A University of Minnesota researcher is using travel data to rank the best areas in the state to live based on access to vital destinations.
The University’s Accessibility Observatory is evaluating transportation destinations, such as jobs, schools and hospitals in the state in order to measure accessibility.
The data could shape how entities like the Minnesota Department of Transportation plan future transit projects.
Andrew Owen, lead researcher and director of the observatory in the University’s Department of Civil, Environmental and Geo-Engineering, said the research identifies where jobs are concentrated.
“Focusing on accessibility gives a way to look at how well we are achieving the goals of transportation systems,” he said.
The program uses bus, rail, car and pedestrian travel times combined with census data to measure the number of jobs that can be reached within 30 minutes of a person’s home, Owen said. The data can be adjusted to give information about any type of destination from anywhere in the state.
…
David Levinson, a professor in the Department of Civil, Environmental and Geo-Engineering, said this information can also explain why people choose a certain mode of transportation.
“In places with higher transit accessibility, people are more likely to use [public] transit,” he said.
Levinson said the research also focuses on how frequently public transportation is available at a certain location.
“Transit accessibility varies by time of day,” Levinson said. “If the bus just left and won’t be back for another 30 minutes, you can’t reach very many places.”
Something unprecedented has happened to Americans’ travel patterns. Even before the recent recession, total distance traveled per person had started to decline, and the rate of total vehicle travel had begun to steadily decrease as well.
In a new five-part series of research reports sponsored by the Minnesota Department of Transportation and the Metropolitan Council, U of M researchers are delving into a set of rich data encompassing more than four decades of travel behavior surveys to enable the region’s transportation planners to better understand how its residents make decisions about whether, when, where, and why to travel.
In the first study, researchers examined how changes in the accessibility of destinations—such as jobs, shopping, and leisure activities—have changed travel behavior in the past 20 years.
“We started with a detailed analysis of travel surveys conducted by the Metropolitan Council in 1990, 2000, and 2010,” says David Levinson, the study’s principal investigator and RP Braun/CTS Chair in the Department of Civil, Environmental, and Geo- Engineering. “We found that people are spending slightly less time in motion and more time at home. We also found that accessibility is a significant factor in determining not only travel behavior but overall time budgeting in general. In short, each person has to decide how they will use the time allotted to them each day, and many of those decisions are directly related to the transportation and land-use systems in place.”
A deeper look into the data sheds additional light on the relationship between accessibility and travel behavior. For example, trip durations for workers have gone up for all activities between 1990 and 2010. More noticeably, distances for trips have increased markedly: workers take jobs farther from their homes and shop farther from their homes. Travel speeds also increased for the average worker, due to more travel on faster suburban roadways that carry a larger share of all travel. In contrast, for non-workers, trip durations and overall travel time have gone down.
“Interestingly, although time, distance, and speed per trip has generally risen for workers, the number of those trips is declining,” Levinson says. “As a result, overall, fewer miles are being traveled and less time is being allocated to travel.”
Total time spent shopping also decreased for workers and for males, likely caused in part by an increase in online commerce. “The Internet has provided electronic accessibility, much as the transportation network has in the material world,” Levinson explains. “It helps to facilitate commerce, communication, education, and leisure. This may lead to a decreased need for people to travel, and account for more time spent at home.”
Jonathan Ehrlich, planning analyst with the Metropolitan Council, says the research “helps us get more value from our travel surveys and will aid in understanding how travel is changing, and what the risks are in the assumptions and models we use for planning and forecasting.”
The findings will prove useful not just for Twin Cities transportation planners but for planners and engineers worldwide. “Our models can be easily adapted to data from other cities or for other activities besides work,” Levinson says. “This creates an approach that can be used to gauge the impact of a transportation project from an accessibility standpoint and determine how that project will translate into time allocation.”
Other parts of the study will look at changes in telecommuting behavior over time, the effect of transit quality of service on people’s activity choices and time allocation, changes in travel behavior by age cohort, and analysis of bicycling and walking in light of land-use and transportation system changes. The Catalyst will feature coverage of these projects as they are completed.
“Excellent accessibility ranking by Dr. David Levinson (the Transportationist) and the Nexus Research Group at University of Minnesota:
California has the #1 and #2 most accessible cities, and they provide an interesting contrast in two ways to get accessibility. Accessibility can come from density (everything is close together so it doesn’t take long to go from one place to another) or mobility (everything is far apart but there are huge highways so you can traverse long distances).
I’m excited to see academics using visual media to put across point about public policy.”
David Levinson, a transportation economist at the University of Minnesota, has emphasized that one of the key issues in infrastructure investment is improving accessibility, or the ease of reaching valued destinations. One way to improve accessibility is make it easier to traverse long distances, so you can reach a larger number of jobs and consumption opportunities, etc., in a given amount of travel time from home. Another way to improve accessibility is to bunch up jobs and consumption opportunities and homes, i.e., by increasing density. Levinson finds that while accessibility has deteriorated relative to 1990, it has improved relative to 2000. My sense is that the best way to increase accessibility is to focus on implementing peak road-user fees and using the resulting revenue stream to carefully add capacity at bottlenecks, and also to ease local land use regulations that have proven a barrier to increased density in high-productivity regions. These strategies ought to be pursued in tandem. One crude way of putting this is that while we tend to fixate on the “hardware” layer of infrastructure, we should devote more attention to the “software” layer, i.e., the systems governing the allocation of infrastructure resources. Focusing on accessibility rather than infrastructure spending levels as such will get us much closer to tackling the frustrations that plague commuters.
New Study Ranks Access to Jobs via Auto Commuting
Transportation is not an end in itself; it’s a means to other ends, such as getting to and from work. Taking this point to heart, a growing number of researchers in recent years have promoted the concept of “access” as being more important than speed or travel time, per se. One of the leaders in this field, David Levinson of the University of Minnesota, defines accessibility as “the number of destinations reachable within a given travel time” by a particular mode of transportation. He is the author of a new study called “Access Across America,” released last month by U of M’s Center for Transportation Studies.
In this study, Levinson estimated the accessibility to jobs by car for the 51 largest U.S. metro areas. His data are for 1990, 2000, and 2010, so in addition to providing a snapshot of conditions as of 2010, the data also allow him to document trends over the past two decades. The results may surprise many of those concerned about traffic congestion in the largest metro areas, because Levinson finds that the 10 metro areas that provide the greatest accessibility to jobs via auto commuting are, in order: Los Angeles, San Francisco, New York, Chicago, Minneapolis, San Jose, Washington, Dallas, Boston, and Houston. And over the past two decades, the places with the largest increases in accessibility by car are Las Vegas, Jacksonville, Austin, Orlando, and Phoenix. Those with the largest decreases are Cleveland, Detroit, Honolulu, and Los Angeles.
What accounts for these findings? Although Levinson doesn’t really get into the details, I think one of the most important factors is the ongoing suburbanization of jobs. Remember, Levinson’s data are for entire metro areas, and there has been a huge dispersion of jobs throughout these metros over the past 50 years. A good summary of the data was provided last month by Wendell Cox in “Job Dispersion in Major US Metropolitan Areas, 1960-2010.” (www.newgeography.com/content/003663-job-dispersion-major-us-metropolitan-areas-1960-2010) For example, in 1960 54% of employment in 35 major metro areas was in the historical core municipalities—but by 2010, that figure had dropped to 30%, with 70% in suburban and exurban areas. The suburbanization of jobs has made huge numbers of workplaces more accessible by car than before, leading to shorter average work-trip travel times than in Canada or Europe.
Levinson’s data show that in 31of the 51 metro areas, all the jobs can be reached by car in 30 minutes or less; upping the limit to 40 minutes brings the total to 39 of the 51, and at 60 minutes, almost everyone can reach nearly every job in every one of the 51 metro areas. That’s pretty outstanding performance by the highway system, despite the existence of serious congestion.
It’s instructive to contrast Levinson’s auto accessibility figures with the findings of a Brookings Institution study from 2011 on accessibility to jobs via transit (“Missed Opportunity: Transit and Jobs in Metropolitan America”). Using a 45-minute transit commute time, that study found that only 7% of jobs could be reached, in the 100 largest metro areas. Even at 60 minutes, transit could get people to only 13% of the area’s jobs. To reach 30% of the jobs, you need an average travel time of 90 minutes, which is more than three times the duration of the average U.S. auto commute.
Knowing this, some advocates of Smart Growth therefore disparage the suburbanization of employment as “jobs sprawl” and seek to promote public policies that would reverse it, so that transit could do a better job. But that confuses means with ends. If the purpose of an urban transportation system is accessibility, we should work to make the system serve that goal, not engage in a utopian quest to massively reshape the urban landscape. And, as I have written in previous issues of this newsletter, the implication for transit is to develop more flexible systems that can link more people cost-effectively to jobs. That argues for grid-based bus systems as opposed to radial bus and rail systems focused on what used to be the “central business district.”
A study recently released by the University of Minnesota presents an interesting alternative to the TTI’s metrics. UMN Transportation Engineering Professor David Levinson recently analyzed metropolitan commuting according to a very different criterion: accessibility, or “the ease of reaching desired destinations.”
Levinson attempted to improve on the TTI report by tracking the time it takes for people in the 51 largest U.S. metro areas to reach jobs. His findings stand in stark contrast to the TTI’s report. Large metros like Los Angeles, San Francisco, New York and Chicago offered the greatest number of jobs within a 10-minute car commute, Levinson found.
While TTI’s methodology penalizes cities for locating homes and businesses close together, because that increases congestion, in Levinson’s analysis, higher concentrations of destinations are rewarded for helping to reduce travel times.
“There are two ways for cities to improve accessibility—by making transportation faster and more direct or increasing the density of activities, such as locating jobs closer together and closer to workers,” Levinson writes.
“Accessibility is not a new idea,” he adds. But his is the first study that uses it to systematically attempts to measure how different metro areas compare. The report focuses only on auto access, but the same concepts could be applied to walking, biking, or transit access, he says.
To measure accessibility, Levinson factored in average job density, the average speed of car traffic in the transportation network (from the TTI analysis), and the circuity of trips (how indirect they are). The analysis also looked at the number of destinations within 10-, 20-, 30-, and 40-minute “donuts” around the city.
Levinson found that his measure of “accessibility” is linked to a number of positive economic indicators. For example, he found that home prices in a metro area increase 0.23 percent with every 1 percent increase in accessibility. He also found that doubling accessibility leads to a 6.5 percent increase in real average wages.
There are environmental and quality-of-life connections, as well. Levinson found that a 1 percent increase in accessibility is linked to a 0.06 percent reduction in the share of commuters who drive. He also found that accessibility tends to be linked to shorter overall commute times. A 1 percent increase in “accessibility,” he found, is correlated to a 90-second reduction in average commute time each way.
All of this suggests that prioritizing “accessibility” in transportation investment — rather than alleviating congestion – might be more economically beneficial for metro regions.
Levinson found that accessibility, or the ease of reaching important destinations, has declined in the United States over the past two decades. Image: University of Minnesota
Levinson also measured how accessibility has changed in metro areas over time, finding that it has worsened in American regions overall, both since 1990 and since 2000.
I was interviewed a few weeks ago by Dale Connelly for KFAI Community Radio. The edited interview (aired April 22, 2013) is below as an MP3 (5:51)
Every so often we get a discouraging report on traffic congestion in metropolitan areas.
The latest rating from Texas A & M’s Transportation Institute gave Washington DC the worst rating for congestion, followed by Los Angeles, San Francisco, New York and Boston. No big surprise there.
But one University of Minnesota professor says counting stationary cars is only part of the story. David Levinson is a professor of civil engineering and author of the Access Across America Study. He told KFAI’s Dale Connelly there’s more to consider when looking at the problem of traffic congestion.
David Levinson holds the Richard P. Braun/Center for Transportation Studies Chair in Transportation at the University of Minnesota. His study, Access Across America, says some of the cities regularly identified as most congested actually have transportation networks that provide good access to jobs. You can see the study online at http://cts.umn.edu.
CTS Catalyst summarizes our Access Across America report:
Moving beyond mobility: measuring accessibility in U.S. cities
Every year, Americans face a steady stream of discouraging news. We’re spending more time stuck in traffic. Congestion in our metro areas is on the rise. Yet these reports focus almost exclusively on traffic mobility—how quickly travelers can move between any two points via automobile or transit. But according to a new University of Minnesota study, there’s much more to the story.
“Focusing solely on mobility and traffic delay doesn’t provide a complete picture of how the traffic system is functioning,” says Professor David Levinson, the R.P. Braun/CTS Chair in Transportation Engineering. “Travelers in many of these cities have the ability to reach their desired destinations, such as shopping, jobs, and recreation, in a reasonable amount of time despite congestion and slower travel because these cities have greater density of activities. In short, these travelers enjoy better access to destinations.”
A new study, Access Across America, goes beyond congestion rankings to focus on accessibility: a measure that examines both land use and the transportation system. The study is the first systematic comparison of trends in accessibility to jobs by car within the U.S. By comparing accessibility to jobs by automobile during the morning peak period for 51 metropolitan areas, the study shows which cities are performing well in terms of accessibility and which have seen the greatest change.
To generate the rankings for this study, Levinson created a weighted average of accessibility, giving a higher weight to closer jobs. Jobs reachable within 10 minutes are weighted most heavily, and jobs are given decreasing weight as travel time increases up to 60 minutes. Based on this measure, the 10 metro areas that provide the greatest average accessibility to jobs are Los Angeles, San Francisco, New York, Chicago, Minneapolis, San Jose, Washington, Dallas, Boston, and Houston.
“It can be surprising to see that some of the cities often ranked as the most congested also have the highest levels of job accessibility,” Levinson says. “This is due to the density of jobs those urban areas offer.”
Levinson also found that job accessibility has changed over time. In the past two decades, Las Vegas, Jacksonville, Austin, Orlando, and Phoenix have seen the largest percentage gains in job accessibility while Cleveland, Detroit, Honolulu, and Los Angeles have seen the largest percentage drops.
According to Levinson, this research offers an important takeaway for metro areas interested in increasing accessibility. “There are two ways for cities to improve accessibility—by making transportation faster and more direct or by increasing the density of activities, such as locating jobs closer together and closer to workers. While neither of these things can easily be shifted overnight, they can make a significant impact over the long term.”
This report extends the Access to Destinations study, an interdisciplinary research and outreach effort coordinated by CTS with support from multiple sponsors.
“… David Levinson, a transportation economist at the University of Minnesota, has emphasized that one of the key issues in infrastructure investment is improving accessibility, or the ease of reaching valued destinations. One way to improve accessibility is make it easier to traverse long distances, so you can reach a larger number of jobs and consumption opportunities, etc., in a given amount of travel time from home. Another way to improve accessibility is to bunch up jobs and consumption opportunities and homes, i.e., by increasing density. Levinson finds that while accessibility has deteriorated relative to 1990, it has improved relative to 2000. My sense is that the best way to increase accessibility is to focus on implementing peak road-user fees and using the resulting revenue stream to carefully add capacity at bottlenecks, and also to ease local land use regulations that have proven a barrier to increased density in high-productivity regions. These strategies ought to be pursued in tandem. One crude way of putting this is that while we tend to fixate on the ‘hardware’ layer of infrastructure, we should devote more attention to the ‘software’ layer, i.e., the systems governing the allocation of infrastructure resources. Focusing on accessibility rather than infrastructure spending levels as such will get us much closer to tackling the frustrations that plague commuters. “
Today we released a new report describing the level and change of accessibility across the 51 metropolitan areas in the United States: Access Across America. The website includes an interactive map, the report can be downloaded directly here.
Executive Summary
Accessibility is the ease of reaching valued destinations. It can be measured across different times of day (accessibility in the morning rush might be lower than the less-congested midday period). It can be measured for each mode (accessibility by walking is usually lower than accessibility by transit, which is usually lower than accessibility by car). There are a variety of ways to define accessibility, but the number of destinations reachable within a given travel time is the most comprehensible and transparent as well as the most directly comparable across cities. This report focuses on accessibility to jobs by car. Jobs are the most significant non-home destination, but it is also possible to measure accessibility to other types of destinations. The automobile remains the most widely used mode for commuting trips in the United States.
This study estimates the accessibility in the 51 largest metropolitan areas in the United States for 2010, and compares results with 2000 and 1990.
Rankings are determined by a weighted average of accessibility, giving a higher weight to closer jobs. Jobs reachable within ten minutes are weighted most heavily, and jobs are given decreasing weight as travel time increases up to 60 minutes. Based on this measure, the ten metro areas that provide the greatest average accessibility to jobs are Los Angeles, San Francisco, New York, Chicago, Minneapolis, San Jose, Washington, Dallas, Boston, and Houston.
Job accessibility has changed over time. In the past two decades, Las Vegas, Jacksonville, Austin, Orlando and Phoenix have seen the largest percentage gains in job accessibility while Cleveland, Detroit, Honolulu and Los Angeles have seen the largest percentage drops. Key findings
In 2010, the average American living in the top-51 metro areas could reach slightly fewer jobs by automobile than in 1990, but more jobs than in 2000.
Automobile speeds were faster in 2010 than in 2000 (and about where they were in 1990).
Overall job losses in these 51 areas have limited accessibility gains associated with faster networks.
The average American city is slightly more circuitous in 2010 than in 1990 because roads in newer areas (suburban growth) are not as well connected as those in older areas of the metropolitan region.
The overall most accessible metropolitan areas in 2010 were (in order): Los Angeles, San Francisco, New York, Chicago, Minneapolis, San Jose, Washington, Boston, Dallas, and Houston.
There have been significant changes among accessibility leaders since 1990, when New York, Philadelphia, Chicago, Miami, Los Angeles, Boston, Cleveland, Detroit, Washington, and Dallas made up the top 10.
People living in many smaller metropolitan areas can reach as many jobs by car as people living in much larger areas within both the 10- and 20-minute time frames. For instance New Orleans, Salt Lake City, and Jacksonville are all among the top 10 for number of jobs that can be reached within 10 minutes. Jacksonville, Milwaukee, and Las Vegas are among the top 10 for number of jobs that can be reached within 20 minutes.
There are two ways for cities to improve accessibility—by making transportation faster and more direct or increasing the density of activities, such as locating jobs closer together and closer to workers. While neither of these things can easily be shifted overnight, they can make a significant impact over the long term.
Andrew Owen will represent the Nexus group at the FOSS4G (Free and Open Source Software
for Geospatial – North America 2013) conference happening in Minneapolis, May 22-24.
FOSS Experiences in Transportation and Land Use Research
Andrew Owen, University of Minnesota Nexus Research Group

The Nexus Research Group at the University of Minnesota focuses on understanding the intersections of transportation and land use. In this presentation, we will examine case studies of how open source geospatial software has fit into specific research projects. We will discuss why and how open source software was chosen, how it strengthened our research, what areas we see as most important for development, and offer suggestions for increasing the use of open source geospatial software in transportation and land use research. Over the past two years, we have begun incorporating open source geospatial data and analysis tools into a research workflow that had been dominated by commercial packages. Most significantly, we implemented an instance of OpenTripPlanner Analyst for calculation of transit travel time matrices, and deployed QGIS and PostGIS for data manipulation and analysis. The project achieved a completely open research workflow, though this brought both benefits and challenges. Strengths of open source software in this research context include cutting edge transit analysis tools, efficient parallel processing of large data sets, and default creation of open data formats. We hope that our experience will encourage research users to adopt open source geospatial research tools, and inspire developers to target enhancements that can specifically benefit research users.
“As with stores, houses too are getting larger over the long run. New suburban homes have more space to store goods in-house. While urban residents export storage to common stores, suburban residents more likely to have second freezers, have more space to store stuff.”
Just-in-time consumption: Does the `pint of milk test’ hold water?
Just-in-time production revolutionized manufacturing, enabling both a reduction in inventories as supplies arrive only shortly before needed, and an improvement in quality as poorly made inputs are no longer stored for long periods of time, but can be quickly identified and feedback provided to the supplier. The widespread adoption of the just-in-time process is itself the product of the logistics revolution, information and communications technologies, containerization in shipping, and the modern freeway system. It has seen a concomitant change in the retail sector, which has brought about fewer and larger stores at a greater distance from the end consumer.The notion of “just-in-time consumption”, (acquisition of a good by the end consumer shortly before its use, rather than being acquired and stored for future use) though seemingly a natural mirror to the more widely used “just-in-time production” has not received the same attention. The phrase itself, only generates 212K hits in Google, (significantly higher than 2007, and of which the original version of this post rates #2) of which only a few are on-point, in comparison to over 2.23M for “just-in-time production”.Yet many goods and services are already consumed in a just-in-time manner. Most notable is energy, which is delivered on-demand to users, who no longer store coal at home for the furnace, but instead buy natural gas or electricity as needed. (The slowly vanishing home heating oil remains an exception). Other services that are provided on-demand or just-in-time include water and sewer, communications (internet, telephony, and television). What is in common about these disparate technologies is their network nature, the large infrastructure required to enable using the flows on-demand. While sewer is a continuous service for most people (those who do not have septic tanks), garbage is typically only collected periodically (e.g. once a week), and recycling less so (e.g. fortnightly).Other goods once saw regular to-the-house delivery, especially in suburban areas. Foxell [2005] writes of goods and services found in Metro-land, the idyllic north London suburbs built by the Metropolitan railway in the early twentieth century:
“This service economy is illustrated by the variety of tradesmen that called at our home: the milkman twice a day, with a horse-drawn cart; the baker once a day, with a large upright barrow on two wheels, the handles of which lifted him off the ground when going down hill; the postman thrice; the butcher’s boy by bicycle twice a week; and the grocer twice a week. Others like the coalman or the Gas, Light & Coke Co. in their steam-powered Sentinel lorry also made regular deliveries. Over a longer period, visits could be expected from the men from the Prudential [insurance], Hoover [vacuum cleaners], Singer [sewing machines] and the like – all using a service call to take the opportunity to sell new products. There was something reassuring about seeing such familiar faces and catching up with the latest gossip. In addition there were the itinerant callers such as Walls Ice Cream man on his tricycle as well as the French onion sellers, gypsies with pegs and posies, rag and bone men, tinkers [metalsmiths] and the knife-sharpeners with their pedal-driving grinding wheels.”
Today, the vast majority of those goods are not acquired at home but in stores or online. Delivery services have replaced salesmen, as the two functions (delivery and sales) are now distinct and specialized. Today’s visitors might be the post office, FedEx or UPS, and the pizza delivery boy.
Just-in-time does not require delivery to the residence, it can involve ubiquity in the placement of stores, so that they are near the end consumer. Traditionally the retail store was just that, a place where a community could store goods, and individuals could take or buy them as needed. A new model of temporary lockers may emerge to fill the gap.
Many planners would like to make the ability to acquire goods just-in-time without the use of a vehicle a normative planning standard. For instance, a report, Beyond 2010: A Holistic Approach to Road Safety in Great Britain calls for the “pint of milk test”, for all new developments, whereby a resident can get to a shop to buy a pint of milk in 10 minutes or less without getting in their car [Parliamentary Advisory Council on Transport Safety, 2007]. The idea of 10 minutes comes from people’s willingness to walk, people are less willing to walk longer distances than shorter, and 10 minutes (or one-half mile (0.8 km)) seems to be a threshold over which walking tolerance seems to drop. This distance was derived from several empirical studies, including Pushkarev and Zupan [n.d.], who showed the median walk by travelers accessing the New York subway was 0.35 mi (0.57 km), while the median walk to access commuter rail stations in suburban New Jersey was 0.5 to 0.6 mi (0.8 – 1.0 km). Results from the 1983/84 National Personal Transportation Survey reported by Unterman [1990] found shorter distances: 70 percent of Americans will walk 500 feet (0.15 km) for normal daily trips, 40 percent walk 1,000 feet (0.31 km), and only 10 percent walk a half-mile (0.8 km).The pint of milk refers to a standard quantity of a highly perishable and frequently consumed good. The objective of avoiding car use is obvious for a group advocating road safety. The pint of milk test has received some currency in England, being noted by several studies in recent years [Bennett and Morris, 2006, Marsh, 2004]. This is a particular issue in a crowded city like London, where auto ownership is lower than suburban areas, roads are more crowded, and parking more difficult even for those with a car.
The trends in retailing have been clear in the United States for a long time. Stores are over the long term getting larger and gaining larger market areas [Yim, 1990]. Small stores serving local areas have been losing market share to larger stores which bring with them economies of scale. Efforts to reverse this trend have met with resistance from retailers, consumers, and neighbors [Nelson and Niles, 1999].
Illustrating this trend, the Food Marketing Institute reports in 2011 there were 36,569 supermarkets (with $2 million in sales or more, noting the median annual sales for a supermarket was $17 million, and average size was 46,000 sq. ft. (slightly down form a 2005 peak of 48,058, indicating an increasing number of smaller markets in recent years, but nowhere near retracing the long march upward). The average number of trips per week consumers make to the supermarket was up to 2.2. (from 1.9 in 2006).
In 1930, The Great Atlantic and Pacific Tea Company, at the time the leading US supermarket, alone had 16,000 stores with a combined revenue of $1 billion (or per store revenue of $62,500 in 1930 dollars, estimated to be $754,000 today) [The Rise and Decline of the Great Atlantic and Pacific Tea company, n.d.]
Handy [1993] claims “the automobile instigated a collapse of the retail hierarchy by encouraging the growth of community and regional centers at the expense of local shops and the central business district. The result has been a cycle of dependence, in which suburban communities are designed for the automobile leaving residents little choice but to drive.”
As with stores, houses too are getting larger over the long run. New suburban homes have more space to store goods in-house. While urban residents export storage to common stores, suburban residents more likely to have second freezers, have more space to store stuff.
While the number of freezers per household in the United States is declining as second freezers are being retired and not replaced, the number of refrigerators is increasing slightly, due to households obtaining second refrigerators. [Wenzel et al., 1997]. While no immediate inference can be made about this, other trends are also at work. Total refrigerated and frozen space has not been computed, though the average size of a house’s primary refrigerator or freezer is likely increasing. Food may last longer in refrigerators than it used to due to the addition of preservatives (though the trend of increased consumption of organic foods may reverse this). Further globalization may mean that fewer goods are seasonal and need to be accumulated prior to their being out-of-season.
Persson and Bratt [2001] note that e-shopping may induce the installation of a second set of fridge/freezers per household to receive delivered goods. This additional electricity consumption has environmental consequences; already, there are 2.2 refrigerators and freezers per household in New Zealand (Roke, 2006) cited in [New Zealand Ministry for the Environment, n.d.].
If urban residents do undertake more just-in-time consumption than suburbanites both because of the higher storage costs associated with smaller houses, and the greater opportunity afforded by more stores nearby, we would expect to see this show up in the travel behavior data that is collected by urban regions.
variable
Minneapolis
St. Paul
Remainder of Hennepin County
Year Structure Built
1926
1929
1970
Sq. Ft.
1773
1826
2152
Sq. Ft. per Person
822
755
810
Households with No Cars
5900
2800
2500
Table 1 illustrates some of the differences between the City of Minneapolis, suburban Hennepin County (Hennepin excluding the City of Minneapolis), and the City of St. Paul in neighboring Ramsey County. Residents of Minneapolis live in older houses (average year built of 1926 vs. 1970 in the suburbs) with 1773 square feet vs. 2152 in the suburbs. However because of the lower household size, city residents actually have slightly more area per person. Further Minneapolis residents are more likely to be carless.
According to the 2000/2001 Twin Cities Travel Behavior Inventory among residents of the City of Minneapolis, 12.8 percent of daily trips were for shopping 5 while for Hennepin County excluding the City of Minneapolis the number was 12.2 percent. Thus Minneapolitans devote 5 percent more of their trips to shopping than suburban Hennepin County residents.
Minneapolitans also make slightly more trips than their suburban brethren, 3.81 per day vs. 3.70 for suburban Hennepin. (The unpublished 2011 TBI will likely show significantly lower numbers here). Given the small differences and their temporal instability, it probably is unreasonable to make much of them.
The evidence supports the hypothesis that city residents who have somewhat higher accessibility (see Figure) to neighborhood stores and somewhat reduced storage space at home shop more frequently.
Broadly, there are two types of places, those that satisfy the pint of milk test, and those that don’t. Similarly, there are two kinds of people, those who care about the pint of milk test and those who don’t. The problem comes from the mismatch of those who care but live in places that are unsatisfactory. (Those who don’t care but live in places passing the test are probably okay). If self-selection is at work, these cells are not randomly distributed, but people who want to live in particular environments do so. People who prefer milk-accessible areas bid up prices in those areas, while those who are indifferent (or perhaps lactose-intolerant) move out. However, if preferences change faster than spatial structure, there may be a mismatch.
Policy that excludes mixture of residential and commercial development may also foster a mismatch.
Evidence from the Twin Cities bears on the issue (Figures 2 to 5). According to the American Housing Survey [US Census Bureau, n.d.], over 80 percent of residents in the City of Minneapolis report satisfactory neighborhood shopping within a mile of home, compared with 70 percent of those in suburban Hennepin County (Figure 2). Despite that positive assessment of shopping, suburban Hennepin residents have a better opinion of their own neighborhood than those in the City of Minneapolis (Figure 3). The problems these urbanites report in greater numbers than their suburban counterparts are noise and traffic, crime, and odors (Figure 4).
When people move, they are doing so to places they believe are better, but for all residents it is the home that is better than previous much more so than the neighborhood, and in Minneapolis, only a third rate their current neighborhood as better than their previous (in contrast to half of suburban residents) (Figure 5).
To the extent neighborhood shopping enabling just-in-time consumption of the pint of milk is important to people, cities fare better than their suburbs, but if the cost of that neighborhood shopping is other urban ills, people will make the trade-off, sacrificing access to retail to have access to quiet and congestion free, safe, and pleasantly smelling suburban environments.
Whether this is a social good is another question entirely, and depends on relative efficiency of urban goods delivery services, energy efficiency of in-store displays vs. at-home refrigeration units, and numerous other questions.
References
Bennett, J. Morris, J. 2006 , Gateway people, Technical report, Institute for Public Policy Research.
Foxell, C. 2005 , Rails to Metro-Land., Clive Foxell, Chesham, Bucks, England.
Handy, S. 1993 , “A Cycle of Dependence: Automobiles, Accessibility, and the Evolution of the Transportation and Retail Hierarchies”, Berkeley Planning Journal , Vol. 8, pp. 21-43.
Heiskanen, E. Jalas, M. 2000 , Dematerialization Through Services: A Review and Evaluation of the Debate, Ministry of Environment: Edita, jakaja.
Herman, R., Ardekani, S. Ausubel, J. 1990 , “Dematerialization”, Technological Forecasting and Social Change , Vol. 38(3), pp. 333-347.
Marsh, G. 2004 , “Tesco piles ‘em high: Flats above supermarkets are a good buy”, The Times , Vol. June 18, 2004.
Nelson, D. Niles, J. 1999 , “Market Dynamics and Nonwork Travel Patterns; Obstacles to Transit-Oriented Development?”, Transportation Research Record , Vol. 1669, Transportation Research Board of the National Academies, pp. 13-21.
New Zealand Ministry for the Environment n.d. , Technical report.
Parliamentary Advisory Council on Transport Safety 2007 , Beyond 2010: A Holistic Approach to Road Safety in Great Britain, Technical report, Parliamentary Advisory Council on Transport Safety. Parliamentary Advisory Council for Transport Safety http://news.bbc.co.uk/1/hi/uk/7046200.stm.
Persson, A. Bratt, M. 2001 , “Future CO 2 savings from on-line shopping jeopardised by bad planning”, Proceedings of the 2001 ECEEE Summer Study ÔFurther than Ever from Kyoto .
Pushkarev, B. Zupan, J. n.d. , “Where Transit Works: Urban Densities for Public Transportation”, Urban Transportation: Perspectives and Prospects , pp. 341-344.
The Rise and Decline of the Great Atlantic and Pacific Tea company n.d.
Unterman, D. 1990 , `Accommodating the Pedestrian: Adapting Towns and Neighborhoods for Walking and Bicycling. Personal Travel in the US, Vol. II: A Report of the Findings from 1983-1984 NPTS, Source Control Programs’.
US Census Bureau n.d. , American Housing Survey for the Minneapolis St. Paul Metropolitan Area, Technical report, US Census Bureau. 1998AHS: Minneapolis h170-98-9.
Wenzel, T., Koomey, J., Rosenquist, G., Sanchez, M. Hanford, J. 1997 , “Energy Data Sourcebook for the US Residential Sector”, Lawrence Berkeley National Laboratory, Report , Vol. 40297.
Wernick, I., Herman, R., Govind, S. Ausubel, J. 1996 , “Materialization and Dematerialization: Measures and Trends.”, Daedalus , Vol. 125(3), American Academy of Arts and Sciences.
Yim, Y. 1990 , The Relationship Between Transportation Services and Urban Activities: The Food Retail Distribution Case, PhD thesis, University of California, Institute of Transportation Studies.
Adapted and updated from a post on The Transportationist Nov. 8 2007.
David Levinson, Transportist 
You must be logged in to post a comment.