Abstract
In this study we estimate a variable cost function on a panel of English and Welsh Water and Sewerage companies, observed over two full regulatory periods (1995–2004). The main aim of the paper is to investigate the presence of a ratchet effect in the cost cutting activity associated with the price cap regulatory regime. By applying the Generalized Index of Technical Change approach suggested by Baltagi and Griffin (J Polit Econ 96(1):20–41, (1988)), we provide some empirical evidence consistent with the existence of regulatory cycles. In particular, firms’ cost cutting activity tends to increase in the early phase of the regulatory cycle, while it weakens as the price review approaches. D24, L51, L95.
This is a preview of subscription content, log in via an institution to check access.
Similar content being viewed by others
Notes
The other side of the coin is that the regulator has to leave costly rents to the firm and therefore a sub-optimal degree of allocative inefficiency would arise.
In particular in the neoclassical framework it is not possible to separately identify movements towards the frontier from movements of the frontier. We thank an anonymous referee for a useful suggestion on this issue.
Ofwat uses econometric analysis to disentangle cost differences among firms due to different operating environment from those due to inefficiency. After having formed a view about the relative efficiency of each water company, Ofwat groups water companies into five efficiency bands and differentiates the X factor by requiring water companies belonging to the different bands to close a given percentage of the estimated gap with the industry frontier over the regulatory period ahead.
Similar mechanisms have been put in place by Ofgem. See Ofwat (2003) for an accurate description of Ofwat’s methodology applied over the 2000–2004 regulatory period.
Anglian Water’s view on the public consutation “Ofwat forward programme 2007–2008 to 2009–2010” available on www.ofwat.gov.uk/aptrix/ofwat/publish.nsf/AttachmentsByTitle/fpres070307anh.doc/$FILE.
The interested reader is referred to Saal et al. (2007) for a more comprehensive review of those studies that have been conducted on efficiency, technical change and productivity growth in the case of the English and Welsh water and sewerage industry.
For a robustness check on this issue see Sect. 5.
In our empirical specification we have also included other technical variables, such as the average pumping head (which is a proxy for the consumption of energy); the densities of the water and sewerge networks, expressed in both cases as properties connected per km of network; the trade effluent intensity (which is a proxy for the importance of industrial users for the sewerage service). However, none of these variables turned out to be statistically significant, and therefore we have dropped them from the analysis.
Homogeneity can be imposed by normalizing the dependent variable and factor prices with the price of one of the inputs: we normalized for the price of other costs (this normalization procedure is equivalent to impose the following restrictions: \( \sum\nolimits_{j = 1}^{J} {\beta_{j} = 1} \); \( \sum\nolimits_{j = 1}^{J} {\beta_{js} = 0} \); \( \sum\nolimits_{j = 1}^{J} {\rho_{jn} } = 0 \); \( \sum\nolimits_{j = 1}^{J} {\phi_{jv} } = 0 \)), thus reducing the components of the P vector to one. Symmetry of the cost function is imposed by assuming that \( \beta_{js} = \beta_{sj} \), \( \gamma_{np} = \gamma_{pn} \);\( \delta_{vx} = \delta_{xv} \); \( \rho_{jn} = \rho_{nj} \); \( \phi_{jv} = \phi_{vj} \) and \( \chi_{nv} = \chi_{vn} \) before estimation. Concavity of the cost function is verified if the Hessian is a negative semi-definite matrix, while monotonicity in factor prices requires that costs rise as a factor price rises; finally monotonicity in output requires positive marginal costs. Both monotonicy and concavity conditions are checked after the estimation.
Given the potential sensitivity of the estimates to the starting values, we carried out some robustness analysis by experimenting with different starting values, which reassured us that the parameter estimatates were robust to the starting values used in the estimation routines. All the regressions have been run with the software EViews5.
The joint modelling of the Wascs and Wocs presents some problems. In fact, as the Wocs are not active in the sewerage sectors, conventional flexible functional forms like the translog cannot be used. One possibility would be to estimate more flexible functional forms, like the Composite, that allow for the possibility that some firms do not produce some outputs (Bottasso et al. 2007). Moreover, Saal and Parker (2005) have shown that neglecting the joint production of water and sewerage services by the Wascs could lead to biases in the coefficients. In this work we have therefore decided to focus on the ten Wascs which, however, represent most of the industry turnover.
In 2000 Welsh Water contracted out the operations of the water and sewerage services to a neighbouring Wasc. As a result, Welsh Water included in the statutory accounts only the salaries paid to its own employees, which are however not representative of the true price for the labour input. In order to avoid possible distortions, we dropped Welsh Water from the sample for the last 4 years of our panel.
During our sample period, other mergers involved the Wascs, but because of average size differentials as large as eighty times, we thought that it was reasonable not to create “new Wascs” after those mergers.
Other sources of data employed in this study are the Wascs statutory accounts.
The basic assumption of this normalisation is that the expenditure on materials, consumables, and power is roughly proportional to the capital stock. This is confirmed by a correlation index of 0.98 among capital stock and the aggregate input measure. See, for the water industry, Saal and Reid (2004) and Filippini et al. (2008) and Fraquelli et al. (2004) in a multi-utility setting.
Our results are virtually unaltered when we use different price deflators such as a wage index for the UK economy and a price index for intermediate inputs for the UK water industry both taken from the University of Groningen database.
A similar choice for the outputs of the water and sewerage services is found in Saal and Reid (2004).
We also used the RPI index. However, our main results did not seem to have been materially affected by the choice of the price index. Therefore, we decided to report the results obtained with the Construction Output Price Index.
For some robustness analysis on this issue see the Results section.
Each Wasc’s water supply area is divided into water supply zones. In each water supply zone Ofwat considers a set of tests on a series of indicators related to drinking water quality. Each water company has to report, for each of its water supply zones, the percentage of tests that meets a given threshold for eight key parameters, such as faecal coliforms, taste, odor, nitrate, aluminium, iron, lead, pesticides.
Each variable has been normalized by its sample median so that the first order coefficients can be interpreted as the elasticity of variable costs with respect to each variable. In particular, the first order coefficients of the input prices, capital stocks and outputs should be interpreted as the cost elasticities in the first year of the panel.
The most common explanations are that the positive capital elasticity is the result of an Averch-Johnson effect due to the rate of return regulation features of the regulatory regime as well as of the structural features of the water and sewerage industry, where most infrastructures are built in order to meet future demand -which is assumed to grow. Filippini (1996) argues that the positive elasticity of the capital stock might be due to a multicollinearity problem that would arise when there is a positive correlation between the capital stock and variable costs.
The point estimate for the long run scale economies computed as \( {\frac{{1 - \sum {\partial \ln {\text{VC}}/\partial \ln K} }}{{\sum {\partial \ln {\text{VC}}/\partial \ln y} }}} \) is about 0.91.
All technical change figures we have computed are referred to an hypothetical median company in each year (i.e., the figures used for the capital stocks, factor prices and outputs in equation 8 are referred to the median Wasc in each year).
Saal et al. (2007) report a technical change figure of about 2.4% and 1.95% for the regulatory periods 1990–95 and 1995–00, respectively. Therefore, technical change, although positive, exhibits a declining trend which started at the beginning of the 1990s.
Indeed technical change takes on a negative value also in 2001, which is not significantly different from zero.
We have also estimated a model without the companies’ fixed effects, and the results were virtually unchanged.
In particular, in the last year technical change is not significantly different from zero: this implies that there has been an improvement with respect to the test year, when it was significantly negative.
Saal and Reid (2004) estimate a variable cost function with a time trend specification for the ten Wascs for the period 1992–2002 (a partially overlapping time period with that used in our study) and end up with technical change estimates that decline over time—although at a slower pace than that identified in our paper—and that remain in the last year of their sample (2002/2003) as high as 1.7%, which is one full percentage point higher than in our time trend specification. The reason for this discrepancy might be in the different time span and in the partially different control variables and model specifications employed in their paper but, perhaps more importantly, in the dramatic fall in technical change in 2004 identified by the Generalized Index approach.
References
Baltagi BH, Griffin JM (1988) A general index of technical change. J Polit Econ 96(1):20–41. doi:10.1086/261522
Baltagi BH, Griffin JM (1995) The measurement of firm specific indexes of technical change. Rev Econ Stat 77(4):654–663. doi:10.2307/2109813
Baltagi BH, Rich DP (2005) Skil-biased technical change in US manufacturing: a general index approach. J Econom 126:549–570. doi:10.1016/j.jeconom.2004.05.013
Bloch H, Madden G, Grant Coble N, Savage SJ (2001) The cost structure of Australian telecommunications. Econ Rec 77(239):338–350. doi:10.1111/1475-4932.00027
Bottasso A, Conti M (2009) Scale Economies. Evidence from the English water only sector. regional science and urban economics, Technology and Technical Change in the Water Industry. Reg Sci Urban Econ 39(2):138–147. doi:10.1016/j.regsciurbeco.2008.05.017
Bottasso A, Conti M, Piacenza M, Vannoni D (2007) The appropriateness of the poolability assumption for multiproduct technologies. Hermes working papers 6
Dalen DM (1995) Efficiency improvement investment and the ratchet effect. Eur Econ Rev 39:1511–1522. doi:10.1016/0014-2921(95)00007-Z
Diewert WE, Wales TJ (1992) Quadratic spline models for producer’s supply and demand functions. International Econ Rev 33(3):705–722. doi:10.2307/2527134
Dyck IJA, Di Tella R (2002) Cost reductions, cost padding and stock market prices: the Chilean experience with price cap regulation. Harvard NOM research paper 22 (3)
Europe Economics (2003) Rolling schemes in price control reviews. London
Filippini M, Hrovatin N, Zoric J (2008) Cost efficiency of Slovenian water distribution utilities: an application of stochastic frontier methods. J Prod Anal 29(2):169–182. doi:10.1007/s11123-007-0069-z
Fraquelli G, Piacenza M, Vannoni D (2004) Scope and scale economies in multi-utilities: evidence from gas, water and electricity combinations. Appl Econ 36(18):2045–2057. doi:10.1080/0003684042000245543
Giulietti M, Waddams Price C (2005) Incentive regulation and efficient pricing. Ann Public and Cooperatives Econ 76(1):121–149. doi:10.1111/j.1370-4788.2005.00273.x
Hunt L, Lynk E (1995) Privatization and efficiency in the UK water industry: an empirical analysis. Oxford Bull Econ Stat 57(3):371–388
Joskow PJ (2006) Incentive regulation in theory and practice: electricity distribution and transmission networks. Centre for Energy and Policy Research
Kumbhakar S (2004) Productivity and technical change: measurement and testing. Empirical Economics 29(1):185–191. doi:10.1007/s00181-003-0185-1
Laffont JJ (1994) The new economics of regulation ten years after. Econometrica 62(3):507–537. doi:10.2307/2951658
Laffont JJ, Tirole J (1993) A theory of incentives in procurement and regulation. MIT Press, Cambridge
Meyer MA, Vickers J (1997) Performance comparisons and dynamic incentives. J Polit Econ 3(105):547–581. doi:10.1086/262082
Ofwat (2003) A further consultation on incentives mechanisms: rewarding future out-performance and handling under-performance of regulatory expectations. Office of Water Services, Birmingham
Pint EM (1992) Price-cap versus rate-of-return regulation in a stochastic cost model. Rand J Econ 4(23):564–578. doi:10.2307/2555906
Saal D, Parker D (2005) Assessing the performance of the water operations in the English and Welsh water industry: a panel input distance function approach. Aston business school working paper 0502
Saal D, Reid S (2004) Estimating Ofwat productivity growth in English and Welsh water and sewerage companies: 1993–2003. Aston business school working paper 0434
Saal D, Parker D, Weyman-Jones T (2007) Determining the contribution of technical change, efficiency change and scale change to productivity growth in the privatized English and Welsh water and sewerage industry: 1985–2000. J Prod Anal 28(1–2):127–139. doi:10.1007/s11123-007-0040-z
Author information
Authors and Affiliations
Corresponding author
Rights and permissions
About this article
Cite this article
Bottasso, A., Conti, M. Price cap regulation and the ratchet effect: a generalized index approach. J Prod Anal 32, 191–201 (2009). https://doi.org/10.1007/s11123-009-0140-z
Published:
Issue Date:
DOI: https://doi.org/10.1007/s11123-009-0140-z