Austin Contrarian

Back and forth

It never surprises me to hear suburban commuters bash congestion tolls.  But I am surprised when a sophisticated transit blogger bashes them.  Yonah Freemark, who writes the very good the transport politic, is the guilty party.  He and Ryan Avent have been going back and forth and back again.  

Yonah's objection is equity:

A huge percentage of the U.S. population pays far too much for transportation; to put it simply, most working adults have no choice other than to own a vehicle and often to drive it dozens of miles every day. Making driving more expensive is a great way to devastate the already impoverished.

It’s true, tolling highways would save “money, time, lives, and emissions.” But it would also sacrifice the mobility of a large segment of America, because the reduced congestion would be a result of the poor and the middle class choosing not to drive because of expense, not because of choices made by the wealthy.

He argues that our public transit systems are too underdeveloped in most places to cure this inequity.

Ryan responds with several good points, but I want to elaborate on why the regressivity argument leaves me cold.

First, congestion tolls are less regressive than toll opponents claims.   The low-income do pay tolls, which means they value the time savings more than the price of the toll.   Plenty of low-income drivers take the toll lanes on SR 91 in California, for example.  The reason many do is that they have little flexibility in their schedules.  They have to be at work or at the daycare center at a specific time.  To compensate for congestion, they must leave much earlier to guarantee a timely arrival.  They thus suffer two types of costs:  scheduling delays and travel delays.  (The value of scheduling delays is very hard to measure, which means that studies like the TTI report almost surely understate the cost of congestion.)  Affluent professionals often have more flexibility.

Second, tolls encourage a number of shifts.  Yes, shifts to transit, which seems to be Yonah's main concern, at least when the transit system is underdeveloped.  But they encourage other shifts, too.  Shifts to other routes and shifts to other times.   Commuters are the least likely to be nudged to other routes or times.  The most sensitive are those who use congested roads for local trips.  Take the soccer mom who hops in the SUV and enters a congested highway to get to the grocery store a mile down the road.  She imposes enormous costs on others.  Tolls make her internalize those costs and nudge her to use the local streets. 

Perhaps this is regressive.  But it doesn't evoke much sympathy from me.  A congestion toll is a charge for getting in everyone else's way.   Behind every claim of regressivity is the assumption that drivers are entitled to get in everyone else's way even when it is not worth all that much to them. 

And that leads to my last point.  The regressivity argument wouldn't move me even if it were true. This is one of those cases in which our desire for efficiency should trump our concern with regressivity.

Here's an apt analogy.  An amusement park owner decides to throw open the gate to all comers on a first-come, first-serve.  Naturally, a long line forms.  But this isn't a typical line where newcomers go to the back of the line.  No, in this queue, newcomers elbow their way to the front of the line and force everyone behind them to wait a little longer.  

Naturally, chaos ensues.  No one leaves for the park knowing how long it will take to get in.   There is "queue rage" and general aggravation.  The park owner loses business.  And many low-income parents find themselves worse off because they have a smaller window of free time and dislike the chaos and aggravation as much as anyone else.  (I've always thought it patronizing to assume that low-income drivers put such a low value on their time and aggravation.)

Now, in the real world, the queue would never work this way.  Those at the rear of the line would use informal sanctions (fist fights) to deter queue jumpers.  And, in fact, there would be only a short  line to get into the park in the first place because the owner would charge for admission.

But this is exactly the crazy system we use to ration highway access.  For whatever reason, our cultural norms have evolved to tolerate a free-for-all.  In almost every other situation we ration scarce goods using price.  Sure, that's regressive in the sense that the poor have less money to spend on things.  But we tolerate some regressivity elsewhere because we recognize that (1) the gains in efficiency outweigh the equity concerns; and (2) there are better ways to ensure an egalitarian distribution of wealth than creating artificial shortages, chaos and mayhem.

Regressivity is not the be all and end all.

Cross-posted at Urban Returns.

Sorting

I've written recently about Glaeser and Resseger's research showing that workers in skilled cities tend to become more productive as their cities grow while workers in unskilled cities do not.   I've also written about Abel et al.'s paper showing that while workers in all cities tend to become more productive as their cities grow denser, workers in skilled cities are especially likely to benefit.  The latter  receive, on average, three times the productivity gains of workers in the least skilled cities.  Growth and densification are good if yours is a highly skilled city.  

But the Glaeser and Abel papers also raise another possibility:  perhaps skilled cities are better off keeping out the unskilled than growing without bound.   

This is arguably an implication of their work. (And one I don't like, by the way.)   Both Glaeser and Abel define highly skilled cities to be those with high percentages of college-educated workers.  The higher the ratio of B.A.s to high school grads, the more skilled the city under their definition -- and the more the workers benefit as the city grows.

One way to keep a high ratio of skilled to unskilled is to price out the unskilled.  The most effective way to price unskilled workers out of a city is to keep the cost of housing high.  And, indeed, the highly-skilled cities on the coasts are adept at (and notorious for) using rigid land-use regulations to inflate home prices.  (Or were until recently; more on that later.)   The incumbent homeowners in these cities of course benefit from a tight housing supply since it raises the value of their properties.  But perhaps their skilled workers benefit too.

I speculated about this  last October without the benefit of Glaeser ir Abel's research:

Perhaps the elites in the superstar cities sense it is better to surround themselves with a uniformly high-quality workforce.  They want to spend their time with other elites; letting in lots of less-skilled workers would introduce so much static, just as if Harvard were to throw open its doors to anyone with a half-decent high school transcript.  In other words, an influx of less-skilled workers might dilute the experience for the high-skilled.

I got my undergraduate degree from Ole Miss and my law degree from Yale.   I studied harder at Ole Miss, took harder classes and spent a lot more time with my professors.  I thought law school was mostly boring and had a suburban home and the premium cable channels by my third year.  But I probably got a better education at Yale than Ole Miss merely by hanging around a bunch of people who were smarter than me.  There were economists, physicians, mathematicians, political scientists, a Broadway actor, an Olympic miler  and, of course, a bunch of smart new college graduates  from all over the country (but mostly from the Ivy League, Standford and Berkeley.).  It was a small class of 175.  It was impossible not to keep up with new ideas or trends because someone you bumped into would know about them.  I wasn't paying a gazillion dollars for the law school classes, I was paying for the spillovers.  It was a fair price.

The housing markets in  LA, San Francisco and New York function as giant sorting machines.  By setting home prices so high, they weed out all but the most skilled -- except for the class of relatively unskilled who are supported by price controls or subsidies.  They're using the  Ivy League model.

So are they better off?  Yes, clearly, if they can shunt the less skilled elsewhere and  continue to grow and densify.  But it's hard to do both.  High prices, in fact, require that the housing supply be kept tight  in order for prices to stay high.  Otherwise, the new supply of housing would drop home prices and the riff-raff (from their perspective) would come flooding back in.  But perhaps merely increasing the density of skilled workers will yield the productivity gains yielded by  growth and densification.  Their motto could be, "We get smarter, not bigger."

I think this is probably wrong, though.  What matters is being close to other people with the same skill set; that's where the knowledge and productivity spills occur.  A software developer benefits from having a lot of other smart software developers around; a musician benefits from having a lot of other good musicians around.  Adding more software developers or musicians who aren't quite as skilled shouldn't dilute this benefit because we city-dwellers largely control whom we interact with.  Firms, in fact, segregate themselves by quality all the time.  Deepening the labor pool simply can't hurt.  I think cities like San Francisco and New York would be even more productive if they loosened their growth restrictions and allowed in more people.

And even if were true that the San Francisco model can achieve greater productivity without growth, there's always the risk that its  sorting machine might break. To wit, a housing market collapse.  Skilled workers cash out what they can and move elsewhere.  Lower home prices attract the less-skilled who had been priced out of the market.  A political and regulatory climate makes the city less hospitable entrepreneurs, who benefit from cheap space and light regulation.

It will be interesting to see what happens to productivity in these wealthy, slow-growth cities as the recession wears on.  My guess is that they will have to begin growing vigorously again if they want to recapture the productivity they had at the peak of the housing boom.

Skilled cities II

While I'm in this nerdy mood, let me point to this paper by Jaison Abel, Ishita Dey and Todd Gabe:

We estimate a model of urban productivity in which the agglomeration effect of density is enhanced by a metropolitan area’s stock of human capital. Using new measures of output per worker for U.S. metropolitan areas along with two measures of density that account for different aspects of the spatial distribution of population, we find that a doubling of density increases productivity by 10 to 20 percent. Consistent with theories of learning and knowledge spillovers in cities, we demonstrate that the elasticity of average labor productivity with respect to density increases with human capital. Metropolitan areas with a human capital stock that is one standard deviation below the mean level realize around half of the average productivity gain, while doubling density in metropolitan areas with a human capital stock that is one standard deviation above the mean level yields productivity benefits that are about 1.5 times larger than average.

This is a little technical, so let me translate:   Cities tend to become more productive as they grow denser.   On average, a city's workforce becomes 10% more productive when the city doubles in density.   But that average obscures the importance of skills.   Less skilled cities benefit a lot less than skilled cities from densification.   In fact, skilled cities, on average, enjoy three times the productivity gain from denser growth than less skilled cities.   This is yet more evidence of the increasing returns and agglomeration benefits from density.

This is a nice complement to the Glaeser and Resseger paper.  Glaeser and Resseger found that workers in skilled cities become more productive as the city grows, while workers in unskilled cities do not.   Abel et al find that workers in skilled cities become more productive as the city grows denser; workers in unskilled cities, less so.

The authors also test their conclusions using a variant of weighted density. They find even greater productivity gains (20% on average) when a city doubles its weighted density.

The authors use a coarse form of weighted density.  They weight urbanized area density by county subdivisions.  But how you chop up a city matters when calculating weighted density.  In order to calculate weighted density, you first divide the city into a bunch of smaller regions.  You then assign each region's density a weight equal to its share of the populations.  In general, weighted density increases as you chop up the city into smaller regions.

I used census tracts for my weighted densities.  There are many more census tracts than county subdivisions.  I thus got a lot more stratification than they did -- e.g., their top weighted density was 19,000 ppsm, while mine was 33,000.  How you divide a city for calculating weighted density is somewhat arbitrary, but I think using census tracts makes more sense than county subdivisions, which are more or less arbitrary.  I suspect the authors would have found even greater returns to density had they weighted density by census tracts rather than by county subdivisions.  

H/t Richard Florida.

$160 per trip

That's the social cost that Charles Kamonoff, an NYC environmental/transportation analysist, believes each driver who enters Manhattan's central business district imposes on other drivers.  Via Felix Salmon: 

After crunching the numbers, he calculates that on a weekday, the average car driven into Manhattan south of 60th Street causes a total of 3.26 hours of delays to everybody else. (At weekends, the equivalent number is just over 2 hours.) No one car is likely to suffer excess delays of more than a few seconds, of course, but if you add up all those seconds for the thousands of affected cars and trucks, it comes to a significant amount of time.

Many of those hours are very valuable things, especially when you consider big trucks, staffed with two or three professionals, just idling in traffic. Komanoff calculates (check out the “Value of Time” tab) that the average vehicle has 1.97 people in it, and that the average value of an hour of saved vehicle time south of 60th Street in Manhattan on a weekday is $48.89. Which means, basically, that driving a car into Manhattan on a weekday causes about $160 of negative externalities to everybody else.

I can't vouch for Komanoff's numbers or his methodology. Given the sheer complexity of the traffic dynamics in Manhattan -- which are much more complicated than the traffic dynamics of a single highway -- I would treat his estimate skeptically.

That said, his estimate doesn't seem outrageously high to me. The basic point is sound:  we severely underestimate how many people we delay when we enter a congested network of roads.  If you've ever tried to make the trip crosstown Manhattan in the middle of the day, you understand just how much delay one driver can cause.

Komanoff recommends congestion pricing.  A good idea.  But he also proposes making buses free, which is a bad idea (and one floated in Austin occasionally). 

One problem is that a free bus would be a magnet for the homeless looking for a cool place to hang out.  That sounds callous, I know, and perhaps it is, but in that case you'd never see a large switch from cars to buses, so what would be the point?

The other problem is that both bus and rail can be congested, too.  If you've ever taken a packed No. 3 around ACL, you understand this perfectly well.  There is a cost to standing up on a jerky bus for 20 minutes, squeezed in among a bunch of strangers who may or may not practice good hygiene.  This is a genuine cost of congestion.  In fact, it is probably too high a cost for most of the drivers around here, who would much rather be stuck in traffic in a cool, comfortable car.

If we price roads but not buses or trains, then the buses or trains will be too crowded (at least in Manhattan and DC, if not in Austin) and we simply will have shifted the cost of externalities from drivers to bus and train riders.  This certainly does not mean that buses and trains must pay their own way.  But it does mean pricing them with the interests of both drivers and non-drivers in mind.

Skilled cities

Ed Glaeser and Matthew Resseger have a new paper out examining how worker productivity varies with city size.  They find that workers in skilled cities become more productive as the cities grow; workers in less skilled cities do not:

There is a strong connection between per worker productivity and metropolitan area population, which is commonly interpreted as evidence for the existence of agglomeration economies.  This correlation is particularly strong in cities with higher levels of skill and virtually non-existent in less skilled metropolitan areas.  This fact is particularly compatible with the view that urban density is important because proximity spreads knowledge, which either makes workers more skilled or entrepreneurs more productive.  Bigger cities certainly attract more skilled workers, and there is some evidence suggesting that human capital accumulates more quickly in urban areas.

No one is quite sure exactly why (skilled) city growth makes workers more productive (Glaeser and Resseger included).  One theory is that workers learn more quickly from one another in larger cities; there is more trade-specific information "in the air."  For example,  software engineers and musicians develop their skills more quickly in a city with lots of software engineers and musicians.

The other theory is that a large city has a deeper pool of entrepreneurs and others with "high human capital" (i.e., smart, creative people).  It generates more innovations, which make the city's population more productive, which attracts more workers, who generate more innovation, etc. (This one better matches Jane Jacobs’ theory.)

These are genuinely different explanations. If the first is accurate, then workers’ pay should rise more steeply with experience in large, skilled cities than in small, skilled cities.  The second is a sort of “creative destruction” theory; it means that workers in large, skilled cities face a higher risk of being displaced by the latest innovation.  I think both are true, which (partly) explains why big cities attract some highly skilled people while repelling others.

The Texas Triangle

Ryan Avent writes about Texas's strong growth, and the possibility that this strong growth will beget more growth:

Austin’s strength is pretty remarkable, as well, though I shouldn’t be surprised given the way Texan metropolitan areas have held up, in terms of home construction and prices at least, throughout this crisis. Interstate migration doesn’t necessarily imply anything about preferences; as I’ve mentioned many times before, housing supply restrictions in high demand markets make housing there unaffordable, diverting people to places with loose supply and low home prices (Texas!). But agglomerations are powerful, and whatever advantages a growing place began with, it will eventually develop a serious attractive force when it reaches a large size.

It’s surely a coincidence, but California’s population loss in 2008 due to domestic migration was almost exactly the same as Texas’ gain due to same — about 140,000. With economies in California, Arizona, and Nevada withering, it’s not difficult to imagine a shift in population taking on its own momentum. 

Agglomerations are indeed a powerful thing.  A growing population fuels demand for more specialized services, which spurs the development of larger, more sophisticated and more efficient industries, which in turn attract more people.  On the supply side, a growing city means a deepening labor pool, which makes the city more attractive to employers.

For the four large Texas metropolitan areas, the interesting question is, "What is the agglomeration?"

There are two possibilities here.  One is that Houston, Dallas, Austin and San Antonio are just four large, independent metropolitan areas, each an independent agglomeration.  If this is the case, the growth of one does not particularly benefit the others.  On the contrary, the growth of one could hurt the others if they are competitors.  If Houston and Dallas were vying to become major banking centers, for example, then a bank gained by Dallas would strengthen its banking industry in absolute terms and relative to Houston's.  Perhaps more importantly, a banking industry split between two competing cities would not enjoy the economies of scale or increasing returns of a banking industry concentrated in one or the other.  

The other possibility is that the four metropolitan areas collectively comprise a single agglomeration.  Texans have started talking up that idea the last few years (although most refer to it as the "Texas Triangle" rather than the "Texas Agglomeration").

Fortunately for Texas (if not for its rivals), the latter appears to be the case.  I just stumbled upon a 2004 report from the Dallas Fed's Houston branch (!) which looked at the four cities' concentrations of industry and concluded that they are complements, not competitors:

The Texas Triangle cities developed as economic complements, providing unique goods to the other Triangle cities and importing goods that represented strength elsewhere. Why is this important? First, it means that the Texas Triangle is in fact a megalopolis in the sense that we can add the pieces together with a minimum of duplication. It is spread over a triangular area of roughly 250 miles on each side. Second, it implies that despite traditional rivalries and competition among these cities, especially Houston and Dallas, they don’t really overlap much in their economic roles. We could isolate only a few areas where meaningful rivalry might take place—oil and gas extraction and semiconductors (Austin and Dallas) and heavy construction (San Antonio and Houston).

By and large, however, one or two Triangle cities have such a secure niche in each export industry that others are unable to compete effectively. Given the lack of competition across cities, a cooperative effort at industrial recruitment and economic development programs makes sense, even though the cities are spread over an area as large and diverse as the Texas Triangle.

Put differently, the four Triangle cities export goods and services to one another.  They largely export different goods and service, which means that Houston benefits when Dallas grows and vice versa.  Houston's growth provides Dallas firms a larger market for their "exports," and a growing cluster of Dallas firms provide Houston with more specialized goods and services.

All four metropolitan areas are among the fastest growing in the country.  Housing is cheap in all four metropolitan areas (although not in Austin proper), and cheap housing and lots of jobs explain some of that rapid growth.  But it is probably also the case that each city is growing rapidly because the other three cities are growing rapidly, too.  

Liveability rankings are bunk

Someday I will make my own list of the ten dumbest "Best Cities" lists.   This Economist ranking of the "most liveable" cities will make the cut.  (I'll put it on my list even though I haven't read the full report and don't intend to -- the Economist is asking $250 for it.)

From the summary:

The Economist Intelligence Unit’s liveability rating quantifies the challenges that might be presented to an individual's lifestyle in 140 cities worldwide. Each city is assigned a score for over 30 qualitative and quantitative factors across five broad categories: stability, health care, culture and environment, education, and infrastructure. The categories are compiled and weighted to provide an overall rating of 1–100, where 1 is considered intolerable and 100 is considered ideal.

The top cities are Vancouver, Vienna, Melbourne, Toronto, Perth, Calgary, Helsinki, Geneva, Sydney and Zurich.  According to Richard Florida (whom I sure was sent a copy gratis), Pittsburgh is the top U.S. city, at 29th.

This kind of list is silly because we already have a much more accurate method of gauging "liveability":  just compare prices.

Home prices in a metropolitan area are determined by area wages and amenities.  All else being equal, residents must pay more for home prices when a city offers higher wages.  Ditto with amenities.  If City A is a lot nicer place to live than City B but they offer the same wages, then City A has to be more expensive than City B; otherwise, residents of B would migrate to A until they bid up the home prices there.

We can't measure the value of amenities directly, but we can infer it from wages and home prices.  When a city's home prices are low relative to wages, we know there are disamenities lurking about.  When home prices are high relative to wages, we know the city must be, all things considered, a very pleasant place to live.  (That is, unless home prices are bubbly.  Irrationally high home prices can suggest that a place is more pleasant than it really is.)

The ratio of median home prices to median household incomes (see, too, the linked report) tells us quickly whether a city is cheap or expensive relative to wages.  It is thus a pretty good indicator of city amenities.

In 2007, the median home in Pittsburgh was priced at 2.6 times the median household income.  By comparison, the ratio for Honolulu was 9.1; San Francisco, 8.0; Los Angeles, 7.2; New York, 7.0; Seattle, 5.2; Milwaukee, 4.0; Austin, 3.3; Houston, 2.9.  The cheapest was Youngstown at 1.8.  Pittsburgh's companions include Rochester, Buffalo, Vincent, St. Louis and Atlanta (the traffic really, really sucks in Atlanta).

Pittsburgh is cheap relative to wages.  That tells us a lot about how pleasant it is to live there.  Perhaps the Economist ignored weather.

Some of these ratios have doubtless been inflated by the housing bubble. I otherwise would have a hard time understanding why Providence's ratio is 4.4 while Denver's is 3.7. But the bubble would not have artificially deflated home prices in places like Pittsburgh and Youngstown. On the contrary, the easy credit of the mid-2000s likely propped up home prices there like everywhere else.

If we look at consumer preferences as revealed by price, Pittsburgh is not the most attractive or pleasant city in the United States.  It's not even close to the top.  And if "liveability" does not mean "attractive," then what do we care?  Otherwise, the Economist is just quantifying a random combination of factors that it thinks are important but which do not necessarily matter much to city residents.

I do not question that Vancouver and Perth are highly desirable cities.  (This is hearsay; I haven't been to either).  But if the Economist's methodology doesn't get things right in the U.S., I have no reason to believe that it gets things right anywhere else.

Suburb math

During the current recession, suburban home prices have tended to fall by larger percentages than (some) central-city home prices.  People have offered lots of explanations.  Suburbs have lost their luster because of changing tastes.  Suburbs' higher transportation costs have finally caught up with their homeowners.  Suburbanites, perhaps, are less secure financially than city dwellers, or are more likely to have to stretch to buy a home.  Some or all of these could be true.

But everyone seems to assume that a steeper decline in the price of a home necessarily implies a steeper decline in the demand for the place.  It does not.  Even if demand for each place in a city drops by the same percentage, suburban home prices must drop by a larger percentage than central-city home prices, assuming transportation costs do not also decline.   More precisely, home prices must drop by a larger percentage in places with higher relative transportation costs.  

People confuse demand for a home in a given place and demand for the place itself.  In a recession, what actually declines is demand for the place.  Residents are poorer, the city offers fewer jobs and amenities, and city services decline.  Residents, on average, are not willing to pay as much to live in the city -- or any given place within the city.

The cost of living in a place is not just the cost of housing.  The cost includes transportation. This cost varies by place.  Suburban home owners pay less for homes and more for cars, while residents of swank central-city neighborhoods, close to jobs and city amenities, pay lots for housing and relatively little for transportation (especially after considering the time cost of commuting).  

This is a trite observation, I know.  But it matters.  It means that when demand drops for a place, the drop in demand must be reflected in either lower housing costs, lower transportation costs, or both.  When demand for a place drops but transportation costs do not, then the cost of housing must absorb all of the decline.  Which means housing costs must drop by a larger percentage in places where housing costs are a smaller part of the budget.   This is true even if each place in the city sees the same percentage decline in demand.  The demand for a house in a given place is not equivalent to the demand for the place itself.

The algebra is not hard, but it’s easier to explain with a (grossly oversimplified) example.

Let's assume that during a recession the demand for each place in a given city drops by 10%.  Consider two neighborhoods:

In neighborhood 1, an expensive neighborhood near the city core, the typical household's transportation costs are negligible compared to the cost of the home itself; for all practical purposes, each household spends 100% of its "housing" budget on the home.  In neighborhood 2, an inexpensive exurb far from jobs, the typical household spends only 50% of its housing budget on the home itself and spends the other 50% on transportation.

If transportation costs hold up, then housing costs must absorb all of the drop in demand.

This is a simple calculation for neighborhood 1:  the price of each home drops by 10%.

Demand for each place in neighborhood 2 likewise drops by 10%.  Again, housing costs must absorb all of the decline because transportation costs do not change.  But if home prices in neighborhood 2 were to drop by just 10%, as in the central neighborhood, the total cost of living in neighborhood 2 would drop by just 5%.  In order to get a 10% drop in the total cost, home prices must drop by 20%.

It looks to all the world as if demand for the suburban neighborhood has dropped more than demand for the central-city neighborhood, but the demand for each has dropped by the same percentage.  The suburbanite has simply experienced a steeper drop in a smaller percentage of his total cost for the place.

Let’s consider three other scenarios.

Expensive suburbs.  In a suburban neighborhood of $1 million mansions, the cost of housing might be a very large percentage of the total budget. Home prices in these neighorhoods will react like home prices in central-city neighborhoods.

Poor, central-city neighborhoods.  Transportation is cheap in poor, central-city neighborhoods.  But housing sometimes is even cheaper.  Homes can be so cheap that housing makes up a relatively small percentage of the total cost of the place.  Then a drop in demand would have the same effect as in the ‘burbs:  home prices would drop by disproportionately large percentages.  Counter-intuitively, the cheaper the housing, the larger the percentage drop in home prices.  (This should sound familiar.)

Polycentric cities.  Average transportation costs do not vary as much from neighborhood to neighborhood in a polycentric city.  (Note this does not mean the city's average transportation costs are lower than another city's.)  At the extreme -- when jobs and amenities are evenly distributed within the city -- average transportation costs are exactly the same everywhere.  In such a city, all that matters is the cost of housing.  The cheaper the housing, the smaller its percentage of the housing budget -- and the steeper the decline in the price of the home.

We thus don’t have to posit any change in housing preferences or demographics or the geographical distribution of subprime borrowers to explain why suburban home prices decline more steeply than (some) central-city home prices during a recession.  It’s just algebra.

None of this is to suggest that other factors aren’t at play.  We might very well be seeing a shift in the type of communities people prefer -- although any such shift surely began long before this recession.  And in some cities, the credit-risky and financially precarious have been relegated to the suburbs; these were the first to feel the recession's pangs, and the first to default on their mortgages en masse.

But do note that these alternative explanations imply alternative predictions.  If there has been a fundamental shift in preferences, then suburban home prices will remain (relatively) depressed even after the economy recovers.  If this little model is right, suburban home prices will recover by a larger percentage than central-city home prices.   

One could argue with my assumption that transportation costs hold steady.  It's not a bad approximation in the short run.  Residents (at least those who keep their jobs) must make the same commutes as before, make the same drives to the grocery store, etc.  The time cost of commuting is a big part of the overall cost.  But if, say, gas prices do decline -- they have in this recession –- then that decline will offset the decline in suburban housing costs somewhat.

One last point.  I'm using "housing cost" interchangeably with “home value.”  This is  sloppy.  For suburban renters in my first example, a 10% drop in demand means a 20% drop in rent, holding transportation costs constant.  The suburban renter really is no better or worse off than a central-city resident who pays nothing for transportation but sees only a 10% drop in rent.

It’s the suburban landlord who takes the real hit.  She loses 20% of her asset's value; it is little solace to her that her renter pays such a small percentage of her budget for housing.

Homeowners, of course, are both renters and landlords.  They seem to worry only about the value of their asset — understandably, since they are highly leveraged.  But the decline in the value of their asset is offset by the decline in the imputed rent they pay, which is why those who lose their homes to foreclosure often can afford to rent them back from the bank.  As renters, they are no worse off than anyone else in the city; they just get burned as investors.

A creative idea for pricing roads

Ryan points to this interesting proposal for a combined congestion/vehicle miles traveled ("VMT") tax:

The system might work like this: Vehicles would be fitted with a GPS device to record distance traveled, time and location of travel, and type of vehicle. This data would be sorted into various toll categories, and the device would wirelessly upload the totals to the gas pump when the motorist refueled. The pump would significantly discount the gas tax and add the appropriate road-use fees to the fuel bill. To further protect privacy, only category totals would be communicated to the governing agency via the pump. Tourists and others lacking the transponder would pay the full gas tax. Travel outside the area would not be recorded.

Resources for the Future calculated that the charges under a similar policy for the Washington area would average 9.3 cents per mile. They estimated an 11 percent reduction in VMT, 19 percent less emissions of volatile organic compounds and 17 percent less carbon monoxide. The estimated social welfare benefits (reduced congestion, pollution, accidents, etc.) of this reduction in driving were estimated at the equivalent of $1.1 billion — even before the revenues were disbursed.

This is an elegant solution to the practical problems of collecting a VMT tax and charging for congestion.

A couple of off-the-cuff thoughts:

First, the nominal gasoline tax would have to be set very high in order to prevent drivers from switching to fuel efficient cars merely to avoid the VMT/congestion charges.  For example, if the gasoline tax were set below $3.00/per gallon, drivers could lower their effective VMT tax -- and avoid the congestion premium -- merely by  switching to vehicles that got 35 mpg.  Yes, encouraging fuel efficiency is a good thing, but that's not really the purpose of either a VMT tax or a congestion charge.

Second, such a pricing scheme, counterintuitively, would almost certainly increase the number of SUVs on the road, at least as a percentage of all cars.  The reason is that it would reduce the relative cost advantage of smaller, more fuel-efficient cars.  If we were to discount gas taxes below current levels but charge everyone the same average per-mile cost, then driving an SUV would suddenly be cheaper, while driving a smaller car would suddenly be more expensive.

Or think of it this way.  Suppose the nominal gas tax were set at $3.00/per gallon.  Someone driving a car that got 30 mpg would receive no discount by switching to the VMT tax.  But the owner of an SUV that got 15 mpg would receive a net discount of ten cents per mile, assuming the average VMT tax worked out to ten cents per mile.  The SUV would become a relatively better deal, which would mean more SUVs.

This isn't really a bug of the proposal.  Any VMT tax will reduce the relative cost advantage of small cars.  That's the point, in a sense.  Large cars and small cars both impose externalities that are not fully captured by the gas tax.  Small-car drivers should be encouraged to avoid congested roads, too.  And we want drivers who are heavy users of roads (local streets in particular) to pay more than drivers who are light users.

I think this would be an interesting experiment, not only to evaluate its efficacy, but to determine its counterintuitive effects.

The City's Comprehensive Housing Market Survey

In the fall of 2008, the City Council, concerned about the lack of affordable housing in the city, commissioned a comprehensive housing market study from BBC Research & Consulting.  BBC presented its final report to Council in March.  (H/t Katherine Gregor.)

I haven't had time to work through the entire 142-page report.  I disagree with some of what I've read, but there is much here to like -- and much that echoes this blog's themes.

BBC's very first recommendation is that the city "reevaluate the zoning and development process."  BBC recognizes that part of the problem is the role played by neighborhood groups and part of the problem is the lack of density; not surprisingly, these are linked:

Austin’s current process of evaluating applications for residential development is community based. The city’s zoning and land use regulations also reflect the city’s dedication to environmental preservation and commitment to smart growth.

These principles are part of what makes Austin a great city. However, they can conflict with providing affordable housing for residents and workforce.  In desirable areas where there is much demand for housing, anything that constrains the supply leads to increased housing costs.

We have identified several opportunities for the city to modernize its current development process that will reduce the barriers to affordable housing development in Austin.  These include:

• Reconsider the role that many neighborhoods groups are playing in development decisions.
• Develop a strong, citywide Comprehensive Plan that guides development and forms the basis for the acceptance or denial of development applications.
• Increase density by approving dense developments that offer opportunities for affordable, attached housing products.
• Educate residents about the need for workforce housing in Austin and the consequences of not meeting current and future needs for housing.

(Italics mine.)

And more on neighborhood groups and density two pages later:

The city’s current neighborhood-based planning process does very little to facilitate the development of affordable housing on a citywide basis. Some of the neighborhood plans have affordable housing as a goal; others do not. We were also told many times in our focus groups with more than 100 stakeholders that Austin has lost many affordable units to neighborhood resistance.

Austin is not unusual in this regard. Residents in every city and town are notoriously resistant to density, and the more affordable the project and the greater the density, the higher the resistance. Neighborhoods often forget that a desirable city will grow; they cannot stop this momentum. Restricting workers from obtaining housing in an area does not mean these workers will go away— they may live farther away, but they still need to drive to work. Growth limits almost always lead to increased traffic congestion and the leapfrog effect of affordable housing being pushed farther and farther from employment centers.

Neighborhoods often use declining property values as successful arguments to fight affordable housing developments. Many academic studies have adeptly demonstrated that the effect of density and affordable developments on property values is not negative.

As I said, there is a lot more to this study, and it flags other problems with Austin's housing market.  I intend to cover some of the other points after I've finished working through the study.  But it's good to see the city hired consultants willing to provide a clear-eyed assessment of the city's problems.