WI-FI FOR THE PEOPLE
Deployment of the kiosks wasn’t as straightforward as either the City or CityBridge had assumed. First, as it turned out the energy service and conduit for fiber optic cable to the payphone sites turned out to be degraded and entirely useless for purposes of the LinkNYC kiosks. As a result, CityBridge would have to do more trenching than it anticipated and Con Ed would need to be called upon to provide electrical service to all of the Link sites. That became a pinch point in the speed of deployment because Con Ed had a lot of demands on the staff that provided new service to business sties (requiring street openings), and because of the swiftly developed bad blood between CityBridge and Con Ed staff, CityBridge requests for service went straight to the bottom of the request pile.
Con Ed then decided that CityBridge needed to pay its maximize charge for delivery of service because the LinkNYC kiosks were “permanent” (like buildings), and that each kiosks would have to be individually metered and billed (unlike street lights). This was a large unanticipated additional cost to CityBridge, which it challenged unsuccessfully at the State Public Service Commission (a decision incorrect on the legal merits, in my judgement).
Siting the kiosks also turned out to be more complicated than anyone expected – because the kiosks were larger and differently shaped from phone kiosks they couldn’t just be replaced one-for-one. Also, site conditions had changed from the time the phone kiosks were installed, and by the time CityBridge surveyed sites for Link kiosks. Plans for each site had to be review by DoITT staff, and then inspected by DoITT inspectors (who were retrained from having been payphone inspectors for decades) for compliance with various siting criteria – particularly clearances from other street furniture. Pre-COVID outdoor cafes proved to be a significant obstacle to Link kiosk siting. CityBridge never made the deployment targets (each Borough had its own targeted minimum number of kiosks to be deployed each year) required by its franchise agreement.
Most of the publicity about the Link program during its yearly years arose out of concerns that they were a magnet for the homeless. There was widespread publicity that homeless individuals were watching porn on the internet tablets built into the kiosks. The charging ports were also seen as drawing folks looking to recharge their phones (although the question of how it was that apparently destitute people had handheld devices and monthly phone service plans was never raised or explained). In a number of places in midtown Manhattan, homeless people did congregate around the kiosks, it seemed to me for reasons unrelated to their functionality that I couldn’t figure out (we also found people sleeping under derelict telephone kiosks that had no functionality). For a while, we were reasonably successful in working with the outreach teams under contract to the City’s Department of Homeless Services to get services to people congregating around the kiosks. But, fundamentally, the situation wasn’t a problem generated by the LinkNYC program, but by the Administration’s failed multi-billion dollar strategy for dealing with street homelessness (but that is the subject for another long blog). The basic fact was that the LinkNYC kiosks did not cause homelessness.
I spent a fair amount of my time when working for DoITT trying to understand who was using the Wi-Fi and the other features of the kiosks and how they were using them. CityBridge provided data on usage (which is available on the City’s open data portal) which I reviewed monthly. I also tried to spend a couple days a month wandering around city neighborhoods, checking on their functionality (as did our terrific team of six inspectors), and observing, a al my guru, William H. Whyte, how people were using them in real time. I made sure my site visits were in neighborhoods far from the city’s center and were ones I hadn’t been to before.
The system was highly utilized – with millions of users and millions of terabytes of annual downloads. I concluded that the bulk of the use in Manhattan was by tourists and other visitors. Most New Yorkers have data service. That’s not to say that making the service available to that user group isn’t of value – but it wasn’t exactly advancing the goal of closing the supposed digital divide. What was interesting was how the kiosks were being used outside Manhattan. They were becoming a kind of social center. I saw people setting up chairs near them, using the Wi-Fi or charging their phones. Businesses put out tables and chairs in front of their storefronts for Wi-Fi users. I observed people using the telephone functionality – making very long calls, much to my surprise. The data showed that the top websites accessed and phone numbers called had to do with contacting social services – finding out about missing SNAP and other transfer payments. Clearly, the Link kiosks were becoming an important resource in otherwise underserved communities and had a great deal of value there. They were enlivening the sidewalks in commercial corridors in Brooklyn and The Bronx. This is something that should continue to be a focus of the Link program.
At one point I recommended, following the Holly Whyte formula, that we work to place movable chairs and tables near the kiosks, since many of the complaints about their use were about people sitting near them on the ground. My thinking was that this would “normalize” this use and support the vitality of commercial corridors around the city. This suggestion was not well received, despite the near universal acknowledgement among urbanists of the utility of movable chairs to animate public space, as evidenced by their success in Bryant Park.
MORE PROBLEMS ARISE
In 2017, CityBidge began to complain that it wasn’t making its advertising sales targets, in part because of deployment delays and in part because ad sales weren’t living up to expectations. As a result, the franchise agreement was being amended when I arrived at DoITT around that time, to provide a slower roll out of kiosk deployment and a deferment of payment of a portion of guaranteed minimum payments. A forensic audit of the financial records of CityBridge was performed for the City by a highly respected CPA firm to determine its financial condition. All were assured that the terms of the second amendment would solve CityBridge’s financial problems, based in part on representations made by CityBridge and Sidewalk Labs staff, and in part on the audit results.
Immediately after the approval of the amendment to the franchise agreement (by City Hall, the Office of Management and Budget and the City’s Franchise and Concession Review Board, made up of representatives of the Administration, the Comptroller and the Borough Presidents. The FCRC is an interesting vestige of the City’s powerful Board of Estimate, made up of the Mayor and Borough Presidents and which once controlled the City’s zoning and franchising functions until it was found to be unconstitutional on the basis of being undemocratic – given that each Borough President had an equal vote), CityBridge requested of the City permission to amend its privacy policy to be able to collect the Mobile Advertising Identifier generated by peoples’ mobile phones and use it in order to increase the value of its ad panels (and by the way, you can turn off the generation of an MAI by an iPhone by going into its privacy settings).
There was some degree of bad faith in this request, since the claim of CityBridge was that the ability to collect the MAI was essential to the economic successful of the franchise – a success that was supposedly assured by the adoption of the Second Amendment to the Franchise Agreement (the first amendment involved approval of a change in the members of the consortium making up CityBridge). The request caused tremendous ructions within the De Blasio administration which had privacy hard-liners in charge of its electronic privacy policy. The Administration had also gone through a laborious process of gaining the approval of the electronic privacy advocacy groups, like the New York Civil Liberties Union (also ardent privacy hawks) to not oppose the LinkNYC program. The reality of the privacy protections built into the franchise was that the Wi-Fi use of the LinkNYC signal was more secure and had higher standards of privacy than that provided by the any of the cable companies’ internet service to homes. The Link program collected the email addresses of the users in order for them to sign up for using the program, and nothing else about the users. Those email addresses were prohibited to be used for any other purpose or to be shared with any third party. The privacy protections were, as far as I could determine, rock solid.
Much was made of the potential for privacy invasions from the kiosks. They had three cameras. One was for use with embedded tablet for telephonic and emergency communications, and two on either side of the kiosks were for the monitoring and prevention of vandalism of the kiosks. To the best of my knowledge those side cameras were used only occasionally, and the videos generated were kept for a maximum of seven days and then erased. The kiosks also had a range of sensors that, to my best knowledge, were never used. These sensors had the capacity to collect pedestrian traffic, vehicle traffic, weather and seismological data – all which might have been of great value to the city and none of which were ever used. My understanding was that since the data was found by Sidewalk Labs to have little economic value (nobody wanted to pay for it), the sensors were never turned on. There was nothing the slightest bit “big brotherish” about City use of LinkNYC as far as I could tell, and I would have known.
I am actually something of a privacy dove. The MAI had the potential for huge value for both CityBridge and the City. It was anonymized, so it had a very limited impact on user privacy, but it upped the efficacy of the value of the ad panels. Most interesting to me was that the MAI could enable the Link system to relate the passing of an ad by a pedestrian to a sale to that person minutes later through their mobile phone. Let’s say an ad for Brooks Brothers was shown on a panel and an individual walked by it. If within a short period of time that individual went into a Brooks store a purchased a shirt, the Link system could get a commission on the sale. This was the kind of effectiveness/outcome metric that is so difficult to determine in the out of home advertising market – and would have provided the system with a huge competitive advantage. It would also have allowed Intersection to precisely determine how many viewers each ad panel had, something of great interest to advertisers, and would enable a more precise value to be placed on each ad.
Finally, much of the internet advertising sales world is driven by algorisms and a complex computerized marketplace by which ads are sold through instantaneous electronic auctions. Billboards, obviously, unlike internet banner ads, can’t participate in that electronic pricing system and as a result, are limited in their value. Tapping into the MAIs captured by the Link system would enable those ads to be sold through the electronic auction marketplace. The system would know how many people were seeing an ad at a particular location at a particular time and could auction off an ad at that time and place. Without the MAI data, the Link ads were not able to easily participate in this process (although I did argue internally that using data available from smartphones collected by Google without the MAI, such sales were possible, without success).
The City’s Chief Privacy Officer and the DoITT General Counsel were adamant. No way was CityBridge was going to be able to amend its privacy policy to collect the MAI under this administration. This was communicated to CityBridge quite clearly in June of 2017.