Category Archives: Transportation

Transportation and “catalog” retailing

The last post discussed the role of urban transportation improvements leading to the development of the department store in the late nineteenth century. And of course the role of the automobile in shaping retail developments in the twentieth century is obvious. This got me thinking about the role of transportation (and communication) improvements in the evolution of retailing where the customer orders goods from a remote vendor and those goods are delivered to the customer.

We tend to think of modern developments such as e-commerce as novel developments. However, I’m going to start again in the late nineteenth century. But first, a brief excursion into the pros and cons of this type of purchase from the standpoint of the consumer. The major advantage is the selection of goods available, the ability to purchase things that are not available in local retail establishments, along with the convenience of being able to purchase the merchandise without having to travel to a store. The major cons are the inability to physically view the items to be purchased and the delay associated with the need for delivery, the lack of the instant gratification associated with physical purchase. Both introduce some uncertainties into the transaction. I am not mentioning price. The vendor saves money by not having brick-and-mortar stores, but this will be offset at least to some extent by the costs of shipping. This could go either way.

The mail-order catalog business emerged in the late nineteenth century with the major vendors being Sears, Roebuck and Montgomery Ward. The retailers made available a variety of merchandise to residents of rural areas that they otherwise would have been unable to acquire. The development of the railroads along with express freight services and parcel post to deliver the merchandise was undoubtedly a prerequisite. On the communications side, regular reliable mail service had been available for some time. But this mail-order business also required the development of printing technology that enabled production of the catalogs at a reasonable cost. (I don’t know just when this threshold may have been crossed, but I seriously doubt that Ben Franklin could have printed large numbers of Sears catalogs economically.) Just as transportation improvements enabled the rise of the general mail-order catalog, widespread use of the automobile made physical stores accessible to rural residents and led to its decline.

Another wave of remote shopping expanded in the second half of the twentieth century with the growth of specialized catalog shopping with telphone ordering, ranging from clothing (Lands End, L.L. Bean, etc.) to gourmet foods (Dean & Deluca). The attraction to the consumer was access to a wider selection and to specialized goods they could not purchase locally. Some improvements to delivery services helped. UPS did far better than the very long delivery times the post office provided, especially in the past. The role of improvements in communication should not be discounted. Toll-free 800 numbers made the calls free. I imagine customers would not have relished the idea of paying the expensive long-distance charges of the past to make purchases. Again, I don’t know at what point the costs of high-quality color printing for catalogs became reasonable, especially since they send out huge numbers. But it does seem that I saw a lot less color printing in the mid-twentieth century. These catalog retailers also innovated to minimize the risks associated with remote purchasing, offering no-questions-asked returns if something didn’t fit or even if you just didn’t like it.

We finally get to today’s e-commerce. It is noteworthy that Amazon started with books, which have two features favoring this model. First, for any given author and title, all books are the same. There is not the guessing that would be involved in choosing among several green sweaters. And second, with books, the breadth of selection is everything. No brick-and-mortar bookstore can possibly approach the inventory of an online retailer. Amazon and the other online retailers also adopted the policies of the catalog retailers (now, of course, also online) with easy returns and high levels of customer service. Zappos has no problem with your ordering multiple pairs of shoes in order to pick the one pair you want and send the others back.

Obviously the World Wide Web was the innovation on the communications side, making both obtaining informtion on available items and ordering quick and easy. And on the transportation side, the expansion of e-commerce is driving improvements in delivery services, with 2-day and even 1-day delivery becoming commonplace without excessive charges. This, of course, reduces the penalty of having to wait for delivery. Indeed, considering the likelihood of a lag between wanting to purchase an item and having the time to go out to a store, online ordering may be quicker.

Given the rapid developments in e-commerce and speedy delivery, we may be seeing only the first stages in the effects on physical retailers and therefore our urban areas.

Transportation and economies of scale in retailing

The automobile may be blamed for the evolution of big-box retailers, but the effect of improvements to intraurban transportation on retailing began much earlier. This can be seen clearly with the development of the department store in the latter part of the nineteenth century.

In the walking city of the early nineteenth century, most urban residents could only move around on foot. This necessarily limited the distances they could travel and the amounts of goods they could carry. Stores tended to be small and rather limited.

Transportation improvements–horsecars, cable cars, electric streetcars, and more–dramatically increased mobility in urban areas. Cities greatly expanded as residents took advantage of the greater ease of travel. Going to the developing central business districts several miles away became feasible.

A larger number of potential customers could travel to a store located in the downtown area, creating a greater market. This allowed the emergence of the modern department store carrying a far larger range of goods with greater selections. More volume provided greater economies of scale to the store in the sale of its merchandise. But these were also economies of scale from the perspective of their customers, who benefited from the convenience, wider selection, and lower prices.

Coming to shop at the department stores via public transportation did have one limitation, however. Customers purchasing large numbers of items or very large items could find it difficult or impossible to carry their purchases back home with them. The stores recognized this problem and offered delivery of merchandise purchased in the various departments of the store.

This evolution depended solely on the transportation improvements made in the late nineteenth century. It had nothing to do with the automobile. Indeed, at least some department stores continued to assume that significant numbers of their customers would come to their downtown stores using public transportation at least into the 1950s. When growing up and shopping at the large downtown department stores in Milwaukee during that decade, the stores were continuing to offer their delivery services. Of course now, the assumption more often is that customers will be arriving by automobile and can take all but the largest items home themselves.

The joy of cul de sacs

Every day, I go out walking with our little dog, for a half hour, an hour, or even more. We just walk around our neighborhood, going different ways each day, enjoying being outside, being together, and getting exercise.

In the beginning, I was the one taking Daisy for the walk, choosing where we would go each day. We would pass cul de sacs, but I would never turn into one. You’d just have to come back out. You wouldn’t be getting anywhere.



Over time, when we got to a corner, Daisy would indicate a preference for going one way or the other, and since we were not going anywhere in particular, I would let her choose the way. This evolved until now she is taking me for the walk, deciding where to go and when to head home.

Soon after she took over, Daisy decided to turn into a cul de sac. My initial reaction mirrored the reason I never did that: We would just have to come back out. But I followed along…and then started thinking about it. Yes, we weren’t getting anywhere. And that was the point. We were just out walking. We weren’t trying to get anywhere. Walking into a cul de sac was a perfect example of that.

Sometimes Daisy turns into a cul de sac and sometimes she doesn’t, except for one favorite cul de sac which she almost never skips when we are in that area. And whenever she does turn in, I now just smile and think, this is just great. We’re not getting anywhere. And that’s just what I like.

Alternatives to connected streets

Walkable urban environments are good. One characteristic that makes an area walkable is the provision of means to walk from one location to another in a fairly direct manner. Having a highly connected street network is advocated as a means to accomplish this. And it does.

What the advocates of connected streets fail to see, however, is that having a connected street network is not necessarily the only way to ahieve the desired objective. Connected streets are required only to the extent that the only paths for walking are sidewalks at the sides of the streets.

Paths can be provided for walking that are separate from the street network. Such paths have advantages and disadvantages compared with sidewalks along the streets. But they definitely represent an alternative to connected streets for providing more direct ways of walking from one place to another.

Radburn, New Jersey, was designed with a system of pedestrian paths that were grade-separated from major roads. Greenbelt, Maryland, the depression-era planned community, likewise included such paths. Moving forward several decades, the planned communities of Columbia, Maryland, and Reston, Virginia, included extensive systems of walking paths separate from the street system.

When I first moved to Irvine, California, in 1974, I lived in an apartment adjacent to one end of University Park, one of the early single-family residential developments in that planned city. At the other end was a small shopping center with a modest supermarket and a few other establishments. University Park had an extensive system of interior walking paths. I would frequently use those paths to walk to the market, as I found it more pleasant than walking along the streets.

Walking along separated paths or along a street each have advantages and disadvantages, depending on the situation. The separated path, especially if nicely landscaped, can be pleasant and tranquil. On the other hand, it may not have much activity. Some streetscapes can be more interesting to walk along, while others bordering a major road without other interesting activity can be far less pleasant.

One more thing. It can be easy to turn cul de sacs, the bane of the connected street advocates, into walkable settings. This is especially easy if the cul de sac is extended close to an adjacent street with no house built at the end of the cul de sac. Simply put in a short path from the end of the cul de sac to the adjacent street. This alternative is fairly common in my neighborhood in Upland, California, which is east of Los Angeles. Here is the satellite view from Google Maps showing the path (obscured at several points by trees) from the end of the cul de sac  to the sidewalk along Mountain Avenue (that sidewalk is also obscured by the continuous line of trees along Mountain Avenue):

Cul de sac with walking path at end, from Google Maps

Cul de sac with walking path at end, from Google Maps

And here is a ground view of the path from the end of the cul de sac looking west towards Mountain Avenue:

Cul de sac with walking path at end, ground-level view

Cul de sac with walking path at end, ground-level view

Connected streets

It is currently broadly accepted that a highly connected street network is a good thing, that it represents good design. Connected streets promote walkability by providing more direct paths between more origins and destinations. This is extremely reasonable. I think it makes sense. I have no quarrel with this.

Where I do have a problem is when the proponents of connected streets assert that layouts with curving streets and cul de sacs necessarily always represent bad design. And I have even more of a problem with those who have asserted that designs with curving streets and cul de sacs arose from some kind of nefarious plot to keep people from walking and force them into using automobiles.

This is nonsense. Subdivison design with curving streets and cul de sacs was adopted to minimize through traffic in residential areas and to slow down vehicles that did drive on those streets. The purpose of doing this was to enhance safety, especially for the children living in the areas. This strikes me as a very laudable goal.

During the middle of the twentieth century, some of the best examples of planning in the United States were considered to be the planned communities of Columbia, Maryland, and Reston, Virginia. And residential areas in both places made ample use of curving streets and cul de sacs.

A negative effect of such design was, of course, that it limited options for walking, making many possible trips longer and less direct. Just like connected streets may result in more through traffic in some areas. In both instances, the negative effects may be reduced by making good design choices.

I have three takeaways from this:

First, most of the choices we make, whether in planning or many other spheres, involve tradeoffs. Emphasizing one objective is often likely to reduce the achievement of another. We can work to address both, but usually something has to give if we seek more of something else.

Second, historical perspective is important. In this case, understanding the bases for design choices 5 or 6 decades ago tells us how they came to consider those good designs. It is unfair to simply dismiss the ideas of those who came before when they did not necessarily know what we know now and they did not necessarily have the same priorities that we have now.

And third, perhaps be a bit more humble. We have our ideas of what is best, and we should certainly advocate for those. But don’t go so far as to convince yourself that we have discovered the ultimate truth. Remember that in a few decades, people may be looking back on our ideas and observing how totally wrong we were.

Major transportation infrastructure and metropolitan extent

For defining both MSAs and CSAs, the Office of Management and Budget specifies (different) minimum commuting thresholds. Obviously the choice of the exact value is somewhat arbitrary. But OMB has to make definitions that use data that the Census Bureau collects for the entire country and must use a uniform, consistent standard.

CSAs involve the combination of MSAs (and Micropolitan Statistical Areas) into a single area. This got me to thinking whether other factors might be considered in judging whether areas should be combined and be considered part of a larger metropolitan area. I would like to offer one suggestion: Combined or shared major transportation infrastructure, specifically commuter rail systems and commercial airports.

Going through the major CSA combinations I discussed in the previous post, New York is connected by commuter rail service to its Connecticut suburbs. San Francisco has commuter rail service down the peninsula to San Jose and beyond. And there is additional service from San Jose extending to the east. The three major cities of the Riverside-San Bernardino-Ontario MSA are served by no less than three commuter rail lines, two to downtown Los Angeles and one to Orange County, a part of the Los Angeles MSA.

For the two large combinations about which I initially expressed some skepticism: Commuter rail lines connect Washington with Baltimore and Boston with Providence.

When a single airport provides all or most commercial airline service to two MSAs, it is reasonable to consider them to be part of a single, larger metropolitan area. Raleigh-Durham International Airport serves its two namesake MSAs, of course, as does Greenville-Spartanburg International. The Piedmont Triad International Airport serves the Greensboro, Winston-Salem, and High Point MSAs, hence the name.

The Provo and Ogden MSAs are combined with Salt Lake City into a single CSA. Provo and Ogden each have an airport providing service to just a handful of destination (3 for Provo, 1 for Ogden). So it seems clear that the Salt Lake City airport provides much of the airline service to those MSAs.

The very largest MSAs and CSAs can be served by multiple airports, so one would not expect a single airport to serve both MSAs in those situations. But there are at least two examples of airports located in one MSA that are also seen as providing significant service to another MSA in the CSA. The first one I’d mention is Baltimore’s airport, Baltimore-Washington International. The name says it all. And I’ve flown into it multiple time when going to Washington.

In the greater Los Angeles area, Ontario International is obviously located in the Riverside-San Bernardino-Ontario MSA. Ontario International is literally owned and operated by Los Angeles World Airports, the authority that is also responsible for LAX and one other airport in the Los Angeles MSA. (As an aside, ownership and control of Ontario International is due to be transferred back to the City of Ontario and San Bernardino County later this year. They fought for the change believing that Los Angeles World Airports was favoring LAX over the Ontario airport.)

Common or shared major transportation infrastructure should not necessarily be the sole basis for determining that two areas should be considered part of a single, larger metropolitan area. But I believe it is a strong indicator.

New York versus Los Angeles — transportation

After doing the comparison of population densities in the New York and Los Angeles areas, I thought it could be interesting to do some comparisons related to transportation.

New York is the area with the well-developed and highly used public transit system. Los Angeles is famous for its dependence on the automobile. So let’s begin by looking at the percentages traveling to work using transit versus private motor vehicles. (The journey-to-work statistics are from the Census American Community Survey, 2012, 1-year data for the MSAs). Here is the modal split:

  • Transit
    • New York: 31 percent
    • Los Angeles: 6 percent
  • Car, truck, van
    • New York: 57 percent
    • Los Angeles: 84 percent

No surprises here. New York has a high 31 percent using public transit versus only 6 percent in Los Angeles. Fully 84 percent of workers in Los Angeles traveled to work in a car, truck, or van. So it’s clear that workers in Los Angeles are extremely auto-dependent.

But look again at the New York figures. Well over half of the workers in New York also drove or carpooled to work, nearly one and one-half times the number taking transit. So it’s hard to say that New York is not, to a considerable extent, also auto-dependent.

While I was getting these modal split data, I also got the data on the average times for the journey-to-work by mode of travel. So I am comparing these as well:

  • Transit
    • New York: 50.5 minutes
    • Los Angeles: 50.2 minutes
  • Car, truck, van
    • New York: 29.4 minutes
    • Los Angeles: 27.9 minutes

Almost exactly the same for New York and Los Angeles for both transit riders and those driving or carpooling. Transit trips take much longer, on average, than the car, truck, and van trips. I’m not sure what to make of the lack of differences, but it strikes me as interesting.

Finally, Los Angeles is known as the land of the freeway, which is the iconic symbol of the region. Here are the miles of interstates and of other freeways and expressways for the two Urbanized Areas (from Federal Highway Administration, Highway Statistics 2014):

  • New York: 1,258 miles
  • Los Angeles: 594 miles

The New York area has twice as many miles of these highways compared to Los Angeles! But this is not quite a fair comparison, since the New York area is much larger than Los Angeles. So we take this into account by calculating the miles per million population:

  • New York: 70.7 miles per million
  • Los Angeles: 50.4 miles per million

Even taking population size into account, New York has 40 percent more Interstates, and other freeways and expressways, than Los Angeles. This is so much at variance with our accepted ideas about the two areas that I am not sure what to say.

For a visual comparison, here are links to Google Maps showing, at approximately the same scale, the freeways as bold orange lines for the New York and Los Angeles areas.