The INS now expects to spend at least $2.8 billion on its automation programs. Despite uncertain funding sources and rising costs that could significantly delay completion of its automation projects, INS still had not developed a contingency plan to ensure that mission-essential programs are implemented. In addition, the ultimate cost for INS' automation programs was uncertain because actual costs incurred were unreliable and projected cost estimates were unsupported. Further complicating these problems was the breakdown of OIRM's primary cost reporting system for its $2.8 billion automation programs.
In our initial audit, we found that INS expected to spend about $2.6 billion on its automation programs through FY 2001, with an additional $260 million required to complete these programs beyond FY 2001. We observed that if the additional funding did not materialize, rising operations and maintenance (O&M) costs could significantly delay completion of projects in support of automation programs. We also questioned the accuracy of funding and cost data reported by INS in its quarterly funding profiles.
To fix these problems, we recommended that INS develop a contingency plan to ensure that at least its mission-essential automation programs, which should be its top priority, are implemented. We also recommended that INS assess the benefits of submitting semiannual, as opposed to quarterly, funding profiles. We further recommended that INS maximize funding of the automation programs to the fullest extent possible through the use of user fees.
The INS responded by stating that through a combination of in-place and planned management reporting mechanisms, it would provide the comprehensive management strategy necessary to ensure that essential programs were implemented. It also agreed to work with the Justice Management Division to consider the benefits of semiannual reporting, and further stated that the extent to which automation programs would be funded from user fees would be based on the results of a planned FY 1998 cost allocation study.
During our current audit, we found that as of June 1997, INS expected to spend about $2.6 billion on its automation programs through FY 2001, and an additional $234 million beyond FY 2001, for a total anticipated cost of about $2.8 billion. In our judgment, in view of the fact that uncertain funding and rising operations and maintenance costs could significantly delay completion of its automation projects, INS still needs to develop a contingency plan to ensure mission-essential programs are implemented.
We also found that the ultimate cost of INS' automation programs is uncertain because actual costs incurred are unreliable and projected cost estimates are unsupported. Further complicating these problems is the breakdown of OIRM's Integrated Budget Execution (IBEX) system, INS' primary cost reporting system for its $2.8 billion automation programs. Additionally, management solutions for the cost reporting problems are questionable. Further, although OIRM made some effort to transition to semiannual cost reporting, progress has been stalled due to personnel changes and the IBEX system breakdown. Finally, the proposed OIRM cost allocation study should ensure that funding of the automation programs from user fees is maximized to the fullest extent possible.
A. INS Reported Costs for Automation Programs
Costs for the automation programs include: (1) investment costs consisting of one-time expenditures to develop and deploy the automation programs, (2) recurring O&M costs consisting of costs for supporting and maintaining automation programs after they are brought on-line, and (3) overhead costs such as those for OIRM operations.
Based on the June 1997 OIRM quarterly funding profiles, INS expects to spend approximately $2.8 billion on its automation programs through FY 2001 and beyond. Of this amount, approximately $1.1 billion will be spent on investment costs, $1 billion will be spent on O&M costs, and $.7 billion on overhead costs.
The June 1997 funding profiles also reported that INS plans to spend approximately $203 million on investment costs during FYs 1999 to 2001.5 However, based on the same report, INS projects that it will need a total of $437 million in investment costs to complete its automation projects. As a result, it appears that after FY 2001 INS will still need $234 million to fully develop and deploy its automation programs.
B. Reliability of Reported Costs for Automation Programs
To be successful, INS' automation programs must be based to a large extent on decisions driven by accurate and up-to-date cost information. Using a combination of actual costs incurred and future cost estimates, INS managers must make critical, ongoing decisions as to whether systems under development will remain economically viable. Actual costs used in making these decisions must be generated from reliable information systems. Estimated future costs must be documented and supported by reasonably accurate projections of workload and capacity requirements.
During our current audit, we found that INS' reported investment costs for its automation programs were either unsupported or generated from unreliable sources. For instance, of the $437 million that INS projects to spend on investment costs during FYs 1999 to 2001, $363 million (83 percent) was completely unsupported. The OIRM management insisted that supporting documents existed but could not be located because the individuals who prepared them had since left INS. Management added that poor filing by administrative staff also contributed to this problem.
In our judgment, management's explanations for these missing documents are unacceptable because project managers are responsible for maintaining these critical documents in project folders. Further, since these documents must be continually updated to support ongoing management reviews and decisions, the absence of the documents further attests to management's lack of adequate oversight of the system development process.
We also found that OIRM's compilations of actual costs incurred for the automation programs during FYs 1995 to 1997 were unreliable. This occurred because the existing account code structure did not delineate between costs for automation programs and other expenditures. Moreover, during FY 1997, OIRM's unique IBEX system proved to be increasingly unreliable in generating accurate cost data. According to a September 29, 1997 internal OIRM memorandum, the IBEX system was considered a failure and was ultimately abandoned two weeks prior to the close of FY 1997. This action left INS without its primary cost reporting system for its $2.8 billion automation programs.
According to OIRM senior management, INS' solution to this problem includes a totally refurbished "IBEX 98" system based on Access 8.0 software. However, according to the same September 29, 1997 memorandum referred to earlier, user needs and suggested enhancements seem to have been ignored in developing IBEX 98. The memorandum further states that if this is so, then IBEX 98 will generate the same inaccurate data as before -- only much faster.
C. Allocating Funds to Ensure Implementation of Essential Systems
In our judgment, government agencies should develop a funding allocation process that ensures implementation of critical or mission important automation projects. Regarding INS' automation projects, this requirement is all the more challenging due to the uncertainty associated with future funding. According to INS' June 1997 funding profile, annual funding levels are projected to remain at a constant $394 million until FY 2001. The single largest source of that funding, 46 percent annually, will be provided by the Violent Crime Reduction Trust Fund (VCRTF). However, OIRM senior managers expressed concern to us that VCRTF funds may not be available beyond FY 1999.
Adding to the funding concerns is the fact that as more systems are brought on-line, annual O&M costs will increase accordingly. In turn, as more funds are allocated for higher priority O&M costs, fewer funds will be available for system investment (development and deployment) costs. In our judgment, the uncertainty of future VCRTF funding, the increasing O&M costs as more systems are brought on-line, and the estimated $234 million in additional investment funding required beyond FY 2001, all seriously threaten significant delays in the full implementation of the automation projects. For these reasons, we believe INS must develop a contingency plan to ensure full implementation of mission-essential automation projects.
The INS, however, maintains that a contingency plan is unnecessary since it has other alternative measures in place that address these concerns. These alternative measures include a contractor assessment of INS' automation programs, a one-page guide of parameters for the OIRM budget execution plan, and the creation of an information technology investment review board. We believe that none of these alternative measures provides adequate assurance that INS will develop the funding allocation process needed to implement essential automation projects.
For instance, the contractor assessment of INS' automation programs is limited to only 14 out of a potential universe of 96 INS automation projects. The assessment is silent as to how funds should be allocated, nor did we see any evidence that OIRM is using the assessment as a basis for allocating funds. In addition, the one page guide of budget parameters is for FY 1998 only, and the list contains no guidance as to how INS should allocate funding to ensure implementation of mission-essential automation projects. Finally, we support OIRM's decision to establish an investment review board to provide a "formal mechanism for key organizational decision makers to plan and request funding for IT [investment technology] investments." While OIRM did hold several meetings of the board during FY 1998, none of the meetings specifically dealt with developing a contingency plan to ensure implementation of essential systems. Consequently, INS still has no assurance that mission-essential automation projects will be implemented in the event of funding shortfalls.
D. Semiannual Funding Profiles
The Information Technology Management Reform Act of 1996 requires INS to establish information systems that are designed, developed, and used effectively to compile actual costs incurred for inclusion in financial statements. Costs incurred must be provided on a reliable, consistent, and timely basis. Within OIRM, the IBEX system has been used since FY 1995 to compile the actual costs incurred by INS in developing its automation projects. Until June 1997, compiled costs were reported in OIRM's quarterly funding profiles, which assist INS managers in making decisions on the continuation, modification or dismantlement of its automation projects. However, with the IBEX system failure, cost reporting was suspended indefinitely.
In our initial audit, we questioned the accuracy of funding and expenditure data reported by INS in its quarterly funding profiles because INS estimates for the costs of the automation programs fluctuated significantly between reporting periods, and automation program expenditures, as reported by INS, were not comparable to the sum of individual project costs within a program. We recommended that the INS Commissioner assess the benefits of reporting cost data semiannually at the conclusion of the first and third quarters. By doing so, INS would report fewer cost fluctuations for the automation programs and reduce the OIRM administrative burden.
The INS responded by stating that cost fluctuations between reporting periods occurred because the numbers represent the most accurate budget figures available at the time; nevertheless, they did agree to work with JMD to consider the benefits of semiannual reporting. Since then, OIRM has decided to implement semiannual cost reporting. Our discussions with OIRM managers disclosed that the basic format for the semiannual reports has been agreed upon; however, beyond that, progress appeared to have stalled. While OIRM was responsible within INS for preparing funding profiles, recent personnel changes within OIRM and the INS Management Resource Staff obscured responsibility for preparing the funding profiles. At the time of our audit, neither the OIRM Deputy Associate Commissioner nor the Management Resource Staff Director knew who would ultimately be responsible for completing preparations for semiannual cost reporting.
Complicating the issue was the lack of FY 1997 actual cost information because of the IBEX system breakdown. Since funding profiles are based on cumulative costs, semiannual reporting could not begin until FY 1997 transactions were manually reentered in the refurbished IBEX 98 system. At the time of our audit, the OIRM managers were not sure how long that would take.
As a result of the delays in producing timely and accurate semiannual funding profiles, INS continues to lack the critical financial data needed to make informed decisions about the future of its $2.8 billion automation program.
E. Maximizing User Fees
In our initial audit, we recommended that INS maximize funding of the automation programs through user fees. The INS responded by stating that it recently completed a fee study that established a rate-setting methodology. In addition, INS plans to perform a cost allocation study in FY 1998 to determine the appropriate allocation of costs to the various fee-related accounts.
Based on our review of the statement of work for the proposed cost allocation study, we believe the study should, when implemented, sufficiently address the concerns cited in our prior report. The study, however, will not be completed until FY 1999, with implementation of the results probably not occurring until at least FY 2000.
F. Recent INS Developments to Initiate Corrective Actions
Subsequent to completion of our field work, INS stated that it does not believe that a contingency plan is the answer to the problem of funding uncertainties. Instead, INS says it has implemented new, industry-accepted Information Technology (IT) investment management practices that are designed to ensure that available IT funding is used for the highest INS priorities, and to support its most critical mission needs.
The INS also stated that the IT Investment Review Board (IRB) is now in the process of finalizing FY 1999 IT allocations, and is responsible for setting INS-wide priorities for all INS IT spending. This board is made up of the INS Deputy Commissioner and the Executive Associate Commissioners, and will ensure that informed decisions will be made about how to allocate increasingly scarce IT resources at the highest levels of INS management. The INS further stated that its Office of Budget is now requiring IRB approval for all requests for IT budget enhancements, beginning with the upcoming FY 2001 spring budget call. This means that future IT budget enhancement requests will have to be consistent with overall INS priorities as established by the senior agency officials who are members of the IRB.
Regarding its IBEX system, INS stated that technical budgetary problems not related to IBEX prevented INS from using the IBEX system to track expenditures at the end of FY 1997. According to INS, these problems were resolved, all the data for FY 1997 was subsequently entered, and the system continues to be an important tool.
We recommend that the Commissioner, INS:
At the time of this audit, the INS was unsuccessful in procuring independent audits of its Information Technology Partnership (ITP) contract costs. However, subsequent to our audit, INS stated that the Defense Contract Audit Agency (DCAA) has completed several audits resulting in over $2.5 million in credit adjustments and/or proposed rate reductions. Additionally, INS made significant improvements in obtaining timely, written task order authorizations. In addition, INS took sufficient actions to plan for the expiration of its major contract for the automation programs.
The INS relies heavily on contracting to develop and maintain projects associated with its automation programs. As of January 30, 1998, INS had 29 active contracts with awards totaling over $1.6 billion. The total obligations and expenditures for these 29 contracts exceeded $684 million.
In August 1994, INS awarded the ITP contract to Electronic Data Systems (EDS). This contract, which provides for technical services in all stages of systems development, had a total estimated award value of $342.2 million, which is the largest of the 29 active contracts. The contract is a cost reimbursement contract providing for the contractor to be paid for a level of effort, which is generally presented in labor hours, rather than in terms of a finished product. Under this type of contract, the scope of the work required is described in general terms and the contractor is only obligated to devote a specified level of effort for a specified time period. This type of contract requires a high degree of oversight and monitoring by INS.
In May 1998, INS awarded a number of new contracts as a follow-on to the current ITP support services contract with EDS. The ITP contract with EDS will end after a 90-day transition to the new contracts. The new contracts, referred to by INS as the STARS contracts, consist of a systems management/integration contract, multiple development and maintenance contracts, and an independent verification and validation contract. The STARS contracts will consist of a mixture of cost reimbursement and fixed price contracts.
A. Monitoring Contract Costs
In our prior audit, we noted that INS had taken action to correct weaknesses in verifying contractor invoices prior to payment. We also noted that INS also needed to ensure that the EDS contract was adequately audited. Information provided by independent audits helps INS determine whether contractor invoices should be paid. In our judgment, the scope of prior DCAA audits was not sufficient to determine the accuracy of billings from EDS. Accordingly, we recommended that INS ensure that future audits of contract costs are procured in a timely manner to determine if charges are allowable, allocable and reasonable.
During our current audit, we determined that despite its most recent efforts, INS is still unsuccessful in procuring DCAA audits of its ITP contract costs. Our review of OIRM records indicated that in June 1997, INS forwarded a statement of work to DCAA under the provisions of an existing DCAA-INS interagency agreement. In September 1997, DCAA responded by proposing several alternative audits, along with its detailed rationale of why it believes the alternative audits were necessary. However, as of March 1998, INS still had not reached an agreement with DCAA on the scope of audit work to be performed. Although INS conducts its own reviews of contractor performance, it had no assurance from independent audits concerning the validity of ITP contractor invoices totaling $280 million as of December 30, 1997.
B. Task Order Approvals
Under the Federal Acquisition Regulation, INS is required to closely monitor cost reimbursement contracts to ensure the validity of payments to contractors for work performed. One activity that must be closely monitored involves proper approvals of contractor task orders. To control this critical aspect of contractor activity, INS' original contract with EDS stated that: (1) the contractor is not allowed to incur costs without the express written direction of the Contracting Officer, and (2) the task orders shall be reviewed and approved by the Contracting Officer prior to issuance to the contractor.
In our prior audit, we determined that INS risked paying contractors for unwanted or unnecessary work because INS had allowed contractors to begin and sometimes complete work on task orders without prior approval. Our review of 79 FYs 1994 and 1995 task orders revealed that 96 percent were not formally approved prior to the task order start date. To fix this problem, we recommended that INS ensure that task orders are approved in writing prior to initiating work.
The INS' remedy to correct this problem began with a January 1997 modification to the EDS contract that allowed the contracting officer to place oral orders under the contract as allowed by the Federal Acquisition Regulation. In effect, this modification allowed the contracting officer to verbally approve task orders prior to signing them. However, INS also stressed that written task order approvals should be obtained as early in the process as possible.
Our review of 17 FY 1998 task orders showed significant INS improvement in obtaining more timely approvals since our last audit. On average, initial approvals occurred 35 days prior to the task order start date, and final approvals occurred, on average, 79 days after the task order start date. In contrast, our review of 14 FY 1997 task orders showed that, on average, initial approval occurred 59 days after the task order start date, with final approval not occurring until 159 days after the task order start date.
In our judgment, this significant improvement in the timeliness of INS' contractor task orders should minimize the risk of unwanted or unnecessary work being performed by ITP contractors.
C. Contract Cost Management
A cost reimbursement contract provides for the payment of allowable incurred costs to the extent prescribed in the contract. However, a contract cost ceiling must be established in each contract to prevent the contractor from exceeding the ceiling without the prior approval of the contracting officer.
In our initial audit, we found that although the ITP contract was awarded for a one-year period with four one-year options, INS was approaching the total contract funding ceiling of $295.2 million after only three years. To avoid exceeding the contract funding ceiling, JMD recommended that INS move tasks from the existing EDS contract to other contracts. We expressed concern that INS could incur significant delays and added costs in moving these tasks, and recommended that INS assess the impact of moving tasks from the EDS contract.
During our current audit, we determined that INS took sufficient actions to move the tasks to other contracts without exceeding the EDS contract funding ceiling or incurring significant delays in completing project tasks. The INS' actions, which in effect increased the contract capacity of the information technology partnership by $170 million, included the following:
In our opinion, these actions appear sufficient to allow work under the contract to continue until the STARS transition without exceeding the EDS contract ceiling. As of December 31, 1997, the contractor's estimate of the total cost to complete all task orders under the contract was approximately $318 million, which was well below the total revised contract ceiling of $342.2 million.
D. Recent INS Developments to Initiate Corrective Actions
Subsequent to completion of field work, INS stated that since our audit, DCAA has completed several incurred cost, rate adjustment, and specific ad hoc audits resulting in over $2.5 million in credit adjustments and/or proposal rate reductions.
We recommend that the Commissioner, INS:
The INS had not implemented adequate safeguards to ensure the accuracy of existing data to be used by systems currently being developed or re-engineered. Moreover, INS had not implemented adequate safeguards to ensure the adequacy of future data inputs. As a result, existing or new INS systems could rely on inaccurate or unreliable data accessed from existing databases or future data inputs.
In our prior audit, we determined that INS had not implemented adequate safeguards to ensure the accuracy of existing data to be used by systems currently being developed or re-engineered. We noted that INS did develop several promising proposals and guidelines to help ensure data integrity; however, we expressed concern that these activities were not fully implemented at the time of our review. As a result, some of INS' newly developed or re-engineered systems were in danger of using incomplete or inaccurate data from existing systems. To fix this problem, we recommended that INS ensure the accuracy and reliability of data entered into or accessed by new systems.
The INS did not concur with our finding and stated it was actively addressing the data accuracy issue. This was evidenced, they stated, by ongoing re-engineering efforts that would ensure the accuracy of day-forward data inputs, and by the Records and Processes Improvement Design (RAPID) project that would ensure the accuracy of already existing data.
In our current audit, we determined that INS still had not implemented adequate safeguards to ensure the accuracy of existing data to be used by systems currently being developed or re-engineered. We further determined that it had not implemented adequate safeguards to ensure the accuracy of future data.
The INS' efforts to improve the accuracy of existing data focus almost entirely on the RAPID project. With a five-year life span and a projected cost of $133 million, the project would include a reconciliation of hard copy alien files with the corresponding computer files to ensure that the computer files contain accurate data. However, as ambitious as the project sounds, it falls short of a comprehensive solution to INS' data integrity problems. Additional solutions, such as those recommended in a September 1996 Coopers & Lybrand contract study, must also be addressed.6 Moreover, implementation of the RAPID system is still uncertain since project funding had not been obtained as of March 1998. Currently, INS has no alternative strategy to improve the accuracy of existing data.
We were told by INS managers that efforts to improve the accuracy of future data focus on re-engineering efforts that employ various technologies. However, despite our repeated efforts, we were unable to obtain from OIRM personnel a clear explanation of how these technologies will be applied, which systems are included, what the time frames are for implementation, and what the costs will be.
Computer specialists from the OIRM Corporate Systems Branch referred us to INS' Five Year Reengineering Plan to obtain answers to these questions. Our review of that plan disclosed that it did not contain information specifically related to data integrity or reliability. Other OIRM personnel referred us to the FY 1997 Results of the INS Commissioner's Priorities Report. Although each of the Commissioner's seven priorities contained data integrity objectives, there was nothing in the report that addressed INS' specific plans to ensure the accuracy or integrity of future data.
The OMB Circular A-130 requires government agencies to establish and maintain strategic information resources management planning processes which preserve data integrity. Since INS lacks such a documented planning process, some of its existing, newly developed, or re-engineered systems are in danger of using incomplete or inaccurate data.
The Executive Associate Commissioner for Management told us that the integrity and reliability of INS data was a valid area of concern but that solutions to existing problems would need to be addressed across the INS organizational structure.
We recommend the Commissioner, INS:
The INS took significant steps since our previous audit to better manage its inventory of IDENT equipment. As a result, on-hand quantities of IDENT equipment, which totaled $2.4 million in February 1997, were reduced to about $975,000 by the end of January 1998. However, we did note that IDENT equipment costing $520,000 remained in the INS inventory for more than one year. This occurred because equipment items were not consistently shipped from the INS warehouse inventory in the first-in first-out (FIFO) sequence. Further, technically obsolete equipment had not been exchanged for newer equipment.
In our initial audit, we noted that INS maintained a larger inventory of IDENT equipment than was necessary. This occurred because IDENT equipment was purchased too far in advance of deployment dates. Seventy-six percent of the IDENT equipment inventory, with an estimated value of about $1.8 million, was stored for six months or more prior to deployment. Purchasing equipment too far in advance of scheduled deployment dates can result in unnecessary storage costs, increased maintenance costs due to expiration of the warranties before the equipment is deployed and operational, and obsolescence of equipment prior to deployment. We recommended that INS ensure that equipment purchases are coordinated with scheduled deployment dates to reduce the aged surplus.
In our current audit, we determined that INS took significant steps to better manage its IDENT equipment inventory. We found that the IDENT inventory, which totaled $2.4 million in February 1997, was reduced to approximately $975,000 by the end of January 1998. To reduce on-hand balances, the IDENT project manager required the warehouse contractor to follow FIFO inventory procedures, instituted procedures to monitor the contractor's automated inventory reports, and minimized new purchases of IDENT equipment.
On the other hand, IDENT equipment costing $519,764 remained in the inventory for longer than one year. This occurred primarily because the warehouse contractor did not consistently ship equipment to IDENT field locations in the FIFO sequence. During October 1, 1997 through January 28, 1998, the warehouse contractor shipped 978 IDENT inventory items. However, 74 percent of these items were shipped when older but identical items remained in the inventory as of the shipment dates.
Contributing to this problem was the fact that some technically obsolete equipment had not been exchanged for newer equipment. We learned from the IDENT project manager that the Videolabs Flex Cameras, version 3.1, were no longer technically compatible with the ENFORCE/IDENT system and needed to be exchanged for the newer version 6.0 of the cameras. The manufacturer agreed to exchange the cameras, and the IDENT Project Manager had planned to send the cameras back to the manufacturer in December 1997. However, as of January 28, 1998, 163 obsolete cameras costing $57,213 had been in the INS inventory for more than one year. Although the manufacturer did not state how long the exchange offer would remain in effect, prudent business practice dictates that INS should act upon the manufacturer's offer as soon as possible.
Had INS followed FIFO shipment procedures and exchanged the obsolete cameras for new cameras, the on-hand balances of equipment that remained in the inventory for longer than one year would have been reduced from 726 items costing $519,764, to 88 items costing $113,228. Also, by following FIFO inventory procedures, INS would have greatly minimized the risk of deploying equipment for which the manufacturers warranty had already expired.
A. Recent INS Developments to Initiate Corrective Actions
Subsequent to completing our field work, INS stated that it has continued to make progress managing its equipment inventory. Specifically, the contractor managing the Largo, Maryland facility enhanced the software used to manage the inventory. The inventory software is now capable of providing project managers with accurate, up-to-date information, and generates deployment documents automatically based upon first-in, first-out rules. These G-504s are automatically populated with the oldest equipment of the type requested, and warehouse personnel use these forms to pull inventory for deployment.
With respect to the inventory of Videolabs Flex Cameras, INS stated that all version 3.1 cameras in its inventory at the time of the audit have been returned to the vendor and exchanged for version 6.0 cameras. Version 3.1 (and earlier) cameras returned from the field since that time as damaged may still be in the inventory but have been flagged as undeployable, are in need of maintenance, or are in the process of being tested to determine what to do with them.
We recommend that the Commissioner, INS:
The INS had no viable option for electronically exchanging data with the Executive Office for Immigration Review (EOIR). This occurred because EOIR was developing a new scheduling system based on a stand-alone computer system rather than on the DOJ Dallas Data Center mainframe computer. Consequently, the only option INS had for obtaining EOIR data was through direct dial-up modem access. However, INS could improve the reliability of this service by obtaining more telecommunications lines.
The EOIR, an independent Department component, is responsible for the administrative adjudication of immigration cases and appeals. The EOIR primarily adjudicates deportation and exclusion cases, including asylum and criminal alien cases. The INS is required to record EOIR's hearing dates and times, which are contained in EOIR's Automated Nation-Wide System for Immigration Review (ANSIR), on its charging documents.
In our initial audit, we noted that EOIR hearing dates and times cannot be transferred electronically from EOIR to INS because the computer system used by EOIR is not compatible with the systems used by INS. The INS can only obtain the data by dialing directly into EOIR's ANSIR system. Once obtained, the data must be entered manually into the INS Refugees, Asylum and Parole System (RAPS) as well as the INS Deportable Alien Control (DACS) systems. However, INS officials consider this procedure undesirable since the connection is slow and disconnects occur frequently.
In FY 1997, EOIR received funding to upgrade its ANSIR system. In the interim, INS and EOIR formed a working group to resolve the system compatibility issues. Together, they developed a number of interim methods for exchanging data. We recommended that the INS ensure that its systems are compatible with EOIR's ANSIR system.
In our current audit, we learned that EOIR no longer intends to implement its modified ANSIR system on the DOJ mainframe computer. Instead, EOIR is now developing a new ANSIR system to operate on an IBM RISC 6000 mini-computer independent of the DOJ mainframe computer. Consequently, current solutions for exchanging data between the two systems are no longer viable. The only option available to INS at the current time is direct dialing into EOIR's ANSIR system, obtaining the required data via modem access, and manually entering it into INS' RAPS and DACS systems.
We did learn, however, that INS could improve the modem access by expanding its dial-out capacity. Within EOIR there are 60 telecommunication lines allowing access to the ANSIR system. However, according to the EOIR Information and Program Analysis Deputy Associate Director, no more than three or four lines are in use at one time.
5 We excluded the infrastructure program from our calculations since estimated completion costs for that program were not available.
6 Coopers & Lybrand, LLP, Records Reengineering Phase 1: Needs Assessment and Change Opportunities, September 1996.