Reproduced and distributed with the compliments of the University of
TfZTl f^"^^ Extension Division for educational purposes, and
pLn^ ""^^ ^^^''^ expressed in this report are those of
the RAND Corporation .
A CITYAND ITS SUBURBS
PREPARED FOR THE NATIONAL SCIENCE FOUNDATION
BARBARA R. WILLIAMS
R-1353-NSF
AUGUST 1973
SANIA MOMCA, CA. 90406
PREFACE
This report summarizes the results and implications of a year of analytical work
concerning the decline of St. Louis as the central city of a metropolitan area, the
implications for the city's future, and policy strategies for improving these future
prospects. Although the work is primarily that of The Rand Corporation, vital
assistance was provided by faculty members at three major universities in the St.
Louis area: Washington University, St. Louis University, and the University of
Missouri, St. Louis. The work was supported by the National Science Foundation.
The project was one of a series of three urban studies that also includes an analysis
of causes, effects, and control of two decades of explosive growth in San Jose, and
adaptation to acute aerospace recession in Seattle.
This report is oriented primarily toward the policy implications of the St. Louis
analysis, and is intended for use by policymakers concerned with the future of the
area.
iii
SUMMARY
This report summarizes the research findings and policy imphcations of a series
of studies conducted under the St. Louis project of the Rand Urban Pohcy Analj^^
Program. The analysis in St. Louis has been directed toward evaluatmg alternative*
for decisionmakers at local, state, and federal levels who must deal with urban
Sie central concern in St. Louis is the city's significant decline in popiUation
and economic activity that occurred in the 1960s, and the rapid rate ofbmWing «id
neighborhood abandonment that accompanied it. Because abandonment itse f might
open possibilities for new development, the report asks what those possibiliti^^e
how probable their achievement is. and how policymakers might encourage the
reali^tion of desired changes. Three possible futures for the "ty are posed: con^
tinued decline; stabilization in a new role as an increasingly black suburb aiid
return to a former role as the center of economic activity in ""^^^^P^^;^^^^^^^
As things stand, the most likely prognosis is for contmued decline. G ven outage
revenue sources, however, the city of St. Louis might assume a new « /^^^
suburb among many other suburbs, making the transition easier for its population
and institutions.
ORIGINS OF ST. LOUIS'S PRESENT SITUATION
. Decline in St. Louis is mainly a function of the same trends that have stimu-
lated movement from central cities to suburbs across the United States The fast-
paced decentralization of all American urban areas since World War II has been
stimulated mainly by the desire for living amenities and productive facilities that
could be provided most easily and cheaply at the periphery of cities. Rising incomes
and improved transportation systems have facilitated the move to the suburbs.
Certain federal policies-real estate tax incentives, interstate highway develoj^
ment, FHA and VA mortgage programs-have accelerated these trends.
. However, St. Louis differs from some central cities in manifesting rapid and
absolute declines in central city population and business activity. is. rapid
metropolitan growth has brought explosive suburban development to some Ameri-
can ur^an areS-mostly in the West and parts of the South-without inducing high
rates of abandonment in their central cities.
V
VI
• jS^. Louis's unusual rate of decline has come about because many phenomena
that appear to accelerate central city decline in older metropolitan areas combine in
unusual strength there. Many central cities of the East and Midwest contain a large
stock of housing and industrial capital that is old and expensive to maintain and
restore, further increasing the advantages of suburban location. This is particularly
evident in St. Louis, a large portion of which was urbanized before 1900.
The large amounts of flat farmland around St. Louis also made decentralization
easier. Such land was readily developed for industrial and residential uses.
St. Louis is a small city of 61 square miles. Its political boundaries, frozen since
1876, have prevented the city from expanding its resource base as its proportion of
disadvantaged residents increased.
Large in-migrations of groups that vary from the existing population— such as
rural, low-income families— appear to hasten the departure of more affluent families
to the suburbs. St. Louis has been an important portof entry into urban life for rural
migrants. The fact that many of these migrants are black seems to have precipitated
the rapid departure of whites from particular neighborhoods; however, our research
suggests that departures of people from the city are more class-related than race-
related. At most income levels, blacks and whites left at about the same rate during
the 1960s.
PROGNOSES FOR ST. LOUIS
The report argues that without major policy changes beyond the local level, the
city will most likely continue to decline. It is unlikely either to become a stable,
increasingly black suburb or to return to its former central economic function. Several
demographic and economic trends induce this conclusion:
• Heavy and prolonged out-migration of the city's younger white residents has
left behind an elderly population in which death rates exceed birth rates. Growth
of the white population therefore depends on massive in-migration- an improbable
development.
• While the city's black population continues to grow through natural in-
crease, it began to decline in 1969, indicating a net migratory loss severe enough to
offset its natural increase.
• The city and its suburbs received unequal shares of metropolitan economic
growth during the 1960s, with most industrial sectors declining in the city and all
industrial sectors growing in the suburbs. If industrial location trends during the
last half of the 1960s continue for another five years, St. Louis County* will contain
that share of metropolitan business activity usually characteristic of a central city.
Neither a survey of industrial developers in the area, a 1967 survey of people's
expressed preferences for residential location in the area, nor more episodic evi-
dence about industrial and residential location since 1970 supported the hypothesis
that past decline has created new conditions in the city that will mitigate or even
reverse past trends.
The city has not '^bottomed out" so that large blocks of inexpensive empty land
will readily stimulate new forms of investment. Rather, land remains relatively
•The City of St. Louis is entirely separate in area and jurisdiction from the County of St. Louis.
vii
expensive to develop in the city. Nor have reduced numbers of people and businesses
made public goods and services unequivocally easier to provide. Indeed, metropoli-
tan decentralization has reduced the city's share of more affluent residents and
increased its share of disadvantaged ones. Economic growth, as well, has gravitated
to the suburbs. Thus, public revenues have become progressively more difficult to
generate locally. New resources^ available to the city from sources outside the city,
are essential to any improvement
ALTERNATIVE STRATEGIES FOR THE FUTURE
Just as local policies did not cause the decline, local policies cannot readily
change the trends and characteristics that did cause it and are still operative. The
analysis suggests that, among the alternatives open to the city, promoting a new role
for St. Louis as one of many large suburban centers of economic and residential life
holds more promise than reviving the traditional central city functions.
One strategy for assuming a more suburban role is to entertain administrative
or jurisdictional changes that would allow municipal services and regulations to be
geared to varying neighborhood needs. In this way, the city's large, heterogeneous
population might capture some of the benefits of small homogeneous (and affluent)
suburban municipalities where residents can purchase and control the public goods
and services they want.
To succeed, however, this strategy will require new outside resources— new
mechanisms for generating revenues that make the poor a smaller financial burden
for the jurisdictions where they live. Several mechanisms for doing this are offered
for consideration:
• A more substantial federal revenue-sharing program.
• A state revenue-sharing program to support selected public goods.
• A metropolitan revenue program, sharing revenue generated by industry in the
metropolitan area.
• A metropolitan earnings tax.
ACKNOWLEDGMENTS
This report has been reviewed by Stephen Crocker, Paul Jordan, John Koehler,
Robert Levine, Don Rice, and Gus Shubert of The Rand Corporation; by Professor
William Alonso, University of California, Berkeley; by Norman Murdoch, Director
of the St. Louis City Plan Commission; by Dempster Holland and George Wendel of
St. Louis University; by Peter GrandstafT, Robert Markland, and Hugh Nourse of the
University of Missouri, St. Louis; and by James Little of Washington University. All
of the above have helped formulate the findings reported in this document— in many
cases by arguing against them. However, members of the St. Louis research project
developed these findings, and the author takes full responsibility for the implica-
tions drawn from them.
The author wishes to thank those who carried out the analysis on which this
report is based: Sinclair Coleman, Leola Cutler, Peter deLeon, John Enns, Cyrus
Gardner, Peter GrandstafT, Marie Hoeppner, Dempster Holland, Charles Leven,
James Little, Robert Markland, Peter Morrison, Hugh Nourse, Gerald Payne, Betsy
Schmidt, Richard Slitor, and George Wendel.
ix
CONTENTS
PREFACE • iii
SUMMARY V
ACKNOWLEDGMENTS ix
Section
I. INTRODUCTION 1
n. METHODOLOGY 4
Conceptual Design 4
Research Design 7
III. RESEARCH FINDINGS 10
Models of Urban Change 10
Demographic Analysis 14
Economic Changes 24
Racial Factors 28
Policy Accelerators to Central
City Decline 31
IV. ALTERNATIVE STRATEGIES FOR THE FUTURE 39
Alternative Futures for St. Louis City 39
Living with the Future 43
Strategies for Local Policy 45
BIBLIOGRAPHY
I. Documentation 49
II. Selected Background References 51
xi
I. INTRODUCTION
The St. Louis Standard Metropolitan Statistical Area" encompasses the City of
St Louis and six counties lying on both sides of the Mississippi River: St. Louis, St.
Charles, Franklin, and Jefferson Counties in Missouri, and St. Clair and Madison
Counties in Illinois.^ Most of the SMSA lies in Missouri, three-fifths of it west of the
City of St. Louis; the Mississippi forms the eastern boundary of its central city. The
City of St. Louis is entirely separate in area and jurisdiction from the County of St.
. Louis. The area surrounding metropolitan St. Louis is semirural, dotted with medi-
um-sized towns. The closest metropolitan area of comparable size is Kansas City,
about 275 miles away.
In 1970, the population of the St. Louis SMSA was about two and a half million
From 1960 to 1970 it had increased by only 12 percent, a rate lower than the average
national metropolitan increase of 17 percent. Economic growth in the area has also
been slow. As shown in Table 1. in the late 1960s the St. Louis area lagged behind
the rest of the nation in growth of total income, per capita income, and employment,
c^o * . ^""^ ^"'^ together contain about three-quarters of the
SMSA s population-622,000 in the city and 951,000 in the county.' A strong and
persistent westward progression in the area's settlement pattern has steadily
drained St. Louis's share of area population and economic activity. From 1960 to
1970, the city's population declined 17 percent while the suburban population in-
creased by nearly a third; jobs declined close to 15 percent in the city but nearly
doubled in the suburbs.
Stark reminders of the city's demographic and economic losses are 2200 vacant
and vandalized buildings, occupying an average of one-tenth of an acre each. Under-
standably, St. Louis believes it is plagued by that set of problems widely lumped
together as 'the urban crisis": a declining tax base, rising costs for providing ser-
vices, a high crime rate, a problematic school system, high unemployment, racial
mequities, and a spectacular rate of building and neighborhood abandonment.
Uur analysis does not attempt to diagnose completely this formulation of the
urban crisis. We give slight attention to problems that occur with great frequency
in th.?rej^rt"" "metropolitan St. Uuis" are used interchangeably
been add'^J °t!! fh.^^T"'' ''^^i.^A "^"""''^'^ ""^^ M""^o«> have since
S of mo ^^'^ '"^ ^^'^ however, are based on the SMSA definition
» Hereafter. "St. Louis" will refer to the city. Arhile St. Louis County will be so designated.
1
2
Table 1
COMPARISON OF NATIONAL AND ST. LOUIS
SMSA ECONOMIC GROWTH, 1966-1970
Item
Average Annual
Growth. 1966-1970 (%)
U.S.
St. Louis SMSA
Total employment
1.9
1.0
Total income
2.8
2.2
Per capita disposable income
3.2
l.A
Data provided by Regional Economics In-
formation System, Bureau of Economic Analysis, Office
of Business Economics, U.S. Department of Commerce;
and from Eaonomic Report of the President, 1971.
not only in cities, but in other places as well: crime, poor schooling, unemployment,
racial inequities. Instead, we concentrate on events more peculiar to the structural
change cities are undergoing, exploring the implications for a central city of the
redistribution of population and economic activity in its metropolitan area. For this
purpose, we chose building and neighborhood abandonment in St. Louis as the
initial focus of our study.
Abandonment is both a symptom and a cause of problems, but by making land
available for new uses, it also offers possibilities of new directions for the future. Our
objectives were to determine what those possibilities are, how probable their
achievement is, and how policymakers might encourage the realization of desired
changes. For purposes of analysis, we pose three alternative futures for St. Louis-
continued decline; stabilization in a new role as an increasingly black suburb; and
return to a former role as the center of economic activity in the metropolitan area *
Which of these futures is the likeliest, and how policymakers might affect future
directions, are analyzed by examining past change and current growth potential in
St. Louis as a guide to the future.
The analysis has required methods and concepts from more than one discipline.
We have tried here-as in other Rand urban studies-to assemble a broad spectrum
of evidence bearing on the reasons for St. Louis's current condition and prognoses
for Its future from a number of disciplines, including demography, economics, soci-
ology, and political science. To interpret this evidence, we have used statistical
techniques and explored various lines of argument to form a coherent picture of
where the weight of the analysis and prognosis lies.
The picture, in brief, suggests that decline in St. Louis is mainly a function of
the same trends that have stimulated movement from central cities to suburbs
across the United States. The rate of decline is more acute in St. Louis because
several factors that accelerate decline combine in unusual strength there. The
problems of decline, however, do not lie in the fact that population and business have
aonLT^fK* fl^ M "''^ «"CO™Pa*' a'' possible futures for the city. They include those futures that
appeared both feasible and representative of the range of possibilities after initial data analysis.
redistributed themselves within the metropolitan area, with the city losing and its
suburbs gaining. Rather, the problem for the city is that it wants to remain attrac-
tive to its current residents, to its metropolitan population, and to visitors-offering
well-maintained public goods and high levels of public services geared to a variety
of tastes. Its resources to do this, however, are more and more diminished because
Its more affluent citizens have moved to the suburbs in massive numbers, and
economic growth has gravitated with them.
These two issues— metropolitan population and economic redistribution, and
city redevelopment-are separable. However, our question is whether the redistri-
bution has created new conditions in the city (e.g., more available land) that will
attract new investment essential to central city redevelopment. Our analysis leads
us to believe this is unlikely: relative to its own conditions in the past, St. Louis is
better off in certain ways; relative to its suburbs in the present, St. Louis remains
less attractive to new investment in important ways. We suggest that instead of
focusing on the city as a geographic entity needing restoration, local policymakera
focus on St. Louis's assuming a somewhat different role, functioning more like a
suburb or a set of suburbs within the metropolitan area. However, as with all other
policy directions we considered, this too requires additional revenues not now avail-
able to the city.
Section II of this report describes our research methodology. Section III— the
keystone of the report-then applies the methodology: St. Louis is examined succes-
sively in terms of demographic trends, both within the parts of the metropolitan
area and between it and other areas; in similar economic terms, examining both the
economy of the area and the division of the economy between the central city and
Its suburbs; in regard to racial hypotheses about urban change, such as "white
flight"; and finally, in terms of the additional impetus to decline imparted by federal
and local policies.
The final section of the report turns to policy strategies for the future. These
strategies do not take the form of specific policy recommendations for St. Louis.
Rather, they form more of an agenda for policymakers. Time and budget constraints
required us to be selective in the analysis of the complex syndrome of urban decline-
important parts of the research we think necessary to decisionmaking in St. Loui^
either remain to be done or are still in process at local universities in the area. Our
analysis has been directed toward examining alternatives for decisionmakers at
local, state, and federal levels who must deal with urban decline
V
II. METHODOLOGY
CONCEPTUAL DESIGN
Initial View: Abandonment as a Problem
Lou,?w«?hf V' •? '^P'^ neighborhood change and abandonment in St.
^.Z r f '"1 u ''''y other knowledgeable
2nHo ; abandonment to be the city's most serious problem. Second.
tuXrr ^'f 'r^'^'" ^^^^^^^ appropriate concejv
tual counterpoint to the phenomenon of rapid growth central to our research in the
ban Jose metropolitan area.
In the St. Louis research, abandonment is considered to have occurred wherever
there are empty, vandalized buildings and empty spaces within older neighborhoods
For our purposes empty buildings and land are seen as part of the chy's spatial
crnr."^ ~'"l t "'"'^ ''"'^ of inventory as goods held in the anticipation of
changing demand. Empty land is most appropriately called inventory, since it pro- "
v^es space for the expansion of existing land uses or the introduction of new land
oT; ^^^^^^-^ be thought
Citv Ra?rorr' "^^ 1" 1936, the St. Louis
City Plan Commission published a study of shifts in land use that said:
tekeJIhfcitv'^sSj ''Yf' terms-if adequate measures are not
Sntral areas o^^^ gradual economic and social collapse. The older
Sntinue until Ih 9 ^emg abandoned, and this insidious trend will
continue until the entire city is engulfed.'
S^eL'^ ^^^^'"'"endations of that Commission report could be fully implement-
t.^Z u ^"'"^ " -P'^ly accelerated job opportuni-
les at the same time that new housing construction was narrowly restricted. By
1950 these conditions swelled the city's population to 857,000. close to its peak of the
century a few years later. With the end of the war, new housing construction grew
Ix.Jis!S:r[:MaS"l9T2' Jif'' Washington University. St.
4
vigorously— but in St. Louis County, not in the city— and existing densities in the
city placed the county at a greater competitive advantage.
During the 1960s, abandonment in St. Louis was widespread enough to warrant
three studies funded by the U.S. Department of Housing and Urban Development,'
a study by the Urban League, and a survey sponsored by the City Plan Commission
on the problem of residential blight." It also stimulated a variety of research inter-
ests and projects at local universities on problems of neighborhood transition. We
have relied on this past and continuing research to provide richness of detail in the
microanalysis of neighborhood change, and deployed our resources toward analyz-
ing the possibilities of alternative land use for the future— viewing abandonment as
inventory.
2. Although abandonment figured as a major problem in most of our early
discussions with local officials and citizens, many people understood that abandon-
ment could also be viewed as an opportunity. We therefore reexamined a basic
question: For whom is abandonment a problem? Since the voluntary relocation of
households or businesses from one area to another ordinarily suggests improvement,
abandonment may be viewed as an indicator of rising incomes, better housing, and
land use opportunities for some people and businesses.
For people left behind, however, some problems worsen. Public services decline
—at the very least, in effectiveness— as the private incentives to housing and neigh-
borhood maintenance diminish. And large blighted areas not only uglify a city but
can mtensify old problems and generate new ones, such as increased vulnerability
to fire.
While the problems of declining areas often dominate local concern, many St.
Louis decisionmakers had begun to view abandonment as a possibility for improve-
ments. As evidence, several significant municipal codes were changed in the 19608:
until the early 1960s, owner-protective code requirements caused demolition of a
building to lag behind notification of condemnation by six months to a year; by 1970,
code revisions had reduced that time to one week.* Since October 1970, the city has
been engaged in a massive program, financed primarily by the federal government,
to remove condemned buildings."
Restructuring: The Future of the Central City
Given the growing view of abandonment as opportunity, it seemed that policy-
makers in St. Louis might benefit from research oriented toward future expecta-
tions, asking how the city might manage the inventory it is accumulating How a city
manages its inventory largely depends, of course, on what it can anticipate about
//niynf wT'^'"^"'n!^"K"''"^.^"'^ Development. A Study of the Problems of Abandoned
SvD7rlthrD^\^^^^^ """-^ ^''^y bourse and Lmc. LiUle.
' The Center for Community Change and the National Urban League, The National Survey of Hous-
ing Abandonment. .\pn\
* Alan .M. Voorhees and Associates, Inc., Technical Report on a Residential Blight Analysis for St.
Louus. Missouri. Washington, D.C., March 1969.
» City of St, Louis. Ordinance 55681, Section 2126.1, approved July 15, 1970.
• In January 1972, MUD dwlared a moratorium on federal funds (or demolition. In addition, there
trim^ri"^, f °" 9'"'* •"°"''>' ^'"^ "sed for demolition. The city has now allocated
?.i.4m),ouo Irom rcu'nue-sharinj; money for continuation of a lart,'e-.scale demolition effort. Despite re-
moval 01 much ol the oldest housing stock in the city, median rents fell from $66 to $57 and median
housmg values fell from $12,000 to $11,000 (in 1967 dollars) between 1960 and 1970
6
future change. Thus, Rand's research in St. Louis has been structured to answer the
following questions:
• Without major policy changes, which of the following possibilities will most
likely occur in the St. Louis metropolitan area over the next ten to twenty years?
Continuation of past trends. The rates at which people and jobs depart from the
city accelerate, and their rates of entry remain sluggish. The city's inventory ac-
cumulates but competes poorly with surrounding suburban inventory for either
business or residential investment. Selective out-migration causes the city s resident
population to become smaller and older, with a growing proportion of disadvantaged
persons, many of them blacks. Under these conditions, the city would presumably
go into bankruptcy^ and become something like a ward of the federal government— a
jurisdiction incapable of generating locally the revenues with which to manage
itself
Stabilization of current growth potential. As inventory accumulates in the city,
it fails to compete with suburban locations for most types of industry. However, the
black middle- and lower-income residential population in the city shows growth
through natural increase. One of many suburbs, St. Louis retains the usual comple-
ment of people-serving industries that prosper as its population increases either in
size or affluence.
Reversal of past trends. The city's accumulating inventory exhibits a selective
competitive advantage over suburban inventory with respect to industrial develop-
ment for which central location is a dominant consideration. Economic decline
gradually "bottoms out" as past decline creates new conditions— available land,
decreasing population densities— that attract new growth. The city again becomes
an active hub of economic exchange within the metropolitan area.
• Within the next ten to twenty years, can specified policy changes at the local,
state, or federal level alter the likelihood that one or another of the above possibilities
will occur?
In answering these questions, we have drawn on the conceptual tools and ana-
lytic methods of economics, sociology, demography, political science, and statistics.
Our diverse research efforts, however, shared a common perception of basic urban
processes. We assumed that metropolitan areas represent a set of political bound-
aries (central cities, counties, smaller municipalities, etc.) normally subject to a
more or less continuous procession of people and jobs entering and leaving. Any
jurisdiction's population grows as it attracts more migrants than it loses and as it
experiences more births than deaths. Its economy grows as its firms expand, produc-
tivity increases, and jobs show a net increase.
Although population and employment have been suburban izing for many
decades, these changes have been especially pronounced since World War II. During
the 1960s, an unprecedented number of the nation's central cities not only ceased
to grow but lost population. Fifteen of the 21 central cities with over half a million
residents in 1960 ended up losers, and 6 reported losses of 10 percent or more. The
degree of decline in St. Louis may be exceptional, but St. Louis is no exception to
the rule.
' Bankruptcy means that the city would no longer perform the existing level of services because of
an inability to pay bills, meet payrolls, etc. This form of fiscal crisis is discusst^d in the Advisory Commis-
sion on Intergovernmental Relations, City Financial Emergencies: The Intergovernmental Dimension
(forthcoming).
exXaUrs for /^^^^ ^^'"^^ "'^^ yards, have been the usual
falHn^f . ^ decentralization ofurban population. Changing technology and
w TdtXTicfe^ r ^"^^^^ industriafdeLtralizatTon a.
for es Nat oLl^^^^^^^^ accelerated the trends set in motion by these market
creased homeowners' acce'ss to1he!l"'bs '
in J!"^ f^^t^o^' these market forces and federal incentives should be much alike
^ aiThlTuf ''T' ^^^^ -"-^ --'^-b'y from one a ea
i^Cnces to l: f'^^ We understand these different outcomes, despite common
popuLtion coin V . '"'"'■P^"^ ^'^"'^^"'•^J in local
population composition, industrial mix. governmental makeup, age topography
and region; and (b) exogenous shocks peculiar to certain areas (e g part cE^^J
areas, and legation propensities of major export industries).
f«IHn!V ^"'^ research, we assumed that market forces (rising incomes
fallmg transportation costs,' changing tastes, technological change) have provS
city to suburb. Further, we have striven to identify the marginal accelerators of
urban change more explicitly-the additional forces that have Le erat^d the^
national trends operative in the St. Louis metropolitan area.
they promifet be th' ^^'^^^J^^''^^' '""^'^--^ policymakers to understand, as
level To the ex^n ,7^"^ 'T'' "lodify-though perhaps not at the l<Kal
juniiSons S urL ^'^^f '"^^^^ sources exogenous to local
s one onhe reas^^^^^^^^^^^ ""'^ ^'"'^^ ^« '^^^"^^ ^hem.» This
will L addressed to n.K , '^'"^^T ^P^^i^c cities
De addressed to national as well as local policymakers.
RESEARCH DESIGN
To assess the likelihood of alternative futures for St Loni. T^.^A , a ^ .
sen. of specinc research designed U, .ns.:/^!]::!:,^^^ ' '
• 'n^: „7;r:rsr r::: r -tardea
8
Selection of Variables
Three sources guided our selection of conditions and policies to explore- the
ZrZZ '''''' ^^^--^ '° R-d-s con!
^tTknn M c'^t'^'''-" ^''^^"^'"^ informal discussion
with knowledgeable St. Louis citizens and local officials
d^lli; ''''''^".'"^''u^* conditions have been unique in determining St. Louis's
fatZi. f P«'"t^d to the city's aging physical stock- and the
Xd to h " '^^'^'^ '"""^ ''^^ '^^^ 1950s. Local policies
Sfedtold ' ' ZT ''""^ ^"'^'^ ^^^""^ "^"-^ ^hose related to the city's
Zlr^^^!l ?f o^-it« banking community. We repeatedly
heard about the set of local decisions that, in effect, have frozen the city's boundaHe^
to encompass 61 square miles since 1876. exacerbating the subsequent effect of
ly dTsSbeTr'-r f'T f-^her out. City banks, frequent'
acquTre oLllT ^ 7Tf '"'^'^ ^""^ '"^^^ "^'^ ^^P'^^l difficult to
acquire locally. We subjected each of these conditions and policies to as thorough an
empirical test as we could devise with available data sources
esta^tJf T' '"^^"'•es of the incentives that various real
ari%^ • ^TJ*"^" of investment in the St. Louis metropolitan
IZLZ J T 'r™"""* J'"" have influenced industrial and
ITother ^h""^"S ''•^^^^ fro- one part of the area to
economv totef '""''^'''f ' ^'''"'^ ^^^^^^ ^^^"'^'"^^ «^'he metropolitan
economy to be an appropriate complement to the detailed demographic anal^is we
rfC^T'""' 'T' T''''' "° d^^'^'-^-^^ between the structure
Inomv w^lJh ""'"'T ^'^^"^"^y (or changes in it) and that of the national
economy which might be responsible for the high rate of decline in the central city.
An analy IS that would go beyond the documentation of structural changes in the
bwTnve' r «^^hose changes (i.e.. tracing decline in some sector to
Z ^^''^ '^"^ of investment) required more re-
sources than we had available."
prof^slttlhrer"'-'"' ^^^^"^^ ^t-^-d
professors at three universities in the St. Louis area: Washington University St
^uis University, and the University of Missouri. St. Louis.- Our univeS'co t
o:^:s^'^ad^"^''"^^''^ '^^^'^^"^^"^ analysis of pnC dat^
sources. In addition, concurrent research funded by the Department of Housing
See Sec. II of Bibliography,
tion'.' m^S^NSF^Sli^jt''' """"^'^ •''^-^ 5.;nma;^Po/,c,5/a^.;„.„, The Rand Corpora-
Lol Urv^tf ^ItsT^"^^^^^^^^ b":'^ 't'-"- ^^''O- De-PSter Holland. St.
portions of several dtiesandudSrSu^^^^^^^^^ ^'^^ abandonment and those
Midwestern Cities," 1973 (unpublished pajr) ^"^ P°P"'^"°" Change in Seven
Moill^fSo'^,?;:^^^^^ ^^P'^r^ F. Fisher, et al.. "A
1972. See especially' thVf^compln^ r'Sment." 2- March
HolL'J^S^TnXnSt L^TsSiveSvVpT"^"'^.'"^^^ ^^'^''^'^^ Dempster
9
and Urban Development^® has furnished data germane to our research interests. We
have also relied heavily on secondary data sources.'^
Projections of Trends and Current Composition
It is a tricky business to project the future of anything so complex as a city.
Simple extrapolations of trends are particularly vulnerable to unforeseen shocks
such as technological innovations or new federal policies. And today's linear trend
may become exponential tomorrow.^® For example, the normal filtering of housing
in a metropolitan area — the orderly passage of successively lower income groups
through the housing stock — may turn disorderly and set off large-scale disinvest-
ment if income distributions between successive groups vary sharply.
In various ways, we have sought to strengthen the projections discussed below
with types of data that explain, rather than simply describe, the salient trends. For
example, in assessing the future trend in migration away from the city, we rely not
merely on descriptive census data but also on survey data that reveal people's
intentions and expectations about moving. And in considering the crucial role of the
automobile in patterns of metropolitan settlement, we have weighed the possible
effect of the rising price of gasoline on automobile use. We cannot foretell every
exogenous shock to the metropolitan area, however. While we take a systematic
approach to those contingencies we can now identify (e.g., the rising price of gaso-
line), our projections are firmly anchored in the caveat, 'If everything else remains
the same. .
Corporation, R-1358-NSF (forthcoming). Charles Leven and James Little have received a separate grant
from the National Science Foundation to do a survey of movers within selected migration corridors of
the SMSA, aimed at developing a model of residential preference, entitled, A Study of Determinants of
Inter-Neighborhood Mobility, GI-37861-NSF.
HUD Grant M0PD4, 1972.
The U.S. Census; Office of Business Economics, U.S. Department of Commerce; interviews carried
out as part of the City Plan Commission's survey of residential blight; and vital statistics data furnished
by the City's Department of Health.
Or it may change direction as well.
III. RESEARCH FINDINGS
Starting with a model that provides the underlying explanation of population
redistribution in all major metropolitan areas, we then examine conditions associat-
ed with dilferential growth rates in central cities. After that, we analyze change in
St. Louis from a number of standpoints:
• Demographic, including trends in the city and the suburbs, changing re-
placement capacity of various components of the city population, implications of
demographic trends, and effects of interurban migration.
• Economic, including implications for the city of slow growth in the met-
ropolitan area, changes between the city and the suburbs, and the results of a survey
of industrial developers taken specifically for this analysis.
• Racial, involving various tests of the accelerating effects of racial aversions
on jurisdictional and neighborhood change.
• Financial and legal, including contributions of current policies that acceler-
ate other trends, especially federal highway and income tax policies; local jurisdic-
tional boundaries; and the effects of local banking conservatism.
MODELS OF URBAN CHANGE
St. Louis is by no means unique among American central cities in showing
absolute declines in population and jobs in the 1960s. Two events of the last twenty
yearsr— rising incomes and falling transportation costs— have affected central cities
in the United States in such a way that all of them should either be growing more
slowly than their suburbs or experiencing absolute declines in population and jobs.
That is, it can be argued that in concert, the desire to provide and consume public
services collectively (because they are cheaper that way) creates powerful incentives
for firms and households to locate near one another. But countering factors make
for dispersion: land is cheaper away from the central city, and more space is avail-
able for modern spread-out, low-rise industrial and commercial operations. Further,
cities typically provide a fixed bundle of public services intended to be uniform
across neighborhoods. However, varying demands for public services are imposed by
diflTerent subgroups (young families, the aged, higher income people, lower income
people). To get the level and mix of public services they want, households must move
to or create jurisdictions containing people of similar needs or tastes. Thus, the
10
11
heterogeneity of a city's population itself creates incentives for subgroups to disperse
and regroup in more homogeneous jurisdictions.
Although cities, under many conditions, maintain a tenuous equilibrium be-
tween the opposing forces that impel clustering and dispersion, rising incomes and
falling transportation costs have tipped the scales in favor of dispersion to the
suburbs.
Within that broad truth, growth rates in central cities still vary. To compare
the strength of conditions associated with differential growth, Emmett Keeler and
William Rogers devised a simple three-equation structural model of central city
change in metropolitan areas of over 250,000 population.' The variables used in the
model, with their means and standard deviations, are given in Table 2. This model
assumes that SMSA total income growth and population growth are jointly deter-
mined, and that SMSA population growth together with other exogenous variables
determine central city population growth (see Fig. 1). Applied to 124 urban areas,
the analysis shows SMSA income and population growth closely linked. It can be
seen from the first equation in Table 3 that Congressional power, stronger city
governments, and manufacturing have added to SMSA income growth. Natural
increase and a good climate independently add to SMSA population growth (second
equation). Central city change is mainly related to SMSA population change, but
older cities with more old or black citizens lost more population, even with SMSA
population change taken into account (third equation).
Applying this model to St. Louis,^ we found that the mo/or phenomena associat-
ed with the city's 2 percent annual rate of population decline were the slow growth
of the metropolitan area (-0.33%), the city's age (-0.7%), its high percentage of
black population in 1960 (-0.3%), a limited-power mayor form of city government
(-0.2%), high median age of the population in 1960 (-0.2%), and a high density
(-0.2%). The city's age showed a stronger association with decline than did any of
the other variables. This finding supports other evidence (discussed below in the
subsection on racial change) that associates decline with an old housing stock. It
should also be noted, however, that the model explained only 60 percent of St.
Louis's rate of decline (the actual rate was 2.5 percent a year; the model explained
1 .5 percent),^ so that other factors, not included among the explanatory variables,
are also important in St. Louis.
We now turn to other analyses that examine factors included in the models we
have just discussed, and additional factors useful in understanding how St. Louis got
where it is. We also discuss the implications of these factors for projecting the city's
future.
p 1 9^1m^ iJr 1 D^o c^^^' ^ Classification of Large American Urban Areas. The Rand Corporation,
K-1 J46-Nbt , May 1973. See Appendix D therein for applications of the model to 124 urban areas, 59 urban
areas with central cities over 200,000, and St. Louis.
' To estimate the effect of different St. Louis characteristics on the city's growth rate, values of the
dependent vanab es for bt. Louis were multiplied by the parameters estimated in the regression for 124
urban areas (Table 3).
' For some other declining cities-e.g., Pittsburgh, Baltimore, and Newark-the model explained
almost all the decline.
12
Table 2 i
SELECTED STATISTICS FOR 124 METROPOLITAN AREAS^
Variable
Abbreviat ion
Mean
Standard
Deviat ion
Total Income growth"
EcGro
0.0709
0.015
bnSA population change"
SMSCh
0.0175
0.014
Central city population change^
CC Ch
0.0070
0.019
Congressional power^
CONG P
0. 2A
0. 45
Type of city government^
C GOV
2.60
1.30
Manufacturing ratio^
MANUF
0.3A
0.08
High school graduates^
HSG
0.53
0.05
Age of cityg
Age C
2.03
0.31
Ginl coefficient^
Gini
0.341
0.028
Federal employees^
FEDEM .
0.07
0.03
South (dummy)
SOUTH
0.32
0.47
Birth rate minus death rate, 1968
Nat Inc
0.0089
0.003
Unemployment, 1970J
Unerap
0.13
0.018
Climate^
Clim
-6.8
9.5
Old in central city, 1960 (decile)
CC Old
4.9
2.5
Density central city (log)
DENSC
3.73
0.26
Black in central city, 1960^
Black
0.24
0.12
Data base minus Honolulu, which does not seem to fit into the
same pattern. All variables in percents are transformed by variance
preserving transformation in the calculations,
b
Annual rate of growth, 1960-1970.
c
This variable is 1 if a key Congressman or Senator has city as a
base.
^Weak mayor « 0, stronger « 1.2, comm. « 3, manager ■ 4.
Percent of nonagricultural employed.
^Percent of population.
^Population of city in (1890 + 1910 1930)/1960.
A large value for inequality.
^Percent of population, 1969.
■^Percent of work force.
K
Ten people were asked to rank ten major cities on a scale of 1-5
for desirability of climate. The average ranking was regressed
against summer and winter mean noon temperature and annual rainfall.
13
Congressional Power
Manufacturir>g Ratio
% High School Graduates
% Federal Employees
Uncmplo/ment
South
Gini (Inequality)
Cllrnate
Birth - Deaths
1968
Total
Income
Growth
1960-1970
SMSA
Population
Growth
1960-1970
Age of City
City Government
Central City
Population
Growth
1960-1970
Black, Spanish in City 1960
Old People in City 1960
Fig. 1 — Model of metropolitan growth
14
Table 3
A MODEL OF GROWTH AND DECLINE^
124 Urban Areas^
EcGro « 0.95 SMSCh + 0.0035 CONG P + 0.0015 C GOV + 0.025 MANUF - 0.023 HSG + 0.0072 Age C
(2.9) (3.5) (1.9) (l.A) (2.2)
0.082 Gini + 0.0055 SOUTH + 0.017 FEDEM - 0.039 Unemp + 0.13.
(2-3) (2.8) (0.7) (1.1) (0.6)
Standard error « 0.00565.
SMSCh = 0.93 EcGro - 0.042 Gini - 0.007 SOUTH + 0.37 Nat Inc + 0.0002 Clim - 0.0342.
(11) (+1.2) (3.4) (1.9) (2.25) (2.2)
Standard error = 0.0063.
CC Ch = 0.60 SMSCh - 0.018 Age C - 0.026 Black + 0.001 C GOV - 0.0012 CC Old
(^•^) (2.7) (+2.3) (1.1) (2.3)
+ 0.0084 DENSC + 0.0114. '
(1.7) (0.5)
Standard error = 0.0114.
SOURCE: Keeler and Rogers. A Classification of Large American Urban Areas,
NOTE: See Table 2 for explanation of the variables, their abbreviations, means, and
standard deviations.
Estimated by two-stage least squares.
All but Honolulu.
c * • . • . • • •
Values in parentheses are t-ratios.
DEMOGRAPHIC ANALYSIS^
The population of metropolitan St. Louis, like that of other metropolitan areas,
changes through natural increase (the difference between births and deaths) and
through migration. A continual process of redistribution is under way in the area
as people move into and out of particular neighborhoods. During the 1960s, dis-
similarities between population change in St. Louis and its metropolitan ring
became intensified.
Comparative Trends in the City and Metropolitan Ring
St. Louis is a city of 600,000 in a metropolitan area of about 2.4 million people.
During the 1960s, St. Louis's population declined 17 percent while its suburban ring
* Taken from Peter A. Morrison, San Jose and St. Louis in the 1960s: A Case Study of Changing Urban
Popzi/a/tonjy, The Rand Corporation, R.1313.NSF (forthcoming).
15
population increased 29 p^ercent. The central city decline was acute, compared with
that of most cities. Examination of the demographic change components (Table 4)
reveals why.
The white population declined mostly because of massive outward migration,
chiefly to the suburbs. Between 1960 and 1970, 34 percent of the white city-dwellers
moved away. But whites also declined because their death rate steadily approached
their birth rate, and since 1965 has exceeded it. Those who remained in the city
added only 2 percent to their numbers (nationally, the increase in the white met-
ropolitan population was 11 percent).
It was a different picture for blacks. There was no gain or loss through net
migration during the 1960s, but the black population rose 19 percent through natu-
ral increase, very close to its national rate of 21.6 percent. Annual population
Table 4
COMPONENTS OF POPULATION CHANGE IN ST. LOUIS, 1960-1970
(Rates per hundred 1960 residents)
Area
Total
Change
Natural
Increase
Net
Migration
Both Races
St. Louis SMSA
St. Louis City
Remainder of SMSA
(suburban ring)
12.3
-17.0
28.5
11.5
7.3
13.8
0.8
-2A.4
U.7
Whites
St. Louis SMSA
St. Louis City
Remainder of SMSA
(suburban ring)
9.4
-31.6
26.6
10.1
2. A
13.3
-0.7
-34.0
13.3
Nonwhites
b
St. Louis SMSA
St. Louis City
Remainder of SMSA
(suburban ring)
28.2
18.6
53.8
20.2
19.5
22.0
9.7 «
-0.4
37.2
SOURCE: U.S. Bureau of the Census, Census of
Population and Housing: 1970; General Demographic
Trerids for Metropolitan Areas, 1960 to 1970, Final
Report PHC(2)-U Tables 10-12; PHC(2)'27, Table 3;
PHC(2)-n, Table 3, Government Printing Office,
Washington, D.C. , 1971.
^Rate of increase attributed to excess of births
over deaths.
^In this section of the table, "Total Change"
applies only to the black population. "Natural In-
crease" and "Net Migration" apply to the nonwhite
population as a whole, but in the St. Louis SMSA,
virtually all nonwhites are black.
16
estimates, however, show the black population in St. Louis to have peaked in 1968
at around 269,000. By 1972, it was estimated to have dropped below 250,000. In view
of the black population's positive natural increase, the only explanation is that
blacks have been migrating out of the city since at least 1968 (and almost certainly
before).
The redistribution of St. Louis's population during the 1960s and early 1970s
was marked by a sharp withdrawal of residents from areas adjacent to the original
central business district, and racial turnover in an area north of the city's center.
To examine these changes in greater detail. Rand developed a model for estimating
annually the population of city health districts.* (This model will enable city depart-
ments to continue to monitor changes in the population's size and racial composition
at the health-district scale throughout the 1970s.)*
As can be seen from Fig. 2, substantial numerical increases in black population
were registered in five health districts north and west of the city's center. In fact,
four-fifths of the total citywide increase in black population occurred in this area.
Since 1970, however, there has been no appreciable increase of blacks anywhere in
St. Louis except in health districts 7 and 9.
In districts where blacks increased substantially during the 1960s, the white
population registered sharp declines (Fig. 3)— in some cases falling to less than
one-fifth of its 1960 numbers.^ Since 1970, the white population in health district 23
has stopped declining. In the eight other darkly shaded districts in Fig. 3, however,
the white population has declined 15 percent annually from 1970 through 1972.
Population in districts on the city's south side declined moderately or slightly during
the 1960s and remains almost totally white today.
TTiese hardly random changes in racial location in St. Louis continue to reflect
the city's long history of residential segregation. (Until 1962, city newspapers car-
ried separate advertisements for real estate open to blacks and whites.)
Trends in the Suburbs
Demographic trends were somewhat more uniform outside the city (Table 4).
Natural increase and net migration contributed equally to the white population's
26.6 percent increase during the 1960s. The black population's 53.8 percent subur-
ban growth was attributable more to net migration than to natural increase.' St.
Louis's suburbs attracted migrants largely from the city but also from outside the
metropolitan area. Increasingly, migrants of both races entering the St. Louis SMSA
bypassed the city and settled in the suburbs (mainly in St. Louis County). It can be
seen in Fig. 4 that the total stream of new arrivals to St. Louis City between 1965
and 1970 was smaller (both absolutely and relatively) than it had been a decade
earlier. For blacks, the inbound stream was numerically about the same; but in
relative terms, newly arriving blacks increasingly favored the suburbs.
• The city is divided into 26 health districts, which range in population from about 10,(X)0 to 50,000.
uli^,'^^^ ^' ^"^^'l-'^^ Population Estimates for the City of St. Louis, 1960-1972, with a
Model for Updating Them, The Rand Corporation. R-1373-NSF (forthcoming).
' Whereas Fig. 2 shows numerical increases, Fig. 3 shows percentage increa.se8. The two figures are
necessanly mcompatible: in Fig. 2, there are many instances where the numerical base is very small and
growing, whereas m Fig. 3 the base is typically large and shrinking.
* Suburban blacks register a high overall rate of growth between 1960 and 1970 because their 1960
population base was miniscule (81,000).
NOTE: Cifywide, the block population Increased 39,814
befwcen I960 and 1970.
Substantial increase ( > 8,000)
Moderate increose ( 2,000-6,000)
I I Slight Increase (< 2, 000)
r% j;°o| Decline .
< 1,000 blacks either year
Major park or cemetery
Fig. 2— Districts gaining and losing black population, St Louis City
Health Districts, 1960-1970
oefween I960 and 1970.
^^ir^l Major park or cemetety
Fig. 3— Decline of white population, St. Louis City Health Districts,
1960-1970
°Dafa shown for 1955-1960 refer to nonwhites.
Fig. 4— Destination of migrants entering the St. Louis SMSA, 1955-1960
and 1965-1970 (persons 5 years old and over, residing outside SMSA
or abroad five years previously)
Changing Replacement Capacity
The importance of these sharply divergent growth dynamics reaches beyond the
mere decline of the city's population to the cumulatively weakening effects of pro-
longed and severe out-migration. These effects are evident in the white population:
heavy and prolonged out-migration has drawn away potential parents and left
behind an elderly population that no longer regenerates itself
The severity of out-migration by young adults can be gauged by following in-
dividual age cohorts from 1960 to 1970 (Fig. 5). For example, if there were no net
migration, the number of persons 5 to 14 years old in 1960 would appear as persons
15 to 24 years old in 1970, less a small allowance for mortality during the decade.
Below age 45, this allowance is minimal (at most 5 percent), so any sizable discrepan-
cy between a young adult cohort in 1960 and 1970 indicates the extent of migration
that has taken place. Figure 5 gives stark evidence of extensive out-migration in the
early adult years. For example, in 1960 there were 37,900 white females aged 15-24,
but by 1970, only 17,900 aged 25-34 remained— a 53 percent reduction. There were
31,100 males aged 25-34 in 1960, but only 15,900 aged 35-44 in 1970— a 49 percent
20
Age
1960
I I 1970
75 +
^^^^^^^^^^
65-74
35 40
Fig. 5 — Age distribution of white population, St. Louis City,
1960 and 1970
reduction. Overall, 46 percent of whites aged 15-34 in 1960 were gone by 1970,
leaving St. Louis with a sharply diminished pool of prospective parents.
This diminished replacement capacity is illustrated more directly in Table 5,
which shows:
• Women in the middle and later childbearing years have grown more scarce.
In 1960 white women 25 to 44 years old made up 22.1 percent of all white women
in the city; by 1970 the figure had dropped to 17.6 percent. (Part of this drop stemmed
from the changing national age distribution.)®
• The proportion of elderly whites has risen. Whites 65 and over made up 14.5
percent of the population in 1960, but 19.2 percent in 1970. (The corresponding
figure nationally was 10 percent in both years.)
• Partially as a result of these changes in age structure, the crude birth rate
per thousand whites declined from 22.1 in 1960 to 12.0 in 1972; and the crude death
rate per thousand whites rose from 14.8 to 18.0. (Part of the decline in the birth rate,
of course, was a consequence of the national trend in the birth rate, which dropped
nearly 25 percent during the 1960s.)
• For white women nationally, this age group declined from 26.4 percent to 23.5 percent of the total
population between 1960 and 1970.
21
Table 5
BLACK AND WHITE REPLACEMENT CAPACITY IN ST. LOUIS,
1960-1972
i. IIU a L OL
1 Q<ir»
lybu
1970
1972
Women in later childbearing
years (25-44)
White
Black
1
22. 1%
27.1%
17.6%
22.7%
NA^
NA
Population 65 and over
White
Black
14.5%
6.8%
19.2%
8.3%
NA
NA
Crude birth rate
White
Black
22.1
34.4
14.5
25.1
12.0
24.9
Crude death rate
White
Black
14.8
11.4
17.7
11.3
18.0
11.2
Since 1965, the white population has ceased to replace itself, its death rate
having exceeded its birth rate. By 1972, deaths exceeded births by a margin of 3 to
2. Since it is now undergoing natural decrease, St. Louis's white population will
continue to shrink whether or not net out-migration continues. Only a dramatic rise
m fertility or a massive influx of young adults can alter this situation.
The city's black population has not undergone severe migratory change and
retains its strong replacement capacity: in 1972, its crude birth rate was 24.9 per
thousand, but its crude death rate was only 11.2. Nevertheless, the black population
began to decline in 1969, indicating a net migratory loss severe enough to offset its
natural increase." This recent shift could signify an increase in departing migrants,
a reduction in entering migrants, or a combination of both. What weak indications
we have favor the first of these explanations."
General statements about an entire city invariably mask specific neighborhood
exceptions. This is true of St. Louis, where certain areas are registering growth by
attracting new residents. Two important questions are: Where are these new resi-
dents coming from— outside the city or other parts of the city? At what rate is the
change occurring? If the change is internal, or if new residents are coming from
outside, but at a slow rate, then the significance is small.
althou^!t"!! hi IT" ^""^ difficult to forecast, a dramatic rise cannot be entirely ruled out.
plSon-s'^e^^^^^^^^ - -^^'^'y - ^PP-^'^e bearing on the
Uf^dafi^^ThemSet"" ^«"'»'"^« City of St. Louis. 1970-1972. with a Model for
lOfiS onH 1o7r'" u'^- ^ i"'*'^^'* ^^^^ "''•''ack migrants entering St. Louis between
in rrr.; 1J70 was about the same as between 1955 and 196(>-around 10,000. Thus, only an increase
in gross out-niigration could account for the change in net migration.
22
latit^rL S '"^ ^'■""^^^ '"^^"'"^ considerable
lat tude for judgment. One local view we encountered repeatedly was that young
^H^ Z T '•^•"""^"^^ -'^h suburban living and attracted to inexpensively
.unn ^l'^;^ h°"«'ng..are beginning to move back to St. Louis. Evidence offered in
suppor of this view is anecdotal: a south-side realtor claims to have more buyers
than sellers; a particular parish reports that whereas few new families moved in last
SLlTd'hu^H '^i'.r'' ''^^^""'^ ^^P^^P'^ ^"^"^ « neighborhood
festival, and hundreds of would-be newcomers inquire about buying a home there.
1 V „nr? ^^yP°th«s's that a return to the city has commenced cannot be tested direct-
ly until the next census is taken. Nevertheless, some signs of this alleged reversal
should appear in our specially prepared population estimates for health districts on
dltr cril nt J'"' ^^"'^ ^^'^^ P«P"'^^'- of health
distncte 1, 3, and 14 has increased since 1970. The increases are small (3, 7, and 5
Tre^rXn n ^*TV''' ''''' '^^^^^ '''' --'^^^^ health district
disS 3 tnd \T"^" ' J '^^^ ""'^ ^^^^ 1971 it declined. Health
incr^rl T'T^''^ l^^l- W« '=«""<'t ^'^'^^ whether these
^Zrhe c7ty ' " - ^ P-P^« -1-ting
Hp '^''"'''^ "^^^ speculate about whether and to what
J^rtualW r vrT'.'' ^" ''^^ P'^P^^^^O" °" the south side
liZnn\ ^hite^"d advanced in age. In coming years, more and more houses
TrZ T ""r ^ population dies off". Some of this housing may well
d^nf /T- T"^' ''''' '''^''^^ ^o-th in some neighborLods
despite declining population citywide.
oc^mrdt';^''%S;T ^^^^^^^^^^Phic evidence we have, we can assert that the
oi^mZ/decline of St. Louis's population will continue and may well accelerate We
acJuirSlt. Z 'u" ^"''^^""^l cumulative loss of city-dwellers' has
acquired it^ own dynamic: the elderly now die off- faster than the young are born.
fK. f-T' ^!"hstantial proportion of whites are either entering or already within
thP P'-o.^^^tive parents are becoming scarce among St. Louis's whites, and
Lounf ?htwt;r" ? f ^'^"'^ ^^^^ -ntinu "to
mount The white population's crude birth rate is therefore likely to fall, barring a
" '"'"^'^ " ^ ^^--^"^ ^"^^--^ -fl- of child-bearing
tiallv °True ^h^ '"^'T? "^"^^'^ ''^^^'^ P°P^'^«°" ^^w substan-
hnl l ^'"^'.^^^ hlack population is expanding steadily through natural increase
but black migration out of the city is more than enough to cancel that increlT'
Accumulation of Disadvantaged Citizens
T^n!tl'"'f f °" '^^"^^"^ metropolitan-Wide distribution of population, St.
. V ^^"'^ V^'' '""'P''^^- P«P"'^t'0" has come to be comprised of
those citizens who are disadvantaged, as the following comparisons show
23
• Between 1960 and 1970, the city*s black population rose from 29 percent to
41 percent, but only from 6 percent to 7 percent in the rest of the metropolitan area.
• The city's residents 65 years and older increased from 12 percent of the
population to 15 percent; they stayed at 8 percent in the remainder of the metropoli-
tan area.
• For families and unrelated individuals, median income in the city was 79
percent of that for the SMSA in 1959; city income was only 68 percent of SMSA
income by 1969.
• The proportion of relatively high-income families declined sharply. In 1959,
1 1 percent of families in the city had incomes at least double the city's median family
income; by 1969 only 4 percent had such incomes.
• The proportion of relatively low-income families rose slightly. In 1959, 16
percent of families in the city had incomes below half the city's median family
income; by 1969, 21 percent had such incomes.
Through selective out-migration, problems of dependency and poverty— not ex-
clusively problems of St. Louis— have come increasingly to be located mSt. Louis.
Migration
In this context, it is important to clarify how migration contributes to or allevi-
ates the problems facing St. Louis and its residents. Like other metropolitan areas,
St. Louis is linked with urban and rural areas throughout the country by migratory
interchange. Among white migrants, this is a broadly connected system, indicative
of metropolitan St. Louis's niche in a national system of manpower exchange. The
migration of blacks, however, is more of an urbanizing process: incoming migrants
enter metropolitan St. Louis mostly from rural origins in such states as Mississippi,
Missouri, and Arkansas. Outgoing migrants go to metropolitan destinations, often
large centers such as Los Angeles, Kansas City, and Chicago. For many blacks, St.
Louis serves as an entry point into urban life.
The city, however, is where most black in-migrants to the metropolitan area
settle. An important question here is: How do these incoming migrants fare com-
pared to the St. Louis residents they join? Although we lack the requisite data for
exploring this point thoroughly, it is possible to examine the unemployment experi-
ence of recent in-migrants after their arrival in St. Louis and compare it with St.
Louis residents they join. »^ Data in Table 6 show that among blacks, recent migrants
differ little from long-term residents with respect to unemployment at any age.
Among whites, recent migrants also have unemployment rates similar to those of
long-term residents (except for the 20-24 year age group). There is, however, a sizable
difference between blacks and whites in every category: blacks are substantially
more unemployed than whites.
The effects of migration, then, have to be judged cautiously. In trying to analyze
these effects, a major difficulty is that standard social and economic statistics are
compiled and organized mostly by areas rather than by groups of people. Conse-
Specifically, we compared the 1969 unemployment experience of recent migrants (defined as per-
sons entering St. Louis between 1965 and 1970) with that of long-term residents (natives and earlier
migrants). The source of these data was the 1970 Census Public Use Sample.
24
Table 6
UNEMPLOYMENT RATES IN ST. LOUIS FOR RECENT MIGRANTS
AND LONG-TERM RESIDENTS, 1969
Age and Migration Status, 1970^
Percent
Unemployed, 1969
Whites
Blacks
20-24 years old
Recent migrants
6
11
Long-term residents
3
10
25 years old and over
Recent migrants
1.6
3.5
Long-term residents
1.5
3.6
SOURCE: U.S. Bureau of the Census, 1970 Census
of Population^ Public Use Sample of Basic Records.
All persons 20 years old and over in the labor
force. Recent migrants are persons who moved to
St. Louis between 1965 and 1970. Long-term resi-
dents are persons who lived in St. Louis in both
1965 and 1970.
quently, we can observe the experience of places, but not of people. These experi-
ences can differ sharply. For instance, black in-migrants from impoverished rural
areas in states like Mississippi may be less affluent or employable than the mostly
white population they join in St. Louis. If this is true in St. Louis (as it is in other
cities), then area indicators (e.g., unemployment or poverty in St. Louis) may register
a worsening of local conditions. But measures of individuals' experiences (e.g., their
unemployment experience or poverty now, compared with what it was before they
came to St. Louis) may show marked improvement.
ECONOMIC CHANGES
Slow Economic Growth in Metropolitan St. Louis
We have seen that economic and population growth are slower in metropolitan
St. Louis than in the nation as a whole. To what extent does this slow growth account
for the central city's decline? There is certainly reason to believe that faster growth
in the metropolitan area would be useful to the central city, but the Keeler and
Rogers analysis discussed earlier indicated that slow metropolitan growth is not
the variable most strongly associated with central city decline in St. Louis.
Further, the policy implications of stimulating metropolitan growth to gain
positive effects for a central city are curious. Built into an "average" metropolitan
growth rate for the nation is the fiact that some areas fall below, some above that
A Classification of Large American Urban Areas.
25
average. The most obvious way to change relative rankings is for metropolitan areas
to compete v^ith one another for jobs and people. While proponents of local growth
are accustomed to such competition/^ there is no compelling rationale for public
policy at a higher level to artificially equalize the economic performance of met-
ropolitan areas by redirecting people and jobs toward certain areas and away from
others.
Even if a local metropolitan area competes and wins, what are the relative
benefits of so doing? In California, San Jose's rapid economic and population expan-
sion produced its own set of problems:^® affluence increased, but its distribution did
not become notably more equitable (Chicanos did not appear to benefit as much as
Anglos); certain environmental amenities deteriorated as tract housing develop-
ments and freeways destroyed orchards and serene vistas; and while San Jose's
residents seem less concerned today with what urban planners regard as the aesthet-
ic outrages of rapid growth, local policymakers were sufficiently skeptical of the
benefits of rapid growth to ask Rand whether continuation of such growth was
essential to economic well-being. Similar problems beset some of the suburbs of
metropolitan St. Louis: traffic congestion, sudden new demands on municipal ser-
vices, unplanned and inefficient land use are much more characteristic in the grow-
ing suburbs than in the central city.
Economic Growth: City Versus Suburbs
Economic decentralization has paralleled the movement of population in met-
ropolitan St. Louis. Between 1960 and 1970, the city's share of SMSA population
shrank from 39 to 26 percent. Its share of area jobs declined from 61 to 42 percent.
The figures in Table 7*^ illustrate how sharply the city and suburban economies
were diverging during the latter half of the 1960s. In St. Louis, earnings grew only
in the government and service sectors; all other sectors declined. In the suburban
ring, all sectors registered positive growth. Table 8 shows that within the service
sector, city employment grew slower than suburban ring employment; indeed St.
Louis lost employment in ''hotels" and "personal" services and showed only a minis-
cule increase in "legal" services and "amusement." These growth rates, combined
with declining earnings in industrial and commercial sectors, strongly suggest that
the center of economic activity is shifting away from the central city. We estimate
that if these trends continue until 1975, St. Louis County will have captured a share
of business activity approaching that usually associated with the central city of a
metropolitan area (see Table 9).*®
Survey of Industrial Developers
To supplement the projections based on past aggregate data. Professors D. K.
Holland and G. D. Wendel of St. Louis University carried out a survey of eight
X'iljorous effbrts to recruit business at the regional level are planned by the St. Louis Regional
Commerce and Growth Ass^x:iation formed in 1971.
These rosuhi; are reported fully in Levine and Alesch, Growth in San Jose.
" Taken f rom C. Gardner and G. Payne, An Economic Analysis of Central City Decline, The Rand
Corporation, R-135a-NSF (forthcoming).
Ibid.
26
Table 7
ANNUAL GROWTH RATES OF EARNINGS BY SECTOR AND INDUSTRY,
ST. LOUIS CITY AND SUBURBS, 1966-1970
(In percent)
St. Louis
Sector
City
Suburbs
Government
5.3
7.1
Private
-3.1
6.1
Manufacturing
-A. A
3.5
Transportation, communications,
Utilities
-1.9
9.2
Trade
-5.1
11.7
Finance, insurance, real estate
-2.0
7.1
Services
3.0
7.1
Total
-2.3
6.3
SOURCE: Data provided by Regional Economics Infor-
mation System, Bureau of Economic Analysis, Office of
Business Economics, U.S. Department of Commerce.
Table 8
EMPLOYMENT IN SERVICES, ST. LOUIS CITY AND SUBURBS,
1959 AND 1970
St.
Louis City
Suburbs
Number
of Jobs
Held
Number
of Jobs
Held
Service
1959
1970
AGR^
1959
1970
AGR
Hotels
Personal
Auto
Repair
Amusement
Medical
Legal
Education
Nonprofit
Miscellaneous
5,672
7,86A
2,190
1,184
2,123
13,299
1,072
3,177
8,396
4,687
7,025
2,609
1,690
2,177
20,317
1,173
10,244
10,486
-1.91
-1.13
1.75
3.56
0.25
4.24
0.90
11.71
2.22
962
4.039
1,185
705
2,056
6,030
352
1,134
4,029
3,351
7,043
2,626
1,077
3,611
21,712
946
5,905
8,442
12.48
5.56
7.96
4.24
5.98
12.81
9.89
16.50
7.40
business
7.144
14,744
7.27
1,114
6,211
17.18
SOURCE: County Business Patterns, U.S. Department of
Commerce, 1959 and 1970.
Average growth rate.
27
Table 9
PERCENTAGE OF SMSA EARNINGS BY SECTOR, ST. LOUIS CITY,
ST. LOUIS COUNTY, REMAINING SUBURBAN RING,
1970 AND 1975
Remaining
St. Louis
St.
Louis
Suburban
City
County
Ring
Sector
1970
1975
1970
1975
1970
1975
Total nonfarm
43
33
35
47
22
20
Government
38
36
30
32
32
32
Federal government
50
46
15
25
35
29
Private
44
33
36
50
20
17
Manufacturing
42
33
36
47
22
20
Transportation, communications.
utilities
55
41
23
41
22
18
Wholesale and retail
46
27
37
69
17
4
Finance, insurance, real estate
51
39-
35
47
14
14
Services
43
33
35
47
22
20
SOURCE: Gardner and Payne, An Economic Analysis of Central City
Decline.
industrial developers in the St. Louis area.*® On the whole, industrial developers
substantiated the conclusions of other analyses: that industrial dispersion in met-
ropolitan St. Louis had been stimulated by the search for more space and by the
construction of interstate highways. At one stroke, these highways made available
large, relatively inexpensive tracts of suburban land and lowered transportation
costs. The developers also indicated that high crime rates in the city gave added
impetus to dispersion.
Interviews with these developers pointed up two objective limitations facing
any concerted effort to reverse industrial suburbanization:
• More land is available in the suburbs than in the city. At present, about 1300
acres (not all zoned industrial) are available for development in St. Louis, whereas
4200 acres already zoned for industrial use are available for development in St.
Louis County. Furthermore, despite widespread abandonment in St. Louis, develop-
ment remains more expensive there than in the suburbs.^° Moreover, recent cut-
backs in urban renewal funds have virtually eliminated the land write-down feature
that formerly made the cost of city land development nearly competitive with subur-
ban land development.
• High crime rates have reduced the attractiveness of the land available for
redevelopment in St. Louis. Yet crime rates depend heavily on the income level and
*• These firms manage 50 percent of the area's industrial parks. A full description of the interview
and analysis method used in the study, the sample of developers interviewed, and the full report of
interview results are given in Holland and Wendel, Development of Industrial Parks.
*° That is, in general, it remains expensive to buy abandoned property, clear it, and develop it or sell
it to developers. Even with subsidized demolition (most of the clearance that has taken place is federally
subsidized), the potential redeveloper must iixke into account the negative effects of surrounding, deteri-
orating neighborhoods.
k
28
age structure of the local population, factors that municipal governments can do
little to change.*'
From the survey of developers, we know that the most attractive land in the city
for industrial development is distant from low-income residential neighborhoods
and accessible to highways. From the developer's viewpoint, then, accumulating
spatial inventory in the city is a necessary but not sufficient condition for future
business investment.
RACIAL FACTORS
Several data sources we have examined show a strong relationship between the
presence of blacks and a rapid exodus of whites at the neighborhood levelboth within
St. Louis and in surrounding suburbs. However, the hypothesis that city-wide popu-
lation decline is largely a matter of "white flight" (i.e., racially motivated departure)
was not substantiated. This finding fits with the analysis by Keeler and Rogers,"
which shows a weaker relationship between race and central city decline than
between the city's age and its decline.
Race and Central City Decline: The White FUght Hypothesis
Precipitous neighborhood change may be explained in part by racial transition,
but the aggregate population decline of the city appears to be a response to other
factors. In one attempt to account for the pattern of residential and industrial
dispersion evident in the St. Louis metropolitan area, we tested three hypotheses
that seemed to be plausible explanations of trends that have left St. Louis City with
a population composed increasingly of blacks."
• Industries in which whites are overrepresented have been suburbanizing
more than other industries. A high proportion of whites have been choosing to live
close to their jobs.
. Other things being equal, nearly everyone prefers suburban residence to city
residence. If whites' incomes have been rising more than blacks', a higher proportion
of whites will have moved to the suburbs.
• The white population has been leaving the city to escape the black popula-
tion (the "white flight" hypothesis).
We examined black and white departure rates from the central city adjusted for
interracial differences in income and job location within the metropolitan area."
^JuJl^^l^l^iT^A^^^^J"^'' ^^rK^^- ^"'^ ^ ^ ^■^^ ^^f" °«j<"- "iine reduction program
sponsored by the U.S. Department of Justice.
" A Classification of Large American Urban Areas.
nof'l^f*" n of the hypotheses might account for all of the city's demographic change, they are
not^mutually exclusive. Indeed, it is unlikely that a single mechanism has been operating
i« k-t" '"j*" constructed that allows us to calculate an expected value for the 1960-1970 change
rJrfn?, A city residenu. allowing for city job losses (gains)-holding income constant. See
Oardner and Payne. An Economic Analysis of Central City Decline, Appendix A.
29
There were two findings. First, within every income bracket for both blacks and
whites, fewer people live in St. Louis City than would be expected, given the spatial
distribution of jobs. Second, at only the lowest and highest income levels did whites
leave the city at faster rates than their black cohorts. That is, for most of the income
distribution, blacks and whites were leaving the city at the same rate during the
1960s.
m
Reasons for Residential Change
A survey conducted by the City Plan Commission in 1967" offers clues about
why these moves took place. Interviews suggest that people's desire to own a home
or enlarge living space is as important as their desire to escape repellent neighbor-
hood conditions in motivating movement within and away from the city. Of respond-
ents who intended to move, both blacks and whites who gave priority to becoming
homeowners tended to favor St. Louis County as their destination. For prospective
homeowners, then, St. Louis's housing stock is less competitive than housing in the
suburbs.^® On the other hand, respondents who intended to move and who gave
priority to enlarged living space tended to designate locations within the city.
Thus, St. Louis's housing stock is competitive mostly as it offers space for rent.
Blacks and whites who expected to move, however, were found to designate mutually
separate areas of the city. While we hesitate to accept these expressions of intent
without question, we note that they are borne out by the actual patterns of white
and black movement, to which we now turn our attention.
Analysis of Census Tract Population Changes
During the 1960s, rates of population change varied widely among different
sections of St. Louis. Census tracts— small and relatively homogeneous areas into
which cities are subdivided— are a useful scale at which to examine these variations.
Regression analyses of population changes in St. Louis census tracts from 1960 to
1970 reveal consistent patterns beneath this variability and offer clear indications
of the contrary racial trends just noted.
Not surprisingly, the population changes at this small-area scale fit well with
the somewhat larger health district scale analysis reported above, showing continu-
ing differentiation of the black and white populations within the city. Blacks exhibit
a strong tendency to move into tracts where blacks already reside, rather than
disperse evenly throughout the city. (As noted earlier in this section, under the
heading "Comparative Trends in the City and Metropolitan Ring," four-fifths of the
black population's citywide increase was concentrated in 5 of the city's 26 health
districts.) Predominantly white tracts tend to retain white residents if most white
households own their homes. White population declines more severely where most
whites rent, or where blacks— particularly new arrivals— make up a significant
" A reanalysis of selected parts of that survey is reported in Morrison, San Jose and St. Louis in the
1960s.
" A more detailed analysis of the reasons for residential choices will be available when C. Leven and
J. Little publish their model of residential preference. See Sec. II, footnote 15. above.
" A full discussion of the data and regression coefficients on which the analysis in this subsection is
based is reported in Morrison, San Jose and St. Louis in the 1960^. , ,
90
fraction of residents. Although our regression analysis cannot shed light on causa-
tion or underlying motives, it documents the powerful continuities in racial separa-
tion in St. Louis. These tendencies reflect a long history of overt racial segregation
in housing. Today, however, the behavioral mechanisms at work may also involve
income differences that affect the filtering of housing. To investigate this possibility,
we examined the dynamics of the housing market.
Neighborhood Level Analysis of the Housing Market"
An arbitrage model of household locational decisions can often be used to pre-
dict and explain the response of the housing market to changes of race and income
in neighborhood household composition. This model assumes that the housing mar-
ket is segregated by race and income, reflecting people's preference to cluster in
homogeneous groups. High-income families who can afford new housing receive
discounts if they live in neighborhoods adjacent to low-income families. Low-income
families pay premiums to live in neighborhoods adjacent to high-income families
and capture their amenities. Equilibrium in the market occurs when the price of
housing along the boundary between these two groups is the same for each.
Under these conditions, increased housing demand by poor families can provide
an incentive for houses to change from use by high-income families to use by low-
income families; as the price of housing goes up for low-income families, the contrac-
tion of housing supply for high-income families increases the cost of their housing
and provides an incentive to new construction. Racial prejudice would affect the
housing market similarly, although two boundaries should develop: one between
high-income and middle-income black families, and one between black families and
low-income families. In this case, as black families demand more housing and drive
the price up, there is an incentive to shift some housing from use by high-income
whites to use by middle-income blacks.
Initial evidence based on rental market data for selected city neighborhoods
shows the following pattern: vacancies rise as black and low-income family occupan-
cy approaches a particular block. Near the peak of the vacancies, rents begin to
declme. Then, once the boundary is passed, vacancies decline and rents rise briefly
to their previous level. Finally, rents decline continuously until they reach a floor,
at which time the units are removed from the market.
This same pattern has since been found to prevail in suburban neighborhoods.
Additional data on the housing market (i.e., data on the owner-occupied as well as
the rental housing market) suggest that middle-income blacks— themselves moving
from lower-income black areas— have been the leading edge of suburban neighbor-
hood transition even where their incomes are somewhat higher than those of their
" ™f analysis has been developed from several different sources. The "arbitrage model" of
househo d locational decisions was first used in Urban Decay in St. Louis (Institute for Urban and
KegionaJ btudies, Washington University) to describe and explain events irt the city's housing market
U)mprehensive housing market data to test the model are being developed by Hugh Nourse and James
UtUe under HUD grant MOPEM. In this latter study, the arbitrage model will be tested in nine neighbor-
hoods located in the city and m suburbs surrounding the city: University City, Wellston, Jennings.
Normandy, River View Gardens, Bayden. Walnut Park. The Hill, Lafayette^ulard. An application of
the model to additional data developed for University City in response to Rand's request is reported in
James Little, Housing Market Behavior and Mobility Patterns in a Transitional Neighborhood, Institute
for Urban and Regional Studies, Washington University, St. Louis, June 1973.
31
new white neighbors. Vacancy rates increase in such neighborhoods, and housing
prices deflate ovtT time to the point where less affluent black families can enter the
area. White neighborhoods in the path of such movement anticipate transition,
showing higher vacancy rates even before black entry — and the process continues.
These latter data support the hypothesis that precipitous neighborhood change
is stimulated by racial transition. Even so, it can be argued that white response at
the neighborhood level is less a direct flight from the first middle-income black
families than a response to expectations that lower-income blacks will follow.
Policy Implications
To repeat, racial aversions do not appear to be strongly associated with St. Louis
City's population decline; in fact, at most income levels, blacks and whites have been
leaving the city at about the same rate. On the other hand, there is some evidence
that racial transition causes precipitous neighborhood change u;i7/im jurisdictions.
But whether race, income, or some combination of differences causes people to flee
certain neighborhoods and certain jurisdictions, communities of high transition are
typically left with lowered income distributions. Local policymakers are then less
able to intervene in the transition process, to the extent that public services depend
on tax revenues collected from residents.
POLICY ACCELERATORS TO CENTRAL CTTY DECLINE
The decline of St. Louis remains most strongly associated with the demographic
and economic factors discussed above, factors that local policy can do little to control.
Nor can St. Louis easily control federal policies that contribute to decline. However,
certain federal and local policies not only contribute to decline; they accelerate it.
Some of these policies are discussed in this section.
Federal Highway Policy**
During the 1960s the federal government supported the construction of five
major interstate highways in metropolitan St. Louis (Fig. 6).^° Total capital expendi-
tures of the interstate system during this period exceeded $250 million dollars, of
which the state paid 10 percent.'*
" Calculations in this section are reported ir. J. Enns and P. deLeon, The Effect of Highways upon
Metropolitan Dispersion: St. Louis, The Rand Corporation, P-5061, September 1973.
Three of the links in this system (1-70, 1-44, and 1-55) are radial routes extending west from the
Central Business District; the other two links (1-270 and 1-244) make up the outer beltway that connects
the northern and southern portions of St. Louis County. These routes were completed on the following
dates: 1-70 (July 1961), 1-270 (June 1964), 1-55 (July 1967), 1-244 (November 1968), and 1-44 (December
1972). Two other major arteries also were developed during the 1960s. SUite Highway 40 (the Daniel
Boone Expressways was substantially improved, while St. Louis <:k)unty completed a portion of the inner
beltway that connects Highway 40 with 1-70.
The expenditure data were obUined from U.S. Department of Transportation, Federal Highway
Administration, Highway Statistics, Table SF-15, annual volumes 1965-1970. The data cover JelTerson,
St. Charles, and St. Louis Counties and St. Louis City.
32
Source: Holland and Wendel, Development of Industrial Parka
Fig. 6 — Location of major highways and industrial parks
in the St. Louis SMSA
33
Industrial park locations (the dots in Fig. 6) show that this form of industrial
development is sensitive to the location of the interstate beltway, a fact borne out
by the survey of industrial developers discussed above. To gain a rough numerical
picture of how beltways influence land use patterns, we examined residential and
employment density changes that occurred between 1965 and 1970 for three rings
of the SMSA,'^ designated by the letters A, B, and C in Fig. 6. Ring A is bounded
by city limits on the west and includes all traffic zones within the city. Ring B
contains all zones in St. Louis County within the outer beltway (1-270, 1-244). Ring
C includes the remaining zones west of the outer beltway and portions of St. Charies
and Jefferson Counties (represented by dashed lines in Fig. 6).
Land use densities in 1965 and 1970, shown in Table 10, reveal several interest-
ing trends. First, population density is roughly three times higher in St. Louis City
(Ring A) than in inner St. Louis County (Ring B); but in both areas population
density declined from 1965 to 1970. In outer St. Louis County and portions of
Jefferson and St. Charles Counties (Ring C), population density increased about a
third, but the absolute density of this ring is still far lower than that of the inner
two rings. Second, industrial employment density (persons employed per industrial
acre) declined in Ring A and increased in both Rings B and C, supporting other
observations about the direction of industrial expansion. By contrast, commercial
employment density increased slightly in Ring A while falling in Ring B; it rose
sharply in Ring C, although from a small initial base. These trends suggest that the
urbanization process is continuing well beyond the previous county suburban bound-
ary (represented by the inner beltway).
We also explored the influence of radial highway routes on changing land use
for industrial purposes in one portion of the metropolitan area. Our expectation was
that areas experiencing large decreases in travel time to St. Louis's Central Business
District would display the greatest relative increases in land use density." Our
results were mixed. For that portion of the SMS A lying north of State Highway 40,
density changes tended to be greatest where travel time changes were smallest-
opposite to our expectations. This finding may reflect important time lags in the
adjustment of urban activities to transportation change. We note that the major
northern radial route, 1-70, was completed in 1961; thus its impact on changing
travel time is not captured by our 1965-1970 data. The changing densities we do
capture may reflect continuing response to those earlier time changes.'*
In the southern portion of metropolitan St. Louis, the data confirmed our an-
ticipated effect of travel time change on density. Industrial employment density
changes were directly influenced by improved access to the CBD. The estimated
" This was the only time span for which appropriate data were available.
• "u^u ^^^^ °^ decreasing travel time to the CBD, a simple regression model was used
in which absolute changes m industrial employment density formed the dependent variable; the model's
explanatory variables consist of absolute change in travel time to the CBD plus some initial density
variables to allow for differing levels of land use between regions at the start of the time period. Traffic
zone data on population, employment, and CBD travel time for the Missouri portion of the SMSA were
obtained for two years (1965 and 1970) from the Mi.csouri State Highway Commission; the observations
used to estimate the regression equations were then formed by calculating the changes in density and
Ln^au'sis""* observations (traffic zones) available for the regression
" y*^** conclusion is partially supported with regard to industrial employment by an inspection of
the pattern of industrial park development that occurred during the 1960-1970 period. Of the 25 parks
developed during the decade in North St. Louis County, 13 were opened aRer 1965. Data on opening dates
ot industrial parks were obtained from Holland and Wendel, Development of Industrial Parks
34
Table 10
POPULATION AND EMPLOYMENT DENSITIES IN
THE ST. LOUIS SMSA, 1965-1970
Ring
Population
Density
Indus trial
Employment
Density
Cnmmp vr 1 1
Employment
Density
Ring A
1965
1970
33.5
31. A
12.7
12.0
10.5
10.8
Ring B
1965
1970
11.7
10.5
3.1
3.4
4.7
4.2
Ring C
1965
1970
2.3
3.2
0.75
1.0
1.3
2.4
SOURCE: Traffic zone data obtained from
Missouri State Highway Commission.
NOTE: Population density figures refer'
to persons per residential acre. Employ-
ment densities refer to persons employed per
industrial or commercial acre.
coefficient for travel time change (the major explanatory variable) was near unity,
indicating that for the typical zone, a one-minute decline in CBD travel time is
associated with an increase of one industrial job per industrial acre.^'^
No doubt, the size of this coefficient is inflated because our model excludes other
important explanatory variables, such as land value. Nevertheless, it is clear that
the response of industrial firms to declining transportation costs, as measured by
CBD travel-time changes, is significant.
To summarize, both radial highways and beltways have stimulated industrial
and population dispersion from St. Louis. If these patterns continue during the
remainder of the 1970s, the interstate beltway may be a catalyst for further west-
ward movement of population and employment. Indeed, our survey of industrial
developers indicates that this beltway has already sparked a substantial increase of
industrial activity in the outer western ring of metropolitan St. Louis.
Federal Real Estate Tax Incentives^'
Federal real estate tax incentives do not determine the jurisdictional locations
(e.g., city versus suburb) in which money is invested. But they accelerate dispersion
by ofiering advantages to types of investment that are simply more available in
" The jiata used represented actual employment between 1965 and 1970 (not jobs available). This
result IS thus affected to some unknown degree by the level of unemployment prevailinjr in 1965 and 1970.
" Material in this section is taken from R. Slitor, "Tax Effects on Urban Growth in Three Cities: San
Jose, St. Louis, and Seattle" (unpublished Rand document).
36
suburbs than in central cities. For example, the benefits to home ownership encour-
age middle- and upperrincome families to purchase new housing which appears
mostly in the suburbs. At the same time, laws that do not allow deductibility of
capital losses on owner-occupied homes, but that do tax capital gains, hasten disin-
vestment in central city housing that is comparatively older and more likely to
decline in value, and discourage improvements likely to be reflected in capital gains.
In combination, these laws encourage panic selling to avoid loss, and worsen the
instability of neighborhoods undergoing racial or income transition; they deter
capital improvements, thereby hastening the deterioration of housing stock; and
they encourage conversion of homes to rental property for a period before sale, thus
accelerating neighborhood change. For metropolitan St. Louis, Richard Slitor has
estimated that there is a $67 million annual tax break for home ownership, a $22
million capital gains incentive for real estate speculation, and $13 million in capital
gains unrealized at death. These figures represent different sorts of incentives for
private actions, not estimates of the market effects of such actions. Thus, they are
not additive. Nonetheless, they represent an impressive set of incentives for invest-
ment in suburban housing and disinvestment in central city housing.
Other Federal Policies
A number of other contributory federal policies warrant mention in an overall
assessment of St. Louis's decline. While we have not carried out specific analyses
here, we can oflTer the following observations based on research by others.
The housing policies of the 1960s had important eflTects in St. Louis. The Pruitt-
Igoe public housing development, built in the 1950s, abandoned during the 1960s,
and partially demolished in the 1970s, is well known— a classic example of how
federal high-rise, low-amenity, problem-concentrating public housing fails.
Less dramatic but more important than public housing has been the effect of
mortgage reinsurance by the FHA and VA, far and away the most powerful federal
policy affecting housing in the post-World War II era. To be sure, decentralization
of population has stemmed from an overwhelming popular desire for suburban
housing; but FHA and VA reinsurance had two important contributory effects. First,
it enabled people to buy suburban housing with no down payment and at low interest
rates. Second, it created a national mortgage market. Both of these eflfects accelerat-
ed the outward movement of families. Moreover, in the case of St. Louis, the FHA
has oflen refused insurance on inner-city mortgages, making it even less likely that
private owners would maintain the existing housing stock. More recently, the De-
partment of Housing and Urban Development has sharply curtailed urban renewal
funds, thereby removing the land write-down feature — the only remaining factor
that could make city land development nearly competitive with suburban land
development.
Housing is by no means the only realm in which the powerful side effects of
federal policy have hastened St. Louis's decline. Earlier we pointed out that one
reason for differential decline is the ample availability of cheap agricultural land
that can readily be developed for industrial or residential uses. Although partly a
result of St. Louis's natural geography, suburbanization may be substantially ac-
celerated by federal flood control policy under the Corps of Engineers, which contin-
ues to create more land.
36
A number of federal policies, then , have contributed to St. Louis's decline. Other
possible policies, such as revenue-sharing and income maintenance, might have
helped to slow this decline or soften its effects on people, but such policies have not
been in force.
Local Policies
Jurisdictional Boundaries. St. Louis is one of the five geographically small-
est cities in the United States with over half a million population. " At the peak of
Its twentieth-century population, more than 880,000 residents lived within its 61
square miles. Although there are fewer than 600,000 residents today, St. Louis
retains a high density of land use.'» Indeed, our survey of industrial developers
shows them in unanimous agreement that the desire for additional space has been
an overwhelming (though not exclusive) motivation for business departures from
the city.
Some local analysts contend that sharp decline registered in St. Louis's popula-
tion is, to a degree, a statistical artifact arising out of the political decisions that have
kept city boundaries fixed for so long. According to this view, the corporate entity
called "St. Louis" is more artificially defined than most other central cities, the
result being to render it an exaggerated example of the typical older U.S. city.^"'
To measure the eflfect of this size restriction more exactly, we calculated how
far St. Louis boundaries would have to extend for its decline in 1960-1970 population
to equal the a yero^e population decline for comparable U.S. cities.*" We found that
city boundaries would have to extend about six miles farther west, taking in two-
thirds of St. Louis County's population and about one-third of the county's land area
(see Fig. 7). Of course, it is possible to extend city boundaries far enough to create
a jurisdiction that shows no population or job loss between 1960 and 1970. In that
case, what we call the city would become a larger jurisdiction's concentration of
low-income population. It is not clear, however, that the jurisdiction we might create
this way would have developed in the same way had such annexation actually
occurred.
Though we cannot argue that restricted boundaries have accelerated St. Louis's
decline, they have distributed the consequences of decline to the city's disadvantage.
Reduced revenues, coupled with the necessity of maintaining an older physical stock
and the necessity of providing services to a growing proportion of aged and poor,
engender problems that are less and less capable of solution from within those
political boundaries.
Conservative Banking Community. People we spoke with often mentioned
the conservatism of the city's banking community as a barrier to new growth in St.
Louis. Using recently developed techniques of portfolio analysis, we tested this
K.Z.'^!' f ^'■*"<='^">' Pittsburgh, and Washington (we exclude the individual
boroughs of New York from this comparison).
" See Table 10.
»• Dempster Holland's unpublished paper illustrates, for several older cities, the strong relationship
between the proportion of the city urbanized before 1900 and the proportion of the city with the highest
ratw of abandonment. Of the cities compared (Cincinnati. Chicago. Pittsburgh, Columbus. Detroit. Cleve-
lana. M. U)ms\ bt. Louis had the largest portion of its land area urbanized before 1900.
" The calculation appears in Morrison. San Jose and St. Louis in the 1960s.
37
Area hypothetical ly annexed
Fig. 7— Hypothetically different St. Louis City boundaries
belief.** The approach used was suggested by the method of the Capital Asset Pricing
Model, developed by William Sharpe*^ and others.
Applying this model to the four largest banks in St. Louis City— The First
National Bank, Bank of St. Louis, Boatmen's National Bank, Mercantile Trust— we
found conventional wisdom to be true. These banks are conservative when compared
with the rest of the nation's big-city banks. And between 1940 and 1970, the banks
showed increasing conservativeness. But what is the import of this for urban
growth?
It can be argued theoretically that in cities where banks finance higher-risk,
higher-payoff investments, the wealth of the population will grow faster than in
cities where banks maintain lower-risk preferences. The regulation of entry into
banking reduces the probability that the full spectrum of risk preferences will
develop among banks, a phenomenon that can be expected to develop in an open
competitive market. Unit-rule banking, a very strict Ibrjn of entry regulation, may
tend over time to reinforce any disequilibrium of risk preferences (e.g., where all
A discussion of this test appears in Cyrus J. Gardner, Bankinfi Rcffulation and Urban Growth, The
Rand Corporation, P-5057, August 1973.
William Sharpe, "Capital Asset Prices: A Theory of Market Equilibrium under Conditions of Risk,"
Journal of Finance, Vol. 19, No. 3, September 1964, p. 425.
38
banks in a community are conservative or all have high-risk preferences) by retard-
ing the entry of new banks into a community/^
While we cannot estimate the quantitative effect of a conservative banking
community on economic growth in St. Louis City, we can argue with confidence that
it has not helped the city's economic viability.
« Unit-rule banking involves a state regulation that forbids the formation of branch banks. Both
Missouri and Illinois have unit-rule banking. For an empirical study showing that creation of new banks
isretarded "^o^e under unit-rule banking regulations than under regulations that allow branch banking,
aeeR Pakonen. The Differential Effect of Branch Uw Regulation on Commercial Bank Entry." Ph.D.
dissertation. Washington State University. Pullman, Washington. 1971
IV. ALTERNATIVE STRATEGIES FOR THE FUTURE
ALTERNATIVE FUTURES FOR ST. LOUIS CITY
Because St. Louis has already undergone major economic and population de-
cline, it is possible that the attendant accumulating inventory may initiate new
conditions in the city that will gradually mitigate or even reverse the downward
trends of the past. Theoretically, any urban jurisdiction can "bottom out," as large
blocks of inexpensive empty land stimulate new forms of investment.
Today, of course, St. Louis is far from emptied out. It still contains some 600,000
residents and 40 percent of the SMSA's business activity. However, its population
is on a course that cannot easily change: the white population will not cease declin-
ing without net in-migration, and the black population will not continue growing
unless out-migration ceases. If industrial location trends during the latter half of the
1960s continue for another five years, the city will be only one other center of
business activity, as opposed to the chief center. St. Louis County will contain more
economic activity than the city. Public revenues have become progressively more
difficult to generate locally: receipts from the earnings tax are falling in real terms;
the statutory limit on the property tax rate has been reached; and assessed valuation
is not increasing faster than inflation. In both 1971 and 1972, sales tax receipts were
disappointingly less than had been expected.*
Current and Future Inventory
Nevertheless, the city now has approximately 1300 acres available for develop-
ment. This figure could rise to 2200 by the year 1990, if current trends in land
clearance continue (through continuing population and business dispersion, along
with the present rate of building demolition). We can arrive at a rough estimate of
what new investment in these areas might mean for the city as follows:*
If almost 60 percent of the available land is allocated to industrial-commercial
use and the remainder to streets, alleys, and residential uses, then about 1300 acres
will be available for industrial-commercial development by 1990. Translating this
acreage into jobs— 67 percent allotted to industrial and 33 percent to commercial
' Annual Report of Comptroller, City of St. Louis. 1971, and conversation with Mayor John Poelker,
June 22. 1972.
' These estimates were made by Dempster Holland of St. Louis University and appear in Holland and
Wendel, Development of Industrial Parks.
39
40
jobs, as is now the case— and then allocating 26 employees to each industrial acre
and 85 employees to each commercial acre (using 1971 statistics), some 56,000 jobs
could be developed by 1990. Next, since the earnings tax is one of the most significant
sources of current city revenues,^' we estimate average salaries for the jobs to be
developed* and the consequent earnings tax: by 1980, approximately $3 million, and
by 1990 an additional $4.5 million, might be generated in city earnings tax.
These obviously crude estimates indicate possibilities, not probabilities; they do
not represent net increases; they assume short temporal lags between abandonment
and remvestment; nor are they consistent with the results of our survey of industrial
developers. If those results are indicative of a solid frame of mind on the part of the
private interests making the relevant decisions, the simple availability of land in the
city will not suffice.
Federal Decisions
Before we turn to specific local policy efforts to induce redevelopment, it is
important to mention three pending decisions. Each will be made primarily at the
federal level, and any one of them could reduce the effectiveness of local attempts
to brmg economic activity and residences back to the city. They are:
(1) The development of bottomland. About 5000 additional acres could become
available for industrial development in St. Louis County alone with construction of
the Meramec Basin Dam by the Army Corps of Engineers.
(2) A projected interstate highway link from Kansas City to Chicago. This high-
way will cut two hours and over 100 miles from the present route through St. Louis.
Many argue that the new highway will hurt both the trucking and tourism indus-
tries in the St. Louis metropolitan area, and particularly in the city.
(3) A new airport site. The development of a new airport to the southwest of
the city, in addition to Lambert Field on the northwest, would further act to draw
new mdustrial development away from the city limits. (This is not to imply the
converse— that an airport located to the east in Illinois would draw business back
mto the city.) However, any new attraction for development west of the city is likely
to hurt, with more western development reinforcing a strong existing tendency.*
Local Policies: Redevelopment Possibilities
Meanwhile, the St. Louis City Plan Commission has just published a new 15-
year development program" intended to reestablish the city as a viable working and
thPv S! * l-pcrcent tax on salaries of residents wherever they work and employees wherever
they live, which accounte for approximately one-third of city revenues.
r^rLi°^S""^ broken down by industrial classification and average salary for 1970 (according to current
n^,^? '" categories). Average salaries for 1980 and 1990 were then computed, assuming
a 5-percent increase compounded annually, using current wage guidelines.
current dispute about airport location, involving expansion of Lambert Field versus construc-
TLl^ fl'^^l'" '""'f' clear^ut. Lambert exists, and much ot what mav be attracted to
AfThi . . ' ^ ^'"■^"'^y attracted. However. Missouri locations to the wes't and southwest
01 the city were suggested several years ago. and could be revived as candidates for new airport location
St. Louis Development Program (A Summary). St. Louis City Plan Commission, January 1973
41
residential community. Although emphasizing physical development, particularly
of residential neighborhoods, the program also focuses on controlling crime, improv-
ing educational opportunity, and restructuring some parts of city government. The
plan envisions differential treatment of neighborhoods aimed at retaining the stabil-
ity of neighborhoods that are presently sound, rehabilitating other neighborhoods,
and continuing demolition in still others.
For the short-range phase of the development program that stresses residential
betterment, estimated financial requirements are $154 million. The program recom-
mends, in addition, a four-year public improvement effort involving: $6 million for
demolition; $7.2 million for waste disposal and pollution abatement; $29 million for
facilities to encourage economic development; $68 million for transportation im-
provements; $16 million for major recreational facilities; and $40 million for educa-
tional facilities. Accomplishment of the long-range (15-year) plan is estimated to
require close to $1.5 billion.^
In addition to urban redevelopment carried out by the St. Louis Land Clearance
for Redevelopment Agency (both federally assisted and nonfederally assisted), three
policies are being used in concert to enhance the prospects for private investment
in the city:
(1) The Missouri Urban Redevelopment Corporation Law, which provides the
power of eminent domain to corporations planning expansion or redevelopment.
This allows more efficient accumulation of land and is accompanied by a 25-year
schedule of tax abatement.
(2) Planned Industrial Expansion, which allows industrial revenue bonds to be
used for industrial development.
(3) The Land Reutilization Act, which permits the city to foreclose on tax-
delinquent property, thereby enabling the city to accumulate property for purposes
of restoration or rezoning for new uses.
Recent private investment has been substantial, according to the Plan Commis-
sion. For example, they point out that the Mercantile Trust Company announced
plans this year for a $150-million Mercantile Center. The Boatmen's National Bank
has announced plans for a $23-million project. Design of a $25-million public Con-
vention Center is nearing completion. Official approval is near for a $75-million
Convention Center Plaza private redevelopment effort. Construction has begun on
an $8-million addition to Stouffers Inn. Breckenridge Hotels Corporation has re-
quested approval for a $10-million hotel development over the vacant Spanish Pavil-
ion. General American Life Insurance Corporation has announced plans to build a
new headquarters at an estimated cost of $10 million. A Florida developer has
announced plans to renovate the city's old Post Office building, and other plans are
in progress.
Further, local officials see great promise in a new consciousness of neighborhood
identity among many of its residents, and a renewed interest in city dwelling among
young families. In the last five years, neighborhood corporations have burgeoned,
and neighborhood festivals — drawing from 10,000 to 50,000 people — have been tak-
' Ibid., pp. 3a-36.
42
ing place in increasing number. These festivals consciously promote the amenities
of in-city living, and encourage potential homebuyers to sign up to be contacted
when housing comes on the market. According to some estimates, between 2000 and
4000 young families have been attracted to city residence since 1970, either recruit-
ed by conscious neighborhood effort or drawn by their own tastes for city residence.
The Probabilities
No one can deny that local policy has taken an active and vigorous posture
toward reviving city life. But there are major uncertainties as to whether the under-
lying causes of the urban crisis in St. Louis can be effectively changed or reversed
by measures envisioned in the new plan:
Will the current revival of private investment in the city continue?
Will middle-income families and businesses be attracted back to the city in
significant numbers by these measures alone?
Will they manage to generate sufficient revenues to support municipal goods
and services, despite having to share these revenues with the city's disadvantaged
population?
Our analysis makes us doubt that the present policies alone can sufficiently
attract new investment to the central city. The city is capturing a dwindling fraction
of the new industrial and commercial development occurring in metropolitan St.
Louis. In 1968, approximately 56 percent of new investment in projects involving
$100,000 or more was in the city. In 1970, the figure was only 23 percent; and in 1972,
11 percent.® Although investors are betting that the scheduled new ofiice space will
be filled, their bets are cautiously hedged. Thus, the Mercantile Center development
is staged over a decade. It will begin with a $25-million building to house existing
bank facilities, requiring that only 50 percent of the space be leased on the open
market. Subsequent buildings— a luxury hotel and three more office buildings— will
be developed sequentially. As first steps are justified by new demand, next steps can
be taken. And caution is not unwarranted: in the city, utilization of general ofiice
space remained at 9 million square feet from 1955 to 1971, in spite of net additions
to supply of office space of over 2 million square feet.®
The city's power to hold current private economic investment and to attract still
further investment is somewhat compromised by the frequent requirement of feder-
al resources to force down the price of land. This necessity makes the city's future
economic development quite vulnerable to changing federal decisions (freezing HUD
funds, changing urban renewal policies). Indeed, one major company in the city has
been working with local and federal resources for five years to develop 44 acres of
surrounding land; they estimate that achievement of the development will take at
least another seven years. Thus, while city locations can be made attractive for
certain types of private investment, the encumbrances of so doing assure that subur-
ban locations will remain strongly competitive in the foreseeable future.
• Figures as listed in various issues of St. Louis Commerce. These figures do not include construction
by churches, schools, housing developments, and government. Also, they exclude listed investments
where no dollar value was shown or where locations were not specifically designated as being in or not
in the city of St. Louis (e.g., "ten restaurants, various locations," or "Mississippi River Transmission
Corporation — pipeline expansion").
• Institute for Urban and Regional Studies. Urban Decay in St. Louis, p. 35.
43
Another argument sometimes made is that fuel shortages will reduce the rate
of metropolitan decentralization, inducing higher densities of population and busi-
ness which will be more acceptable to the smaller families anticipated in the next
two decades. But transportation research in progress at Rand indicates that the
price of gasoline will have to triple to induce a 9-percent decline in vehicle miles
traveled. In addition, income elasticities are shown to be high for both automobiles
and gasoline, suggesting that as. incomes rise, the purchase and use of automobiles
will increase as well.
Even if decentralization is slowed, the city cannot expect to be sole beneficiary
of these trends. Under conditions that reduce decentralization, ei^ery jurisdiction in
a metropolitan area might expect to house a larger proportion of its labor force, as
some communities attract more employers of their residents, while others attract
as residents those who also work within their boundaries. Already, in the fifteen
largest U.S. metropolitan areas, an average of 72 percent of workers both live and
work in the suburbs. (St. Louis is close to the average with 70 percent.) And in nine
of those fifteen metropolitan areas, suburbs have equaled or far exceeded their
central cities as the principal location of jobs.*
Our conclusion is that no current policy available to the city can induce the rate
of private investment that would return the city to a position of economic dominance
in the metropolitan area. Nor does St. Louis appear to be moving in the direction of
becoming a predominantly black, self-supporting suburb.
Rather, what is happening now is that the major causes of the urban crisis are
stimulated and accelerated by conditions and policies beyond the reach of local
policy; local policy is left mainly to ameliorate their consequences— and left with
reduced sources of revenue to do even that. This position forces local policymakers
to devise short-term solutions, because they simply cannot finance long-term solu-
tions. Yet, paradoxically, the short-term solutions can worsen the longer-term prob-
lems.
St. Louis City's earnings tax is a case in point. In the short term, it captures as
much revenue as possible in an equitable way. The city cannot afford to eliminate
this revenue source until a very different municipal financing system is in place.
Nevertheless, in the longer view, the earnings tax falls most heavily on the use of
land in business districts and can be escaped by removing the activity. It falls most
heavily on residents who work outside the community and can escape the tax by
leaving the city. It creates a systematic incentive to live and work outside the city.
Tobe sure, the lower property values that may induce new residential and industrial
investment in the city can be traded off against the earnings tax. But as property
values rise with stimulated investment, the trade-off becomes less advantageous for
later investors, enhancing once again the competitiveness of suburban locations.
UVING WITH THE FUTURE
St. Louis does appear to have the opportunity to reduce the rate of its decline,
hut eiH^n this reduction requires new sources of revenue outside its own jurisdiction.
TTie New York Times, October 15, 1972.
44
Since the legislation to achieve that is clearly long-term, the city remains locked in
short-term strategies as described above. In our opinion, these strategies must con-
tmue to be developed in the understanding that for the most part the historical
functions of central cities are technologically obsolete today." Clustering of people
and of economic activity is no longer paramount to the degree it once was. St. Louis's
age and location within an outwardly sprawling urban region render it increasingly
just "another part of town." Making the best of what it has to offer means catering
more deliberately to the diversity of interests that lie within its boundaries. How
might this be done?
Our proposal is to engage in jurisdictional or administrative changes designed
to enable groups of common interests, tastes, and needs (e.g., neighborhoods) to
define and receive public goods and services tailored to those needs. Certain plans
of local policymakers are already aimed in this direction: the current city develop-
ment program proposes different strategies of intervention, depending primarily on
housing conditions in different neighborhoods. The full development of this strategy,
however, is dependent on a mechanism for generating revenues that will allow low-
income residents to live where they want to, without requiring the jurisdictions they
choose to depend predominantly on internal sources of financing
We have already noted in Sec. Ill that with rising incomes and diminishing
transportation costs, people disperse and regroup into homogeneous jurisdictions
where public services tailored to their particular desires and needs are provided. The
larger and more heterogeneous a taxing and service-delivering jurisdiction, the
more likely it will be that current forms of municipal financing and allocation of
public goods'* will return a lower proportion of the tax dollar to relatively affluent
citizens than to the less affluent in the form of goods especially tailored to their own
needs and tastes. To be sure, less affluent citizens might choose municipal expendi-
ture patterns quite different from those selected by wealthier citizens, and in that
sense public goods directed to the needs of disadvantaged groups provide less accept-
able returns on their tax dollars as well. However, since the total revenue available
from wealthier citizens is greater than that available from poorer citizens, the latter
receive greater benefits from the affluent than they could support out of their own
resources.
Thus, the more affluent have strong incentives to support their desired services
in separate smaller jurisdictions— much stronger than the incentives of the poor to
isolate themselves from more affluent neighbors. And rising incomes and diminish-
mg transportation costs increase the ability of more affluent city residents to form
new, more homogeneous jurisdictions— i.e., to suburbanize.
According to this argument, St. Louis City would not be the only municipality
tin the metropolitan area subject to departures of the better-off. And indeed, the
population of University City (a ring suburb) declined by 10 percent between 1960
sand 1970;'' and median income fell, though in no sense is University City a low-
#ncome community, even now. In this smaller community, where jurisdictional
" J^. "nay not be true of all those functions in all cities. Finance and communications may continue
to cluster in New York to a substantial degree. But our point is illustrated by the very different type of
sprawling central city represented by San Jose. rentiypeoi
" Including public services such as police, fire, sanitation, library facilities, and schools, as well as
public streets, parks, hospitals, auditoriums, etc.
" Brentwood is down 12 percent; Wellston is down 11 percent.
45
boundaries can be escaped by even shorter moves, there is some evidence to suggest
that racial as well as income transition accelerates movement from jurisdiction to
jurisdiction. On the basis of this understanding of urban processes, we would argue
that until some form of revenue-sharing— federal state, or metropolitan— makes the
poor a smaller financial burden for any single jurisdiction, rising incomes will contin-
ue to encourage the more affluent to flee and their amenities will encourage the less
affluent to pursue them. And if whites cannot tolerate sharing the same bundle of
public services with blacks, movement from jurisdiction to jurisdiction within the
metropolitan area will be further accelerated.
We suggest widening the jurisdictional boundaries at which revenues are collect'
ed. Federal revenue-sharing is an example of what we mean, though presently it is
neither substantial enough nor is it perceived as permanent enough to represent a
tenable solution to municipal financing problems. True, other levels of revenue-
sharing are feasible for some areas— metropolitan or state revenue-sharing. How-
ever, for the St. Louis metropolitan area, past voting records suggest that metropoli-
tan revenue-sharing has little likelihood of acceptance. Perhaps changes in suburbs
like University City will make possible a metropolitan coalition; perhaps the federal
government will come up with an effective incentive program for metropolitaniza-
tion (but we would not recommend such a program on the basis of this single study).
In any case, most metropolitan solutions for greater St. Louis seem out of current
reach.
If, however, ways were presented to make municipal financing less dependent on
the ability of current residents to pay, municipal governments would be more free to
experiment with different modes of service provision. At least some goods and services
might best be ordered and provided at very narrow jurisdictional levels, e.g., neigh-
borhoods. In this way, cities with heterogeneous populations might capture some of
the benefits of small homogeneous (and affluent) suburban municipalities where
residents can purchase and control the public goods and services they want.
STRATEGIES FOR LOCAL POUCY
Our analysis of St. Louis has discouraged us from emphasizing local policy
changes. In many ways, the city is already handling its inventory in ways our
analysis would suggest: subjecting hung-up inventory to demolition; accumulating
contiguous parcels of land in a land bank; discouraging small scattered develop-
ments where empty land has promise of accumulating; attempting to reduce the
price of city land. We have opinions about the consequences of certain local policy
issues that have been informed by our research:
• Branch banking would appear to promise more beneficial than negative
consequences for growth throughout the metropolitan area.
• Proposed Missouri sites for a new airport will reinforce the already strong
westward development in the area.
This initial evidence comes from early findings by Charles Leven and James Little, as reported in
Little, Housing Market Behavior and Mobility Patterns in a Transitional Neighborhood.
46
. Continuing development of bottom land in the metropolitan area will pro-
vide substantial new suburban inventory with which city sites must compete.
• The proposed interstate highway connecting Kansas City to Chicago may
hurt such industries as tourism and trucking in the metropolitan area and especially
in the city.
Local policies may have beneficial effects, but the most significant steps for
ameliorating the city's decline rest on policies that must be developed outside its
jurisdiction at either the state or federal level. Yet recommendations to state or
federal officials for major changes in urban policy must necessarily be tentative
when derived from the analysis of one city. Thus, rather than make recommenda-
tions, we present examples of policies that could make the poor a smaller financial
burden for any single jurisdiction:
• At the federal level, this calls for a much more substantial revenue-sharing
program that takes into account the large proportion of public goods (streets, hospi-
tals, parks), as well as services that cities currently support. Formulas for distribut-
ing revenues should provide higher than current returns for proportions of low-
income citizens.
• At the state level, a more limited form of state revenue-sharing could sup-
port selected public goods in cities— for example, public hospitals.
• At the metropolitan level, even limited revenue-sharing would help. For
example, revenue generated by industry in the metropolitan area might be appor-
tioned to municipalities in the area by a formula that would grant higher returns
to jurisdictions with high proportions of residents in poverty." This would reduce
the competition for industry between metropolitan jurisdictions and would promote
industrial location more suited to the environmental concerns of the whole met-
ropolitan area. We recognize that revenue-sharing of this type would be extremely
complex to accomplish across states; for that reason decisionmakers might consider
limiting such a plan to the Missouri portion of the SMSA.
• Alternatively, a metropolitan earnings tax would be possible. Once again,
revenues would be apportioned to area municipalities by a formula that would grant
higher returns to jurisdictions with high proportions of residents in poverty. '«
But what might the city do under current forms of generating revenue to lessen
the incentives that encourage affluent citizens to move to other jurisdictions?
(Though we address this strategy to the city, it would apply to any municipality.)
Our analysis suggests that the most helpful strategy toward this end would be to
gear the administration of municipal services and regulations to varying neighbor-
hood needs."
This practice is not new to municipal policymakers, though most jurisdictions
mamtain the principleoC providing the same set of public services and enforcing the
metropSitM a^T ^""^ "^^^ ^^'^ revenues from new development in the Minneapolis^. Paul
" CoMideration of the implications of this alternative should be aided by research in progress on the
earmngs tax under the direction of Norton Long at the University of Missouri. St. Louis,
-wfu- ^^"^"^■.^•l^'"* are several ways to accomplish this. One that considers jurisdictional changes
w^thm the city is discussed in Gardner and Payne, An Economic Analysis of Central City Decline. Other
T^TaT^ . ^ done appear in Center for Urban Programs. Recommendalions on
Ugal-Admintstrative Policies for the City of St. Louis. St. Louis University, 1971
47
same regulations in all neighborhoods. However, since housing stock varies consid-
erably as a municipality ages, it is not unreasonable to consider local policies that
impose different housing codes on varying stock. In the past, makers of home loans
(FHA, commercial and savings and loan banks), insurers of property, and owners of
property have acted upon their individual expectations of the changing future of
particular neighborhoods, escalating that change as they did so.*®
However, if cities could show clearly how municipal services and codes will
respond or ane responding to neighborhoods undergoing racial or income transition,
the anxieties of present or potential residents, anxieties that now lead to precipitous
neighborhood change, might well be reduced. Clear public prescriptions of this kind
could also lend support to citizens who seek financing for homes in transitional
neighborhoods. Research in progress at Washington University*® should be useful
in determining the services that public policies should stress for transitional areas.
*• See Institute for Urban and Regional Studies. Urban Decay in St Louis,
Leven and Little's Grant GI-37861-NSF, A Study of Determinants of Inter-Neighborhood Mobility.
BIBLIOGRAPHY
I. DOCUMENTATION
The Center for Community Change and the National Urban League, The National
Survey of Housing Abandonment, April 1971.
Center for Urban Programs, Recommendations on Legal-Administrative Policies for
the City of St, Louis, St. Louis University, St. Louis. Mo., 1971.
City of St. Louis, Annual Report of Comptroller, 1971.
City of St. Louis, Ordinance 55681, Section 2126.1, July 15, 1970.
Enns, J., and P. deLeon, The Effect of Highways upon Metropolitan Dispersion: St.
Louis, The Rand Corporation, P-5601, September 1973.
Gardner, C, Banking Regulation and Urban Growth, The Rand Corporation. P-
5057, August 1973.
, and G. Payne, An Economic Analysis of Central City Decline, The Rand Corpo-
ration, R-1350-NSF (forthcoming).
Holland, Dempster, "Population Change in Seven Midwestern Qties," St. Louis
University, unpublished paper, 1973.
, and George Wendel, Development of Industrial Parks, The Rand Corporation,
R-1358-NSF (forthcoming).
Institute for Urban and Regional Studies, Urban Decay in St. Louis, Washington
University, St. Louis, Mo., March 1972.
Keeler, Emmett, and W. Rogers, A Classification of Large American Urban Areas,
The Rand Corporation, R-1246-NSF, May 1973.
Little, James, Housing Market Behavior and Mobility Patterns in a Transitional
Neighborhood, Institute for Urban and Regional Studies, Washington Univer-
sity, St. Louis, June 1973.
Levine, Robert, San Jose, The Urban Crisis, and the Feds, The Rand Corporation,
P-4839, May 1972.
, and D. Alesch, Growth in San Jose: A Summary Policy Statement, The Rand
Corporation, R-1235-NSF, May 1973.
Morrison, Peter A.. The Impact and Significance of Rural- Urban Migration in the
United States. The Rand Corporation, P-4752, March 1972.
, Population Movements: Where the Public Interest and Private Interests Con-
flict. The Rand Corporation, R-987-CPG, August 1972.
, San Jose and St. Ixyuis in the 1960s: A Case Study of Changing Urban Popula-
tions, R-1313-NSF (forthcoming).
49
50
, Small-Area Population Estimates for the City of St, Louis, 1960-1972, with a
Model for Updating Them, The Rand Corporation, R-1373-NSF (forthcoming).
The New York Times, October 15, 1972.
Nixon, Richard M., Economic Report of the President, transmitted to the Congress,
February 1971.
Pakonen, R., "The Differential Effect of Branch Law Regulation on Commercial
Bank Entry," Ph.D. dissertation, Washington State University, Pullman,
Washington, 1971.
Rothenberg, J., F. Fisher, et al, "A Model of Metropolis," Papers and Proceedings
of the American Economic Review, Vol. LXII, No. 2, March 1972.
Sharpe, William, "Capital Asset Prices: A Theory of Market Equilibrium under
Conditions of Risk," Journal of Finance, Vol. 19, No. 3, September 1964, p. 425.
Slitor, Richard, "Tax Effects on Urban Growth in Three Cities: San Jose, St. Louis,
and Seattle," The Rand Corporation, unpublished document.
St. Louis City Plan Commission, St. Louis Development Program (A Summary), St.
Louis, January 1973.
SL Louis Commerce, various issues.
U.S. Bureau of the Census, Census of Population and Housing: 1970; General Demo-
graphic Trends for Metropolitan Areas, 1960-1970, Final Report PHC(2)-1,
PHC(2)-27, PHC(2)-15, Government Printing Office, Washington, D.C., 1971.
U.S. Department of Commerce, County Business Patterns, Washington, D.C., 1959,
1970.
U.S. Department of Housing and Urban Development, A Study of the Problems of
Abandoned Housing, Washington, D.C., November 1971.
U.S. Department of Transportation, Federal Highway Administration, Highway
Statistics, Washington, D.C., annual volumes, 1965-1970.
Voorhees, Alan M., and Associates, Inc., Technical Report on a Residential Blight
Analysis for St, Louis, Missouri, Washington, D.C., March 1969.
11. SELECTED BACKGROUND REFERENCES
Articles
Conforti, Joseph M., "Newark: Ghetto or City," Society, Vol. 9, No. 10, Special Issue,
September/October 1972, pp. 19-32.
Ganz, Alexander, and Thomas O'Brien, "The City: Sandbox, Reservation, or Dyna-
mo?" Public Policy, Vol. 21, No. 1, Winter 1973, pp. 107-123.
Von Furstenberg, George M., "Place of Residence and Employment Opportunities
Within a Metropolitan Area," Journal of Economic Issues, Vol. 5, June 1971,
pp. 101-116.
Books
Alonso, William, and John Friedmann, Jr. (eds.). Regional Development and Plan-
ning, The M.I.T. Press, Cambridge, Mass., 1964.
Banfield, Edward C, Big City Politics: A Comparative Guide to the Political Systems
. of Atlanta, Boston, Detroit, El Paso, Los Angeles, Miami, Philadelphia, St
Louis, Seattle, Studies in Political Science, Random House, New York, 1965.
, The Unheavenly City, Little, Brown and Company, Boston, 1970.
Beckman, Martin, Location Theory, Random House, New York, 1968.
Bergsman, J., P. Greenson, and R. Healy, Explaining the Economic Structures of
Metropolitan Areas, Urban Institute Working Paper 200-1, The Urban Insti-
tute, Washington, D.C., 1971.
Birch, David L., The Economic Future of City and Suburb, Urban Studies Series,
Committee for Economic Development, New York, 1970.
BoUens, John C. (ed.). Exploring the Metropolitan Community, University of Cali-
fornia Press, Berkeley and Los Angeles, 1961.
Burgess, Ernest W., and Donald J. Bogue (eds.). Contributions in Urban Sociology,
University of Chicago Press, Chicago, 1963.
Dean, R., W. Leaky, and D. McKee, Spatial Economic Theory, The Free Press,
Riverside, N.J., 1970.
Downs, Anthony, Urban Problems and Prospects, Markham Publishing Company,
Chicago, 111., 1970.
Duncan, Otis D., et al.. Metropolis and Region, The Johns Hopkins Press, Baltimore,
Md., 1960.
Forrester, Jay W., Urban Dynamics, The M.I.T. Press, Cambridge, Mass., 1969.
Ganz, Alexander, Our Large Cities, New Light on their Recent Transformations-
Elements of a Development Strategy; a Prototype Program for Boston, The M.I.T.
Press, Cambridge, Mass., 1969.
51
52
Goldsmith. Harold R. and Elizabeth L. Unger. Differentiation of Urban Sub-Areas,
Mental Health Study Center, National Institute of Mental Health. Bethesda,
Md., 1970.
Gordon. Mitchell. Sick Cities. The Macmillan Company and Collier-MacmiUan Ltd..
New York and London, 1963.
Greer Scott, The Urbane View, Oxford University Press, New York 1972
Grigsby. William G., Housing Markets and Public Policy. University of Pennsyl-
vania Press, Philadelphia, Pa., 1963.
Hughes, Janies W Urban Indicators. Metropolitan Evolution, and Public Policy.
Center for Urban Policy Research, Rutgers University. The State University
of New Jersey, New Brunswick, N.J., 1963.
Karsch. Robert F.. The Government of Missouri, 11th ed.. Lucas Brothers Publish-
ers, Columbia, Mo., 1971.
Nourse. Hugh, Regional Economics. McGraw-Hill Book Company, New York, 1968
Perloff. Harvey and Lowdon Wingo. Jr. (eds.). Issues in Urban Economics. The
Johns Hopkins Press, Baltimore, Md., 1968
Rossi Peter H., and Robert H. Dentler, The Politics of Urban Renewal. The Free
Press. New York, 1961.
Schmandt. Hen^^ J Paul G. Steinbicker. and George D. Wendel, Metrr>politan
Reform m St Louis: A Case Study. Holt. Rinehart. and Winston. New York,
Schnore. Leo F.. The Urban Scene. The Free Press. New York. 1965
Schuman. Howard, and Barry Gruenberg. Impact of Cities on Racial Attitudes,
Survey Research Center. Institute for Social Research. The University of Mich-
igan. Ann Arbor, Mich., 1970.
Stemlieb, George, The Tenement Landlord. Center for Urban Policy Research, Rut-
gere University, The State University of New Jersey, New Brunswick. N.J..
p'^eS^ Chtc^o^m2^^^^ C9/«<rucrto« of Communities. University of Chicago
federal Activities Aff-ecting Location of Economic
if. Offi w' ^'P*"^' ^' ^"'"'"^'^ P^"^ I-V. Government Print-
ing Office, Washington, D.C., 1970.
Yin, Rol^rt K. (ed.). The City in the Seventies. F. E. Peacock Publishers. Itasca. III..
Publications
Alliance for Re^onal Community Health. Inc. (ARCH), The Final Report of the
Health Facilities Planning Committee:Wo\. II, Profiles of the ARCH Planning
^^e^n(Presented to the ARCH Board of Directors), St. Louis. Mo.. October 20.
Bureau of the Census. Census of Population and Housing. 1970. PHC(3)-21. Employ
^"l^^fil^^ of Selected Low-Income Areas. Government Printing Office.
Washington, D.C., 1970.
East-West Gateway Coordinating Council, Industry in the St. Louis Region: Indus-
trial Expansion. 1966-69. St. Louis, Mo., 1971.
53
Eisenstadt, Karen, Factors Affecting Maintenance and Operating Costs in Private
Rental Housing The Rand Corporation, R-1055-NYC, August 1972.
Flax, Michael J., A Study in Comparative Urban Indicators: Conditions in 18 Large
Metropolitan Areas, The Urban Institute, Washington, D.C., April 1972.
Gladstone, Robert, and Associates (a series of Community Renewal Program techni-
cal background reports prepared for the St. Louis City Plan Commission to
provide a basis for a St. Louis Development Program), Washington, D.C., 1968-
1969.
, Analysis of Housing within St Louis, November 1968.
, Basic Economic Indicators for SL Louis, 1968.
, City-State Fiscal Relations, June 1969.
, Commercial Development Potentials, 1968.
, Housing Policy Guidelines for St. Louis, November 1968.
, Housing Trends and Outlook for St, Louis, November 1968.
, Industrial Development Potentials, 1968.
Governor's Advisory Council on Local Government Law, Missouri Local Govern-
ment at the Crossroads (Report presented to the Honorable Warren E.
Hearnes), Jefferson City, Mo., November 1968.
Heckman, Jan L., Housing Conditions, Needs, and Programs in the City of St, Louis
(East- West Gateway (^rdinating Council Staff Technical Report), St. Louis,
Mo., December 1970.
Mendelson, Robert E., and David C. Ranney, East St Louis Fiscal Crisis— 1970,
Regional and Urban Development Studies and Services, RUD Report No. 4,
Southern Illinois University, Edwardsville, 111., October 1970.
President's Committee on Urban Housing, A Decent Home, Government Printing
Office, Washington, D.C., 1969.
Rainwater, Lee, The Problem of Lower Class Culture (Prepared for Sociology De-
partment Colloquium, University of Wisconsin), Washington University, St.
Louis, Mo., 1966.
Ranney, David C, Fiscal Crisis in East St Louis, Illinois, Public Administration
and Metropolitan Affairs Program, Southern Illinois University, Edwardsville,
111., February 1967.
St. Louis City Plan Commission, A Housing Program for the City of St Louis 1970-
1980, St. Louis, Mo., April 1971.
, Model City Land Use Plan, St. Louis, Mo., June 19, 1972.
St. Louis Commission on Crime and Law Enforcement, St Louis High Impact Anti-
Crime Program Plan (Prepared for the Mayor of the City of St. Louis, Alfonso
J. Cervantes), St. Louis, Mo., April 24, 1972.
St. Louis Development Program (CRP) Staff, History of Renewal for St Louis, Mis-
souri (one of a series of Community Renewal Program technical background
reports to provide a basis for a St. Louis Development Program), St. Louis, Mo.,
February 1971.
Tobin, Gary A., The St Louis School Crisis: Population Shifts and Voting Patterns
(Published by Washington University for the Department of History through
the Mayor Raymond R. Tucker St. Louis Award Fund), St. Louis, Mo., June
1970.
U.S. Commission on Civil Rights, Hearing Before the US, Commission on Civil
64
Rights. St. Louis, January 14-17, 1970, Government Printing Office, Washing-
ton, D.C., 1971.
Voorhees, Alan M., and Associates, Inc. (a series of Community Renewal Program
technical background reports prepared for the St. Louis City Plan Commission
to provide a basis for a St. Louis Development Program), McLean, Va., 1970.
, Industrial Facilities Survey and Analysis, 1970.
, Residential Quality Within the City of St. Louis, Missouri, March 1970.
