The calculations feature in a study for the Drinking Water Inspectorate (DWI) by consultants WRc and Oxera. The work, which examined both the costs and benefits of meeting the incoming EC standards for lead in drinking water, appears to be the first of its kind in Europe.
The current limit for lead is 50µg/l. The new EC Directive on drinking water, now close to adoption, would cut this to 25µg/l in 2003, and then to 10 µg/l in 2013 (ENDS Report 273, pp 34-37 ). The exact implications will not be clear until a sampling protocol for lead is agreed later this year.
The DWI's consultants compared the costs and benefits of meeting the standards with a "do nothing" scenario. Separate evaluations were carried out for the standards as absolute values and as averages. Costs and benefits from 1998 to 2034 were discounted at a 6% interest rate and expressed in 1995 prices.
An important influence on costs will be the extent to which orthophosphate dosing will reduce the plumbsolvency of drinking water sufficiently to reduce lead levels below 10µg/l. The alternative, replacing lead plumbing, is much more costly.
Confident predictions on the success of orthophosphate treatment are not yet possible for all water supply zones. WRc drew on the best available data, and assumed that all lead pipes would have to be replaced where treatment failed.
The study projects that the limit of 25µg/l will not be met in 2-10% of supply zones in England and Wales without extra measures, with lead pipe replacement being needed in 39-58% of these. An estimated 20-63% of zones will fail to meet the 10µ g/l limit without extra measures, requiring replacement of lead pipes in 5.5-9.8 million properties. The figures display wide variation because of different assumptions about the sampling protocol yet to be agreed at EC level.
The above figures apply where the lead limits are maxima. In that case, the total cost of meeting the 10µg/l limit would be about £6.5 billion, with householders paying about three times more than water companies. Costs would fall if the limits were expressed as averages, possibly to as little as £1.7 billion, depending on the sampling protocol used.
The cost data are subject to large uncertainties, but provide the basis for a reasonable estimate of costs once the compliance regime for lead is defined. In contrast, major assumptions and uncertainties figure in the study's estimates of health benefits.
The relatively low blood lead levels now found in the UK population are linked to a variety of health effects. No attempt was made to value the benefits of reducing lead in water for some of these - such as effects on gestational age, and on hearing impairment and growth in children - because models linking them with blood lead levels are not well developed.
The two main health impacts considered were IQ and blood pressure. A large body of evidence suggests that reductions in blood lead from 20µg/dl - now rarely exceeded in the UK - to 10µg/dl are associated with an IQ increase of about two points, and the study assumes that a similar relationship prevails below 10µg/dl. Good evidence is also available to develop a similar association between changes in blood lead and blood pressure.
Significant assumptions had to be made in quantifying the relationship between reductions in lead levels in drinking water and lead levels in blood. The latest studies on this point were carried out in the early 1980s, since when there have been large reductions in blood lead levels in the UK as well as changes in the sources of exposure to the metal.
The consultants went on to place a value on the benefit of increased IQ by quantifying the links between IQ, years of schooling and performance in the labour market - in terms of both the probability of employment and level of earnings. Most of the data used in this exercise were drawn from US studies and are not necessarily transferable to the UK. In addition, as the study concedes, most links in this chain are subject to large uncertainties.
The report puts the IQ benefits of the new standards at £930-1,500 million if they are interpreted as maxima, falling to £560-940 million when the standards are treated as averages. The IQ benefits accounted for at least two-thirds of the overall health benefits.
The consultants also attempted to place values on a series of non-health effects. The most significant was reduced water leakage resulting from replacement of lead pipes, valued at a maximum of £269 million - an underestimate because the "externalities" associated with water abstraction were not included. The value of recovered lead pipes was put at up to £118 million.
The results are summarised in the table below. They show that the 25µg/l limit, if applied as a mean, gives net benefits regardless of the success of water treatment. However, the final standard of 10µg/l gives net benefits only if water treatment is successful.
If standards are applied as maxima, only the best case would yield a net benefit from the 25µg/l standard, while net costs are incurred in any scenario for the 10µg/l limit. Successful application of orthophosphate treatment could bring costs down significantly.
The above figures were derived using a discount rate of 6%. With rates of 2% and 4%, a net benefit was achieved from bringing lead levels down to 10mg/l, though only in the best case scenario.
The study raises many important questions. Some are methodological. For instance, it assumes a smooth relationship between IQ and years of schooling, and this in turn is critical to its estimates of individuals' earning capacity. However, in practice individuals do not in general experience the part-years of schooling which the assumption implies. This is important because only a relatively large IQ variation of about ten points changes an individual's time in school by a year.
There is also the wider question of whether changes in individuals' earning capacity are the only or even the most important consequence of changes in IQ. The report alludes to this, noting that "it could...be argued that increases in productivity are not the only benefit arising from a more highly educated population, which may generate wider social benefits.
"Such intangible social benefits are not captured through the increased money value of wages commanded by individuals with higher IQ. To the extent that such positive externalities exist, then the values presented here could be regarded as underestimates." But some people may believe that the consultants' cautious approach to the merits of education is unwarranted - and that the study is a poor basis for policy decisions precisely because of its narrow focus on the relatively quantifiable effects of IQ changes.