Can low income households locate closer to employment?

The location of people throughout the metropolis is constantly in flux, in terms of who locates where. This means transportation investments to serve certain place-based communities or populations can often be misguided. Looking at the bigger picture, what are the goals of new transportation investment? Reducing congestion is the one that gets the most press. Opening up new land for development gets less but is still a very real objective for local governments in particular. One that is not explicitly talked about as much is to increase accessibility for people in general and especially for lower-income/mobility populations. Reducing congestion is part of doing so by reducing travel times. One may speculate that those with highest incomes value time the most, and have the financial resources to locate wherever in the metropolis will maximize their accessibility/minimize their time in transport. So do lower-income populations inevitably just get the short end of the stick?

This paper talks about the effect of income on commuting time. It also refers to the tension between the high value of time for high-income households, making them locate near job centers and minimize commute times, and the neoclassical theory of higher incomes leading them to consume more land and housing, causing them to locate further out. Through least-squares regression and other controls, it attempts to control for all other variables other than income. The paper mentions that commute times increase slightly as income does (18-24 minutes on average from lowest to highest income quartile). The coefficient for time in the model is positive, showing commute times increased as time went on (i.e. as people got older). Therefore, older people may tend to commute longer as well. The end result is that income has a positive (increasing) effect on commute times, but not much of one – the demand for more housing pushing them outwards and value of time pulling them closer virtually cancel each other out.

This is a good study and produces valuable information. What one must remember is that high and low incomes simply do not occur in isolation. There are many other factors including household size, presence of children and even personal taste for living space and commute time. Higher-income workers are unlikely to agglomerate in large numbers near job centers. With widely varying household characteristics and the financial ability to live almost anywhere, there is a low probability of getting the majority of them to live in close proximity to where they work. However, it’s interesting to note that while they have higher commute times than lower-income workers, they don’t commute much longer. Why is this?

In part, lower-income individuals face a different tradeoff. Having fewer financial resources, they are more likely to not own a car and thus have to use transit, limiting their access to jobs. Furthermore, the increased commute times created by taking transit make some jobs not worth it, and price some low-income households out of the labor market. Most wouldn’t consider commuting three hours one way for a minimum wage job. Research indeed shows (Ihlanfeldt 1993) that lower income households have a lower propensity to commute. This may suggest it’s more important – and more efficient – for lower-income households to be able to locate near their workplaces. Put differently, perhaps higher-income households, as a whole, are as close to their jobs as they wish to be, while low-income households probably aren’t.

Spatial mismatch literature suggests that the poor do have lower access to jobs. While this alone may not lead to significantly higher unemployment among this population, skill mismatch, in which jobs located near an area are poorly matched to the occupational skills of its residents, comes into play as a big issue as well (Immergluck 1998). Furthermore, another study (Covington 2009) suggests that most of the improvement in the poor’s access to jobs during the 1990s stemmed from the movement of that population closer to jobs, as opposed to jobs moving closer to them.

This would strengthen the argument for vastly increasing transit service in metro areas. This would be necessary because closely targeted transit investments may not be very effective for lower-income populations, given the unlikely event that these jobs cluster/agglomerate in small areas (as opposed to higher-income professions, such as finance). Such investments, which would include things like commuter rail, are probably more likely to favor higher-income groups more than anything. While this may not be politically possible at this point, there is a lot to suggest that offering extensive transit service throughout the metro is simply the most effective way to give the poor better access to jobs. This would reduce transportation costs for individuals and make any given location in the metro more accessible. It would potentially reduce overall costs for low-income households by reducing the need to move so often. If it allowed households to move less often, it might also stabilize and strengthen low-income neighborhoods.

Most of this does not speak to the congestion problem, of course. Nor does it actually speak to whether it would reduce vehicle miles traveled overall. It seems that the eternal challenge for transportation investment is simply figuring out ways to move everyone more efficiently. With the massive allocation of resources likely necessary to accomplish this, we probably won’t make progress on this front anytime soon. However, promoting equality in transportation and accessibility is certainly possible. Further research, which is needed, may indicate that bringing jobs within easier reach of low-income households may even improve congestion somewhat, and may lead to better economic outcomes overall.

Still around

Does anyone still read my blog? I’m not sure, but I am working on a new post which will hopefully be ready within a few days. If you’re reading this, you might be interested so please check back.

Agencies vs. interest groups vs. agencies: transportation planning in Texas

A recent article (now a few weeks old) by Ben Wear in the American-Statesman details the I-35 widening project between San Antonio and DFW.  Many, probably including myself, would say this project is overdue.  However, additional planning for the corridor, with additional improvements being considered, is currently underway as well.  The website for the project identifies it as “a different approach to transportation planning.”  So what’s different about it?

The process claims to get local officials, stakeholders, etc. involved in the process from the beginning, which is often not done.  While this may be a step in the right direction, it’s not exactly a fundamental change in the planning process. Still, it’s hard to complain about the effort.

However, for what all this is worth, the perspective of TXDOT on building roads doesn’t seem to have changed.  The plan is full of plans for eight- and ten-lane roadways, managed (HOV/toll) lanes through the countryside, freeway extensions, and so forth.  An example:  a project being considered is widening I-35W between downtown Fort Worth and SH 174 in Burleson to eight lanes with four managed lanes, and continuing the eight lanes (without managed lanes) all the way to Alvarado.  35W from downtown to I-20 is already eight lanes, and 35W south of I-20 is not a particularly congested road, even at rush hour.  While suburbs to the south have grown quite a bit in recent years, until recently they weren’t suburbs at all – they were probably more accurately described as smaller rural towns on the outskirts of a major metro area.  In short, the area south of Fort Worth is growing, but certainly not booming, and shouldn’t require a roadway of this width plus toll or HOV lanes to boot.

Plans also include a western bypass around Waco and a new limited-access freeway from central Arlington south to Hillsboro, to name two examples.  My guess is that citizen opposition will derail many of these projects (which would require a lot of new right-of-way).  However, it’s somewhat distressing that TXDOT seems to see itself as the proponent of building any and all new roads, unless the citizens can stop them.  TXDOT isn’t the only guilty party here (although rightly or wrongly, it’s acquired more power than most); most agencies seem to try to justify their existence by constantly pushing for more and more of whatever they do.  These agencies fight it out with other agencies for funding and with community and interest groups who want to stop them.  This has become the status quo, but one has to question if it is really the best way to do things.

Promising new developments at the federal level may help to change this.  The recent joining together of HUD, the DOT and the EPA will hopefully reduce tension between the objectives of these respective agencies and, in doing so, fully leverage the expertise and knowledge present in each to create solutions to problems relating to all, especially relating to sustainability and livability.  The partnership created a sustainable communities and regional planning grant program, providing funding for planning at a geographic level, the region, that is too often ignored by politicians and planners. Collaboration at the regional level also often falls victim to exclusionary policies by local municipalities, and to competition for retail and other businesses between these cities.  Hopefully this is just one example of good things to come from this and other interagency partnerships.

Birmingham’s sprawl problem

In recent years, Birmingham has struggled mightily with issues of public transit and other issues related to urban livability.  Perhaps most telling about the state of the metro area is the population change in the central city and county (Jefferson) over the past 30 years or so.  Between 1980 and 2000, Birmingham and Jefferson County’s population actually fell, by 14.6% and 1.4% respectively.

This can be easily chalked up to poor leadership, race issues, or both, and many do so.  But is the situation really that simple?  Perhaps Birmingham’s rapid suburbanization is to blame.  The Birmingham MSA actually added three counties (Bibb, Chilton and Walker) between 2000 and 2009, and grew in population mostly because of this fact.  (Jefferson County actually lost population between 2000 and 2008, falling from about 662,000 to 659,000.)  To be sure, pretty much every metro in the country has seen significant suburban growth during this period, even those that have declined overall.  As the growth pole of a traditionally rural state, however, Birmingham may have been positioned for more rapid suburbanization than most.  Has this been the case?  Is Birmingham’s rapid sprawl a major, or even the main source of its problems?

This report from 2002 on measuring sprawl places the Birmingham MSA relatively high on the list of most sprawling metros, but at number 23 out of 83 metros studied, it’s certainly not the worst.  However, breaking down the sprawl index created by the authors of the study should shed more light on Birmingham’s situation in particular.

The study combines many economic, socioeconomic, housing and travel/land use variables in a principal components analysis to separate out each particular contributing factor to sprawl.  They end up with components for residential density, street connectivity, mix of land uses and overall centeredness of the metro (the strength of its center and subcenters).  Birmingham fares relatively well on the centeredness factor and, to a lesser extent, the street connectivity factor.  Its ranking is dragged down by low scores on the residential density and land use mix factors.

Indeed, a look at the density numbers for the Birmingham area compared to some similarly sized metros confirms its relatively low residential density.

MSA	Population density (persons per square mile)
Birmingham, AL	289.1
Knoxville, TN	280.6
Jacksonville, FL	417.6
Oklahoma City, OK	255.1
Louisville, KY	495.0
Columbia, SC	368.7
Richmond, VA	338.4

This might help to explain Birmingham’s struggles with providing public transit.  Its very low residential density makes efficient and effective transit service very difficult.  Its relatively high score on the centeredness factor in the study may be somewhat misleading as well.  The low residential density has likely put a high percentage of residents so far away from the central city that regular interaction with it is not an option.  The Birmingham MSA likely includes a number of subcenters detected by the study’s methodology that have very large service areas.  The separation of land uses would corroborate this as well, as the average subcenter would have a larger than average concentration of commercial, industrial and other non-residential land uses, to compensate for the strict separation of residential land uses, and would thus likely serve a larger area.

Density is not all that matters, but is probably the most important factor.  To bring this into focus, consider that a metro with a very dense center and low density outlying areas isn’t necessarily sprawling.  The concept of weighted density, which weights densities of sub-areas within a larger area based on the percentage of population they contain, can be used to illustrate this.

	Weighted density (persons per square mile)	Central city vacancy rate
Birmingham	1,666	11.7%
Knoxville, TN	1,151	9.8%
Jacksonville, FL	2,242	7.9%
Oklahoma City	2,561	10.4%
Louisville	2,888	8.1%
Columbia, SC	1,660	8.4%
Richmond, VA	2,327	8.4%
Atlanta	1,873	10.0%

At about 1,666 persons per square mi, the Birmingham MSA’s weighted density is very low.  Larger MSA’s do tend to have higher weighted densities, but Birmingham’s is low compared even to MSAs of a similar size.

Interestingly, Birmingham’s weighted density is only a little lower than Atlanta’s. However, one might consider Atlanta a bit more successful on the livability front (its awful traffic notwithstanding).  Atlanta central city has a much higher median household income than Birmingham; the gap between it and median income for the MSA is also smaller.  This results in a larger residential and commercial tax base.  The vacancy rate for Birmingham’s central city further illustrates the disinvestment that has occurred.  At 11.7%, it is higher than any of its counterparts in the Southeast.  Political initiatives that would help the central city in Birmingham have little support.  Atlanta’s public transit agency has a little sway, though not much, in the state legislature.  Birmingham’s, of course, has none.

All is not necessarily lost for the Birmingham area, though.  Its strong center and subcenters should be considered something to build on; attracting jobs to subcenters throughout the area to bring them closer to residents is a possibility.  This would also allow for a more viable and effective transportation strategy, including better mass transit, linking these centers.  Discussions and studies getting underway regarding the future of U.S. 280 through Jefferson and Shelby counties are good opportunities to start moving in this direction.

Continuing to invest in the knowledge and high-tech economies in Birmingham and Alabama as a whole will also likely help the central city.  Such industries often favor inner-city locations and can occupy spaces left vacant by industrial land uses that have long since departed.  This will make the inner city more attractive and attract more skilled and educated residents who often favor a more urban type of living.

Good leadership will be needed to make legitimate efforts to pursue these objectives.  That, of course, is another matter entirely.

I-35 and Austin’s possible urban rail

This article from the American-Statesman on Austin’s expected bond election this November mentions an interesting study a small portion of the proceeds will be spent on:  what to do about traffic problems on I-35.  The SH 130 toll road around the city now complete and still rather sparsely used, the city has apparently realized the problem has not been solved, and isn’t waiting for TXDOT to figure it out.

The challenge is accommodating all of that through traffic combined with local traffic, which probably made the toll road sound like a nice idea (albeit one I doubt many are surprised isn’t working).  Unfortunately, better transit service, at least in the short term, probably wouldn’t alleviate local traffic much, as it wouldn’t replace many longer trips for which people are far more likely to choose or need to use 35, to speak nothing of the ability of land uses and densities to support such service.  All of this has to be done with minimal damage to the urban fabric in central Austin, a point which will inevitably be reinforced by neighborhood groups and probably some downtown boosters as well.  (In its current state, I-35 through downtown doesn’t occupy a huge footprint; this facilitates some pedestrian travel between East Austin and downtown, and has probably played a small role in continuing reinvestment in near eastside neighborhoods.)  It seems extremely doubtful that removing it in favor of a parkway-type road would pass muster.  Even if the city thought it would be good from an economic development/livability type standpoint, the importance of I-35 as a through route means TXDOT would be extremely unlikely to agree, with the only other options around Austin being toll roads. It’ll be interesting to see what they come up with; I’m no transportation planner/engineer, but I’m not seeing many good options there.

Speaking of transportation, some will notice a certain omission from the project list for the bond election: the city’s planned streetcar line, which was getting some attention several months ago as Capital Metro’s Red Line struggled to even begin operations.  I’m glad I procrastinated as usual on this post, because now the city has announced it’s set to pass a resolution identifying a 2012 bond election as the funding source for the urban rail.  As the city’s work on it continues, this project appears to still be a priority, but don’t hold your breath waiting for construction to begin.

Scotland’s cities: creative centers?

Migration statistics provide one of the keys to evaluating theories of urban development such as Richard Florida’s creative class theory. (Maybe find quote here to illustrate that creative class theory hinges on migration.) This data (http://www.gro-scotland.gov.uk/files2/stats/gros-mid-2008-population-estimates-scotland-population-estimates-by-sex-age-and-administrative-area/j1075011.htm) from Scotland is interesting to consider in the context of his theories, which haven’t been discussed much (to my knowledge) in the context of countries other than the U.S. Florida’s theory relates a city’s diversity, including its proportion of immigrants, to its population and economic growth. Scotland’s largest cities, which in order are Edinburgh, Glasgow, Aberdeen and Dundee, all saw positive net migration from 2007-2008. Most of this positive migration came from overseas, as denoted in the column on the far right. Migration within Scotland tended to be negative for these cities. Furthermore, the data indicates that overall net migration to and from abroad was negative for most other areas in Scotland. (U.S. Census data doesn’t provide net migration data for overseas.) Put simply, the largest magnets for immigrants in Scotland were the largest cities.

This pattern doesn’t vary from cities in the U.S. as much as one might think. MSAs (which is the level of geography Florida uses for his analysis) with high inflows of immigrants in 2009 tended to be near the top in terms of size, as seen in the table below. The top four destinations for immigrants were New York, Los Angeles, Chicago and Miami; all of these cities (MSAs) had negative net domestic migration. Overall, seven of the top 15 cities on this list had negative net domestic migration for 2009.

CBSA Code		Total population change /1		Natural increase	Vital events		Net migration
		Number	%		Births	Deaths	Total	International /2	Domestic
35620	New York-Northern New Jersey-Long Island, NY-NJ-PA	101,295	0.5	108,389	250,337	141,948	-9,609	100,669	-110,278
31100	Los Angeles-Long Beach-Santa Ana, CA	106,402	0.8	112,033	190,001	77,968	-4,838	75,062	-79,900
33100	Miami-Fort Lauderdale-Pompano Beach, FL	45,299	0.8	23,238	70,298	47,060	22,227	51,548	-29,321
16980	Chicago-Naperville-Joliet, IL-IN-WI	64,931	0.7	71,553	139,860	68,307	-7,026	33,363	-40,389
47900	Washington-Arlington-Alexandria, DC-VA-MD-WV	98,305	1.8	48,153	79,101	30,948	50,093	31,904	18,189
19100	Dallas-Fort Worth-Arlington, TX	146,530	2.3	69,479	105,621	36,142	76,812	31,571	45,241
26420	Houston-Sugar Land-Baytown, TX	140,784	2.5	63,191	96,404	33,213	77,658	27,996	49,662
41860	San Francisco-Oakland-Fremont, CA	57,617	1.4	25,927	55,577	29,650	32,353	24,376	7,977
38060	Phoenix-Mesa-Scottsdale, AZ	76,771	1.8	42,539	71,553	29,014	34,274	21,833	12,441
12060	Atlanta-Sandy Springs-Marietta, GA	89,627	1.7	51,412	83,003	31,591	37,767	20,288	17,479
14460	Boston-Cambridge-Quincy, MA-NH	43,975	1.0	18,505	53,616	35,111	26,063	19,250	6,813
41940	San Jose-Sunnyvale-Santa Clara, CA	29,054	1.6	18,313	27,733	9,420	10,986	16,347	-5,361
37980	Philadelphia-Camden-Wilmington, PA-NJ-DE-MD	27,756	0.5	22,620	75,747	53,127	5,367	12,944	-7,577
42660	Seattle-Tacoma-Bellevue, WA	51,211	1.5	21,685	43,392	21,707	29,954	12,919	17,035
36740	Orlando-Kissimmee, FL	21,453	1.0	13,757	29,057	15,300	7,441	11,720	-4,279

Immigration is one aspect of creative class theory that doesn’t seem very well articulated. Florida includes it in his diversity index in his book The Rise of the Creative Class, and cites that index’s significant correlation with population and income growth. However, when immigration is broken out, the relationship is not significant. Regardless, diversity, including immigration, has a prominent place in his theory. Semi-recent blog posts (http://www.newgeography.com/content/001110-the-white-city) only confirm that there is much more to the creative city’s relationship with immigrants and diversity in general.

Furthermore, this article (http://www.nytimes.com/2010/04/16/us/16skilled.html) from the New York Times points out that metro areas attracting immigrants of all skill levels are the ones that grew the most between 1990 and 2008. It’s hard to exactly define the relationship here; a fast-growing city might simply attract more immigrants to work in the construction industry and similar areas. Regardless, the data examined here indicates that the relationship may not be exactly as Florida surmised.

This data also highlights the concentration of immigrants in urban centers of the UK, as opposed to a more widespread suburbanization of immigrants in the U.S (Waldinger 2001). Immigrants are a smaller percentage of in-migrants for principal cities in the U.S. (about 13% compared to about 30%). Maybe Scotland’s larger cities, along with the rest of the U.K.’s cities, would fit into a slightly different model for creative cities. That their migration within Scotland tended to be negative suggests an outmigration of mainly families with children, while in-migrants probably tend to be single or without children.

There has been lots of discussion on the limitations of creative class theory in describing U.S. cities. It would be interesting to see a more thorough discussion of urban systems abroad and how they relate to it.

Reference: Waldinger, R. (ed.) (2001) Strangers at the gates: new immigrants in urban America. Los Angeles: University of California Press.

New post soon

I wanted to let the few people who read this blog know that I plan to post something new sometime this weekend, after a somewhat long layoff. Mostly, this layoff has been due to my recent move to Tuscaloosa, Alabama, to take a job here. For anyone primarily interested in Austin stuff, I expect to still post things on Austin, as I will be visiting from time to time and keeping up with developments there. So, I hope you’ll keep reading.

Thanks to everyone who reads this blog. I look forward to hearing all of your thoughts on upcoming posts as well as past ones.

The value of Houston light rail

A post about the value of passenger mile measures in determining a transit system’s effectiveness or level or service by Michael Lewyn at Planetizen brought to mind another post by Randal O’Toole a couple of months ago about the “failure” of Houston’s Main Street light rail line.  In Randal’s view, construction of the line killed the growth in total passenger miles, and this coupled with an increase in bus operating costs despite an overall reduction in service spells failure for this venture.

This line of reasoning begins with a pretty big assumption – that passenger miles were going to keep growing at the same rate as the previous four years or so, and beginning construction of a light rail line is what choked off that growth.  Randal says nothing about the reason for the growth during that time, or why it should have continued.

Furthermore, there actually was substantial growth in overall ridership and passenger miles in the first full year the light rail was open.  There was a slight decrease in both the second year, which is troubling to be sure, but Randal doesn’t discuss why this may be – he simply assumes that, for whatever reason, it’s because of the light rail line.  I can’t speak intelligently on the reason for the decrease, but it certainly isn’t clear to me that the light rail is to blame.  Bus vehicle revenue miles decreased along with these – why did Metro reduce those?  It’s another pertinent question that remains unanswered.

He also cites the fact that bus operating costs increased 42 percent from 2001 to 2006 despite a slight reduction in vehicle revenue miles.  However, the National Transit Database shows an increase of about 28 percent nationwide for the same time period ($6.49 to $8.30 in operating costs per vehicle revenue mile).  While Houston’s figure is a bit higher, the two are comparable, and this suggests that the increase in operating costs are mainly due to some factor or set of factors affecting all transit systems (increased fuel costs are likely at least partly to blame).

Additionally, it’s not clear what connection there would be between construction of a rail line and operating costs for bus services.  Costs relating to construction shouldn’t even figure into operating costs.

And finally, Michael Lewyn’s post sums it up about passenger miles.  Isn’t reducing passenger miles kind of the point?  It’s been shown that transit users in central cities with good access to transit (including bus and rail) travel shorter distances than does anyone driving a car.  One of the goals of such a line should be to more effectively and efficiently link places, and one way it could do that would be to require current and future transit users to travel fewer miles to get to where they’re going.  A new light rail line through the heart of the city could certainly explain some of the reductions in passenger miles and VRM, if it helped many find shorter routes.  Possibly more importantly, it has the potential to influence land use patterns to effectively link more people and their destinations in the future.

Light rail, especially if not done well, is no godsend for our cities.  However, a more evenhanded discussion of the costs, benefits, and effects on overall mobility of this and any other proposed line is necessary to make the tough decisions about where to allocate federal funding and local tax dollars.

What is urban, and does it matter?

I recently saw a dispute between commenters on a blog post over what exactly constituted “urban.” This dispute raised a good point in my mind. Such a discussion is an interesting one, and an important one, on some level. However, discussion of whether or not our cities’ development is “urban” is in dire need of context. It was cited in a comment at this post that urban areas must primarily serve pedestrians, with cars only a secondary concern or not a concern at all. Density would obviously figure prominently into this. Others, including Jane Jacobs, have suggested that density alone is sufficient. Still others would indicate that urbanity is signaled by street activity, not just pedestrian friendliness or density.

Regarding the development of cities, these aren’t the questions that should be asked. Rather, is “urban” in fact what our cities need, and in what form? What parts of our lives does such development improve, and what interests does it serve? How does it affect our lifestyles? What role does the market play – i.e. is the market inclined to provide and/or accept such development, and why or why not?

At this point in time, our urban landscapes are quite varied. For example, in Austin, they include these widely disparate areas:

• Downtown, primarily an employment center and a regional destination for restaurants and other entertainment, but with a couple thousand residents (most of whom I’m guessing have cars) as well;

• Central Austin neighborhoods (such as Brykerwoods, Brentwood, and Allandale) with a lot of single-family housing, but some multifamily as well, some, but not strict, separation of uses, often characterized by walkable grid-like street systems and strip-oriented commercial development (but not entirely), which in many cases is set up (or retrofitted as the case may be) to serve both pedestrians and cars somewhat well;

• New urbanist developments like Mueller, denser and more walkable than most single-family neighborhoods, with some, but not too much, mixing of housing types, and little proximity to non-residential development at this time (but Mueller has plans for it in a residential area, albeit more or less concentrated in a centralized location)

• Larger suburban single-family subdivisions, such as Anderson Mill in Northwest Austin, many of which have large lots, but even in subdivisions that don’t, street grids that don’t lend themselves to walking, entirely separate from other types of development – and their counterparts, the large commercial developments on arterials and highways, concentrations of many businesses and lots of commercial space, serving almost exclusively cars and not accessible or friendly, for the most part, to pedestrians.

Most planners, architects, and even people who read these kinds of blogs would agree the last type is something we’re trying to avoid at this point. However, all of these types of neighborhoods remain popular, and all except the last still are admired by a wide range of groups, including planners, politicians, developers, and urban boosters (sources for this?). Perhaps this variety is what our society needs. It ties in closely to Richard Florida’s creative class theory – according to Florida, “successful places do not just provide one thing; rather they provide a range of quality of place options for different kinds of people at different stages in the life course. Great cities are not monoliths; as Jane Jacobs said long ago, they are federations of neighborhoods.” In short, successful cities have something for everyone. We are becoming more racially diverse as a nation, a higher percentage of people are getting married and having children later in life, if at all; some are having fewer children. Our elderly population is increasing as well due to the aging of the baby boomers. Thus, perhaps this variety is what will best serve our society as well as our cities in general.

New Urbanist developments such as Mueller are highly walkable and feature parkland and public spaces prominently. These developments tend to be isolated from other uses, effectively separate land uses, or are not designed to connect well with nearby areas, limiting their true walkability. Instead, many have a more insular approach, and appeal to those who like the idea of “community” and believe their designs add to it; most likely, many residents of New Urbanist developments prefer that their community doesn’t relate well to neighboring land. This is reflected in the housing stock in developments like Mueller, which is mostly large single-family homes, albeit with some townhomes (which are still pretty large) and an apartment building all the way at the other end of the development. These developments likely appeal more to young married couples, many of which have very young children (I live in Mueller and seem to be one of the few who isn’t married with children).

Neighborhoods like those in central Austin appeal to such young families as well, but are likely to present a more diverse housing stock, in terms of size and age, and due to this, may be more accessible and desirable to singles or older people. They are also likely to be denser than suburban areas at a variety of scales, adding to their walkability and access to public transit.

Areas like downtown seem to appeal to higher-income people, most of whom are single or at least without children. Residents of such areas can afford the high rents they must pay to live in the area’s central commercial district, and value their proximity to urban amenities, which they are able to make more use of due to far fewer demands at home. However, despite the fact that access to public transit is usually better in downtown areas than anywhere else in the metro area, downtown’s overwhelmingly high-income residents likely drive anywhere more than a few minutes’ walk away.

How much are we self-selecting into these neighborhoods (i.e. we move to them because they suit our preferences/needs)? In contrast, how much is our behavior driven by where we live?

Central to even beginning to answer these questions is the transportation/land use connection. That is, a central question is how much do our environments affect our travel patterns? In certain ones, we need to travel by car less, and less far, but do we? Another related question is, to the extent environments can’t induce less motorized travel, what percentage of people would like to be living in more dense, walkable environments, and aren’t already?

A paper by Cervero and Kockelman (1997) illuminates the complexity of the transportation-land use relationship. Cervero and Kockelman separate into factors variables based on vehicle miles traveled, and put together predictive models for VMT and transportation mode choice. Using the best travel data they can find, their work is still subject to aggregation bias, compromises to achieve decent sample size, and variables that provide imperfect descriptions of what they’re trying to measure. However, it does confirm to some extent the effect of density on VMT (as well as the effect of pedestrian-friendly design on VMT as well, but to a lesser extent). As much as anything in the predictive models, the control variables, which included some demographic variables, provided the predictive power. (Only a slight increase in r-square occurred from addition of the design, density and diversity variables.) This would support the idea of a variety of urban environments within a city – provide those who are inclined to walk, bicycle, or take transit (lower-income, younger, childless, etc.) dense, walkable environments, and don’t worry too much about traditional suburban-style subdivisions for those who will drive anyway.

Questions still remain after this short and incomplete analysis, though. First, are such environments available in terms of affordability and accessibility to all those who would fully utilize the opportunities for non-motorized travel such areas afford them? Second, for individuals that will drive anyway, are suburban subdivisions actually better, for them and for everyone else? Finally, in the words of the authors of the article, “any relationships that are covered are necessarily associative rather than casual.” How do we measure a neighborhood’s, let alone a city’s, overall success in making places accessible for everyone, while keeping factors like VMT to a minimum?

Thus, in evaluating what types of neighborhoods to build where, there are several factors to consider. What is its perceived aim (a speculation is that Mueller aims to be self-contained, with walkable streets, recreational areas and plans for additional stores and services, but mediocre connectivity to adjacent areas)? To what extent could it feasibly achieve that aim (does it truly allow for a certain travel pattern or behavior pattern, or make compromises for whatever reason)? Finally, to what extent is it likely to achieve that aim (i.e. will the behavior of its residents actually follow)?

More importantly, to find the answers to creating better cities, we must ask the right questions. That doesn’t happen nearly often enough.

Reference: Cervero, R. & Kockelman, K. (1997) Travel demand and the 3ds: density, diversity and design. Transportation Research Part D, 2, pp. 199-217.

Taking the long view on high-speed rail

In the recent debate over high-speed rail, a lot of consideration has been given to the “last-mile” problem. The idea is that most cities are not dense enough or do not offer good enough transit to allow for transportation of the vast majority of passengers needing to travel to parts of the metro area other than that immediately in the vicinity of the train station. This is a legitimate concern.

A possible solution to this problem is better interagency cooperation to allow for better intracity transit connections. My hometown of Fort Worth has combined the train station serving both Amtrak and the commuter rail between Fort Worth and Dallas with the hub/transfer point for its bus system. A similar arrangement exists in Greensboro, NC. Putting the transit hub at the rail station does open up a far greater number of destinations throughout the city for rail passengers.

This probably doesn’t adequately solve the last mile problem, however, given the poor transit service overall in most cities, Fort Worth included.

However, there’s reason to believe that land use could intensify significantly around train stations if appropriate investments in train service are made (see a good post on this at the Transport Politic, who beat me to this idea). In Fort Worth, the rail station is located at the edge of downtown, in or near what may be referred to as the “zone of discard,” near a lot of vacant lots and surface parking. This is likely the case in many cities, given the effect of rail lines on land use (i.e., their tendency, in the past, to attract relatively large warehouse and industrial land uses). Many inner cities across the country are marked by these uses, many of which have been left vacant by shrinking industries. In addition, obviously, they usually lie adjacent to a large, pedestrian-friendly downtown area full of shops, offices, and restaurants, and sometimes even residences. Thus, redevelopment is a real possibility around these stations, with the additions of hotels and office space being likely. While not taking care of the last-mile problem, it would serve to pull travelers to destinations and amenities near rail stations to high-speed rail. It might also improve transit a little, in that it would conceivably create one more area of dense activity in the inner city and encourage increased or more frequent transit connections to that locale.

Related to this, investments in rail overall may not be immediately justifiable, as Glaeser showed in his cost-benefit analysis of a hypothetical line between Dallas and Houston. However, when one takes the long view it’s quite possible they could pay off in the long run, in that they break patterns of path dependency (read a post on this at Austin Contrarian). Such patterns are why they don’t compete well with road investments right now, and may blind us to increasing returns on high speed rail further down the road.

My guess is that any cost-benefit analysis over a longer time horizon would paint a far different picture of high-speed or other kinds of rail. For instance, Glaeser has used a 30-year depreciation schedule in costing out the capital outlay necessary for his hypothetical rail line (which is to say, in his analysis, each year’s costs includes 1/30th of the total cost of constructing the line, including inflation). While this is common in the development field, it may not be the appropriate approach for what would basically be a piece of public infrastructure. Another salient point is that a train trip is far different from one by plane or car, in that almost all of the travel time is useful. This adds to its appeal for people whose time is very valuable (some of whom work or live in the aforementioned downtown areas), those with unconventional or erratic schedules or work habits (think “creative class”), or those who just want to enjoy travel more. In my opinion, Glaser pretty heavily underestimates this point (he says the need to arrive at an airport early may add an hour, and largely neglects the idea of people switching to train travel from driving).

Furthermore, high-speed rail is a step in the right direction in terms of planning on a regional scale. “Megaregions” throughout the country appear to be increasing not just in size but in importance. High-speed rail initiatives have required collaboration between leaders of metropolitan areas between states, which will likely become increasingly necessary given the number of economic and social connections between mid-size to large metro areas 200 to 400 miles apart. It may also encourage efficient economic linkages by providing another viable, stable travel choice between metro areas (for example, a company with higher-level functions in one metro area and its back-office functions in another city 250 miles away).

While it’s no godsend, high-speed rail certainly has potential to reshape travel and become a meaningful, viable alternative for significant parts of the population. In light of this, calls to cancel plans for funding high-speed initiatives seem misguided; attempting to create the best possible rail system through discussion and analysis would be far more beneficial for the country’s urban areas.