Houston (Street), we have a problem.

A lesson in the elasticity of demand, prices and urban congestion. It looks like Uber, Lyft and other ride sharing services are swamping the capacity of New York City streets

Every day, we’re being told, we’re on the verge of a technological revolution that will remedy our persistent urban transportation problems. Smart cities, replete with sensors, command centers, and linked to an Internet of Things, will move us every more quickly and effortlessly to our destinations, traffic lights and even traffic will become a thing of the past. The most fully deployed harbinger of the change that such technology has wrought is clearly the app-based ride-hailing services, including Uber, Lyft and others. A few weeks back, we took to task claims from one study that a couple thousand autonomously piloted 10-passenger vans could virtually eliminate traffic congestion in Manhattan.

A new report from New York should give the techno-optimists reason to pause. Transportation consultant Bruce Schaller, who previously served as New York City’s deputy transportation commissioner, has sifted through the detailed records of New York City’s Taxi Licensing Commission and painted a stark picture of growing traffic congestion in the city–thanks to to increasing proliferation of app-based ride hailing services, which the report refers to as transportation network companies (TNCs). Schaller has a shorter op-ed summarizing his report in the New York Daily News, but you’ll want to have a look at the entire 38-page report, entitled “UNSUSTAINABLE? The Growth of App-Based Ride Services and Traffic, Travel and the Future of New York City,” The full report is worth a read, but here are some highlights.

  • The total number of miles driven in New York City by TNCs vehicles has increased by 600 million miles in the past three years.
  • Trips taken by the combination of taxis and TNCs has outpaced the increase in trips taken by transit; for the precdeing 24 years (since 1990) growth in person trips has been led by increased transit ridership.
  • The additional vehicle traffic associated with TNCs amounts to about a 7 percent increase in traffic levels in Manhattan, about the same amount of traffic that the city’s cordon-pricing proposal was supposed to reduce.
  • The bulk of the increase in traffic associated with TNCs has been in the morning and evening peak hours.
  • While TNCs initially grew mostly by taking traffic from yellow cabs, increasingly they are taking riders from transit, and in some cases stimulating additional travel.

Bottom line: the growth of TNCs is increasing the volume of vehicles on New York City streets, and adding to congestion and delays. Interestingly, this key finding is a turnaround for Schaller, who acknowledges that just a year ago (in January 2016), he was part of a team that concluded on behalf of Mayor de Blasio that the added traffic from ride hailing services wasn’t increasing the city’s congestion. The growth in TNC volumes over the past year has apparently changed his mind.

What’s changed? In short, the price of car travel has fallen in New York City, relative to the alternatives. Previously, a combination of high taxi fares and limited entry (a fixed number of medallions) coupled with prohibitively expensive parking rates in most of Manhattan, meant that taking a private car cost four or five times more times as much as the average transit fare. But with Uber and Lyft charging much lower rates, and flooding the market with additional vehicles, there’s been a noticeable uptick in traffic volumes. This is a fundamental lesson in economics: increasing the supply of vehicles and lowering prices is going to trigger additional demand. And in the face of limited street capacity, congestion is likely to increase. And we should keep in mind that the TNCs are just a dress rehearsal for fleets of autonomous vehicles. Subtract the cost of paying drivers, and they promise to be even cheaper and more plentiful than Uber and Lyft are today, especially in high density urban markets.

Ultimately, the solution to this problem will come from correctly pricing the use of the city’s scarce and valuable street space, particularly at rush hour. Schaller makes this very clear:

As they steadily cut fares, TNCs are erasing these longstanding financial disincentives for traveling by motor vehicle in Manhattan. If TNC growth continues at the current pace (and there is no sign of it leveling off), the necessity of some type of road pricing will become more and more evident.

The detailed data from Uber and Lyft however, point up the major limitations of the proposed cordon-pricing scheme that was suggested for New York City a decade ago (under cordon pricing, vehicles entering lower Manhattan, below 96th or 110th street, or crossing the Hudson or the East River would pay a daily toll). But because so much traffic and so many rides begin and end within those boundaries, the cordon pricing scheme does nothing to disincentivize travel in the center, once the vehicle has paid the toll.

Consequently, if pricing is going to work in Manhattan, it will probably have to be some kind of zoned, time-of-day pricing, charging higher rates for travel in and through Manhattan during peak hours, which much lower fees for travel in out-lying boroughs and at off-peak hours. In effect, Uber’s much maligned “surge” pricing is proof of concept for this model; it just has to have prices reflect the scarcity and value of the publicly owned roadway rather than just the momentary scarcity of Uber’s privately owned vehicles.  The GPS and mobile Internet technology that’s now been proven in taxis and TNC vehicles, shows a such a system is technically quite feasible.

Without some form of road pricing, the high concentration of profitable fares in denser neighborhoods and at peak hours, coupled with the additional financial inducement of surge pricing bonuses could lead ever greater volumes of TNC vehicles to clog city streets. As New York transportation expert Charles Komanoff puts it, the Schaller report settles the question that TNCs are making the city’s gridlock worst. He effectively calls the report a must read:

It touches on virtually every consequential transportation trend and policy question facing the five boroughs and stands as the most thoughtful and thorough analysis of New York City traffic and transportation issues since the Bloomberg years.

But New York is just on the leading edge of a range of technological and policy issues that every city is going to have to confront in the years ahead.  If you want to get ahead of the curve–and think about where the expansion of TNCs, and ultimately autonomous vehicles is taking us–here is a good place to start.

Yet another flawed congestion report from Inrix

Big data provides little insight

Cue the telephoto lens compressed photo of freeway traffic; it’s time for yet another report painting a picture of the horrors inflicted on modern society by traffic congestion. This latest installment comes from traffic data firm Inrix, which uses cell phone, vehicle tracking and GPS data to estimate the speed at which traffic moves in cities around the world.

Two words summarize our reaction to the new Inrix report: tantalizing and aggravating. The tantalizing part is the amazing data here: Inrix has astonishingly copious and detailed information about how fast traffic is moving, almost everywhere. The aggravating part:  its essentially just being used to generate scary–and inflated–statistics about traffic that shed precious little light on what we might do to actually solve real transportation problems. Its main purpose seems to be to generate press headlines: “Los Angeles Tops Inrix Global Congestion Rankings,” “Atlanta Traffic Among Worst in the World, Study Finds,” and other scary stories.

 

Same as for the last congestion report. (Atlanta Journal-Constitution)

One one level, its a truly impressive display of big data. Inrix has compiled 500 terabytes of data, for hundreds of thousands of roadway segments, from hundreds of millions of sources on more than a thousand cities around the globe. That’s a real wealth of information. Inrix casually slips in the factoid that average speeds on New York streets are 8.23 mph, versus 11.07 mph and 11.54 mph in L.A. and San Francisco respectively. But unfortunately, in this particular report, it has chosen to process, filter and present this data in a way that chiefly serves to generate heat, rather than shed any light on the nature, causes and solutions to urban traffic problems. If “big data” and “smart cities” are really going to amount to anything substantial, it has to be more than just generating high tech scare stories.

We’ve read through the report, examining its key findings and comparing it to previous work by others. We think there are four fundamental problems that readers should be aware of: the report has a new methodology, which while more detailed than previous reports, is neither comparable to them, nor a major improvement. Like other reports, the definition of congestion is unrealistic, and its cost estimates are exaggerated (with no acknowledgement that building enough capacity to reduce congestion would be even more expensive, and likely be ineffective). Most importantly, like all travel time index measures, the Inrix methodology ignores differences between average travel distances in cities, which effectively penalizes denser, more compact cities. Its disappointing to see so much data providing so little insight into what we might do to understand and solve these problems.

Methodology: New and non-comparable, but not significantly different or better

First off, Inrix has revised its methodology and definitions for computing and presenting metro level congestion statistics. They’ve segmented their data by time of day and trip characteristics, which in theory provides a more nuanced view than earlier work. But it also means that results in this year’s studies aren’t comparable at all to the data (and claims) made in earlier Inrix reports (which we’ve raised some questions about). Inrix has also changed the geographic definition of what constitutes a city or urbanized area. As a result, this report can’t tell us whether congestion is getting worse or better, or shed light on the strategies and investments that may have actually led to reductions in congestion in different cities around the world.

The failure to report data on a consistent basis over time undercuts our ability to use it to make sense of the world. For a long while, Inrix actually gathered and reported its data on a consistent monthly basis in a way that allowed independent observers to view congestion and travel time trends. This data actually showed traffic congestion easing in most metropolitan areas in the United States from 2009 2014. Inrix stopped reporting this data in 2014, and scrubbed the links to it from its website (although the original data still live on in a Tableau server you can see here).

But while the numbers are new and non-comparable, they appear to mostly be saying same thing that we were told in earlier congestion reports.  So for example, we’ve graphed the 2016 Inrix Congestion Index (ICI) against the 2012 Texas Transportation Institute Travel Time Index (TTI). The two indices are calculated in different ways, and TTI runs from about 1.15 to 1.45 for each metro area while ICI runs from 1 to 19. (In both cases, the travel time index measures how much longer a trip takes due to congestion than it would during free-flow conditions; ICI omits an implied 1.0.)  But the two measures are highly correlated (r-squared of .69), which says they’re really measuring the same thing in pretty much the same way, making ICI in many ways simply a gussied up, really-big-data version of the Texas Urban Mobility Report’s Travel Time Index, with all its attendant flaws.

An unrealistic definition of congestion

Second, the definition of congestion is a novel and expansive one:  Any time travel speeds fall below 65 percent of free-flow speeds, Inrix regards this as being “congested.” Inrix says it determines free-flow speeds using actual traffic data. As we and other have noted, this approach often results in using speeds that exceed the posted legal speed limit for a roadway as the baseline for determining whether a road is congested.  For example, if “free flow” speeds on a posted 55 mile per hour road are 60 or 65 miles per hour, Inrix would presumably use this higher baseline for computing congestion. This has the curious implication that the inability of a motorist to engage in an illegal behavior constitutes a “cost.” Also: its worth noting that roadways achieve their maximum throughput (number of vehicles moved per hour on a roadway segment) and speeds that are usually much lower than free flow speeds.  (At higher speeds, drivers increase their following distance and the road carries fewer cars per hour). So in many cases, these lower speeds (say 40 miles per hour on a 55 mile per hour roadway, where free-flow speeds are 60 miles per hour), may actually be more efficient.  As we pointed out in our essay “The Cappuccino Congestion Index,” no one expects businesses to have has much capacity to provide the same service at peak hours that they do in slack times.

Exaggerating costs

Third, while Inrix claims to have estimated the “cost” of congestion to travelers, these estimates are suspect for a number of reasons. Inrix uses a value higher than most other studies–almost $20 per hour for commuter travel time (a $12.81 wage rate, multiplied by 1.13 occupants per vehicle multiplied by 1.37 to reflect the aggravation of congestion). But real world experience shows that commuters actually value travel time savings at something more like $3 per hour. It also appears that there’s been a major shift in the monetization of congestion costs: Older studies like TTI, estimated dollar costs based on the additional time spent on a trip due to congestion: So if a trip that took ten minutes in un-congested traffic took a total of 15 minutes in a congested time period, they would monetize the value of the five minutes of additional time spent. The Inrix report appears to monetize the total value of time spent in congested conditions, i.e. anytime travel speeds fell below 65 percent of free flow speeds.  It’s actually hard to tell exactly what they’ve done, because their explanation is at best somewhat cryptic:

The direct costs are borne directly by the car driver through their use of the roads in congestion, and include the value or opportunity cost of the time they spent needlessly in congestion, plus the additional fuel cost and the social cost of emissions released by the vehicle. (Page 8)

In our example, if congestion were evenly distributed over this same 15 minute journey, it appears than Inrix would monetize the entire 15 minutes as “time spent in congestion.” This has the effect of greatly increasing the estimated “cost” of congestion.

As we’ve pointed out before, despite the impressive sounding estimates of the value of time lost to congestion, the key question that the Inrix report begs is whether the the cost of building enough roadway capacity to eliminate congestion would somehow be less expensive than the supposed value of lost time due to congestion. Its likely the cost of building enough capacity to eliminate congestion would dwarf travel time savings–and that’s before considering the induced traffic that added capacity would add. We know from thorough academic studies like Duranton and Turner’s fundamental law of road congestion, and practical experience with freeway widening projects in Los Angeles and Houston, that spending billions of dollars on more capacity doesn’t reduce congestion, it increases traffic.

Ignoring distance, discounting accessibility

Fourth, the Inrix Congestion Index, like the TTI travel time index still has the major flaw of overlooking the differences in average travel distances between cities. Some cities have much shorter commutes than others (usually because of much more compact development patterns), and while a larger percentage of trips may occur during “congested” time periods, the total duration of trips is far shorter in these more compact metro areas. Consider two cities, one with a five mile average commute and the second with a ten mile average commute.  Suppose that in both cities, drivers drove an average of two miles in “congested” conditions: Inrix would tell us that in the first city 40 percent of commuter travel was “congested” while in the second only 20 percent (2 of 10 miles) was congested. Even though both sets of commuters experienced the same amount of congestion, the more compact city had trips that were half as long. As we’ve shown this percentage-based congestion index is profoundly biased against compact cities with short average travel distances. Its frequently the cases that the average commuter in a city with a high congestion score will have a shorter duration commute time than someone living in a city with a low congestion score because they don’t have to travel as far.

Why can’t big data tell us something useful?

Think of Inrix as a test case for big data and smart cities. It has to do something more than simply serve as a high powered tool for p.r. and pro-road talking points. It should be an analytical tool which helps us figure out what works, and what doesn’t, what to do more of and what to stop doing altogether.

For example, this kind of data , should help us judge which cities are doing well, and why, and how we can learn from their successes.  According to Inrix, Birmingham and Oklahoma City have some of the lowest levels of traffic congestion (at least as measured by their flawed ICI) of any of the 50 or so largest metro areas.  Is there actually something that these cities are doing, or some keen insight other cities can learn that will show how to reduce traffic congestion? Absent a framework for connecting this data to policy–and for correcting the biases against compact, accessible development that are implicit in the travel time index/ICI–this data isn’t terribly useful for setting transportation policy or deciding on how best to invest in transportation infrastructure.

As we’ve argued at City Observatory, you can’t address transportation policy without a clear model of why we have congestion in the first place. There’s overwhelming evidence that roads get congested at peak hours because we’ve set too low a price for road use. When we actually charge even a modest price for road use, congestion problems evaporate (see our story about Louisville). The report’s authors, economists Graham Cookson and Bob Pishue, clearly understand that there’s something more to the traffic problem than the scary stories and big numbers presented in this report.  In blogs at Inrix, both highlighted the importance of demand side strategies, specifically including road pricing.  Graham Cookson wrote that we need to be “encouraging the efficient use of our roads through wider adoption of road user pricing.” Pishue acknowledges that pricing roads would reduce congestion, but apparently frets that this requires a change in behavior: “Demand-side strategies like road user pricing and flexible work schedules can be effective, yet rely on changing driver and economic behavior.” Unfortunately there’s no reference to these policies in the Inrix report itself; the word “pricing” simply doesn’t appear anywhere in its 44 pages.

Last year, we gave Inrix a grade of D for the last iteration of this report. This year we’re dropping that down to an incomplete. Inrix clearly has a wealth of data that could tell us a lot about how well our transportation systems perform, but so far, it appears that they’re chiefly interested in generating headlines, rather than providing the kind of analytical tools that could help inform policy choices. We hope they’ll do better in the future.

 

The Week Observed, February 24, 2017

What City Observatory did this week

1.Busting the urban myth about high income housing and affordability. One of the most widespread beliefs about housing is that the construction of new high income housing somehow makes the housing affordability problem worse. Widely believed, but wrong. We marshal the economic evidence for filtering–how as apartments and houses age, they decline in relative price and become more affordable to progressively lower income households. More importantly, failing to building more high income housing has the effect of pushing higher income households down-market, leading them to bid up the price of more modest housing that would otherwise be occupied at lower prices or rents by the less affluent.

2. Playing Apart.  In our report Less in Common, we described the multi-faceted trend of American’s living their lives apart from one another, and segregating by income, interests and social class. This trend is most evident in recreation, as Americans have fitted their homes and yards with pools, play courts, and swing sets, and foregone the public park. But as Samsung highlighted in its commercial “A Perfect Day” there’s still a longing or a lingering aspiration to have the kinds of public spaces that we can all share.

3. Cursing the candle. There’s growing evidence that some areas of Detroit are experiencing a rebound. As we chronicled at City Observatory, Wayne County has racked up five years of consecutive job gains. But some critics are saying that claims of a rebound are overblown, and argue that the uneven geography of job and population growth within the city is aggravating inequality. We consider these claims, but note that any turnaround its going to start small, and at least initially take hold only in a few places. Indeed: building critical mass in a few locations is essential to achieving the kind of urban spillovers that make development viable and self-sustaining. Its not realistic to expect that there’s any solution to Detroit’s woes that will fix everything, everywhere for everyone all at once.

Must read

1. Taxing Parking. We’re a little late to the party here, but several cities in the UK are taking a look at copying a very successful policy implemented by the city of Nottingham. They’ve imposed a tax on private, employer provided parking and used the proceeds to help subsidize transit service, expand the city’s tram (light rail) system, and make improvements to bike and pedestrian facilities. Like the purely private efforts of employer Lyft in San Francisco, the net effect of this system is to encourage those with alternatives to driving to work to use them, and provides a ready source of funds for encouraging greener commuting.

2. Want to reduce traffic?  Build more housing. Writing in Bethesda Magazine, Ben Ross makes the case that this Washington DC suburb is basing its future planning efforts on outdated projections of automobile use. As the city’s demographic profile changes and as more people are seeking to live in walkable, mixed use communities, increasing population is no longer associated with greater and greater levels of driving. Rather than increasing as projected by traditional traffic models, traffic levels on some of the city’s key arterials have actually gone down. Instead of ignoring this trend, the city ought to take advantage of it by encouraging more housing in locations that enable people to drive less.

3. Ford’s CEOs says the future of cities has almost nothing to do with cars. We singled out Ford for criticism a couple of weeks back, noting that their view of the future of cities was very vehicle and travel centric. Despite the hopeful sound the the title of this article in Business Insider, while Ford’s Mark Fields is de-emphasizing cars, the core of their vision is about moving people and vehicles by throwing more technology at cities. As he puts it: there goals is shifting from getting cars through an area to “how do we maximize people getting through the areas.” But as we pointed out, maybe the objective of city planning ought to be to prioritize the experience of the people in cities rather than those who are simply passing through. If we arranged land uses so that we didn’t have to travel as far or as often, we could avoid a lot of the needless and disruptive investment in having to move so many people in the first place.

New Research

1. The economic implications of housing supply. Two of the nation’s leading urban economists, Ed Glaeser and Joe Gyourko, have collaborated on an essay exploring housing’s link to urban prosperity for a forthcoming issue of the Journal of Economic Perspectives. They make the key point that the variations in housing costs among locations in the US have relatively little to do with variations in construction costs, and almost everything to do with land costs. While the price of land is influenced to some degree by geography (elevation and bodies of water restrict where cities can expand), the big variable is local land use policies: how much  is permitted to be built and in what locations. They conclude that about three-quarters of American homeowners live in housing that is priced at or below its marginal cost of profitable production, but one quarter of homes, mostly in tight coastal markets command prices much higher than typical construction costs. In addition to addressing housing affordability, Glaeser and Gyourko also examine the claim by Enrico Moretti that constrained housing supplies in highly productive metropolitan areas (like San Francisco) have the effect of reducing national income. While they concur in the logic of this analysis, they tend to think that the magnitude of the cost to the national economy is less than estimated by Moretti.

2. Visualizing neighborhood change. Ken Steif of Urban Spatial has used longitudinal census data from the period 1990 through 2010 to compare neighborhood level changes in income, housing values and educational attainment in a series of 29 mostly “rust belt” cities. You’ll find a series of maps that illustrate which neighborhoods are gaining or losing ground relative to city-wide average changes. And Spatial Data also presents regression results that show the which neighborhood characteristics in 1990 are most closely correlated with relative neighborhood improvement in each variable over the succeeding two decades. For example, low rents, a high fraction of “non-family households”, welfare recipiency and transit use were negatively associated with changes in home values over the 1990 to 2010 period.

City Observatory in the News

1. Last week’s commentary on not demonizing driving was named one of the five best ideas of the day for February 21 by Time.

2. That same commentary was translated into Portugese–“Não é preciso demonizar automóveis — basta parar de subsidiá-los”–for the Brazilian publication Caos Planejdo.

 

Cursing the candle

How should we view the early signs of a turnaround in Detroit?

Better to light a single candle than simply curse the darkness. The past decades have been full of dark days for Detroit, but there are finally signs of a turnaround, a first few glimmers that the city is stemming the downward spiral of economic and social decline. But for at least a few critics that’s not good enough: not content with cursing the darkness, they’re also cursing the first few candles that have been lit, for the sin of failing to resolve the city’s entire crushing legacy of decline everywhere, for everyone, and all at once.

Flickr: Uetchy

Michigan State political scientist Laura Reese and Wayne State urban affairs expert Gary Sands have written an essay “Detroit’s recovery: The glass is half-full at best,” for Conversation which was reprinted at CityLab as “Is Detroit really making a comeback?” The article is based on a longer academic treatment of this subject by Reese, Sanders and co-authors, entitled “It’s safe to come, we’ve got lattes,” in the journal Cities.  (This is one of those rare cases where the mass media version of an article is more measured and less snarky than the title of the companion academic piece, but I digress.)

Reese and Sands set about the apparently obligatory task of offering a contrarian view to stories in the popular press suggesting that Detroit has somehow turned the corner on its economic troubles and is starting to come back. We, too, are wary of glib claims that everything is fine in Detroit.  It isn’t. The city still bears the deep scars of decades of industrial decline coupled with dramatic failure of urban governance. The nascent rebound is evident only in a few places.

There’s a kind of straw man argument here.  Is Detroit “back?” As best I can tell, no one’s making that argument. The likelihood that the city will restore the industrial heyday of the U.S. auto industry, replete with a profitable oligopoly and powerful unions that negotiate high wages for modestly skilled work, just isn’t in the cards.  As Ed Glaeser has pointed out, it’s rare that cities reinvent their economies.  But when they do–as in the cases of Boston and New York–it’s because they’ve managed to do an extraordinary job of educating their local populations, and that base of talent has served as the critical resource for generating new economic activity. Detroit’s still far from that point.

And there’s no one who should think a renaissance will happen quickly if it happens at all. History is littered with examples of once flourishing cities that failed for centuries to find a second act: Athens was long deserted, Venice had its empire and economy collapse, Bruges had its harbor silt-up. In each case, these city’s early economies lived hard, died young and left a beautiful (architectural) corpse.  It’s really only been in the 20th century that each of these cities revived to any degree after their historical decline.

That said, there’s clear evidence that Detroit has stanched the economic hemorrhage. After a decade of year over year of job losses, Wayne County has chalked up five consecutive years of year-over-year job growth. True, the county is still down more than 150,000 jobs from its peak but has gained back 50,000 jobs in the past five years.

While this article presents a number of useful facts that remind us how far Detroit has to go, there are a lot of unresolved contradictions here.  In successive paragraphs, the authors decry the lackluster performance of Detroit home prices (they’re still way below housing bubble levels and haven’t rebounded nearly as well as in other cities), but then go on to decry the unfolding gentrification of the city.  You can’t have it both ways.  Either housing is cheap and devalued, or the city is becoming more expensive.

Why neighborhood level equality is a misleading metric for urban well being

Reese and Sands seem to be upset that Detroit’s nascent recovery is somehow unequal; that some parts of the city are rebounding while others still decline.

“. . . within the city recovery has been highly uneven, resulting in greater inequality.”

Detroit’s problem is not inequality, it’s poverty.  As the Brookings Institution’s Alan Berube put it:

“Detroit does not have an income inequality problem—it has a poverty problem. It’s hard to imagine that the city will do better over time without more high-income individuals.”

To be sure, more higher income residents and new restaurants, condos and office buildings may bring poverty into sharper contrast, but had those same higher income jobs and households located in the suburbs (or some other city), its far from obvious that poor Detroiters would somehow be better off.

As a result, the only way that Detroit is likely to improve its economy is to become at least somewhat less equal.  The city has a relatively high degree of equality at a very low level of income. The reason this has occurred, in large part, is those upper and middle income households—those with the means to do so—have exited the city in large numbers, leaving poor people behind.

We’ve long called out the misleading nature of inequality statistics when applied to small geographies. What’s called “inequality” at the neighborhood level is actually a sign of economic mixing, or economic integration—a neighborhood where high, middle and low income families live in close proximity and where there are housing opportunities at a range of price points.

Incomes in central cities are almost always more unequally distributed than in the metropolitan areas in which they are located, but this is because cities are more diverse and inclusive. At small geographies, this statistic says more about integration than it does about inequality.  

At a highly local level, “equality” is generally achieved in one of two ways:  by having a community that is so undesirable that no one with the means to live elsewhere chooses to stay, leaving an “equal” but very poor neighborhood. Alternatively, high levels of neighborhood equality can be achieved through the application of exclusionary zoning laws that make it illegal and impossible for low (and in some cases even middle income) families to live in an area. As in some exclusive suburbs, such as Flower Mound, Texas and Bethesda, Maryland — two of the highest scoring cities on equality — the equality is only for high income families.

Indeed, the big problem in American cities, as we’ve documented in our report “Lost in Place” is that in poor neighborhoods income is actually too equal. Neighborhoods of concentrated poverty, where more than 30 percent of the population lives below the poverty line, have tripled in the past 40 years. As the work of Raj Chetty and others has shown, neighborhoods of concentrated poverty permanently lower the lifetime earnings prospects of poor kids. Otherwise similar children who grow up in more mixed income (meaning unequal) neighborhoods have higher lifetime earnings.

As a practical matter, the only way forward for the Detroit economy is if more middle income and even upper income families choose to move to the city (or stay there as their fortunes improve). That will nominally make some of the income numbers look less “equal” but will play a critical role in creating the tax base and the local consumption spending that will —gradually— lead to further improvements in Detroit’s nascent economic rebound.

Where do we start? Achieving critical mass

The second fundamental critique in the City Lab piece is an argument that the city’s redevelopment efforts are failures because they aren’t producing improvements for everyone, everywhere in the city all at once. To date, the city’s successes have been recorded in downtown, Midtown and a few nearby neighborhoods, but because other parts of the city have continued to deteriorate and depopulate, the assumption is Detroit must be failing.

This critique ignores the fundamental fact that development and city economies are highly dependent on spatial spillovers. Neighborhoods rebound by reaching a critical mass of residents, stores, job opportunities and amenities.  The synergy of these actions in specific places is mutually self-reinforcing and leads to further growth. If growth were evenly spread over the entire city, no neighborhood would benefit from these spillovers. And make no mistake, this kind of spillover or interaction is fundamental to urban economics; it is what unifies the stories of city success from Jane Jacobs to Ed Glaser.  Without a minimum amount of density in specific places, the urban economy can’t flourish.  Detroit’s rebound will happen by recording some small successes in some places and then building outward and upward from these, not gradually raising the level of every part of the city.

Scale: Making the perfect the enemy of the good

Anyone familiar with Detroit knows that the city’s most overwhelming problem is one of operating and paying for a city built for two million people with a population (and consequently a tax base) less than half that size. The city is still wrestling with the difficult challenge of triage—reducing its footprint and shrinking its service obligations to match its resources.

And that’s the final point that’s so disturbing about the CityLab critique. Reese and Sands  argue that Detroit needs more more jobs  and resources for, among other things, educating its kids. No one doubts this. But where will that money come from? Certainly not from federal or state governments. It will have to come in large part from growing a local tax base, which is contingent on creating viable job centers and attracting and retaining more residents, including more middle and upper income residents.

Make no mistake: the scale here is daunting. The authors offer the helpful observation that if Detroit just somehow had another 100,000 jobs paying $10 per hour, it would pump more than $2 billion a year into the city’s economy. (Keep in mind that from 2001 through 2010, the Wayne County lost about twice that many jobs). While their math is impeccable, their economics are mystical.  This is the academic equivalent of the old Steve Martin joke about how to get a million dollars tax free:  “Okay, first, get a million dollars.”

It would be great if we could craft a sudden solution that would immediately create hundreds of thousands of jobs and drop billions of dollars in wages and money for schools and public services into Detroit. But that’s simply not going to happen. Instead, progress on a smaller scale has to start somewhere, has to involve new jobs, new residents and new investment in a few neighborhoods and then build from there. Businesses will start-up or move in, a few at a time, more in some neighborhoods than others, and then over time grow, providing more jobs and paying more taxes.

It’s going to be a long, hard road ahead for Detroit. And that road will lead to a different and smaller Detroit than existed in, say, the 1950s.  That road is made even harder by critics that damn the first few candles for shedding too little light.

 

Playing Apart

Our City Observatory report, Less in Common, catalogs the ways that we as a nation have been growing increasingly separated from one another.  Changes in technology, the economy and society have all coalesced to create more fragmentation and division.

As Robert Putnam described this trend in his 2000 book, we are “Bowling Alone.”  And while work, housing and shopping have become more stratified and dispersed, there still ought to be the opportunity for us to play together. Sports fandom is one of the few countervailing trends: within metropolitan areas popular support for the “home team” whether in pro-sports or college athletics is cuts across demographic and geographic boundaries.

But in our personal lives our recreation is becoming more isolated, chiefly through the privatization of leisure.

Consider:  instead of going to public parks and playgrounds, more children play in the copious backyards of suburban homes. This trend is amplified by helicopter parents.  Free range children are an anomaly, and the combination of sprawl and insecurity adds to the chauffering burden of adults–which in turn means spending more time in cocooned private vehicles. And as we know, the decline in physical exercise among the nations children has been a key factor in the explosive growth of juvenile obesity.

One of the hallmarks of the decline in the public recreational commons is swimming.  In the early part of the 20th century, swimming pools were almost exclusively in the public domain.  Prior to World War II it was estimated that there were fewer than 2,500 homes with private, in-ground swimming pools.  Today, there are more than 5 million.

That’s one of the reasons we found Samsung’s television commercial “A Perfect Day” so compelling. It highlighted the adventures of a group of kids, cycling around New York City, and ending up spending time at a public pool.  Its encouraging that a private company can make our aspirations for living life in public a central part of its marketing message.

That’s certainly a contrast to the trend of commoditization of leisure. Increasingly, we pay to play, and play in the private realm.  The number of persons who belong to private gyms has increased from about 13 million in 1981 to more than 50 million today.  While gyms provide a great experience for those who join, they tend to draw disproportionately from wealthier and younger demographic groups–again contributing to our self-segregation by common background and interest.

Over just the past five years, the number of Americans classified as “physically inactive”–not participating in sports, recreation or exercise, has increased from 75 million to 83 million, according to the Physical Activity Council.  And youth participation in the most common team sports — soccer, basketball, football and baseball — has declined 4 percent since 2008.

As we think about ways to strengthen and restore the civic commons, we will probably want to place special emphasis on parks and recreation.  Public parks are one of the places where people of different races, ethnicities and incomes can come together and share experiences.

Urban myth busting: Why building more high income housing helps affordability

After fourteen seasons, Discovery Channel’s always entertaining “Mythbusters” series ended last year. If you didn’t see the show-and it lives on at Youtube, of course–co-hosts Adam Savage and Jamie Hyneman constructed elaborate (often explosive) experiments to test whether something you see on television or in the movies could actually happen in real life. (Sadly, it turns out that you can’t make a bullet curve no matter how fast you flick your arm.)  

Adam-Savage-and-Jamie-Hyneman-in-Mythbusters

At City Observatory, we feel compelled to enter into this void, and we’ll start by doing our own urban myth-busting. First up: Does building new high-priced apartments, affordable only by middle- and upper-income families, make housing less affordable for lower income households?

We’ve heard this claim time and again in public hearings: new rental housing charges higher rents than existing apartments, and must therefore be making affordability problems worse. Business Week’s Noah Smith shared the lament of the misunderstood economist who confided in a progressive friend that he favored building new market-rate (i.e. high priced) housing in San Francisco:

I was talking to a friend the other day, a San Francisco anti-eviction activist, and said that allowing more housing construction in the city would be a great way to lower rents. She looked at me in horror, blinked and asked “Market rate?” I nodded. She was speechless.

My experience was far from unusual. To my friend and many others, it has become an article of faith that building market-rate housing raises rents, rather than lowers them.  The logic of Econ 101 — that an increase in supply lowers price — is alien to many progressives, both in the Bay Area and around the country.

Even Harvard University’s Joint Center on Housing Studies reprised this line in one of their recent reports: “50 percent of rental households make less than $34,000 per year, but only 10 percent of new multi-family units are affordable at this income.”

From this statistical observation, it’s a short leap to the conclusion that building new housing is part of the affordability problem. The Wall Street Journal reported that “much of the new supply is aimed at higher-income renters.” In May, the Journal ran a story claiming: “A focus by builders on high-end apartments helps explain why rents are soaring across the country.”

New construction in San Francisco. Credit: torbakhopper, Flickr
New construction in San Francisco. Credit: torbakhopper, Flickr

 

On its surface, this sounds terrible. But the key context missing here is that in the United States, we have almost never built new market-rate housing for low-income households. New housing—rental and owner-occupied—overwhelmingly tends to get built for middle- and upper-income households. So how do affordable market-rate housing units get created? As new housing ages, it depreciates, and prices and rents decline, relative to newer houses. (At some point, usually after half a century or more, the process reverses, as surviving houses—which are often those of the highest quality—become increasingly historic, and then appreciate.)

What really matters is not whether new housing is created at a price point that low- and moderate-income households can afford, but rather, whether the overall housing supply increases enough that the existing housing stock can “filter down” to low and moderate income households. As we’ve written, that process depends on wealthier people moving into newer, more desirable homes. Where the construction of those homes is highly constrained, those wealthier households end up bidding up the price of older housing—preventing it from filtering down to lower income households and providing for more affordability.

This isn’t theoretical: As we’ve discussed before at City Observatory, the vast majority of today’s actually existing affordable housing is not subsidized below-market housing, but market-rate housing that has depreciated, or “filtered.” Syracuse economist Stuart Rosenthal estimates that the median value of rental housing declines by about 2.2% per year. As its price falls, lower-income people move in. Rosenthal estimates that rental housing that is 20 years old is occupied, on average, by households with incomes about half the level of incomes of those who occupy new rental housing.

Screen Shot 2015-11-09 at 9.55.56 AM
Apartments get cheaper up until they’re about 50 years old.

 

In its 2016 report on the state’s housing crisis, the California Legislative Analyst’s Office noted that as housing ages, it becomes more affordable. Housing that likely was considered “luxury” when first built declined to the middle of the housing market within 25 years. Take the 1960s-era apartments built in Marietta, a suburb of Atlanta: When they were new, they were middle to upper income housing, occupied by single professionals, gradually, as they aged, they slid down-market, to the point where the city passed an $85 million bond issue to acquire and demolish them as a way of reducing a concentration of low income households in the Franklin Road neighborhood. 

Another critical point is that if we don’t build more housing at the high end of the market, those households don’t just disappear, they take their demand “down-market” and bid up the price of housing that would otherwise filter down to middle and lower income households. That’s exactly what the Montgomery County Maryland housing department reports is happening there:

The shortage of rental housing at the high end of the market creates downward pressure on less affluent renters, the study found, because when higher-income households rent less expensive units, lower-income renters have fewer affordable choices. Cost-burdening is linked with this unbalanced market, especially at the lower end of the income spectrum.

Ironically, this problem persists in Montgomery County in spite of its widely touted inclusionary housing requirement that forces builders of new apartments to set aside a portion of them for low and moderate income households.

New Cars are Unaffordable to Low Income Households, too

Here’s another way to look at the connection between affordability and the price of new things: cars. (After houses, cars are frequently the most expensive consumer durable that most American’s purchase.)

Exactly the same thing could be said of new car purchases: Most new cars aren’t affordable to the typical household either—the average sale price of a new car is nearly $34,000.

Credit: Brian Timmermeister, Flickr
Credit: Brian Timmermeister, Flickr

 

In fact, using the same kind of approach that Harvard’s Joint Center for Housing Studies used to assess rental affordability, Interest.com reports that the median family can afford to buy the typical new car in only one large metropolitan area. Similar to the “30 percent of income” rule widely—and in our view inappropriately—used to gauge housing affordability, they assume that the typical household makes an 20 percent down payment, finances its purchase over four years and pays no more than 10 percent of its income for a car payment. They report in most metros that the typical family falls 30 to 40 percent short of being able to afford a new car. So most households deal with car affordability pretty much like they deal with housing affordability: by buying used.

When it comes to anything new and long-lived, higher-income households buy most of the output. According to Bureau of Labor Statistics data, households in the two highest income quintiles accounted for about 67 percent of the purchase of new cars in the US in 2001. New car buyers are getting progressively older, and are more likely to be high income. According to the National Automobile Dealers Association, the median new car buyer is 52 years old and has an income of about $80,000, compared to an average age of 37 and an income of $50,000 for the overall population.

But there’s no outcry about America’s “affordable car crisis.” The reason: high-income households buy newer cars; most of the rest of us buy used cars—which are more affordable after they’ve depreciated for a while.That’s even more true of housing, which is much longer lived. Nationally, 68 percent of the nation’s rental housing is more than 30 years old—so only about 10 percent of the nation’s renters live in apartments built in the last decade.

New houses, like new cars, are sold primarily to higher income households—and affordability comes from getting a bargain when the car (or house or apartment) has depreciated. Building more high priced new apartments, in fact, is critical to generate the filtering down of older housing that constitutes the affordable housing supply.

This myth is busted: building more high end housing doesn’t make housing less affordable.

The Week Observed, February 17, 2017

What City Observatory did this week

1. Anti-social capital. You’re probably familiar with the term “social capital” which Robert Putnam popularized with his book Bowling Alone. In it Putnam devised a series of indicators that show the extent to which we associate with and trust one another, ranging from membership in clubs and civic organizations to regularly hosting our neighbors for dinner. We’ve taken the opposite tack and developed measure of anti-social capital, the number of security guards per capita in each of the nation’s large metro areas. The more people we pay to watch each other and our our stuff, the less trust we have in our communities. See where your city rates on this metric.

2. Are young adults moving less? Data presented by the Pew Research Center shows a significant reduction in migration by 25 to 35 year olds over the past two decades. But these results aren’t borne out by other survey data, which show that young adult migration rates actually went up during the housing bubble and then returned to pre-bubble levels in the past few years. These conflicting data mean the case for lower migration is less clear than it first appears.

3. Postcard from Louisville: Tolls Trump Traffic. We’ve been following the opening of the new tolled I-65 river crossing in Louisville. Indiana and Kentucky spent more than a billion dollars to widen the crossing to twelve lanes to reduce congestion, and the early traffic data show that they’ve succeeded beyond anyone’s wildest dreams: vehicle counts have fallen by almost half, to just 66,000 cars per day. The new bridges are mostly empty, even at rush hour. Which raises the question: Why did they need to expand the bridges when simply tolling them would eliminate the congestion problem?

4. Let’s not demonize driving, just stop subsidizing it. A lot of the rhetoric about the problems with cars takes a pretty high moral tone. But in our view, the problem is not that cars (or the people who drive them) are evil, but that we use them too much, and in dangerous ways. And that’s because we’ve put in place incentives and infrastructure that encourage, or even require, us to do so. When we subsidize roads, socialize the costs of pollution, crashes and parking, and even legally require that our communities be built in ways that make it impossible to live without a car, we send people strong signals to buy and own cars and to drive—a lot. As a result, we drive too much, and frequently at unsafe speeds given the urban environment.

Must read

1. Rents are heading for a fall in Germany. Writing at CityLab, Feargus O’Sullivan tells of a new report looking at the housing market in leading German cities. The national landlords association is predicting a decline in rents in the year’s ahead. The City Lab story buries the lede in the fifth paragraph: the reason that the association is forecasting rent decreases is because of a big increase in supply (3000,000 new apartments and houses coming on line) coupled with weaker than expected immigration. While it takes time for supply to catch up with demand, when it does, there’s evidence that rent inflation eases.

Nikolaiviertel, Berlin (Flickr: oh_berlin).

2. More people are dying on the roads. The New York Times and other national news outlets report the grim news that for the second straight year the death toll from traffic crashes has increased. Last year, 40,200 Americans died on the roads, a six percent increase from 2015. While the article points to an improving economy, speeding and distracted driving at causes, At City Observatory, we’ve emphasized that much of the increase in deaths is related to the greater amount of driving associated with cheaper gas (since prices declined in 2014), and that there’s strong evidence that the added driving induced by cheaper gas is by riskier drivers traveling at more dangerous times.

3. Why declining migration may be good news. A recently published paper on interstate migration by Federal Reserve economists has a provocative theory as to why the rate at which Americans move across state boundaries has declined over the past several decades. Its a long and technical paper, but Lyman Stone has a lively and readable analysis. A key factor: as state economies have become more similar over time, there’s little need for people looking for jobs a in particular industry or occupation to move very far to find them. While we used to have big regional variations in wage levels and job opportunities, those differences are more muted now, and so people are less likely to have to move for career reasons.

New Research

1.  Keeping up with the Joneses is making homes bigger and driving up debt. American homes have become increasingly large as American families have gotten smaller. And the amount of debt we’ve collectively shouldered to pay for houses has increased as well. A new study from Clement Bellet of the London-based Centre for Economic Performance shows that a major part of the reason for these increases may be that we’re constantly trying to have houses as large as our neighbors. Using data from Zillow and the American Housing Survey, the authors find that suburban homeowners were are less happy with their homes if they live in a county where the new homes built since they purchased their home are much larger. These homeowners who experienced a relative “downscaling” because large homes were built nearby then disproportionately tended to expand their houses or move to new and larger homes–taking on more debt to finance larger houses. Bellet estimates that the US housing debt to income ratio would have been about 25 percentage points lower in 2008 absent this “keeping up with the Joneses” behavior.

2. Arts and Gentrification. One of the most durable creation myths about gentrification is story of how starving artists colonize once-poor neighborhoods, in the process making them attractive to others and triggering a wholesale transformation. A new paper published in Urban Studies “Gentrification, displacement and the arts: Untangling the relationship between arts industries and place change,” tries to test this theory looking at the concentration of arts related industries (think galleries and museums) and commercial arts  (like film, music and design firms) and subsequent gentrification. The study relies on the Census Bureau’s Zip Code Business patterns data, which only capture the location of firms with paid employees; so they don’t necessarily reflect the residential locations of artists, or the work-places of self-employed artists. The study finds only relatively weak associates between the arts businesses and gentrification; if anything gentrification (measured by rising income levels and home prices) seems to precede the growth of arts businesses. But these data may simply be confirming that larger scale arts activity (like galleries) come later in the gentrification process,.

 

Let’s not demonize driving—just stop subsidizing it

At City Observatory, we try to stick to a wonky, data-driven approach to all things urban. But numbers don’t mean much without a framework to explain them, and so today we want to quickly talk about one of those rhetorical frameworks: specifically, how we talk about driving.

Our wonky perspective tells us that there are lots of problems that stem from the way we use cars: We price roads wrong, so people over use them. Cars are a major source of air pollution, including the carbon emissions that are causing climate change. Car crashes kill tens of thousands of Americans every year, injure many more, and cost us billions in medical costs and property damage. And building our cities to accommodate cars leads to sprawl that pushes us further apart from one another.

But the problem is not that cars (or the people who drive them) are evil, but that we use them too much, and in dangerous ways. And that’s because we’ve put in place incentives and infrastructure that encourage, or even require, us to do so. When we subsidize roads, socialize the costs of pollution, crashes and parking, and even legally require that our communities be built in ways that make it impossible to live without a car, we send people strong signals to buy and own cars and to drive—a lot. As a result, we drive too much, and frequently at unsafe speeds given the urban environment.

This car might be evil, though. Credit: Michael Coghlan, Flickr
This car might be evil, though. Credit: Michael Coghlan, Flickr

 

Many people—transit boosters, cyclists, planners, environmentalists, safety advocates—look at the end result of all this, and understandably reach the conclusion that cars are the enemy. The overriding policy question, then, becomes: “How do we get people out of their cars?”

In this December 2015 story in The New Republic, for example, Emily Badger quotes Daniel Piatowski, a planning PhD presenting a paper on “carrots and sticks” at the Transportation Research Board conference, saying: “The crucial component that’s missing is that we’re not implementing any policies that disincentivize driving.”

“Getting people out of their cars” is a rallying cry and a mission statement that’s guaranteed to provoke a formidable opposition. That’s because most people, correctly, can’t imagine any time soon when they won’t need to use a car for most—even all—of their daily trips. As a practical matter, the fact that for seven or eight decades the entire built environment and most transportation investments have been predicated on car travel means that we can’t quickly move away from auto dependence. For most Americans, driving isn’t attributable to an irrational fondness for cars. In many places, it’s simply impossible to live and work without one.

But there’s good news. The first is that incentives matter. We learned that higher gas prices, for example, had a large and sustained impact on driving behavior. After growing steadily for decades, vehicle miles traveled per person peaked and declined after 2005 (as gas prices shot up). This produced knock-on changes in housing markets, and helped accelerate the move back to cities. And the decline in gas prices since 2014 has triggered more driving. “This shows that more intentional kinds of pricing schemes, like congestion pricing or parking pricing, could have similar effects.”

The second point is that small changes matter. Even slight reductions in car use and car ownership will pay big dividends. Traffic congestion is subject to non-linear effects: small reductions in traffic volumes produce big reductions in traffic congestion. Travel monitoring firm Inrix reported that in 2008, the 3 percent decline in vehicle miles traveled led to a 30 percent decline in traffic congestion. . As driving declined, carbon emissions declined and so too, did crashes and traffic deaths.

Moralizing about mode choice is a recipe for policy gridlock

Bitter and acrimonious flamewars between people who are convinced that one side or the other is trying to run us off the road will surely be unproductive. We agree with most of the policies advocates like Piatowski want, including the “sticks” like parking and congestion fees—but not the way they’re being described.

Credit: Steve Snodgrass, Flickr
Credit: Steve Snodgrass, Flickr

 

Rather than being framed as a punishment, it should be more about responsibility. Drivers should pay for the roads that they drive on. They should be regulated in a way that protects the safety of other users of the right of way. Trucks ought to pay for the damage they do to roads. Every car driver ought to pay for their parking space they use—whether it’s in the public or the private realm. All cars and trucks should be responsible for the carbon pollution they emit. We shouldn’t require third parties such as homebuilders or renters or local businesses to subsidize car travel and parking. This isn’t about creating a “disincentive for car use,” but, as a matter of fairness and practicality, dropping what have essentially been subsidies for financially and socially expensive and dangerous behavior.

Driving is a choice, and provided that drivers pay all the costs associated with making that choice, there’s little reason to object to that. After all, very few people think that a zero car world is one that makes a lot of sense. Low-car makes much more sense that non-car as a policy talking point. How do we get people to make these choices. There’s an analogy here to alcohol. We tried prohibition in the twenties. It was moral absolutism, zero tolerance. Alcohol in any amount was evil. That didn’t work.

When we experienced the epidemic of drunk driving, we didn’t go back to prohibition. Instead, we raised penalties to make drivers more responsible, set tougher limits on blood alcohol content, and put more money into enforcement. People still drink—but there’s a different level of understanding of responsibility and consequences, and fewer people drive drunk.

Postcard from Louisville: Tolls Trump Traffic

Tolls cut traffic levels on I-65 in half; So did we really need 6 more lanes?

Last month, we wrote about Louisville’s newly opened toll bridges across the Ohio River.  As you may recall, Ohio and Indiana completed a major expansion of highway capacity across the Ohio, doubling the I-65 freeway crossing from six lanes to twelve near downtown and adding a beltway bridge to complete the freeway loop around the region. To pay help pay for the multi-billion dollar project, the two states imposed tolls ranging from $1-4 for cars using I-65 and the new East End/Lewis & Clark Bridge.

But travelers through Louisville still have a toll-free river crossing alternative, the old four-lane  Clark Memorial (Second Street) Bridge, just a few hundred yards downstream from the larger newer toll bridges, continues to be free. 

In our earlier commentary, we speculated that the addition of tolls to the freeway bridge, coupled with the presence of a nearby free alternative would dramatically reduce car traffic.  In the absence of any actual data, we took our first cues on what was happening from rush hour traffic cams showing cars on the I-65 and Second Street Bridges. The tolled freeway bridges were nearly empty, and the Second Street Bridge looked pretty busy. That seemed to confirm our hypothesis.  Nonetheless, several readers took us to task for not waiting until we had actual traffic counts. Well, now we do.

And they confirm what the traffic cameras were telling us: traffic is very light on the newly tolled freeway crossing.  The data from Riverlink (the joint Indiana-Kentucky toll operator) show that average weekday traffic on the two bridges was about 66,000 vehicles per day in January.  To put that number in context, it is about half the level of traffic (122,000 vehicles per day) that used I-65 in 2012–before the new Lincoln Bridge opened. To put that a slightly different way: the two states spent over a billion dollars to double the capacity of the I-65 crossing, and now it is used by about half as many vehicles as used it before.

 

Source: Riverlink, February 1, 2017

This is important because the principal objective of the Ohio River Bridges Project, according to the “purpose and need”: spelled out in its environmental impact statement was to reduce traffic congestion. Unless the river crossing was expanded, the EIS claimed, traffic would have increased to 155,000 vehicles per day, more than 23 percent over capacity, by 2030. Now its apparent that with tolling, traffic is actually unlikely to rebound to pre-construction levels.

There’s a lot more to dig into here. One of the big unanswered questions is has tolling I-65 worsened traffic congestion elsewhere in Louisville, particularly on the un-tolled Second Street bridge and nearby streets.  Because Riverlink doesn’t toll this bridge, they don’t have data on traffic levels–so we’ll have to wait until traffic counts are published by the two states.

But for the record:  one more set of photos of traffic over the I-65 and Second Street Bridges.  These were taken Monday afternoon February 13 at about 5:20 PM, from the local “TRIMARC” traffic monitoring website.

First, traffic on the tolled Kennedy and Lincoln Bridges (Camera 0333A).

Next, here’s an image of traffic coming off the Second Street Bridge as it enters downtown Louisville.  This picture was captured at 5:25 pm EST (despite the time-stamp on the camera).

 

The data and the traffic cam photos suggest that Louisville has demonstrated a powerful, fast-acting solution for reducing traffic congestion: charge a toll. It’s too bad they found out only after spending in excess of a billion dollars building new road capacity that apparently wasn’t needed or valued by those who travel across the Ohio River each day. Maybe other cities can learn from the Louisville’s expensive experiment.

 

Are young adults moving less?

Conflicting data sources present very different pictures of young adult migration rates

The Pew Research Center presented an analysis of census data reporting that today’s young adults are less likely to move in a given year than were their predecessors. A new article from Pew concludes: “Americans are moving at historically low rates, in part because Millennials are staying put.”  Only about 20 percent of 25-35 year olds moved in the previous year, a rate far lower than the 26 percent of GenXers who moved when they were the same age and 27 percent of late boomers, who moved when they were in this age group.

The data for this finding come from the Census Bureau’s Current Population Survey, which each March asks a battery of detailed demographic questions, in what’s called the Annual Social and Economic Supplement (ASEC).

The Census Bureau also asks a nearly identical question as part of its ongoing American Community Survey.  And, as we’ll see in a minute, the ACS presents a very different picture of migration trends for young adults than presented by the CPS. There are  a couple of important differences between the two surveys: the ACS is much larger (about 3 million sampled households, as opposed to about 100,000 in the CPS).  The CPS uses a slightly different sampling strategy (it includes only non-institutional households), and is administered solely via telephone. The CPS is superior for long-run research because it goes back to the 1960s; the ACS has been gathered only since 2001.

There’s little question that overall migration rates have declined in the US in the past several decades. There’s a good body of research exploring the reasons for this.  Some of them have to do with an aging population (older people are less likely to move). Others have conjectured that declining migration is associated with growing occupational homogeneity among different regions (meaning there’s little reason to migrate to a different region for employment). We reviewed much of this literature in our report on the Young and Restless.

Others have identified some concerns using CPS data to chart migration patterns of young adults.  The University of Virginia’s Luke Juday pointed out that CPS tended to significantly undercount the migration of young adults–“Migration data miscounts millennials, confuses the media.” Juday reports a discrepancy of more than a million migrants between CPS estimates of city-suburb migration and Census population estimates. Looking specifically at the question of interstate migration, in a 2011 article entitled “A Sharp Drop in Interstate Migration? Not really,” Greg Kaplan Sam Schulhofer-Wohl of the Federal Reserve Bank of Minneapolis found that imputation procedures used in processing CPS sample data tended to inflate the estimated numbers of interstate migrants (they also found that this particular problem didn’t affect the overall estimate of migration rates).

Given this track record of problems with the ACS data, it probably makes sense to double-check Pew’s findings with a second source of data, which we can do with the American Community Survey.  We downloaded data from the University of Minnesota’s Integrated Public Use Microdata Series (IPUMS), for both the CPS, and the ACS.  (See full citations below).  For each data source, we computed the percentage of 25 to 35 year old respondents in each year who reported living in a different home than they did in the previous year.

 

The CPS data (red) show the migration rates declining fairly steadily from about 24 percent in 2001 to less than 20 percent in 2016–mirroring the findings reported in the Pew report.  While the ACS estimates of 25-35 year old migration rates are remarkably similar (within about half a percentage point) in 2001 and 2002, they diverge substantially thereafter.  The ACS estimates suggest that migration rates for young adults rose from 2001 through 2006, peaking about the same time as the housing bubble, and then declining somewhat thereafter.  The 2015 ACS 1 year migration estimate for 25 to 35 year olds (24.1 percent) is within about one-half percentage point of its 2001 level (24.6 percent).

These two series paint very different pictures, especially on the effect of the housing market on migration rates.  If one believes the Pew/CPS story, the housing bubble had essentially no effect on the long term trend in migration rates; the likelihood that a young adult moved declined even as the housing market was super-heated–and rates kept on declining through 2013.  In contrast, the story in the ACS data is that at least temporarily, at a time coincident with the housing bubble, the rate of migration for 25-35 year olds surged, and then started declining after the bubble collapsed.

We don’t have a definitive means of choosing one of these data series and its attendant explanations over the others.  Given the larger sample size of the ACS, and its freedom from some of the doubts that have arisen about other uses of the CPS for migration studies, we’re inclined to put more weight on the ACS. It also seems plausible to us that the housing bubble and its attendant collapse would produce the kind of “surge and purge” pattern seen in the ACS data. But in the end, we may have to defer to the Magic 8-ball for our answer:

 

References

Steven Ruggles, Katie Genadek, Ronald Goeken, Josiah Grover, and Matthew Sobek. Integrated Public Use Microdata Series: Version 6.0 [dataset]. Minneapolis: University of Minnesota, 2015. http://doi.org/10.18128/D010.V6.0.

Sarah Flood, Miriam King, Steven Ruggles, and J. Robert Warren. Integrated Public Use Microdata Series, Current Population Survey: Version 4.0. [dataset]. Minneapolis: University of Minnesota, 2015. http://doi.org/10.18128/D030.V4.0.

This post was revised to correct a mis-statement about the differences in the sampling universes of the two surveys.

 

 

Anti-Social Capital?

In his book Bowling Alone, Robert Putnam popularized the term “social capital.” Putnam also developed a clever series of statistics for measuring social capital. He looked at survey data about interpersonal trust (can most people be trusted?) as well as behavioral data (do people regularly visit neighbors, attend public meetings, belong to civic organizations?). Putnam’s measures try to capture the extent to which social interaction is underpinned by widely shared norms of openness and reciprocity.

It seems logical to assume that there are some characteristics of place which signify the absence of social capital. One of these is the amount of effort that people spend to protect their lives and property. In a trusting utopia, we might give little thought to locking our doors or thinking about a “safe” route to travel. In a more troubled community, we have to devote more of our time, energy, and work to looking over our shoulders and protecting what we have.

The presence of security guards in a place is arguably a good indicator of this “negative social capital.” Guards are needed because a place otherwise lacks the norms of reciprocity that are needed to assure good order and behavior. The steady increase in the number of security guards and the number of places (apartments, dormitories, public buildings) to which access is secured by guards indicates the absence of trust.

The number of security guards in the United States has increased from about 600,000 in 1980 to more than 1,000,000 in 2000 (Strom et al., 2010). These figures represent a steep increase from earlier years. In 1960, there were only about 250,000 guards, watchmen and doormen, according to the Census (which used a different occupational classification scheme than is used today).

This trend has led Sam Bowles and Arjun Jayadev to argue that we have become “one nation under guard” and that the growth of guard labor is symptomatic of growing inequality. The U.S. has the dubious distinction of employing a larger share of its workers as guards than other industrialized nations and there seems to be a correlation between national income inequality and guard labor.

Just as the U.S. has a higher fraction of security guards than other nations, some cities have more security guards than others. To understand these patterns, we’ve compiled Bureau of Labor Statistics data from the Occupational Employment Survey on private security guards. BLS defines security guards as persons who guard, patrol, or monitor premises to prevent theft, violence, or infractions of rules, and whom may operate x-ray and metal detector equipment. (The definition excludes TSA airport security workers). In 2015, there were more than 1,050,000 security guards in the US.

This occupational data reports the number of security guards in every large metropolitan area in the country. Adjusting these counts by the size of the population in each metro area tells us which places have proportionately the most security guards– which are arguably the least trusting, and which places have the fewest security guards. This may be an indicator of higher levels of social trust.

Here are the data:

At the top of the list is Las Vegas. While the typical large metro area has about 39 security guards per 10,000 population, these Miami has more than  twice as many (86 per 10,000). Other cities with high ratios of security guards to population are Memphis, New Orleans, Miami and Baltimore. Washington D.C., with its high concentration of government offices, defense and intelligence agencies, and federal contractors, also has a high proportion of security guards.

At the other end of the spectrum are a number of cities in which the ratio of security guards to population is one-third lower than in the typical metro area. At the bottom of the list are Minneapolis-St. Paul, Providence and Portland. (The Twin Cities and Portland also do well on most of Putnam’s measures of social capital)

It seems somewhat paradoxical, but the salaries paid to security guards get treated as a net contribution to gross domestic product. Yet, in many important senses, security guards don’t add to the overall value of goods and services so much as they serve to keep the ownership of those goods and services from being rearranged. As Nobel prize winning economist Douglass North has argued, we ought to view the cost of enforcing property rights as a “transaction cost.” In that sense, cities that require lots of guards to assure that property isn’t stolen or damaged and that residents, workers, or customers aren’t victimized, actually have higher costs of living and doing business than other places. These limits on easy interaction may stifle some of the key advantages to being in cities.

The Week Observed, February 10, 2017

What City Observatory did this week

1. The persistence of talent. City Observatory regularly stresses the strong connection between educational attainment and economic success at the metro level. We step back and look at how education attainment has influenced state level economic success over the past 25 years. The data shows that the fraction of the adult population with a four-year degree explains about two-thirds of the variation in per capita income among states, a relationship which has strengthened since 1990. And the effect is long term: 1990 education levels also are strongly correlated with 2015 incomes.

2. Visions of a future city. Our premise is that narratives and visions matter to the kind of cities we build. For most of the past century, we’ve built cities with the explicit vision of accommodating car travel, with devastating results. Going forward its critical that we select a narrative that reflects the kind of places we want to live in. We started our three-part examination of our visions of future city living by reviewing Ford’s presentation to the Consumer Electronics Show. Like General Motors’ 1930s vintage World’s Fair program, Ford offered up an auto-centric view of urban living–with a few half-baked and largely decorative nods to transit, bikes and pedestrians.

3. A perfect day. A contrasting narrative about idealized urban living comes from consumer electronics giant Samsung, in a long-form commercial called “A Perfect Day.” It shows an energetic group of teenagers on bikes and skateboards ranging over New York City and drinking in a wide variety of decidedly urban experiences. Its these experiences and the personal interactions the kids have with each other and the people that they meet–and not the technology–that make this “a perfect day.”

4. You don’t own me. Our third narrative is a 30 second micro-drama from Toyota, which only in passing features cars. Instead in focuses on a young chef who sees her creativity dashed by a corporate restaurant and instead quits this straight job to open her own food truck. She’s joined by a chorus of other young people–hiking, playing basketball, cycling, crashing headlong in roller derby–who are similarly rebellious. Again, the narrative here is that the city is about experiences, not how we travel. Toyota’s tag-line: Let’s go places.

Must read

1. More evidence that housing supply is moderating rents. Two of the nation’s hottest apartment markets have been Washington and San Francisco. In the past few weeks there have been reports from both cities that the surge in new construction is finally starting to have a visible effect on rents. According to Axiometrics, rents in San Francisco have declined 2.3 percent in the past year. Meanwhile, in the nation’s capital, Greater Greater Washington reports, a record number of apartments have been built in the District of Columbia. They add that rents are up 2.6 percent in the past year, notably below the national average of 4.0 percent.

John Ricco, Greater Greater Washington

2. Portland’s inclusionary zoning land rush. The imposition of inclusionary zoning in Portland (which took effect on February 1, led to a last minute rush to file building permit applications. In the two months before the new requirements took effect, developers applied for permits to build nearly 6,000 new apartments, more than the number of new apartments built in Portland in either 2013 or 2014. That adds to a pipeline that already swelled to more than 20,000 units. Now the city expects a long drought before developers propose any new housing under the new law, which requires them to set aside 10 or 20 percent of new units as discounted affordable housing. For the next few years, the surge in apartment supply is much more likely to provide rent relief for Portlanders will any affordable inclusionary units.

New Research

Young homebuyers prefer more central locations. Research from the Atlanta Federal Reserve looks at the locational preferences of young homebuyers, shedding additional light on question of whether today’s young adults are more likely to purchase homes in cities. Researchers Elora Raymond and Jessica Dill  find that after controlling for other factors influencing buying decisions, younger first time buyers are significantly more likely to live in the heart of downtown. For each additional year, the odds that a buyer will decide to live within one mile of the city center drop by 6 percent. This is further confirmation of the trend that we and others have documented.

 

 

Visions of the City Part III: You don’t own me

What kind of future do we want to live in? While that question gets asked by planners and futurists in an abstract and technical way, some of the most powerful and interesting conversations about our future aspirations are reflected in the mass media. Lately, we’ve been struck by the visions embedded in recent television commercials.  Some of these visions are explicit, but others are a bit more subtle.

Earlier we took a close look at one car maker’s image of what a future city might look like. Ford’s vision, prepared for the Consumer Electronics Show, was a computer-generated video simulation of what it might be like in the near future, and then sometime later, to live in a city full of autonomous vehicles.  It was undoubtedly developed by the company’s engineers, and was used to polish the cred of the auto companies with the tech community and investors.

Until that future arrives, big auto companies have to survive, as they always have, by selling cars today. The world’s largest automaker, Toyota, released a new television ad to try and sell its vehicles. And implicit in its commercials is another view of the future (or perhaps the present) of what cities and city living should be like. And its actually much more interesting, compelling and human than Ford’s.

The commercial is nominally a pitch for Toyota’s Corolla, but if you watch the video–embedded below–its apparent that the car is really just a bit player in a thirty-second drama that’s really about what millennials do to achieve personal fulfillment.  The Corolla is Toyota’s entry level vehicle and is the world’s largest selling car model (nearly 50 million units to a mere 21 million for the Volkswagen Beetle). Like all car makers, Toyota aims to build a lifelong relationship with consumers, starting them out with a little Corolla in their twenties, graduating to a Camry as they get older, and then a Sienna minivan when they have kids. And if they’re successful, in later life they’ll graduate to a Lexus. So the purpose of advertising entry level vehicles is to establish an affinity with these customers early on. That’s not an easy task given the much lower rate at which young adults are getting drivers licenses, driving and buying new cars.

“You don’t own me” is a miniature drama. Its protagonist is a young twenty-something chef (picture a tattooed aspiring Top Chef contestant). In the opening scene of the 30-second commercial, she’s shown cooking away in an established tony restaurant, and then having her signature creation summarily tossed in the trash by a dismissive–and considerably older and male–head-chef). She immediately quits, throwing in her apron, and driving off in her Corolla, singing along to Leslie Gore’s 1964 hit “You don’t own me.”

In act two of this mini-drama there are a series of vignettes of other young adults. An African-American woman is ditching her dress shoes and tying on hiking boots at a state park, a peloton of cyclists pedals by, female roller derby players repeatedly crash into one other on an oval track. A group of young men is playing basketball. Trade magazine CampaignLife highlighted how the ad and song resonate with millennial aspirations:

The ad features young people triumphantly, you might even say defiantly, singing along to the song as they go about various activities like bonfires, group bike rides, and roller derby contests. Created by Saatchi & Saatchi LA, the ad’s best performance is among 21- to 35-year-olds, and interestingly displays stronger Desire and Relevance scores among males in that age group.

Strikingly, for a car ad, the activities highlighted in these vignettes mostly don’t involve driving. They’re set in urban spaces (parks, playgrounds, bike paths).  The car and its technology are essentially featured only once, in passing, as a lane alert signals a group of singing passengers that they’ve drifted across the centerline.

[table id=6 /]

The final act this little drama shows our young chef cooking in her new food truck, and dishing up one of her creations to another millennial standing outside. She’s been transformed from oppressed and disrespected, to an independent, creative entrepreneur. No one owns her.

Of course, this is a fable: most twenty-something start-up food truck owners would probably be maxing out their credit to make the payments on even a second-hand food truck; it’s likely that if they owned a car at all (rather than relying on their bike as a principal means of transportation) that they’d buy that used as well.

Its worth reflecting how different this drama is than Ford’s CG vision of cities of the future. Toyota appreciates that its potential customers are people who are more interested and engaged by all of the things that they can do in cities when they’re not in a car.

Unlike Ford’s futuristic vision, Toyota’s vision of a slightly idealized present focuses on people and how they live. Fittingly, the tag-line of the commercial is “Toyota:  Let’s go places.” The emphasis here is on “places.” And ultimately, that’s the difference between the Ford (and other techno-futurist) view of transportation and the view offered here. We attach value to the places we want to be. What we value in this “near” future is not being owned; being independent and engaging with other people.

Superficially, one might see a subliminal car-sharing message embedded in this ad: “you won’t own me” is essentially what the cars of the future are saying. Instead, there’ll be some combination of fleets of autonomous vehicles, along with much more widespread availability of “on-demand” rental cars like Car2Go, ReachNow and ZipCar. The more serious issue is that a key problems with cars, and our auto-dependent transportation is that in a sense “our cars do own us.” In many places its simply impossible to be a first class citizen without owning one. “You don’t own me” has been a kind of feminist anthem on and off over the years, and maybe that same slogan, applied to privately owned cars, should be a guiding principle for urban planning.

There’s a subtle but profound shift in what’s being sold here: Ultimately this moves us in the direction of transportation as a service. And transportation is just a means to an end (or set of ends) rather than an end in itself. Its good because it enables us to get and do the other things we want. It ceases to be an object of status or consumption good in its own right.

As we think about the future, and the kind of cities we want, maybe we should spend less time fetishizing modes of transportation, and more generally technology-and think about the kind of places we want to be in the the kind of experiences that they enable. Crafting the right kind of narrative about the cities and the lives we desire is an indispensable part of creating a better future.

Visions of the City Part II: A Perfect Day

Yesterday we took a close look at Ford’s vision for the future of cities. Our take: Ford’s preferred narrative of the places we’ll live is all about optimizing city life for vehicles. But is that the narrative that should guide us?

Another big global corporation has, perhaps unwittingly, given us a very different vision of cities–and life.  This other vision comes from Samsung, the Korea-based technology company.  They’ve been running a long form (60 second) television commercial called “A Perfect Day.” It follows the exploits of a half dozen kids–armed just with bikes, skateboards, and of course Samsung Galaxy smart phones–as they roam around New York City.  There’s a lot going on here, so let’s see if we can’t unpack all the different, and in many ways radical narrative its proposing.

A Perfect Day

First of all, they are in  a city. New York is front and center. This is not an anonymous or sanitized CG landscape. Its authentically and identifiably a city–a real city. And its shown from the perspective of actual humans experiencing it on the ground.

They are traveling by bike. The first scene of this micro-drama shows a platoon of cyclists (and one lagging skateboarder) set out in the morning, traveling in a marked bike lane on a residential street (in Queens or Brooklyn). They round a corner onto a busy arterial, and then ride across the Williamsburg bridge to Manhattan.

They are un-supervised by adults. The demographics of the group are just a little too perfect: teens and tweens, black, brown and white, boys and girls. But strikingly no adult authority figure is present. A parent calls only as dusk is falling (call answered via wrist-watch, naturally), only to be somewhat dismissively told “almost home,” with that message punctuated with a chorus of “Love you, Mom!” from the ensemble.

[table id=4 /]

 

They are hanging out in public spaces.They’re not in a den, a great room, a tech-laden suburban bedroom, or even a cosseted back yard.  They’re on the streets of the big city. They’re taking their own 3D photos and then sharing their virtual reality headset with a complete stranger they meet on the street. They’re at a skatepark under another towering bridge. They spend the afternoon hanging out at a public pool.

They are having experiences. The kids are recording and sharing their experiences with their Samsung devices. But in every case, the technology is incidental or subservient to the experience.

So here, in a nutshell, we have something that actually resembles a compelling future vision of cities. It includes technology, a little. But it isn’t about autonomous self-driving cars, or about side-walk internet kiosks or ubiquitous electronic surveillance.

Our vision of cities ought to be about the joy and wonder of the experiences we can have in them, not obsessing about the plumbing of moving people and stuff to and fro. For too long we’ve optimized our cities for the vehicles moving through them, rather than the people living in them. Samsung, or at least its creative agency, Weiden and Kennedy get this.

We’re not the only ones who were struck by this ad. Writing at her blog, Free Range Kids, Lenore Skenazy asked “What is this amazing Samsung ad trying to tell us?” The answer is pretty clear: If a city is a place where kids can roam and play, what else does it need to do?

Why narrative matters

In his Presidential Address to the American Economics Association two weeks ago, Nobelist Robert Shiller presented his thoughts on what he called “narrative economics.” Human beings are not the cold rational calculators they’re made out to be in traditional economic modeling. Instead, Shiller argued, human’s are hard-wired to visualize and understand the world through story-telling: We really ought to be called “Homo Narans.” That’s why getting the story right matters so much. If we have a story that centers on technology, vehicles and frenetic movement, we can remake our world in that image. If, instead, we have a story that embraces experience, and place and freedom, we’ll get a very different world.

It’s ultimately debatable whether Samsung’s version of “a perfect day” is one that everyone would agree with. But its an example of the kind of vision that might guide us, as we think about the kind of places we want to build. We should be deliberate in choosing our preferred narrative.

Visions of a future city, Part I

What stories do we tell ourselves about the kind of world we want to live in?

In his recent presidential address to the American Economics Association, Nobel Laureate Robert Shiller talked about “narrative economics.”   He argues that economists, like other disciplines need to begin to recognize that human cognition is structured around story-telling.

” . . . narratives, stories that seem outwardly to be of entertainment value only, are really central to human thinking and motivation.”

There’s something to this, and this week at City Observatory we’re exploring the question of what kinds of narratives are evident in the popular culture about cities and in particular the future of city living. Images can be a powerful way to communicate, and to sell ideas. Over the next few days, we will want to break away from our usual statistics heavy focus and talk explicitly about the images that are being used to describe cities and imagine their futures.

In the recent history of the American city, General Motors famous Futurama exhibit at the 1939 New York World’s Fair captured the imagination of Americans, and served as an iconic model of a new, auto-centric lifestyle that promised an end to traffic congestion and urban crowding. There’s no doubt that this image of a bright, mobile future appealed to a nation just recovering from the Great Depression. That image ultimately got reflected in policy–and pavement–with the enactment of the federal interstate highway program in the 1950s.

Vision statement, circa 1939.

At this year’s Consumer Electronics Show–which is now the place for automobile companies to roll out their newest ideas, technologies and models–Ford presented its remake of the Futurama, which it called “The City of Tomorrow”.  A short video illustrates their vision. It starts out with an immediate future (presumably 2020s world) where there are autonomous cars, and bikeways, and forests built right up the edge of high-rises (we think they’re subconsciously channelling Vancouver BC as the model here).

See? The future is multi-modal.

Right off the bat, you’ll notice that this is a road and vehicle-centered view of urban space.  Cars dominate. Sure, there’s a sop thrown to biking, pedestrians and transit, but notice this:  All of the pedestrians, and cyclists are shown traveling in parallel to the cars.  Walking and biking are just alternative ways of doing the same thing one would do, if one only had a car.  And never mind that a cycle-track in the middle of a four-lane arterial would be a hellish experience, and difficult to access, or that the bulb-outs for the cross walk are grass, rather than paving (this is after all, a car company’s visualization of what it would be like to be a pedestrian).  Its a striking consistency that illustrations of what future cities are like show individual human beings as tiny specks, and show a perspective of what the city will look like from someone floating somewhere in mid-air.

As the video progresses, Ford looks even further ahead, imagining what cities will look like decades from now when the full vision of vehicle autonomy is achieved. Now, small self-driving pods weave in and out, with computer-controls making sure that they don’t collide with one another (or with pedestrians, who simply walk across the street, sans signals.

In the not too distant future?

 

Clearly Ford is channeling some earlier thinking by MIT proposing that we do away entirely with traffic signals, and have cars flow steadily through intersections in all directions, with the speeds of individual vehicles controlled to allow gaps to open up and traffic to cross, marching-band style.

But who’s to say that Ford’s vision doesn’t actually morph into something more like the dystopia depicted in the 2008 Disney-Pixar movie, Wall-E.  Blob-like humans carted everywhere in floating electric reclining chairs with view screens permanently set in their line of vision.

Advance images from next years CES?

 

This is one image of what the future of transportation might look like. But there’s a real question as to whether anyone would want to live in such a place. As we’ve pointed out at City Observatory, Americans are now paying an increasingly large premium to live in places that are highly walkable, where they don’t have to drive so much. And more people, especially well-educated young adults are choosing to live in close-in urban neighborhoods. A city awash in vehicles and optimized for movement may have precious little reason for people to come, to stay, to work, or to live.

What this all boils down to is whether we build our cities as places to “be in” or places to “move through.” The automobile companies, and by and large technology companies and the engineering profession are all optimizing cities for moving vehicles. But as we’ve learned, a place that you can move through quickly, especially in a car (or private pod) is not a place that people want to spend time in. And while technologists in the car world are intent on pushing a vision of remaking the city in the image of ever more sophisticated vehicles, the folks at car companies that actually have to sell something today are much more attuned to the fact that people really don’t want to live in this kind of place.  We’ll present their images of how cities should work, and how people might live in them tomorrow.

As the largest car company in the world tries to sell the best-selling car in the world to the next generation of consumers, it recognizes that it has to tell a story not about cars and driving, but about entrepreneurship, eating, hiking, biking, and hanging out in urban places.

 

The enduring effect of education on regional economies

One of the themes we stress at City Observatory is the large and growing importance of talent (the education and skills of the population) to determining regional and local economic success. As we shift more and more to a knowledge-based economy, the places that will do well, and that are resilient in the face of change are the one’s with the best-educated populations.

One of the most robust statistical relationships we’ve observed is the strong correlation between the four-year college attainment rate (measuring the fraction of the adult population that’s completed at least a four-year college degree) and a place’s per capita income (its total income divided by the total population). Places with better educated populations have higher incomes. While we generally focus on the relationship at the metro level, we thought we’d step back this week and look at state level data over the past quarter century to see how this relationship has evolved.

Let’s start in 1990, and look at the correlation between state per capita income and the percent of the state’s adults with at least a four-year degree. To facilitate comparisons with more recent data, we’ve expressed 1990 per capita incomes in 2015 dollars using the Implicit Price Deflator for Personal Consumption Expenditures.  This table shows how a state’s educational attainment in 1990 was correlated with its average income in 1990.

The data show a strong positive correlation. Each one percentage point increase in the adult population with a four year degree was associated with an increase of about $950 dollars (expressed in 2015 dollars).  In a statistical sense, variations in education explain about 6x percent of the variation in per capita income among states.

Now move forward to 2015, the latest year for which we have state level data on educational attainment and per capita income. This chart has the same setup, but now shows 2015 adult educational attainment compared to 2015 levels of per capita income. Its similar, but now the relationship is stepper (each one percentage point increase in educational attainment in a state is now equal to a $1,070 increase in per capita income, and education alone explains 66 percent of the variation in income between states.  Keep in mind that idiosyncratic factors beyond educational attainment matter to state income: still flush from the oil boom, Wyoming, North Dakota and Alaska all had much higher per capita income in 2015 than one would expect based on education alone.

Together these two charts show that education is a strong and increasingly important factor related to state income growth. But what’s interesting is to go a step further and put these two analyses together. What we’ve done in the following chart is to compare 1990 levels of educational attainment (i.e. what fraction of a state’s adults had a four-year degree 25 years ago, and compare it to today’s level of per capita income. That relationship is shown here.

The key finding here is that 1990 educational attainment was a better predictors (i.e. higher coefficient of determination ) of 2015 income that it was of 1990 income. That is, educational attainment in 1990 was more strongly correlated with income 25 years later than in the current year. Each one percentage point increase in the four-year attainment rate in 1990 was associated with $1,600 higher per capita income in 2015. The most obvious reason for this is that many of the 1990 college educated citizens in a state were still around 25 years later. More generally, though, states with high levels of education (and income) may be more likely to invest in education, and send their children to college, and also be places that attract even more college graduates. Either way, the persistence of educational attainment seems to be a key factor correlated with long-term economic well-being.

The reason to pursue a talent-centered economic development strategy should be a strong motivation to reap the long-term benefits of a well-educated population.

A hat-tip to our colleague Phineas Baxendall at the Massachusetts Budget & Policy Center points us to similar work done by his organization which looks at the median wage for all workers by state, and compares it to each state’s college attainment rate. As with our analysis, there’s a very strong correlation.

The Week Observed, February 3, 2017

What City Observatory did this week

1.What HOT Lanes tell us about the value of travel time. The economic underpinning of claims that traffic congestion costs Americans billions and billions of dollars each year is the assumption that travelers would value time savings at about half their average wage rate, or around $15 per hour. But when given a chance to actually vote with their wallets, do travelers actually pay that much to save travel time? The adoption of High Occupancy Toll (HOT) lanes provides a natural experiment. Two researchers from the University of Washington and Louisiana State University have looked at the actual behavior of travelers on Seattle area HOT lanes and estimated that the willingness to pay for time savings works out to about $3 per hour. Travelers put a much higher value on reliability than time savings. These findings suggest that project’s ostensibly undertaken to save traveler’s time may not be economically justifiable.

2. How openness to immigration powers metropolitan economies. President Trump’s order closing the US border to persons from seven majority Islamic countries has produced widespread outcry as they conflict with fundamental American values and the constitution. But regardless of this order’s legality, the economic case for remaining open to immigration is powerful. America has flourished because its long been a haven to immigrants–many of them smart, entrepreneurial and refugees–who’ve helped generate new ideas and build new businesses. The data show that there’s a statistically strong relationship between the nation’s most productive metropolitan economies and those with the largest fraction of college-educated workers born outside the US.

3. Outdated “rules of thumb” distort our transportation planning decisions. Much of the policy making for transportation is clothed in the pseudo-scientific industry standards that measure levels of service, trip generation, lane widths and parking standards. But as we point out, many of these measures are out-dated, based on flawed data or analysis, or are simply biased in favor of promoting auto-mobility over other transportation modes and competing priorities.

4. Happy Groundhog’s Day, Oregon. Like many state’s and cities, Oregon has adopted an aggressive goal to reduce greenhouse gases–mostly in the far future. But with cheap gas and an uptick in driving, carbon emissions are actually increasing, meaning the state’s no longer on track to meet its modest 2020 goal, and stand’s little chance of reducing emissions 75 percent by 2050, as it law requires. Meanwhile, in the face of a big budget deficit, the Legislature is gearing up to pass measures to spend more money on expanding roads.

Oregon Global Warming Commission, 2017

 

Must read

1. America’s Great Divergence. The increasingly apparent cleavage in American society and politics is along educational lines, argues Alana Semuels in The Atlantic. Her reporting highlights the very dissimilar geographies and life experiences of those who get college degree compared to their peers who don’t. She contrasts the fate of those living in rural Connersville, Indiana, where manufacturing is in seemingly permanent decline, with the relatively much stronger economy of Indianpolis, where urban revival and a growing set of amenities are attracting and anchoring talent. Along the way she weaves in research insights from Bill Bishop’s Big Sort and findings from Enrico Moretti’s research on urban economies.

2. Whither infrastructure? The early signs from Washington are that infrastructure might be the one policy area where the new administration may move ahead with a major domestic policy initiative. The University of Pennsylvania Institute for Urban Research has assembled short essays from eighteen of the nation’s leading thinkers on urban affairs about what the nation should be thinking about. Mark Zandi calls for broadening the definition of infrastructure to include affordable housing, predicting that housing supply will increasingly become a constraint on economic growth. Gilles Duranton argues that we need to shift priorities in transportation spending from building new capacity to fixing the infrastructure we already have. Bob Yaro urges more reliance on pricing, including eliminating the current federal limits on tolling existing Interstate highways.

3. One  more infrastructure viewpoint. David Levinson, the University of Minnesota’s iconoclastic “Transportist,” has also weighed in on infrastructure policy. He argues that massive investment in infrastructure at this point is bad policy, for several reasons. First, like other observers, he argues we need to shift funding to maintenance, rather that spending money on expensive “ribbon-cutting” projects. He also argues that funding and project selection responsibility need to be devolved to the local level because most benefits are local. In addition, because the economy is approaching full employment, there aren’t the slack resources to build new infrastructure the way that there were during the depths of the recession, and as a result, infrastructure will compete with and displace other investment in the economy.

 

New Research

1. Which colleges promote economic mobility. Raj Chetty and his colleagues at the Equality of Opportunity project have produced another stunning big-data analysis, this time tracking the economic outcomes of graduates of the nation’s college’s and universities. There work uses anonymized tax records to identify the family income of college students, by institution and then to track their post-college incomes. The results show which colleges serve students from throughout the income distribution, and which do the best job of helping their students rise economically. Among the institutions who serve the biggest fraction of students from the bottom quintile of the income distribution, and then see those students end up in the top fifth of the income distribution include  Cal State Los Angeles, the City University of New York (CUNY), and the University of Texas El Paso (UTEP).  The report provides data for every major college and university in the US, and gives us a much more detailed picture than ever before of who’s served and how graduates fare. A must read.

Equality of Opportunity Project

2.  Mandated parking spaces sit unused. Smart Growth America published a new report “Empty Spaces” looking at the actual levels of utilization of parking in five transit-oriented developments around the country. Their analysis shows that these developments generated between one-third and two-thirds as many car trips as predicted by the Institute of Transportation Engineers handbook (one of those deeply flawed “rules of thumb” we mentioned this week). Because so many more people used transit, walked or rode bikes than predicted, the parking spaces mandated by the handbook and local rules were significantly under utilized. Peak parking occupancy for these projects was between 19 percent and 46 percent of the levels predicted by ITE. The report is powerful additional evidence that parking requirements based on ITE standards lead to expensive wasted space. If we’re looking for regulations to reduce or eliminate, these would be a good place to start.

 

 

Happy Groundhog’s Day, Oregon

Climate change gets lip service, highways get billions.

Like many states and cities, Oregon has been a leader in setting its own local goals for reducing greenhouse gases. In a law adopted in 2007, the state set the goal of reducing its greenhouse gas emissions by 10 percent from 1990 levels by the year 2020, and the further goal of reducing them by 75 percent by 2050. In light of the Trump Administration’s denial of the scientific evidence on global warming, many environmental activists are pinning their hopes for progress on this kind of local effort.

Sadly, there’s a deep flaw in this approach. Despite the high-minded and quantitative nature of these goals, the actual date for their achievement is set far in the future, typically beyond the expected political lifetime of any of the public officials adopting these goals. And there’s little if any mechanism, aside from moral suasion, to require accomplishment of these goals. So in practice, what they may do is simply give politicians cover for expressing concern about climate change, without actually having to do anything substantive or difficult to attain it.

Purple mountain tragedy. Oregon’s iconic Mt. Hood sees its snowpack disappearing.

That fear was brought home by the release of the Oregon Global Warming Commission’s biennial report to the state Legislature. In addition to its goal, Oregon has a tiny citizen commission charged with riding herd on the state’s emissions inventory, and looking to see what, if any progress the state is making in reducing greenhouse gases. The news is not good. Oregon’s emissions are rising. The commission warns:

Key Takeaway: Rising transportation emissions are driving increases in statewide emissions.

As the updated greenhouse gas inventory data clearly indicate, Oregon’s emissions had been declining or holding relatively steady through 2014 but recorded a non-trivial increase between 2014 and 2015. The majority of this increase (60%) was due to increased emissions from the transportation sector, specifically the use of gasoline and diesel. The reversal of the recent trend in emissions declines, both in the transportation sector and statewide, likely means that Oregon will not meet its 2020 emission reduction goal. More action is needed, particularly in the transportation sector, if the state is to meet our longer-term GHG reduction goals.

This finding confirms exactly what we’ve pointed out at City Observatory over the past year: the decline in gasoline prices in mid-2014 prompted an increasing in driving and with it, an increase in crashes and carbon pollution.  Oregon’s vehicle miles traveled, which had been declining steadily, ticked up in 2015, as did its fatality rate.

As a result of the growth in driving and related emissions, and slower than expected progress in reducing emissions from other sources, it looks like there’s no way the state is anywhere close to the path it needs to be on to reach its 2050 goal. (The blue line shows progress to date, the yellow line is the glide slope to achieving the state’s adopted 2050 goal, and the red line is the commission’s estimate of where the state is headed).

Oregon Global Warming Commission, 2017

In the best of all possible worlds, this warning would prompt the Governor and legislators–ever mindful of their legally enacted commitment to reduce greenhouse gas emissions–to redouble their efforts and figure out ways to discourage carbon pollution, especially from transportation.

In the real world, legislators are actually poised to do pretty much the opposite. One of the leading priorities for the 2017 Oregon legislature is the enactment of a major new transportation package. Among its chief priorities: adding additional capacity to the state’s freeways. A new legislative report identifies a “need” for $1.3 billion annually in additional transportation funds for state and local highway projects.  As has been demonstrated time and again, additional capacity produces additional driving and increased emissions. And ironically, this proposal comes forward as the state faces a $1.7 billion shortfall in its general fund budget for the coming biennium, which is likely to lead to significant cuts to funding for schools and other public services.

Climate data now shows that 2016 was the warmest year on record. Global warming is an increasingly obvious reality. We’re going to need something more than soothing rhetoric and distant goals to avoid dramatically altering our planet. We’ll check in again next Groundhog’s day to see if anything’s different.

 

Our old planning rules of thumb are “all thumbs”

We all know and use rules of thumb. They’re handy for simplifying otherwise difficult problems and quickly making reasonably prudent decisions. We know that we should measure twice and cut once, that a stitch in time saves nine, and that we should allow a little extra following distance when the roads are slick.

What purport to be “standards” in the worlds of transportation and land use are in many cases just elaborate rules of thumb. And while they might have made sense in some limited or original context, the cumulative effect of these rules is that we have a transportation system which is by regulation, practice, and received wisdom, “all thumbs.”

How we feel about bad rules of thumb. Get it? Credit: Jesper Ronn-Jensen, Flickr.
How we feel about bad rules of thumb. Get it? Credit: Jesper Ronn-Jensen, Flickr.

 

One of the problems with rules of thumb (or the more academic term, “heuristics”) is that while they may work well in many cases, they may work very poorly in others – and they may be subject to important cognitive biases that lead us to make bad decisions.

Here are five rules of thumb that have led to a distorted view of our transportation problems and their appropriate solutions.

Old rule of thumb #1: We should have a high “level of service” on our streets. Around the country, traffic engineers have long assigned one of six letter grades A through F to describe traffic flow on streets. (A is free-flowing traffic, F is highly congested.) Many planning decisions emphasize the need to maintain high levels of service, which means that roads are designed to be much bigger (and more expensive) than they need to be most of the time. And level of service only measures car travel time on a particular road, ignoring non-car travelers, and – importantly – the effect of more roads on sprawl and overall trip lengths. These flaws have lead California to eliminate level of service as a factor in environmental analyses of traffic impacts.

Old rule of thumb #2: Wider streets are safer streets. It’s long been an engineering axiom that wider roads are safer, because they give cars and others more space to avoid collisions. But the behavioral effects of wider roads overwhelm the supposed safety advantages. Wider lanes encourage vehicles to drive faster, and higher speeds produce deadlier consequences—especially for cyclists and pedestrians. New research shows that the optimal lane width for minimizing crashes and injuries is something like 10 or 11 feet, not the 12-14 feet of many travel lanes in streets around the country.

The wider the lanes, the easier it is to speed. Credit: Pier-Luc Bergeron, Flickr.
The wider the lanes, the easier it is to speed. Credit: Pier-Luc Bergeron, Flickr.

 

Old rule of thumb #3: We should require “enough” off-street parking for every use. As Donald Shoup has shown, parking requirements spelled out in zoning codes—often based on formidably inaccurate estimates prepared by the ITE (Institute of Traffic Engineers) lead to a situation where every business’s parking lot is sized for the peak hour of the peak day of the year (holiday shopping season at the Mall, example). Not only does this produce more parking than is needed the rest of the year, it turns out that parking “requirements” grossly overstate demand even in peak periods, and especially for urban uses where more people arrive by other means, and park for shorter periods of time. As Smart Growth America’s report “Empty Spaces shows, when developments have density, transit access and mixed uses, spaces mandated by parking requirements simply sit unused.. The product of this rule of thumb is that parking is over-supplied, destinations are further apart than they would otherwise be, and walking, transit and cycling are non –functional.

Old rule of thumb #4: We should plan for a certain number of car trips to be generated by every land use, no matter where it is. Another rule of thumb for planning is that every land use “creates” or generates a certain number of trips. But it isn’t necessarily so: the studies used to make these esimates are drawn from large-scale suburban development where proportionately more trips are by auto. A careful analysis of the data shows that trip generation estimates for most uses are overstated by a factor of 2, leading local governments to require even more transportation capacity than is needed—driving up development costs, and inducing additional travel.

Old rule of thumb #5: We should have a hierarchy of streets. The street hierarchy makes an explicit analogy to the human circulatory system. Just as we have an increasingly fine array of arteries, veins and capillaries, so too does the transportation system have freeways, arterials, collectors and local streets. And we’ve abandoned the traditional street grid for a dendritic pattern. It turns out that these hierarchical street systems are less resilient to disruption and have less capacity than the old-fashioned grids they replace, and are especially hostile to non-automotive modes of travel (pedestrians and bikes are forced to take circuitous routes and are hard to accommodate at the intersections of major arterials that have limited “green” time to accommodate cross=traffic and turning movements. The hierarchal system of “arterials, collectors, and local roads that we’ve adopted in place of the traditional street grid has had the effect of making the average trip between any two points longer. Over the past two decades the “circuity” of trips has increased by 3.7 percent in the nation’s 50 largest metropolitan areas. This increase is on top of the increase in trip distance due to sprawl and decentralization.

 

Our “all thumbs” approach to transportation planning leads to a specific pattern of development that is as inefficient for cars as it is hostile for persons traveling on foot, via bicycle and on transit.

What is needed are a new set of rules of thumb. Like all heuristics, this isn’t meant to be taken as a final set of “standards” to fit every situation – but there are some emerging ideas about what we might emphasize.

New rule of thumb #1: Closer is better. Having more different destinations close at hand facilitates a wide range of mode choices, especially walking and cycling. Mixing uses, which is often anathema under traditional zoning codes turns out to be desirable for consumers and expeditious for transportation.

New rule of thumb #2: Slower is safer. When cars and people on foot and on bikes interact, safety comes from slow speeds even more than separation. Local streets that move traffic slowly are friendlier—and safer—for non-auto modes of transportation.

15106162164_1b7fc8ce47_z
Source: NYC DOT, Flickr.

 

New rule of thumb #3: Sharing is efficient. Rather than require every use to provide parking for the peak hour of the year, arranging uses so people can park once, and walk mostly leads to less traffic, greater safety and more congenial, fine-grained development patterns.

New rule of thumb #4: Our objective should be accessibility, not mobility. Many transportation heuristics emphasize speed: how do we make things move faster. But what we really care about is getting to (or being at) our destinations, not rapidly traveling among them. There’s great new work being done on how to measure accessibility, and use it as a guide to policy.  Speed should be secondary to choice.

Of course, these new “rules of thumb” are just a beginning. There’s a lot of work to be done to un-learn and re-think the unfortunate heuristics we’ve employed in thinking about transportation planning and land use. But as these examples illustrate, re-thinking these issues isn’t a purely technical matter: it depends critically on re-imagining the way we visualize and tell stories about how our transportation system works.