Sunday, February 23, 2014

Cross Border Urbanism: From Texas to Tamaulipas

Although the American and Mexican urban traditions have at various times shared some fundamental similarities, including a heavy use of orthogonal grids and an apparent tendency toward single-family homes sitting on their own lots, the different manner in which these have been carried out in each country have produced strikingly dissimilar urban landscapes.

Nowhere is this more evident than with the sister cities of Brownsville and Matamoros, facing each other across the Rio Grande river. Demographically and economically they have much in common: the 2010 Census indicates that over 85% of Brownsville's population is of Mexican ancestry and Spanish-speaking, and although Brownsville is wealthier than Matamoros, it is very poor by American standards, having recently replaced McAllen as the poorest city in the nation. The Mexican state of Tamaulipas, by contrast, is wealthier than the Mexican average. Indications are that Brownsville's per capita GDP may only be around $14,000 to Matamoros' $10,000.

Matamoros at left, Brownsville at right. From Google Maps.
The economic and demographic convergence at the border might suggest a gradual transition in urban form, but instead there is abrupt break at the Rio Grande river, as can easily be seen above (the map has been rotated clockwise). Matamoros is a city that makes almost exclusive use of attached or nearly attached dwellings on small lots, while Brownsville has a quintessentially American pattern of detached houses sitting on much larger lots.

A typical house lot in a newer neighborhood of Matamoros has only around 1200 square feet, while in Brownsville, new subdivisions have lots around 5400 square feet. Interestingly, the lot size trends appear to be diverging: Matamoros' new lots are smaller than those in older parts of the city even as car ownership has soared, while Brownsville's are much larger. Older areas of both cities, dating back to the late 1800s or early 1900s, both tend to have house lots of around 3,000 square feet, although in different configurations (Matamoros, for instance, never used rear alleys).

Contemporary subdivision street in Brownsville. Google Maps.
New homes in Matamoros are often a tiny, in some cases little more than 500 square feet, but appear to be mere placeholders for expansion. Usually set back about 15-20 feet from the lot line, these homes are swiftly expanded forward into the setback and up a story or two using simple construction techniques, with the result that after no more than a decade or so, the street's appearance is completely transformed, and no longer appears mass-produced. (Single-use zoning seems to be unknown or unenforced, as numerous small commercial establishments can be seen cropping up, mid block, along these streets.)

Mexican "snout house" in Matamoros, expanded from the developer's
original house of the type still visible to the right. Google Maps.
Expanded in this way, a very small home can grow to perhaps as much as 1000-1500 square feet. Persons desiring even larger homes could, presumably, simply buy adjoining lots and combine them.

This fundamental difference in lot size has major implications for the size of the urban area. Matamoros, with a 2010 population of 489,000, actually occupies slightly less area (28 sq. mi.) than Brownsville's 175,000 (29.9 sq. mi.), based on a mapping out of residential urban boundaries in each case. It also must have some implications for home prices: while Brownsville homes are much cheaper than the American average with a median home price of $130,000 for what is typically a three bedroom home, a relatively large Matamoros starter home (915 square feet over two stories on an 1130 square foot lot) can be had for only $34,000. The tiniest of the starter homes, no bigger than a studio or micro-apartment, are as little as $10-15,000. This suggests that a new two-bedroom home in Matamoros can be purchased for approximately what the down payment would be for a typical three bedroom home in Brownsville, and helps explain Mexico's very high homeownership rate. Finally, it has implications for mixing of uses, since the dense packing of houses allows for businesses to thrive on foot traffic, reducing the political pressure for parking minimums that would, even in the absence of zoning, effectively ban businesses on tiny lots.

The economic, financial, geographic, historical, cultural and legal factors that have led to such divergent patterns of urban growth aren't easily or quickly summarized, but the different outcomes are clear enough.

Related posts: Check out Apex Urbanism, which has a series of posts featuring various Mexican cities.

7 comments:

  1. We made a few family trips through Mexico when I was a kid, including Matamoros, and they were extremely formative in my life-long interest in urbanism. They were totally unlike the American places I'd experienced up to that point.

    ReplyDelete
  2. Maybe we can learn something from observing how others do things. The concept of "mixed use" urban areas once was common here in the USA. This was back in a time where you didn't need a car to go or do anything! I can recall walking to various stores to buy things back in the 1940's and 1950's. Today of course this sort of social order no longer exists. Which I do think is a loss that only us old people understand.

    ReplyDelete
  3. This was really interesting. Thanks! I didn't know the housing prices could be so divergent in a the same urban area - all those zoning laws and housing regulations make a difference!

    ReplyDelete
  4. It's surprising how cheap the homes are in Matamoros. The standard $100/sf used to estimate construction costs in the United States doesn't apply. The example you gave is only $37/sf.

    I wouldn't be surprised if the higher density saved a few tens of thousands just for local infrastructure though (streets and utilities), not to mention more savings on regional infrastructure (highways, arterials, fire protection).

    I certainly thing an old urbanist development pattern could have a huge change on the dynamics in Toronto. Buying a small house newly built 1000sf house doesn't sound so appealing if you're going to pay $350k vs $500k for a 2000sf house. However, if you can get a 1000sf house with similar lot size as in Matamoros, with the old urbanist pattern of development, it would surprise me if you could do that for around $170k in Toronto (greenfield). The 2000sf size might be more around $400k with say, a 2500 sf lot and narrow streets. If you're going to dedicate so much land and construction related resources on wide streets, plus driveways and garages, plus mandatory front setbacks, you might as well go for the whole package, with double garage, double width driveway, moderate sized backyard (1000 sf), 5 beds, 5 baths with jacuzzis, marble floors...

    ReplyDelete
    Replies
    1. Thanks Nick -- I'm working on a follow up post trying to crunch some related numbers about small houses in the United States in the context of changes in household size. Why American (and Canadian) developers so rarely aim to construct very modest houses given what should in theory be strong demand is something I haven't been able to answer definitively, but the influence of minimum lot sizes and gold-plated infrastructure far in excess of what is needed for SFDR, as you say, may be contributing to it.

      Really enjoying your blog, BTW.

      Delete
  5. I think you used the wrong figures which make the relative wealth look more similar. As you used Real Per Capita Income for Brownsville Texas which is different to GNI per capita used for Tamaulipas. The USA has only a Real Per Capita Income of $27,319 according to your source, were as the USA has a GNI per capita of $50,000. There for I performed this calculation which will provide a rough estimate of Brownsville GNI, $50,000/$27,000 * $14,000=$26,000 which is 2.6 times bigger Tamaulipas GNI.

    ReplyDelete
    Replies
    1. Thanks anonymous -- it is not easy to find comparable figures here, especially comparing municipalities across national borders. Nonetheless, the basic point remains that the difference in incomes must be less here than between almost any other American and Mexican city, even if not to the extent I suggested. There would be a very strong incentive, if possible, to work in the US while living in Mexico, although I have no idea if that is commonly done in this region.

      Delete