Published on: 04/02/2015
It's hard to predict what impact investments and innovations in the water sector will have on citizens' access to services. Understanding underlying mechanisms and potential bottlenecks of change can help decide how and where to invest resources, while also giving a more realistic picture of the time scale required. In this blog we describe IRC's innovative work on understanding how water service delivery systems evolve.
By Carmen da Silva Wells and Deirdre Casella
At the heart of IRC’s approach to change is a vision of how the rural water sector needs to function if sustainable water services are to be provided to everyone. IRC has done a lot of practical work on investigating, documenting and facilitating a shift towards better services and a sector which learns and adapts. In the past years, IRC has made an effort to better ground our understanding of systems change in the academic literature and methods from the field of complex adaptive systems, and to learn from sector change processes in energy, mobile communications and education.
Between June 2013 and April 2014, Simone van Tongeren worked on her MSc thesis project under guidance from Deirdre Casella, learning coordinator of the Triple-S project. Simone’s research combines existing theories and approaches in a new application for understanding water services delivery in rural areas. This framework lays the foundation for an agent-based model for exploring how change (policy, or innovation) arises in rural water service delivery.
Rural water service delivery is a 'complex adaptive socio-technical system'
The rural water sector is a complex, dynamic, adaptive, constantly changing system, where control tends to be highly dispersed and decentralized. The delivery of water services is dependent on multiple actors and organisations operating at different hierarchical levels and the water infrastructure- which encompasses a range of technologies. Constellations of individuals and organisations are constantly involved in some facet of policy making, financing and financial management, operations, maintenance and consumption of water (or sanitation) services. In IRC's Thematic Overview Paper 23, Régis Garandeau summarises the similarities between complex adaptive systems and the water, sanitation and hygiene sector (see figure 1, similarities between complex adaptive systems and the WASH sector, source Garandeau, 2009: p 30).
Simone developed a detailed conceptualisation of an agent-based model, which starts with the individual- the agent – and adds further layers of actions and interactions to describe the whole socio-technical system of rural water services using information from the Republic of Uganda.
Key elements of the framework are Universal Darwinism, Nobel laureate Elinor Ostrom's Institutional Analysis and Development (IAD) framework and Memetics, the study of how people select and transmit information.
Universal Darwinism provides a generic description of how social and economic systems change over time through the evolution mechanisms of selection, variation and heredity.
The IAD framework helps identify the influence of physical and material conditions, rules-in-use and community attributes, such as cultural values, on individuals' behaviour. Like the genes of society, information (or, a 'meme') about rules of behaviour (institutions) in a given system, is passed on from person to person. Much of what people do is based on beliefs and implicit knowledge –that does not change overnight. In fact, following Williamson's model for institutional analysis (see figure 2), governance structures change over a period of 10 years, policy change take between 10 and 100 years while customs and social norms change over centuries.
These theories about change in socio-technical systems form the basis of a model for simulating how individual agents' behaviours and choices such as adopting new information, or adapting how an existing technology is used in an innovative manner to conduct a routine activity, give rise to new macro-level patterns, or emergent outcomes that are otherwise difficult to predict.
Putting the model to use
Imagine that the Government of Uganda wants to test the use of mobile telephone technology for monitoring the water service delivery levels in rural communities to assess whether this innovation will contribute to the national vision of sustainable water services for all rural residents...
A new application of an existing technology (the use of mobile telephones to collect monitoring data about water point functionality) is conceived of as a potentially promising innovation that will contribute to improving rural water service levels at scale. To test this idea and learn about the conditions under which it might provide the desired results, a pilot experiment is planned. Establishing a multi-stakeholder team, setting up the conditions for a pilot project within the existing rules-of-game that regulate national mobile communications, rural water services delivery and development sector interventions and commencing with pilot activities requires a period of about one year. This period is about identifying and mobilising human, infrastructure, knowledge and information, financial and other resources.
The process of pilot testing and refining the approach for gathering, transmitting, synthesising and analysing data requires a period of two to five years. This period of action research is characterised by a process of stakeholder engagement, capacity building, adaptive management of the pilot experiment components, process documentation and scrutiny of the potential outcomes and limitations of this new application of mobile technology to address a defined problem. If deemed worthwhile to take to scale, sector policy, strategy, budgets and implementation plans are adopted, amended or introduced across the different administrative levels. This period of aligning governance and resources arrangements to be able to roll a new approach out at scale requires a period of up to ten years.
Each phase is characterised by dynamic, non-linear pathways of change as individuals' actions and interactions throughout the 'system' result in unpredictable macro-level patters and outcomes. It is also challenging to define clear boundaries of such a 'system'. Testing and rolling out the innovative application of mobile communication technology in the rural water sector is in fact only possible because of the pre-existence of an established, regulated country-wide mobile communications network. The development and introduction of mobile communication as an institutionalised mode of communication took over twenty years in-country and was under development as a communication technology for nearly forty years internationally.
Agents' behaviour impacts upon the effectiveness of resource allocation, governance and the institutional environment. Individuals select, vary or inherit information that informs their decisions and actions. The prevailing societal norms or rituals – the informal rules that guide people's behaviours - pertaining to use of a water point such as beliefs about water as a 'free' good or the use of a private mobile phone to send information about the breakdown of a communal water point may require generations to change.
We don't have a crystal ball that will give a single answer but, we do have a sound framework that can help people learn, identify trends and influencing factors for failure and success.
This timeline gives an indication of the rate of systemic change. Similarly, an agent-based model can enable policy makers and practitioners to visualise and explore how individual actions give rise to macro-level patterns. And how these may ultimately influence water service levels in light of changes to how water services are monitored over long periods of time. Ultimately, this work provides us with a tool for decision makers to reflect upon the effects of a whole system change to delivering services without having to wait the 10 – 100 year period to understand potential outcomes and impacts.
A model can visualise how future policies might influence these trends help decision makers better estimate the effects of certain decisions. And this can save time and money. But, the value of any model depends upon the quality of information that is used to create it, and the quality of the interaction with actors who have an interest in seeing change happen in the real world.
Scroll down for a 4-pager about Agent- Based Modeling and for Simone van Tongeren's thesis report.
At IRC we have strong opinions and we value honest and frank discussion, so you won't be surprised to hear that not all the opinions on this site represent our official policy.