Abstract: Smart and data-driven technologies seek to create urban environments and systems that can operate efficiently and effortlessly. Yet, the design and implementation of such technical solutions are full of frictions, producing unanticipated consequences and generating turbulence that foreclose the creation of friction-free city solutions. In this paper, we examine the development of solutions for wait time predictions in the context of civic hacking to argue that a focus on frictions is important for establishing a critical understanding of innovation for urban everyday life. The empirical study adopted an ethnographically informed mobile methods approach to follow how frictions emerge and linger in the design and production of queue predictions developed through the civic hacking initiative, Code for Ireland. In so doing, the paper charts how solutions have to be worked up and strategies re-negotiated when a shared motivation meets different data sources, technical expertise, frames of understanding, urban imaginaries and organisational practices; and how solutions are contingently stabilised in technological, motivational, spatiotemporal and organisational specificities rather than unfolding in a smooth, linear, progressive trajectory.
Up until relatively recently tracking the location and movement of individuals was a slow, labour-intensive, partial and difficult process. The only way to spatially track an individual was to follow them in person and to quiz those with whom they interacted. As a result, people’s movement was undocumented unless there was a specific reason to focus on them through the deployment of costly resources. Even if a person was tracked, the records tended to be partial, bulky, difficult to cross-tabulate, aggregate and analyze, and expensive to store.
A range of new technologies has transformed geo-location tracking to a situation where the monitoring of location is pervasive, continuous, automatic and relatively cheap, it is straightforward to process and store data, and easy to build up travel profiles and histories. This is especially the case in cities, where these technologies are mostly deployed, though some operate pretty much everywhere. Here are eleven (updated from 7 in original post) examples.
1. Many cities are saturated with remote controllable digital CCTV cameras that can zoom, move and track individual pedestrians. In addition, large parts of the road network and the movement of vehicles are surveyed by traffic, red-light, congestion and toll cameras. Analysis and interpretation of CCTV footage is increasingly aided by facial, gait and automatic number plate recognition (ANPR) using machine vision algorithms. Several police forces in cities in the UK have rolled out CCTV facial recognition programmes (1,2), as have cities in the U.S., including New York and Chicago (each with over 24,000 cameras) and San Diego (who are also using smartphones with facial recognition installed) (3). ANPR cameras are installed in many cities for monitoring traffic flow, but also for administrating traffic violations such as the non-payment of road tolls and congestion charging. There are an estimated 8,300 ANPR cameras across the UK capturing 30 million number plates each day (15).
2. Smart phones continuously communicate their location to telecommunications providers, either through the cell masts they connect to, or the sending of GPS coordinates, or their connections to wifi hotspots. Likewise, smart phone apps can access and transfer such information and also share them to third parties. With respect to the latter Leszczynski’s analysis (14) of the data generated by The Wall Street Journal in 2011 (4) details that 25 out 50 iPhone apps, and 21 of 50 Android apps transmitted location data to a third party other than the app developer. Of these, 19 of the iPhone apps and 13 of the Android apps did not require locational data as a functional requirement. Half the iPhone and a third of the Android apps did not request consent for passing on the locational data. These locational data are shared with advertisers and utilised by data brokers to create user profiles. For example, ‘Verizon have a product called Precision Market Insights that let businesses track cell phone users in particular locations’ (5). It sells data ‘about its cell phone users’ “age range, gender and zip codes for where they live, work, shop and more” as well as information about mobile-device habits’ including URL visits, app downloads and usage, browsing trends and more’ (5).
3. In a number of cities sensor networks have been deployed across street infrastructure such as bins and lampposts to capture and track phone identifiers such as MAC addresses. In London, Renew installed such sensors on 200 bins, capturing in a single week in 2014 identifiers from 4,009,676 devices and tracking these as they moved from bin to bin (6). The company reported that they could measure the proximity, speed, and manufacturer of a device and track the stores individuals visited, how long they stayed there, and how loyal customers are to particular shops, using the information to show contextual adverts on LCD screens installed on the bins (6). The same technology is also used within malls and shops to track shoppers, sometimes linking with CCTV to capture basic demographic information such as age and gender (7, 8).
4. Similarly, some cities have installed a wifi mesh, either to provide public wifi or to create a privileged emergency response and relief communication system in the event of an urban disaster or for general surveillance. In the case of public wifi the IDs of the devices which access the networked are captured and can be tracked between wifi points. In the case of an emergency/police mesh access might not be granted to the network, however each network access point can capture the device IDs, device type, apps installed, as well as the locational history (9). Such a wifi mesh, with 160 nodes, was installed by the Seattle Police Department in 2013 (9). The locational history of previous wifi access points is revealed because a wifi-enabled device broadcasts the name of every network it has connected to previously in order to try and find one it can connect to automatically. Such data reveals the movement of device owners between locations, revealing the sites of popular spots such as home, work, and where they shop. Beyond a wifi-mesh, anyone with a wi-fi adapter in monitor mode and a packet capture utility can capture such data (12).
5. Many buildings use smart card tracking, with unique identifiers installed either through barcodes or embedded RFID chips. Cards are used for access control to different parts of the building and to register attendance, but can also be used as an electronic purse to pay for items within the facility. Smart card tracking is becoming increasingly common in many schools to track and trace student movements, activities and food consumption (10). Smart cards are also used to access and pay for public transport, such as the Leapcard in Dublin or the Oyster Card in London. Each reading of the card adds to the database of movement within a campus or across a city.
6. New vehicles are routinely fitted with GPS that enables the on-board computers to track location, movement, and speed. These devices can be passive and store data locally to be downloaded for analysis at a latter point, or be active, communicating in real-time via cellular or satellite networks to another device or data centre. Active GPS tracking is commonly used in fleet management to track goods vehicles, public transport and hire cars, or to monitor cars on a payment plan to ensure that it can be traced and recovered in cases of default, or in private cars as a means of theft recovery. Moreover, cars are increasingly being fitted with unique ID transponders that are used for the automated operation and payment of road tolls and car parking. Again, each use of the transponder is logged, creating a movement data trail, though with a larger spatial and temporal granularity (at selected locations).
7. There are also many other staging points where we might leave an occasional trace of our movement and activities, such as using ATMs, or a credit card in a store, or checking a book out of a library.
UPDATE: I’ve had three further ways of tracking people pointed out to me (thanks Linda, Stephen, Jim) and I also thought of one more. Plus I’ve updated method 4 (thanks Paul-Olivier).
8. Selected populations — such as people on probation, prisoners on home leave, people with dementia, children — are being electronically tagged to enable tracking. Typically this done using a GPS-enabled bracelet that periodically transmits location and status information via a wireless telephone network to a monitoring system. In other cases, it is possible to install tracker apps onto a phone (of say children) so the phone location can be tracked, or to buy a family tracking service from telecoms providers (11)
9. Another form of staging point is the use of the Internet, such as browsing or sending email, where the IP address of the computer reveals the approximate location from which it is connected. Typically this does not have a fine spatial resolution (mile to city or region scale), but does show sizable shifts of location between places.
10. Another set of staging points can be revealed from the geotagging (using the device GPS) and time/date stamping of photos and social media posted on the internet and recorded in their associated metadata. This has more spatial resolution than IP addresses and is also accompanied with other contextual information such as the content of the photo/post. Such data can be used in interesting ways such as tackling cyber-bullying by revealing the location of posters (13).
11. Location and movement can also be voluntarily shared by individuals through online calenders, most of which are private but nonetheless stored in the cloud, and some of which are shared openly or with colleagues.
As these examples demonstrate, those companies and agencies who run these technologies possess a vast quantity of highly detailed spatial behaviour data from which lots of other insights can be inferred (such as mode of travel, activity, and lifestyle). These data can also be shared between data brokers and third parties and combined with other personal and contextual information. For example, Angwin (5) has identified 58 data brokers in the mobile and location tracking business in the US, only 11 of which offered opt-outs (in total she found 212 data brokers operating in the US that consolidated and traded data about people, only 92 of which allowed opt-outs – 65 of which required handing over additional data to secure the opt-out). Moreover, these data can be accessed by the police and security forces through warrants or more surreptitiously. The consequence is that individuals are no longer lost in the crowd, but rather they are being tracked and traced at different scales of spatial and temporal resolution, and are increasingly becoming open to geo-targeted profiling for advertising and social sorting.
If you can think of other ways location/mobility is being tracked, please leave a comment – thanks.
Through the development and adoption of technical solutions to address city issues the smart city seeks to create effortless and friction-free environments and systems. Yet, the design and implementation of such technical solutions are friction-rich endeavours which produce unanticipated consequences and generate turbulence that foreclose the creation of friction-free city solutions. In this paper we argue that a focus on frictions is important for understanding civic hacking and the role of social smart citizens, providing an account of frictions in the development of a smart city app. The empirical study adopted an ethnographically informed mobile methods approach to follow how frictions emerge and linger in the design and production of a queuing app developed through civic hacking. In so doing, the paper charts how solutions have to be worked up and strategies re-negotiated when a shared motivation meets differing skills, perspectives, codes or designs; how solutions are contingently stabilised in technological, motivational, spatiotemporal and organisational specificities rather than unfolding in a smooth, linear, progressive trajectory.
Moving applications: A multilayered approach to mobile computing Jim Merricks White, National University of Ireland, Maynooth
Mobile computing plays an increasingly important role in the way that space is experienced in the city. This has political consequences, both at the micro level of everyday production and consumption, and at the macro level of institutional and political economy. While geographers have explored the ontological role which might be played by hardware, software, data and mapping within this spatial paradigm, there remains little concerted effort to explore mobile computing as a technological system which incorporates all of these socio-technical assemblages. By drawing on adjacent disciplines of science and technology studies (STS) and media and communication studies, this essay proposes a multilayered model for such a holistic inquiry: hardware—software—data(base)—GUI (graphical user interface).
By applying this model to a self-reflexive exploration of the taxi service Hailo and the mobility tracking application Moves, I attempt to demonstrate how it might be put to work as a heuristic tool. Following on from my desire to expose and explore the politics of mobile computing, the model is used to draw attention to the networks of power which make up these mobile computing services.
Digital urbanism in crises: A hopeful monster? Monika Büscher, With Michael Liegl, Katrina Petersen, Mobilities.Lab, Lancaster University, UK
Intersecting mobilities of data, people and resources are an integral part of a new digital urbanism. Thrift speaks of Lifeworld.Inc, a new entertainment-security sector driven contexture where people’s everyday activities, movements, physiological data, thoughts, desires and fears are so richly documented in real time that commercial enterprise as well as urban services (transport, energy, security) can dynamically anticipate and shape them ‘just-in-time’ (2011). While this opens up novel opportunities for more efficiency, comfort, and sustainability in networked urban mobilities, it also provides new leverage for mobilizing disaster response. In a ‘century of disasters’ (eScience 2012), where urbanization has increased vulnerability and climate change contributes to increased frequency and severity of disasters, this opens up a perspicuous site for investigations of post-human practices, phenomenologies and ethics. Big data analytics and information sharing for risk prevention and disaster response can exacerbate the unprecedented surveillance contemporary societies practice (Harding 2014), Kafka-eske transformations of privacy and civil liberties (Solove 2004) and a splintering urbanism (Graham & Marvin 2001). At the heart of these transformations is a digital phenomenology of invisibility, immateriality and ‘intelligence’ that does not lend itself to human control. ‘Smart cities’ may depend on smart citizens (Greenfield 2013), but the technologies contemporary societies produce do not support human intelligence. We report from ‘inside the belly of the beast’ of innovation in mobilizing Lifeworld.Inc data for disaster response (Balka 2006). Drawing on experience from collaborative research and design projects (e.g. http://www.bridgeproject.eu/en), we discuss the relationship between lived cyborg practice, phenomenology and ethics in networked urban mobilities. Using a disaster perspective for a disclosive ethical investigation (Introna 2007) does disclose some potentially disastrous transformations, but it also highlights avenues for alternative, radically careful as well as carefully radical design (Latour 2009).
The urban riots of the USA in the late 1960s were some of the most powerful political events of that era. As well as drawing numerous responses from media, the civil rights movement, black nationalists, and groups such as the Situationist International, the uprising also triggered a range of research responses including some of the first computational models of cities. T.C. Schelling’s “Models of Segregation” attempted to provide a logical model for racial segregation and laid much of the groundwork for what later became agent-based modeling. Such work is expressed contemporarily for instance in the riot and insurgency modeling of J.M. Epstein and others. For the state, such events mark a schizophrenic relationship to the contingency of riot and how the algorithms play out in such a scenario. How can it govern events that both demonstrate and excite its power and also undermine it? This paper will propose a tracing of the genealogy of such models alongside a reading of other ways of using urban modeling in relation to the urban riots of that era and now. A parrallel reference point here will be the work of W. Bunge a quantitative geographer and spatial theorist. Bunge consistently argued that geometrical patterns and morphological laws express disadvantage and injustice under contemporary capitalism, and that identified patterns could be remedied by rational methods.
The history of computing, from G.W. Leibniz onwards, tangles with the problematic of developing rational approaches to complex, multi-dimensional problems with a high-degree of what J. Law describes as “messiness”. This paper will examine the ways in which rationality, or ratio, is positioned in relation to urban conflict as a means of discussing the relations between the city and software. The paper will develop a discussion of ratio in relation to questions of abstraction, reduction and empiricism. We are especially concerned to find a relationship between abstraction and the empirical that, by working with the materiality of computational systems recognises, and perhaps works with, the tendency to reduction(ism) but through which modes of abstraction may also work with the highly and complexly empirical.
As part of the Programmable City project, some of our research involves looking at hackathons, hack nights, and their role in city governance. Hackathon’s are increasingly being used by city governments as a way to tap into the creativity of its citizens and make use of open data to help manage the city and address issues that citizen’s may find important.
Meyer and Ermoshima (2013, 3) categorise hackathons into three types: “issue oriented”, centred around a problem or set of problems; “tech oriented”, focused on developing systems; and “data oriented”, where the data sets required to be worked upon are supplied by the organisers. These events are becoming increasingly appealing to entrepreneurs, venture capitalists, and governments who see potential in these events to “reform their structure, renew their methods of functioning, and attract the attention of developers” (Meyer and Ermoshima 2013, 3). The social and ethical benefits derived from the use of open governmental data has seen an increase in “civic hackathons”. Following the American “coding for democracy” movement, Meyer and Ermoshima described civic hackers as “technologists, civil servants, designers, entrepreneurs, engineers – anybody – who is willing to collaborate with others to create, build, and invent to address challenges relevant to our neighbourhoods, our cities, our states and our country… a hacker is someone who uses a minimum of resources and a maximum of brainpower and ingenuity to create, enhance or fix something” (2013, 3).
Thus as Meyer and Ermoshima (2013, 5) note hackathons contain an experimental element of bricolage as well as being collaborative, heterogeneous and constituted by hybrid networks, through which they question divisions of technical experts and others. This creates an innovation and problem solving tool that creates “appropriate conditions to work on a social challenge, to develop software and hardware solutions and to create a sustainable community or ecosystem of technical and non-technical experts, lawyers, activists and citizens” (Meyer and Ermoshima 2013, 6)
With regards to civic hackathons or those encouraged by the government, the belief is that hackathons are methods of connecting communities, developing relationships between communities, governments and tech people, and creates a better city for all. It is reflective of the ideal of letting “the collective energy of the people in the room come together and really take that data and solve things in creative and imaginative ways” (Llewellyn 2012).
But do they actually live up to the promise? Are there any success stories where hackathon’s have produced apps which have made a significant change to the city? Or is it more rhetoric with limited real world application and difference? As part of our research we are attending regular hack meet ups such as Code for All Ireland, Coding Grace and Pyladies to get an idea about the dynamics of the groups, what they do for participants and what participants in turn give back to the group and the broader community.
As our research within Dublin hacking spaces is a work in progress, I will look at another example of hacking oriented problem solving for cities. Let’s have a look at where the government has run app competitions to try and provide solutions to city-based problems. A perennial favourite in these contexts and contests, is public transport. How can an app help negotiate the challenges of the public transport commute?
I’ll use Sydney as an example here. In 2012 the New South Wales (NSW) Government held an app hacking competition for developing public transport apps that used real time data and which were intended to help commuters better plan their journey and create a better public transport experience. The three successful apps, TripView, TripGo and Arrivo Sydney were launched in December 2012. According to the NSW Government, the apps contain data spanning 8,200 stops, more than 1,900 buses and close to 1,200 routes throughout Sydney. It is claimed that TripView is one of the most popular apps with close to 1 million requests each week. According to the Minister for Transport, Gladys Berejiklian,
Real Time information for buses is changing the way public transport customers travel – they are now able to plan ahead, not just by looking at the timetable but by actually seeing where their bus is located on the route, and its estimated time of arrival… This is just one of the many improvements we are making to improve technology across the transport network to make customer journeys easier.
The apps do seem to be useful. On one morning commute, I overheard a conversation about the realtime apps. This fellow commuter reasoned the apps helped you decide when you needed to leave the house in order to catch your bus on time or let you know how long you had to wait, and observed that “It adds value to the waiting experience”. A happy customer it seems.
The apps were the product of a Transport for NSW (TfNSW) App Hot House which was facilitated by PWC’s Digital Change team. The Hot House was held over a two day period and during this time the teams worked with General Transit Feed Specification (GTFS) data and API feed that delivered real time bus information to create prototypes of consumer products for mobile phones which they had to pitch to judging panel consisting of Transport for NSW, industry experts and PWC’s Digital Change Team. It was particularly important for teams to demonstrate that their business model could work in the real world and that it would improve the travel experience. The winning teams received:
First access to the real time information
The opportunity to collaborate with TfNSW stakeholders
Promotional support from TfNSW for their apps
You can find a video of the app competition here:
There was some criticism of Transport NSW’s App Hot House project however. The data was made available to only a selected group of developers and there were claims that this created an anti-competitive market which did not benefit consumers. Transport NSW commented that they are considering opening up real time data to other than the approved app users but need to make sure that their internal infrastructure for data delivery is robust and secure enough to support all subscribers and deliver a consistent level of service to customers. Additionally, the real time information was not available to all Sydney commuters:
The real time information will initially be available in the Sydney CBD, the eastern suburbs, the inner west, southern suburbs, northwestern suburbs, the northern beaches and lower North Shore. The apps contain real-time data spanning some 8200 stops, more than 1900 buses and almost 1200 routes across the Sydney Bus Network.
The success of the real time bus apps prompted the State Government to run a similar competition to develop real time train apps for Sydney trains. RailCorp developed a data feed of real time train locations which was made available to developers. Live information from the trains, received by markers placed at stations, was immediately fed back to customers via apps on their handheld devices. According to the Minister for Transport, Gladys Berejiklian, Transport NSW was
keen to improve the customer experience on the rail network and we know that a Real Time Train App will give customers the information they need to make decisions about their journey, all in the palm of their hand.
But it was not always a government that was promoting citizen use of data to create tools that would help city service users. There were examples of apps made prior to the App Hot House which the government shut down citing the argument that the transport data system could not yet provide reliable and sustainable feed and therefore the apps couldn’t offer a reliable service. There is also some controversy about the non-universal access to the data when the government did eventually provide it in real time.
Prior to the government endorsed apps, some developers had created apps without support from Transport NSW. In 2012 the government and Google announced that transport data would be available on Google maps, joining 400 other cities which were already doing this. However, the information remained static timetable data, updated once a week, and it seemed the government was resisting releasing real-time data to third parties.
In 2011 Ben Hosken created a real-time app from data released by the government for a two week period before an “app day”. The app was a success receiving 200,000 views in the two weeks it was available before Hosken’s was asked by the NSW government to take it down. The government justified their request to remove the app by saying the system was not yet reliable enough to provide a reliable and sustainable feed, while maintaining that they were committed to providing data to developers. A similar reason was also given to another developer, Marcus Schappi, in 2010.
Accessing the data would however require developer’s signing agreements with the state government. According to Berejiklian, “That’s a formal relationship, and obviously if we do embark on that in the future we want to make sure the organisation we are dealing with is going to respect the integrity of the data we are providing them and is going to make sure that they are used in a good way that’s going to help people.”
Hosken and Schappi, were not the only developers who were ahead of Transport NSW’s ideas on the use of their data. TripView developer, Nick Maher was also threatened with copyright infringement, along with other developers who created the apps, Metro Sydney and Transit Sydney, by NSW government agency, Railcorp. This was despite Railcorp not offering a comparable service.
Maher developed TrainView in 2007 and TripView in 2008 had to stop selling both apps after threats from RailCorp. Maher claims that he had asked them at the time whether they had any problems using the data and they had said they didn’t have any. He continued to sell the applications before recently contacting them about some updates. RailCorp then said they had changed their stance with regard to copyright and that they weren’t giving people permission to use their data in third-party apps.
RailCorp has contacted about four developers requesting them to remove from sale mobile applications that breach RailCorp’s copyright over its timetables because these applications were providing out-of-date timetables that had the potential to confuse and mislead our customers…Copyright in all CityRail timetables is owned by RailCorp. Any unlicensed republication of the timetables represents a breach of this copyright. We have not pursued any legal action to date.
Railcorp did however open up its data through a competition in 2013. According to Maher, this was a good thing: “They could have built their own app but instead opened it up to competition which is good because you get more innovation that way.” He added that the concern with the earlier apps had been because the government was worried the data was incorrect. NSW premier, Nathan Rees, had to intervene after a social media storm, and gave his support to the developers.
Intellectual property lawyer, Trevor Choy, said that even though RailCorp was a public service, copyright law was “biased” in favour of the Government and did not make any distinction between information that should be a public service, like train timetables, and private information. “Government agencies are supposed to use their powers wisely, but here they are behaving exactly like a private company preventing a competitor from launching a ‘competing product’” said Choy.
The Sydney Morning Herald quoted Dr Nicholas Gruen, former chairperson of the Australian federal government’s Web 2.0 taskforce, who welcomed Berejiklian and the transport’s department’s move to open up the transport data to Google, noting that people have been attempting to access that data in NSW for five years. He cautioned however, that in Sydney’s case there was evidence bureaucracy was weighing the minister down, and referred to the fact that there “are over 400 cities in the world that provide that sort of data to Google and others, and many of them do it on an open basis”. Gruen suggested that to fully utilise government data, required lowering hurdles to its use which means “making it as easy as possible for the developer’s computer to tap into a stream of information provided by the government without, if possible, stopping to ask permission”.
There is clearly a debate about the openness of the data and the way the government utilises both this data and the skills of developers to manage city systems. This example is not true of all cities as Gruen alluded to. Cities have their own approaches to data, whom they make it accessible to, and to which areas it is applied. The Sydney transport apps example does however provide insight into the complexity and legal issues surrounding the use of data. It is also an example of where the product of the hackathon has been successfully adopted by city users.
These interfaces to the smart city suggest that we’ve traded in our environmental wisdom, political agency and social responsibility for corporately-managed situational information, instrumental rationality and personal consumption and convenience. We seem ready to translate our messy city into my efficient city.
Hackathons and the apps they produce may not therefore serve all or benefit the majority of the community and are often inspired by an individual’s perspective on what they think would make the city better and their own life within that city more pleasurable. While the Sydney transport apps discussed above do seem to have been enthusiastically adopted and have no doubt improved many commuters transport experience, it should also be remembered that not all areas in the wider Sydney region were given access to real time transport information, and that not all transport users utilise the technology that enables the app. When discussing hackathons as part of the wider programmable city discourse we should continue to ask what and who’s vision of the city does it present and how does this shape the city spaces and urban experience for everyone.
We’ll be trying to gain further insight into the role of hackathons in the city and whether they are successful and live up to their hype through our ongoing research and by attending regular hackathons and hack meet ups. We’ll keep you posted on the work that hackathons do in the smart city.
Llewellyn, A. (2012, June 29). The power of hackathons in government. (S. Herron, Editor, & NASA) Retrieved 2 May , 2014, from open.NASA: http://open.nasa.gov/blog/2012/06/29/the-power-of- hackathons-in-government/
Mattern, S. 2014. Interfacing Urban Intelligence. Places: Design Observer. http://places.designobserver.com/feature/how-do-we-interface-with-smart-cities/38443/ (accessed 17 June 2014).
Meyer, M. and K. Ermoshina. 2013 Bricolage as collaborative exploration: transforming matter, citizens and politics. Draft paper for the i3 Conference Cooperating for innovation: devices for collective exploration Telecom ParisTech 2.12.2013 accessed 15/7/14 from http://www.i-3.fr/wp-content/uploads/2013/04/Meyer_conferenceI32013.pdf
This afternoon’s black cab blockade of London comes in response to car ride apps that are changing the character of the city’s taxi industry. While there has been little visible backlash against comparable services in Ireland, it seems only a matter of time before these struggles find their way to our shores.
The market-leading smart phone taxi service in Ireland is Hailo. Self-described as “the evolution of the hail”, Hailo was founded in November 2011 by three London taxi drivers and three entrepreneurs. It launched in Dublin, its second city, in July 2012 and as of mid-2014, provides sporadic coverage across the country with a specific attention on areas of high population density.
To use Hailo as a potential passenger, one needs simply to download and launch their free smart phone application. To use the Hailo as a driver, things are a little more complicated. One must sign up using a geographically specific online portal (such as exists for London, Ireland, New York, Boston and Tokyo). Registration is restricted to licensed taxi drivers such that Hailo is effectively leveraging the screening procedures of existing small public service vehicle (SPSV) infrastructure in Ireland. Of Dublin’s 12,000 registered taxi drivers, I’ve been told that about half are using the service.
In this post I will describe two observations on the role Hailo plays in Dublin: that it competes with existing taxi infrastructure, and that it capitalises on and potentially extends the deregulation of transportation in the city. I will briefly compare the service to Uber and Lyft, and argue why their competition will likely bring the taxi wars to Dublin.
Hailo competes with existing taxi infrastructure
In addition to plying for hire (being in motion and available for hire) and standing for hire (being stationary and available for hire), taxi drivers can increase their number of fares by enrolling to a radio service. When a customer phones a taxi company, this company then leverages its radio-enabled network to source an available taxi driver. Cab drivers working in Dublin have told me that subscription to such a service can cost as much as €5,000 per year. This service is dependent upon a radio communications unit being physically installed into driver’s vehicle, the hire of which is presumably incorporated into the cost of the service.
Hailo, in drawing upon already existing smart phone usage, does not need to install any radio communications infrastructure. This means lower fixed capital costs and lower associated installation and maintenance costs. Furthermore, by having software perform the role of the radio operator, there are presumably fewer attendant labour costs.
These factors lead to a considerably different pricing model. Rather than charge a yearly subscription fee, Hailo is free to install and use, but charges a 12% commission on every fare sourced through the application. In order to compete with these rates, a €5,000 per year radio service would need to direct €41,667 worth of fares to each driver. Understandably, Hailo poses a considerable threat to Dublin’s taxi radio companies.
There is however an important geographical caveat which needs to be made. In cities such as Dublin – where there is a confluence of high population density, a high number of taxis per head and a high usage of smart phones – the benefits of Hailo to both drivers and passengers outweigh those offered by taxi radio companies. In a geographical location where taxi or smart phone use is more sparsely distributed however, Hailo has less opportunity to draw upon existing infrastructure. Where I work in Maynooth – a small university town 20 kilometres west of Dublin – it is very difficult to find a taxi using Hailo. In such locations both spatial scarcity and community loyalty lead me to suppose far less competition between Hailo and existing taxi radio services. In these instances, the volume of jobs rather than rate of commission is the important factor.
Competing SPSV smart phone service Uber, which has been recently valued at a truly astounding $18.2bn1, launched in Dublin in January 2014. Uber operates under a business model which is far more challenging to existing taxi infrastructures as a whole. Rather than recruit taxi drivers exclusively, Uber is open to private hire vehicle (or limousine) drivers as well. Less regulation on the cost of limousine services allows the company to employ a surge pricing model, so that fares can cost considerably more than a standard rate (up to as much as eight times more expensive in rare incidences of peak demand). The cost of an SPSV licence for limousines in Ireland is only €250, compared to the €6,300 for a taxi licence. While cars must still be deemed as fit for such a purpose, and drivers must similarly undergo clearance by An Garda Síochána, Uber hopes that its matchmaking infrastructure for limousine services will allow its service to compete favourably with the taxi industry as a whole. It’s probably too soon to draw any conclusions on the success of the company in Dublin, but you can be sure they are here for the long run. Uber have money to burn and their CEO Travis Kalanick has a combative attitude toward vested interests.
Hailo capitalises on and potentially extends the deregulation of transportation in the city
As described above Hailo is most effective in urban areas where there is already considerable competition amongst taxi drivers. Uptake amongst drivers is dependent upon a pull effect, whereby not using the service would render them excluded from a proportion of the passenger market. There is, as such, a supply-and-demand-like positive feedback loop between the driver and passenger applications. Increased use of the application by one group would be expected to result in an increase in use by the other.
It was not always so easy to hail a taxi in Dublin. Between 1978 and 2000 local authorities in Ireland were entitled to limit the number of taxi licences issued in their area. In Dublin, the number of licences in circulation between 1978 and 1988 was fixed at 1,800. This number was increased slowly through the late 1980s and 1990s to around 2,800 by the end of the decade. Supply was not kept up with demand however. By the late 1990s the cost of purchasing a licence on the open market was as much as €100,000. In line with its commitment to improving Dublin’s taxi services, the Action Programme for the Millennium encouraged the issuance of 3,100 new licences in the city in November 1999. This was not found to be enough however, as one year later, on November 21, 2000, S.I No. 367/20002 lifted licensing regulations in Ireland. This had the immediate effect of devaluing existing licences and subsequent enquiries and legal cases have been undertaken to assess the fairness of this deregulation. The change has had its intended effect however. By 2008, union estimates placed the number of taxi licences nationwide at 19,000 with 12,000 of those being held in Dublin. Locals tell me that these days it is much easier to get a cab home after a night out than it was in the 1990s.
Under conditions where taxi drivers do not have to compete so vigorously for fares (under a licence-restricted, consortium-based or heavily unionised model rather than the predominantly individualised system currently in operation in Ireland), there would be little pressure for them to use Hailo. And as I’ve already argued, no drivers using the service would translate to no passengers using it either. Indeed it is quite clear from various interviews and media appearances that Hailo have made tactical decisions about how and in which cities they introduce their service.
While Hailo takes advantage of such spaces of deregulation, it has been careful so far to adhere to local legislation. The service does not on its own necessitate a further deregulation of the SPSV sector in Ireland. Hailo is not unique however. Competitors include the globe-spanning Uber, Lyft, Sidecar, Wingz, Summon and Flywheel in the USA, cab:app, Cab4Now, Get Taxi and Kabbee in London and Chauffeur-Privé in France. These transportation network companies compete on a range of business models, services and geographical coverage. This marketplace of competing car ride apps has the potential to push against regulations which are in place to control regional SPSV sectors in Ireland.
Consider the following examples from what’s been called the taxi wars.
Uber arrived in London, it’s first European city, in mid-2012. London’s taxi industry consists of over 25,000 hackney carriage taxis serving a daytime population of around 10,000,000. In order to obtain a taxi licence, or Green Badge, drivers must pass to a knowledge test covering a 113 square mile area of the city which is purported to require up to five years of study. Uber, in accepting any driver with a private hire vehicle licence, poses a clear threat to the time and money that London’s taxi drivers have invested in their profession. In recent weeks, The Licensed Taxi Drivers Association has summoned six Uber drivers to court, alleging that the Uber smartphone app is equivalent to a taximeter and therefore illegal under the 1998 Private Hire Act which reserves that right for licensed taxi drivers. In addition, the London taxi-driver’s union lambast Transport for London for failing to pursue the illegal activities which Uber facilitate, suggesting that the state body is afraid of the money backing Uber. In response, Transport for London has sought a binding decision from the high court regarding the legality of the smart phone taximeter.
The way in which other transportation network companies have responded to Uber’s entrance into the London market is important in the case of Ireland. Hailo have announced that they will extend their services to private hire vehicles in addition to taxis. This was news was not well accepted by the industry, with the word ‘scabs’ being gratified on the wall of Hailo’s London office. The company responded on their website, insisting that they are simply following consumer demand. Assuming the move doesn’t too badly damage its position among existing service users, we can expect to see something similar in Ireland, either to pre-empt or combat Uber’s growing market share. Not all apps are going the way of private vehicle hire however. Competitor Cab:app has pushed back against the trend set by Uber and Hailo, asserting publicly its commitment to the taxi industry.
Protests in London and throughout Europe today (June 11, 2014) are in direct opposition to Uber’s perceived infringement upon the taxi industry. In the past, strike action has escalated to involve the destruction of property. In January 2014, Parisian taxi drivers struck in opposition to Uber’s unregulated market competition. Confrontations between unionised taxi drivers and Uber drivers resulted in “Smashed windows, tires, vandalized vehicle[s], and bleeding hands”.
The Irish SPSV sector is admittedly quite different from comparable industries in America, the UK and France. Hailo does not face the same kind of regulatory barriers in Ireland as it would in parts of the US. The National Taxi Drivers’ Union is less powerful than taxi unions in London, and certainly less militant than those in Paris. Hailo is a relatively well behaved company in this sector however. As competitors such as Uber and Lyft seek to expand their more aggressive strategies throughout Europe, it is highly likely, given the success of Hailo and Dublin’s reputation as a high-tech-friendly city, that a real push will be made to establish a foothold in Ireland. Spokesperson for Uber’s international operations Anthony El-Khoury has told the Irish Times: “We see a lot of potential in the Irish market and a lot of demand”. As this market heats up, pressure to further deregulate the SPSV industry will be put on Dublin City Council and the Irish government.
The future of the SPSV sector in Ireland
These two preliminary observations on the role of Hailo in Dublin are commensurate with larger tendencies of state deregulation and market-oriented competition. While the emergence of car ride apps may be positioned as a form of Schumpeterian creative-destruction, there are larger political and economic forces within which these applications should be contextualised. I have attempted to sketch some of these forces in this blog post.
The taxi industry is regulated for a reason. The screening and approval of drivers is important in ensuring accountability and the safety of passengers. Regular metering and Ireland’s nationwide fare system (which was instituted in September 2006) make certain that taxis provide an honest and consistent service that does not gouge rural communities or those in need. While Hailo operates under a minimally challenging model to these standards, it does not exist in a vacuum. By exploiting the freer regulation of limousines, Uber is effectively side-stepping many but not all of them. By doing away with accreditation altogether, Lyft would throw these standards onto the wills of the market. Car ride apps have the potential to put huge pressure on the taxi and limousine business. If other European cities are anything to go by Ireland’s SPSV sector is likely to be forced to deregulate further in coming years. This impacts the price, availability and safety of taxi services, and undercuts the ability of workers in the industry to bargain for fair pay and work conditions. It is important that any changes to the sector be submitted to proper public scrutiny and debate.
June 12 – An article on RTE alerted me to the recent launch of Wundercar in Dublin. The app seems to follow the Lyft or UberX business model, whereby unlicensed drivers give passengers a lift somewhere in the city for “tips”. Drivers are screened by Wundercar rather than An Garda Síochána.
June 22 – The Independent reported on June 20 that Hailo is offering a suite of new services targeting the business sector. Most significant to my discussion here is the introduction of the company’s limousine option to Ireland. Limousine licences are considerably cheaper than taxi licences and there are fewer regulations on fare pricing.