Below is the uncorrected machine-read text of this chapter, intended to provide our own search engines and external engines with highly rich, chapter-representative searchable text of each book. Because it is UNCORRECTED material, please consider the following text as a useful but insufficient proxy for the authoritative book pages.
Chapter 2 ORGAN IZAT IONAL EVOLUT ION Growth and Organizat tonal Change NIH has experienced phenomenal growth in the past four decades. Its budget has grown from 31.3 million in 1943 to almost 34.5 billion in 1984,4 and the number of its employees has grown from Slightly over 1,000 to over 16,000 during the same period.5 The number of extramural research grants made by NIH went from 79 in 1946 (the first full year in which NIH had authority to make external grants for research) to about 20,000 in the 1983 fiscal year.6 Few, if any, other federal agencies have enjoyed such a rapid and sustained rate of growth. Growth necessarily brings increasing organizational differentiation into institutes, divisions, and 80 forth. At NIH such change has generally taken the form of categorical inatitutea, organized around a particular health problem or issue. Beyond the concept of categorical institutes, NIH has never had a systematic design for its organizational development. As the nation's concern for health research grew in the years after World War Il. Congress and the Executive Branch created new institutes to cover major areas of human diaease.7 As a result, NIH now has four bureaus, eleven research institutes ~ three of which are counted among the bureaus), and Six other research and support divisions. ~ Appendix B describes the relationship among these entities more fully. 'the foci of the research institutes are anything but uniform: come are disease- oriented ( Cancer; Neurological and Communicative Disorders and Stroke; Arthritis, Diabetes, ant Digestive and Kidney Diseases ; Allergy and Infectious Diseases), others are anatomical (Heart, Lung and Blood; Dental; Eye), others relate to stages in human development (Child Health and Human Development; Aging) and still others have broad themes (Environmental Health Sciences; General Medical Sciences). This mixture does, however, allow NIH to cover all important research areas and simultaneously accommodate public interest in particular health issues. In this categorical system, the public, acting through Congress, can indicate broad priorities among health problems for which they want research answers. NIH and the scientific community then determine how to address these public desires scientifically. In addition, the scientific community, operating on the basis of perceived opportunities, also influences the direction of research programs. The committee believes that the interplay of categorical
institutes and disciplinary study Sections introduces a creative tension between national needs and scientific autonomy that leads to productive and useful research. This balance between apparently opposing factors has had a great deal to do with NIH's well-deserved reputation for excellence, as well as with the countless contributions to the public's health that can be traced to the efforts of NIB. However, if the tension between these forces is too great it can impede rather than stimulate; if it is unrelenting it becomes~wearing rather than inspiring. The balance is a delicate one, and attempts to shift it markedly in one direction or the other are likely to yield less, not more, valuable results. One consequence of a categorical system is that demands for increased emphasis in particular areas often turn into proposals for changes in the NIH organizational structure. Between 1943 and 196S, the total appropriations for NIH grew from $~.3 million to tI.l billion.9 This is nearly a thousand-fold increase, and constitutes an average growth rate of 27 percent per year. (Even with an adjustment for inflation, the average growth rate during this period was 24 percent.l°) Since then growth has slowed, but compared with other federal research agencies, NIH continues to receive close attention and strong financial support from Congress. Its 1984 appropriation was $4.5 billion, contributing to an average annual growth rate of 9 percent (2 percent after adjustment for inflation) between 1968 and 1984.li As the overall rate of growth slowed, pressure increased for more activity in particular areas. In response to public interest, members of Congress and others continue to propose new institutes or other organizational changes. Since 1970, there have been at least 23 new institutes proposed for NIH, 13 through the legislative process and 10 by other means.l2 Although many instituter were proposed, only three new ones have been added to NIH since 1968. NIH has thus acquired an image of being opposed to innovations proposed from outside. However, many of the proposals have led to me jar alternative organizational changes: three institutes changed their title and mission (twice each), and three institutes were granted bureau status.13 The committee believes that the current structure is sufficiently flexible to respond to most needs. However, the time has come to articulate an attitude about responsiveness to national health needs and to define a process and a set of criteria for future organizational growth and change. Before doing so, it is important to under Stand an fully as possible the effect of organizational change on research budgets and research programs. Overall NIH Appropriations Since the 1930s, NIH has seen the establishment of many new institutes and divisions. These entities grew out of existing institutes, were transferred into NIH from elsewhere, or were built up from the start. In addition, existing instituter have changed their
12 names, their research focus, and have had their organizational status upgraded. Figure la shows the total NIH appropriation in the years 1940 through 1982. It also shows when new institutes were added and other organizational changes (new divisions, namP changes, elevation to bureau status) took place. Figure Ib shows the same data, but on a logarithmic scale. On this scale ~ exponential growth appears an a straight line, and the details of NIH's early budgetary history are eas ier to see. Because the effects of organizational ant budgetary changes are difficult to isolate in a single year, it is best to analyze longer periods of growth and organizational change. The figures indicate that there have been three phases in the growth of NIH. During the late 1940s many new institutes appeared and the budget grew drama- tically. Between 1955 and 1965 there were few organizational changes but the budget again grew dramatically. Since then there have been many more organizational changes but bees growth. Thus, there is no evidence that the addition of new institutes and other organizational changes, in themaelvea, have led to increased budgetary growth for NIH as a whole. There may seem to be one exception to this conclusion. Between 1971 and 1973, the NIH budget grew by more than 40 percent in constant dollars. This growth was concentrated almost exclusively in the appropriations for the National Cancer Institute (NCI) and the National Heart and Lung Institute (NHLI), and coincided with the National Cancer Act of 1971 and the National Heart, Blood Vessel, Lung, and Blood Act of 1972. It is important to note that this legislation did not create new institutes but rather raised the status of existing institutes to bureaus, gave them expanded responsibilities, and made them subject to periodic reauthorization. Appropriations for the Affected Instituter The committee has examined whether there is a cause and effect relationship between creation of new institutes or elevation to bureau status ant funding patterns. While we could not find conclusive evidence of sloth a relationship, we report the following observations. The longest period of sustained relative growth for a new institute occurred for the National Institute on Aging (NIA), which has grown faster than the remainder of NIH in every year since its first appropriation in 1976. NIA's growth did taper off, however, from an average annual differential (between it and the rest of NIH) of 22 percentage points between 1976 and 1980, to 6 percentage points between 1980 and 1984. Between 1976 and 1980, the National Eye Institute (NEI) and the National Institute of Environmental Health Sciences (NIGHS) also grew at a rate substantially above the remainder of NIH. Although these three are the most recently established institutes, NIEHS and NEI were, respectively, eight and nine years old in 1977. Thus, the relative increase in the NIEHS and NET budgets,
13 2.0 1.8 1.6 1.4 o - z 1.0 o < 0.8 - 0.e 0.4 1.2 0.2 ~ ~ 0~ O it_ ~ 11340 1~5 1 9S0 1~6 . 1 i 1 1 i i 1 1 1 1 I1900 191S6 YEAR o ~E. t~l7=E O Cal - G. Of h~E · BUREAU STATS ~E. D'Ve~lm * IN"I=TE LEAVES No i 1 1970 19~ 1~ FIGURE la. Total congressional appropriations for NIN (excluding all programs that were eventually transferred out of NIB), in 1972 dollars, from 1940 to 1982. The occurrence and nature of major o An anizational changes are also indicated. The two new divisions noted are those that later became institutes. 'I . - o _ 1 00 E o - 10 ! 1~0 191#; ~ j1--~=T · BUREAU STOW · NEW mvlS8 - 1 1 1 | ~ t~" LEAVES Ale 1 1 1 1 I I . i ,~ 1~ 1970 lots 1 YEAR FIGURE lb. The game data represented on a logarithmic scale. SOURCE: Calculated from data in the 1983 NIN Almanac.
14 and perhaps the NIA budget as well, probably reflected increased Congressional interest, but was not necessarily related to their "newness." Because reliable figures are not available for comparable research expenditures in the area of a new institute before it was established, it is impossible to measure the full effect of establishing a new institute on the funding of research in the affected field. However, the effect of establishing a new institute out of a previously existing one can be portrayed. Figure 2 shows that NET, which was split off from the National Institute of Neurological Diseases and Blindness in 1968 (but did not have its own budget until 1970), grew faster than the rest of NIH for only one year. The "remainder" institute, which is now called the National Institute of Neurological and C" ~unicative Disorders and Stroke, grew more slowly than the rest of NIH for four years. The total of the two grew more slowly than the rest of NIH for six years. In the case of NLA, Figure 3 shows that the new institute grew faster than the rest of NIH, but the remainder institute, the National Institute of Child Health and Human Development (NICHD), grew at about the same rate as NIH during the same period. The total of the two grew faster than the rest of NIH. NCI provides one example of an organizational change--elevation to bureau status--that was associated with a major redistribution of funds within NIH. In the years before and after the National Cancer Act, NCI grew relative to the rest of NIH. Since 1976, however, it has been shrinking relative to the rest of NIH. This can be seen in Figures 2 and 3 as a general tendency for the other institutes to exhibit negative relative growth in the early 19708 and positive re let ive growth in later years . In summary, although there have been substantial changes in the distribut ion of funds within NIH, the committee f inds no evidence of consistent, sustained causal effect of organizational changes on the dist ribut ion of the budget . The Ef f ec t on Sc fence and Management The remaining question is the extent to which different organiza- tional forms facilitate the NIH objectives of maintaining scientific excellence ant being responsive to the needs of society. It is extremely difficult to assemble objective data on the impact of new institutes or other organizational changes on the scientific productivity and the generation of health research in these fields. The health research enterprise is extraordinarily complex, with extensive interaction between different fields and interdependence of developments in basic science that emerge from laboratories of all disciplines. As a result, even a thorough analysis of the research output of an institute could not give an unequivocal picture of its precise contribution to the research advances. The difficulties may be insurmountable in trying to achieve cause-and~effect conclusions.
15 fir 4OI - 3 at - -10 at Lot U. - C) 30 20 _ -20 _ -30 _ -40 _ _ ~ Nl NC~ · NEI O N'NCDS ~ NE. \~ -50 _ . 19~ 19158 1970 1972 1 1 1 1 1 1 1 1 1 1 1 1 1 1 _1 1 1974 1976 1978 1~ 1~2 YEAR FIGURE 2: Differential growth rates of congressional appropriations for the National Institute of Neurological and Communicative Disorders and Stroke (NINCDS ~ ant its predecessors, the National Eye Institute (NET), and the total of the two, 1955-1982, in 1972 dollars. The figures represent the annual growth rate of the given institute minus that of the rest of the NIH. SOURCE: Calculated from data in the 1983 NIH Almanac. 50r 40 30 - 3 In us lo: . ~ o a UJ cut -20 at: flu ILL - 20R -10 ~ · NIA ~ NICHO O NICHD ~ NIA t972 1974 1916 1978 1~ 1982 YEAR FIGURE 3: Di f ferent is l growth rates of congres a tonal appropriat ions for the National Institute of Child Health and Oman Development (NICHD), the National Institute on Aging (NIA), and the total of the two, 1965-1982, in 1972 dollars. The figures represent the annual growth rate of the given institute minus that of the rest of the NIH. SOURCE: Calculated from data in the 1983 NIH Almanac. _ . _
16 Based upon the beat evidence available, however, the committee concludes that establishing a new institute at least has not hampered the scientific effort, and may have helped it considerably. The National Eye Institute appears to be an example where the creation of a new institute had major qualitative effects on the research field. For example, by expanding the intramural vision research program and developing workshops and programs emphasizing an interdisciplinary approach to visual disorders, the NEI has aignifi- cantly expanded the Scope of vision research to encompass developments in such arena as immunology and molecular biology. And by attracting eminent epidemiologists and statisticians, the NEI was able to expand epidemiological vision research efforts which led to advances in the evaluation and demonstration of effective treatments for diabetic retinopathy and macular degeneration, two of the leading causes of blindness in the United States.15 On the other hand the establishment of the Lung Division and the addition of "Lung" to the title of the National Heart Inatitute led to similar effects without establishing a new institute. Through a series of actions taken to increase the visibility of and resources for this area, lung research wee transformer from a relatively undeveloped area into a thriving one that underwent a scientific revolution in the 19708.16 Frommer characterizes this as a shift from "studying the lung simply as a bellows and gee exchange organ to Studying it as a complex physiological organ involved in numerous other processes and to studying it at the cellular and subcellular levels and also by biochemical, immunological, and other techniquea.''l7 In the case of the Lung Division, a major reason for the successful outcome was the active cooperation of institute leadership with the forces urging greater attention and resources for lung research. What both of these cases had in common was a scientific readiness for major advances and dynamic leadership from committed individuals with a vision for the field. These elements, together with increased resources and visibility, may be more critical to scientific progress than institute status itself. Responding to Needs and Opportunities In the last decade the research management style of NIH has shifted from one driven primarily by scientific developments and individual researchers to one in which NIB staff cooperate with outside Scientists in a planning process and, on this basis, take a more active role in setting research priorities and developing new research area. Although the planning process varies from institute to institute, as described in Appendix B. it generally involves interactions with the larger research community to determine promising research poasibilities.~9 The institutes then stimulate activity in those areas through requests for grant applications, contract proposals, and other means. The goal is to plan the research environment, not specific research projects. All applications and proposals are
17 reviewed by scientific peers. Close interactions both with the public and the scientific community have enabled NIH to structure scientifi- cally appropriate and effective research programs that expand basic research and attack specific health problems. ~ Two examples suffice to illustrate this point. In genetics research, the early efforts of NIH were driven almost exclusively by scientific opportunities and were oriented toward grants for basic research. But as the premise of this field grew in the minds of clinical researchers and the public, the program Stuff developed a portfolio of activities linking basic scientists and clinicians. The main focus remained on investigator-initiated research grants, but the staff, with advice from external advisors, became more active in stimulating proposals in needed areas and in developing a short-tenm research contract program to complement the grant program. 20 The Sudden Infant Death Syndrome research program, which was established because of public concern about "crib death," illustrates a similar approach to a more applied problem. In this case, NIH program staff worked with the scientific community to develop a research agenda and draw more scientists into the field, and worked with the public to develop educational programs and to develop support for basic research in the field.21 Simultaneously, most of the institutes have switched from an internal structure organized around research mechanisms to one organized around research programs.22 Before the 1970s most institutes had divisions or offices of intramural research, research grants, research contracts, and no on. But partly as the result of an internal study completed in 1973, most institutes have since adopted a program-orientet structure.23 The National Institute of Arthritis, Diabetes, and Digestive and Kidney Diseases, for instance, now has divisions corresponding to each of the disease groups in the title, and under them nine branches in rheumatic diseases, musculoskeletal diseases, skin diseases, diabetes, endocrine ant metabolic diseases, digestive diseases, nutrition, kidney and urologic diseases, and blood diseases. This program orientation enables coordinated efforts in identified research areas, committed ant accountable program leader- ship, and organizational responses below the institute level. As is argued below, NIB now is at a stage in which the addition of new institutes tends to be a costly and inappropriate way to implement changes in research priorities. As discussed above, however, a shift in organizational emphasis now enables NIB to respond to health needs and scientific opportunities without resorting to major organizational changes. The committee believes that all of the components of NIH, and especially the Director's office, must adopt an active approach and a visible mechanism for assessing and responding to needs and opportunities when they first arise. Such an approach, suggested in the following set of recommendations, should reduce sharply the need for new institutes and perhaps the pressure for them.
18 Sensing the Need for Action The committee recommends : The Director of NIH should establish an NIH-wide mechanism to (1) keep abreast of the views, concerns, and proposals of the NIH scientific and public constituencies; ( 2) a88e88 their potential implications for _ ~ Ian responses. NIH has numerous points of contact with its constituencies, but no office at a high level whose primary job is to monitor the emerging concerns and demands for more focus, greater investment, or faster progress in particular areas of health research. - The individual or office charged with this surveillance function should keep track of scientific developments ant public perceptions 80 as to alert the Director when strategies of response are required. This mechanism would allow NIH to seize the initiative in interacting with the external world, rather than simply reacting to its pressures. Research Planning The committee recommends: NIH should strengthen its planning efforts, particularly at the Institute level and on the coordination of NIH-wide aci iiic ies- . . Having sensed a need or an opportunity, NIH must ensure that it addressed. Program planning provides a means for doing 80. Overall NIH planning involves the Director, his planning staff, and the individual institute directors. In addition, each institute has its own planning process, respons ive both to its own needs and also to a colon NIH process. However, the involvement of extramural scientists, Congress, and the public in the planning process is still uneven. The planning process can be strengthened in a number of ways. Effective planning must obtain and rely on the views of the scientific community and the public, and be linked with budgeting. At the institute level, national advisory councils and boards of scientific counselors must be involved. By now, each of the institutes has developed its own unique planning process; each should examine and benefit from the successes of others. At the NIH level, there should be better planning and coordination of research activities that cross institute boundaries .
19 Range of Effective Responses -The committee recommends: NIH should avail itself Of a range of activities, short of establishing new institutes, to respond to health needs and opp-ortunttzes. NIB has at its disposal a continuum of possible responses to the needs and opportunities it identifies, matches to the magnitude and urgency of the demand. Proponents of new institutes often are not aware of these options, and should explore them carefully. Some examples of these responses are : . publicizing what scientific research has accomplished, and the status and promise o f current research in the area; disseminating the resules of research widely to centers that can make immediate clinical use of them; . accelerating research by such means as soliciting applications for grants, schedul ing spec is 1 peer review sessions, appointing special study sections, and utilizing "high relevance ~cores" for applications in the area concerned; holding major conferences of the leaders of the scientists and practitioners concerned to exchange ideas and discuss future actions, and in appropriate situations, creating such groups by action of the Secretary, or the White House, or by joint action with Congress; naming special panels or subcommittees of the national advisory councils , boards of Scientific counselors, review committees, and study sections, as appropriate, to examine both extramural and intramural programs and opportunities; forming an NlH-wide or inter-agency committee, preferably led by an institute director, to coordinate and develop research in the area; providing higher visibility and stronger leadership over the area of concern by organizational upgrading; for example, elevating sections to branches, branches to programs, or programs to divisions.
20 The Basis for Major Organizational Change The evidence about the effect of organizational change and therefore about its desirability, as we have seen, is mixed. Many of the changes in the past have been auccesaful. The committee believes, however, that NIH has reached a point at which there should be a pre~umption--to be overridden only in exceptional circumetances-- against additions at the institute level, whether they are proposed to occur by fission or by transfer from outside NIH. Although a particular research area may benefit from the establishment of a new institute, the overall manageability of NIH suffers from the cumulative effect of such additions. This is because new institutes (1) fragment the scientific effort and diminish effective communication with key scientists in other institutes, (2) add to the burden and difficulty of effective program coordination by the NIH Director and his top staff, and (3) add to administrative C08~8 without ensuring increased appropriations. Furthermore, as discussed above, there are effective alternatives to establishing new institutes. Organizational structure has imp! ications for long-range institutional dynamics, so thought must be given before any research area is singled out for increased organizat tonal attent ion . On the other hand, there have been cases in the pant in which organizational changes have brought new vitality to a scientific f ield, and there are likely to be such cases in the future. The committee in not opposed to major organizational changes, but believes that such measure. should be taken sparingly. Of the many proposals for new institutes, only a small fraction are likely either to be necessary or appropriate . Therefore, the committee proposes the following process and set of criteria for assessing the need for major organizational changes. A Process to Assess Proposed Organizational Changes The committee recommends: There should be a formal process to assess proposed major organiza- tional changes in NIH. This process must involve the entire range of interested parties. Proposals for institutional change always have emerged, and will continue to emerge, from a number of different sources. NIH should consider them all seriously. As part of any request for institute status there should be documentation that the appropriate alternative mechanisms for meeting perceived needs have been explored and were found to be inadequate. All proposals should be analyzed at a level of broader perspective than that of NIH to encourage ant focus the debate along the lines of the criteria stated below, to evaluate the arguments presented by NIH and the various interested parties, and to analyze the pros and cons for decision makers in the Administration
21 and Congress. To Serve this as well as other purposes, the next chapter recommends the establishment of a Health Science Board in the Department of Health and Human Services. An entity outaite of NIH would be especially important in considering proposed transfers into or out of NIH. These cases must be examined from three perspectives: that of NIH itself, that of the entity that is proposed to be moved, and that of the overall Strategy of the Public Health Service for improving the nation's health. No single agency, or advisory board associated with a single agency, would have the broad perspective and the public credibility that are needed to address transfer issues. Criteria to Assess Proposed Organizational Changes The criteria described below identify a set of issues to be addressed in the discussion about the need for a new institute or other major organizational change in NIH. Only when most of these factors are positive are the benefits from the creation of a new institute likely to exceed the costs. Because the decision must reflect an assessment of the organizational costs and benefits of a proposed change, and because the particulars of every individual proposal are complex, it is not possible to develop simple formulas to determine the need for change. Instead, the criteria must be adapted to the situation and interpreted. The Health Science Board described in the next chapter can provide such judgment. 1. The activity of a new institute or other organizational entity must be compatible with the research and research . training mission of NIH. If a major emphasis of the proposed new entity is in regulation in the delivery of services or . _ ~ , in other non-re~earch activities, it is not appropriate for incorporation in NIH. ~ These considerations would be most clearly applicable to the case of a proposed transfer into NIH of an existing function currently located elsewhere in the Public Health Service, or of a proposed new institute. In either event, the formal statement of mission would have to be consistent with the overall mission of NIH. One reason for this criterion is that activities do not flourish unless they exist in a supportive environment. Another is that the need to manage non- research activities necessarily distracts the attention of the NIH leadership from the primary mission of NIH, that of research. The National Institute of Occupational Safety and Health, which currently has significant direct and indirect responsibilities in support of regulatory activities, for example, does not match the NIH research mission. Similarly, the major service mission of the National Institute of Mental Health was a critical factor in its removal from NIH in 1967.24
22 2. It must be demonstrable that the research area of a new . institute or other major organizational entity (defined . . . either as a disease or health Problem or as a biomedical or , behavioral orocese related to a health Droblem is not already receiving adequate or appropriate a~]cion7~~ As a case in point, calculations of the burden of illness associated with arthritis formed one of the main arguments uset by proponents of the National Institute of Arthritis and Musculoskeletal Diseases. These calculations include the number of people affected by the disease and economic costs such as medical expenses, disability payments, and loss of earning power. In arguments for institutes focused on other diseases, mortality statistics (annual number of deaths attributable to the disease, years of life lost, and so on) are of ten c ited . The problem with such arguments is that, although they often sound convincing in particular canes, they are extremely difficult to construct on a consistent basis across all institutes and disease groups. First, the burden of illness includes mortality, morbidity, economic impact, and quality of life; diseases have differential impacts on each of these components. Furthermore, there are many ways to measure each of the components, and some are obviously more favorable to the argument than others. Second, even if there was a unique index of the burden of illness, resources probably should not be allocated strictly on this basis. According to economic theory, to maximize the overall benefit, resources ought to be allocated so as to equalize the marginal benefit of the last dollar spent on each of the alternatives. Such a calculation, if it could be made, would have to include the scientific potential of a field and the potential reduction in the burden of illness (not the total burden). Thus, comparative calculations are extremely difficult, and probably cannot be done for one institute at a time. 3. There must be reasonable prospects for scientific growth in a research area to justify the investment in a new institute or other mailer organizational entity. Critical mass arguments cut both ways. In the case of vision research, proponents of a new eye institute claimed that the scientific personnel and research ideas already existed, but were not getting enough attention.25 The National Institute on Aging, on the other hand, was established, in part, in order to develop a cadre of researchers on aging, and personnel development continues to be one of the institute' s top priorities.26 The committee feels that the research area to be emphasized must have a potent ial for growth, both in terms of the underly ing sc fence and in the number of scientists whose efforts could prof itably be directed toward the area. This does not mean that the research ideas have already been fully worked out or that there already exists a critical mass of scientists working in the area. It simply means that upgrading the organization is likely to promote such growth.
23 4. There must be reasonable prospects of sufficient funding for a new institute or other major or~anizationa Organizational upgrading involves the creation of new functions and positions, and so must require additional funds for administration. Such additional funds ens ~ be regarded as the price to pay for the benefits of the new organizational structure, and thus should not come out of research budge te. Furthermore, although budgets are set on a yearly basis, there must be reasonable prospects for adequate long-range financial support, if scientists are to shift their research interests. 5. A proposed change in the NIH organizational structure should, on balance, Improve communication, management, priority setting, and accountability. , Organizational changes that are intended to focus and enhance the effort of NIH in one area, especially those that divide existing insti- tutes into component parts, necessarily interfere with communication and research management in other areas. Thus the need for visibility and focus must be balanced against the need for a broad-based approach that enhances scientific communication and allows research areas to cross-fertilize each other. Research on the lung, for instance, has much in common with research on the heart and blood, and thus there are substantial benefits from the continuing integration of these research areas in one institute. Organizational structures define the trade-offs in the budget process, and thus have an important impact on priority setting. In this respect, an organizational change is an improvement if it leads to a more appropriate set of trade-offs, for instance, one in which the alternatives are different scientific approaches toward the same health goals. In addition, because institute appropriations are determined outside of NIH, narrowly defined institutes may lead to excessively detailed external control. Similarly, organizational structures define the loci of authority and accountability. Creating a new institute involves establishing a potentially powerful new institute director who can be held account- able for progress in the new field. But at the same time, it dilutes the authority necessary for managing NIH's overall research program. Prevention research provides an example of the importance of this final criterion. Because of a perceived need to sating then prevention research, some have proposed the creation of a separate National Institute of Public Health in NIH. The problem with such a proposal is that prevention research is and should be an essential part of every institute's research mission. Establishing a new institute would disrupt critical lines of communication between population-based and laboratory-ba~ed researchers working on the same problems. It would create inappropriate budgetary trade-offs between budgets for "prevention" and "disease-specific" research rather than between population-based and laboratory-based approaches toward the same goals. And finally, it would tend to absolve the other institutes of responsibility for prevention research, a critical part of NIH's @ @ mass con r@
