 1 (Alquist) and granted the authority to suspend or terminate information technology projects. DOIT's job is to monitor projects and to work collaboratively with the department involved to mitigate risks to the State.

DOIT officials said they have monitored the SACSS implementation, and were part of the decision to suspend implementation in early 1997 and to hire an outside consultant to assess the system's viability.

More generally, DOIT officials said SACSS is one of six large computer systems, representing investments worth hundreds of millions of dollars, that have some welfare-related applications that must be changed to conform to federal welfare reforms. DOIT is concerned about whether some of the systems, including SACSS, can be completed at the same time they are modified to perform new functions in time to satisfy federal deadlines.

DOIT officials said they will make a decision about SACSS by mid-summer 1997 and that decision will be guided by the progress Lockheed has made on the corrective action plan. DOIT officials said one problem in making that assessment is a shortage nationwide of technical experts capable of evaluating large complex systems.

California is rich in expertise, and state officials do have at their disposal a tool for harnessing public and private industry expertise to help make important and difficult technology-related decisions. The California Council on Science and Technology is one conduit for that expertise. The Council, which is made up of representatives from public and private universities and colleges, was set up by the Legislature in 1988 to analyze public policy issues and provide recommendations in the area of science and technology. A list of the members is included in the Appendices.

While the Logicon review focused on the technical viability of SACSS, how or whether to proceed with SACSS will require a broader judgment call -- whether it can be made to work efficiently at a reasonable cost and in a reasonable time frame.

Central Case Registry

While state officials have struggled to implement SACSS, the State has forgone significant benefits that could have been realized from basic computerization. The Legislature in 1992, 1993 and 1994 directed DSS to develop a centralized case registry that would provide a single source of information for all cases -- names, addresses, dates and the amounts of orders.

The central registry would be a comparatively simple computerized tool that would provide counties with a reliable and unified source of fundamental information about welfare-related and non-welfare related cases. The California District Attorneys Association and public advocates have long agreed that a case registry would ease some of the problems associated with a county-based enforcement program and would be needed even after SACSS is fully implemented.

The department, in a feasibility study ordered in 1993 by the Legislature and completed in 1997, found that the registry would improve the State's compliance with federal laws, reduce duplicated efforts and unnecessary work by county family support divisions, and improve enforcement actions and collections -- all without significantly increasing costs. The feasibility study concluded the registry would cost $2 million to construct and $3.5 million a year to operate.

Despite the low costs and high benefits, the department has put the project off -- primarily because of its desire to implement SACSS first. The federal welfare reforms, however, are now requiring the State to do something that program directors and policy makers have known for a long time makes sense.

A Question of Due Process

In the last decade, child support policies have focused on using all of the resources available to the government to catch missing parents who are hard to find and reluctant to voluntarily comply with child support orders. The reforms have focused on both the front end and the back end of the process.

To expedite order establishment, district attorneys have expanded their use of "default judgments." After being served with a summons and complaint to establish paternity and a support order, a non-custodial parent has 30 days to respond. If the parent fails to respond, the court can enter a default judgment establishing paternity, ordering support to be paid based on estimated earnings and attaching wages without the non-custodial parent ever showing up in court.

There is little sympathy for the citizen who ignores a summons. But once legal paternity is established the action of the court is permanent. And once declared the father, whether factually accurate or not, the financial obligations last until the child reaches 18, and the debt that accumulates cannot be relieved.

The Child Support Task Force estimates that statewide 50 percent of the orders are established by default. In some counties, as many as 80 percent of the orders are established through default. And the support obligations are retroactive to the time when cases are filed in court. So it is not uncommon for a non-custodial parent -- if they ignore the summons -- to be thousands of dollars in debt before the enforcement tools kick in.

At the enforcement end, the government can take away licenses and divert lottery winnings, tax returns and worker compensation payments. It can seize bank accounts, real property and personal property. Some of these actions can be taken with lower notice requirements than were used to establish the order.

Winning political support for these harsh consequences has not been difficult for program directors. Irresponsible parents contribute to the poverty of children and the swelling of government debt -- and seldom show up to defend themselves in public forums.

But as the child support dragnet becomes wider and more efficient, the chances increase for errors to be made and unintended consequences to develop. These enforcement tools have the ability to deny rights and privileges that allow parents to earn an income and impose procedural costs that could drain away resources that might otherwise go to pay child support.

The driver's license match program is a good example. The Department of Motor Vehicles matches the names of delinquent parents with licensed California motorists. To those motorists, DMV sends certified letters informing them their license will be revoked in 150 days if they do not make arrangements to pay owed child support and file the correct paperwork to stop the revocation. DMV also sends a temporary license -- even though technically the permanent license is not revoked unless the motorist fails to take action.(111)

Of all the certified mail sent by the department, 40 percent is returned as undeliverable, presumably because the person moved without notifying DMV. Nevertheless, the revocation process continues. DMV and DSS officials acknowledge that in many of those cases, the motorist does not know that the license will be revoked, and often finds out after it has been revoked. Some motorists have been notified by their insurance companies, which have refused to renew the policy because of the revocation. Some insurers -- including one large auto insurer -- have treated the first notice as a revocation, refusing to renew insurance policies even in cases where the parent has agreed to a payment schedule and the license was never revoked. DSS is aware of these unintended and unanticipated consequences. But more importantly, the department did not take the initiative to work out the problem, either directly with the insurers or with the state Department of Insurance.

Default judgments and enforcement tools further the goal of getting needed financial support to children. But along with making child support enforcement better and faster, the State should affirm its commitment to be fair -- to hear complaints, identify errors and streamline remedies. A sense of fairness is essential to maintaining public confidence in government and in the child support enforcement program in particular.

Finding Balance

There have been some efforts to balance the heavy hand that automation and default judgments can bring with provisions to reopen or set aside decisions once non-custodial parents fully understand the seriousness of their obligations and the government's commitment to enforce them.

SB 1058 (Speier), which implemented the Court Task Force recommendations, allowed for judges to use "imputed" or estimated earnings in setting an order. Estimating income is necessary to impose wage assignments when the judge does not know how much a parent earns. The due process was provided by allowing non-custodial parents to object to the order and seek a modification within 90 days of the first wage assignment.

The next step down this path of accelerated order establishment would be to provide for service of the original summons and complaint by mail. Given the ease of obtaining default judgments, the highest hurdle in securing an order has become serving the alleged parent with legal notice. Under existing law, the district attorneys must have the non-custodial parent personally served with the summons. Under some conditions, the summons can be left with a roommate or spouse. And in cases where the DA can show the parent is avoiding service, the legal notice can be accomplished by publishing the information in the newspaper.

Many district attorneys want the ability to provide service with first class mail using DMV addresses -- the same addresses that DMV knows are inaccurate in 40 percent of the cases. Practically speaking, many alleged non-custodial parents would not know that they were sued for paternity and financial support until after the court had found against them and their wages are attached. Proponents believe that service by mail can be made fair by providing lenient rules for reopening cases within a period of time after a wage is assigned as in cases where wages are imputed.

The Value of Notice

The Franchise Tax Board has demonstrated that when consequences are plain, people are more likely to respond. Among the most cost-effective tools the FTB has used in collecting child support has been the seriously worded demand letter that tells parents to pay up, or else. In fiscal year, 1995-96 the board sent out 170,000 such letters to parents who had skipped out on child support; 8,100 responded by making payments.

The Legislature has recognized the need for clear notice to custodial parents about future court hearings. The intent of the provision was to give custodial parents a chance to advocate on their behalf. In unusual specificity, the Legislature stated the precise language and even the size of type (14-point) that must be used:

Similarly, the Governor's Court Task Force concluded that despite efforts in this regard, the process was still not simple enough. It recommended:

Adequate notice encourages the participation that is fundamental to a fair legal system. The drivers license match program, for instance, was approved by policy makers with the understanding that parents would have an opportunity to pay back support before losing their license.

But of equal importance, the most cost-effective use of the numerous enforcement tools that have been enacted recently would be a deterrent to delinquency. Voluntary compliance is far cheaper than enforcement. Some counties have made some effort to tell non-custodial parents with newly established orders all of the potential consequences involved in falling behind in support payments -- from the accrual of interest to liens against property, to revocation of professional licenses. The FTB has demonstrated the value of clearly worded notices that actually reach non-custodial parents. Comprehensive efforts to let non-custodial parents know all of the risks and costs involved in not making support payments would make the system more fair and could be expected to increase compliance.

Resolving Complaints

One of the traditional problems plaguing counties has been the thousands of calls that inundate family support divisions each month -- many of the calls coming from the same parents, calling repeatedly to find out the status of a check.

Critics maintain that if the district attorneys were more efficient, fewer parents would have to call so many times -- to either get action taken in their case or to inquire about a support payment. The district attorneys complain that they could put more resources into processing cases if caseworkers did not have to spend so much time on the telephones.

One benefit of local automation efforts has been the ability of family support divisions to install voice response units (VRU) -- sophisticated answering machines that allow parents to call and check on developments in their cases. SACSS is suppose to provide this service when it comes on line. The automated information systems give caseworkers more time to deal with individuals whose questions cannot be answered by the VRU. In Los Angeles County a special team of operators is assigned to handle case inquiries from the politically connected -- the district attorney's main office, the mayor's office, legislative representatives.

But advocates for children and parents maintain that voice response units provide inaccurate information and cannot by themselves resolve the communication problems between authorities and parents. They want a process that ensures their complaints are heard and their issues resolved. State and county officials maintain the existing complaint procedures are adequate. But there is no state policy or process that allows for parents whose cases have languished for months or years to determine if the district attorneys have done all they could or should to enforce the law in their cases.

A uniform complaint process -- with DSS or the Attorney General in the information loop or even acting as an independent reviewer -- would improve service to custodial parents, increase accountability and provide state program managers with another source of information about the effectiveness of a given county and the program overall. A uniform process also could provide a venue for resolving complaints from non-custodial parents -- whose frustrations with the process can reduce voluntary compliance.

Summary

Automation is essential to managing millions of child support cases. But after significant effort and cost, there is evidence that SACSS will never perform as intended. To that end California needs to think about the possibility that SACSS will never work and to find ways to meet basic automation needs first. In addition, as automation does become more efficient, and as millions of Californians get involved one way or another in the child support enforcement program, procedures ensuring fairness will be essential to maintaining public confidence and support for the program.

Recommendation 4: Given the high stakes involved in child support, the State should prepare for the possibility that SACSS will never function properly. The State also should rigorously review the existing oversight provided by the Department of Information Technology. And the State should craft policies that enhance automation while maintaining basic fairness.

The frustrating reality is that several counties in California, independently of SACSS, have automated routine steps in securing and enforcing child support orders. What those counties needed -- and what eventually all counties could have benefited from -- was a centralized case registry and easy access to other databases that can provide information on the location of missing parents and their assets. The State was led down the road to SACSS with specific directions from the federal government, but that does not mean that it cannot pro-actively devise strategies that will meet California's business needs. Specifically, the State should take the following measures:







Next Section   Previous Section   Table of Contents