And this is an issue that has been talked about. And I know the Constitution reads "residents." If you are a resident in the country. And it is an issue that has been discussed in this Congress I know over the years. There are some cases in the courts right now with respect to Utah and North Carolina with Americans that are outside the country and not being counted.

Anyway, I think what Congress did do a number of years ago, was say we would count all the military officersNo? They have not done that?

You want to respond to that, Bill?

Mr. BARRON. At the time of the 1990 Census, there was a considerable interest in Americans stationed overseas, and in responding to that, there was an agreement that the Census Bureau, in the 1990 Census, would count overseas stationed military as well as federal employees.

That is another issue where we need to report back to Congress. I think it is by September 30, 2001, that we have been asked to do an assessment of what we can do in terms of counting all Americans overseas. It is a very difficult problem. I am not real optimistic about what we are going to do, but we will provide a report and sort that out with folks up here as soon as we can.

Mr. SERRANO. Well, Mr. Chairman, I am going to give up the mike for this round.

But let me just say, I would make a request, Mr. Secretary, if you could have your legal minds get together and maybe look at the question and advise us

Secretary EVANS. You bet.

Mr. SERRANO [continuing]. On the question of whether the Constitution says count people who live within the States, or whether it could be interpreted to say count all folks living under the American flag.

Secretary EVANS. You bet, Congressman. We will do that by September 30, 2001, as mentioned above, and we also would like to get with your Staff on the Cuba issue.

Mr. SERRANO. Sure. Thank you.

Mr. WOLF. Mr. Latham.

Mr. LATHAM. Thank you, Mr. Chairman.

Secretary EVANS. Good morning, Congressman.

Mr. LATHAM. And welcome.

Secretary EVANS. Thank you, sir.

Mr. LATHAM. And I know you will do a good job because you have Brenda Becker on your staff, so congratulations.

Secretary EVANS. She said she does not want to come up to the mike right now.

Mr. LATHAM. Oh, okay.

METROPOLITAN STATISTICAL AREAS

I have a question of you that I think several areas have problems with, and in Sioux City, Iowa, we are in what they call the Sioux Land, because it includes Sioux City, Iowa, South Sioux City, Nebraska, North Sioux City, South Dakota, and currently the metropolitan statistical area just includes South Sioux City, Nebraska, and Sioux City, Iowa, and does not include North Sioux City, South Dakota, or Union County, South Dakota, which includes the head

quarters for IBP includes Gateway's home facility with 5,000, 6,000 employees and a new, huge housing development up there.

And we have been trying for several years to try and get all of this included in the MSA, and I think we are making some progress but apparently there is a concern with other statistical areas which are going to take more time and more information to establish, that all of the designations are going to be held up probably until the year 2003, which is extremely unfortunate for that region as far, as you know, the basis for the economy and what they can talk about bringing in business and also for federal grants and those type of programs.

I just wondered, and maybe we need to refer this, I would like to visit with you sometime or your staff, on what we can do to expedite that, but it is a huge issue for us, and very important.

Secretary EVANS. I would like to learn more about it. I am aware of the problem and the issue. I have been informed that we will not be providing the data that OMB needs until March of 2002, and then I do not know how long they need with the data to massage it and maybe that gets you out to 2003.

And so I have asked our staff to look at that, to see if there is a way to compress the time line by which we can submit the data to OMB, which is the body that will set the SMSA.

Mr. LATHAM. Right.

Secretary EVANS. So we will work with your staff on that, and we will get back to you to give you a hard answer of yes, we think we can speed it up, or, unfortunately, it is impossible.

Mr. LATHAM. The biggest concern is that you cannot do the MSAs until all of the areas are designated and different criteria goes with those. And apparently the information they have, everyone knows that it will be included, but it is just a matter of being able to get it done.

And it is a huge issue and of course, I have got about 80 people from Sioux Land coming out here Monday; they want to hear about this issue.

Secretary EVANS. Sure. We will try-we will get back to you this week on that, so certainly you will be prepared for them with at least the best information we have at that point in time.

[The information follows:]

I have been asked by the Secretary of Commerce, the Honorable Donald L. Evans, to respond to you concerning questions that remained outstanding from the hearing held May 3, 2001, before the Commerce, Justice, State and Judiciary Appropriations Subcommittee. You requested further information on the timely delivery of Census Bureau data derived from long form questions on commuting and place of work that are used by the Office of Management and Budget (OMB) to define Metropolitan Statistical Areas (MSA's).

I have enclosed a detailed description of the process necessary to prepare the long form data for delivery to OMB, and a brief description of OMB's process following the delivery. After review of the enclosure, you will see that the process does not lend itself well to compression in order to speed delivery of the data. Nonetheless, we will further review the present timetables and provide you with an update.

your support of the Census Bureau, and we look forward to working with you.

Domy

William G. Barron, Jr.
Acting Director

Enclosure

JCaldwell/CAO 5/4/01

USCENSUS BUREAU

Helping You-Make informed Decisions

www.census.gov

Defining Statistical Areas Based on the Standards for Defining Metropolitan and Micropolitan Statistical Areas and Census 2000 Data

Defining statistical areas, identifying principal cities, and titling statistical areas under the Office of Management and Budget's (OMB's) new Standards for Defining Metropolitan and Micropolitan Statistical Areas depend on data produced by the U.S. Census Bureau. The standards will use population data, delineations of urban areas, and commuting data from Census 2000. Of these data, the commuting data will be the last to become available and therefore be a key determinant of the overall schedule. This review addresses the preparation of the commuting data and the process that will yield

metropolitan and micropolitan statistical areas and other geographic entities covered by the standards.

Census Bureau Work in Producing Commuting Data That Will Be Used in Defining
Metropolitan and Micropolitan Statistical Areas

The commuting data that will be used in implementing the Standards for Defining Metropolitan and Micropolitan Statistical Areas have their source in the Census 2000 "long form" question on place of work. The geographic coding of the write-in responses to that question has been under way since mid-December 2000. Currently, more than 400 coders are working on the project and are on track to finish the last step of the process (problem referral coding) by the end of July. Final processing of the 40 million-plus records will take place during August.

Codes being returned from all (not just geographic) coding operations must be matched back to original files so the data can be prepared for edit and imputation procedures. The goal of editing is to correct obviously erroneous reporting by respondents and to make the data internally consistent. Imputation is the process by which data values are assigned to questionnaire items that were left blank by respondents. Programming of the edit and imputation algorithms for the more than 50 items collected on the census long form is under way.

Once the programming is complete, the edit and imputation processing can begin. There is a logical order in which the sample items are edited, reflecting a complex web of interdependencies in the data items. That is, some questions must be edited before others, because the latter rely on the presence of legal values in previously edited items. For example, the count of workers (for whom places of work will be required) is dependent upon each person's age having been edited, because a worker is defined to be at least 16 years of age.

In addition to the constraints imposed by the order in which the sample items must be edited and imputed, the place of work data are unique in that they also have a geographic dependency within the edit and imputation process. Other sample items can be edited one state at a time, using the responses that were completely filled in by persons living in the state as the basis for the imputation of incomplete data. For place of work, however, the work locations of people living in the state do not necessarily provide a complete picture of employment there. People can commute into a state from outside it and constitute an integral part of the total employment picture. Thus, when we impute

77-308 D-01--2

values for missing place of work responses, we require that people who work in the state, regardless of where they live, be included in the process. This means that the whole country must be available for place of work imputation and that the residents of one state cannot be pulled out and processed separately.

The limitations dictated by the order of the sample edits and the geographic requirements of place-of-work allocation for all practical purposes make it impossible to rearrange the procedures or to pull out a particular area for expedited processing. To attempt to do so would disrupt the planning and scheduling that are currently in place for the entirety of the sample data products program, introducing unacceptable levels of risk to that program.

The current schedules call for the edit and imputation process to be completed by mid-February 2002. This is followed by a review and approval process, in which the subject matter experts for each topic examine the diagnostic output from the edit and imputation programs as a quality control check. This process is scheduled to end in midto-late April and is to be followed by the creation of the final sample edited detail file (SEDF), which is the base file for further Census 2000 processing. The target date for completing SEDFs for all states is late May 2002.

Assuming that the SEDFs are available by the end of May, work would begin then to create the county-to-county, minor civil division-to-minor civil division, and place summary commuting data necessary for applying the new Standards for Defining Metropolitan and Micropolitan Statistical Areas. The programs to produce these data will have already been written and tested by that time, but we have planned a quality control review of the output files by subject matter specialists. We expect the commuting data to be released by the end of June 2002.

Census Bureau Work (in Support of OMB) in Applying the Standards for Defining
Metropolitan and Micropolitan Statistical Areas With Census 2000 Data

Work can begin on defining the statistical areas at issue once the Census 2000 county-to-county commuting data file is available. Under OMB's new standards, the definitions of metropolitan statistical areas, micropolitan statistical areas, combined statistical areas, and metropolitan divisions all depend in part on commuting data, as does the identification of principal cities.

Census Bureau staff will use previously prepared and tested computer programming to define metropolitan and micropolitan statistical areas associated with all urban areas. A second stage of programming and review will identify all cases where these metropolitan and micropolitan statistical areas overlap and assign counties to single areas based on strongest commuting patterns. A third stage will identify cases where contiguous metropolitan and micropolitan statistical areas should be merged, based on strong commuting relationships between the central counties of those neighboring areas; when areas merge, the areas merging lose their separate status.

Each metropolitan and micropolitan statistical area will contain at least one principal city. The largest city in each area will qualify as a principal city, but additional cities can qualify for principal city status depending in some instances on commuting patterns. Assignment of titles to metropolitan and micropolitan statistical areas will

2