Oversight Hearing on "Perspectives on America's Transit Needs."

Thank you for inviting me to testify today.

My name is Wendell Cox. I am an independent consultant headquartered in Belleville, Illinois, in the St. Louis area. I am also a visiting fellow at The Heritage Foundation. I must stress, however, that the views I express are entirely my own, and should not be construed as representing the position of The Heritage Foundation.

I was appointed to three terms on the Los Angeles Country Transportation Commission by Mayor Tom Bradley and was appointed to the Amtrak Reform Council by Speaker Gingrich. Earlier this year I served an assignment as a visiting professor at the Conservatoire National des Arts ET Metiers (CNAM), a French national university in Paris, conducting seminars and research on urban planning and transport.

I will share perspectives that you may not have heard before --- about how transit has little or no potential to address traffic congestion and how so-called "smart growth" promises to worsen traffic congestion while making housing less affordable especially for the nation’s lower income households who are disproportionately minority. These views are held by other professionals and academics as well, and they challenge what is considered to be the conventional wisdom in both transport and urban planning. I will, of course, be pleased to supply the Committee with additional details as requested.

It is painfully obvious to commuters in virtually all US urban areas that traffic is getting worse. This has been going on for some time, but has become much more critical in recent years. For example, the US Census Bureau reports that average work trip travel time increased 3.1 minutes nationally from 1990 to 2000, a rate four times that of 1980 to 1990. And, things are likely to get much worse (Figure #1).

Figure 1

For some time there has been a widely held view that transit has the potential to reduce urban traffic congestion. Indeed, that sentiment was part of the rationale behind making highway user fees available to transit in the 1982 reauthorization.

Yet, despite spending nearly $500 billion in subsidies at the federal, state and local level since 1960, transit’s share of urban trips has continued to trend downward. This is confirmed by the 2000 Census, which shows that transit’s share of work trips has reached a new low --- 4.6 percent (Figure #2), down more than 10 percent from 1990 (Table #1). While employment was increasing 13.2 million, transit work trip use declined nearly 23,000. Only two metropolitan areas with more than 1,000,000 population maintain a transit work trip market share of more than 10 percent (Table A-1).

This is not to suggest that transit does not play an important role. Make no mistake about it --- where transit provides auto-competitive service, people use it. To be auto-competitive, transit must be, at a minimum, time competitive with the automobile. The 2000 Census data indicates that transit work trips take considerably longer than auto work trips. The average transit work trip was 43 minutes, which compares to other modes (mainly auto) at 24.8 minutes. Transit work trips take longer than auto trips in all metropolitan areas with more than 1,000,000 population (Table A-2).

But where transit is auto-competitive, it is very successful. For example:

• In the Tokyo area, nearly 60 percent of travel is on transit. Why? The reason is that transit is generally faster than the automobile, with the dense network of urban rail services and frequent connecting bus service making transit service attractive throughout the urban area. This is illustrated by the fact that Tokyo, with approximately 1.5 times the population of the New York urban area, carries approximately 4.5 times as many riders annually, and 60 percent more riders than all of the transit systems in the United States combined. Nonetheless, the automobile is making significant inroads in the Tokyo area. In suburban areas, where the auto is more competitive, transit market share is nearly 40 percent below that of the central city. And, throughout the area, transit market share has fallen nearly 30 percent since 1970.
• In the Paris area, 24 percent of travel is by transit. There, the commuter rail and metro systems provide generally faster travel than the automobile, with more than 60 percent of central city oriented travel on transit. But the situation is much different in the suburbs, where 80 percent of Parisians live and work. There, only 15 percent of travel is on transit, because transit provides little suburb to suburb service that is auto-competitive. As in Tokyo, transit’s market share has fallen in the Paris area, approximately 40 percent since 1970. The situation and experience in Paris is typical of other urban areas in Western Europe.
• In the New York area, approximately 10 percent of travel is on public transit. Much of this travel is focused in the city of New York, and especially to the Manhattan central business district (below 59^th Street). In 1990, approximately 75 percent of work trips to this area were on transit. In fact, approximately 22 percent of the nation’s transit work trips were to the Manhattancentral business district, an area of less than 10 square miles. Nearly 73 percent of work trips to locations within the city of New York were on transit, representing more than one-third of the nation’s transit work trips. Yet, in the vast sprawling suburbs of New York, only four percent of work trips were by transit. The difference is that substantial levels of auto-competitive service is provided to Manhattan and within the city of New York. It is possible to provide auto-competitive service because of the city’s high population and density, and the slower automobile operating speeds that result from the intense traffic congestion. Other than travel in and to the city, however, few trips can be made by auto-competitive transit.
• Transit accounts for more than 50 percent of work trip travel to the Chicago central business district, the nation’s second largest downtown. Again, the key is auto-competitive service available to this concentrated area from most locations in the urban area. Within the city itself, transit’s work trip market share is nearly one-third. But in the suburbs, transit’s work trip market share is barely three percent. Travel by transit from suburban residence to suburban employment locations can average more than two hours, in an urban area where automobile work trips average less than 30 minutes.
• In most other major metropolitan areas, transit competitive service is largely limited to downtown locations, and often only during peak commuting hours. Outside the New York and Chicago metropolitan areas, overall transit market shares tend to be approximately one percent (Figure #3)

Figure 3

What all of this says is that transit is largely about downtown and to a lesser degree a small number of urban cores (Figure #4). Overall, only the New York and Chicago metropolitan areas maintain a transit work trip market share of more than 10 percent, with most of that concentrated in the core areas. In fact, more than one-half of the nations’ transit work trips are to locations within the central cities of New York, Chicago and the ten next largest central business districts. Thus, 3.1 million of the nation’s transit work trips are to a gross area of less than 600 square miles, while the balance of 2.8 million transit work trips are to the other more than 86,000 square miles of urbanized land. Thus, outside trips to downtown, transit is able to make little or no difference with respect to traffic congestion.

In most major metropolitan areas, auto-competitive transit service is limited to downtown service. This is illustrated by Portland (Oregon), which has the nation’s most aggressive smart growth policies and has nearly doubled transit service in the last decade. Transit competitive service is provided from approximately 70 percent of the urban area to downtown (where transit competitive is defined as 1.5 times auto travel times). Outside downtown employment locations are accessible to only five percent of the urban area by transit competitive service. People are not going to forsake their cars for transit service that takes too long, or transit service that doesn’t even exist.

Figure 4

It is not surprising, therefore, that people who use transit to non-downtown locations have much lower incomes than those able to access the auto-competitive services to downtown. In 1990, downtown transit commuters had an average household income within six percent of the national average. Non-downtown transit commuters had an average household income 40 percent below average. It would appear that transit is used for non-downtown work trips only by those who don’t have a choice (those who have no automobile available).

The key to getting people out of their cars is to provide automobile competitive service --- service that is competitive in travel time. But, as noted above, there is little auto-competitive service in the United States and little more planned to areas other than downtowns. And, despite their vertical impressiveness, downtowns represent a small and declining share of metropolitan employment in the United States. In 1990, the average downtown area accounted for barely 10 percent of metropolitan employment (Figure #5). Even Manhattan’s central business district, the second largest in the world, accounted for barely 20 percent of metropolitan New York’s employment.

The Union of International Public Transport is hardly the type of organization that would be expected to make critical comments about public transit. But this organization, the international equivalent of the American Public Transportation Association (APTA) put it this way:

Actually, this is something of an overstatement. Transit plays an indispensable role in providing auto-competitive service to a few much focused areas of the nation. But outside these areas, the potential for transit to attract people out of cars is nearly non-existent.

This is illustrated by the record of metropolitan areas that have built new rail systems...

• Between 1990 and 2000, Dallas opened a commuter rail line and three branches of a light rail system. Yet, overall transit work trip ridership decreased (Table #2).
• Between 1990 and 2000, St. Louis opened a new light rail line. Yet, transit work trip ridership decreased (Table #3).
• Between 1970 and 2000, Washington opened approximately 100 miles of its Metro system. Yet, transit’s work trip market share dropped 29 percent.
• Transit’s work trip market share continues to trail pre-light rail levels in Portland, though minor gains were made in the 1990s. Street and highway traffic has grown faster than transit use since 1985, the year before the first of two light rail lines was opened. Traffic delays, as measured by the Travel Time Index have grown 31 percent, the second largest increase in the nation (Figure #6).

The problem is further illustrated by the case of Minneapolis-St. Paul, which is currently building a light rail line (the "Hiawatha Line") and seeks to build a commuter rail line (the "Northstar Line"). There, the Texas Transportation Institute estimates that it would take an addition of 84,000 annual transit one-way peak period riders just to stop to growth of traffic congestion. The two lines would add fewer than 9,000 one-way transit riders over a period of 20 years or more. During the 1990s, transit work trip use increased approximately 200 annually, a small fraction of what would be required to materially impact traffic congestion (Figure #7).

Urban rail systems are exceedingly expensive. Often the annual cost per commuter attracted from the automobile exceeds the recurring lease cost for a new automobile.

The modest returns from the nation’s new urban rail systems are evident when measured in terms of the cost per new passenger mile. From 1980 to 2000, incremental government expenditures (federal, state and local) on transit were $1.19 per incremental passenger mile, nearly 40 times that of streets and highways (Figure #8). And, while highway user fees and special imposts accounted for 75 percent or more of highway expenditures, transit user fees accounted for less than 30 percent of expenditures.

This spending imbalance is even more significant in some of the nation’s major urban areas. For example, through 2025, the Atlanta region will spend 55 percent of its transportation resources on public transit, while transit’s share of trips is expected to grow from only 2.6 percent to 3.4 percent (Figure #9).

The defining factor with respect to urban transport is that virtually all new travel demand is expected to be automobile related. Even Portland’s land use-transport planning agency, Metro, acknowledges this (Figure #10). The fundamental problem is that there is no transit system that can provide auto-competitive service to a significant share of destinations outside downtowns. This is true not only in the United States, and to a somewhat lesser degree even in Western Europe. Our research indicates that a transit system that provides auto-competitive service throughout the modern American urban area could cost as much as a metropolitan areas’ gross regional product.

But this does not mean that there is not a cost efficient role for strategies other than the single occupant automobile. The recent Census data indicates rays of hope. While overall transit work trip ridership was declining slightly, progress was made in much less costly modes, namely working at home (telecommuting) and car pooling. . Working at home increased 778,000 and car pooling increased 256,000, though registering a 3.4 percent market share loss (Figure #11).

• In eight metropolitan areas, telecommuting increased more than 25,000. In each of these areas, the increase was greater than that of transit (Table #4). Telecommuting increased more than transit in 40 of the nation’s 49 metropolitan areas over 1,000,000, and increased 19 times as much as transit (Table A-3). Little, if any government subsidy has been used to encourage telecommuting.
• In 11 metropolitan areas, car pooling increased more than 25,000. Again, in each such area, the increase was greater than that of transit (Table #5). In both Atlanta and Dallas-Fort Worth, which have made major urban rail investments, the 1990s increase in car pooling alone exceeded the total transit work trip ridership. Car pooling increased more than transit in 36 of the nation’s 49 metropolitan areas over 1,000,000 and increased 16 times as much as transit (Table A-3). Some of the metropolitan areas, such as Atlanta, Phoenix, Dallas-Fort Worth, Seattle and Houston employed aggressive high occupancy vehicle (HOV) lane programs during the 1990s. Regional HOV lane systems can provide opportunities for improved mobility to the entire area, not just to downtown. Further, they can and are being used by transit agencies to provide comparatively low cost express bus services. The concept can be expanded even further through use of HOT lanes (high occupancy toll lanes), which permit single occupant vehicles to use HOV lanes for a price, which can be used to expand the system even further.

One of the most promising developments has been the recognition by the Federal Transit Administration and some transit agencies of the much more cost effective options for rapid transit using buses. USDOT research has indicated that bus rapid transit can be five times as cost efficient per passenger mile.

The "Smart Growth" movement seeks to stop or control urban sprawl. Proponents claim that it will reduce traffic congestion, reduce air pollution and reduce costs. As a result, there are proposals to impose land use regulations for controlling urban sprawl as in the federal transportation program. It is fundamental that smart growth and containing sprawl require higher densities. Smart growth’s goals simply are unattainable without much higher densities.

US urban areas tend to be less densely populated than those in Western Europe and Japan (Figure #12). But, contrary to the popular view, sprawl is not an American phenomenon. Sprawl occurs wherever there is population growth and rising affluence, and European urban areas have seen their urban densities decline at an even greater rate than in the United States (Figure #13).

Figure 12

Figure 13

I do not favor sprawl. I favor allowing people to live and work where and how they like. And, there is no reason not to allow it. Even today, nearly 400 years after Jamestown, urbanization accounts for less only 2.6 percent of the nation’s land area.

The claims of the smart growth movement simply do not hold up.

National and international data clearly indicates that traffic congestion rises with population density. The higher density European and Asia urban areas, with their much higher public transit market shares also have much worse traffic (Figure #14). Research commissioned by the United States Department of Transportation indicates that at current US urban densities, vehicle miles rise more than 80 percent when population density is doubled. Now, admittedly, that means that per capita driving declines marginally, but it means that there are more miles in a defined area --- traffic congestion is worse.

Figure 14

More driving per square mile means that traffic slows down and that people must spend more time in their cars. Not surprisingly, journey to work travel times tend to be longer where population densities are higher --- whether in the United States or internationally.

And, as traffic volumes in a particular area increase, there is also an increase in stop and go driving. Slower speeds and stop and go driving mean greater production of air pollution. So, not surprisingly, air pollution production tends to be higher where densities are higher. And, it is well to consider the great progress that has been made in air pollution abatement in the United States. In the last 30 years, driving has increased substantially, while criteria air pollution production has decreased --- not just per capita --- but overall.

So, smart growth increases traffic congestion, travel times and air pollution.

Some months ago research was published showing that transportation costs are higher in more sprawling areas. This is to be expected. But what may be surprising is that overall household expenditures tend to be lower where densities are lower. The big factor in this equation is housing costs. Housing costs are less where densities are less, and they tend to be less to such a great degree that the transportation cost disadvantage is more than canceled.

But, the worst impact of all is social. Home ownership is lower where densities are higher. Thus, smart growth works to make home ownership more difficult for lower income households. Recent decades shows than minority home ownership, (African-American and Hispanic), is rising faster than that of non-Hispanic whites (Figure #15). At the same time, minority home ownership levels still remain well below that of non-Hispanic whites, which is why the Bush Administration has undertaken steps to more greatly expand minority home ownership.

Figure 15

By raising the price of housing, smart growth promotes social inequity. Smart growth rations land and development. It is a fundamental principle of economics that when valuable goods are rationed, their prices rise. When prices rise, it is the lower end of the income spectrum that is driven away from the market. The lower income spectrum has a disproportionate representation of minorities. As a result, smart growth reduces home ownership opportunities for lower income households, especially African-Americans and Hispanics. There is a raging debate between supporters and opponents of smart growth about the extent to which home ownership is reduced by smart growth. We often hear from smart growth supporters that they way to compensate for smart growths reduction of home ownership is to provide greater amounts of affordable housing. Such proposals are no more than empty platitudes in view of the fact that, by some reports, current public resources are sufficient to provide housing assistance to barely one third of eligible recipients. In fact, recent research by Matthew Kahn of Tufts University indicates that African-American home-ownership tends to be higher in more sprawling urban areas (Figure #16). Further, research by Edward L. Glaeser and Joseph Gyourko, published by Harvard University found that much of the difference in housing affordability around the nation can be attributed to land regulation. It is not surprising that Oregon, with the nation’s most comprehensive smart growth regulations, experienced by far the greatest increase in housing values between 1990 and 2000 (Figure #17).

Figure 16

Figure 17

Thus, smart growth is promises to produce a more traffic impacted urban area and one that is less economically inclusive. It would be a mistake for the federal government to encourage such measures through the transportation program.

Three conclusions and three recommendations are suggested by the current situation and recent trends in urban transport.

• The 1990 to 2000 Census data makes it clear that telecommuting and car pooling options can be far more cost effective than transit. The General Accounting Office should be asked to review the potential for more effectively using new transit investment funds to encourage additional telecommuting and car pooling.
• There is little potential for reducing traffic congestion or increasing transportation choice for all but a few through transit. There are no material successes, US or international. The nation should move toward a surface transportation program in which new investments are based upon a policy goal that all can identify with --- the reduction of actual travel times. Generally, new investments should be made based upon their cost per reduced hour of actual travel delay.
• Smart growth strategies tend to intensify the very problems they are purported to solve, especially by increasing traffic congestion. There should be no federal mandates with respect to land use or smart growth.

The Heritage Foundation is a public policy, research, and educational organization operating under Section 501(C)(3). It is privately supported, and receives no funds from any government at any level, nor does it perform any government or other contract work.

The Heritage Foundation is the most broadly supported think tank in the United States. During 2001, it had more than 200,000 individual, foundation, and corporate supporters representing every state in the U.S. Its 2001 contributions came from the following sources:

The top five corporate givers provided The Heritage Foundation with less than 3.5% of its 2001 income. The Heritage Foundation’s books are audited annually by the national accounting firm of Deloitte & Touche. A list of major donors is available from The Heritage Foundation upon request.

Members of The Heritage Foundation staff testify as individuals discussing their own independent research. The views expressed are their own, and do not reflect an institutional position for The Heritage Foundation or its board of trustees.

Home | Menu | Links | Info | Chairman's Page