The Innovation Deficit in Urban Water: The Need for an Integrated Perspective on Institutions,Organizations,and Technology

Defining and characterizing innovation

The articles in this special issue of Environmental Engineering Science, and the field of environmental engineering in general, rightly focus on technological aspects of innovation. Such hardware technology systems comprise the material core of urban water services provision, such as sewers, toilets, and water treatment facilities. In contrast, innovation systems focus on different, but equally important aspects of innovation, namely, the actors, networks, and institutions that develop new products and technologies. The distinction between the two and the resulting analysis of the importance of context for technological innovation underlie a key concept: technological inventions are necessary, but not sufficient, for innovation, and the success of each is strongly influenced by context.

Invention and innovation are related but different (Lemelson-MIT Program, 2004). Invention refers to the process of devising something useful that was not previously known or existing. Innovations provide creative drivers of change in markets and society (Schumpeter, 1947). Innovation includes the products of invention plus their commercialization and introduction into markets and practice. So, while viable polymeric membranes were invented by 1963 (Loeb and Sourirajan, 1963; Lee et al., 2011), they did not become an innovation until the 1990s, when the first large-scale desalination plants began to be built. This example illustrates the distinction between invention and innovation, but also suggests that defining, measuring, tracking, and characterizing innovation can be a long-term process.

Making even coarse comparisons of the importance of different innovations can be challenging. Various classification schemes for innovation have been proposed, but one often-used distinction is made between radical innovation and incremental innovation. Radical innovations can represent “clear departures from existing practice” as opposed to “minor improvements or simple adjustments in current technology” (Dewar and Dutton, 1986). In some cases, the distinction can be made by subjective perception (Dewar and Dutton, 1986), but even extreme examples can present problems due to the existence of multiple simultaneously relevant continuums for most products, such as price, performance, potential for substitution, or functional elements along a value chain (Markard and Truffer, 2006). For example, membrane bioreactors are radical technology in the sense of relying on fundamentally different physical and chemical principles from the filtration systems they can replace (Melin et al., 2006). However, in practice, membrane bioreactors could be considered incremental if retrofitted into existing treatment plants to meet requirements for more effective filtration (Ahn et al., 1999).

A given technology can move in a trajectory from radical to incremental over time. For example, desalination was initially a radical invention as it opened possibilities for entirely new sources of drinking water. Over the past decades, with its increasing acceptance as a potential part of water supply portfolios, desalination has benefitted from incremental technical improvements, to the point where it is a mature and commercially viable technology (Elimelech and Phillip, 2011). Even in the absence of major technological shifts, most water systems are not static, as incremental change comes from production engineers, technicians, and others involved in day-to-day operation. Radical and incremental changes are each important to innovation, and cumulatively both have substantial roles in the evolution of technology over time (Freeman, 1994).

Empirical analyses of technological innovation on the scale of market sectors, such as agriculture (Griliches, 1960; Sunding and Zilberman, 2001) or industry (Fisher and Pry, 1971), suggest that diffusion of new technologies into widespread use follows a predictable conceptual pattern.1 In a classic technology diffusion framework (Rogers, 1962) (Fig. 1), initial adoption happens slowly as early adopters risk unproven reliability for the promise of novel benefits. Later, the rate of adoption increases as solutions to initial issues arise and acceptance is gained among early and late majority adopters. The rate of adoption then tapers as the sector reaches saturation.

OPEN IN VIEWER

FIG. 1.

Stylized diffusion curves. (Reproduced with permission from Meade and Islam, 2006.)

Note that in water resources, both inventors of new technologies and the utilities that adopt them could be considered innovators. That is, adopting a new technology can be an innovative act in itself, as it is a key enabler of diffusion.

Although determining appropriate indicators for assessment of innovation is not always straightforward, diffusion has been measured using total installed capacity, market share, or proxies such as patents, bibliometric indicators, or revealed preferences (Markard and Truffer, 2008). For example, global desalination capacity has grown exponentially since the 1960s (Fig. 2). As technology improves and becomes more energy-efficient and cost-effective, desalination capacity is projected to double over the period from 2008 to 2016 (The Economist, 2008). From the perspective of market share, existing desalination plants in operation account for only 0.5% total global water use as of 2008. However, such metrics are sensitive to how a market is defined. If the market for desalination were segmented further, such as geographically into coastal areas or those with ready sources of brackish groundwater, or by sector, such as for industries requiring high-quality water, such metrics might appear different.

OPEN IN VIEWER

FIG. 2.

Cumulative capacity of installed desalination plants. The capacity of desalination plants that are online or presumed online in 2006 is shown as point data. Source: NRC (2008).

The impact of an innovation depends heavily on how much it disrupts existing modes of operation. The core question is whether established capabilities in a firm are suited to support a technological novelty or whether entirely new capabilities and routines will have to be built up. Disruptiveness can be independent from technological advancement. For example, Christensen (1997) drew a distinction between sustaining and disrupting innovations, using a case study of the computer disk drive industry. Sustaining innovations continue existing trajectories of product development. For disk drives, sustaining innovations refined performance within an existing format. Disruptive innovations, in contrast, generate new standards and trajectories for performance, perhaps adding additional metrics, enabling new and different uses and creating new markets. Perhaps surprisingly, in the case study of the disk drive industry, disruptive technologies were actually minor technological advances that initially generated quantitatively inferior performance on accepted metrics such as total storage capacity. The key to their disruptive nature was that in each case they enabled new markets and ways of using the products, such as when they reduced in size to enable the development of desktop or laptop computers. Established companies were aggressive and innovative in developing technological advances. However, they did so within existing frameworks, such as mainframe or microcomputer manufacturers. Entrant companies did not contribute to cutting edge engineering. Rather their success as disruptive innovators was based on developing new paradigms for use of disk drives—creating new markets that eventually developed to take over incumbent mainframe and microcomputer business. A potential lesson here for urban water innovators is that entry points for innovation may be found not only through technological refinements of process efficiencies, but also through conceptual disruption of established ways of operating. For example, community or building-scale water and wastewater treatment many create opportunities for new entities, such as private water service providers, to operate under different business models from that of centralized utilities. Similarly, disruptive institutional innovations have begun to shift centralized decision-making through more transparent decision-making and the incorporation of stakeholders in collaborative processes.

An additional distinction can be made between incremental innovation for specific technologies as described above, and broader transitions to new states. Transition management deals with co-evolution between technological and institutional aspects of new sociotechnical regimes, and often spans time frames of several decades. Both are important for urban water management: incremental technological change is ongoing, but the durable nature of water infrastructure makes other institutional timeframes such as electoral and business cycles less applicable to state changes as broad as those envisioned for holistic reinvention of urban water systems.

To address the problem of a large scale sectoral transformation toward more sustainable future service provision, a whole set of radical technological innovations would have to take place causing disruptive innovations within an industry, or even the entry of new industry actors (Tushman and Anderson, 1986) and associated new institutional frameworks. Large-scale, long-term transformation processes (for example, a shift toward an electricity system based largely on renewables) have been analyzed in the recent literature on sustainability transitions (Smith et al., 2010; Van Den Bergh et al., 2011; Markard et al., 2012). One core element of transitions research considers the emergence of new industry structures that provide new products and services, becoming important drivers of the transition processes.

The innovation system literature emphasizes the interplay of different actors (firms, but also research institutes, regulators, associations and other intermediaries, civil society and end users) in diverse network structures (Markard and Truffer, 2008), which is useful for looking both at management of incremental innovation or sectoral transitions. In particular, the approach of technological innovation systems enables the identification of necessary conditions for radical innovations to succeed by overcoming barriers stemming from deficiencies in firm capabilities or mismatch in institutional structures between new technologies and incumbent systems (Jacobsson and Bergek, 2011). A technological innovation system can be thought of as “a network of agents interacting in the economic/industrial area under a particular institutional infrastructure […] and involved in the generation, diffusion, and utilization of technology” (Carlsson and Stankiewicz, 1991). It is important to note that the emphasis here is on the actors, networks, and institutions that enable innovation, and not on the nuts and bolts of the hardware itself.2 This emphasis reflects the importance of these nontechnical elements in the generation of innovation.

In urban water management, transitions research has drawn from innovation studies in analyzing historical transitions (Geels, 2005, 2006) as well as potential pathways toward water sensitive urban design (Brown et al., 2011; Van De Meene et al., 2011). Further, emerging technological innovation systems have been viewed through this lens, include onsite treatment in urban water management (Hegger et al., 2007; Truffer et al., 2012) and the possibility for leapfrogging to new system designs in emerging economies like China (Binz et al., 2012).

Drivers of innovation

In coming decades, increasing stresses on urban systems will influence supply, demand, quality, cost, environmental, economic, and distributive elements of urban water (Gleick, 1993; Zimmerman et al., 2008). The perennial challenges of climate variability will grow in importance for urban water as it is exacerbated by climate change. Historical expectations for surface water (Bates et al., 2008; Kundzewicz et al., 2008) and groundwater (Döll, 2009; Earman and Dettinger, 2011; Treidel et al., 2011) supplies and flooding (Ntelekos et al., 2010) can no longer be relied upon, leading to unanswered questions about future water and stormwater management (Milly et al., 2008; Brown, 2010). Population growth and urbanization trends promise to increase the spatial extent of urban areas as well as their population density (Martine, 2007; Seto et al., 2011), putting pressure on sources of water supply (Vörösmarty et al., 2000; Sun et al., 2008). The energy and greenhouse gas intensity of water and wastewater systems (EPRI, 2002; Klein et al., 2005; Hall et al., 2011) will need to be addressed with both new technologies and decision-making strategies as costs and regulations for emissions increase. Emerging contaminants, such as endocrine disruptors and pharmaceuticals and personal care products, threaten ecological and public health and promise to complicate water treatment (Kolpin et al., 2002; Snyder et al., 2003). Urban water systems are linked to deteriorating environmental quality within cities as well as upstream and downstream of them (Grimm et al., 2008). Increasing societal recognition of the benefits of aquatic ecosystem services (Brauman et al., 2007) puts corresponding pressure for restoring and reducing impacts to aquatic systems. The inexorable deterioration of even long-lived infrastructure suggests impending challenges even absent these stressors (American Society of Civil Engineers [ASCE], 2011). On top of all of these pressures, funding investment and maintenance in urban water systems is increasingly challenged with tightening budgets at all governmental levels (OECD, 2006, 2007; UNESCO/World Water Assessment Program, 2006; Urban Land Institute and Ernst and Young, 2007), and recurring debates about organizational reform in utilities complicate efforts for future planning.

The primary unifying theme of this nonexhaustive list is that of change and uncertainty—the only thing we can say for sure about the future of urban water is that it will look quite different from the past. Water crises are increasingly viewed by the business (World Economic Forum, 2012) and national security (Clapper, 2012) communities as among the greatest social risks of coming years, and urban water systems will be called on to ameliorate these risks.

Existing practices will need to change in ways not yet anticipated, and simply doing more of the same will not be economically efficient or societally acceptable. New ways of doing things—innovation—will be critical to the future of urban water systems.

Can technology promise solutions commensurate to the impending stresses on urban water systems? To some extent it can. Much theoretical room remains for futuristic advances in the efficiency of water-related technologies (Shannon et al., 2008; Elimelech and Phillip, 2011). However, paradigm-shifting concepts promise innovation in all aspects of the urban water cycle (Daigger, 2011). Green infrastructure, distributed water reuse, recovery of energy and nutrients from wastewater (Guest et al., 2009), potable water reuse (NRC,2012), source separation of wastewater components (Larsen et al., 2009), and other concepts all have potential to change the physical structure and financial outlook for urban water systems. Many of these technologies are the subject of intensive research (www.urbanwatererc.com), and some are reviewed in companion articles in this issue.

However, in spite of much promise, radical technological advances have not been implemented at large scale. While technological inventions are necessary to the future of urban water systems, they are not in and of themselves sufficient. Rather, the potential for innovation depends on a blend of technological and institutional factors that operate in concert.

Processes behind technological innovation

Broad observations of trends in technology change have led to research on how and why innovation happens. Scholarship on innovation has moved to look inside the “black box” of technology development (Freeman, 1994), paying increased attention to the processes of innovation.

Whether innovation is driven more by the supply of new technological options (“technology-push”) or by market demand for solutions to known problems (“demand-pull”) has not been explored in depth for water resources, although the question has been explored with some nuance in areas such as renewable energy (Taylor, 2008). Freeman (1994) suggests that the answer depends in part on the stage and type of innovation. For example, there is unlikely to be preexisting market demand for a given radical, early-stage innovation. Some innovators understand this intuitively, as illustrated by the way Steve Jobs eschewed market research while creating a new market category with the Apple iPad, saying, “It's not the consumers' job to know what they want” (Markoff, 2011). Conversely, demand for improvement of existing technologies is more likely to spur incremental innovations. The development of desalination illustrates both points. The initial commercialization of cellulose acetate membranes can be classified as a technology push. Since then, innovations in water permeability and salt rejection of reverse osmosis membranes (Fig. 3) have resulted in gradual improvements in the efficiency of desalination technology (Lee et al., 2011). Over time, a demand for better performing membranes has led to a pull that motivates efforts to increase economic viability of desalination plants.

OPEN IN VIEWER

FIG. 3.

Incremental innovation in asymmetric reverse osmosis membrane. (a) Early developments and (b) example of recent increases in efficiency, in Dow membranes. (Reproduced with permission from Lee et al., 2011.)

Distinguishing between technology-push and demand-pull models of innovation may often be a false choice over the lifetime of a technology. Such formulations imply linear models (Freeman, 1994) with smooth, one-way flows of information from research to development to product marketing. Such linear models have been roundly criticized not only as simplistic, but as simply wrong, and replaced by more sophisticated views of innovation process. A progenitor of current models, the “chain linked model” (Kline and Rosenberg, 1986) serves to illustrate the paradigmatic shift. It presents innovation as not a single process driven by scientific discovery, but a chain of innovation activities including market identification, detailed design and testing, production, and so forth. Each element of the chain is connected to the others via feedback loops, and also directly or indirectly to research and knowledge production. The key point emerging from the chain linked model is that a view of innovation that places science as its central driving force is not only simplistic, but potentially misleading. Research is crucial to innovation, but exists only within a larger set of processes.

“Hard” functions (Galli and Teubal, 1997), such as basic and applied research, development and testing of prototypes and products, and the provision of technical services, all of which are core elements of engineering training and practice, are directly involved in development of new technology. However, while technological advances are central to innovation, innovation also includes “soft” functions such as information exchange, policies, financial methods, management strategies, or ways of doing business.3 Such functions provide essential complements to the “hard” ones more directly related to technology development. For example, Gebauer et al. (2012) detail how business networks may need to change dramatically to act on new business opportunities opened by the changing European wastewater industry. They argue that successful development and deployment of novel technical solutions usually cannot be the provenance of a single (technologically focused) firm. Rather, networks with overlapping expertise are required. Understanding how the broad range of innovation activities do, or should, work together for reinvention of urban water systems remains an unanswered challenge.

Innovation in context

Scholarship on innovation studies (Fagerberg and Verspagen, 2009) takes as its starting point the premise that innovation is determined not only within a given firm or research group, but also by the context in which they are embedded. Myriad elements, including but not limited to technical ones, that influence the potential for successful innovation. A novel technological development alone is not sufficient to disrupt an existing technological regime, nor can it be developed in isolation from the institutional and social context in which its entrepreneurial researchers operate. Innovation results from complex interactions among people and organizations. For example, a collaborative group of firms may combine their offerings into an interrelated package, forming an innovation ecosystem (Adner, 2006) that can offer otherwise untenable products, but creating a new set of risks for the now interdependent enterprises. An example would be separate firms making hardware and software for mobile phones. Factors such as this can complicate understanding or forecasting of technological change. They can also lead to insights about what a specific invention needs to consider in order to become a successful innovation.

Technological merits alone cannot explain the ultimate fate of new technologies. A superior but unfamiliar technology needs to overcome the advantages of the ecosystem that develops around an incumbent technology. Existing technologies benefit from knowledge accumulated by users, capital outlays and infrastructure, available skills, and other aspects that collectively support an inertial dominance (Kemp, 1994). For example, in considering whether to change from sand filtration to membrane filtration, a water system operator could not simply evaluate the decreased risk to public health and space efficiency. In addition, they would need to weigh the potential costs of retraining its workforce and developing a new supply chain. Often, broader effects on interconnected systems would have to be evaluated. For example, while the “theoretical potential for desalination is virtually unlimited,” economic, social, environmental, and political complications “are far more important than technological desalination process constraints in limiting the potential for desalination to help meet anticipated water supply needs” (NRC, 2008). Similarly, water organizations have been slow to adopt promising emerging approaches such as in-line energy recovery (Klein et al., 2005; House, 2010) or leak detection and reduction (ECONorthwest, 2011) that promise to plug into existing systems, reducing existing inefficiencies to deliver potentially rapid return on investment. The explanation may stem in part from an inability of the new technology to overcome the advantages of existing technology, or from institutional context, such as lack of familiarity, embedded best management practices, or risk aversion among decision-makers faced with unproven options.

Inertia and path dependency are particularly relevant for water systems. Infrastructure built around the centralized system paradigm can have long design lives, especially compared to other technologies, such as computers, where generation times are orders of magnitude smaller (Christensen, 1997; Christensen et al., 1998). Because of large sunk costs and challenges associated with integrating new technologies into existing systems, wholesale disruptive changes will continue to be rare within established urban cores. Alternatively, urban developments where water efficiency is a design priority may be better venues for radical new approaches. New developments focused on shifting from linear water systems to closed-loop configurations have resulted in 50%–80% reductions in potable water consumption and 70%–100% reductions in discharges to local waterways compared to other regional communities (Apostolidis et al., 2011; Apostolidis, 2012). Such achievements within legacy systems would require dramatic measures, but integrated design can be constructed at similar cost to traditional systems in new developments (Apostolidis et al., 2011). This does not imply that such achievements are easy in new developments. On the contrary, numerous institutional, regulatory, and technical hurdles abound.

One lens through which to view innovation systems is that of evolutionary theory and evolutionary economics. McKelvey (1997, cited in Hekkert et al., 2007) describes the net processes of innovation systems as analogous to biological systems, where necessary elements of evolution include (1) retaining and transmitting information; (2) generation of novelty leading to diversity; and (3) selection among alternatives. Information sharing, policymaking, design and enforcement of patent laws and other related standards, and professional coordination (Galli and Teubal, 1997) are critical to the information functions. An example would be scientific and professional conferences geared toward highlighting new available technology. Generation of technological diversity happens through research and development, manufacturing, and also by end-use. In the context of urban water infrastructure, consulting engineers will often serve the role of information brokers who add value by understanding and presenting a range of technological options that may fit a given utility's needs. Selection among alternatives can happen through the search process where customers look for solutions to existing issues, through allocation of resources for research and development, through exchange of information and vision, and through external facilitation of new market formation.

Other key functions of innovation systems have been enumerated by Hekkert et al. (2007). Entrepreneurial activities help develop new knowledge, networks, and markets into business opportunities. Presently, support for water entrepreneurs is limited. Water technology entrepreneurs have some access to attention and funding from a small number of environmentally focused venture capital firms, but the capital deployed under the cleantech/greentech umbrella for water resources pales in comparison to that for renewable energy. In 2011, venture capital investments in water technologies totaled about $224 million in 40 deals (Kho, 2012), or about 5% of the $4.3 billion invested in 323 deals in the cleantech sector overall (PwC/NVCA, 2012). Awards for innovative water products (imagineh2o.org, artemistop50.com) provide avenues for increasing entrepreneurial visibility, but there is much room for growth. In part, the lack of entrepreneurial activity is tied to limitations on direct profit potential from water. These include its provision by public agencies rather than for-profit entities, limitations on water markets, and the expense of and limited scope of facilities for bulk storage and conveyance of water that limit the geographic potential of water markets.

Knowledge development can include “learning by doing” and “learning by searching.” For example, East Bay Municipal Utilities District (EBMUD) has attempted both approaches to attaining its goal of an energy-positive wastewater treatment facility (Hake et al., 2006). Motivated by California's energy crisis in 2000–2001 as well as the adoption of a Renewable Portfolio Standard by the state, but lacking an accepted set of best management practices for a new goal, EBMUD has had to mix engineering analysis with experimentation to further its efforts in innovative energy management.

Knowledge diffusion often occurs through social and business networks, through learning by interacting or learning by using. Trade organizations such as International Water Association provide forums where industry representatives and public agency representatives can engage through conferences to build networks of knowledge and practice. Regional forums, or groups for those with more specialized needs or specific goals provide more targeted opportunities for knowledge diffusion.

Guidance of search refers to defining and increasing the visibility of demands for new functions among users. In a pull model, innovation producers recognize the need to understand market demands. Goal setting by policymakers is one mechanism for such signaling. For example, recently passed California legislation sends clear signals by requiring 20% reduction in urban water use from year 2000 levels by the year 2020 (State Water Resources Control Board, 2008, 2009). Analysis and communication that incorporates stakeholders can also clarify market demands. Such boundary work (Guston, 2001; Cash et al., 2003) recognizes the gaps between producers of and potential consumers of knowledge. Empirical and theoretical studies have shown that the use of knowledge in decision-making in policy contexts is made more effective through early, active, and ongoing collaboration between both groups, and such concepts likely transfer to innovation systems as well. Boundary work can be carried out by boundary organizations that have a defined mission to serve as a bridge between different sectors, and may be most effective when multiple stakeholders are involved in ongoing processes (Jacobs et al., 2010). Research foundations can also serve a boundary role as part of their mission. For example, the Water Research Foundation, the Water Environment Research Federation, and the Water Reuse Foundation not only fund primary research, but also conduct outreach to bridge technology development and science communication. The role of boundary organizations and boundary work is implicit in innovation studies, but new research formally joining the two areas of inquiry could yield useful insights (Taylor, 2008).

The need for market formation recognizes that incubation of potentially disruptive new technologies in protected spaces or niche markets can enable them to mature to a point where they have a greater chance of being competitive. In niche markets, actors are willing to accept such teething problems as higher costs and will invest in improvements in a new technology, in order to work toward specific functionalities (Markard and Truffer, 2008). Protected spaces can be developed through policies such as tax regimes and minimal consumption levels. This model has helped enable development of increasingly efficient and cost-competitive renewable energy sources. For example, Germany has become a market leader in photovoltaics over the past two decades (Dewald and Truffer, 2012). The proximate enablers were subsidies in the form of a feed-in tariff providing a protected space. However, the ultimate origin of support for these subsidies can be traced to an evolving range of actors, starting with local pro-solar initiatives and moving across a range including local and state governments and corporations, illustrating the importance of a broad range of actors in a technical innovation system. Closely related to this example, mobilization of resources through industry or government funded R&D programs can enable high-risk, early stage actions that would be difficult to fund through market-based means alone.

Creation of legitimacy is also necessary to put a new technology on a policy agenda and break down resistance to disruptive change. One could argue that potable water reuse in the United States appears to be in the midst of a process of legitimacy creation, as technical understanding of its performance grows, and political and social resistance breaks down as nonpotable reuse and potable reuse projects operate for extended periods without adverse health effects (NRC, 2012). Ultimately, all of these functions described by Hekkert et al. (2007) are most effective when part of an iterative process.

The scholarship discussed above helps peel away layers of complexity for better conceptual understanding of the ways in which context affects innovation. Translation of such insights into predictive models of innovation systems is in its infancy, and much room remains for further research toward this goal. Arguably, innovation is inherently unpredictable, but attempts have been made to quantitatively model diffusion of new water technologies within multifaceted innovation systems. For example, Kotz and Hiessl (2005) use Agent Based Modeling (ABM) to represent choices and decisions by water suppliers and water consumers within an urban water management system, although with coarsely parameterized and incomplete implementation. Their results show how ABM can reflect feedbacks between agents operating on different timescales and with different incentives and how they complicate system-wide technology adoption, such as where conservation actions by households can impact the economics of sewer system operation. Detailed, citywide ABM representation also holds promise as an integrator of multiple modeling approaches. Galán et al. (2009) model technologies for household water conservation, showing that urban population dynamics and other sociogeographic effects influence technology spread in a way that may not be apparent in aggregate modeling. Such efforts highlight the different motivations, different decision frames, and different timescales relevant to water providers, sewerage providers, and water consumers, and how each can affect decisions in a dynamic context. Although ABM in particular has yet to fulfill its promise (Bankes 2002), as the field of innovation studies matures, descriptive and predictive quantification of innovation trajectories will be an important area for further work.

Technological, behavioral, and institutional changes need to be explicitly situated within existing infrastructures and institutions. Recognizing innovation as the product of a multitude of influences and dissecting the different elements that influence innovation promises to help reveal bottlenecks and inform decision-making.

Institutional influences

Institutional context can be a key hindrance or enabler for innovation in urban water systems. Defining institutions broadly as the rules, norms, and practices that govern decision-making enables inclusion of the multiplicity of factors that shape water systems. Such factors include public health regulations and laws, but also economic, social, cultural, and other nontechnical aspects. Even where technology with demonstrated potential for improving urban water systems is available, institutional challenges may prove to be the greatest hurdles to cross. Learning how institutional elements enable innovation in urban water will be a key enabler of transforming technological inventions into technological innovations.

In many cases institutions hinder possibilities for innovation, as Roy et al. (2008) discuss in the context of sustainable stormwater management. Uncertainties in performance and cost relate provide challenges (Roy et al., 2008). This can be exacerbated by risk aversion and resistance of change by decision-makers who may prefer proven technologies to a potentially better one. Insufficient engineering standards and guidelines (Roy et al., 2008) may apply particularly to radical innovations. Fragmented responsibilities, lack of institutional capacity, and lack of legislative mandate (Roy et al., 2008) all provide challenges for decision-making in water systems that cross existing jurisdictions, and institutional design that can address such challenges presents opportunities for new research. Lack of funding and effective market incentives exacerbate water's dual status as a public and private good (Hanemann, 2006). With the possible exception of performance uncertainties, these are not technical impediments per se, but rather institutional challenges that affect the potential for implementation of existing technologies.

In the face of such challenges, there may be the potential to adapt institutions in support of innovation. Edquist and Johnson (1997) suggest that institutions can function in three ways to foster innovation. First, institutions can serve to reduce uncertainty by providing information. Information provision can be required by laws, such as the U.S. National Environmental Protection Act reporting requirements. It can also function through informal means, such a networking among practitioners at industry conferences. Second, institutions can actively manage conflicts and foster cooperation, such as when stakeholder processes allow communities to engage with and develop more durable solutions that take multiple interests into account (Dominguez et al., 2009). Third, institutions can provide incentives for innovation. For example, patent laws generally protect new intellectual property such that entrepreneurs can profit from taking risk on technology development. More specific drivers can be provided by targeted research funding to encourage development in desired areas. Such funding exists from sources including government programs and venture capitalists, but have typically been small for water technologies relative to other natural resources, with U.S. government and private sector investment in water technology lagging renewable energy funding by orders of magnitude in recent years (Moresco, 2009).

Finance, pricing, and innovation

Water systems are capital intensive, but access to capital required for investment in technology faces hurdles related to water's unique characteristics (Hanemann, 2006). Access to capital markets is influenced by agencies' status as public entities or heavily regulated private ones. Privatization of water utilities has yielded mixed results, and it is not clear to what extent market efficiency can be brought to bear without undue risks for consumers. The limited ability of public entities to take financial risks has lead to creative attempts in private sector financing of infrastructure projects. For example, Poseidon Resources' efforts to finance a desalination plant in San Diego represent an attempt to shift risk from the city to outside investors. The Poseidon group has proposed a plan to take on initial capital costs, betting that increases in the price of imported water over the 30-year project lifetime will result in profitable sales of desalinated water to the city (Wilson, 2010). However, the project remains controversial in part because of the lack of overall transparency as well as subsidies attained from other organizations. Thus, even before technological risks are realized in construction and operation, business risk arises from political factors unrelated to technology.

One challenge for water innovation is that costs for innovation and the accrual of benefits can be misaligned. The scale of water problems and solutions often does not coincide with the scale of relevant institutions. For example, distributed, modular methods for water reuse have different costs and benefits at the sub-system level than what aggregates at the whole-system level. In Germany, experience with building-scale water recycling has benefits from reduced water use at multiple scales, but unintended consequences of increased deterioration of pipes because of lower flows and greater concentrations of solids in the waste stream. Short-term infrastructure costs to municipalities thus need to be balanced with long-term public good benefits from decreased water withdrawals. Similarly, reductions in imported water, cost savings from lowering pumping costs, reductions in carbon intensity of water systems, and other benefits often accrue to the water provider in response to conservation, but may be hard to measure and credit (Venkatesh, 2012). Aligning incentives properly at different scales is an ongoing challenge for technocratic, deliberative, participatory, collaborative, and market-based methods.

Culture and governance of organizations

Organizations, including agencies that make and implement decisions about water resources, are particularly important for shaping urban water. Beyond their mandates, their design and culture can have profound effects on whether and how innovation results.

Legal control of water resources often forms a primary constraint on all elements of water management, including supply availability, treatment, and disposal needs. Historical development of water rights has in some areas become out of sync with economic and social priorities, as the latter evolve more quickly than the former (Doremus and Hanemann, 2008). Further, in places with static prioritization systems for water allocation, changing social values may result in misalignment between the best supply sources and the uses with the highest values to society (Sax, 2007). Since urban areas were often settled more recently than the agricultural areas from which their water may be sourced, low priority of use can influence the risk for urban water supply. Further tensions arise where environmental services from in-stream flows have become increasingly valued by society in areas where stream flow is by and large allocated to other uses. All of these factors set the stage and help define the degrees of freedom within which urban managers can make decisions.

The number of organizations with overlapping and fragmented responsibilities in the urban water sector, and the inconsistency in their jurisdictions and methods of decision-making make for complex governance structures (Adler, 2009). Urban water often lies within the domain of organizations at all levels of government, from federal to local. Fragmentation takes multiple forms, categorized by Adler (1995) as “(1) political fragmentation—the overlapping and conflicting division of responsibilities among multiple levels of government and agencies; (2) issue fragmentation—the artificial division of related water issues into separate programs (such as water quality and quantity, land and water use, and surface and groundwater); and (3) gaps in program design and implementation.” Further, fragmentation results from spatial mismatch between watershed and political boundaries. Institutional fragmentation is one of the key drivers for the need for coordination and collaboration, which can affect the potential for change. For example, joint management of energy and water could potentially have net benefits for both sectors, financially and environmentally. However, regulatory mismatch between the two closely related sectors can shift a problem previously framed in technical terms into a regulatory challenge, as mechanisms do not exist for valuing costs and benefits across the two sectors and disparate spatial scales.

Organizational capacity and culture also play a role in ability to innovate, and risk aversion is a central facet. Water resource managers are often described as conservative and risk averse (O'Connor et al., 2005; Rayner et al., 2005; Lemos, 2008). It is important to emphasize that for many water managers, such risk aversion is entirely sensible and consistent with the context in which they operate. For example, in civil service, managers can be punished for failure, yet not rewarded for success. This is especially true when the cost of failure is measured in outbreaks of waterborne disease, sewage spills, and flooding of residential properties. Whereas responsibility for increases in efficiency in the corporate world can result in raises, corresponding success in meeting water quality regulations or avoiding supply shortfalls is expected, but where bonuses exist they are rarely substantial. Downside risk, or the risk of bad outcomes, drives conservatism in the public sector, where the negative outcomes are not offset by corresponding rewards for strong upside performance. Decisions involving new technologies will often fall in a particularly tricky category for risk-averse decision-makers because they combined high risk with low data availability. Thus, the reluctance among decision makers to take risks on unproven technologies is unsurprising, since the overall direction of large public works projects is determined by relatively few large and impactful decisions. Reluctance to adopt potable water reuse in spite of the scientific evidence that it can offer safe and cost-effective new supply (NRC, 2012) exemplifies how resistance to change can result from risk aversion. In the aggregate, risk aversion likely slows progress for the water industry as a whole. The role of incentives, organizational culture, and individual behavior in enforcing (or possibly reducing) risk-averse decision-making among water managers represents an opportunity for research.

Capacity for innovative decision-making may be related to the capacity to reflect on future options. Some inertia results from existing infrastructure systems with long design lifetimes, and the entrenched institutions that have evolved to support them. However, water utilities often have not invested in capacity to conduct strategic planning in house. When they explicitly or implicitly relegate strategic planning to consulting engineers, there is greater possibility of restricting the range of technological and organizational options considered (Dominguez et al., 2009), potentially damping innovation by perpetuating established technologies and organizational structures.