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 121
BIOMEDICAL/BE~VIORAL COHORT MODEL: A TECHNICAL PAPER
Joe G. Baker*
INTRODUCTION
The legislation prompting this study requires the assessment
of ". . . the nation's overall need for biomedical and behavioral
research personnel. . . ."~ Past committees have defined this
"need" in labor market terms--that is, how many biomedical and
behavioral researchers will be "needed" in the future? The
purpose of this paper is to document the model used to project
the future labor market for biomedical and behavioral scientists.
Future labor market conditions are defined in terms of
improvements or deteriorations from historical market conditions,
and future demand conditions are examined in the context of the
appropriate National Research Service Awards Acts variable--that
is, Ph.D. production and postdoctoral study. Fo1 lowing a
descriptive overview of the model used are detailed discussions
of the various model components, data, and coefficient estimates.
THE MODEL: AN O1JERVTEW
Past committee projections of the future need for biomedical
and behavioral scientists have focused on academic demand. Job
openings were developed based upon growth in academic positions
and openings resulting from faculty death, retirement, and field
switching. These projections were developed for the near term;
the 1985 committee report included projections to 1990.
This model expands the earlier analysis in several ways:
1. In almost every biomedical and behavioral field, the
major source of historical and projected employment
growth is in nonacademic sectors, primarily private
industry. Thus, this model includes industry,
government, hospital, and other nonacademic sources of
demand for biomedical and behavioral scientists.
* The opinions expressed in this paper are the author's and do
not necessarily reflect those of either the Committee on
Biomedical Research Personnel or the National Research Council.
iSection 489 of P.L. 99-158.
OCR for page 122
2. The focus of the NRSA program is research personnel.
This analysis develops separate projections for the
labor market in general and that subsection of the
labor market associated with scientists whose primary
work activity is research and development (R&D) or the
management of R&D.
Given concern over "graying" of the work force, this
analysis includes a demographic/economic model for
estimating scientists' attrition due to death,
retirement, and net occupational movement based upon
the age and experience structure of the scientific work
force.
4. This analysis includes a model of labor supply.
5. Given that the median time to complete the biomedical
sciences Ph.D. has grown from seven years in the late
1970s to eight years in 1987, the 1997 biomedical
scientist labor market will be influenced by student
decisions and NRSA policy in 1989. In the behavioral
sciences, median time to Ph.D. has increased from
approximately 8.5 years to 10.5 years during the same
period. These types of lags argue for a longer time
horizon of analysis; the current study projects labor
market variables to the year 2000.
Figure 1 is a schematic of the labor market assessment
model. The stock of scientists in time period t is characterized
by biological age (years since birth) and career age (years since
degree). Historical data provide estimates of the number of
deaths and retirements by biological age; these scientists are
removed from the scientist stock. Those who do not retire or die
can leave the field for other employment; this is assumed to be a
function of career age. These estimates of outmigration are net
of immigration from other fields and are estimated from
historical data. The surviving scientist stock is available for
employment in period t+1. The required scientist stock in period
t+1 is estimated from submodels that link demand for scientists
to the demand for the good or service that scientists produce--
for example, R&D or graduate students. These demand submodels
vary by discipline (biomedical, behavioral, clinical), work
duties (R&D and non-R&D), and sector (academic, industrial,
governmental). The difference between the surviving scientist
stock and the required scientist stock in period t+1 is the job
openings (vacancies) that must be filled by new entrants. These
2See National Research Council, Summary Report 198 7
Doctorate Recipients from United States Universities (Washington,
D.C.: National Academy Press, 1989), Table P.
OCR for page 123
Scientists in time t
Characterized by
1. Biological Age
2. Career Age
Figure 1.
~1
Net Outmigration
dependent upon
Career Age ~ ~
"Surviving" Scientists
in time t+1
Scientists in time t+1
Characterized by
1. Biological Age
2. Career Age
Schematic
Demand for Scientists
1. R&D Spending
2. Enrollments
3. Other
Deaths & Retirement
dependent upon
Biological Age
New Hires by
1. Biological Age
2. Career Age
.F
"Required" Scientists
Stock t+1
1. Academic
2. Industry
3. Government
4. Hospitals
5. Other
New Hires=
"Required" -
"Surviving"
Supply of Scientists - `
( 1 Readiness
of the Labor Market Assessment Model
123
\~ 2. Recruitment )
,~ Retention
OCR for page 124
job vacancies are compared to supply to compute "vacancy ratios"
--that is, the number of vacancies per new Ph.D. or
postdoctorate. Increases or decreases in future vacancy ratios
from historical ratios give one a sense of tightening or
loosening of the projected demand/supply balance in the scientist
labor market. This same basic analysis is replicated for the R&D
subsector.
SCIE=IST ATTRITION:
DEATH, RETIREMENT, AND NET MOBILITY
The model used to estimate attrition is a demographic/
economic cohort survival model that follows closely the work of
Kuh, Radner, and Fernandez. 3 The nature of a cohort survival
model is to essentially "march" each cohort of scientists through
time and apply age-specific rates of death, retirement, and net
outmobility as the cohort ages. The population described here is
that of basic biomedical scientists. Separate coefficients were
estimated for behavioral scientists and are included in the
Appendix tables.
Three types of annual scientist attrition are addressed by
the model: death, retirement, and net outmobility. The annual
number of deaths and retirements are assumed to be a function of
an individual's "biological" age (B). Annual net outmobility is
defined as the number of scientists who leave biomedical science
for other occupations minus the number of workers who enter
biomedical science from other occupations. Net outmobility is
assumed to be a function of a scientist's "career" age (C). It
is unlikely, for example, to see a large net outmobility rate in
the younger career ages because scientists are unlikely to move
out of an occupation that has consumed an average of 8 years of
training. Later in their careers, a segment of the scientists
will more likely move out of science and into administration or
management.
In addition to career patterns, scientist mobility could be
influenced by labor market characteristics. For example, if the
labor market deteriorated and job opportunities and wage growth
were depressed, scientist outmobility would probably increase.
However, historical data indicate very little variation in
scientist unemployment and underemployment rates through time.
Also, it could be argued that scientists would be unlikely to
3Chariotte Kuh and Roy Radner, Mathematicians in Academia :
1975-2000 (Washington, D.C.: Conference Board on the
Mathematical Sciences, 1980~; and Luis Fernandez, Project on
Quantitative Policy Analysis Motels of Demand and Supply in
Higher Education (Washington, D.C.: Carnegie Council on Policy
Studies in Higher Education, 1978~.
124
OCR for page 125
leave a career in reaction to short term cyclical economic
conditions. In the mode J described here, these labor market
influences are assumed to be small compared to career patterns as
determinants of net mobility, i.e., they are ignored.
Each scientist in the model is defined by biological age and
career age. Although scientists can enter the system at any
career age, those who enter at career ages other than one are
incorporated into the net outmobility rates. For practical
purposes, all new entrants enter the system at career age one.
The parameters of the model include:
o Initial biological and career age distribution,
o Biological age-specific death and retirement rates, and
o Career age-specific net outmobility rates.
Obviously, there is a high degree of correlation between
biological age and career age. Because of this, one can estimate
fairly accurately the biological age distribution of a group of
scientists given their career ages. If one assumes that the
biological age distribution of newly minted Ph.D.s is stable, one
can construct a model whose states (biological age and career
age) depend only upon the career age distribution. At any time,
the biological age distribution can be estimated from the career
age distribution and the model can be simplified.4
Data
Except where noted, all data are from the Survey of
Doctorate Recipients (SDR) sponsored by the National Science
Foundation and the National Institutes of Health. These data are
collected by the National Research Council biannually from a
cross-section sample of approximately 10 percent of the U.S.
Ph.D. scientist population.
Initial Career and Biological Distribution
Table 1 is a cross-tabulation of the distribution of
biomedical scientists in 1987 by career age and biological age.
Given that a biomedical scientist is C<=5 ; then P(B<30 ~ = .115
These probabilities allow the estimation of biological age from
given career age.5
4See Fernandez, op. cit. , pp.132-33.
5Career and biological age cross-tabulations for behavioral
scientists are contained in the appendix.
125
OCR for page 126
Table 1. Career Age and Biological Age Distribution of Biomedical
Scientists, 1983-1987
Bio
Age =41 Total
<30 1918 21 0 0 0 0 0 0 0 1939
30-34 9463 4650 43 0 0 0 0 0 0 14156
35-39 3748 8547 4908 8 0 0 0 0 0 17211
40-44 977 3298 8618 5161 0 0 0 0 0 18054
45 -49 280 678 2535 6340 2357 57 0 0 0 12247
50-54 124 204 907 2084 2943 1151 0 0 0 7413
55-59 81 101 274 744 1330 2037 1228 77 0 5872
60-64 3 61 77 181 511 1048 1702 474 6 4063
65-69 10 0 16 84 91 95 376 566 219 1457
>=70 0 0 8 0 13 22 57 174 88 362
Total 16604 17560 17386 14602 7245 4410
Relative Distribution
Career Age
3363 1291
313 82774
Bio
Age
<=5 6-10 11-15
26-30 31-35
36-40 >=41
o
o
<30 0.1155 0.0011 0
30-34 0.5699 0.2648 0.0024
35-39 0.2257 0.4867 0.2822 0.0005 0 0
40-44 0.0588 0.1878 0.4956 0.3534 0 0
45-49 0.0168 0.0386 0.1458 0.4341 0.3253 0.0129 0 0 0
50-54 0.0074 0.0116 0.0521 0.1427 0.4062 0.2609 0 0 0
55-59 0.0048 0.0057 0.0157 0.0509 0.1835 0.4619 0.3651 0.0596 0
60-64 0.0001 0.0034 0.0044 0.0123 0.0705 0.2376 0.5060 0.3671 0.0191
65-69 0.0006 0 0.0009 0.0057 0.0125 0.0215 0.1118 0.4384 0.6996
>=70 0 0 0.0004 0 0.0017 0.0049 0.0169 0.1347 0.2811
O O
O O
Total 1 1 1 1 1 1
o
o
, O
o
O O
O O
O O
o
SOURCE: 1983, 1985 and 1987, Survey of Doctorate Recipients, NRC.
Table 2. Death and Retirement Rates
Bio
Age
o
' O
o
o
o
o
0.0026
0.0753
0.1714
1
Retire Death
Rate Rate
0.0011
0.0017
0.0027
0.0038
0.0063
0.0114
0.0179
0.0271
0.0378
0.05
SOURCE: Charlotte V. Kuh and Roy Radner,
Mathematicians i n Academia: 1975 - 2000
A Rew rt to the National Science Foundation,
Washington, D.C.: Conference Board of the
Mathematical Sciences, 1980, pp. 84-86.
126
OCR for page 127
Death and Retirement Rapes
Table 2 contains estimates of death and retirement rates.
These rates are for academicians in toto.
Net Mobility Rates
Conceptually, the net outmobility rates of scientists are
simply the gross occupational exits minus the gross occupational
entrants by career age. Practically, these longitudinal
movements are extremely difficult to measure given existing data
sets. Therefore, a shortcut method based upon a cross-section of
scientists by career age was used6 to estimate these movements.
As shown in Table 3 the values of each cell are the sum of SDR
survey years 1983 , 1985 , and 1987. For example, the number of
biomedical scientists of C<=5= 31,538: C<=5 for 1987 (10,823)
C<=5 for 1985 (10,720) + C<=5 for 1983 (9,995~. Putting aside
debate of whether postdoctorates are students or employed and
counting them as employed, the total pool of biomedical
scientists C<=5 is 48,847 (31,538 ~ 17,309~. The total number of
Ph.D.s in all fields except humanities and engineering with C<=5
is 202,517; this is assumed to be the primary supply source for
biomedical scientists. Because these estimates are for attrition
other than death and retirement, all scientists who will retire
must be removed from the labor supply pool (column 6, Table 3),
giving a labor supply pool of 202,378 for C<=5. Thus, for every
1,000 scientists of C<=5, 241 are biomedical scientists L(31538 +
17309~/~202378~. For every 1,000 scientists of Cow, 231 are
biomedical scientists. Assuming that this cross-section reflects
the longitudinal movement of scientists, then the net outmobility
rate of C<=6 is 4.1 percent r (241-231~/241~. This rate is for a
five-year period and must be annualized: assuming that the
average time of transition for a given scientist C<=6 to C6~0 is
2.5 years, then the annualized net outmobility rate is 1.79
percent (i.e., 98.21 percent of biomedical scientists of C<=6
"survive" into the next year, while 1.79 percent move into
nonbiomedical fields). 7
are biomedical scientists L(31538
1,000 scientists of Cow, 231 are
Mode} Structure
Given the above parameters, the operation of the model is
fairly straightforward. At a point in time i, a group of
scientists S are identified by their career age C. Death and
retirement are estimated from this group by the following
the
6This shortcut method was suggested by Robert Dauffenbach, a
member of the Committee on Biomedical and Behavioral Research
Personnel.
70f course, this survival rate is the dif ference between
1 arger gross outmobility and gross immobility
127
OCR for page 128
equation:
(1) PUDGY- E(n) P(Bn~Cr)*P(DRn~Bn)
Where
P(D~) -
P(Bn~Cr) =
the probability of death or retirement given
career age r;
the probability of biological age = n given career
age = r;
P(DRn|Bn) = the probability of death or retirement given
biological age = n;
and E(n) sums these probabilities across all biological ages n.
Given that an individual does not retire or die, he or she can
now leave the biomedical work force for other occupations. The
total attrition rate for career age = r is thus:
(2) P(~) = P(O~*~1-P(D~]
where
P (~)
P (O~)
conditional probability of total attrition given
career age r;
simple probability of net outmobility for career
age = r
C1-P(D~] = probability of not retiring or dying given career
age r;
. . . the total attrition estimate for all employed biomedical
scientists is:
(3) TA = E(r) Sr*P(~)
where
TA = total attrition from death, retirement, and net
outmobility and
the number of scientists career age r.
S =
Thus, for each group of scientists career age r, the estimated
attrition leaves a surviving group of scientists in period t+1
that moves to the next career age r+l. Summing across all career
ages gives the total attrition for the stock of scientists S and
allows for the computation of the total surviving scientists in
period t+1 (Figure 1~.
128
OCR for page 129
DEMAND FOR SCIENTISTS
As shown in-Figure 1, the demand for scientists is derived
from the services that they produce--for example, R&D and
teaching graduate students. Other factors, in particular wage
rates, can also affect the level of scientist demand. The demand
models used in this analysis--simple forms of this conceptual
models estimated for each field (biomedical and behavioral),
sector (academic, industrial, hospital,) and work activity (R&D
and non-R&D)--are included in the Appendix. Table 4 contains the
detailed projections for biomedical scientists; the Appendix
contains the detailed tables for nonclinical psychologists and
other behavioral scientists. The projection model assumptions
were developed by the Committee on Biomedical and Behavioral
Research Personnel.
Table 3. Estimated Quit Rates for Biomedical Scientists,
1983-1987
(1) (2) (3) (4) (5) (6) (7) (8) (9)
Biomed Total Bio Sci Annual Annual
Career R&D & Postdoc Total Phones per Quit R&D Quit
. Age Biomed Mgt R&D Biomed Retired Ph.D.s -retire 1000 Sci Rate Rate
<=5 31538 19695 17309 139 202517 202378 241
6-10 48171 2997S 2419 373 219649 219276 231 -1.79°X -8.18%
11-15 49162 26230 531 990 216355 215365 231 0.00% -6.69%
16-20 34190 16674 135 2279 156309 154030 223 -1.38% -5.06%
21-25 18257 8406 24 4041 90762 86721 211 -2.20% -4.52%
26-30 12726 5383 36 6428 66917 60489 211 0.03% -3~21%
31-35 8567 3790 24 13162 55630 42468 202 -1.6rX 0.10X
36-40 2743 1095 7 11772 26291 14519 189 -2.60% -6.51%
41+ 1295 495 0 6507 11286 4779 271 15.40% 13.24%
Total 206649 111743 20485 45691 1045716 1000025
Average annual quit rate
-0.8%
SOURCE: 1983, 1985, arid 1987 Survey of Doctorate Recipients, National Research Council.
Postdoctorates
There are no explicit projections for postdoctorates. The .
level of postdoctorate employment is assumed to be an
institutional variable determined by public policy and, as a
simplifying assumption, the level of postdoctorate employment was
8For a model of biomedical scientist demand that includes
factor prices in both the demand and supply equations, see Joe G.
Baker, "The Ph.D. Supply Crisis: A Look at the Biomedical
Sciences," paper given at the June 21, 1989, Western Economics
Association Meeting, Lake Tahoe, Nevada. It should be noted that
the inclusion of factor prices did not change the salient results
of the analysis.
129
OCR for page 130
held constant during the time frame of analysis. Thus, the level
of postdoctoral employment does not affect the level of total
employment; and given that most postdoctorates are young, very
little labor supply is assumed to be lost from death, retirement,
or outmobility.
If the postdoctorate "pool" is assumed constant with little
or no attrition, then one can further assume that the annual
exits from this pool equal entrants, and the net effect on labor
supply to positions outside the pool is small. If the pool is
changing in size, then entrants and exits will not be equal. As
can be seen in Figure 2, the postdoctoral pool has been flat
since 1981 at approximately 8,100 biomedical scientists. In the
behavioral sciences,
postdoctoral employment has
been flat since 1979 at
approximately 1,100
scientists.
Analysis: Vacancy Ratios
The model discussed above
estimates annual job openings
from death, retirement, net
outmobility, ant] growth. As a
means to summarize the
relationship between these job
openings and labor supply and
to provide information about
the future labor market, the
concept of "vacancy ratios"
was developed. Simply put,
the vacancy ratio is the
average number of job openings Figure 2. Postdoctoral
per new Ph.D. For the period Employment in Biomedical and
1983-1987, the number of job Behavioral Sciences, 1973-1987
openings from death,
retirement and growth averaged
4846 annually (Table 5~. The average number of new Ph.D.s in the
biomedical sciences produced annually for this same period was
3862. Thus, the "vacancy ratio" was 1.25 openings per each new
Ph.D. An "R&D vacancy ratio" is calculated by comparing R&D job
openings to postdoctorate production.
SOURCE: Survey of Doctorate
Recipients.
Obviously, not all new Ph.D.s in biomedical science go into
the biomedical field; also the field draws Ph.D.s from other
areas (e.g., physical sciences anti other life sciences) including
foreign scientists. However, the primary source of new Ph.D.s is
U.S. graduates in the field and, thus, the vacancy ratio gives
one a sense of the historical relationship between this supply
source and demand. No value judgments are made in terms of what
the "correct" vacancy ratio is ; the pro j ected vacancy ratios
130
OCR for page 131
s imply provide information about the relative state of the labor
market uncier differing assumptions.
Table 4. Summary Projections in Biomedical Sciences
Biomedical Sensitivity Model
Model Assumptions High Mid Low
1. Federal Health R&D Funding Growth 4.0%
2. Private Health R&D Funding Growth 13.0%
3. Other Health R&D Funding Growth 4.0%
Grad and Undergrad Biomed Enrollment 1.0%
5. "Other" blamed R&D Employment Growth 3.5%
6. "Other" biomed non-A&D Employment Growth 10.0%
2.7% 0.2%
9.0% 5.0%
3.0% 2.0%
0.0% - 1 .0%
2.5% 1.5%
8.0% 5.0%
Projected Employment of Biomedical Scientists, 1973-2000
(in 100s of workers)
Low Case
Year Total R&D % R&D
~-
Mid Case
Total R&D % R&D
-
High Case
Total R&D % R&D
1973 396.4 197.8 49.9% 396.4 197.8 49.9% 396.4 197.8 49.9X
1975 447.2 211.8 47.4% 447.2 211.8 47.4% 447.2 211.8 47.4%
1977 479.2 237.9 49.6% 479.2 237.9 49.6% 479.2 237.9 49.6%
1979 543.8 284.5 52.3% 543.8 284.5 52.3% 543.8 284.5 52.3%
1981 600.1 316.2 52.7% 600.1 316.2 52.7% 600.1 316.2 52.7%
1983 626.4 322.1 51 .4% 626.4 322. 1 51 .4% 626.4 322.1 51 .4%
1985 703.2 366.7 52.1% 703.2 366.7 52.1% 703.2 366.7 52.1%
1987 762.6 437.6 57.4% 762.6 437.6 57.4% 762.6 437.6 57.4%
1988 786.4 451.2 57.4% 802.4 464.0 57.8% 816.6 475.3 58.2%
1989 791 .6 459. 0 58. 0% 823 .3 484 . 1 58. 8% 851 .6 506. 5 59. 5%
1990 811.1 466.8 57.6% 858.7 504.2 58.rX 901.7 538.1 59.7%
1991 825.0 474.6 57.5% 888.8 524.4 59.0% 947.8 570.9 60.2%
1992 839.3 482.4 57.5% 919.9 545.2 59.3% 996.8 605.9 60.8%
1993 855 . 2 490.4 57. 3% 953 .6 566.9 59. 4% 1050. 7 643 .9 61 . 3%
1994 869.6 498.5 57.3% 987.2 589.8 59.7% 1107.5 685.7 61.9%
1995 886.5 506.8 57.2% 1024.6 614.1 59.9% 1171.5 732.1 62.5%
1996 894 . 9 515 . 4 57. 6% 1055 . 3 640 . 1 60 . 6% 1232 . 7 783 . 7 63 . 6%
1997 913.2 524.3 57.4% 1097.9 668.1 60.8% 1310.3 841.4 64.2%
1998 928.8 533.5 57.4% 1140.0 698.3 61.3% 1392.5 906.0 65.1%
1999 947.4 543.1 57.3% 1187.7 731.0 61 .6% 1486.0 978.4 65.8%
2000 973.6 553.1 56.8% 1245.6 766.5 61.5% 1596.4 1059.6 66.4%
Growth Rates:
73-87 4.8% 5.8% 4.8% 5.8% 4.8% 5.8%
87-91 2.0% 2.0% 3.9% 4.6% 5.6% 6.9°X
87-20 1.9% 1.8% 3.8% 4.4% 5.8% 7.0%
NOTE: This table does not include postdoctoral employment or unemployment.
SOURCE: Estimated by National Research Council.
131
OCR for page 144
Table A-7. Government Projection Models, Nonclinical Psychology
Employment Proiected
Non- Total Non- %
Year Total R&D R&D Gov R&D R&D R&D
Assumption
73 1083 426 657 1083 657 426 60.7
75 1170 345 825 1170 825 345 70.5
77 1404 495 909 1404 909 495 64.7
79 1164 334 830 1164 830 334 71.3
81 1235 487 748 1235 748 487 60.5
83 1320 670 650 1320 650 670 49.3
85 1189 511 678 1189 678 511 57.0
87 1724 711 1013 1724 1013 711 58.7
88 1741 1072 669 61.6
89 1759 1083 675 61.6
90 1776 1094 682 61.6
91 1794 1105 689 61.6
92 1812 1116 696 61.6
93 1830 1127 703 61.6
94 1848 1139 710 61.6
95 1867 1150 717 61.6
96 1884 1160 723 61.6
97 1901 1171 730 61.6
98 1918 1181 736 61.6
99 1935 1192 743 61.6
100 1952 1203 750 61.6
NOTE: The Government Projection Model is a simple trend model
that uses the BLS estimates of growth in total
psychologists (1 percent from 1987-1995; .09 percent
from 1995-2000). The average proportion of scientists
involved in R&D for the 1973-87 period (61.6 percent)
was assigned constant over the projection period.
144
OCR for page 145
Table A-8. All Other Employment, Nonclinical Psychology
Employment
Proiected
Non- Total Non- %
Year Total R&D R&D Other R&D R&D R&D
Assumption 3.4% -0.5% 7.2%
73 1043 407 636 1043 636 407 61.0
75 1159 520 639 1159 639 520 55.2
77 1221 684 537 1221 537 684 44.0
79 1409 833 576 1409 576 833 40.9
81 1433 875 558 1433 558 875 38.9
83 1413 1022 391 1413 391 1022 27.7
85 1497 1111 386 1497 386 1111 25.8
87 1662 1075 587 1662 587 1075 35.3
88 0.0338 0.0718 -0.005 1736 584 1152 33.6
89 1795 581 1235 32.4
90 1856 578 1324 31.2
91 1918 575 1419 30.0
92 1983 572 1521 28.9
93 2050 569 1630 27.8
94 2120 567 1747 26.7
95 2191 564 1872 25.7
96 2265 561 2007 24.8
97 2342 558 2151 23.8
98 2421 555 2305 22.9
99 2503 553 2471 22.1
100 2587 550 2648 21.2
NOTE: The "All Other" sector is a simple trend model that
projects assigned rates of growth for the R&D and
non-A&D sectors through the year 2000. The rates of
growth are the 1973-1987 rates for non-A&D (7.2 percent
per annex) and R&D (-0.5 percent per annex).
145
OCR for page 146
Table A-9. Academic Projection Models, Other Behavioral Sciences
Other Behavioral Sciences
Prolected
Employment Total
Non- Beh Sc Total Non- %
Year Total R&D R&D Fac. Fac. R&D R&D R&D
Assu I pt i on -2. 0%
73 6135 5201 934 19900 6135 5201 934 15.2
75 7621 6559 1062 23600 7621 6559 1062 13.9
77 9239 7784 1455 25600 9239 7784 1455 15 .8
79 9568 7915 1653 26900 9568 7915 1653 17.3
81 9477 8074 1403 28200 9477 8074 1403 14.8
83 1 1002 9546 1456 29800 1 1002 9546 1456 13.2
85 11009 9530 1479 31700 11009 9530 1479 13.4
87 10767 9079 1688 31800 10767 9079 1688 15.7
88 31732 10741 1602 9139 14.9
89 31635 10708 1596 9113 14.9
90 31444 10644 1586 9058 14 .9
91 31166 10550 1572 8978 14.9
92 30806 10428 1554 8874 14 .9
93 30369 10280 1532 8748 14 . 9
94 29860 10108 1506 8602 14.9
95 29283 9912 1477 8435 14 .9
96 28644 9696 1445 8251 14.9
97 27945 9459 1409 8050 14.9
98 27192 9204 1371 7833 14 .9
99 26388 8932 1331 7601 14.9
100 25536 8644 1288 7356 14.9
NOTE: The Academic Projection Model uses a quadratic regression of
total behavioral faculty = f (behavioral graduate students).
1. Growth assumptions in behavioral student enrollment
provided by the committee.
The portion of total behavioral faculty that is clinical
psychology, nonclinical psychology, and "Other Behavioral
Sciences" remains fixed at 1987 levels.
3. The portion of other behavioral sciences in R&D, which
has average 14.9 percent from 1973-1987, remains fixed.
146
OCR for page 147
Table A-10. All Other EmployTent, Other Behavioral Sciences
Employment
Non-
Year Total R&D R&D
Proi ected
Total Non- %
Ind R&D R&D R&D
Assurnpt i on 3.0% 0.07 0.1491
73 456 179 277 456 277 179 60.8
75 579 201 378 579 378 201 65.4
77 837 328 509 837 509 328 60.9
79 1277 450 827 1277 827 450 64.8
81 1343 524 819 1343 819 524 61.0
83 1476 1028 448 1476 448 1028 30.3
85 1680 1308 372 1680 372 1308 22.2
87 1969 1254 715 1969 715 1254 36.3
88 0.1101 0.1491 0.0700 2206 765 1441 34.7
89 2474 819 1656 33.1
90 2779 876 1903 31.5
91 3124 937 2186 30.0
92 3515 1003 2512 28.5
93 3960 1073 2887 27.1
94 4465 1148 3317 25.7
95 5040 1229 3812 24.4
96 5695 1315 4380 23.1
97 6440 1407 5033 21.8
98 7289 1505 5783 20.7
99 8256 1611 6646 19.5
100 9360 1723 7637 18.4
NOTE: The "All Other" projection model is a simple trend model that
grows 1987 ernployTent to the year 2000 assuming that 1973-1987
historical growth rates in the R&D (7 percent) and non-A&D
(14.9 percent) would continue through 2000.
147
OCR for page 148
Table A-11. Behavioral Enrollment Faculty, All Behavioral
Sciences
Assumption -2.0% -2.0%
Year
Est
Behav.
Undergrad
Enroll
Est
Behav.
Grad
Enroll
Estimated Total
Total Fac Student
1972
1973
1974
1975
1976
1977
1978
1979
1980
1981
1982
1983
1984
1985
1986
1987
1988
1989
1990
1991
1992
1993
1994
1995
1996
1997
1998
1999
2000
705.8
695.6
672.6
668.1
671.8
663.5
648.9
612.8
604.8
593.6
589.8
574.5
563.0
551.7
540.7
529.9
519.3
508.9
498.7
488.8
479.0
469.4
460.0
450.8
441.8
433.0
424.3
415.8
407.5
44.7
48.2
51.7
55.4
59.1
58.3
63.8
63.8
63.8
64.8
64.3
63.1
63.7
65.6
65.3
65.1
63.8
62.5
61.3
60.0
58.8
57.7
56.5
55.4
54.3
53.2
52.1
51.1
50.1
19.9
21.7
23.6
24.6
25.6
26.2
26.9
27.5
28.2
29.0
29.8
30.7
31.7
31.7
31.8
31.7
31.6
31.4
31.2
30.8
30.4
29.9
29.3
28.6
27.9
27.2
26.4
25.5
750.5
743.8
724.3
723.5
no.s
721.8
712.7
676.6
668.6
658.4
654~1
637.6
626.7
617.3
606~0
595.0
583~1
571.4
560.0
548.8
537.8
527.1
516.5
506.2
496.1
486.2
476.4
466.9
457.6
.
NOTE: Growth assumptions regarding enrollments in
behavioral sciences were provided by the
corr~nittee.
Faculty = f (behavioral graduate students)
Regression Output:
Constant
Std Err of Y Est
R Squared
NOa of Observations
Degrees of Freedom
X Coefficient
Standard Error
of Coefficient
3.5676722079
0.0207556027
0.9032859059
15
12
0.0001001 -0.0000000014
0.0001713 0.0000000088
148
OCR for page 149
Table A-12. Summary Projections for Nonclinical Psychology
Nonclinical Sensitivity Model
Model Assumptions (in percent)
1. Graduate Student Enrollment
2. Industrial Employment Growth
High Mid Low
1.0 0.0 -1~0
4.0 3.0 2.0
Projected Employment of Nonclinical Psychologists, 1973-2000
(in 100s of workers)
Low Case
Mid Case
Year Total R&D % R&D Total R&D % R&D
High Case
Total R&D % R&D
1973 129~0 38~8 30~0 129~0 38~8 30~0 129~0 38~8 30~0
1975 148~1 39~5 26~6 148~1 39~5 26~6 148~1 39~5 26~6
1977 153~2 43~1 28~1 153~2 43~1 28~1 153~2 43~1 28~1
1979 157~9 49~4 31~3 157~9 49~4 31~3 157~9 49~4 31~3
1981 179~9 50~2 27~9 179~9 50~2 27a9 179~9 50~2 27~9
1983 187~2 47~7 25~5 187~2 47~7 25~5 187~2 47~7 25~5
1985 191~0 46~8 24~5 191~0 46~8 24~5 191~0 46~8 24~5
1987 198~4 55~7 28~1 198~4 55~7 28~1 198~4 55~7 28~1
1988 200 ~ 0 54 ~ 9 27.5 200 ~ 3 55 ~ 0 27e 5 200 ~ 7 55 ~ 2 27a 5
1989 195~5 54~0 27~6 202~2 55~5 27~4 204~9 56~1 27~4
1990 195~2 53~9 27~6 204~0 55~9 27~4 209~2 57.1 27~3
1991 194~9 53~9 27~6 206~0 56~3 27~4 213~6 58~1 27~2
1992 194~8 53~8 27~6 207~9 56~8 27~3 218~2 59~2 27~1
1993 194~7 53~8 27~6 210~0 57~2 27~3 223~0 60~2 27~0
1994 194~7 53~8 27~6 212~1 57~7 27~2 227~9 61~3 26~9
1995 194~8 53~7 27~6 214~2 58~2 27~2 232~9 62~5 26~8
1996 194~9 53~7 27~6 216~4 58~7 27~1 238~1 63~6 26~7
1997 195~1 53~7 27~5 218~7 59~2 27~0 243~4 64~8 26~6
1998 195~4 53~8 27~5 221~0 59~7 27~0 249~0 66~1 26~5
1999 195~8 53~8 27~5 223~4 60~2 26~9 254~7 67~4 26~5
2000 199 ~ 2 54 ~ 5 27e 3 225 ~ 9 60 ~ 7 26 ~ 9 260 ~ 6 68 ~ 7 26 ~ 4
Growth Rates (in percent):
1973-1987 3~1 2~6 3~1 2~6 3~1 2~6
1987-1991 ~0~4 ~0~9 0~9 0~3 1~9 1~0
1987-2000 0~0 -0~2 1.0 0~7 2~1 1~6
NOTE: This table does not include postdoctoral employment or unemployment.
SOURCE: Estimated by National Research Council.
149
OCR for page 150
o
o s
o 4J
~J 3
1 0
0
C~
C~ S
0` ~
3
1 0
r~
0
0`
I"1
Ql
"1
i Q
I CY
o 0 m1
o o
o o
0
o~ ~
o
o
4~
a,
~n
cn
JJ _
cn cn
o
~n
o ~ o~
o~
o
s
~ ~n
cn ~
~ o
L) 3
._ ~
._
-
~ 1
o
~ ~ ~ iff! ~ ~ ~ ~ ~ ~ ~
0` ~) ~ ~ U' ~ 1~ ~ la~ ~o
· a · · · a · · · · ~
~ O O ~ ~ ~ ~ O O O ~
1 1 1 1
~a~ O U' O O O O ~ O O
· a · e · · · · · · ~
O ~ t~J ~a~ ~ ~ ~ O ~ t~l
1
<) 0 ~ 1 ` 0 0 t~l ~ 0` ~a~1 0
· a a a a a · a a a a
`0 1` 0 ~ O O O
1 1 1 1 1
0% 0 ~
a a a
O
1
O O O O ~ ~ C~ ~
a a a · · ~ ·
~a`~ O O
1
~a~ ~ ~J ~ 0
· · a · · a a · · a
a~~
~J t~ ~J ~ 1~ ~ <) ~J ~J N t~J
~J `0 1^ 0 `0 1` 0
· a a a a a a
'~, la~_ a~
- 1 - 1 ~1 '
0 ~ 0 0` `0 U~
a a a a a · ~
~ O 00 ~ 0% ~ 0
O ~ Lr, ~ ~ U'
U~ ~ ~ ~
· a · ~
0 0
Ut `0 `0
0~
a
U~
~ 0% %0
a a a
~ O
0~ ~ ~
I~ ~ ~a'~ ~ ~a~ `0 0 `°
°iS a a a a a a a a
cy ~ ~ ~ ~ ~ ~ ~ O |a~ b_ la~
~ ~ ~ ea~] ~ ~a~ ~ ~ ~ ~J
u~ ~0 0~ ln 0 I~ ~ 0 C~ ~a~ ~
C) a a a a a a · · · a a
I~ 0 ~ ~ ~ ~ U~ ~ `0 0
~ ~ ~ ~ ~ ~ ~ ~ ~ U' ~
I_ 0 `0 1a~ 0 ~1 0 ~ ~
a a a a · a · a
0 0~ 1a~ o~
0~ 0 `0
a a a
~a~
0` 0 ~
~ O
L) 3 _
._ ~ .
._ ~
_ a ~
~ C~ %0 `0 ~0 ~ ~ ~ ~ ~ ~ ~ ~ ~ O
t~ 08 a a a a a · · · · · a G~ C_
O CY ~a~ ~a~~ ~ ~80 ~0 ~) 0 ~ 0 0 0 ~
Z S~ ~ ~~ ~~ ~~~ ~~~ <~J ~~ ~a~ ~ ~ ~J O S
O ~ 0~ 0~ ~ 0~ 0~ ~ 0~ ~_ ~_ 1a~ G) ~
C.O 8a`_ C) a a a a a a a a a · ·
~ 0 ~ O O 0 0 0 0 0 U' U~ ~ U~ ~ ~
O 0` O! ~ ~ ~ ~ U' 1^ 1^ tY
._ ~ ~
4J ~ _
a~ _ ~ o 0 0 0 ~ ~o ~ ~ ~ ~
'_ ~ · a a · a · a a e · a ~ O
O 4~ O O O ~ ~ ~ ~ %0 0 0 0 ~ ~
O ~ ~ ~ ~1 ~ ~ ~ ~ 0~ ~ J~
O Z
JJ
~ · 0 n
~ ~ JJ cn
._ ~ cn ~
4~ ~ Q O
~ . cn ~
_# ~ . ~ a,
a' a) ~ ~ e_
U) C~~ Cl) ~ O
a ~ 4) ~ 4) {~ ~ O ~ ~
t~l (1) C~ C'/) a) C-) ~ ~ ~ c~ Cl)
cn ~ ~ ~n ~ ~ ~ ~ ~ x
E ~ ~ ~ E c~ ~ ~ ~ E c~ tu
~ :, ~ ~ ~ E s ~ ~ aa
E ~s ~ ~s ~ ~ ~ s UJ
~ O G) 3 ~ ~) cn 3 ~ ~ ~ O _ 3 ~ ~ · a O
_ ~ ~ 0 a~ ·_ ~ 0 a~ c_ ~v to 0 a~ ~ L~l
n ~ ~ _ ~ s ~ ~ ~: s ~ _ ~ _ ~ s ~ ~
a, ~ ~ 0 0 0 0
cn ~ ~ ~ - : ~ z cn
150
OCR for page 151
Table A-13. Estimated Quit Rates for All Behavioral Scientists, 1983-1987
(1) (2) (3) (4) (5) (6) (7) (8) (9)
Behav Total Beh Sci Annual Annual
Career R&D & Postdoc Total Ph.D.s per Quit R&D Quit
Age Behav Mgt R&D _ Behav Retired Ph.D.s -Retire 1000 Sci Rate Rate
<=5 41948 5607 1834 139 201127 200988 218
6-10 43488 6282 343 373 217900 217527 201 -3.1% -7.5X
11-15 36561 4612 127 990 215219 214229 171 -6.3% -12.0%
16-20 23923 2915 68 2279 156711 154432 155 -3.8% -5.3XO
21-25 13410 1463 51 4041 91497 87456 154 -0.4% -4.3%
26-30 10323 1271 38 6428 67207 60779 170 4.2% 9.1%
31-35 6155 950 8 13162 55931 42769 144 -6.5% 1.6%
36-40 1456 261 0 11772 26355 14583 100 -13.rX -8.6%
41+ 303 43 0 6507 11240 4733 64 -16.3% -23.8%
SOURCE: 1983, 1985, and 1987 Survey of Doctorate Recipients, National Research Council.
Table A-14. Historical and Projected Vacancy Ratios, Nonclinical Psychology,
1973-2000
Annual Averages
Total
Year Vacancies Ph.D.s Ratio
Vacancies Ph.D.s
R&D
Post -
Vacancies does Ratio
1973-78 1189 1592 0.75 385 196 1.96
1978-83 1386 1555 0.89 359 226 1.59
1983-87 1291 1435 0.90 422 261 1.61
1990-95
Low 995 1260 0.79 320 300 1.07
Mid 1252 1366 0.92 380 300 1.27
High 1551 1479 1.05 449 300 1.50
NOTE: Assumes that Ph.D. production changes in proportion to enrollment
assumptions and postdoctoral production remains constant.
SOURCE: National Research Council.
1 5
OCR for page 152
Table A-15. Summary Projections for Other Behavioral Scientists
Model Assumptions High Mid Low
1. Graduate Student Enrollment 1.0% 0.0X -1.0%
2. "All other" Employment Growth
R&D 7.0% 5.0% 3.0%
NonR&D 0.15 0.1 0.07
Projected Employment of Other Behavioral Scientists, 1973-2000
(in 100s of workers) (Grad enrollment model)
Low Case
Year Total R&D % R&D
Mid Case
Total R&D % R&D
High Case
Total R&D % R&D
1973 65.9 12.1 18.4 65.9 12.1 18.4 65.9 12.1 18.4
1975 82.0 14.4 17.6 82.0 14.4 17.6 82.0 14.4 17.6
1977 100.8 19.6 19.5 100.8 19.6 19.5 100.8 19.6 19.5
1979 108.5 24.8 22.9 108.5 24.8 22.9 108.5 24.8 22.9
1981 108.2 22.2 20.5 108.2 22.2 20.5 108.2 22.2 20.5
1983 124.8 19.0 15.3 124.8 19.0 15.3 124.8 19.0 15.3
1985 126.9 18.5 14.6 126.9 18.5 14.6 126.9 18.5 14.6
1987 127.4 24.0 18.9 127.4 24.0 18.9 127.4 24.0 18.9
1988 126.8 23.2 18.3 129.0 23.6 18.3 131.4 24.0 18.2
1989 126.4 23.2 18.3 130.7 23.9 18.3 135.8 24.7 18.2
1990 126.1 23.1 18.4 132.6 24.3 18.3 140.6 25.6 18.2
1991 125.9 23.2 18.4 134.7 24.7 18.4 145.9 26.4 18.1
1992 125.8 23.2 18.4 137.0 25.2 18.4 151.6 27.4 18.1
1993 125.9 23.2 18.4 139.5 25.6 18.4 158.0 28.4 17.9
1994 126.0 23.3 18.5 142.2 26.1 18.4 165.0 29.4 17.8
1995 126.4 23.3 18.5 145.1 26.6 18.3 172.8 30.5 17.6
1996 126.8 23.4 18.5 148.3 27.1 18.3 181.5 31.7 17.4
1997 127.4 23.5 18.4 151.8 27.7 18.2 191.1 32.9 17.2
1998 128.2 23.6 18.4 155.7 28.3 18.2 201.9 34.2 16.9
1999 129.1 23.7 18.4 159.9 28.9 18.1 213.9 35.6 16.6
2000 130.2 23.8 18.3 164.4 29.5 18.0 227.4 37.1 16.3
Growth Rates:
73-87 4.8% 5.0°/O 4.8% 5.0% 4.8% 5.~%
87-91 -0.3% -0.9% 1.4% o.rx 3.4% 2.4%
87-;20 0.2% -0.1% 2.0% 1.6% 4.6% 3.4%
NOTE: This table does not include postdoctoral employment or unernployrnent.
SOURCE: Estimated by National Research Council.
152
OCR for page 153
to
to
to
J 3
1 0
0
0
I~ ~ ~ ~ ~ sat
C) 0 ~ ~ o o o ~ lo
o~t · a a a a a a · ~
CY ~ O ~ ~ U. ~ O ~ r~
_ ~ ~ ~ ~ ~ ~ ~ ~ X
~ ~ 0 `0 ~ U~ 0 ~ O %0
~ · · · e a ~ a a ~
O ~ O ~ U' 0 ~ O
1~ ~ ~
S~ X
r~ ~ 0 0 0 o, r_
~ 08 · a a · · · · · a
=: ~ ~ O ~ 1^ 1~- 0 0 t~l
0
~ 3
J I o
0 C~ _
0` ~ U. O `0 `0
··· ··· ··~
O - o ~ ~ 0 ~ O
~ ~ ~ 1
o
O C] ~ 0% 0~ 0 0 ~ ~1 0 r~
O ~ ... ... ...
~ CY ~ ~ ~ U' ~ 0 0 0 ~
o0`
0 u~ u' ~ 0 IO ~
O ~ ... ... ...
r~ ~ 0 ~ ~ ~ C~ ~
O
O
4
~ _ ~ ~ O ~ 0 ~ ~ ~ ~
{D ~ · · · · a · a a a
~ 0~ ~ ~ O ~0 ~ O '
>. O 0 0 tO -t U-\ 0~ ~' ' ~ ~
' .~
cn
4
B_
O C) 0` 0~ C~ ~ ~ 0` ~ ~ ~
t~ 0 °is a a ~ ~ ~ a a a a
~ ~ ~ ~ ~ ~ ~ ~ 0` 0 ~ 0
c_ .
~_
~ u) ~ O ~ 0~ - ~ ` -
t~ ~ ~ a a a ~ a ~ a a ~
~ a, o~ ~ u' ~o ~ 0 0 0~ ~ ~ `0
O ~ ~ ~ - ~ ~J ~I N
O
3 a
_# _ ~ ~ %0 U~ ~ ~ 0K ~ 0~ _
~) ~ ~ ~ ~ a ~ a a a ~ c_
m 4 ~ ~ ~ ~ ~ ~ u' ~ u~
0 0 0 . ~ ~ r~ ~ r~
. ~ ·
a) ~ 0
~ a
JJ
0
1 ~ ~ ~ ~ ~ ~ o~ o~ t).
cY u~ u~ u~ ~0 ~0 %0 0 0 0 a~ a
o~
._ ~
o~ ~ o~ ~ ~ ~ 0 0 0 2~ _
|~~ ~ a a
a, 0 06 ~ ~o ~o ~ ~ h_ ~ ~ ~ ~
._ 0~ ~ ~ ~ ~ ~ ~ ~ ~ O
O ~ _ e_
Q _ t_ ~ ~ ~ ~ ~ ~ ~ ~ O
... ... ...
4- ~ ~ r~ o~
g' O O O O ~ ~ ~ - 1 ~J ~J O Q
:' 1~ ~ ~ ~ ~ ~ ~ ~
~ ~ cn
~ ~n ~
~ ~ ._
8 ~ ~
_ ~ ~
a' a, ~ ~ ~
cn cn cn ~ O
a~ ~ a) a~ a~ ~
41) C~ ~ 0) ~ cl~ 4) (~ tO O Q
Cl) tlJ ~ ~ ~ ~ ~ X
. .O ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ..
~ 0 G, 3 ~ ~ 0 3 ~ C~ 3 ~ ~ ·e C~
_ ~ {~ 0 a~ a_ 0 a~ ~ a~ 0 6)~ llJ
Q ~ ~S J ~ ~ = J ~ O O ~
~ CD eS ~C ~ Z CO
153
OCR for page 154
Table A-16. Historical and Projected Vacancy Ratios, Other Behavioral
Sciences, 1973-2000
Annual Averages
Post-
Year Vacancies PhDs Ratio Vacancies does Ratio
1973-78 1152 1211 0.95 289 97 2.97
1978-83 930 1093 0.85 120 124 0.97
1983-87 875 943 0.93 185 120 1.54
1987-95
Low 649 814 0.80 142 120 1.18
Mid 927 882 1.05 191 120 1.59
High 1353 955 1.42 250 120 2.08
NOTE: Assumes that Ph.D. production changes in proportion to enrollment
assumptions and postdoctoral production remains constant. Grad model.
SOURCE: National Research Council.
154
Representative terms from entire chapter:
labor market
