Openness to immigration drives economic success

Last Friday, President Trump signed an Executive Order effectively blocking entry to the US for nationals of seven countries—Iraq, Iran, Libya, Somalia, Sudan, Syria, and Yemen. We’ll leave aside the fearful, xenophobic and anti-American aspects of this policy: others have addressed them far more eloquently than we can at City Observatory.  And while there’s no question that the moral, ethical and constitutional problems with this order are more that sufficient to invalidate it, to these we’ll add an economic angle, which though secondary, is hardly minor.

Lighting the way to a stronger US economy since 1886.


America is a nation of immigrants, and its economy is propelled and activated by its openness to immigration and the new ideas and entrepreneurial energy that immigrants provide. Its commonplace to remind ourselves that many of the nation’s greatest thinkers and entrepreneurs, Andrew Carnegie, Albert Einstein, Andy Grove and hundreds of others, were immigrants, if not refugees. All six of America’s 2016 Nobel Laureates were immigrants. The fact that America stood as a beacon of freedom, and a haven from hate and oppression, has continually renewed and added to the nation’s talent and ideas. Immigration has also played a critical role in helping revitalize many previously depressed urban areas and neighborhoods. As Joel Mokyr explains in his terrific new book “A Culture of Growth,” the key factor triggering the Enlightenment and the Industrial Revolution was the ease with which heterodox and creative thinkers could find sanctuary in other countries and spread their thinking across borders. The US was founded on the kind of openness and tolerance than underpinned this process, and flourished accordingly.

The critical role of immigration is abundantly clear when we look at the health and productivity of the nation’s urban economies. The metro areas with the highest fractions of foreign-born well-educated workers are among the nation’s most productive.

Metros with the most foreign born talent

Our benchmark for measuring foreign-born talent is to look at the proportion of a region’s college-educated population born outside the United States. We tap data from the Census Bureau’s American Community Survey, which tells us what share of those aged 25 and older who have at least a four-year college degree were born outside the United States. (This tabulation doesn’t distinguish between those who came to the US as children and were educated here, and those who may have immigrated to the US later in life as adults, but shows the gross effect of all immigration). In the typical large metropolitan area in the United States, about one in seven college educated adults was born outside the nation. And in some of our largest and most economically important metropolitan areas, the share is much higher: a majority of those with four-year or higher degrees in Silicon Valley are from elsewhere, as are a third of the best educated in New York, Los Angeles, and Miami.


Foreign born talent and productivity

We’ve plotted the relationship between the share of a metropolitan area’s college-educated population born outside the United States and its productivity, as measured by gross metropolitan product per capita.  Gross metropolitan product is the regional analog of gross domestic product, the total value of goods and services produced, and is calculated by the Bureau of Economic Analysis.  The sizes of the circles shown in this chart are proportional to the population of each of these metropolitan areas.

These data show a clear positive relationship between the presence of foreign-born talent and productivity.  Several of the nation’s most productive metropolitan areas–San Jose, San Francisco, New York and Seattle–all have above average levels of foreign-born persons among their best educated.

Of course, these data represent only a correlation, and there are good reasons to believe that the arrows of causality run in both directions: more well-educated immigrants make an area more productive and more productive areas tend to attract (and retain) more talented immigrants. But it’s striking that some of the nation’s most vibrant economies, places that are at the forefront of generating the new ideas and technology that sustain US global economic leadership, are places that are open and welcoming to the best and brightest from around the world.

There are a lot of reasons to oppose President Trump’s ban on immigration from these Islamic countries. The most important reasons are moral, ethical and legal. But on top of them, there’s a strongly pragmatic, economic rationale as well: the health and dynamism of the US economy, and of the metropolitan areas that power the knowledge-driven sectors of that economy, depend critically on the openness to smart people from around the world.



Constant change and gentrification

A new study of gentrification sheds light on how neighborhoods change.  Here are the takeaways:

  • The population of urban neighborhoods is always changing because moving is so common, especially for renters.
  • There’s little evidence that gentrification causes overall rates of moving to increase, either for homeowners or renters.
  • Homeowners don’t seem to be affected at all, and there’s no evidence that higher property taxes (or property tax breaks) influence moving decisions.
  • While involuntary moves for renters increase slightly in gentrified neighborhoods, there’s no significant change in total moves

We’ve been closely reading a new study on gentrification and neighborhood change. In an article published in Urban Affairs Review, “Gentrification, Property Tax Limitation and Displacement,” Isaac William Martin and Keven Beck present their analysis of longitudinal data from the Panel Survey of Income Dynamics that track family moves over more than a decade.  An un-gated version of the paper is available here. One of the challenges of studying gentrification and neighborhood change is that most data simply provides snapshots of a neighborhood’s population at a given point in time, and provides little information about the comings and goings of different households. The PSID sample is unusual, in that in tracks households and individuals over a period of decades–this study uses data on the movement of household heads from 1987 through 2009. Martin and Beck were able to access confidential data that reports neighborhood location and enables them to identify the movement of households to different neighborhoods.

Richard Florida reviewed the Martin and Beck paper at City Lab and highlighted two of the study’s key findings:  that homeowners don’t seem to be displaced by gentrification and a subsidiary finding that property taxes (and tax breaks for homeowners) don’t seem to affect displacement.  These are both significant findings, but we want to step back and look at the broader picture this study paints of how neighborhoods change, because this study provides a useful context for understanding the complex dynamics of migration that are often left out of discussions of gentrification.

Change is a constant–Most renters have moved on after two years

One of the most striking findings from this study is how frequently renters move. These data show than in any given two-year period a majority (54 percent) of renter households had moved to a different neighborhood. The average tenure (length of time they’ve lived in their current residence) is on average 1.7 years. (Table 1).  Moving rates are lower (16 percent over two years) for homeowners, and average tenures are considerably longer (4.9 years, on average).  But the important thing to keep in mind is just how much volatility and turnover there is in neighborhood populations. Statistically, if about half of all renters move out of a neighborhood every two years, the probability than any current renter will live in that neighborhood ten years hence is about 3 percent (0.5 raised to the fifth power).

Many of the public discussions of gentrification assume that somehow, in the absence of gentrification, neighborhoods would somehow remain just the same, and that few or no residents would move away. This study shows reminds us that this isn’t true. In addition, we know that for poor neighborhoods that don’t see reductions in poverty rates, that population steadily declines. Our own study of poor neighborhoods shows that over 4 decades, the three-quarters of poor neighborhoods that didn’t rebound lost 40 percent of their population.

Most moves are voluntary

Unlike many other studies, the Martin and Beck paper is able to use survey data to try and discern the motivations for household moves. Broadly speaking they divide moves into “voluntary” and “involuntary” moves.  The PSID asks movers why they moved, and those that respond to this open-ended question with answers coded as “moved in response to outside events including being evicted, health reasons, divorce, joining the armed services, or other involuntary reasons” are treated as involuntary moves.As they note, the distinction isn’t always as sharp as one would like, and it may be that some respondents rationalize some involuntary moves as voluntary ones, but the self-reported data are clear:  among renters, voluntary moves dramatically outnumber involuntary ones.  About 54 percent of all renters moved in the last two years; about 13 percent of all renters reported an involuntary move.  That means that about 75 percent of all renter moves were voluntary and about 25 percent of renter moves were involuntary.  As Margery Turner and her colleagues at the Urban Institute have shown, moving to another neighborhood is often the way poor families get better access to jobs, better quality schools, safer neighborhoods and better housing.

Gentrification has no impact on overall renter moves, but is associated with a small increase in involuntary moves

One of the most important studies of gentrification is Lance Freeman’s 2005 paper “Displacement or Succession?: Residential Mobility in Gentrifying Neighborhoods” which found that gentrification had essentially no effect on the rate at which households moved out of gentrifying neighborhoods.  Martin and Beck replicate this finding for all moves by renter households, they write:

Consistent with Freeman’s findings, Model 2 indicates that we cannot be confident that the average effect of gentrification on the probability of moving out is different from zero.

Graphically, Martin and Beck’s findings are can be depicted as follows.  About 54 percent of all renters move within two years. According to Martin and Beck’s modeling, the probability that a person in a gentrifying neighborhood moves in two years is about 1.7 percentage points  greater than for the typical person (after controlling for individual household characteristics). That suggests that for a typical resident, their probability of moving in a gentrifying neighborhood is about 55.7 percent, but that estimate in not statistically significant.

When they look just at “involuntary” moves, however, they find that there is a statistically significant effect of gentrification on the probability of moving.  Specifically, they find that rental households in in gentrifying neighborhoods are about 2.6 percent points more likely to report an in “involuntary move” in the past two years than those who don’t live in gentrifying neighborhoods.  Its important to put that in context.  According to the paper, about 54% of all renters moved in the last two years, and about 13 percent of them experienced an “involuntary move.”  The estimate in the paper is that the effect of living in a gentrifying neighborhood is about a 2.6 percentage point increase in the likelihood of an “involuntary” move.  That means if the average renter has a 13 percent chance of an involuntary move, a renter in a gentrifying neighborhood has a 15.6 percent chance of such a move.  These results are shown below:

Here, the estimate that a renter makes an involuntary move from a gentrifying neighborhood  (+2.6 percentage points) is greater than the 95 percent confidence interval, which suggest that there is a statistically significant difference between the share of the population experiencing involuntary moves in gentrifying neighborhoods as compared to all neighborhoods.

What would that look like in a typical neighborhood?  If you have a neighborhood with 2,000 households (about 5,000 people, with about 2.5 persons per household), and about half are renters and half are homeowners, you would expect of the 1,000 renting households that about 130 households would experience an involuntary move over a two year period.  If that tract gentrified, you would expect an additional 26 households to experience an “involuntary move.” But you would also expect 530 total households to have moved out of the neighborhood in that time, for all reasons, voluntary and involuntary.  These data put the scale of the gentrification effect in perspective. Whether or not they gentrify, there’s going to be enormous change in the renter population of any given urban neighborhood.

Gentrification has no impact on homeowner moves

Martin and Beck find no evidence that homeowners in gentrifying neighobrhoods are more likely to move, either in the aggregate, or involuntarily.  They test a number of different models of the connection between gentrification and moving: none produce statistically significant correlations between gentrification and moving; in some cases (though statistically insignificant) the correlation is negative: gentrification is associated with fewer homeowners moving from a neighborhood.  Their conclusion: for homeowners, their study “produces no evidence of displacement from gentrifying neighborhoods.”

Property taxes (and tax breaks) seem to have no connection with homeowner movement from gentrifying neighborhoods

One popular argument is that gentrification pushes up property values and results in higher property taxes for homeowners, and that especially for households with a fixed income, the burden of higher property taxes is likely to force them to move. Martin and Beck look closely at this question, and examine how changes in property assessments and property taxes correlate with the probability of moving. They find that there’s no statistically significant link between property taxes and moving in gentrifying neighborhoods.  Several states and localities have enacted property tax or assessment limitations, in part with the objective of lessening the financial exposure of fixed income households to the burden of higher property taxes. Martin and Beck look at the relationship between such limits and the probability of moving, and find that such limits don’t seem to have any effect on whether homeowners move out of gentrifying neighborhoods or not.

While homeowners in gentrifying neighborhoods have to shoulder the burden of paying higher property taxes, its typically only because their homes have appreciated more in value. In most cities, property taxes are levied at a rate equal to about 1 to 2 percent of a property’s market value, so the wealth effect of property appreciation dwarfs the negative income effect of having to pay higher property taxes.

Urban renters are a highly mobile group. Most renting households are likely to have changed neighborhoods in the past two years. We observe the same overall level of movement out of neighborhoods whether they gentrify or not.  This study suggests that somewhat more of those moves would be involuntary rather than voluntary.


This post has been revised to correct typographical errors, and to replace an earlier data table with charts illustrating the same information.

How urban geometry creates neighborhood identity

Does geometry bias our view of how neighborhoods work?

Imagine a neighborhood that looks like this:

On any given block, there might be a handful of small apartment buildings—three-flats—which are usually clustered near intersections and on major streets. Everything else is modest single-family homes, built on lots the same size as the three-flats.

What kind of community is this? Well, if you were to walk, bike, or drive around it, you would spend most of your time in front of these bungalows, which make up, on the block pictured above, fully 75 percent of the buildings. Visually, they define the landscape; the three-flats are accents, notable but clearly in the minority.

If you lived in this community—particularly if you lived in one of the bungalows—this visual character might be something you’re attached to, and identify with. You might begin to define your neighborhood by these bungalows, and expect the neighborhood’s future changes to conform with this identity.

And yet there’s something curious here: equal numbers of families live in bungalows and three-flats in the neighborhood pictured above. There are nine bungalows, each with one family; and three three-flats, each with three. (And if any of those three-flats have converted garden apartments, there are more people in the three-flats!)

But basic rules of geometry mean that if there are equal numbers of people in higher-density and lower-density housing types in the same neighborhood, the people in the lower-density housing will take up much more space—and, maybe, have an advantage in defining the identity of their neighborhood. (You’ve certainly noticed a similar dynamic with maps of the presidential race by county: a sea of low-density counties in red visually swamps the fewer, but much higher-density, counties in blue.)

Does this matter? I think yes, because the power to define a neighborhood’s publicly accepted identity also brings with it a great amount of power in shaping its future development. That’s especially the case in cities like Chicago, where local aldermen representing relatively small areas have near-veto power over new housing, businesses, and many transportation decisions within their wards. A group of people who manage to convince their alderman that a particular development, or streetscape, is “out of character” with the neighborhood’s identity is often able to defeat it.

This is especially relevant because the low-density/high-density housing usually corresponds to other axes of unbalanced power: within any given neighborhood, people in higher-density housing usually have lower average incomes, and are more likely to be people of color. What’s more, they’re also likely to be younger and renters rather than owners—and so statistically less likely to sit on a neighborhood board, or attend public meetings. A dynamic that privileges the ability of people in low-density housing to define and shape their neighborhood, then, is likely to reinforce some of the most basic inequalities of American society.

Nor is this only a theoretical issue. I thought of it after reading articles like this one, about Jefferson Park on the far northwest side of Chicago. “Should Jefferson Park Keep Suburban Vibe?” the headline asks, referring to some locals’ opposition to any new multifamily housing. Much of Jefferson Park looks a good deal like my imaginary neighborhood above; it’s generally identified with the city’s much-loved “bungalow belt” of early twentieth century single-family homes. Thus its identity as “suburban,” relative to the denser neighborhoods to the east.

But this widely accepted identity—one taken for granted in the headline of a story about whether the neighborhood ought to accept new high-density residents—is an artifact of urban geometry. According to the Chicago area’s metropolitan planning organization, 72 percent of the residential land in Jefferson Park is taken up with single-family homes. But most people who live in Jefferson Park—52 percent—actually live in an apartment or condo.

There’s obviously no smoking gun here about the power to define the future of the neighborhood. Can a neighborhood where most people live in multifamily housing be said to have a “suburban vibe” in this sense? If not, does that mean any of the people who strongly oppose new multifamily housing, and the people who would live in it, would change their minds? Or would their rhetoric be less powerful to those (probably the vast majority) who don’t have a strong opinion? To the alderman?

It’s hard to know. But it seems unlikely—especially if you believe any of the arguments made by people like Sonia Hirt about the cultural power of the idea of the single-family home—that these sorts of constructed identities don’t have some kind of effect on the paths that neighborhoods take.

Of course, there is a flip side to the way that urban geometry distorts people’s perceptions of how most of their neighbors live. And that’s that it’s possible to add much more housing without changing the visual character of the neighborhood in the same proportion. The question is whether it’s possible to add that housing—contributing, on average, to more diverse, affordable, and sustainable cities—when people believe (rightly or wrongly) that the character of their neighborhood must change to accommodate it.


What HOT lanes reveal about the value of travel time

Every year, the Texas Transportation Institute, and traffic monitoring firms like Inrix and Tom-Tom trot out scary sounding reports that claim that Americans lose billions or tens of billions of dollars worth of time sitting in traffic. And just as regularly, highway advocates parrot these dire sounding numbers as the justification for spending billions and billions on highway expansion projects (like a Texas plan to spend $70 billion in the next decade).

All these estimates and this case hinge on one crucial assumption:  What dollar value travelers attach to time savings?  But as a lawyer might argue, all of these studies are “assuming facts that are not in evidence.” Everyone complains about how awful and burdensome traffic congestion is, but when presented with an opportunity to buy a quicker trip, do people put their money where their mouth is? As we saw earlier this month in Louisville, given the opportunity to save several minutes time by paying a toll to use that city’s new, wider I-65 bridges, many travelers opted to save a few bucks by driving over the older, parallel–and un-tolled–highway bridge.

There’s a rule of thumb that travelers value their time at something close to half their wage rate.  The 50 percent rule is based on statistical inference from studies of traveler behavior (“revealed preference”) and surveys of what people say they would do when confronted with a range of hypothetical situations (“stated preference”).  There’s a range of estimates, centered mostly on 50 percent of the wage rate, which is the official standard used by the U.S. Department of Transportation, and works out to about $12.50 per hour (in 2009).

But what’s been lacking is solid, experimental evidence: when people are actually presented with a choice to spend some specific dollar amount to save a few minutes of travel time, do they actually pony up?

The implementation of high occupancy toll (HOT) lanes in a growing number of metropolitan areas provide us with exactly the kind of experimental evidence to evaluate travel times.  A recent paper from the University of Washington’s Austin Gross and Louisiana State University’s Daniel Brent explores driver behavior on high occupancy toll lanes.

SR 167 High Occupancy Toll Lanes (WSDOT).

An important feature of these lanes is that they use dynamic pricing:  the amount of the toll varies depending on the level of traffic.  The system is calibrated to adjust the price every five minutes, and raises (and lowers) the price so that traffic in the HOT lane moves along at at least 45 miles per hour.  When there’s light traffic, the toll is low; when traffic is heavier–and the HOT lane offers a faster trip relative to the road’s un-tolled lanes–the toll is higher.

Three years ago, Zachary Howard and Clark Williams-Derry at the Sightline Institute pointed out that preliminary data from Washington’s Highway 167 HOT lanes implied that travel time values were a lot lower than everyone suggested. They noted that most drivers were unwilling to pay any amount to speed up their trip, and the very small fraction of travelers that were using the HOT lanes were paying a toll that valued their time savings at something like $12 an hour, a figure that they came up with by dividing the number of minutes saved by the toll rate.

But Gross and Brent point out that the time savings are only part of the value:  It’s not just a quick trip people want, it’s a reduction in uncertainty about when they’ll arrive.  If you need to catch a plane, start a meeting, or open your shop at a certain hour, you’ll pay extra for the certainty of arriving on-time.  It’s worth noting, for example than many more users pay tolls to use the HOT lanes in the morning–i.e. going to work–than they do in the afternoon–going home; this difference reflects the distinction between reliability as opposed to time savings. Reliability trumps time savings in most toll-payers calculus:

The main result is that VOR (value of reliability) is vastly more important than VOT (value of time); in fact in several specifications the VOT is insignificant. In the base specification the reliability ratio (VOR/VOT), is 7.5 indicating that reliability is much more important in using the HOT lane than time savings. These results suggest that the simple estimates of VOT on HOT lanes vastly overestimate the true VOT, and that much of the purchase decision is actually based on improved reliability.

Gross and Brent use data on about a million observed trips to estimate how much value the users of the system attach to their travel time savings and to the increased reliability of their trip.  They estimate that the typical user values travel time savings at about $3 per hour, and reliability improvements at about $23 per hour. This estimate, particularly of the value of  travel time savings is much lower than is regularly used in analyzing and justifying transportation investments.

Gross and Brent also use surveys of transponder owners, combined with census data to estimate the typical income of HOT lane users. Unsurprisingly, higher income households are much more likely to be HOT lane users–the median income of HOT lane users is over $100,000 annually, much higher than for commuters in the corridor. This is also important because it signals that even higher income households have a much lower value of time than is generally assumed in modeling and project evaluation.

Using too low a value of time means that DOTs overestimate the economic utility of projects–and also over-estimate the revenue that tolled projects produce. In the case of the Highway 167 HOT lanes, much lower value of travel time means that the HOT lanes have produced only about a third of the revenue that was projected before the lanes were opened.

If travelers attach so little value travel time savings, this calls into question the rationale for investing public funds in highway projects. Benefit-cost analyses used to justify highway projects count the estimated travel time savings, often valued at around $15 per hour, as the benefit of the project. If the real value of travel time savings is something like $3 an hour, that reduces the benefits by about 80 percent. So if a project had a benefit cost ratio of 5 to 1 under a $15 value of travel time assumption, it would essentially have no net value if travel time were valued at only $3 per hour.

As with Louisville’s new tolled bridges, the experience with HOT lanes is a kind of natural experiment that tells us how much value people attach to travel time savings. In the case of the HOT lanes, the economic evidence is that the tolling promotes greater efficiency, by giving those who value travel time savings–and actually travel time reliability–the option of paying for and getting a higher level of service. But it also suggests that many investments of scarce public resources in additional unpriced road capacity isn’t economically worthwhile for the travelers who use it.



The Week Observed, January 20, 2017

What City Observatory did this week

1. The long journey toward greater equity in transportation. The observance of Dr. Martin Luther King’s birthday got us thinking about how far we’ve come–and how far we have yet to go–having a truly equitable society. We reviewed two recent studies that address lingering racial disparities in transportation. The first sheds some evidence on “driving while black”: given their higher probability of being pulled over for traffic infractions, blacks seem to be quite cautious drivers, with average speeds 8 percent slower than other drivers. A second study shows that there are racial disparities in ride-hailing services; passengers with black-sounding names were more likely to have rides cancelled.

2. Beer and crowd-sourced data. The number of breweries in the US has nearly tripled in the past decade. We’ve used a variety of data sources, including crowd-sourced directories of microbreweries to plot the density of microbreweries in cities across the country. To this, we’ve added Treasury Department data on the total number of breweries per capita in each state. Some clear regional patterns emerge, with the most breweries in the Northeast and Pacific Northwest and the fewest in the South.

3. Has Louisville figured out how to eliminate traffic congestion? Louisville has just started charging tolls to vehicles using its new multi-billion dollar bridges over the Ohio River. But the tolls only apply to three of the five bridges–which creates incentives for drivers to use the non-tolled routes. While we don’t have actual data on traffic patterns, we were able to take a quick look at area traffic cameras. They show that the 10-travel lanes of the new I-65 bridge were nearly empty at rush hour on a recent Tuesday afternoon. Meanwhile, an older, slower, narrower bridge–without tolls–had substantial traffic.

4. An experiment in transportation economics. It turns out that Louisville’s decision to toll some bridges and not toll others creates a kind of natural experiment that will provide some useful data on how much value travelers attach to the time savings associated with highway improvement projects. The economic justification for these investments is that the value of travel time savings exceeds the costs of building new capacity. If users aren’t willing to pay a toll to get these time savings, its a signal that there may be no good economic reason to expand the roads.

Must read

1. What have we learned about the causes of gentrification?  Two economists for the Philadelphia Federal Reserve have written a review of recent literature on neighborhood change, published in the latest edition of Cityscape. Jacklyn Hwang and Jeremy Lin explore the causes of gentrification, and offer a comprehensive tour of recent scholarship on the subject. They conclude that the answer to the question posed by their article remains elusive. Socioeconomic upgrading appears correlated with central city job growth, improved urban amenities and declining crime, but the patterns of causality are complicated and unclear.

2. How housing choices impair adult friendships. Writing at, David Roberts considers the connections between our low density land use patterns and the nature of our adult relationships. Our lower density communities and auto-oriented lifestyles work against three of the critical ingredients that sociologist believe are key to forming friendships: proximity, repeated and unplanned interactions, and an environment that permits us to let our guard down. The antidote? Wallkable communities and lively public spaces that facilitate spontaneous interactions.

3. Infographic on road-pricing equity. Economists love road pricing, but often seem to be the the only ones. Nobody likes to pay for something they perceive to be free, but frequently the most powerful argument is that road pricing will hurt the poor, because tolls represent a larger economic burden for low income households. A new infographic from the UCLA Institute for Transportation Studies tackles that question head on, comparing who pays and how much for a road widening depending on whether its paid for with road pricing or a sales tax increase. The poor actually pay more–and benefit less–when new roads are paid for from sales taxes. (And the USC analysis leaves out the fact that the poor are much less likely to own cars in the first place). A useful tool for talking about the equity implications of road pricing.


New Research

1.  The Changing Shape of Metropolitan America. Luke Juday of the University of Virginia has updated is invaluable dashboard of radius-based measures of metropolitan socio-economic characteristics to include the latest 2015 American Community Survey data. Juday’s work summarizes census data on educational attainment, income, age, and race and ethnicity by distance from the center of the metropolitan area for each of the 50 largest US metro areas. The dashboard shows the difference between 1990 and 2015 values for each of these variables.  Here’s the composite analysis of the 50 largest metros of educational attainment by distance from the central business district (you can also see similar charts for individual metropolitan areas, as well).

This chart shows the growing concentration of well-educated adults within 3 miles of the center of US metro areas. Adult educational attainment is higher now than in 1990, but has grown most rapidly in these close-in neighborhoods. Not surprisingly, there aren’t large differences between the 2012 and 2015 data; the two series are based on overlapping samples from the American Community Survey. This is an terrific resource for neatly summarizing the geography of change within metro areas.

2. The 2015 American Housing Survey. The Census Bureau and Department of Housing and Urban Development released the results of the 2015 American Housing Survey which provides detailed data on the housing stock and neighborhood characteristics of 25 of the nation’s largest metropolitan areas. The Census website offers a table creator that lets you generate your own cross tabulations of data for particular metropolitan areas. Variables include everything from income, rents, reasons for moving, to the size and amenities of housing units, to data on litter, abandoned properties and whether nearby houses have bars on their windows.


The Week Observed, January 27, 2017

What City Observatory did this week

1.How urban geometry creates neighborhood identity. Our colleague Daniel Hertz is back this week with an examination of the way we look at and think about neighborhood identities. He points out that in many urban neighborhoods the amount of land taken up by single family homes creates the impression that most families must live in such buildings. But because they are so space efficient and usually concentrated, multi-family homes–though less visually apparent–may  actually be home to a larger fraction of all households. This subtle perceptual bias is important because it underscores widely repeated notions of what kinds of development are and aren’t “in keeping with the character of the neighborhood.”

2. Peak millennial?  Not yet. Not soon. And not about to undermine urban revival. The New York Times published a contrarian article suggesting that the nation’s recent urban rebound might be undermined by the cresting of a demographic wave. We pushed back at City Observatory, pointing out that the number of 25 to 34 year olds in the nation’s population will continue to increase through 2024 (and plateau but not decline thereafter). More importantly, the growth in the number of young adults in urban centers has been propelled not primarily by the size of this age cohort, but instead by the growing relative affinity of young adults for urban living. Compared to the 1980s and 1990s, 25 to 34 year olds today are much more likely to live in urban neighborhoods–a trend that shows no signs of abating.

3. Suburban renewal. Its been a long time since big cities pursued heavy-handed strategies of urban renewal, demolishing whole neighborhoods–usually of modest income households, and communities of color. In Marietta, Georgia, however, that old tactic got a new lease on life when the city government used local tax revenues to buy up, then demolish hundreds of affordable apartments. In their place, the city leased out the land to Atlanta’s new major league soccer franchise to use as a practice facility. While such a policy would undoubtedly create enormous outcry in a city, it goes almost unremarked upon in the suburbs. Perhaps this reflects a deeply ingrained but seldom-voiced bias in our views about place: Suburbs are for rich, mostly white people. Cities are for poorer people and people of color. Anything change that runs counter to this worldview (like gentrification of a Brooklyn neighborhood, or efforts to build affordable apartments in suburbs like Marin County) is an affront to the order of things.

4. Does gentrification lead to increasing movement out of neighborhoods? It’s often assumed that when a neighborhood gentrifies, we see a large increase in movement out of the neighborhood by existing residents. A new study published in the Urban Affairs Review compares out-migration rates from gentrifying and non-gentrifying neighborhoods. It finds that the overall migration by renters from gentrifying neighborhoods is no higher than in non-gentrifying neighborhoods. The study does detect a small increase in the number of involuntary moves by renters, finding about 2.6 percent more involuntary moves over a two-year period in gentrifying neighborhoods.  But to put that number in context, in gentrifying and non-gentrifying neighobrhoods alike, more than half of all renters (54 percent) move out after two years. The study shows that gentrification has no effect on the rate at which homeowners move from neighborhoods.

Must read

1. How to Price Parking. Don Shoup has famously written an entire tome on how and why to price parking. But where to start?  Next City has published the tale of one company’s successful effort to apply Shoupian principles to its own parking. The story is told by Evan Goldin, who was a product manager (and parking lot manager) for transportation network company Lyft. When Lyft moved its operations to San Francisco in 2014, it decided to start charging for parking, and used the net revenues to subsidize other employee transportation choices, including transit. It took some tweaking with prices and policies, but the result has been positive for everyone: those who need and value a dedicated parking space are assured of having one, and those who don’t drive have some of their commute costs subsidized by those who drive. While the article is entitled “How I helped to change the commuting culture at Lyft,” it’s clear that the commuting changed in response to a change in prices, not because of a kumbaya transformation of the company or its employees.

2. Infrastructure “Clash of the Titans” at Brookings. Two of the nation’s leading economists–both from Harvard–former Treasury Secretary Larry Summers and Triumph of the City author Ed Glaeser, debated the economic merits of a national infrastructure investment push.  Summers stressed the macroeconomic argument in favor:  more government spending at a time of historically low interest rates is likely to help stimulate long-term economic growth. Glaeser took a more skeptical position, and stressed the microeconomic view: infrastructure projects ought to be subjected to clear cost-benefit tests, and it makes sense to insist that most projects be financed by user fees that place the costs of building such projects on those who will benefit. Pricing not only pays for projects by provides strong incentives to build only those projects that have an economic return. You can read Brookings synopsis of the event or watch their speeches on-line at the Brooking website. Also worth reading: Matt Kahn’s take on the debate: (he’s the one who came up with “Clash of the Titans.”)

3. The Crane Count. We’re always on the lookout for great urban indicators. Here’s one from the Seattle Times: the count of high rise construction cranes working in leading cities around the country. Its a simple visual indicator of how much central construction activity there is in different cities. Some 62 cranes are currently working in Seattle and 58 in Chicago. Two of the nation’s most expensive (and vertical) markets have many fewer cranes–San Francisco has 24 and New York just 21. They’re currently rivaled by much smaller cities including Portland (25), Denver (19) and Austin (20). One way we’ll cope with our shortage of cities is by building up, and the crane count is one way of tracking where that’s happening.

4.  The unintended consequences of inclusionary zoning. Seattle’s in the process of developing a set of inclusionary zoning requirements as part of its Housing Affordability and Livability Agenda (HALA). Sightline’s Dan Bertolet has been crunching the numbers to assess how these requirements will affect the profitability and probability of building new apartments in Seattle.  The news isn’t good: for many types of development, the requirements are so costly that its likely that developers won’t go ahead with projects. If fewer market rate units get built, the tenants that would have occupied them will likely end up competing for housing with low and moderate income renters, bidding up the price of existing apartments. Bertolet estimates that losing just one 250 unit apartment building would have the net effect of wiping out about half a year’s progress towards improving the city’s affordable housing supply. While well-intended, a badly designed inclusionary zoning program could easily make the city’s housing affordability problems worse.

New Research

1.  Is your smartphone cutting you off from face-to-face interactions? A new study from the University of Milan looks at the connection between smartphone adoption and use and face-to-face interactions and life satisfaction. One of the regular findings of the happiness literature–using survey data on self-reported well-being is that a person’s happiness generally increases with amount of time they spend interacting with friends. Using data from Italy’s version of the general social survey, Valentina Rotondi, Luca Stanca, and Miriam Tomasuolo show that for smart phone users, the positive effect of spending time with friends is significantly reduced.

2.  The best city websites.  If you’re like us, you’re regularly searching the web for interesting information about cities. Kyle Zheng  has compiled a useful and well-organized directory of many of the best sites worldwide. His “Master City” website is organized both by area of interest (placemaking, walkability, sustainability) and had a city-by-city listing of websites.


Louisville’s experiment in transportation economics

As we pointed out yesterday, there’s some initial visual evidence–from peak hour traffic cameras–suggesting that Louisville’s decision to toll its downtown freeway bridges but leave a parallel four-lane bridge un-tolled has produced a significant diversion of traffic away from the freeway. Perhaps without knowing it, Louisville has embarked on an interesting and useful economic experiment.

One of the big questions in transportation economics is what value people attach to travel time savings: How much is it worth to me to shave five or ten minutes off my daily commute?  There are a lot of theoretical arguments about the value, but there’s nothing quite like an actual experiment which gives people real world choices and observes the results. And that’s just what Louisville has done. If you’re traveling across the Ohio River between Jeffersonville, Indiana and Louisville, Kentucky, you have a couple of choices: you can pay between $1-$4 and drive across the shiny new multi-lane I-65 bridges on the freeway, or you can use the old US 31 route, and take the 1930s-era Second Street Bridge for free.

At least as of last Tuesday, it looked like a lot of people were choosing the “free” way, instead of the “freeway.” The following photographs were taken by Louisville’s traffic cameras at shortly after 5 o’clock local time on January 17th. The lefthand photo shows the new freeway bridges; the right hand shows the Second Street bridge.

Toll: $1-$4.Free.

“I just take Second Street,” said Tijuan Howard, who lives in Louisville. “That’s just common sense. And whatever way you need to get to Jeffersonville, you can just take the back streets. People are going to figure that out. It’s not hard.”


The choices that actual travelers like Mr. Howard make will tell us a lot about how much monetary value people attach to travel time savings.  In traffic forecasting parlance this decision comes down to the “value of time”: how much do travelers value their time on an hourly basis. For a $3 toll to justify a two minute time savings, one’s time has to be worth about $90 an hour; to justify a $4 toll for two minutes of time savings, one’s time has to be worth $120 an hour. If you face the standard $3 toll, and your time is worth $15 an hour—a standard estimate in travel time studies—you’d wouldn’t find it economically worthwhile to use the toll crossing unless it saved you about 12 minutes of travel time.

This calculus suggests that is very likely that the tolls on the I-65 bridge will prompt many motorists to pull off the freeway and use the free crossing. Of course, if a large number of travelers exit the freeway, and take the parallel Second Street route, that will produce congestion, and increase travel times for those avoiding the toll. The limiting factor on this toll-related diversion is likely to be the capacity of the Second Street Bridge (which has two lanes in each direction) and the capacity of the surface streets connecting the Second Street Bridge to I-65 on each side of the river. We would expect delays to increase on this route at peak hours.

We can estimate how long the delays on the Second Street Bridge route are likely to be, given the value of commuter time. If commuters value their time at $15 per hour, then those facing a $4 toll would be willing to put up with a about 16 minutes of delay (4/15*60 = 16) before they’d be willing to pay a toll to use the bridge; those who had to pay a $3 toll would be willing to tolerate about 12 minutes of delay. Many motorists will attach value to the convenience and certainty of the tolled route, but it would be surprising if the diversion to the Second Avenue Bridge didn’t result in delays of ten minutes or more compared to using the tolled I-65 crossing at rush hours.

At off-peak hours, the diversion rate is likely to be even higher. Because the toll doesn’t vary by time of day, I-65 users will have to pay the same toll regardless of when they cross. While it might be a comparatively good deal to pay the toll during rush hours when the Second Street Bridge is crowded, at off peak hours, the additional time penalty will be closer to the traffic-free two minute estimate provided by Google Maps.

There’s one more wrinkle here as well: The $2-$4 tolls are for cars; medium and heavy trucks pay much higher tolls–up to $10 to $12 for large trucks.  While for some kinds of deliveries, the time savings may be worth the cost of the toll, in most cases, neither shippers or truck drivers are willing to pay extra to save just a few minutes. And the same math applies to trucks as to cars. We can assume that the value of time of a truck is in the $35/hour range (reflecting the compensation of the driver and the operating cost of the truck itself). If the toll is $12, a commercial driver will find it more profitable to get off the freeway and use the Second Street Bridge rather than pay a toll, even if the trip takes as much as 15 or 20 minutes longer than the freeway. According to the project’s Supplemental Environmental Impact Statement, the Second Street Bridge carried about 22,000 vehicles per day, and was operating at about 58 percent of its capacity in 2010.

In addition, many if not most commercial drivers have no financial or operational advantage from using tolled roads. Independent truckers and many shipping companies are paid a fixed amount per load (based on distance) and as long as they meet delivery deadlines, don’t get paid any extra for time saved in transit. Effectively, an independent trucker may have to pay the toll out of his profit on the trip. Unless he’s under considerable pressure to make a delivery deadline, he may prefer to spend an extra few minutes taking the free route, rather than pay the toll out of his own pocket. A study prepared for the Transportation Research Board concluded:

“ . . . truck drivers stated an extremely low willingness to pay even a token toll for different time savings scenarios . . . a large cross section of the trucking business cannot monetize toll road benefits.”

Also, unlike car traffic, which is highly concentrated during morning and evening rush hours (and in the case of the I-65 bridge, very much a morning-southbound, evening-northbound peak), truck traffic is much more evenly spread throughout the day. Proportionately more trucks will be crossing the river when the Second Street Bridge is less congested, and therefore will be a more attractive route.

There are more than a few ironies to this situation: First, after spending a billion and a half dollars on the new Lincoln Bridge, and doubling the number of freeway lanes crossing the river on I-65, the new bridge will likely carry fewer vehicles for the foreseeable future (through at least 2030) that it did in 2005. Second, a project avowedly designed to reduce congestion will actually lead to regular congestion of the Second Street Bridge—which is expected to see an 20 percent increase in traffic according to the project’s own estimates (but we believe this figure is probably a substantial under-estimate). What toll diversion—and the permanently depressed level of traffic predicted for the I-65 bridges signals is that a significant fraction of bridge users don’t value the time savings provided by the project to pay for them—even though tolls cover less than half of the cost of the bridge improvement project. In short, this is clear economic evidence that the project isn’t economically warranted.

By tolling some of the bridges across the Ohio, and leaving others toll-free, Indiana and Kentucky are conducting a real-world behavioral economics experiment. Over the next few months we’ll be watching to see how travelers respond to the financial incentives they’ve been provided, and how their travel behavior shifts in response. The results will be interesting.

Who pays the price of inclusionary zoning?

Requiring inclusionary housing seems free, but could mean less money for schools and local services

Last month, the Portland City Council voted 5-0 to adopt a sweeping new inclusionary housing requirement for new apartment buildings. The unanimous decision came with the usual round of self-congratulatory comments about how they were doing something to address the city’s housing affordability problem. No one mentioned that they were also in effect voting to cut funding for schools or other local government services. But in Portland’s case, that’s exactly what the inclusionary program is likely to do, according to the city’s own budget office.

On the surface, one of the compelling policy attractions of inclusionary zoning (“IZ”)is that it doesn’t seem to cost any money:  You require developers to build 1 or 2 affordable housing units for every ten new apartments that they build.  Maybe your city offers up a density bonus, or expedites permit handling, but unlike conventional public housing, the city doesn’t have to lay out any of its cash to get more new affordable housing. That’s why Evan Roberts of StreetsMN described it as politically understandable, though terrible policy.

Portland new ordinance is one of the nation’s most demanding inclusionary housing requirements.  Basically, the city will require that all new apartment buildings of 20 or more units set aside 20 percent of their units for renters with no more than 80 percent of the region’s median household income (about $56,000).  Alternatively, developers could set aside 10 percent of their units for households earning less than 60 percent of the region’s multifamily housing. (As we’ve noted at City Observatory, most cities have far lower inclusionary requirements, offer exceptions, or only apply the requirement to newly up-zoned properties or those projects receiving city subsidies).

The city’s plan includes the usual list of non-cash aid to developers–lighter parking requirements, faster permit processing and density bonuses–although there’s considerable dispute as to whether these effectively allow developers to build more than they would have otherwise.

All of the analyses of the city’s inclusionary zoning plan concluded that unless the city offset the cost to developers, fewer units would get built. As a result, in addition to regulatory concessions, the city is also assuming that new apartment buildings would also get subsidized via a property tax exemption. And, as it turns out, this is where inclusionary housing starts get to costly for the public sector.

Portland’s plan offers up two levels of property tax exemption. For most new apartment buildings, the property tax exemption would apply only to the affordable housing units. For developments with a FAR of 5.0 or more (meaning for example, that a developer is building a 50,000 square foot or larger building on a 10,000 square foot lot), which would typically be an apartment tower, developers would get a property tax exemption for all of the apartments in the building.)

The amount of revenue foregone due to the tax exemptions is difficult to estimate.  It depends on whether developers choose to set-aside 10 percent of their units for families at 60 percent of median income or 20 percent of their units for families at 80 percent of median income. It also depends on how many units are actually built.  Even with property tax breaks and other incentives, many developers argue that it will no longer be financially attractive to build new apartments in Portland.  The City Budget Office has developed estimates of lost property revenue from the inclusionary housing program based on the assumption that developers will mostly go the 10/60 route (which minimizes their construction costs and gets them the largest property tax benefit per affordable unit). They also assume that the IZ program doesn’t impair housing construction–that the city builds as many units as its Comprehensive Plan calls for between now and 2035.  Under these assumptions the IZ program will cost the city $15.8 million in tax and fee revenue per year.

The total cost of the property tax breaks and city fee waivers per unit of affordable housing ranges from about $21,000 to almost $220,000 per unit.  But the cost to the City of Portland is far less than this amount, because most of the foregone property tax revenue is lost to other local property taxing entities, including K-12 schools, community colleges, Multnomah County and a handful of other local governments.  In essence, the City Council voted to have these other taxing entities pay about three-fourths of the fiscal costs associated with inclusionary housing, with the result that these other governments will have less revenue to pay for schools and other local services.  Here’s the takeaway quote from the City Budget Office:

CBO Analysis: The proposed policy would result in an estimated per-affordable unit cost that ranges from $20,787/unit to $218,663/unit, depending on project location and incentive package selected. The cost to the City General Fund is less – ranging from $4,674/unit to $57,529/unit – due to the property tax exemption costs being spread across schools, the County and other local public agencies.

In essense, the City has voted for other taxing jurisdictions to pay three-quarters of the public fiscal cost associated with inclusionary zoning.

Whether or not developers will get this volume of tax exemptions is still in doubt. The city doesn’t have exclusive jurisdiction over property tax exemptions, and by mutual agreement with Multnomah County, the city has agreed to abide by a $3 million annual cap on revenue lost to property tax exemptions.  The City and County will face a major dilemma in the months ahead as the program goes in to effect. If the $3 million cap isn’t lifted, there won’t be the necessary subsidies to make the inclusionary zoning program attractive enough for developers, and the city’s housing supply will suffer. If the $3 million cap is lifted, the city–plus the county, schools, and other local governments-will have a significant revenue shortfall to make up.

The time-tested adage of economists is “There’s no such thing as a free lunch.”  And when it comes to inclusionary zoning we might well add:  “There’s no such thing as free affordable housing.”

Has Louisville figured out how to eliminate traffic congestion?

Louisville is in the transportation world spotlight just now.  It has formally opened two big new freeway bridges across the Ohio River, and also rebuilt its famous (or infamous) “spaghetti junction” interchange in downtown Louisville. A story at Vox excoriated the decision to rebuild the interchange rather than tear out the riverfront freeway as a “a testament to how cars degrade cities.” In our first story about the Louisville area’s new $3 billion bridge project, we described how a bizarre toll structure will actually encourage wasteful driving, and probably lead to periodic congestion

But there’s another problem with the Louisville bridges: Only half the highway crossings are tolled. The main, north-south I-65 crossing, now encompassing twelve traffic lanes (split between the new Abraham Lincoln Bridge and a renovated Kennedy Bridge) will be tolled as will a new “East End Crossing,” about eight miles to the north. But two other highway bridges, the Sherman Minton bridge carrying I-64 from the West, and the venerable 1920s era Clark Memorial Bridge (usually referred to as the Second Street Bridge) continues to be free.

Louisville’s Clark Memorial aka Second Street Bridge (Wikipedia)

The other tolled route is the East End Bridge new crossing, far from existing routes. But the I-65 crossing–carried on the parallel Kennedy and Lincoln bridges–is just a few hundred feet upriver from the Second Street Bridge. While regular commuters will pay just a dollar a trip to use I-65 (provided they cross it at least forty times a month), occasional users will pay three or four dollars each way if they don’t sign up for a transponder, and instead, rely on toll-operator RiverLink to capture an image of their license plate.

According to Google Maps, taking the Second Street Bridge rather than the I-65 bridge adds about two minutes to a car trip from Jeffersonville Indiana, across the river to Louisville. So the question many motorists will face is: is it worth $3 or $4 to save two minutes on their trip.


The big question here is:  How many commuters will take the Second Street Bridge to avoid paying tolls on the new Lincoln and Kennedy I-65 bridges? Of course, it will be months–or longer–before we get the detailed and definitive traffic counts that will enable us to answer that question, but given our curiosity about the results of this economic experiment, we decided to take a quick peak at the traffic cams for the two facilities. This is of course an inexact and unscientific comparison, but the results are interesting.

In Louisville, Kentucky’s Transportation Cabinet operates a website called “TRIMARC“–”Traffic Response and Incident Management Assisting the River City”–which has a network of traffic cameras monitoring the region’s principal arterials and freeways. On January 17, shortly after 5pm we captured images of the traffic on the Second Street Bridge and the Kennedy and Lincoln Bridges from this website. Tuesday, January 17th was an ordinary business day, following Monday’s Martin Luther King Day holiday, and the weather was clear and dry, with temperatures in the mid-50s. Tolls had been in effect on the bridges for almost three weeks–since December 30.

First, here’s an image of traffic just North of the Kennedy and Lincoln I-65 bridges at milepost 0.1 in Indiana. This view is from Camera 3869, looking North toward Indiana, with the Southbound travel lanes (entering Louisville) on the viewer’s left and and the Northbound travel lanes (leaving Louisville) on the viewer’s right. The metal lattice structure stretching across the freeway in the foreground is the gantry holding the bridge’s toll collection cameras and sensors. This picture was taken at 5:07 pm EST.

I-65 Bridges at 5:07PM January 17 (TRIMARC Camera 3869).

Each of these bridges is striped for  five lanes of traffic at this point. As you can see there traffic is extremely light. (For you highway buffs, this is Level of Service “A”).

Next, here’s an image of traffic coming off the Second Avenue Bridge as it enters downtown Louisville. The vehicles on the left are Southbound into the city, the vehicles on the right are Northbound leaving the city. This picture was captured at 5:05 pm EST.

Second Street Bridge at 5:05PM, January 17, (TRIMARC Camera 065)

The Second Street Bridge has four travel lanes–two in each direction.  As the two Southbound lanes enter downtown Louisville (on the left of this image) they branch into two right turn lanes and two through lanes.  Although this camera is set at a much lower height, and therefore shows a shorter segment of roadway than the I-65 camera above, there are clearly more vehicles shown crossing the Second Street Bridge than are crossing the I-65 bridge at nearly the same time. (The truck in the center of the picture is saving $12 compared to the cost of driving over the nearby I-65 bridge).

This very anecdotal, if visual, information suggests a couple of hypotheses.  First, it does appear that a fairly large segment of traffic crossing the Ohio River in Louisville on this particular afternoon chose the older, slower and non-tolled route over the newer, faster and more expensive tolled freeway bridges. Second, it seems like there is plenty of capacity crossing the Ohio River at this particular point to accommodate all these vehicles. Not only is the freeway nearly deserted, but traffic appears to be well below capacity on the Second Street Bridge as well.

Of course, this is a small sample and a highly unscientific set of observations. But taken at face value, these pictures call into question the decision to spend several billion dollars to increase highway capacity over the Ohio River. If so few cars are actually crossing the river at peak hour on a typical business day, and if such a relatively large proportion of them apparently prefer to use the older, slower route rather than pay to use the fancy new crossing, did the states of Kentucky and Indiana have any reason to spend so much to build a giant new bridge?  And will the two states, who are counting on toll revenues to pay back a major share of the cost of the project, be able to cover their debts?

There may be some very good news here. The image of I-65 is compelling evidence of how to alleviate, if not completely eliminate peak hour traffic congestion:  Charge a toll. When faced with a positive price for driving (the toll is $1 each way for regular commuters) apparently very few people want to drive on the freeway. As we’ve long maintained at City Observatory, traffic congestion is a direct consequence of charging too low a price to road users. Kentucky and Indiana have apparently demonstrated that a very modest level of tolling can work wonders for alleviating traffic congestion. It’s just too bad that they also had to spend several billion dollars for highway capacity that motorists apparently don’t want in order to find that out.



Race & transportation: Still a long way to go

January 17 is the day we celebrate the life and dream of Dr. Martin Luther King, Jr.  This year is also the first year that we’re observing a national day of racial healing. We thought we’d take a minute to reflect on two recent studies that provide some strong statistical evidence for the unfortunate persistence of racial discrimination in transportation: “Driving while Black,” and discrimination in taxi hailing.

Driving While Black

“DWB.”  Driving while black.  Its a sad and scary fact of life in many communities in the United States.  African-Americans are much more likely than other drivers to be pulled over for routine traffic infractions.  Federal data on traffic stops show a clear difference by race: Black drivers are about 31 percent more likely to be pulled over for a traffic infraction than white drivers.

If you’re white, it might not be something you notice.  In theory, it might be possible to discount this, if there were evidence that African-Americans were somehow more reckless than other drivers (not that we think that’s a plausible explanation).  But a new analysis of travel data suggests that if anything, African-American drivers are significantly more cautious than others (an observation that is fully consistent with DWB).

The evidence comes–almost in passing–in a new paper on  inter-metropolitan differences in transportation system performance entitled “Speed.”  Written by economists Matthew Couture, Gilles Duranton, and Matthew Turner, it presents a  systematic set of estimates of travel speeds in different metro areas. One of the complicating factors of speed estimation is that speeds vary by length of trip, time of day, and trip purpose.  In general, shorter trips involve lower speed travel. That makes sense:  if you’re just traveling a mile or two, especially between your home and some other destination, its likely you’ll travel mostly on local streets, with stop signs and traffic signals. But for longer trips, it makes more sense, even if its not the shortest distance, to travel part way on higher speed arterials or limited access freeways. Couture and his co-authors use detailed micro-data on trip taking in the National Household Transportation Survey to estimate variations in speed across metropolitan areas, after controlling for differences in trip distances and other demographic factors.

While its almost a footnote to the study, the author’s report the variations in driving speed according to a range of demographic characteristics.  Some of their findings correspond with our stereotypes about driving behavior.  Men drive faster than women.  Older people drive more slowly.  And as they note, race has a significant relationship with average driving speeds.  From their report:

Women are about 0.5% slower than men. Age is more important. A year of age is associated with 0.3% slower speed. Black drivers drive about 8% slower.

Its interesting to compare the differences in driving behavior by age and race.  If each year of age reduces your average speed by about three-tenths of one percent, that means that the typical black driver travels at about the same speed as a white driver who is about 26 years older.  For example, a 25 year-old black driver would, according to these estimates, be expected to drive about as fast as a 51 year-old white driver.

These results are powerful evidence that DWB is real, that it affects driving behavior, and that the disparity in traffic stops is probably even greater than indicated by the gross statistics.  Black drivers travel at speeds that are essentially the same as drivers that are a generation or more older than they.  And if black drivers are going much more slowly, it seems like they ought to be less likely, all other things equal, to commit traffic infractions, than other drivers.

Discrimination in Taxi Hailing

Its a classic story of discrimination. A yellow taxi drives past an African-American whose arm is outstretched to hail a cab, and instead pulls in a few dozen feet further down the block to pick up a white fare. In theory, at least one of the advantages of transportation network companies–the computer-based ride-hailing services such as Lyft and Uber is that they should be more race blind. Indeed, there’ some evidence that these services have expanded for-hire transportation to neighborhoods that have often been under-served by traditional taxis.

But despite the technological advances, it appears that racial discrimination still persists. A new paper “Racial and Gender Discrimination in Transportation Network Companies,” from economists Yanbo Ge, Christopher R. Knittel, Don MacKenzie, Stephen Zoepf looks at discrimination by ride sharing services. The researchers constructed an extensive experiment using transportation network companies in Seattle and Boston. They sent a team of testers including men and women, Blacks and whites, to take about 1,500 rides in each city. They found found that drivers for ride-sharing services are prone to discriminate against African Americans, making blacks wait longer for rides when they can identify the race of the ride-hailer and more frequently cancelling rides when alerted to African American-sounding names.

Photo: WNYC

In Seattle, blacks waited an average of 35 percent longer than whites. In Boston, they found that customers who had black-sounding names were about twice as likely to have their ride requests cancelled, compared to customers with white sounding names. There was also evidence of discrimination by gender as well: women tended to be taken on more circuitous–i.e. more expensive and profitable–routes to their destinations.

Conversations about race and discrimination are frequently made more difficult by the very different lived experiences of blacks and whites. For those who haven’t been discriminated against in a traffic stop or in trying to hail a taxi because of the paler shade of their skin, its hard to understand whether these experiences are real and commonplace. The data presented in these two studies is a quantitative reminder that it is still quite real.

Housing supply is catching up to demand

As Noah Smith observed, economists invariably encounter monumental resistance to the proposition that increasing housing supply will do anything meaningful to address the problem of rising rents–especially because new units are so costly. One of the frustrations that we (and increasingly cost-burdened) renters share is the “temporal mismatch” between supply and demand.  Demand can change quickly, while supply responds only slowly, thanks to the long time it takes to detect a market need, plan for, permit, finance and then finally build new apartments. Rents go up quickly, and economists can only counsel patience while this process is unfolding.

But it’s increasingly apparent that housing supply is now responding. Several cities are recording impressive increases in the number of new apartments permitted and under construction. Take Seattle, which saw rents increase nearly 10 percent in 2015. Like the rest of the country, the city saw a fall off in new construction after 2007 with the advent of the Great Recession. But since then, developers and investors have been pouring money into the local housing market. Nearly 10,000 new apartments are expected to be completed in just the next year. As the Seattle Times highlighted, more apartments will open this decade in Seattle than in the previous half century.

Source: Seattle Times

Even though these apartments haven’t been completed yet, the growing supply, and the prospect of more units coming on line is already having an effect on prices. According to the Seattle Times, one local market analyst is saying apartment rents have reached a turning point; rents in some of that city’s hottest neighborhoods have declined between 3 and 4 percent in the past year.

Just about three hours south, down Interstate 5, Portland’s housing market has seen a similar cycle, with events lagging about six to a year or more behind Seattle.  Just a year ago, Portland had the highest rent increases in the nation–12.4 percent year-over-year in October 2015–according to real estate firm Axiometrics. But since then builders have been moving aggressively to add more apartments. The city has permitted between 3,000 and 4,000 apartments in each of the last two years.  According to local market analysts Barry and Associates, 7,000 units are currently under construction and and additional 15,000 are in the planning stages. In Portland, as in other cities, supply is responding, in a big way, to past rent increases.

To be sure, the supply response hasn’t actually produced declines in rent levels–yet. But as more and more new units come on line, and their owners seek tenants, the balance of power in the rental marketplace will shift from sellers to buyers.

Rising rents have generated a tangible response.

The recent sharp growth in supply comes, ironically, just as the City of Portland has enacted one of the nation’s most demanding inclusionary zoning requirements. Developers building 20 or more apartments will be required to set aside 20 percent of their units for households earning no more than 80 percent of the region’s median household income. (Indeed, a big part of the reason for the recent surge of applications to build new apartments has been developers seeking to obtain building permission prior to the February 1, 2017 effective date of the new regulations). Almost exactly the same thing happened in San Francisco, as developers scrambled to beat the deadline for that city’s more stringent inclusionary zoning requirements.

There’s a growing body of evidence that as more housing gets built, rent inflation moderates, and rents even decline. Nationally, Axiometrics sees a slow down in rental growth, with actual declines in some markets. In Houston, New York and the San Francisco Bay, growing inventories have produced “negative rent growth.” Last month the Chicago Tribune reported that while rents were up about 2.3 percent citywide, they were down about 0.3 percent inside “The Loop.”  The New York Times even floated the notion that 2017 will be “The Year of the Renter” as growing supply takes the edge of rent increases in New York.

For the moment, then, the good news is that in these coastal cities, where rents have been rising at alarming rates, that the forces of supply and demand are operating. As more apartments are completed–even at the high end of the market–the added supply is dampening the rate of rent increases. The key question going forward is whether the demanding inclusionary requirements enacted in places like Portland and San Francisco–and which are still pending in Seattle–will prompt developers to pull back from the current frenzied pace of building.

Nothing’s worse than funky beer, except funky beer data

You know the feeling: you’re thirsty, you’re primed for a cool, refreshing beverage, and the anticipation has your taste buds tingling. But you pop the cap on the bottle only to find that the beer has turned skunky.  It’s very disappointing.  Well, we had a small taste of funky beer a couple of week back at City Observatory. Today, we want to wipe that taste out of our mouths, and present some other data that helps triangulate the geography of US microbreweries. First, a look back.

(Flickr: Base Camp Baker)

Many of you read our post ranking the density of microbreweries in the nation’s 50 largest metropolitan areas. To construct this index, we relied on a GIS tool constructed by (which allows us to do radius searches for microbreweries). The tool is, in turn, built on top of a database constructed by breweryDB. And while it appears to be one of the most comprehensive sources of detailed information on microbreweries, it is far from perfect. A number of CityObservatory readers, looking at our rankings, and then clicking through to their local cities, found that listings in many cases included breweries that were closed, mis-located and in a few rare cases non-existent (maybe not yet open).

The underlying data for BreweryMap comes from crowd-sourced data. BreweryDB makes it clear on their website:

The team at is aware that the database is less than perfect. They invite users to send corrections to breweryDB:

BreweryMap uses BreweryDB to provide information about the beers and breweries. If you find a brewery that is missing or incorrect, please click the edit button on the brewery info window to go to and add or correct info. The world will appreciate your contribution and you’ll get a nice warm fuzzy feeling in your heart!

State level brewery data

While there’s something deeply democratic about crowd-sourced data, it inevitably comes with a degree of noise and a margin of error that is larger than we find in more professionally curated and quality-checked data. Its always a good idea to cross-check crowd-sourced data against other independent information sources (which is why we dug up some market survey data on microbrewed beer consumption by media market). Another way to triangulate our findings is by looking at data from the Treasury Department’s Alcohol and Tobacco Trade and Taxation Bureau. Breweries have to be licensed by the federal government–they pay a tax on every barrel brewed–and this produces a reliable and comprehensive count of breweries. Unfortunately, the published aggregations of this data re available from ATTTB only at the state level. While these data may be more consistent than crowd-sourced counts, that doesn’t mean they’re perfect: it’s likely that some breweries that closed during a year are still in the database, and some licensees may be counted even though they haven’t yet brewed beer; others may be temporarily inactive.

The federal data show a striking increase in the number of breweries in the US, especially in the past eight years or so. While the microbrewery revolution traces its roots to the 1980s, there was a period of fast growth in the late 1990s (as the total number of breweries rose from just a few hundred to nearly 2,000). That number plateaued for most of the decade of the 2000s, but since 2009, has taken off again. The number of breweries in the US has tripled, to nearly 7,000.

This growth has been widespread. Today, every state has at least a dozen breweries. Overall, California has the most breweries (927) and the District of Columbia, Mississippi, and North Dakota have the fewest (13, 14, and 15 respectively). If we normalize these counts by population, we can identify the states where breweries are most common. The following chart uses the state level data prepared by the ATTTB to compute the number of microbreweries per capita in the US.  Vermont leads the list, with more than 116 breweries per million population, followed by Maine, Montana, and Oregon, which each have more than 70 breweries per million population. The typical state has about 20 breweries per million population.

When we map these data, some clear regional patterns emerge. In the following map, darker colors correspond to higher number of breweries per capita, and lighter colors correspond to fewer breweries per capita. Breweries are most common in the Pacific Northwest and in the Northeast, and are least likely to be found in the middle South. (You can scroll over each state to see the number of breweries and our count of breweries per million population.)

Its great, of course, to see the flourishing and spread of local craft brewing. At City Observatory, we really enjoy tracking these numbers (despite their limitations). But like you, we suspect there’s one thing we enjoy more, and it comes in a pint glass. Prost!

Pollyanna’s ride-sharing breakthrough

A new study says ride-sharing apps cut cut traffic 85 percent. We’re skeptical

We’ve developed a calloused disregard for the uncritical techno-optimism that surrounds most media stories about self-driving cars and how fleets of shared-ride vehicles will neatly solve all of our urban transportation problems.

But a new story last week re-kindled our annoyance, because it so neatly captures three distinct fallacies that suggest that fleets of shared autonomous vehicles, if actually deployed in the real world, would produce a dramatically different outcome than the one imagined.  The story in question last week described a new study which reportedly proved that ride-hailing apps could reduce traffic congestion by 85 percent. The headline at Mashable was typical:

And don’t blame the writers at Mashable. Their interpretation closely mirrors a florid press release produced by MIT, which summarized the study as follows: “One way to improve traffic is through ride-sharing – and a new MIT study suggests that using carpooling options from companies like Uber and Lyft could reduce the number of vehicles on the road 75 percent without significantly impacting travel time.”

These can be replaced. Mathematics proves it.

The press stories were based on a new paper entitled “On-demand high-capacity ride-sharing via dynamic trip-vehicle assignment,” written by Javier Alonso-Mora, Samitha Samaranayake, Alex Wallar, Emilio Frazzoli and Danila Rus, a team of engineers and computer scientists from MIT and Cornell. Their key finding:

Our results show that 2,000 vehicles (15% of the taxi fleet) of capacity 10 or 3,000 of capacity 4 can serve 98% of the demand within a mean waiting time of 2.8 min and mean trip delay of 3.5 min.

Despite the press release announcing the study–entitled “Study: Carpooling apps could reduce traffic 75 percent”–the study isn’t referring to all traffic, its just referring to New York City’s existing taxicab fleet. What the study is really saying is that 3,000 ten passenger vehicles with a sophisticated real time ride-hailing and vehicle assignment system could provide just as many trips as 14,000 yellow taxis, many of  which are idle or simply cruising Manhattan looking for one- or two-person fares headed to a single destination. Some of the gain comes from higher occupancy (up to 10 passengers, rather than mostly a single passenger), and other of the gain comes from better dispatching (fewer miles driven with an empty vehicle). So far, so good. But going from this observation to the conclusion that this will solve our urban congestion problems quickly runs afoul of three major fallacies.

The fixed demand fallacy

Autonomous vehicle designers may be using LIDAR, imaging, vehicle-to-vehicle communication and prodigious computing power, but down deep they’re still engineers, and they’ve apparently given no thought whatsoever to induced demand.  Just as highway engineers have assumed that there’s a fixed demand for travel and that highways need to be sized accordingly, and ignored the effect of new capacity on in stimulating added travel, the MIT study assumed that the current level of taxi use exactly captures future travel demand. Its worth noting that the demand for taxis is limited, in large part, because New York City has long regulated the number of licensed cabs via its medallion system.

There’s no reason to believe the demand for 10- and 4 passenger vehicles would be restricted to just those who currently patronize cabs. Taxis handle about 360,000 rides in the Manhattan daily. About 2.8 million travel to or from Manhattan by public transit.  If their were suddenly a viable on-street ride sharing option–especially if it were cheaper–the system could have much more demand–which could swamp the congestion reducing benefits.

The big urban transportation challenge is not simply optimizing a pre-determined set of trips, its coping with the complex feedback loops that produce a fundamental law of road congestion. This study glosses over that inconvenient truth.

The big data fallacy

A big part of what propels the illusion of fixed demand is our second fallacy:  big data.  Thanks to GIS systems in taxis, cheap telecommunications, and abundant computing power, we now live in a world where we can easily access copious data on the origin and destination of everyone of the several million annual taxi trips in New York City.  While the data is massive, it isn’t infallible or immutable: it simply reflects that decisions that travelers made with a particular technology (taxis), a particular set of prices and a set of land uses and congestions levels and alternatives that were in place at the time. It may be richly detailed, but it’s dumb: it tells us nothing about how people would behave in a different set of circumstances, with different technology and different prices.  And as big as the dataset is that’s used here, it leaves out the overwhelming majority of travelers and trips in New York who travel by train, bus, bike and foot. As we’ve suggested at City Observatory, the presence of highly selective forms of “big data” is a classic “drunk under the streetlamp” problem that focuses our attention on a few selected forms of travel, to the detriment of others. Optimizing the travel system for a relatively small segment of the population–the one for which we have rich data–doesn’t prove that this will work in the real world.

The mathematical model fallacy

The third fallacy is the mathematical model fallacy. A mathematical model can be a useful tool for sussing out the scale of problems. But in this case, it involves abstracting from and greatly simplifying the nature of the system at work. Yet press accounts, like those at Mashable are awestruck by the author’s use of a mathematical model:

This study, released Monday, used a mathematical model to figure out exactly how vehicles could best meet demand through ride-sharing.


“There’s a mathematical model for the autonomous future.”

The authors have constructed a very sophisticated route-setting and ride matching algorithm (taking that big data on origins and destinations as a starting point) and figured out how many 10 person person and how many 2 person vehicles it might take to handle all the 14 million trips. This requires formidable math skills, to be sure. And let’s take nothing away from the authors’ technical prowess: they’ve figured out how to solve a very complicated and huge math problem almost in real time. But simply using math to model how this might work doesn’t prove that people would actually use such a system.

Consider another possibility. Suppose we banned private cars and taxis in Manhattan and increased the number and frequency of city buses five-fold.  A bus would go by every bus stop every two or three minutes. And bus travel times would be faster, because there’d be no private cars on the road. We could probably carry all those taxi trips with even fewer vehicles.  One could even construct a mathematical model to evaluate the efficiency of this system. If we extrapolate from the MIT paper, it seems likely that if  2000, 10-person vehicles or 3,000 4-person vehicles cars can eliminate 85  percent of the trips percent of the trips, then its probably not a stretch to suggest that maybe 1,000 40-person buses could do the same thing.

Would the dogs actually eat this dog food?

Its worth asking a simple real world question: Would all or many of the people currently traveling in taxis agree to travel in ten-person or even four-person shared vehicles? If all that people were paying taxi fares for was travel between points A and B, maybe so. But there are many reasons to think they wouldn’t. For starters, travelers in pooled vehicles  get a slower trip. On average their pooled trip is going to take three and half minutes longer according to the MIT study. And people taking taxis are often paying for much more than just getting from A to B. In addition to getting to their destination as quickly as possible, they may also want the amenities of a dedicated vehicle, such as privacy–they don’t want to share their ride with other persons (even if it costs less). As David King notes at CityLab, taxis are a premium service in the world of urban transportation. And especially, what riders may be paying for is greater certainty that they are getting reliable and high priority service. The authors report the average wait time for a shared vehicle would be 2.8 minutes and the average added travel time would be 3.5 minutes, but some riders would face longer waits and greater delays. And avoiding that uncertainty or variability is a big part of why people pay for a solo taxi. And finally, it may be about status–being driven as the only occupant in a car. Often times, its faster to travel between most points in Manhattan by subway, yet many people take a taxi or uber because of just these other considerations such as comfort, convenience, priority, privacy and status.

Here’s an analogy: If Per Se or Momofuku Ko (two of New York’s swankier restaurants) served all of their food cafeteria style, they could dispense with waiters and serve 200 percent more diners than they do today. A good MBA student could probably produce a pro forma that could calculate, to the penny, how much more profit the owners would make. But it’s likely that high-end restaurant patrons–like taxi customers–are paying for something more than just the basics, and that in the real world, this wouldn’t attract many customers.

Its unfortunately too easy to oversimplify the nature of the urban transportation problem. We tend to be beguiled by new technology and blinded by big data in ways that lead us to overlook some fundamental questions about, for example, geometry. As we look to implement new technologies, like autonomous vehicles and shared ride services, we need to remember some of the hard earned lessons about things like induced demand.


Brownstone Brooklyn and the challenges of urban change

In the middle of The Invention of Brownstone Brooklyn—a book published in 2011, but no less relevant today—Suleiman Osman turns the tables on the people who have long been the heroes of urbanist lore.

Speaking of the insurgent middle-class professionals who, starting in the 1950s and 60s, began to organize to stop to the massive urban renewal projects of Robert Moses and others, Osman says: “Perhaps the anti-Moses movement deserves an inverted version of the charge thrown at modern planners. If Brooklyn’s new white-collar professionals loved people, they hated the public.”

What does it mean to hate the public? And how could the people who stood up to clearance for highways be villains?

The illusive New York apartment (Flickr: Sharona Gott)

Brownstone Brooklyn tells the story of the first generation of gentrifiers in the rowhouse neighborhoods across the East River from Lower Manhattan. For Osman, it is an essentially countercultural movement. The “brownstoners” struck out for Brooklyn because the Upper East Side and Greenwich Village were getting too expensive, sure. But they also rejected those places—and the people who lived there—as part of a corrupt modern society, caught up in a deracinated, atomized culture increasingly dominated by the impersonal global market. Brooklyn was a place where you could live an authentic life in a “real neighborhood”: one deeply connected to local history (even if, as Osman writes, much of that history had to be rediscovered or simply made up), and where commerce meant a corner bodega owned by a family on the next street instead of a national chain.

But though some brownstoners may have claimed a kind of grassroots populism, from the very beginning they had a tense relationship with the working class Italians and Puerto Ricans they encountered in Brooklyn Heights and Park Slope. On the one hand, their presence created the kind of “diversity” and “authentic” characters that made Brooklyn different from Manhattan to begin with.

On the other, brownstoners’ project of neighborhood renewal often meant restoring townhomes that had been broken up into boarding houses to their original single-family conditions, evicting several low-income households in the process. Many of the people who thought of themselves as refugees from the harsh real estate market of Manhattan may have been anguished to realize they were landlords and villains in Brooklyn. But the fact that their movement was commonly referred to as “unslumming” shows that a kind of economic restructuring was integral, not incidental, to the project.

Moreover, the move from Greenwich Village to Boerum Hill was an escape from larger market forces only in a very incomplete way. After all, most of the new arrivals still commuted back to white collar jobs in Manhattan; a 1971 survey of brownstone owners found they were 99 percent white, 60 percent held graduate degrees, and 98.7 percent had earnings in the top fifth of New York City households. As they brought their purchasing power to Brooklyn, they more deeply implicated their neighbors in the very markets they had supposedly been trying to flee. And in some cases—as in their long campaign to convince skittish banks to make loans to support the purchase and rehabbing of “obsolete” 19th century townhomes—they explicitly lobbied to bring to bear the massive power of Manhattan capital to transform their new neighborhoods.

At this point, brownstoning might have reevaluated what kind of movement it was. As this process pushed the frontier of “authenticity” (and affordability) further south, they might have realized that “modernity” was not something that could be left on the other side of the East River; that their idyllic urban villages were in fact sites of increasing competition for housing and cultural expression.

Instead, in Osman’s telling—and in passages that will ring true to many observers of local politics today—the movement doubled down on a politics of authenticity and local purity. Sometimes this politics had a progressive gloss, as when they opposed the mass displacement of locals for urban renewal clearance. But that reading became harder to sustain when they opposed large middle- and low-income housing developments in an area with rapidly increasing housing prices, or when they lobbied against school desegregation. The common thread here is a commitment to local control as a way of preserving cultural distance, and local authenticity, from mass society—as represented by both freeways and public housing.

And this is where, finally, we get to what it means to love people but hate the public. Brownstoning was a movement that romanticized the participants in Jane Jacobs’ “street ballet,” but which was committed to holding the city as a whole at arm’s length. Ironically—and unfortunately—this way of relating to urban life is the progenitor of much of contemporary gentrification and anti-gentrification thinking. When Osman quotes the very same people speaking of their roles both as “frontiersmen” transforming “wild,” “undiscovered” Brooklyn and as defenders of these wilds from the rapacious, homogenizing forces of capitalism and government-led development, he could just as easily be capturing Crown Heights in 2016 as Brooklyn Heights in 1959.

That is, the people who are most invested in building their ideal environment—which almost always means reshaping an existing community to new specifications—are also the most sensitive to new forces whose influence threatens their own. That’s not just an observation about gentrification: it also applies to the family that bulldozes a farm to build a dream home on the edge of the city, only to complain bitterly when someone else ruins their pastoral view when they bulldoze the next farm over. But it certainly applies to the people who invented Brownstone Brooklyn, “reclaiming” a very particular (and often ahistorical) historical identity, and then just a few years later fought furiously to retain their particular interpretation of the neighborhood against the next wave of newcomers. Maybe debates about gentrification feel so circular and unproductive because what presents as two opposing sides is in fact a snake’s head and its tail.

There are two mistakes holding each other up here. The first is defining “urbanism” primarily as a countercultural project, which inevitably means rejecting any number of “incompatible” cultures and the people who identify with them. In some cases, this looks like brownstoners telling their working class neighbors that the longstanding practice of modifying their homes had to end in deference to professional class ideas about historic preservation. At other times, it looks like a 1970s anti-Burger King campaign that alternated left critiques of mass consumer culture with warnings about people who eat fast food being “purse-snatchers, muggers, and criminals.”

The second, related mistake is retreating to neighborhood-level politics, rather than broader political movements to reshape the city. Osman argues convincingly that the brownstone movement, in both its pro- and anti-gentrification forms, contributed to the breakup of the New Deal coalition, as much of the “new middle class” decided that unions and working class political machines were part of the establishment against which they were rebelling.

Of course, many of their critiques—and those of the unlikely sometimes-allies they found in radical black and Puerto Rican political organizations—were correct and necessary. Modernist planning really was profoundly indifferent to the “collateral damage” of its great public works, and the New Deal coalition really was profoundly influenced by racism.

But the new localist order they helped to create didn’t shed these problems. By closing down boarding houses and preventing the construction of new housing, both market and subsidized, the brownstoners’ “renewal” ultimately created neighborhoods nearly as homogenous as if they had been swept clean by Robert Moses’ bulldozers. (In fact, those modernist campuses are now some of the more demographically diverse parts of New York’s wealthy areas. The Cadman Plaza development, which was furiously fought by new Brooklyn Heights residents in the 1960s, sits in a Census tract that is 64 percent white; across the street, a Census tract with pristinely preserved brownstones is 82 percent white.) And in reinforcing the local power of those increasingly homogenous communities, they created durable coalitions against change, whether in the form of new housing or school integration.

The question Brownstone Brooklyn raises is whether this movement might have taken a different path. Could they have joined an urbanist movement whose first principles were not about a personal search for authenticity, but rather a commitment to a common well-being—the kind of housing, schools, public spaces, and access to regional amenities that anyone might need to build their own version of the good life? Could they have helped to transform Moses-era New Dealism to be democratically accountable without giving up on powerful citywide policy tools—ones capable of reshaping, rather than reinforcing, massive inter-neighborhood inequality?

Most importantly: Can it be done today?

The Week Observed, January 6, 2017

What City Observatory did this week

1. A Toast to 2017:  Beer and Cities. Its traditional to begin the New Year with a delicious beverage, and more and more Americans are choosing to celebrate with a locally brewed ale. That’s gotten much easier in the past decade, as microbreweries have flourished around the country. Microbreweries are a decided urban affair. The website claims to have inventoried more than 1,000 microbreweries within five miles of the center of the nation’s 50 largest cities.  Using their tabulations, we’ve computed the density of microbreweries for each of these cities, and produced a ranking.  See where your city stands on this list.

2. How not to make housing more affordable. In the face of rising affordability problems, there’s enormous pressure on elected leaders to do something–anything–to address the situation. Two recent proposals in the Pacific Northwest caught our eye. Both would provide subsidies to households. British Columbia’s Liberal government has proposed matching first-time homebuyer’s down payments–up to $37,500, while a landlord’s group in Oregon has suggested that the state create a $20 million fund to provide rent subsidies (rather than enacting rent control). While both proposals would provide some relief for their direct recipients, economics suggests that in the face of a tightly constrained housing supply, both measures would mostly serve to bid up rents and housing prices–actually making overall affordability problems worse.

3. Brownstone Brooklyn and the challenge of urban change. Daniel Kay Hertz offers his review of Suleiman Osman’s “The Invention of Brownstone Brooklyn.” The book chronicles the movement of middle-class, mostly white professionals from Manhattan in the 1950s and 1960s, looking for authentically urban neighborhoods in Brooklyn. Paradoxically, it turns out that while seeking authenticity, these migrants loved the people, but disliked the public. Brownstoning was a movement that romanticized the participants in Jane Jacobs’ “street ballet,” but which was committed to holding the city as a whole at arm’s length. Ironically—and unfortunately—this way of relating to urban life is the progenitor of much of contemporary gentrification and anti-gentrification thinking.

4. Pollyanna’s ridesharing breakthrough.  You may have seen the headlines last week about a new study claiming that with a ridesharing app and appropriate route assignment software it would be possible to reduce traffic 85 percent in Manhattan. It’s not quite that simple of course. The study in question looks only at replacing current taxi trips with pooled 4-passenger or 10-passenger vehicles. The real problem with this study is that it falls prey to three major fallacies that bedevil much of the techno-optimism about urban transportation, including a failure to consider induced demand, putting too much faith in big data, and assuming that a relatively simple mathematical model can fairly simulate the behavioral responses to a different transportation set-up.

Must read

1. High Transit construction costs.  Last week, with great fanfare, New York City finally inaugurated its long anticipated Second Avenue subway. We have a two-fer hear. At Vox, Matt Yglesias writes about the tragedy of the new subway: which has taken vastly longer, cost vastly more, and produced far less than subway development projects in other advanced countries, something that’s long been documented by Alon Levy at Pedestrian Observations. Building a kilometer of subway in New York costs six times as much as it does in Paris, London, or Copenhagen. Yglesias explores the many reasons for the higher costs: unnecessarily larger stations, deeper tunnels, intransigent unions, poor oversight, and more. Writing at the Frontier Group’s blog, Tony Dutzik piles on, adding that we shouldn’t overlook the fact that highway projects tend to be hugely over-priced as well, and in many cases produce minimal or negative social benefits. And he notes that cost overruns, as well as high price tags, serve as a formidable obstacle to thinking we can reasonably expect to do things differently, in a big way. To these diagnoses we can add the “hitch your wagon to a sacred cow” theory: when multi-billion dollar projects get going, they tend to come with a long list of elaborate add-ons to placate a wide range of constituencies as a way of selling the project.

2. Federal infrastructure spending makes us poorer.  Writing at Strong Towns, Chuck Marohn challenges the conventional wisdom that a big influx of federal infrastructure funding is a good thing for the nation, especially struggling small towns. While it may seem like free money, generous federal capital grants for road construction often lead local governments to borrow to come up with matching funds and have led to sprawling development patterns, which directly raise utility costs for cities and counties, and in the long run, saddle them with prodigious maintenance costs. The emphasis on funding “new” capacity creates competing, peripheral business locations that effectively undercut the economies and tax bases of older, more central locations.

3. Change local incentives to get more affordable housing. The California Apartment Association convened a policy forum to explore ways to address the Golden State’s growing  housing affordability problem, and heard from invited speakers including Trulia’s Ralph McLaughlin, Sonja Trauss of the Bay Area Renters Federation, and Mac Taylor the Legislative Analysts Office. There’s a 19-page report that summarizes the key issues they identified. We would highlight key finding number 4: changing local incentives. A central problem is that local governments have little incentive to approve affordable housing in their jurisdiction. Cities get much more tax revenue from commercial rather than residential development, and also do better fiscally with expensive single family homes than affordable apartments. Individual local governments also face a kind of prisoner’s dilemma: allowing apartments when neighboring jurisdictions don’t means that your community may bear the brunt of the impacts. While these incentive effects may be more prominent in California thanks to the state’s arcane Proposition 13 property tax system, this incentive problem underlies affordability challenges in many cities around the country.

New Research

1. City Street Grid Visualizations.  Streets are the skeletal structure of the city, and the network of streets–their width, spacing, number of intersections–defines the fabric of urban interaction. The University of California, Berkeley’s Geoff Boeing has created a series of same-scale diagrams of city street grids that allow an easy, intuitive comparison of different cities.  Each of the following blocks shows a single mile-square segment in the center of a large city.

Geoff Boeing, UC Berkeley

Boeing built these maps using Open Street Maps and Python. And he’s posted the code for creating similar maps of any other city in the world.  This is a great tool for urban planners to use to better understand and communicate the way streets shape the way we experience our urban environments.

2. Metropolitan Economic Conditions Index.  Last month, the St. Louis Federal Reserve‘s indispensable FRED economic data website expanded the list of metropolitan areas for which it computes an economic conditions index.  The index, developed by Fed economists Maria Arias, Charles Gascon, and David Rapach, illustrates how a region’s gross metropolitan product (the local analog of gross domestic product) is performing relative to other metropolitan areas. Their analysis shows that different recessions affect metropolitan areas in different ways. Some metros experienced severe downturns in 1990 and 2001, while others largely avoided contraction. Nearly all metros felt the effects of the Great Recession, but the effects were more severe and prolonged in some areas than others. The site allows users to display data for any of the 50 largest metropolitan areas for the period from 1991 onward. The chart below compares the economic trajectories of Portland and Minneapolis-St. Paul.  It shows Portland’s economy has overall grown faster, but has been more “boom-and-bust”–growing faster in expansions, and declining more sharply in recessions. This is a useful tool for understanding the cyclical sensitivity of metropolitan economies.

The Week Observed, January 13, 2017

What City Observatory did this week

1. How diverse are the neighborhoods white people live in? Data from the newly released 5-year American Community Survey tabulations give us an updated picture of the demographics of urban neighborhoods. A new report from the Brookings Insitutiton’s Bill Frey shows that the typical US metropolitan area is continually becoming more diverse. But it’s still the case that the typical white resident of a large metropolitan area lives in a much less diverse neighborhood. We take a close look at Frey’s data and plot the relationship between metro level diversity and the experienced neighborhood level diversity of the typical white resident. There are surprisingly wide variations among metropolitan areas; in some very diverse metropolitan areas, the typical white resident lives in a neighborhood with a relatively low level of diversity. These results reflect the continuing importance of localized patterns of segregation.

2. Tying it all together. As we look forward to 2017, we reflect on some of the key lessons that we think we’ve learned at City Observatory.  Here we’ve highlighted some of the principal insights in four broad areas: the growing importance of city centers, the shortage of cities, the need to re-think transportation policy and the challenge of segregation, integration and neighborhood change. We have links to some of our most useful commentaries in each of these areas.

3. Housing supply is catching up to demand. We know that ordinary humans find little solace in the economist’s admonition that rising rents will ultimately be counteracted by an increase in housing supply. The big problem is timing: demand can change quickly, and supply only slowly, so rents can go up until supply catches up. In some particularly hot markets, there’s growing evidence that this is just what is happening. In one of the hottest markets–Seattle–the city is on track to build as many apartments this year as in the preceding half century. And rent increases there and in some other large cities have slowed and in a few neighborhoods, actually declined. As more supply is added in the months and years ahead, its likely to have a further effect on rents.

Rising rents have begun to generate a response..

4. Who pays the price of inclusionary zoning? One of the political attractions of inclusionary zoning is that it seems to be a way to force developers to pay for affordable housing without requiring scarce public funds. Portland’s newly adopted inclusionary zoning requirements include the kinds of density bonuses and other concessions that are common in inclusionary housing programs, but also include a tax break for developers. Plans to give developers a tax exemption for all of the units (not just affordable ones) in high rise apartments could push the subsidy per unit over $200,000, according to city estimates. And property tax exemptions don’t just cost the city forgone revenue: three-quarters of the revenue loss will be felt by schools and other local governments.

Must read

1. Making America great again isn’t about money and power, says Robert Shiller in the New York Times. Many urbanists are leery of the implications of the incoming administration’s promise to “Make America Great Again.” Nostalgia for a poorly remembered bygone era is unlikely to be a good guide to policy in the 21st Century. Economist Robert Shiller, (recipient of the 2013 Nobel in his field), has some potent observations on what leads to greatness: and in his view, its has a lot to do with the kind of communities we build. He specifically invokes Jane Jacobs, and her view of how cities generate the kind of innovation and prosperity that drive our economy: “Cities grow organically, she said, capturing a certain dynamic, a virtuous circle, a specialized culture of expertise, with one industry leading to another, and with a reputation that attracts motivated and capable immigrants.” Shiller warns that the Trump Administration needs to avoid undercutting the conditions that have enabled cities to grow, thrive and create opportunity.

2. Silicon Valley’s self-serving vision for self driving cars. Writing in the Fiscal Times, David Dayen takes a critical view of the hype about the disruptive potential of self-driving cars. Many of the imagined futures assume that the public sector will make massive investments in new infrastructure–like dedicated lanes for high-speed autonomous vehicles. If the utopian vision of fleets of self-driving cars hinges on a massive public subsidy, that transforms this new technology into a kind of scam, according to Dayen: “I think normal people would call what we have here a grift. The car companies want to commandeer public infrastructure as a massive subsidy for their business model. And in the zero-sum world of government spending, such a scheme necessarily crowds out transportation that everyone can afford to use. ” The point is strongly made, but its helpful to remember that plain-old human driven cars also depended for their success on massive public investment, plus a fundamental re-writing of the rules of the road in a way that literally gave the right of way to vehicles over humans. It’s probably worth thinking about that as we move forward.

3. The devilish details of implementing an inclusionary zoning process. The Sightline Insitute’s Dan Bertolet sifts through the byzantine details of Seattle’s proposed mandatory inclusionary housing requirements. In its much ballyhooed Housing Affordability and Livability Agenda (HALA) civic leaders in Seattle reached a handshake deal to implement mandatory inclusion for new development in the city–but left the details to be hammered out by the city’s planning department. Bertolet looks at the complex system they’ve drafted, which divides the city into three broad areas (based on rents), requires three different levels of inclusion (ranging from 5 percent to 11 percent in each), and which offers varying density bonuses in different zones. There’s a lot to digest here in terms of the practical, real-world difficulties of implementing such a program. One such insight: density bonuses may be useless to developers in situations where adding another story would require them to shift from relatively cheap wood construction to all-concrete and steel. Ultimately, he argues, the success of the program will depend on striking a delicate balance between inclusionary requirements and regulatory relief. Just how delicate a balance? Bertolet writes: “Just two 200-unit apartment buildings rendered infeasible by MHA per year would effectively negate the benefits of all of the subsidized units produced by the program.”

New Research

1. Where walking is hazardous to your health. For several years, Smart Growth America has annually produced “Dangerous by Design,” a report taking on the grim but necessary task of detailing the number of pedestrians killed by cars in the US.  Over the past decade, the toll is 46,149 deaths. As is now well understood using the word “accident” to describe these pedestrian deaths amounts to Orwellian blame-shifting. It turns out there’s little random about the pattern of deaths, either in their locations or demographics. A series of metropolitan areas in the US South–especially in Florida–again top the charts in pedestrian fatality rates. These are the same places that have arterial street networks that were mostly designed and built in the automobile era, and which consign pedestrians to second-class status. Not owning a car is clearly a serious risk factor. As the report points out, the elderly, the poor and people of color–who are less likely to own cars and drive–are disproportionately the victims of vehicular death. Designing cities to move cars quickly kills people who walk.

2. How cities promote social interaction. Economist Ed Glaeser, author of Triumph of the City famously described cities as “the absence of space between people.” A key aspect of the urban experience is facilitating interaction between different people. Konstantin Buchel and Maximilian von Ehrlich of the University of Bern undertook an analysis of  anonymized phone records from Switzerland looks into the ways that urban density influences the range and nature of networks of social contacts among residents. Not surprisingly, their study finds that the likelihood of contacts between any two people decreases with distance. But the authors find that density doesn’t so much increase the number of contacts, but the diversity and richness of those contacts. They confirm the idea that denser locations promote the strength of so-called weak ties (that people have networks with less overlap from their friends and colleagues and thereby have greater reach).


How not to fix housing affordability

Plans to subsidize renters and homebuyers will likely just fuel housing cost inflation

Rising rents and home prices are becoming unbearable–or at least politically unpalatable–in cities around North America.  Over the past year, two Pacific Northwest cities, Portland and Vancouver, have seen some of the biggest rent and home price increases anywhere. Portland’s reported double digit rent increases in early 2016 were among the highest in the nation. By one reckoning, Vancouver’s home prices were up 24 percent over the previous year, although, as we’ve noted, there’s been a good deal of noise in home price estimates there. The political pressure that rising prices and rents have produced is generating some novel policy suggestions.

Vancouver (Flickr: Gord McKenna)

In Vancouver, the British Columbia provincial government has announced that it will match the down-payment that first-time homebuyers make to purchase a new residence. The program, announced by Premier Christy Clark–whose ruling Liberal government faces the voters in the new year–would give first time home buyers up to $37,500 to match their down payments.  Up to 40,000 households might qualify for the credit.

Portland just adopted a poorly designed inclusionary housing requirement, and now state legislators are looking at repealing the state’s ban on city-imposed rent control programs. In response, a local landlords group has proposed that the state instead enact a rental subsidy, roughly modeled on the federal Section 8 voucher program. The proposal is that the state would allocate something on the order of about $20 million per year to provide rent subsidies of around $100 per month to as many as 20,000 low and moderate income households.

Both of these proposals intervene on the demand side of the marketplace, giving some homebuyers (Vancouver) or renters (Portland) more buying power to help them better afford new homes or apartments. While its hard to argue that such subsidies won’t be of some benefit to their direct recipients (households that get subsidies will be able to more easily afford housing than otherwise), the economics of these proposals are clear and ugly.  By giving more households more money to bid for housing, these measures amp up demand, but do nothing–at least not directly or in the short term–to raise supply. Greater demand in the face of fixed supply means one thing: even higher prices and rents.

The underlying economic problem in both Vancouver and Portland is that the growth in demand for housing–for urban living, especially in this part of the world–has dramatically outstripped the supply of housing. Adding even more demand in the form of subsidies is like pouring gasoline on the fire.  More prospective renters and buyers, with more money in their pockets will bid up the price of housing still further.  UBC economist Joshua Gottlieb called the BC proposal “shockingly illogical”:

“It’s a pretty bad idea. It is counter-productive and if you give people more purchasing power, they will be able to bid up the prices of the homes that they’re looking at, and that price increase will eat up any benefit from the subsidy. The same people who were going to compete for a home at $500,000 are still going to be competing for that same unit, but at a higher price. They’re just going to compete away the benefit. The only people who gain are existing homeowners or developers who benefit from these higher prices. . . . it’s not going to increase supply and it’s supposed to increase demand, so how is it not going to increase prices?  . . .  There’s no way there was any serious economic analysis.

The same problem holds for the Oregon rental voucher program.  Giving thousands of renters more buying power in the face of an already tight market and a constrained supply of rental units is likely to accelerate rent increases. In both Oregon and British Columbia, the necessary solution is on the supply side: building more housing.  And the money that would be spent on homebuyer or renter subsidies would be better spent on measures to add to the housing stock, especially for low income households.

As we’ve said, the housing affordability crisis ought to be a teachable moment for economists. At its root, the problem cities are experiencing are about supply and demand.  The only way to craft effective solutions is to apply some fundamental economic analysis to the problem. Rent and home purchase subsidies may help out a few individuals, but only at the cost of making the affordability problem worse for everyone else.

Beer and cities: A toast to 2017

Celebrating the new year, city-style, with a local brew

Champagne may be the traditional beverage for ringing in the new year, but we suspect that a locally brewed ale may be the drink of choice for many urbanists today. Much has changed about American beer in the past two decades. Most of the post-prohibition era was characterized by the industrialization of beer-making and the consolidation of the brewing industry; the number of breweries in the US fell from roughly 500 at the close of World War II to only about 100 in 1980. But since then, consumers have turned away from the big national brands and increasingly patronize local, even neighborhood micro-brewers, who offer a wide array of ales, stouts, porters and other distinctive brews, often featuring local ingredients.

’tis the Saison – Happy New Year.

While the overall market for beer grew only about 0.6 percent in 2015, the consumption of craft beer grew by more than 10 percent, according to Nielsen. In the past five years, the market share of craft brewers has doubled, from less than 6 percent of beer sold, to more than 12 percent, according to the Brewer’s Association. Now avid beer drinkers in cities around the country have dozens–and in some cases hundreds–of local brews to choose from. And at City Observatory, that set us to asking:  which city has the largest concentration of micro-breweries.

Which cities have the greatest density of microbreweries?

To answer this question, we turned to which has mapped the locations of a comprehensive database of US microbreweries. The site allows users to search for breweries by location, and we used their on-line search tool to look for all microbreweries within 5 miles of the city center of each of the 53 principal cities of those metropolitan areas with more than a million population. Using a standard five-mile radius allows us to eliminate the variation of computed density that would arise from the very different geographic areas included in city boundaries. Here’s the BreweryMap of Portland.

Portland has 89 microbreweries within this 5 mile circle, the highest density of microbreweries of any large city in the United States.  The following table shows the density (number of microbreweries within 5 miles of the city center) for each of the 53 largest metro areas in the US.  Denver and Seattle rank second and third, respectively, followed by Chicago and New York.  Virginia Beach and Hartford have the lowest density of microbreweries in their urban cores.

Consumer survey data confirm the same geographic patterns of preferences for micro-brewed beer.  Scarborough Research, an arm of Nielsen, reports that Denver, Portland and Seattle had the highest reported rates of micro-brewed consumption of any US metropolitan areas. About 5 percent of American adults have had a micro-brewed beer in the past month: In Denver and Portland, the figure is about two and a half times higher: 13 percent.

In an era in which so much of what we consume is commoditized and globalized, its nice to see a distinctive local product flourishing in so many places around the country. Here’s to a Happy New Year!


How diverse are the neighborhoods white people live in?

Overall, America is becoming more diverse, but in many places the neighborhoods we live in remain quite segregated. The population of the typical US metropolitan area has a much more ethnically and racially mixed composition than it did just a few decades ago. Overall, measured levels of segregation between racial and ethnic groups are declining. But change at the neighborhood level, particularly in the neighborhoods that are home to the “typical” white family, have changed more in some places than others.

Our interest in this subject was kindled last month with an analysis of the latest American Community Survey data, released last month, prepared by the Brookings Institution’s venerable demographer, Bill Frey. In a post entitled, “White neighborhoods get modestly more diverse, new census data show,” Frey looked at the racial and ethnic composition of the nation’s largest metropolitan areas, and gave us his first blush analysis of the unfolding trends of growing diversity and gradually receding racial and ethnic segregation.

The big picture is that American metro areas are becoming more diverse. In the 100 largest metro areas, the share of the population that is white and non-Hispanic has declined from 64 percent in 2000 to 56 percent in 2011-15.  And conversely the share that is Latino, Black, Asian or some other racial-ethnic category has increased from 36 percent to 44 percent. But while metro areas are becoming more diverse, the neighborhood in which the typical white resident lives is much less diverse than the overall metro area. In 2011-15, the typical white resident in a metro area lived in a neighborhood than was 72 percent white, down slightly from a level of 79 percent in 2000.  In 2000, the average white resident lived in a neighborhood that has 15 percentage points (79% – 64%) more “white” than the metro area, in 2011-15, the typical white resident lived in a neighborhood than was 16 percentage points (72% – 56%) more white than the overall metro area.


Frey’s key point is that while America’s metro population is becoming increasingly diverse, especially with the growth of the Latino and Asian populations, most white Americans still live in neighborhoods that are disproportionately white, especially when compared to the overall racial and ethnic composition of the metropolitan area in which they are located. The best way to neatly summarize the complex relationship between neighborhood and metropolitan racial ethnic composition for this purpose is to look at the share of the population categorized as white at the metropolitan level, and then compare it the the share of the population that is white in the neighborhood in which the typical white resident in a particular metropolitan area lives.  Statistically, by “typical” we mean median.  Frey computes the share of the white population in the census tract in each metropolitan area which includes the population-weighted median white resident, with tracts sorted by white share of the population.  Essentially, this means that half of the metro area’s white population lives in a tract with a white share of population higher than this “typical” number and half lives in a tract that has a white share of population lower than this number.

Of course, at City Observatory, we wanted to dig deeper.  And Brookings and Frey have publicly posted their tabulation of the ACS data  (you can download the spreadsheets here).

Metro versus neighborhood

A major factor influencing the demographic composition of the typical white neighborhood is the metropolitan area in which it is located. In more diverse metro areas, the typical white resident tends to live in a neighborhood with a smaller share of white population.  We’ve plotted the relationship between the white share of the metropolitan area population (shown on the horizontal axis of the chart) against the share of the white population in which the typical white resident lives.  The upward sloping line and strong correlation confirms that metro diversity influences neighborhood diversity.

You can think of the line as showing the typical relationship between the share of the population in a metropolitan area that is white and the average share of the population in a typical white neighborhood that is white.  Metropolitan areas above that line have a higher fraction of whites in the typical white neighborhood than you would expect given their demographics, while metropolitan areas below the line are ones where the typical white resident lives in a neighborhood with a smaller share of white residents than you would expect, given the national pattern.  So, for example, consider the difference between Portland and St. Louis.  The two metro areas have a nearly identical share of white population (74 percent in St. Louis, 75 percent in Portland). But the average white St. Louis resident lives in a neighborhood that is 85 percent white, while the average Portland resident lives in a neighborhood that is 77 percent white.  A second example: Memphis and Las Vegas have a similar share of white population (45 percent and 46 percent respectively). Yet the average white Memphis resident lives in a neighborhood that is 66 percent white, while the average Las Vegas white resident lives in a neighborhood that is 54 percent white.

Measuring the neighborhood effect

The difference between the overall share of the metro area population that is white and the white population’s share of the typical white neighborhood is a good indicator of the neighborhood effect:  the extent to which the white population is segregated into neighborhoods that are whiter than the metro area itself.  In effect, the difference between these two neighborhoods is an indicator of how segregated whites are from non-whites in the metropolitan area.  If every census tract had the same white/non-white shares as the metropolitan area as a whole, there would be zero white/non-white segregation.  (Notice that mathematically, the median share white in white census tracts cannot exceed the share white in the entire metropolitan area).   In the following table, we rank metropolitan areas according to the difference in share white in the typical white neighborhood minus the metro area share that is white.

In the median large metropolitan area (for example Hartford, St. Louis or Charlotte), the typical white resident lives in a neighborhood that is about ten percentage points whiter than the metropolitan area as a whole.  The places with the largest difference between the white share of the metro population and the white share of the typical white neighborhood include Miami, New York and Los Angeles, where white residents live in neighborhoods that are about 22 percent more white than the metropolitan area.  The places with the smallest difference between the typical white neighborhood and the metro area average include Portland (2 percentage points), Pittsburgh and Salt Lake City (both about 4 percentage points).

The upshot is that when it comes to the lived experience of diversity, some factors are global, but important ones are still local. Its possible to live in a very diverse metropolitan area, with a high fraction of non-Hispanic white residents, and still have a high level of segregation, so that white residents live in places where they have very high levels of white neighbors. And the converse is also true, in some


Throughout this commentary, we follow Bill Frey’s condensed version of Census Bureau racial and ethnic categories.  By white, we mean persons who report to Census that they have a single race (white) and who report they are not of Hispanic origin.  Frey’s report includes data for African-Americans and Asian-Americans who report a single race, persons of Hispanic origin, regardless of race, and a final category including all other persons.  As we do regularly here at City Observatory, Frey reports data for metropolitan areas with a population of 1 million or more.

Suburban Renewal: Marietta demolishes affordable housing

Just say the words “urban renewal” and you immediately conjure up images of whole neighborhoods–usually populated by poor families and people of color being dislocated by big new publicly funded development projects. It seems like a relic of the past.  But it appears to be getting a new lease on life in the suburbs. For a couple of years now, we’ve been following the story of Marietta, Georgia, where local officials used $65 million in taxpayer funds to buy up and begin demolishing some 1,300 apartments along Franklin Road. This is a striking case where the displacement of low income families was an explicit objective of public policy, rather than the side-effect of a changing real estate market. The tale raising some interesting questions about how we talk about neighborhood change, and whether we’re really open to economic integration in all places–city and suburb.

In a few weeks, the new Atlanta United franchise of Major League Soccer will kick off its inaugural season. The team’s been preparing at the verdant practice facility in the Atlanta suburbs.

Atlanta United’s Marietta Practice Facility (Google Maps)

What isn’t apparent from this picture is the fact that hundreds of old but serviceable apartments were demolished to make way for these spacious fields. The Marietta apartment complexes had been built in the 1960s, and, when new, were a preferred upscale location for single professionals and young married couples. Over the decades, the apartments aged and the mix of occupants changed. Franklin Road shifted to families with increasingly modest incomes—a process housing economists call filtering, which is the primary source of affordable housing throughout the nation. Along the way, the economic and and racial makeup of the apartments transformed from nearly 90 percent white in 1980, with a poverty rate around five percent, to 20 percent white in 2010, with a poverty rate of nearly 25 percent.

Despite the usefulness of filtering, which increased the diversity of suburban Marietta, the city perceived these units as growing concentrations of poverty and, thus, a problem. So it used the proceeds of a voter-approved bond measure to purchase and begin demolishing the housing complexes. It’s worth noting that no one ever claimed that the buildings themselves were a problem, despite their age. Rather, it had everything to do with the demographics of their occupants. In all, the city’s plan calls for acquiring and demolishing almost 10 percent of the city’s multi-family housing stock.

Caution:  This post contains graphic images of housing displacement. Viewer discretion is advised.

Marietta’s plan is proceeding apace. The city re-christened Franklin Road as “Franklin Gateway” to signal change. It began demolishing the apartments last year. Here, we’ve used imagery from Google Maps show what’s happened where one of these complexes—the Woodland Park Apartments—once stood.  This is what they looked like in 2011.

2011 (Google Streetview)
Google Maps, September 2011

Early this year, the demolition was nearly complete. All that remains of  the old apartment complex is its driveway, a partial brick wall and metal gate, and two patches of evergreen shrubbery, flanked by a stand of pine trees.

Google Maps, March 2016
Google Maps, March 2016

The latest Google imagery shows the city’s unfolding plan for development. The site will become the training facility for Atlanta United, the area’s new major league soccer team. For use of the 32 acres formerly covered by apartments, the team will pay the city $1 per year for the first ten years of a thirty-year lease.

Google Maps, May 2016
Google Maps, May 2016

The demolition of the apartments on Franklin Road represents a kind of national blind spot when it comes to talking about neighborhood change. In any large city—say New York, Los Angeles or Washington—the wholesale demolition of affordable housing to provide discounted land for new businesses would undoubtedly be treated as the most pernicious form of gentrification. But because it happens in a suburb, somehow it doesn’t count, or at least isn’t objectionable.

Perhaps this reflects a deeply ingrained but seldom-voiced bias in our views about place: Suburbs are for rich, mostly white people. Cities are for poorer people and people of color. Anything change that runs counter to this worldview (like gentrification of a Brooklyn neighborhood, or efforts to build affordable apartments in suburbs like Marin County) is an affront to the order of things. The apparent prevalence of this outlook shows just how hard it will be to make progress on economic integration.

Flood tide–not ebb tide–for young adults in cities

The number of young adults is increasing, not declining, and a larger share of them are living in cities.

Yesterday’s New York Times Upshot features a story from Conor Dougherty–”Peak Millennial? Cities Can’t Assume a Continued Boost from the Young.” It questions whether the revival in city living is going to ebb as millennials age, and the number of persons turning 25 decreases in the years ahead.

In our view, the story offers two mistaken premises: first, that the growth of young adults in cities has been driven primarily by the large size of the Millennial generation, and second, that the affinity of young adults for cities is waning. Our research shows that neither of these premises is true. The movement of young adults to the city has been gathering steam for more than 25 years, and the number of young adults in cities was actually increasing during the 1990s–at a time when the number of 25 to 34 year olds nationally was actually declining. And the relative preference of young adults for city living continues to increase.

The argument that as millennials age they will increasingly move to the suburbs mistakenly conflates life-cycle effects with generational change. As individuals age, the likelihood that they’ll live in different places changes. After high school, there’s a large migration to college towns. Young adults, starting out in their careers are disproportionately likely to rent and live in cities compared to other Americans. As they get older, find partners and have children, they’re more likely to own homes and live in the suburbs. This general pattern of succession holds for most recent generations as they age. But what’s different–and important–is how many people are in each generation, and how long they remain in each stage in this process. What’s happening now is that today’s young adults–the so-called Millennial generation–are both more numerous than the immediately preceding generation and are demonstrating a greater propensity to spend a larger share of their early adult years living in cities.  That’s the essence of what our research (and that of many others) shows has been happening.

The Upshot story takes issue with this thesis in two respects.  First, building on an argument advanced by USC’s Dowell Meyers––which we addressed when it first came out–Upshot says that the fortunes of cities will wane because the number of persons turning 25 years of age will decline slightly in the next decade. Second, Upshot says that as individual millennials get older, they’ll tend to move to the suburbs in greater numbers.

The essence of the Upshot story is two claims: (1) that the impact of millennials on cities will decline because their numbers will decrease, and (2) that their propensity to choose to live in urban settings will decline. Let’s consider each of these ideas in turn.


The number of 25-34 year old millennials will increase by about 3 million over the next 7 years; this is the stage in the life cycle when they are most likely to live in cities.

Is the move to cities being buoyed by the rising number of millennials, and will their numbers decline soon, and therefore cause a decline in cities?

What’s interesting is that in 2015—for the first time—Millennials (those born between 1980 and 2000) constitute all of the persons aged 25-34. At the time of the last decennial census (2010), about half of the 25- to 34-year-old age group was composed of people in the tail end of generation X, and half were the early wave of Millennials. So strikingly, what the Census data shows is that the total number of 25 to 34 year olds in the US will increase from now through 2024. This chart shows the Census Bureau’s estimates of population aged 25 to 34 based on historical data through 2014 and its projections through 2035.

Screen Shot 2016-03-31 at 12.51.15 PM

During the 1990s the number of 25 to 34 year olds actually declined as Baby Boomers aged out of this age group and were gradually replaced by the numerically smaller Generation X. Between 2000 and 2010 the number of Gen-Xer 25-34 year-olds increased slowly, with all of the aggregate increase in this age group being recorded after 2008. So, to the extent there was a movement back to the city in the 1990s, and the first half of the 2000-2010 decade, it was propelled not by an aggregate increase in the number of 25-to-34 year olds in the nation, but the changing relative preference of young adults for urban locations.

So, what we—and others—have recorded as the movement of young adults back into cities has, until very recently, little to do with the size (or preferences) of the Millennial generation. In fact, as they turn 25, and as they now dominate this age cohort, the next decade will be the time when the Millennial generation’s effect on cities will be most fully felt. Rather than declining, the number of 25-to 34 year olds in the United States will increase each year from now through 2024, rising from 44.1 million in 2015 to 47.6 million in 2024. In reality, the Millennial wave of urbanism is just now hitting the beach.

The outlook after 2024 (when the 25-34 year olds will increasingly be “post-Millennials”) is not for a dramatic demographic collapse. Rather than a peak, the young adult population stabilizes at a fairly high plateau above 47 million 25 to 34 year olds through 2035. So there is little basis for forecasting a decline in the key population group that has driven urban growth.


With each passing year, 25-34 year olds, especially those with a four-year college degree or more education are more likely to live in close-in urban neighborhoods than other Americans.

Are young adults becoming less likely to live in cities?

At City Observatory, we’ve tracked the carefully tracked the location of urban residents in America by age group over the past three decades.  We’ve measured the relative preferences of young adults for close-in urban neighborhoods (census tracts within three miles of the center of the central business district).  The relative preference is the probability that a young adult will live in a close in neighborhood compared to the probability than any other resident of any age would live in such a neighborhood. These figures are drawn from Table 5 of our Young and Restless report; we’ve computed relative preference by dividing the probability that a person aged 25 to 34 lives within a three-mile radius of the center of the CBD of one of the 51 largest metropolitan areas, and compared it to the probability that the average resident of a metropolitan area live in this radius. If 11 percent of 25 to 34 year olds live in the 3 mile radius, and 10 percent of the population as a whole lives inside that radius, the relative preference is 10 percent (11 percent/10 percent)=110 percent, meaning that a 25 to 34 year old is 10 percent more likely than the typical resident to live in this area.

Since 1980, the relative preference of young adults for close-in neighborhoods has increased steadily. In 1980, young adults were 10 percent more likely than all metro residents to live in these neighborhoods; in 1990 12 percent more likely, in 2000 32 percent more likely; and in 2010, 25 to 34 year olds were fully 51 percent more likely to live in close-in neighborhoods than other metro residents. The relative preference of 25 to 34 year olds with a four-year degree to live in such neighborhoods was even higher:  over 100 percent in 2010.

Another way of looking at this is examining census data on where young adults live.  The University of Virginia’s Luke Juday has prepared a revealing set of charts that show the concentration of the young adult population by distance to the city center for all of the nation’s 50 largest metro areas.  He has comparative data for 1990, 2012 and 2015.  The data–just as in the statistics cited above–show that in the aggregate, the share of young adults living in close-in neighborhoods has increased and the share living in more distant neighborhoods has decreased. The orange line shows the share of young adults (age 22 to 34) as a share of the population in 1990.  The magenta and teal a lines show the share of young adults in 2012 and 2015 (respectively).  The share of young adults living in close in neighobrhoods (which had been higher in the center in 1990) increased over the next two decades. The share living in the suburbs declined. The steepening of this gradient is clear evidence of a growing relative preference of young adults for central locations.

This essential finding is actually the core of several well regarded academic papers.  Edlund, Machado and Siviatchi show that well-educated prime-age workers are increasingly concentrated in neighborhoods closer to the center of the metropolitan area. Couture and Handbury have replicated our findings, reporting:

A recent report by CEO for Cities (Cortright (2014)) – and covered extensively by the New York Times (Miller (2014)) – also uses 2000 census data and 2007-2012 ACS data, and shows that the 25-34 college-educated population are growing faster downtown than in the suburbs in the majority of the 51 largest MSAs. We confirm and expand this narrative to the older 35-44 college-educated group . . . Strikingly,we find that the college-educated 25-34 age group grows faster in the urban area of 23 of the 25 largest CBSAs. The exceptions are Riverside, which essentially lacks a downtown, and Detroit, which is famously struggling.

These papers, and Rebecca Diamond’s research show that the attraction of cities is amplified by the growth of and growing demand for urban amenities.

More young adults are migrating to cities

Finally, almost in passing, the Upshot article resurrects a claim made in 2015 by’s Ben Casselman, which had asserted that more young adults were now moving from cities to suburbs than vice versa, and claiming “Whether by choice or economic circumstance, young Americans are still more likely to leave the city for the suburbs than the other way around.”  We had a lengthy back-and-forth with Casselman at the time, but the University of Virginia’s Luke Juday pointed out that the data that Casselman used from the Currently Population Survey effectively missed millions of young adults, and disproportionately missing those living in cities. Juday’s analysis shows that actually the reverse is true–young adults are increasingly moving to cities:

Over the past 5 years about 3 million more Americans age 20-29 moved from suburbs to principal cities than from cities to suburbs, with last year being the largest net gain for cities yet.

The Takeaway: More Young Adult Urban Growth is Coming

The number of 25 to 34 year olds—the key group driving urban living, will not decline, but will grow between now and 2024. The urban wave we’ve experienced starting in the 1990s, and accelerating in the past decade wasn’t propelled by generational growth, so much as by a growing preference for urban living by young adults. The shift of young adults to cities, drawn by urban amenities, is increasingly confirmed by academic researchers, and is borne out by the latest Census data. 

Data Notes

The Census data for our of estimates of the 25 to 34 year old population come from three sources. Data for the period prior to 2010 comes from the archive of historical Census population estimates (

Data for the period 2011 to 2014 comes from Annual Estimates of the Resident Population for Selected Age Groups by Sex for the United States, States, Counties, and Puerto Rico Commonwealth and Municipios: April 1, 2010 to July 1, 2014, Source: U.S. Census Bureau, Population Division, Release Date: June 2015. See: Census projections of the population by year and age for the period 2015 through 2035 come from the 2014 Census Population Estimates series

* Note the discontinuity in the data between 1999 and 2000 reflects the disparity between the Census Bureau’s annual intercensal estimates of the 25-34 year-old population and the actually higher number of 25-34 year olds enumerated by the 2000 decennial Census. It’s likely that the actual number of 25-34 year olds was underestimated in the intercensal estimates, during a period of significant immigration).

Pulling it all together

At City Observatory, we post several new commentaries each week on a variety of urban themes, and aim to provide discrete, coherent analyses of specific questions, and contributing to the policy dialog about cities. At the start of a new year, we’d like to pull back a bit, and reflect on what we think we’ve learned, and how these varied pieces add up to a cohesive vision.

So what follows is not a manifesto, but more an outline of of the knowledge assembled in our work at City Observatory, since we started in October 2014. Here’s a list—not entirely exhaustive—of what we came up with, grouped into four big themes:

The growing economic importance of city centers

Credit: Jonathan Miske, Flickr
Credit: Jonathan Miske, Flickr

The demand for cities is rising.

Talented young people are increasingly choosing to live in urban centers.

People are increasingly seeking dense, diverse, interesting, transit-served, bikeable and walkable communities.

This is leading to a surge in city center jobs.

Cities are powering the nation’s economic growth in this cycle.

Cities are cleaner, greener, and safer than ever before.

The economic advantage of cities is growing in providing convenience and experiences.

The shortage of cities

We have a shortage of cities, the growing demand for city living is outstripping the supply of great urban spaces, which is producing higher housing costs.

Our land use planning systems, dominated by homevoters, make it too hard to build new housing, especially in the most desirable locations, driving up housing prices.

Hyper-local decision-making can shut out important voices and lead to more segregated cities.

We’ve effectively made the most desirable kinds of housing—dense, diverse mixed use neighborhoods with narrow streets, and a varied range of housing types—illegal.

Other prosperous countries with attractive cities have very different ways of zoning that allow more traditional urban neighborhoods.

The need to rethink transportation policy

Credit: Montgomery County Planning Commission, Flickr
Credit: Montgomery County Planning Commission, Flickr


The big subsidies to parking—socialism for car storage in the public right of way—undermines biking and walking and drives up the price of housing.

There’s no such thing as a free way—taxpayers subsidize car ownership significantly, causing people to drive much more than they would otherwise.

The engineering rules of thumb that are used to forecast traffic, set road widths, require parking are pseudo-science, with perverse effects on cities and humans.

The way we design our roads costs thousands of lives a year. It’s time for another road safety revolution.

When it comes to public transit, what matters is reliability and convenience—not whether it’s rail or bus.

Land use is as important to public transit as the actual transit infrastructure. It’s especially important to have destination density—of jobs, amenities, homes—near transit stations.

The challenge of segregation, integration, and neighborhood change

Economic segregation is growing, the rich and the poor live apart from one another in our cities, this is a product both of the secession of the rich, especially to exclusive suburban enclaves, and by the concentration of poverty.

Gentrification, though rare, is actually reducing economic segregation.

Poor households living in gentrifying neighborhoods are no more likely to move away than poor households in non-gentrifying neighborhoods and report higher incomes and greater satisfaction with their neighborhoods than those living in non-gentrifying neighborhoods.

Unless housing supply increases in high demand locations, rents and home values will rise, and the poor will be priced out of neighborhoods.

High-inequality neighborhoods actually reduce inequality at the city and metro level.

Obstructing new development, even new higher-income development, is a recipe for aggravating problems of affordability and displacement.

Policies that aim to put the burden of paying for affordable housing on developers are unlikely to work. Their margin is too small, and the incentive effects will lead to less housing being built.

The scale of public investment in affordable housing is dwarfed by the housing market—but we can do better.

In the coming year, we’ll look to dig deeper into each of these propositions, and add others to our list. If you take issue with the positions we’ve staked out here, can offer relevant evidence that confirms, denies or sharpens these propositions, or have other ideas that are candidates for this list let us know. We look forward to continuing this conversation in 2017.