Policy making is a group process in which multiple stakeholders identify issues, develop alternative options to resolve the issues, collect citizens’ opinions on the options through public consultation, and publish new changes in policies. Due to the nature of the tasks of the process, policy makers in practice employ unstructured decision making methods including focused group interviews, the Delphi method, and nominal techniques. While such methods have the advantage of collecting qualitative and subjective expert opinion, they have limitations in supporting evidence based policy making which require objective and quantitative data that exists in public on various data sources.
Fuzzy Cognitive Maps (FCMs) allow decision makers to integrate domain knowledge from multiple experts and conduct ‘what-if analysis’ through simulation. Fuzzy cognitive maps (FCMs) have been widely applied to various applications including engineering, industrial marketing etc. FCMs were introduced by Bart Kosko in 1986, where he developed a way to structure expert knowledge using a software programming approach. An FCM can be understood as a graphical representation of the knowledge or the perception of a given model. An FCM is a combination of Fuzzy logic and cognitive map (CM). A CM consists of concept nodes, causal relation and causal relation weighting factor. The causal relation weighting factor is assigned with directed fuzzy value that shows the strength of the causal condition between concepts. This feature marks one of the primary differences between a FCM and a CM.
In the Policy Compass project, FCMs will be applied in two pilot cases; one of a local and one of a regional scope. The lines below focus on the local pilot case of Policy Compass (Cambridgeshire County Council – CCC), while details on the second pilot case (St. Petersburg) will be provided in a future blog post.
In Cambridgeshire County Council, FCMs will be applied to the community learning and skill development (CLSD) funding issue. Cambridgeshire County Council has to respond to the UK Government policy on community learning (which is focused on assisting skills development within the local community) on a regular basis. For this purpose, the government allocates financial resources to the council through a Community Learning Fund which is managed by the national Skills Funding Agency (SFA). The council responds to this public policy by assigning a Community Learning Trust (CLT) Fund which is used to distribute resources to local training agencies that specialise in adult learning. The CLT aims to commission, deliver and support learning in ways that contribute directly to the objectives below, including:
- bringing together people from all backgrounds, cultures and income groups, including people who can/cannot afford to pay
- using effective local partnerships to bring together key providers and relevant local agencies and services
- devolving planning and accountability to neighbourhood/parish level, with local people involved in decisions about the learning offered
- involving volunteers and Voluntary and Community Sector organisation (VCSO) groups, shifting long term, ‘blocked’ classes into learning clubs, growing self-organised learning groups, and encouraging employers to support informal learning in the workplace
- supporting the wide use of online information and learning resources
- minimising overheads, bureaucracy and administration.
To achieve the aforementioned objectives, the CLT objectives are defined in the Cambridgeshire County Council Adult Learning and Skills Strategy (Skills strategy framework). The skills strategy is implemented through different action plans according to local priorities in four different districts in Cambridgeshire. Each district has Community Learning and Skills (CLAS) partnership which identify local priorities for funding. The priorities for each district are identified annually by Partnership members using a range of information such as DBIS policy, SFA funding rules, CCC skills strategy, data on deprivation, unemployment, current availability of provision, historical provision, local knowledge of stakeholders, facilities etc. This process is identified in the action plan as a local needs analysis. The funding decision is made based on scorecards which are marked by proposal evaluators.
The problems with the current process for CLT funding include the lack of ‘learner voice’ in the decision making and local Learner Advisory Panels (LAP) are being developed to address this. Also, the priority setting in local district is still conducted based on qualitative opinion of participants despite of the existence of quantitative data due to the lack of analytic tool. Also, the evaluators of proposals are lacking tools to conduct direct impact analysis of the proposals toward the local priorities and skills strategy.
Policy Compass will provide decision makers with user friendly graphic interface for analysing different indices in comparison with multiple regions within the district. Also, the policy model based on fuzzy cognitive map is expected to allow the proposal evaluators to conduct impact analysis which shows how much impact a proposal can make to the local priorities and skills strategy of the council.
The following figure illustrates how FCM can be applied to model the policy impact of funding proposals. The concepts on the right hand side include strategic and goal concepts while the ones on the left include the concepts derived from submitted proposals. The diagram show how each selected funding proposal can make impacts to the strategic goals of the programme.
Moreover, the following figure shows a potential simulation result which will show how the strategic goal concepts will be changing during the simulation periods when one of the proposal concept on the left hand side is selected. The graphs will show how strategic goal concepts will be changing by funding of one or more proposals.