Extraordinary Observations

NPR's Morning Edition aired a nice story on Tuesday about the Obama administration's interest in high speed rail investment. Given my enthusiasm for urbanism and transit, it may surprise some readers to find out that I do not get particularly excited about the prospect of high speed rail. It's not that I'm opposed to it in principal, but I don't do nearly as much traveling from city to city as I do around cities. I don't think the air travel is nearly as bad as people make it out to be (although this could stem from my uncharacteristic loyalty to Southwest Airlines) and I question whether a high speed rail network really makes logistical sense.

One of the biggest downfalls of air travel (and biggest potential advantages of high speed rail) in our country is that airports are almost always located on the fringes of cities - making it particularly inconvenient to travel from a city's airport to the city itself. To me this is an opportunity to improve rail transit access to airports; by which I mean transit stations within walking distance of airport terminals. Airports that can be used as models would be both of Chicago's airports, Cleveland, Washington National, Atlanta, San Fransisco International and Portland. The ability to step off of a plane and ride a single rail line to the city is convenient, easy and affordable. As more cities build transit lines to connect their cities to their airports, one of the biggest inconveniences to flying will be slowly eliminated.

I hope progress toward better passenger rail service is made where it makes sense, and if all this enthusiasm for passenger rail investment spurs more discussion and action for local transit systems, then I really can't complain.

Felix Salmon's new piece in Wired is an extremely interesting look at an aspect of the financial meltdown I didn't know much about. The situation raises a lot of questions about the advantages and the limitations of data - or more accurately, how we apply data. The reliance we now put on numbers is actually sort of amazing. A hundred years ago very little data even existed. What did was difficult to utilize, as any calculations had to be made by hand, with nothing more than pencil and paper. Advanced statistical techniques either hadn't been invented or were hardly reasonable to carry out.

It's not that having a lot of data is inherently a bad thing; it's that there are seemingly endless mathematical functions and formulas we can apply to them, and then even more interpretations and conclusions to draw. I always cringe a little when people use phrases like "it's all in the numbers" or "this is what the data says" because numbers themselves can't talk - it still takes a fallible person to draw a conclusion.

In preparation for this year's March Madness tournament, I've been building a data set with historic statistics from college basketball teams that played in the NCAA tournament since 2002. I originally intended to crunch the numbers myself and see if they could be of any use in filling out my bracket, but the more I look at these basketball statistics, the more I realize that there is no perfect method for crunching these numbers. I'm thinking a more interesting social experiment would be to give a handful of math and economics PhDs their own copy of the data set and a blank bracket the morning after Selection Sunday and see what they come up with. My guess is that if ten of PhDs entered my pool, there would be ten unique brackets; and of course, they can only be a single winner.

Last Friday's episode of NOW is one of the best I've seen on the politics of transit systems in America.



There are a ton of takeaways from this episode, but one that I drew attention to recently is the issue of local leadership. I don't know much about Pat McCrory, but along with other mayors like John Hickenlooper of Denver, the issue of transit development has been at the heart of their agendas. When we look at various cities that are building new transit systems and pushing urban development and wonder why some are excelling and others continue to flounder, it may be as simple as taking a look at who is running city hall.

This brings up a dillemma for transit advocates. Given that some cities have leaders that are more willing to embrace transit and properly develop their cities around these systems, is it beneficial to concede projects in some cities if it means more progress in others? In other words, is it better to build two mediocre transit systems, one in a place with supportive local leadership and the other in a place with indifferent leadership? Or should federal funding be distributed in some way taking into consideration how effectively local leaders will make use of these systems?

I've come across a few bloggers recently linking to a Pew Research report about whether Americans would prefer to live in a place with more McDonald's or Starbucks stores. For whatever reason, we are obsessed with categorically based data. What do women like? What do young people prefer? How about minorities? The problem, of course, is that a single person falls into multiple categories, and not all of them are always good descriptive fits.

If someone from Pew were to call me and ask whether I would prefer to live in a place with a Starbucks or a McDonald's, I would without a doubt answer Starbucks. Now, I'm not particularly likely to visit either of these chains, but I think the only time I've eaten at a McDonald's in the last year or two has been at 6am in an airport when there is basically nothing else available. Starbucks, on the other hand, isn't a place where I would buy a cup of coffee on a regular basis, but it is a place where I would go to meet a friend or type some blogs or just get out of the house. I can't imagine ever wanting to hang out at a McDonald's just for the sake of it.

But if you look at the different categories in the pew survey and try to apply them to me, you probably won't be able to come to a conclusive answer as to whether I would prefer Starbucks or McDonald's. These are the categories to which I belong and the group preferences:

Male (McDonald's +16)
18-29 (Starbucks +13)
White (McDonald's +8)
Some College (McDonald's +2)
Less than $30k (McDonald's +23)
Liberal (Starbucks +13)
Midwest (McDonald's +19)

In this case is my age more important than my gender? Does my political affiliation trump my geographic location? And why? In the next few years I might move from the education category of "some college" to being a "college grad" and hopefully will make a little more money when I have a career - both of which would make me more likely to prefer Starbucks. But at the same time, the older I get, the less likely I am to prefer Starbucks based on age. I might move from the Midwest, but more likely east than west, which would make me only marginally less inclined to prefer McDonald's.

Of course, even if every category I fell into prefers McDonald's, it shouldn't be surprising if I still choose Starbucks; these categories aren't all or nothing. But that's the extrapolation that a lot of people want to take from studies like these. If more liberals prefer Starbucks than McDonald's, all of a sudden all liberals prefer Starbucks. It just isn't the case. While I appreciate reports like these for the interesting perspective they provide, the conclusions drawn as a result of them are often intellectually bankrupt.

I didn't think I'd be able to go without eventually writing a response to David Brooks's recent column about suburban sprawl, car dependence, and America's love affair with them, but plenty of people have already beaten me to the punch; Ben Fried's and Alex Marshall's pieces are both particularly good, so I will leave you with them to provide you with a dose of Friday afternoon reading material.

PBS's new Frontline documentary on the brief history of the financial meltdown is awesome. Seriously. This is by far one of the best hours you can spend during your week.



Admittedly, before seeing this, I only knew this story from bits and pieces I'd collected along the way. Now, the whole chain of events is a lot more clear.

I didn't think I would love Tom Vanderbilt's new book Traffic when I picked it up a few weeks ago; after all, I can't stand driving, so what value would there be for me in a book with the subtitle "Why We Drive the Way We Do (and What It Says About Us)"? As it turns out, Vanderbilt's book is one of the best I've read this year, and has really got me thinking about the psychology of driving in a lot of situations.

When it comes to opinions about police, some people have extreme opinions one way or the other, and they tend to be rather outspoken; but most would accept that they are more or less necessary to keep society functioning. That is, of course, unless they're in a car.

If we're on foot, we like seeing officers on the street. We appreciate the fact that if we see something we don't like we can just yell out across the street and help will be right on the way. And we like the fact that the presence of police should theoretically deter crime from happening in the first place.

But when we're in a car - cops are a nuisance. They are an annoying bunch of bureaucrats who are out giving tickets just to meet some quota. We wonder why they are busy chasing down poor drivers and not our dealing with "real crime". Some have been conditioned in a Pavlovian manner to have strong emotional reactions when the drive by a police car at all.

One unintended consequence of America's car culture is that police on foot patrol in many cities have been moved to cars. Unfortunately, that may not have done much to better their image.

Richard Florida's cover story in the March issue of the Atlantic is worth reading for a lot of reasons, but my favorite takeaway is a paragraph buried toward the middle:

Suburbanization—and the sprawling growth it propelled—made sense for a time. The cities of the early and mid-20th century were dirty, sooty, smelly, and crowded, and commuting from the first, close-in suburbs was fast and easy. And as manufacturing became more technologically stable and product lines matured during the postwar boom, suburban growth dovetailed nicely with the pattern of industrial growth. Businesses began opening new plants in green-field locations that featured cheaper land and labor; management saw no reason to continue making now-standardized products in the expensive urban locations where they’d first been developed and sold... But that was then; the economy is different now. It no longer revolves around simply making and moving things. Instead, it depends on generating and transporting ideas. The places that thrive today are those with the highest velocity of ideas, the highest density of talented and creative people, the highest rate of metabolism. Velocity and density are not words that many people use when describing the suburbs. The economy is driven by key urban areas...

Suburbs, in a lot of ways, were always inevitable. A growing population has to go somewhere, and the inner-rings that surround cities make sense. The problem, it seems, is that the sprawl simply got out of control, creating what a commenter recently called "doughnut cities" where the once dense population centers are now abandoned and blight-ridden. Metro areas with growing populations may be able to fill in these holes in the future - but the fate of metro areas with declining populations looks a lot scarier. If Florida is correct, then these doughnut cities are toast, as the suburbs that circle them simply won't be able to keep pace with the economy.

After the discussion from a few weeks ago about the role that sprawl has played in the USPS's financial situation, I caught a letter to the editor in last Thursday's Plain Dealer that I thought deserved a response. The letter comes from Susan DiFranco of Columbia Station, Ohio. A town of about 7,000 on the outer-suburban fringe of Cleveland.

I recently read that the U.S. Postal Service is having financial difficulty and is weighing its options. One option is to deliver five days a week, which would save approximately $3.5 billion. A lot of businesses operate Monday through Friday; I don't see a huge problem with a five-day-a-week delivery system.

The other option is to close down rural stations, leaving their residents no post offices at all. This option would be the same as shutting down all of the post offices in the city and having all of the mail coming and going from the main post office on Orange Avenue. How many people would want to drive down to Orange Avenue to pick up a package or drop something off to be mailed? Not very many, I'm guessing!..

The reality of mail service is that the USPS has always subsidized rural (and recently suburban) areas at the expense of cities. The reason is simple: a stamp costs a fixed price regardless of whether a piece of mail is going to the middle of a city or the middle of nowhere. The difference, of course, is that a single mail carrier, driving a single mail truck, can only deliver a fraction of the mail to a rural area or a low-density suburban area than a mail carrier in a city can deliver.

Cutting all mail service to 5 days per week would save the USPS enough money to balance it's books, which is OK if society decides we ought to treat urban, suburban and rural areas equally - and there is good reason to think we shouldn't any more. When the the Postal Service started delivering mail early in America's history, people either lived in rural areas because agriculture was their livelihood or in urban areas because that's where they made a living. Since everyone lived where they did for strict economic reasons, it made sense offer similarly priced mail service to everyone. Only recently have people started living in places like Columbia Station "just because". If we decide that we ought to stop subsidizing rural and suburban mail service, here are some options to consider:

We could charge senders more based on the distance a piece of mail is traveling and what type of area it is going. The private companies are already doing this. I used to have to do a lot of overnight shipping via FedEx. Most of the time we were shipping to clients in major cities, but occasionally we had to ship to an area well outside of a major metro area. In those instances, FedEx charged a big premium for deliveries. Unfortunately, on first class letters, expecting senders to calculate the cost of postage based on the zip code the piece is going might not be entirely reasonable.

Or we could charge receivers based on where they live. This could be implemented in a few different ways. We could make individuals pay with their time and the cost of driving to pick up their mail from a central mail center. Or the USPS could charge individual households a monthly fee for daily service; some might opt-in, others might not. Local municipalities might also pay the USPS for the service, either by collecting revenue through a special tax or using general revenue funds to pay for the service. Or we could implement a combination of these ideas.

In the end, DiFranco's claim that imposing a cost for postal service on rural dwellers is as bogus as doing the same for city residents isn't an entirely fair comparison. I obviously don't have enough data to know how much each of the above options could help the USPS balance its books, but they would at least make different areas more appropriately bear the costs they impose on the postal system.

Tom Friedman's column from earlier this week proposes an interesting idea for stimulating our economy:

Leave it to a brainy Indian to come up with the cheapest and surest way to stimulate our economy: immigration. “All you need to do is grant visas to two million Indians, Chinese and Koreans,” said Shekhar Gupta, editor of The Indian Express newspaper. “We will buy up all the subprime homes. We will work 18 hours a day to pay for them. We will immediately improve your savings rate — no Indian bank today has more than 2 percent nonperforming loans because not paying your mortgage is considered shameful here. And we will start new companies to create our own jobs and jobs for more Americans.”

I mentioned this idea to a co-worker, originally from Pakistan. He said he personally knew of more than 100 people who would jump at an offer like this, even if buying a home was a pre-condition to receiving the Visa. That seems pretty amazing to me.

A hot new report from the Urban Land Institute posits that residents in the Washington DC metro area can't necessarily find more affordable living in far-flung suburbs like they used to be able to. The reason is relatively simple: any savings in housing are offset or even surpassed by commuting costs; non-car alternatives and a high number of District and inner-suburb residents who get around by transit, bicycle, and foot brings down the average cost of getting around for those close to and in the city.

On a recent trip to Washington, DC a friend of mine remarked that high rents in the area are overrated because the costs are easily offset by the fact that you can live a fulfilling life without a car. The point may seem obvious, but becomes more clear when you actually do the calculation. I've pointed out the statistic from AAA that the average cost of car ownership is now about $8000 per year. Given that cost structure, perhaps you can only afford monthly rent of about $700. Without owning a car, and without incurring any additional costs, you might be able to afford a place at $1300 a month.

An interesting consequence of the ULI report is that it should alter the way we think about "affordable housing". Public policy that strives to provide more affordable housing doesn't necessarily mean rents need to be cheap - policy provides solid non-driving alternatives could have a similar effect.

A creative metaphor for the decline of walkability in America:

Last week the Center for American Progress rounded up various rankings of the best cities in America to be without a car. For the most part, there is a pretty strong consensus about which American cities are best for non-drivers, and a lot of the places are ones I identified last summer.

With all the talk about federal stimulus dollars and what projects the money should and should not be funding, it can be easy to forget that good, walkable urbanism can't exactly be forced down the throats of cities that don't want to embrace it. The places identified above got to where they are primarily thanks to local leadership committed to creating and maintaining these types of places. Some of them may have been built in part with federal dollars, but so have highways and roads the line America's least pedestrian-friendly cities. At the end of the day, the old adage that "all politics is local politics" still rings true.

Ryan Avent's new piece in the American Prospect is excellent in detailing how stimulus provisions that exclude funding for transit and rail are a step in the wrong direction; one paragraph in particular caught my attention:

Perhaps worst of all, the Senate, like the House, declined to specifically
direct funding toward operating costs for transit systems. While capital
spending to repair and enlarge transit systems is absolutely necessary to meet
long-term environmental (and economic goals), those investments do nothing to
keep trains and buses running right now. With gas tax and general budget
revenues plummeting, systems nationwide are cutting service, increasing fares,
and sacking employees. And while grants to state governments may be used to
cover some of the shortfall, state officials will face strong pressure to plug
other holes first, stimulus concerns aside. Multi-jurisdictional systems in
particular may be out of luck, as governments prove reluctant to devote money to
systems that serve non-constituents.

Political rhetoric frequently mentions "creating new jobs" and "putting people back to work" but often overlooks preserving the jobs that people already have. In the wake of the economic meltdown, transit agencies across the country are looking for ways to slash costs, including letting some of their paid employees go. If ten-thousand transit workers become unemployed because of budget cuts and they all find work in other industries thanks to government's spending, there technically would be no net increase in jobs (or net decrease in unemployment).

But politically the distinction is huge. When transit workers get tossed, the blame gets chalked up to the terrible economy, decreasing tax revenues, or something else seemingly out of anyone's control. When a "new" job gets "created" thanks to government spending on a new capital project, for example, congress expects credit for the accomplishment. In some perverse way, it is more politically beneficial to let a thousand people lose their jobs, go through the stress of unemployment and put them to work elsewhere than to simply preserve the positions they already hold.

Of course, this sort of creative destruction is necessary to prevent progress from being hindered. When the typewriter industry started dying, it would have been foolish for government to artificially prop up typewriter companies when the personal computer was revolutionizing technology. Things are different with transit, though. Demand is at its decade's long high and it is one of the most proactive tools we can use to clean up the environment and fight global warming. Thanks to skewed political incentives, unfortunately, it is proving to be difficult to find a sympathetic ear in congress.

Tom Vanderbilt links to a interesting new study by Ohio State University psychologist Dennis Shaffer. When asked to describe the length of the white stripes on streets and highways, people significantly underestimated reality:

Each dashed line measures 10 feet, and the empty spaces in-between measure 30 feet. So every time a car passes a new dashed line, the car has traveled 40 feet. But in this study, people consistently judged the lines and the empty spaces to be the same size, claiming that both were two feet.

This is an amazing finding to me and shows that we aren't just bad at estimating distance, we're terrible. For every 40 feet that we travel, the brain apparently thinks we travel a mere 4 feet.

I wonder if this is the same reason why people typically describe distance with non-constant variables? When I lived in Dallas, people almost never described distance in miles, it was always in minutes or some otherwise subjective unit of measurement. It drove me crazy; but not just because you can't measure distance in minutes, but because people consistently underestimated the number of minutes it took to get from point A to point B. For instance, something "5 minutes away" was often a 15 minute drive. A "half hour drive" could be closer to 45 minutes or more. Something "just down the street" could be 5 or 10 miles away. At first I thought people were purposely doing this so that it would be easier to convince their friends to travel farther than they might want to; but after a while I realized that it was a consistent behavior, and people's sense of time and distance truly seemed to be warped.

I finally got a chance to watch the abbreviated version of the I.O.U.S.A. movie. If nothing else, it is an eye-opening look at an issue few people know much about. Check it out:

I came across this interesting tidbit about HDTV over the weekend:

In-Stat reports that 17 million of the 39 million U.S. households with HDTV, or 43.6 percent, don't watch in HD. The finding is based on a recent survey of consumers and defines "HD programming" as paid high definition services from cable and satellite providers and free broadcasting over the air. Packaged media, such as Blu-ray disc and video games, are not included.

Michael Paxton, an analyst for In-Stat, told CDFreaks that there are two main reasons why so many HDTV owners are still watching in standard definition. Cost was a major factor, with consumers saying they didn't want to fork over extra fees to lease an HD set-top box or to get HD channels. Also, consumers often said the amount of high definition programming wasn't enough to justify the extra effort.

I wonder if these results are skewed either by people who think they are watching shows in HD when they aren't or by people who don't want to admit they own an HDTV but watch it in sub-HD quality.

Where I live, NBC has been on Channel 3 since 1953, but if you tune to channel 3 on an HD-equipped cable box, you don't get any HD content; for that, you need to tune to channel 403. While this may seem like an insignificant detail, I am confident that there are plenty of people who are simply not technologically savvy enough to understand the distinction. Not to mention the fact that only a handful of the shows show on the 400-range channels are broadcast in HD. It seems plausible that someone could tune into Seinfeld on channel 443 one night, realize it looks exactly the same as the Seinfeld episode on channel 33, and conclude that both stations are showing content in HD (in reality, neither is). There may even be folks who have the HDTV, the correct cable box, and are tuning to the appropriate channels, but they don't have an HD-compatible cable connecting the box to the TV, so what they think is HD content is merely standard stuff.

I started watching shows in HD about four years ago and frankly I think that it is one of the best and worst things to happen to TV. The 'best' in the sense that plenty of shows are now broadcast in amazing quality - the 'worst' in the sense that there is an even greater incentive to watch them than to do something productive. Regardless, I suspect HD adoption rates are lower than these official numbers report. Unless they find a way to make it easier for viewers to access HD content, I think it will be longer than they might estimate for the technology to fully take over what we have now.

Bloggers have been linking to this Consumerist post from last week and describing it with words like “sad”, “depressing”, and “wow”. I’m not trying to argue that the economy isn't in bad shape, but judging the health of the economy based on how many resumes recruiters are receiving for open positions is a poor metric to use.

Imagine a simple hypothetical economy with 1000 people – all of whom are in the workforce. At 5% unemployment, things look OK. Out-of-work individuals are searching for jobs; others apply to new positions while gainfully employed or out of fear of losing their current job. Perhaps 100 out of the 1000 people in this imaginary economy are applying for work. Further, they send out an average of 10 resumes each. Thus, there are about a thousand resumes in circulation.

Now the hypothetical economy delves into deep recession and unemployment doubles. Desperate for work, people start sending out an average of 20 resumes each. So now there are 200 people sending an average of 20 resumes; the number of resumes in circulation quadruples to 4000. As a result, stories start popping up on the internet about absurd numbers of resumes being received for this position or that, causing out-of-work individuals to become even more frantic. Fearing that the situation is rapidly deteriorating, they start applying anywhere and everywhere; the average number of resumes they send jumps again to 40 per person. 8000 resumes are now being sent.

So, this hypothetical economy went into a horrible recession and unemployment increased 100%, but the number of resumes in circulation jumped an overwhelming 700% - to the point where the numbers now look like this:

1000 = Total population
200 = Looking for work (including unemployed)
100 = Unemployed
8000 = Resumes in circulation

Given the events that transpired, there are now 7 times as many resumes as people in this economy, 39 times as many resumes as people looking for work, and 79 times as many resumes as unemployed individuals.

In the end, this behavior hurts everyone. Employers receive an excessive number of resumes, from people who are wildly overqualified to those who apparently didn’t even bother to read the “qualifications” section of the job posting. Recruiters are bogged down by so many resumes that it becomes incredibly difficult to analyze each candidate, which in turn hurts the candidates themselves, who are more likely to have their resume overlooked because it is just a piece of paper in a stack of dozens/hundreds/thousands.

In other words, in this simple model, whether there are 4000 resumes in circulation or 8000 doesn’t change the fact that there are 100 unemployed individuals and 200 people looking for work. Articles that use these resume metrics and employ phrases like, “you know the economy is bad when..” and “what does it say about the economy when..” are misleading and only exacerbate the problem by subliminally encouraging people to send out more and more resumes to any place that is hiring, regardless of whether it makes sense or not. A more appropriate hypothetical question to ask as it relates to this dilemma is, "what does it say about people's behavior when.."

Of course, the more resumes that people send, the more value there is to interacting with recruiters directly or getting to know someone who already works at a company. If nothing else, it allows employers to bring in at least partially-known candidates for interviews and alleviates recruiters from digging through thousands of resumes. Those who aren't aware of this phenomenon will continue blindly sending out resumes, and then in an interview with the local newspaper or TV news station, will dismally report, "I sent out 150 resumes and didn't hear back about a single one!" and the vicious cycle begins again.