Appendix D
Demographic Projections of the Research Workforce, 2001–2011

Are universities educating enough research scientists in the biomedical, clinical, and behavioral fields, or are they educating too many or, perhaps, too few? These questions underlie this attempt to project the research workforce. To determine how this workforce will grow, projections are created by taking the current workforce data in these fields and extrapolating over 10 years. Then this projected workforce is compared with estimates of the demand for future researchers.

This analysis only deals with the workforce in the aggregate form looking at the major fields of biomedical, clinical, and behavioral research and does not seek to provide complete answers. In particular, it does not tell whether the workforce will be adequate in specific disciplines. It says nothing about the quality of the workforce, although that is a critical variable. It can be used to evaluate projected quantity in relation to the expected numbers of positions, which could be altered were it essential to national goals.

A demographic model for the current workforce in each field is created by adding graduates year by year (these numbers are estimated with regression models based on data from past decades) and then subtracting retirees and decedents. From these calculations a picture of how the workforce may evolve is developed. Provision is also made in the model for immigrants who arrive with Ph.D.s in these fields, who move from employment to unemployment and back, and for graduate students, who take nonscience jobs and possibly move into science jobs late. The original model is described fully in Appendix D of the last report on the scientific workforce in these fields, Addressing the Nation’s Changing Needs for Biomedical and Behavioral Scientists (National Research Council, 2000).1 For this report, the model has been updated with new data and the addition of new groups.

This appendix will address the following issues:

  1. fields covered and the data used: more than those in the previous application of the model;

  2. current workforce: its components and characteristics, with some attention to recent trends and how they might be extrapolated;

  3. workforce projections: over a decade; and

  4. parallel projections of research positions: by sector that may become available.

COVERAGE AND DATA

For this report the definition of the behavioral field has been expanded to include clinical psychologists—a substantial expansion which means that current results for this field cannot be compared with previous results because the projections for 1995–2005 in the previous report did not include this category. The major field of clinical research was also added and includes Ph.D.s in clinical areas (except psychology) and M.D.s engaged in clinical research. Data for the clinical Ph.D.s are available from the same sources as for biomedical and behavioral Ph.D.s. For the M.D.s, only limited data are available; thus, it is only possible to speculate about orders of magnitude.

The disciplines included in biomedical and behavioral sciences include fields from anatomy to zoology; the behavioral sciences cover psychology, sociology, anthropology, demography, and speech-language pathology and audiology. The clinical sciences include all medical disciplines that are not considered biomedical. These categories have not changed from those covered in the previous report except for the addition of clinical psychology in the behavioral sciences.

Data on the U.S.-trained Ph.D. workforce in these fields were obtained from the Survey of Doctorate Recipients (SDR), a longitudinal biennial survey. Data on graduates were obtained from the Survey of Earned Doctorates—an

1  

National Research Council. 2000. Addressing the Nation’s Changing Needs for Biomedical and Behavioral Scientists. Committee on National Needs for Biomedical and Behavioral Scientists, Education and Career Studies Unit, Office of Scientific and Engineering Personnel. Washington, D.C.: National Academy Press



The National Academies | 500 Fifth St. N.W. | Washington, D.C. 20001
Copyright © National Academy of Sciences. All rights reserved.
Terms of Use and Privacy Statement



Below are the first 10 and last 10 pages of uncorrected machine-read text (when available) of this chapter, followed by the top 30 algorithmically extracted key phrases from the chapter as a whole.
Intended to provide our own search engines and external engines with highly rich, chapter-representative searchable text on the opening pages of each chapter. Because it is UNCORRECTED material, please consider the following text as a useful but insufficient proxy for the authoritative book pages.

Do not use for reproduction, copying, pasting, or reading; exclusively for search engines.

OCR for page 120
Advancing the Nation’s Health Needs Appendix D Demographic Projections of the Research Workforce, 2001–2011 Are universities educating enough research scientists in the biomedical, clinical, and behavioral fields, or are they educating too many or, perhaps, too few? These questions underlie this attempt to project the research workforce. To determine how this workforce will grow, projections are created by taking the current workforce data in these fields and extrapolating over 10 years. Then this projected workforce is compared with estimates of the demand for future researchers. This analysis only deals with the workforce in the aggregate form looking at the major fields of biomedical, clinical, and behavioral research and does not seek to provide complete answers. In particular, it does not tell whether the workforce will be adequate in specific disciplines. It says nothing about the quality of the workforce, although that is a critical variable. It can be used to evaluate projected quantity in relation to the expected numbers of positions, which could be altered were it essential to national goals. A demographic model for the current workforce in each field is created by adding graduates year by year (these numbers are estimated with regression models based on data from past decades) and then subtracting retirees and decedents. From these calculations a picture of how the workforce may evolve is developed. Provision is also made in the model for immigrants who arrive with Ph.D.s in these fields, who move from employment to unemployment and back, and for graduate students, who take nonscience jobs and possibly move into science jobs late. The original model is described fully in Appendix D of the last report on the scientific workforce in these fields, Addressing the Nation’s Changing Needs for Biomedical and Behavioral Scientists (National Research Council, 2000).1 For this report, the model has been updated with new data and the addition of new groups. This appendix will address the following issues: fields covered and the data used: more than those in the previous application of the model; current workforce: its components and characteristics, with some attention to recent trends and how they might be extrapolated; workforce projections: over a decade; and parallel projections of research positions: by sector that may become available. COVERAGE AND DATA For this report the definition of the behavioral field has been expanded to include clinical psychologists—a substantial expansion which means that current results for this field cannot be compared with previous results because the projections for 1995–2005 in the previous report did not include this category. The major field of clinical research was also added and includes Ph.D.s in clinical areas (except psychology) and M.D.s engaged in clinical research. Data for the clinical Ph.D.s are available from the same sources as for biomedical and behavioral Ph.D.s. For the M.D.s, only limited data are available; thus, it is only possible to speculate about orders of magnitude. The disciplines included in biomedical and behavioral sciences include fields from anatomy to zoology; the behavioral sciences cover psychology, sociology, anthropology, demography, and speech-language pathology and audiology. The clinical sciences include all medical disciplines that are not considered biomedical. These categories have not changed from those covered in the previous report except for the addition of clinical psychology in the behavioral sciences. Data on the U.S.-trained Ph.D. workforce in these fields were obtained from the Survey of Doctorate Recipients (SDR), a longitudinal biennial survey. Data on graduates were obtained from the Survey of Earned Doctorates—an 1   National Research Council. 2000. Addressing the Nation’s Changing Needs for Biomedical and Behavioral Scientists. Committee on National Needs for Biomedical and Behavioral Scientists, Education and Career Studies Unit, Office of Scientific and Engineering Personnel. Washington, D.C.: National Academy Press

OCR for page 120
Advancing the Nation’s Health Needs FIGURE D-1 The potential research workforce. annual, virtually complete census of doctoral graduates in the United States.2 Data on foreign-trained Ph.Ds in the United States were obtained from the National Survey of College Graduates—rounds of this survey were conducted in 1993, 1995, 1997, and 1999, following up the migrants originally identified in the 1990 census. For clinical M.D.s, data were gleaned from the national roster of medical school faculty from the American Association of Medical Colleges (AAMC). These AAMC data allow some inferences about numbers of researchers, but projections are generally not possible. Therefore issues relating to this group are discussed further below. CURRENT WORKFORCE AND RECENT TRENDS The active workforce of Ph.D. and M.D. researchers includes all those employed in research. The potential workforce is broader and includes those unemployed, those neither employed nor looking for work (provided they are not retired), and those with jobs outside science (see Figure D-1). Some proportion of each of these groups returns to the active research workforce every year. To avoid confusion, the term “employed” is used instead of “active workforce” and refers to those employed in science jobs. Those employed in nonscience jobs are not counted in the employed group but are referred to as being outside science. The unemployed and those not in the labor force are referred to together as “not working” and combined with those outside science as “not active” in research. The potential workforce is incremented—in this model on an annual basis—by entering graduates and migrants with 2   See Appendix D in Addressing the Nation’s Changing Needs for Biomedical and Behavioral Scientists (National Research Council 2000a) for more details on these surveys.

OCR for page 120
Advancing the Nation’s Health Needs BOX D-1 How Accurate Were Previous Projections? Are the previous projections, for 1995–2005, consistent with current workforce data? The answer to this question is limited by the following: 2001 is the last date for which reported data are available at this writing. Biomedical researchers only are included because the definition of behavioral researchers has changed and clinical researchers were not previously projected. U.S.-trained researchers only are included because (as is discussed below) there is no solid current data on the foreign-trained postdoctorates for comparison purposes. Those employed outside science who were not previously counted in the potential workforce are excluded. To make the comparison, the earlier projections are rerun excluding the foreign-trained researchers but not changing any other assumptions. The projections, shown in Table D-A, up to 2001 for male and female biomedical researchers were 2-3 percent too low. By comparison with actual growth, this is not a large error. For males the error is smaller than growth in the workforce between 2000 and 2001; and, for females only a sixth of 2000-2001 growth. The error is also small relative to the uncertain number of foreign-trained researchers in the workforce. Nevertheless as small as this error is, it is not easy to explain. Looking at the number of graduates, the most important component of the projections, it can be seen that the medium projection TABLE D-A Comparison of Projected with Reported 2001 Workforce in Biomedical Research Biomedical researchers Males Females Reported, excluding nonscience employment 71,209 33,988 Projected from 1995 76,604 37,455 Projected excluding migrants 68,956 33,350 Ratio of projected (without migrants) to reported 0.968 0.981   SOURCE: Addressing the Nation’s Changing Needs for Biomedical and Behavioral Scientists. Washington, D.C.: National Academy Press (National Research Council 2000a); and the National Science Foundation Survey of Doctorate Recipients from 1995 did not capture the fluctuations in reported numbers but is reasonably close. The estimates of male graduates are slightly too high rather than too low. Another possible explanation for projection error has to do with stay rates, the proportions of graduates assumed to stay and work in the U.S., based on their stated intentions. The 1995 projections assumed constant stay rates, whereas rates actually rose, so that by the year 2001 males were 93 percent instead of the projected 90 percent. This still would not account for more than a quarter of the error. Other possibilities may include retirement or death rates being slightly too high, relatively too much assumed movement toward jobs outside science, or even errors in the data (such as errors in age distribution) that could affect projections. Because the projections seem generally accurate, many of these errors, if they exist, probably cancel each other out. doctoral degrees. It is decremented by retirements and deaths. Movements within the workforce are also modeled, such as those between employment and unemployment. The most common movements are those in both directions between science employment and nonscience employment. The biomedical and behavioral fields include only Ph.D.s. The clinical field includes Ph.D.s and M.D.s. The distinction between Ph.D.s and M.D.s is relevant partly for methodological reasons as the kinds of data available for each are quite different. Therefore, these groups are discussed separately. Similarly, a distinction between U.S.-trained and foreign-trained researchers is important because the types of data differ. The components of the current potential workforce, as illustrated in Figure D-1, are given in the following order: U.S.-trained Ph.D. researchers; M.D. researchers; foreign-trained Ph.D. researchers; entrants into the workforce, meaning migrants and graduates; and movements within the workforce and exits from it. Current levels as well as recent trends that suggest possible approaches to projection are

OCR for page 120
Advancing the Nation’s Health Needs FIGURE D-A Reported biomedical graduates and high, medium, and low projections from 1995, by sex. SOURCE: Addressing the Nation’s Changing Needs for Biomedical and Behavioral Scientists. Washington, D.C.: National Academy Press (National Research Council 2000a); and the National Science Foundation Survey of Doctorate Recipients. taken into consideration in the discussion. The characteristics of each group will be addressed below. U.S.-Trained Ph.D. Workforce Numbers Table D-1 shows the U.S.-trained potential workforce of Ph.D.s in 2001 as determined from the Survey of Doctoral Recipients. The size of the workforce is essentially equal in both the biomedical and behavioral fields, at 113,000 to 114,000 each, but much smaller in the clinical field, at 19,100. Between 87 and 90 percent of the potential workforce in each major field is employed in science, and 7 to 9 percent have jobs outside science. Nonscience jobs are more common among behavioral scientists than the other two groups. Those who are not working, whether they are unemployed or simply not looking for work, comprise 3 to 5 percent in each field. The numbers of Ph.D.s not working are

OCR for page 120
Advancing the Nation’s Health Needs TABLE D-1 Potential Workforce of U.S.-Trained Ph.D. Graduates in Three Major Fields, by Employment Status and Sex, 2001 Field and employment status Number % of Workforce Males Females Total Males Females Total Biomedical Potential workforce 75,866 37,422 113,288 67.0 33.0 100.0 Employed 69,156 31,106 100,262 61.0 27.5 88.5 Postdoctorates 6,342 5,338 11,680 5.6 4.7 10.3 Unemployed 713 306 1,019 0.6 0.3 0.9 Not in labor forcea 1,340 2,576 3,916 1.2 2.3 3.5 Out of science 4,657 3,434 8,091 4.1 3.0 7.1 Clinical Potential workforce 8,149 10,956 19,105 42.7 57.3 100.0 Employed 7,526 9,654 17,180 39.4 50.5 89.9 Postdoctorates 136 279 415 0.7 1.5 2.2 Unemployed 3 124 127 0.0 0.6 0.7 Not in labor forcea 74 330 404 0.4 1.7 2.1 Out of science 546 848 1,394 2.9 4.4 7.3 Behavioral and Social Potential workforce 59,175 54,822 113,997 51.9 48.1 100.0 Employed 52,606 46,540 99,146 46.1 40.8 87.0 Postdoctorates 759 1,377 2,136 0.7 1.2 1.9 Unemployed 280 509 789 0.2 0.4 0.7 Not in labor forcea 647 2,771 3,418 0.6 2.4 3.0 Out of science 5,642 5,002 10,644 4.9 4.4 9.3 aNot employed, not looking for work, but not retired. SOURCE: National Science Foundation Survey of Doctorate Recipients. higher among women: 2 to 3 percent of the total workforce compared to less than 2 percent of men. Trends Over the past three decades, the U.S.-trained workforce in each field has grown steadily. From 1973–2001, the average annual workforce growth rates have been 4.1 percent for biomedical researchers, 6.8 percent for clinical researchers, and 4.8 percent for behavioral researchers (see Figure D-2). Growth in employment outside science has contributed to this (see Figure D-3). When only those employed in science are considered, growth rates would be 0.2 to 0.5 percentage points lower. More recently, from 1991 to 2001, growth rates were lower for the workforce as a whole: 3.7 percent for biomedical, 6.3 percent for clinical, and 2.7 percent for behavioral researchers. As a proportion of the potential workforce, the percentage of those not working has increased. Their proportion varied roughly between 1.5 and 2.5 percent up to 1990, but the range has since increased to 3.0 to 4.5 percent. These percentages appear small, but the numbers involved have risen rapidly. For example, in the biomedical field, those not working numbered 1,500 in 1989. This increased to 4,900 in 2001. Two major factors contributed to this increase. First, the proportion not working was consistently higher in each field among females, who comprised only 10 to 20 percent of the workforce in the 1970s and increased to 30 to 60 percent in 2001. Second, around 1990, the proportion of males not working began to increase. Nevertheless, total unemployment has not increased by much, neither among males nor females. Notably, most of the increase in those not working has involved people staying out of the labor force. This not-working group represents a small proportion of the workforce, but the absolute numbers have grown rapidly among females, and males in the not-working category began to outnumber the unemployed in the early 1990s (see Figure D-4). People may stay out of the labor force to care for chil-

OCR for page 120
Advancing the Nation’s Health Needs FIGURE D-2 Trends in the U.S.-trained potential workforce by major field, 1973–2001. SOURCE: National Science Foundation Survey of Doctorate Recipients. dren, to take a break between jobs, or for other personal reasons, but the extent to which this is voluntary is not evident from the data. M.D. Workforce Numbers These data include M.D.s if they also have a Ph.D. from a U.S. institution. There are few data on clinical researchers with only M.D. degrees. Medical school faculty in the year 2000 totaled 90,678, but only a small minority of these should be considered clinical researchers (see Table D-2). of the total, only 64 percent have M.D.s, rather than Ph.D.s, M.D./Ph.D.s, or no doctoral degree (M.D./Ph.D.s are grouped with Ph.D.s because they are included in datasets on Ph.D.s). of the M.D.s, 96 percent are in clinical departments and the rest are in biomedical departments (anatomy, biochemistry, microbiology, basic pathology, pharmacology, physiology, and other basic sciences). of the clinical M.D.s, only 2,879, or 5.2 percent, have ever had an R01 grant. Therefore, it is feasible to add the 2,879 M.D. researchers to the 19,105 clinical researchers in the workforce. This number allows for M.D.s who are not currently doing research but have done so in the past, yet it is still a substantial underestimate for three reasons. First, R01 grants are not the only possible type of grant; research funds from pharmaceutical companies, foundations, and other sources compose a sig-

OCR for page 120
Advancing the Nation’s Health Needs FIGURE D-3 Percent in conscience jobs by major field, 1973–2001. SOURCE: National Science Foundation Survey of Doctorate Recipients. nificant part of the research enterprise. Second, M.D. researchers may not be on medical faculties. Third, some additional (though very small) number currently may not be working or may be employed outside science. While one could argue that all M.D.s not in research should be categorized as “outside science” in the workforce, the discussion here is limited by the available data. Fourth, the criterion that requires evidence of a research grant may be too strict as Ph.D.s may not have had a research grant but are still counted in the potential workforce. Of the Ph.D.s and M.D./Ph.D.s in clinical departments, 3,860 have had R01 grants. These individuals were included among those counted in the previous U.S.-trained Ph.D. discussion. However, the 13,577 who have not had an R01 grant were also counted. Ph.D.s and M.D./Ph.D.s in clinical departments make up 90 percent of the clinical Ph.D. workforce, although the proportion is somewhat uncertain due to lack of data on the foreign-trained Ph.D. component. These Ph.D.s and M.D./Ph.D.s in clinical departments total 4.5 times the number that have had R01 grants. When this same ratio is applied to the number counted as M.D. clinical researchers, it results in a total of 13,000. While there is some ambiguity in defining the M.D. clinical research workforce, it could be assumed that their numbers are between 3,000 and 13,000. This means that the number of clinical researchers regardless of degree type would have to be increased from the earlier estimate by as little as 16 or as much as 68 percent to incorporate the M.D. researchers who would, however, still

OCR for page 120
Advancing the Nation’s Health Needs FIGURE D-4 Numbers unemployed and not in labor force for all fields combined by sex, 1973–2001. SOURCE: National Science Foundation Survey of Doctorate Recipients. be in the minority. Given the considerable uncertainty in these numbers, no attempt is made to project clinical M.D. researchers other than to make the following observations. Trends and Characteristics The faculty roster data show that M.D.s in clinical departments who have ever had an R01 grant increased from 2,482 in 1993 to 3,090 in 2003, which is an annual rate of increase of 2.4 percent. This is much slower than the rate of increase for clinical Ph.D. researchers of 6.3 percent annually (from 1991 to 2001) and slightly slower than faculty growth in clinical departments. However, M.D. clinical researchers with other types of grants or those not on medical faculties might have increased more rapidly. There is a striking contrast between the male-female ratios for both M.D. and Ph.D. clinical researchers. Ph.D. clinical researchers are 57 percent female (see Table D-1). However, M.D. clinical researchers on medical faculties who have had R01s are 84 percent male. In each clinical discipline, male faculty with M.D.s and R01 experience outnumber equivalent female faculty by at least 2 to 1 and often much more. This is true from anesthesiology to surgery, including such fields as family medicine, pediatrics, and obstetrics and gynecology. Since Ph.D. clinical faculty with R01s are also largely male (71 percent), the mechanisms by

OCR for page 120
Advancing the Nation’s Health Needs TABLE D-2 Medical School Faculty, by Major Field and Degree, and Number and Percent Having Had an R01 Grant, 2000   Total Biomedical Clinical Behavioral Faculty Total 90,678 14,583 76,073 22 M.D.s 58,104 2,285 55,819 0 Ph.D.s 23,759 11,030 12,715 14 M.D./Ph.D.s 5,706 983 4,722 1 Other 3,109 285 2,817 7 Faculty who have ever had an R01 grant Total 11,824 5,033 6,787 4 M.D.s 3,183 304 2,879 0 Ph.D.s 6,960 4,191 2,767 2 M.D./Ph.D.s 1,585 490 1,093 2 Other 96 48 48 0 Percent who have ever had an R01 Total 13.0 34.5 8.9 18.2 M.D.s 5.5 13.3 5.2 a Ph.D.s 29.3 38.0 21.8 a M.D./Ph.D.s 27.8 49.8 23.1 a Other 3.1 16.8 1.7 a aFewer than 20 base cases. NOTE: Major field is defined by department, which may not agree exactly with field definitions in Table D-1. SOURCE: American Association of Medical Colleges Faculty Roster. which medical faculty obtain R01s may have something to do with the small proportion of females. Foreign-trained M.D.s on faculty rosters represented 15 percent of those with R01s in 1993, rising slightly to 17 percent by 2002. It is important to distinguish between the data on Ph.D.s discussed in the previous section, which only included U.S.-trained Ph.D.s, and this section, which integrates an estimate of foreign-trained Ph.D.s. The following discussion will address this issue of the number to add for foreign-trained Ph.D.s. Foreign-Trained Ph.D. Workforce Numbers and Trends Foreign-trained Ph.D.s are an important component of the workforce. However, the data are limited and provide uncertain results. The data for foreign-trained Ph.D.s come from the National Survey of College Graduates. This survey covers only those identified in the 1990 U.S. census. Therefore the numbers for those who migrated since 1990 and for attrition up to 2001 needed to be determined. This was done in several steps. First, those identified in the 1993 survey by date of entry into the United States are broken down and then are “reverse survived” by age group for each field and sex to obtain estimates of their original numbers at entry going back as far as 1980. Second, this flow of migrants, from 1980 to 1990, is projected forward to 2001 using regression models. Third, by taking the distribution of the foreign trained in 1993 as roughly equivalent to the 1990 Census cohort, they are projected forward from 1990 to 2001 by adding, year by year, the estimated flow of migrants. This same projection model is used just as in the main exercise, although in this case it deals only with migrants and not U.S. graduates. This model allows for the accumulating numbers of migrants and for attrition. These rates are similar to those described below for the workforce as a whole. Figure D-5 shows data and results in the biomedical field. The stock of foreign-trained biomedical Ph.D.s in the 1993 survey (at the left end of the middle line) was 8,800. This stock declined in subsequent surveys since migrants after 1990 were not added to the sample. The slight increase in 1999 occurred because survey nonresponse was reduced. Working backward from these levels, the migrant flow over the 1980s is estimated, shown by the line on the lower left. This flow, referred to as “entrants” to distinguish them from the stock of migrants, rose from 400 to over 700 annually during the 1980s. These entrants are added to the stock to give the projected stock of foreign-trained Ph.D.s, which doubled from 1993 to 2001, when it reached 17,400. This projected stock analysis relies on the trends of the 1980s entrants. There are no data that allow direct assessment of whether changes in the flow took place in the 1990s. However, there are data on postdoctorates that permit a similar exercise to be conducted. Assuming that foreign-trained postdoctorates entered the United States just before accepting their posts, their dates of entry must have been more recent, on average, than those of migrants captured in the 1990 Census. Again, the data used are from the Graduate Student Survey, which includes numbers of postdoctorates by calendar year and field. Those with M.D.s rather than Ph.D.s are eliminated. Those with U.S. rather than foreign doctoral training are determined by applying ratios of temporary residents to U.S. citizens and permanent residents. In addition, stay rates are estimated for graduates. Both of these will be discussed further below. An estimate is made of when each postdoctorate started, using data on the duration an individual spent as a postdoctorate, and the assumption is made that this date is identical to the date of migration. Figure D-6 resembles Figure D-5 (and is drawn to the same scale to facilitate comparison). The middle line shows the stock of foreign-trained postdoctorates, and the bottom line shows entrants estimated, as described above, from this stock. The stock numbers in Figure D-6 are consistent with those in Figure D-5, at least to the extent of indicating a smaller number of foreign-trained postdoctorates than foreign-trained researchers generally. The entrant numbers, however, are inconsistent, which suggests that more foreign-

OCR for page 120
Advancing the Nation’s Health Needs FIGURE D-5 Proportion of temporary residents among graduates and proportion of each group intending to stay in the United States: Means for 1997–2001. SOURCE: National Science Foundation Survey of Earned Doctorates. trained Ph.D.s entered the U.S. to take postdoctorates than the total number of entering foreign-trained Ph.D.s. The reasons for this inconsistency are not clear. Either the data or the assumptions must be incorrect in Figure D-5 or Figure D-6, but it is not possible to determine which. One scenario would be to take the estimated entrants in Figure D-6, add a minimal 10 percent to allow for entrants who do not take postdoctorates, and apply the trend to 1993 stock numbers for all foreign-trained Ph.D.s. This results in a projected stock shown in the top line in Figure D-6. Note that these numbers are higher than those in Figure D-5 by 40 percent in the year 2001. Consequently, both estimates referring to the projected stock in Figure D-5 will be termed “medium” estimates, while those in Figure D-6 will be termed “high” estimates. Similar results can be produced for foreign-trained clinical Ph.D.s. As for biomedical Ph.D.s, estimates based on surveyed migrants are lower than estimates based on reported postdoctorates. The former are obtained by working backwards from a stock of 2,200 foreign-trained researchers in 1993. This results in just under 200 annual entrants in the 1980s but follows an upward trend. Projections based on this trend would result in foreign-trained researchers tripling during the 1990s and reaching 6,200 by 2001. However, using reported postdoctorates results in an estimate that is 30 percent higher by 2001. Again, the former is considered a medium estimate and the latter a high estimate. In contrast, foreign-trained behavioral researchers have a different picture. In 1993 this group was almost 50 percent more than clinical researchers, but proportionally more of

OCR for page 120
Advancing the Nation’s Health Needs FIGURE D-6 Foreign-trained biomedical Ph.D.s: postdoctorates, postdoctoral entrants, and projected total stock. SOURCE: National Science Foundation Survey of Graduate Students and Postdoctorates in Science and Engineering; and National Science Foundation National Survey of College Graduates. them arrived before 1980, and during the 1980s entrants were few topping off at 100 annually. Projections based on this flow would indicate an increase in the stock to 4,300 by 2001. Foreign-trained postdoctorates are very few. As a result, the stock averages about 100 over two decades. For this group the projected stock from the migrant survey gives the high estimate, and the medium estimate is created by assuming a constant inflow of 100 per year. Characteristics The estimates for 2001 of foreign-trained Ph.D.s in each field-broken down by sex and employment status based on the 1993 survey-are shown in Table D-3. In the biomedical field there are between 17,400 and 24,800 foreign-trained Ph.D.s who make up 13 to 18 percent of the entire workforce. Although the number of foreign-trained Ph.D.s in the clinical field is smaller, they comprise between 25 and 30 percent of the workforce, which is a larger proportion; only 17 percent of the M.D. clinical researchers are on medical faculties. Finally, in the behavioral field they are much less consequential, at 3.0 to 3.6 percent of the entire workforce. The distribution of foreign-trained Ph.D.s by employment status resembles that for U.S.-trained Ph.D.s with one exception: female biomedical researchers appear substantially more likely to be out of the labor force. The balance between males and females differs somewhat by field. In clinical research, foreign-trained Ph.D.s are more likely to be male than the U.S.-trained Ph.D.s, but the reverse appears to be the case in the other two fields (among M.D. clinical re-

OCR for page 120
Advancing the Nation’s Health Needs TABLE D-12 Projected U.S.-Trained and Foreign-Trained Workforce and Employed Researchers by Field, Medium Scenario, 2001–2011 Field and Year Potential Workforce Employed U.S. Trained Foreign Trained % Foreign U.S. Trained Foreign Trained % Foreign Biomedical 2001 113,289 17,437 13.3 100,262 14,627 12.7 2002 117,175 18,330 13.5 103,148 15,231 12.9 2003 120,953 19,250 13.7 105,851 16,061 13.2 2004 124,661 20,179 13.9 108,544 16,987 13.5 2005 128,335 21,111 14.1 111,305 17,931 13.9 2006 131,992 22,057 14.3 114,147 18,883 14.2 2007 135,590 23,015 14.5 117,010 19,833 14.5 2008 139,135 23,978 14.7 119,875 20,774 14.8 2009 142,632 24,952 14.9 122,730 21,706 15.0 2010 146,082 25,937 15.1 125,564 22,627 15.3 2011 149,482 26,918 15.3 128,361 23,528 15.5 Clinical 2001 19,104 6,178 24.4 17,180 5,840 25.4 2002 20,177 6,661 24.8 18,281 6,299 25.6 2003 21,273 7,166 25.2 19,344 6,781 26.0 2004 22,368 7,691 25.6 20,405 7,281 26.3 2005 23,454 8,232 26.0 21,459 7,791 26.6 2006 24,536 8,796 26.4 22,504 8,314 27.0 2007 25,621 9,381 26.8 23,546 8,853 27.3 2008 26,694 9,985 27.2 24,563 9,415 27.7 2009 27,744 10,608 27.7 25,541 10,001 28.1 2010 28,785 11,246 28.1 26,498 10,604 28.6 2011 29,818 11,898 28.5 27,444 11,218 29.0 Behavioral 2001 113,997 3,469 3.0 99,146 3,047 3.0 2002 116,267 3,470 2.9 101,902 3,049 2.9 2003 118,368 3,465 2.8 104,256 3,046 2.8 2004 120,408 3,454 2.8 106,418 3,046 2.8 2005 122,375 3,438 2.7 108,416 3,041 2.7 2006 124,247 3,423 2.7 110,281 3,038 2.7 2007 125,960 3,411 2.6 111,957 3,036 2.6 2008 127,497 3,395 2.6 113,432 3,030 2.6 2009 128,868 3,376 2.6 114,729 3,018 2.6 2010 130,101 3,349 2.5 115,887 2,997 2.5 2011 131,154 3,312 2.5 116,874 2,966 2.5   SOURCE: NRC analysis. pathologists. To compare with behavioral researchers, these categories are totaled without weighting. In the last two BLS projections, the number of employed implies an average annual growth rate for biological scientists of 1.8 to 1.9 percent; for medical scientists a rate of 2.4 percent; and among psychologists, other social scientists, and speech-language pathologists a combined rate of 2.1 to 2.2 percent. These growth rates are reasonably close to those estimated from 20-year trends. Note, however, that BLS constructs employment projections differently. BLS assumes a full-employment economy, takes labor supply into account, assesses growth in industries’ outputs and intermediate inputs, and applies occupational staffing patterns by industry, coupled with expert assessments of likely trends.6 The confluence of results from the BLS approach with at least our longer-term trend may provide some additional confidence. In the past, BLS has underestimated growth in professional as well as service occupations.7 For 1988-2000 the Bureau’s projected annual growth rate for biological scientists of 2 percent was well below the actual growth rate of 3.3 percent, and the rate for psychologists of 2 percent was also well below the actual rate of 2.8 percent (for all occupations combined, the BLS projection was for 1.2 percent annual growth compared to the actual 1.6 percent growth rate). Furthermore, the BLS projected growth rates for 2000–2010 and 2002–2012 are even smaller than the rates projected for 1988–2000. If the new projections are similarly too conservative, the higher rates, at least for biomedical and clinical researchers, estimated from our five-year trend may be appropriate. Applying BLS growth rates to the reported number of positions provides alternative projections of the positions to become available. These are shown in Table D-15. These projections do not represent actual BLS numbers but rather the numbers that would be obtained by applying their implied growth rates beginning in 2001. Estimates by sector are derived by partitioning total increases in proportion to increases in the 20-year trend projections. The projections from the five-year trend clearly stand out. Over the decade the BLS growth rates, as well as the 20-year trend, imply an increase of about 20,000 biomedical researchers, but the five-year trend implies an increase of almost 30,000. Similarly, the BLS and the 20-year trend imply an increase of about 4,700 clinical researchers, as opposed to 7,700 from the five-year trend. How these projections compare with the projected workforce is the critical issue. WORKFORCE VERSUS POSITIONS Comparisons are made with projected U.S.-trained researchers because research positions have been projected from survey data limited to U.S. graduates. (The BLS projections presumably allow for migrants, but their growth rates are applied to our own base-sector employment num- 6   Finn, M. G. 2003. Stay rates of foreign doctorate recipients from U.S. universities, 2001. Paper prepared at the Oak Ridge Institute for Science and Education for the Division of Science Resources Studies, National Science Foundation. 7   Alpert, A., and J. Auyer. 2003. The 1988–2000 employment projections: How accurate were they? Occupational Outlook Quarterly (Spring): 3–21.

OCR for page 120
Advancing the Nation’s Health Needs TABLE D-13 Percent Foreign Trained in Alternative Scenarios, Projected 2011 Total Workforce, and Comparisons of Total to Medium Scenario, by Field Scenario % Foreign Trained Among Entrants (2002–2011) 2011 Total Workforce Ratio to Medium Scenario Total Workforce Biomedical Clinical Behavioral Biomedical Clinical Behavioral Biomedical Clinical Behavioral High 23.3 31.2 4.5 198,703 48,181 139,863 1.13 1.15 1.04 Rising 15.8 26.6 2.0 178,203 42,011 135,139 1.01 1.01 1.01 Medium 15.8 26.6 2.0 176,400 41,716 134,466 1.00 1.00 1.00 90% 13.2 21.5 1.8 172,702 39,843 134,180 0.98 0.96 1.00 50% 8.7 14.2 1.0 163,699 36,886 133,105 0.93 0.88 0.99 0% 0.0 0.0 0.0 152,346 33,184 131,791 0.86 0.80 0.98 Low 15.3 24.9 1.9 168,770 38,355 131,149 0.96 0.92 0.98   SOURCE: NRC analysis. FIGURE D-15 Trends in employment by major field and sector, 1973–2001. SOURCE: National Science Foundation Survey of Doctorate Recipients.

OCR for page 120
Advancing the Nation’s Health Needs TABLE D-14 Linear Regressions for Trends in Employment in Various Sectors   1997–2001 Data 1981–2001 Data Constant B t-test R2 Constant B t-test R2 Biomedical researchers Faculty –1050820 544.25 8.44 0.986 –943405 490.08 6.88 0.840 Other academics –961154 485.00 5.92 0.972 –550611 279.69 15.65 0.965 Industry –3618579 1823.00 6.81 0.979 –1942012 984.15 16.98 0.970 Government –329386 168.50 1.89 0.781 –300883 154.16 12.02 0.941 Other –238397 121.25 12.18 0.993 20757 –8.42 –0.29 0.009 Postdoctorates 430600 –209.25 –1.02 0.511 –784436 398.21 7.11 0.849 All employed –5773736 2935.75 7.12 0.981 –4500066 2297.61 19.18 0.976 Clinical researchers Faculty –891776 449.75 10.42 0.991 –529538 268.42 15.27 0.963 Other academics –207292 104.25 4.49 0.953 –103572 52.32 9.65 0.912 Industry –418385 211.25 2.82 0.888 –292163 148.06 19.47 0.977 Government –67025 34.25 0.42 0.147 –104107 52.80 11.18 0.933 Other –12660 7.00 1.01 0.505 –85473 43.38 8.65 0.893 Postdoctorates 57922 –28.75 –1.52 0.698 –33401 16.92 3.55 0.583 All employed –1539217 777.75 7.56 0.983 –1148253 581.90 18.02 0.973 Behavioral researchers Faculty –740856 387.50 11.09 0.992 –750028 391.74 5.89 0.794 Other academics –1009329 510.00 46.49 1.000 –716760 363.48 15.09 0.962 Industry –1881172 956.50 4.95 0.961 –2296361 1164.63 19.79 0.978 Government –82699 46.25 1.81 0.766 –683510 346.82 8.89 0.898 Other –192626 100.75 7.67 0.983 69222 –30.17 –0.38 0.016 Postdoctorates 247152 –122.50 –2.22 0.832 –137383 69.70 3.06 0.509 All employed –3648552 1873.00 12.95 0.994 –4512779 2305.16 31.26 0.991   SOURCE: Analysis based on National Science Foundation Survey of Earned Doctorates. bers; consequently the resulting employment projections share the same limitation). Comparisons are made not with the potential workforce as a whole but with the numbers employed assuming that some proportion of the future workforce (as at present) will take jobs outside science, choose not to work, or be unemployed (possibly temporarily between jobs). Table D-17 shows the comparisons between the projected numbers employed in the workforce and the projected numbers of positions from sector employment trends and BLS projections. Results vary by field, and Figure D-16 focuses on some key comparisons. On the one hand, U.S.-trained biomedical researchers will be close to the available number of positions in 2006 and 2011 if the five-year trend projections are correct. On the other hand, they will be about 4 percent too numerous in 2006 and 6 to 7 percent too numerous in 2011 if the 20-year projections and the BLS projections are correct. Possibly, a small deficit would not be an issue since it could presumably be filled by those projected not to be active in science or by foreign-trained Ph.D.s. A possible excess, however, could be more of an issue, particularly since the foreign-trained Ph.D.s, who are not in the comparison, appear to be increasing more rapidly than U.S.-trained biomedical researchers. Where clinical researchers are concerned, each comparison indicates that the workforce will exceed future available positions. By 2011 the excess will be 10 percent if the five-year trend continues and about 25 percent whether the 20-year trend continues or the BLS is accurate. This is a relatively small field that is growing rapidly. Some adjustment to slower workforce growth can be expected in the future but at what point and what level is not possible to say. However, the situation for behavioral researchers is different. By 2006 little by way of excess or deficit will be evident. By 2011 this will still hold if the five-year trend is correct. If the 20-year trend or the BLS projections are accurate, there will be a deficit in the number of U.S.-trained Ph.D.s of around 5 percent. Because there are fewer foreign-

OCR for page 120
Advancing the Nation’s Health Needs TABLE D-15 Positions by Sector and Increases from 2001–2011 In Alternative Projections, by Major Field Field and Sector Positions in 2001 5-Year Trend 20-Year Trend Projected 10-Year Increase Based on BLS Rates for 2000–2010 2002–2012 Biomedical Facultya 38,299 5,368 3,842 3,915 3,810 Other academic 9,236 4,945 2,612 2,662 2,591 Industry 28,935 18,539 8,170 8,325 8,103 Government 7,886 1,582 1,254 1,278 1,244 Other 4,213 1,224 –385 –392 –382 Postdoctorate 11,655 –1,856 4,716 4,806 4,678 All sectors 100,224 29,801 20,210 20,594 20,045 Clinical Facultya 8,124 4,547 2,127 2,068 2,098 Other academic 1,285 1,069 364 354 359 Industry 4,413 2,026 1,171 1,138 1,155 Government 1,604 247 479 466 473 Other 1,339 78 429 417 423 Postdoctorate 415 –309 207 201 204 All sectors 17,180 7,659 4,777 4,643 4,711 Behavioral Facultya 34,491 3,915 3,263 3,251 3,324 Other academic 11,194 5,087 2,998 2,987 3,055 Industry 32,561 9,788 13,143 13,093 13,390 Government 9,877 433 4,073 4,058 4,149 Other 8,960 1,023 –415 –413 –423 Postdoctorate 2,093 –1,289 691 688 704 All sectors 99,176 18,958 23,754 23,663 24,199 NOTE: BLS rates are Bureau of Labor Statistics employment projections. aTenured or on tenure track—those on nontenure track are counted under “Other academic.” SOURCE: Analysis based on National Science Foundation Survey of Earned Doctorates and Bureau of Labor Statistics employment projections (Hecker, 2001, 2004). trained migrant Ph.D.s, their numbers are unlikely to be sufficient to fill any gap. In principle, a deficit could be filled from the portion of the workforce with jobs outside science because this field is larger than the other two. However, whether this is practical would have to be investigated. Therefore, the three fields have different prospects. Nor is each field homogeneous, being composed of a variety of specific disciplines and separate sectors. Although they deserve attention disciplines cannot be considered here. Instead sectoral trends are briefly examined. SECTORAL PROSPECTS AND POSTDOCTORATES Sectors If employment grows according to the projections given here, the distribution by sector of those employed in science could show one important change: the balance could shift in the other direction of those employed as faculty (either tenured or on a tenure track) in the biomedical and social and behavioral fields in 2001 who outnumbered those in industry or self-employed by 2011. In the social and behavioral field this will happen whether growth is as slow as projected from the five-year trend or as fast as projected from the 20-year trend. But in the biomedical field it will happen only with the faster growth scenario from the five-year trend. However, academic employment in each field will still be higher than industrial employment if nonfaculty appointments and postdoctorates are taken into account. Nontenure appointments are a growing part of academic employment. The growth has been steady since the 1970s. Those who are neither tenured nor on a tenure track are close to becoming half of all academic employees in the biomedical field. This proportion is lower in the other fields but is TABLE D-16 Projected and Actual Growth Rates (%) for Various Occupational Groups 2001–2011 Projections Biomedical Researchers Clinical Researchers Behavioral Researchers   1997–2001 Data 2.60 3.69 1.75   1981–2001 Data 1.84 2.45 2.15 Bureau of Labor Statistics Biological Scientists Medical Scientists All Social Scientistsa Psychologists Other Social Scientists Speech-Language Pathologists 2002–2012 Projection 1.82 2.42 2.18 2.19 0.57 2.44 2000–2010 Projection 1.87 2.39 2.14 1.62 1.25 3.27 1988–2000 Projection 1.95 b b 1.99 b 2.08 1988–2000 Actual 3.33 b b 2.77 b 5.78 aCombines psychologists, other social scientists, and speech-language pathologists. bNo reported data. SOURCES: Bureau of Labor Statistics rates are estimated from projections for 2002–2012 (Hecker, 2004), 2000–2010 (Hecker, 2001), and 1988–2000 (Alpert and Auyer, 2003) and data reported in Alpert and Auyer (2003).

OCR for page 120
Advancing the Nation’s Health Needs TABLE D-17 Excess or Deficit of U.S.-Trained Researchers in Relation to Various Projections of Research Positions Field and Year Excess or Deficit Relative to Positions Excess or Deficit as % of Positions Projected from Trend Over Based on BLS Growth Rates to Projected from Trend Over Based on BLS Growth Rates to 5 Years 20 Years 2010 2012 5 Years 20 Years 2010 2012 Biomedical 2002 –483 3,395 1,033 1,080 –0.5 3.4 1.0 1.1 2003 –712 3,800 1,810 1,905 –0.7 3.7 1.7 1.8 2004 –952 4,196 2,541 2,685 –0.9 4.0 2.4 2.5 2005 –1,124 4,659 3,302 3,499 –1.0 4.4 3.1 3.2 2006 –1,215 5,203 4,107 4,357 –1.1 4.8 3.7 4.0 2007 –1,284 5,768 4,894 5,200 –1.1 5.2 4.4 4.7 2008 –1,352 6,335 5,644 6,008 –1.1 5.6 4.9 5.3 2009 –1,430 6,892 6,344 6,768 –1.2 5.9 5.5 5.8 2010 –1,529 7,428 6,983 7,468 –1.2 6.3 5.9 6.3 2011 –1,664 7,927 7,543 8,092 –1.3 6.6 6.2 6.7 Clinical 2002 442 1,561 685 680 2.5 9.3 3.9 3.9 2003 727 2,042 1,322 1,311 3.9 11.8 7.3 7.3 2004 1,011 2,521 1,947 1,930 5.2 14.1 10.5 10.4 2005 1,287 2,993 2,554 2,530 6.4 16.2 13.5 13.4 2006 1,554 3,456 3,141 3,111 7.4 18.1 16.2 16.0 2007 1,818 3,916 3,714 3,678 8.4 20.0 18.7 18.5 2008 2,058 4,351 4,251 4,207 9.1 21.5 20.9 20.7 2009 2,258 4,747 4,737 4,686 9.7 22.8 22.8 22.5 2010 2,437 5,122 5,191 5,131 10.1 24.0 24.4 24.0 2011 2,605 5,487 5,621 5,553 10.5 25.0 25.8 25.4 Behavioral 2002 675 –272 581 537 0.7 –0.3 0.6 0.5 2003 1,150 –224 744 653 1.1 –0.2 0.7 0.6 2004 1,434 –369 667 529 1.4 –0.3 0.6 0.5 2005 1,553 –677 378 189 1.5 –0.6 0.3 0.2 2006 1,540 –1,118 –94 –335 1.4 –1.0 –0.1 –0.3 2007 1,337 –1,748 –805 –1,100 1.2 –1.5 –0.7 –1.0 2008 934 –2,579 –1,769 –2,121 0.8 –2.2 –1.5 –1.8 2009 352 –3,589 –2,964 –3,375 0.3 –3.0 –2.5 –2.9 2010 –368 –4,737 –4,351 –4,823 –0.3 –3.9 –3.6 –4.0 2011 –1,260 –6,056 –5,965 –6,501 –1.1 –4.9 –4.9 –5.3   SOURCE: NRC analysis. rising just as insistently. The biomedical field is ahead in this trend mainly because of a high number of postdoctorates, which was rising until the late 1990s. Postdoctorates A specific look at postdoctorates is warranted, first those who are U.S. trained, since the foreign-trained postdoctorates data are not included in the sectoral data. Though many Ph.D. graduates may spend some time as postdoctorates, this period is generally short in comparison to other employment, so that at any given time postdoctorates are few relative to total employment. Among clinical and behavioral Ph.D.s, postdoctorates are only 1.5 to 3.5 percent of the total. In the biomedical field, though, they compose about 10 percent of total employment. U.S.-trained postdoctorates are not necessarily young. Roughly 40 percent tend to be age 25 to 34, but about the

OCR for page 120
Advancing the Nation’s Health Needs FIGURE D-16 Percent excess or deficit of U.S.-trained researchers in relation to various projections of research positions, 2006 and 2011. SOURCE: NRC analysis and Bureau of Labor Statistics. same proportion are 35 to 44. In the clinical and behavioral fields, about 20 to 25 percent are 45 and older. This is for years 1997, 1999, and 2001, across which there is little change. At younger ages, proportionally more of those employed are postdoctorates; this is particularly true in the biomedical field, where, under age 35, 50 percent of employed males and 60 percent of employed females are postdoctorates. For the remainder of the postdoctoral pool, more limited information is available from department reports in the Graduate Student Survey, which was used to estimate the foreign-trained workforce. These reports indicate that postdoctorates with U.S. Ph.D.s, who have been considered first because of better information, are actually a minority in two fields, making up only 49 percent of biomedical postdoctorates and 40 percent of clinical postdoctorates in 2001. Among the much smaller contingent of behavioral postdoctorates they form a majority of 80 percent (see Figure D-17). M.D.s and foreign-trained Ph.D.s take the remaining positions, M.D.s being more important among clinical postdoctorates and foreign-trained Ph.D.s more important among biomedical and behavioral postdoctorates. The trends for M.D.s and foreign-trained Ph.D.s are opposite. The percentage of clinical postdoctorates taken by M.D.s has fallen by more than a third in two decades, from 59 percent in 1983 to 37 percent by 2001. Over the same period, the percentage of postdoctorates taken by foreign-trained Ph.D.s roughly doubled in each of the three fields, reaching 39 percent in biomedical research, 22 percent in clinical research, and 18 percent in behavioral research. The importance of noncitizen researchers may be even greater than these figures indicate. Among U.S.-trained Ph.D.s with postdoctorates, a number are on temporary resident visas: 25 percent in biomedical research, 23 percent in clinical research, and 8 percent in behavioral research. These percentages have risen about as fast as those for foreign-trained Ph.D.s. Note, however, that department reports are not entirely consistent with sector data. Figure D-18 matches comparable groups: academic postdoctorates in sector data (who must, because of the survey sample, have U.S. Ph.D.s) and postdoctorates with U.S. Ph.D.s in department data (who

OCR for page 120
Advancing the Nation’s Health Needs FIGURE D-17 Postdoctorates reported by graduate departments, by field, degree, and visa status, 1983–2001. SOURCE: National Science Foundation Survey of Graduate Students and Postdoctorates in Engineering and Science, 2001. must be academic). The sector data suggest a surge in academic postdoctorates in the 1990s that does not appear in the department data, except among clinical researchers. In the clinical field, however, the department data indicate far more postdoctorates (even with M.D.s excluded) than the sector data. The definition of clinical disciplines may be different between surveys, and some of the postdoctorates in clinical departments might have biomedical Ph.D.s. It is also possible that our partitioning of temporary residents into U.S. trained and foreign trained is inaccurate and that the foreign trained are even more common than they appear among clinical Ph.D.s. However, these possible explanations cannot be verified, and the inconsistencies need further investigation. If sector data (which provide ages) is relied upon and it is assumed that the proportions of Ph.D.s on postdoctorates stay constant by age, these proportions can be applied to the projected workforce to estimate the number likely to be seeking postdoctorates in the near future. This can be done only for U.S.-trained Ph.D.s—a significant but unavoidable limitation. Table D-18 shows their numbers as well as the numbers of projected postdoctorate positions, taken from the preceding sectoral projections. Numbers and positions are compared in Figure D-19. U.S.-trained graduates seeking postdoctorates may or may not outnumber the positions available for them. In the biomedical field, individuals and positions may in fact be in balance if the 20-year trend continues, but individuals will be far too numerous—by 5,600 by 2001 or 57 percent more than the available positions—if the more recent downturn reflected in the five-year trend continues. In clinical research, individuals will outnumber positions by 15 percent even given the more optimistic 20-year trend in positions, and if the more pessimistic five-year trend comes to pass, the number of individuals will be a startling eight times greater than the number of positions available. Similarly, in behavioral research, there will be either 13 percent more individuals than there are positions available or four times the number of positions, depending on which projection is used. A projected downturn in positions derived from the five-year trend is responsible for the large shortfalls; this essentially extrapolates from the recent decline in the number of positions taken by U.S.-trained Ph.D.s. This decline at least partly reflects the rising share of postdoctorates taken by foreign-trained Ph.D.s. Whether the foreign inflows—which also include

OCR for page 120
Advancing the Nation’s Health Needs FIGURE D-18 Postdoctorates compared from sector data and department reports, by field 1973–2001. SOURCES: National Science Foundation Survey of Earned Doctorates, 2001, and National Science Foundation Survey of Graduate Students and Postdoctorates in Engineering and Science, 2001. rising U.S.-trained temporary residents—will continue is uncertain. CONCLUSION This discussion began with the question of whether the right numbers of people are being trained for the future research workforce. To answer this question, first data on the workforce and its components were examined and then recent trends were modeled and projected into the future. Such workforce projections appear to have worked in the past. Projections from 1995 were generally accurate up to 2001, in the partial comparisons that are possible with current data. Nevertheless, the projections have limitations, rooted especially in inadequate data. Projections have been made only for Ph.D. researchers and not for M.D. researchers since there was only a rough estimate of the numbers of the latter. Our projections for foreign-trained Ph.D.s are based on dated and inconsistent data. Rates were assumed for movement between statuses that are constant within age groups, and adequate information was not available on trends in rates. A basic assumption could be made that future trends will resemble past trends. Trends in graduates (and to a degree in prior enrollment levels) are assumed to follow a future course modeled on past trends. Migration is assumed to follow previous trends, though alternatives were also modeled. Retirement ages are assumed to maintain their current distribution. Graduation, migration, and retirement are discretionary behaviors that could be postponed at will, within limits, or possibly accelerated, but no attempt was made to model the factors underlying such decisions. The projections have been extended in another dimension

OCR for page 120
Advancing the Nation’s Health Needs TABLE D-18 Projected U.S.-Trained Postdoctorates and Positions for U.S.-Trained Postdoctorates by Field, Alternative Scenarios, 2001–2011 Field and Year Postdocs in Workforce Projected Positions from Trend Over Medium High Low 5 years 20 years Biomedical 2001 12,711 12,711 12,711 11,655 11,655 2002 12,726 12,820 12,627 11,682 12,787 2003 12,819 13,025 12,613 11,473 13,186 2004 12,950 13,278 12,617 11,263 13,584 2005 13,214 13,686 12,739 11,054 13,982 2006 13,515 14,149 12,882 10,845 14,380 2007 13,916 14,721 13,111 10,636 14,779 2008 14,291 15,282 13,309 10,426 15,177 2009 14,695 15,871 13,524 10,217 15,575 2010 15,052 16,416 13,690 10,008 15,973 2011 15,392 16,948 13,843 9,799 16,371 Clinical 2001 507 507 507 415 415 2002 546 554 540 364 470 2003 585 601 570 336 487 2004 620 643 594 307 503 2005 659 694 623 278 520 2006 705 751 658 249 537 2007 750 808 690 221 554 2008 785 855 713 192 571 2009 813 895 727 163 588 2010 834 930 735 134 605 2011 859 967 744 106 622 Behavioral 2001 2,391 2,391 2,391 2,093 2,093 2002 2,527 2,543 2,512 1,907 2,157 2003 2,651 2,685 2,617 1,784 2,226 2004 2,746 2,800 2,693 1,662 2,296 2005 2,842 2,917 2,765 1,539 2,366 2006 2,905 3,002 2,806 1,417 2,435 2007 2,968 3,090 2,845 1,294 2,505 2008 3,021 3,167 2,873 1,172 2,575 2009 3,072 3,245 2,900 1,049 2,645 2010 3,109 3,308 2,912 927 2,714 2011 3,144 3,369 2,919 804 2,784   SOURCE: NRC analysis. to cover not only the workforce but also the positions that will become available. Positions are projected based on employment trends by sector and give results that appear comparable to occupational projections from the BLS. However, whether past employment trends will be sustained in the future is not known. Societal changes, such as the aging of the population, the increasing need for and complexity of health care, and the growing importance of science-based decision making could affect the demand for health research. Without attempting to assess such factors, these projections of positions rely on past sectoral trends—which, it should be noted, presumably also reflect substantial and sometimes unexpected societal changes. Comparing these projections with workforce projections gives some limited and tentative answers to the initial question. The answers differ by field. Biomedical researchers are probably being graduated in sufficient numbers for the next decade (2001–2011), though perhaps they will be a few percentage points too numerous. Clinical Ph.D.s, however, seem to be headed for a situation of substantial oversupply. This field includes substantial numbers of M.D.s, who are not projected, that category is smaller but growing more rapidly than the other fields, so the trajectory is somewhat uncertain. Finally, behavioral researchers are on a trend that will lead largely to balance until the end of the decade, at which time a slight deficit is possible. These statements refer only to U.S.-trained researchers. Foreign-trained researchers are an important and increasing part of the workforce in the biomedical and clinical fields. The research positions they fill are not distinguished from those taken by U.S.-trained researchers (but are not reflected in our sectoral employment data). Therefore the balance, or lack of balance, between the U.S.-trained workforce and positions filled by foreign-trained researchers presents a partial and somewhat artificial picture. The complications introduced by foreign-trained researchers are evident from the perspective of postdoctorates. U.S.-trained researchers on postdoctorates declined in the late 1990s, and if this trend continues, a substantial oversupply of those seeking postdoctorates may emerge. However,

OCR for page 120
Advancing the Nation’s Health Needs FIGURE D-19 Reported and projected U.S.-trained postdoctorates and postdoctorate positions, 1973–2011. SOURCES: National Science Foundation Survey of Doctorate Recipients; and NRC analysis. this decline may actually indicate a replacement of U.S.-trained with foreign-trained researchers, who are increasing rapidly in the postdoctorate pool in the biomedical and clinical fields. If this proves to be the case, the uncertain prospects for future migrants may substantially affect the prospects for bringing the numbers of U.S.-trained researchers seeking postdoctorates into balance with the numbers of positions that, from past trends, will be available to them.

OCR for page 120
Advancing the Nation’s Health Needs This page intentionally left blank.