In order to discuss whether urban water markets can work, it is necessary to consider the institutional framework in which the proposed markets will operate. Worldwide, there have to date only been tentative reforms directed at introducing competition. While these reforms will increase flexibility and efficiency, they only address partially the above problems.

In Australia, Part IIIA of the Trade Practices Act allows for third-party access to an urban water networks, provided a ‘declaration’ of that service is made. So far only one such declaration has been sought — by Services Sydney in 2004 trying to gain access to the sewerage infrastructure of Sydney Water. This access was resisted by both Sydney Water and the New South Wales premier. This ultimately led to the development of the NSW Water Industry Competition Act (2006), which regulates third-party access to urban water infrastructure in the Sydney and Hunter regions. More recently, the ACCC has made a determination on the methodology used to set access prices to water infrastructure (for a detailed discussion of these issues, see Gray and Gardner 2008). To date, no third-party access to urban water networks has occurred in Australia.

In the UK the Water Act (2003) allowed the implementation of the current ‘Water Supply Licensing’ regime. This regime (like that in Australia) has not, to date, produced significant competition in UK water supplies. The UK water regulator, Ofwat, identifies the current cost of access, including current access, as the reason for this (Ofwat 2007). Ofwat has recommended changes to the way in which the urban water industry is structured so as to increase the level of competition. In particular, it argues for a separation of contestable from non-contestable markets (sewerage and retail sectors are specifically identified) and the development of new market models for upstream competition (Ofwat 2008: 8).

Sibly and Tooth (forthcoming) and Young, McColl and Fisher (2006) propose methods by which the role of urban water markets could be expanded in Australia. These proposals involve, to some extent, a decoupling of water and network ownership (as do all that suggest the implementation of competition). Young, McColl and Fisher (2006) propose a model in which the market is supplied by a monopoly water authority, but one which allows trading between its customers. To generate urban water trading, they suggest allocating each household a tradable allowance of 200 kilolitres per annum. Households would need to purchase additional water to consume beyond that limit and could sell any unused allocation. Water use beyond 200 kl/yr could be regulated by (i) a scarcity price or (ii) a cap-and-trade system. Young, McColl and Fisher propose that the scarcity price would be related to dam levels. If such a price were regulated, it need not necessarily be set at the market clearing level. In this event, water restrictions would remain necessary. In the cap-and-trade system a limit on total water consumption above 200 kl/year/household would be set, and households could bid for a share of the available water. Such a process would eliminate excess demand for water, and thus could eliminate water restrictions. However, to ensure the efficient inter-temporal allocation of water (storage) the cap must be set appropriately in each period.

Sibly and Tooth (forthcoming) propose a more comprehensive change in the way urban water markets function, with an aim to introduce competition at every point in the value chain. In particular, their proposal removes the presence of the monopoly water supplier. It suggests that private parties, so-called virtual water suppliers, be given the right to own the water stock held in dams. New inflows into the storages are proposed to be periodically auctioned off to these private water owners. Virtual suppliers compete with one another in the provision of bulk water. The implementation of this proposal, when combined with effective third-party access to urban water networks, would lead to a competitive bulk water market. Retailers would coordinate the supply of bulk water in each period with consumer demand. Entry into the retail market would have relatively low costs. Thus the retail market would also be competitive. In this way the volumetric price would be efficient at every point in the value chain, giving all market participants appropriate price signals. Note that under these arrangements consumers need only negotiate supply with retailers, and do not require the hydrological expertise of the participants in rural water markets.

Introducing competition in urban water markets, as proposed by Ofwat, Young, McColl and Fisher, or Sibly and Tooth has not been attempted before. It may be advisable to introduce these reforms gradually, so that unforeseen difficulties can be identified early before they cause too much disruption. Indeed, Ofwat (2008) presents a timetable for the gradual introduction of competition into UK urban water markets. Similarly, a variant on the Sibly and Tooth approach would be to allow some limited private ownership of stored water in parallel with the existing allocation system. The role of the private market could be expanded gradually over time, until eventually all urban water was traded on an open market.