The real $3.4 billion hole in the I-5 bridge project

The Oregon and Washington transportation departments understated the funding gap for a revived I-5 Columbia River Bridge by more than $1 billion

Correcting for an arithmetic error increases the gap between identified revenues and potential costs from $2.3 billion to $3.4 billion.

ODOT & WSDOT also used too low an inflation factor for escalating project costs, understating total costs by a further $680 to $860 million.

Preliminary estimates appear to leave out costs of the revived planning effort and compensation to river users for the new bridge’s lower navigation clearance.

And ODOT and WSDOT have a history of under-estimating costs and over-estimating revenues.

The dreadful transportation news from the Pacific Northwest this month is the continuing plans of the Washington and Oregon transportation departments to revive the years-dead Columbia River Crossing project.  That project foundered in 2014, after nearly a decade of planning—and roughly $200 million spent on staff and consultants—because neither state had the money to pay for the project.

This past year, the two states have scraped up another $50 million and are hiring consultants and dusting off the old CRC plans, with the thought of reviving the project.  They’ve concocted a totally false story that if they don’t start construction on a new bridge by 2025, they’ll have to repay the US Department of Transportation the $140 million in federal money they spent earlier. As we demonstrated a year ago, that claim is incorrect, Federal Highway Administration regulations provide that if the states select the “No-Build” alternative at the end of the NEPA review process, there is no repayment liability.

But now, as the last time round, the big issue is who will pay for the project.  And neither state has the money.  Last week, the two state DOTs released their “Draft Conceptual Finance Plan,” which isn’t so much a plan as it is a picture of giant hole in the ground that they’d like to fill with money, if they can find some. As Clark County Today reported, money is still the sticking point for this project:

Sen. Lee Beyer (Oregon’s 6th District) stated he believed the major problem of the project will be “an inability to fund the project.”

Bottom line on the plan:  The two DOT’s told the legislators overseeing their work that the estimated size of the hole is between $1.8 and $2.3 billion.  Senator Beyer seemed to express considerable skepticism that even this range was within reach.

But even if one believes their cost estimates and revenue projections—and one shouldn’t as we’ll explore in a minute—the two agencies couldn’t even do the arithmetic correctly to state the actual range of estimates of the size of the hole in their finance “plan.”  The real gap, according to the two DOTs’ own numbers ranges as high as $3.4 billion, more than a billion dollars higher than the maximum gap they estimated.

Here’s how they made their mistake.  The DOTs constructed “high” and “low” estimates of revenues and expenditures for each of two alternatives (a widened freeway plus bus rapid transit system, and a freeway plus light rail line).  For each cost and each revenue estimate has a separate high and low estimate.  They computed the range of estimates by combining the low estimate of costs to the low estimate of revenues and comparing it with the high estimate of costs and high estimate of revenues for each alternative.  (The red arrows on the table below show how they’ve lined up the low estimates of cost with the low estimates of funding, and likewise with the two high estimates).

ODOT’s incorrect table claiming a $1.8 to $2.3 billion gap.

But that’s wrong:  The true range is illustrated by combining the low range of costs and the high range of revenues and comparing it to the high range of costs and the low range of revenues. The following chart fixes the error in ODOT’s estimates, and now compares a combination of the low range of revenues with the high range of expenses, showing that there’s a potential funding gap of up to $3.4 billion—fully a billion dollars more than acknowledged in ODOT presentation. We’ve literally just re-arranged the “funding assumption” rows in the ODOT chart above to line up the “low cost” assumption with the “high revenue” assumption, and the “high cost” assumption with the “low revenue” assumption, and then recalculated the values in the right-most column to reflect this change.  This generates the correct range of estimates of the gap implied by these figures.

Our corrected table shows the real gap could be as high as $3.4 billion.

The truly risky case here, and the one the states need to plan for if they’re moving forward, is a project that has costs at the high end, and that has revenues at the low end.  And in that case the gap is roughly $3.4 billion.

Now it’s true, that under the most optimistic reading, (high revenues and low costs) the gap might be only $800 million), but what this presentation has done is greatly overstate the precision and understate the financial risk associated with this project. Rather than being a relatively narrow gap of $2 billion plus or minus a couple of hundred million, the range of possible estimates of the gap are from a little less than a billion to nearly three and a half billion. But that’s not all.

But even these figures can’t be trusted.

There is, of course, another shoe (or two) that will drop here:  The two DOTs, ODOT in particular, have a lousy record in accurately forecasting project costs.  Less than a year ago, ODOT bumped up its cost estimate for the I-5 Rose Quarter project (just a few miles south of this proposed bridge) by 75 percent, from $450 million to as much as $800 million.  The estimates presented here are described as “initial”—they’ve simply been extracted from 2012 vintage CRC estimates, and inflated to current and year-of-expenditure dollars.  Already, these estimates are probably low.

For starters, in inflating the estimates from 2012 levels to 2020 levels, the two DOTs have pegged annual construction cost inflation to 2 percent per year.  But in the real world, according to the US Department of Transportation, highway construction costs have risen at an annual rate of 3.0 percent from the third quarter of 2012 to the third quarter of 2019 (the latest period for which data are available.  This higher actual increase in highway costs already recorded between 2012 and 2020 means that ODOT and WSDOT have underestimated the current (2020) cost of the base CRC project by about a quarter of a billion dollars.  Assuming that they similarly under-estimate inflation over the next decade (using a 2.0 percent inflation factor, rather than the 3.0 percent inflation factor we’ve experienced over the past decade, increases the year of expenditure cost of the project to between $4.0 billion and $5.7 billion, an increase of between $680 million and $860 million from the estimates made by the DOTs.  These changes represent an increase in the revenue “gap” for the project.  Combined with the earlier arithmetic error in calculating the range (above) that means that the funding gap could easily be more than $4.2 billion.

But there are other problems with the gap estimate.  While the new revenue estimates count the $50 million Oregon and Washington have chipped in for the renewed planning effort, but there’s nothing that indicates that those costs were added to the old 2012 estimates of construction costs (which assumed planning was essentially complete).  It’s also unclear whether the project costs include any of the promised $86 million compensation payments for river users whose access to the river will be impaired by the proposed bridge’s new fixed span. And since the agencies haven’t designed or selected the project, its impossible to say what the real cost might be. But we know one thing for sure:  the state DOTs almost never guess low.

Both ODOT and WSDOT routinely go over budget on major projects.  WSDOT is already $223 million over budget on the Alaskan Way Viaduct’s replacement tunnel, and faces an unresolved lawsuit for $480 million more in costs associated with a failed tunnel boring machine.  ODOT’s largest recent project, a five-mile widening of Highway 20 was 5 years late and more than $200 million over budget.  A series of major ODOT projects over the past two decades have experienced cost overruns averaging 200 percent.  When something is called a “draft conceptual” finance plan, that’s simply bureaucratic code for “we’re low-balling the cost estimate now and we’ll certainly raise it later.”

Similarly, the two agency’s have equally lousy in estimating revenues, and most notably, toll revenues.  Back when the CRC was being planned, WSDOT was in the process of developing a financial plan for the SR 99 deep bore tunnel under Seattle (which replaced the now-demolished Alaskan Way viaduct).  WSDOT confidently told the state legislature that tolling the new tunnel would generate $400 million toward its construction costs.  Even before the tunnel was built, that number was cut in half; current experience shows that the revenue could be even less than that.

So, all in all, the financial exposure to Oregon and Washington is easily in the range of $4 billion, but is likely to go up from there, as the project will no doubt cost more, and tolls will likely generate less revenue than forecast.

It’s alarming that with a $50 million budget, and with responsibility for a multi-billion dollar project, Oregon and Washington’s DOT’s could make a billion dollar math error in a such a vitally important calculation.  In doing so, they understated the liability the two states face, and created a false illusion that there’s a narrow range of uncertainty about the size of the gap the project faces.  Is this incompetence or deceit?  Can it be both?

More cynical greenwashing from the highway industry

There’s no shortage of cynical greenwashing to sell climate-killing highway widening projects

GeorgiaDOT and AASHTO have a new PR gimmick to promote the same old product

AASHTO—the American Association of State Highway and Transportation Officials—was touting one of their innovative environmental programs, something called “Planning and Environmental Linkages.”  Its currently being deployed by the Georgia Department of Transportation (GDOT).

Planning and Environmental Linkages or PEL represents a collaborative and integrated approach to transportation decision-making that 1) considers environmental, community, and economic goals early in the transportation planning process, and 2) uses the information, analysis, and products developed during planning to inform the environmental review process.

The benefits include: Improved relationships with stakeholders; improved project delivery timelines; and better transportation programs and projects.

The description is replete with the appropriate buzzwords—collaborative, stakeholders, improved—but its clear from the description that this greenwash pure and simple.  Note that PEL only gives procedural weight to environmental and community factors—it will “consider” them—but offers no substantive metrics, like reducing vehicle miles traveled, or lower greenhouse gas emissions.  It says it will “use” this information in the planning process, but doesn’t say how. (In our experience, it’s usually for denial and obfuscation). Ultimately, there’s no question of the objective here:  delivering transportation projects:  The only tool we have is a hammer, and by God, all our problems are nails.

The greenwash is epitomized by the PEL logo, which looks more like something designed for the Sierra Club, than a highway agency.


There’s a turquoise blue sky, white mountains, a green forest and a blue highway.  There’s a single, soaring bird, but nary a car to be seen anywhere.

But the reality of what the GDOT is planning and what it will look like on the ground is dramatically different. This becomes clear when one digs deeper into the example of “Planning and Environmental Linkages” offered by AASHTO. Its a huge project to widen I-85 in the Atlanta metro area.  The plan is plainly to add vastly more car-carrying capacity on I-85, through a combination of additional travel lanes, shoulder driving, “auxiliary lanes,” the construction of “collector/distributor” side-roads along the freeway, and mammoth new car-oriented intersections, like this diverging diamond interchange.

To get a more detailed idea of how they’d integrate environmental concerns into project planning, we took a quickly look at the “I-85 PLANNING AND ENVIRONMENTAL LINKAGES (PEL) STUDY Corridor Strategies Memo.  Strangely for a report that promises to address “environmental linkages” there’s virtually no mention of the environment, or the highway’s negative impact on the climate and adjacent communities.  The words “climate,” “pollution,” “greenhouse gas,” “emissions” do not appear anywhere 30-page report. There are two mentions of “bicycles”, one of them in the context of diverging diamond interchanges, which are notoriously hostile to non-auto travelers.  The message of the document is clear:  the word “environment” might be on the cover, but the needs of the environment and non-auto users are simply irrelevant to the final outcome.

If you really cared about environment, you might pay attention to the fact that greenhouse gas emissions from transportation have been increasing steadily and sharply in Georgia.  Since 2012, metro Atlanta’s average emissions per capita are up more than 1,000 pounds per person.

In short, this is all about generating the perception that GDOT cares about the environment.  It doesn’t.  It includes environmental concerns in only the vaguest and most procedural senses, not specifying any substantive or measurable goals.  It focuses on options that are all about making driving easier and expanding capacity, with lip service given to alternative modes.  And it doesn’t even bother to acknowledge pressing environmental problems.

Why—and where—Metro’s $5 billion transportation bond measure failed

Portland voters resoundingly defeated a proposed multi-billion dollar payroll tax to pay for transportation projects

The two areas slated for the biggest benefits voted against the measure:  The Southwest Corridor and East Portland both opposed the measure

A generous electorate didn’t want to spend billions on transportation

A few months back, we laid out the case against a $5 billion transportation bond measure proposed by Portland’s regional government, Metro. In the November election, voters rejected the measure by a 58 percent to 42 percent margin.

It’s difficult to argue that the measure failed because of widespread anti-tax sentiment:  Voters in the Portland and Multnomah County, in the center of the region, approved every single money measure on the ballot, except for Metro’s transportation tax.  Portland voters approved a $400 million parks levy (64 percent yes), a $1.2 billion school bond (75 percent yes), a $400 million library bond (60 percent yes). They voted to impose a new high earner tax to fund tuition-free preschool education.  These votes come on top of other measures they approved strongly in the May primary election, including multi-billion dollar regional tax to fund homeless services and extension of a 10 cent a gallon Portland city gas tax—which passed with an 77 to 23 percent margin of victory.

Opponents of the measure, chiefly the area’s business community, mounted a  $2 million campaign that mostly emphasized a series of anti-tax messages. But many in the community shared our concerns that the measure did nothing to reduce greenhouse gases. More generally, it appears that the public wasn’t convinced of the benefits of this proposed spending package, which would have been the most expensive single such effort in the region’s history. You’d think that $5 billion worth of projects, chosen after a year-long involvement process, would produce strong support, especially among the likely beneficiaries of these projects. But the evidence suggests that the measure failed to generate much support in two sets of neighborhoods that were singled out for major expenditures, in Southwest Portland and in East Portland.

The pattern of support and opposition in Multnomah County

While there hasn’t been any post-election polling to lay out the reasons for the measure’s crushing defeat, there are some clear clues in the geography of election returns. In particular, election returns for Multnomah County, which are available at the precinct level show, the distinct geography of support for and opposition to the Metro measure.  (Precinct data for Clackamas and Washington counties weren’t available as of publication date).

This map shows the overall pattern of votes in Multnomah County, with areas supporting the measure shaded green, those opposed shaded yellow, and those extremely opposed (fewer than 35% “yes” votes), shaded red.

Support came from a green core. Within the county, the precincts which voted in favor of the measure are in the city’s close-in urban neighborhoods, including downtown, and adjacent neighborhoods in Southeast, Northeast and parts of North Portland.  Outside these areas, voters were opposed. The support for the measure came from a handful of precincts on Portland’s westside, in and near downtown, and a swath of close-in urban neighborhoods in North, Northeast and Southeast Portland. These neighborhoods tend to have the highest levels of transit ridership and bike commuting in the region, as well as being politically liberal. The pattern of voting in favor of the measure closely resembles the split between incumbent Mayor Ted Wheeler (who was re-elected) and his challenger, Sarah Iannarone; Iannarone, the progressive, did best in the same precincts that voted for the measure.

Source: Multnomah County Elections.

The map of the left shows the electoral split for the Metro Bond, with the light colors representing yes votes and the darker colors “No” votes.  The map of the right shows the plurality winner of the Portland Mayor’s race, with moderate incumbent Ted Wheeler in light blue, and progressive challenger Sarah Iannarone in dark gray.  The Metro measure did best in those precincts most favorable to Iannarone.  More liberal-leaning neighborhoods supported the measure, but it wasn’t enough to produce a majority even in Multnomah County.

Neighborhoods that stood to get projects voted against the measure

In theory, at least two geographic constituencies should have been big beneficiaries of the measure.  The biggest single project in the package, the SouthWest Light Rail, earmarked for nearly $1 billion of Metro money, would have built a new light rail line with stations in neighborhoods in Southwest Portland.  But apparently this area didn’t care:  All of the Multnomah County precincts through which the project would have run voted against the measure.

Another constituency that was supposed to benefit from the measure were residents of East Portland. Metro stressed that the measure was designed to improve equity, by investing in transit, safety and pedestrian improvements, especially in low income neighborhoods like those along 82nd Avenue in East Portland.  This area, formerly part of unincorporated East Multnomah County, has some of the region’s weakest transportation infrastructure, with many unpaved streets and relatively few sidewalks. In recent decades, income levels in this area have slipped relative to the region, and advocates of the measure made a particularly strong point that spending in East Portland would address serious equity concerns. But the election returns show that there was no outpouring of support from these neighborhoods. East Portland is generally regarded as the area East of 82nd Avenue.  The map below zooms in on Portland’s East side, and shows a clear dividing line at 82nd Avenue (the red line on the map). Only 2 precincts East of 82nd Avenue voted in favor of the measure; nearly all were opposed. Many of the precincts in the area were strongly opposed, with fewer than 35 percent “Yes” votes. Claims that this measure would advance equity and improve safety generated no support in the very communities would ostensibly benefit from the measure.

No support in the suburbs, either

The Metro electorate includes voters in three counties, Multnomah (which includes nearly all of Portland) and two suburban counties:  Clackamas and Washington.  The measure lost in Multnomah County by a 54/46 margin, and was defeated by an even larger margin in each of the two counties. Neither of these counties has yet released precinct-level returns through the Secretary of State’s website, so we haven’t analyzed their geography here.


Limits of the “Christmas Tree” approach

From the outset, the Metro measure was fashioned as a kind of giant Christmas tree, with a raft of projects each designed to appeal to a specific constituency: a billion dollars for Tri-Met’s next light rail line, a freeway interchange for the Port of Portland’s airport, new highways in suburban Clackamas County, a big contribution to Multnomah County’s plan to replace the Burnside Bridge, sidewalk and safety projects to appeal to pedestrian advocates, and transit fare subsidies for students.  Seemingly everyone participating in Metro’s process got a little something and endorsed the package. Metro’s effort’s seemed dominated by an at times cynical political perspective. Early on, they fielded misleading push-polls falsely claiming that measures to reduce traffic congestion would reduce greenhouse gas emissions.

It was all well and good to come up with a list of projects, but Metro’s political calculus doomed this effort when it came to deciding how to pay for it. Supposedly because a gas tax didn’t poll well, Metro’s leader’s tried to stick it to “someone else” —the someone else being the region’s larger employers. The actual funding, a selective payroll tax, was decided only at the last minute, and with little public debate, and Metro chose to exclude itself another other local governments from paying the tax. In theory, a measure that only taxed big businesses and exexempted politically popular small businesses, and which was pitched as a tax that someone else would pay, should poll well. But in the end, it generated considerable ire with the business community, which opposed it, and the “no” campaign’s messages clearly resonated with the voters.

In our view, it was unwise from a policy standpoint to choose a method of paying for this package with little or no relation to transportation. The payroll tax amounted to the equivalent of a 30 cent a gallon subsidy to gasoline purchases, compared to the more conventional way of paying for road improvements through the fuel taxes. Ironically, as the 77 percent “yes” vote for extending the City of Portland’s 10 cent per gallon just six months ago demonstrates, voters and the business community had little problem, when asked,  in paying for transportation in a direct and visible way. Moreover, Portland passed a gas tax with a margin of more than 100,000 votes, a block that would have dramatically bolstered chances for regional success even if suburbs were intransigent.

For whatever reason, 2020 was a year when Portland area voters were willing to vote for big tax increases for a wide ranges for services, from dealing with homelessness, to parks, to libraries, to schools and pre-school education. The failure of the Metro measure in this environment shows considerable miscalculation, both in terms of what the measure proposed to fund, and how it proposed to pay for it.




Systemic racism and automobile insurance

Does geographic rating of car insurance amount to 21st Century redlining?

Car insurance rates vary more based on who your neighbors are than on your driving record

The premium penalty for living in a Black neighborhood is twice as large as for being an aggressive driver.

States should ban using small geographies, like zip codes, to set insurance rates.

In the the past several months, there’s been increased attention paid to whether our transportation system is equitable.  Public policy debates have hinged on whether or not building bike lanes is equitable, and whether major regional transportation packages benefit disadvantaged communities.  Huge and hidden subsidies that chiefly benefit higher income households, like nearly universal free parking, seldom get questioned from an equity perspective.

As we’ve stressed, its impossible to understand the equity aspects of transportation without looking comprehensively at the whole system. And one important aspect of transportation is the nearly universal requirement that everyone who drives a car purchase car insurance. After depreciation, insurance, along with fuel, is the largest cost of operating a car.  (If your vehicle gets 20 miles per gallon and you drive 12,000 miles per year, you buy about 600 gallons of gas; at current prices, of about $2.12 per gallon, you spend roughly $1,272 on fuel, which is very much in the same ballpark as the typical auto insurance premium ($1,463) reported by Insurify.

But are car insurance premiums fair and equitable? Insurers vary rates widely depending on what neighborhood you live in, for reasons that are far from clear. Insurance firms claim that geography is correlated with loss, but car insurers are famously opaque about the data and algorithms they use to set geographic rates, claiming that this information represents trade secrets.

But just as with housing discrimination, we can observe the effect of these algorithms in practice.  And there are some consistent patterns.  A growing body of evidence suggests people who live in predominantly Black neighborhoods pay higher prices than otherwise similar drivers who live in predominantly white neighborhoods.

Higher auto insurance rates in Black neighborhoods, even for safe drivers

Most recently, insurance website Insurify analyzed the variables that most influence automobile insurance rates. It found that drivers with good records living in predominantly African-American neighborhoods paid higher rates, on average, than drivers with “incidents” — i.e. a previous crash or serious violation — in predominantly white neighborhoods.  As Insurify’s chart below shows, safe drivers in Black neighborhoods pay about $2,100 per year on average, which is more than the roughly $1,700 to $1,800 that drivers with bad records pay, on average, in predominantly white neighborhoods.

Overall, the rate premium you pay for being a bad driver is less than for living in a Black neighborhood.  Insurify looks in detail at what it calls “aggressive drivers”–those with infractions like reckless driving, failure to yield or stop, street-racing, tail-gating, and hit-and-run driving.  Overall, fewer than 5 percent of all drivers fall into this “aggressive driver” category.  On average, according to Insurify, these aggressive drivers pay about $350 more more than safe drivers.

That’s a far smaller premium than insurance companies charge to drivers living in predominantly Black neighborhoods compared to white neighborhoods. Insurify’s data show that a safe driver living in a Black neighborhood pays about $700 more than a safe driver living in a predominantly white neighborhood.  Insurance rates impose about twice as large a penalty for living in a Black neighborhood than they do for being an aggressive driver. What makes this particularly egregious and ironic is other independent data that suggests that Black drivers are less likely to speed than other drivers.

Insurify’s findings echo what others have reported. There’s a formidable body of evidence documenting these consistent patterns of discrimination.

In 2017, ProPublica and Consumer Reports conducted a detailed comparison of automobile insurance rates charged in different neighborhoods in California, Illinois, Missouri and Texas.  It found that on average, otherwise similar drivers paid 30 percent higher premiums if they lived in predominantly minority zip codes.

A 2007 analysis of insurances rates in Los Angeles by Ong and Stoll found that even after accounting for differences in risk of loss at the neighborhood level, rates varied substantially according to the racial/ethnic and economic status of area residents, with higher rates especially in Black and Latino neighborhoods.

It’s likely that these trends can be amplified by the increasing sophistication with which insurance companies vary rates charged to different persons, based on how they shop. The rates that insurance companies charge to different individuals are essentially opaque to outsiders, meaning that the its difficult or impossible to hold companies accountable for the discriminatory effect of their rate-setting.

The solution:  Aggregate to much larger geographies

The best way to address this problem is to prohibit automobile insurance companies from setting rates based on excessively small geographies.  In theory, it might make sense to vary rates, by say, zip code, if car owners did all of their driving only in the zip code in which they lived.  But people drive over much larger geographies. According to the Brookings Institution, the average trip in large US metro areas is seven miles—a distance that spans many zip codes. While some insurance firms are offering “pay by the mile plans,” most insurance places little if any weight in the distance traveled, and to the extent it pays attention to any geographic factors, it’s only where a car is regularly parked at night, not where its actually driven. Even those who live in low claim zip codes may spend much of their time driving in areas with higher claims.

A good first step would be to aggregate rating territories to much larger areas, counties or even better groups of counties like metropolitan areas.  While people spend relatively little time driving in their own zip code, metropolitan areas are defined by the federal government to encompass commuting sheds, and most people do most of their driving in the metro area in which they work.  There’s clear precedent for this kind of measure:  insurance companies have, for example, been barred from using gender to set rates for auto insurance in California and several other states.

Instead of being based on who your neighbors are, your insurance rates ought to be based on how you drive, and how much you drive.  Some insurers have developed pay-by-the-mile insurance rates, and some have plans that monitor driver behavior, but these are optional and offer modest discounts.  You still get a better deal if you live in the right zip code than if you are a hyper-cautious, low-mileage driver, and that’s not equitable.

Covid & Cities: Reasons for optimism

There are several  compelling reasons—the seven “C’s”—to believe cities will thrive and prosper in a post-pandemic world:

  • Competition: Zooming it in works when everyone has to do it, but if you work remotely while others are in the office, you are at a competitive disadvantage in contributing to and advancing in your work, especially if you are early in your career.
  • Consumption: Cities are about more than work.  They provide us with varied, abundant and diverse experiences and opportunities for social interaction and consumption. 
  • Couples: Young people are drawn to cities because they are  the best place to find life partners.  Once partnered, cities offer more opportunities for both partners to pursue their careers.
  • Careers: Cities are still the best place to find your way in life and build skills, networks and a reputation that enable you to be as successful and fulfilled as possible.
  • Creativity: The serendipitous interaction that happens most and best in cities is what fuels our knowledge-based economy.
  • Camaraderie and Commitment:  Being there matters. Face-to-face in place shows you care, and you’re committed. It’s about being in the room where it happened, not in the Zoom where it happened.
  • Civic commons: Cities are still the place we come together for collective experiences, from concerts and celebrations to rallies and protests. The pandemic has rekindled our awareness of how important public spaces are for enabling us to connect.

To hear many tell it, cities are either doomed or in for a prodigious reset.  They claim the lingering and resurgent Covid-19 virus, and the heightened fear of future pandemics has soured people on working and living in cities. Especially for those fortunate workers and firms who’ve been able to work remotely, the theory goes, there’s no reason to go back to the office ever.

If you’re zooming it in from your desktop, these people think, it doesn’t matter at all where that desktop is.

It’s a short step from that observation to the conjecture that there’s no reason for people or businesses to pay the premium in rents for downtown offices or urban locations.  Businesses can decamp to cheaper suburban offices or small towns, or dispense with offices altogether. And their workers can move to distant suburbs or rural hamlets, mostly working via telepresence, coming to the office once or twice a week, or even more rarely. In this world, who needs cities?

The idea that centrifugal forces are about to tear cities apart is an old one. Back in the Twenties, the advent of electrification prompted urbanist Lewis Mumford to predict that manufacturing would fly from cities to small towns and rural areas, as electric machinery would free manufacturing from the need to build massive factories. At the dawn of the Internet in the 1990s, techno-soothsayers proclaimed the “death of distance” and predicted a wave of decentralization.  Neither of those things happened.

Nobel economics laureate Robert Lucas posed the question “Why don’t cities fly apart?” He observed there’s nothing in the theory of production to hold them together. Land is always less expensive outside cities.  His answer, in the form of a rhetorical question:  “What can people be paying Chicago or New York rents for except to be near other people?” And being close to other people, which is so hard now, is the core reason why we should be optimistic about the future of cities in a post-pandemic world.

No one can say with certainty that Covid won’t somehow be a new Black Death that depopulates whole cities and inverts the economic order, but there are good reasons to be sanguine about the prospects for cities.  I group these signs of hope into seven “C’s”.


To be sure, the Covid pandemic has proved the efficacy of remote work at a scale never before achieved. According to the Dallas Federal Reserve, in the early days of the pandemic about 35 percent of those still employed were working exclusively at home. While many firms have repeatedly postponed the date at which workers will return to offices, the share of workers working exclusively at home has already declined in recent months to about 24 percent of all employed workers. That figure may seem low to some, but that’s likely because college-educated workers are far more likely to be working at home; about half of all college-educated workers were working remotely in April, compared to about 40 percent today.  Over time, we can expect more work and more workers to return to the office.

As those who suffer Zoom fatigue can testify, there are profound limits to the utility of remote work. It’s the kiss-through-the-screen door, lacking the breadth, warmth, and serendipity of in-person interactions. Regardless of how fast the internet connection or the number of pixels, there’s just so much more bandwidth to interpersonal human communication than can be delivered through a computer screen.  But in addition to the limits of the technology, there’s another key factor:  competition.

As long as everyone has to work remotely, every worker is in the same boat.  But someday soon, more and more of us will start working back in the office.  Those who work in the office will be better connected to the business, in the stream of more information; they’ll also be seen as having more commitment, and be able to discern the subtle cues (and bolster the in-person relationships) that privilege them over other workers. In short, those who work at a distance will be less competitive for pay, promotions, responsibility, and opportunity.

As humans, we’ll always attach more importance to the knowledge, processes and decisions that are tempered by face-to-face interaction. You want to be in the room where it happened, not the zoom where it happened, because the really big decisions are going to get made face-to-face, not on a zoom call.

While everyone’s zooming it in, you’re not at a disadvantage. But when if most, or even some of the people are in the office, you have less information, connection, and power than they do. That’s true immediately and becomes even more important over time.

There’s a reason why most of the stories we hear about people leaving the city involve mid-career or older professionals. It’s the same reason why young people, just starting their careers, gravitate toward cities:  cities are great places to develop skills, build networks and establish a professional reputation. Once you’ve established all those things, usually after a couple of decades of hard work, then maybe you can think about down-shifting and working remotely from Boulder or Bend. But if you’re just starting out, small towns and rural areas offer few ways to find the challenging opportunities, hone professional skills, network with peers and mentors, and build a compelling resume.

The importance of “being there” is critically important when it comes to promotions and layoffs. When looking to promote, organizations prize workers who show zeal, enthusiasm and commitment. If nothing else, it is hard to overlook those who are in the office every day. And the reverse is true:  out-of-sight—or off-site—is out of mind, making you more vulnerable to downsizing. Doubt that’s real? In the early 80s I worked for a public agency in Portland. Herb, one of the agency’s long-time employees, had qualified for a six-month professional sabbatical after two decades of service.  He left the office to do detailed research in his field. Meanwhile, in the midst of a recession, the agency had to slash staff.  It was a tough, tense time for supervisors to decide who’d be let go.  It was little surprise that Herb, someone they didn’t see regularly, ended up on the layoff list.

In short, we shouldn’t confuse the temporary, make-do nature of current work-at-home measures with the environment we’ll find ourselves in when we re-open. To always or primarily work at a distance when your colleagues are in the office puts you at a competitive disadvantage. The role of competition is muted now when we’re all forced to work at a distance—but it will assuredly reassert itself in the days ahead.


The underlying assumption of the urban dispersion argument is that the only reason people live in and near cities is to be close to jobs. Cities are more job-rich and productive than rural areas. That’s one reason people prefer cities, to be sure, but hardly the only one. Economists have increasingly pointed to the role that social, environmental, cultural and commercial amenities play in attracting people to cities. As we’ve documented, well-educated young adults have been moving to cities in increasing numbers.  And the evidence suggests that their migration to cities is because that’s where they desire to live; if anything, jobs are following people to cities, rather than the other way around.

It’s hard to see now, in the midst of social distancing, but the big attraction of cities, as economist Robert Lucas described, is being near other people, both to directly interact with them, but then to enjoy the panoply of services, opportunity and activities that exist when there are lots of other people with similar interests in the same area.  The density of cities provides the minimum customer base needed to support the things we value.

And to be sure, many forms of consumption have gone online.  You can get Netflix anywhere, and anything on Amazon is equally available to everyone.  But what’s ubiquitous on the Internet is, by definition, irrelevant to location; it makes no difference.  What does differ is all of the non-Internet goods, services and activities, and cities have unusually rich and diverse arrays of these things.

In the wake of the pandemic, there’s a huge pent-up demand for the kind of personal and social interaction we used to take for granted in the “before times.” Given the slightest hint that it might be safe, people are again flocking to bars and restaurants, to parties and concerts.  While many are foolish to do so now, at some point, as the pandemic subsides, we’ll see a renewed surge to these high touch experiences.  City advantage in retailing and social activities stems from its ability to bundle experiences with the activities of daily life. And in cities, we’re constantly inventing and discovering new experiences. Remember:  the years immediately after the 1918 pandemic were the Roaring 20’s replete with jazz, flappers, and prohibition; it was a period of urban libertine excess by comparison to previous times. People don’t just live in cities because they are close to jobs; they live in cities because of the wealth of consumption opportunities, consuming not just private goods and services, but public spaces, amenities and opportunities for social interaction.  Those advantages will reassert themselves in the post-Covid world, regardless of the scope of remote work.


While consumption may be overlooked, work is still important.  But “work” isn’t just about the ability to do your existing job.  It’s your ability to build an entire career. That means finding a job, and also finding a path to subsequent jobs that offer successively higher pay and satisfaction.  One of the decisive economic advantages of cities is their thick labor markets.  There are many jobs, many different kinds of jobs, and many different employers. This array of opportunities and “ladders” enable workers and firms to make good matches, for firms to find those who can meet their special needs, for workers to find employers who tap their skills. This process is dynamic and ongoing, fueled by the scale of cities.

If we visualize the economy as a kind of large-scale task rabbit or atomistic gig-economy, where everyone is evaluated instantaneously and one-off for a particular short-term task, then arguably, the Internet-mediated labor market may be all we need. In the longer run, though, people are looking for careers, and, at minimum, for the opportunity to steadily progress to better jobs. The wide array of jobs and potential networks in cities make them a far richer territory for pursuing one’s career.

If no one ever changed jobs, if firms didn’t hire new workers, if tasks were well-defined and unchanging, and workers didn’t expect to develop new skills, then perhaps all these interactions could happen virtually.  But workers, especially the professionals who can work remotely don’t work at a single job their whole career. Remote work is, in many ways, a “make-do” solution for a static situation. Likewise, reputation and networks, which still have a substantial geographic component, amplify the value of place.  For a time, a worker or organization can take all of its established routines and do them at a distance. But the process of on-boarding new workers and integrating them into existing teams, and developing new skills and new routines and new products, is vastly more complicated at a distance. Organizations that depend exclusively on remote work will likely find themselves to be less nimble and competitive than those who can more quickly build effective teams with face-to-face interaction.


Location decisions aren’t just individual decisions, nor are they fixed.  Most of us find and live with partners, and cities offer decisive advantages in finding partners and meeting the shared needs of a couple.  Cities are, both because of their size and diverse social and cultural options, “meet markets,” where there are more potential partners to find.  The increasing concentration of well-educated young adults in cities both fuels, and is fueled by the meeting and mating dynamic.

Sex and the City is a real thing.  Cities are both a place to find your partner and a place where both you and your partner can simultaneously realize your professional dreams. As Matt Kahn and Dora Costa have shown in their research on “power couples”—couples where both partners have a four-year degree or higher—cities offer another advantage.  City labor markets are sufficiently large and diverse as to more reliably offer employment and career prospects for both ambitious partners than are smaller and more rural locations. Anyone who works for a college, hospital or large firm located in a small rural town well knows the “trailing spouse” problem:  these locations are at a disadvantage in recruiting talent because both the candidate and their spouse will be looking for long-term employment opportunities.

If anything, the pandemic has created a pent-up demand for people seeking partners. That alone will be a substantial impetus to urban economies as we overcome the Covid virus.


Intense and serendipitous interaction seems to be the key to creativity.  In his book the Triumph of the City, Ed Glaeser cites numerous examples where the concentration of talent in a particular location produces a kind of critical mass of innovation. Other economists who’ve studied the geography of scientific advances in high tech industries find that proximity is a powerful explainer in the productivity and resilience of places like Silicon Valley and Cambridge, Massachusetts.

And urbanist Jane Jacobs reminds us that the ability of cities to throw diverse people together in unexpected combinations is the dynamic force that gives rise to “new work”—the stream of innovations that advance culture and the economy. Economists studying these “knowledge spillovers” have concluded that close physical proximity plays a key role.  The productivity and innovation benefits from being in a cluster of similar firms dissipates rapidly with distance. Rosenthal and Strange cite studies showing that most of the benefits flow from being within a few miles, to less than 1,000 feet, and for some industries, the benefits come from being in the same building or on adjacent floors.

Knowledge creation is a fundamentally social activity.  We make knowledge and apply it to real needs in society by combining what we know with what others know, and through that process coming to understand not how brilliant we are but what it is that we don’t yet know. And again, creativity thrives in many different environments, but it really goes nuts in places where lots of folks are doing lots of things and creating combinations no one ever thought of. That fact has given cities and enduring economic edge, and will continue to do so.

Camaraderie and Commitment

Oregon legislator and one-time forest products industry executive Bernie Agrons once told me, “The world is run by those who show up.”  Being there matters.  Those who are on site have a disproportionate impact on the directions of an event, a company, a movement.  Showing up demonstrates to others in a tangible way the depth of your commitment and builds an unspoken network of bonds and obligations to others.  There is no more profound dismissal of someone’s engagement than to say that “they’re phoning it in.”  In the future, those who commit to being there will have a decisive advantage over those who simply “zoom it in.”

The choice isn’t “home or office,” but “home and office.”  

In the long run, distance work isn’t a substitute for working in the office, lab or factory. It’s a complement.  Everyone who works in the office also has all of the networks and information that are available through Zoom, or the internet, or through whatever web-based application an organization uses to structure its processes.  But the point is that they have these telepresence assets, plus all of the advantages of being in the heart of things. Those who work remotely will always be “remote” in critical ways from their peers and organization.

Civic space

Stay at home orders and the need for social distancing have made us more acutely aware of the vital role that public spaces play in giving us opportunities to associate, to see and be seen, and to mix with others. Barred from gyms and health clubs and bored at being home-bound, more Americans have taken to public streets and public parks to walk and recreate, and to witness signs of life that seem almost normal. And in recent months, city parks, squares and streets have been the place where we have assembled to seek redress of our grievances and to declare that Black Lives Matter. Even as we had to social distance and wear masks, we gathered together in the urban public realm to make our voices heard; we didn’t (and couldn’t) do so simply by clicking like on a Tweet or a Facebook post.

National Public Radio

Public parks in US cities are in part a legacy of the pandemics of the 19th Century, when shared open space was viewed as a way to bolster public health. Cities around the nation, like Oakland, have created “slow streets” that prioritize biking and walking and slow traffic, improving neighborhood livability and opening the public realm to more diverse uses. Restaurants have spilled onto city sidewalks and into spaces formerly used primarily for parking cars. The density and diversity of cities means that there are more opportunities to rethink public space to support these kinds of activities.

Cities are still compelling

Mired as we are in the depths of a third wave of the pandemic, it is hard to think about a post-Covid world.  We have our heads down and faces masked in our socially distanced quarantine foxholes and can’t imagine a time when we can stick our heads up, much less walk around.  It’s still a shocking departure from the blissfully ignorant world where we crowded together, shared the air in cramped spaces, and touched freely. But cities have long endured these kinds of crises and have always flourished when they passed. As Canadian economist Amine Ouzad has shown, despite severe short term shocks, enduring fundamentals ultimately propel urban success.

The pessimism about cities ignores the vital role cities play in bringing us together. Fundamentally, we’re social animals, and we want to be with other people.  Cities are emblematic of more than a grudging willingness to live together.  Rather, they are perhaps the greatest human achievement and the manifestation of how we are most truly human, together.  It’s time to stop regarding cities as consolation prizes and recognize them as the prize.

Achieving equitable mobility: Reallocate road space, price driving

Reallocating street space to buses is inherently equitable

Charging a very high price to cars for using scarce road space promotes equity

Just a year ago, New York took the bold step of of restricting traffic on 14th Street in Manhattan to buses and a relative handful of local deliveries. The improvement in local travel conditions was immediate and sustained. Crosstown buses on 14th Street, which had been the city’s slowest, are now blowing past their schedules. According to the latest evaluation report from Sam Schwartz Engineering (completed just prior to the Covid pandemic) bus travel times between have been cut by 15 to 25 percent, benefitting thousands of cross-town travelers.

M14D and M14A going crosstown at 14th Street, Manhattan.

Because the buses are moving so much faster, more people are riding them. The street itself is a more lively, people-oriented place because there are fewer cars. The buses themselves are more efficient and productive (each driver carriers more passengers per hour, because their buses are both faster and fuller, which ultimately benefits taxpayers). And contrary to the folk wisdom about traffic displacement, there’s been no detectable increase in car traffic or significant slowdown on the side-streets paralleling 14th Street.

This “Miracle on 14th Street” should be an object lesson to transportation planners in New York and around the country. Dedicating more of our scarce and valuable public space to highly efficient means of moving people can make cities more pleasant, can make transit more productive and efficient and can make overall transportation better.

Those are reasons enough for cities to look for opportunities to create their own car-free, bus-only streets. But the experience on 14th Street has important implications for another issue:  transportation equity.

Let’s consider the distribution of costs and benefits of the 14th Street project. To a first approximation, the beneficiaries are bus riders (who get faster travel times) and those who bear the costs are those who can no longer drive there (people in private cars, and those traveling by taxi or other for hire vehicle).  There are second order beneficiaries as well:  people and businesses on 14th Street have a more pleasant environment, the MTA has more productive buses, faster cross town connections probably make the whole city transit system more attractive, and so on.

What 14th Street tells us about the equity of tolling

There’s one other angle to consider. While no one describes it this way, but in effect the ban on most car traffic on 14th Street is the equivalent of a very high toll for cars using the street. (You’ll pay a $50 fine for your first offense of driving on 14th Street, escalating by $50 for each subsequent infraction). So what 14th Street shows us is in a nutshell is the equity implications of the most expensive road pricing regime you can imagine. (Any other toll, including the upcoming congestion pricing scheme for Manhattan will involve much lower charges than the current fine for driving on 14th Street).  So, is this prohibitively high toll for 14th Street “equitable?”

The answer is, of course.  As we’ve stated before, no urban transportation policy is more equitable than measures that enable buses to move faster. Bus riders (as the New York data bear out) have lower incomes than other urban travelers, and importantly, many have few or no options. If you don’t own or can’t drive a car, and can’t afford a taxi or ride-hailed vehicle, you’re stuck with whatever level of service buses provide. And on shared streets, buses take a back seat to general traffic. Giving buses their own lane or their own street produces big benefits for those who are most disadvantaged by our car-dominated transportation system.  Nothing is more equitable than that.

There’s a growing emphasis on “equitable transportation” in policy discussions around the country.  Thinking about equity is long overdue. For too long, transportation planning exercises have simply been about maximizing throughput (of cars) with little or no consideration given to how that affects different people.

But too much of what’s being said about equity is either hopelessly vague or seemingly obsessed with planning processes rather than substantive results.

We’re told that road pricing is inequitable because its regressive (that any given toll is a higher fraction of income for low income people than it is for the overall population). That’s also true, as it turns out, of the price of just about everything, from pizza to coffee to haircuts to parking meters. The common underlying issue in this case is the inequity in the distribution of income, not some intrinsic unfairness in prices. A toll for using a road isn’t a tax, it’s payment for a specific service.

The focus on the supposed inequity of road pricing is routinely divorced from the larger context of the prices we pay for transportation.  We don’t give cars away for free (even though they’re very expensive), we don’t give fuel away for free (even though that’s more of a burden for the poor),  We require everyone who owns or drives a car to buy insurance (which is both expensive, and often proportionately more expensive for low income people). You can’t use a public street unless you can afford a car, afford to pay for fuel, and pay for government-mandated insurance, which in California costs and average of $150 per month and in Michigan average more than $200 a month (and even as much as $400 per month in Detroit).

Engaging in price-fixing to try to address equity concerns is a recipe for failure, and wastefully distributes most of the benefits to higher income households. Car ownership and use is highest among the highest income households; subsidizing roads, car travel and parking confers much more benefits on these high income households than low income households. In effect, complaints about pricing being inequitable work amount to a convenient shield to let the higher income people who are getting an expensive public service continue to pay little or nothing for it.

In what way is a car-dependent transportation system “equitable?”

Ultimately, the under-pricing of transportation leads to a world that is even more sprawling and car-dependent. Distant suburban big-box stores can benefit from the large market for car shopping, while increased travel by car undermines the viability of neighborhood retailers. In contrast, urban environments with transportation systems that allow people to reach common destinations by walking, by transit, and by bicycle are inherently more equitable than car-dependent transportation systems. Equity has to be viewed as a system condition, and not measured at the level of microscopic features transport.

And we can’t divorce equitable transportation from equitable land use.  If we make it prohibitively expensive for people to live in areas that are walkable and well-served by transit, the problem is really one of housing affordability, not of transportation equity. Continuing to subsidize car travel has the perverse effect of creating more sprawl and more car-dependence which makes the transportation system even more inequitable.

In America, we have a caste system for transportation:  There is a privileged class, those who own cars and can drive them, and their is an under-class, those who can’t afford or aren’t able to drive a car, who must endure a slow, expensive and inadequate system of transit or walking to their destinations. Anything than subsidizes or facilitates car travel generally tends to be inherently inequitable, as it works to the advantage of those who have cars, and makes the world a more difficult and worse place for those who don’t.

And that’s why New York’s Miracle on 14th Street is so important:  Reallocating a tiny fraction of Manhattan’s road space from car traffic, to bus travel is a classic example of how to make the transportation system more equitable (as well as more efficient and productive).  It’s the sort of thing, if we’re interested in promoting equity, that we should be doing more of.


The Week Observed, November 6, 2020

What City Observatory did this week

1. Achieving equitable transportation: Reallocate road space and price car travel. New York has recorded a kind of “Miracle on 14th Street.” By largely banning through car traffic, its speeded bus travel times 15 to 25 percent, with virtually no effect on traffic on adjacent streets. Buses now run faster, attract more passengers and are more efficient. And in effect, New York has priced car travel on 14th Street:  You’ll pay a $50 fine for driving there increasing by $50 for each subsequent violation. This miracle illustrates why reallocated and pricing road space is inherently equitable.  First, by making buses run faster (and carry more people) it helps all those who can’t afford to travel by car, which tends to be low income people.

2. Institutionalized racism in car insurance. It’s long been recognized that the real estate and mortgage lending industries did considerable damage to many neighborhoods through the process of red-lining, which dates back to the 1930s. Its now largely illegal to use geographic classifications to effectively discriminate against neighborhoods based on their racial composition. But a very similar practice still persists in automobile insurance. Auto insurance is generally mandatory, and roughly as costly as vehicle fuel, but the amount you pay depends significantly on where you live. A new study from Insurify shows that residents of Black neighborhoods pay considerably more than those in white neighborhoods; so much so that driver’s with clean records in Black neighborhoods pay more than bad drivers in white neighborhoods.

Insurance rating schemes have a higher up-charge for neighborhood characteristics than for bad driving:  the 5 percent of drivers classifed as “aggressive” pay about $350 more than safe drivers; those who live in Black neighborhoods pay about $700 more for their insurance. States could make automobile insurance more equitable by requiring insurers to use larger geographies that correspond to actual transportation markets, rather than using zip codes to effectively price discriminate.

Must read

1. Cities and Restaurants in the Wake of the Covid-19 pandemic.  New York Times economics columnist Eduardo Porter has a wide-ranging, well-informed and provocative analysis of how the pandemic is affecting the restaurant business, and why this is critically linked to the health of city economies. He provides a good overview of the economic literature demonstrating how urban amenities, particularly the opportunities for social interaction provided by bars and restaurants, have been a key to drawing well-educated workers to cities.

The New York Times

The need for social distancing due fight the spread of the Coronavirus has devastated restaurant sales, leading to widespread closures. With the advent of a third wave of infections, and the waning of federal stimulus payments, many restaurants are closing for good.  Porter explores how this is likely to affect city economies, and contemplates how long it will likely take to repair the damage.

2. A 15-minute city is going to require rethinking a lot of our policies.  Paris Mayor Anne Hidalgo has turned heads locally and around the world with her efforts to lessen urban car use and promote greater livability. Her headline initiative, “the 15 minute city” is getting considerable interest. Seattle’s Mike Eliason has an essay at Publicola pointing out that this will require Mayor’s and other city leaders to move beyond simply rhetorical flourishes. It’s easy to embrace the consumption convenience and clever illustrations of the 15-minute city messaging.

But realizing this vision will require bolder, broader and faster action. While cities like Seattle are implementing some slow streets, few if any are in commercial hubs. And simply including the 15-minute criteria in the long list of items to be addressed in future rounds of land use planning lacks scale and immediacy. Moreover, as Eliason argues, realizing actual 15 minute living implies a radical decentralization of a range of activities, ranging from health care to retailing, and will probably only be achieved with considerable increases in urban density, and if equity is to be achieved, much more social housing. It’s good to have a compelling goal and vision for urban growth, but we need to back it up with this kind of comprehensive thinking.

New Knowledge

Shocks vs. Fundamentals. In the wake of the Covid-19 pandemic there’s widespread speculation about the future of cities. The need for social distancing, coupled with much broader adoption of remote work has led some to speculate that urban locations will either decline or revive slowly.

How do cities respond to shocks like pandemics?  The historic record suggests that even severe shocks are mostly transitory, and that once the initial shock has subsided, the underlying fundamentals that shape urban growth (or decline) again predominate. Economist Amine Ouzad looks at two particular examples of severe shocks to city economies: the 2001 terrorist attacks on New York City, and the 1987 Loma Prieta Earthquake that damaged the San Francisco Bay Area.  In both cases the shocks produced considerable property damage, and for a time changed consumer, and investor perceptions of these markets.

Ouzad uses detailed statistics on metropolitan growth over the past few decades to analyze the relative impact of negative shocks—like natural disasters and civil disturbances—and long term fundamentals. He finds no significant impact of either disasters or disturbances on growth.  In contrast, fundamentals, like the educational attainment of the population, the industrial composition of the economy, and whether a region suffers from racial segregation, tend to be consistent and significant predictors of future growth.

Housing market data for New York and San Francisco show that shocks do have observable short-run effects, driving up vacancy rates and holding down prices.  But vacancies and prices tend to quickly revert to pre-shock trends and levels.

Ouzad concludes:

. . . over the span of four decades, metropolitan areas are remarkably resilient to shocks – fundamentals rather than short-run shocks drive long-run population trends. Such resilience of urban housing markets suggests that the benefits of agglomeration play a key role in residents’ welfare; sharing, matching, and learning are key motives that explain the desirability of urban living. These benefits have, over the long run, arguably been greater than the negative externalities of agglomeration. High levels of education, a diversified industrial composition, and racially integrated neighborhoods are keys to the resilience of metropolitan areas.

Amine Ouzad, Resilient Urban Housing Markets: Shocks vs. Fundamentals, Center for Interuniversity Research and Analysis on Organizations, Montreal. Cahier Scientifique 2020S-53, October 2020.


The Week Observed, November 13, 2020

What City Observatory did this week

1. Seven reasons you should be optimistic about cities in a post-pandemic world. There’s widespread pessimism about the future of cities. With the pandemic-induced advent of work-at-home, many people reason that soon there won’t be any reason to go into the office, or have offices, or even cities. We disagree. Not only are cities more than just about access to jobs, the pandemic is masking (sorry!) a deeper truth about how labor markets work. As long as everyone has to work remotely, no one is at a competitive disadvantage in the workplace, we’re all in the same Zoom-limited boat.  But as we gradually return to work, even a few days a week, those who are present will have a competitive advantage, better able to contribute to the organization, tap into informal communication, demonstrate commitment, and build networks.  Competition is the first of seven “c’s” that we think are key reasons to believe that city’s will not just survive, but actually flourish again in the post pandemic world.  The full list is as follows:

  • Competition: Zooming it in works when everyone has to do it, but if you work remotely while others are in the office, you are at a competitive disadvantage in contributing to and advancing in your work, especially if you are early in your career.
  • Consumption: Cities are about more than work.  They provide us with varied, abundant and diverse experiences and opportunities for social interaction and consumption. 
  • Couples: Young people are drawn to cities because they are  the best place to find life partners.  Once partnered, cities offer more opportunities for both partners to pursue their careers.
  • Careers: Cities are still the best place to find your way in life and build skills, networks and a reputation that enable you to be as successful and fulfilled as possible.
  • Creativity: The serendipitous interaction that happens most and best in cities is what fuels our knowledge-based economy.
  • Camaraderie and Commitment:  Being there matters. Face-to-face in place shows you care, and you’re committed. It’s about being in the room where it happened, not in the Zoom where it happened.
  • Civic commons: Cities are still the place we come together for collective experiences, from concerts and celebrations to rallies and protests. The pandemic has rekindled our awareness of how important public spaces are for enabling us to connect.

2. Why—and where—Metro’s $5 billion transportation tax measure failed.  Portland voters resoundingly defeated a ballot measure that would have raised about $5 billion from increased payroll taxes to pay for slate of transportation projects, including an extension of the area’s light rail system. Strikingly, the measure failed on the same ballot where voters approved billions in additional taxes for parks, libraries, pre-school, and school construction. We look in detail at the geography of voter sentiment.


The measure passed chiefly in close-in urban neighborhoods that also voted for an unsuccessful progressive mayoral challenger. The measure was rejected in the neighborhoods that would have been served by the proposed light rail extension, and was also rejected in East Portland, a relatively low income area where the measure’s proponents argued that pedestrian and safety improvements would redress long-term inequities.

Must read

1. The geography of the 2020 presidential election. The standard 50-state red/blue maps used to illustrate the geography of voter preferences are wildly misleading, principally because they make big, sparsely populated (and mostly red) areas seem much more important than they are (electorally).  The Washington Post has a fine antidote for this kind of visual lie:  a map that shows the results for every county, with a dot corresponding to the size of the electorate in each county.

Mapping election returns by population, rather than by acreage presents a much more realistic picture of the weight and distribution of political opinion across the landscape.

2. The economic geography of the 2020 presidential election. Our friends at the Brookings Institution have revisited a theme they explored four years ago:  How does the distribution of the presidential vote compare to the distribution of Gross Domestic Product (GDP).  Preliminary data show that the counties that voted for Joe Biden and Kamala Harris account for 70 percent of the nation’s gross domestic product.

That’s an increase from 2016, when the counties voting for the democratic ticket accounted for about 64 percent of the nation’s GDP. The increase reflects both the growing importance of urban economies to national economic output, and also the shift of a couple of key metro areas, like Phoenix, into the blue side of the ledger. But the message is the same:  voters in the nation’s most productive communities voted Democratic.

3. Rent control in a complex and dynamic market.  The District of Columbia is considering an ordinance to expand its system of rent control to housing built in the last 15 years. (The city currently has rent control for units built before 1976).

The DC Policy Center’s Yessim Taylor has a thoughtful analysis of how the expansion of rent control is likely to affect the housing market. Taylor points out that over time, rent control tends to lead to the removal of housing from the rental market place, either through condominium conversion, shifts to other uses or demolition. Historical experience in DC shows that in the years immediately following the imposition of wider rent control, about two-and-half percent of previously rented units moved out of the rental pool each year.

And rent control has long term effects on the marketplace, with diminished incentives to maintain housing, and limited incentives to build more units, rent control can lead the overall housing supply to shrink, driving up rents for everyone. In DC, its likely that expanded rent control will drive up rents in the “shadow” rental market (where homeowners rent out single-family homes or condominiums).

Over time, rent control would also tend to reduce local property tax collections as it drives down the market value of rental housing. On its face, rent control seems like a simple policy, but as Taylor’s analysis shows, it will have widespread, enduring and counter-productive effects on housing affordability.

New Knowledge

A trend to watch:  Declining central city apartment rents. At City Observatory, we firmly believe the steady, long-term increase in the relative rents commanded by central city apartments relative to those in the suburbs is a key indicator of the growing demand for cities (and a signal we’ve not building enough housing in desirable dense urban neighborhoods).  So we keep a close eye on current market indicators of rent.

One of the best sources of data on market trends is ApartmentList. Rob Warnock, an economist for Apartment List, has a new research report comparing city and suburban apartment rent trends for large metro areas for the first 9 months of the year. His data shows that across the 30 largest markets, apartment rents have continued to soften in central cities, while they have rebounded in suburbs. Since January, suburban rents are up slightly (about half a percent), while rents in their central cities have declined about 5 percent.  (It’s worth remembering that “principal cities” are usually the single most populous municipality in a metro area, and encompasses the entire city limits, and not just downtown or even dense urban neighborhoods).

This pattern holds in nearly all of the 30 markets, where suburban rents have either increased more or decreased less than rents in the central cities they surround. This suggests, in the summer and early fall at least, that urban markets are not growing as robustly as suburbs.  Warnock offers some reasons as to why this may be taking place, partly having to do with the recession, partly to do with pandemic, partly having to do with the mix of apartments in cities compared to suburbs:

There are also differences in the types of apartments available in each type of city; dense urban centers are more likely to contain newer, more-expensive, more-luxurious apartments that are positioned to see more vacancies and steeper rent drops during an economic recession. Meanwhile, suburbs tend to have a greater share of cheaper, lower-density homes that remain in high demand even as renters look to cut costs.

It’s important to note that city apartments still generally command a substantial rent premium compared to suburban apartments, but this divergence from the longer term trend of relatively increasing urban rents bears close watching. It may be, at least for the moment, that apartment supply in cities is no longer being outstripped by demand (deliveries of new apartments, typically years in the making, don’t change as quickly as demand for housing).  And as Warnock notes, some of what we may be observing may reflect a short term movement of renters to ownership spurred by low interest rates.

Warnock’s report has detailed data and interactive charts for each of the 30 largest US metro areas, showing rental trends for central cities and their suburbs from January through September, 2020.

Rob Warnock. “The suburban rent rebound,” ApartmentList.Com, November 10, 2020

In the News

City Observatory’s Joe Cortright is quoted in a Portland Business Journal article, “The departed” chronicling the demise of many long-established local restaurants.



The Week Observed, November 30, 2020

What City Observatory did this week

Black Friday, Cyber Monday, Gridlock Tuesday?  The day after a nation celebrates its socially distanced “Zoom Thanksgiving” we’ll look to see how the pandemic affects the traditional “Black Friday” shopping spree. It seems likely that more retail sales than ever will gravitate to on-line shopping. That’s got many self-styled smart city futurists predicting gridlock from more and more delivery vehicles.

But paradoxically, buying stuff on-line and having it delivered actually reduces vehicle miles traveled, because each mile traveled by a delivery truck wipes out 30 (or more) miles driven by shoppers in their cars.  And because delivery density increases the more packages Amazon or Fedex or UPS deliver—their trucks are traveling shorter and shorter distances between deliveries, making them greener and more energy efficient.

Must read

1. Brookings ideas on climate policy for the Biden Administration.  As the presidential transition proceeds, the policy experts at the Brookings Institution are offering their thoughts on what the federal government might due to reinvigorate the nation’s climate efforts.  We were struck by suggestions from two Brookings scholars, Jenny Schuetz and Adie Tomer.  Schuetz made a strong case for reforming land use and making it easier to build more housing in cities, where living is greener. For too long, federal policy has encouraged sprawling development that’s only made greenhouse gases increase:

Each year, we see more evidence of the devastating financial and human costs of building homes in the “wrong” places—and yet we continue to do it.

Adie Tomer echoes this diagnosis:

Neighborhoods designed at a human-scale—ones with greater proximity between homes and destinations—lead to far shorter trips. The problem is most of metropolitan America stopped building these neighborhoods long ago.

And, as Tomer notes, subsidizing electric vehicles would just repeat the mistakes of past federal policies, incentivizing more sprawl and long-distance trips. Instead, we need to look for ways to build more human scale neighborhoods.  That, and not a technical fix, is a critical choice for the climate.

2. A transportation parable:  Gas lines.  Michael Manville of UCLA takes an historical event, America’s gas lines from the “Energy Crisis” of the early and late 1970s,” and deploys it as a parable of why we have traffic congestion, and what we can do to solve it.  For a time in the seventies, America’s energy policy was stuck by the imagined political impossibility of allowing gas prices to rise in the face of global oil shortages. Rather than allow prices to rise, we resorted to official and unofficial rationing, everything from alternate day purchases, to gallon limits, to simply having long lines at the pump.

National Public Radio.
Briefly, the reason we have traffic congestion today is the same reason that, back in the 1970s, we had gasoline lines. We fixed the price of gas too low, so people had to line up to get it.  Since 1979, we haven’t had price controls on gas, and when there have been price increases, we haven’t had shortages, because demand quickly adjusted. When we price something correctly, people make different choices and figure out ways to use less (or in the case of traffic congestion, travel at different times). As Manville argues, we should be applying the same lesson to our streets:  If we price peak hour travel, we can eliminate congestion.
3.  Transit for all.  Yonah Freemark, famously of Transport Politic, and now with the Urban Institute starts the conversation about what the Biden Administration might do to improve the nation’s transit systems with a back-of-the-envelope set of estimates of what it might cost to increase transit service in all of the nation’s cities to the levels deployed in relatively high performing places.  Freemark uses the National Transit Database to estimate the number of hours of transit service provided per capita in every city, and then works out what it would cost to get every city up to say, Chicago levels of transit provision.  The short answer is, not that much.  Freemark estimates that it would cost $16.7 billion to provide every city of 100,000 or more with that higher level of transit service.  (Hours of service is a bit of gloss on the actual quality of service provided, principally because densely populated places get much better service per hour of transit operation than do sprawling ones, but even with that qualification, this is a great way to begin thinking about what we might do to improve transit everywhere.


Black Friday, Cyber-Monday and the myth of gridlock Tuesday

Far from increasing traffic congestion, more on-line shopping reduces it, by reducing personal shopping trips

Delivery trucks generate 30 times less travel than people traveling to stores to make the same purchases

The more deliveries they make, the more efficient delivery services become

The day after a nation celebrates its socially distanced “Zoom Thanksgiving” we’ll look to see how the pandemic affects the traditional “Black Friday” shopping spree. Last year, it is estimated that online shoppers orders billions worth of merchandise on this single day, and the expectation is this will grow even further this year.

The steady growth of e-commerce has many people worrying that urban streets will be overwhelmed by Amazon, UPS and Fedex delivery trucks ferrying cardboard boxes from warehouses to homes.  One of these jeremiads was published by Quartz:  “Our Amazon addiction is clogging up our cities—and bikes might be the best solution.”  Benjamin Reider notes—correctly—that UPS and other are delivering an increasing volume of packages, and asserts—without any actual data—that truck deliveries are responsible for growing urban traffic congestion.

While there’s no question that it’s really irritating when there’s a UPS truck doubled-parked in front of you—it’s actually the case that on-balance, online shopping reduces traffic congestion.  The simple reason:  Online shopping reduces the number of car trips to stores.  Shoppers who buy online aren’t driving to stores, so more packages delivered by UPS and Fedex and the USPS mean fewer cars on the road to the mall and local stores. And here’s the bonus: this trend benefits from increased scale.  The more packages these companies deliver, the greater their deliver density–meaning that they travel fewer miles per package. So if we look at the whole picture, shifting to e-commerce actually reduces congestion.

                                               Delivering packages and reducing urban traffic congestion!

To the fleets of UPS, USPS and Fedex delivery trucks, we can now add tens of thousands of Amazon trucks. The e-commerce giant has even contracted for 100,000 electric vans from startup Rivian. The rise of e-commerce and attendant residential deliveries has led to predictions that urban streets will be choked to gridlock by delivery trucks. A recent article in Forbes predicted that package deliveries would triple in a few years, adding to growing traffic congestion in cities around the world. Added to this, the flood of on-line shoppingt triggered by the pandemic—there was a 30 percent increase in e-commerce sales in the second quarter—seems to mean we’re approaching delivery truck gridlock on city streets.

In our view, such fears are wildly overblown.  If anything they have the relationship between urban traffic patterns and e-commerce exactly backwards.  The evidence to date suggests that not only has the growth of e-commerce done nothing to fuel more urban truck trips, but on net, e-commerce coupled with package delivery is actually reducing total urban VMT and traffic congestion, as it cuts into the number and length of shopping trips that people take in urban areas. The first point is that, despite the rapid growth of e-commerce, truck traffic has been essentially flat.

Shopping on line substitutes for personal shopping trips and actually reduces traffic congestion

It actually seems like that increased deliveries will reduce urban traffic congestion, for two reasons.  First, in many cases, ordering on line substitutes for shopping trips.  Customers who get goods delivered at home forego personal car shopping trips.  And because the typical UPS delivery truck makes 120 or so deliveries a day, each delivery truck may be responsible for dozens of fewer car-based shopping trips.  At least one study suggests that the shift to e-commerce may reduce total VMT and carbon emissions.  And transportation scholars have noted a significant decrease in shopping trips and time spent shopping.

There are already signs that e-commerce is reducing the amount of travel associated with shopping.  The National Household Travel Survey, conducted in 2009 and 2017, shows a decrease in travel-related shopping.  The US Department of Transportation concludes:

In 2017 people made fewer everyday trips than previously. The decline in travel for shopping and running errands was primarily due to the increase in online shopping and home deliveries.

The decline in vehicle miles traveled per person per day was greatest for younger adults–the group that reports the most frequent use of on-line shopping.  On-line shopping creates some travel for delivery, but reduces the number of consumer shopping trips.  And there are vastly more consumers than delivery trucks; and each delivery truck makes many deliveries.  Professor William Wheaton of MIT estimates that $100 spent on line generates about eight-tenths of a mile of vehicle travel for UPS delivery trucks; while the same amount of consumer spending at brick and mortar retailers generates about 28 miles of vehicle travel.  This means that on-line shopping produces 30 times less vehicle travel than personal shopping.

Source: William Wheaton, MIT Center for Real Estate, 2019.

It’s likely that, just as more fuel efficient cars generated a “rebound” effect of more driving, that convenient and seemingly “free” delivery will prompt more frequent shopping.  But given the scale of the differences in VMT generated by e-commerce and auto shopping trips, there’s still plenty of room for more frequent shopping and still having vastly less traffic.

The more deliveries, the more efficient they become

But there’s a second reason to welcome–and not fear–an expansion of e-commerce from a transportation perspective.  The efficiency of urban trucks is driven by “delivery density”–basically how closely spaced are each of a truck’s stops.  One of the industry’s key efficiency metrics is “stops per mile.”  The more stops per mile, according to the Institute for Supply Management, the greater the efficiency and the lower the cost of delivery.  As delivery volumes increase, delivery becomes progressively more efficient.  In the last several years, thanks to increased volumes — coupled with computerized routing algorithms — UPS has increased its number of stops per mile–stops increased by 3.6 percent but miles traveled increased by only about half as much, 1.9 percent.  UPS estimates that higher stops per mile saved an estimated 20 million vehicle miles of travel.  Or consider the experience of the U.S. Postal Service:  since 2008, its increased the number of packages it delivers by 700 million per year (up 21 percent) while its delivery fleet has decreased by 10,000 vehicles (about 5 percent). As e-commerce and delivery volumes grow, stop density will increase and freight transport will become more efficient.

So, far from putative cause of worry about transportation system capacity—and inevitably, a stalking horse for highway expansion projects in urban areas—the growth of e-commerce should be seen as another force that is likely to reduce total vehicle miles of travel, both by households (as they substitute on-line shopping for car travel) and as greater delivery density improves the efficiency of urban freight delivery. If you don’t need a car for so many shopping trips, then owning just one car, rather than two, or going carless, becomes that much more attractive. A study of the shopping and travel habits in the United Kingdom showed that those who used on-line shopping reduced the total number of shopping trips that they took, suggesting that package delivery stops substitute for personal shopping trips. The study concludes:

Crucially, having shopped online since the last shopping trip significantly reduces the likelihood of a physical shopping trip.

As David Levinson reports, data from detailed metropolitan level travel surveys and the national American Time Use Study show that time spent shopping  has declined by about a third in the past decade.    As Levinson concludes “. . . our 20th century retail infrastructure and supporting transportation system of roads and parking is overbuilt for the 21st century last-mile delivery problems in an era with growing internet shopping.”

So the next time you see one of those white or brown package delivery trucks, think about how many car based shopping trips its taking off the road.



William Wheaton, The IT-Energy Transportation Revolution: Implications for Urban Form

Department of Economics, Center for Real Estate, MIT May, 2019