Transport Policy 22 (2012) 2935

Contents lists available at SciVerse ScienceDirect

Transport Policy
journal homepage: www.elsevier.com/locate/tranpol

Consequences of differences in cost-benet methodology in railway

infrastructure appraisalA comparison between selected countries
Nils O.E. Olsson a,n, Andreas kland b, Siri B. Halvorsen b
a
b

Department of Architectural Design and Management, Norwegian University of Science and Technology, N-7491 Trondheim, Norway
SINTEF Technology and Society, Postboks 4760 Sluppen, N-7465 Trondheim, Norway

a r t i c l e i n f o

abstract

Available online 15 June 2012

This paper presents the cost-benet methodology used in the appraisal of railway infrastructure in
Norway, Sweden, Denmark, the UK, France, Germany and Switzerland. The consequences of differences
in methodology are illustrated by a case-study undertaken with the methodology from each of the
seven countries. Differences in methodology means that results from the analyses are far from similar.
The case project has a positive net present value based on Swiss and British methodology, but negative
net present value using methodology from any of the other ve countries.
& 2012 Published by Elsevier Ltd.

Keywords:
Cost-benet analysis
Railway infrastructure
Project appraisal
National methodology

1. Introduction
The purpose of this paper is to compare cost-benet methodology used in Norway, Sweden, Denmark, Britain, France,
Germany and Switzerland for railway infrastructure projects
and to illustrate the consequences these differences produce on
a practical example. This type of comparison has been asked for
by for example Nakamura (2000). Even though the fundamentals
and the aims are the same in the cost-benet analyses in the
seven countries, the methodological differences may result in
different conclusions analysing the same project, depending on
the applied methodology. One rationale for such a study is that
recent analyses of high speed trains in Norway triggered a public
debate about the methodology used to study infrastructure
investments. We therefore wanted to investigate to what extent
the results of a cost-benet analysis of a railway project are
depending on the applied national methodology.
The cost-benet methodology is used to assess public investments in infrastructure in a large number of countries (Kelly and
Laird, 2005; COWI, 2002). Still, as the analyses often come to be
regarded as a black box by the people not directly involved in
them, the results of the analyses are questioned (Kaufman et al.,
2008). The existence of non-valued costs and benets also
contribute to question the abilities of the cost-benet analysis
(Tavasszy, et al., 2005).
There have been a number of studies over the last decade
comparing appraisal methodologies for infrastructure in Europe,

Corresponding author. Tel.: 47 977 13 628; fax: 47 45025251.

E-mail addresses: nils.olsson@ntnu.no (N.O.E. Olsson),
andreas.okland@sintef.no (A. kland), siri.halvorsen@sintef.no (S.B. Halvorsen).
0967-070X/$ - see front matter & 2012 Published by Elsevier Ltd.
http://dx.doi.org/10.1016/j.tranpol.2012.03.005

including Odegaard, Kelly and Laird (2005), Bristow and Nellthorp

(1998), COWI (2002); Lyk-Jensen (2007). Transport Policy Volume
7, Issue 1, January 2000 was devoted to the topic, including a
comparison by Hayashia and Morisugib (2000). The work of
comparing and synchronising cost-benet methodologies in Europe
has been promoted by the EU through the EUNET and HEATCO
research projects.

2. A theoretical comparison of the cost-benet methodology

in railway infrastructure
This paper is based on an existing proposal to connect Norways capital Oslo and the suburban town of Ski with new 22 km
double track-railroad, called the Follo line. The project is
described by Jernbaneverket (2011) and Musus et al. (2009).
The railway line goes through some of the most densely
populated areas in Norway. New tracks will be an addition to
the existing double-tracks connecting the two. The existing
double tracks are operating at above 100% capacity, and delays
are commonplace. Improving capacity south of Oslo is vital in
order to increase rail transport locally and nationally for both
freight and passenger trafc. By separating express trains from
the slower, local trains, the Follo line will shorten travel time,
improve punctuality and more than double capacity. Rock tunnels
will constitute as much as 19 km of the projected line. Apart from
the rock tunnel twin tubes, the project includes a new station at
and a cut-off at Bryn, which eventually
Ski, a passing loop at As
will connect the Follo line with the Alnabru freight terminal. Both
TBM and drill and blast are considered as excavation methods for
the tunnel. Feeding two new tracks into Oslo Central Station, the
busiest part of the Norwegian rail system, will be a complex

30

N.O.E. Olsson et al. / Transport Policy 22 (2012) 2935

challenge. With maximum effort and resources, Jernbaneverket

(2011) claims that construction is likely to take 56 years. Journey
time for the fastest SkiOslo service is expected to be cut from 22
to 11 min. The new line will enable 11,000 more public transport
journeys daily. The investment cost used in this study is 1352
million Euros.

2.1. Appraisal period, discount rates and other issues

The authorities in each country provide the underlying parameters that strongly inuence the outcome of all cost benet
analyses. One of these is the length of the period of appraisal. In
railway projects, the life span of equipment and infrastructure
vary to a large extent. Actual rails have, depending on type of
trafc and frequencies, a life span of about 30 years, whilst a
tunnel may have close to indenite life. Many modern railwaylines in Scandinavia still follow the same trajectories as when
they were rst constructed over hundred years ago. Some
European countries assume the same for new railway-lines and
use an indenite period of appraisal (COWI, 2002). The period of
appraisal is set to 40 years in the HEATCO recommendations.
Sweden have adapted to the period recommended by HEATCO,
whilst Denmark and Britain still keep their old values of 50 and 60
years. France has 50 years, Germany 36 and an indenite period is
used in Switzerland. In Norway the period is 25 years. For
equipment with a shorter life span, the period of appraisal is set
equal to the life expectancy. To reect cases where there is
residual value in the infrastructure after the end of the period
of appraisal, this value is included in the analysis in all the
countries except for Germany and Switzerland. The size of the
residual value is strongly dependent on the number of years from
the base year of the analysis and the discount rate. All countries
assume linear reduction in the value of the equipment through its
life-span. The life-span of equipment varies little between the
countries, with the exception of tunnels, bridges, and land which
varies from 60 to 100 years. As the appraisal period of Norway is
by far the shortest of the countries in the study, the residual value
plays the most signicant role in Norwegian analyses.
As costs and benets spread out over the period of appraisal,
discounting is used to relate values to present value. The discount
rate varies between 2% (Norway) and 8% (France) and can be
regarded as the opportunity cost of capital. Investment in infrastructure should give a rate of return on investment. An addition
to the discounting rate can also be justied to include a general
risk premium. The risk premium is an addition to the demanded
rate of return on investment to compensate for the risky nature of

infrastructure projects. In Norway, a risk premium is added to the

demanded rate of return, with an average addition of 2.5%. In
Switzerland, a risk premium of 5% is added on construction cost.
Most large-scale infrastructure projects are nanced by government funds. Government is nanced by the collection of taxes
directly and indirectly. The re-prioritisation of the allocation of
resources induced by taxes leads to an efciency loss in the
economy. The size of the efciency loss is hard to measure exactly.
Denmark and Norway add a cost to offset the efciency loss,
through the marginal cost of public funds. In Swedish methodology, the efciency loss is now abandoned altogether (from
30% before 2008). The extra cost of nancing investment over
the public budget consists only of the tax distortion factor. The
value is set to 21% to represent the average value added tax on
private goods. Denmark operates with a tax correction factor of
17% in addition to the efciency loss set equal to 20% of the cost.
Table 1 shows a summary of the discussed economic parameters
in the studied countries, along with HEATCO recommendations.

2.2. Main CBA posts methodology comparison

Normally, the most important benet is travel-time savings. In
UK road schemes as much as 80% of the benets come in the form
of travel-time savings (Mackie et al., 2001). The appraisal of travel
time is traditionally founded on the belief that travel time is wasted
time in between real activities like work or leisure (Lyons and
Urry, 2004; Lyons et al., 2007; Hjorthol and Gripsrud, 2008).
Another view has surfaced more recently that claims attention
must be given also to the activity of travelling in itself (Mokhtarian
and Salomon, 2001), in addition to the well-established work of
Hensher (including Hensher, 1977) that focuses on the valuation
of productive time while travelling. The countries in the study
calculate travel-time savings by grouping the travellers into
distinguished groups; business travellers, commuters, and other
travellers. Each group is assigned a travel-time value based on
parameters such as travel mode, travel distance, average pay,
willingness to pay and for the case of business travel, the
Hensher-equation of relative productivity. As shown by Table 2,
values of travel-time savings differ between the countries in the
study. To some extent the large differences illustrates the inuence of culture when dealing with willingness to pay.
Reductions in the number of accidents and their induced cost
are one benet from many investments in infrastructure on both
road and rail. Railways are however in a special position when it
comes to low accident probability rate (Assum, 1998). Accident
costs are calculated from the Value of Statistical Life (VSL) and

Table 1
General economic parameters.

Appraisal period
Discount rate
Risk premium
Cost of public funds/tax correction factor

Norway

Sweden

Denmark

UK

France

Germany

Switzerland

HEATCO

25 years
2%
2.5%
20%/ 

40 years
4%

 /21%

50 years
6%

20%/17%

60 years
3.53%

 /21%

50 years
8%

36 years
3%

(uses factor prices)

Innite
22.5%

(uses factor prices)

40 years
3%

(uses factor prices)

Table 2
Values for travel time savings (h/hour).
VTTS rail travel (shortlong)

Norway

Sweden

Denmark

UK

France

Germany

Switzerland

Business travel
Commuting
Other

19.521.8
7.116.8
4.411.4

30.2
5.611.2
5.611.2

39.3
10.0
10.0

44.5
8.9
7.9

12.8 (15.8)
11.5 (14.1)
6.3 (7.7)

29.7
8.1
8.1

21.1
12.34
6.7

N.O.E. Olsson et al. / Transport Policy 22 (2012) 2935

31

Table 3
Values for a statistical life (VSL) in Euros.

Fatal
Serious
Slight

Norway

Sweden

Denmark

UK

France

Germany

Switzerland

3 296 715
2 251 703 (very serious)
746 472 (serious)
99 513

2 453 882
455 904

1 633 182
170 287

2 110 424
238 583

1 730 246
259 537

1 752 043
121 868

21 898

46 299

18 394

38 065

5 332

2 603 137
895 197 (very serious)
232 592 (serious)

Table 4
Overview over results from the cost benet analyses (mill. h).
Mill 2008-Euro adjusted to 2018

Norway

Sweden

Denmark

UK

France

Switzerland

Germany

Benet from travel-time reduction

Operators public transport (excluding altered subsidies)
Operators goods
Cost to public sector (excluding altered subsidies)
Third party members
Net present cost/benet society

532.1
245.0
13.1
 1633.1
17.5
 580.3

467.9
333.5
17.9
 1508.8
23.1
 666.5

828.8
255.1
24.2
 2032.8
35.5
 889.3

2976.6
265.5
23.5
 1701.8
62.2
1626.0

647.2
159.9
8.5
 1518.4
11.3
 691.4

1231.7
540.8
29.0
 1329.5
62.2
534.1

750.7
268.8
14.4
 1319.7
17.3
 271.6

historic records of number of accidents. The VSL-value incorporates both direct and indirect costs of accidents, such as material
damage, welfare loss (for the injured and family/relatives), administrative costs, and the loss of future tax proceedings for society.
Both VSL-values and the historic records of accidents differ
between the countries in the study. Table 3 shows values for a
statistical life (VLS) in Euros in the studied countries.
The transfer of trafc from road to rail is a declared goal of
both the EU and many individual governments (Comission of the
European Communities, 2001). Especially the focus on global
consequences of climate gasses like CO2 has motivated the
transfer from road to rail. The countries in the study value these
global effects in various manners. The British methodology uses
the shadow price of carbon as presented by DEFRA (DEFRA, 2007).
The valuation is given by HMT (Stern, 2007) and gives the current
shadow price of carbon. The cost of future emission is assumed to
rise with 2% per year as the concentration of carbon in the
atmosphere rises. HEATCO recommends a methodology similar
to the British methodology (Bickel, et al., 2006). Their valuation
use the social price of carbon, based on the work of Watkiss, et al.
(2005a), Watkiss, et al. (2005b).

3. Cost-benet methodologies for railway infrastructure A

case study
As shown in Table 4, the study illustrates major consequences
of methodological differences between the seven countries. The
analysis with the original Norwegian methodology shows a
project with negative net present value. Using Swedish, Danish,
French and German methodology also gives a negative value.
Applying British and Swiss methodology on the other hand, the
project turns out with a positive net present value.
3.1. Method
The original cost-benet analysis of the new railroad has been
executed by an independent Norwegian consultancy rm. The
reports are available to the public (Musus et al., 2009; Andersen
and Rusten, 2008; Rekdal et al., 2008; Andersen and Andersen, 2008).
The calculations are based on the number of travellers and the travel
time reductions calculated by a set of Norwegian transport models
(RTM23 and NTM5). Based on the data made available, we have
made new cost-benet analyses using Swedish, Danish, British,

French, German and Swiss methodology and parameters. Where

available, parameters from the other countries are then used to
calculate costs and benets for society. These include values for
travel-time savings, waiting time and changes for mode of transport,
value for reductions of accidents and emissions for all the countries
in the study. Where valuations have only been available for some of
the countries, the Norwegian values have been adjusted with PPP
values from OECD (March 2009) and used. This is the case with
operating costs and maintenance cost for Denmark and the UK. This
is also the case for the benet for goods transporters for all countries,
as precise data even for the Norwegian case were not available. The
methodologies are continually evolving. This paper is based on
available information as per 2009 (including Finansdepartementet,
2005; DfT, 2006; Jernbaneverket, 2006; SIKA, 2008; Banverket, 2005;
Trakministeriet, 2003; HEATCO, 2006).
The base investment cost is the same in all calculations. The
methodology and the results of the analysis have been subject to
quality assurance with representatives from the specialist community in the involved countries to guarantee the trustworthiness of the results. All monetary values are given in 2008-h.
The countries in the study all have comparable living standards and wage rates, and they are close to one another culturally,
and were therefore suitable for comparison. The Scandinavian
countries were natural to include because it is a widespread
collaboration between the countries on many areas, including
transport research. Consultants and experiences from Great
Britain, France and Germany have been used in the Norwegian
high speed evaluation carried out from 2010 to 2012. The
research presented in this paper was initiated based on discussion
on which method to apply in the Norwegian high speed evaluation. It was therefore interesting to include these countries.
Switzerland was chosen because the Norwegian railway community have used this country as an example of a successful strategic
approach to infrastructure investments.
Even though these countries have several similarities, there is
a concern about how valid it is to transfer parameters between
countries. We have approached this issue in two ways. The rst
approach was to compare parameters and methodologies used in
the studied countries, and then apply them on the Norwegian
case. The second approach was to apply discount rates, appraisal
period and GDP growth adjustments from the different countries
on the Norwegian case with Norwegian parameters. The second
approach highlights the effects of parts of the methodology, and
avoids the concern related to parameter transfer.

32

N.O.E. Olsson et al. / Transport Policy 22 (2012) 2935

equally large benet for government for every year of the

appraisal period.

3.2. Benet from travel-time reductions

Travellers benet is calculated from simulations of the trafc
in 2025, performed by a Norwegian regional transport model and
a national model. The general trafc growth is adapted to the
growth in the respective countries. The values of travel time are
those used in cost-benet analysis used in the respective countries, and the resulting travellers benet is shown in Table 5.
The Swedish travel-time values are 65% of the ones used in
Norway, the Danish almost identical, and the British values are
210% of the Norwegian values. German values are similar as for
Norway, Swiss somewhat higher and the French values are the
second highest. The total benet is calculated for a period of 25
years for Norway, 36 years in Germany, 40 years for Sweden, 50
for Denmark and France, 60 for Britain, and innite number of
years in Switzerland but in practice 50 years. The Danish and
British methodologies continually adjust the travel-time value
throughout the period. The forecasted growth in GDP per capita is
however higher in the UK than in Denmark.

3.4. Financial consequences operators in transport of goods

The proposed project has only minor consequences for the
transportation of goods (benet goodso2% of investment cost),
as there are bottlenecks located elsewhere in the rail system. If
these bottlenecks where removed in the future, the benets of the
new double tracks would signicantly increase. None of the
countries in the study assign any value to a possible future
change of this kind as long as no plans exist. The benet would
rather be included in a cost-benet analysis of removing the
bottlenecks at some later stage.
3.5. Financial consequences for government
The biggest post among the nancial consequences for government (as shown in Table 7) is clearly the investment cost. The
differences in investment cost reect only differences in discount
rate, and in the British case, optimism bias (Flyvbjerg and COWI,
2004; Mott MacDonald, 2002). British methodology also includes
increasing wage cost as a result of GDP-growth. The cost of using
public funds enters the analysis as a nancial consequence for
government even though it is a cost borne by society as a whole,
not just by government.

3.3. Financial consequences for operators of public transport

The proposed new railway line will result in a large number of
busses being replaced by a train-service. This results in a transfer
from the bus companies to the train-operators. Table 6 shows the
nancial consequences for the operators.
Subsidies are still necessary for the operators to be in balance,
as is common in many countries (Rothengatter, 2001). The
amount of subsidies are however lower than in the do-nothing
scenario. The reduction in subsidies is represented by a cost for
the operators in this post in the analysis. It is however offset by an

3.6. Costs and benets for society in general

As shown in Table 8, costs and benets for third-party are
small compared with other big posts in the analysis. Accidents, of

Table 5
Travellers benet (mill. h).

Travellers benet 2025

Total travellers benet

Norway

Sweden

Denmark

UK

France

Switzerland

Germany

33.3
532.0

21.5
467.9

33.5
828.8

69.8
2976.6

46.6
647.2

37.0
1231.7

30.9
750.7

Table 6
Financial consequences operators of public transport (mill. h).

Increased revenue
Operating cost busses
Personnel expenses
Maintenance and energy
Administrative cost
Capital cost
Other cost
Altered subsidies

Norway

Sweden

Denmark

UK

France

Switzerland

Germany

364.4
17.9
 17.4
 33.3
 41.5
 38.1
 7.1
 245.0

496.0
24.5
 23.6
 45.4
 56.4
 51.8
 9.6
 333.5

452.2
33.1
 32.0
 61.4
 76.5
 47.2
 13.0
 255.1

459.1
32.1
 31.0
 59.9
 74.3
 47.9
 12.7
 265.5

237.7
11.7
 11.3
 21.8
 4.6
 27.1
 24.8
 159.9

804.2
39.5
 38.3
 73.6
 15.6
 91.6
 83.9
 540.8

399.8
19.7
 19.0
 36.6
 7.8
 45.5
 41.7
 268.8

Table 7
Financial consequences for Government (mill. h).

Investment cost (construction)

Residual value
Operation cost rail infrastructure
Operation cost road infrastructure
Altered road toll
Altered charges/taxes
Cost of public funds/efciency loss
Altered subsidies
Sum n.cons. Government

Norway

Sweden

Denmark

UK

France

Switzerland

Germany

 1351.7
229.2
 11.9
22.4
 9.1
 28.0
 238.8
245.0
 1388.1

 1330.9
88.2
 1.6
1.2
 12.5
 23.4
 229.8
333.5
 1175.3

 1605.2
118.0
 17.3
1.9
 23.8
 66.2
 440.3
255.1
 1777.7

 1515.8
0.0
 21.4
40.1
 23.5
 92.9
 88.1
265.5
 1436.4

 1504.7
3.8
 7.8
14.6
 6.0
 18.2
0.0
159.9
 1518.4

 1270.4
0.0
 26.4
49.4
 20.2
 61.8
0.0
540.8
 1329.5

 1290.4
0.0
 13.1
24.6
 10.1
 30.7
0.0
268.8
 1319.7

N.O.E. Olsson et al. / Transport Policy 22 (2012) 2935

33

Table 8
External costs (mill. h).

Accident reductions
Global emissions
Local emissions
Noise
Emissions construction
Crowding
Total

Norway

Sweden

Denmark

UK

France

Switzerland

Germany

11.1
3.0
1.7
3.5
 1.9

17.5

12.5
16.4
1.6
4.7
 12.3

23.1

7.3
0.7
2.2
8.9
 1.3
17.9
35.5

44.0
10.0
3.3
6.8
 1.9

62.2

5.0
2.7
3.3
2.3
 1.9

11.3

20.7
20.7
27.1
 4.4
 1.9

62.2

8.8
3.4
0.4
3.9
 1.9

17.3

Table 9
Applying different methodologies, dened as appraisal period, discount rate, and adjustments for GDP growth, tax cost, residual value, congestion cost and adjustments for
global pollution from different countries on the Norwegian baseline study (mill. h).

Appraisal period
Discount rate
Mill 2008 NOK adjusted to 2018
Benet from travel-time reduction
Operators public transport (exluding altered
subsidies)
Operators freight
Cost to public sector (excluding altered subsidies)
Residual value
Cost of public funds/efciency loss
Third party members
Net present cost/benet society
NPV/I

Norway

Sweden

25 years
4.5%

40 years
4%

532.2
245.0

743.6
342.4

Denmark Denmark
(GDP)
50 years
50 years
6%
6%

600.3
276.4

891.2
231.9

13.1
18.4
14.8
21.9
 1388.2
 1528.8
 1896.2
 2042.5
229.2
67.9
4.7
11.6
 238.8
 228.2
 455.1
 399.9
17.5
25.2
37.8
37.8
 580.3
 399.3
 967.0
 859.6
 0.43
 0.30
 0.70
 0.56

which reductions have got important political impact, do not

affect the analysis to any important degree. The increasing values
of accident cost using British methodology that includes GDP
growth, also contributes to make the reductions in accident cost
so much larger for Britain.
The proposed project will lead to reductions in global emissions, local pollution and noise emissions. These are, however, not
signicant compared to the huge investment. Much of the reduction in global emissions is offset by large emissions during the
construction period, because emissions during the lifespan of the
project are discounted to give their present values.
The British methodology uses increasing values for carbon
emissions throughout the calculation period, as the accumulation
of carbon in the atmosphere grows. The environmental effects are
valued highest using Swiss and British methodology.

4. Applying different countries methodology on different sets

of parameter values
To further visualise the effect of different methodologies, we
have varied the methodology and values from different countries.
What can be dened as methodology and values can be challenged. We have used the following distinction. In methodology, we applied the different countries use of appraisal period,
discount rate, GDP based adjustments (used in the UK and
Denmark), tax cost, residual value, congestion cost (only used in
Denmark) and adjustments for global pollution (used in the UK).
Other parameters, such as value of time, cost of accidents etc. are
treated as values.
First, different countries methodology is applied on the
Norwegian base case study. The results are shown in Table 9.
Table 9 is based on Norwegian monetary values, such as value of
time, accidents, pollution etc., but applying the factors dened as
methodology above from the different countries included in the
study. The British and Danish approach is illustrated with, and

UK
60 years
3%

1107.5
509.9

UK
( GDP)
60 yrs
3%

1483.4
424.5

27.3
41.0
 1534.4
 1643.3
0.0
0.0
 188.3
 112.3
38.4
38.4
149.8
345.4
0.12
0.25

France

Switzerland Germany

50 years
8%

50 years
2.5%

462.9
213.2

1108.0
510.1

11.4
27.3
 1526.1
 1326.1
1.9
0.0
0.0
0.0
15.0
38.4
 823.8
357.7
 0.55
0.28

36 years
3%

810.8
373.3
20.0
 1331.1
0.0
0.0
27.6
 99.5
 0.08

without GDP adjustments, in order to show the effect of such

adjustments. Adjustments for GDP, discount rate and calculation
period do inuence the results.
Table 10 is another approach. We vary both methodology and
values (based on the distinction between the two mentioned
above) from all countries in the study. The line for Norway is the
same calculation as shown in more detail in Table 9.

5. Discussion
The results of the study illustrate how large consequences
differences in methodology create. Some of the differences in
methodology are within the scope of regular sensitivity analyses
that accompany larger infrastructure projects. Still, sensitivity
analyses rarely combine alterations of all key parameters.
As the cost-benet analysis includes the welfare of the
members of society, cultural differences play a large part in how
highly transport is valued. Benet from travel-time savings
accounts for between 56% (Sweden) via 66% (Switzerland and
Norway), 72 (Denmark and Germany), 78 (France) to 89% (Britain)
of total benet. This is consistent with previous studies. According
to IER (2006), British methodology gives rise to almost double
NPV in comparison with HEATCO guidelines. In our study, the
higher valuation of travel-time in Britain by itself accounts only
for about a doubling of benets for travellers compared to the
original Norwegian analysis. Combined with the other parameters, the total benet from reduced travel-time grows to more
than ve times the valuation in the original Norwegian analysis.
The British methodology gives higher results than the other
countries. One explanation is the inclusion of GDP per capita
growth. Although Denmark includes GDP growth, their predictions are more conservative. By the introduction of appraisal
based on GDP per capita growth and lower discount rate, cost and
benets later in the appraisal period gain more weight in the
overall analysis. Combined with GDP growth and diminishing

34

N.O.E. Olsson et al. / Transport Policy 22 (2012) 2935

Table 10
Applying different methodologies, dened as appraisal period, discount rate, adjustments for GDP growth, tax cost, residual value, congestion cost and adjustments for
global pollution from different countries on value used in the different countries (mill. h).
Methodology

Values (other parameters than in those

varied in the methodology section)

Norway
Sweden
Denmark
UK
France
Switzerland
Germany

Norway

Sweden

Denmark

UK

France

Switzerland

Germany

 543.8
 758.6
 551.0
 674.3
 141.5
 256.8
 693.2

 399.1
 668.2
 370.6
229.8
 64.8
 52.8
 558.8

 859.6
 1204.2
 949.6
 685.8
 441.5
 759.7
 602.5

345.4
196.4
635.8
1601.3
871.2
865.0
 261.3

 823.8
 1001.2
 838.7
 1033.5
 691.4
 750.1
 922.0

357.7
 91.0
340.5
789.0
669.9
534.0
122.6

 99.5
 409.9
 118.9
71.2
129.0
173.7
 271.6

discount rate, an appraisal period of 60 years (as in Britain) have

large inuence over the results of the CBA. There are however
increasing uncertainty in the forecasts from the transport models
further away from the base year.
The green benets of the project are small compared to the
costs, no matter which countrys methodology is being used.
Accident costs also only play a marginal role in the overall
analyses. British statistics lead to more than triple benet compared with the others, still the total benet is only equal to
between 0.8% (Denmark) and 2.5% (Sweden) of the benet from
travel-time savings.
In Table 9 we apply different countries methodology on the
Norwegian base case. As expected, long calculation period, low
discount rate and upgraded values based on GDP growth
increases the benet/cost ratio. Britain and Switzerland both have
long calculation periods and low discount rates. This contributes
to making the project protable in socio-economic terms based
on these two countries methodology.
Transferring parameter values such as value of time between
countries can be questioned. In Table 10, we varied both parameter values and methodology. This gives an opportunity to
study the effect of different sets of parameter values, as well as
different methodologies. Using Norwegian and Danish methodology, the project has negative value based on all sets of value
parameters. Norway and Denmark are the only countries with
such consistent results in Table 10 (the Swedish methodology
gives negative value using all but British value sets). In the other
end, Table 10 shows that the British and Swiss methodologies
give positive values of the project using all but one value set
(German and Swedish, respectively). The pattern in Table 10
appears more consistent in the columns (for methodology) than
in the lines (for values).

million Euros. The project also has positive net present value
using Swiss methodology.
British and Swiss approaches generate positive net present
value for the case studied. Switzerland applies a low discount rate
and long appraisal period. The UK has relatively high values for
travel-time savings, long appraisal period and adjustments for
GDP growth. Table 10 shows that the project would had positive
net present value using British and Swiss methodologies and most
of the included countries parameter values. Table 10 also indicates that methodology, as the concept is used in this paper, has a
more consistent impact on the net present value than the
different countries parameter values. Costs and benets later in
the appraisal period play a larger role in British methodology. The
Scandinavian countries have closely related values even though
methodological differences on single posts still are substantial.
Further studies could investigate a number of issues. The effect
of different approaches to cost estimates is an interesting topic for
further study. In the UK, costs are added to compensate for
optimism bias. In Norway, contingencies are evaluated in external
project evaluations. Other countries may have other practices.
Related to appraisal period, a topic for further study is to what
extent renewals should be included. Especially for the longest
appraisal periods, it is not obvious that a long period always will
result in a higher benet/cost ratio. Combined with adjustments
for GDP growth, future cost for renewal can inuence the total
cost. We have only studied a single project, and not the relative
difference between several projects, which is a common situation
in decision making. Studies are recommended to analyse to what
extent the relative ranking of a set of projects varies depending on
the applied methodology. Finally, it would be interesting to do
similar studies on other projects, to investigate to what extent our
results are inuenced by the particular project, or specic features
of the Norwegian methods.

6. Conclusion
References
This paper has presented and compared the cost-benet
methodology used in appraisal of railway infrastructure in
Norway, Sweden, Denmark, the UK, France, Germany and
Switzerland. The consequences of differences in methodology
have been illustrated by a case study. Even though the basic
principles are the same in the cost-benet analyses in the seven
countries, the methodological differences means that the result
for the cost-benet analysis is depending on which countrys
methodology that is used. The studied project has a benet cost
ratio between 0.45 and 0.57 using Scandinavian approach and a
benet cost ratio of 2 using British approach. Similarly, the project
has a net present value between  580 and  889 million Euro
using Scandinavian and French approaches. The net present value
using German approach is  272 million Euros. Using British
approach, the project gets a positive net present value of 1626

Andersen, C., Rusten, G., 2008. Nytt dobbelspor Oslo SSki Samfunnsgeograske
perspektiver. Samfunn og Nringslivsforskning AS, Bergen.
Andersen, C., Andersen, T., 2008. Nyttekostnadsanalyse av nytt dobbeltspor Oslo
SSki. Working paper, Samfunns og Nringslivsforskning AS, Bergen.
Assum, T., 1998. Dds- og Personskaderisiko i Persontransport. Transportkonomisk institutt, Oslo.

samhallsekonomiska

Banverket, 2005. Berakningshandledning

Hjalpmedel
for

bedomningar
inom jarnv
agsektorn.
Banverket.
Bickel, P., Friedrich, R., Burgess, A., Fagiani, P., Hunt, A., De Jong, G., Laird, J., Lieb, C.,
Lindberg, G., Mackie, P., Navrud, S., Odgaard, T., Ricci, A., Shires, J., Tavasszy, L.,
2006. Proposal for harmonised guidelines. HEATCO Deliverable 5, IER,
Germany.
Bristow, A.L., Nellthorp, J., 1998. Transport project appraisal in the European
Union. Transport Policy 7 (1), 5160.
Comission of the European Communities, 2001. White paper European transport policy for 2010: time to decide. Comission of the European Communities,
Brussels.
COWI, 2002. Brug afsamfundskonomiske i udvalgte lande. Trakministeriet.

N.O.E. Olsson et al. / Transport Policy 22 (2012) 2935

DEFRA, 2007. The social cost of carbon and the shadow price of carbon: What They
Are And How To Use Them In Economic Appraisal In The UK. Economics
Group, Defra.
Department for Transport, 2006. Transport, wider economic benets, and impacts
on GDP. Technical Paper, Department for Transport.
Finansdepartementet, 2005. Veileder i samfunnskonomiske analyser.
Finansdepartementet.
Flyvbjerg, B., COWI, 2004. Procedures for dealing with optimism bias in transport
planning. Guidance Document, Department for Transport.
Hayashia, Y., Morisugib, H., 2000. International comparison of background concept
and methodology of transportation project appraisal. Transport Policy 7 (1),
7388.
HEATCO, 2006. Developing Harmonised European Approaches for Transport
Costing and Project Assessment Last accessed 15.11.2011 url: /http://
heatco.ier.uni-stuttgart.de/S.
Hensher, D., 1977. Value of BusinessTravel Time. Pergamon Press, Oxford.
Hjorthol, R., Gripsrud, M., 2008. Bruk av reisetid ombord pa toget. TI.
IER, 2006. Case Study Results. HEATCO Deliverable 6, IER, Germany.
Jernbaneverket, 2006. Samfunnskonomiske analyser for jernbanen. Jernbaneverket.
Jernbaneverket, 2011. New Double Track OsloSki Last accessed 15.11.2011 url:
/http://jernbaneverket2.no-ip.org/en/Startpage/Projects/New-double-track-O
slo-Ski/S.
Kaufmann, V., Jemelin, C., Pieger, G., Pattaroni, L., 2008. Soci-political analysis of
French transport policies: The state of the practices. Transport Policy 15 (1),
1222.
Lyk-Jensen, S.V., 2007. Appraisal Methods in the Nordic Countries. Danish Transport Research Institute, Lyngby.
Lyons, G., Urry, J., 2004. The Use and Value of Travel Time. Unpublished Paper.
Lyons, G., Jain, J., Holley, D., 2007. The use of travel time by rail passengers in Great
Britain. Transportation Research Part A 41 (1), 107120.
Mackie, P.J., Jara-Daz, S., Fowkes, A.S., 2001. The value of travel time savings in
evaluation. Transport Research Part E 37 (23), 91106.

35

Mokhtarian, P.L., Salomon, I., 2001. How derived is the demand for travel? Some
conceptual and measurement considerations. Transportation Research Part A
35 (8), 695719.
Mott MacDonald, 2002. Review of large public procurement in the UK. HM
Treasury.
Musus, S.U., Andersen, C., Rekdal, J., Larsen, O., Andersen, T., 2009. Utredning nytt
dobbelspo Oslo SSki Hovedrapport. Det Norske Veritas.
Nakamura, H., 2000. The economic evaluation of transport infrastructure: needs
for international comparisons. Transport Policy 7 (1), 36.
OECD Comparative Price Levels March 2009. Last accessed 08.06.2009 url:
/http://www.oecd.org/dataoecd/48/18/18598721.pdfS.
Odgaard, T., Kelly, C., Laird, J., 2005. HEATCO deliverable 1 current practise in
project appraisal in Europe. European Commission.
Rekdal, J., Larsen, O.I., Jansson, B.K., 2008. Trakkanalyse for nytt dobbelspor Oslo
SSki. Mreforskning Molde AS.
Rothengatter, W., 2001. Transport subsidies. In: Hensher, K.J. (Ed.), D.A. Handbook
of Transport Systems and Trafc Control. Emerald Group Publishing. Button.

transportsektorn:
SIKA, 2008. Samhallsekonomiska
principer och kalkylvarden
for
ASEK 4. SIKA.
Stern, N., 2007. The Economics of Climate Change. Cambridge University Press,
Cambridge.
Tavasszy, L., Jonkhoff, W., Burgess, A., Rustenburg, M., Hunt, A., 2005. Key issues in
the development of harmonised guidelines for project assessment and transport costing. HEATCO Deliverable D3.
Trakministeriet, 2003. Samfundskonomisk manual anvendt metode og praksis

pa transportomradet.
Trakministeriet.
Watkiss, P., Anthoff, D., Downing, T., Hepburn, C., Hope, C., Hunt, A., Tol, R., 2005a.
The Social Cost of Carbon (SCC) Review Methodological Approaches for
Using SCC Estimates in Policy Assessment. AEA Technology for DEFRA, London.
Watkiss, P., Downing, T., Handley, C., og Buttereld, R., 2005b. The Impacts and
Costs of Climate Change. AEA Technology Environment, Stockholm Environment Institute, Oxford.