You are on page 1of 12

Land Values in a Newly Zoned City

Author(s): Daniel P. McMillen and John F. McDonald


Source: The Review of Economics and Statistics, Vol. 84, No. 1 (Feb., 2002), pp. 62-72
Published by: The MIT Press
Stable URL: http://www.jstor.org/stable/3211739
Accessed: 01-06-2017 19:29 UTC

JSTOR is a not-for-profit service that helps scholars, researchers, and students discover, use, and build upon a wide range of content in a trusted
digital archive. We use information technology and tools to increase productivity and facilitate new forms of scholarship. For more information about
JSTOR, please contact support@jstor.org.

Your use of the JSTOR archive indicates your acceptance of the Terms & Conditions of Use, available at
http://about.jstor.org/terms

The MIT Press is collaborating with JSTOR to digitize, preserve and extend access to The Review of
Economics and Statistics

This content downloaded from 190.116.36.68 on Thu, 01 Jun 2017 19:29:49 UTC
All use subject to http://about.jstor.org/terms
LAND VALUES IN A NEWLY ZONED CITY

Daniel P. McMillen and John F. McDonald*

Abstract-The introduction of a new zoning ordinance to Chicago in 1923 Zoning's effects would be far easier to detect with a
offers a natural experiment that allows us to determine the effects of
natural experiment. Suppose we were to observe a sample of
zoning on relative land-value growth rates. Policymakers claimed at the
otherwise identical parcels of land, all of which are initially
time that land-use zoning would raise aggregate land values by reducing
negative externalities associated with mixed land use. After controlling for
unzoned. A new ordinance is adopted under which some
initial land use and the endogeneity of zoning decisions, we find that
parcels are zoned residential and others are zoned commer-
residential zoning led to higher land value growth rates than commercial
zoning. cial. If we know land values before and after the new
ordinance, then the difference in the land-value growth rate
I. Introduction between residential and commercial parcels provides a
clean measure of residential zoning's effect relative to
potentially conflicting land uses often are close to one zoning. Of course, if no parcels remain un-
commercial
another in densely populated urban areas. Stores, gas
zoned, we cannot determine whether growth rates are higher
stations, and even factories are sometimes in the same block
overall after the zoning ordinance than would have pur-
as houses. It is possible that everyone is better off by the
veyed without the new law. But if the zoning ordinance is
arrangement: businesses get a ready supply of workers and
necessary to eliminate unwanted intrusions of commercial
customers, who in turn have their transport costs reduced.
land into residential areas, we should find that land-value
But it also is possible that businesses create nuisances-
growth rates are relatively high in areas that are zoned
such as noise and traffic-to households, and that the
exclusively residential.
negative externalities are ignored when firms make their
This difference-in-differences approach cannot be imple-
location decisions. The potential market failure may be
mented using recent data because nearly all American cities
averted through land-use zoning: if negative externalities
were zoned for the first time in the 1920s, and large-scale
are limited to a small area around businesses, then separat-
revisions of ordinances are rare.2 Moreover, the necessary
ing business and residential areas may improve on the
panel of land-value data is rarely available. In this study, we
market allocation of land. Zoning then will increase resi-
use a remarkable data set to determine the effects of zoning
dential land values because homeowners are willing to pay
on relative land-value growth rates in Chicago. Chicago
a premium for exclusively residential areas.
adopted its first zoning ordinance in 1923. Land use was
Verifying the effect of zoning on land values has proved
completely unregulated before this date, but every block of
extraordinarily difficult.1 The reason is twofold. First, the
the city was zoned as of April 16, 1923. Chicago has a
competitive market may already have separated land uses.
unique data source, Olcott's Land Values Blue Book of
Large tracts of land are often developed by firms that would
Chicago, which provides annual estimates of land value for
forego profits if they ignored homeowners' preferences for
every block in the city. We obtained maps showing the use
exclusivity. Siegan (1972) is the standard reference. He
of every building in 1921 and the new zoning category for
argues that Houston, which is the only major American city
every block. For a sample of 1,013 blocks, we recorded the
without a zoning ordinance, has similar land-use patterns to
zoning category, the 1921 and 1924 land values, and the
other cities although there all land-use decisions are made
initial use, along with standard control variables such as
privately. Similarly, McMillen and McDonald (1999)
distance from the city center. All the pieces are in place to
present evidence that land use was highly segregated in
determine the effect of zoning on relative land-value growth
Chicago even before the city adopted its first zoning ordi- rates.
nance. The second reason empirical studies have failed to
The potential endogeneity of the zoning decisions poses
establish conclusively that zoning affects land values is that
econometric challenges. Zoning officials may have been
zoning may be flexible, mimicking the competitive market
aware in 1923 that land values were growing slowly in some
(although probably with a lag). McMillen and McDonald
areas. If blocks with high land-value growth rates were
(1991) present evidence for this view using data from
more likely to be zoned residential, then residential zoning
suburban Chicago. They find that changes in relative land
would appear to cause higher growth rates even if zoning
values lead to changes in zoning categories. But no study
had no independent effect on land values. The availability of
has documented the frequency of variances or simple toler-
land value data before and after zoning permits us to use
ance of zoning code violations that many researchers sus-
recent extensions (Heckman, Smith, & Clements, 1997;
pect make zoning far from binding.
Heckman, Ichimura, & Todd, 1997, 1998; Heckman et al.,
1998) of Rosenbaum and Rubin's (1983) propensity score
Received for publication April 2, 1998. Revision accepted for publica-
tion February 13, 2001.
* University of Illinois-Chicago. 2 The first comprehensive zoning ordinance was passed in New York
City in 1916. Mills (1979) reports that by 1925 nearly 500 American cities
We thank Dan Black, Chris Bollinger, John Engberg, Robert Moffitt, and
the referees for helpful suggestions. had zoning ordinances, and by 1930 three-fifths of the urban population
1 See Fischel (1990) and Pogodzinski and Sass (1991) for reviews. lived in zoned communities.

The Review of Economics and Statistics, February 2002, 84(1): 62-72


? 2002 by the President and Fellows of Harvard College and the Massachusetts Institute of Technology

This content downloaded from 190.116.36.68 on Thu, 01 Jun 2017 19:29:49 UTC
All use subject to http://about.jstor.org/terms
LAND VALUES IN A NEWLY ZONED CITY 63

matching procedure to control for selection bias caused


includes by
"manufactures and storage of explosives and high
pressure
zoning endogeneity. The estimated probability from gases."
a probit
model of residential zoning forms the basis for McMillen
a compari- and McDonald (1999) present a detailed anal-
son of changes in land values across zoning classifications.
ysis of the land-use patterns implied by these maps. They
Highest weight is placed on blocks receiving showdifferent
that land use was heavily segregated even before
Chicago adopted
zoning but which have similar estimated probabilities ofa zoning ordinance. Houses were some-
times next to stores, but were very rarely next to manufac-
being zoning residential. The idea is to compare differences
turing establishments. The base category of manufactur-
in outcomes among sites that are otherwise similar.
ing-Category
Our evidence suggests that residential zoning led toA-was often mixed with commercial
higher land values than did commercial zoning. establishments,
This result but the segregation of other manufacturing
categories
is somewhat surprising because the zoning ordinance in- was nearly complete. McMillen and McDonald
present logit
cluded grandfather clauses that permitted conflicting uses to models of land use and find that land-use
patterns
continue indefinitely. Furthermore, commercial zoning did are quite predictable. Mixed land use was most
not preclude residential use because early zoning likely to occur on major streets and closer to the city center.
ordinances
Commercial
were hierarchical, with residential land use at the top of the establishments were concentrated on major
hierarchy. Our results suggest that residentialstreets, and manufacturers located near rail lines and water-
landowners
ways. Although
valued the insurance that zoning provides against futuremixed residential and commercial blocks
intrusions of conflicting commercial land uses.were not uncommon, the overall land-use pattern was not
Residential
zoning may also have been beneficial to some substantially
commercial different from a stylized moder city.
landowners because the grandfather clause providedIt seems unlikely
local that zoning would have much effect on
land values
monopoly positions for commercial establishments when land-use separation was so pervasive.
in resi-
dential zones. Nevertheless, a simple partial equilibrium model is suffi-
cient to show that residential zoning can potentially raise
land values if homeowners pay a premium to live in areas
II. The 1923 Chicago Zoning Ordinance that are exclusively residential. In this section, we present a
graphical treatment of the more general nonlinear program-
Early zoning ordinances were less restrictive than modem
ming model of land values and zoning developed for the
codes. The courts had only recently viewed land-use zoning
two-category case in McMillen and McDonald (1993).3
as an undue restriction on a landowner's property rights. To
Even the simple partial equilibrium approach yields few
ensure that new codes would pass court review, zoning unambiguous predictions for the effect of zoning on land
ordinances covered an entire city at once and were hierar-values.
chical. In April 1923, the entire city of Chicago was newly The two land-use categories are residential (r) and com-
zoned for one of three primary uses: residential, commer- mercial (c). We assume that land values are linear functions
cial, or manufacturing. The residential category was at theof the proportion of land in the area that is residential:
top of the hierarchy: once a block was zoned residential, no
commercial or manufacturing development would be per- Vr = tr + r,p
mitted in the future. Commercial zoning permitted all resi- (1)
dential and commercial uses, but excluded any future man- Vc = tc + pcP,
ufacturing development. No uses were excluded under
manufacturing zoning, which was at the bottom of the where v denotes land value, p is the proportion residential,
and ar, Pr, at, and 3?c are parameters. Other characteristics
hierarchy. Grandfather clauses allowed any existing con-
of the parcel-such as distance from the city center and
flicting use to continue indefinitely, although limits were
placed on the extent to which an establishment could be location on a major street-are subsumed in the parameters
Oxr and ac. The linearity simplifies the graphs, but it is not
expanded.
critical.
Maps were drawn showing the use of every existing
The best case for hierarchical zoning can be made if both
building in the city. The detail is impressive. Housing
functions are upward sloping (Pr > 0 and 3c > 0). In this
includes single-family dwellings, two-flats, apartments,
case, homeowners prefer to live in areas that are predomi-
boarding or lodging houses, and vacant residential build-
nantly residential, but businesses are also attracted to resi-
ings. Eight types of commercial uses are shown. Manufac-
dential areas. Three different scenarios are illustrated in
turing is divided into four categories based on the degree of
figure 1. C is the commercial land-value function, and R1,
externality generated. Buildings in the base manufacturing
R2, and R3 represent three different residential functions.
category are nearly identical to many commercial establish-
Consider R1 first. In this case, land is worth more in
ments: the base includes "general manufacturing which is
carried on in [a] building without yard storage and without
3 We extend the model to include multiple land-use categories in Mc-
noise, smoke, odors, danger, excessive fire risk, or other Donald and McMillen (1998). The two land-use case corresponds to our
nuisance features." In contrast, manufacturing Category D subsequent empirical analysis and is easy to present diagrammatically.

This content downloaded from 190.116.36.68 on Thu, 01 Jun 2017 19:29:49 UTC
All use subject to http://about.jstor.org/terms
64 THE REVIEW OF ECONOMICS AND STATISTICS

FIGURE 1.-LAND VALUES AND ZONING


A2 under commercial zoning but would rise to A2* in the
long run under residential zoning.
Land Value In a fourth case that we do not illustrate, the residential
per Acre function is below the commercial function at both p = 0
and p = 1, but the residential land value is higher at p =
1 than the commercial land value at p = 0. In this case, the
market assigns all land to commercial use, but land values
would be higher under residential zoning.
The analysis suggests that different existing land-use
patterns-residential, mixed, or commercial-lead to differ-
ent expectations for subsequent land-value changes. Zoning
is irrelevant when the area is all residential. When an area

A3 combines residential with commercial uses, land values rise


under residential zoning but are unchanged when the area is
zoned commercial. Zoning is irrelevant when an area is all
commercial and the residential land-value function is very
low. If an area is all commercial but the residential land-
value function is close to the commercial function, residen-
tial zoning may increase land values. Land values will never
fall under commercial zoning because land-use patterns are
unaffected.

Proportion Residential
The simple partial equilibrium model is entirely static.
Suppose that landowners are forward looking. Grandfather
clauses permitted existing uses to continue indefinitely. The
zoning ordinance created local monopolies for commercial
residential than in commercial use for any level of p. A enterprises in the middle of areas zoned residential (Bab-
competitive market allocates the land to residential use cock, 1957). Because these establishments could be sold,
(point A 1). The opposite situation occurs when the residen-
land values might rise in mixed land-use areas that were
tial land-value function falls to R3. Commercial firms
zoned residential. In contrast, landowners might have ex-
always pay more for the land than do homeowners, and pected
the that the grandfather clauses would eventually fade
competitive market assigns all land to commercial useleading the new zoning provisions to be binding.5 Land
out,
(point A3). The most interesting case occurs when the then would fall in mixed land-use areas that were
values
residential land-value function is R2. Here, businesses out-
zoned residential if commercial firms expected eventually to
bid homeowners for parcels on exclusively residential
leave. In a growing city, it is reasonable to expect that
blocks, but homeowners outbid businesses for parcels in
commercial firms may eventually seek out sites in areas that
exclusively commercial areas. Point A2 then is a stable
currently are predominantly residential. Land values might
equilibrium, and the competitive market's tendencyfalltoin residential areas that are zoned commercial if current
equalize land values leads to mixed land use.4 landowners are wary of the possibility of invasion by
In the 1923 Chicago zoning ordinance, new businesses
conflicting land uses in the future, and residential landown-
were not permitted in areas zoned residential, but either use
ers may pay a premium for the insurance of neighborhood
was permitted in areas zoned commercial. Ignore grandfa-
stability provided by residential zoning.
ther clauses for now, and assume that a block zoned resi-
We must also allow for the potential endogeneity of
dential would subsequently have p = 1. When the residen-
zoning decisions. Zoning authorities were probably aware
tial land-value function is as high as R1, residential zoning
of the effects of their decisions on land values. They
is redundant: the market has already led to exclusively
claimed at the time that eliminating nuisances would raise
residential land use. Commercial zoning also is meaningless
land values, which suggests that they did not want to take
in a hierarchical zoning system because houses are permit-
actions that would lower land-value growth. An area with
ted and the land is worth more in residential use. In contrast,
commercial land would be zoned residential only if its land
suppose that the residential land-value function is as low as
R3 and that the area is currently all commercial. Land
5 Indeed, the city reneged on the grandfather clause in 1957 when the
values remain at A3 if the area is zoned commercial, butordinance
fall was revised. Conflicting businesses had ten years to sell, after
to A3* under residential zoning. The best case for zoning
which all land had to conform with the new zoning. The new zoning
occurs under mixed land use. Here, land values remainordinance
at was not hierarchical, and residential uses also were required to
leave when an area was zoned commercial. The 1957 ordinance would
have been viewed as Draconian in 1923, and would not have passed court
review.
4 The mixed-use equilibrium is stable if both ,r and P, are positive, but Still, there may have been some early suspicion that the grandfa-
3, > P,. See McDonald and McMillen (1998). ther clauses would eventually be phased out.

This content downloaded from 190.116.36.68 on Thu, 01 Jun 2017 19:29:49 UTC
All use subject to http://about.jstor.org/terms
LAND VALUES IN A NEWLY ZONED CITY 65

value was not expected to fall significantly underOur objective


residential is to determ
by
zoning. To the extent that zoning simply follows zoning
the market, category for a
analogous
it will have little or no effect on land-value growth in the to the literature
short term. view residential zoning as
zoning
The net effect of all these factors is ambiguity. Official as the alternative.
when
publications claimed at the time that the zoning ordinancea block is zoned re
Smith,
would increase aggregate property value by a billion dollars and Clements (1997
as opposed
by eliminating conflicting land uses. Their claims may well to commercial
be correct if residential landowners were willing to pay for
the insurance that zoning creates. But zoning may also
decrease land values by removing an option for aE(ylID
future= land
1) - E(y(ID = 0) = (I - )'X
use that may be quite valuable. The effect of zoning on land (4)
values is ultimately an empirical issue.
+ E(uI\D = 1) - E(u}ID = 0).

III. Empirical Procedure


Estimating the mean effect is complicated because the error
terms are unlikely to have zero expectations. Zoning deci-
Our data set includes land values before and after the
1923 zoning ordinance. Before 1923, land values variedsions
by are not made randomly. Official documents state
the market-determined land use. Our theoretical model explicitly that zoning would increase land values, which
implies that relative land values influenced zoning. Site
suggests that we must distinguish among three different
characteristics that are more highly valued in residential
land uses: residential, commercial, and mixed. To simplify
than in commercial use would tend to be associated with
the presentation, we suppress an implicit subscript for land
residential zoning.
use and assume that the following functional form accounts
for variations in land values for a given use: y? = 3P'XiAssume
+ that the underlying propensity for residential
u?, where y? is the natural logarithm of land value zoning
at can be represented by the latent variable D* =
y'Zi
location i in the initial time period, Xi is a vector of + vi, where Z is vector of explanatory variables and v
is an error term. Z will include the variables in X if land
explanatory variables, u? is an error term, and P30 is a vector
values influence zoning, but it may include other variables
of coefficients. Xi does not vary by land use or time period
because it includes variables that vary only by site, such also.asAs long as v is correlated with Ur and Uc-a reasonable
distance from the city center. The coefficient vector assumption
is because zoning officials had access to additional
expected to vary by land use, reflecting different returns to information at the time that likely influenced both land
site-specific characteristics across uses. As an example, land values and the zoning category-OLS regressions lead to
values may fall more quickly with distance from the city biased estimates of equation (4). Standard selection bias
center for commercial than for residential use, which im- procedures can be used to estimate the land value functions,
plies that the coefficient for this variable is greater in but they are often not robust to changes in specification and
absolute value for commercial properties. are sensitive to the distributional assumption. Selection bias
In a growing city, the land-value function coefficients willis likely to be present because zoning officials were not
change over time. For example, the intercept should in- conducting an economic experiment.
crease and the absolute value of the coefficient for distance The pre-zoning land value data provide an alternative to
to the city center should fall as a city's population grows, standard selection bias procedures for estimating the mean
income rises, and commuting costs decline. Zoning also effect of residential zoning. Instead of comparing differ-
ences
may alter the coefficients if it creates, for example, a relative in land values across zoning categories, we compare
differences in differences:
scarcity of residential zones along major streets. The land-
value function will likely vary by initial land use and the
new zoning category after 1923. The two zoning categories
we consider are residential and commercial, which are E(yl - y?lD = 1) - E(y' - y?lD = 0) (5)
denoted by r and c. The land value functions are

I = P3'X,
Yri1 Xi ++Uri
ul, (2)
(2) Equation (5) measures the difference in land-value growth
rates for sites that have the same initial land use but were
and subsequently zoned for different uses. In contrast to the
traditional selection bias estimator, which assumes that
, = Pc'Xi + uL. (3) conditioning on Z eliminates selection bias in the post-
zoning time period, equation (5) assumes that conditional
The superscript indicateson the
Z selection bias is constant
time over time soafter
period that it cancelsthe e
of the zoning ordinance.out in differencing.

This content downloaded from 190.116.36.68 on Thu, 01 Jun 2017 19:29:49 UTC
All use subject to http://about.jstor.org/terms
66 THE REVIEW OF ECONOMICS AND STATISTICS

Equation (5) can be estimated using a variantestimate is the standard deviation of the 10,000 new values
on match-
ing procedures (Rosenbaum & Rubin, 1983, 1984). TheWe
of MA. first
also present 90% and 95% confidence intervals
step is to estimate the probability of residential zoning
using by
the percentile method: the values of A are sorted, and
probit: P(Z) = (Dy'Z). The next step is to compare
entries
land-value growth rates across zoning categories, ()B and (1 - 2)B form the confidence interval,
placing
more weight on observations that have similar estimated
where a = 0.05 and 0.10.
probabilities of residential zoning:
Heckman et al. (1998) demonstrate that the matching
estimator is effective at reducing selection bias in the
A = 2o(i)[(y y) - yO) _ W(i,j)(y, - y)], (6)
iEII jEjIo evaluation of a job training program. They decompos
selection bias into three parts. The first two arise because of
where I1 is an index for observations zoned residential and
different densities and different weighting of E( yX, D) for
Io is an index for observations zoned commercial.6 The first
the I1 and Io observations. These two components ar
set of weights, ((i), can be used to account for heteroske-
"selection on observables." The other component arises
dasticity and scale. In our application, o(i) equals 0 when
when differences in outcomes continue to exist after con-
an I1 observation has no counterpart in Io with a similartrolling for Z (which includes X and other exogenous
probability of residential zoning, and w(i) = llnl other-variables), and is the form of selection bias that is handled
wise, where nl is the number of observations with o(i) by: standard estimators. Heckman et al. (1998) show that
0. The second set of weights, W(i, j), sum to 1 and depend
matching estimators are effective at reducing the first two
on the estimated probability of residential zoning.
components of selection bias, and find that empirically these
Following Heckman, Ichimura et al. (1997), we use Fan's
components comprise a much greater proportion of overall
(1992) local linear estimator to construct W(i, j). We use
15
bias than traditional selection on unobservables. Thus, in
their application the matching procedure is a better method
the biweight kernel, K(si) = 1-6(Si - 1)2I(sij < 1),
than traditional IV procedures for controlling for exogenous
where sij = P(Zi) - P(Zj) and h is the bandwidth. The variables.
weights are normalized to sum to 1. The local linear esti-
In our context, the first two components of selection bias
mator is a variant of the nearest neighbor estimator and is
arise when there are few sites with commercial zoning that
less sensitive to boundary effects than standard kernel
have similar values of P(Z) to sites that are zoned residen-
estimators. The estimates are easy to construct by weighted
tial. In the extreme, all sites zoned commercial have P(Z) =
least squares (Pagan & Ullah, 1999). Let Q = ri - y? and
0 and all sites that are zoned residential have P(Z) = 1. A
Qcj = yj - yj. Then, jElo W(i, j)(yj - y?) is equivalent
to the predicted value from a weighted least squares regres- comparison of land values across the two regimes then has
little meaning because the sites are inherently different. In
sion of Qcj on P(Zj) with weights of K(sij)1/2. Thus, the
match for Qri is the predicted value from a weighted least contrast, two sites that both have P(Z) = 0.4 but which are
squares regression of Qc on the estimated probabilities, zoned differently are similar in all measurable ways. If their
using the estimated probability for observation i as the land values differ, the difference is attributable to zoning.
target point. Io observations with estimated probabilities If our concern were only the distributional or functional
that are close to P(Zi) get the heaviest weight in construct- form assumption behind the standard selection model, we
ing the match for observation i. could use a semiparametric estimator such as those of Ahn
The weighted least squares regression has no observa- and Powell (1993), Cosslett (1991), or the approaches
tions, where no is the number of observations with Io = 1. discussed by Pagan and Ullah (1999). We are also con-
Following Silverman (1986), we set the bandwidth using cerned about the plausibility of our exclusion restrictions,
the rule h = 1.066nol/5, where 62 = var(P(Z)). We show which leads us to prefer the matching approach.7 Matching
that the results are not sensitive to moderate variation in the
does not correct for classical selection bias, which focuses
bandwidth. The regression cannot be estimated for obser-on the error terms. It does correct for the influence of
vation i unless there are at least two Io observations withobservables without making strong functional form assump-
K(sij) > 0. In practice, we require three observations with tions as in Barnow, Cain, and Goldberger (1981). As with
K(sij) > 0, and set w(i) to 0 otherwise. other estimators for selection models, the matching proce-
We use a simple bootstrap procedure to construct stan-dure hinges on the specification of Z. Our data set provides
dard errors for RA. We make B = 10,000 independentenough strong predictors of zoning that we are confident
draws from the I, and Io observations, and form new
that we have accounted for the key variables used to draw
estimates of M for each draw. The bootstrap standard error
Chicago's zoning map, and that remaining unobserved in-
fluences are not closely correlated with land-value changes.
6 We do not use Heckman, Ichimura et al.'s (1997) regression-adjusted
local linear matching estimator because our values of X do not differ over
time, which causes the regression adjustment to cancel out of the two 7 Heckman et al. (1999) provide an excellent description of alternative
land-value growth rate terms in equation (6). methods for evaluating the effects of labor market programs.

This content downloaded from 190.116.36.68 on Thu, 01 Jun 2017 19:29:49 UTC
All use subject to http://about.jstor.org/terms
LAND VALUES IN A NEWLY ZONED CITY 67

The matching estimator is better suited to our data set than TABLE 1.-DESCRIPTIVE STATISTICS

other selection model estimators. Variable Residential Mixed Commercial

Distance to the city center 6.673 4.807 5.168


IV. Data (1.756) (1.940) (2.600)
Distance to Lake
Michigan
Our data set is a lattice of 1,116 blocks from the area 4.187 3.049 3.189
(2.385) (1.867) (2.043)
bounded by Devon Ave. on the north side (6400 North),
Distance to 67th
el station 1.412 0.688 0.858
St. on the south, Central Ave. on the west (5600 West), and (1.137) (0.644) (0.837)
Distance to commuter
Lake Michigan on the east. We use the zoning board's maps
train station 0.925 0.859 0.857
to determine the mix of uses-residential, commercial, light (0.524) (0.493) (0.486)
manufacturing, and heavy manufacturing-on Distance to river
each or canal 2.261 1.592 1.637
block
(1.345) (1.091) (1.364)
face in 1921. Using the block face as the unit
Nearof analysis
rail line 0.185 0.132 0.285
greatly simplifies the data collection task, and has
On major the 0.140 0.247 0.503
street
District:
advantage of corresponding to the subsequent mediandeci-
zoning house age 28.331 28.950 29.143
(4.014) (3.385) (4.391)
sions. All block faces were zoned for a single use, which
District: percentage rental 49.915 42.348 43.529
suggests that the zoning board considered the block face the (14.480) (14.311) (15.456)
appropriate unit of analysis. District: average home
value ($1000) 8.798 7.986 8.351
We do not include heavy manufacturing in our analysis (2.422) (2.205) (2.883)
because it was highly segregated and adds little
District: or no
percentage
multifamily 16.652 24.618 22.205
information: 93 of 102 blocks with heavy manufacturing
(12.821) (11.479) (12.010)
were zoned for manufacturing, and only four
Numberwere of zoned
observations 449 385 179
residential. We combine commercial and light manufactur-
Means are followed by standard deviations for continuous v
ing into one category, which we simply call commercial.
Combining the two uses into one category avoids severe
small-sample problems that arise when the data are stratified
by initial land use. The distinction between Our results
commercial andshould thus be inter
light manufacturing appears to have been initial impact
arbitrary, and of zoning on land v
there is little difference in land-use patterns between We supplement
the two the land value, la
uses. with standard explanatory variables
Olcott's Land Values Blue Book of Chicago measures:
provides distances from the city
front-foot land value estimates for each block in our data the nearest elevated train ("el") s
commuter train station. All distances are measured in
set. We recorded estimates from the 1922 and 1925 issues;
the data reflect 1921 and 1924 values because of the pub- straight-line miles. The distinction between el and com
lication lag. We chose these dates because they bracket themuter train stations is necessary because the el is used f
time the ordinance was enacted in mid-1923. Although a intracity travel, whereas commuter trains primarily bri
future zoning ordinance was expected in 1921, the date passengers from the suburbs. We include distance from
when it would take effect and the details of the ordinance navigable river or canal as an explanatory variable becau
manufacturing firms are concentrated near waterways. We
were unknown. It is doubtful that anticipation of the zoning
ordinance affected land values in 1921. The later date expect the probability of residential land use and zoning
be
provides about a year of adjustment to the zoning ordinance. lower near waterways. The next explanatory variable is a
We chose the relatively short post-zoning time because wedummy variable for location along a major street. Chicago's
wanted to isolate as much as possible the effects of zoning system has major streets every mile. We include
grid street
from other changes in the economic environment. four district-level variables to capture differences acros
neighborhoods
Olcott's has a long and productive history of use in in political support for restrictive zonin
practices:
empirical studies, dating to Yeates (1965) and Mills (1969). median house age, percentage rental, average
home
The books are published annually, and we believe they arevalue, and percentage multifamily.8 Our expectatio
is that
accurate when economic conditions are not changing there would be more political pressure for residentia
dras-
tically. Whether they were accurate during the earlyzoning
1920s in
is districts with new homes, few rentals, high hom
unclear. Real estate professionals may or may not have been and a small percentage of multifamily apartment
values,
able to accurately predict the change in land values that
would be caused by the drastic change in conditions 8Thepre-
district data are drawn from Wirth and Bernert (1949). Th
districts are "communities," which were defined in early research b
cipitated by a new zoning ordinance. The widespread pop- of Chicago sociologists and are still in use today. The data
University
ularity of zoning ordinances was at least partly caused by as
back only a far as 1930, but community data do not change rapidly ove
time. The land-value data are drawn from individual blocks, whereas t
genuine belief that exclusive residential zoning would raise
community data represent a much larger area. Thus, the four communit
land values. Land values would rise in the short run in areas
level variables capture neighborhood effects, which are exogenous t
that were zoned residential if this belief were widespread.
landowner on an individual block.

This content downloaded from 190.116.36.68 on Thu, 01 Jun 2017 19:29:49 UTC
All use subject to http://about.jstor.org/terms
68 THE REVIEW OF ECONOMICS AND STATISTICS

TABLE 2.-LAND-VALUE GROWTH RATES


nearly identical across the two zoning categories. Tests for
Residential Mixed Commercial differences in means, presented in the last row of table 2,
Percentage zoned residential 71.492 54.545 14.525 reveal that the differences are significant for blocks that
Mean log land-value change- were exclusively residential or commercial, and that the
zoned residential 0.567 0.396 1.055
difference is not significant for mixed-use blocks.
(0.438) (0.321) (0.905)
Mean log land-value change- Table 2 suggests that landowners on blocks that were
zoned commercial 0.364 0.409 0.415 exclusively residential in 1921 were willing to pay a pre-
(0.475) (0.460) (0.560) mium for residential zoning. It appears that the insurance
Difference in means 0.203 -0.013 0.641
(4.328) (0.337) (4.867) provided by residential zoning led to an increase in values
when compared to the alternative of commercial zoning. In
Standard deviations are in parentheses for the mean log land-value changes. Absolute t-values are in
parentheses for the difference-in-means tests. contrast, zoning appears to have no effect on land values in
blocks that originally combined commercial and residential
zoning, presumably because zoning had no short-term effect
on land use in mixed-use blocks. The results for exclusively
Descriptive statistics for the explanatory variables are pre-
sented in table 1. commercial blocks are more surprising. It is possible that
residential zoning was valuable because it provided local
monopolies to commercial establishments that were permit-
V. Empirical Results
ted to stay under the grandfather clauses. An alternative
Table 2 presents summary data on changes from 1921 toexplanation is that characteristics correlated with high land-
1924 in the natural logarithm of front-foot land values. Landvalue growth caused these blocks to be zoned for residential
values rose by approximately 36.4% for blocks that were use.
exclusively residential in 1921 but were zoned commercial Probit models of residential zoning are the first
in the later period. Land-value growth rates were somewhat adjusting for this form of selection bias. The res
higher at 40.9% and 41.5% for mixed-use and exclusivelypresented in table 3. The most important determi
commercial blocks that received commercial zoning. Resi- zoning are dummy variables representing locations n
dential zoning is associated with higher land-value growth lines and along major streets, both of which sign
rates than commercial zoning for blocks whose initial land decrease the probability of residential zoning. The es
use was exclusively residential or exclusively commercial. models are very successful in predicting zoning. F
For mixed land-use blocks, land-value growth rates are through 4 present kernel density estimates of re

TABLE 3.-PROBIT MODELS OF RESIDENTIAL ZONING

Residential Mixed Commercial

marginal marginal marginal


Variable coef. effect coef. effect coef. effect

Constant 0.792 14.441 0.747 18.936 2.203 19.725


(0.517) (0.410) (0.544)
Distance to the city center -0.059 -1.072 0.046 1.154 0.302 2.701
(0.646) (0.506) (1.499)
Distance to Lake Michigan -0.070 -1.274 0.131 3.322 0.184 1.651
(0.882) (1.618) (0.945)
Distance to el station -0.152 -2.764 0.018 0.458 0.462 4.139
(1.097) (0.097) (1.209)
Distance to commuter train station -0.290 -5.287 -0.095 -2.405 -1.872 -16.761
(1.596) (0.548) (2.531)
Distance to river or canal 0.221 4.032 -0.041 -1.028 -0.260 -2.327
(2.446) (0.407) (1.584)
Near rail line -1.802 -32.833 -1.263 -31.989 -2.942 -26.344
(8.602) (5.289) (3.750)
On major street -1.845 -33.618 -2.118 -53.668 -3.074 -27.525
(7.819) (9.539) (4.668)
District: median house age 0.009 0.165 -0.086 -2.181 -0.189 -1.694
(0.172) (1.390) (1.552)
District: percentage rental 0.031 0.564 0.051 1.287 0.003 0.031
(2.047) (3.279) (0.097)
District: average home value ($1000) -0.014 -0.247 -0.099 -2.500 0.060 0.539
(0.242) (1.535) (0.394)
District: percentage multifamily -0.024 -0.441 0.026 0.665 0.089 0.796
(1.939) (2.230) (2.480)
Asymptotic absolute t-values are in parentheses. The marginal ef
> aP, 100
az, n

This content downloaded from 190.116.36.68 on Thu, 01 Jun 2017 19:29:49 UTC
All use subject to http://about.jstor.org/terms
LAND VALUES IN A NEWLY ZONED CITY 69

FIGURE 2.-KERNEL DENSITY ESTIMATES: RESIDENTIAL LAND USE


zoning probabilities produced by the three probit models.9
The solid lines are the density estimates for blocks
8 that
-
were
commercial
I!
",

actually zoned commercial, and the dashed lines are


---the residential . \
i !

estimates for blocks with residential zoning. The 7 models


- are !
I

successful in predicting zoning if the commercial line has a i I!


I

sharp peak near a 0 probability of residential zoning, and the


6-
residential line has a peak near a probability of 1. The .

double-peaked pattern is evident in all three figures, and is


I I
most pronounced for mixed-use blocks. 5-
i

Figures 2 through 4 show why matching helps to reduce


!

selection bias. Many blocks with commercial zoning have a


/~~~~~
i
4- !
I

low probability of residential zoning, but most blocks with i

residential zoning have a high probability. Zoning


3 -
would I
I

thus provide relatively little explanatory power for relative i

land-value growth rates because the observations differ /


2-
substantially in other ways. The smaller number of obser-
................................... -"~ ~ ~ ~ ~ ~~~~

vations in the middle provide more information. In the


middle, we have blocks in both zoning categories
1 - with
similar probabilities of residential zoning. They are the
same in terms of measurable characteristics, but A they differ
in zoning category. Therefore, the zoning category 0.00 provides 0.25 0.50 0.75 1.00

the most information on relative land values for these probability of resident
observations.
For mixed land use, we have ample numbers of blocks in
use blocks with residential zoning makes a comparison of
each zoning category with comparable estimated probabil- land-value growth rates less reliable. The difference in
ities of residential zoning. The numbers are smaller, butreliability is reflected in larger bootstrap standard errors for
residential-use blocks also have comparable observations in
the commercial-use blocks. It is also reflected in the number
each category. However, the small number of commercial- of observations given no weight (io = 0 in equation (6)) in
constructing the difference-in-difference estimates. Matches
are available for all residential and mixed-use blocks that
9 Following Silverman (1986), we use a Gaussian kernel with bandwidth
h = 1 .066n-5, where 6 is the standard deviation of the probit
were zoned residential, but only six of 26 blocks have o,i
probability estimates and n is the number of observations for the land use
0 for commercial blocks. These six blocks have the lowest
and zoning category under consideration.

FIGURE 3.-KERNEL DENSITY ESTIMATES: MIXED LAND USE FIGURE 4.-KERNEL DENSITY ESTIMATES: COMMERCIAL LAND USE

3.5 35

3.0 30

2.5 25

2.0 20

1.5 15

1.0 10

0.5 5

0.0

0.00 0.25 0.50 0.75 1.00 0.00 0.25 0.50 0.75 1.00

probability of residential
probability
zoningof resi

This content downloaded from 190.116.36.68 on Thu, 01 Jun 2017 19:29:49 UTC
All use subject to http://about.jstor.org/terms
70 THE REVIEW OF ECONOMICS AND STATISTICS

TABLE 4.-NoNPARAMETC DIFFERENCE-IN-DIFFERENCES ESTIMATES


not previously exist. Although the existing nonresidential
Residential Mixed Commercial uses that were mixed with residential use did not reduce

Estimate of residential
residential land values, all blocks faced a risk that an
minus commercial incompatible nonresidential use could move in prior to the
land-value
zoning ordinance. As we have shown, post-zoning land
growth 0.195 0.184 0.477
Bootstrap standard error 0.110 0.095 0.307 values reflect this assurance provided by residential zoning.
Percentile method 90%
confidence interval 0.035, 0.379 0.109,0.246 0.065, 1.002 VI. Spatial Autocorrelation
Percentile method 95%
confidence interval -0.010,0.428 0.063, 0.261 0.015, 1.113
Bandwidth 0.129 0.120 0.107
Autocorrelation is a potential problem in spatial data sets.
Standard probit estimates remain consistent when errors are
autocorrelated, but the standard errors are biased. Bootstrap
estimated probability procedures
of residentialhave not been developed
zoning. for spatially
Thecorrelated
small
ber of residentially zoned error structures.
commercialDriscoll and Kraay (1998)
blockspresent a con-
accou
the large standard error, and is a clear warning sistent covariance matrix for continuous dependent variable th
results for this use should be viewed with caution. models with spatially dependent panel data, but no such
estimator has been developed for cross-sectional spatial data
The results of the nonparametric matching estimator are
presented in table 4. The results are nearly the same asor infor models with discrete dependent variables. Table 6
table 2 for residential blocks: residential zoning increasespresents the results of three alternative model specifications
land values by about 19.5% compared with commercial that help control for spatial autocorrelation. The "spatial
zoning. The bootstrap standard error is high, but we can expansion" model is commonly used by geographers (Ca-
setti,
reject the null of no difference in land-value growth rates at 1972; Jones & Casetti, 1992) to control for spatial
the 10% level. The value of the matching estimator is seen effects. This model adds a low-order polynomial of the
in the last two columns. Table 4 indicates that residential geographic coordinates to the base model. The latent vari-
zoning increases relative land values by 18.4% in mixed-use able for the probit model becomes D* = y'Zi + f(Ni, Ei) +
vi, in
blocks. The effect is statistically significant, whereas where N and E represent distances north and east of the
table 2 we found no difference in land values for this city center. We use a second-order polynomial, adding N, E,
N2, E2, and N X E as explanatory variables. The motivation
land-use category. Table 4 also indicates that residential
behind this procedure is that spatial autocorrelation is usu-
zoning increases land values for blocks that were initially
commercial, but the effect is reduced as compared ally caused by missing variables that are correlated over
to table
2 to a more reasonable figure of 47.7%. The results inspace
table(but hopefully not with Z). The spatial expansion
4 are consistent across land-use category and leadhelps reduce the extent of the correlation in the errors by
to the
taking into
conclusion that residential zoning increased land values in account broad spatial trends.
the period shortly after the adoption of the zoningThe other two alternatives follow the literature on spatial
ordi-
nance. probit models (Case, 1992; McMillen, 1992) by focusing on
The results of nonparametric estimators the are
heteroskedasticity
sometimes that is generally associated with spa-
not robust to changes in the bandwidth. Table 5 presents the autocorrelation per se. Heteroske-
tial models rather than
results when the bandwidth is varied. The results are not dasticity is potentially more serious than autocorrelation
highly sensitive to the bandwidth choice for residential and
mixed-use blocks. The results are more variable for com- TABLE 5.-BANDWIDTH VARIATION

mercial blocks, but they are stable for a fairly wide range of Estimate of Residential Minus Commercial
Land Value Growth
0.05 to 0.11. Table 5 suggests that our results are not an
artifact of the choice of bandwidth. Bandwidth Residential Mixed Commercial

The results of this study are in some contrast to the 0.05 0.162 0.177 0.512
findings in our earlier study of Chicago land values prior to 0.06 0.178 0.183 0.503
0.07 0.191 0.186 0.492
zoning (McMillen & McDonald, 1993). In that study, we 0.08 0.194 0.185 0.484
found that the presence of commercial or light manufactur- 0.09 0.196 0.184 0.480
ing use on a block face did not reduce residential land 0.10 0.198 0.184 0.478
0.11 0.199 0.184 0.477
values in the (as of 1921) unregulated land market. How- 0.12 0.198 0.184 0.740
ever, as we show in table 4, blocks with residential zoning 0.13 0.195 0.183 0.758
increased in value more than did other blocks. What can 0.14 0.192 0.183 0.675
0.15 0.189 0.182 0.660
explain this seeming discrepancy in empirical results ob- 0.16 0.186 0.182 0.632
tained from the same data source (Olcott's)? A simple 0.17 0.185 0.181 0.753
answer is that the residential zoning designation provides a0.18 0.184 0.181 0.754
0.19 0.184 0.180 0.729
guarantee that there will be no additional commercial or0.20 0.184 0.179 0.697
industrial use on the block of any kind, an assurance that did

This content downloaded from 190.116.36.68 on Thu, 01 Jun 2017 19:29:49 UTC
All use subject to http://about.jstor.org/terms
LAND VALUES IN A NEWLY ZONED CITY 71

TABLE 6.-ALTERNATIVE MODEL SPECIFICATIONS

Specification Residential Mixed Commercial

Spatial expansion model 0.204 0.147 0.465


[-0.009, 0.473] [0.044, 0.223] [0.030, 0.975]
h = 0.132 h = 0.127 h = 0.107
Base model with heteroskedasticity 0.293 0.180
[0.069, 0.548] [0.116, 0.241]
h = 0.130 h = 0.122
Spatial expansion model with heteroskedasticity 0.291 0.160
[0.111, 0.653] [0.069, 0.237]
h = 0.133 h = 0.127

The first entry in each cell is the estimate of residential minus commercial land-value gro

because it causes inconsistent estimates. The most com-


plied that zoning does not affect land values for mixed-use
monly estimated spatial model (Anselin, 1988) for contin-
blocks, but all estimated models in tables 4 and 6 suggest
that residential zoning increases land values by between
uous dependent variables is derived under the assumption
14% and
that the errors can be written in matrix form as v = pWv + 19% as compared to commercial zoning. The
e, or v = (I - pW)-l , where p is a parameter, Wresults
is anare less reliable for commercial blocks because
nearly all of these blocks are zoned commercial. The bas
n X n "contiguity matrix" that specifies the relationship
between the errors, and ? is a vector of independent and of table 4 and the spatial expansion model of table
model
identically distributed error terms. The model implies het- that residential zoning raises land values by approx-
suggest
eroskedasticity, and the variance for v depends onimately
the 47%-a high figure, but one that is consistent wi
number of observations near observation i. the later argument that the small number of nonconformin
Probit models are difficult to estimate using this error
uses in residential areas enjoyed local monopoly position
structure. However, the fact that standard models imply that
that might increase in value as land was converted
residential use.
heteroskedasticity is related to the density of nearby obser-
vations suggests the following estimation strategy. First, weThe matching procedure is a partial equilibrium ap-
estimate the density of observations near observation i using
proach. It provides an estimate of the increase in land value
a Gaussian kernel: gi = h-2n- 1 =i1 qp((Ei - for
Ej)la single block when it is zoned residential rather than
h)qp((Ni - Nj)lh), where E and N are normalized to have
commercial. Furthermore, the difference-in-differences ap-
proach estimates the change in land values associated with
unit variance. Again following Silverman (1986), the band-
/0.8 1/6 residential relative to commercial zoning. It does not pro-
width is h = - . We assume that the variance is a
vide an estimate of the level of land values in the absence of

function of gi: o2 = exp(agi), and estimateany


thezoning at all. Thus, our results cannot be used to
implied
estimate
probit model by maximum likelihood procedures. We the overall increase in land values caused by
rees-
Chicago's first
timate both the base model and the spatial expansion modelzoning ordinance. The results do suggest that
using this specification of the variance.10 residential zoning was valued by landowners in 1920s
The idea behind these models is to controlChicago. Zoning provides insurance against future intru-
for spatial
effects in the initial probit estimation to reduce sions of nonresidential
the effects of uses, and landowners were willing to
spatial autocorrelation. We continue to use the pay a premium to
bootstrap for this insurance.
construct standard error estimates. If spatial autocorrelation
remains in the errors, the standard errors will continue to be VII. Conclusion

biased. However, the spatial models establish that the results


Zoning ordinances generally remain in effect fo
for mean effects are robust to these specification changes.
time. Although variances and changes in zoning o
The results of the three alternative model specifications
ally occur, most areas of a city are zoned the same w
are presented in table 6. The simple difference-in-means test
decades. The long run is long in urban areas, and
of table 2, the matching estimator of table 4, and the spatial
the effects of ordinances that seldom change is diff
expansion model of table 6 all suggest that residential
introduction or occasional wholesale change of
zoning raises land values for residential blocks by approx-
zoning ordinance offers a unique laboratory for
imately 20% relative to commercial zoning. The estimate
rises to 29% for the heteroskedastic models, butzoning's effects. Although land use only change
the standard
land values can be used to infer the long-run effects
error rises also. The matching procedure is critical for the
new ordinance because they reflect the present d
mixed-use blocks. The simple difference-in-means test im-
value of the stream of rents.
Nearly all American cities have had zoning ordinances
10 The heteroskedastic models could not be estimated for the commercial
since
properties because of the small number observations withthereliable
1920s, but researchers have not taken advantage of
matches. their introduction to study the effects of zoning. The intro-

This content downloaded from 190.116.36.68 on Thu, 01 Jun 2017 19:29:49 UTC
All use subject to http://about.jstor.org/terms
72 THE REVIEW OF ECONOMICS AND STATISTICS

duction of a new ordinance to Chicago in 1923 offers a


Accounting for Heterogeneity in Programme Impacts," Revie
Economic Studies 64 (1997), 487-535.
natural experiment that allows us to determine the effects of
Heckman, James J., Hidehiko Ichimura, and Petra E. Todd, "Match
zoning on relative land-value growth rates. A simple model
an Econometric Evaluation Estimator: Evidence from Evaluat
of Chicago's hierarchical zoning ordinance suggests that
Job Training Programme," Review of Economic Studies 64 (1
605-654.
residential zoning might raise land values relative to com-
, "Matching as an Econometric Evaluation Estimator," Review of
mercial zoning if the presence of businesses causes negative
Economic Studies 65 (1998), 261-294.
externalities to homeowners. Furthermore, residential
Heckman,zon-
James, Hidehiko Ichimura, Jeffrey Smith, and Petra Todd,
"Characterizing Selection Bias Using Experimental Data," Econo-
ing created valuable local monopoly positions for conflict-
metrica 66:5 (1998), 1017-1098.
ing commercial enterprises that continued underHeckman,
grandfather
James, Robert LaLonde, and Jeffrey Smith, "The Economics
clauses. Alternatively, commercial zoning may prove to be
and Econometrics of Active Labor Market Programs" chapter 31 in
Orley
valuable because it provides more land-use options. Ashenfelter and David Card (Eds.), Handbook of Labor
After
Economics, vol. 3A (Amsterdam: North-Holland, 1999).
controlling for the tendency of the Chicago zoning board
Jones, J. P., to
and Emilio Casetti, Applications of the Expansion Method
assign commercial zoning to blocks with low land-value
(New York: Routledge, 1992).
McDonald, John F., and Daniel P. McMillen, "Land Values, Land Use, and
growth rates, we find that residential zoning led to higher
the First Chicago Zoning Ordinance," Journal of Real Estate
growth rates. Zoning's effect is consistent across initial land
Finance and Economics 16 (March 1998), 135-150.
use and is not sensitive to moderate changes in the band-
McMillen, Daniel P., "Probit with Spatial Autocorrelation," Journal of
width. Regional Science 32 (August 1992), 335-348.
McMillen, Daniel P., and John F. McDonald, "A Simultaneous Equations
REFERENCES Model of Zoning and Land Values," Regional Science and Urban
Economics 21 (May 1991), 55-72.
, "Could
Ahn, Hyungtaik, and James L. Powell, "Semiparametric Estimation of Zoning Have Increased Land Values in Chicago?,"
Censored Selection Models with a Nonparametric SelectionJournal
Mech- of Urban Economics 33 (March 1993), 167-188.
anism," Journal of Econometrics 58:1-2 (1993), 3-29. , "Land Use before Zoning: The Case of 1920s Chicago," Regional
Science
Anselin, Luc, Spatial Econometrics: Methods and Models (Boston: Klu- and Urban Economics 29 (July 1999), 473-489.
wer Academic Publishers, 1988). Mills, Edwin S., "The Value of Urban Land" (pp. 231-253), in Harvey
Perloff (Ed.), The Quality of the Urban Environment (Washington,
Babcock, R., "The Chicago Zoning Ordinance," Northwestern University
Law Review (1957), 174-201. DC: Resources for the Future, 1969).
Barnow, B., G. Cain, and A. Goldberger, "Issues in the Analysis ,"Economic
of Analysis of Urban Land-Use Controls" (pp. 511-
Selectivity Bias" (pp. 43-59) in E. Stromsdorfer and G. 541), in Peter Mieszkowski and Mahlon Straszheim (Eds.), Current
Farkas
Issues
(Eds.), Evaluation Studies Review Annual, vol. 5 (Beverly in Urban Economics (Baltimore: Johns Hopkins University
Hills:
Sage Publications, 1981). Press, 1979).
Pagan, Adrian,
Case, Anne, "Neighborhood Influence and Technological Change," Re- and Aman Ullah, Nonparametric Econometrics (New
gional Science and Urban Economics 22 (1992), 491-508. York: Cambridge University Press, 1999).
Pogodzinski,
Casetti, Emilio, "Generating Models by the Spatial Expansion Method:J. M., and T. R. Sass, "Measuring the Effects of Municipal
Zoning4Regulations: A Survey," Urban Studies 28 (August 1991),
Applications to Geographical Research," Geographical Analysis
(1972), 81-91. 597-621.
Rosenbaum,
Cosslett, Stephen, "Semiparametric Estimation of a Regression ModelPaul R., and Donald B. Rubin, "The Central Role of the
with Sample Selectivity" (pp. 99-144) in W. A. Barnett, J. Propensity
Powell, Score in Observational Studies for Causal Effects,"
and G. M. Tauchen (Eds.), Nonparametric and Semiparametric Biometrika 70:1 (1983), 41-55.
Methods in Econometrics and Statistics (New York: Cambridge , "Reducing Bias in Observational Studies Using Subclassification
University Press, 1991). on the Propensity Score," Journal of the American Statistical
Driscoll, John C., and Aart C. Kraay, "Consistent CovarianceAssociation
Matrix 79 (1984), 516-524.
Siegan, Bernard,
Estimation with Spatially Dependent Panel Data," this REVIEW 80:4 Land Use without Zoning (Lexington, D. C. Heath and
(1998), 549-560. Co., MA: 1972).
Fan, Jianqing, "Design-Adaptive Nonparametric Regression," Silverman,
JournalB.of W., Density Estimation for Statistics and Data Analysis
the American Statistical Association 87 (1992), 998-1004. (New York: Chapman and Hall, 1986).
Wirth, L.,
Fischel, William A., "Do Growth Controls Matter?," (Cambridge, MA:and E. Bernert, Local Community Fact Book of Chicago
Lincoln Institute of Land Policy, 1990). (Chicago: University of Chicago Press, 1949).
Heckman, James J., Jeffrey Smith, and Nancy Clements, "Making
Yeates, the
M., "Some Factors Affecting the Spatial Distribution of Chicago
Most out of Programme Evaluations and Social Experiments: Land Values," Economic Geography 41 (1965), 57-70.

This content downloaded from 190.116.36.68 on Thu, 01 Jun 2017 19:29:49 UTC
All use subject to http://about.jstor.org/terms

You might also like