Later this week, the Antiplanner will review The Triumph of the City, a new book by Harvard economist Edward Glaeser. But because a crucial part of that book is based on a working paper written by Glaeser and UCLA economist Matthew Kahn, I first want to review that paper.
Titled “The Greenness of Cities: Carbon Dioxide Emissions and Urban Development,” the paper attempts to estimate household CO2 emissions from 66 major urban areas. The paper concludes that some urban areas produce substantially less emissions per household and also that suburban emissions are larger than central city emissions, especially in the case of older cities such as New York.
Glaeser and Kahn’s paper attempts to estimate four types of household CO2 emissions: driving, transit, heating with oil or natural gas, and electricity. For most of these, however, they use very crude measures. In the case of driving, the Department of Transportation’s Highway Statistics series has data on miles of driving for more than 400 urban areas. Instead of using these data, Glaeser and Kahn rely on regression equations developed from the National Household Travel Survey to roughly estimate household driving based on population densities, family sizes, and incomes.
The two methods do not produce the same results. Both pre- and post-Katrina Highway Statistics indicate that New Orleans residents drive 10 to 20 percent less than New York metro area residents. But Glaeser and Kahn find that New Orleans residents drive about a third more than New York metro area residents. While Highway Statistics include all driving and not just household driving, I don’t know any reason why non-household driving would be that much different between the two regions. Why rely on crude calculations when more accurate data are actually available?
In the case of public transit, they use the National Transit Database, which are the same data used by the Antiplanner. However, as I’ll show below, they use inappropriate factors for estimating carbon emissions for electrically powered transit.
For home heating, they use the 2000 census to estimate the number of homes in each urban area heated with oil or natural gas and the amount of money these homeowners spent on fuel. Combined with Department of Energy estimates of fuel prices in 2000, they estimate the volume of fuel used. As with the driving data, these calculations are somewhat crude. While I don’t immediately know of any source of data of actual oil and natural gas consumption by urban area, it seems likely that such data are available.
They used the same procedure for household electricity consumption. To estimate CO2 emissions from electrical generation, they averaged CO2 outputs per megawatt of electricity for each of eight regions of the country. The problem with this is that there are huge differences in sources of electricity within each of the eight regions.
Colorado, for example, gets 95 percent of its electricity from burning fossil fuels, while Washington gets 87 percent of its electricity from hydroelectric dams or other sources that generate no CO2. Yet Colorado and Washington are both in the same region so Glaeser and Kahn used the same conversion factors for each of them. This problem also applies to electrically powered transit.
Considering the crude nature of their data sources, their results can hardly be considered reliable. This is particularly disappointing since, for many of the factors, better data were available.
Glaeser & Kahn’s most important results are shown in tables 2 and 5 of their report. Table 2 finds that the least carbon emissions come from urban areas that have mild climates and rely on non-fossil-fuel sources of electricity. California metro areas are supposed to have the least emissions, revealing one of the weaknesses of their regional electricity factors: less than half of California’s electricity comes from renewable sources, so Washington metro areas should score higher. But since Glaeser & Kahn applied the same carbon factors to both states, California looks better than it really is.
Men experiencing this condition can now purchase Penegra online to have an exceptionally positive change in their sexual execution, viagra for uk totally disregarding that disappointing analysis. The scientists of Ajanta cialis 40 mg pharmacy endeavored to find out its solution. Normally, all of us take part in sex. viagra samples in canada Low price, great efficiency and easy availability make discounts on cialis soft tabs first choice of drug to treat impotence amongst young men.
More important, the fact that California scores high while places like Texas, where climates are more extreme and that rely even more on fossil fuels, have the worst scores. In his book, Glaeser writes, “Houston residents, for all the sensible suburban logic of their lives, are some of the biggest carbon emitters in the country. All those 90-degree days and all that humidity means that Houston is a ravenous consumer of electricity.” But climate and sources of electricity are independent of urban form, so this hardly proves Glaeser’s claim that dense cities are greenest.
That claim is based on table 5, which is supposed to show the differences in carbon emissions between suburban and city households for each of the 66 metro areas. The greatest difference is in New York due, no doubt, to the huge difference in densities between New York City and its suburbs. But for some reason, the second-greatest difference is between Nashville and its suburbs, even though Nashville isn’t particularly denser than its suburbs. Los Angeles and Detroit score negative numbers–which doesn’t make much sense considering the huge income differences between these cities and their suburbs.
Remember, Glaeser & Kahn’s estimates of driving as well as some of the other numbers are based on hypothetical calculations, not actual measurements of energy usage or carbon emissions. The density and number of jobs in a region’s central business district is more important for driving than population density. By emphasizing population density and ignoring job density, Glaeser & Kahn are no doubt significantly overestimating driving in some areas (such as low-density New York suburbs that have large central business districts) and underestimating it in others (such as high-density areas cities like Los Angeles with widely dispersed jobs). This error probably exaggerates the differences between many cities and their suburbs.
With respect to heating and electricity, Glaeser knows that multifamily homes in dense cities tend to be significantly smaller than suburban single-family homes. Table 2.3.11 of this Department of Energy report indicates that single-family homes use less energy per square foot than multifamily; the overall savings is gained only by the latter being smaller. Glaeser says that urban multifamily dwellers can live in smaller homes because they “end up sharing common public spaces, like restaurants, bars, and museums.” But he fails to account for the carbon emissions from these spaces. If Glaeser really wanted to reduce carbon emissions, he would advocate that people live in smaller single-family homes, not in multifamily.
This is underscored by a major omission from Glaeser & Kahn’s analysis: the emissions from construction of housing and transportation facilities. Mid-rise and especially high-rise construction requires large amounts of steel and concrete, which are energy intensive and emit lots of greenhouse gases. Low-rise construction is largely wood, which is effectively carbon neutral.
We know that the carbon emissions from rail transit construction can offset scores of years of annual savings from operations. We also know that, because most rail transit lines outside of New York are little used, construction emissions per passenger mile are much higher than for highways. Glaeser & Kahn account for none of these factors.
Estimates of differences in household emissions can also be misleading because households tend to be smaller in dense cities than their suburbs. San Francisco has only 2.3 people per household; its suburbs average nearly 3.0 people per household. It is misleading to credit low household emissions to dense living when the differences in per-capita emissions may be much smaller.
Based on Glaeser & Kahn’s paper, The Triumph of the City concludes that, “Anyone who believes that global warming is a real danger should see dense urban living as a part of the solution.” But this is far from proven. Let’s grant that residents of dense urban areas may emit somewhat less carbon than suburban residents–in Glaeser & Kahn’s paper, as much as 13.5 tons per year in New York. The question is: how cost effective is this compared with other ways of reducing emissions? Recall that McKinsey & Company suggests that we should not pay more than $50 per ton to abate carbon emissions.
Glaeser’s book estimates that, in the absence of land-use regulation, high-rise housing should cost $400 a square foot (page 150). By comparison, Houston housing costs about $75 a square foot. Assuming the urban dweller would be happy in 1,500 square feet while the suburbanite would demand 2,000, the urban home would still cost $450,000 more than the suburban one. Amortizing over 50 years at 5 percent results in an annual difference of $24,000 a year, or about $1,800 per ton of emissions saved. Dense living is clearly not a cost-effective way of reducing carbon emissions.
As I’ll point out in my review of Triumph of the City, Glaeser opposes any government attempts to coerce people out of suburbs or subsidize them to live in dense cities. But his arguments lend support to those who support such policies. This is a disappointment considering that his research is flawed in so many ways.
Some of Kahn’s subsequent work and the work of others show that for low-income quintiles, cities are much lower emitters. For the well-off, they actually consume more energy in cities so they skew the numbers. The infrastructure is there, in other words, but there are no incentives for savings. So changing signals and incentives may help cities realize their potential. Regardless, per capita emissions are higher in rural areas for several reasons, most of them having to do with VMT.
Second, the Houston price/sf is not a viable comparison, as it is a city with low equilibrium rents that has few barriers to sprawl and few pure amenities – no one moves there for the weather or scenery.
Nonetheless, I’m glad to see Randal coming around in his views.
DS
Except or the period from June 1st to September 30, the weather in Houston is really quite pleasant. The weather during the hot months enumerated is really not any worse than the hot months in Philly, or Baltimore or even metro DC.
“…no one moves there for the weather or scenery.”
Nor the substandard public schools, cancer clusters, high violent crime rates, loads of VOC’s and carcinogens in the air, etc, etc, etc.
I suppose that’s the trade off; Loads of petro-chem jobs for a poor quality of life index.
I agree with you on the VOC’s and the carcinogens. I would disagree with you about the schools and violent crime. Otherwise, quality of life is largely in the mind. Where do you live Bennett? and how does it compare on the schools and violent crime front.
I live in Austin, but resided in H-town for a couple of years. The schools are just as bad, just about everywhere in TX has terrible schools(if you’re a resident you know this already. #1 in dropouts, near the bottom in funding per child). In terms of violent crime we’re way better, but size and demographics play a big part in that.
My observations on Houston are that it is a tremendously segregated (by income and race) city, has a big income gap compared to other TX cities and is typical of the old “haves and have nots,” in that the way in which residents experience poor education and harmful environmental conditions is contingent on their income. Also Houston is representative of the TX style urban development, i.e pods of humanity between stretches of highway (there is a reason that drunk driving is such a huge problem). Houston is hotter and muggier than the other Major TX cities. Houston lacks the Hispanic cultural celebration and awesome downtown of San Antonio, the music and charm of Austin, and the arts of big D(FW). Essentially, to me it represents all the bad qualities of a TX city (there are a handful of outstanding eateries. see: Frenchy’s Fired Chicken. Best in the world).
To me (pardon my exaggeration) there are only two reasons why people live in Houston. 1. They’ve never been anywhere nice, so they don’t understand how bad it actually is. 2. Their job requires it.
As for, “quality of life is largely in the mind,” you’re probably right, which is why I got the hell out of Houston. Obviously my analysis is somewhat compromised by subjectivity. You know they say NJ is the armpit of America. Well if Houston is near the bottom, often hot and sweaty and emits more harmful gasses than any other major city, that must make it the…
Dan says, “the Houston price/sf is not a viable comparison, as it is a city with low equilibrium rents that has few barriers to sprawl.” I am glad to learn that he agrees that reducing barriers to sprawl makes housing more affordable. The $400 per square foot for high-rise construction is Glaeser’s estimate of what the cost would be if there were no barriers to construction in New York. He admits that actual costs are at least twice as high.
In Portland, high-rise construction costs start at $150 a square foot. Doing the same calculations as above using that cost results in a cost of $300 per ton of abated greenhouse gases. That’s still quite a bit more than McKinsey’s limit of $50 a ton.
Mr. O’Toole states: “indicates that single-family homes use less energy per square foot than multifamily.”
In a post where you’re lambasting your opponents for their use of data, you trot out this tired old stat?!?!?! Per square ft doesn’t really tell us anything valuable, it’s how those square feet are being used by the residents. In the same report it shows that SF residents use about 2x the amount of energy as their MF counterparts. Those that live in SF homes, on average, consume more energy than the MF dwellers. Period!
Maybe its just me, but I think it rather rude to run down any city in general. There are cities I like, many I hate, but it is because of my own personal desires, and even that changes on my stage in life. Wherever people choose to live, they do so because that is their best choice to them. That should be respected.
I realize planners live and breathe a multitude of indices, and that they probably don’t realize that discussing cities in certain ways is rude to the general public. But life is not a game where the winners live in the “best rated” places. Planners are making more enemies than friends in acting like they know better than others where people should want to live.
There is one overwhelming criteria that most people use to decide where to move their residence. That one reason is why almost everybody in the US ended up here (except for slaves and American Indians).
No need to mischaracterize what I wrote, Randal. As you know, cities that are more attractive and have more amenities have higher equilibrium rents. Choosing a low-amenity and low-attractiveness city like HOU is too low a bar. But hey, Houston is the ideal town because it don’t have zonin, right? Oh, wait: that’s not right either….
DS
Borealis says: “Maybe its just me, but I think it rather rude to run down any city in general.”
Touche. But I do hate Houston all the same.
“I realize planners live and breathe a multitude of indices, and that they probably don’t realize that discussing cities in certain ways is rude to the general public. But life is not a game where the winners live in the “best rated†places. Planners are making more enemies than friends in acting like they know better than others where people should want to live.”
Interesting take. First off, most professional planners are aware of, and sensitive to this. It’s a conundrum of the profession. Much of what planners deal with are the “problems” associated the metropolitan areas. Discussing these problems will ruffle some feathers, and hopefully the planners on the ground (not the Dauny’s and Kunstler’s) will show sensitivity in public involvement (not go off like my Houston rant).
I’m also assuming your comment is in response to my Houston rantings. Let me be clear that it was mostly a subjective argument and while it indicates that I think I “know better than others where people should want to live,” it is the approach I feel comfortable using in this forum (where I’m more of a Duany/Kunstler). As a professional I operate a little differently and leave my chest pounding pissing matches to the blogosphere.
bennett,
Actually I wrote #8 after seeing your #3 without even seeing your #5. I think it is fine to say how you personally feel about any city. I think it is fine to make broad comparisons on a specialized website like this one. But I think it is silly and broadly insulting to rank or run down cities in publications for the general population.
And my main gripe is against the idea that NYC is a model for carbon reduction. NYC is a very unique situation, completely unsustainable and non-repeatable, and will come out first or last in almost any calculation depending on how one chooses to weigh the variables.
If it makes Borealis feel any better; I think as little of Austin as Bennett does of Houston; and I lived in Austin for five years.
The worst part about living in Austin was the traffic and the nimby liberals who keep Austin, too white, too expensive, undriveable, and really liveable only for the well to do. If Houston has pockets of poverty, it is only because the poor move there because they know it provides opportunity which they won’t find in Austin. However, I am grateful that such places as Austin and even Portland, San Francisco, and Boston exist. For that allows people of a certain affect a place to live, be happy and co-exist with others similarly affected. Then they don’t have to live near me or vote in my precinct.
“Jardinero1 says: The worst part about living in Austin was the traffic and the nimby liberals who keep Austin, too white, too expensive, undriveable, and really liveable only for the well to do.”
I must give props where props is due. This is ATX in a nutshell. Nail on the head. Actually, I pretty much agree with all of #12, sans the implicit denial of environmental justice issues in the Houston area. Other than that, I couldn’t agree more.
To be fair, I do love to visit there.
p.s. I’m not that wild about Austin either.
Jardinero1 posted:
The worst part about living in Austin was the traffic and the nimby liberals who keep Austin, too white, too expensive, undriveable, and really liveable only for the well to do.
Here in Maryland, we call them “limousine liberals.”
Montgomery County is notorious for having lots of them, including many who will show up at public meetings and hearings (often in expensive private automobiles) to oppose any and all proposed highway improvements or expansions, making forceful statements in favor of expanding mass transit instead.
Speaking of which, there was this in the N.Y. Times the other day:
Green Development? Not in My (Liberal) Backyard
C. P. Zilliacus says: “Here in Maryland, we call them ‘limousine liberals.'”
In Boulder CO, I believe the term is “Trustafarians,” a combo of Trust fund and rasta. Maybe the only place in America you’ll see dreadlocks behind the wheel of a BMW suv.
There’s a dearth of good and true communiatrians today. People talk the talk…
“I don’t know any reason why non-household driving would be that much different between the two regions.”
Maybe because New Orleans is a vast warehousing, transloading, and shipping hub in a way that NYC is not anymore, especially if Newark-Elizabeth, where the remaining port activity is in the NYC area, is being counted as a seperate Metro area like the Census does.
Chicago and LA also have huge drayage and warehousing operations which would skew non-household driving out of the mean for all cities.
bennett:
“There’s a dearth of good and true communiatrians today.”
I just has to be a choice you want to make with your lifestyle. Too many people cannot fathom acting at the same time out of a sense of respect for the common good and also respecting the rights of private property and freedom of movement and association.
Antiplanner:
“Glaeser’s book estimates that, in the absence of land-use regulation, high-rise housing should cost $400 a square foot (page 150). By comparison, Houston housing costs about $75 a square foot. Assuming the urban dweller would be happy in 1,500 square feet while the suburbanite would demand 2,000, the urban home would still cost $450,000 more than the suburban one.”
It would be nice if other examples were used in between Houston and New York City. Many urban and suburban homes/apartments are much smaller than those numbers. Also, there are other much cheaper but still dense urban development patterns aside from the very costly highrises of NYC. Rowhouses and twins in Philly and Baltimore, tri-levels in Boston, Chicago and Pittsburgh’s single family homes that end 1-2 ft. from the property line to leave a narrow alleyway between it and the neighbors, etc. At 4 people per three bedroom 20 ft. wide rowhome on a 60 ft. deep lot, Philadelphia-style rowhouses allowed a theoretical density of 57,600 people per square mile (this is rarely achieved in reality) including 30 ft. streets with 5 ft. sidewalks. Little brick or wood rowhouses hardly cost anywhere near $400 per sq. ft. to build. NYC’s density is really a farce anyway because of the dearth of families in the city. When one person inhabits a 500-1000 sq. ft. highrise apartment, the per capita density per sq. ft. of housing is less than 4 people living in a 1200-1600 sq. ft. rowhouse or twin. The vast common areas required in apartment buildings for hallways, stairs, elevators, mechanical rooms, etc. are all hidden density killers that force apartment buildings to be very tall before they aproach the density of rowhouses. Many NYC apartments are 4-5 level brownstone type houses where they are subdivided into 10 one bedroom apartments, with perhaps 15 people inhabiting a building that is 5500+ sq. ft. When you work out the developed per capita square footage, its about the same as a rowhouse.