JOURNAL OF REGIONAL SCIENCE, VOL. 11, N O .

I, 1971

A NOTE ON EXTERNALITIES AND

PROPERTY VALUATION*

Lawrence D. Schallt

1. INTRODUCTION
The purpose of this note is to analyze the effects on land and improvement
values of a change in surrounding neighborhood effects. The impacts on land value
and on structure value are distinguished and the significance of the distinction for
public policy is then discussed.
The differential effects on land and improvement values of changes in sur-
rounding externalities, e.g., from an urban renewal or highway project, are gen-
erally ignored or mistreated in the literature and no rigorous analysis of the problem
is available.’ The question is often overlooked by referring only to total (land and
building) property value changes or by concentrating on land value changes alone.2
Rothenberg [S, p. 2611 argues that land and structure values necessarily vary in
the same direction since they are joint inputs.
I n the present paper, the essential relations associated with real property
valuation are first defined and then used to examine land use externalities. It is
observed that land and structure values may shift in opposite directions and that,
in general, this may be the more usual case. Consequently, land value changes
alone may be misleading in terms of the magnitude and even the sign of the effect
of a project’s spillovers on society’s wealth. It is also noted that knowledge of
movements in land and building values separately can be helpful in predicting
the impact of a project on land use patterns.
2. THE VALUATION OF LAND AND IMPROVEMENTS
It is assumed that the market for land use services is competitive. The term
“best use” refers to that use which maximizes the present value of net revenues.
The value of a plot of land, L , at time 0 is equal to the value it would have if it

* The author would like to thank George S. Tolley for helpful comments on the ideas pre-
sented in this paper.
t Assistant Professor of Finance end Business Economics, University of Washington.
1 For a comprehensive treatment of externalities in general see Mishan [5], and for land
use externalities in particular see, for example, Bailey [l]and Davis and Whinston [2].
2 Discussions usually center around the sign and magnitude of the net benefits (or costs)
of spillover effects, implying the use only of total property value changes; the distinction be-
tween land value and structure value changes is ignored. See, for example, Nourse [7]or Bailey
[l]. I n addition to being incomplete, disregarding structure value changes entirely by only con-
sidering land value changes (as in Mohring [S])can, as shown below, be misleading.
Date received: May, 1970.
101
102 JOTJRNAL OF REGIONAL SCIENCE, VOL. 11, NO. 1, 1971

were vacant at time 0. This is so whether or not the land is improved. This value
can be represented as

where r is the discount rate, r is the tax rate and R ( t ) and C ( t ) are the revenue
and cost (including construction, maintenance and demolition) in year t, respec-
tively, associated with the best use of L over time, assuming L is vacant at time 0.8
It follows that the entire value of a property, V T, is
(2) VT = VL -I- v,
where V r is the value of the existing improvements, 1. V T also equals

(3) VT =: c (1 - r)[R'(t)
t -0
- C'(t)l
(1 + r ) $
where R'(t) and C'(t) are the revenue and cost (including construction, main-
tenance and demolition) in year t, respectively, associated with that pattern of
use of L and I which maximizes Equation (3), i.e., which is best given the existence
of I at time 0.4
Rearranging Equation (2)
(4) Vr = V T - VL
Equation (4)reveals an extremely important point. The value of an improvement
is the increment to the present value of net receipts produced by that improvement.
The value of a structure cannot be computed merely by examining its productivity.
The value of the improvement, depends directly upon the productivity of the best
alternative use of the land which is represented by V L and cannot be determined
without a knowledge of V , .
The owner of L and I can demolish and remove I at a cost D and will do so if
the present value of returns from retaining I is less than ( V , - 0).It follows that
the market value of L and I together, VT , cannot be less than (VL - D),i.e.,
TIT 2 VL - D. Therefore, using Equation (4)6
(5) VI 2 -D
Now notice that the net present value of erecting I on L at cost C, would be
( V T - C,) ,where C, is the cost of constructing a structure identical to I in terms
of condition, type, etc. This amount, ( V T - C,), equals VL if 1 is the best use of
L (i.e., would be erected if L were vacant) and is less than V Lif I is not the best
use. Therefore ( V T - C,) I V L and, using Equation (2)

3 T represents the average tax rate per dollar of net revenue and is some composite of

property and income taxes. T may be a function of the land use pattern adopted.
4 As discussed below, it may be that the best use pattern involves the immediate demoli-
tion of I .
6 Relation (5) holds because the property owner has the option of demolishing I . If that
option did not exist, e.g., due to agovernment restriction, then V r 2 - m in Relation (5).
SCHALL : EXTERNALITIES AND PROPERTY VALUATION 103
~~ ~ ~~

(6) Vr ICp

where V I = C , only if I is the best use of L. Combining Relations ( 5 ) and ( 6 )

(7) -D 5 V I 5 Cp
From Relations (2) and (7)) it follows that
(8) VL - D 5 VT IVL + c,
Relations (4) and (7) demonstrate that the value of a structure cannot in
general be computed merely by determining the cost of replacement C , (generally
estimated as current construction costs of I less real depreciation) .6 As just demon-
strated, this approach to determining Vr is only correct if such a replacement of I
is the best use of L, i.e., what a profit maximizing property owner would construct
on L if it were vacant. I n any other case, V I < C, , I n an area with changing ex-
ternalities due to new land use patterns, the best use of a vacant L is likely to differ
from I, and, therefore, Vr is likely to be less than C, .
3. PROPERTY VALUES AND EXTERNALITIES
A technological externality in land use occurs if the land use on one plot of
land alters the productive opportunities available to another plot. Since such an
interdependency characterizes any neighborhood, land use generally involves
externalities. Of concern here are the changes in land and structure values due to a
change in the surrounding “neighborhood effects,” e.g., as might occur due to an
urban renewal or highway project.
The relevant measure of the wealth change due to a spillover is the change in
V T, since VT is the present value of the services produced by the p r ~ p e r t yOnly
.~
if there are no improvements on the land does the change in land value completely
reflect benefits or costs. If a spillover is an economy, i.e., enhances society’s wealth,
then V T of the affected plot increases. Conversely, an external diseconomy involves
a fall in VT .
From Equation (2))it is obvious that an external economy must involve a rise
in either V Lor VI . However, one of VL or VI may fall. For example, assume that,
if a particular plot were not currently improved with a hotel, I, its best use would
be an office building. With a new highway project, returns from I and V Tin Equa-
tion (3) may consequently increase. However, it may also be that the net present
value of revenues that would be yielded by the oEce building, V L, has risen by
even more than has V T,i.e., AVL > AVT . Using Equation (4))it follows that AV,
is negative even though the returns from I have increased. That is

6 For a discussion of replacement cost valuation of improvements and of property valua-

tion concepts in general see, for example, May [4] or Weimer and Hoyt [lo]. See Turvey [9]for
rt discussion of the use of market value magnitudes to estimate structure values.
7 This ignores the net government revenue gains or losses from the area which are not
included in V T .Also, as Rothenberg 18, pp. 138-431notes, the changes in V Tof the properties
affecteddirectly by the project’s spillovers may include value changes due to a shift in loca-
tion factors from the project. Theselocation caused changes in V ~ m u sbe
t subtracted out since
opposite and approximately equal locational effects will occur elsewhere.
104 JOURNAL OF REGIONAL SCIENCE, VOL. 11, NO. 1, 1971

(9) AVr = AVT - AVL < 0

Similar arguments imply that only V r or both V L and V r can rise with a rise in
VT .
Of the three possible situations, it may be that an increase in V Land a fall in
T’, is the most common. The redevelopment of an area, e.g., in a downtown sector,
is very often, if not usually, accompanied by the demolition of revenue producing
structures. Preceding the neighborhood’s revitalization, the same structures remain
because the alternative use of the land is not sufficient to warrant their removal,
i.e., the existence of I adds to the property’s present value at least -D. However,
with the area’s renewal, V Lrises and the buildings add less than -D to the present
value of the net receipts that would be realized from the best use pattern of the
land if the structures were removed. The rise in V Lis accompanied by a fall in V r
and the existing improvements are demolished, to be replaced by the best use of
L. In many cases the argument that land and improvement values move together
is therefore erroneous.
Using similar reasoning, it is easily shown that with a decline in V Tunder exter-
nal diseconomies either one or both of VL and V , could decline. For example, an in-
crease in traffic and noise from a rapid transit system might lower the value of a
hotel. However, if the best use of the land if vacant were a commercial building and if
the transit system were to raise the net returns from that best use, V Lwould rise.
A fall in V Twould occur here if the value of the hotel, V r , were to decline by more
than the rise in V L. On the other hand, both the hotel and office building uses
might consequently yield less or the former might yield more and the latter less.
Both V Land Vr would fall in the former case; only V Lwould fall in the latter.
The outcome is that no generalization can be made regarding the relative direc-
tions of change of land and structure values under either external economies or
diseconomies. Indeed, the notion that such values vary in the same direction may
be false in a majority of cases.
In addition to their analytical interest, the above results have practical implica-
tions. As noted earlier, the change in total property value, AVT , is the correct
measure of the spillover benefits of land use projects, e.g., urban renewal, park,
highway, etc. It follows from the above analysis that examination only of land
value changes to estimate benefits can produce completely misleading results. A
rise in land values may be partially or fully compensated for by a fall in structure
values, i.e., V Tmay decline even though V Lhas risen. Consequently, what appears
to be a benefit on the basis of land valuation may be a net loss. Even if AVL and
AVT are of the same sign, it is clear from Equation ( 2 ) that AVL can be a poor
estimate of net benefits if AVr is significant.
A second implication is that in predicting the spillover impact of a land use
project on any given neighboring property, changes in Vr as well as V T must be
estimated. Although AV, reflects net gains or losses, AVI implies the effect of the
spillover on land use patterns. As noted in the discussion of Relation (5), a struc-
ture must add in terms of present value at least the amount -D to V Lor thestruc-
ture will be demolished. If the spillover causes this addition to present value to
fall below -D,the structure will be removed and a new land use will evolve. There-
SCHALL: EXTERNALITIES AND PROPERTY VALUATION 105

fore, in predicting the impact on land use patterns of a project, the effect on Vr
must be evaluated. Of course, even if it is known that existing structures will be
demolished, the task of predicting the new land use remains. This problem is not
distinct from that of estimating the change in V I since V I itself is dependent upon
the potential use to which the land can be put, i.e., upon VL . Information regarding
this potential use is necessary in predicting AV, and is also a basis for predicting
the new land use if I is to be removed as a consequence of the spillovers.
In conclusion, changes in total property values should be used to estimate
spillover benefits. Land value changes alone may significantly misestimate-prob-
ably overestimate-such benefits. Shifts in improvement vaIues alone, however,
may be useful in predicting land use pattern alterations due to the spillovers.
REFERENCES
[l] Bailey, M. J. “Note on the Economics of Residential Zoning and Urban Renewal,” Land
Economics, 35 (1959), 288-292.
121 Davis, 0. A. and A. €3. Whinston. “The Economics of Urban Renewal,” Law and Con-
temporary ProbZems, 26 (1961), 105-117.
131 Dorfman, R. (ed.) Measuring the Benefits of Government Znvesfments.Washington: Brook-
ings, 1965.
141 May, A. A. The Valuation of Real Estate. New York: Prentice-Hall, 1953.
[5] Mishan, E. J. Welfare Economics. New York: Random House, 1964.
161 Mohring, H. “Land Values and the Measurement of Highway Benefits,” Journal of
Political Economy, 69 (1961), 236-249.
171 Nourse, H. “The Economics of Urban Renewal,” Land Economics, 42 (1966), 65-74.
[8] Rothenberg, J. Economic Evaluation of Urban Renewal. Washington, D. C.: Brookings,
1967.
[9] Turvey, R. The Economics of Real Property. London: George Allen and Unwin, 1957.
[lo] Weimer, A. W. and H. Iloyt. Principles of Real Estate. New York: Ronald, 1966.