The global south is expected to witness significant levels of urbanization in the coming decades which will have implications for the management of water resources (UN DESA, 2014). We see evidence of this in the Ganges Delta, where rapid and largely uncontrolled urban expansion has brought growing concerns of water scarcity, competition, and inequity among users (Kumar et al., 2011; Thissen et al., 2013). The situation is particularly severe in peri-urban areas surrounding Khulna city, Bangladesh, where groundwater is the main source for domestic and livelihood purposes.

Peri-urban areas are dynamic, representing the transition zones in processes of urban expansion with a heterogeneous composition (Allen, 2003; Narain, 2010). In such a context, the management of groundwater resources requires a coordinated approach among actors, based on a shared set of rules. Institutions are defined as formal and informal “rules” that structure interaction and behavior in society (North, 1990). They offer a means for resolving social dilemmas by offering prescriptive guidelines. However, peri-urban villages in Khulna have witnessed the failure of existing institutions in managing water resources in a sustainable and equitable way (Thissen et al., 2013). Achieving effective institutional arrangements in this context, however, is complex. Institutions are expected to evolve with the changing context, yet this is not always possible due to various actor constraints that prevents this. This can result in the creation of a fragmented mix of rural and urban institutions. This research explores how peri-urban institutions and institutional change has contributed to spatial differences in groundwater access in peri-urban Khulna.

The adapted Institutional Analysis and Development (IAD) framework in Fig. 1 focuses on key aspects relevant to the study of institutions and institutional change. The analysis centres around the action arena, consisting of interactions between multiple actors with an ability to influence their system of interest. In this case, the action arena refers to peri-urban groundwater management. Strategic behaviour within this arena and its resulting outcome is based on contextual variables such as the institutions or “rules in use”, community attributes such as socio-economic conditions, and biophysical conditions (Ostrom, 2005)

The institutional context includes consciously designed and codified formal laws, constitutions, or property rights as well as informal, socially constructed norms and values that together comprise a society's culture (North, 1990; Williamson, 1998). They define the participants, their position, resources, possible actions, and the costs and benefits of their associated outcomes within the action arena (Ostrom, 2005). In society, these institutions are organized in a nested structure from local operational rules to increasingly embedded higher order collective choice, constitutional, and meta-constitutional rules (North, 1990; Ostrom, 2005; Williamson, 2000).

Feedback loops from the outcomes are an important part of this framework offering a means to study institutional change. North (1996) describes institutional change to arise through a process of actor learning from the outcomes of their strategic decisions. Here, negative outcomes are a signal to change strategies or alter the institutions in order to achieve their objective. Although institutions are expected to evolve with actor's needs in this manner, this is not always the case. Theories highlight that the embeddedness of rules brings a higher transaction costs for actors investing in institutional change (North, 1990). As a result, institutions may gradually become eroded over time, replaced, or ignored altogether without purposive action (Scharpf, 1997). Moreover, institutions are not socially efficient, so institutional change also depends on the bargaining power of the actors involved (North, 1990, 1996). In Fig. 1, system feedback resulting in observed changes to the biophysical and socio-economic context is differentiated from actor feedback that produces institutional change, given this is a key research objective.

Institutional aspects of groundwater access in peri-urban Khulna is analyzed using the framework in Fig. 1. The villages of Hoglandanga and Matumdanga are selected as case studies given the peri-urban characteristics they display and evidence of groundwater issues. They are currently witnessing urban expansion and are formally located outside Khulna city jurisdiction. Multi-actor interactions in groundwater access, an issue highlighted by both communities, is examined within the contextual variables. Field research reveals three key stages in the evolution of this issue representing the past, present, and future scenario of ground water access. In each stage, the study examines the form of institutional change invested in to manage groundwater access.

Primary data from government websites and published reports is verified and supplemented through field discussions with 18 respondents from government agencies and water users in both villages (Gomes, 2015). Interviewees were selected through structured and snowball sampling methods (Harrell and Bradley, 2009). Interview methods include focus group discussions and key informant interviews. Field data gathered in May–June (2015) was subsequently verified through a de-briefing meeting with local actors or email communication. Formal and informal institutions are analyzed based on their level of operation in the nested structure and the action arena variables they define (Ostrom, 2005).

These results show how peri-urban water access in Khulna is influenced by institutions. Formal institutions in the initial stage produced spatial differences in groundwater access between areas with and without tube well allocations. Although policies specify rules for allocation, lack of adequate enforcing and monitoring mechanisms can be a deterrent to its functioning at the local level. Monitoring and enforcement of rules can help prevent opportunistic behavior by placing a higher risk for those considering defecting (Ostrom, 2005). The need for enforcement, however, depends on whether there is a known willingness for local committees to deviate from rules. There appear to be potential payoffs for some elected members within the committee to support deviation. One could be internal reciprocal norms to appease the constituencies that elect them, which influences their decision. Or it may be the result of conflicting rules such as the Bangladesh Water Act (2013) promoting the right to water and others stating that communities should play a greater role in managing the costs for water services (Local Government Division, 1998; Ministry of Law, Justice and Parlimentary Affairs, 2013; Ministry of Water Resources, 1999). Greif (2006)'s theory on legal pluralism highlights that rules may be complementing, reinforcing, or conflicting to one another. Furthermore, limited understanding of the context in the design of rules or combination of rules can produce unintended or disastrous outcomes (Ostrom, 2005). Thus, research needs to look beyond local allocation rules to higher order rules or rules from other sectors and how they work in relation to one another.

The undesirable allocation outcome prompted local actors to invest in institutional change in order to meet their water needs. In the current situation, we observe that institutional change depends on one's resources. The location of both villages away from rural decision centers, give users limited bargaining power to seek a more equitable distribution of licenses. Thus, actors chose to invest in informal institutional change given the high transaction costs. Here, we see different sets of informal rules for irrigators and domestic users. While irrigators opted for a market system to access groundwater, domestic users share existing tube wells. The latter bears less financial implications but has costs in terms of time, distance travelled, and convenience. Thus, even with informal institutions we see differences in the options between peri-urban users.

In the future, constitutional level rules could again influence groundwater access through urbanization. Formal change by replacing rural with urban service provision creates some uncertainties for users in these communities. For domestic users there is the risk that existing gaps in urban water supply will continue, once again creating spatial difference in access within urban areas. For irrigators, future groundwater dependency is uncertain given the existing changes in land use (Gomes, 2015). When peri-urban villages come under KCC jurisdiction, the fate of peri-urban land will depend upon the master plans. So it is likely that land use will be a bigger uncertainty than water access.

Problems in peri-urban Khulna show that institutions fail to effectively manage groundwater access during urban transitions. It also highlights the challenges with institutional design in this context. The changing needs in peri-urban areas mean existing institutions across a rural-urban divide become ineffective over time causing spatial patterns in water access. Addressing inefficient rules needs to consider how they are applied in practice. Understanding actor motivations and objectives can help in this regard. In Khulna, we observe a shift from formal to informal institutions with the future bringing yet another formal change. Theories suggest that institutions work if they are stable in the long term but this is not possible given the dynamics of peri-urban areas. Thus peri-urban areas are challenged with balancing functionality with stability, the former addressed in this case through informal means. This research signals that improving water management during urban transitions requires focusing on the institutional context, for which the IAD framework offers a means to operationalize key concepts in the peri-urban context.