U.S. Department of Justice
Office of the Inspector General
Inspections Division
The Immigration and Naturalization Service's
Customer Management Information System
Report Number I-98-19
July 1998
TABLE OF CONTENTS
How INS Uses CMIS To Serve Customers
INS Fails To Use Customer Service Data Provided by CMIS
CMIS Data is Not Consistently Reliable Because of Faulty Data Entry
INS Personnel Fail to Enter Data In CMIS at Some Locations
Appendix I Inspection Methodology
Appendix II Immigration and Naturalization Service Response to Draft Report
Appendix III Office of the Inspector General's Analysis of Management's Response
MEMORANDUM FOR DORIS MEISSNER
COMMISSIONER
IMMIGRATION
AND NATURALIZATION SERVICE
FROM:
MICHAEL
R. BROMWICH
INSPECTOR
GENERAL
SUBJECT:
Inspection of the Immigration and Naturalization Service's
Customer Management Information System,
Report Number I-98-19
We have conducted an assessment of the Immigration and Naturalization Service's (INS) Customer Management Information System (CMIS). CMIS is the computer system INS uses to serve walk-in customers who visit information waiting areas in many district offices. When CMIS was initially installed, INS managers believed it could provide district management with a tool to increase the number of customers served, improve statistical reporting, and reduce customer and employee frustration. Two components of the CMIS initiative have been implemented with beneficial results. A triage counter enhances INS' ability to handle brief service requests quickly and to identify customers who require more specialized and lengthier services. The numbering system assures customers that their wait will be orderly and enables the customers to track their progress.
CMIS was designed to collect and report information about the types of services provided to customers, waiting and transaction times, and a myriad of data that could be used by INS to better manage information waiting areas. Based upon our review of the system, we believe that INS managers could use CMIS data to establish baseline information and performance measures through which they could better manage information waiting area operations. However, CMIS' management information capabilities currently provide little value to INS. We found little evidence that CMIS is producing reliable data or that any effort is being made to gather, analyze, or apply such data to enhance INS' use of its resources or improve its delivery of services.
Although CMIS reports could be valuable tools for assessing and improving customer service, we found problems with the reliability of the data contained in CMIS. We observed numerous instances where Immigration Information Officers were either incorrectly entering or failing to enter data in CMIS. These data entry errors and omissions cause the system to report erroneously the total number of customers served, the number of customers served by category, the number of transactions performed, and customers' wait and transaction times. We also surveyed Immigration Information Officers and found that more than half could not identify the correct CMIS category and transaction codes for many services.
We believe INS management should routinely analyze and use CMIS data to improve customer service. Such routine use of CMIS data should assist INS in identifying and correcting data entry problems such as those outlined in our report. CMIS training and data entry practices also need to be improved. Field personnel need to be trained in how to use the system and advised of the system's capabilities. Formal CMIS data entry procedures need to be developed and enforced. Until these data entry problems are corrected, INS should not use CMIS data for decision-making or reporting purposes.
Based upon the information made available to us by INS officials, we estimate that INS has spent approximately $1.3 million on CMIS through FY 1997. These costs are associated with the installation and upgrading of CMIS equipment and the maintenance of related hardware and software. If INS continues to make little or no use of CMIS data, INS should not spend additional funds to install or upgrade CMIS and should consider using less costly customer ticket systems.
We sent copies of the draft report to your office on February 5, 1998, and requested written comments on the findings and recommendation. Your May 26, 1998, response concurred with the single recommendation. We have attached your response as Appendix II.
On the basis of your concurrence with the recommendation and your indication that an assessment of CMIS will be completed within six months, we consider the recommendation resolved and have kept it open pending further action. Appendix III explains what actions are needed to close the one recommendation.
Please respond to the resolved recommendation by December 1, 1998. Your response should provide the additional information requested. If actions have not been completed, please provide projected completion dates. Guidance on report follow-up and resolution can be found in Department of Justice Order 2900.10.
We hope our comments, suggestions, and recommendation in the attached report will be useful in your efforts to address these issues. We appreciate the cooperation extended to our Inspections Division staff during the review. If you have any suggestions how we might improve our review process, or if we can provide you with additional information, please let us know.
Attachment
cc: Kathleen Stanley
Liaison
Immigration and Naturalization Service
Vickie L. Sloan
Director
Audit Liaison Office
The Customer Management Information System (CMIS) is the computer system the Immigration and Naturalization Service (INS) uses to serve walk-in customers visiting information waiting areas in many district offices. Over five million customers annually visit INS information waiting areas. Reasons for visits include: picking up forms, obtaining general information, applying for permanent residency or Employment Authorization Documents, or requesting extensions of status, e.g., students. CMIS is designed to help INS ensure that customers are served through an orderly process. CMIS is also designed to collect and report information about the types of services provided to customers, waiting and transaction times, and a myriad of data that could be used by INS to better manage information waiting areas. Used effectively, CMIS can also allow management to monitor and assess its operations and to reposition staff to more effectively serve customers. The objectives of this inspection were to assess the CMIS' effectiveness in providing management with accurate and useful customer service information and in expediting customer traffic.
INS first installed CMIS in 1988 at its Chicago District Office. Chicago District management stated CMIS provided the district with the ability to increase the number of customers served, improve statistical reporting, and reduce customer and employee frustration. Use of CMIS expanded to other districts because of Chicago's positive experience with the system. INS district and regional managers were responsible for procuring and installing CMIS until 1991, when control of CMIS was moved to INS Headquarters.
Responsibility for CMIS is currently assigned to the Office of Adjudications. Support for CMIS in the Office of Adjudications is limited to a single program manager, who is responsible for supporting 47 of the 51 field sites where CMIS is installed.1 The latest version of CMIS is operational at 39 of the 51 field sites.2 Regional support for CMIS is limited. Each of the three regions has one or two staff members assigned to CMIS-related activities. Regional support staff assigned to CMIS activities estimate that, on average, they devote only 10 to 20 percent of their time to CMIS activities.
CMIS field sites are not currently unified in a single computer network. Each field site operates independently; local management can only access its own data base. INS Headquarters and Regional Offices have the capability to access CMIS at field sites, but only one field site at a time.
Based on the information provided to us by INS officials, we estimate that INS has spent approximately $1.3 million on CMIS through FY 1997. These costs are associated with the installation and upgrading of CMIS equipment and the maintenance of related hardware and software.
How INS Uses CMIS To Serve Customers
When entering an INS information waiting area that uses CMIS, customers are referred to a "triage" counter. At the triage counter, the nature of the customer's visit is established by an Immigration Information Officer (IIO). Simple requests for information or forms are usually handled at the triage counter so these customers can be served quickly and depart the information waiting area. Customers requiring more extensive services are issued sequentially numbered tickets for various categories of services. Customers issued tickets remain in the waiting area until their numbers appear on a CMIS display board. The customers then proceed to the designated service counter to receive assistance on all work required. Figure 1 depicts a typical information waiting area operation.
Figure 1: Typical Information Waiting Area
Source: OIG Observations
Information waiting areas using the latest version of CMIS typically contain the following equipment:
· Ticket Printer - located at the triage counter, this printer is used to issue a ticket to a customer showing a letter identifying the customer service category, a number showing the customer's position in the category, and an estimated wait time.
· Counter Terminals - located at the triage counter and all service counters, these keypads are used to categorize customers, enter data into the system, and call customers to be served.
· Main and Counter Displays - located at strategic locations in the information waiting area and above each service counter, these lighted boards provide customers with their wait status by displaying the customer number being served in each category, the customer number being served at each counter, and the next number due for service.
· Personal Computer - located in the supervisor's office, this computer is intended to monitor and control the information waiting area operation, assign priorities to customer categories, redirect INS personnel as needed, and store statistics for display, analysis, and report retrieval.
CMIS captures and reports data on four basic variables: wait times, transaction times, customers served by category of service, and types of transactions. CMIS reports can be used by management to monitor current information waiting area workflow as well as provide analysis of past performance. CMIS reports can cover time periods ranging from an hour to a year. Some of the types and the capabilities of CMIS reports include:
· Transaction Time Reports. These reports reflect the number of customers served and the time required to complete each transaction. These reports could be used by management to monitor customer service goals, e.g., to provide a specific category of service in less than four minutes.
· Wait Time Reports. These reports provide the number of customers whose wait time fell within selected time intervals. These reports could also be used by management to monitor customer service goals, e.g., establish an acceptable wait time of thirty minutes or less for any or all services provided.
· Category Reports. For a category, or range of categories, these reports indicate the average and maximum number of customers awaiting service, customers served, and customers who fail to appear for service over a specific time frame.
· Work Station Reports. These reports show the number of customers served at the triage counter or a service counter during a specific block of time.
· Work Station Performance Reports. These reports indicate the time a service counter was open and closed, its operating and idle time, the number of customers served, and the total and average transaction time. These reports could provide management with service counter productivity data to establish office operating standards.
· Transaction Code Reports. These reports list the transaction codes and the number of transactions performed by code. These reports can provide management with data on the specific types of services provided to customers.
The triage and service counters are the two collection points where customer service data is entered into the CMIS data base. INS generally classifies and counts customers entering information waiting areas into one of a basic set of five CMIS service categories:
· Forms - for customers requesting INS forms;
· Info Direct - for customers seeking responses to simple requests for information on any INS subject;
· Information/Questions - for customers seeking responses to more complex inquiries about any immigration subject;
· Applications - for customers submitting applications for INS benefits; and
· Replacement Cards - for customers submitting requests for, or asking for information on, replacement of lost, damaged, or stolen cards.
To accommodate local needs, some field sites have supplemented this basic set with additional categories, e.g., for attorneys and adjudications, or in other instances have reduced the number of categories. Transaction codes are also entered in CMIS for all customers served. There are over 90 different transaction codes used to identify specific customer services performed, and any transaction code can be entered under any category.
CMIS does not capture the time a customer spends waiting to gain access to the information waiting area prior to its opening or waiting in the line at the triage counter. For the purpose of CMIS reporting, a customer's wait time does not begin until the customer is placed in a service category at the triage counter. CMIS can accurately record the customer's wait time and transaction time if the data is entered correctly.
· The customer's wait time begins when the customer is first counted and entered in a service category at the triage counter and ends when the customer's number is called at a service counter.
· The customer's transaction time begins when the customer's number is called at a service counter and ends when all transactions for that customer have been completed.
RESULTS OF THE INSPECTION
INS Fails to Use Customer Service Data Provided by CMIS
CMIS' management information capabilities currently provide little value to INS. INS managers could use CMIS data to establish baseline information and performance measures through which they could better manage information waiting area operations. However, there are no requirements that mandate the analysis or the use of CMIS information, and we found little evidence that INS managers in the field or at Headquarters used CMIS reports. We could not determine whether CMIS had helped reduce wait and transaction times because baseline data prior to CMIS installation was not available from INS. Our observations and interviews with supervisory IIOs confirmed that the CMIS ticketing process contributed to an orderly waiting room environment.
During our field work we determined that INS does not take full advantage of CMIS' capabilities. At most of the field sites we visited, we found that INS management and staff ignore the customer service information collected and reported by CMIS. The system automatically generates a series of standardized reports at the end of the day at each field site. However, most supervisory IIOs we spoke with admitted that they did not review these reports. Based on our discussions with INS officials, we believe that INS' failure to effectively use CMIS results, at least in part, from a lack of understanding of what CMIS can do for INS.
Based on our observations and interviews with INS personnel, INS' only effective use of CMIS is to control and move customer traffic in an orderly manner. IIOs behind the service counters believe CMIS improves their working environment by eliminating the constant line of customers that previously stood in front of their counters over the entire course of the workday. IIOs also believe the customers' frustration level is lowered because the customers can sit and wait for the system to call them for service in an orderly manner.
We believe that CMIS data, if entered properly, could be used to assess INS information waiting area operations and improve the service INS provides to its customers. For example, we conducted an analysis of CMIS reports for the month of January 1997 for the nine field sites we visited.3 Our analysis suggests that triage counters are generally effective in serving a large percentage of INS customers (see Figure 2).4 Our observations at the field sites we visited generally confirmed this analysis.
Figure 2: Percent of Customers Completing Their Transactions at the Triage Counter without Requiring Referral to the Service Counters in January 1997
Source: INS CMIS Reports
CMIS data indicates that over half of the customers coming into INS information waiting areas at the locations we visited completed their transactions at the triage counters. These customers then left the information waiting area, thereby freeing INS resources to serve other customers. If the CMIS data is correct, further study would be necessary to determine the reason for the low customer percentage at the triage counter for Atlanta.
The generally short transaction times shown in the CMIS reports for the Forms and Info Direct categories suggest that triage customers are receiving these services in a reasonable time. The median transaction times for the nine field sites were approximately three minutes for Forms and three and one-half minutes for Info Direct. Figure 3 shows the average transaction times at triage counters. The data in Figure 3 also illustrates a significant anomaly in the triage counter data. While transaction times for Info Direct customers at most field sites fell within approximately three to four minutes, Miami showed an average transaction time of over forty-seven minutes.
Figure 3: Triage Counter Transaction Times for Forms and Info Direct by Office in January 1997
Source: INS CMIS Reports
The report data also showed that while a total of 9,524 customers were served in Miami in January 1997, only 41, less than .5 percent, were placed in the Info Direct category. This extremely low percentage for Info Direct customers in Miami is atypical of most of the field sites visited, where Info Direct constituted a significant percentage of total customers served. However, the report data for Miami also shows that 3,211 customers were served in the Information/Questions category. Service for this category is obtained at the service counters rather than at the triage counter. Generally, this category of questions requires more time on the part of the IIO to answer. The average transaction time for these 3,211 customers was fourteen minutes, and the report data shows that they had to wait an average of two hours and twenty minutes to receive service.
If the CMIS report data for Miami is correct, the triage counter is not used effectively to provide information to customers. The report data appears to indicate that the vast majority of Miami customers needing information are served at the service counters where they must wait a substantial amount of time before having their questions answered. This would appear to defeat the purpose of the triage counter, which is to screen customers who can be served quickly without referral to the service counters.
In order to assess the speed of service at the service counters, we examined the average wait and transaction times for customers requesting replacement cards. We chose the Replacement Cards category because it was the only service counter category that was consistently reflected in CMIS reports for all nine field sites we visited (see Figure 4). Our analysis of CMIS reports showed inconsistent usage of customer categories by field sites, as discussed in more detail in the next section of the report describing CMIS data entry problems. The CMIS reports show that the average transaction times for all field sites fall within the relatively narrow range of six to twelve minutes. Wait times, however, fluctuate from a low of eleven and twelve minutes for Buffalo and El Paso, to almost two hours in San Diego. Wait times for the other six field sites were within the range of approximately one hour to an hour and a half.
Figure 4: Wait and Transaction Times for Replacement Cards by Office January 1997
Source: INS CMIS Reports
Figure 4 shows the wait and transaction times for the nine field sites. If this data is correct, INS should determine how El Paso and Buffalo have reduced waiting times to less than 20 percent of those of the other seven field sites. The lower wait times recorded in El Paso and Buffalo may be due to erroneous data entry, lower than usual customer numbers, or an extremely efficient staff. INS should analyze CMIS reports to help identify the causes of apparent disparities in customer service among district offices. We discuss the reliability of CMIS data in the following section of this report.
CMIS Data is Not Consistently Reliable Because of Faulty Data Entry
Although CMIS reports could be valuable tools for assessing and improving customer service, the data contained in CMIS, and reflected in the reports, is unreliable. We observed numerous instances of IIOs incorrectly entering or failing to enter data. These data entry errors and omissions caused the system to report erroneously the total number of customers served, the number of customers served by category, the number of transactions performed, and customers' wait and transaction times. We confirmed, when possible, our field observations with an analysis of CMIS reports from field sites. We also surveyed IIOs and found that many could not identify the correct CMIS category and transaction codes for specific services.
INS Personnel Fail to Enter Data In CMIS at Some Locations
We observed INS triage and service counter personnel at many of the information waiting areas failing to enter customer service data into CMIS, which prevents the accurate capture and reporting of customer service data. For example, at one field site, Info Direct customers were not counted at the triage counter. The daily report only shows two customers captured in this category for the day. At a second field site we observed certain IIOs failing to enter any data into CMIS at the triage counter. At this second field site we estimate that at least twice as many customers came in for service as were reported by CMIS for that day.
At four field sites we observed IIOs providing assistance to customers at service counters with keypads that had not been activated or were inoperative. This prevented any service data from being captured. At another field site we observed two IIOs failing to enter data into CMIS at triage counters. During a twenty minute time period we observed approximately 30 to 40 customers receiving service. Of the two IIOs staffing the counter, one failed to enter any data, the second IIO only entered forms-related data.
At two field sites we observed that attorneys were allowed to drop off packets of applications at the cashier cage for processing without going to the triage counter to receive a ticket. We were told that members of voluntary agencies were permitted to do the same thing. This practice of allowing certain customers to by-pass the system prevents the number of these applications submitted and the associated processing time from being captured in CMIS.
CMIS Data is Entered Incorrectly
We observed numerous instances where IIOs were incorrectly entering data. At one field site we observed IIOs routinely entering multiple transaction codes at the end of a customer's session. This causes inaccurate transaction times to be recorded.
We found several instances where field management sacrificed accurate CMIS data in order to expedite customer service. We do not believe that INS should impede efficient service in order to merely capture accurate data. However, INS should devise methods to record accurate data when implementing customer service innovations.
At one field site we observed IIOs issuing tickets for a different category of service than the service the customers requested. The supervisory IIO at this field site explained that by placing customers with less time-consuming applications in the Information/Questions category these customers would be called and served more quickly because wait times were shorter for that category than for the applications category.
At another field site, an INS employee working in the Adjudications section generated a series of Adjudications category tickets at the triage counter for customers waiting in line at the triage counter to apply for Employment Authorization Document (EAD) cards. These tickets were not distributed to the customers, but were inexplicably discarded by the employee. These customers were then instructed to turn in their EAD appointment letters to the employee and wait to be called. These letters would determine the order in which the customers would be called to receive service in this section of the information waiting area. After the EAD cards were issued to this group of customers, the employee then entered the transaction code and pushed the "next" button on the keypad for each customer served in rapid succession. Only the first ticket number in this series was recorded in the Adjudications category as a customer served. The remaining numbers in this series were recorded in CMIS as "no shows" (customers failing to show for service) because the INS employee did not wait the required interval before pushing the button to call the next customer.5 The consequence of this unusual method resulted in almost every one of these customers being counted in CMIS as a "no show."
This interval first starts when the button is pushed on the keypad at the service counters to call a customer. This displays the next customer number in the sequence awaiting service. The customer is served and the interval ends when the button is pushed again, calling the next customer number in the sequence. When INS staff fail to wait the required interval before calling the next customer, the system will record this customer number as a "no show" for that category, even though the customer has been served. The CMIS report for that day showed 3 customers served in the Adjudications category, and 52 "no shows."
At another field site, paperwork that was required to start the process for obtaining replacement cards was collected when the customer was issued a ticket at the triage counter. The staff proceeded to conduct and complete the necessary work for the customer while the customer was in the waiting area. Eventually, the customer's number was displayed to start the transaction process when, in actuality, a majority of the process had already been completed. While this practice may help process the customer more quickly, it also distorts the recorded wait and transaction times. The time the staff spent processing the transaction is not accurately captured. This time is actually absorbed in the customer's wait time. Only the abbreviated time that the customer is in front of the IIO picking up the paperwork is captured in the transaction time.
We found that the field sites we visited deviated from the basic set of five category headings to accommodate local needs. Seven sites used six categories and one site used four categories. The remaining field site reduced the number of categories it used to classify its customers from six to two (Information/Questions and Adjudications). We also observed that one district office used different categories than its sub-office. The practice of adjusting the number of categories to facilitate customer service is understandable, but it prevents CMIS from capturing consistent information and comparing data from site to site.
Multiple Counting of Individual Customers
We witnessed numerous instances of individual customers at the triage counters being counted in more than one category. We frequently observed customers at triage counters who requested a form and asked a general question. We then noted the IIO would count the customer twice in CMIS, once as a customer served in the Forms category and once in the Info Direct category. According to the CMIS program manager, each customer should only be counted in one category. This provides the field site with an accurate count of the number of individual customers coming into the office and being served that day.
The multiple counting of customers that we observed is attributable to the triage counter IIOs not following the one-customer-per-category data entry procedure. Multiple counting inflates the number of customers served for the day. The IIOs should only count the customer in one category and enter the correct transaction code(s) to capture any additional services provided and have rules for determining which category a customer who has more than one type of transaction should be sorted.
Two field sites we visited had separate service areas for Forms and Info Direct category customers. It is likely that many customers at these field sites are counted more than once during the day. The customers go to the area designated for forms and information, receive service, and are counted in the appropriate category. The same customers could then go to the other information waiting area in a different location in the building to obtain other types of services and be counted again in another category that same day. CMIS is unable to track and prevent the multiple counting of Forms and Info Direct category customers because they do not receive a ticket.
Survey Results Suggest IIOs Are Incorrectly Entering CMIS Data
Further evidence that CMIS data is inaccurate is supported by the results from the questionnaire we presented to IIOs during our site visits. This questionnaire presented a variety of scenarios that an IIO may encounter at both triage and service counters. The IIOs were asked to enter either the ticket category or the transaction code(s) as they pertained to these scenarios. Only 41 percent of the triage counter responses and 46 percent of the service counter responses were correct.
The questionnaire also identified a large disparity in the way IIOs are entering multiple transaction codes for a single customer. Of the 67 IIOs that responded, only 36 (54 percent) indicated that they correctly entered each transaction code after completing each specific transaction. The other 31 (46 percent) IIOs indicated they entered all of the transaction codes at the very end of the customer's session. This second method of data entry results in inaccurate transaction times.
Based on the questionnaire results, we believe that additional training is needed if INS continues to use CMIS. Past CMIS training has generally consisted of one or two days of hands-on training conducted by the program manager at the time of installation or upgrade of the system. We encountered few INS supervisors and staff who appeared to fully understand the entire range of CMIS capabilities.
Triage Customers are Erroneously Reflected as "No Shows"
Our review of CMIS reports identified a substantial number of "no shows" recorded in the Forms and Info Direct categories at triage counters (see Figure 5). Forms and Info Direct customers are served directly at the triage counters and are not issued a ticket. All remaining customers receive category tickets at the triage counters for transactions to be performed at the service counters. Customers may fail to show for service at the service counters for a variety of reasons correctly resulting in "no shows" being recorded. Forms and Info Direct customers should never be recorded as "no shows."
Figure 5: Percent of Customers Served at Triage Counters Reported as "No Shows" by Office in January 1997
Source: INS CMIS Reports
For eight of the field sites we visited we conducted an analysis of the CMIS reports for the month of January 1997.6 Our analysis for these eight sites reflected a total of 35,850 customers served in the Forms and Info Direct categories. CMIS reports incorrectly identified 9,206 of these 35,850 customers served as "no shows." These 9,206 customers actually received service at the triage counter and therefore should not have been recorded as "no shows." CMIS reports are substantially understating the numbers of customers served at the triage counters. In this case, 26 percent of the customers actually served at triage counters for the eight field sites in January 1997 were not counted as customers served. These customers were recorded as "no shows" because the IIOs failed to wait the required interval before pushing the button on the keypad bringing the next customer to the counter.
CONCLUSION AND RECOMMENDATION
In June 1995, an INS Headquarters memorandum to all Regional Directors stated that CMIS was ". . . designed to expeditiously process the visiting public, improve employee productivity, and provide managers with a useful tool for gathering statistics." Based upon our field observations and our analysis of CMIS reports, we believe that INS is using CMIS to control and serve customers in an orderly fashion. However, our review of CMIS also revealed that INS management does not routinely use or analyze CMIS data to improve or manage waiting room operations. Further, customer service data collected by CMIS is not consistently reliable.
INS has spent approximately $1.3 million on CMIS through FY 1997. We believe INS has not realized the full benefit of this expenditure. INS has used CMIS as a customer service sequencing tool and not as a management information system. A simple, less expensive "take a number" system, without the management information capability may have sufficed.
We believe INS management should routinely analyze and use CMIS data to improve customer service. The routine review and use of CMIS data would also assist INS in identifying and correcting data entry problems and would potentially improve employee productivity. Until these data entry problems are corrected, INS should not spend additional funds to install or upgrade CMIS, or use CMIS data for decision-making or reporting purposes.
The Inspections Division recommends that the Commissioner, Immigration and Naturalization Service:
1. Determine whether the customer service information supplied by CMIS is useful to INS.
· If CMIS information is not useful, INS should stop spending additional funds on CMIS and explore alternative and less-expensive customer ticket systems.
· If INS determines that CMIS information is useful, issue standard procedures and requirements to improve data reliability and require the use of CMIS data to improve INS information waiting area operations.
INSPECTION METHODOLOGY
We conducted our field work from February through May 1997. We interviewed key officials at INS Headquarters and at selected field sites, observed information waiting area operations, and reviewed and analyzed available documentation and reports. Field sites we visited included offices and suboffices in El Paso, San Diego, San Francisco, San Jose, Chicago, Buffalo, Atlanta, Miami, and New York. We visited with the contractor responsible for the CMIS technology and installation at INS field sites to obtain an understanding of the history and features of the system.
The inspection also included a questionnaire administered to INS Immigration Information Officers (IIOs) at field sites we visited who were assigned to information waiting areas. We developed and tested this questionnaire to allow us to assess the IIOs' knowledge of CMIS data input procedures. Management officials from the INS Office of Adjudications, the CMIS program manager, and the Baltimore District's supervisory IIO reviewed this questionnaire for content validity and reliability.
OFFICE OF THE INSPECTOR GENERAL'S ANALYSIS
OF MANAGEMENT'S RESPONSE
On February 5, 1998, the Inspections Division sent copies of the draft report to INS with a request for written comments. INS responded by memorandum from the Commissioner dated May 26, 1998 (APPENDIX II).
1. Recommendation 1 - Resolved-Open. INS concurred with our recommendation and stated in its memorandum that CMIS information is useful and will provide managers with data to improve employee productivity and service to the public. In the next six months INS will reassess its needs and explore whether CMIS is the most cost effective and efficient program to meet those needs.
Please provide the completed assessment by December 1, 1998. In addition, if CMIS is selected as the system to meet INS' future needs, please provide the standard procedures and requirements under which CMIS will operate. If CMIS is not selected as the system, please provide the program plan for the selected system's mission, implementation, and operation.
1 At 4 of the 51 sites (the Detroit International Airport, and 3 El Paso District land ports-of-entry located at Ysleta, Paso del Norte, and the Bridge of the Americas) CMIS is administered locally without the support of the CMIS program manager.
2 INS began upgrading CMIS to its latest version in September 1995. This version provides additional functions to help manage, control, and monitor customer traffic, and offers a wider range of statistical reporting capabilities.
3 Complete data was not available from the New York office.
4 The number of Forms and Info Direct customers served at triage counters in January 1997: EL Paso - 3,225; San Diego - 5,383; San Francisco - 7,934; San Jose - 6,683; Chicago - 6,488; Buffalo -1,073; Atlanta - 619; Miami - 4,445.
5 CMIS allows INS to set a time interval that determines when a customer is to be recorded as a "no show." The purpose of the time interval is to allow CMIS to record a customer as a "no show" if the customer fails to report to the service counter when called. If the elapsed time between the first and the last keypad entries for a customer is less than the designated interval, CMIS records that customer as a "no show." Although CMIS may be set for any time interval, the CMIS program manager decided that twenty seconds was appropriate because every customer transaction should require at least twenty seconds to perform. Calling the next customer in less than twenty seconds tells CMIS that the previous customer failed to report for service.
6 The number of Forms and Info Direct customers reported as "no shows" after being served at triage counters in January 1997: EL Paso - 804; San Diego - 749; San Francisco - 2,062; San Jose - 2,478; Chicago - 1,862; Buffalo -76; Atlanta - 166; Miami - 1,009; complete data was not available for New York.
#####