Accessibility and the choice of network investments in the London Underground

Recently published

In 1863, the Metropolitan Railway of what came to be known as the London Underground successfully opened as the world’s first subway. Its high ridership spawned interest in additional links. Entrepreneurs secured funding and then proposed new lines to Parliament for approval, though only some were actually approved. While putative rail barons may have conducted some economic analysis, the final decision lay with Parliament, which did not have modern transportation, economic, or geographic analysis tools available. How good were the decisions that Parliament made in approving Underground lines? This paper explores the role accessibility played in the decision to approve or reject proposed early London Tube schemes. It finds that maximizing accessibility to population (highly correlated with revenue and ridership) per expenditure largely explains Parliamentary approvals and rejections.

HOT Lanes are on fire

The graph below shows the growth of High Occupancy/Toll lanes in the United States. From a slow start in the 1990s, growth has picked up in the past decade significantly, with a number of projects recently opened and many more underway. The graph shows projects that are open, under construction, or with contracts signed. The number of projects in planning stages are more still.

While still a small share of the total highway mileage in the United States (there are 4 million miles of streets and roads in the US, the National Highway System has 150,000 miles, the Interstates are 46,000 miles of that), it is growing rapidly, and should eventually absorb at a minimum the US HOV network, and perhaps eventually reach a parallel to the congested fraction of the Interstate Highway system.

Growth of HOT Lanes in the United States. Sources various. Thanks to David Ungemah and Mark Burris
Growth of HOT Lanes in the United States. Sources various. Special thanks to David Ungemah and Mark Burris.

We wrote in The Transportation Experience about HOT Lanes

In the late 1990s, High-Occupancy/Toll (HOT) lanes that would allow single occupant vehicles to use HOV lanes (after paying a toll which depends on the level of demand) began to be considered in many US metropolitan areas. HOT lanes were conceived of by Ward Elliott at Claremont McKenna College in the 1970s, and reinvented in the 1990s by Gordon Fielding and Dan Klein as part of a Reason Foundation study. HOT Lanes were introduced on I-15 in San Diego for instance, using the same facility that had been the testbed for the Automated Highway Systems experiment, and have generally been viewed as successful. The revenue generated after paying for operating the facility helps subsidize transit in the corridor.

Seeing the success of I-15, as well as a handful of other HOT lanes, Poole et al. (1999) calls for networks of HOT lanes (or HOT networks) in major US cities. They conclude the benefits of such a system (including congestion reduction for those not using the system) outweigh the costs.

While pricing every congested road using marginal cost pricing may increase system efficiency, providing differentiated services further enhances efficiency. We recognize that different types of freight have different priorities (overnight, two-day, and ground are choices for shipping); that same kind of differentiation applies to drivers. Different drivers have different values of time at different times of the day. Thus the ability to pay a premium and travel at a better level of service during peak times provides a service not currently available, a service enabled by bundling several ideas. The old idea is tolls themselves, whose original intent was simply to raise revenue. Electronic toll collection complements that; its original intent was simply to automate the collection of tolls at traditional tollbooths, reducing both traveler delay and agency operating costs. HOV lanes aimed at giving priority to vehicles carrying more passengers (vehicles which had a higher value of time, since two people are more than one).

The HOT networks also provide the ability to provide bus rapid transit (BRT) services in metropolitan areas, by providing the high speed limited access routes that give transit a travel time advantage over the automobiles not paying for the HOT lanes. In many situations, buses are more cost effective than fixed rail alternatives, but the lack of a fast right-of-way leads people to perceive rail as inherently faster than the bus. With BRT, that perception can be made to disappear. Several cities are now building-out HOT networks through a combination of new construction and conversion of HOV lanes.

Detours and induced demand

Bill Lindeke writes in MinnPost about Detours and induced demand

Probably the craziest detour in Minnesota history was the impromptu rerouting following the Interstate 35W bridge collapse. Needless to say, the 2007 bridge tragedy was completely unexpected, and forced state agencies to reroute 150,000 cars per day into other parts of the Twin Cities’ freeway system.

As it turned out, however, the detours worked surprisingly well. A University of Minnesota study determined that average commute times increased only marginally, less than half a minute on average.

Those unexpected results stemmed from two counterintuitive facets of detour construction. The first is the phenomenon of “induced demand,” or the dynamic relationship between road capacity and people’s willingness to drive. Simply put, the more available capacity on the road, the more likely that people will choose to drive; conversely, the less capacity, the less likely the car trip. That means that expanding roads might create more traffic, while reducing travel lanes might decrease traffic at the margins.

According to David Levinson, the engineering professor at the University of Minnesota who worked on the study, the I-35W bridge was a great example of “reduced demand” (the opposite of “induced demand”) where “about 1/3 of the trips across the river ‘disappeared,’ and were either foregone or went to different destinations.”

The second conclusion is that the Twin Cities’ road system actually has a well-connected road network, relatively speaking. During a detour, most Twin Cities drivers are able to find alternative routes.

Green Line and related bus network improvements increase access to jobs

A new analysis from the University of Minnesota’s Accessibility Observatory has quantified how much the Green Line light-rail transit route and related changes to the regional bus network changed residents’ ability to reach jobs by transit. The study measured access to jobs in Minneapolis and St. Paul before and after the Green Line began service on June 14, 2014.

Change in number of jobs within 30 minutes by transit

“Changes were greatest in St. Paul, where most of the Green Line’s stations are located,” says Andrew Owen, director of the Observatory. “A year after the opening of the Green Line, workers in St. Paul can, on average, reach over 2,000 more jobs within 30 minutes by transit than they could previously—a 5.3 percent increase. In locations near Green Line stations and connecting transit routes, accessibility often increased by over 50 percent, and in a few locations more than doubled.”

In Minneapolis, changes were minor, and in most cases were due to service or schedule changes unrelated to the Green Line project, he says.

For the analysis, researchers measured the number of jobs that can be reached by transit within 30 minutes of travel between 7 and 9 a.m. from each census block in St. Paul and Minneapolis. They evaluated accessibility for three scenarios:

  • The 2014 scenario uses actual transit schedules as of April 2014, prior to the start of Green Line service.
  • The 2015 scenario uses actual transit schedules as of April 2015, which reflects the new Green Line LRT service as well as changes to the regional bus network.
  • A hybrid scenario uses schedules from 2014, modified to include current Green Line service and changes to parallel bus routes, but no changes to other bus routes.

“By comparing the hybrid scenario with both 2014 and 2015, it’s possible to estimate the accessibility impacts of the Green Line itself separately from the impacts of changes to connecting routes in the regional bus network,” Owen explains.

The results suggest that had the Green Line been implemented without any supporting changes to the regional bus network, accessibility benefits would have been limited to areas near the new rail stations. “The relatively low residential density of the area means that a single transit line, whether bus or rail, can reach only a small minority of residents,” Owen says. “But a well-designed network of routes can provide access to a much larger area and can help distribute the benefits of new investments like the Green Line throughout the community.”

Owen notes that the analysis did not look at other effects of the Green Line, such as reliability and passenger comfort. “It’s likely that the Green Line provides far more reliable travel times than did the bus routes it replaced, which operated in mixed traffic,” he says. Additionally, the analysis does not reflect potential travel time improvements to jobs that could already be reached in less than 30 minutes.

The analysis was sponsored by CTS and the Department of Civil, Environmental, and Geo- Engineering.

Read the final report

Originally posted at the Accessibility Observatory

Beneficiaries, Not Taxpayers, Should Fund Modern Streetcars

I wrote a piece for ULI: Beneficiaries, Not Taxpayers, Should Fund Modern Streetcars. An excerpt below:

800levenson

This article is one of a three-part cities on the role of street cars. The opinions expressed here are those of the authors and not the Urban Land Institute or it’s membership as a whole.

Once upon a time (1888 to be precise), the United States and the world launched a huge building boom for urban streetcars. Companies like Twin City Rapid Transit laid miles of track in fast-growing cities, extending well past the built areas to serve greenfield sites for emerging suburbs waiting to be platted and built. They did this because the streetcar promoters benefited directly from the land sales. The availability of a new, fast transit system connecting to downtown made houses much more valuable. The fares from the new passengers covered the operating costs of the system.

Networks continued to grow until the 1920s and 1930s, when the bloom came off the boom. The new motor car served the prospective suburbanite just a bit better than the sluggish streetcar. By 1950, the streetcars were upward of 60 years old and needed a major infusion of capital to be maintained. Instead, they were abandoned en masse across the United States for buses in a process that in the transportation field has been termed “bustitution.”

Continued at ULI

More on StreetCars and America: Freemark: Transportation First, then Economic Development | Klein: Streetcars and the American Commitment to Rail

MnPASS Modeling and Pricing Algorithm Enhancement

Recent report:

Hourdos, John, Michael Janson, David Levinson, Gordon Parikh (2015) MnPASS Modeling and Pricing Algorithm Enhancement. Minnesota Department of Transportation MN/RC 2015-22.

While High Occupancy Vehicle (HOV) lanes have been used for decades as a strategy for mitigating congestion, research has shown that they are not always effective. A 2001 study of the I-394 and I-35W HOV lanes in Minnesota found that the HOV lanes were on average underutilized, moving fewer people than the General-Purpose Lanes (GPL) even with the increased number of passengers per vehicle. To address the issue of underuse, in 2003 the Minnesota Legislature authorized the conversion of the I-394 HOV lanes into High-Occupancy Toll (HOT) lanes, named the MnPASS Express Lanes. The MnPASS lanes operate using a fully dynamic pricing schedule, where pricing is dictated by the level of congestion in the HOT lane. To better understand the nature of HOT lanes and the decisions of their users, this study explored the possibilities for a microscopic traffic simulation-based model of HOT lanes. Based on a series of field studies where the price of the toll was changed while observing changes in demand in the HOT lane, models describing the lane choice behavior of MnPASS users were developed and calibrated. These models interfaced with the traffic simulation software Aimsun through a number of extension modules and tested on the two MnPASS corridors of I-394 and I35W corridors in the west and south suburbs of Minneapolis, Minnesota. The integrated HOT simulation tool was also used to develop and test a number of alternative pricing strategies including a more efficient version of the current strategy.

With driving down, there’s a growing backlash against unnecessary highways | Vox.com

Brad Plumer writes about peak travel, overly optimistic forecasts, and the policy consequences at vox.com in: With driving down, there’s a growing backlash against unnecessary highways. He cites us in the closing paragraph:

Another key point was that states (and the federal government) should probably spend less money on building new roads and more money on maintaining existing roads. Economists have argued that more states should probably be doing this anyway — this Brookings Institution paper by Matthew Kahn and David Levinson makes the most comprehensive case. But the argument gets even stronger if vehicle travel is going to grow more slowly or plateau in the future.

What’s clear is that it doesn’t make much sense for states to keep planning around the idea that driving will grow at 20th century rates forever. Those predictions have been utterly wrong for nearly a decade now — and sticking with them could mean wasting billions on unnecessary roads and highways.

 

Child savings accounts should replace tax credits and deductions

Current federal policy has a Child Tax Credit. This means if you pay taxes, you pay less taxes by $1000 per child (this does phase out at high incomes). If you have a low income and have no taxes due, or less than the credit, you get $0<Credit<$1000. Some of this is a refundable additional tax credit, depending on your income. However, there is no reason to expect someone with a zero income has lower costs per child than someone with a higher income.

In addition to the Child Tax Credit, there is a Child and Dependent Care Credit. This can be worth up to $900 depending on income.

The amount you can deduct from your income per dependent is $3,950. At a tax rate of 20% this is $790.

I therefore suggest we replace all of these with a child savings account of the same value. Each fortnight, a deposit would be made to an account at a federally insured bank or credit union set up on behalf of the child, and managed by his or her guardian. This deposit, on the order $100 every two weeks per child (about $2600/year), is certainly less than the cost of housing, feeding, clothing, and otherwise maintaining a child, but can certainly contribute. If left untouched, at 18 a child could have accumulated, excluding interest, $46,800. That is a nice nest egg to start with.

  • Compared with anything dealing with taxes, which is visible only once per year, this continuous deposit has more salience, since account holders will see the money accrue.
  • It would be continuous rather than a lump sum every April, so would have more positive effects on the economy.
  • It would be a natural mechanism by which an economic stimulus program could be achieved, since amounts could be temporarily increased in the short-term as needed.
  • It would be additive if the account holder didn’t withdraw the money, so by the principle of the default  being savings, would increase personal savings.
  • It increases the relationship between people, especially low income, and the formal banking system, rather than the check-cashing and pawn system that now serves as financing in too many cases.
  • It provides an account to which others could add to help save for the child.
  • It simplifies the tax code.

Withdrawals by guardians would of course need to be for the benefit of the child, but given that (a) money is fungible, and (b) the cost of raising a child exceeds the amount contributed by the government, that should not be hard to establish and in practice would be un-enforced except in the case of other child abuse or financial crimes.

While it might be difficult, and scoring this would be a government expenditure rather than a tax expenditure might have budgetary gimmicks associated with it, this seems like it would be advantageous all around.

 

  1. The UK had a program called the Child Trust Fund, which deposited up to £500 at birth (for low income families) and again at age 7, but this was eliminated in 2010.

Mapping Metropolitan Chicago’s Accessibility

If you like accessibility, or you are interested in Chicago, or better still both, check out urbanaccessibility.com which has web-based interactive multi-modal, multi-purpose cumulative opportunity accessibility maps as well as mobility or isochrone maps.

The report describing the project is here: Mapping Metropolitan Chicago’s Accessibility by Nebiyou Tilahun,  Shi Yin, Moyin Li, Yaye Keita.

Chicago Accessibility by Car, from urbanaccessibility.com
Chicago Accessibility by Car, from urbanaccessibility.com