Content uploaded by Allyson Pollock
Author content
All content in this area was uploaded by Allyson Pollock on Apr 19, 2016
Content may be subject to copyright.
data are available on long term trends in exposure to
sunlight in Britain to confirm that such a reduction has
taken place. Reduced exposure to sunlight could have
adverse effects, but we believe that any advice to increase
exposure to sunlight is premature given the tentative
nature of our review and concerns about the changing
nature of sunlight exposure with the thinning of the
ozone layer.
24
However, we suggest that the basis for cur-
rent advice to reduce exposure to sunlight should be
reviewed in a formal and quantitative manner so that the
potential benefits and harm from exposure to sunlight
can be conveyed to the public. The risk:benefit ratio will
differ between individuals; for many people the small
absolute increase in risk of melanoma could easily be
outweighed by the effect of reduced sunlight on mood.
A recent article in Vogue suggests that lay understanding
is, perhaps again, ahead of medical thinking in attempt-
ing to weigh up factors for and against exposure to sun-
light.
25
Perhaps, while we await the conclusions of such
formal analyses, those of us who enjoy spending time in
the sun can rest (on our deck chair, sun lounger … or
whatever) assured that the chance that we will be one of
the people dying from our tan is small.
We thank Mr Sunil Thakar of the M P Shah Hospital, Jamnagar,
for arranging a visit to the solarium and providing details of its
history.
Competing interests: None declared.
1 Secretary of State for Health. The health of the nation. London:
Department of Health, 1992.
2 Melia J. Skin cancer. Health Hyg 1995;16:153-8.
3 Arthey S, Clarke VA. Suntanning and sun protection: a review of the psy-
chological literature. Soc Sci Med 1995;40:265-74.
4 Bridgwood A, Malbon G, Lader D, Matheson J. Health in England 1995.
What people know, what people think, what people do. A survey of adults aged
16-74 in England carried out by Social Survey Division of ONS on behalf of the
Health Education Authority. London: HMSO, 1996.
5 Carter S. Who wants to be a “peelie wally”? Glaswegian tourists’ attitudes
to sun tans and sun exposures. In: Clift S, Grabowski P, eds. Tourism and
health. London: Pinter, 1997.
6 Frankel SJ, Davison C, Davey Smith G. Lay epidemiology and the ration-
ality of responses to health educators. Br J Gen Pract 1991;41:428-30.
7 Finkel E. Sorting the hype from the facts in melanoma. Lancet
1998;351:1866.
8 Elwood JM, Jopson J. Melanoma and sun exposure: an overview of pub-
lished studies. Int J Cancer 1997;73:198-203.
9 Office for National Statistics. 1995 Mortality statistics. Cause. England and
Wales. London: Stationery Office, 1997.
10 West SK, Duncan DD, Munoz B, Rubin GS, Fried LP, Bandeen-Roche K,
et al. Sunlight exposure and risk of lens opacities in a population-based
study. The Salisbury eye evaluation project. JAMA 1998;280:714-8.
11 Adami J, Frisch M, Yuen J, Glimelius B, Melbye M. Evidence of an associ-
ation between non-Hodgkin’s lymphoma and skin cancer. BMJ
1995;310:1491-5.
12 Freedman DM, Zahm SH, Dosemeci M. Residential and occupational
exposure to sunlight and mortality from non-Hodgkin’s lymphoma:
composite (threefold) case-control study. BMJ 1997;314:1451-5.
13 Grimes DS, Hindle E, Dyer T. Sunlight, cholesterol and coronary heart
disease. Q J Med 1996;89:579-89.
14 Brennan PJ, Greenberg G, Miall WE, Thompson SG. Seasonal variation
in arterial blood pressure. BMJ 1982;285:919-23.
15 Khaw KT. Temperature and cardiovascular mortality. Lancet
1995;345:337-8.
16 Scragg R, Jackson R, Holdaway IM, Lim T, Beaglehole R. Myocardial inf-
arction is inversely associated with plasma 25-hydroxyvitamin D
3
levels: a
community-based study. Int J Epidemiol 1990;19:559-63.
17 Vik T, Try K, Thelle DS, Forde OH. Tromsø heart study: vitamin D
metabolism and myocardial infarction. BMJ 1979;ii:176.
18 Lund B, Badskjaer J, Soerensen OH. Vitamin D and ischaemic heart dis-
ease. Horm Metab Res 1978;10:553-6.
19 Wehr TA, Rosenthal NE. Seasonality and affective illness. Am J Psychiatry
1989;146:829-39.
20 Chew KSY, McCleary R. The spring peak in suicides: a cross-national
analysis. Soc Sci Med 1995;40:223-30.
21 Utiger RD. The need for more vitamin D. N Engl J Med 1998;338:828-9.
22 McMichael AJ, Hall AJ. Does immunosuppressive ultraviolet radiation
explain the latitude gradient for multiple sclerosis? Epidemiology
1997;8:642-5.
23 Marks R, Foley PA, Jolley D, Knight KR, Harrison J, Thompson SC. The
effect of regular sunscreen use on vitamin D levels in an Australian popu-
lation. Arch Dermatol 1995;131:415-21.
24 McMichael AJ, Haines A. Climate change and health: implications for
research, monitoring and policy. BMJ 1997;315:870-4.
25 Hutton D. Health news. Vogue 1998 May:114.
(Accepted 19 April 1999)
The private finance initiative
PFI in the NHS—is there an economic case?
Declan Gaffney, Allyson M Pollock, David Price, Jean Shaoul
The private finance initiative substantially increases the
cost of hospital building. Total costs (construction costs
plus financing costs) in a sample of hospitals built
under the private finance initiative are 18-60% higher
than construction costs alone (table 1). Shareholders in
private finance initiative schemes can expect real
returns of 15-25% a year.
1
The consortiums involved in
these schemes charge the NHS fees equivalent to 11.2-
18.5% of construction costs (table 2). If the Treasury
were to finance new hospitals directly out of its own
borrowing it would pay a real rate of annual interest of
3.0-3.5%. It has been estimated that the £2.7 billion
Scottish private finance initiative programme will cost,
at a conservative estimate, “£2 billion more than if the
Treasury had acquired the assets directly.”
2
The higher
costs will be met locally through cuts in clinical spend-
ing and nationally through subsidies from NHS capital
budgets.
Medical staff are deeply implicated in hospital pri-
vate finance initiative schemes. Clinical directors
approve and medical directors sign off the full business
website
extra
Sources of data in
table and figures
are given on the
BMJ’s website
www.bmj.com
Summary points
Investment under the private finance initiative
costs more than public sector procurement. The
annual charge for the use of privately financed
facilities is between 9.1% and 18% of the original
construction cost, whereas government can
borrow at interest rates of 3.0% to 3.5%
The extra cost of private finance is disguised by
the Treasury’s insistence that NHS trusts discount
costs at 6% per annum when comparing the costs
of the private finance option with public sector
investment
The amount of risk transferred to the private
sector under privately financed deals has been
exaggerated, leading to spurious attributions of
additional value to private sector options
Education and debate
This is the
second of four
articles on
Britain’s
public-private
partnership in
health care
Correspondence to:
Allyson Pollock
allyson.pollock@
ucl.ac.uk
continued over
BMJ 1999;319:116–9
116 BMJ VOLUME 319 10 JULY 1999 www.bmj.com
case, clinical posts are lost, and heroic targets are set for
gains in medical productivity. Clinical concerns are
generally met by assurances that the largely undis-
closed price of the private finance initiative is well
worth paying because schemes approved by the initia-
tive offer better value for money than public
procurement. This claim is based on the fact that, for
approval purposes, all privately financed schemes are
compared with a notional publicly funded equivalent,
the public sector comparator. However, this compari-
son is carried out using an appraisal methodology
under which the cash payments associated with each
option are “discounted,” and costs are adjusted to
reflect “risk transfer.” Both these factors have an influ-
ence on the results of the comparison. The appraisal
methodology is prescribed in government guidance
and plays a crucial part both in the selection of
schemes for the initiative and in making the case for
the private finance initiative as a policy.
The procurement and approval process
The procurement process currently in operation in the
NHS departs from earlier approaches based on
non-financial ranking of options for meeting needs.
3
Now the procurement process requires that NHS
trusts prepare business cases for approval by the NHS
Management Executive and the Treasury. This process
reflects the criteria used by those bodies in appraising
schemes: financial viability and control of public
spending.
45
An NHS trust makes the case for the proposed
investment in an outline business case and gives an
estimate of the capital cost based on standard NHS
costs. Since 1994, all approved outline business cases
have to be tested for inclusion under the private
finance initiative. Final approval of schemes depends
on trusts producing a full business case that includes
an economic appraisal showing that the privately
financed option represents better value for money
than funding by the public sector.
Value for money and discounting (buy
now, pay later)
Calculations of value for money exploit the fact that
under public procurement all the costs of hospital
development are paid in the first few years, whereas
under the private finance initiative they are spread
over 25 or 30 years. To calculate the economic conse-
quences of spreading capital payments evenly
throughout the contract period (under the private
finance option) or paying them all in the first few years
(under the public sector option) a discounted cashflow
analysis is carried out. The principle is that money
spent now or in the near future carries a higher cost
than money spent several years down the road. This is
because more potential to earn interest will be
sacrificed by a sum that is spent immediately than the
same sum spent gradually over 30 or 60 years.In other
words, discounting introduces an interest rate
assumption. Discounting is widely used in the private
sector as it is assumed to maximise value for
shareholders. Its relevance to the public sector, where
profit maximisation is not the objective of investment,
is unclear.
Cost comparisons, once discounted, are expressed
as net present costs. The option with the lowest net
present cost is said to offer the best value for money. In
practice, this single figure is the basis of approval. Value
for money is assumed to provide neutral criticism, but
discounting clearly favours private finance with its pro-
tracted repayment schedule.
The level at which the discount rate is set
determines whether or not private finance option
shows value for money. The higher the discount rate
applied, the lower the value placed now on
expenditure in later years. Treasury guidance imposes
a discount rate of 6%, with the effect that £100 of
expenditure incurred in five years’ time has a “present
value” for appraisal purposes of £74.73, in 10 years of
£55.84, in 20 years of £31.18, and so on (box). Note
Table 1 Capital cost of private finance iniative schemes: construction costs compared
with total capital costs
NHS trust
Construction cost to
private sector (£m)
Total capital cost
(£m) Difference (£m)
Difference as % of
construction cost
Dartford 94.0 115.0 21.5 22.8
Carlisle 64.7 85.0 20.3 31.3
Norfolk 143.5 214.0 70.5 49.1
North Durham 61.0 96.0 37.0 60.6
Greenwich 84.0 109.9 25.9 30.8
Bromley 118.0 155.0 37.0 35.8
Wellhouse 54.0 65.0 11.0 18.5
The difference between construction costs and total capital costs is usually explained as arising from
financing costs incurred by the private sector during the construction period.
Table 2 Construction costs and availability payments* under the private finance
initiative
NHS Trust
Construction cost
(£m)
PFI availability
payment (£m)
PFI availability payment
as % of construction cost
Calderdale 64.6 8.7 13.5
Carlisle Hospitals 64.7 8.0 12.4
Dartford 94.0 10.5 11.2
Greenwich 84.0 11.0 13.1
North Durham 61.0 7.1 11.6
Wellhouse 54.0 10.0 18.5
*The availability payment funds private sector debt service obligations and returns to equity shareholders. A
minor element of the charge also funds maintenance costs over the life of the building.
Trusts with major (>£25m) capital developments under the private
finance initiative
Education and debate
Health Policy and
Health Services
Research Unit,
School of Public
Policy, University
College London,
London
WC1H 9EZ
Declan Gaffney,
research fellow
Allyson Pollock,
head
Social Welfare
Research Unit,
University of
Northumbria,
Newcastle upon
Tyne NE7 7XA
David Price,
research fellow
Department of
Accounting,
University of
Manchester,
Manchester
M13 9PL
Jean Shaoul,
lecturer
117BMJ VOLUME 319 10 JULY 1999 www.bmj.com
that for the purposes of appraisal, the costs to be
discounted are expressed in real terms.
The discount assumption affects fundamentally the
appraisal outcome. Table 3 shows that at 6% the
Carlisle private finance initiative scheme is slightly
cheaper than its public sector equivalent and is thus
held to be better value for money. When the discount
rate is reduced by only 0.5%, the outcome of the
appraisal is reversed and the public sector option
seems preferable. The economic advantage of public
sector procurement continues to increase as the
discount rate is reduced. Thus, economic advantage is
to a large extent a product of the discount rate used.
How is it determined?
Determining economic advantage
The 6% discount rate does not reflect interest rates on
government borrowing, any more than NHS capital
charges reflect the actual cost of public sector capital.
The choice of 6% was a policy decision. According to
Treasury guidance, “the practical choice of 6%, from
the top of the range . . . is an operational judgement,
reflecting, for example, concern to ensure that there is
no inefficient bias against private sector supply.”
4
The 6% discount rate favours private finance and
obscures the central characteristic of private finance:
the higher cost of capital. Therefore, economic
appraisal assumes from the outset what it is held to
prove: the economic advantage of private finance.
Risk transfer
Despite the use of a 6% discount rate, comparisons of
net present costs have been to the advantage of public
sector options in almost all cases. However, in private
finance initiative appraisal, all costs are “risk adjusted.”
The principle of risk adjustment is that, in order to
make a fair comparison of costs between private
finance initiative options and public sector compara-
tors, account needs to be taken of risks which under
public procurement the public sector carries itself, but
which under private finance initiative it pays another
agent, the private investor, to bear. When the cost of
public sector options is adjusted to reflect this transfer
of risk, the apparent cost disadvantage of the private
finance initiative options disappears. In most cases this
is done through adding a lump sum representing the
cost of risk to the net present cost of the public sector
comparator. Table 4 shows that risk transfer is valued at
between £20 million and £73 million in a sample of
schemes and has a decisive effect on the outcome of
economic appraisal. Both discounting and risk adjust-
ment seem to be necessary if the private finance initia-
tive is to show value for money.
There are several problems with the risk adjust-
ment carried out in the appraisal of private finance ini-
tiative schemes. One is that the 6% discount rate
already takes account of an element of risk, as it is set at
a level that is deemed by the Treasury to be higher than
a risk-free interest rate.Interest rates are after all largely
determined by the level of estimated risk associated
with an investment. As the value for money test
involves discounting costs at 6% and then adjusting the
comparator for risk, the cost of risk is effectively
counted twice.
A further problem is the tendency for trusts
seeking approval for private finance initiative develop-
ments to ascribe risks to private finance initiative con-
sortiums that they have not in fact taken on. Risk can
only be transferred through the private finance
initiative contract, by means of financial penalties
imposed on consortiums for failing to meet their obli-
gations. This basic principle is consistently overlooked
in economic appraisals of the private finance initiative.
At Carlisle, one of the risk supposedly transferred was
that targets for clinical cost savings would not be met,
and the cost of this risk was estimated at £5m.
6
The
consortium, however, had no responsibility for
Discounted cashflow analysis
The table below shows a worked example of
discounted cashflow analysis, with a discount rate of
6%. The result is that expenditure of £1000 in annual
instalments of £100 over 10 years, starting next year, is
held to be equivalent to expenditure of £736 this year.
This result depends entirely on the discount rate
used
—
when a 4% discount rate is used, the figure is
£811.
Year Total cash flow (£) Discounted cash flow (£)
000
1 100 94.34
2 100 89.00
3 100 83.96
4 100 79.21
5 100 74.73
6 100 70.50
7 100 66.51
8 100 62.74
9 100 59.19
10 100 55.84
Total 1000
Net present cost 736.02
Table 3 Effect of varying the discount rate on results of economic appraisal in Carlisle
hospitals’ private finance initiative scheme
Discount rate (%)
Public sector option
(PSC) (£000s)
Private sector option
(PFI) (£000s)
Economic advantage of PFI over
PSC (£000s)
6.0 174 337 172 633 1 704
5.5 185 803 186 692 −889
5.0 198 884 202 043 −3 159
4.5 213 900 219 480 −5 580
4.0 231 247 239 388 −8 141
3.0 275 027 288 622 −13 595
0 549 882 577 048 −27 166
Table 4 Risk added to public sector comparator: net present costs over 60 years
Trust
PFI net present
cost (£m)
PSC net present cost
Before risk
adjustment (£m) Risk added (£m)
After risk adjustment
(£m)
Calderdale 1221 1191 73 1264
Carlisle 173 152 22 174
Dartford 928 881 55 937
Durham 177 153 24 177
Wellhouse 1206 1210 20 1230
PFI=private finance initiative; PSC=public sector comparator
Figures may not sum due to rounding. Figures are taken from original full business cases and may not
reflect later adjustments. In some cases, figures represent the entire cost of running the hospital over 60
years, leading to very high net present costs; in others, only the services to be provided under the private
finance initiative contract are included, so that the net present cost relates to the scheme alone and not to
the costs of clinical activity.
Education and debate
118 BMJ VOLUME 319 10 JULY 1999 www.bmj.com
ensuring that these savings would be made, and faced
no penalty if they were not: £5m of additional value
was thus attributed to the private finance initiative
scheme on quite spurious grounds.
Constr uction period risks
The bulk of the risk supposedly transferred relates to
the building period, the first three to five years, rather
than to the operational phase, the subsequent 25-30
years. In the Greenwich Hospital scheme, seven of the
eight risks said to have been transferred to private
investors related to the construction phase of the
project and therefore could not threaten the income
of the investors during the operational phase.
7
Restricting risk to the construction period means that
income during the operational phase is guaranteed by
the Treasury for the life of the contract and is
sufficient to recoup virtually risk free the whole capital
cost of the hospital, together with a return on the
investment. This makes private finance initiatives a
very safe investment.
One of the main risks during the construction
period is that of cost overruns. NHS trusts have based
their estimates of the risk of construction costs
overrunning on historical claims about the level of
cost overruns associated with earlier publicly funded
schemes. One high profile scheme which is regularly
used as evidence of overruns due to public sector inef-
ficiency is the Guy’s Hospital phase III development.
The increase in cost at this scheme was so extreme as
to affect the annual cost performance statistics for the
NHS as a whole. However, the National Audit Office’s
report on the scheme shows that among the major
factors contributing to the overrun were the effects of
inflation and the new liability for value added tax that
came with NHS trust status. The significance of Guy’s
phase III as an exemplar of public sector inefficiency
has thus been exaggerated. The average increase in
cost over approved tender sums for NHS capital
projects has been between 6.3% and 8.4% in the
1990s.
8
Public finance initiative business cases have in
most cases assumed that public sector projects
overrun by 12.5% or more. In costing its public sector
comparator, the Norfolk and Norwich Trust assumed
overruns of 34%.
9
Private finance initiative business cases have tended
to exaggerate the inefficiency of public sector
procurement, leading to overestimation of the eco-
nomic benefit of the private finance initiative.
Moreover, many of the risks associated with public
sector procurement arise from the relationship
between NHS trusts and other parts of the public sec-
tor. Such risks can be managed internally and it is
strange to assume that they should, or indeed can, be
transferred to the private sector. Even if they could, why
should the taxpayer pay to offload a risk that can be
addressed by internal action on the part of the NHS
Executive?
The market’s assessment of risk
The best indicator of the extent of risk actually
transferred in the private finance initiative contract is
the interest rates paid by consortiums to their lenders
(as distinct from the interest rate consortiums charge
NHS trusts). First wave private finance initiative
schemes have achieved extremely favourable borrow-
ing terms on bank debt and bond issues on the basis of
“little inherent risk.” Market interest rates have been
between 4% and 5% in real terms. This suggests that
in the view of funders there has been very little risk
transfer.
This implies that there is little correspondence
between the sums NHS trusts ascribe to “risk transfer”
and the views of capital markets on the risks the
consortiums have actually taken on. In practice,
funders see little likelihood of the consortiums facing
financial penalties under the contracts and have set
interest rates accordingly. The legal advisers on the
North Durham scheme informed the NHS trust that
under the contract “various obstacles have been placed
in the way of . . . deductions being made.”
10
None the
less, the trust valued “risk transfer” under the contract
at £23m,
11
materially affecting the outcome of the
appraisal. Similarly, when Meridian plc launched a
bond issue to finance Greenwich Hospital, its prospec-
tus announced that it had “structured the contractual
arrangements for the Project such that there are
intended to be few risks inherent in the Project which
are retained by the Issuer.” The risk transfer charged
against the public sector comparator in the Greenwich
scheme was £20 million.
12
Conclusion
Formal appraisal of the private finance initiative is not
an objective process: it systematically reduces the
comparative advantage of public sector procurement
and disguises the basis of private sector costs. High
cost is ascribed to risk transfer but little risk is actually
transferred. The discounting method used to compare
the present value of different options is politically
determined and is set well above the government’s
interest rates. The government’s claim that the private
finance initiative represents better value than public
procurement is not supported, and clinicians should
not allow spurious economic arguments to deflect
them from criticising the clinical impact of private
finance initiative developments.
Competing interests: None declared.
1 Chantrey Vellacott DFK. Economics report. London: Chantrey Vellacott,
1999.
2 Fitzpatrick M. Letter to editor. The [Glasgow] Herald 1999 June 1.
3 Cook AN. The appraisal and financing of capital expenditure in the NHS.
London: CIPFA, 1996:3. (HFM/CIPFA occasional paper.)
4 HM Treasury. “The Green Book”: appraisal and evaluation in central govern-
ment. London: HMSO, 1992.
5 NHS Management Executive. Capital investment manual. London:
NHSME, 1994.
6 Carlisle Hospitals NHS Trust. Full business case addendum, appendix 5.
(Available from the trust.)
7 Shaoul J. The private finance initiative: looking glass world of PFI. Public
Finance 1999 Jan 29-Feb 4:14-16.
8 National Audit Office. Cost overruns, funding problems, and delays on Guy’s
Hospital phase III development. London: Stationery Office. 1998.
9 Norfolk and Norwich NHS Trust. Full business case, appendix 4.
(Available from the trust.)
10 North Durham Acute Hospitals NHS Trust. Summary contract
documents, p 40. (Available from the trust.)
11 Gaffney D, Pollock A. Downsizing for the 21st century: a report to Unison
Northern Region on the north Durham acute hospitals PFI scheme. London:
Unison, 1998.
12 Meridian Hospital Company plc. Bond issue prospectus. (Dated 9 July
1998.)
Education and debate
119BMJ VOLUME 319 10 JULY 1999 www.bmj.com
