CHAPTER 8

BENEFIT-COST ANALYSIS FOR MUTUALLY

EXCLUSIVE ALTERNATIVES-PRINCIPLES AND
METHODS

Contents
8-1 Proper Specification Of Mutually Exclusive Alternatives ..................................................... 1
8-2 Basic Principles Underlying Benefit-Cost Analysis .............................................................. 3
8-2-1 Defining The Planning Horizon Or Analysis Period ...................................................... 3
8-2-2 Estimating Benefits And Costs ....................................................................................... 4
8-2-3 Setting The Opportunity Cost Of Capital Or Minimum Attractive Rate Of Return ........ 6
8-3 Benefit-Cost Analysis Methods ............................................................................................. 8
8-3-1 Discounted Versus Equivalent Annual Costs Or Benefits............................................... 8
8-3-2 Net Present Value Method ............................................................................................ 11
8-3-3 Benefit-Cost Ratio Method ........................................................................................... 12
8-4 Problems .............................................................................................................................. 14

The subject of benefit-cost analysis—whether termed that or an engineering economy study or an

alternatives analysis or a feasibility analysis—is not a new one. While the "names" have changed, the
substance and the content have remained unchanged for many years. Nor have the principles changed.

Benefit-cost analysis is little different from long-run economic planning as discussed in previous
chapters. The principles to be used are virtually identical, the major difference being that in benefit-
cost analysis only a subset of specific alternatives is analyzed. That is, benefit-cost analysis is
directed solely at the analysis of a specific set of technological or facility alternatives while long-run
economic planning as discussed earlier is directed at identification of the best alternative among a
much wider range of technological possibilities. Further, benefit-cost analyses will tend to be much
more detailed than the more general long-run economic planning and to take account of other
important aspects—such as variation in demand, either from hour to hour (or intra-temporally) or
from year to year (or inter-temporally).

8-1 PROPER SPECIFICATION OF MUTUALLY EXCLUSIVE ALTERNATIVES

The essential questions for benefit-cost analysis are directed at analyzing the economic desirability
of mutually exclusive alternatives. The term mutually exclusive implies that one and only one of the
alternatives can be undertaken. Importantly, when a list of mutually exclusive alternatives has been
specified for analysis, the null or "do-nothing" alternative should always be part of the process. The
null alternative is simply that of doing nothing. More precisely, it is providing no infrastructure
service and thus incurring no costs and accruing no benefits. When an entirely new facility is being
proposed, there seems to be no misunderstanding about the existence and treatment of the null
alternative. But when an agency is considering the improvement of services on an existing facility or
service, the analyst often and incorrectly specifies the null alternative as the "status quo" or existing
alternative. Sometimes this is done implicitly by virtue of simply analyzing the incremental benefits
and costs between the existing facility or system and its improvement. This is tantamount to
assuming that the existing service will be continued or improved, regardless of its economic merits.
Hendrickson and Matthews 8-1
 There is one exception to our definition of the null alternative which occurs occasionally in the
case of disinvestment in existing facilities. In such a case abandonment may involve shut-down or
demolition costs and, thus, the null alternative as defined above is not available. Examples would
include nuclear power plants (with high waste disposal costs) and mine closures (with acid drainage
issues). These cases are relatively rare, however, since most facilities will have some salvage or
residual land value to offset some or all of the demolition costs.

Also, the usual practice is to overlook consideration of the null (or do-nothing) alternative, as
previously defined. That is, most analysts treat the existing (or status quo) alternative as the null
alternative. In this instance the analyst has implicitly concluded that the benefits associated with the
use of the existing system do outweigh the costs of operating and maintaining it (to include the
foregone opportunity value associated with its fixed way and facilities). In practice, this is equivalent
to saying "No facility or system, once built and in operation, can be abandoned— regardless." In a
rapidly changing society and economy constantly bombarded with new technology and developments,
it would be hard to justify such a position. While we seldom question the wisdom of abandoning
obsolete plants and facilities if privately owned, we nonetheless see no contradiction when we take an
opposite view with respect to abandoning obsolescent publicly owned systems or facilities. The
difference in viewpoint is neither defensible nor understandable.

The proper specification of mutually exclusive alternatives can also be explained by example.
Suppose, for instance, you had attempted to carry out a benefit-cost analysis for BART (the San
Francisco Bay Area Rapid Transit System) prior to its construction. Most analysts would simply
analyze the extra costs incurred to build and operate BART as compared to the extra benefits
stemming from its operation. Other planners would argue that by virtue of building BART certain
additional highways would not have to be built, and thus that an additional benefit item for BART
would be the avoided costs of additional highway construction. In either case the full set of mutually
exclusive alternatives has been mis-specified. For the first the existing highway and transit facilities
(prior to BART) are implicitly assumed to be the null alternative and thus their benefits and costs
are ignored. In the second case two mutually exclusive alternatives are improperly intermeshed. In
contrast, the proper list of mutually exclusive alternatives (involving the existing highway and transit
system, BART, and new highways) might be as follows:
Null Alternative: abandonment of existing transit and highway system, as well as no BART and
no new highways.
Existing (or Status Quo) Alternative: existing transit and highway system without either
BART or new highways.
New Alternative 1: existing transit and highway system plus BART but no new highways.
New Alternative 2: existing transit and highway system plus BART and some new highways.
New Alternative 3: Improved operating and pricing strategies for the existing transit and
highway system.
In turn, the benefits and costs for each of these mutually exclusive alternatives (other than the null,
which has zero benefits and costs) should be evaluated.

The acceptance or rejection of one alternative is not dependent upon another. That is, if one does
not build one alternative, such as "New Alternative 1" above, then one does not have to build
highways or some other transit system instead. Nor does the construction of BART mean that some
new highways are necessarily avoided or "saved."

A story involving former Secretary Robert McNamara highlights this principle. As the story goes,
one day McNamara's son informed his father (a champion of benefit-cost analysis) that he had
saved a dollar by walking home from school rather than taking the bus, whereupon his father
asked why he hadn't decided not to take a taxi home and thus to save $10.00 instead.

Hendrickson and Matthews 8-2

Back to top

8-2 BASIC PRINCIPLES UNDERLYING BENEFIT-COST ANALYSIS

The benefit-cost analysis principles are designed to determine whether any of a set of mutually
exclusive alternatives is economically worthwhile and, if so, which of the alternatives is the most
desirable in an economic sense. Benefit-cost analysis methods are used to insure that (1) no project
will be considered economically acceptable unless its total net benefits are positive and (2) the project
having the highest nonnegative total net benefits is selected as the best.

The analysis methods are designed to take account of the time period in which cost
commitments are made or benefits accrued, and to insure that costs incurred or benefits
accrued during different time periods are placed on a commensurate value scale. In essence,
this is simply to recognize the "time value of resources" and the fact that resources
committed in the present or near future are more costly than those committed farther in the
future.

Back to top

8-2-1 DEFINING THE PLANNING HORIZON OR ANALYSIS PERIOD

All alternative projects must be analyzed for the same analysis period or planning horizon if we are
to properly account for reinvestment of any earnings or benefits accrued prior to the end of the
analysis or planning period, especially when one project may have a shorter terminal date than
another (whether replaced or not). Briefly, the analyst is concerned with (1) examining the benefit
and cost conditions which are expected to occur over the same analysis period or planning horizon for
all alternatives, regardless of when or whether certain capital items are to be replaced or terminated
early, and (2) determining whether any initial capital outlays should be made at the present and, if
so, which level of outlay is best based on expected future benefits and costs. For the first, if one
project among the set of alternatives is terminated early, the analyst must concern himself with the
other opportunities that are available for using the capital funds (which would have been used for
replacement) and what returns (i.e., benefits or revenues) can be accrued from them. Similarly, when
benefits or revenues are accrued in early years, either prior to the end of the analysis period or prior
to the end of any project's terminal date, the analyst cannot ignore the problem of properly accounting
for the reinvestment (or use) of the early-year benefits or revenues for the remainder of the analysis
period. Some of these matters will be clarified in later examples (Sections II-3 and II-4).

There are many ways of designating the analysis period and insuring that alternative projects are
properly compared with respect to the costs and benefits. For one, we can simply adopt an arbitrary
length of time over which the cost and benefit circumstances are to be analyzed. This period might
be chosen to seem suitable in terms of the service or physical lives of the facilities involved or in terms
of other appropriate aspects. For example, sixty years might be used for buildings or pipelines
whereas five years might be used as a planning horizon for wifi electronics.

For another, we may—again, arbitrarily—set the analysis period or planning horizon to be equal to
what we believe to be "the foreseeable future," or the period of time over which we can comfortably
or fairly reliably predict benefits and costs.

A third (and common but undesirable choice) is to set the analysis period equal to the least
common multiple of the physical or service lives of the alternatives being compared. (For instance, if
Hendrickson and Matthews 8-3
alternative 1 will be replaced at the end of 5 years, alternative 2 at the end of 15 years, and
alternative 3 at the end of 25 years, the least common multiple will be 75 years.) The problem with the
third approach is not so much with the length of time established but with the analysis method
usually (though admittedly not necessarily) utilized in combination with the "least common
multiple" approach. Specifically, it usually is employed when the costs are to be expressed in terms
of equivalent annual costs computed by multiplying the initial capital outlay (for each alternative)
times the capital recovery factor for its service or physical life. Accordingly, use of this approach
results in the implicit assumptions that (1) capital items will always be replaced at the end of their
initially designated service life (for however many times as are necessary over the "least common
multiple" life) and (2) the replacement costs of the capital items in future years will be exactly the
same as they were when the project was initiated. Inasmuch as these two assumptions appear to be at
odds with real-world considerations (such as changes in the future with respect to the service lives,
factor prices, and appropriate technology), a different and simpler analysis method seems more
appropriate. Briefly, it appears more straightforward to arbitrarily designate the analysis period
(according to our expectations about the "foreseeable future"), and then to estimate the year-by-
year cost outlays, whether for initial purchase or replacement and whether service lives change or
not, as they are expected to occur over the planning horizon. By so doing, future cost outlays can
easily reflect changes due to variations in factor prices or in the technology employed.

For cases in which discount rates are significant (as described below), the value of benefits and costs over
a lengthy period of time are lower, so the designation of a planning horizon is less important for the eventual
decision of the best alternative.

Back to top

8-2-2 ESTIMATING BENEFITS AND COSTS

Once the alternatives have been specified, estimates must be made of the year-by-year volume they
will experience and, in turn, the year-by-year costs and benefits associated with that pattern of usage.
Importantly, the values for these items should be measured in constant dollars and thus not reflect
inflation or deflation. It is possible to conduct an analysis in inflated or nominal dollar amounts.
However, use of such nominal dollar amounts typically requires applying an inflation factor during
the forecasting phase and then removing the inflation during the evaluation phase. It is
analytically simpler and conceptually clearer to simply restrict the analysis to costs and benefits as
measured in constant dollars.
The yearly costs and benefits calculated for a project should represent the actual benefits and costs
resulting from the project with respect to the viewpoint adopted for analysis. As an example,
suppose that the federal government is considering an investment, and so a national viewpoint is
adopted. In this case dollar amounts which do not represent a commitment of goods or services may
be excluded from the benefit and cost totals since they represent a transfer payment. No tax
payments should be included within the totals in this case unless such taxes represent the
payments for services associated with the construction, maintenance, operation, or usage of the
alternative being analyzed. If for example extra police were hired by the local government to provide
control or security for the alternative, and the costs thereof were to be borne by the local
government and paid out of city taxes, then any city tax payments to that extent should be included
as a cost item. By contrast, any local property taxes levied against motor vehicles probably
should be excluded. Chapters 4 and 5 discussed the definition and measurement of such benefit and
cost items.

The costs and revenues associated with borrowed money deserve particular attention in specifying
and determining the time stream of costs and benefits. Borrowed money has an economic
opportunity cost since these funds cannot thereby be used for alternative investments. Some projects
require borrowed money during the construction phase or later when replacing equipment and

Hendrickson and Matthews 8-4

repairing facilities, such funds then being repaid from later project revenues or from other internal
earnings (such as tax revenues) available to the firm or agency sponsoring the project.

TABLE 8-1. Example of an Alternative's Benefit and Cost Stream with and without
External Borrowing Costs
Cash Flow without Borrowing Cash Flow Combined Cash Flow
Borrowing
Year Benefit Cost Revenue Cost Benefit Cost
0 (Present) 0 600 600 0 600 600
1 512 100 0 412 512 512
2 512 150 0 237 512 387
3 400 250 0 0 400 250
4 560 800 0 0 560 800
5 800 500 0 0 800 500
Note: Dollar amounts are in thousands of constant dollars. A 6% borrowing rate is assumed for
financing.

As an example, suppose a 5-year project has an internal cash flow stream without considering
financing as shown in the first few columns in Table 8-1 and assume the initial outlays are to be
financed by borrowing $600K in year 0, an amount which is paid back in the next 2 years from
project net revenues including a 6% interest charge on unpaid balances. The combined cash flow for
the project, including external financing revenues and payments, is shown in the final two columns
of Table 8-1. While an additional capital infusion of $240K is required in year 4, in this example it is
assumed that the added capital is obtained from internal sources or retained earnings rather than
from external borrowing. Obviously, though, other financing schemes could be considered, thereby
leading to different combined cash flow streams; more attention will be devoted to this prospect in
Section 9-4 of Chapter 9.

To determine the economic acceptability of a project and the best among a set of mutually
exclusive projects, the benefit-cost analysis must consider the total cash flow stream of benefits and
costs, to include any borrowing costs and revenues if those are relevant to the entity conducting the
benefit cost analysis.
.
With the estimation of the time stream or benefits and costs for each alternative, we can define

C Xtt = expected costs or outlays (whether capital or operating) for alternative x during year t

and
B X}t = expected benefits from alternative x during year t.

The stream of costs and benefits over n years might be as shown in Table 8-2. In turn, two
other items of information must be specified: (1) the planning horizon or analysis period (n in
Table 8-2) and (2) the minimum attractive rate of return (MARR, or, say, "cutoff rate") or,
equivalently, the opportunity cost of capital.

Back to top

Hendrickson and Matthews 8-5

8-2-3 SETTING THE OPPORTUNITY COST OF CAPITAL OR MINIMUM ATTRACTIVE
RATE OF RETURN

For all benefit-cost analysis methods it will be necessary to specify an interest rate (or discount rate),
directly or indirectly. Often, and especially when using the internal rate-of-return method, the
interest rate to be specified is referred to as the "minimum attractive rate of return" (MARR), a
rate which reflects the interest which can be earned from other alternative opportunities which are
foregone. This term is equivalent to that used by economists, that is, the opportunity cost of capital
or an interest rate with reflects the earnings which will be foregone from other investment
opportunities if the capital is to be committed to a project in question. To a large extent the
specification of an "appropriate" interest rate, MARR or opportunity cost of capital is arbitrary and
thus open to question. As a consequence, it may be desirable to carry out the analysis for a range of
interest rates (which may reflect private market rates, on one extreme, and judgments about the
social rate of discount, on the other); this range may vary widely, perhaps as much as from 2 to 20%.
Given this wide range of possibilities, and the different judgments with respect to private versus
social rates of discount, it seems appropriate to discuss the basis of these possibilities and differences.

For projects involving United States federal dollars, the discount rate to be used is set by the
Office of Management and Budget (OMB). The OMB typically recommends a discount rate based
upon the federal market borrowing rate. Many private companies adopt a similar strategy, using
their own borrowing rate plus a few percentage points in many instances.

The discount rate can be thought of as an embodiment of a time preference for consumption
versus investment or saving. Resources can be consumed (or enjoyed) now by the current generation
or conserved for future use by either the current or future generations; put another way, programs
can be undertaken principally for the benefit of the current generation or they can be conducted
mainly in the interest of future generations. On the one hand the discount rate reflects the
strengths of peoples' individual preferences with respect to foregoing today's enjoyment until next
year, with a higher rate expressing a higher preference for consumption and enjoyment now relative
to that at a later date. Suppose, for instance, your personal rate of time preference was 10%. Such a
rate of time preference implies that a dollar of enjoyment or consumption 1 year from now is worth
10% less than a dollar of enjoyment now; or stated somewhat differently, you will be willing to
forego a dollar's worth of enjoyment today only if you can receive at least $1.10 worth of enjoyment 1
year hence. On the other hand the discount rate also reflects the productivity of alternative
investments. That is, if you forego consumption and enjoyment today, the resources which you
otherwise would have consumed could be invested in alternative opportunities and thus earn more
resources for more consumption and enjoyment in later years. If, for instance, the rate of productivity
Hendrickson and Matthews 8-6
for some investment is 15%, then each dollar invested now rather than consumed will result in $1.15
worth of resources being available for consumption or reinvestment 1 year hence. In turn, the market
discount or interest rate (in a perfectly competitive economy) would be determined by a balance
between individual's time preferences with regard to substituting future consumption for present-day
consumption and the productivity of alternative investments. Individually, and thus collectively,
people would continue to invest (and thus forego consumption until a later date) so long as the rate
of productivity for increments of investment was larger than their rate of time preference for
increments of present versus future consumption. The market discount rate would be determined
by that rate which just balances these two rates—that is, when the marginal rate of productivity (or
rate of productivity from the last dollar of investment) is equal to the marginal rate of time
preference (or rate of time preference for the last dollar of foregone consumption).

An alternative approach to using market rates of interest to determine a discount rate is to

decide upon a ‘social’ or ‘collective’ discount as a political decision. An often used example to
illustrate such collective action and social considerations would be the California redwood trees.
On the one hand purely economic considerations (in all likelihood) would indicate the wisdom of
chopping down the commercially important Sequoias and using the wood for homes and
furniture, in addition to lumber by-products. To do so, however, would mean that many future
generations as well as some members of the current generation would completely forego the
opportunity of enjoying the beauty, majesty, and grandeur of these giants for perhaps four to five
centuries. As a consequence, some argue that it is socially desirable to forego some of our economic
gains merely to preserve this option for those yet unborn and unable to express their preferences in
the marketplace.

Another matter to be considered in the selection of the "proper" interest rate is that of risk. Two
aspects of risk (and uncertainty) might be accounted for: (1) the uncertainties of estimating
accurately the future costs and benefits of a project and (2) aversion to assuming risk on the part of
individuals or organizations. While the uncertainties of cost and benefit prediction are often
implicitly accounted for by increasing the discount rate over what it would be with no risk or
uncertainty, a more appropriate way of handling the problem would be to incorporate the
uncertainties into the computation of the year-to-year estimates of cost and benefit and to use the
expected values. With uncertainties varying from year to year and generally increasing over time,
these adjustments should clearly be made year by year rather than on some arbitrary, constant, and
overall interest rate increase basis. Furthermore, this type of treatment will permit differentiation
between uncertainties (or risks) and the time value of money, rather than combine the two aspects on
some implicit and unidentifiable basis.

In addition to accounting for inaccuracies in estimating costs or benefits because of risk or

uncertainty, it may also be necessary to make adjustments for "risk aversion"—either in
calculating the true market value aside from "risk aversion" or in specifying a proper discount rate
where "risk aversion" is preferable. Risk aversion applies to the preferences of individuals (or firms)
with respect to undertaking investments with differing degrees of risk; some people, for example, are
"risk avert-ers" and are unwilling to invest in situations unless there is minimal or no risk, regardless
of how high the expected return might be, while others or "risk takers" would hesitate to invest in
situations unless there is at least some chance of a very large return (relative to the expected return).

More specifically, risk averters generally would be unwilling to undertake investments unless the
quoted return is higher than the expected return (or so-called risk aversion premium), while
risk takers might be willing to invest in situations having quoted returns less than the
expected rate if high returns were a reasonable possibility. A simple example of the latter
would be gamblers who are willing to place bets in a house crap game (e.g., on each roll
of the dice, those betting on "boxcars" or 12 receive a payoff of 30 to 1; however, they can
expect a payoff only once every 36 rolls, thus producing a negative expected value).

Hendrickson and Matthews 8-7

 As noted earlier, in practice, an analyst may have little choice in the discount rate used for
analysis: the applicable rate is often prescribed by a decision maker or a higher level of government.
In the absence of such restrictions, analysts would be well advised to conduct their analyses at
different discount rates to determine the sensitivity of investment choices to the discount rate.
Sensitivity analysis of this type is described in Chapter 10.

Back to top

8-3 BENEFIT-COST ANALYSIS METHODS

Among the many available benefit-cost analysis methods, the following will be discussed: (1) net
present value (or net present worth) method; (2) benefit-cost ratio method; and (3) internal rate-
of-return method. For these methods discounted benefits and costs will be used; however,
since some analysts use equivalent annual costs and benefits instead, the distinction between
the two measures will be discussed below. Also, a later section (14-1) will touch upon another type
of analysis procedure, the cost-effectiveness method.

Economists almost universally find the net present value method superior to all others, both
because it is simple and because it is unambiguous in indicating which alternative has the highest
economic potential. None of the others is so straightforward.

Throughout, the following terms and definitions will be used:

i = interest or discount rate (i.e., the minimum attractive rate of return or opportunity cost of
capital), expressed in decimal form
n = length of analysis period or planning horizon, in years
Cx,t = expected costs (capital or operating) to be incurred for project x during year t
Bx,t - expected benefits (or revenues) to be accrued from project x during year t
For convenience it will be assumed that the benefits or costs, B xt or C Xtt , occurring during year t will
be accrued or committed in lump sum at the end of year t. Typically, for other than the "do-nothing"
or abandonment alternative (i.e., the alternative for which C X t t - = B xt = 0 for all t), there will be some
initial cost outlays at the beginning of the first year (i.e., when t = 0), although the benefits or
revenues will not usually (but not necessarily) begin to accrue until at least a year later (i.e., when t >
1). In any case, though, the formulation is perfectly general and will apply to all situations. The cost
and benefit streams during the n-year planning horizon for any project x will look as shown in Table
8-2.

A year-by-year cash flow tabulation of the benefits and costs for all alternatives (where, say, there
are m alternatives and thus x varies from x = 1, 2,..., m) could be displayed in much the same
manner as that indicated for project x in Table 8-2, and then ordered for analysis. Commonly, the m
alternatives are ordered in ascending order such that the alternative having the lowest initial cost in
year t = 0 is the first or alternative 1 (i.e., x = 1 corresponds to the lowest initial cost alternative), the
alternative having next lowest initial cost in year t = 0 is that for which x = 2, and so forth, until
the alternative having highest initial cost in year t = 0 is alternative m (i.e., x = m for it).

Back to top

8-3-1 DISCOUNTED VERSUS EQUIVALENT ANNUAL COSTS OR BENEFITS

Calculations for all benefit-cost analysis methods can be carried out while using either discounted or
equivalent annual costs or benefits. While the final decisions—as to which alternatives are
Hendrickson and Matthews 8-8
worthwhile and which is the most economical—will not differ for the two measures of costs and bene-
fits, numerical results will not be the same. However, since some analysts and textbooks stress the use
of one method, and others emphasize the second, it seems useful to review the two techniques. At the
outset, though, we should note that for most benefit-cost analysis the use of discounted benefits and
costs will be simpler and require fewer calculations.

With either discounted or equivalent annual costs and benefits, the objective is simply to account
for the time value of money. That is, the objective is to account for the fact that a dollar of
resources expended today is more costly than one expended in later years. Similarly, a dollar of benefit
accrued today is more valuable than one to be accrued in some future year.

The first and simplest way to account for the time value of money and to place all present and
future costs or benefits on a commensurate value scale is to discount them to their present value or
present worth (i.e., their value today or at year t-0). Thus, the discounted value of the costs or
benefits occurring in some year t for alternative x would be as follows:

[PVC x , t ] t = (P\F, i, t)C x,t = {pwf’ i,t )C X,t (8-1)

and

[PVB x , t ] t = {P\F, i, t)B Xft = (pwf’ i,t )B, >t (8-2)

where [PVC X,t ]t is the discounted present value of the costs during year t, [PVB x,t ]t is the
discounted present value of the benefits during year t, and pwf’ i,t or (P\F, i, t) is the (single
payment) present value factor for a cost or benefit occurring during year t when the interest rate
(or opportunity cost of capital) is i. The present value factor is:

This formulation assumes annual compounding of interest at a per annum rate of i (though i is
expressed in the formula as a decimal fraction).
For a project x the discounted values for a stream of costs and benefits occurring over n years
such as those shown in Table 8-1 would be

where [TPVC Xfn ]i and [TPVB Xtn ]i are the total discounted costs and benefits, respectively, for an n
-year period and interest rate i.

The second way to account for the time value of money and to place all present and future
costs or benefits on a commensurate value scale is to convert the stream of costs and benefits
to an equivalent annual cost or benefit figure. Essentially, the problem is to determine an equal
annual amount which is exactly equivalent to either the year-by-year cash flow stream or the
discounted value as computed by Equation (8-1), (8-2), (8-4), or (8-5). This is entirely analogous
to the procedures used to determine the equal payments which are charged by banks or credit

Hendrickson and Matthews 8-9

 firms for mortgages or installment loans, though usually the creditor will make monthly rather
than annual payments—the latter being the case here.

The most straightforward technique for determining the equivalent annual cost or benefit
for a stream of benefits and costs would be to multiply the total discounted costs or benefits,
as computed with Equation (8-4) or (8-5), times the capital recovery factor. Thus,

where crf i,n or (A\P, i, n) is the capital recovery factor for an interest rate i and an n-year analysis
period, and [EAC x , n ]i and [EAB x,n ] i respectively, are the equivalent annual cost and equivalent
annual benefit for n and i.

The capital recovery factor used in Equations (8-6) and (8-7) can be computed as follows:

However, the capital recovery factor shown in Equation (8-8), as well as its use in (8-6) and (8-7),
applies only to equivalent annual payments and to annual compounding of interest.

An example using both the discounted and equivalent annual payment methods of representing a 5-
year cash flow stream for an interest rate of 5% is shown in Table 8-3. Essentially, we can represent
the 5-year cash flow stream of costs shown in column 2 either by the total discounted cost of $542.13K
in a lump sum now or by an equal amount of $125.22K in 5 successive years beginning 1 year from
now. That is, if we determine the present value of the five equal annual payments, as shown in
column 5, the total discounted value is $542.14K, the small $0.01K difference being due to rounding.

TABLE 8-3. Discounted and Equivalent Annual Costs for a 5-Year Cash Flow Stream and an Interest Rate
of 5%
Cx,t [PVCx, tt ] b% [EACX,t ]5% Present Value
Costs Incurred Discounted or Equivalent Annual of Equivalent
Year at End of Year Present Value Cost for the Annual Cost
t t for Alternative x of C Xjt 5-Year Stream in Year t
0 $200K $200K 0 0
1 60K 57.14K $125.22K $119.26K
2 70K 63.49K 125.22K 113.58K
3 80K 69.11K 125.22K 108.17K
4 90K 74.04K 125.22K 103.02K
5 100K 78.35K 125.22K 98.11K
Total $542.13K $542.14K

Hendrickson and Matthews 8-10

The two methods for placing present and future costs or benefits on a commensurate value scale, as
described above, are entirely general and will apply to any stream of costs or benefits. It will apply
whether or not there are any costs or benefits occurring in year 0 or now, as well as in situations
when the year-to-year costs or benefits vary, as they did in Table 8-3.

Back to top

8-3-2 NET PRESENT VALUE METHOD

With the net present value method the stream of benefits and costs are discounted to their present
value or present worth (that is, to their value now) and then netted to determine the resultant net
present value. Thus, for any alternative x the net present value for the n -year analysis period when
the interest rate is i or [NPV Xtn ]i would be

[NPV x, n]i = [TPVB x , n ] i -[TPVC x , n ] i (8-10)

Note: (P\A, 5%, 10) = uniform series present value factor = 1/(A\P, 5%, 10). Also, it can be shown
that (A\P, 5%, 10) + (P\F, 5%, 5)(A\P, 5%, 10) = (A\P, 5%, 5). To show this, substitute
10
for the5
capital recovery factors using the formula given in Equation (8-8). Hint: (1 + i) - 1 = [(1 + i)
+ 1][(1 + i)5 - 1].

The net present value must be determined for each alternative from x = 1, 2,..., m. All
alternatives which have a nonnegative net present value can be regarded as economically feasible
while the best alternative will be that having the highest nonnegative net present value.

The method is straightforward and will guarantee that public or private agencies maximize net
social benefits, however these are measured and for whatever planning horizon or interest rate is
chosen. Where the opportunity cost of capital (i.e., the interest rate for other foregone investments)
is unknown or subject to question, the calculations can be repeated for different rates and the final
results compared for similarities or differences in ranking or acceptability. Also, if the net present
value increases when moving from a lower initial (or total) cost alternative to a higher initial (or
total) cost one, then one may be certain that the discounted incremental or extra benefits outweigh
the discounted extra costs; otherwise the net present value would not have increased.

Back to top

Hendrickson and Matthews 8-11

8-3-3 BENEFIT-COST RATIO METHOD

In some sense the benefit-cost ratio method— when properly applied—is little different than the net
present value method. The identical benefit and cost measures are used to compute the benefit-cost
ratios and proper interpretation will invariably lead to the same decisions about which alternatives
are economically feasible and about which one is the best, economically speaking. The only
differences are that extra computations are required for the benefit-cost ratio method and that proper
interpretation of the ratios is confusing in some cases.

To begin, one may order alternatives for analysis in a number of different ways, though we will
assume that alternatives are placed in ascending order with respect to the initial-year cost or C x>0 for
x = 1,2,..., m. After the alternatives are ordered, there are two steps in applying the benefit-cost
ratio method.
The first step is undertaken in order to determine whether any alternative is economically
worthwhile. Simply, we determine the benefit-cost ratio for the lowest-ordered alternative as
follows:

[BCRhn]i is the benefit-cost ratio for alternative 1 over an ft-year analysis period for an interest rate i.
If the ratio is equal to or larger than 1, then alternative 1 can be regarded as acceptable. That is, if
the ratio is at least as large as 1.0, then we know that the total discounted benefits are at least as
large as the total discounted costs and that the net present value is nonnegative. If, however, the
ratio for the lowest cost alternative is less than 1.0, then it will be rejected and the ratio will be
computed for the next higher initial cost alternative. This process is repeated until we identify the
lowest ordered alternative having a benefit-cost ratio equal to or greater than 1.0. If all alternatives
have ratios less than 1.0, then all should be rejected, economically speaking. Let us assume,
however, that alternative x is the lowest-ordered alternative having an acceptable benefit-cost ratio;
thus,

Put differently, we know that it is better to undertake alternative x than any of the lower-ordered
alternatives.

The second step is to determine whether or not it is worthwhile to undertake an even higher-
ordered alternative. That is, we must justify any additional increments of cost. For this purpose we
compute the incremental benefit-cost ratio for the additional expenditures. Pairwise comparisons
are made between successively higher-ordered alternatives, starting with the lowest-ordered
alternative which is acceptable, as indicated by Equation (8-13). Specifically, for an M-year analysis
period and an interest rate of i, the incremental benefit-cost ratio for the increments in benefit and
cost between the lowest-ordered acceptable alternative x and the next higher-ordered one will be
[IBCRx/x+,ln]i, where x/x + 1 simply means alternative x + 1 as compared to alternative x, as follows:

If the incremental ratio is equal to or larger than 1.0, then alternative x + 1 will be more
desirable than alternative x and, in turn, the incremental ratio will be computed for alternative x + 2
Hendrickson and Matthews 8-12
as compared to x + 1. On the other hand, if the incremental ratio for alternatives x and x + 1 is less
than 1.0, then alternative x + 1 will be rejected and the incremental ratio for alternative x + 2 as
compared to alternative x will be computed, As we shall later see, however, the above rules strictly
apply only when both the numerator and denominator are positive.

TABLE 8-4. Example for the Benefit-Cost Ratio Analysis Method Benefits and
Costs of Alternatives (in $1000s)
Alternatives
1 2 3 4
Cxfi 50 55 60 65
[TPVB Xfn ]ia 175 258 360 320
[TPVC x ,n\ib 180 200 300 250
[BCR Xtn]ie 0.97 1.29 1.20 1.28
[NPV x , n ]id -5 58 60 70
"Present value of benefits.
6
Present value of costs.
c
Benefit cost ratio = [TPVB X , n] t /[TPVC x , n ] £ .
d
Net present value = [TPVB Xtn ]i - [TPVC x Ji.

Pairwise comparisons are continued until we identify the highest-ordered alternative which
satisfies both of the criteria set forth by Equations (8-13) and (8-14). Importantly, do not use the
highest benefit-cost ratio—as computed by Equation (8-13)—as the criterion for choosing the best
alternative. Choosing the alternative with the largest ratio results in maximizing the return per
dollar of cost, but this is not the same as maximizing the net present value or benefits less costs.

For incremental ratios one should not be confused by the fact that the numerator and/or
denominator of the incremental benefit-cost ratio will sometimes be negative but should recognize
that in such cases the criterion (of being at least as large as 1.0) can change. Specifically, when both the
numerator and the denominator are negative, the next higher-ordered alternative is preferable to the
lower-ordered one whenever the incremental benefit-cost ratio is equal to or less than 1.0, but when
just the denominator is negative, the higher-ordered alternative is always preferable; when just the
numerator is negative, the lower-ordered alternative is always preferable. The example shown in
Table 8-4 will highlight these principles.

Four alternatives have been analyzed and have initial-year costs, as well as total discounted costs
and benefits, all as shown in Table 8-5. The analysis should proceed as follows:
1. The benefit-cost ratio for the lowest-ordered alternative (x =1) is 0.97 or less than 1.0 and,
therefore, it is rejected, according to the criterion in equation (8-12);
2. The benefit-cost ratio for alternative 2 is 1.29, thereby indicating
that alternative 2 is the lowest-ordered acceptable alternative;
3. The incremental benefit-cost ratio for alternative 3 as compared to
alternative 2 is equal to ($360 - $258) divided by ($300 - $200) or
1.02; thus, alternative 3 is more desirable than alternative 2, using the
criterion shown in equation (8-14);
4. The incremental benefit-cost ratio for alternative 4 as compared to
alternative 3 is equal to ($320 - $360) divided by ($250 - $300), or
-$40/-$50, which is equal to 0.8; however, since both the numerator
and denominator are negative and since the incremental ratio is less
than 1.0, alternative 4 is preferable to alternative 3. (More simply, in
this case the benefits were reduced when we moved from alternative 3
to alternative 4, but the costs were reduced even more, thus resulting
in overall economies and a gain in net benefits.)
Hendrickson and Matthews 8-13
Overall, alternative 4 is the highest-ordered alternative for which both sets of ratios satisfied the
criteria and therefore is the best or most economically acceptable alternative. Moreover, we should
note that the alternative with the highest benefit-cost ratio (alternative 2 with a ratio of 1.29) is
clearly not the best alternative. In fact, to have used the highest benefit-cost ratio as the choice
criterion would have resulted in foregoing two alternatives which would have brought about higher
total net benefits. Note, further, that both the net present value and benefit-cost ratio analysis
methods would have resulted in the choice of alternative 4 as the best alternative, the difference
between the methods being simply that the latter requires more computation and is less
straightforward in application. Properly and fully applied, though, the benefit-cost ratio method will
always identify the most economical alternative.

Back to top

8-4. Problems

P8.1 Suppose an agency has an opportunity to make a new investment with benefits and costs over a six
year planning horizon as follows:
Year Benefits Costs
0 0 600
1 200 150
2 250 180
3 300 200
4 360 250
5 350 200
6 360 200
a. At a 3% (0.03) discount rate, is this investment economically worthwhile?
b. At a 5% discount rate, is this investment economically worthwhile?
c. What is the Equivalent Uniform Annual Value of this investment at a 5% discount rate?

P8.2 Suppose a farmer is considering a bid to turn some land in a wind farm for generating electricity.
The farmer would provide the land in a twenty year contract and receive a portion of the electricity sales
from the wind turbines. The wind farm would preclude continuing to use the land for farming.
a. What alternatives might the farmer define to perform a benefit/cost analysis of this bid?
b. What are the primary benefit and cost categories for each alternative?

Back to top

Hendrickson and Matthews 8-14