More and more public sector organisations are embracing predictive analytics to improve their forecasting, planning and spending decisions.
As Simon Dennis, central government director at SAS UK, says: "Using predictive analytics, the public sector can combine internal and external data to derive valuable insights that inform decision-making".
Dennis adds that "This type of analysis is used extensively in the private sector, and in the age of austerity it's vital that resources are used in the most efficient way possible, rather than a very simplified approach that sees services cut in arbitrary ways to save costs."
Breaking that down further, here are seven key ways that predictive analytics can help to improve public services:
Evidence based decision-making
Analytics tools allow organisations to allocate their resources based on hard evidence, rather than best guesses and assumptions. This proved crucial when The Higher Education Funding Council for England (HEFCE) was tasked with identifying new areas for higher education centres – part of the government's A New University Challenge initiative, which sought to widen opportunities for higher education in low-income neighbourhoods (known as HE cold spots).
No analytical work had been carried out to establish where these cold spots might be. Rather than relying on information provided by areas bidding for funding for a higher education centre, HEFCE conducted its own research, analysing UCAS application data to identify people who wanted to study locally and whether or not they could afford to live at home and travel to university. The 'cold spots' concept was thus backed by solid evidence, which helped to determine the bids that received funding.
Unlocking the potential of siloed data
Public sector organisations tend to generate reams of data. But without the right analytics tools, it can end up trapped in silos instead of providing valuable insights. For an example of what can be achieved when data is put to work, look no further than a project carried out by South Wales Police. After migrating more than 15 years' worth of crime data to a single location, the force was able to perform detailed analysis on records covering a million plus crimes, as well as everything linked to them – including photographs.
The results of this initiative are impressive. In 2010, Home Office statistics showed that crime in the South Wales Police area had fallen by over 11.8% compared to the 2008-09 figures. Nationally, South Wales Police was also found to be the 8th most improved force in terms of overall crime reduction, with 14,000 fewer victims of criminal activity.
Better collaboration & feedback for partner agencies
As well as giving an organisation greater insight, analytics tools can also help it work more effectively with its partners, as the Croydon Drug & Alcohol Action Team (DAAT) has found. Since automating its statutory reporting with SAS analytics, DAAT has made significant time savings. This extra time has been used to improve collaborative working at a local level.
DAAT also checks the data of any agencies it commissions, providing them with reports on their work. Treatment dropout rates are analysed to determine what stage a client decided to leave, and conclusions about what elements are working well, and what might need improvement, are matched against Key Performance Indicators. In this way both DAAT and its partners are able to ensure their work meets the needs of the local community.
Maximising tight resources
One of the biggest benefits of analytics is that it allows organisations to target their resources for maximum impact. In today's economic climate that ability is more valuable than ever, as it allows cash-strapped public sector bodies to do more with less.
The Vehicle and Operator Services Agency (VOSA), which is responsible for ensuring the roadworthiness of vehicles and the regulatory compliance of fleet operators, is already doing just this thanks to its own analytics programme. SAS advanced analytics tools have allowed it to identify those operators with the highest risk of non-compliance, and resources are then targeted accordingly.
Within a year of using SAS analytics VOSA saw 15% fewer investigations, but yielded a 32% higher prohibition rate. In addition, revenue collection jumped 22%, and customer satisfaction by 20%.
Planning for future service demands
As the population of the UK ages, the impact on the economy could be profound. Hardly surprising then that the Department for Work & Pensions uses some very sophisticated software to predict what might happen to pensioner income distribution in the years and decades ahead, and forecast benefit expenditure.
Using both internal and external data, plus SAS code, the DWP is able to experiment using 'what if' scenarios based on different policies and approaches to pensions, such as the impact of proposed changes on particular groups, individuals or public expenditure. This kind of predictive modeling allows the DWP to reliably and accurately forecast future pensioner incomes, and plan benefit expenditure accordingly, thereby avoiding any shortfalls. No prior knowledge of SAS analytics is needed to use the platform, as all the modeling is carried out within Excel spreadsheets.
The Croydon DAAT also analysis its quarterly reports to predict future behaviours, which enables field workers to take preventative action before problems arise – another key benefit of analytics tools.
Improving customer satisfaction
The insights provided by analytics tools can play a big part in helping public sector organisations to improve customer satisfaction, allowing them to quickly identify areas for improvement, as well as areas where they are doing well. This can be crucial when organisations find they have to compete for limited resources, or justify the importance of their services – as is often the case today.
For South Wales Police public satisfaction also forms part of their People First initiative. A large decrease in violence and robbery, aided by excellent detection rates and a more streamlined approach to data analysis in general, has all served to improve public confidence in the force: a British Crime Survey showed it had risen from 45.8% to 51.5%.
For UK Transplant, a division of NHS Blood and Transplant, it has meant improving tissue matching to reduce the time adults with rare blood groups must wait for a donor organ. SAS analytics was used to investigate over 30 simulation models for the allocation process before one was settled upon. Once implemented, this model sped up the transplant process for NHS patients.
Revealing patterns and identifying anomalies
The need to identify patterns is often a driver for adopting analytics. For example, this type of work has become considerably easier for the HMRC since the advent of Connect, a platform that allows it to easily see all the information it holds on a taxpayer. It analyses and cross-matches around one billion items of HMRC and third party data for anomalies and commonalities. Staff can then target the businesses and individuals that are most likely to commit fraud.
Work that would have taken weeks or months of painstaking research can now be done in minutes, and 'spider diagram' functionality gives tax investigators a visual representation of the taxpayer and all their data links, further speeding up the analysis process.
Graphics and data maps are also used by the HEFCE and London Fire Brigade to explain results and reveal patterns, again highlighting how analytics provides evidence-based intelligence for public sector organisations.
In summary, says Dennis: "There are many areas where analytics can aid decision-making, as these examples show. In addition, more widespread use by other departments and agencies would allow this information to inform their own forecasting and planning, removing the guesswork that leads to inefficiencies and 'hidden surprises'."
Content on this page is provided by SAS, partner on the Transparency hub
