Introduction Background NewberyPollitt Counterfactual CEGBplan Efficiency OtherEfficiency Restructuring Environment Summary Distribution Conclusions

At the time of the British utility privatisations in the 1980s there was great debate as to what their consequences would or could be, but until recently there has been little rigorous appraisal of what the effects have actually been..Among the most serious and systematic studies have been two social cost benefit analyses of the British electricity industry privatisations. Newbery and Pollitt examined the effects of restructuring and privatising the Central Electricity Generation Board (CEGB). The CEGB was previously responsible for generation and transmission in England and Wales. Domah and Pollitt examined the effects of privatising the twelve Area Boards that were previously responsible for local distribution and retail supply in England and Wales.

Newbery and Pollitt concluded that there had been a net gain of nearly £10 billions in generation and transmission, equivalent to a permanent cost reduction of 5% a year in generation. However, they calculated that more than all of this had gone to producers: consumers were worse off. Domah and Pollitt concluded that there had been a gain of £6 billions in distribution and supply. Of this, £1 billion would go to consumers. However, this latter consumer gain was all in prospect - up to 1998 consumers were estimated to have lost.

ISNIE founding president Ronald Coase has remarked that important new empirical results in the natural sciences are soon tested and retested to check their validity, whereas this seems not to happen so frequently in economics. Since privatisation was a major characteristic of economic policy in the 1990s, in the UK and elsewhere, and remains potentially significant in many other countries, some further testing of these calculations and conclusions is in order.

This paper is a report on research in progress that seeks to re-examine the evidence on which these calculations and conclusions are based. It focuses on the Newbery-Pollitt study of the CEGB privatisation. The present paper looks at three main items. First, the counterfactual assumptions on plant build and operating costs; second, the valuation of the environmental gains; third, the required rate of return and the value of assets to which this is applied. The first two affect the total net benefit of privatisation; the third has important implications for the distribution of these benefits.

Future research will attempt to update the previous studies in other respects, taking into account substantial reductions in costs and prices to customers since those studies were conducted, but this has not yet been attempted.

The main aim of the Newbery-Pollitt paper was to assess the simple question whether electricity privatisation and restructuring was beneficial or not. It deliberately took a conservative approach, and showed that the policy had a positive net present value, even on assumptions most favourable to the public sector counterfactual. The aim of the present research is to obtain more accurate and more up-to-date estimates of the extent and distribution of costs and benefits than was possible at the early stage when the Newbery-Pollitt research was carried out. In particular, the present research will reassess the controversial calculation that customers did not gain from privatisation. In due course it will also assess the claim by Branston - that observed electricity prices are significantly higher than they would have been had privatisation not occurred.

The next section summarises the Newbery-Pollitt paper (henceforth NP). The subsequent sections take in turn some of the key assumptions and indicate the alternative assumptions that we have made, and discuss the results that follow from this. It should be emphasised that this paper is a preliminary report on work in progress, and much of the anticipated research is yet to be undertaken. Nevertheless, the initial results seem of interest, and comments are welcomed.

From 1948 to 1990 the British electricity supply industry was under public ownership. The CEGB was a vertically integrated statutory monopoly that operated all transmission and generation in England and Wales. At Vesting on 31 March 1990, the CEGB was formally split into four successor organisations: three generators National Power, PowerGen and Nuclear Electric, and the transmission business National Grid Company (NGC). (In practice the split began to take effect earlier.)

Until 1990 there were twelve Area Boards in England and Wales, each a vertically integrated statutory monopoly responsible for distribution and retail supply in its own area. At Vesting these were privatised as Regional Electricity Companies (or RECs).

NGC had certain duties with respect to operating the newly created Pool. NGC was initially jointly owned by the twelve RECs, but in 1995 NGC was privatised as an independent company. Nuclear Electric was initially kept in the public sector, but the more modern nuclear stations were privatised in 1996 as British Energy.

After privatisation, new entry and competition were allowed in generation. There has been considerable entry of new generating plants, mostly using Combined Cycle Gas Turbine (CCGT) technology; some of these plants were initially part-owned by the RECs. The two successor companies National Power and PowerGen inherited all the CEGB’s coal-fired and oil-fired plant; they have since built CCGTs and have also disposed of many of their own coal and oil-fired stations.

NP introduced their paper as follows.

NP’s approach was to estimate the efficiency gains from restructuring and privatisation (which they call R&P) and competition by calculating the difference in costs of generation and transmission between two scenarios: the actual outcome and a counterfactual outcome in which the CEGB retained responsibility for these functions. The period studied includes the period before restructuring, effectively up to 31 March 1988, and the period from which the new structure was legally in place from 1 April 1990. NP had data up to April 1996, and made projections until March 2010. This included calculating the capital value of generating plant at that date, which in turn depends on post-2010 cost assumptions. They argued (plausibly) that any effect of R&P on price made negligible difference to the efficiency benefits of privatisation, given the low elasticity of demand, though such price differences would affect income distribution.

NP considered that the main benefits of R&P were threefold. First was the direct impact on the efficiency with which existing plant and fuels were used to generate electricity - that is, the extent of operating cost reductions. Second was the effect of relaxing the CEGB’s obligation to use high cost British coal, and the effect of new entry. These factors could be expected to influence the pattern of fuel use and the nature of investment in new plant, and hence impact on capital and operating costs. Third was the impact on the environment of changes in fuel use, particularly of reduced emissions of acid rain and carbon dioxide. Against these three types of benefits NP set the costs of initially restructuring the industry.

Since several of these items are matters of conjecture, NP "attempt to bracket the likely range of plausible scenarios with a case most favourable to R&P, and one least favourable." (p. 286) NP’s preferred scenario is the "pro-privatisation scenario" reflecting "business as usual" in the counterfactual CEGB, but they also calculate a "pro-CEGB scenario" in which the CEGB management pursue a more commercial approach than hitherto.

We share NP’s preference as to the relative plausibility of the two scenarios. It is difficult to accept certain key aspects of the "pro-CEGB scenario" - for example, that the CEGB would have abandoned its planned nuclear programme, and that coal prices would have been the same as if there had been no electricity privatisation. In what follows, therefore, we concentrate on NP’s preferred pro-privatisation scenario. However, NP’s pro-CEGB scenario is of interest in that NP find that privatisation was beneficial even under the pro-CEGB assumptions that are particularly unfavourable to it. Also, we find it easier at this stage of our research to carry out certain calculations with respect to that scenario.

NP make calculations according to two discount rates: the then-obtaining public sector test discount rate of 6%, and a higher rate of 10%. Their preferred rate is the former, and we too concentrate on that case.

On these assumptions, NP estimate that the efficiency benefits of R&P (in the pro-privatisation scenario) would have a present value of £8.8 billions (at 1994/5 prices [in April 1995]). Against these efficiency savings NP set restructuring costs of £2.8 billions. These were partly redundancy costs and partly provisions for possible future costs. NP estimate that the fuel savings and lower investment costs from switching fuels would amount to £3.6 billions.

The net total of all these benefits and costs is £(8.8 - 2.8 + 3.6) billions, making a total net benefit of £9.6 billions.

NP’s preference is to ignore environmental gains whose cash value is hard to measure. However, if such gains were included, NP calculate that there would be a further £2.3 billion associated with lower sulphur and carbon emissions. So the total NPV of benefits of electricity privatisation including externalities would be £(9.6 + 2.3) = £ 11.9 billions.

For comparison, NP’s pro-CEGB case has estimated efficiency benefits of £7.6 billions less the same restructuring costs of £2.8 billions. NP estimate that fuel and investment costs would be £0.7 billion higher (rather than lower) in the event of privatisation. This gives a total NPV of benefits of £(7.6 – 2.8 – 0.7) = £4.1 billions. The estimated environmental gains in this scenario are £1.9 billions.

In sum, NP’s estimate is that, on the preferred pro-privatisation assumption, net benefits total nearly £10 billions without environmental externalities and £12 billions with these externalities. On the pro-CEGB assumption the corresponding figures are about £4 billions and £6 billions. Broadly speaking, then, NP suggest that the net benefits of privatising the CEGB lie in the range £4 to £10 billions, with an additional £2 billions for environmental gains. NP also report the effects of certain sensitivity analyses that we discuss further below. Table 1 summarises the above figures

(£ billions NPV in 1995. Source: NP Table IV)

Efficiency benefits (cost savings) 7.6 8.8

Restructuring costs -2.8 -2.8

Fuel and investment savings -0.7 3.6

NPV of benefits 4.1 9.6

Environmental externalities 1.9 2.3

NPV of benefits incl externalities 6.0 11.9

Before privatisation, the CEGB primarily built and operated coal-fired and nuclear generating stations. NP note that after privatisation, "the pattern of investment and fuel choice changed dramatically, with the ‘dash for gas’, the abandonment of the planned programme of building a series of standard PWR nuclear power stations, and the collapse of the market for British coal". (pp. 285-6)

What stations would the CEGB have built if privatisation had not taken place? NP use information about the objectives and plans of the CEGB taken from its evidence presented to the Hinkley Point Inquiries and published in 1990. (These Inquiries concerned planning permission for a new nuclear station Hinkley C on the Somerset coast.) The CEGB’s central forecast was that it would need 12.5GW of new plant by the end of the century to meet the increase in demand and the replacement of 5.5 GW of aging coal plant.

On the basis of the Inquiries evidence, NP’s (preferred) counterfactual assumption is that, in the absence of privatisation, the CEGB would have built two 1.3 GW PWR nuclear plants and 2 GW of large coal plants equipped with Flue Gas Desulphurisation (FGD). It would have retrofitted a further 2 GW of existing coal plant with FGD, and built CCGT plants to preserve the reserve margin and to meet sulphur emissions limits. It would have kept imports from France at the 1988/9 levels (5 TWh less than under privatisation).

As noted, NP’s estimated savings in fuel costs and investment costs from R&P amount to a total of £3.6 billions. These savings vary considerably by fuel. There is a saving of £3.3 billions from not building so much nuclear plant. There is an extra cost of £2.6 billions from importing more French electricity, at market prices rather than (as before) at a price midway between the avoidable costs on the two systems. The French also have to be paid a premium by virtue of their exemption from the fossil fuel levy that funded the non-fossil fuel subsidy. There is a saving of £16.4 billions from not building so much coal plant, from lower coal prices and from lower coal purchases. More gas plant (CCGT) is built and operated than under the CEGB, at an extra cost of £13 billions, even though gas prices are assumed lower than they otherwise would have been. There is a small increase in the use of oil and orimulsion plant, costing £0.5 billion. These savings total £(3.3 – 2.6 + 16.4 – 13.0 – 0.5) = £3.6 billions.

The Hinkley Point Inquiries opened in October 1988, and the Inspector’s Report was published in [month] 1990. The Report provides a useful indication of the CEGB’s thinking over that period. However, it is important to recognise that before the Inquiries started the Government had already indicated its intention to privatise the electricity industry, that the CEGB plans and arguments initially submitted to the Inquiries already reflect a change in Government and CEGB thinking that seems unlikely to have occurred in the absence of impending privatisation, that as a result of impending privatisation the Government’s thinking on nuclear matters was evolving throughout the Inquiries, and that as a consequence the CEGB’s own submissions changed during the course of the Inquiries. Much of this is documented in the Report itself.

The Report also indicates how some of the nuclear aspects of policy developed.

The Report provides some further background detail about this change of policy.

The background to this changing policy was that the Government was grappling with how best to implement its manifesto commitment to privatize electricity alongside its commitment in the same manifesto "to proceed with the next phase of our nuclear programme … to go on playing a leading role in the task of developing abundant, low-cost supplies of nuclear electricity". On closer investigation with a view to privatisation, it had become increasingly apparent that the nuclear stations were far from a low-cost supply.

It is now generally well known that the proposed 70-30 structure reflected a view that, because of their costs and risks, it would not be feasible to privatise the nuclear stations as a separate entity, but that they could be privatised as a minor component of a much larger company with a predominantly non-nuclear asset base. After further consideration, the Government concluded that it was not feasible to privatise the older Magnox nuclear stations. After yet further consideration, the Government had to withdraw all the nuclear stations from the privatisation.

The overall picture is thus one of a steadily increasing realisation that nuclear stations were uneconomic. The Government was driven to this conclusion, against its will, by the commercial assessments of its financial advisers consequent on its decision to privatise. There seems no reason why it would have come to this conclusion otherwise.

In consequence, the true nature of the policy that would have been carried out in the absence of privatisation has to be found in earlier statements than those presented to the Hinkley Point Inquiries. The CEGB had in fact submitted investment plans to Government on an annual basis for many years before that. At the time when the Government first came to consider privatisation in detail, our understanding is that, to maintain the momentum of the nuclear programme, the CEGB had plans for completing Sizewell B and five further PWRs by 2000. A further six PWRs were envisaged by 2003. There were also plans for two or three large coal-fired plants. [either this or last generation plan as per below - documentation in either case?]

[The CEGB’s 1988 (and last) generation plan assumed a far greater build of PWR and coal plant than its proposal to the Hinkley Point Inquiry scenario. This last annual plan assumed 5.4GW of new coal-fired plant (three stations, each with two 900MW sets) coming on stream 1995-97, plus a total of eight PWRs with one coming on stream each year from 1997 to 2004. ]

The CEGB considered that CCGTs were an untried technology. Moreover, CCGTs burned what Britain and the European Community labelled a premium fuel, and the EC at that time banned CCGTs from use in generation. The CEGB expected that CCGTs would contribute at most 1GW by 2000.

[need to relate to large plant directive and ending of ban on use of gas]

In practice, the CEGB’s annual plan would have been reassessed during the course of implementation, and planning difficulties or the innovative nature of the very large coal plants might have delayed the plant-building programme. The counterfactual is necessarily uncertain. However, the pro-nuclear and pro-coal sentiment at the CEGB should not be underestimated. In our view the likelihood is that significantly more nuclear and large coal sets would have been built than indicated by the CEGB’s evidence to the Hinkley Point Inquiry, and the CEGB’s last annual plan [if we have it!] is a more reliable basis for a counterfactual assumption.

We have simulated the CEGB’s last annual plan using the same outputs and plant parameters as NP. The CEGB expected the larger coal sets to be cheaper overall than existing sets. However, we have not made any adjustments to the cost figures on the grounds that teething problems associated with new technology might well have offset such cost savings. Like NP, we assume that residual output requirements are met by new CCGT plant. This implies 1GW of CCGT capacity by 2000, 6GW by 2005 and 26GW by 2010, in contrast to NP assumptions of about 5 GW, 15 GW and 40GW of CCGT at those dates. It is conceivable that the CEGB’s plan might have been revised to bring in more CCGTs before 2000, but it is also arguably unrealistic to assume that the CEGB would have built no more coal-fired stations after 1997, even though environmental considerations might have begun to weigh more heavily against coal. Future modelling will assume that the CEGB would have made a more gradual shift to CCGTs.

The NPV of generation cost and fuel savings of our counterfactual CEGB scenario is £[11.8] billions more expensive than NP’s actual (plus projected) privatisation scenario, even assuming that counterfactual fuel prices were the same as under privatisation. This is in contrast to NP’s calculation that the counterfactual based on the Hinkley Point Inquiry plan would be £3.6 billions more expensive than the actual (plus projected) privatisation scenario, even assuming that counterfactual fuel prices were lower as under the actual privatisation scenario. In other words, on our assumptions, privatisation reduced generation and fuel costs by some £[8.1] billions more than NP calculate.

NP’s pro-privatisation scenario assumed that domestic coal prices would have been 5% higher at each level of output than under privatisation (because of less pressure to increase productivity), and that gas prices would have been 10% higher (because there would have been less ‘dash for gas’ and a more carefully managed market).

This seems unduly optimistic. Electricity privatisation exerted – and was intended to exert – major pressure to bring the UK coal industry into line with international efficiency and prices. We ran a further variant of the model assuming that UK coal prices declined 1% a year in nominal terms to 1994/95 and 1% a year in real terms thereafter. This provided a further £[0.7] billion saving in addition to the £11.8 billions saving (in which prices were assumed to be the same as under privatisation). Updating late 1990s gas prices, which were slightly lower than assumed by NP, provided a further £[250] million saving. These two modifications suggest that electricity privatisation saved in total some £13.3 billions of fuel and investment cost, or some £[5.6] billions more than NP estimate.

NP note that, as a result of privatisation, Electricite de France (EdF) "is now able to sell at market prices rather than at a price mid-way between the avoidable costs on the two systems, in addition to enjoying a non-fossil fuel subsidy of £97 million per year." (pp. 292-3) The present value of this subsidy (in the form of exemption from the fossil fuel levy) in 1995 was £727 million. (fn. 21 p. 293) The extra cost of French imports is £2.6 billions. It is not clear what proportion of this reflects the higher price charged, as opposed to the production cost of the increased level of output.

NP treat the extra revenue paid to the French as a social cost. This may be valid from a British perspective but it understates the total net benefit of privatisation. For better or worse, a portion of the benefits accrued to overseas interests. (This reflected, amongst other things, the fact that EdF was joint owner of the interconnector to France.) It therefore seems appropriate to note the benefits of privatisation accruing to French interests via EdF. The change in payments to EdF on the assumption of the same output is shown as £1.5 billion in NP’s pro-CEGB scenario

As to the efficiency of using existing plant, NP note that "before R&P, the CEGB experienced the same productivity growth as UK industry, but after R&P the rates of growth in each successor generating company (including Nuclear Electric) dramatically accelerated. National Grid… also increased its rate of growth of productivity, but not so sharply." (p. 284)

NP assume as a counterfactual that unit (non-labour) costs remain constant at their average value in 1985/6 - 1987/8, but that unit labour costs fall in proportion to the expansion of nuclear power (i.e. the total labour costs of nuclear generation remain constant and are spread over an increasing volume of nuclear output). NP’s central estimate is that the CEGB would have achieved the same increase in nuclear output as Nuclear Electric until 1995/96. This means that the CEGB would have achieved a cost reduction of 1% a year during this period.

NP further assume (p 288) that after 1995/96 (looking into the future at the time of NP’s study) the annual productivity gain in the counterfactual case grows at the rate of growth of electricity supply (1% a year). This implies that the counterfactual CEGB would have increased productivity just as rapidly as the privatised industry would.

On the basis of these assumptions NP estimate that privatisation led to efficiency savings of £8.8 billions (or £7.6 billions in the pro-CEGB case).

NP acknowledge that there was reason to doubt whether the improvement in nuclear output would have taken place as rapidly in the absence of privatisation. It is questionable whether there would have been any increase at all. It is also very unlikely that the counterfactual CEGB would have increased productivity after 1996 as fast as the privatised industry did. We hope shortly to model something more realistic in both these respects.

Pending our revised calculations, we note that NP carried out a sensitivity analysis by assuming that CEGB costs would have fallen by 1% p.a. instead of the assumed 1.5% p.a.(p. 289) This adds another £3.6 billions to the previously calculated £8.8 billions of efficiency savings, making a total of £12.4 billions for this element.

NP subtract from efficiency gains an estimated £2.8 billion costs of restructuring. NP comment on these as follows.

Future research will explore this point. In the meantime, we leave the estimated restructuring costs at the original level of £2.8 billions, though if NP’s judgement is correct, this is an overstatement of costs or an understatement of benefits.

NP estimate the environmental impact of privatisation by showing that the successor companies produced a substantially lower volume of harmful emissions, mainly as a result of the switch from coal to nuclear and gas. They then evaluate these reductions, relative to the counterfactuals, by valuing reductions in SO2 emissions at £125/tonne and CO2 emissions at £12/tonne carbon. In the pro-privatisation case with 6% discount rate this gives benefits of £1.0 billion for sulphur and £1.4 billion for carbon, making a total of £2.3 billions (evidently some rounding in the separate figures). For the pro-CEGB case they calculate total environmental benefits of £1.9 billions.

The revised counterfactual that we have proposed would have more nuclear stations, though less output from each station as a result of lower productivity. It would also have more new coal stations and fewer CCGTs. The overall impact of this is to produce a similar quantity of emissions (slightly less SO₂ and more CO₂) compared to NP’s pro-CEGB scenario, but less emissions than in NP’s pro-privatisation scenario.

There is considerable uncertainty about the appropriate values to attach to the projected emissions. We take in turn the values of sulphur and carbon.

1. Sulphur

NP derive the value of £125/tonne for sulphur emissions from Newbery’s earlier calculation of the value needed to justify the costs of fitting FGD. NP suggest that this figure is arguably on the low side (p. 290) compared to Department of the Environment figures for the incremental damage of SO2 in the range £1329 to £2372/tonne. NP do not adopt such figures because on that basis the environmental gains of privatisation would exceed all other benefits. NP also note that one reason for scepticism about the figures is that the larger part of the damage is extra mortality costs, and there is epidemiological reason to think that this element may have been overstated by about two orders of magnitude.

This suggests that, rather than ignoring the damage figures, there is a case for revising them to take account of the overestimate mentioned. However, neither NP’s paper nor the Dept of the Environment study to which it refers indicate what proportion of the figures reflects the suggested over-estimated extra mortality costs.

Recent EU modelling puts even higher figures on the marginal external costs of SO2 emissions than the figures cited by the Dept of the Environment. They suggest euro 4500/tonne SO2 in rural areas of the UK (against an EU average of euro 5200), and euro 6000/tonne for a city of 100,000 people anywhere in the EU. This last figure is increased by a factor of 5 for a city of 500,000 , 7.5 for a city of 1 million, and 15 for a city of several million people. These figures are of the order of double the UK figures cited.

In contrast, prices have been much lower in the SO2 emissions trading scheme already established in the US. Prices have varied considerably, from a minimum price of $65/ton SO2 in early 1996 to a maximum price of slightly over $200 observed from roughly mid-1998 through mid-1999 and then briefly in 2001. From the end of 2001 until mid-2002, SO2 prices remained around $160 but have since fallen to around  $130 in November 2002.

At an exchange rate of $1.4 to £1, NP’s assumed value of £125/tonne SO2 in 1995 money is equivalent to about £125 x 2000/2240 x 1.4 x [1.23] = $190/short ton SO2 in 2002 money. This is towards the top end of the range of observed in the US traded market.

If we follow NP in using the cost of avoiding the SO2 damage, rather than the cost of the damage itself, then the traded values will be a better indication. Many factors could affect the market value of traded SO2 emissions in the event that a market was created in the UK, not least the severity of the underlying emissions constraints. However, for the moment there does not seem sufficient justification for assuming a value per unit SO2 saved that is different from the values observed in the US market.

We therefore take a range of values from one third below that which NP assumed, corresponding to about $130 in the US market today, to the same level as NP assumed, corresponding to about $200 in the US market. These correspond to total values of SO2 emissions savings ranging from £0.7 bn to £1 bn, or in other words up to £0.3 bn less than NP assumed.

(b) Carbon

NP’s value of £12/tonne for carbon emissions is "a rough estimate of the possible externality cost of global warming" (fn. 18, p. 290, based on calculations by Pearce).

Since that time, there have been several EU studies of the marginal abatement costs of CO2 emissions. The commonly used figure is euro 20/tonne CO abated, in 1999 money. This seems to be the estimated value that would be established in a traded EU market sufficient to achieve the Kyoto target of an 8% reduction in each country’s emissions in 2008 - 12, relative to 1990 emissions levels. This calculation is based on the assumption that the ACEA agreement is implemented. [check meaning] The figure is of course subject to several caveats. If the ACEA agreement is not implemented, the cost would rise to euros 33. If there were no trading, the values could range up to euros 100, with an average of about euros 43, although the estimate for the UK is about euros 33 in those circumstances. More generally, it is estimated that the UK would be a net exporter of CO2 emissions savings, because the marginal abatement cost would be lower there than in much of the rest of Europe. In the absence of EU trade, therefore, the market value of CO2 in the UK could be below euros 20/tonne CO2. All these values were estimated some three years ago and are due to be revised in mid-2003.

At 1.56 euro to £1, the EU benchmark of euro 20/tonne CO2 in 1999 money corresponds to about £11.57/tonne CO2 in the 1994/5 money of the NP study. Adjusting for the difference in atomic weights, £11.57/tonne CO2 is equivalent to 11.57 x 44/12 = £42.42/tonne carbon. This is about three and a half times the NP assumed value of £12/tonne carbon.

There is now tangible evidence of UK market prices for carbon saving. The UK emissions trading scheme auction that first took place in March 2002 cleared at £53.37 per tonne of CO2. However, this was for a total of three tonnes per bid, so the effective price was 53.37/3 = £ 17.8/tonne CO2. Because of certain constraints at the time [explain], this was regarded as on the high side. .Subsequent trading opened at around £5/tonne CO2. The market price rose eventually to about £12/tonne CO2, but fell back. A trade was reported to us at £7.50/tonne CO2 in November 2002.

The bottom (£5) end of the traded range is equivalent to £15.14/tonne carbon in 1995 money, the upper (£12) end to £36.34/tonne carbon.

The former value is about 25% higher than the NP assumption, the latter value is approximately the same as the EU benchmark, at three times the NP assumption. For present purposes we take this as the revised range for the value of carbon savings. Given NP’s calculated total present value of carbon savings as £1.4 bn, the range we use is 1.4 x 1.25 = £1.8 bn to 1.4 x 3 = £4.2 bn.

Combing the two emissions, this gives a total present value of emissions savings ranging from 0.7 + 1.8 = £2.5 bn to 1.0 + 4.2 = £5.2 bn.

[NOx emissions?]

Our findings so far are summarised in the third column of Table 2 The total NPV of benefits without considering externalities is nearly £23 bn, which is more than double the NP calculation in the most favourable (pro-privatisation) case. Emissions externalities add a further £2.5 to £5 bn. depending on the assumptions made.

(£ billions NPV in 1995. Source: NP Table IV plus present text for revised)

Efficiency benefits 7.6 8.8 12.4

Restructuring costs -2.8 -2.8 -2.8

Fuel and investment savings -0.7 3.6 13.3

NPV of benefits 4.1 9.6 22.9

Environmental externalities 1.9 2.3 2.5 to 5.2

NPV of benefits incl. externalities 6.0 11.9 25.4 to 28.1

NP make some calculations of the possible distribution of benefits of privatisation, with a view to calculating in particular the effects on consumers. They use their assumptions about counterfactual costs, make further assumptions about counterfactual profits, and derive prices and revenues that they compare with the privatisation outcome and scenario. In deriving profits they say, "We start from a projection of costs and the CEGB’s requirement to earn 3.75% on CCA assets (which at 1994/5 prices are £25 billion), to give the CEGB’s forecast price." (p. 294)

We suggest two modifications to this calculation. First, the CEGB asset value had been written down by £7-8 billion (in prices of the day) in 1988/89. This was largely in response to the scrutiny resulting from the privatisation process, and in the light of the likely proceeds from selling the assets on the open market. It seems unlikely that the assets would have been revalued if privatisation had not taken place. The relevant asset value for counterfactual calculations would therefore be about £35 billions rather than £25 billions (in 1994/5 prices). Using this higher value for purposes of calculating the counterfactual price level would increase CEGB prices and revenues by about £0.9 billion per year. The net present value of this over the period would be £11.1 bn.

Second, there is the question of the counterfactual rate of return on these assets. The Hinkley Report summarises the situation as follows.

1983/4 - 1984/5 (2 years) 1.4%

1985/86 - 1987/88 (3 years) 2.75%

1988/89 3.75%

1989/90 4.75%

The key questions are whether these rises in required rate of return would have taken place in the absence of privatisation and, if so, whether they would have continued to an even higher level.

A familiar argument is that the industry was simply "being fattened for privatisation" so that the Treasury could get the higher present value from the higher consequent stream of revenues. On this basis a plausible counterfactual is a return towards the lower end of the above spectrum. Thus Branston

On this basis, Branston assumes a return of 3% from 1988/89 to 1992/93 and 3.25% from 1993/94 to 1997/98.

Similarly DP say

DP nonetheless assume a counterfactual return of 4.75%, equal to the required return in 1989/90.

An alternative view is that the return on assets in the electricity supply industry as a whole had been unusually low during the early 1970s and early 1980s. Reasons for the low return included the pressure on costs from high fuel prices [dates?] and from the miners’ strikes in 1974/5 and 1984/5. [This was also a time of relatively high inflation.] The Government was reluctant to see these cost increases passed through to customers at the time. However, as these pressures eased, the Treasury was very keen for the industry to secure a higher and more economic return, and would have been keen even in the absence of privatisation.

This concern about under-pricing was not confined to the electricity industry. It reflected a wider and growing concern within the Treasury. The (Labour Government’s) 1978 White Paper (Cmnd 7131) introduced the concept of the Required Rate of Return (RRR) on each enterprise’s new investment programme as a whole. This was to be 5 per cent, based primarily on the real pre-tax returns achieved in the private sector. The Government also proposed to set a financial target for profitable industries in terms of pre-interest (and tax) return on average net assets employed. In addition, there was reference to publication of performance indicators. A subsequent (Conservative) Treasury Memorandum in June 1981 confirmed the above approach.

In 1984 the Treasury asked Ian Byatt to chair a committee to advise the government on the accounting policies of the nationalised industries. The first main finding of the Byatt report, published in 1986, was that "Current Cost Accounting (CCA) is especially important in nationalised industries, because of the long lives of their assets and because their economic performance has to be assessed in the absence of fully competitive markets." (page 4) It pointed out that in 1983 the CCA return on UK private sector companies was 9 per cent, whereas the CCA return on the water and electricity industries was under 3 per cent. (page 5)

The Government reflected this in its policy.

Since 1984 the Treasury has provided guidance on economic appraisal for Government departments, in a manual known as the Green Book. The 1991 Green Book, which superceded the guide issued in 1984, specified that a figure of 6 per cent in real terms should be used as a discount rate, reflecting the cost of capital based on the pre-tax long-term cost of capital for low risk purposes in the private sector. However, it noted that the RRR for the nationalised industries was usually set at 8 per cent, and commented "The figure of 8 per cent is higher than 6 per cent because it is an average return, as opposed to a marginal cut-off return. It is based on average returns on assets achieved in the private sector for activities with low cyclical year by year variability." (Annex G, para 6, page 77) The 1997 Green Book updated the 1991 edition, but did not change the guidance in relevant respects for present purposes.

To summarise, during the 1980s there was an increasing determination that, in the absence of reasons for subsidy, nationalised industry rates of return should reflect those in the private sector, and be based on CCA asset valuations. The Required Rate of Return for nationalised industries, set in 1978 at 5 per cent on new CCA investment, was increased to 8 per cent in 1989 to reflect an increase in comparable private sector returns. Financial targets for individual nationalised industries were generally less than the RRR, and it was recognised that increases in them might have to be phased in over time.

All this suggests that the policy of increasing the required return on assets in the electricity industry would have happened anyway, even without privatisation of that industry. Whether and if so how far it would have continued beyond 4.75% is more difficult to judge. Most of the potential comparator nationalised industries had been privatised. The main industry remaining in public ownership was the Post Office. As the following brief account shows, its required returns were indeed increased over the period after the electricity supply industry was privatised.

In 1988/89 the Post Office group as a whole had a target of 3.25% and achieved 2.5%. Its main component, the Letters business, had a 2.8% target in 1987/88 and 1988/89, and achieved 3.4% and 2.6% in those two years. In January 1990 the Government announced profit targets for the three years to 1991/92. "The target returns in the last year of the period equate to a real rate of return of 8% on capital employed. The Government announced, in March 1989, that public sector industries should aim to achieve an 8% real return on new investment. The Government accepted that the Post Office should progress to the new rate of return over the target period." The targets in the three years were 2.4%, 6.3% and 11.0% for the Post Office as a whole, and 1.0%, 6.8% and 12.2% for Royal Mail (previously known as Letters). Royal The Group as a whole failed to meet the second year target, but met the other two. Royal Mail met all three targets comfortably.

Thereafter no further targets were set for the Group as a whole. Royal Mail’s targets for the next three years 1992/93 to 1994/95 were 15.0%, 15.2% and 14.5%, all of which were again comfortably achieved (no less than 29.2% in the third of these years). Since the end of that period, it appears that targets have been set on an annual basis with no overall guideline. From 1995/96 to 1998/99 they were 16.8%, [?], 20.0% and 20.0%. All these targets were met with a margin to spare. After this time the concept of the required return seems to have been superseded by the concept of the [public sector?] dividend.

There is thus evidence that a major un-privatised but potentially privatisable nationalised industry was indeed set targets that gradually increased to (and perhaps beyond) the level of 8% real return on investment. Whether required returns in the counterfactual electricity supply industry could and would have increased at such a rate is another matter. Much must have depended on the implications for prices, which in turn depended also on the pattern of operating costs and capital investment. Over the period discussed the Post Office was being set, and to a large extent apparently achieving, targets for real unit cost reduction, and this would have had implications for the political feasibility of the increased returns on investment.[check extent of price increases]

For the present, we calculate that if the electricity industry had been set and had achieved a target rate of 4.75% during the whole period of the study, then the NPV of the extra payments from customers to the government (in addition to the extra payments on the assets not written down) would have been about £10.1 bn higher than NP calculate. The possibility of an increased return beyond 4.75% will be explored in conjunction with the performance of the counterfactual CEGB on costs and the implications for prices.

NP assumed that prices under privatisation and under continued nationalisation would converge at some date after privatisation. They then calculated a gain or loss to customers by comparing prices in their scenarios with those under privatisation, for the period until that date. On the basis of their calculated return on capital, plus projected operating costs, and in their pro-privatisation case, NP calculate that consumers were worse off by £1.3 billions in the event that prices in the privatised industry converged to the (assumed lower) level of those in the counterfactual CEGB by the year 2000, and thereafter remained aligned with such prices. If the convergence did not take place until 2010 then consumers would be worse off by £6.1 billions. The corresponding figures for the pro-CEGB case were £5.4 billions and £9.4 billions, respectively.

The difference between a 3.75% return on £25 billions and a 4.75% return on £35 billions is almost £[ ] billions a year (= £0.9 + [ ] billion). The present value (at 6% discount rate) in 1995 of that sum each year from 1990-2010 is £21.5 billion. In other words, if privatisation had not gone ahead, and if these revised assumptions about asset valuation and return on assets had been used as the basis for estimating prices in the counterfactual, consumers would have paid some £21.5 billion higher prices than NP assume.

Whether and when an 4.75% return would have been set and obtained in the absence of privatisation is of course debatable. We estimate that to achieve a 4.75% return on pre-privatisation asset values would have required about a [5]% real increase in final prices at the time of privatisation. It is unlikely that the CEGB would have been required or allowed to achieve this in one step. Some phasing over time is more probable, perhaps offsetting what would otherwise have been falling prices. However, it is very probable that a higher return than 3.75% would have been sought and achieved, and on a value of assets that had not been written down to facilitate privatisation.

On this basis, NP’s calculated losses to consumers from privatisation, ranging from about £1 billion to about £9 billions depending on the scenario, would have been gains ranging from about £1 billion to £9 billions if the assets had not been written down and the required return stayed at 3.75%, and would have been gains ranging from about £[12-20] billions if, in addition, the required return had increased to 4.75%.

Such gains could exceed NP’s calculated total net benefits from privatisation. However, they would not exceed the net benefits we have calculated . It is therefore quite possible that customers and the government (taxpayers) could have benefited as well as investors.

The detail of these assumptions and outcomes is of course arguable, and as the research progresses we intend to examine the distributional impact in more detail. Nonetheless, if it is accepted either that CEGB assets would not have been written down in the absence of privatisation, or that a 4.75% rather than 3.75% return on assets would eventually have been achieved, it follows that, on the basis of NP’s other assumptions, privatisation of the CEGB led to gains and not losses to consumers, relative to what would otherwise have occurred. Making both assumptions increases those gains.

NP ask whether the attempts of the companies to diversify after privatisation have led to benefits. They comment that the evidence so far does not suggest a startlingly successful investment record.

It would be very time-consuming, if indeed it were possible at all, to track down each investment made by each successor company. Some diversifications may turn out to be loss-making, as NP suggest. However, it seems doubtful that private investors would accept losses on a continuing basis. Is there reason to think that the eventual outcome of successor company diversification should be significantly different from performance of the stock market as a whole?

It is nonetheless possible to update the history of the first and most striking example given by NP, and the only one they cite that relates to successor companies of the CEGB.

Energis had a chequered history. It was subsequently floated on the stock market, and at the height of the "dot.com" boom had a reported value of around £14 billions. However, Energis crashed with many other telecommunications investments after the collapse of the "dot.com" boom. What if any value it might have in future remains to be seen.

One way of valuing Energis for present purposes is to compare the investment that NGC put into it with the value that NGC received from it. NGC sold two-thirds of its Energis shares in several tranches at the time of the Energis flotation and thereafter. We are informed (personal communication) that NGC’s total investment was of the order of £500 millions and that its proceeds from sale of Energis shares were of the order of £2.2 billions. This suggests a net value from this particular diversification of some £1.7 billions, rather than a loss. It might be argued that this is a private gain, reflecting a transfer, rather than a social gain. For the moment we omit this item from our calculations.

Newbery and Pollitt set out a pioneering calculation of the costs and benefits of privatising the Central Electricity Generation Board in England and Wales. They found that net benefits ranged between about £4 billions and £10 billions excluding environmental benefits, depending on the counterfactual assumptions. Their preferred scenario gave the higher value. They indicated that they had erred on the side of caution in estimating the gains from privatisation, and their main sensitivity analysis suggested that net benefits might be between £1 and £3 billions higher.

We have reconsidered some of the assumptions used by Newbery and Pollitt, principally their counterfactual assumptions about investment in generation. We suggest that, in the absence of privatisation, the CEGB would have built more nuclear and coal-fired plant, and that this would have been more costly. We have adopted some of the benefits that Newbery and Pollitt calculated but did not incorporate in their preferred scenario. We have also calculated an updated value for the main diversification example to which they referred. The outcome of our calculations so far is a net benefit from CEGB privatisation exceeding £20 billions, more than double NP’s calculation for their pro-privatisation case.

Newbery and Pollitt estimated that the environmental benefits of this privatisation, in terms of reduced emissions of sulphur and carbon, were worth about £2 billions. Our assumptions yield similar amounts of these emissions. However, we have been able to update the values of these emissions in the light of recent trading. On these revised assumptions, the savings from reduced emissions are worth something in the range £2.5 to £5 billions.

Newbery and Pollitt made some rough calculations of the distribution of the benefits of this privatisation. They found that net benefits to customers were negative. Customers were between £1 billion and £9 billions worse off as a result of privatising the CEGB.

We suggest here that the prices to customers in the absence of privatisation would have been higher than Newbery and Pollitt assume. This is because the Government would have required a higher return than had hitherto been received, on a value of assets that had not been written down. The extent and timing of such a higher return is necessarily debateable. However, if the return were kept at 3.75% and there were no write down of assets, the costs to customers of non-privatisation – and hence the benefits to customers of privatisation - would be some £11 billions higher than Newbery and Pollitt assumed. If in addition there were an increase in the return to 4.75%, as was actually announced, the benefits to customers would be a further £10 billions. On NP’s range of other assumptions, customers would be significantly better off as a result of privatisation - by up to about £20 billions - rather than worse off.

We emphasise how difficult it is to assess the net benefits of privatisation, how sensitive the numbers are to different assumptions, and how uncertain any conclusion must be. We also emphasise that our results are provisional and need checking and will be extended. However, we believe that it is reasonable to say that, considered relative to the course of action on which the CEGB was engaged, privatisation of generation and transmission may have yielded net benefits of over £20 billions plus environmental benefits of about £2 to 5 billions. Moreover, on Newbery and Pollitt’s other assumptions, a good part of these benefits went to customers.

We intend to widen our analysis to other areas and to update the work using information on recent years that was not available to the earlier studies. Our expectation is that widening and updating will yield further net benefits to electricity privatisation.

Updating is not as straightforward as it might seem. In previous studies it was comparatively straightforward to calculate what had actually happened to date, and the challenge lay in constructing a plausible counterfactual. Most of this paper has been concerned with the specification of an alternative counterfactual scenario. In contrast, the main problem in updating the data is in quantifying what has actually happened. Many firms have entered the industry and they do not necessarily produce accounting information that identifies their electricity activities separately. Even the ESI successor companies are no longer required to produce separate accounts for their activities in competitive markets such as generation and supply. Updating therefore requires either extrapolation of previous data using general market price reports or disaggregation of final electricity prices by deducting information on prices charged by the transmission and distribution network businesses in order to estimate the costs and prices of the generation and supply businesses. Neither approach is simple.