What does a house look like?

Find a little kid and ask them to draw a house. Go ahead, I’ll wait.

Finished? Alright, chances are, your kid’s drawing looks something like this:


Single-family, detached house with a pitched roof and a chimney in the middle of a yard with no neighbors in sight.

This makes sense in a lot of places; for a lot of American kids, that’s probably the only type of house they’ve ever lived in. The funny thing is, I think this image is so archetypal that even kids who live in a city like Philadelphia, where there are almost no houses that look like this, would draw something like this if asked to draw a house. So I wanted to look and see; in America’s biggest cities, what does a standard house look like?

I made a list of America’s 50 largest cities, and looked up information on housing units from the American Community Survey (ACS) conducted by the US Census Bureau. The ACS classifies housing types into ten groups, based on the number of housing units in the building: 1, detached (single-family detached homes); 1, attached (rowhouses, townhouses, and twins); 2 (duplexes); 3 to 4, 5 to 9, 10 to 19 (small apartment or condo buildings); 20 to 49, 50 or more (large apartment or condo buildings); mobile homes; and other (boat, RV, van down by the river, etc.). I included the national numbers for comparison. The results are below, and a spreadsheet with the information can be found here.

House Type Graph

Of the top 50 cities in America, 39 had a lower percentage of single-family detached housing than the nation as a whole; however, detached housing is still the predominant form of housing in all but six of the cities, and made up more than half of all housing units in 27 cities. Of the six cities where detached housing was not the primary form, there were three types.

In New York, Miami, and Washington, DC, the primary housing type is large apartment buildings such as towers. In New York this building type is very dominant, whereas in Miami it is only about 4% higher than detached housing, and in DC it is less than 1% more common than rowhouses.

Philadelphia and Baltimore are the only cities in America where rowhouses are the predominant housing type. In both of these cities, they make up more than 50% of housing units. Philadelphia has the lowest percentage of detached housing on the list, at just over 8%.

And all on its own, Boston’s predominant housing type is 3-4 unit apartment buildings. These come mostly in the form of the New England triple-decker, a three-story apartment building with one unit per floor.


The other thing I notice looking at the results is that it shows the lack of small-scale multi-family housing, also known as “missing middle” housing. This housing type is important for providing affordable housing without some of the negative consequences of the highest density forms of housing. They also provide a smoother transition between detached and high-rise forms of development, and allow the density that is necessary for mixed-use development.

So is the first image what a “house” looks like? Well, for a lot of people, yes, but not for everyone. For some people it looks like a rowhouse, or a garden apartment building, or a condo tower. But it’s important for cities to have a better mix of these types, so that there is room for anyone regardless of what sort of house they call home.

A Gondola System for Johnstown

When it comes to public transit systems, gondolas are sort of like Google Glass: they seem really cool, but no one really knows what to do with them or how to make them work. In the last decade or so they’ve been growing quickly, particularly in South America, and some places in the United States have been considering them as an alternative to more traditional forms of transit. While gondolas have certain advantages, they also come with a number of challenges, making their wider dissemination difficult. But in certain situations, such as those present in the small town of Johnstown, Pennsylvania, their unique advantages can provide better transportation service.

The first predecessor to the public transit gondola was an aerial tram system built in 1644 in Gdansk, Poland, which was used to move soil across a river to build fortifications. Early cable car systems were similarly used in mining operations, and the first people to use them for transport were probably miners. In 1893 the first aerial tram system exclusively for moving people was built in Hong Kong, and was used to transport workers to and from a mine. The first recreational cable car was built in 1907 at Mount Ulia near San Sebastian, Spain. After that development, the system was employed at other peaks throughout the alps, and from there became a mainstay at ski resorts around the world.


One of the earliest systems designed to be used for urban commuters was the Roosevelt Island Tramway, built in New York City in 1976. It was intended as a temporary connection between Roosevelt Island and Manhattan until a subway connection was completed, but due to delays in building the subway, it became a permanent fixture. While the Roosevelt Island Tramway and other aerial tram systems like it have a fairly high capacity and can move fairly quickly, they only have two passenger cabins (leading to less frequent service), their design only allows for two terminals and no intermediate stations, and their lines can’t turn. All of this adds up to their application as a form of urban transit being fairly limited.


From Jim Henderson via wikipedia.org.


The first true gondola public transit system, and to this day probably the most famous, is the Metrocable of Medellín, Colombia. Since 2004, Medellín has built three gondola lines (with two more in the works) that connect to the Metro system and run up into the barrios on the steep hillsides of the Aburra Valley. These barrios are so steep and so dense that regular buses simply couldn’t reach them, and residents were commuting over two hours by foot each way to work. Gondolas are able to travel over the community, rather than through it like a bus on a road would, so the development pattern below isn’t a problem. Using gondola technology rather than the older aerial trams also allows for intermediate stations and allows the line to turn, giving it much more flexibility both in its geometry and in how it serves the residents. Since it’s implementation, Metrocable has inspired dozens of similar systems in South America, Africa, Asia, and Europe, as documented by the Gondola Project.


From Steven Dale via flickr.com.


Metrocable illustrates two situations where gondola public transit is particularly strong: extreme topography and irregular street networks. Gondolas aren’t for everyone; according to this paper by Baha Alshalalfeh, et al, they cost more to build than a standard bus system and have a much lower capacity than more expensive forms of transit. This is one reason why some large system proposals such as that in relatively flat and gridironed  Austin may not work; for the same cost, you could build a tram system that could move about six times as many people.

Johnstown, in a number of ways, is not Austin. If you’ve heard of Johnstown at all, it’s probably because of the 1889 Johnstown Flood, the greatest single-day loss of civilian life in America before 9/11 and the source of it’s unfortunate moniker “Flood City” (though subsequent devastating floods in 1936 and 1977 didn’t help). Johnstown unfortunately has the perfect topography for severe flooding: steep mountainsides above and narrow river valleys with small pockets of flat, developable areas (many of them built on fill) below. Johnstown is divided into several somewhat discontinuous areas in the valleys along the Conemaugh and Stony Creek Rivers, as well as a few on the tops of the plateaus above. the discontinuous nature of the developable area in Johnstown makes navigating the city quite complicated.


Johnstown Terrain. Base from Google.


If we were to design a gondola transit system for Johnstown, our first step would be to identify which areas are dense enough to support transit. We can do that by counting the number of housing units in a given block group and dividing that by the area of the block group.


What we can see is that there are many areas in Johnstown dense enough, according to the Victoria Transport Policy Institute,  to support an occasional commuter form of transit, as well as a few (Downtown, Morrellville, and Moxham) that are dense enough to support some form of local bus. These should be the major hubs for our transit system.

The next step is to identify coverage. While there is a lot of variability when it comes to distances between gondola stops, in general they are about half a mile apart. We can calculate the center point for each block group and then measure a half mile radius from that center point.


From here we can take the block groups with the highest density and eliminate the other center points within a half mile radius from the list of potential gondola stops. As we do this, we arrive at the following system.


Alshalalfeh argued that gondola systems have similar capacity to bus transit, meaning that these lines could probably move the same amount of people that are currently using buses in this area quite easily. In addition, the ability to move in a straight line rather than following the circuitous routes that buses have to take saves a lot of time. For instance, a bus traveling from Oakhurst to downtown today would take about 18 minutes, while a standard MDG gondola moving at about 6 meters per second could do it in just 12. So while you are moving about the same amount of people, you are able to do it much faster in a gondola.

The entire system is about 5.8 miles long, and at a cost of about $8-16 million per mile, you’re looking at a total cost of $45-90 million for the entire project. While that is a lot more than you would pay for a bus system, it is cheaper than what you would pay for any other form of advanced transit. And while you would have a lot of land to purchase for most other forms of surface transit, you would only have to purchase the land for the stations and possibly an easement for the intermediate towers. Since Johnstown is a rustbelt city, it already has a lot of vacant or underutilized land which could be used to site stations.

While there are several technical reasons that a gondola system could work in Johnstown, there is also the intangible reason that it would just be really cool. Johnstown is a lovely little town tucked into some very dramatic scenery. There’s a reason that gondolas were used for recreational purposes before they were used as a means of transportation, and that is because the views from the gondolas are amazing. And while the scenes from the stops along the valley from Ferndale to Downtown would be great, the view as you go from Morrellville to Downtown would be breathtaking. If you maintain a straight line between the two stations, you would go over a rise that towers 400 feet above the valley floor. If the “Gringo Problem” that Medellín has experienced is any indication, people may travel to Johnstown just to ride the gondola over that peak.

This slideshow requires JavaScript.

Gentrification and Market-based Zoning

A few months ago we moved back to Philadelphia from DC. Washington is a very nice place, but we only make a little bit more than the average American household (which makes about $52,000 a year), while the average income in DC is about $90,000, which means that we couldn’t afford anything. For instance, in Philadelphia in 2013, we lived 1.4 miles from the center of town for $1,200 a month. When we moved to DC, we lived 4.5 miles from the center of town for $1,350 a month. There were several reasons why we chose to move back, but largely, it was because


When we came back to Philly, we knew that we wanted to live south of Market Street, and beyond that, we didn’t care too much. We spent several weeks looking all over South Philadelphia and toured a number of apartments including a very nice one in Point Breeze. Several neighborhoods in Philadelphia are experiencing rapid gentrification, but Point Breeze is in many ways at the forefront. If you don’t know what gentrification is, it’s basically when rich(er) people start moving into a poor(er) area, redeveloping it and driving up rents and property taxes, and driving out existing residents. Though this is strictly speaking an economic issue, because minorities in America tend to live in poorer neighborhoods, it often affects them disproportionately.

Planners, architects and developers have a mixed relationship with gentrification, and so do I personally. I mean, I don’t want to hurt poor minorities, but I also want access to affordable housing close to where I work (by the way, we ended up deciding against Point Breeze; we found a cheaper apartment in South Philly). But why do I have to go to neighborhoods like Point Breeze to find affordable housing in the first place? Why can’t I find housing close to my office in Center City? The obvious answer is that because it is too expensive, but why is that the case?

An interesting argument was presented in this article by Kriston Capps of Citylab. Capps points out that tech, a common boogeyman in the discussion of gentrification in San Francisco, is actually not the problem; it’s anti-development/NIMBY residents of rich neighborhoods. These folks hold much more influence in City Hall than their poorer neighbors, so they can get zoning ordinances and other restrictions passed that keep development from happening in their back yards. But people still want to move to San Francisco, developers will still build new housing to provide for them, and since they can’t do it in the rich neighborhoods where they actually want to live, they develop in the closest poor neighborhood. John Mangin, in his article The New Exclusionary Zoning, says, “Don’t blame in-movers or developers for gentrification—they’d rather be in the high-cost neighborhoods. Blame the exclusionary practices of people in the high-cost neighborhoods.” Mangin argues that in addition to zoning, the lengthy approval processes required by many desirable cities increases the cost of building housing, as well as increasing the time to develop it, and privileges large, savvy, politically-connected developers over smaller neighborhood builders.

Capps proposes that development should be expanded in richer neighborhoods, amending zoning laws and making decisions that are best for the city (or the region) and not necessarily for individual neighborhoods. Mangin argues that we need to pursue policies that either increase the supply of housing or decrease the demand for it in poorer neighborhoods. This includes allowing some development (because not allowing any development simply drives up the prices of the existing housing stock until they cost too much for the current residents and are bought up by more affluent move-ins), while at the same time advocating for more development in the high-demand areas where rich NIMBYs are keeping new folks from moving in. Governments could also impose regulations on new development that would help existing residents, such as requiring some of the taxes assessed on new development to go toward investing in the neighborhood; or creating something of a cap-and-trade market for density, where if people want to exclude development from their neighborhood, they have to help pay for it to happen elsewhere.

What is argued by both of these authors is that zoning is out of sync with housing demand, particularly in rich areas, leading to spillover of new move-ins in poorer neighborhoods. So what would it look like if a city’s zoning were rewritten to reflect market demand? Let’s take a look at Philadelphia.

Below we have the existing housing density, in units per acre, of the Philadelphia Metropolitan Statistical Region by census tract. Density is concentrated in the City of Philadelphia, as well as some of it’s suburbs, particularly the string running southwest through Delaware County down to Wilmington, Delaware, and beyond; and running southeast through Camden County, New Jersey. (The land use categories shown correspond to the densities required for different types of transit service, according to the Victoria Transport Policy Institute: 0-2 supports no transit, 2-4 supports regional rail, 4-7 supports minimal local bus, 7-9 supports intermediate local bus, 9-12 supports light rail, 12-15 supports rapid transit, and over 15 supports frequent local bus service)

Density Existing-01

This is our baseline, and lets us know how many housing units already exist in an area. The next thing we need to know about a tract is the median housing value.

Median Value-01

This shows us that housing value is highest along the Main Line; in some of the suburbs of Wilmington; and around New Hope, PA and Moorestown, NJ. Now that we have the median value, we can multiply that by the number of housing units to get an idea of the total value of housing in a tract, which will indicate to us the demand for housing in that area.

Total Value-01

We see a similar pattern to median value, as well as a slight concentration in Center City, Philadelphia (while the median value is a bit lower than in some of the rich suburbs, there are so many more units that the total value is quite high). Now that we have a measurement for the demand, we need to translate that back into housing units. The average value of a home in the United States is $175,700, so we can divide the total value of each tract by that number to get an idea of how many units that area should supply.

New Density-01

What we can see from this is that many of the suburbs, particularly in affluent Montgomery County, are not pulling their weight, and should take on a greater share of the region’s new development. Housing density in Philadelphia remains high; however, taking a look at the change in units per acre reveals some interesting patterns.

Change-01Philly Change-01

Moving from the edge of the region to the center, we see that some of the outermost communities could actually afford to lose a few units per acre. However, by and large, density should increase as you get closer to the city, particularly along the affluent Main Line to the west. At the same time, there are several small cities in the area that are overbuilt, including Pottstown, PA, Norristown, PA, Upper Darby, PA, Darby, PA, Chester, PA, Camden, NJ, Salem, NJ, and Wilmington, DE. As we look at Philadelphia itself, the neighborhoods of Northeast and South Philadelphia should grow modestly, while the more affluent neighborhoods in the northwest should grow more steeply. This is followed by a ring of overbuilt areas in North and West Philadelphia, as well as Point Breeze and parts of South Philly. However, Center City and University City are underdeveloped, and should grow considerably. This underdevelopment is what is fueling the gentrification of areas where the green and the red meet, such as Point Breeze, Mantua, and Kensington.

Percent Change-01Philly Percent Change-01

It is also interesting to look at the percent change, rather than total units per acre, to tell you something about the degree to which these neighborhoods will be affected by change. Some areas, such as Piedmont, DE, and New Hope, PA, would only see a modest change in the absolute numbers, but because their existing densities are so low it may feel like a large shift. Others, such as Center City, would see significant growth in total units per acre, but because their existing density is already high it will not have as significant an impact on the area. And while there was a lot of red on that first map, the area most impacted by a decline in density is actually more limited when you look at the percentage, which the heaviest impacts in Norristown, Darby, Chester, Camden, and North and West Philadelphia.

So, who would support a plan like this and who would oppose it? The most obvious answer is that rich homeowners, both in the city and the suburbs, may not take kindly to a plan like this. They would see this, not entirely inaccurately, as a threat to their property values and their way of life. Mangin advocates several “smaller scale reforms that preserve a space for sub-local [neighborhood] politics while altering, sometimes subtly, the incentives that political actors face and the procedures by which they arrive at decisions,” to try and get richer residents on board as much as possible.

Initially, many residents of poorer neighborhoods might also oppose it, because zoning by demand would mean severely downzoning several poorer areas, which may look to some residents like “benign neglect” or, even worse, the sort of problems that arose with urban renewal and the use of eminent domain in the middle of the last century. This sort of “depletion” or “neighborhood triage” has been strongly opposed by neighborhood groups who want to preserve their communities and see it as a method for removing poor residents to make room for future development. This sort of opposition was seen in the defeat of the “Team Four Plan” in St. Louis (If you want to pay for it, you can read Patrick Cooper-McCann’s recent article on it in the Journal of Planning History here, or if you’re a cheapskate like me you can get the gist of the article by reading his master’s thesis here for free). It would be important to implement this strategy in steps, such that new housing in desirable areas was available early so that anyone who wished to move out of poorer neighborhoods may have an opportunity to do so, while those who wished to stay behind could do so, safe in the knowledge that restrictive zoning would prevent new development from infringing on their community while they would be allowed to stay there as long as they wished.

The people that a plan like this would really be good for would be middle-class people wanting to move in from outside the city or to move up to a more desirable neighborhood. As it is now, the urban middle class is squeezed between neighborhoods they can’t afford and neighborhoods where they are seen as unwanted agents of change and distress. Opening up more development opportunities, both in Center City and in densified suburbs along the Main Line and elsewhere, would provide for more opportunities for affordable urban living without the guilt of hurting those lower down on the economic ladder.

Gentrification is a hard nut to crack, but it’s important to look at it as a problem of restricted housing supply in affluent areas not being able to meet the demand for development, which then spills over into less affluent neighborhoods. Changing our zoning laws to better reflect the demand for housing in desirable neighborhoods would help ameliorate gentrification and allow more options for middle-income families in cities. While efforts like this would face an uphill battle against entrenched interests, bureaucratic roadblocks, and NIMBYism, in the words of John Mangin, “The options are pretty clear: build more, or stand by as low-income and middle-class people get priced out of ever-wider swaths of the country.”

Walkable Does Not Necessarily Mean Big

People I talk to about urbanism tend to think that I’m a “city person.” and I can see why they would think that, since I eventually learned to love Philadelphia, live in DC (okay, Arlington, but I would live in DC if I could afford it), and generally disdain suburbs. But people who know me better know that New York or Los Angeles is not my ideal. When I think of a perfect place, the one that made me want to be an urban designer and the one I would like to replicate in my work, I think of Northampton, Massachusetts.

From ictir2015.org.

Northampton isn’t big. It’s population is approximately 28,592, and the way that towns are set up in Massachusetts, that number includes a lot of people who live out in the countryside and not “in town.” But even though it isn’t big, Northampton feels urban, because you can walk to anything you would need on a daily basis and could live quite comfortably without owning a car.

There is a strong correlation between a place feeling urban and it having a high Walkscore. I’ve mentioned Walkscore before, but to sum it up, it is a measure of how easily one can reach their everyday needs on foot. It goes from zero to 100, and a score below 50 being car-dependent, 50-69 being somewhat walkable, 70-89 very walkable, and over 90 a walker’s paradise.

Parts of Northampton are walker’s paradises, as were all the neighborhoods in Philadelphia that I lived in and all the neighborhoods in DC where I would live if I could afford it. I decided to look and see where one could find walker’s paradises, so I searched the whole country for apartments with a Walkscore over 90 (the apartments are important because no matter how many shops and restaurants you have, if no one can walk to them from their home, you essentially have a mall). I mapped the results, noting that many places may have an apartment building or two with a Walkscore of 90 while the neighborhood as a whole is below that, and that other places are “true” walker’s paradises, where the entire neighborhood has a Walkscore above 90.

Click to enlarge.

Click to enlarge.

“Now wait a second,” you might be thinking, “New York is the biggest dot!” And that’s true, but New York is so big that it has the most of many things, including walkable neighborhoods. What’s important is that Los Angeles, the second biggest city in the United States, is not the second biggest dot, nor is Chicago, Houston, or any other city larger than the one that actually is second biggest, San Francisco. In fact, I think Houston is the best example of how big and urban/walkable are not the same thing. Houston, despite its population of 2,239,559 and its size of 627.8 square miles, only has three walker’s paradises, none of which are “true” walker’s paradises. This means that in urbanism terms it is not the equivalent of Chicago (population 2,695,598 with 17 walker’s paradises), but of Lawrence, Massachusetts (population 77,657 with three walker’s paradises).

CorrelationIn fact, as the graph shows, population explains about 60% of how walkable a place is. While a large city does allow for more services, it’s size has nothing to do with how those services are laid out, which has a huge impact on how urban a place is. That is why San Francisco (second highest on the graph above) is so walkable, even more so than simple population projections would predict, while Los Angeles (second furthest to the left on the graph) is actually less walkable than one would project a city of its size to be. San Francisco was built around the pedestrian and the streetcar; Los Angeles was built around the automobile.

So small cities, don’t think that you can’t be great urban places just because you’re not very big. Great urbanism comes from putting the pedestrian first, from planning great streets with a mix of housing, working, and services, and from making a pleasant and vibrant environment for people. Make these a priority and you will be urban, regardless of size.

Finding the Best Country in the World

I’m a bit paranoid. Like when I get home, stick my head in the door, look around, and tell my wife, “Yep, no racoons,” I’m only sort of kidding (a racoon did break into my house when I was a kid so that’s not a totally crazy thing to say). One way this paranoia manifests itself is in creating certain contingency plans. For instance, if I had to flee the country where would I go? The easy answer would be the other English-speaking countries (Canada first because it’s close, the UK second because it’s further but I have friends there, New Zealand third because it looked nice in Lord of the Rings and Australia last because I’m sorry but I do not trust your fauna). But what if, say, all the English speaking countries got nuked or something? Or if we find out that global warming only affects them and they get wiped off the face of the earth? Or they all get conquered by North Korea? Well, if that’s the case, it’s time to flee to a non-English speaking country and learn another language. But which one?

At one point I couldn’t decide, and I was studying Danish, Dutch, Irish, Spanish and Swedish on Duolingo all at the same time (yes, I know next to no one speaks Irish, that was just for kicks). Eventually I cut out Irish and Spanish, but also picked up German, so I wasn’t much better off. So I wanted to figure out, if I’m going to have to flee to another country, including non-English speaking countries, which ones should I focus on so that I can actually study their language effectively?

I thought the Human Development Index (HDI) was a good place to start. It measures a country’s well-being based on a combination of several factors, including life expectancy, education and income. Its top five countries are Norway, Australia, Switzerland, the Netherlands, and the United States. However, HDI doesn’t measure everything. In fact, in 2010 they introduced a new index, the Inequality-Adjusted Human Development Index (IHDI) to try and fill in some of the gaps. This new adjustment changes the top five to Norway, Australia, the Netherlands, Switzerland, and Germany, and kicked the US down to 28th, three spots below Greece.

My thought was that the best way to cover as many factors as I could was to combine as many different indices as I could to create one super index that would bring in enough factors to create a truly well-rounded indicator of what the best country was. I combined the IHDI, Democracy Index, Human Poverty Index, Social Progress Index, World Happiness Report, Global Peace Index, Legatum Prosperity Index, Where-to-be-born Index, Satisfaction with Life Index, and OECD Better Life Index to create what I call the Best Country Index.

Index Index2

So the top five are Norway, Sweden, Switzerland, Denmark, and Finland. Well, right now Norwegian isn’t offered on Duolingo, so even if it’s number one I’m not studying it until they finish creating the course for it, so for now I’ll just have to assume it’s somewhere between Danish and Swedish. And I’m not even going to try with Finnish; I had a hard enough time with Irish, and that’s still an Indo-European language, I can only imagine the sort of trouble I would have learning a Finno-Ugric one. So that leaves me (until Duolingo adds Norwegian) with Swedish, German, and Danish. Well, at least I can stop studying Dutch for now.

For those of you who would like a searchable version of the list above, here it is:

  1. Norway, 2.9877
  2. Sweden, 2.9732
  3. Switzerland, 2.9724
  4. Denmark, 2.9702
  5. Finland, 2.8872
  6. Canada, 2.8858
  7. Netherlands, 2.8778
  8. Australia, 2.8257
  9. New Zealand, 2.7938
  10. Austria, 2.7792
  11. Germany, 2.6754
  12. Ireland, 2.6643
  13. Iceland, 2.6580
  14. Belgium, 2.5993
  15. United Kingdom, 2.4850
  16. United States, 2.4315
  17. France, 2.3879
  18. Japan, 2.3759
  19. Spain, 2.3667
  20. Slovenia, 2.2898
  21. Czech Republic, 2.2789
  22. Italy, 2.2484
  23. Costa Rica, 2.2419
  24. Chile, 2.1414
  25. Luxembourg, 2.1216
  26. South Korea, 2.1083
  27. Singapore, 2.0700
  28. Poland, 2.0554
  29. Slovakia, 2.0191
  30. Portugal, 2.0180
  31. Israel, 2.0131
  32. United Arab Emirates, 1.9953
  33. Argentina, 1.9745
  34. Uruguay, 1.9719
  35. Kuwait, 1.9342
  36. Cyprus, 1.9248
  37. Panama, 1.9047
  38. Estonia, 1.8984
  39. Hungary, 1.8743
  40. Croatia, 1.8356
  41. Malaysia, 1.8159
  42. Taiwan, 1.8096
  43. Malta, 1.8054
  44. Greece, 1.7843
  45. Brazil, 1.7838
  46. Trinidad and Tobago, 1.7417
  47. Mauritius, 1.7076
  48. Hong Kong, 1.7074
  49. Mexico, 1.6853
  50. Latvia, 1.6692
  51. Lithuania, 1.6575
  52. Saudi Arabia, 1.6445
  53. Colombia, 1.5608
  54. Romania, 1.5560
  55. Bulgaria, 1.5293
  56. Mongolia, 1.5285
  57. Jamaica, 1.5272
  58. Venezuela, 1.5111
  59. Serbia, 1.4994
  60. Dominican Republic, 1.4682
  61. Thailand, 1.4682
  62. Qatar, 1.4612
  63. Indonesia, 1.4586
  64. Ecuador, 1.3933
  65. Montenegro, 1.3921
  66. Paraguay, 1.3908
  67. Botswana, 1.3800
  68. Peru, 1.3728
  69. El Salvador, 1.3698
  70. Brunei, 1.3360
  71. Bhutan, 1.3356
  72. Philippines, 1.3135
  73. Bahamas, 1.2969
  74. Suriname, 1.2959
  75. Nicaragua, 1.2713
  76. Guatemala, 1.2650
  77. Sri Lanka, 1.2616
  78. Albania, 1.2463
  79. Vietnam, 1.2340
  80. Macedonia, 1.2286
  81. Cuba, 1.2284
  82. Guyana, 1.2281
  83. Bolivia, 1.2224
  84. Namibia, 1.2217
  85. Kazakhstan, 1.2140
  86. Tunisia, 1.2068
  87. Bosnia and Herzegovina, 1.1818
  88. Oman, 1.1790
  89. Turkey, 1.1783
  90. South Africa, 1.1738
  91. Antigua and Barbuda, 1.1625
  92. Belarus, 1.1555
  93. Jordan, 1.1534
  94. East Timor, 1.1433
  95. Moldova, 1.1385
  96. Seychelles, 1.1310
  97. St. Kitts and Nevis, 1.1231
  98. Honduras, 1.1190
  99. Ghana, 1.1158
  100. Liechtenstein, 1.1140
  101. Uzbekistan, 1.0997
  102. Vanuatu, 1.0995
  103. Barbados, 1.0759
  104. Kyrgyzstan, 1.0725
  105. Dominica, 1.0601
  106. Morocco, 1.0518
  107. Belize, 1.0164
  108. Fiji, 1.0143
  109. Ukraine, 1.0137
  110. China, 1.0126
  111. Georgia, 1.0109
  112. Bahrain, 0.9991
  113. Cape Verde, 0.9768
  114. St. Vincent and the Grenadines, 0.9734
  115. Russia, 0.9629
  116. Andorra, 0.9617
  117. India, 0.9617
  118. Armenia, 0.9466
  119. Bangladesh, 0.9446
  120. Zambia, 0.9422
  121. Senegal, 0.9332
  122. Azerbaijan, 0.9238
  123. St. Lucia, 0.9183
  124. Lebanon, 0.8938
  125. Gabon, 0.8898
  126. Nepal, 0.8886
  127. Samoa, 0.8789
  128. Papua New Guinea, 0.8788
  129. Tanzania, 0.8685
  130. Algeria, 0.8633
  131. Cambodia, 0.8305
  132. Maldives, 0.8239
  133. Lesotho, 0.8086
  134. Iran, 0.8053
  135. Laos, 0.7953
  136. Madagascar, 0.7853
  137. Kenya, 0.7636
  138. Benin, 0.7603
  139. Tajikistan, 0.7579
  140. Solomon Islands, 0.7452
  141. Tonga, 0.7371
  142. Sao Tome and Principe, 0.7316
  143. Grenada, 0.7213
  144. Egypt, 0.7050
  145. Palestine, 0.6883
  146. Malawi, 0.6800
  147. Mozambique, 0.6666
  148. Burkina Faso, 0.6468
  149. Congo, 0.6432
  150. Uganda, 0.6267
  151. Mali, 0.6129
  152. Angola, 0.6028
  153. Liberia, 0.5733
  154. Sierra Leone, 0.5716
  155. Libya, 0.5652
  156. Cameroon, 0.5604
  157. Haiti, 0.5543
  158. Gambia, 0.5511
  159. Djibouti, 0.5483
  160. Mauritania, 0.5472
  161. Nigeria, 0.5384
  162. Togo, 0.5369
  163. Swaziland, 0.5051
  164. Pakistan, 0.5007
  165. Rwanda, 0.4921
  166. Comoros, 0.4904
  167. Iraq, 0.4849
  168. Yemen, 0.4505
  169. Kiribati, 0.4407
  170. Turkmenistan, 0.4333
  171. Niger, 0.4307
  172. Syria, 0.4233
  173. Guinea, 0.3683
  174. Ethiopia, 0.3662
  175. Equatorial Guinea, 0.3662
  176. Cote d’Ivoire, 0.3357
  177. Burma, 0.3104
  178. Zimbabwe, 0.2542
  179. Burundi, 0.2502
  180. Guinea-Bissau, 0.2489
  181. Eritrea, 0.1878
  182. Afghanistan, 0.1861
  183. Central African Republic, 0.1709
  184. Chad, 0.1657
  185. Sudan, 0.1458
  186. Democratic Republic of the Congo, 0.0675
  187. Somalia, 0.0205
  188. North Korea, 0.0169

Where I Should Live, According to Math

I don’t live in Washington, DC. I live near Washington, DC. I would like to live in it, but it’s an expensive city, and my income, while above the national average, is well below the regional average, and finding a two bedroom in our price range is difficult.

This got me thinking about affordable housing more broadly. For instance, where could I find a good, walkable neighborhood, anywhere in the country, that is within my price range? That got me started on my current project.

Using census data, I decided to map the variables of affordability and walkability. Affordability wasn’t hard; I mapped all the census tracts in the country that had a median income within $10,000 of mine, both above and below.Affordability

This shows me where I can afford to live, but a lot of the areas are rural places that I would never want to live in. My next task was to map walkability.

Walkability was harder to map. Even though Walkscore covers everywhere in America, it only offers it’s data in downloadable form for Washington, DC. So I downloaded the data, calculated the average Walkscore for census tracts in DC, downloaded virtually the entire American Community Survey, and compared the data therein to the average Walkscore to look for correlation. I found nineteen variables that had some significant correlation with Walkscore.VariablesI took each of these variables and gave them a score of one or zero, one if Walkscore would be above 70 at the value, or a zero if it would be below. Then I multiplied that score by each variables’ R-Square value, and added all the variables together to get a weighted Walkability score. I eliminated the bottom 50% of these values, and added the remainder to the map.Walkability

I was pretty happy with the result. With the exception of a few large tracts in western states, walkable places are where you would expect them to be; densely concentrated around major metropolitan areas

I intersected the two layers to get tracts that were both affordable and walkable.Intersected

This led to an interesting pattern: a few small, walkable town centers on the edge of metropolitan areas, but mostly urban neighborhoods outside of the downtown or in inner-ring suburbs.

However, it was still too many places to look at as a group, so I assigned a score to each tract based on how walkable and how affordable they are. I added these two together to get a combined score for what neighborhood would be best for us, based on these two criteria. In case you wanted the full equation for this score, it is

Combined score = (a – |a – b|) / a + ((if(c ≥ 373.6958, 1, 0) * 0.3153) + (if(d ≥ 21.2983, 1, 0) * 0.2725) + (if(e ≤ 38.8903, 1, 0) * 0.2803) + (if(f ≥ 68.0899, 1, 0) * 0.2971) + (if(g ≥ 67.4557, 1, 0) * 0.3350) + (if(h ≥ 59.9592, 1, 0) * 0.4048) + (if(i ≤31.4668, 1, 0) * 0.2529) + (if(j ≥ 65.5846, 1, 0) * 0.2734) + (if(k ≥ 65.3918, 1, 0) * 0.2839) + (if(l ≥ 58.6467, 1, 0) * 0.3533) + (if(m ≤35.7247, 1, 0) * 0.2576) + (if(n ≥226.8280, 1, 0) * 0.2763) + (if(o ≥78.1848,1, 0) * 0.2779) + (if(p ≥3.8273, 1, 0) * 0.2943) + (if(q ≥ 602.4307, 1, 0) * 0.2795) + (if(r ≤ 4.1293, 1, 0) * 0.2698) + (if(s ≥ 732.9079, 1, 0) * 0.2573) + (if(t ≥ 21.1155, 1, 0) *0.3974) + (if(u ≥82.4877, 1, 0) * 0.2810)) / 5.6596


a = Your Personal Income

Data for Each Tract from the American Community Survey:

b = Median Income

c = Nonrelatives in Household

d = % with at Least a Bachelor’s Degree

e = % Born in State of Residence

f = % 16 and Older in Labor Force

g = % 16 and Older in Civilian Labor Force

h = % 16 and Older Employed in Civilian Labor Force

i = % 16 and Older Not in Labor Force

j = % Females 16 and Older in Labor Force

k = % Females 16 and Older in Civilian Labor Force

l = % Females 16 and Older Employed in Civilian Labor Force

m = % 16 and Older Driving to Work Alone

n = Workers 16 and Older Walking to Work

o = Workers 16 and Older Commuting to Work by Other Means

p = % 16 and Older Commuting to Work by Other Means

q = Houses Built 1939 or Earlier

r = % 10-14 Years Old

s = Population 25-34 Years Old

t = % 25-34 Years Old

u = % 18 Years and Older

So, what got the highest score?

Capital Hill-01Capitol Hill, Seattle led the pack. To be honest, I was expecting something like a smaller, affordable Midwest town or something, but it the highest scoring areas were usually just outside of major downtowns. Other top areas included Cambridge and Somerville outside of Boston, and the South End in Boston; Columbia Heights, Washington, DC; The Mission District, Lower Haight, and Russian Hill, San Francisco; Midtown, Atlanta; Greenwood, Dyker Heights, Kensington, and Sheepshead Bay, Brooklyn; Graduate Hospital in Philadelphia, where we used to live; Lake View, Chicago; and Five Points, Denver.

Holly and I won’t be moving out of the region any time soon, but it’s good to have some idea of where to look if we decide to. And good to know that Columbia Heights is probably the neighborhood in DC for us, when the time comes. The formula isn’t perfect; it’s hard to control for things like how much of people’s income goes toward housing, and there is still a lot of wiggle room in these walkability measures. But it is a reasonable guideline that has provided interesting results.

UPDATE: I’ve gotten a few special requests for specific data on various areas, so I decided to make a little gallery of them below.

Why No One Rides the Train in Phoenix

Phoenix’s new TOD districts. From theatlanticcities.com.

I saw this article from Eric Jaffe a few days ago and it sent me on a bit of a journey that I hope you will find interesting. In the article, Jaffe discusses how Phoenix, a city which would not exist if not for massive water projects, cheap housing, and abundant air conditioning, is addressing the fact that nobody wants to ride its fancy light rail train. While light rail projects in Salt Lake City and Denver have been overwhelmingly successful, Phoenix’s venture into improved transit has languished. So the city has organized a series of five districts along the corridor and gotten federal funding to promote transit-oriented development and find out why no one is riding.

There are a couple of reasons. Unfortunately, Phoenix’s light rail line opened in 2008. Yes, that 2008. And boomburgs like Phoenix took the hit even harder than the rest of us (which is partially why Philly was able to take back the number five spot from Phoenix in the largest cities in the US after the 2010 census). But there are deeper-rooted issues that have kept TOD from blossoming like a desert rose, and those are principally density and urban design.

Light rail, although more affordable than a subway or elevated train system, is still quite expensive, and needs a certain concentration of people to be viable. According to John Renne, that magic number is nine dwelling units per acre. As you can see in the above map, very little of Phoenix meets this threshold. While Phoenix does have over 1.4 million people, those folks are sprawled out over more than 500 square miles. I wonder if Phoenix planners were hoping that the light rail system would encourage density to develop along the corridor, but really they should have planned the density in before installing the rail.

The other main issue is that Phoenix is in the middle of the desert, and is quite an unpleasant place to be outside in. And unfortunately, the development patterns of the city have only exacerbated the issue, creating a heat island effect over a huge area. This happened because Phoenix is an entire city designed like a post-war American suburb, which was designed largely for the moderate climates of higher latitudes and not the blistering heat of Maricopa County.

But it’s not like cities have never been built in deserts before. The arid climates of Mexico, Northern Africa, and the Middle East have hosted hundreds of large cities, many larger than Phoenix and some that have been around for thousands of years. But these cities weren’t designed the was Phoenix was, and their adaptations allowed them to beat the desert heat.

The name of the game in desert urban design is shade and wind. After combing through over a hundred cities and comparing them to Phoenix based on climate, World City status, population, location characteristics, and presence of rail transit, I came on two examples which are particularly apt examples: Dubai, UAE and Monterrey, Mexico. Both cities share a climate zone with Phoenix (which stradles the boundary between Hot Desert and Semi-Arid climate classifications), are both World Cities of higher rank than Phoenix, have similar populations (with Dubai at 2.1 million and Monterrey at 1.1 million), both have extensive rail transit systems, and Monterrey shares Phoenix’s interior location (whereas Dubai is on the coast). Let’s take a look at these cities and see what Phoenix can learn.

Dubai is an Medieval Arabian seaport. Whereas many Arab/Muslim cities are built around a dense medina, where the widest streets are often only spacious enough to accommodate two passing camels, Dubai, which up until recently had more money than you could shake all the sticks at, adapted its center, building larger buildings and a system of streets that serve cars while still allowing for pedestrian paths within blocks. The tall buildings and narrow passageways create shade which is intensified in some areas by souks, or covered pathways, which often are a gathering place of commercial activity. Many of the buildings are also light neutrals in color, which helps to reflect heat, rather than absorb it like dark materials, such as asphalt.

Monterrey is a Mexican city of Spanish imperial origin, and as such its urban design was based on the Law of the Indies, which allowed for rapid expansion in a roughly gridiron pattern. While not as tortuous as the Arabian medina, the streets of Monterrey are relatively narrow, and the buildings make extensive use of awnings, window covers, and other shade devices, while the public realm is full of trees and shade structures.

While both of these cities feature some large roads with high-speed traffic, these streets are the norm in Phoenix, even along the light rail line. Most of the roads it parallels have four or six lanes of traffic, and widen at corners with dedicated turning lanes. In many cases, buildings along the roads are set back behind a parking lot. There are even numerous vacant lots along the rail line. These conditions need to change.Web

First, narrow everything up as much as possible. Remove the buffers around the trains and let cars come right up next to the curbs. These buffers are “necessary” when you have high-speed traffic, but that’s the last thing you want on a pedestrian/transit corridor. Additionally, remove the dedicated turn lanes at intersections. They are there to make life easier for cars, and increase the distance and time necessary for pedestrians to cross the street. Screw the cars. Put the pedestrians first. Next, narrow the travel lanes to ten feet, and never have more than four. Again, this road is for pedestrians and transit users, not cars. Include a row of on-street parking, which buffers pedestrians from traffic and further encourages drivers to slow down. Include buffered bike lanes on the outside of the parking lane, so cyclists don’t have to compete with cars and can avoid being doored. Provide shade on the sidewalk with either low-maintenance shade structures or trees (I personally love the local Palo Verde trees). Bring the buildings right up to the sidewalk, and put the parking behind. In these areas, you want people to be able to see your storefront window, not your ample free parking.

Now that we’ve scrunched everything together, it’s time to go up. Buildings should be, at the very least, two stories, to allow for vertical mixed uses. When it comes to creating a sense of enclosure on a street, the bare minimum ratio of height to street width is 1:6. 1:2 is better. 1:1 is probably best. Beyond that you start getting into the enclosure territory that only a New Yorker can love.

Building height can also be manipulated to create microclimates and cool an area down. For instance, if you have taller buildings on the south side of a street, the shadow they cast to the north will keep the street cooler. Since Phoenix is at the fairly low latitude of 33.5 degrees north, the sun is only ten degrees south of directly overhead in the summer. So to create shadows, especially across a wide street, you need some really tall buildings. Of course, if you narrow the street, you can get by with smaller ones. Tall buildings on one side of a street can also catch winds and force them to ground level. Since the prevailing winds in Phoenix blow pretty much due east, putting tall buildings on the east side of north-south running streets would be the best way to catch and divert wind.

Finally, architectural details can do a lot to make being in Phoenix more pleasant. Make extensive use of awnings and other shade structures, as even small Casa Grande has done in their downtown. Although you want a lot of visible glass on the ground floor for stores, consider screens and shades for upper floors. Make sure windows can open to allow for natural ventilation. The principle colors of a building should be low albedo, to avoid heat gain, but that doesn’t mean bright colors can’t be used for accents.

Phoenix’s light rail has a lot of things going against it right now, but if the city can learn something from its desert ancestors and take advantage of density and urban design, then they can create a wonderful transit-oriented corridor that will breathe new life into the city.

%d bloggers like this: