Most of us have probably seen Steven Spielberg's 2002 sci-fi movie Minority Report, starring Tom Cruise. In an “absurd” future society, a special police unit hunts future criminals before they commit their crimes and the illegal activity takes place. The police use sci-fi tools, such as “pre-visualisation,” to stop the lawbreakers of the future before anyone is hurt. The film's story takes place in 2054, but the plot is actually not so absurd. Monitoring agencies in 2018 – in practice – could apply similar principles to improve monitoring work by analysing and joining accessible data with the assistance of machine learning.

Since the start of the financial crisis in the previous decade, Nordic countries have been keen to create initiatives and reforms that contribute to stimulating economic activity, thereby creating growth and better opportunities to maintain our welfare models. Concurrently with the fiscal policy initiatives, there has been a persistent focus on increasing productivity in the public services in order to maintain and increase services on offer while demand for them is growing.

Public sector tasks such as monitoring and control of work environments, food safety, disbursement of welfare services, etc. have also been under political pressure to increase productivity and efficiency and to deliver better results.

Together with the implementation of large reforms in areas related to employment, education and social/integration areas, the focus has also shifted more towards the effects of the policy work.

From the political side, the last few years have been focused on quantity in the control and monitoring work. It seems as if the thinking has been “the more monitoring, the better.” As the focus grows on the effect of the initiatives, the quality of the monitoring and control task becomes more important. For example, the Minister for Employment in Denmark has focused on improving the control and monitoring of unemployment benefits in 2018. The question is, however, what does it mean to improve the control and monitoring task?

The auditors general have all said the same thing in the three Nordic countries: The risk-based control can be improved both in regard to method and cost-effectiveness. Too many scheduled inspections are not really risk-based, and are only scheduled based on a rule-based model and method.

‘Pre-visualisation’ with analytics

In our experience at SAS, if you expand the rule-based method that many agencies use to also include an analytic approach for constructing a risk-based monitoring system, the agency will find that there is great potential for improvement. This relates not only to the effect of the monitoring efforts, but also to cost-effectiveness.

The use of data mining, including analysing unstructured data such as, for example, earlier monitoring reports, network analyses of behaviour on social media and the internet in general, can improve the monitoring effort's ability to more precisely identify the individuals or companies that do not comply with the law.

Furthermore, experience has shown that if you use machine learning algorithms on data about behaviour and monitoring, you do not only succeed in improving the conditions in which the monitoring is planned. Unlike with the rule-based approach, you also get the opportunity to find out what (and where) the reasons are for companies and individuals not complying with the legislation.

At the same time, this new approach using more advanced analytical models simultaneously provides more opportunities to stay on top of organised crime. In other words, the rule-based model will “fog up” and be incapable of keeping up and plugging the holes for those who carry out organised crime.

Our experiences at SAS show that it is possible to achieve efficiency gains of up to 25 percent when measured in terms of how many lawbreakers, both companies and individuals, are exposed by a targeted monitoring effort. This means that the same effect of the monitoring can be achieved with significantly less effort.

It is possible to achieve efficiency gains of up to 25% measured in terms of how many lawbreakers are exposed by a targeted monitoring effort. #risk #monitoring Click To Tweet

The good news

In the Nordic countries, we currently have a hidden potential that can be found and exploited if the monitoring tasks are conducted based on an analytic approach. Such an approach will provide more financial resources for better being able to plan a preventive effort and communication and thereby increase the efficiency of the invested public funds.

An analytic approach to risk-based monitoring will benefit everyone by, over time, improving the work environment for both you and me, improving food safety for everyone, and also ensuring less exploitation of our common resources – welfare benefits, student grants and social services – by those not entitled to use them. These are all good examples of the improved effectiveness of our agencies' efforts that lead to better quality of and satisfaction with public services.

Let us find the hidden potential of Tom Cruise's “pre-visualisation tools.” You can read more about how SAS can support you with risk-based monitoring and control in the public sector here.

This blog was first published in a Danish version at Den offentlige.dk