PERSONNEL ACCESSIONS (U)

• Document Type: CREST
• Collection: General CIA Records
• Document Number (FOIA) /ESDN (CREST): CIA-RDP81-00142R000400030027-9
• Release Decision: RIPPUB
• Original Classification: S
• Document Page Count: 18
• Document Creation Date: December 9, 2016
• Document Release Date: April 5, 2001
• Sequence Number: 27
• Publication Date: December 18, 1978
• Content Type: MF

Approved For Release'2001/07/12 : 0}L1 -00142R00040003m9 -
SlBJEC `
Director of Central Intel'
F. 4. M.Ja y
Director Ptrsowr-l
Personnel Accessiom, (U)
Mmo for Df s fr CI 4t 4
PF-rsa reel
1. Action Ptwmsted: ~ _ This .
i
s
t ut, particularly in the 7LO. ii)
2. ~Txvd:
a. Each J)iree rate renartA its ,,rs
i
s mac
sions, tit a fiscal yr-ar basis t 2 P
`
sgnel Ply AJ
MW 197'
of i s t r f-
~t to VM Ma on22 k -^1
s ~r of September 1974. T ruts reflect .
breaks 1.
do this data into s: i T ? ,cal sn dClei and
b. The Off ice of Personnel has *-mio .a- and used stetis I
dell. te6niques the debt of t `ms's 1 el r td r ,
is. The process that his beexi demionftj fnr t ) Is serf bye
1 h
bii
d
e
t
ng paragraphs (U)
, 3. Professional A .ssicsr s into the
s. The DID st elssion for ti Y ' o7n e~,,A1 n----.t
f
pro
essional zccmsims (interim Srxl t xte 1) of 78;
40 are Av- t C
Y
areer
ni ng 1-Mrs t ltt if
i .aer trainees c leti last year's profit suie ; nn t~k
the Directorate. (S)
h. I oftssi el. seParat,ioss fi ~ he P C er Service are
currently at an annual rata of a ?0. 11?is is a T ` '
outflow from llp l.sr !+ts p i l talatt ee shoe s 4
Approved For Release 2001/07/1 k P81-00142R000460030-0'27-9 ` ` '
 Approved For Release 2001/07/12 : CIA-RDP81-00142R00040003,,0027-9
c. The Directorate is aware o.' the desirability of ac .:w:ir
of 120 career trainees plus specialists i-nd conversions. InTc~n'
previous years, when more than half the.1ra#essianal input came r-i #
internal conversions (from technicians ard clericals), the Director-ox
places primary emphasis upon the Career ' raise a Program. (S)
d. The Office of Personnel ia~ studie in 1974, 1975, 1.j r:?
-
has lied simple models to a and r
- ---- -
applied age grace flows over five year I)er}o,
.
It applied an early Systems Dynamic model to the problem in 1-075. 4/11
e. An irportant conclusion min these shies was the
bility of aiming at a long-term entry 1ev31 input averaging 120 ~r
early year attrition, a net input of 500. This size of class over -9v-t
years compares favorably with the age di.stributien of the Career S'
which averages 500 in a five-year block, )lus or r nus 200 acco
past periods of increasing or decreasing
inputs. Such entry level
would be aupnented.by conversions and aec~ssion of specialists.
f. Another way to cross-check wring plates is to apply :.r
attrition-type of formula. The key to this approach is to derive ir,
appropriate long-term average rate of attrition based upon. career
terns in a particular Directorate. The attrition rates are partl c :R ar
term average rates may differ from current: rates, which are raise
_
substantial age hump in the DIX) and. OR grid lows ed by the lengthmied
retirement age permitted under CSC retir lent prevalent in the POS61- ai d
NFAC. Given the appropriate attrition fi:rure, the "steady flow" rite f
accessions would be found by the equation
Input - (Target level) tip*,. (Attrition rate)
lire sophisticated models make allowance or the impact of difffererlce
in age structure upon attrition rates, v ch are very sensitive to t%-1
measure of years of service, which in tur correlates highly with g. Table 1 displays data (on a Directorate basis) for tio
target 1979 level of professionals, the re?placei t inputs for a1t? A it
specified levels of attrition (long-term) and the snualiaed astir arc
CY 1978 separations (based on 11 months). Table 4-1 compares the, I 111
Annual Personnel Plan data on profession: inputs with the caappar4
long-term levels derived from Table 1. As is e3gz cted, the MO nTlfta;
are low but the total for all Directorate- is quite in line. NPAC arul
DDS&T are clearly in a strength-building 1, osture. (S)
Approved For Release 2001/07/12: CIA Ot- 142R000400030027-9
 Approved For Release- 2001/07/12 'C kE !r1-00142R0004( 030027-9
4. Follow:
a. I Agree with your craaa t x n is the passible a lic o
bility of models to this area. Attached are two 1976 articles prepare
by personnelists detailing early work ri such applics items. 00
b. We shall date the Systems Dynan is tel. for DOD profess
to see what additional insights it reveals. (U)
Janney
Attac n ts:
As Stated
Distribution:
Orig - Addressee
1 - DDCI
1-ER
2-DfA
STATINTL 2 - D/Pers (I OP DCI File)
1 - OP/Ps
OP/P4C/ Cnc (13 Dec 78)
RETYPED: - OD/Pers;ri (15 Dec 78)
Approved For Release 2001/07/12 : CIA-ROP81-00142R000400030027-9
 Approved For Release 2001/07/12 : CIA-RDP81-00142R000400030027-9
Approved For Release 2001/07/12 : CIA-RDP81-00142R000400030027-9
 Approved For Release 2001/07/12 : CIA-RDP81-00142R000400030027-9
Next 1 Page(s) In Document Exempt
Approved For Release 2001/07/12 : CIA-RDP81-00142R000400030027-9
 Approved For Release 2001/07/12 : CIA-RDP81-00142R000400030027-9
Approved For Release 2001/07/12 : CIA-RDP81-00142R000400030027-9
 Approved For Release 601/07/12: CIA-RDP81-00142R00040000027-9
IS-January 1976
MEMORANDUM FOR THE RECORD
SUBJECT : Description of Age Distribution Model
REFERENCE: Memo for the Record dtd 17 Nov 75, subj: Age Distribution,
Projection Model
1. The age distribution model is based on the system dynamics
approach; that is, a focus on the structure and behavior of systems
composed of interacting feed-back loops. System dynamics utilizes
the DYNAMJ pro aging languTage, flow diagraming and causal-loop
diagraming to define and simulate t~ie`n,odel of the system un er
consideration. The use of DYNAMO flow diagraming and causal-loop
diagraming provides a convenient way to complete y descr- i~ie-our
model before introducing the system equations.
2. The causal-loop diagram for our age distribution model
identifies the principal feedback loops without describing the inter-
connection process of the variables. In Figure A a causal-loop diayrtrn
describes the basic assumptions of the model: the number of persons -Ti
an age group is dependent on the plus side by hires, conversions-in, and
aging-in. On the negative side the impact is brought about by sepa-ralions,
conversion-out, and aging-out.
3. Definitions of the variables are as follows:
Level in Age Group (L) - Total number of persons in an age
group. For the purpose of tTie model these will be professionals on]y_
Hires (E) - Direct hires from outside of Agency.
Conversion=In (E) - Persons who through change of status er to r?
age groupTe.g,, clerical to professional).
Separatees (S) - Separation from the Agency.
Conversion -Out-- - Opposite process of conversion-in.
Age-In (U) - Persons from a younger age group who through the
aging process are increasing the total number of persons in the as?e 'Pr-nip.
Approved For Release 2001/07/12 : CIA-RDP81-00142R000400030027-9
 Approved For Release-2001/07/12 : CIA-RDP81-00142R00040QO30027-9
e-Out (U - Persons who are.- aging out of the age group.
Note that the same variable has been used for both Age-In and Age-Out.
This is to imply a movement up and out of the system. If mirror inakos
of the causal-loop diagram are generated, (see Figure B) then the Age..
Out for a previous level becomes the Age-In for the current level. In
the actual model subscripts are used to distinguish between the Ulf. for
level I or level II or level III.
4. The next step in the process is to define the DYNAMO flow
diagram. The flow diagram in Figure C consists of two levels and seven
rates. Levels may be thought of as holding tanks and rates as the
controling valves on pipes (solid lines) between the tanks. Arrows
indicate the direction of flow through the pipes. The "clouds" in the
diagram represent sources or sinks depending on if something is entering
or leaving the system respectively. Dotted lines represent flows c.f
information, and in the case of this model information on how many persons
are currently in the level is used in conjunction with a constant to
calculate the rate. For example, if the level represents a five-year
age group, then the rate (yearly) at which persons age-out of that gr3up
would be a constant one-fifth (20%) times the number of persons in that
age group. That is to say the valve (rate) on the holding tank (level)
would be opened so that at any given time the flow through the valve
would represent one-fifth (at a yearly rate) of the level in the tank.
5. The next step in the process is to replicate Figure C o thai. it
will encompass ten levels and thirty rates. These rates and levels
correspond to five-year age groups covering 19 under to 60-64. FigurS
D represents a portion of this replication with the variable names co.i-
sist with the equation names.
6. Generating equations follows directly from Figure D. in
Figure E the equations for two levels are given. The equations for tre
other levels are the same with the appropriate changes to the ":;ubscr;pts"
of the variables. Supplementary equations are used to calculate level
totals, average, age, and the number of EOD's which occur below age 31.
7. Output from the model. is specified in the usual way througi he
use of print and plot cards. Runs were usually limited to 10 periods
with period 0 representing the base period and period 5 representing --he
live-year projection.
8. Input to the model was in the form of constants, thus allo'aii:
multiple runs to be made with new vectors for each run. This enabled
a single copy of the model to represent any desired directorate without
major brain surgery to the model. It also made the model more comp-icl
by eliminating multiple copies and made the input data easier to check - -
it was available in a single vector rather than scattered through the
model.
Approved For Release 2001/07/12 : CIA-RDP81-00142R000400030027-9
 ' Approved For Releas2001/07/12 : CIA-RDP81-00142R0004( 130027-9
9. The vectors were designed to represent two situations: that
of the continued hiring of young professionals, and the case where no
young professionals under the age of 34 would be hired. As DYNAj90 doe,'
not allow the selective carrying of con,tants from one run to the next.
it was necessary to replicate the vectors twice; with one vector
representing continued hiring and the other with no hiring. This of
course presents a possible modification to the model: the use of a
switch to cut off hiring on a second run. This modification was not
made in the course of our work, but should the requirement for this
kind of model come up again the introduction of a switch will be made.
10. Other possible modifications to the model which are czar renal
being looked at include random noise variations in the rate values zinc
the introduction of switches to represent changes in management pol:_ci!s.
STATI
Plans Staff
Office of Personnel
Approved For Release 2001/07/12 : CIA-RDP81-00142R000400030027-9
 Approved For Release,2A01/07/12 : CIA-RDP81-00142R00040(0027-9
FIGURE A
Hires
Conversions
In
(E)
LEVEL IN AGE GROUP (L)
Separatees
Conversions
Out
(S)
FIGURE B
LEVEL IN AGE GROUP I
LEVEL IN AGE GROUP II
LEVEL IN AGE G1ZOUP III
IJI)
Age (Ju t
Av,e - In
12
\c~e -O t
Ave
(U)
1)3)
Out
Approved For Release 2001/07/12 : CIA-RDP81-00142R000400030027-9
 Approved For Release?2001/07/12: CIA-RDP81-00142R00040Q 30027-9
FIGURE C
Accession
Constant
Age-Out
Constant
S
Accession
Constant
Age-Out
Constant
Separation
Constant
I r
Separation
Constant
Approved For Release 2001/07/12 : CIA-RDP81-00142R000400030027-9
 Approved For Release 2601/07/12 : CIA-RDP81-00142R00040OW027-9
FIGURE D I
U19UC
N
E19UC
C>
E19U
U5559C
Approved For Release 2001/07/12 : CIA-RDP81-00142R000400030027-9
 Approved For Release-12001/07/12 : CIA-RDP81-00142R000400 027-9
FI GURE E
L
N
C
R
C
R
C
R
C
L5559.K=L5559.J+DT*(-S5559.JK,U5559.JK+E5559.JK+U5054.JI:)
L5559=L5559C
L5559C=O
S5559.KL=S5559C*L5559.K
S5559C=0
U5559.KL=U5559C*L5S59.K
U5559C=.20
E5S59.KL=E5559C*L5559.K
E5559C=0.0
NOTE
NOTE
L
L5054.K=L5054.J+DT*(-S5054..TK-U5054.JK+E5054.JK+U4549.JIt)
N
L5054=LSOS4C
C
L5054C=O
R
S5054.KL=SS054C*L5054.K
C
S5054C=O
R
U5054.KL=U5054C*L5054.K
C
U5054C=.20
R
E5054.KL=E5054C*L5054.K
C
E5054C=0.0
Approved For Release 2001/07/12 : CIA-RDP81-00142R000400030027-9
 Approved For Release 2001/07/12 : CIA-RDP81-00142R000400030027-9
apri 11976
Approved For Release 2001/07/12 : CIA-RDP81-00142R000400030027-9
 Approved For Release 2001/07/12 : CIA-RDP81-00142R000400030027-9
STATINTL
The current terrorist threat appears to
pervade all phases of the overseas en-
vironment. This thrust and the increased
tempo have caused OS to undertake a
program to address the problem on a priori-
ty basis in all of its facets. The result is that
OS is now dispatching two-man teams on a
world-wide basis to visit Agency residences,
to counsel personnel, to study personnel
habit patterns which might play into the
hands of terrorists. Eight such teams were in
the field as of 1 April. The program has been
weicomea with consioeraoie enchusiasm oy
those stations visited thus far.
personnel
THE PROJECTION OF AGE DISTRIBUTION
MODEL
The Office of Personnel periodically
analyzes on-duty professional personnel to
determine the age cistribut on of the
professional population. Frequently, the
analysis has involved some tedious hand
counts and computations. During the past
year, however, a computer-based model
programed in the Dynamo language has
been used to facilitate the analysis.
The System Dynamic methodology was
used to design the model. It is based on the
assumption that there is a relationship
between the number of persons in an age
group and the number of persons who flow
:n ard out c' that age group as a result of
aging, accessions and separations.
The model was used first to project the
effect of several different management
policies on the age distribution of the on-
duty professional personnel. For example,
OP wanted to find out what the impact would
be if a particular directorate hired various
numbers of professionals. The managers in
that directorate were considering three alter-
natives they could follow between 1 July 75
and 30 June 80:
hire 50 in FY 76 and 70 each year
thereafter;
Approved For Release 2001/07/12 : CIA-RDP81-00142R000400030027-9
 Approved For Release 2001/07/12 : CIA-RDP81-00142ROO0400030027-9
1 100
June 1975-June 1980
AGE DISTRIBUTION OF "Y" PROFESSIONAL PERSONNEL
employees
600 r
400
2C-24 25-29 30-34 35-39 40-44 45.49
age Qrol1D~
Approved For Release 2001/07/12 : CIA-RDP81-00142ROO0400030027-9
 ppproved For Release 2001/07/12 - :C1A-RDP81-001428000400030027-9
hire 50 in FY 76 and 125 each year
thereafter; or
hire 125 each year.
OP was able to provide them with the infor-
mation which helped them to identify a
desirable option, given objectives and con-
straints.
The model was used again to project what
the age distribution of professional per-
sonne' d t at ''e end v1 Fv ?15 would be
in FY 80.,This ~ model profile was compared
with a modified profile which had been pro-
jected to show the age distribution under a
policy that would not permit the hiring of
professionals 34 years old or younger during
a five-year period. Again, OP was able to
provide management with information which
they needed to consider in balancing the im-
mediate or short-range problem of being
over strength with the more long-range re-
quirement to have qualified personnel to
meet manpower needs at the managerial
level. See chart, page 9.
Many personnel policies can affect future
age cistributions in the Agency. Conse-
quently, it is useful for the manager to have a
model that can simulate the effects of
various alternate policies and thus facilitate
the choice of preferred policies. The System
Dynamic methodology provides an excellent
vehicle on which models for the testing of
policies can be built. The current model is an
example of this method and is capable of
further development once we have a better
understanding of the causal relationships
affecting separations and accessions.
This model is an important member of a
family of analytic tools that can assist
managers to make personnel management
decisions.
STATINTL
Approved For Release 2001/07/12 : CIA-RDP81-00142R000400030027-9
 Approved For Release 2001/07/12 : CIA-RDP81-00142R000400030027-9
r December 1978
MEMORANDUM FOR: Director of Personnel
FROM: Director of Central Intelligence
SUBJECT: Personnel
1. For some time now I have been unable to obtain adequate
information on which to base estimates of how many people we need
to take in in each Directorate each year in order to ensure a
reasonably steady flow, especially in the DDO where lack of adequate
input in any given period of time will lead to definite gaps in the
future because of the lack of lateral entry. I have just had a
glance at the dynamic simulation of personnel systems work that has
been done by the Information Science Center. Could we not use these
models to our advantage in this area?
2. Again, my most specific interest is what is the right number
of people to bring into the professional areas of the DDO every year.
we could solve that one, I think all of the others are probably
basically more simple. Please let me know if you've looked at this
possibility.
STATINTL
S!ANSFIELD TU
Approved For Release 2001/07/12 : CIA-RDP81-00142R000400030027-9