THE PROGRAM HYPOTHESIS—AN INITIAL STEP IN THE

THE PROGRAM HYPOTHESIS—AN INITIAL STEP IN
THE PROGRAM, PLANNING A N D BUDGET
SYSTEM APPROACH
by
Director of Research,
S
MARTIN
GREENBERG
Council of Jewish Federations
O much, attention has been given to
& Welfare
Funds, New York
subjecting them to empirical confirma­
the challenge of computing benefit/
tion (in the same way that theoretical
cost ratios, that we tend to overlook the
hypotheses are confirmed or rejected).
usefulness of some of the less sophisti­
To the extent that this confirmation takes
cated preparatory steps which are also an
place we increase our ability to make wise
integral part of the Program Planning
decisions in communal affairs.
Budgeting System ( P P B S ) approach.
One of these initial steps is the form­
ulation of what might be called the "pro­
gram hypothesis",
a set of
estimates
which could be useful in considering any
program proposal.
In this paper we de­
scribe these estimates, give
illustrative
examples, and show how they might be
useful
in
community
planning
and
There are four components in the
'' program hypothesis'':
1. Estimate of
the
"program uni­
verse. ' '
2. Estimate of the number of persons
(within the "program universe") who
are expected to participate in the pro­
gram.
3. Expected "outcomes" for the pro­
agency administration.
gram participants.
The Built-in Hypothesis in
maining people in the program universe,
4. Expected "outcomes" for the re­
(i.e. those who do not participate in the
Every Program Proposal
program).
Every
proposal for a
new program,
All of these elements are familiar and
whether it be " demonstration," " pilot,''
intuitively understood.
"exploratory," or "experimental," has
receive a very substantial amount of
built-in expectations or assumptions re­
thought and attention by the lay and
garding both the population to be served,
professional leaders who formulate new
and
the benefits
they,
and the com­
munity, are likely to receive.
to give these
expectations
label "program hypothesis."
(Too frequently, however,
when the serious negotiations for financ­
There are two reasons why it may be
useful
programs.
Generally they
the
First, the
ing
a proposal take
place
with
the
central planning and financing organiza­
tion, some of these elements tend to be
fact,
given a "once over lightly" treatment.
hypothesized figures (based on relation­
Instead, much more emphasis is placed
ships between outcomes and activities in
on such administrative considerations as
other program areas).
Secondly, this
how large a staff will be required, how
label calls attention to the desirability of
much will the new program cost, where
expectations
are
frequently,
in
will the money be found, what will be the
tion of whether there are enough people
structure for the new program). From
"out there" to sustain the proposed pro­
the viewpoint of PPBS, the community
gram.
planning process will be enhanced if all
of these elements received explicit
atten­
establish a Golden Age program in a
new area.
tion.
Let us examine what goes into each of
them.
Here is an illustration;
The community center proposes to
They may appear to be simple,
straightforward
and
reasonable,
but
It is to be administered by
a staff of one, and have the capacity to
serve 125 people.
The number of older
people in the area may fall anywhere
there are a number of difficulties in ad­
between
vancing productive estimates of size, or
ageney feels confident that it will be able
1,000
and
10,000,
and
the
hypothesized outcomes.
to attract at least 125 participants.
The Program Universe
planning body is a " y e s " or " n o " re­
If the only decision to be made by the
sponse to this one program proposal, then
The "program universe" consists of all
a very crude estimate of the "program
the people who conceivably would be
universe"
"consumers" (or "participants," or be
make a big investment in determining
"exposed to the treatment") if the pro­
whether the program universe is more
gram were to have universal
likely to be 5,850 (or 10,000) when, for
coverage.
would be sufficient.
Why
Thus, for a Golden A g e program the
administrative reasons, the program to
"program universe" might consist of all
be financed will have a capacity to serve
ambulatory persons, aged 60 years or
only 125?
more,
area.
living
within
the
geographical
The P P B S approach, however, sug­
For a medical home care program
gests that attention be constantly fo­
it might include all patients who cur­
cused on the entire potential universe-
rently have certain specified medical and
to-be-served. Other programs could then
social characteristics.
be developed (if they do not already ex­
For a fund-rais­
ing campaign, it might include all per­
ist), and each proposal could be con­
sons who might conceivably be asked to
sidered as alternatives aimed at the same
contribute.
'' program universe.''
Most of the time the characteristics of
Moreover, if the focus is primarily on
the "program universe" are inferred
how many people the agency wants to
from the many bits and pieces of in­
serve—with no counter-balancing atten­
formation which normally serve as our
tion to how many people are '' out there''
"indicators" of social trends. Occasion­
needing
ally a special study may be undertaken
drift
to get estimates of the size, geographical
(rather than being people-oriented).
service—planning
can
into an agency-oriented
soon
process
location or social characteristics of the
"program universe."
It is only on rare
Estimate of Program Participants
occasions that a voluntary agency may
have a full census of the "program uni­
Few programs are expected to provide
verse" and detailed information about
100 percent coverage of the "program
the characteristics of individuals in it.
universe" (in the way, for example, that
Curiously enough, when a new pro­
every
child
gets
vaccinated
against
gram is proposed, few agencies find it
smallpox).
essential to get a narrowly defined esti­
built-in
mate of the '' program universe.''
limit the number of participants.
Most
frequently there is just a determina­
Most of the time there are
processes
of
selection
which
This
selection may be exercised by the agency
through its eligibility requirements, or
by the clients (through their exercise of
freedom of choice).
possibility of inconsistency if rated by
different judges, as for example: family
tension reduced; student now working
Moreover, the "program universe" is
rarely a homogeneous set of individuals.
Most of the time it consists of a con­
tinuum of people all of whom could be
considered as "fair game" for the pro­
gram, but some much more so than
others.
at full capacity; group achieved fuller
Occasionally a completely new pro­
gram comes along, and all an agency
may be able to say, in good conscience,
is something like this; "we have no ex­
perience from which to estimate how
many people are likely to participate in
the program, or from what part of the
continuum they will come. W e don't
even know how many people our staff
will be able to serve during the initial
phase. But we hope to be kept busy."
worker goals; degree of resolution of
A P P B S approach would suggest,
however, that even in these extreme sit­
uations it would be desirable to offer
"guesses"; these serve as a basis for a
preliminary assessment of the signif­
icance of the program, and as a reference
point for follow-up attention.
integration of members. (The literature
incidentally is giving increased atten­
tion to the various dimensions, or char­
acteristics, whieh could be used in clas­
sifying client outcome.
degree of
crises,
or
These include:
achievement of
client and
problem-solving;
absolute
movement from one specified level of
functioning to another specified level;
relative amount of movement, regardless
of the absolute "before" and "after"
levels; and so forth.)
Most practitioners have an impression
of what each participant got from the
service.
This impression may be purely
intuitive and global, or elegantly con­
ceptualized and compartmentalized. But
in either case, PPBS would suggest that
it is better to have the benefit of existing
practioner judgments rather than work
without any information regarding the
practitioner's appraisal of client out­
comes.
Practioner judgments are not neces­
The Expected Outcomes for the
sarily an evaluation of the impact of the
Program Participants
agency's program.
Smallpox vaccination may give 100 per­
cent immunity to the disease, with no
adverse side effects. But few other pro­
grams hit the bull's eye in the same way.
Most other programs have outcomes that
include drop-outs, low achievers, and
gratifying successes.
Those who propose programs gener­
ally have both a yardstick for describing
client outcome and some expectation of
the distribution of outcomes that is likely
to occur.
Some yardsticks are objec­
tively defined and readily measured, as
for example: client got job, student got
diploma, child placed in adoption. Other
outcomes may be more abstract, with the
For the most part
they just give a "before" and "after"
comparison of the participant.
Change
may, of course, be in response to the
agency's program.
But changes may
also have occurred in response to other
factors, not related to the agency's pro­
gram, such as: assistance given by other
persons or agencies; normal growth and
maturation; changes in the social en­
vironment
But even if we cannot say,
with scientific rigor, that the client out­
comes were the direct or indirect result
of participation in the agency's pro­
grams, it is nevertheless useful to know
whether changes took place and their
nature.
tive vocational fields, who have demon­
strated leadership ability in one or more or­
ganizations.
Expected Outcomes for the
Other People in the
Program Universe
There is an obvious rationale
for all
programs: something good happens to
those
who
participate,
and
something
not-so-good happens to those who do not
participate.
program,
The measure of success of a
therefore,
pivots largely
on
how much of a gain is achieved for par­
ticipants
as compared with
non-partici­
pants.
Here too, those who propose a new
program invariably have some distinct
ideas about what is happening to people
"out there" in the program universe,
and why it is so essential to establish the
program they propose.
From the point
of view of P P B S , it would also be very
useful to have an explicit statement of
what they see as the outcomes for those
who would not participate in the pro­
posed program.
Program participants—The program will
be limited to 30 men with priority selection
of those who show the most "predisposi­
tion" for federation activity, but are not
currently active.
Expected outcomes for program partici­
pants—five will drop out because of pres­
sure of other commitments, three will lose
interest because of personality factors or
disagreement over how federation activi­
ties are conducted; 15 will become active
on committees and special projects; and
seven will become board members within
two years.
Expected outcomes for the others: Of the
estimated 410—420 who do not participate
in the program: 15 will be recruited for
board and committee positions in the mem­
ber agencies; 15 will become active on
federation committees or special projects,
and one will become a member of the fed­
eration board within two years. The bal­
ance will not show any increased participa­
tion in federation activities.
Examples of Program Hypotheses
Perhaps the easiest way to gauge the
potential
usefulness
of
a
"program
hypothesis" is to look at concrete ex­
amples which might be formulated from
situations which arise in agencies.
Four
are cited below. They deal with a leader­
ship development progam in a federa­
tion, a Golden Age Group, a social work
recruitment program and a treatment
center
for emotionally
disturbed
chil­
dren:
A—Leadership Training Program—This is to
consist of a group, meeting monthly, under
the sponsorship of the federation president.
The program objective is to give the mem­
bers an "inside story" of how the federa­
tion works, a motivation to grapple with
some of its problems, and a feeling of per­
sonal ties with key federation leaders.
Program universe—Based on review of
federation's master file, there are an esti­
mated 440-450 men, between 30-40 years
of age, in the top echelon in their respec­
B—Golden Age Group—Concerned over the
older people in Area who might feel lone­
some or isolated, the program objective
is to offer three varied opportunities a
week for group participation.
These include: ( 1 ) a weekly group discus­
sion, or other activity fostering member­
ship participation, ( 2 ) visits to places of
interest and ( 3 ) lounge-type of activity.
Program universe—Based on (a) census
data, (b) a recent Jewish population study,
and (c) judgments of persons who know
the area, the estimated number of persons
who conceivably could participate in the
program is as follows:
Age
Total
Men
60-79
70-79
80 and over
1,150
900
450
500
400
200
650
500
250
2,500
1,100
1,400
Total
Women
Estimate of program participants—Enroll­
ment will be limited to 150 persons (this
being the optimum number which can be
accommodated by available staff). Selec­
tion to be made on a "first come first
served" basis.
Expected outcomes for the program par­
ticipants—10 enrollees will come twice or
less during the entire program year; 20
will come an average of once a month; 30
will come an average of two or three times
a month; 40 will come an average of 4
times a month; and 50 will come an aver­
age of five or more times a month.
Expected outcomes for the other people—
Of the 2,350 other people in the program
universe: 500 participate in programs
sponsored by other ageneies or organiza­
tions, coming to events on the average of
twice a month—and do not seek additional
opportunities; 600 also participate in other
programs, but seek additional opportu­
nities; the remaining 1,250 do not partici­
pate in any organizational program but feel
that they already have a "full schedule"
of activities.
C!—Social Work Recruitment Program—This
is a series of five-session seminars, con­
ducted by different social agencies. Par­
ticipants are to include college sophomores
and juniors. The program objective is to
give them a first-hand view of the work
of an agency, and to motivate them to
consider social work as a career.
Program universe—Based on school enroll­
ment data, an estimated 5,400-5,600 stu­
dents are in their sophomore or junior
year at colleges in our town.
Program participants—Seek participation
of 200 students, with priority attention
given to those whose "profile" places them
in the groups that are most likely to pur­
sue » social work career.
Expeited outcomes for program partici­
pants—Of the 200 participants: 20 will
attend only one or two sessions; 80 will
attend 3-5 sessions, but not show any fur­
ther interest; 50 will attend 3-5 sessions
and accept offer of an individual confer­
ence on career opportunities; the remaining
50 will also attend 3-5 sessions, ask for
an individual conference on career oppor­
tunities, and enroll in a school of social
work within one year after graduation.
25 will enroll in a school of social work
in response to other recruitment activities
in the community; an additional 10 will
enroll in a school of social work without
having participated in any other special
recruitment program; 15 will accept em­
ployment with a social agency (without
enrolling in a school of social work); the
balance will not show any interest in
social work as a career.
D—Treatment Center For Emotionally Dis­
turbed Children—This is for emotionally
disturbed children aged 13-17, suspended
from school because of "anti-social" be­
havior, who are likely to respond to the
Treatment Center's program. They would
live in group residences, receive intensive
psychotherapy and a variety of other ser­
vices for themselves and their families. The
program objective is to restore them to
public school.
Program universe—Based on census data
that there are an estimated 12,500 children
in the 13-17 age group in Our town. From
information received from the schools,
court, other social agencies, religious lead­
ers, etc., there are between 125 and 550
children showing "potentially serious anti­
social behavior" who might be referred to
such a treatment center.
Program Participants—The program would
be available to 15 children at a time, with
priority selection given to those who pre­
sumably have the greatest potentiality to
benefit from the service.
Expected outcome for the program partici­
pants: For every 15 accepted for care:
one will drop out (without any sign of
improvement) within one month of admis­
sion; two will continue beyond one month
but will be transferred (without any sign
of improvement) to other forms of inten­
sive treatment within one year of admis­
sion; six will receive care for twelve to
eighteen months and show moderate im­
provement (i.e. reduction of symptoms,
some reorientation of personal goals, but
a likelihood that further help would be
required during crises) and return to school
under "supervision"; the remaining six
would also continue for 12 to 24 months,
show a substantial reduction of symptoms
and ability to cope independently with the
full range of problems and social situa-
tions—and return to school without requir­
ing "supervision."
Expected outcome for the others—For those
who do not participate in the program, it
is expected that about % will show no
change or become worse and continue in
the "suspended from sehool" category;
% will show minor improvement but bounce
in and out of school; and the remaining
% will show substantial improvement and
continue in school without interruption.
fundamental question of effectiveness of
agency service.
The Golden Age group, for example,
was established out of concern for older
people "who might feel lonesome or iso­
lated."
client
But the suggested indicator of
outcome deals exclusively
with
frequency of participation, as if one had
already demonstrated that high attend­
ance reduces feelings of loneliness and
isolation.
These figures might tell us more, how­
Comments on Examples
ever, if the agency had two Golden Age
These examples illustrate the range of
groups.
problems in trying to formulate a pro­
both groups five percent of the enrollees
gram hypothesis.
dropped out after only two sessions. But
They also suggest a
Suppose, for example, that in
estimates
in Group A , 85 percent of the members
whieh could be used until more precise
came an average of five times a month
variety
of
interim
working
or more, while in Group B only 25 per­
data become available.
For
some
programs,
reasonably
cent came this often.
"hard" data may be obtained as esti­
mates of the '' program universe.''
example,
the
Federation
master
For
file
probably contains the name of every upand-coming younger executive
community.
in
the
On the other hand, esti­
mating the number of emotionally dis­
Of course, there might be a Group C
organized by the agency as a bridge and
canasta league.
Smaller in size, Group
C meets not at the agency's building,
but rotates its sessions at the homes of
the members.
Thus each year a partici­
pant takes a few turns in serving as the
turbed youth is doubly difficult because
hostess of her group.
of the widely ranging views on what con­
figures for Group C may be comparable
stitutes
to Group A .
potentially
"anti-social"
be-
disturbance,
and
havor, or emotional
The attendance
But the quality of partici­
pation may be quite different, and an in­
what patterns of disturbance are likely
dicator of client outcome based only on
to respond to intensive psychotherapy in
attendance figures would be "blind" to
a residential treatment center.
the difference in personal meaning of the
Similarly, some client outcomes can
group.
be recorded very easily, as for example,
Similarly, for the leadership training
the frequency of attendance at events
program, gross figures on participation
sponsored by the Golden Age Group.
in committees, or membership in
Outcomes in other programs, however,
board, would be insensitive to differences
may require the passage of time (as for
in quality of leadership participation.
the
example, the social work recruitment
Yet with all these limitations, it is
program) or a worker's prediction of
better to get some figures on frequency
future behavior (as, for example, what
of
a young person is likely to do when he
client outcome) rather than no figures at
leaves a residential treatment center to
all, which is frequently what happens
move to a less protected setting).
participation
(as
an
indicator
of
when an agency keeps only records of
total enrollment.
Crude as they may be,
tions used for client outcomes clearly do
these
can
not provide definitive
"middle ground" gain: they spot gross
In these illustrations, the
classifica­
answers to the
indicators
offer
a sort
of
differences in patterns of outcomes, and
thus point the way to more selectivity in
follow-up studies.
Frequently a functional agency will
have very full information about its own
program participants and the outcomes
which appeared for them. However, the
functional agency may have very sparse
information about the "program uni­
verse," or the outcomes for those who
do not participate in their programs.
For these two items, the central organi­
zation may be of special help since it
generally
accepts
responsibility
for
monitoring trends in social problems in
the community.
Assets and Liabilities
Formulating the program hypothesis
may require extra work—and may in­
troduce "extra controversy", partic­
ularly if the estimates of the "program
universe" or of client outcome cannot be
based on prior experience or information
drawn from comparable situations.
What pluses would we get from the
extra effort?
First, the program hypothesis is a per­
sistant reminder to keep in constant view
all the people living "out there" for
whom services might be considered. This
increases the likelihood of a better
balance between agency-oriented plan­
ning (which is essential) and peopleoriented planning (which is also essen­
tial).
Secondly, the program hypothesis
nudges us to be explicit in communicat­
ing our ideas about the people who might
be served, and the nature of the benefits
they and the community can be expected
to receive. There may be good reasons
for having hazy views: the program pro­
posal may by its very nature be a pion­
eering effort, and almost anything might
happen. Or the outcome might be dif­
ficult to express in objective terms (rely­
ing instead on practitioner judgments
which might be biased or show low statis­
tical reliability). Or the assessment of
outcome may require the passage of a
considerable amount of time before a
judgment can be made. But the require­
ment to write down our "best guesses"
will inevitably encourage us to find some
way of getting better empirical data. In
time, this should lead to progressively
more useful working estimates which
could be used for planning.
The "program hypothesis" also paves
the way for attracting wider participa­
tion by research personnel. They are
accustomed to analyzing problems in
terms of (a) variations in outcomes and
(b) the different factors which are as­
sociated with these variations. Their
collaboration could be tremendously
helpful in refining the different measures
which could be used to describe client
outcome. They could also help us im­
prove our use of samples; this could
reduce the cost of our studies and, at the
same time, increase the ability to make
generalizations.
They also have the
skills to tell us whether differences (be­
tween expectations and actual findings)
may
be due to chance or reflect the
influence of special factors.
The program hypothesis also provides
a logical outline for reports on demon­
stration
projects.
Essentially,
the
demonstration project is a promise: if
the community makes available certain
specified resources, a program could be
sustained which would produce the kind
of benefits which the community would
want to have on a continued basis. By
specifying the program promise in ad­
vance, in the form of a "program hy­
pothesis," the planners and the agency
administrators get a clear, mutually
understood, chart of the critical check­
points.
Finally, by focusing attention on
client outcome (and encouraging the
development of progressively more use­
ful techniques to categorize them) the
"program hypothesis" paves the way
for further analysis. One possibility is
to add data on program costs and move
in the direction of benefit/cost ratios.
Another possibility is to do studies in
depth on the dynamics of the relation­
ship between client outcome and ele­
ments of the agency program, and thus
move in the direction of examining
agency impact and developing guidelines
for improving practice.
become of secondary importance during
negotiations with the
"sources of
funds." A s soon as the conclusion is
reached that an adequate pool of "poten­
tial consumers" exists, there is a tendancy to give primary attention to the
administrative aspects of a program pro­
posal—and the size, structure, location,
and so forth are often determined by
considerations of what would be admin­
istratively most effective or most effi­
cient.
Summary
These administrative considerations
are essential and desirable, but in time,
they tend to produce a quality of social
planning which is agency-oriented. The
"program hypothesis" recaptures the
original thinking about people and their
problems and provides an orderly way in
which this orientation can be kept in the
mainstream of community planning.
The "program hypothesis" can also
pave the way for benefit/cost studies
and development of guidelines for im­
proved practice.
"Program hypothesis" is suggested as
a useful initial step for agencies which
wish to explore the potential usefulness
of P P B S approach.
Essentially, the
"program hypothesis" is a set of esti­
mates which most agencies normally
make when they formulate a proposal
for a new program, or a substantial ex­
tension of an existing one. While most
agencies consider these estimates in their
own preliminary planning, they often