<DOC> [106th Congress House Hearings] [From the U.S. Government Printing Office via GPO Access] [DOCID: f:67019.wais] OVERSIGHT OF THE 2000 CENSUS: STATUS OF KEY OPERATIONS ======================================================================= HEARING before the SUBCOMMITTEE ON THE CENSUS of the COMMITTEE ON GOVERNMENT REFORM HOUSE OF REPRESENTATIVES ONE HUNDRED SIXTH CONGRESS SECOND SESSION __________ MARCH 14, 2000 __________ Serial No. 106-161 __________ Printed for the use of the Committee on Government Reform Available via the World Wide Web: http://www.gpo.gov/congress/house http://www.house.gov/reform U.S. GOVERNMENT PRINTING OFFICE 67-019 WASHINGTON : 2001 _______________________________________________________________________ For sale by the Superintendent of Documents, U.S. Government Printing Office Internet: bookstore.gpo.gov Phone: (202) 512-1800 Fax: (202) 512-2250 Mail: Stop SSOP, Washington, DC 20402-0001 COMMITTEE ON GOVERNMENT REFORM DAN BURTON, Indiana, Chairman BENJAMIN A. GILMAN, New York HENRY A. WAXMAN, California CONSTANCE A. MORELLA, Maryland TOM LANTOS, California CHRISTOPHER SHAYS, Connecticut ROBERT E. WISE, Jr., West Virginia ILEANA ROS-LEHTINEN, Florida MAJOR R. OWENS, New York JOHN M. McHUGH, New York EDOLPHUS TOWNS, New York STEPHEN HORN, California PAUL E. KANJORSKI, Pennsylvania JOHN L. MICA, Florida PATSY T. MINK, Hawaii THOMAS M. DAVIS, Virginia CAROLYN B. MALONEY, New York DAVID M. McINTOSH, Indiana ELEANOR HOLMES NORTON, Washington, MARK E. SOUDER, Indiana DC JOE SCARBOROUGH, Florida CHAKA FATTAH, Pennsylvania STEVEN C. LaTOURETTE, Ohio ELIJAH E. CUMMINGS, Maryland MARSHALL ``MARK'' SANFORD, South DENNIS J. KUCINICH, Ohio Carolina ROD R. BLAGOJEVICH, Illinois BOB BARR, Georgia DANNY K. DAVIS, Illinois DAN MILLER, Florida JOHN F. TIERNEY, Massachusetts ASA HUTCHINSON, Arkansas JIM TURNER, Texas LEE TERRY, Nebraska THOMAS H. ALLEN, Maine JUDY BIGGERT, Illinois HAROLD E. FORD, Jr., Tennessee GREG WALDEN, Oregon JANICE D. SCHAKOWSKY, Illinois DOUG OSE, California ------ PAUL RYAN, Wisconsin BERNARD SANDERS, Vermont HELEN CHENOWETH-HAGE, Idaho (Independent) DAVID VITTER, Louisiana Kevin Binger, Staff Director Daniel R. Moll, Deputy Staff Director David A. Kass, Deputy Counsel and Parliamentarian Lisa Smith Arafune, Chief Clerk Phil Schiliro, Minority Staff Director ------ Subcommittee on the Census DAN MILLER, Florida, Chairman THOMAS M. DAVIS, Virginia CAROLYN B. MALONEY, New York PAUL RYAN, Wisconsin DANNY K. DAVIS, Illinois MARK E. SOUDER, Indiana HAROLD E. FORD, Jr., Tennessee ------ ------ Ex Officio DAN BURTON, Indiana HENRY A. WAXMAN, California Jane Cobb, Staff Director Laura Chamberlain, Professional Staff Member Amy Althoff, Clerk Michelle Ash, Minority Counsel C O N T E N T S ---------- Page Hearing held on March 14, 2000................................... 1 Statement of: Mihm, J. Christopher, Associate Director, Federal Management and Workforce Issues, U.S. General Accounting Office, accompanied by Randolph C. Hite, Associate Director, Accounting and Information Management Division, U.S. General Accounting Office; and Robert Goldenkoff, Assistant Director, General Government Division, U.S. General Accounting Office.......................................... 18 Letters, statements, etc., submitted for the record by: Maloney, Hon. Carolyn B., a Representative in Congress from the State of New York: Letter dated August 20, 1999............................. 11 Prepared statement of.................................... 15 Mihm, J. Christopher, Associate Director, Federal Management and Workforce Issues, U.S. General Accounting Office, prepared statement of...................................... 23 Miller, Hon. Dan, a Representative in Congress from the State of Florida, prepared statement of.......................... 5 OVERSIGHT OF THE 2000 CENSUS: STATUS OF KEY OPERATIONS ---------- TUESDAY, MARCH 14, 2000 House of Representatives, Subcommittee on the Census, Committee on Government Reform, Washington, DC. The subcommittee met, pursuant to notice, at 2 p.m., in room 2203, Rayburn House Office Building, Hon. Dan Miller (chairman of the subcommittee) presiding. Present: Representatives Miller and Maloney. Staff present: Jane Cobb, staff director; Timothy J. Maney, chief investigator; Chip Walker, communications director; Erin Yeatman, press secretary; Lara Chamberlain, professional staff member; Amy Althoff, clerk; Michelle Ash, minority counsel; David McMillen and Mark Stephenson, minority professional staff members; and Earley Green, minority assistant clerk. Mr. Miller. Good morning. The Census Subcommittee has a quorum present. We will start with my opening statement, Mrs. Maloney and Mr. Mihm's. The census questionnaires have been mailed, and many people have already received them. Please fill out the form and mail it back as soon as possible. I can't overemphasize this enough to those sitting here and those watching on TV. Your friends, neighbors, local and State officials, local school boards and State universities, all areas of the private sector are dependent upon an accurate census. So when you get the questionnaire, short form and long form, take a few minutes to sit down and answer the questions and mail it back. America is counting on you. Today we again welcome the nonpartisan General Accounting Office before the Subcommittee on the Census. As I mentioned previously, the GAO's mission is to help the Congress oversee Federal programs and operations to assure accountability to the American people. GAO evaluators, auditors, lawyers, economists, public policy analysts, information technology specialists and other multidisciplinary professionals seek to enhance the economy, efficiency, effectiveness and credibility of the Federal Government both in fact and in the eyes of the American people. GAO accomplishes this mission through a variety of activities, including financial audits, program reviews, investigations, legal support and policy program analyses. GAO is dedicated to good government through its commitment to the values of accountability, integrity and reliability. Last week the Director of the Census Bureau was before the subcommittee and questioned the nonpartisan GAO and other oversight entities for what he termed real-time oversight that, in his mind, was not understandable. However, I would remind everyone that in the first major mailing of census materials to the general public, all 120 million prenotification letters were misaddressed, and the letter itself did not explain fully the purpose of the enclosed envelope to those who only spoke English, obviously a majority of the population. The Director has termed both of these errors as major embarrassments, and, as Mr. Ryan noted, the national 800 number was not printed on the forms. According to the Census Bureau, the 800 number was not available at the time the forms were printed. I must admit that I find this explanation highly improbable. In the wake of these errors, a strong argument can be made for more oversight, not less. And these errors also call into question the ability of the Bureau to conduct the ACE or estimation adjustment. Today the GAO will have its turn to defend its actions, all of which have been sanctioned by our subcommittee either jointly or independently. While Mrs. Maloney attacked our level of oversight and thought that it may be intruding on the Census Bureau's ability to conduct the census, many of the GAO reports have been jointly requested with her support or that of Mr. Waxman, the ranking member of the full committee. I also found it ironic that while Mrs. Maloney criticized the level of oversight and called it intrusive and burdensome, in her next breath she asked the Director to provide the subcommittee with yet another report, this time on the level of oversight and the amount of time that the Bureau takes to comply with the various requests for information. The stories in the wake of last week's hearing said that the spirit of bipartisanship had been broken. That is not entirely accurate. Both Mrs. Maloney and I have tirelessly promoted the census and will continue to do so. However, I am not in a position to tirelessly defend the Bureau at all costs. When deserving of praise, the Bureau should receive it, but when deserving criticism, it should also receive it. Those of us who sit on this subcommittee in Congress, and the President, are ultimately responsible for the census. The American people, and rightly so, hold the elected officials responsible for the actions of their government. This is the people's census, and we are the people's representatives. This subcommittee requests information because it believes that it is needed to make an informed judgment on the success of the 2000 census. On a regular basis the subcommittee is questioned on the status of operations by Members of Congress, constituents from around the Nation, and the press. Let's say a reporter or local government official calls and asks, for example, how, in my opinion, hiring is proceeding? Imagine the shock if my answer was, well, the Census Bureau tells me everything is fine. That would be not be the sign of a well- informed chairman or a subcommittee doing its job. While the Bureau prefers to talk about the census in national terms, such as, ``Hiring is 4 percent ahead of schedule,'' that does not mean anything to a reporter or city official that calls from Dayton, OH, or San Antonio, TX. They justifiably want to know how they are doing locally. The Bureau may want to reevaluate the level of information that it provides on the local level. It seems that the media is in need of this type of information. Of course, it is not that the Census Bureau cannot provide local information. Soon every local government will have access to their mail response rates daily. It is hoped that this information will encourage local involvement and raise the 2000 mail response rate 5 percent above that of 1990. I believe that because hiring is such an integral part of the census, everybody should have access to this information. This information is not meant to embarrass the Census Bureau. To the contrary, it is meant to spur action. An example is what delegate Eleanor Holmes Norton has done right here in the Nation's Capital. After press reports on the hiring shortage here in D.C., she organized a job fair. That is a positive action by a local official who was made aware of a problem. This census has been a highly contentious one since the start of the decade. Both parties in Congress are examining it much closer than in previous decades. In fact, I would say the Nation as a whole is taking a much closer look. But if the 2000 census is to be truly transparent, everyone should be able to see clearly through the window without obstruction. I am pleased to report that the Census Bureau has been very responsive to my requests for a meeting among the oversight parties, and that this meeting will take place after the hearing today. I hope all of the remaining issues will be resolved to everyone's satisfaction. I am also pleased to report that since I raised the issue with Secretary Daley almost 2 weeks ago, GAO has reported significant progress in obtaining the information it feels it needs to conduct thorough oversight. Today we will hear testimony in a number of key areas. The Census Bureau had to reconfigure the data capture system in order to capture 1.5 billion pages of data from 119 million households. I am very concerned that testing and development of this system has been behind from the start. I am also concerned that the new software and hardware has not been used in a simulated census environment. A key ingredient to the local census outreach efforts are the Complete Count Committees [CCC]. These committees are designed to do local outreach and promotion. However, many of the Complete Count Committees that we visited are frustrated from a lack of resources. One such CCC in a major county told the subcommittee that they were shocked when they were contacted by the Census Bureau's partnership specialists and asked to supply materials for an upcoming event and given a 2- day deadline. Imagine the shock of the CCC. They understood that it was the Census Bureau that was supposed to supply the materials, not the other way around. And, of course, hiring remains a concern on this subcommittee despite the Director's assertion that ``I don't lose any sleep over it.'' I hope my concerns never bear fruit and, in fact, the local census offices are fully staffed for a nonresponse followup. Mr. Mihm, thank you for coming before the committee, and I look forward to your testimony. [The prepared statement of Hon. Dan Miller follows:] [GRAPHIC] [TIFF OMITTED] T7019.001 [GRAPHIC] [TIFF OMITTED] T7019.002 [GRAPHIC] [TIFF OMITTED] T7019.003 [GRAPHIC] [TIFF OMITTED] T7019.004 Mr. Miller. Mrs. Maloney, do you have an opening statement? Mrs. Maloney. Thank you, Mr. Chairman. Census day may be 18 days away, but the census has begun. Almost 100 million questionnaires are in the mail, and 22 million more are being delivered by hand in rural areas. I received mine yesterday, and I urge all Americans to fill out their questionnaires and mail it back. As has been the case in our recent hearings, the news on preparations for the census is good, a point which can easily be lost in the details of a hearing. But if we look at the forest and not the trees, things are going pretty well. Particularly noteworthy is a new USA Today/CNN/Gallup Poll out just yesterday in which 96 percent of the respondents say they will mail back their questionnaires. I doubt it will be that high, but that is certainly an important indication of the all- important mail response rate, and it is very good news. Beyond that, and as the testimony today shows, things are on track. All 520 offices are open and running. Though there are localized problems, recruiting is actually ahead of schedule nationwide at about 75 percent of the total needed. The questionnaires--all 120 million--are printed and actually being delivered by U.S. Post Office and Census Bureau personnel even as we speak. As Director Prewitt has emphasized, unexpected problems could develop tomorrow, but as of today things are running pretty well. Now, the chairman mentioned the problem in recruitment. I want to give the number to the public, 1-888-325-7733, as displayed here for people to call if they want jobs and want to help the Census Bureau. At our hearing last week, the issue of appropriate access of oversight entities to census 2000 activities and information was a major point of discussion. I do not want to belabor those issues, but I do want to clarify the record on a couple of points. First, the General Accounting Office and the Census Bureau were well on their way toward reaching an agreement regarding GAO's access to Bureau information before our last hearing. If this fact had been clear to all concerned, I think much of the discussion we had would have been avoided. There was not last week nor is there today a disagreement over access between the Bureau and GAO. I will let Mr. Mihm speak for himself, but I understand that all of these issues have been resolved. I also suspect that in a project of this size, scope and complexity, it is normal to have differences that need to be worked out and reviewed. Second, the guidelines on oversight which the Bureau has implemented were sent to the oversight bodies on December 16, 1999, almost 3 months ago. As best I can tell, they represent the continuation of policies which have been in place for over 2 years, and I am somewhat surprised that they have become an issue this late in the process. If there was a problem with these guidelines, and they are only guidelines, it should have been addressed long ago. Mr. Chairman, you have raised concerns about the access of our own staff and that of the Census Monitoring Board to field offices. While I would note that similar visits never happened during the 1990 census, they may have some value. But it is also important to understand that GAO and the Inspector General's staff are highly trained auditors and evaluators, working under strict professional standards and their own guidelines on how to conduct themselves in the field. Although these agencies act in a strictly nonpartisan manner, I have real concerns regarding the conduct of the Monitoring Board staff given their activities in the field to date and the fact that they are not subject to any similar guidelines for their conduct. I know that the chairman mentioned his concern regarding the need for representatives of the regions or headquarters staff accompanying subcommittee staff on their visits to local census offices. I just want to point out that this is far from unusual. The chairman and I both liken the census to a military operation, and I think that is a good analogy. I just want to point out that when Members of Congress or their staff go into the field to visit military installations, they are usually accompanied by half of the Pentagon, so I do not think that it is unusual or inappropriate to have representatives accompany our own staff. I know my staff has found the presence of regional staff helpful in understanding the census operations since many times they can answer questions that the local staff cannot. I do want to compliment the chairman on his idea of getting all of the principals together from the Monitoring Board, cochairs Ken Blackwell and Gil Casellas; the GAO, Mr. Mihm and perhaps Mr. Walker; the Commerce IG, Mr. Frazier; and ourselves, to personally resolve any issues that remain. As you know, that is exactly what Director Prewitt suggested in his letter of August 26 of last year to you in which he expressed his concerns regarding the demands of various oversight bodies and their impact on the Bureau's ability to conduct the census. I would like to put that letter in the record and, if I could, put this letter in and the one from February 8. Mr. Miller. Without objection, so ordered. [The information referred to follows:] [GRAPHIC] [TIFF OMITTED] T7019.005 [GRAPHIC] [TIFF OMITTED] T7019.006 [GRAPHIC] [TIFF OMITTED] T7019.007 Mrs. Maloney. That letter of February 8 was directed to cochairs Blackwell and Casellas, asking for a meeting to ensure that their information requests were met. These hardly seem like actions of someone trying to, ``hide something.'' Even if these issues are settled at the staff level, I think a meeting of the principals could be very useful for all concerned and personally am ready to attend such a meeting. Mr. Chairman, while we are reviewing the issue of oversight, I want to pose a question. What are the oversight goals of this committee with respect to the census? Oversight to what end? Are we trying to make this census better, to develop plans for the next census in 2010? If we are trying to make sure that this census is the best it can be, then why hasn't the subcommittee responded to the major recommendations GAO had in their December report? The GAO gave us some concrete statutory steps to improve the pool of possible enumerators, which you have pointed out is still a concern in some small pockets around the country. I know that you strongly supported Mrs. Meeks' bill, H.R. 683, which would have allowed current welfare recipients to receive their benefits and work for the census at the same time, but the Majority Leader has refused to bring it to a vote on the floor. I also know that recruitment is still a concern in the LCOs in both our districts. I think responding to the GAO's suggestion incorporated in my bill, H.R. 3581, would make sense. If this subcommittee is committed to constructive oversight, we should act on those recommendations. Of course, the alternative to constructive oversight is to use it to play gotcha with the census in a continuing effort to try to stop the use of modern statistical methods. Thank you very much, Mr. Chairman. [The prepared statement of Hon. Carolyn B. Maloney follows:] [GRAPHIC] [TIFF OMITTED] T7019.008 [GRAPHIC] [TIFF OMITTED] T7019.009 [GRAPHIC] [TIFF OMITTED] T7019.010 Mr. Miller. Mr. Mihm, if you would stand, Mr. Hite and Mr. Goldenkoff, and raise your right hands, and I will swear you all in. [Witnesses sworn.] Mr. Miller. Let the record acknowledge that they have answered in the affirmative. Mr. Mihm. STATEMENTS OF J. CHRISTOPHER MIHM, ASSOCIATE DIRECTOR, FEDERAL MANAGEMENT AND WORKFORCE ISSUES, U.S. GENERAL ACCOUNTING OFFICE, ACCOMPANIED BY RANDOLPH C. HITE, ASSOCIATE DIRECTOR, ACCOUNTING AND INFORMATION MANAGEMENT DIVISION, U.S. GENERAL ACCOUNTING OFFICE; AND ROBERT GOLDENKOFF, ASSISTANT DIRECTOR, GENERAL GOVERNMENT DIVISION, U.S. GENERAL ACCOUNTING OFFICE Mr. Mihm. Thank you, Mr. Chairman and Mrs. Maloney. It is once again a great honor and pleasure to appear before you to discuss the status of the 2000 census. March is one of the most crucial periods in the 10-year census cycle. Over the course of this month, the Bureau will deliver census questionnaires to the vast majority of the Nation's 120 million households. I also received my questionnaire yesterday. Knowing I would appear before you today, I made sure that I filled it out and mailed it in today. I would have mailed it in in any case, but I wanted to get it in today. The Bureau will begin to process millions of completed questionnaires at its four data capture centers located across the country this month. Outreach and promotion efforts will be at their greatest intensity. The Bureau's temporary work force will approach peak levels, and coverage improvement programs are to get under way. I am fortunate to be joined by two of my colleagues who have managed GAO's work on the 2000 census. Randy Hite leads GAO's work on the wide range of Federal technology issues, including the census; and Robert Goldenkoff has day-to-day responsibility for much of our work on census operations. Our statement focuses on developments that have occurred since we last testified before this subcommittee in February regarding essential activities such as, first, the outreach and promotion program; second, field followup operations, including staffing and coverage improvement; and third, data capture. In addition, I will discuss the steps the Bureau has taken to ensure that the census questionnaires do not contain the same misprint as was in the mailing addresses in the notification letters. However, before turning to those issues, I want to spend just a moment on the question of GAO's access to census operational information. Mr. Chairman, as you and Mrs. Maloney mentioned in your opening statements, we have reached agreement with the Bureau and implemented a process that I am confident will allow us to fulfill our role in supporting the bipartisan oversight needs of this subcommittee. This agreement with the Bureau provides us with access to the routine management information on the status of the census while importantly minimizing the burden on the Bureau. It is a new process that they established to make sure that it reduces the burden from the other process that they had been using. In reaching this agreement I want to thank the Bureau for its willingness to work with us on access issues, and the subcommittee in its efforts, and in particular you, Mr. Chairman, for the attention and support you gave us during this time. I deeply appreciate your efforts on our behalf on this issue. Turning now to the census outreach and promotion program, the Bureau has formed partnerships with organizations across the country to help promote the census. The Complete Count Committees, as you mentioned, Mr. Chairman, are a key component of the partnership program. They consist of local government, religious, media and other community leaders. Not surprisingly, in our conversation with members of these committees across the country, we are finding significant differences regarding the resources that they have available to promote the census and their level of activity. Regarding the support that the Bureau is providing these committees, the situation appears to be equally mixed. On the one hand, committee representatives we contacted were generally pleased with the assistance and guidance that they were receiving from the Bureau's partnership staff. The local committee representatives were also satisfied with the quality and quantity of the English-language material that they received from the Bureau. On the other hand, however, several committee representatives we spoke with said the amount of foreign-language materials, especially Spanish and Asian languages, were insufficient to meet their needs. In addition to the Complete Count Committees, the Bureau Census in Schools Program is intended to increase response rates by encouraging students to remind their parents or guardians to respond to the census. The Complete Count Committee and local census representatives we spoke to spoke highly favorably of the idea of promoting the census through the schools. Mr. Chairman, I know that you had very positive experiences at home in that regard. To date the Bureau has fulfilled orders for about 1.5 million teacher kits for elementary, middle and high school teachers. However, problems have occurred in distribution of these materials. Orders for Census in Schools materials are taking between 2 and 4 weeks to be filled, according to the Bureau officials responsible for the day-to-day management of this program. An effective publicity and outreach program is important to boost mail response rates, which contributes to higher census data quality and reduces the staff needs and schedule burdens on the census. This leads to the second topic I would like to cover, census field operations. The Bureau's update/leave operation, which began on March 3, was the first test of the Bureau's ability to staff its operations at near peak activities. Over 700,000 enumerators and other staff are now in the field conducting update/leave. However, to meet its nonresponse followup staffing needs, the much larger operation that is coming in just a few weeks, the Bureau needs to recruit an additional 700,000 qualified applicants to meet its overall goal of 2.4 qualified applicants by mid-April. Bureau data show that nationally, as of March 2, the Bureau had recruited 74 percent of the applicants that it needed. This was slightly ahead of the Bureau's national 71 percent goal at that point. However, 7 of the Bureau's 12 regional offices fell short of the 71 percent benchmark. Table 1 in my written statement shows the progress that census regions have made in meeting recruiting goals in February and in March. Most important, at the local level, 270 of the Bureau's 511 local census offices fell below the Bureau benchmark of 71 percent. Of these 270, 22 had recruited fewer than half of the qualified applicants that the Bureau needed as of March 2. The regional and local census officials continue to aggressively recruit applicants. For example, the Census Bureau is working with communities to set up testing, meeting with local leaders, the Atlanta and several other regional offices are mailing postcards to targeted ZIP codes that have been identified as hard-to-recruit areas, and in some cases they have increased pay rates for enumerators and other staff. In addition to the followup efforts--efforts to followup on nonresponding households, the Bureau has included coverage improvement programs in the 2000 census that are aimed at increasing the count of the hard-to-enumerate populations. Two of these programs are the Bureau's walk-in Questionnaire Assistance Centers and the Be Counted Program. Bureau data as of March 1 show that a combined total of 46,000 Be Counted sites and Questionnaire Assistance Centers have been committed to be established. For perspective, this is about 3\1/2\ times the 12,600 McDonald's restaurants that are in the Nation. Being the father of small children, I am well-acquainted, I think, with about half of those restaurants. As we discussed in our February report to the subcommittee, the Bureau appears to be taking the steps needed to ensure a successful Be Counted/Questionnaire Assistance Center effort. They learned some key lessons from the dress rehearsal and are taking appropriate actions. Turning to the third topic I will discuss today, census data capture. As of today the Bureau reports that it has implemented the near-term system changes and completed all scheduled test events, including an operational test at each data capture center and a test of the four centers operating concurrently. Further, as of today the data capture centers have been operating for about a week, the first 2 days of which they checked in over 117,000 questionnaires. This workload represents about 8 percent of the daily workload expected later this month when peak operations kick in, when they will be processing 1.5 million questionnaires per day. Thus the actual operations and the data that we have thus far demonstrate that the centers are up and running, but they do not demonstrate those centers' readiness to operate at expected production- level workloads. Moreover, the information that we have seen on actual operations does not address whether recent changes to the data capture system are functioning correctly. We, therefore, remain uncertain about the centers' readiness to meet the full production workload anticipated to begin in about 2 weeks. Our prepared statement details uncertainty about the results of the Jeffersonville operational test, recent software changes, the Bureau's four-site operational test in late February, and other ongoing changes. In the interest of brevity, I will highlight the software changes and the four- site operational test. As you know, Mr. Chairman and Mrs. Maloney, the Bureau has decided to adopt a two-pass approach to its data capture operations. To implement this two-pass data capture solution, two sets of software changes or releases were required. The first release, designed to support the first pass, that is when they will get the 100 percent data which will be used for apportionment purposes, was completed in early February. This work involved modifying software to write the long form to a storage unit rather than presenting it to keyers for immediate action. This release was in place for the final four-site operational test. While Bureau officials have stated that the changes were successfully implemented, the Bureau has not yet provided us with the procedures used for testing those changes and the results. As a result, we do not have the information needed to know with any certainty whether these software changes are performing as intended. The second issue deals with the four-site operational test. To help prepare for the actual data capture operations during the census, the Bureau and its contractors conducted a final operation test from February 22 to the 25th. The test was important because it involved production-level workloads at all four data capture centers as well as centralized corporations and Bureau headquarters simultaneously. However, the test was limited in that it did not include all of the data capture center operations such as the center's ability to sort, check in and prepare questionnaires for processing. Most of the questionnaires used in the test had machine-printed rather than handwritten responses with the same answers on each questionnaire, which would, of course, simplify keying demands. Our overall point is some data capture operations have not been verified with production-level workload at all DCCs. Specifically neither the Baltimore nor Pomona center have successfully conducted operational tests of their ability to support a production load for sorting or documentation. Finally, let me comment very briefly on the address list problem and the notification letter. Since this problem was discovered by the Postal Service last month, we have been examining the Bureau's check to make sure that the problem was not repeated on census questionnaires. Once they learned of the error, the Bureau officials said that both the Bureau and its contractors checked a sample of six types of questionnaires with the preprinted addresses to ensure that they did not contain the same misprint that was on the notification letters. No such errors were found by the Bureau. My written statement details the types of checks that the Bureau went through. In addition, we independently reviewed a very small sample of questionnaires. We went to four post offices in northern Virginia and found that they did not contain any of the same misprints as the advance letters. In short, it appears that the Bureau is well justified in its confidence that the address letter error in the notification letter is not present in the census questionnaires. In summary, Mr. Chairman and Mrs. Maloney, let me echo a statement that both of you made far better in your opening statements about the importance of public cooperation with the census. With census day just over 2 weeks away and census questionnaires already out in the mail, one of the themes that we have been highlighting throughout the decade has been the importance of public cooperation with the census. The mail response rate, a key measure of that cooperation, will be central to determining the overall accuracy and cost of the census. The response rate will, therefore, provide the first indication of the success of the 2000 census. A high mail response rate will reduce the Bureau's followup workload and relieve some of the staffing and schedule pressures the Bureau confronts. On behalf of this subcommittee, we look forward to continuing to track these and other census operations for you, and we will be available to report at any point. Mr. Chairman, this concludes my statement. My colleagues and I would be pleased to answer any questions you may have. [Note.--The GAO report entitled, ``2000 Census, Actions Taken to Improve the Be Counted and Questionnaire Assistance Center Programs,'' may be found in subcommittee files.] [The prepared statement of Mr. Mihm follows:] [GRAPHIC] [TIFF OMITTED] T7019.011 [GRAPHIC] [TIFF OMITTED] T7019.012 [GRAPHIC] [TIFF OMITTED] T7019.013 [GRAPHIC] [TIFF OMITTED] T7019.014 [GRAPHIC] [TIFF OMITTED] T7019.015 [GRAPHIC] [TIFF OMITTED] T7019.016 [GRAPHIC] [TIFF OMITTED] T7019.017 [GRAPHIC] [TIFF OMITTED] T7019.018 [GRAPHIC] [TIFF OMITTED] T7019.019 [GRAPHIC] [TIFF OMITTED] T7019.020 [GRAPHIC] [TIFF OMITTED] T7019.021 [GRAPHIC] [TIFF OMITTED] T7019.022 [GRAPHIC] [TIFF OMITTED] T7019.023 [GRAPHIC] [TIFF OMITTED] T7019.024 [GRAPHIC] [TIFF OMITTED] T7019.025 [GRAPHIC] [TIFF OMITTED] T7019.026 [GRAPHIC] [TIFF OMITTED] T7019.027 Mr. Miller. Thank you for being here today and giving us a chance to discuss the status of the upcoming decennial census. Oversight is a critical responsibility of Congress, and there are four different agencies of the government involved in it: the General Accounting Office; the Inspector General of the Commerce Department itself; and from Congress, our staffs; and the Monitoring Board, which was created by Congress and the President back in 1998. Of taxpayers' money, $17 billion is involved in this constitutionally mandated requirement. We have a responsibility to make sure that the money is spent wisely and we get the best job possible, and that is what we are all concerned about. We appreciate your being here. I am delighted that there is going to be a meeting this afternoon to make sure that everybody is comfortable that they have access, because this has to be a transparent system. If there is not trust in the census, it threatens the whole system of government. Thank you for explaining why everybody should complete their form, because it does save the government money. The more people respond by mail, the less costly it is to followup. Otherwise you have to send people knocking on the doors. It is critical to every community, whether it is my hometown in Florida--what money flows to that city from Washington--or Tallahassee--is based on census data. So our own individual communities, whether it is education money, transportation or health care dollars, is based on census data or influenced by it, and so we need to do it, and so I encourage everybody to complete those forms. The first major mailing of the 2000 census was almost a failure. We are getting calls from people who are confused about the prenotification mailing. The letters were misaddressed, and there were no directions in English about what to do with the envelope. Also, the national 800 number was not included in the mailing. Director Prewitt called the first two problems an embarrassment for the Bureau. Let me ask you, what is your level of confidence that this problem won't occur again? Did the Bureau use focus groups to evaluate the particulars of that letter as far as the 800 number? And what grade would you give the Bureau for its attention to detail and quality control at this stage? Mr. Mihm. If I heard them correctly, there are at least three questions in there. Let me deal first with the issue of the focus groups and then the level of confidence that we have and then get back to the grading. Dealing first with the question of the focus groups, it is our understanding, and we will be happy to do more on this, that focus groups were not used in this. The second mailing, or the decision to combine the notification and the notice that there was the availability of a non-English language questionnaire, came rather late in the census process. It was after the dress rehearsal, and so there was not extensive testing of this. One of the lessons that comes out of this, and certainly one that we have been urging throughout the decade, when you are doing the census, you don't want to be using untested procedures. That is why we have census tests and a dress rehearsal so that only the smallest fine-tuning is done for the actual census. I attended your hearing last week and I heard, too, the Director say that he was embarrassed and how unfortunate it was that those errors crept in. For us the lesson is the importance of testing, making sure all procedures and operations have been thoroughly tested before the census. In terms of the level of confidence that the Bureau should have in other census mailings, dealing specifically with the questionnaires, the Bureau checked six different types of questionnaires--and this was after they discovered the problem with the notification letters--checked six different types of questionnaires, short and long forms, and did not find the problem in any of these questionnaires. Given that it appears that the problem with the notification letter was a systematic error, that is, a programming error that occurred in all 120 million letters, they didn't have to do randomized samples, you just had to do enough to make sure that it didn't occur in any additional cases. We didn't find any problems, and so it appears to us that the Bureau is well justified in its confidence that this has not crept into the census questionnaires. In terms of your third question on the grade, clearly there is a need for when all is said and done and the census is over--and we are going to be, at the request of this subcommittee and others, looking at the lessons learned from the 2000 census--to think about the quality control procedures, how they are constructed, and what sort of things that the Bureau looks at as part of its quality control. I think at this point it is probably an incomplete on the grade of the quality control standards, but that is the type of thing that we need to continue to look at, and we will on your behalf be vigilant as the census moves forward. Mr. Miller. I agree with the testing. In the pretest in Sacramento and Columbia, they did the second questionnaire? Mr. Mihm. Yes. Mr. Miller. We think that they should have had a second questionnaire, raise response rate 7\1/2\ to 15 percent, but a decision was made prior to Director Prewitt's tenure at the Bureau. The question is did they ever just pretest the prenotification letter by itself, or was it only at the test sites? Mr. Mihm. I will have to get you that information for the record. Robert, do you know offhand? Mr. Goldenkoff. No. Mr. Mihm. We will check on that. The decision to combine the notification letter and the availability--the notification that there be a non-English language form available did come coincident with the decision not to do a second mailing of the questionnaire. Mr. Miller. Why did they decide not to? Mr. Mihm. There are a number of reasons, according to the Bureau. One is that they say they received a large number of second questionnaires that were duplicates, and that it really overburdened the ability of their system to sort out and to check those duplicates. I know the Bureau was very influenced by a number of press articles that showed up in South Carolina in which people were quoted as complaining, ``I had just sent my form in, or I had received my form and right away they come back and hit me with another form.'' This is something that is going to bear some scrutiny from us as an option that needs to be seriously examined for the 2010 census. It is hard to imagine that a publicity and outreach campaign could not have been developed that would have said to people, ``If you got the first form, don't mail back the second.'' We all subscribe to magazines, and routinely get letters that say, ``If your payment and this bill have crossed in the mail, please throw this away.'' Thus, it was both the policy aspect as well as some technical aspects that caused the Bureau to be cautious on this. Mr. Miller. You mentioned in your opening statement that many of the access problems you experienced have now been resolved, and I am glad to hear that. However, can you please outline some of the problems that you encountered prior to their resolution and explain when and how these were resolved? Mr. Mihm. Our fundamental concern with the discussions that we were having with the Bureau was that it was seeming to take an inordinate amount of time and negotiation between us and very senior people at the Census Bureau. I heard and was taken by Dr. Prewitt's comments last week that he had to spend a third of his time on oversight issues, including requests that came from us. That is unacceptable, and we realize that. So our concern was that routine information that was readily available--that we knew from 1990 was readily available within the Bureau was taking a lot of give and take and a lot of negotiation. We had never been denied access to anything that we felt was important to get. We always came to a resolution. That is why I am so pleased with the new process that the Bureau has in place, we will be able to routinely get the information that we need with very little burden on senior census managers. I am very concerned in making sure that they don't view our data requests as burdensome to them. Mr. Miller. Right. No one wants to have the burden. But this seems like a bureaucratic tie-up every time you or anybody wanted information. They had to jump so many hoops. Why did you have to waste their time and your time when information was fairly readily available? Since you were involved in the 1990 census, would you compare your staff and access now to the way that it worked in 1990? Mr. Mihm. In 1990, we had about similar levels of headquarters staff, about seven or eight people working full time. As I mentioned in introductions, I am very, very fortunate to have my colleague Randy and some of his staff be able to help out on the data capture aspects. The headquarters complement was about the same time as last time. In 1990, however, GAO was 35 percent larger than we are now. In reviewing the 1990 census, we had in five different regions about two and a half people working 3 or 4 months on the census, primarily during peak operations, that were responsible for looking at the implementation of the census. We had a team in New York, a team in Philadelphia, a team in Dallas, and a team in Kansas City and Los Angeles where we had staff, so there was a larger field presence last time. In terms of the access, the source of some of our frustration was that when we were negotiating with the Bureau, we knew from 1990 that the information that we were asking for--and they were not denying this, I don't want to imply that there was a disagreement of this--but we knew that this information was available in a readily consumable form. We were having trouble getting equivalent levels of access that we had in 1990. Now with the agreement that has been reached and the new process in place, we have much more access than we had in 1990. Mr. Miller. I am going to let you go next, Mrs. Maloney. Mrs. Maloney. First of all, I would like to ask the audience and the panelists how many people received their census forms? Pretty good. How many mailed it back? We have got to get the response rate up here. We know that $2 trillion over the next 10 years is tied to census numbers. We know that it is very important for building our roads, bridges, and mass transit. All of government's funding formulas are tied to these numbers, so filling it out really is important to not only yourself and your own family, but to your neighbors and communities that they be counted in the census and counted in the funding formulas. First of all, I would like to make sure that there is no misunderstanding. Do you currently, Mr. Mihm, enjoy all of the access that you need? Are your questions being answered? Mr. Mihm. Yes, ma'am. As I mentioned, we are confident that the new process that has been put in place just within the last few days, because of the efforts of this subcommittee and certainly the Census Bureau, will provide us with the access we need to meet the oversight needs of this subcommittee. There will also be give and take, as you mentioned in your opening statement, on an operation this large about what is available and when we get it. But the routine operational information that describe how the census is going, is now available to us, and I am very pleased with that. Mrs. Maloney. You were involved in 1990? Mr. Mihm. At GAO. Mrs. Maloney. What was your role then? Mr. Mihm. I was the senior evaluator responsible for our reviews of the decennial census. Mrs. Maloney. Could you compare the access in 1990 to the access that you have now? Mr. Mihm. The access that is as a result of the new agreement is much greater than we had in 1990. Mrs. Maloney. Much greater? Mr. Mihm. Yes. As I mentioned earlier, part of our frustration when we were negotiating with the Bureau was that we were asking for information that we knew was readily available, and that was what was taking so long to get to us. Now, with this new agreement, it has minimized the burden on the Bureau and us, and allowing much greater access than we had in 1990. Mrs. Maloney. Could you elaborate on your greater access now? What didn't you have in 1990 that you felt like you needed? Mr. Mihm. Before I answer it directly, I don't want to give the impression that there was information in 1990 that we needed that we didn't think that we were getting. The Bureau with its new procedures this time has been very forthcoming in saying, here is not only what you are asking for, but here is some additional information which may be informative. They have been helpful in making sure and working with our people to make sure that we understand their cost and progress system, that is, what data will be available, what data will not be available and why that data won't be available. So it is just-- it has been a pleasure the last few days. Mrs. Maloney. I also understand that you feel that the guidelines for observers released by the Bureau about 3 months ago present no problems for your auditors since GAO has established protocols for dealing with these types of issues. Is that correct? Mr. Mihm. Yes, ma'am. The guidelines for us are not particularly relevant, and I don't mean that in a critical way about the guidelines. They seem to me to cut on a couple of categories. First, there is the series of things concerning basic collegiality, ``Don't expect rides from the airport, show up on time,'' we would hope that anyone would do that in any circumstance. Second, there is the separate set of points concerning how to conduct business, including making sure to not interfere with an enumeration while it is going on. We have a set of professional standards, the vernacular is in our yellow book, which allow us to certify that every job is done in conformance with generally accepted government auditing standards. These standards create a higher bar of professionalism and independence than the Bureau's Guidelines. There has never been a time when we have been talking to the Bureau about the importance of a field visit that the guidelines have come up or been referred to by us or Census Bureau. They are not a particularly relevant document for us. Mrs. Maloney. Did the Bureau ever use their guidelines in any way to deny you, GAO, or anyone else access to information? Mr. Mihm. I can't speak for anyone else. I know in our case we are quite confident and comfortable with the access that we have been able to get at the field level. We are able to make the trips that we need. One of the things that they ask for generally is a 2-week window before we make a visit. We have been able to give them those 2 weeks. If nonresponse followup is problematic in some isolated areas, and the subcommittee wants us to go to some areas, those aspects may be tested. But generally we have not had a problem with them, no. Mrs. Maloney. So they are generally just pro forma organizing tools. Show up on time, and you don't get a ride from the airport. You have never been denied information that you were trying to get in your professional---- Mr. Mihm. In the case of the GAO, we have not had a problem with those. Mrs. Maloney. You have gotten all of the information that you have wanted? Mr. Mihm. Within the last few days, as a result of the efforts of this subcommittee and the efforts of the Census Bureau, we have come to an agreement which will allow us to assist you in your oversight efforts. Mrs. Maloney. The General Accounting Office is a nonpartisan accounting and evaluating arm of Congress, the principal watchdog of the executive branch. Your staff is highly trained and is subject to professional and governmental standards. Would you explain these standards to us? Mr. Mihm. Yes, ma'am. In order to certify reports as being done in conformance with generally accepted government auditing standards, very similar to the point that the chairman made last week to Mr. Prewitt about private sector auditing standards, there is a whole series of requirements that we have to go through embodied, as I mentioned a moment ago, in what we call our yellow book for the very sophisticated reason that it has a yellow cover, and it lays out for us as auditors and for other auditors who want to do things in conformance with government auditing standards the very specific practices that we need to go through. In addition, we have a set of core values that we have to adhere to, which the Comptroller General has reinforced, of accountability, integrity, and reliability. We have a set of congressional protocols, which I know have been shared with your offices, on how we deal with the Hill and how we deal with requests, and so we have a large body of requirements that we have to follow in that regard. Mrs. Maloney. On the substance of your report, Mr. Mihm, your testimony reflects GAO's usual thoughtful job. It points out a number of what I would call minor problems. Certainly there doesn't seem to be anything which could threaten the success of the 2000 census? Mr. Mihm. As I mentioned at the end of my statement, the single greatest thing which has us concerned is fundamentally beyond the control of the Bureau, and that is the mail response rate. If we can, and the Census Bureau, and the efforts of this subcommittee and everyone else can get that mail response rate high, then the Bureau is in very good shape. I know I don't need to remind the subcommittee, but when we are dealing with such large numbers, even small marginal differences can have huge implications for the census. A 1- percentage point difference in mail response rate is about 1.2 million cases, and as Director Prewitt testified a couple of hearings ago, if they are off more than 2 or 3 percentage points, then there is trouble. Now, that is one area that is not within the Bureau's control. Within the Bureau's control there is the issue of the DCS 2000, which is their data capture system. I am going to ask Randy Hite, who is the expert on this, to comment on it. Mr. Hite. I would just add one point, the second pass operation for the data capture process, that is the changes that need to be made to DCS 2000, have not been made yet, so that is a development effort that still remains to be done. And as with any software development effort, you have risks associated with it, so that it is an unknown right now, and that is still an item that would give us some concern at this point in time. Mrs. Maloney. Well, overall would you say--when I read your testimony last night and heard it today, it seemed like everything seemed to be on target and going well and going forward. Is that your assessment of the census today? Mr. Mihm. I think in key areas things are going well. The Questionnaire Assistance Centers and Be Counted Program, that seems to be working very well. They have learned some lessons. They have localized staffing shortage. Over half of the district offices or local census offices have not met recruiting guidelines, but they are taking actions to address that, and that is the important point there. As I just mentioned, though, on the other hand, there is reason for caution in the mail response rate, if we can keep that high, and as Randy mentioned with the data capture system, there is still some real uncertainty there. We are just now beginning to see the processing of the data, and they are not at anywhere near processing peak data, the questionnaires are not there. It is not a fault of the system, so we don't know yet how they are going to work. There are a number of uncertainties with data processing. I don't know if you want to add some more on that. Mr. Hite. We don't have data now that shows that there is a problem, but part of our concern rests with the fact that there are data that we have not seen yet, not that the Bureau has not been forthright in providing it, it is not available for some of the tests which have been completed and data associated with the performance of the system, DCS 2000, during actual operations. Part of GAO's job is turning hearsay into facts so you can do the oversight that you do, and right now we don't have that--the basis for drawing those kinds of conclusions for you at this point. Mrs. Maloney. I understand from the chairman that we are going to take a field trip to the data processing center so we can see firsthand for ourselves. Mr. Miller. I hope to. Mrs. Maloney. Mr. Mihm, your testimony does not mention many activities currently under way, such as the update/leave, the telephone assistance, the Internet response, activities in remote Alaska and other update enumerate areas, preparations for the accuracy and coverage evaluation, and all of these are currently in progress. I assume the fact that because you didn't mention these activities, that they are on schedule and going forward appropriately? Mr. Mihm. No, ma'am. These are very important operations, update/leave and remote enumeration in Alaska. We have been focusing on what in the past have been the key vulnerabilities to a successful census. The mail response rate for the vast majority--the 96 percent of the households that have a mail- out/mail-back--the nonresponse followup efforts, staffing for the nonresponse followup processing, for overall, if the Census Bureau is successful in those efforts, the census will be successful. If they are not successful in those efforts, unfortunately, overwhelming success in some of these other efforts will not pull them through. So it is more of a reflection of what are the key vulnerabilities to the census. Mrs. Maloney. Thank you. Mr. Miller. Mr. Mihm, I am glad to hear that the Bureau, after, I think, my discussion with Secretary Daley and with Director Prewitt, has brought about a new openness. There is nothing to be hidden in here. We need to have the transparency. I hope this meeting will resolve any concerns. One of the concerns that was raised early on was when I saw these guidelines--I can see the case where all of a sudden there is a lot of fraudulent filling out of forms and other things going on, and if you have to give 2 weeks' notice to visit a community, wait a minute. You have to plan your flight schedules and hotel rooms and things like that, but you need to have the flexibility to respond where there is a problem. If there is a problem in Jeffersonville, you need to identify that. How do you sense these guidelines compare? Would Arthur Andersen have accepted these guidelines? Mr. Mihm. Well, the Comptroller General is a former managing partner at Arthur Andersen, so I have to be careful how I answer that. In my experience they are unusual for other agencies. I don't often, when I examine other agencies, and in our discussions I mentioned I have looked at quite a few in other aspects of my work, see something like that. However, the census is a very unusual undertaking in that it is so nationally diverse, the Bureau's dealing with so many temporary employees at various levels, and this is not to excuse or justify something if you have concerns, it is just to say that there may be explanations why they have it in this case and I don't see it in other cases. Another way that it is different in assessing the census is the real-time aspect. Much of GAO work is less real time than working on the decennial census. Perhaps the closest analogous situation could be the work that we do during filing season for the IRS. We do a great deal of real-time assesments of tax filing. I haven't spoken to my colleagues whether they have anything similar to this in their work. Mr. Miller. The Census Bureau made it out of the field ahead of schedule during the dress rehearsal. This seemed to be at the expense of accuracy since the level of proxy data was as high as 20 percent. What was the level of proxy data in 1990, and has the Bureau decided on a maximum level for 2000? What do you think are acceptable levels of proxy data? Mr. Mihm. One of the things that is unfortunate from the data capture operations in 1990 is we don't have a good number for the amount of proxy data in 1990. The Bureau did record what they called ``last resort data,'' which is their final attempt to get data, and presumably some of that or even a lot of that may have been proxy data, but nobody knows for sure. We do know, though, in some large urban local census offices, in 1990, upwards to 20 percent of the nonresponse followup enumeration was done using the proxy data. Unfortunately, the highest office in the Nation was the northeast Manhattan office where 42 percent of the nonresponse universe was enumerated using last resort and presumably proxy data in 1990. The Bureau's goal for 2000 is to have 6 percent of the universe be proxy data, and this is why it was such a concern that in all three locations during the dress rehearsal they were significantly outside of that number during the dress rehearsal--as you mentioned, 20 percent in Sacramento. Mr. Miller. Why did they allow such high proxy data during the dress rehearsal? Mr. Mihm. It is an issue that we are still talking to the Census Bureau about, and there has not been a thorough evaluation, in our view, of the causes of this high usage of last resort or proxy data during the dress rehearsal. As you mentioned in your question, it at least was helpful in getting out of the field early in the various locations or on time in the various locations. The Bureau believes, though, that part of it was just a failure at the enumerator and crew leader level to adequately follow procedures. So what they are trying to do is reinforce the procedures and the importance of going through the six contacts, at least three of those must be personal visits, before you do proxy data. That should be helpful. We would have preferred to see a fuller examination, I think, of really what were the causes to make sure that it was a lack of following established procedures. Mr. Miller. As you know, the use of proxy data can affect the quality of the census that we have, so it is very important as we go through this, if we are using too much proxy data, we need to be aware, and that is one of the things that we need to find out as we approach that proxy data period, which is in June. Mr. Mihm. Toward the end of nonresponse followup. Mr. Miller. You stated that with the update/leave you have fewer problems with recruiting because they can hire immediately. Mr. Mihm. Yes, sir. Mr. Miller. How long have some individuals sat in the applicant pool, and why is this important for hiring, because I am hearing from last week that people get hired in January, and they don't get the phone call until May. How much of a problem is that shelf life of that applicant pool? Mr. Mihm. It can be a significant problem or challenge for the Bureau in that they establish very ambitious recruitment goals which require a ramp-up in order to meet. Referring to our statement, we found in discussions with local census offices, where that office's update/leave operations is under way, they are the ones that are most ahead on recruitment because they are able to offer a job immediately. If you come and take the test and you pass the test, the Bureau would be able to say the job is available right now. For a nonresponse office the wait may be up to several weeks. This is at least part of the reason that the Census Bureau needs 9, 10, 11 applicants for every position. When they begin to hire, they burn through those lists of qualified applicants very, very quickly. The Bureau tends to find that if an applicant was looking for a second job, he or she will have found a job somewhere else, will have lost interest in the census, or will have forgotten that they have even applied to the Census Bureau. That is one of the reasons that they burn through the list of applicants very, very quickly. Mr. Miller. Let me switch to the data capture system, because you have been expressing concern about that for a month now. What is your current assessment of the risk of developing the new two-pass system for data capture? Mr. Hite. I will deal with the second pass first. Part of our concern there deals with any software development effort where you have a cadre of core software engineers that are very intimate with the behavior of the software, and those people are invaluable, and you want those individuals involved in any changes to the system as well as involved in the operations of the system. Well, the approach that the Census Bureau was going to take for data capture operations was to have those parties available as part of technical support during data capture operations because they are so important. But now with the move to the two-pass, concurrent with first-pass data operations, census will have to be developing the second-pass software. So those core software engineers are going to be diverted to the development effort for the second pass. The Census Bureau has said their primary priority is going to be supporting ongoing data capture operations, but nevertheless, you are taking a group of very important resources, and you are spreading them across two activities, and so you are stretching your resources. There is a resource risk there. If you look at the schedule that has been established for the development and the conduct of the first pass, the testing of the second pass, and the initiation of the second pass, there is very little tolerance for any slippage in the schedule. As soon as first-pass operations conclude, then census needs to have tested the second pass, so that it is ready to begin second-pass operations. So, for example, a lower response rate that could cause first-pass operations to be extended is going to have an impact on second-pass operations beginning on time. That is the primary risk that I see with regard to the second pass. With respect to the first-pass operations, I don't have data now that shows me that I shouldn't have confidence in the performance of the first-pass software, but then again, I don't have the data that I need in order to have that confidence. I need to see the results of the software integration test or the system integration test, and we have not seen that yet. The first-pass software was exercised as part of the four-site OTDR, the operational test, and we haven't yet gotten the report on the results of that test. So we need to see that kind of data in order to put us in a position to have confidence about the performance of the two- pass software. Mr. Miller. Why haven't you received that? Is there a delay, or is that a problem? Mr. Hite. There is a normal period that transpires from the conclusion of the test and the development of the test report, and I believe in the case of Census it has been about a 30-day period for that, and that is not unusual. My understanding is that, for example, the Jeffersonville OTDR report was to be available or sent to Census Headquarters May 6. We have not received that yet. I don't have a date off the top of my head when the four-site OTDR report was to be available, but we expect to have that request fulfilled when that report is available. Mr. Miller. The second-pass data, is that the long form data? Mr. Hite. The way it works, the second pass will retrieve the long form images from a disk storage, and from those long form images they will present to the image keyers the fields where there is low confidence so that those keyers can correct that data. So it does deal with the long form images. However, during first-pass short form as well as long form, images are optically read, and that data are forwarded to Census Headquarters. And they are the correct data that are produced during the second pass that are also forwarded to headquarters and will supplant the data that was forwarded to headquarters as part of the first pass. Mr. Miller. How comfortable are you that the optical reading, the whole process from A to Z, unloading it from the trucks, processing them through, scanning, filing, capturing the data and the handwritten forms, how comfortable do you feel that it is going to work? Mr. Hite. You have to break it up into each of the steps. With some items I am quite confident. The optical character, the optical mark recognition, I think the testing has proved that to be within accepted bounds. With regard to the check-in, the sorting at production levels, that depends on which site we are talking about, because at some sites it was tested, and it was successfully tested, and they demonstrated their ability at those levels. It was not successfully tested at Pomona and Baltimore. There we have questions.I21The scanning of the images at production levels, that is what the four-site OTDR was to give us. We understand that there were production levels put through the system. There are issues about the key from image productivity rates at some OTDRs. For example, at Phoenix, those rates were achieved. At Pomona they were not. At the four-site OTDR, because of the nature of the forms which were used, which were basically forms with preprinted information, all the answers being the same, you are not going to get a true read of the key from image rates because it reduces the demands on key for image such that the keyer sees the same section of the form to correct any keying from. So it is variable depending on what steps in the process we are talking about. It is variable depending on what sites we are talking about. Mr. Miller. Thank you. Mrs. Maloney. Mrs. Maloney. Throughout this, shall we say, effort to get an accurate census, recruitment has been a tremendous concern, given the extremely strong economy and the low unemployment rate, and, in fact, the Census Bureau is still hiring, and you can call if anyone is unemployed and needs a job at 1-888-325- 7733, or you can check the Census Bureau's Internet site at www.census.gov/jobs2000, so we can try to get more people informed to go to these jobs to bring this recruitment effort up. But as of March 2, 2000, recruiting nationally is slightly ahead of schedule; that seven regions are below their goals and five are above; that 22 of 520 local census offices are experiencing severe recruitment problems. Is this fairly good, or how does this compare to 1990, the recruitment levels? Mr. Mihm. I think it is a good story at least thus far, and we will have to see how things pan out when they actually begin to hire. Certainly the number of offices that appear to be in the severe category, which is, granted, not a term of art, is smaller this time around. They vary in numbers, we are looking at 22 offices, and the Census Bureau is tracking a bottom 50. Last time in 1990, there were over 100 offices that were experiencing some difficulties, or it appeared that they would be experiencing some severe difficulties. I guess the concern is and has always been that at a national level you are going to do well. Nationally you can meet these goals in recruiting and hiring. The challenge to the census is to be able to hire people across the country. There is no other employer that has to hire people in every neighborhood in the country. That is the tough thing that they face. They have made a strong commitment, which we completely agree with, for data quality reasons they want to hire people to enumerate their own neighborhoods. Thus, extra applicants in one part of town will not help the Bureau in another part of town. That is the challenge that they faced in 1990, and they still face in 2000. Mrs. Maloney. The Bureau takes a number of steps to address local recruiting problems which you outlined in your testimony. Do they seem adequate to you? Mr. Mihm. Yes. They seem to be taking important and aggressive action to address these, and I should have added this is one of the important differences I think this time compared with 1990. Certainly in 1990, when they had problems, they would go out and take actions. But this time there seems to be a much more focused effort in identifying low-performing local census offices going in and using targeted recruitment mailings, in some cases raising pay rates for enumerators--that is something that they obviously only want to do as a last resort--bringing in swat teams to help with that recruitment. There are a whole series of efforts that they are undertaking at a local and regional level. What we will see over the next 3 weeks is whether this extra intense effort is actually going to make a difference. We will be tracking those offices as well the Bureau. Mrs. Maloney. So it seems that they have learned from the problems of the 1990's and have taken concrete steps to address it. Can you think of anything that the Bureau should be doing that it is not doing to help in recruitment? Mr. Mihm. We do not have an open series of recommendations on what they should be doing in the recruitment area. The only matter that we offered up for Congress's consideration was possible legislative exemptions. The Bureau, as I mentioned in my statement, is pursuing at least some of those dealing with welfare or TANF down with State governments, and it is reportedly making some progress in that regard. The fundamental challenge that the Bureau faces in terms of recruitment is the nature of the census job, it is not the most attractive. It is short-term and temporary, obviously. It doesn't have benefits. Although it is very important for our Nation that this work be done, it is a hard sell due to the combination of the nature of the job and labor markets. Mrs. Maloney. I introduced legislation to respond to your last report, H.R. 3581. You mentioned the TANF problem, which Congresswoman Meeks has worked hard on in trying to allow current welfare recipients to receive their benefits and work for the Census Bureau. One of the problems of dealing through a State system, some States will not take those steps. My own State has not taken that step, so Federal legislation would greatly help. I know that the chairman supports these efforts. We could do this in a bipartisan way and expand the field of possible workers for the Census Bureau, by also allowing military personnel to take these jobs. So that is one thing that we could do in a positive way to help and to respond to GAO, and I hope we will be able to move that this year. Anyway, Mr. Hite, to go back to you on the data capture, you discussed the decision to move to a two-pass system and some of its implications. My understanding is that the decision to move to the two-pass system resulted from the key from image productivity rate. It seems like a sound decision, but what is your view? Mr. Hite. The two-pass solution is one of three options that the Census Bureau considered, one option of which is to do nothing, which is always an option, but was unacceptable. The two-pass approach is technically feasible. It is a reasonable approach. The key to any type of approach or any type of plan or solution is the implementation of it. With any implementation there is risk associated with it. I have no objection to the approach that they have taken. The key will be to make sure that it is implemented according to schedule and according to specifications so that it functions as it is intended to do. Mrs. Maloney. Wasn't the February test a full load on the testing? Mr. Hite. The four-site test yes, it was production-level workloads first being scanned through the system to create the digital images, and then for the data to be captured through the optical recognition, and then forwarded on to headquarters. It also involved, as I mentioned, the changes for the first- pass software so that the images for the long forms--I'm sorry, the sample data associated with the long forms could be written to disk for storage until they would be needed for the second pass. They also tested some centralized operations, the central coordination center, the technical support, the individual site support. So it tested a number of things. The test is very important, and hence we look forward to seeing the test report. Mrs. Maloney. Earlier you stated that you were uncertain about the data capture centers' readiness to meet full production-level workloads. Can you describe what is a full production-level workload? Mr. Hite. The volumes that the data capture centers expect to be able to process during peak production are roughly 1.5 million forms per day. Thus far we are looking at 17,000 based on the first 2 days of operation. The 1.5 million are the production-level volumes, and, as I understand it, for the four-site OTDR, the volumes were upwards of 2 million, so they were pushing production-level volumes through the system beginning with the scanning process, but it didn't include the check-in and the sorting and the preparation of the forms prior to the scanning beginning. Mrs. Maloney. I was under the impression that all of the four centers were supposed to be as identical as possible. Is that correct? Mr. Hite. The centers are identical in terms of their configuration. They will all rely on clusters, but the number of clusters vary depending on the expected workloads. In the case of any type of operation that relies on human beings, the centers are going to be different in that the people operating the system are different. And in the case of some centers, some forms will be processed--for example, the Jeffersonville data capture center--that are not being processed at other centers. So there are some differences, but from a physical configuration standpoint, yes, they are the same. Mrs. Maloney. Based on what you know today, do you have any reason to believe that the data capture centers or systems will not perform as expected? Mr. Hite. No, ma'am. As I mentioned, I don't have the basis to believe that they are not going to perform as expected. I need to see information for me to say that I have confidence that they will perform as expected, and I haven't seen that information yet. Mrs. Maloney. What additional information do you need? Mr. Hite. I would like to see the software integration test results and procedures for the first pass, the systems integration test procedures and results. I would like to see the test report for the Jeffersonville OTDR, and the test report for the four-site OTDR. Mrs. Maloney. Would you describe that test for us? Mr. Hite. The four-site OTDR? Mrs. Maloney. Yes. Mr. Hite. This was a test of all four DCCs operating in conjunction with headquarters operations in processing production load volumes of questionnaires. The test began with the scanning operations, the forms being fed into the scanner to create a digital image of those forms, the forms to be optically read for the marks and the characters, and for those data then to be prepared and formatted for transmission to census headquarters and confirmation of the receipt of that information. It also involved a number of support activities associated with the data capture operations, including onsite technical support and centralized technical support, the operational control center and headquarters. Mrs. Maloney. Thank you. It is good to hear that you expect them to perform as expected. Thank you, Mr. Chairman. Mr. Miller. I will ask a couple of short questions. Of the two-pass system, what impact does that have on the cost and any delay of data release? Mr. Hite. With regard to cost, there is a marginal cost associated with the two-pass operation. We do not--we have not yet received an estimate on the total cost associated with that, but I can tell you the cost number that I have seen associated with the development contractor making the changes to the system, and in that case I have seen a $35.2 million figure, which would include $10 million for the disk storage, for storing the digital images awaiting the two-pass process, and also for the software changes associated with the two-pass approach. There is also cost associated with a tape verification back-up system. There are additional costs associated with keeping the data centers open and operating longer, but we have yet to receive a cost estimate on that. Mr. Miller. And the delay of the data release, is that going to be impacted? Mr. Hite. Chris, I don't know if you want to comment on this or not, but we have been told by Census that it is not going to have an impact. We don't have any independent information on that. Mr. Mihm. We still need to look into that, Mr. Chairman. Mr. Miller. One response to Mrs. Maloney's proposed bill. I do support Mrs. Meeks' bill allowing people on welfare to work. However, the reaction we are receiving is that once we passed welfare reform in 1996, and it has had a great deal of success, one of the ideas is that we allow States to make decisions. And if we start opening it up and mandating things, there is real concern by many that we are breaking that original agreement. That is the concern that we are having. I am glad to hear you say that some States are looking at the TANF requirements to make it possible. I understand the concerns about mandating something out of Washington. In some areas like working military--INS and IRS people cannot work for the census, for good reason. And the same way with military, so we have to be careful with any current Federal employees. Let me ask two more questions. You mentioned in your testimony that there have been multiple delays in the mailing of both recruiting and teaching material for the Census in Schools Program since early fall. What is the cause of these delays, could they have been prevented, and what do you feel the impacts of the delays will be? I have enjoyed the opportunity to go to schools, and I think it is a good program. What are the delays and why? Mr. Mihm. The story of the delays is--it is a long and complex one that we won't go into at this time. We are still trying to disentangle it from the Bureau. Robert has focused on this. Mr. Goldenkoff. The cause of the delays, as Chris said, is unclear. I think it is something that the Bureau would like to know as well. We are still trying to unravel it. We do know first that there were multiple actors involved. There was Scholastic, the company that was responsible for creating the material. There was another contractor which was responsible for distributing the material. There was the Government Printing Office, and there was the Bureau itself. So there was a web of these different actors involved that were not always going in the same direction. In October 1999, the distributor ran out of some of the teacher kits, and you have probably seen them, they are very eyecatching and glossy, and the local partners we have spoken to have all given them high marks. The Bureau was not immediately notified that the distributor had run out of some of the teacher kits. In the meantime, in fall 1999, there were some delays in developing some of the materials that were used for the full recruiting package that went out to the principals. There was one period of recruiting in the spring of 1999 when teachers in hard-to-enumerate areas were invited to participate in the program; but then there was a second mailing to principals and others which was supposed to take place in September. Because of the delays within Scholastic, that mailing did not go out until December 1999. That is the framework for delays. Some of those--the problem was cleared up. As of mid- December, the distributor resumed filling orders for the teaching kits, and principals and others started to get their kits beginning in January. As was said earlier, the turnaround time is now 2 to 4 weeks. As to whether they get them in time, it depends partly on when the material is requested. We are in that 2 to 4-week timeframe right now. But also, I should say that the material is available off the Bureau's website. It can be downloaded; however, it is not the same level of quality. That is one of the local Complete Count Committees did, they actually downloaded it. Mr. Miller. Thank you. I am sorry that the information is not getting out as quickly as we had hoped. We are big supporters of the Census in School Program. Mr. Bass has gone to schools, Mrs. Maloney has, and I have, and I know that they have been very positively received by students, and they hopefully get their parents motivated. One final question, and you will be back again next month. Would you describe what type of oversight the GAO has planned for the next 6 months? Mr. Mihm. Subject to your approval, because we are a staff agency of the Congress, and so we work on your behalf, but over the next several weeks, as we ramp up for nonresponse followup, we will be taking full advantage of the access to information that this subcommittee has been helping us secure in focusing on mail response rates by local census office, looking and comparing that to where they are on their recruitment efforts, and trying to get a sense is there a subset of census offices where we are going to have the greatest trouble in meeting schedule and data quality. And, Mr. Chairman, you have been particularly clear in making sure that we are focused on proxy data. We will also be looking very closely at the ACE. We are looking at a series of key operational indicators that will tell us and hopefully inform the subcommittee on how the ACE is going while it is going on. One of the things that was a bit problematic last time, that is in 1990, was that we only get a sense of the quality of the ACE long after a decision has been made to adjust--in that case, obviously, not to adjust--census data, after much of the interest has dissipated and the public's attention goes on to other things. Only then do we get a full view of the quality of the adjusted number. We are going to be looking at operational indicators at the ACE. We fully expect, putting my colleague on the hook here, that we will continue to look very, very closely at data processing operations and following up on these reports as they become available from the Census Bureau and seeing how they work, and I guess at the largest sense being available for the subcommittee to look into things that can best support you in your oversight efforts. Mr. Miller. Thank you. Mrs. Maloney. Mrs. Maloney. Census in the Schools is one of our favorite programs. I have gone to a number of schools and had lesson plans with students. It is inspiring to watch them learn about the census and informing their own families about the census and its importance. Weren't the delays as a result of the decision really to reach out to 100 percent of the schools, not just 40 percent of the schools? Wasn't that one of the reasons for the delays? But I just want to make clear that there are no continuing problems now, are there, with Census in the Schools? Mr. Mihm. Let me deal with the second part and ask Robert to respond to the first. Our information from the people that are running the Census in Schools Programs is that they are still experiencing these delays in getting the kits. If you are looking at a 4-week delay at this point, it is not going to make sense for a teacher to submit a request for a kit. It is in response to these delays that some of the Complete Count Committees and regional census staff that we have talked to have been taking action, either downloading the information off of the Internet in the case of the Dade County Complete Count Committee. They were disappointed. They said, we basically missed the window of opportunity in Dade County for Census in the Schools to be effective. So there is still a bit of a delay. The Census Bureau has made clear to us that they will continue to process any requests as soon as they can as long as they have kits. So there is no drop-dead deadline that requests must be in by. We can still hope for some response or kits to be met out there through nonresponse followup. Mr. Goldenkoff. The other part of your question as to the reason for what--the expansion of the program, we are still looking into that. Mrs. Maloney. Well, I want to go back to the recruitment challenge and really to compliment GAO not with just coming forward with rhetoric, but with concrete proposals that we in Congress and our communities can enact to increase the pool of people that can go out and be enumerators and help get the count. The chairman said we should not interfere with the States, but the census is a Federal program. It is a Federal project. It is not about politics, it is about people and making sure that every person literally counts. We have heard from GAO concrete examples of how we can have a contingency plan to help get the count up and to make sure that the recruitment levels are met by allowing welfare recipients to have these jobs and not lose their TANF benefits, and by allowing military personnel to work for the census. These are important contingencies, and I feel we should act on them because we know what is at stake. We know in the last census 8 million Americans were missed, and over 4 million were counted twice, and we also know that there is a disturbing civil rights trend that the people that are missed are overwhelmingly the poor in rural and urban areas, African-Americans, Latinos, Asians, Native Americans, children. These are the people that the Census Bureau, the scientists have told us are overwhelmingly missed, and it seems to me going back to the stated purpose of hiring from neighborhoods, in these poor neighborhoods a very likely labor pool would be welfare recipients who could go out and get job experience to help them move from welfare to permanent work, but also be part of helping their neighbors and their communities get an accurate count, because this census is really the basis of virtually all demographic information that is used not only in government, but by journalists and community leaders, educators, policymakers, and businesspeople when they plan where they should put their businesses. Everyone relies on accurate census data, so I really-- although we usually agree on many things, I really feel that we should on a Federal level act since the Federal census is a Federal program and so much is at stake, literally $2 trillion over the next 10 years, literally dollars that are needed for schools and emergency response, for education. We should followup on GAO's concrete proposals that they put before us on ways to assist the professionals at the Census Bureau in reaching their recruitment levels. That's my last question if this is the last round. If not, I have a lot more to say. Mr. Miller. We all have the same goal and objective: to have the best count possible and have, in effect, no differential undercount. The Bureau has developed a plan to count those undercounted populations. We are anxious for the results. We have been very supportive of the total dollars spent on the census. It is going to cost $6.8 billion. My understanding is that--I know that the Bureau has been working very closely with the welfare-to-work population. They developed a program with Goodwill. My understanding is that some 37 States and the Virgin Islands have granted exemptions, so we are moving in the right direction. I feel confident we are going to have a good census. We appreciate your oversight responsibilities. I am happy to see that there has been increased transparency made available through the Census Bureau, and I look forward to hearing that from all of the other agencies involved. Thank you for being here today. Mr. Mihm. Thank you for your efforts on the access issues, Mr. Chairman. Mr. Miller. I ask unanimous consent that all Members' and witnesses' written opening statements be included in the record. Without objection, so ordered. In case there are additional questions that Members may have, I ask unanimous consent that the record remain open for 2 weeks for Members to submit questions and for the witnesses to submit written answers as soon as practicable. The meeting is adjourned. [Whereupon, at 3:35 p.m., the subcommittee was adjourned.]