UNIT 5: BUDGET ANALYSIS
I.
BUDGET APPROVAL & THE LEVEL OF GOVERNMENT
A. The Differing Patterns of Executive Budgeting
B. Legislative Characteristics of State & Local Budgeting
II.
PROGRAM ANALYSIS
A. Analyze That
B. Tools & Techniques
1
10/26/2019
UNIT 5: BUDGET ANALYSIS
I.
BUDGET APPROVAL & THE LEVEL OF GOVERNMENT
C. The Differing Patterns of Executive Budgeting
A key difference between the development of executive budgeting at the state level
from the federal is the presence of multiple executives with independent power bases. This
complicates both the preparation and approval processes.
Forty-four of the fifty U.S. states give their governors some form of line-item veto
power; Indiana, Nevada, New Hampshire, North Carolina, Rhode Island, and Vermont
are the exceptions. The Mayor of Washington, D.C. also has this power. The formal
powers of governors vary from state to state, and this includes their power over budgeting.
In addition to formal power, political culture plays a role in determining actual power.
North Carolina, for example, is toward the bottom of the formal powers given a governor;
yet the state has a history of strong, powerful governors.
IGR further complicates budgeting at the state level. States are both recipients and
providers of intergovernmental transfers. Much of a state's budget, for example, may go to
pay schoolteachers who are employees of a local school district.
Local Patterns (multiple executives, fragmentation of power, IGR; discuss rest with
legislative below)
D. Legislative Characteristics of State & Local Budgeting
1. State Patterns (Bicameral w/ 1 exception, great variation in quality of staff, annual
& biannual patterns)
2. Local Patterns (strong mayor, weak mayor, city manager, commission)
II.
PROGRAM ANALYSIS
Budgets are the most concrete plan produced by any unit of government. You may
have a collection of marvelous ideas concerning where you want to be in the future and how
you intend to get there, but until you’ve actually committed resources in a budget, they
remain dreams.
At the same time budgets should themselves be reviewed in terms of how they do or
do not advance the programs desired by the relevant community.
A. Analyze That
There are two ways of looking at program evaluation: process and impact
evaluation. Budgeting is involved in both, as we discussed in the last unit. The line item
budget remains a key component of process evaluation.
2
10/26/2019
B. Tools & Techniques
Simply, this simply implies that the benefits outweigh the costs for a program to be
adopted. A Pareto Optimum decision guide argues that a policy should leave both parties
better off, or one party better off and no one worse off. This would have those who benefit
from a policy always reimburse any losers. As long as you can do this and still have a net
gain, it is a good policy. This is impractical in actual program design, so the generally
accepted rule is that a policy is good as long as winners could pay off losers in theory, and
then the program in fact does compensate the easily identified losers. (Kaldor-Hicks
principle, named after two British economists) In other words, as long as benefits outweigh
gains, and no one is made seriously worse off, and society has other mechanisms to deal
with inequities, the program is good.
Using the Kaldor-Hicks principal, an agency or government should select the option that
produces the greatest net benefit. This is the principle behind benefit/cost analysis. Let's
assume that an agency has the following programs under consideration:
Gains to
Loss to
Costs to
Net
Benefit/Cost
Project
Consumers
Suppliers
Taxpayers
Benefits Ratio
A
200
50
100
50
1.50
B
200
50
200
-50
.75
C
450
50
300
100
1.33
D
100
10
100
-10
.90
E
650
0
500
150
1.30
_________________________________________________________________
Given this data, what is the optimum budget for the agency? You would simply
select those projects where the benefits were larger than the costs: A, C, and E. (A [100] C
[300] E [500] = 900—remember, the cost to taxpayers is the budget cost, but loss to
suppliers should also be factored in as an equity issue). The benefits are A (200) + C (450)
+ E (650) = 1,300 for a benefit/cost ratio of 1.44.
What if you have an allocation of $1,000,000? Many of you will say that doesn’t
change anything, as the other proposed projects have a negative benefit/cost (ratio below
1.0). You should simply not spend the remaining $100,000. That is the responsible thing to
do, right? It also means next year’s budget will be cut by the $100,000 you didn’t spend.
Thus, most agencies will add in project D so they can spend the entire $1,000,000. When
you add up the benefits and costs, you still have a net benefit/cost ration (costs = 100 + 300
+ 500 +100 = 1,000; benefits now equal 200 + 450 +650 + 100 = 1,400 for a benefit/cost
ratio or 1.40). The agency will spend the $1,000,000 and report an overall benefit/cost ratio
of 1.4, and feel very virtuous. This is one of the systemic limits on rationality I didn’t
discuss earlier.
A simple, yet very useful device in a first-cut benefit/cost analysis is the benefit/cost
matrix. You divide the matrix between benefits and costs, and then between real (tangible
and intangible) and pecuniary costs. Tangible benefits and costs are those that you can
measure and place a value on; intangible cost and benefits are more subjective. Pecuniary
costs and benefits are those costs and benefits that are transferred to others, such as spillover
effects. We will go back to our dam problem for an example. We need to know the benefits
3
10/26/2019
and costs of the dam, one key component of which is irrigation water. For the purpose of
this illustration we will further assume that the dam will cover a valley that is currently a
wilderness area popular with hunters and fishermen. You construct a benefit/cost matrix as
indicated below:
Benefit/Cost Matrix for Irrigation Water
Real
Direct
Tangible
Intangible
Indirect
Indirect
Tangible
Pecuniary
Benefits
Increased farm output
Costs
Cost of dam
Beautification of area
Loss of wilderness
Reduced soil erosion
Diversion of water from other uses
(opportunity costs)
Relative loss to hunting & fishing
suppliers
Relative improvement in position of
farm industry
In an actual analysis this would be much more complex.
Benefit/cost analysis is a key component of any policy study, but it should never
make the decision—it should be one factor in making a decision. For one thing, it is usually
I took an exam in my master’s program where I had to use a decision tree
to decide whether to buy a gas or diesel powered van for a meals-on-wheels program.
The key variable that was uncertain was how the relative cost of the two fuels would
change over the life of the van. After working the problem through it turned out that the
diesel-powered van had a higher expected value than the gas-powered. I showed the
computations, then answered the question by stating that I’d buy the gas-powered van
anyway, because I didn’t like the way the diesel van sounded and I thought its exhaust
had a funny smell. I got extra credit for the answer.
impossible to attach dollar figures to each important aspect of a proposed policy—although
you may attach them to more variables than you might at first think using shadow pricing,
etc. Ethical issues must be factored in, along with a number of other factors. We’ll discuss
this a bit more below.
A. A Modified Goeller Scorecard: Cost/Benefit Analysis Made Practical
A Goeller scorecard is a widely-used took to select between different alternatives. It
is a relatively simple device—the hard part is obtaining the data to plug into it, something
we’ll gloss over a bit in this course. I use a modified Goeller scorecard for pedagogic
4
10/26/2019
purposes. Called a weighted criteria matrix, it is useful for the type of quick and dirty
analysis that public managers often must do in their careers. The matrix is shown below:
_________________________________________________________________________
Selection Criteria
(Ranked by Weights)
Mission Element #
Alternative #
1
2
3
4
Let’s say that you’re putting together a program to educate children of migratory
farm workers, most of whom are Hispanic. You’ve developed your plan, but you’ve come
up with a problem: you don’t have enough people who are eligible to serve as instructors in
the program who can speak Spanish. You have been told by your superiors (a term you use
loosely) that you have to go with your current pool of people, even if it means that they have
to learn Spanish. You conduct a blue-sky brainstorming session with your colleagues, and
come up with the following list of ways to teach your potential instructors to speak Spanish:
1. enroll them in Spanish courses at VSU
2. have them marry a Spanish speaking spouse
3. immerse them in Spanish by sending them to Spain for a year
4. put them through a Berlitz course
5. have them enroll in a UGA correspondence course
6. have them sleep with Spanish lessons playing under their pillows
You put these into your matrix, along with your selection criteria. Two selection
criteria that must be included in any study are cost and effectiveness. You come up with
others that are relevant to you, and you then weight them by importance. This gives you the
following matrix:
5
10/26/2019
Selection Criteria
(Ranked by Weights)
Mission Element # X
Alternative #
1
2
3
4
5
6
Cost Political
Effectiveness Time
Total
Acceptability
(10)
(7)
(9)
(8)
4 (40)
10 (70)
8 (72) 7 (56)
238
X
X
9 (81)
1 (8)
89
1 (10)
10 (70)
10 (90 5 (40)
210
5 (50)
10 (70)
8 (72) 10 (80)
272
3 (30)
10(70)
4 (36) 8 (64)
200
10 (100)
10 (70)
1 (9)
1 (8)
187
The number in parentheses under each criterion is the weight you’ve given that
criterion on a scale of 1 to 10. You then weigh the various options on a scale of 1 to ten,
with 10 being the highest in most instances, but not for cost or time. Since you want the
lowest cost and time, you invert your values, making the least costly 10 and so forth. You
then simply multiply the two weights, total them, and come up with a “best” alternative.
You will enroll them in a Berlitz program.
Note that I placed an “X” in some of the cells. This indicates that the alternative is
beyond the acceptable parameters on that criterion. If that alternative still comes out well,
you may want to see if you can modify the criterion parameters. For example, if you have
an alternative that looks good on each of your criteria, but is outside your budget you may
wish to use this matrix to try to get the money placed in your budget.
One key concept everyone should be familiar with when completing budgets is
present value. We’ll go over this briefly, using a reclamation project as our example.
The essence of discounting is that it reduces a stream of costs and/or benefits to a
single amount, termed the present value, by using the method of compound interest. The
calculation of present value is thus a calculation of net benefit that takes explicit account
of the timing of costs and benefits. The benefit-cost criteria that were developed in the
last chapter may then be applied, with the term net benefit understood to mean "the
present value of the discounted stream of net benefits."
The basic rationale is this: everyone, under almost any circumstances, would prefer
$1 now to $1 a year from now. A sum of money in hand is worth more than a promise of
the same sum at a specified time in the future, because the money may be invested so as
to produce earnings in the intervening time. This is true whether the money is to be
invested by an individual or by a business, or by a government that must raise the
necessary funds through taxation or borrowing, although to be sure the uses for the
money will differ. Take the simplest case, that of an individual. Suppose Mrs.
Robinson is to be paid $100 a year from now. There is some lesser sum that she can
invest today, for instance by depositing it in a savings bank, that will accumulate to
$100 by the time the year has passed. This lesser sum is the present value of the
payment a year hence of $100.
Note carefully that we are not saying anything about risk; we assume no risk is
involved here and that Mrs. Robinson's $100 is as certain as anything can be. Rather, we
6
10/26/2019
are saying that having to wait for payment means forgoing the income that could be
earned on the money in the meantime. In other words, waiting carries a cost in the form
of a lost opportunity. This is not to suggest that risk should be ignored in analyzing a
project-far from it-but merely that this is not where it enters the analysis. The problem of
decision making under conditions of risk and uncertainty will be considered later in this
book. In the real world, uncertainty and pure waiting are often entangled; it is important
that we understand that they are separate phenomena.
Although the calculation of present value ordinarily involves flows of dollar
amounts, the method is applicable to any stream where all the returns are measured in the
same units. (Thus, we could discount recreation days or acre-feet of water, or additions to
food stocks.) We discuss below both the actual mechanics of the calculation and the
theoretical and practical problems that arise in the choice of a discount rate.
The Mechanics: The Arithmetic of Present Value
We have just seen that waiting involves an opportunity cost; naturally the
question is "How much?" What is the present value of a payment one year from now of
$100? Clearly it depends on our assumption about what return a sum of money invested
today will earn over the coming year. Suppose it will earn 5 percent; if we invest $100
today, it will accumulate to $105 by a year from now. If we invest $90 now, we will have
$94.50 then. And there is some amount X that we can invest today at 5 percent interest
that will give us exactly $100 a year from now:
X(1 + .05) = 100,
or
100
X = 1.05 = 95.24.
We say that $95.24 is the present value of $100 payable a year from now at an annual
discount rate of 5 percent. If the rate that could be earned were
10 percent, the present value would be:
$100/ 110 = $90.91.
At this point you may ask, "What is this discount rate? It looks like an interest rate!"
Indeed it does; interest rate and discount rate are different names for what is
arithmetically the same thing. It is customary, however, to use the two terms in different
contexts. When we start with a sum of money and calculate the earnings on it forward
into the future, we speak of the interest rate. When we start with a given sum at some
time in the future and calculate back in time to the present to determine the present value,
we speak of the discount rate. As far as the arithmetic is concerned it's almost a
distinction without a difference. We saw in the section on difference equations that when
we compute compound interest we write:
7
10/26/2019
S1 = (1 + r)So
where S1 is the sum of money after one year, r is the rate of interest, and So is the initial
sum. When we discount, we simply move the discount factor (1 + r) over to the other side
of the equation and write:
S0 = S1/ 1+r
Although r is still a rate of interest, we now call it the discount rate.
Compute the present value of obtaining $100 in year 5 at a discount rate of 5%.
S0 = 100/(1.05)5 = 100/ 1.276281562 = 78.35
A proposed reclamation project has an estimated cost of $10 million and an effective life of
ten years, with no residual value. Annual operating costs are estimated to be $500,000 and
annual revenues are estimated to be $2 million. The construction costs will be paid in a
lump sum at the beginning of the first year of operation.
Prepare:
a) a cash flow schedule
b) discounted cash flow schedule using a 10% discount rate and taking the
beginning of operations as the starting date.
PV = -10 +2 +2 +2 +2 +2 +2 +2 +2 +2 +2 = -10 +1.5 +1.5 +1.5 +1.5 +1.5 +1.5 …
-.5 -.5 -.5 -.5 -.5 -.5 -.5 -.5 -.5 -.5
Using our trusty table: PV = -10 + (1.5)(6.1477) = -10 + 9.22155 = -0.77845
This project is sensitive to the discount rate (sensitivity analysis). At 5% it would be:
PV = -10 + (1.5)(7.7217) = -10 + 11.58255 = 1.58255
There is good reason to continue to use both terms. The interest rate, or to be more
exact the whole complex of interest rates, including rates paid by savings banks on
deposits, rates paid by homeowners on their mortgages, rates paid by businesses on
commercial paper and on long-term loans, and rates paid by governments on their debts,
is determined in the capital markets. As far as we are concerned, it is given to us
exogenously. The discount rate, on the other hand, is deliberately (but not arbitrarily)
chosen by the person performing an analysis. He may be making this choice according to
very strict guidelines, and the various interest rates currently available in the market will
bear heavily on his choice. Nevertheless it is a choice, and a very sensitive one at that, as
we shall shortly see. Consequently it is useful to preserve the parallel terminology, and
you should soon find yourself comfortable with it.
To get back to the main thread of this discussion, it should be clear that the general
formula for the present value (PV) of a sum of money Sl payable a year from now,
assuming a discount rate r, is
8
10/26/2019
PV=S1 / 1 + r
And after all our work on difference equations you should expect the present value of a
sum Sn payable n years from now to be
PV = Sn/(1 + r)n
In other words, we can treat a present value calculated as of some time in the future
exactly as if it were a payment occurring at that time.
Furthermore, we don't care whether the sum is payable to us or by us. The formula
still holds. You might think a little about the implications of this. It means that if we owe
money, we would prefer to pay later rather than sooner, because the later we pay, the less
will be the present value of the amount paid. If we are owed money, the opposite will be
true. Extending this idea to public projects, it means that all other things being equal, we
should prefer a project with early benefits and deferred costs to one where the reverse
holds.
Ordinarily, public projects do not result in a single benefit at some time in the
future. Rather, a stream of costs and benefits is generated over time. The present value of
such a stream is simply the sum of the present values of the individual items.
Lecture notes developed by Nolan Argyle and modified by Robert Kellner
9
10/26/2019
Purchase answer to see full
attachment