Dr. J. Steven Landefeld Director, Bureau of Economic Analysis Dr. Landefeld has been Director of the Bureau of Economic Analysis (BEA) since 1995. Prior to becoming Director, he served in a number of other capacities at the Bureau, including Acting Director, Deputy Director, and Associate Director for International Economics. While at BEA, he has led a number of pioneering efforts in statistics, including the introduction of unbiased estimates of real Gross Domestic Product (GDP) and prices, the development of monthly estimates of trade in goods and services, alternative balance of payments accounts, integrated economic and environmental accounts, and the use of data exchanges with foreign banks to improve international capital estimates. Dr. Landefeld aso has led a number of managerial improvements at the Bureau including the introduction of a performance-based personnel system, the development of "private-sector" financial accounts (BEA was one of the first Bureaus in the Department to receive an unqualified opinion from an outside auditor on its financial statement), and the move from an antiquated mainframe to an integrated micro-computer network (BEA was the first major statistical agency to successfully make such a move). Before coming to BEA, Dr. Landefeld held a number of positions, including Chief of Staff for President Reagan and Bush's Council of Economic Advisers, Director of the Business Issues Analysis Division at the Department of Commerce, and Research Assistant Professor at Georgetown University. He has authored numerous professional articles and has received the Henri Willem Methorst Medal from the International Statistical Institute, two Abramson Scroll Awards from the National Association of Business Economists and the Gold, Silver and Bronze Awards from the Department of Commerce for his work. Dr. Landefeld has served on numerous professional committees and working groups including those of the Organization for Economic Co-operation and Development, the International Monetary Fund, and the Conference on Research in Income and Wealth. GENERAL DISCUSSION Mr. ROGERS. Thank you both for your statements. This is a pitched battle that we are into here. It has been for several years now. After the 1990 census, as you say, no one was happy with it, and here we are on the verge of I think another disaster in the making. You two gentlemen are not new at the Department, but you are new in the positions that you are holding. You are acting personnel. So it is going to be hard for us to blame you for what is happening. You are a little bit like the old fellow that was hauled into court, charged with setting his bed on fire at the rooming house where he was staying and, also, being drunk. He came into court and he pled guilty to being drunk. The judge was trying to find him also guilty of arson for setting the bed on fire, but he had this excuse. He said, "Now, wait a minute, Judge." He said, "I admit I was drinking, but that bed was on fire when I got into it." [Laughter.] Mr. HOLMES. Your point is well taken, Mr. Chairman. Well taken. [Laughter.] Mr. PRICE. There was heat before I got there. Mr. ROGERS. You knew what a hot seat you were getting ready to sit down in, though, didn't you? Mr. PRICE. I did. VULNERABILITIES IN CURRENT CENSUS 2000 PLAN Mr. ROGERS. We have got the IG report, as of December 30, 1997, essentially saying that every major component of the Census' own plan is at risk because the Bureau had not made major decisions on design. The plans they did have, such as the address list development were not working, and the plan had such a tight time schedule that it probably would not be executable without major cost growth and/or degradation in quality. And then, just yesterday, the GAO issued their report, and they came to many of the same conclusions, identified the same areas and issues. However, the GAO report is even more important because, while the IG recommended trying to fix and then test the problems in the dress rehearsal, the GAO report looked at how the Bureau was fixing and testing these problems in the dress rehearsal, and they said-in fact, the title of their report is, "Preparations for the Dress Rehearsal Leave Many Unanswered Questions." GAO said that the dress rehearsal is, "The last opportunity for an operational test of its overall design of the 2000 Census and to demonstrate to Congress and other stakeholders the feasibility of its plans." They found problems in every key activity of the census plan, including address list development, local outreach and promotion, staffing, and statistical sampling. They said, "All of these activities are still facing the developmental and/or implementation challenges that led GAO in 1997 to raise concerns about the high risk of a failed census in 2000.” And GAO concluded, they said, "With the decennial census just two years away, we find it troubling that the Bureau finds itself facing several ongoing and newly emerging operational challenges." Here are some of the areas they say are vulnerable: One, they say, "The accuracy of the Bureau's address lists and maps is uncertain and local reviews may be too sporadic to greatly improve them." GAO essentially questioned whether the Bureau's fixes for its address list problems will work and notes they will not even be tested in the dress rehearsal. Two, quote, they say, "The Bureau's outreach and promotion efforts face obstacles that could impede its ability to achieve its mail response rate objective." They found that in the preparations for the dress rehearsal key components of the Bureau's plan to increase mail response rates were not working as planned. Three, quote, "The Bureau's sampling and statistical estimation design faces methodological, technological, and quality control challenges." "Most disturbing, like the IG, the GAO found the Bureau's plans and procedures to 'adjust' the census numbers face serious questions and may not be able to effectively be implemented" and found that the Bureau was already running into serious problems in preparing to conduct the adjustment procedure; that is, the integrated coverage measurement program in the dress rehearsal. GAO found, "The Bureau has made several missteps in drawing the ICM sample because these errors went undetected until relatively late in the sample selection process. GAO is concerned about the Bureau's ability to catch and correct problems.' All of that is calculated to cause me, at least, a good deal of nerv ousness. Now, the IG recommended that you do five things immediately. This was in December 1997. One, prioritize and access the readiness of its major design components; two, simplify the design; three, realistically reassess costs; four, communicate the results both internally and externally; and, five, redirect the dress rehearsal. Have you done any of those things and what has changed? Mr. HOLMES. I guess, first and foremost, Mr. Chairman, the challenges that are detailed in both of those reports are challenges that most of us are aware of. I think, as Lee Price mentioned, that this is the largest peacetime activity that this country gets involved in. I can say, from a personal perspective, having lived at the end of the food chain on three of those, there is absolutely nothing as large and as complex as doing a census. But I think it is also important to keep in mind that, while there is an enormous amount of criticism associated with things that we do not know about the dress rehearsal or things that we do not know about the plan, it is important to keep in mind that we have been doing dress rehearsals since the 1930s, and the whole idea behind doing a dress rehearsal is to get some idea what components of the plan are working and which ones are not working and to make adjustments accordingly. Yes, we recognize that there are some problems with the address list. That was the main reason that back last year the Census Bureau decided to re-engineer the process and came to the Congress and asked for $108 million more to do that because it became very clear that the route that we were traveling to get the address list ready for the Year 2000 was not going to get us the level of quality that we wanted. So it would have been irresponsible on our part to continue down a path that we knew was going to fail. Again, that is the main reason for doing that. The GAO expressed some concerns about local involvement in the address list. That has always been a problem. If my memory serves me correctly, in the dress rehearsal, roughly, 50 percent of the Governments, at least in the South Carolina area, participated in the program. That is a tremendous increase from what we had in 1990, where it was around 30 percent. It is not mandatory for Governments to participate, but we do ask them to participate because it is important to them CENSUS' RESPONSE TO IG RECOMMENDATIONS Mr. ROGERS. I understand. But my question was did you follow the IG's five things that he recommended you do? Mr. HOLMES. Well, in terms of the address list, yes, we did. Mr. ROGERS. How about all of the others that I have mentioned, the five that I mentioned to you. Mr. HOLMES. Would you mind repeating the five again, sir. Mr. ROGERS. Prioritize and access the readiness of its major design components, simplify the design, realistically reassess costs, communicate the results both internally and externally, and redirect the dress rehearsal. The IG told you to do that in December 1997, about three months ago. Mr. HOLMES. One of the things, if I am not mistaken, that was done to prioritize the activities was the development of the master activity schedule, which lists some 3,000 or so discrete activities that, essentially, spell out all of the major activities that are associated with the census and their interconnection. So if something goes wrong on one, it plays out through the Mr. ROGERS. Well, that was done before the IG told you to fix things, what you just described. Mr. HOLMES. The master activity schedule? GAO RECOMMENDATIONS Mr. ROGERS. That was being done before the IG's report. Well, let me go on. Have you studied the GAO report as of yesterday? Mr. HOLMES. No, sir, I got a copy of it this morning, and I was trying to get myself prepared for this hearing today. Mr. PRICE. I did look at it. It looks very similar to the draft that we had a couple of weeks ago, and I have read that. Mr. ROGERS. Have you taken into account what they suggest and recommended? Mr. PRICE. Well, yes. There are really four main categories that they raise in each of their long chapters. The one that overlaps with several issues you raised of the IG about statistical design, prioritizing, and simplifying it, the Census Bureau staff, I am told, is working on that, has been simplifying and prioritizing. They did complete, since that December 30th report to Senator McCain, they have completed the design for the dress rehearsal, and they will soon be completing the proposed design for 2000. As committed last July, that will be completed this summer for public review and peer professional scrutiny over the course of the next year before it is actually finalized for 2000. GAO raised several issues that are really serious challenges in any kind of census we do. One of them is the employment challenge. They are saying it is going to be hard to hire enough people to do it Mr. ROGERS. No, I am talking about they have severe criticisms of your methodology and even statistical sampling. This is a harsh. report, and you are not responding to what they are suggesting in there. GAO in their report raised the same problems that were raised by the Inspector General four months ago, that are still around and still not being addressed, and here we go again. I was here for the 1990 census on this Subcommittee, and that was under the Bush Administration, and it was botched, and I swore at that time that I would do everything I needed to do not to have it happen again in 2000 and here we go again. I think this is going to be worse than 1990, to be frank with you. And you are aware of the problems and you will not address them, and I do not know what else we can do. Address list development, for example. Everybody in the world will tell you that when you mail that first questionnaire out that the best way to get response from the non-responders is to mail a second letter out. It is the most cost-effective way, right or wrong? ADDRESS LIST DEVELOPMENT Mr. PRICE. Yes. That is what we found in our tests and that is what we plan to do in the dress rehearsal and 2000. Mr. ROGERS. And, apparently, now you are dropping the idea of mailing the second letter. Mr. HOLMES. No, sir, that is not the case. What we are doing is we are mailing an initial questionnaire. There is a letter that actually goes out before that, and there is a reminder card. What we are not doing is we are not blanket mailing a second questionnaire to everyone, at least that is not the plan for 2000 at this point, but we are testing that in the dress rehearsal. If it works, then I think it is safe to say that we will come back to the Congress and ask for additional funds to do that, but we did not think that it was appropriate to spend I think it is $42 million to have questionnaires printed up when there are some concerns about whether or not that could cause additional problems for us in terms of duplication or people may just have a less-than-positive reaction to it. So what we are doing is, again, testing it in the dress rehearsal. Mr. ROGERS. You are testing what in the dress rehearsal? Mr. HOLMES. We are testing the concept of doing just what you were just asking us to do; mailing a second questionnaire, yes, sir. Mr. PRICE. The address list is a very serious issue, and I would like to respond, if I could, on that. We have done a mail-back census since 1970. So the three previous census' we have had a mail-out census. We have also done 49-025 98-18 |