greater demand for health care from veterans and the possible need for more money. It also informs the Congress about the extent to which some veterans could be displaced or denied care if the bill was enacted and appropriations were not increased.

CBO estimates that expanding the provision of long-term care to veterans, as specified in section 101 of the bill, would ultimately increase the VA's resource requirements by about a billion dollars a year. Similarly, expanding the Department's authority to pay for emergency care, as provided in section 102, would increase the VA's resource needs by about $400 million a year. Again, whether Federal outlays would actually increase as a result of enacting those provisions would depend on the extent to which additional appropriations were provided.

În contrast to those provisions affecting veterans' medical care, the establishment of the Veterans' Tobacco Trust Fund under section 203 of the bill would create direct, or mandatory, spending. If that section of the bill was enacted, no further legislation would be required to allow the VA to spend its share of any proceeds the Government might receive from the tobacco industry. Because the amounts that the Federal Government might receive from the tobacco industry could be substantial, the spending authority created by this bill could be significant.

To develop an estimate of that authority, CBO had to answer three questions. First, what is the likelihood that the Government will win or settle a lawsuit? Second, how much would the Government recover if it did win or settle such a suit? And third, what proportion of the funds recovered would be attributed to the VA? Clearly, none of those questions can be answered with any precision, and the range of possible outcomes is large. Equally clear, however, is that the provision cannot reduce spending, only increase it. In such a situation, CBO attempts to estimate the expected value of the proposal's budgetary effect-that is, the weighted average of the cost of the proposal under a variety of circumstances, taking account of their respective probabilities.

For this estimate, CBO has assumed that there is a 10 percent chance that the Federal Government would win or settle a lawsuit with the tobacco companies. All things considered, CBO estimates that section 203 could be expected to increase mandatory outlays by about $600 million over the 2000-2009 period. Those outlays could either supplement or supplant discretionary spending for veterans' medical care.

In sum, CBO estimates that H.R. 2116 could have a significant budgetary impact on both spending subject to appropriation and spending that occurs outside the annual appropriation process. Assuming appropriation of the necessary amounts, CBO estimates that the bill would raise discretionary spending by about $200 million in 2000 and about $1.4 billion by 2004. Assuming that those amounts are not appropriated, those figures are estimates of the extent to which other activities or beneficiaries would be displaced. In addition, the provision to spend the VA's proceeds from tobacco litigation would create significant authority for direct spending.

That concludes my prepared statement, Mr. Chairman. I am looking forward to answering your questions.

[The prepared statement of Mr. Van de Water, with attachment, appears at p. 31.]

Mr. STEARNS. I thank the gentleman.

Dr. Garthwaite, Deputy Under Secretary for Health, Department of Veterans Affairs. Your opening statement, please.

STATEMENT OF THOMAS L. GARTHWAITE, M.D., DEPUTY UNDER SECRETARY FOR HEALTH, DEPARTMENT OF VETERANS AFFAIRS; ACCOMPANIED BY DANIEL J. SCHOEPS, COMMUNITY BASED CARE SERVICE LINE, GERIATRICS AND EXTENDED CARE STRATEGIC HEALTHCARE GROUP, DEPARTMENT OF VETERANS AFFAIRS; AND WALTER A. HALL, ASSISTANT GENERAL COUNSEL, DEPARTMENT OF VETERANS AFFAIRS

Dr. GARTHWAITE. We did not realize or believe we were supposed to prepare one. We have none. But you had asked Mr. Schoeps to review his methodology.

Mr. STEARNS. All right. Daniel Schoeps, Geriatrics and Extended Care Strategic Healthcare Group, Department of Veterans Affairs. Welcome, Daniel. We appreciate your opening comment.

STATEMENT OF DANIEL J. SCHOEPS

Mr. SCHOEPS. Thank you, Mr. Chairman. It's an honor to be here. I would like to review each of the points that you mentioned in your opening remarks and just talk about them briefly.

VA began its planning for long-term care in earnest in 1995, when it brought together a group of VA experts in long-term care and planning, researchers from the University of Michigan, and from the Department of Health and Human Services, to establish a planning model for long-term care, how would we project and plan into the future the long-term care needs of veterans who might come to VA for those services.

That effort was based on a concern that VA had, that we were not planning well for home and community-based care services, that we were underrepresenting the need for those services, and that we needed a model that would help us promote that and bring it to an equal footing with planning for nursing home care, which had been well established and was up and running since 1981.

That model was developed based on the National Medical Expenditure Survey, which looks at utilization of long-term care services, both nursing home and home and community-based care services, and VA adopted that model for its use to project against its own population, sensitive to the fact that need changes with age and need changes with level of disability.

We took that model and brought it to the Federal Advisory Committee on the Future of Long-Term Care in 1997, and asked them to validate whether or not that model was the correct one to use. They had very serious discussions about that model, its strengths and weaknesses. They felt that, since it was reality based, since it was based on projecting need based on people actually using services, as opposed to some theoretical construct of people who might come to use services, that it would a strong model.

The downside of the model was that it was old. It was based on 1987 data and new data was not quite available, although there had been a new survey done in 1996.

So the Federal Advisory Committee validated for our use this long-term care planning model, and we have been using it since that time. That is what we used to develop, at a staff level, projections on the effect of this bill on long-term care and on the serviceconnected population rated 50 percent or more.

Our findings are, number one, that we serve a large portion of that population already. Two-thirds of the veterans that we estimate need long-term care services, who are service-connected 50 percent or more, are already receiving those services from VA. The model estimates that 15,000 veterans in that group need services on any given day, and we're already serving 10,000 of them.

The second thing the model found when we estimated that 5,000 more veterans would come to VA for care was our cost estimates, again at the staff level, depending on some adjustments in the assumptions that we made, ranged in the first year from $115 million to $184 million, and ranged upward in the fifth year from $148 million to $237 million, which is again with differing assumptions. The copayments that we estimated were at $143 million, and remain constant throughout the time.

I could not characterize CBO's work as having serious flaws. I think there might be a question of level of disquietude. On that subject, there would be just two points that I would make-and they differ because CBO used different models for nursing home and for home care. I think we have more serious concerns about the model for nursing home care, which generates this very large number. I feel that there is not enough reality basis, a basis of real experience, in the model that they're using. We are looking at the need for this service based on actual utilization, and I think their approach, quite legitimate, is more theoretical than ours.

On the issue of home care, they do use an experience-based model for their home care projections. It is different than the model that we use. It focuses on a more specific, chronically ill population than the model that we use. I think, if we were to use that modeland we have discussed using that model-we would need some changes. It's a very well respected data set. Dr. Ken Manton at Duke has helped us in our office many times over the years. He is kind of the guru of that data set.

If we were to use that model, I think we would make some modifications, some adjustments to using that approach that they did. But it is experienced based and, for that reason, we could talk about the adjustments.

That would conclude my comments.

Mr. STEARNS. I thank the gentleman.

Let me offer my ranking member, Mr. Gutierrez, an opportunity for an opening statement.

Mr. GUTIERREZ. Mr. Chairman, I would like my opening statement to be included in the record.

Mr. STEARNS. By unanimous consent, so ordered.

[The prepared statement of Congressman Gutierrez appears at p. 26.]

Mr. GUTIERREZ. I thank the chairman for bringing this hearing together, with members of the CBO and Department of Veterans Affairs, and I thank all my colleagues on my side of the aisle and your side for showing up today.

Since we are so well represented with the ranking member, with five other colleagues on this side of the aisle, I'm going to ask to be excused, Mr. Chairman, as I have a very bad cold. But I wasn't sure how many people were going to show up, but it seems to be very well attended.

Mr. STEARNS. You are very well represented.

Mr. GUTIERREZ. Yes, it looks like I'm going to be well represented.

Mr. STEARNS. Mr. Buyer is representing us. (Laughter.)

Mr. GUTIERREZ. Again, thank you. I ask to be excused, Mr. Chairman.

Mr. STEARNS. Thank you for coming.

We will now open up for questions, and I'll start. I think this is one of the rare opportunities I have had to hear CBO estimates and then hear Administration representatives discuss them and perhaps rebut them or agree. I think for members this is a great opportunity for us here on the Veterans' Affairs Committee to understand the assumptions made by CBO and test these assumptions through the experience of those who administer the programs in question.

Mr. Schoeps, you're saying that CBO is incorrect in assuming that most veterans who would qualify for VA nursing home care would seek it, when I guess the reality, the predilection, is that the people themselves would not want to go into a nursing home other than as a last resort. Is that what I hear you saying?

So the difference in the estimate of CBO and the administration, and actually what we see in practice, is that the CBO is assuming that most eligible under this bill would seek complete nursing home care, whereas the reality is that it's not true?

Mr. SCHOEPS. As I read their testimony and analysis, my understanding is that they looked at veterans in this sub-population of service-connected 50 percent or more, and looked at their disability levels in terms of activities of daily living, their ability to dress and bathe and go to the toilet and transfer and walk, and made a determination that if people had three or more or four or more disabilities in ADL's, deficiencies in ADL's, that they would be nursing home eligible and would go into a nursing home.

We were unwilling to make that assumption. That was an assumption that the Federal Advisory Board warned us against. That represents the major difference between our projection at the staff level and CBO's projection.

Mr. STEARNS. You indicated that the long-term copayment would help to offset this estimate that you come up with?

Mr. SCHOEPS. Yes, sir. We estimated at the staff level copayments totalling $143 million.

Mr. STEARNS. Which would practically offset it.

Mr. SCHOEPS. Depending on, in the first year, the range of costs was $115 million to $184 million. So it would offset, or almost offset.

Mr. STEARNS. In looking at this, you refer to the fact that you had some history on this, and looked at different models. I gather you have been working on this question with the advisory committee for, I think, two years. I should point out to my colleagues that CBO has worked on this estimate for a much shorter period of time.

Is that correct, that your experience has been much longer than CBO's?

Mr. SCHOEPS. I don't know the folks that worked on this. I can't speak of their experience. In our office, we have been involved in long-term care planning since VA's first efforts. When I came to extended care in 1980, the first assignment that Dr. Paul Haber, my boss, gave me was to work on a planning model for long-term care. Over that period of time, we have become more sophisticated and have spent a lot of time on this issue, and have called in experts frequently to help us.

Mr. STEARNS. So you're saying, since 1980, you have had experience with developing this modeling for long-term care? You've worked on it for almost 20 years?

Mr. SCHOEPS. Yes, sir.

Mr. STEARNS. Mr. Van de Water, you say that based on age, marital status and functional impairment, 45,000 additional serviceconnected veterans would get nursing home care under this bill by the year 2010. Perhaps you might want to comment on what Mr. Schoeps said, particularly dealing with the human element. I think your assumption is that most of those eligible under the bill, this 45,000, would go into nursing home care, and his modeling reaches a very different conclusion. So you might want to comment on what he said, and perhaps anything else he has said that you would like to comment on.

Mr. VAN DE WATER. Yes, Mr. Chairman. Mr. Schoeps's comments were, I think, very helpful.

I would start by qualifying only one thing that he said: that the VA's approach is experience-based and ours is not. Clearly, in addressing a bill such as this, both of us are trying to project what would happen under circumstances that no one has yet observed. If either of us had experience with the situation, we wouldn't be in this discussion at all. We're trying to guess how people would respond to a change in legislation, so we're both having to extrapolate from past experience.

One of the differences, as Mr. Schoeps indicated, is that they relied on one particular database, the National Medical Expenditure Survey, whereas we have tended to rely on two other databases, the National Survey of Veterans and the National Long-Term Care Survey. That's point number one.

The second point, as you said, is that we have had a limited amount of time to evaluate this particular bill. And I have to say that the estimate that Mr. Schoeps has discussed this morning is one we have only just heard about for the first time. Our previous impression-based on Dr. Kizer's testimony here on the 19th of May, in which he stated that the cost of this provision would be substantial-was that the VA's views were more consistent with

ours.