Jon Poole is Research and Intelligence Manager at Bath and North East Somerset Council, and works with the IPR on the project Connecting data across public services in Bath & North East Somerset
In 2012, the Harvard Business Review, seemingly without irony, declared the Data Scientist the “sexiest job of the 21st century”. This showed that complex data analysis had truly arrived as a tool of the future. That there is more data than ever before is surely the most self-evident truth facing anyone working in any field, let alone one in which data is an important currency. This creates opportunities, possibilities and a popular narrative that something should be done about it all. Analysts are asked to anticipate this demand and help decision makers understand the opportunities, challenges and risks in all this data.
In local government, every service (from abandoned shopping trolleys to zoos) adds realms of transactional data to the corpus of local knowledge on a daily, if not instantaneous, basis. Cities are already using algorithms to predict crime, re-route traffic and think about social care risk. It can sometimes feel as if any lamppost which isn’t wi-fi enabled is a waste. This execution has not been consistent, and the ability of city managers and local governments to realise the analytical opportunities inherent in this data is incredibly varied, with a number of different models emerging.
Analysing data in local civic spaces
Traditionally local government has run in-house analysis, often employing qualified specialists in fields such as demography, market research or public health. Teams are normally federated across an organisation or run through centralised units, and the relationship with IT services (the facilitators of access to data) is often unstructured. In-house analysis is currently characterised, alongside much of the sector, by increasing demand in the face of reductions in resources.
Some cities have already been thinking about the opportunities inherent in this more applied form of data for some time. Most famously, the New York model of an 'Office of Data Analytics' created a central space, outside of normal service delivery, to apply data science to public problems. The successes of this programme are widely documented. It is notable how similar in branding and involvement, how similar the look and feel of these narratives is, to those adopted by larger consultancy firms. Within the UK, NESTA have notably adopted the model and are rolling it out amongst a number of local authorities.
The use of consultancies is nothing new in the sector, and although expensive, is often preferable to employing rare skills in an organisation. Many of the larger consultancy firms are well established in the field of data analytics, often in association with an emerging ‘smart city’ agenda. These activities are often typified by a ‘black box’ approach to methodologies, accessed on a proprietary and paid-for basis. In some cases academic organisations themselves are incentivised to replicate this model, developing intellectual property and new methodologies with an eye to relicensing.
Finally, there are those areas adopting a more open model; the Data Mill North, originally in Leeds, but expanding fast; the Trafford Innovation Lab; and closer to home, Bath:Hacked have all provided a space localities can release data to empower the wider community to use local data. These models focus on an open data infrastructure, with Bath:Hacked being notable as having a local datastore owned entirely by the local community, rather than government itself.
It’s clear when looking at the exemplars in this area; that these innovations are often the preserve of larger or well-off cities. Innovation scales well, and those who have the money can develop, brand and sell their concepts and ultimately sustain the work through ongoing external investment. Those who can’t, won’t. Data ‘haves and have-nots’ may inevitably develop.
As public sector austerity remains one of the overriding forces of contemporary urban governance, there is a question about how it is possible to realise these benefits in places which might not have the advantages of scale or funding. It is too easy to finish any summary such as this with a passionate plea for government or some undefined higher power to provide funding for good data works, but in practice that is likely to be little more than an optimistic hope; instead we must continue to question if there may be another way.
An alternative way - towards partnerships of data and analytics?
In Bath and North East Somerset we think we have an interesting local approach, one which could perhaps scale without relying on ‘black boxes’ or extensive investment.
In 2013 a collaboration was established between the Council, Bath and North East Somerset Clinical Commissioning Group and the Institute for Policy Research. The aim of this collaboration was to try to help the Council access academic skills and knowledge and realise actual policy impacts. Simultaneously the Council supported the development of Bath:Hacked.
This collaboration is based open principles of open data and open source and makes a virtue of the varied skills held across our civic space. No one organisation has the ownership of the entire process, nor are the benefits realised by one partner alone.
The collaboration has been successful in attracting external funding for projects as wide-ranging as school’s energy efficiency through to uses of new technology in Social Prescribing services. This approach has helped all partners use their analytical resources differently without requiring significant additional investment.
Such endeavours also come with some degree of risk. Our work is reliant on the passion and dedication of individuals, within organisations and in the community at large and is vulnerable to people changing jobs, leaving the area or finding themselves with less spare time to commit to this work. In addition, as the collaboration emerged from fixed term funding, the work itself was constrained by time. In addition we have faced various issues of timeliness, access and relevance in ensuring that analytical outputs reach their desired audience.
The future of collaboration.
For 2017 the collaboration will look to move onto a more formal partnership footing and examine options to do this. They seem to range from increasing the number and range of informal networks to a structured ‘front door’ model.
Focussing more on an informal footing means we increase the sustainability of the collaboration and its reach, but increases the reliance on goodwill and personal commitment from individuals involved.
As we progress, the main question remains whether this model can scale, not just in terms of engaging beyond the traditionally active data community, but also whether it might provide a model for other areas, potentially those without the scale or resource to attract high-cost innovation.