As PhD researcher working on projects somewhat similar to Kustsuplus, the debate around its demise has some familiar themes. Kutsuplus was an Uber style service in Helsinki allowing citizens to summon a minibus using a smartphone, developed by Aalto university as part of the city’s commendable public transport aims. It has lost its funding.
Evgeny Morozov, long time critic of Silicon Valley companies, painted this event as part of a larger narrative: Californian tech behemoths wriggle out of paying their taxes, in turn the governments cannot afford to pay for public sector digital innovation projects. The result is that we’re stuck with a for-profit San Francisco monoculture.
Stian Westlake responded by pointing out that corporate tax take as a percentage of GDP hasn’t fallen, and suggests instead that the problem lies with the difficulty of public sector innovation more broadly.
The PhD program of which I’m a part (The Creative Exchange) is looking at ‘digital public space’, and could easily have attempted a project like Kutsuplus. The type of design research we do has a history of democratising digital technologies going back to the 1970s. The project I’m working on – localnets.org – is a web app to help local government understand communities using social media data. I certainly relate to some of the issues mentioned in Stian Westlakes’ blog.
This post mulls some of my experience of public sector digital innovation, and the necessity of getting it right. The argument has been rehearsed a lot before, but for clarity here’s some reasons why you might want to build something reminiscent of a California startup in the public sector…
Does getting in an Uber, or staying at an AirBnB make any of the parties unsafe? Will these arrangements lead to workers becoming every more precarious, as they are no longer afforded official employment status? Will AirBnBs be used as brothels? Are Uber drivers lonely? Does AirBnB undermine public housing?
There’s plenty written about these issues, but most of the concerns can be placed in one of two categories. Firstly, the idea that these services bypass regulations agreed at a local level, and however imperfect that local democracy might be, that doesn’t seem right. Secondly, these Silicon Valley companies have stratospheric valuations – Uber is currently the most valuable private company in the world. These valuations are probably based on the presumption that they will become monopolies in the long run, and will then start charging rates that are unfair, but also highly lucrative for shareholders.
One answer is to build your own tech platform locally and make it function according to whatever the local rules are – for example ensuring taxi drivers receive minimum wage through the design of the software. That solves the monopoly problem too – any profits from the system can be redistributed to citizens.
Alternatively, you might just find Uber, or another innovative company, inspires some new system for public services.
Who could help a public sector organisation build such a thing? It’s hard to turn to the creative energy of entrepreneurs, who are enticed by the prospect of billion dollar valuations rather than the more modest rewards of a well executed public infrastructure project. Universities are supposed to be centres of expertise – perhaps they can help instead?
Kutsuplus is an example of this model, a collaboration between Aalto University Computer Science and Engineering Department, the design Faculty and Helsinki’s transport authority. I’m not that familiar with the project – many of the articles I can find are in Finish – but the project’s demise made me reflect some of the complexities I’ve experienced.
University based projects are often supposed to serve an incredible diversity of goals. My work aims to provide a functional public consultation tool for community-focused institutions. While I’m doing that I’m supposed to be doing research of sufficient quality to warrant a PhD. At the same time, we are encouraged to have an eye on commercialisation. Through all of the this, leaders of the PhD program will be seeking to find evidence that our work has invigorated an innovation ecosystem. Alongside doing whatever project we are doing, we also supposed to be documenting transferable knowledge – how could future projects be informed by what we are doing? It’s not that any of these things are a bad, it’s just a lot to think about – contrast all this with the laser focus of a startup.
I don’t know if this was a problem for Kutsuplus, but I notice one first papers on Goolge Scholar about Kutsuplus is the snappily titled Intermediation For Eco-Innovations: Aalto Centre For Entrepreneurship In The Context Of A University Innovation Ecosystem, which is not about Kutsuplus’ core goal of running minibuses and perhaps a distraction. Anecdotally, I’ve heard that Streetlife, a UK hyperlocal startup, refuses all requests for access by researchers, presumably to avoid loss of focus that might entail.
I’ve found it’s difficult to keep up with communicating in two quite distinct ways. For academic audiences you need papers, workshops, posters, conferences, to develop written communication in an academic style and carefully referencing a wider literature. Unsurprisingly, there is also a whole other language required for explaining my product to the public sector, with polished graphics, websites and succinct explanation. Communicating your message twice is another burden for a project that seeks to do socially minded tech from within the university.
The commercial sector’s demand for developers has driven their wages upwards, meaning it’s hard for academic projects to afford experienced developers, or, even more lavish, software development agencies – yet this is probably exactly what’s needed.
As a result, many research projects get so far with informal development work carried out at below market rate by friends or colleagues, or just by individual freelance developers, who inevitably move on or get distracted mid way through, leaving the investigator unable to complete. Even computer science departments might not have the relevant expertise. Looking at the publications from Kutsuplus, I notice a lot of them focus on the algorithmics for routing the buses, which is exactly the kind of problem that computer scientists love. Delivering a beautiful mobile web experience, for example, isn’t so likely to be an academic computer scientists’ skill set. To build consumer facing tech you might have be able to afford to buy in that kind of expertise. (A point reinforced here in the context of data science.)
The Economist poked fun at the idea of the public sector doing digital, but we can see from the success of UK’s Government Digital Services that it is possible for first class web services to be developed outside of Silicon Valley.
Regulating monopolies and dealing with technologies that undermine existing legislation are classic reasons for state intervention – look how fast the government moves to change the law when someone invents a new kind of recreational drug.
On the one side, handing ownership of important infrastructure – and monopoly profits – to corporations thousands of miles away isn’t going to be politically acceptable for ever and in every context. After Snowden, China’s reluctance to open the door to US tech companies seemed prescient. On the other side, businesses like Uber are inspiring new approaches to public services like Kutsuplus – both of which could have worked together towards Helsinki’s transport goal of reducing car ownership.
Academia can take on projects of any scale, think of ambition of the research at CERN. Universities are also responsible for all kinds of important software projects that we can’t imagine being without, like the operating system running your iPhone. There is no reason a university can’t play a part in developing digital public transport services – at a recent D-CENT event there was lots of ambitious talk to that effect. From my perspective, more realism about the resources required and a narrower set of goals could only help.