Critique of UIC-UTC “Highways and Urban Decentralization” Study

Introduction

The University of Illinois Urban Transportation Center was awarded a contract from the Illinois State Toll Highway Authority (ISTHA) to assess whether and how highways have promoted sprawl in the Chicago region. The contract was awarded in a politically charged atmosphere. The Toll Highway Authority has been under pressure from the state legislature to reform certain of their policies and practices to become more accountable for their actions to the public and to the elected leadership of the state. ISTHA thus far has been unsuccessful in attempts to build two new roadways, I-53 in Lake County and the I-355 extension in Will County, due in large part to citizen groups that object to the sprawl that inevitably follows new road construction. This report was apparently intended to negate the concerns of citizens and legislators about the sprawl-inducing effects of new tollways. The Urban Transportation Center has made the report available online.

Ironically, the ISTHA-UIC report avoids examining the relationship between sprawl and transportation by being overly broad in its conclusions. Useful data, such as the relative timing of road and housing construction, were not made available in the report. The Center for Neighborhood Technology is able to make that information available here:

Data Source: Bell Federal Savings and Loan Association Survey of Building 1951-1990

The methodology and data used in this study are misleading, confusing, poorly referenced, and often irrelevant. The conclusions drawn from the data are often erroneous. Furthermore, even if one were to accept the methodology and conclusions, the public policy direction which the report seems to imply does not logically follow.

Broadly Constructed Fallacies

Apparently, the authors of the study intended to draw two conclusions: (1) that highway construction does not cause sprawl, and, (2) that sprawling development is economically beneficial and naturally accompanies prosperity. The study fails on both counts.

Several general concepts were given only cursory analysis, and outdated reports were used because they were the only ones which supported the conclusions the authors of the study intended to draw. In many cases the conclusions contradict known phenomena; results are also in direct conflict with the statements of the Illinois Department of Transportation.

Sprawl and Economy

FICTION Comparing Chicago to Lagos, Nigeria, and other developing nations, the study concocts a direct relationship between density and Gross Domestic Product (GDP) (p.2). The implication is that the densest cities in the world are the poorest and that sprawl is therefore good unless you prefer third-world living conditions. The table compares only the national data for the given cities. The three U.S. cities listed, New York, Los Angeles and Chicago, all appear as having the same GDP.

FACT Absent any other variables, this information is simplistic and deliberately misleading. What is the impact of natural resource availability, geographic barriers to sprawl, form of government, type of economic system, educational level of workforce, availability of high quality nutrition, and other factors that contribute to the U.S.’s affluence?

If there is any validity at all to using these two measures in isolation from other variables, a more useful comparison would be between U.S. cities, where the excluded variables do not vary between such extremes. The three U.S. cities in the UIC chart all show the same GDP because the figures are national, not regional. A close approximation for GDP for a city is per capita personal income. In fact, personal income is more reflective of the wealth the researchers claim leads to lower density development. If one compares this statistic for New York, Los Angeles, and Chicago, one might draw the opposite conclusion: the New York City metropolitan area, the densest of the three, is also the wealthiest.

In fact, if we examine several metropolitan areas around the U.S., no definitive pattern emerges. The selection of only New York, Los Angeles, and Chicago is not justified in the report. For a broader perspective, we examined ten U.S. cities to compare their wealth and population density. The rank given is in relation to 318 Metropolitan Statistical Areas as defined by the Census Bureau.

In fact, if any rough conclusions can be drawn from this table, they would be opposite those found in the report. If the population density and per capita personal income of these cities is plotted on a scatterplot, we see a general trend correlating higher personal income with greater population density. The two least dense cities, Las Vegas and Tucson, also have the lowest personal income of the group. A more in depth examination of this sort would be far more valuable than comparing three American cities to a host of third-world countries.

These personal income and density statistics are available from the Census Bureau. The UIC study fails to reference its density data.

An additional fallacy in using international GDP figures converted into dollars arises from differences in exchange rates and buying power in different economies. A more useful economic measure would have been purchasing power. In fact, the report itself claims to attribute decentralization to increased “incomes relative to prices” (p.2) yet lists no figures that indicate local prices.

The disparities between urban and rural dwellers in some developing nations creates more distortion of the GDP per capita figures, because in many of those nations the rural population is dramatically poorer, on average, than the urban populations. A comparison that used urban regional figures, rather than national figures might have been marginally informative.

FICTION The UIC researchers say that affordable housing is made more available by highways opening up new land.

FACT The price of land near expressways always increases dramatically after construction is completed. This phenomenon was recognized nationally as early as 1939 by the Roosevelt Administration, and continues to hold true today. (See a study on increased property values in areas served by new transportation infrastructure.) Two other factors never considered in the study are the increase in household transportation costs when families in sprawling locations must own two or more cars, and the subsidy provided by the state to suburban developers in the form of new infrastructure like sewer lines which hides part of the cost of the new housing. Commonly-practiced exclusionary zoning further limits the affordability of newly developed land.

FICTION “Highways tend to reduce the price of land in areas previously well served and renders the central city more attractive to firms in other cities.” (p. 6)

FACT The phenomenon referred to here is called “disinvestment” and has been a destructive, rather than constructive, force in urban evolution. A recent presentation by Professor Marlon Boarnet at the Brookings Institute describes this phenomenon: “Recent evidence suggests that, at the margin, highway infrastructure contributes little to state or national productivity … Yet the idea that highways enhance economic health is common in the policy and planning communities. … Part of the reason for this disconnect is that when growth occurs along a new highway, people fail to acknowledge that part of that activity shifted from somewhere else within a region.” This findings are described in detail in M. Boarnet, “Highways and Economic Productivity: Interpreting Recent Evidence,” Journal of Planning Literature, volume 11, number 4, pp. 476-486, May, 1997.

Sprawl and Highways

FICTION “Transportation is a necessary but not sufficient condition for economic development.” (p 27).

FACT Of course transportation cannot spur economic development in a vacuum; other basic infrastructure elements like sewer lines, a water supply, and a livable climate are also needed. The role of transportation in development can hardly be overstated, however: Chicago’s whole economic history is built on its central role in the national transportation networks, from water, through rail, to roads and on to airports. The Illinois Department of Transportation in its March 1997 report, “The State of Transportation in Illinois: Lifelines to the Economy” (emphasis added) states, “An adequate transportation network is indispensable to a healthy economy. It is essential for the distribution of products and services to local, regional, state, national and international markets. It is essential for getting people to work and to the other places where they participate in the economy …”

FICTION The UIC study does not focus on the actual toll highways or specific growth patterns in the decade(s) following construction, but takes a sweeping view of development over the last century. The two maps (pp. 10 & 11) included in the report confuse the issue by showing highways that were not yet built during the decades the maps purport to represent. The study states “This analysis … revealed no evidence of a relationship between the regionwide rate of decentralization and the time of completion of the expressways in the Chicago region.” (p. 6)

FACT UIC’s researchers refer to decentralization since the turn of the century as a singular process, when it is well known (see Losing Ground, a report released in the fall of 1998 by the Openlands Project) that decentralization in the early to mid-1900s occurred compactly along rail lines, but has been broadly dispersed throughout the region since the highway era began. See the map above for a visual representation of the timing of highway construction as it relates to housing starts. Although the relationship between highway construction and residential growth is complex, these maps clearly show that they are interdependent and that often highways preceed housing development.

FICTION The report mentions several “activity centers” outside of the central business district, such as the I-88 corridor, O’Hare Airport, Schaumburg, Oak Brook and Old Orchard, all of which are immediately adjacent to highways. Other examples, such as downtown Evanston, Orland Park, Stratford Square, Fox Valley, Randhurst, Hawthorne, Franklin Park and Elk Grove Village “grew without immediate access to the expressway systems,” (p. 18) and this demonstrates that highways do not cause sprawl.

FACT Observing that certain “activity centers” have developed in locations not immediately adjacent to highways does not demonstrate anything. The thesis that highways do not cause sprawl would only be validated by the repeated observation of areas with immediate access to the expressway systems that do not show signs of economic activity.

Furthermore, the examples given do not illustrate the intended point. Evanston, which was first settled in 1835 and officially incorporated in 1863, is hardly typical of the dispersed development patterns that have occurred in the region since the mid-1950s. Development in Evanston occurred compactly along rail lines before the onset of the highway era.

Orland Park is referenced later (p. 23); the Orland Square Shopping Center is “located in a triangle formed by I-80, I-55 and I-294 … The closest expressway interchange is approximately four miles to the south on I-80.” Apparently the researchers have defined “access to” an expressway as being immediately adjacent to an interchange. The other four shopping centers mentioned are three to four miles from interchanges:Stratford Square (from I-355), Fox Valley (I-88), Randhurst (I-294), and Hawthorne (I-294).

In the case of Elk Grove Village there is a tollway immediately adjacent to the village and the CNT analysis of new housing starts cited above shows that all the communities adjacent to Elk Grove Village experienced their peak building years in the decade immediately following construction of I-90. Franklin Park and Elk Grove Village are two communities that experienced their peak building years during or before construction of I-90; they are also adjacent to another major transportation facility, O’Hare Airport. It is disingenuous to use the examples of Elk Grove Village and Franklin Park without mentioning the many examples of nearby communities that grew after I-90 was built.

Inconsistencies in Details

This report includes lengthy data manipulations that give a cursory reader the impression of a lot of statistical detail; unfortunately most of the tables, charts and data are extraneous and unresponsive to the question of whether roads contribute to sprawl, as well as poorly referenced. Examples include:

• Using data that has no source cited, the researchers say that “average travel time even shows a slight decrease” between 1970 and 1990. (p. 14) The highly respected Texas Transportation Institute released a report on highway expansion and congestion in November of 1998 that found a significant increase in annual hours spent in congestion from 19 in 1982 to 42 in 1996, in the Chicago region.
• “We did not include employment because of a lack of necessary data.” (p. 1) The Urban Transportation Center at UIC is well aware of data made available by the Northeastern Illinois Planning Commission on the location and number of jobs in the region, and has used employment data in previous publications. Omitting employment data results in an incomplete picture of development in the region.
• An arcane computation of a decentralization measure for the last seven decades shows nothing about how and where the population of the region has dispersed. (p. 7)
• Another section tries to ascribe a numerical value to the ratio of population, employment, number of passenger vehicles, and number of retail sales employees to the number of lane miles in each county. The section concludes: “Even though rural counties have traditionally had the highest ratios, this has not caused these counties to increase in population, in fact many rural counties are experiencing population losses.” (p. 23) What rural counties they are referring to is unclear; the table immediately following this statement lists the six counties in the Chicago region, none of which has lost population. Furthermore, the ratio itself is deceptive: since population is very low in agricultural areas, even a single highway running through the county would present a high ratio. Sprawl occurs at the outer fringes of the metropolitan area, not in distant rural land.
• “Another way to examine the conventional wisdom that decentralization and suburbanization are only post WWII phenomena is to calculate when the city of Chicago would have started to lose population if its boundaries were smaller or larger than the current limits,” proving that “Had Chicago only had 47 square miles, similar to San Francisco and Boston, it would have started losing population in the 1920’s. If it were the size of Jacksonville then suburbanization would have started around 1970.” Not only is this extraneous data, but the authors have explicitly stated (p. 7 & p. 12) that decentralization was occurring by the turn of the century, so they are contradicting themselves.
• Eight pages of text, tables, and graphics are devoted to a recitation of statistics on travel behavior, trip chaining, trip purpose, trip-length frequency distributions, vehicle miles traveled, and other information marginally related to the question of whether construction of highways causes sprawl.

Conclusion

Sprawl is a complex phenomenon, with myriad contributing factors and uncountable consequences. The UIC Urban Transportation Center has the resources and technical competence to treat the subject appropriately. Unfortunately, even the easily performed analyses such as those in this critique were omitted, resulting in an insufficiently developed report. The study purports to address the role of transportation in sprawling development, yet sidesteps the crucial issue by only examining other economic factors. We can only hope that future research will make better use of the Urban Transportation Center’s exceptional resources.

For an alternative vision of transportation in Chicagoland, see the Changing Direction: Transportation Choices for 2030.