https://ntrs.nasa.gov/search.jsp?

R=19770004965 2019-09-29T16:08:22+00:00Z

N A S A TECHNICAL NASA TM X-3453

MEMORANDUM

COST-BENEFIT ANALYSIS
OF SPACE TECHNOLOGY

Gerald F. Hem, Steven M. Stevenson,

and Joseph N. Sivo
Lewis Research Center
Cleveland, Ohio 44135 I

V
NATIONAL AERONAUTICS AND SPACE ADMINISTRATION • WASHINGTON, 0. C. • NOVEMBER 1976
1. Report No. 2. Government Accession No. 3. Recipient's Catalog No.
NASA TMX-3453
4. Title and Subtitle 5. Report Date
COST -BENEFIT ANALYSIS OF SPACE TECHNOLOGY November 1976
6. Performing Organization Code

7. Author(s) 8. Performing Organization Report No.

Gerald F. Hein, Steven M. Stevenson, and Joseph N. Sivo E-8796
10. Work Unit No.
9. Performing Organization Name and Address 644-04
Lewis Research Center
11. Contract or Grant No.
National Aeronautics and Space Administration
Cleveland, Ohio 44135 13. Type of Report and Period Covered
12. Sponsoring Agency Name and Address Technical Memorandum
National Aeronautics and Space Administration 14. Sponsoring Agency Code
Washington, D.C. 20546
15. Supplementary Notes

16. Abstract
A discussion of the implications and problems associated with the use of cost -benefit techniques
is presented. Knowledge of these problems is useful in the structure of a decision making proc-
ess. A methodology of cost -benefit analysis is presented for the evaluation of space technology.
The use of the methodology is demonstrated with an evaluation of ion thrusters for north -south
stationkeeping aboard geosynchronous communication satellites. A critique of the concept of
consumers surplus for measuring benefits is also presented.

17. Key Words (Suggested by Author(s)) 18. Distribution Statement

Cost benefit analysis Unclassified - unlimited
Technology assessment STAR Category 83
Value analysis

19. Security Classif. (of this report) 20. Security Classif. (of this pagel 21. No. of Pages 22 Price*
Unclassified Unclassified 25 $4. 00

* For sale by the National Technical Information Service, Springfield, Virginia 22161
 COST-BENEFIT ANALYSIS OF SPACE TECHNOLOGY

by Gerald F. Hein, Steven M. Stevenson, and Joseph N. Sivo

Lewis Research Center

SUMMARY

A discussion of the implications and problems associated with the use of cost-benefit
techniques is presented. Knowledge of these problems is useful in the structure of a de-
cision making process. A methodology of cost-benefit analysis is presented for the eval-
uation of space technology. The use of the methodology is demonstrated with an evalua-
tion of ion thrusters for north-south stationkeeping aboard geosynchronous communica-
tion satellites. A critique of the concept of consumers surplus for measuring benefits is
also presented.

INTRODUCTION

The question of whether government programs are worth the price first appeared in
literature in 1844 in a paper by Jules Dupuit a French engineer. Since that time there
have been many developments and refinements of methods used in what is now called cost-
benefit analysis, but in many ways Dupuit's criticism remains valid. His complaint was
that "... political economy has not yet defined in any precise manner the conditions
which these (public) works must fulfill in order to be really useful" (ref. 1).
The initial works dealing with the evaluation of public projects began in the U. S. with
the passage of the River and Harbors Act of 1902 which required a board of engineers to
report on the desirability of Army Corps of Engineers' river and harbor projects, taking
into account the amount of commerce benefited and the cost (ref. 2). From an economic
point of view, a significant event occurred in 1936 with the passage of the Flood Control
Act. Federal participation in flood control programs was authorized "if the benefits to
whomsoever they may accrue are in excess of the estimated costs. " It is important to
note several consequences of this particular law. The first was that the Congress stip-
ulated a necessary condition for federal participation: the benefits must exceed the costs.
But the act does not demand participation when benefits exceed costs; it only authorizes
participation. Thus in 1936 Congress recognized the importance of leaving a final
decision with a decision maker rather than with the outcome of an analytical technique.
The second consequence of the 1936 legislation was that Congress approved the use of a
decision criterion derived from the branch of economics dealing with public or govern-
ment expenditures. The criterion allows for the fact that individual citizens of the
United States will incur costs and receive benefits from public programs. The criterion
further allows for the compensation of individuals for costs directly incurred beyond tax-
ation but does not require compensation for indirect effects such as loss of trade result-
ing from changes in demography. On the other hand, individuals are not prohibited from
benefiting directly or indirectly from the effects of public expenditures regardless of
social status.
Since 1936 there have been many accomplishments in the development of cost-benefit
methodology, but, when the need to use cost-benefit analysis arises today, it is usually
not apparent what should be included in such an analysis. As resources become more
scarce in the public domain, there will be a continuing need to determine whether ben-
efits exceed costs. In addition to making cost estimates and determinations of monetary
benefits, there is a need for estimating the monetary value of benefits and costs of things
more difficult to quantify in dollar analogues. Social impact assessment is the deter-
mination of the effects of public programs on society beyond those things which may be
measured in monetary terms. What used to occupy government analysts and researchers
from the social and economic areas of academic institutions has slowly evolved into
rather controversial issues at a time when resources are becoming scarce; and decision
makers in the public domain are in need of evaluative techniques for assessing the merits
of the government programs competing for these scarce resources.
One of the primary difficulties of cost-benefit analysis, economics, and social im-
pact assessments is that they are not exact sciences and are therefore somewhat lacking
in the consistency which is attainable in engineering and other exact sciences. This lack
of exactness is not a weakness, although many engineers and scientists label the social
sciences as being inferior for this reason. If the task of modeling society were as sim-
ple as that of modeling a transformer, it would have been accomplished by now. Despite
these difficulties, the need exists for information with which to make decisions in the
public domain. This need goes beyond the evaluation of technical plans. Unfortunately
no one has yet determined how far and exactly what should be considered when perform-
ing cost-benefit and social impact assessments for the expenditure of public funds. Some
of these considerations and the mechanics will be discussed further in the next section.

A DISCUSSION OF SOME OF THE PROBLEMS

Within the past few years, the terms cost-benefit analysis and impact assessment
have begun to receive widespread attention. Much of the literature devoted to these
topic deals with methodology, but a considerable portion deals with criticism of the many
techniques that have been proposed for use in the process of allocating expenditures.
Much of the criticism is formulated with perspectives that reflect an extreme point of
view, whether from a position of advocacy or adversity. There are those who believe
that the methodology is useless because it is so eclectic as to be incomprehensible.
Others assail the techniques because benefits and costs can be manipulated to yield a
predetermined answer or because the role of the decision maker is rendered superfluous.
At the other extreme are those who would allow the quantification of all costs and benefits
in order to develop a benefit-cost ratio, and those who say that cost-benefit analysis may
be used in lieu of a decision.
A more reasonable perspective is that cost-benefit analysis is an aid to decision
making. The analysis is based primarily on economic science and may be used to es-
timate whether, and to what extent, the economic benefits of a proposed project outweigh
the economic costs. Not all costs and benefits can be transformed to dollar analogues,
and so this analysis constitutes only a segment of the information required to make a de-
cision. Cost-benefit analysis might more properly be termed a state of mind rather than
a method. Within this perspective, an examination will be made of some of the problems
associated with this type of analysis. Some examples will be developed to demonstrate
that despite shortcomings, this type of analysis can be used to evaluate some of the eco-
nomic value to be gained from the expenditure of some resources.
Two of the most difficult areas in which to apply cost-benefit analysis are the eval-
uation of the benefits of research and development programs and the evaluation of social
programs. There are fundamental reasons for these difficulties. Rather than resort to
criticism of research and social programs or of cost-benefit methods, it is imperative
to first examine the nature of each in order to determine the source of some of the crit-
icism and difficulties encountered in the application of cost-benefit analysis in these
areas.
The branch of economics that deals with the optimal allocation of public (or govern-
ment) funds is known as welfare economics. Despite the connotations associated with
the term welfare, it is used here in a broader sense than is normally construed; welfare
economics relates to the spending of all government funds. With this in mind, let us ex-
amine some of the problems encountered in the process of managing the distribution of
public funds for federal programs.
During the past few years, many forms of technology have been criticized because of
the inability of anyone to foresee the adverse consequences of the implementation of the
technology. As a result of some of these inadvertently created problems, managers have
had to contend with demands to demonstrate beforehand that a program is worthwhile and
that the deleterious effects are minimal. They are asked to deliver in addition to the
usual technical plans such things as benefit-cost ratios, social impact scenarios,
quality-of-life considerations, risk assessments, and other reports that have little ap-
parent relevance to what is being proposed.
On the other hand, there are those who object to the reduction of a technical plan to
a number (the benefit-cost ratio) which is then compared with the ratios assigned to other
plans in order to select what should be done with the resources at hand. The benefit -
cost ratio is subject to criticism from different directions and for different reasons.
Cost-benefit analysis is criticized because many important considerations are over-
looked in the generation of the benefit-cost ratio. There simply is no way to generate a
number that can consider the effects of polluting a stream or of saving a person's life.
Many managers object to the comparison of their projects with others because the
competition is decided by the best salesman rather than the merits of the proposed work.
The skepticism is typified by the words of a retired Army colonel which appeared in the
New York Times in 1972 and again in a journal article: (referring to the b-c ratio)
"Its nothing but a ritual. They come down the aisle swinging their incense and chanting
TDenefit-cost. ' You can adjust the b-c ratio to justify any project. I did it myself a few
times" (ref. 3). It is unfortunate that the colonel completely misunderstood the purpose
of the methodology.
Another alleged weakness of cost-benefit analysis is that the same agency that pro-
poses and sponsors a particular project performs the benefit-cost analysis (ref. 3).
This criticism likewise indicates a misunderstanding of the purpose of using such tech-
niques to evaluate project proposals. It is a matter of good business to gather informa-
tion and use it to rank alternatives when making a decision concerning the expenditure of
resources.
In addition to technical evaluations and consideration of other factors, it is reason-
able to expect a decision maker to formally consider the economics of various proposals
for the allocation and spending of public monies. Because the goals of the nation are
considerably less tangible than maximizing profit or minimizing cost, the decision cri-
teria and methods used in the evaluation of public expenditures are more complicated
than the methods encountered in private enterprise.
To begin an elaboration of methods and some of their implications, it is necessary
to define the principal caveats associated with cost-benefit analysis or the numerous
designations and names associated with this portion of policy analysis. Whether one is
a practitioner, a manager, or a bystander, it is well to remember that there is no sub-
stitute for good judgment, common sense, and logic. Cost-benefit analysis is one of
many basic information requirements a decision maker may use. The information itself
should not force a particular decision, but merely serve as a source of information and
an explicit declaration of the assumptions made by an analyst.
The use of this information to construct a ratio of benefits to costs for comparison
with other proposals is, in general, not recommended. The results of such use of cost-
benefit analysis techniques are not generally valid when used to compare benefit-cost
4
ratios of many different types of programs. Inevitably such use will result in the selec-
tion of those programs with the largest ratio of benefits to cost. The implication of such
use is that this ratio is a sole criterion and thus there is little need for a decision maker.
The logical extension of comparing the benefit-cost ratio of one program with another is
that all benefits and needs are of equal weight, which is not generally true. Suppose a
policy analyst has determined that two programs, A and B, have quantifiable benefit-cost
ratios of 10 and 5. Suppose further that program A is a new and innovative method of
providing remote health care delivery and that program B is the development of a hybrid
food staple. Since the benefit-cost ratios considered only those aspects of each program
that could be quantified in dollar values, a comparison of the ratios on this basis will
only result in the implication that A should be funded before B. However, the decision
should not be made before considering program priorities, the needs of the target pop-
ulations, the impact of implementing or not implementing either program, and the policy
objectives of the funding organization. In actual practice, the priorities and needs are
not obvious; hence, the requirement for the decision maker.
Another implication of the benefit-cost ratio is that the magnitude of costs involved
is not important. Thus a very costly project may be preferred over one costing much
less if the more costly project has a higher benefit-cost ratio. Again, the need for a de-
cision maker is obvious to prevent the methodology from dictating policy to the organiza-
tion. The purpose of these naive examples is to demonstrate that there are many things
to consider in a decision making process and that cost-benefit or economic analysis is
only a part of making the decision.
A distinction often overlooked is that there are certain endeavors that simply do not
submit to economic analysis. Some programs are so exploratory in nature that there is
no reasonable way of estimating the monetary value of the outcome. The conduct of basic
research provides a good example. This is not to say that basic research should not be
goal directed (quite the contrary). It simply means that basic research is exploratory in
nature whereas cost-benefit analysis of necessity is product oriented. Basic research is
a fundamental part of the business of operating a laboratory or R & D organization. It
should require justification only to the extent that other overhead items do. It is no more
feasible to justify exploratory research with cost-benefit analysis than it is to apply these
procedures to a justification for turning on the lights. The alternative is to abandon re-
search activities and wait for development and applications to stop.
In addition to the conceptual problems associated with placing economic value of ex-
ploratory research, there exist similar problems in the social and political mileau. One
easily comprehends the necessity of comparing apples with apples and oranges with or-
anges, but the dictum loses significance when the apples are changed to dollars and the
oranges are relegated to units of measure not readily transf err able to dollar analogues.
An example of this is the attempt to display the monetary merits of social programs that
save lives. The literature is replete with methods of assigning a dollar value to human
5
 life such as pareto optimality and compensating variation. But the use of such tech-
niques assumes a priori that there is a value to saving human life before giving any con-
sideration to more fundamental questions. The result is a contest of one-upsmanship in
which none of the methods of valuing human life provides an acceptable measure. A
more reasonable approach for an organization might be to pose the question: Is there
value to human life? Or, should resources be expended for saving lives? Going to the
fundamental question and defining it in a way that is amenable to solution within the con-
fines of organizational policy should result in a definitive answer. If the outcome is pos-
itive, an approach to defining effectiveness should be in units of lives saved since this is
the common denominator, which is also irreducible to dollar analogues.
The point of this example is that considerable thought should be given to defining the
problem. When problems are not defined accurately, it is possible to make assumptions
which imply a solution to the genuine problem. The implied solution may be incorrect
but without a proper definition, there is no way of knowing that an error has been made.
Another item requiring care is the selection of units of measure. Many things may
be valued in terms of money, but many more cannot. These may be termed incommen-
surables and intangibles. Incommensurables are those things that have valid units of
measure but which cannot be equated with monetary values. Examples include lives
saved and level of achievement. Intangibles consist of those things that cannot be
measured validly with any system of units. Examples here include patriotism, good will,
and national prestige.
When measuring incommensurables, the unit of measure is valid,but,irreducible.
This does not detract from the value of the unit; it implies that regardless of the value
scale attached to the unit, the unit itself is a meaningful measure and may be used as
such. Intangibles may also have value but there is no meaningful unit of measure. If
one is dealing with a problem involving incommensurables or intangibles that are the pri-
mary output of a program, these things must be dealt with directly and without trans-
formation to dollar analogues.
In the event that there exist secondary effects or spillover, there is often a tendency
to include such effects in the calculation of benefits and costs, whether the effects are
measurable in monetary units or consist of incommensurables and intangibles. Benefits
and costs may generally be divided into project costs and benefits and other costs and
benefits. This division gives rise to two classification schemes known as internal and
external effects and direct and indirect effects.
Internal effects are those which are directly attributable to the project objective.
External effects are neither deliberately produced nor deliberately consumed and may be
valued but not priced. External benefits are involuntarily received by others for which
they are not charged. External costs are imposed on others without compensation.
Direct benefits from a project are increased real values of output associated with the
project. Indirect benefits reflect the impact of the project on the rest of the economy.
 No one would deny that economic impact should be a consideration in the evaluation
of large public programs, but one should consider the implication of including secondary
benefits in the calculation of a benefit-cost ratio. An excellent example comes from a
technology program designed to increase the availability of a certain type of ground
transportation. This particular mode of transportation was the least expensive mode
when compared with two others. The policy analyst calculated a value for the increased
availability and then reasoned that this increased availability would represent a net gain
in income for the geographic region of analysis. This additional regional income would
then have a multiplier effect on the regional economy. The total benefits included the
amounts saved from the increased availability of the lower cost transportation plus the
increase in regional income multiplied by a number representing the multiplier effect.
The secondary regional benefits were enormous compared with the primary benefits.
Several aspects of this analysis merit closer scrutiny. The stated objective of the
analysis was to determine the benefits associated with increased availability of a certain
mode of transportation. However, calculated benefits included the effect of this increase
on the regional economy thereby implicitly changing the stated objective to one of max-
imizing regional income. The analysis fails to consider whether this new income is a
growth in total income or comes at the expense of other modes of transportation. It is
possible that other modes of transportation may incur displacements which would result
in serious economic problems for the entire region. Finally, there appears to be a fal-
lacy in the reasoning used to calculate benefits. The analyst assumed that regional in-
come would be maximized by using the advocated method of transportation, and that
transportation income should be maximized in order to maximize the benefits to the re-
gion. This occurs when transportation charges are maximized. Therefore, all shipping
should be accomplished by the most expensive form of transportation rather than the form
advocated by the analyst. If one chooses to include secondary benefits in support of an
objective, then the true objective will usually be altered.
Although many analyses include secondary benefits in the calculations, secondary
costs are often overlooked because it is difficult to anticipate deleterious effects 5 or 10
years before they occur. A reasonable approach to secondary costs and benefits is to at-
tempt to calculate them but not to include them directly in the economic evaluation. A
decision maker should consider the effects of his decision on the economy both in terms
of benefits and costs, but only if the project is of such magnitude as to have a significant
social or economic impact. In the event that social and economic impact assessments
are conducted, the starting points lie in an analysis of all institutions and commodities af-
fected by the decision. For very large programs social and economic researchers
would be required to perform some of the analysis. When multidisciplinary research of
this kind is conducted, the project manager must carefully structure the problem in order
to obtain the information he is seeking.
 Example of Cost-Benefit Methodology

It is possible to use the methodology of cost-benefit analysis as a consistent struc-

ture to develop information required for policy decisions regarding government invest-
ment in technology programs. The purpose of such an analysis is to estimate the poten-
tial economic and social impact of government investment in technology. In many cases
the methodology has been used as a device to convince government agencies and legisla-
tive bodies of the need to fund particular programs. Rather than use the methodology as
an internal mechanism to supplement decision information, the methodology has been
used as a marketing tool. Although there is nothing objectionable about such practice,
many cost-benefit studies used for marketing purposes tend to concentrate on "making
the system look good" rather than being an objective assessment of the potential benefits
and costs.
The example presented here was developed because of an interest in determining po-
tential benefits and costs of various technologies used in space communications. The
subject technology appears to have more benefits than costs, and so it was selected to
demonstrate the benefits associated with funding a technology development program.
Ion thrusters are engines that generate a very low thrust but are capable of sustain-
ing this thrust for very long periods of time. Chemical propulsion works on the principle
of releasing large masses at high velocities through an exhaust. Because of the principle
of conservation of momentum, the vehicle is propelled in the opposite direction. Ion
propulsion likewise is an application of the conservation of momentum. Instead of burn-
ing and releasing pressurized gases as in chemical propulsion, an element such as
mercury is ionized and then accelerated in an electric field. The ions are then neutral-
ized in charge, and the element is ejected through an exhaust. Because the mass flow
rate is so much less than in chemical propulsion, the propellant lasts much longer. How-
ever, the individual ions are accelerated to much greater velocities than are the particles
in chemical propulsion. This results in a much higher specific impulse which is a per-
formance measure for rocket propulsion systems. Despite having a greater specific im-
pulse, the ion thruster is of no value for overcoming the gravitational field of the Earth
at low altitude since the thrust generated is on the order of hundredths of a pound. For
deep space probes or the thrust required to perform the stationkeeping on a geosynchro-
nous orbit satellite, the thrust of an ion engine is quite sufficient.
Although ion engines have been available for a decade or so, these engines have not
been used in an operational propulsion system aboard a spacecraft. Instead, hydrazine
thrusters have supplied the stationkeeping requirements for geosynchronous orbit com-
munication satellites.
Because of the limitations of the launch vehicles used to deliver commercial com-
munications satellites to geosynchronous orbit, these satellites have been limited in
weight to approximately 726 kg (1600 Ib). Although the weight constraint may be relaxed
8
in future years, this constraint presently limits the communications capacity of a com-
mercial satellite.

Quantifying Benefits

The primary difficulty encountered in quantifying the benefits of space technology is

that the technology has never been deployed and thus cannot fully be assessed. With the
proper qualifications, however, it is possible to gain some information regarding the po-
tential impact of introducing new flight hardware. The ion thruster has been selected to
serve as an example for this procedure. A relatively conservative position must be
taken because the ion thruster is competing against hardware that has proven itself to be
reliable. Thus it is necessary to place the burden of proof on the ion thruster.
Since the cost incurred for a space mission is in dollars, it would simplify the ben-
efit assessment if dollar benefits could be specified. At present, most of the spacecraft
producing dollar benefits are those operated by Intelsat and its U. S. subsidiary Comsat.
The example developed here will be the comparison of ion thrusters with the hy-
drazine thrusters, that are presently used on Intelsat spacecraft. The expendable
launch vehicle used for Intelsat IV spacecraft imposes a 726-kilogram (1600-lb) weight
limit on the satellite. This weight limit places a limit on the revenue generating capa-
bility of each satellite. Increases in profits for Intelsat are transformed into rate re-
ductions for customers because of legal limits on the rate of return. However, the con-
cept of transforming a weight savings into increased profit is valid since this must be ac-
complished before rate reductions can be made. One method for obtaining the increased
profit would be to use fewer satellites. In 1974 the Atlantic region path consisted of
three Intelsat IV series geosynchronous satellites located at 19. 5°, 24. 5°, and 30° west
longitudes. The satellite at 30° west longitude serves as the major path satellite, han-
dling large volume traffic (ref. 4). Communications satellites are designed so that ca-
pacity use is normally reached in the last year of the satellite design life. But for a ma-
jor path satellite, full capacity use may be reached early in the satellite life; thus the
major path satellite is the prime generator of revenue. The major path system offers a
means of examining the effect of using the ion thruster. Because weight savings may be
transformed to additional communications capability, which in turn may generate addi-
tional revenue, the Intelsat IV over the Atlantic path offers a means of examining the
benefits of one application of an ion thruster system if it were to replace the hydrazine
thruster system presently in use.
The example presented here will provide information relating to the use of an ion
thruster in one specific application. The benefits of ion thrusters cannot be measured as
such since there are many possible applications of these thrusters. It would be very dif-
ficult to measure the value of using an ion thruster for a deep space probe because the
9
benefits of such a probe would be scientific knowledge (an intangible). However, it is
possible to measure the effectiveness when compared with conventional thrusters. For
those missions that produce revenue, a product oriented analysis may be constructed to
examine the potential value of a thruster system, but to measure the total benefits, all
possible products would have to be specified. A more reasonable approach is to develop
a lower bound estimate based on a single use. The analysis does not necessarily repre-
sent the manner in which an individual corporation would implement technology, but
rather one rational policy of many.
The example deals with a proposal to fund a flight program designed to gain infor-
mation about various aspects of spacecraft technology and to flight qualify and ion
thruster for north-south stationkeeping in geosynchronous orbit. There are many ben-
efits to be gained in terms of knowledge but the primary measurable benefit would be the
flight qualification of the ion thruster. Until this is accomplished, the thruster will prob-
ably not be used because of the risk associated with a thruster failing in orbit if deployed
on a commercial satellite.
Before proceeding with the analysis, a digression into the nature of decisions is in
order. Figure 1 is a summary of this discussion. Before one is faced with a decision,
there is the status quo or what presently exists. If a proposal is made to alter the status
quo, the decision must fall into one of three types of selection with respect to the new
proposal: either accept or reject the proposal; select one of several proposals; or se-
lect several of many proposals. The next consideration is whether there is a capital
constraint. If there is none, then from an economic point of view any proposal whose
benefits exceed costs is acceptable. The last consideration is that of dependence or
whether the costs and benefits of a proposal are invariant with respect to whether other
proposals are accepted or rejected.
The decision structure of this example is accept or reject. The decision information
process is depicted in figure 2. The economic analysis represents part of the informa-
tion required to make a decision. Assuming that the technical evaluation has been com-
pleted and the external factors have been considered, the remaining information gather-
ing function is the economic analysis. All assumptions will be stated explicitly in order
to establish the results. From that point on the results are irrefutable. Thus if one
chooses to counter the results of the analysis, the assumptions offer the only point of
attack.
When dealing with an accept-or-reject decision structure, one should specify two
scenarios, the status quo and that of the proposed alternative. In this way incremental
costs and benefits may then be specified. The decision rule associated with this struc-
ture is that the new proposal is economically attractive if the net present value (NPV) is
greater than zero. Other criteria may be specified (e. g., return on investment or an-
nual worth), but these criteria are variations of NPV. All three are ranking techniques
and will usually give the same rankings. The NPV equation is the preferred method here.
10
The equation form of net present value is

&+ D)
(1 + D)2 (1 + D)N

where Ct is the cost for period t, Bt is the benefit for period t, D is the discount
rate (or interest rate), and N is the project life. The value of D is usually dictated
by policy. The number presently recommended by the Office of Management and Budget
is 10 percent.
The status quo scenario is relatively simple for the example and has been stated
previously, namely, the use of hydrazine thrusters for north-south stationkeeping re-
quirements of geosynchronous orbit satellites. The scenario for the new proposal is
somewhat more complicated: it is desired, among other things, to flight- qualify ion
thrusters in orbit for north- south stationkeeping to prove their advantages over hydra-
zine thrusters. Other experiments would be carried on the spacecraft. The spacecraft
will be put in orbit on a piggyback flight.
Some launch vehicles carry payloads into orbit with large weight margins; that is,
the launch vehicle has more capability than is used. It is possible to orbit an additional
payload on these flights without incurring additional launch vehicle costs, because the
flights will be made whether the weight margin is used or not. There is an additional
cost for the payload integration. Thus piggyback flights offer a rather cost effective
method of performing flight experiments.
The assumptions and cost considerations for the analysis are as follows:
(1) There is a finite planning horizon of 11 years, 5 of which will be for the test pro-
gram and 6 for the useful Intelsat IV spacecraft life. The design life for these spacecraft
is 7 years, but there are some reasons to believe that this may not be achieved (ref. 4).
(2) The development and flight cost for the test program will be paid by the govern-
ment. After the proof flight of the ion thrusters, it is assumed that ion thruster tech-
nology will be used on only one commercial revenue generating satellite and never used
again, thus establishing a very conservative lower bound for benefits to one corporation.
The return to the government will be additional taxes on the marginal revenue generated.
(3) The weight savings resulting from the use of ion thrusters on the commercial
satellite will be used to add additonal operation and backup transponders thus increasing
the revenue generation capability of the satellite (ref. 5).
(4) A market exists for the additional capability. This assumption is justified by the
fact that growth of satellite communcations has been very large in recent years and is
projected to range from 15 to 35 percent annually for the next decade or so (ref. 4).
(5) The average monthly revenue of a Comsat telephone channel in 1971 was 2900.
There are about 750 channels per transponder. It is assumed that on the average 600

11
were utilized to generate the 2900 per channel month (ref.- 6). It is further assumed that
only half of this revenue will be generated by the additional communications capability in
the first year of use.
(6) Forty percent of marginal revenue is taxable at a rate of 48 percent. The 60 per-
cent not taxable is assumed to cover the hardware and operations costs of the additional
communications capability (ref. 6).
(7) This analysis is based on the Intelsat IV class of satellite.
(8) The operational cost of ion thrusters is commensurate with hydrazine thrusters.
(9) Fifteen million dollars is the estimated cost of the proof flight, including launch
vehicle costs (even though the experiment might be made on a piggyback flight). A 25
million dollar upper bound cost estimate allows for considerable overrun and risk.
(10) No benefits will be obtained from the experiment other than the proof test of the
ion thrusters. It is assumed that the test will be successful because of life testing con-
ducted previously.
(11) Ion thrusters yield a 16 percent increase in communications capability (two oper-
ational transponders, two backup) (ref. 5).
(12) Marginal revenue is generated according to the 1974 Comsat rate schedule
(ref. 6).
The flow of funds is shown in table I.
The net present value of the flow of funds was calculated for both program costs us -
ing the relationship for NPV stated previously. The NPV's are shown in table n. The
present value of the additional taxes to be paid on the new revenue at a 10 percent dis-
count rate was $19. 53 million. The present value of the expenditures for the low and
high program costs are 11. 37 and 18. 95 million dollars, respectively. Thus federal
money spent on the program would be recovered at a 10 percent interest rate in this case.

OPPORTUNITY COST

The opportunity cost does not represent a real cost but rather what is foregone by
not taking advantage of a good investment as quickly as possible. The opportunity cost
can be used to compare one investment with another to determine the benefits delayed by
not taking advantage of a particular investment.
The shaded region of figure 3 represents the bounds of the opportunity cost for high
and low costs of the project. The interpretation given to figure 3 is that if the investment
is delayed 1 year, approximately 8 million present value dollars will be foregone. It is
not an out-of-pocket loss, but rather the opportunity of gaining the return is lost because
of the delay.

12
 The point of the example is to illustrate that the use of ion thrusters for north-south
stationkeeping in a commercial communications satellite is economically productive even
under some rather conservative assumptions. In a more realistic situation where
thruster technology would be used more than once, the benefits to be gained would be
enormous.
In many cases the development of an economic worth for a program is much more
difficult than the example presented here; in some it is impossible. However, one can
always develop an estimate of the change in productivity that may be associated with a
program. It is unreasonable to expect the analysis to render a dollar value in all cases.
The use of incommensurable units may provide all the information required for a cost-
benefit analysis, and the use of the incommensurable unit has the advantage of not being
ambiguous as the case would be if the unit were transformed to a dollar analogue.
In the event that a quantitative cost-benefit analysis cannot be performed, a qualita-
tive analysis is acceptable. The requirement for a quantative cost-benefit analysis
where it is not feasible guarantees the decision maker an erroneous analysis. When
such a situation occurs one must conclude that either there is a lack of understanding of
the nature of the methodology or the program being analyzed or that the methodology is
being used for a purpose other than to examine the benefits and costs associated with the
program.
One final point should be made regarding the measurement of benefits. In recent
years economists have revived Marshall's concept of consumer's surplus as a means of
estimating benefits. Consumer's surplus is defined as "the amount over and above the
price actually paid that one would be willing to pay for a given amount of a commodity
rather than go without it" (ref. 7). The economist derives a price demand relationship
similar to the one in figure 4. If a consumer will purchase Q units of a commodity at a
price of P per unit, then the shaded area under the curve is an estimate of the excess
utility over and above the purchase price that he derives from Q units of the commodity
at a price P per unit. There are several reasons why consumer's surplus is meaning-
less when used to estimate the value of benefits of public expenditures. In the first place
it is impossible to derive a demand schedule which accurately describes the relationship
between price and quantity since data are unavailable. If it were possible to construct a
demand schedule for public expenditures for a particular program, the consumer's sur-
plus would not be a measure of benefits of the program but rather an indication of the
government's willingness to pay for the program.
Another reason the concept is inappropriate in welfare economics is that the concept
requires that the marginal utility of income be a constant. This in laymen's terms
means that there are no priorities in program selection. This fallacy needs no
commentary.
A situation often encountered in the expenditure of public funds is depicted in fig-
ure 5. The quantity Q in many instances is one; that is, the program will either be
13
accepted or rejected. The cost of the program might be a range from PQ to P^. Even
if the concept of consumer's surplus were meaningful as a measure of benefits, it would
be zero in all situations or decisions to either accept or reject a proposal.
A more detailed discussion of consumer surplus appears in appendix A.

SUMMARY AND CONCLUSIONS

Although there many problems associated with the use of cost-benefit analysis in
welfare economics, it is possible to use the methodology to obtain useful information re-
quired for the proper evaluation of public alternatives. The methodology is eclectic and
not fully developed. Much work must still be completed; most of it should be concen-
trated in the areas of incommensurables and intangibles. Of one thing we can be sure:
the future of all nations holds a continuing competition for resources, and thus, a con-
tinuing need for evaluative tools. When properly used, the methods of cost-benefit anal-
ysis or the more comprehensive tools of technology impact assessment will provide a
logical structure for gathering and evaluating the information required for meaningful de-
cisions in the public domain.

Lewis Research Center,

National Aeronautics and Space Administration,
Cleveland, Ohio, July 23, 1976,
644-04.

14
 APPENDIX A

CONSUMER'S SURPLUS

As stated previously, consumer's surplus is "the amount over and above the price
actually paid that one would be willing to pay for a given amount of a commodity rather
than go without it" (ref. 7). It is understandable that economists might wish to use
the concept to measure benefits because if society were to go from economic state A
(where Q« units are demanded at price PQ) to economic state B (where Q^ units are
demanded at price P^, then it would be a simple matter to calculate the value of the
savings to society. In some cases where productivity is increased, this calculation does
indeed.give the value of some of the benefits, but in welfare economics (or government
spending) such is not the case.
To adequately develop the concept of consumer's surplus in the public domain, it is
necessary to review the concepts of marginal utility and diminishing marginal utility.
The law of diminishing marginal utility states that as the amount of a good consumed in-
creases; the marginal utility of the good tends to decrease (ref. 8).
Utility is a measure of the usefulness of a commodity. If utility is negative, one
would not consume the good in question. The law of dimishing marginal utility states
that as more units of a good are consumed, the incremental utility placed on the next unit
is equal to or less than the previous unit. The situation is shown in figure 6 and table IE.
An interesting facet of the law of diminishing marginal utility is that it has never been
proven but rather is an empirical observation derived from original assumptions (ref. 7).
One of 10 requirements for calculating the consumer's surplus is that the marginal
utility of income be a constant. This is a rigid requirement which most economists who
use the concept dismiss with various assumptions, but which nevertheless does not dis-
appear with a wave of the hand. The implications of the requirement are that marginal
utility of income is independent of income (constant) and that total utility of income is
linear since marginal utility is the first derivative of the utility function. In general,
the requirement of constancy of marginal utility of income contradicts the law of dimin-
ishing marginal utility. The only point at which there is no conflict is the limiting value
of marginal utility of income, which must be zero. If and only if this condition of equi-
librium is achieved, consumer's surplus is a measure of these benefits of public
expenditures.
Consumer's surplus may then be obtained from the demand relation. Referring to
figure 7, consumer's surplus is the area ABC and is obtained from

/-A
-A
ABC = / f(P)dP
•s-o
'B

15
where f(P) is the price demand relation. (Note: Economists use the ordinate for the in-
dependent variable and the abscissa for the dependent variable. In this case, quantity
demanded is a function of price.)
The interpretation of the concept is that the shaded area in figure 7 represents the
cumulative value over and above the price that consumers place on D units of the com-
modity in question. To apply the concept to welfare economics, it is necessary to de-
velop the price demand relation that would portray the situation encountered in large
government programs. The first requirement is that marginal utility of income be a
constant (zero). This occurs only when the national debt is zero or when there is a sur-
plus. Otherwise incremental income could be used to reduce the debt, in which case the
marginal utility is positive. The first requirement has not been met in quite a long time.
But let us proceed further. The true situation encountered in government spending for
large programs is a scaled step function. The budget is the predetermined quantity and
Congress makes a go or no-go decision based on a range of estimated costs for a single
program. Congress approves one unit of the program over a range of prices. Even-
tually a price is reached to which Congress says no, and the program is not funded. The
demand schedule is portrayed in figure 8. If B' is the cost of the program, then the
consumer's surplus is given by the area

/-A'
A T B'D'C'= / f(P)dP

where f(P) in this case is a step function U(P). By performing the integration it is
found that

A'B'C'D' = A T -B T

Either A' and BT are equal or they are not. If they are unequal, then there is no prac-
tical way of determining A'. The argument in the event that A* and BT are equal is
that programs are funded with a budget limitation which is the cutoff point in figure 8. If
the budget is not exceeded, it is absurd to say that the amount not spent equals the ben-
efits of the program. Thus in welfare economics consumer's surplus is

/-A1 /-A'
A *B 'D 'C' = J f f (P)dP = J\ f (P)dP = 0

since B' and A' coincide; therefore, consumer's surplus is not a measure of the ben-
efits of public expenditures.

16
 When A' does not equal B', the cutoff cannot be determined and the program may
be funded. Since it is impossible to determine A \J£ f(P)dP is not convergent and
is therefore not defined. The most important aspect of this situation is that the program
can be priced but not valued. Therefore, demand is not an indication of the benefits to
be derived from the program.
Consumer's surplus is further muddled by Alfred Marshall's time varying interpre-
tation of his own concept. The Nobel laureate Paul Samuelson has rigorously developed
mathematical proofs which demonstrate that consumer's surplus is "superfluous"
(ref. 9). .
In addition to these difficulties there remains a question of whether consumer's sur-
plus is a stationary random process. In the strict sense such a process is not affected
by a shift in the time domain, so the statistics of such a process may be recovered de-
spite changes in the time reference. Since the problem has not been addressed, there is
no reason to assume that consumer's surplus is stationary. In fact, many examples
exist in the social and political domain which demonstrate that any surplus in these areas
varies in time and space. Thus what may appear to be a surplus to consumers in the
public domain often becomes a deficit before completion. Moreover, the conjured sur-
plus exists only in the analytical sense and bears little or no resemblance to the reality
the analyst has attempted to portray.
With so many disadvantages it would seem that the concept of consumer's surplus
should be abandoned in favor of more realistic and recoverable concepts of measuring
benefits in the public domain.

17
 APPENDIX B

WEIGHT AND AREA CHANGE IN THE SPACECRAFT

In the example it was stated that the replacement of hydrazine thrusters with ion
thrusters would result in increased communications capability for an Intelsat IV class
satellite. To achieve the increased capability, the satellite will require more trans-
ponder units, additional array area, and a larger battery (in size and area).
For the class of satellite in the analysis, 40 percent of the total weight (726 kg;
1600 Ib) is used for power and communications (290 kg; 640 Ib). The use of the ion
thrusters rather than the hydrazine results in a 7 percent overall weight savings of 58
kilograms (128 Ib). Any changes in the satellite configuration would be constrained to
less than 58 kilograms (128 Ib).

Array Size and Transponder and Battery Weights

The beginning of life (BOL) power in the spacecraft in 660 watts. There are 12 ac-
tive and 12 backup transponders. Each unit has one active and one backup transponder;
thus, each unit requires 660/12 = 55 watts of BOL power. The array area is 20.4 square
o
meters (220 ft ), and the array weight is 72. 6 kilograms (160 Ib). The array density is
3. 55 kilograms per square meter (0. 727 Ib/ft ), and the array power density is 32. 3
watts per square meter (3. 0 W/ftr) at the beginning of life. The additional array area
and weight required to add one unit is then

AA = 55 W/unit = 1 ? m2/unit (or 18 33 ft2/unit)

32.3 W/m2

and

AW = (3. 55 kg/m2)(l. 7 m2/unit) = 6. 04 kg/unit (or 13. 32 Ib/unit)

The change in drum size is given by the following equation:

AA = TrD AZ or AZ =
Trd

where d is the drum diameter and AZ is the incremental length. The diameter is 2. 38
meters, and the calculated AA was 1.7 square meters per unit. Therefore,

18
 17
AZ = ' 0.23 m/unit
7.792?r

The drum length increase for two units is 0.46 meter (18 in.). The launch vehicle used
for Intelsat spacecraft can accommodate this change. The drum length increase for the
Intelsat IV-A was larger than 0.23 meter. Both types are launched with Atlas Centaurs.
The difference in battery weight is obtained by multiplying the present battery weight
by this new capacity:

A(battery weight) = (39.92 kg)(0.167) = 6. 80 kg (or 15 Ib)

The total weight change is

Four transponders (2 active, 2 backup), kg (Ib) . ". . . 27.2 (60)

Additional array (two active transponders), kg (Ib) 12. 25 (27)
Battery weight, kg (Ib) 6. 80 (15)
Total added weight, kg (Ib) 46. 27 (102)

The added weight is 11. 8 kilograms (26 Ib) less than the 58 kilograms (128 Ib) saved by
using ion thrusters.

19
 REFERENCES

1. Dupuit, Jules: On the Measurements of Utility of Public Works. International

Economic Papers, vol. 2 (trans, from French), 1844.
2. Prest, A. R.; and Turvey, R.: Cost-Benefit Analysis: A Survey. The Economic J.,
vol. 75, Dec. 1965, pp. 683-735.
3. Boardman, Sanford L. : Improving the Accuracy of Benefit-Cost Analysis. IEEE
Spectrum, vol. 10, no. 9, Sept. 1973, pp. 72-76.
4. Comsat 1972 - Report to the President and the Congress. Communications Satellite
Corp. , 1972.
5. Free, B. A.: Chemical and Electric Propulsion Tradeoffs for Communication
Satellites Comsat Technical Review, vol. 2, no. 1, Spring 1972, pp. 123-145.
6. Comsat 1974 - Report to the President and the Congress. Communications Satellite
Corp., 1974.
7. Stigler, George J.: The Theory of Price. The MacMillian Co. , Inc. , 1966.
8. Samuelson, Paul A. : Economics. McGraw-Hill Book Co. ,. Inc. , 1955.
9. Samuelson, Paul A. : Foundations of Economic Analysis. Harvard Univ. Press,
1947.

20
 TABLE I. - FLOW OF FUNDS
Program year

1 2 3 4 5 6 7 8 9 10 11 Total

Millions of dollars

Cost of experiment:
Expected 3 3 3 3 3 15
Pessimistic 5 5 5 5 5 25
Revenue generated 21 42 42 42 42 42 231

TABLE E. - NET PRESENT VALUE OF ION

THRUSTER SYSTEM

[At a 10 % interest rate, the present value of taxes

on new revenue is $19. 53 million which exceeds
present value of both cost estimates. ]

Program cost Present value Net present value, a

of costs NPV

Millions of dollars

15 11.4 90.4
25 19.0 82.8
Present value of new revenue minus present
value of program costs.

TABLE m. - TOTAL UTILITY AND

MARGINAL UTILITY

Quantity consumed Total utility Marginal utility

0 0 --, A

1 4 —— • " O

2 7 —•—""""""

3
4 1U
5 10-— "

21
 Structure Criteria

Accept if net present value (NPV)> 0

Select one with maximum NPV

Find affordable sets

Capital constraint rank by NPV or
benefit cost ratios

Rank by NPV or benefit

No constraint
to cost ratios

Find affordable sets

Capital constraint
maximum NPV

Find possible sets

No constraint
maximum NPV

Figure 1. - Problem structure and decision criteria of new proposals.

Identification
of systems

1 Systems
synthesis

Data
collection
and
reduction
Identification
of data

Figure 2. - System engineering perspective.

22
 60xl06

Low investment cost

(15 million dollars)

High investment cost

(25 million dollars)

0 2 4 6
Delay, year
Figure 3. - Present value of opportunity cost of delaying use
of ion engines for north-south stationkeeping at interest
rate of 10 percent.

Consumer surplus

.Demand schedule

Quantity, Q Quantity, Q
Figure 4. - Price demand relationship. Figure 5. - Price demand relation-
ship for a government program.

23
 4 6 0
Quantity, Q
(a) Total utility. (b) Marginal utility

Figure 6. - Total utility and marginal utility.

C1

o>- B

Demand schedule

D
Quantity. Q Quantity, Q
Figure?. - Consumer surplus. Figure 8. - Demand schedule for large
government programs.

24 NASA-Langley, 1976 E-8796

NATIONAL AERONAUTICS AND SPACE ADMINISTRATION
WASHINGTON. D.C. 2O546 POSTAGE AND FEES PAID
NATIONAL AERONAUTICS AND
SPACE ADMINISTRATION
OFFICIAL BUSINESS
451
PENALTY FOR PUPATE USE ,300 SPECIAL FOURTH-CLASS RATE
BOOK

If Undeliverable (Section 158

POSTMASTER : Postal Manual) Do Not Return

"The aeronautical and space activities of the United States shall be

conducted so as to contribute . . . to the expansion of human knowl-
. edge of phenomena in the atmosphere and space. The Administration
shall provide for the widest practicable and appropriate dissemination
of information concerning its activities and the results thereof."
—NATIONAL AERONAUTICS AND SPACE ACT OF 1958

NASA SCIENTIFIC AND TECHNICAL PUBLICATIONS

TECHNICAL REPORTS: Scientific and TECHNICAL TRANSLATIONS: Information
technical information considered important, published in a foreign language considered
complete, and a lasting contribution to existing to merit NASA distribution in English.
knowledge.
TECHNICAL NOTES: Information less broad SPECIAL PUBLICATIONS: Information
in scope but nevertheless of importance as a derived from or of value to NASA activities.
contribution to existing knowledge. Publications include final reports of major
projects, monographs, data compilations,
TECHNICAL MEMORANDUMS:
handbooks, sourcebooks, and special
Information receiving limited distribution
bibliographies.
because of preliminary data, security classifica-
tion, or other reasons. Also includes conference
TECHNOLOGY UTILIZATION
proceedings with either limited or unlimited
PUBLICATIONS: Information on technology
distribution. used by NASA that may be of particular
CONTRACTOR REPORTS: Scientific and interest in commercial and other non-aerospace
technical information generated under a NASA applications. Publications include Tech Briefs,
contract or grant and considered an important Technology Utilization Reports and
contribution to existing knowledge. Technology Surveys.

Details on the availability of these publications may be obtained from:

SCIENTIFIC AND TECHNICAL INFORMATION OFFICE

NATIONAL AERONAUTICS AND SPACE ADMINISTR ATION

Washington, D.C. 20546