Is bigger better for water and sewerage capital programmes?
Cost and efficiency one might assume go hand in hand - the larger the programme the more efficiently it can be delivered and vice versa. Not necessarily true for capital programmes in the water sector, Colm Gibson explainsThere is universal recognition of the importance of efficient delivery of capital programmes in the water sector. Plenty of effort is deployed during the price control cycle to measure, benchmark, review, audit and report on how efficiently water and sewerage companies are delivering their capital programmes. Ofwat, WICS and NIAUR all place significant pressure and incentives on companies to deliver their programmes efficiently and all the companies I have worked with share this aspiration.
Given the total amounts of money involved, this approach is to be expected. What is surprising, however, is that there has been comparatively little effort put into determining the optimal size of such programmes.
Traditional water industry regulatory thinking links big capital programmes with greater efficiency. There are, after all, economies of scale that make the smallest programmes relatively less efficient. On the face of it, everyone wins from larger capital programmes. Larger programmes mean more can be done for the environment and to improve quality, reliability and security of supply for customers.
Shareholders benefit as larger programmes increase the regulatory asset value on which the regulators allow a return to be earned, and it's good for the economy as the utility contracting industry has more work. It's not even expensive for customers (your water undertaker will charge only about 90p to deliver a tonne of water to your house, at any time of the day or night 24/7/365 - there is no other product in that price bracket). Surely then, bigger is better?
It turns out that there are some very real constraints that place limits on the size of capital programmes that water and sewerage companies can deliver efficiently. If companies are asked to deliver particularly large programmes, they may bear additional costs and efficiency penalties that are seldom understood by regulators, who instead regard such companies as being less efficient, or at least slow to improve efficiency.
As part of an assignment for the Water Industry Commission for Scotland (WICS), LECG was asked to establish which factors would constrain the size of capital programme that Scottish Water could deliver efficiently. This was to inform WICS' view as to whether the next capital programme was of a size that Scottish Water could be expected to deliver without incurring a material efficiency penalty.
We drew evidence from a broad range of sources. With remarkable consistency, each piece of analysis reinforced a key message: there was an efficiency penalty for companies expected to deliver a particularly large capital programme (relative to company size) relative to those with a midsized capital programme.
We were able to use our analysis to suggest that the optimal size of the capital programme for Scottish Water for 2010 and beyond should be significantly smaller than Scottish Water's current capital programme, which averages over £600M per year. WICS took account of LECG's findings in the Draft Determination of prices published on 30 June 2009, for a capital programme of around £480M per year, based on a five-year price control.
Key areas of our analysis included improvements in capital efficiency, project completion rates, disruption to customers, planning and third party constraints, skills bottlenecks and inflationary pressures. In each case, the evidence suggested that the largest capital programmes were less efficient than mid-sized programmes.
There are several reasons to believe that there may be a link between the efficiency of capital delivery and the relative size of capital programmes. The larger a capital programme is relative to the size of company, the greater proportion of management time and resources occupied in its delivery.
It seems intuitive that companies could struggle to deliver relatively large programmes efficiently - they would need to focus simply on delivery, without the luxury of diverting management resources to projects to improve or secure efficiency of delivery. Similarly, companies with particularly small programmes may be unable to secure benefits of scale in their procurement operations to the same extent as companies with larger programmes.
Separate analysis required
There are many factors, however, that impinge upon company efficiency - not least the historical efficiency levels of the company. Further, many companies have both sewerage and water capital programmes that tend to be set and delivered largely independently of each other. Any given company could have a relatively large water capital programme and a relatively small sewerage programme or vice versa, and indeed be efficient in one and not the other. Analysis, therefore, needs to be undertaken separately for both services.
This difficulty is compounded by the relatively limited amount of data in the public domain that makes this distinction. Indeed, Ofwat's published efficiency metrics provide only a broad band within which a company's efficiency has been judged to lie. This makes the data very quantised. To complicate matters further, Ofwat changed its efficiency banding mechanism at each periodic review, making comparisons over time harder. To deal with this it is important to adopt a variety of analytical techniques including investigating correlations, efficiency frontier analysis and quartile analysis.
One of the advantages of efficiency frontier analysis is that it can be used to filter out the impact of uncorrelated factors that influence efficiency. The logic of this being that, if a company is on the efficiency frontier itself, then all of the uncorrelated factors must be going in the company's favour. Our results show a clear pattern: namely that the frontier companies with particularly large or particularly small capital programmes did not improve efficiency as fast as those with mid-sized programmes.
Even for companies not on the frontier, similar results are observed. For example, Fig 2 shows that, on average, companies in the third quartile in terms of programme size improved their Ofwat efficiency ranking by 8% more than those with capital programmes in the largest quartile. Companies with relatively large capital programmes should not be expected to improve their capital efficiency as quickly. This has implications for regulators' efficiency assumptions on frontier shift and rates of catch-up.
Basically, if a company is given a particularly large capital programme, Ofwat should not assume that it will improve its efficiency as quickly as it might otherwise. Ofwat may need to assume a different frontier shift for particularly large capital programmes or, at least assume different optimal rates of catch-up for these companies.
Colm Gibson is a principal in LECG's regulation team specialising in advising companies and regulators on price control and efficiency matters.