Amazon is one of the most urbanist employers in America

Who says Jeff Bezos doesn’t do corporate synergy? Amazon’s announcement that they’re looking for a second headquarters in North America has exploded the market for hot takes about cities, which is probably giving at least some benefit to the Bezos-owned Washington Post.

While I do think that Philadelphia would be a very good fit for Amazon, I’m also of the opinion that Amazon already has at most two or three candidates already in mind, and that this process will be a shakedown of the preselected winner, much like Tesla did when “considering” where to site the Gigafactory 1 facility. (My personal prediction is that Philly will lose in the final down-select round to either Chicago or Dallas.) So this post isn’t going to be about all that. Instead, it’s going to be about the specific cultural choices that Amazon has made that make it such a catch for a large, self-respecting city.

When one thinks of tech sector offices, specific images come to mind. Sprawling office park campuses laden down with all manner of creature-comfort amenities, from gourmet cafeterias, to on-site dry cleaning, to nerf guns in every open-plan cubicle. Those images come from the specific circumstances of a specific place: Silicon Valley. OK, the nerf guns actually are just pervasive whimsy, but the on-site, corporately-provided amenities are specific reactions to the fact that Silicon Valley’s office parks were built far away from the transit corridors along Caltrain or El Camino Real, so everyone who doesn’t take a tech shuttle bus drives to work, and leaving the office to run errands off-site would gridlock the roads, if a substantial proportion of workers did so. For that reason, many Silicon Valley companies have agreements with their host municipalities, that they will take extraordinary steps to make sure that as few of of their employees as possible will be driving out of their campuses each day. This means that Silicon Valley workers, the villains of terrible and wrong-headed gentrification takes throughout the Bay Area, are actually systemically under-paid, because so much of what would be their compensation is instead funding these amenities designed to keep them at the office for longer.

Amazon is not a creature of Silicon Valley. Amazon is based in Seattle, and has no appetite for this kind of nonsense. They would much rather give their employees more money, and let them out of the office to meet their needs in the city surrounding their campus, which extends to even eating lunch at the proliferating restaurants and food trucks in South Lake Union. Nor does Amazon want to spend much money on that most common of employee perks, the parking spot. They’ve already built garages, and that was expensive enough, thank you. While there are Amazon-operated tech shuttles in Seattle, most bus-riding employees at the South Lake Union campus are taking King County Metro, which has altered its service patterns to cope with the new demand, most recently to accommodate a wave of summer interns coming south from the UW campus.  Amazon has also covered the capital and operating costs of a fourth vehicle on the South Lake Union Streetcar, which enables that short local circulator route to run every 10 minutes instead of every 15.  According to Amazon’s own figures, 20% of Amazon’s employees in South Lake Union walk to work; something completely unimaginable in the stroad-ridden sprawl of suburban California.  A further 35% bike or take transit.

Here is what Amazon has to say for itself about its present headquarters campus:

Even professional urbanists and planners would have a difficult time assembling such a paean to urban living.

So if Amazon is so proud of its Seattle headquarters, why does it need a new one?  Short answer, at least assuming that the answer makes any sense at all, is that it has run out of room to expand the South Lake Union campus.  Amazon has the largest corporate campus in America, by both percentage of the city’s commercial office space and absolute number of square feet, by wide margins, and even with HQ2 it still expects to grow its physical presence in central Seattle by 50%.  But that growth comes with increasing costs, both directly to Amazon, and indirectly by the increasing difficulty in attracting new recruits to the increasingly expensive Seattle residential real estate market.  Amazon is looking to recreate that urban magic in a place where it won’t wind up costing itself billions of dollars through the sheer disruption of its own scale.  It has no plans to slow down its own growth, after all.  As compelling as Matt Yglesias’s case is that Seattle should compel Amazon to just double down on its present headquarters by abolishing single-family zoning and increasing density throughout King County — indeed, Seattle should do those things anyway — Amazon’s best option to stay concentrated in Puget Sound might be to simply send HQ2 35 miles down the road to Downtown Tacoma.  Seattle is not full, but it is filling up as far as Amazon is concerned.

So not only will the eventual winner of the HQ2 beauty contest win 50,000 high-paying jobs, and billions of dollars of construction, net the billions in tax incentives that Amazon is likely to extort, but there will be massive multiplier implications for neighboring businesses and the wider economy.  This will be especially true in the immediate area (i.e. walking range) of the campus. The tax base implications of that multiplier effect will be very large.  This is a far better deal for a host city than your average tech company, or your average corporate headquarters, assuming of course that the “winning” city doesn’t bankrupt itself by overbidding on the incentives.

But needless to say, there just aren’t that many walkable downtowns in big city metropolitan areas.  As much as I think the best site for Philadelphia for an HQ2 development would be the Uptown Business District surrounding North Philadelphia Station, the location that actually meets more of Amazon’s desires is the Schuylkill Yards and the 30th Street Station District.  The Navy Yard, though it may look tempting on the surface, is right out.  (Also, when I said this post wouldn’t be about a Philadelphia bid, I lied.)

For what it’s worth, we already have one tech company headquarters built along urbanist lines in Center City Philadelphia, and when the Comcast Technology Center opens up next year, that will officially become a campus.  Comcast has also been loathe to run too many of its own amenities; while the cafeteria on the top floor of Comcast 1 is legendary, Comcast doesn’t even run the leasing of the food court in its own basement, and the foot traffic to the Wawa across Arch Street necessitated a midblock crossing.  After Comcast 2 opens, there will be 2,000,000 square feet of Comcast offices between the two buildings, and only 157 accessory parking spaces total; the 70 in Comcast 2 coming after the 87 in Comcast 1 were found to be too many, for a building with a direct entrance to Suburban Station.  Even Amazon can’t beat that, at least not in Seattle.  It’s too bad that Comcast’s fundamental business model is “exploit infrastructural monopolies”, although as more people from the NBC-Universal side percolate into senior roles, maybe that will change.  I’ll believe it when TV shows start being produced here instead of at 30 Rockefeller Center.  Amazon at least has “exploit scale to defeat logistical inefficiencies”, which bodes much better for innovation, since logistics will always be improvable.

So, while I think it’s true that your city will lose the contest for HQ2, and unlike Henry Grabar, I think my own city is not exempt, I think that if Amazon’s philosophy of the relationship between company and polis spreads, it will be very good for large, very dense urban centers like Philadelphia.  It will be especially good for the ones (again, like Philadelphia) that are still relegated to second-tier status.

Advertisements

SEPTA’s Regional Rail fare collection contortions are a symptom of a system that is strategically adrift

Ever since last summer’s Silverliner V equalizer beam crisis, SEPTA has been collecting Center City outbound Regional Rail tickets at the platform stairs in the evening rush.  On Tuesday, SEPTA announced that beginning July 10th, it will modify the procedure to punch tickets only, but have passengers retain their tickets and have them displayed throughout their journey.

The reason is simple: without a subsequent on-board check, it is possible for anyone to ride for any length after presenting a Zone 1 ticket to the fare collectors, or for that matter just a Transpass if your train boards at the same platform location as the Airport Line.  And to make the latter loophole even worse, SEPTA’s ambassadors check SEPTA Key cards for Transpass validity rarely to never.  SEPTA remains completely paranoid about revenue leaks from fare evasion, so they are moving to plug the hole now that trains and platforms are no longer quite so overcrowded.

Unfortunately for SEPTA, rush hour trains remain crowded enough that they cannot be easily swept by conductors for ticket checks until a significant number of riders have disembarked, meaning that the ambassadors checking for some form of fare instrument are still necessary to forestall an even more damaging form of fare evasion that was common before last summer: simply carrying one ticket in the knowledge that conductors will not be able to reach you to punch it.  Even riders without nefarious intent simply weren’t always able, or inclined, to seek out a conductor to get their ticket punched.

SEPTA, like most American commuter rail operators, including all of the Northeastern railroads as well as Metra in Chicago, insist on the obsolete and incredibly expensive method of fare collection where multiple conductors per train manually check every passenger’s fare by hand.  This system has been retained virtually nowhere else in the developed world outside of the United States, cast aside in favor of full faregating, or proof-of-payment, either of which dramatically reduces the employee headcount per train, which in turn allows for much less expensive provision of transit service, which in turn means either lower government subsidies and/or lower fares, and usually also allows more frequent and more useful service for riders.  There is a straight line to be drawn between the massive expense of conductors and the sparseness of our train schedules.

SEPTA is at least making a half-measure towards faregating with the upcoming rollout of SEPTA Key on Regional Rail.  The five Center City stations — University City, 30th Street, Suburban, Jefferson, and Temple — will be equipped with bidirectional faregates, requiring a fare to enter or exit.  Platform validators will be posted at the outer stations; when a Regional Rail rider only taps once, they will be charged the maximum fare.  This defeats the disadvantage of the July 2016-July 2017 system, since the default is to the most expensive fare, and can’t be defeated with a cheaper ticket.

But the main advantage of installing faregates will be lost if SEPTA continues to retain low platforms at so many stations, which necessitate the retention of conductors for their other function, raising and lowering the traps for low-platform stations.  SEPTA is still only installing high platforms at one or two stations per year, rather than the mass installation it needs in order to gain the economies from rationalized train staffing, higher train performance from lower boarding delays, and compliance with the spirit as well as the letter of the ADA.  It has not even prioritized engineering for that program, instead wastefully diverting in-house engineering resources to an unnecessary project to extend Market-Frankford Line stations to eight cars long, on the off chance that this President and this Congress decide to do a helicopter drop of money to shovel-ready infrastructure.  Even though MFL ridership is higher than the entire Regional Rail system combined, the potential financial and ridership return to high-leveling entire Regional Rail lines is much higher, and can be done in phases or batches scaled to the amount of money available.  SEPTA does not even prioritize high-leveling stations with high ridership like Elkins Park, which can shave entire minutes from schedules, nor stations like Eastwick or Fox Chase which are the only low-level stations on the Airport and Fox Chase Lines, respectively, which would allow for fewer conductors on those trains.

That said, it is entirely possible that SEPTA’s experiment with partial faregating will fail.  There are notable station chokepoints at all five Center City stations that may simply be incompatible with masses of humanity trying to filter through faregates, in one direction or the other or both.  Time will tell.  If it does fail, the only reasonable option left will be conversion to proof-of-payment, although the platform/door safety function will necessitate conductor overstaffing for decades at current rates.

Prioritizing the reduction of staffing is not an anti-union or even anti-employment position.  Right now the main obstacle to SEPTA providing real transit service on its Regional Rail lines is the chronic, acute shortage of engineers.  The best and easiest source of ready candidates for training as engineers, which is a months-long and high-attrition process in the best circumstances, is the conductor corps.  SEPTA can use literally anyone they can get through the training program to add service for riders.  The next time you are caught in bad traffic on I-76 or I-95, the failure of the Manayunk/Norristown or Trenton Lines to capture mode share away from driving due to their inadequate, once-hourly schedules, will be the main root cause.  We need approximately the current overall staff levels, just shifted, in order to protect out regional prosperity from predictable congestion.

Meanwhile, we are about to witness the worst of all possible worlds: overstaffed trains, multiple redundant fare checks, and a system that cannot get out of its own way and does not have a sense of urgency about getting better.  This is SEPTA in 2017.  May the Great Maker have mercy on us all.

Philadelphia International Airport’s future as a Transatlantic Gateway

There’s been a lot of anxiety in Philadelphia ever since the announcement of the US Airways/American Airlines merger.  Can the new American Airlines find a place in its network and its business strategy for a hub 90 miles away from New York, where it also has substantial operations?  The answer, so far, has been yes, at least so far as to maintain the status quo.  But with few signs that AA is even paying much attention, the anxiety –in a city where an inferiority complex is part of the cultural identity — remains.

Even for someone who dislikes flying like I do, there are good reasons to want a better selection of destinations from PHL.  As much as flying today is an inconvenient nuisance of modern life, it is still the best option for transcontinental travel, and is the only option for crossing oceans.  Businesses decide where to locate offices based in part on the availability of convenient flights.  A good airport with a broad range of destinations is an important regional asset.  Which makes it all the more frustrating that America’s #5 city has to make do with America’s #19 airport.  It’s mostly due to geographic circumstances and not our fault.  Our domestic air market is missing two of the most lucrative markets, New York and Washington, because you can get most of the way to either city, either on the train or driving, in the time it takes to get to the airport and clear security.  International airlines are slow to add us to their route networks, because serving a different part of the country (like Chicago, or Florida, or Texas), has a greater potential profit than doubling (or tripling) down on the Northeast.  And it wasn’t that long ago that our hometown airline, US Airways, was the seventh-largest of seven legacy airlines in the United States. Continue reading Philadelphia International Airport’s future as a Transatlantic Gateway

I am glad you are here with me. Here at the end of all things.

I have tried, over the life of this blog, to strip away as much of my own personal political views, to present a non-partisan view of the topics I cover here.  Most of them are technical or mathematical in nature to begin with, and many others are the subjects of broad left-right consensus, so this has not been a large handicap.  Where my center-leftism has crept in, it is the inevitable result of America’s main political fissure: the Republican Party is (broadly) the party that opposes cities and the people who live in them, and the Democratic Party is (broadly) the party of the urban archipelago.  This makes some forms of judgment on political affiliation inescapable.  I didn’t turn my Twitter avatar into an “I’m With Her” button, but I highly doubt anyone would have been shocked if I had.

The present circumstances are not normal.  The broad threat to the rule of law and American political and social norms ought to transcend partisan affiliation (but has only done so to a small extent so far).  It is without precedent in American history, at least since 1865.   Continue reading I am glad you are here with me. Here at the end of all things.

Should SEPTA add free Wi-Fi to the Regional Rail fleet?

septa-wifi-signal-symbol

With schedules back to normal after the summer’s Silverliner V debacle, thoughts are turning to how SEPTA can regain the ridership it’s hemorrhaged over the last three months of misery (which isn’t quite over yet; trains are still shorter than quota).  Jason Laughlin, writing for the Philadelphia Inquirer, collected a few suggestions on the Inky’s transportation blog.  A Facebook comment thread discussing his article then brought up the idea of introducing free wireless internet on trains.

Now, normally I’m a skeptic about trying to lure riders on to transit with gimmicky perks.  NY Gov. Andrew Cuomo’s risible plan to lure younger riders onto New York City’s slow and unreliable buses with USB charger ports (and also wi-fi), instead of creating bus lanes and enforcing them, is an egregious example.  But wi-fi on regional rail and intercity trains has a reasonably solid track record of being a popular feature, that doesn’t cost the transit agency very much at all to provide.  The usual complaint is that one gets what one pays for; Amtrak’s wi-fi is notoriously slow internet, slower than one can get tethering through one’s own phone, if you’re paying for that feature.  But at the very least, the experience of slow internet, while frustrating, does not drive away any passengers who would not have also been driven away by the experience of no internet. But bad internet does reinforce public narratives about transit agencies being incompetent and feckless, especially about IT (a reputation that is not entirely undeserved, in many cases).

So, should SEPTA leapfrog a generation of mobile internet solutions (which are now mostly just a cell router in a cabinet)?  My comment on the subject was this:

Amtrak recognizes that its wifi system is considered subpar, especially by Acela customers who want the premium service that they’re paying for. So they’re currently working on divorcing Northeast Corridor wifi from the commercial mobile internet, building our their own network of trackside antennas dedicated to handling their own data. The expected cost of this is expected to be $30 million for the trackside network, and $2 million for upgrades to the Acela fleet, all starting last year and running through 2019, at the end of which the bandwidth caps are expected to come off, so Acela riders can Netflix to their heart’s content.

Using those numbers to estimate what it would take to bring the same system to SEPTA Regional Rail, we’re talking about $11.4 million-$18.5 million for trackside network infrastructure (depending on how much piggybacking on Amtrak’s system is possible and how much redundancy SEPTA will have to build), and a worst-case scenario of $6.667 million to retrofit the entire Regional Rail fleet, which I think is realistically an overestimate, since retrofits are more expensive than including features in the original design, and we’re about to turn over most of the Regional Rail fleet. We’re now talking about a $20-25 million project that would take 3-4 years, after ridership will have already recovered as much as it was going to from this summer’s debacle, and not a $5 million project that takes six months to a year, starting to pull back riders out of their cars then. Is it a sufficiently better alternative that SEPTA should take it? Maybe. I don’t know. But I think either is better than the status quo.

SEPTA capital dollars are still scarce, so neither the cheap route nor the expensive route are slam-dunk cases.  1234 Market Street has been ruthlessly prioritizing its core services, as it rightly should.  But if SEPTA can make more money (whether through increased ridership or increased willingness to pay for fares) by introducing Wi-Fi, than it will cost it to provide, then it should do so.  And by the standards of SEPTA capital projects, which range up to $1.1 billion for the King of Prussia Rail project, a $5 million project, or even a $20 million project, is small.  I predict that arguments over the wisdom of introducing Wi-Fi will continue, even long after the introduction of the service on the SEPTA fleet, should that ever happen.

This, too, is America: Burlington Route edition

I’m on the California Zephyr, on my way home from Denver, and the big stories back home are about the continuing lack of Silverliner Vs, and the announcement by Vice President Joe Biden that the Federal Government will be providing $2.45 billion in loans to Amtrak for the next generation of high speed trains on the Northeast Corridor.  Since I saw plenty of Silverliner Vs running merrily along at Denver Union Station on RTD’s A and B lines, and conventional-speed transcontinental trains are both close cousins and as far from sleek, Pendolino-derived HSTs as you can get, I’ve definitely been feeling this weekend as though I’ve been looking at American passenger railroading through a glass, darkly.

The Acela Express has the dubious distinction of having been such a success that it removed fast train travel in the Northeast from the reach of many ordinary people, since even the Northeast Regional trains that are meant to hold the middle of the market are regularly bid up to the sky and/or sold out.  The Avelia Liberty trainsets, which will be the successor to the original Acelas, are going to be an attempt to implement the aphorism that “the main problems with Amtrak can be solved with more Amtrak”.  28 trainsets (a 40% increase), and 9 passenger cars per trainset (a 50% increase), will result in a doubling of availability for high-margin HST seats.  That will give Amtrak some breathing room to continue making money on the Northeast Corridor, although it may suffer from a lack of ambition (9 cars is barely into the range of respectable length by international standards, but will still require expensive alterations to Amtrak maintenance facilities in Boston).  The tractive power on these new trains will be capable of 165 mph, with ambiguous mention of upgradability beyond that (although that might have been marketing targeted at the California High Speed Rail Authority), but the real trip time improvements will come from replacing the Acela’s “flying bank vault” design, and bespoke tilting mechanism, with mature European designs for both crashworthiness and tilting.  The key to going fast will be not going so slow.

Speaking of going “slow”, my current location obviously indicates that I have no intrinsic problem with it.  Conventional trains, whether day or sleeper, have their place, and will continue to do so even after true HSR begins to roll out across the country.  But this trip has been a painful reminder of American national priorities.  Crossing Iowa on the ex-CB&Q, much of the trip is within sight of US 34, a four lane divided highway with virtually no traffic.  Also frequently in sight is “Old Hwy 34”, a two-lane strip of battered concrete that is nevertheless <em>entirely adequate</em> to handle the observed traffic on both roads, or would be with proper maintenance.  Meanwhile, there is plenty of slack intermodal capacity on the railroad, something easily deduced from the evidence that Amtrak is suddenly and consistently running on time or early, both on 5/6 and nationwide.  That’s always a morbid sign that freight traffic is down significantly, in this case from the Death of Coal.

BNSF’s track department really needs to work on their switch installations.  Every one we went over west of Galesburg felt like a cannon shot fired into the bottom of the train.

I still want my 200+ mph HSR from Chicago to Omaha via the Quad Cities and Des Moines.  That’s where the population is, that’s where the travel markets are.  No offense to the CB&Q, BNSF, or the few, but proud, residents of southern Iowa.  And being able to make the Chicago-Omaha hop in 3 or 4 hours instead of 9 or 10 would be game-changing, including for those continuing to Denver or points west.  As amazing a time as I had in Denver, I don’t think it’s worth it to build HSR from there going either east or west to replace the current conventional service, although a north/south corridor along the Front Range would be promising.

The Silverliner V crisis continues; the latest measure will be express buses from select stations, to relieve pressure on crowded trains after Labor Day weekend.

Advanced tips for how to get in and out of Center City during the SEPTApocalypse: Glenside Combined

After two days of travel woes due to the Silverliner V crisis, some patterns have been established in terms of where the worst delays are, and how best to avoid them.  This series will be a listing of the best strategies to avoid the worst.

These will all assume an origin or destination in or beyond Center City Philadelphia. Directions will be for inbound travel, and will be reversible unless noted.  There will be some assumption made that money is available to exchange for time and/or comfort.

Generic inbound

If you are one of many standees on your train at Fern Rock, change there for the Broad Street Line.  Silverliner IVs are not designed for more than one or two standees, and there will be major delays in offloading at Temple University (at least) if you remain on board.  This will keep you safer, and will keep the system from melting down (further).  This instruction does not work in reverse; outbound trains may skip Fern Rock unannounced.

Glenside

Take the 22 bus on Easton Road.  It will be more crowded than usual since it’s one of the best alternatives for the beleaguered Warminster Line, but there hasn’t been a major reported increase in ridership yet, so you should ultimately be fine.  Change at Olney TC for the Broad Street Line.

Jenkintown

Take the 55 bus on Old York Road.  Change at Olney for the Broad Street Line.

Elkins Park, Melrose Park

Take the 55 to Olney, or the 28 or 70 bus to Fern Rock TC.

Fern Rock TC

Do not attempt to board an inbound RRD train here.  If you drive here, please consider carpooling; parking is massively overstretched.

Wayne Junction

Do not attempt to board an inbound RRD train here.  If you use Wayne Junction as a transfer point, consider using a bus route such as the 26 or the 77 to move circumferentially between branches at farther outlying stations.  All of the surface transit routes that directly serve Wayne Junction (23, 53, 75) connect to the Broad Street Line.