CP

PB
R E P O R T

00/2

Cost-benefit analysis of of only a minor part of the total section: only about 75 kilo-
metres of the total link of 480 kilometres between
high speed rail Amsterdam and Frankfurt. The Dutch government has
already decided to invest some E 0.75 billion in upgrad-
Henri Dijkman, Carl Koopmans and ing the section Amsterdam-Utrecht, whereas in Germany
Martin Vromans* a 300 km/h link between Cologne and Frankfurt will
be constructed, an investment amounting to some
E 4 billion.
Abstract On both sides of the German border – from Utrecht to
The Dutch government is considering the construc- the border, and from the border to Cologne – a final deci- 39
tion of a High-Speed Railway (HSR) between Schiphol sion on investing in infrastructure still has to be made.
Amsterdam Airport and the German Ruhrgebiet (HSL- However, the Dutch and German authorities have set-
East). This note presents a cost-benefit analysis of tled on gearing their decisions to one another in order to
the construction of the railway section. achieve a comparable level of service on both sides of the
The economic effects of this project appear to be border (Warnemünde Agreement).
rather small. This is due mainly to the limited savings
in travel time of a mere ten to 15 minutes. As a result, Alternatives
a new railway seems to be unprofitable in all scenar- The cost-benefit analysis (CBA) considered two project
ios. alternatives, which were then compared with the base
Since the HSL-East may be regarded as a pilot of case. In the base case, no dedicated infrastructure will be
the project evaluation guideline, a number of method- built; nonetheless, HSR-rolling stock will be in service
ological issues were raised during the analysis. In par- on the existing section.The two project alternatives assume
ticular, the assessment of the macroeconomic impact construction of dedicated HSR infrastructure, suitable for
of large projects as well as the evaluation of project a maximum speed of either 200 km/h or 300 km/h.
risks need a more thorough treatment in future cost- Estimated investment and maintenance costs range from
benefit analyses. E 1.8 to 3.4 billion in present-value terms.
Even though the base case limits the speed to 140 km/h,
a considerable improvement in speed and service level
Introduction over the current situation is achieved, since using high-
The Dutch government is considering the construction of speed rolling stock allows for a direct connection between
a High Speed Rail (HSR) section, the “HSL-East”, between Schiphol Amsterdam Airport and Frankfurt. In this way,
Utrecht and the German border, as a part of the HSR-link the full potential of the investments on both the Amster-
connecting Schiphol Amsterdam Airport with the German dam-Utrecht and, especially, the Cologne-Frankfurt sec-
Ruhrgebiet. In preparing the investment decision, the tions can be used, resulting in the base-case in travel time
Ministry ofTransport asked CPB to assess the desirability savings of 1 hour 45 min (about one-third of the current
of the investment.The cost-benefit analysis (CBA) closely value).The dedicated infrastructure suitable for 200 km/h
followed the guidelines for project evaluation that have results in an additional time saving of 17 minutes.The 300
been worked out by a number of economic research insti- km/h alternative offers a gain of three additional minutes.
tutes within the OEEI project (see Eijgenraam et al., 2000a).
The goals of the HSL-East project are manifold:The pro- Economic scenarios
ject should accommodate sustainable growth of national The profitability of the two alternatives has been analysed
and international mobility, and it should divert traffic from using as reference the three CPB background scenarios
road and air to rail. Furthermore, it should improve acces- Divided Europe, European Coordination and Global
sibility of the economically important western part of Competition (Burk and Suijker, 1996).These scenarios pro-
the Netherlands (the ‘Randstad’). From an economic point vide a sufficient band-width to test for the robustness of
of view, however, these are intermediary goals that should the project.
contribute to Dutch welfare in a broad sense, taking into
account also environmental effects. Aspects of the cost-benefit analysis
Following the project evaluation guideline, the main ele-
The Project ments of the cost-benefit analysis are as follows:
The geographical location of the HSL-East is depicted in
figure 1. The project, indicated by the bold line, consists – market and competition analysis;
– investment and maintenance cost assessment;
* For more information, contact Henri Dijkman (tel: +31-70-3383349; – business analysis;
e-mail: [email protected]) – direct economic effects;
CP
PB
R E P O R T Figure 1 The HSL-East as part of the Amsterdam-Frankfurt HSR
00/2 link (bold line indicates the HSL-East section).

as having become obsolete

and were corrected upward
to a range of E 1.4 to 2.6 bil-
lion. Including yearly main-
tenance costs, the net present
value (NPV) of the investment
plus maintenance costs varies
from E 1.8 to 3.4 billion, using
a 4% discount rate.
40 The business analysis,
which confronts operational
200 Kmh Amsterdam revenues with operational
Schiphol costs, shows that no sufficient
Arnhem positive margin is left to pay
Utrecht for the infrastructure invest-
Duisburg ment costs or even mainte-
200 Kmh nance costs alone. The NPV
Düsseldorf of operational revenues is
only E 100 to 170 million, with
operational costs between E
Köln 70 and 110 million.
300 Kmh Besides the operational
revenues, the direct economic
Frankfurt effects consist mainly of the
benefits that accrue to trav-
ellers. The benefits for trav-
– indirect effects and macroeconomic analysis; ellers consist of travel cost differences and travel time sav-
– external effects, including environmental issues. ings. In many cases, a rise in travel costs is offset by a
reduction of travel time. Besides cost- and time differen-
Each topic will be briefly discussed below. tials, another gain is associated with the increased com-
fort of HSL as compared with conventional trains.The cur-
The market and competition analysis was carried out by rent study assumes that these benefits will not exceed
the German consultancy firm Intraplan.To assess the future those resulting from changes in travel-times and costs.
demand for the HSL-East, Intraplan used a fairly detailed, The travel time savings have been valued with value-
integrated modal-split/distribution model.This model pro- of-time estimates obtained from stated-preference sur-
vided estimates of changes in modal-split and generation veys. The NPV of the sum of travel-time and travel-costs
of new traffic, broken down by different trip purposes. In differences is estimated at 80 to 210 million. Due to ben-
this way, network effects form an integral part of the traf- efits related to comfort, this figure might rise to E 160 to
fic analysis. Different timetables were tested by compar- 300 million.
ing the number of passengers with minimal required load- The indirect economic effects do not accrue to the users
factors for rolling stock, thereby ruling out unfeasible com- or operator of the HSL, but may follow from efficiency
binations (from a business point of view). It turned out gains due to, for instance, a more efficient spatial con-
that only in the Global Competition scenario would demand centration of economic activities. In the case of the HSL-
be high enough for two trains per hour. In both of the other East project, these effects were surveyed by Buck
scenarios, one train per hour seems to be preferable. Consultants International (BCI, 2000), who focussed espe-
The total expected number of international passengers cially on the attractiveness of the HSL-stations for inter-
travelling by HSL ranges from 3.0 to 5.2 million in 2020 in national office relocations. Extrapolation of current trends
both the 200 km/h and 300 km/h alternatives. Only half of lead BCI to conclude that the impact of the project would
this volume represents additional train traffic, either sub- be limited to some 50 to 80 extra jobs in 2010. This rela-
stituted from other modes or newly generated traffic. Most tively low figure is, again, mainly explained by the limited
of the substitution stems from car users; the number of travel time savings that result from carrying out the pro-
passengers switching from air travel is fairly small. ject. Besides business relocations, it is argued that the gen-
Despite its importance for the final outcome of the CBA, eration of additional international business trips is con-
the costs involved are only roughly known. Initial calcu- nected with an increase in trade volumes. A slight increase
lations of investment costs, dating from 1996, were judged of imports and exports will result, of E 7 to 14 million in
 CP
PB
R E P O R T

Summary table of CBA HSL-East 00/2

Amounts in net present value (E)

2010. The total NPV of the indirect

economic effects, as estimated by Benefits Financial Related to
BCI, totals E 30 to 70 million. Direct effects
A part of the environmental – Operating revenues E 0.1 to 0.2 bln
impact of the project can be given
– Benefits to travellers E 0.2 to 0.3 bln changes in travel time and
a monetary value. The HSL-East
leads to a modest reduction of travel costs
emissions of gases like CO2 and Indirect economic effects
NOx.This is due not only to the lim- – (Re)location of international
ited reduction of passenger car-kilo- 41
offices, international trade E 0 to 0.1 bln
metres, but also to the fact that high
speed rail uses far more energy – Macroeconomic impact
than conventional rolling stock. In exploitation E -0 to -0.1 bln
the end, this results in only a small
positive balance that is valued at
Environment: emissions
less than NPV E 10 million.
On the local level, the project prevented E 0 bln CO2 and NOx
requires the demolition of some 30 Total benefits E 0.3 to 0.5 bln
to 60 houses and will affect 200 to
250 ha of scenic and agricultural
Costs
areas. Valuing these issues is con-
sidered to be problematic, because Investment E 1.4 to 2.6 bln
cutting through a scenic area might Maintenance E 0.4 to 0.7 bln
have significantly larger impacts Operation E 0.1 bln
than the affected area itself, for
Total costs E 1.9 to 3.5 bln
instance by damaging the natural
habitat of rare flora or fauna. In the
analysis, these were left as PM items Benefits minus costs (financial) E – 1.5 to – 3.2 bln
without an attempt of valuation. IRR *
Most of the topics discussed until
now serve as input for the macro-
economic analysis, to assess higher PM items
order effects of the direct and indi- Distribution effects
rect economic effects. The macro- (between regions) PM1 Some firms will move to the
economic analysis has been carried
HSL corridor
out with CPB’s multi-sector model,
Athena (Vromans, 1998).Travel cost Landscape and nuisance – PM2 200 to 250 hectares, demoli-
differences are modelled as cost dif- tion of 30 to 60 houses
ferences in the transport sector, * Of no significance here, since costs exceed benefits in all years under consideration.
whereas travel time saving of busi-
ness passengers is considered to
raise productivity. Travel time sav- Conclusion: the project is unprofitable
ings of passengers with other
motives, such as commuting or
leisure, are not included in the macroeconomic analysis: counting to some degree, since a fraction of the indirect
these non-pecuniary direct effects are, of course, welfare effects of the BCI estimation will coincide with effects
effects accruing to these user groups; they are, however, within Athena. However, since both effects seem to be
of no significance in the macroeconomic calculations. very limited, it was expected that a detailed analysis would
The indirect economic effects, as estimated by BCI, are not yield a significantly different picture as far as the pro-
modelled as macroeconomic impulses.The effect of addi- ject evaluation as a whole is concerned. However, in future
tional jobs is translated in an increase of export of ser- CBA’s it might be worthwhile to go into more detail here,
vices, the additional import and export are added to the since the double counting obviously results in some over-
balance of payments. estimation of the benefits of the project.
The way in which the indirect effects are included in Besides analysing the macroeconomic effects of the
the macroeconomic analysis might easily lead to double- project during operation, the analysis also took into account
CP
PB
R E P O R T

00/2

the macroeconomic impact of the construction phase. The accuracy of interna-

Here the results appeared to be very sensitive to the way
the project will be financed. If it is assumed that the pro-
tional CPB forecasts
ject replaces some other public investment or govern-
ment consumption, the macroeconomic effects are almost Gerard van Welzenis*
negligible. If the project would be financed by increased
taxation, positive macro-economic effects would be partly
offset by the excess burden. A sensitivity analysis has Abstract
shown that the net result will be considerably smaller than The accuracy of CPB’s short-term projections for domes-
42 0.9 billion. tic variables was analysed in CPB Report 1999/2. An
inadequate assessment of international economic
Integrating the results in the CBA developments proved to be a major source of forecast
All of the above-mentioned results are brought together errors. This study examines the accuracy of CPB fore-
to assess the profitability of the HSL-East project. As can casts for ten international variables over the period
be seen from the summary table, the project appears to 1984-1998. The outcome is compared to the forecast
be unprofitable in all cases. The financial benefits minus accuracy of the OECD and the European Commission1.
costs amount to E –1.5 to –3.2 billion. Calculation of an The various forecasts appear, on average, to be of com-
internal rate of return seems to be of no practical rele- parable quality, if due account is given to the different
vance, since in all cases yearly costs exceed yearly ben- timing of the projections. However, CPB forecasts of
efits over the period considered. variables that are particularly relevant for the Dutch
economy seem to be more accurate.
The role of risks has been analysed in this project in a
superficial way only. A sensitivity analysis with a discount
rate of 8%, including a 4% risk-premium, only resulted Introduction
in lowering the benefits, and reinforced the result that the Initially, CPB’s assessment of the world economy is based
project is unprofitable. on available forecasts made by international and national
organisations. Recent developments are taken into account,
Conclusions as well as our own judgment on the future development
The HSL-East project turns out to be unprofitable in the of some key variables, such as exchange rates, crude oil
circumstances analysed.The base-case, or no-build option, prices and economic policy measures. Given this proce-
is preferable. In the base-case, many benefits can be reaped dure, it is interesting to compare the quality of CPB fore-
when high-speed rolling stock is used on conventional casts with other forecasts available at the same time, to
track. establish whether the additional CPB effort really improves
Although of minor relevance in the case of the HSL- the existing forecasts of the world economy.
East, a number of issues need a more thorough treatment The comparative analysis in this study is restricted to
in future projects. It seems that narrowing the bandwidth the annual forecasts, published twice a year, for the period
of costs would be desirable. Furthermore, a more thor- 1984-1998. The relevant CPB publications are the Central
ough treatment of the risks (see also van Ewijk and Tang Economic Plan, published in March, and the Macro-
in this issue) and of the macroeconomic impact during economic Outlook, published in September. The OECD
the construction phase can contribute to the robustness and EC forecasts are published in December and June,
of the analysis of future projects. A more unified approach i.e. just in-between the CPB forecasts. We analyse six suc-
is desirable – not only for evaluation on a project-by- cessive bi-annual forecasts for a particular year. The last
project basis, as is the case here, but also to enable a more of these ”forecasts” are published in the second half of
consistent comparison between projects. the year following the projection year; they are consid-
ered realisations.
References
BCI (2000), Indirecte ruimtelijk-economische effecten van de Hoge Snelheids The following international variables have been exam-
Lijn Oost (HSL-Oost) (in Dutch), Buck Consultants International, Nijmegen. ined:
Eijgenraam, C.J.J., C.C. Koopmans, P.J. Tang and A.G.P. Verster (2000a),
– GDP volume of the industrial world (1) and that of the
Evaluation of transport infrastructure, CPB Report 2000/1.
Eijgenraam, C.J.J., C.C. Koopmans, P.J. Tang and A.G.P. Verster (2000b),
European Union (2)
Evaluatie van infrastructuurprojecten - Leidraad voor kosten-batenananalyse – Consumer prices in the industrial world (3) and in the
(in Dutch), CPB, NEI, Den Haag. European Union (4)
Burk, K., and F. Suijker (1996), A long-term scenario study for the Dutch econ-
omy, CPB Report 1996/3.
Vromans, M. (1998), ATHENA, The multi-sector model, CPB Report, * For more information, contact Gerard van Welzenis (tel: +31-70-3383316;
1998/3. e-mail: [email protected])