Journal of Transportation Technologies, 2011, 1, 107-115
doi:10.4236/jtts.2011.14014 Published Online October 2011 (http://www.SciRP.org/journal/jtts)
Social Profitability Threshold of High Speed Railway:
A Benefit-Cost Analysis
Pablo Coto-Millán1, Vicente Inglada2, Pedro Casares1
1Departmen t o f Economics, University of Cantabria, Santander, Spain
2Department of Applied Economics, Complutense University, Madrid, Spain
Received July 6, 2011; revised August 24, 2011; accepted S e pt ember 10 , 20 11
Abstract
In this study, the benefit-cost analysis technique for projects evaluation is applied to determine the minimum
level of demand that makes a high-speed rail project economically viable in Spain. To get this goal, it is nec-
essary to take into account not only the costs and benefits of the high speed railway but also the analysis of
the costs and benefits linked to other alternative transport modes, such as plane, private car or bus. The re-
sults show that the high-speed rail studied is socially profitable from a traffic volume of 6.5 million passen-
gers and for a social discount rate of 4%. Given the benefits associated with network effects, which magni-
tude grows parallel to the extension of the high-speed network, this threshold could be significantly reduced
in subsequent projects.
Keywords: High Speed Railway, Benefit-cost Analysis, Passenger Transport
1. Introduction
In parallel to the trend in modern economies towards
greater liberalization in the product and factor markets,
the role of the state has been consolidated, not only as
regulator of the market but also in the implementation of
certain policies embedded in the welfare society.
Investment in transport infrastructure is one of the
policies for which the public sector maintains a leading
role. Particularly since [1], which defined a new role of
public investment as a driver of productivity and ultima-
tely, of the competitiveness of any economy, complem-
entary to the traditional Keynesian perspective based on
demand.
Within the broad range of investment projects, the in-
terest in high-speed train (AVE) has g rown expon en tially
in Europe during the last decades. In Spain, since its in-
auguration in April 1992, the Madrid-Seville AVE has
had great commercial success, being widely accepted by
travelers. Citizens appreciate its advantages: speed,
comfort, punctuality, etc. All this leads to a high occu-
pancy rate and the attainment of positive results for the
operating company. Nevertheless, it is worth noting that
these results do not co nsider either the maintenance co sts
or the depreciation of infrastructures. The high number
of trips generated by this new product and the degree of
absorption of the demand for other modes of transport,
especially air transport, has contributed to make the AVE
become the predominant transport mode in this course,
which is undoubtedly a milesto ne for rail transport.
Therefore, it is not surprising that the investment in
such mode holds a privileged place in the government
investment plans. Such is the case of the proposals con-
tained in the Strategic Plan for Infrastructure and Trans-
port [2], where the amount of railway investments ex-
ceeds those of other modes of transport, including road
investment, which changes the trend maintained during
the last decades. In the same vein, the European Com-
mission (3) calls for the enhancement of rail transport,
including many high-speed rail projects within the TENs
outline.
But despite such beneficial economic and social ef-
fects, the costs that society has to bear in order to finance
these infrastructure projects should not be forgotten . Par-
ticularly relevant is the amount of resources consumed
by the AVE, associated with the high implementation
costs (infrastructure and rolling stock). These two factors
have to be considered when evaluating, from an eco-
nomic perspective, the change in the net social benefit
linked to the adoption of this railway investment policy.
In this sense, the methodological tools provided by the
welfare economics, such as the benefit-cost analysis, are

P. COTO-MILLÁN ET AL.
108
especially useful.
Among the various factors that modulate the AVE so-
cial return, the volume of demand is shown as the most
important [3,4 ,5], limiting the econo mic viability of each
project. The basic aim of this paper is to determine the
volume of demand that makes an AVE project economi-
cally viable. The delimitation of this social breakeven
can serve as a useful tool in planning and designing the
public policy for infrastructures and transport services.
2. AVE Products
As it can be seen in the Madrid-Seville corridor, the
so-called high-speed rail service includes both the "pure"
AVE, which is called long-distance, and two other seg-
ments: shuttles and variable width, which have specific
characteristics.
Table 1 shows the distinguishing features of each of
these segments for the corridor Madrid-Barcelona, which
line became operational in February 2008. This is the
model chosen for the project type of this research.
The main discriminating factors between Shuttle and
Long Distance are demand met and price, while the
segment of variable width differs from the others be-
cause it has to use different rolling stock, due to the need
of using infrastructure with different track gauges. The
use of the existing infrastructure in part of the journey,
along with the necessary width change operation in the
exchanger, makes the average speed, and therefore the
cost reduction, less widespread than in other segments of
the High Speed supply.
3. Methodology Used
The methodology used in this paper for assessing the
social benefit in the benefit-cost analysis of the AVE p-
roject type is a generalization of that used in [8] and ex-
tensively describe d in [9,10] for the AVE Madrid-Sev illa
and in [7] for the case of the Madrid-Barcelona-French
border AVE.
Next, a summary of the basic points in relation to ti me
savings is shown:
For users of conventional train and bus, as shown in
Figure 1, the generalized cost1 (gt) composed of the rate
(pt) and the total time spent on the trip (gt – pt) decreases
to the value of the generalized cost of travelling on the
AVE (gh). The benefits of this reduction can be ex-
pressed for each mode of transport as:
 
12
C C
thtth ht
hh ttth
ggqgg qq
pq pq
 

This expression is equivalent to the areas of the rec-
tangles t and ht less the net cost needed to
obtain such benefits. Such cost corresponds to that of
setting the AVE less the savings resulting from the
elimination of the conventional train and bus services.
gh bg epfpj
However, as shown in the Figure 1, and without con-
sidering who gets the surplus, the social benefit can be
obtained, for trips diverted from other modes, by the time
savings resulting from the introduction of a faster mode
of transport. Therefore, it is sufficient to calculate the
time reduction in travel and access, an d multiply it b y th e
value of time.
Figure 1. AVE benefits for train and car travelers.
Table 1. Discriminating factors of the submarkets of the
AVE Madrid-Barcelona.
Shuttles Long
distance Variable width
[Barcelona-
Tarragona] [Madrid- Barcelona] [Madrid- Soria]
[Barcelona-
Gerona] [Madrid- Zaragoza] [Madrid- Logroño]
Routes
[Barcelona-
Lérida, etc.] Other Other
Material Alsthom Talgo and
Siemens Brava
Infrastruc-
ture New high speed
lines New high speed lines
New and
conventional
line
Price per
passenger /
km (Euro
cents of
2008)
8.11 10.82 9.32
Occupancy 0.60 0.65 0.60
Type of
demand
Suburban train
with high rate of
commuting trips
(conmuter)
Long dis-
tance Long d i s - tance

(1)
1Generalized cost is defined as a lineal combination of three elements:
monetary component of the trip, value of the total time consumed and
valuation of other travel elements such as comfort and safet
y
.Source: [7] and own elaborati on.

109 P. COTO-MILLÁN ET AL.
The total benefit for the trips generated is represented
in this chart by the area under the demand function (the
trapezius bdqhqt) minus the area of the rectangle (edif),
which represents travel time consumed, and which can
be decomposed in two components. The first one, repre-
sented in the graph by the rectangle (fiqtqh) is given by
the income derived from such trips (qh – qt) × (ph) and the
second component, represented by the triangle (bde), is
half the product of the difference of the generalized cost
for the trips generated by the number of such trips (qh – qt)
× (gt – gh)/2.
Thanks to the existence of surveys for the Ma-
drid-Seville corridor, made after its commissioning, it
has been possible to distinguish within the trips gener-
ated, those derived from an increased frequency of travel
(from previous users of other modes of transport ) and
those truly generated. For those who already traveled in
the past, the generalized cost of the original mean of
transport has been taken as a reference, while for the new
ones, the weighted average of the generalized costs of
the different modes has been used. In this case, the
weighing has been determined from data from the survey
conducted on the substitution effect.
This methodology is valid for conventional rail and
bus, but it needs to be adjusted on plane and private car.
As regards private car, the saving on travel diverted re-
sources has to be extended to the operational cost savings
of not travelling by car. Due to the existence of several
alternative modes, the AVE costs should be added to just
one of them, to avoid double counting.
Finally, in the case of the plane, it is noteworthy that
although the generalized cost is lower for the AVE, the
cost reduction is entirely d ue to lower prices as the travel
time component is still less on the plane. So when di-
rectly applying the aforementioned methodology, the
result would be a negative benefit. However, it is neces-
sary to keep in mind other important components of the
generalized cost: comfort, safety, etc. which reduce the
disutility in the AVE in parallel with the price. In this
perspective, it was decided to bring about a reduction in
the AVE widespread cost compared to the other modes,
according to the results of the survey on the reasons to
choose the AVE instead of the plane2.
This framework has been extended with the method-
ology for the valuation of externalities in transport, as
detailed in [11].
In addition, all the characteristic endpoints of the so-
cial assessment of projects, such as shadow prices, ex-
cluding taxes, etc as detailed in [7], have been used.
Table 2 shows the costs and benefits used for evalu-
ating the project type. Therefore the macroeconomic,
sectorial and regional effects have not been directly con-
sidered. Such effects are analyzed in detail by [13] and
[14].
Finally, the benefit implied by the introduction of the
AVE for the territories concerned is quite obviou s, espe-
cially when taking into account the peripheral situation
of Spain regarding the main centers of production and
decision of the European economy. In [15] the author
describes the regional effects produced by the TGV.
According to [7], the methodology to evaluate costs
and benefits is the following:
3.1. Cost
As for any other product, the AVE costs can be divided
into fixed, semi-fixed and variable, naturally depending
on the term considered. Fixed costs are those corre-
sponding to the construction of the infrastructure and its
maintenance (even though in the long term such costs
will probably evolve in parallel to the demand).
Semi-fixed costs correspond to the purchase of rolling
stock and, lastly, variable costs are those commonly
called operating costs, characterized by being highly
sensitive to the evolution of the demand. Taxes are ex-
cluded from the cost sections.
Table 2. Costs and benefits considered in evaluating the
project type.
Costs Benefits
Total cost of infra-
structure, maintenance
and operation
(TC = I – R + RS + IM
+ O)
Total benefits
(IGT + TS + CR + REC)
Infrastructure (I) Income from Generated Trips (IGT)
Residual value (R) Time saving (TS) for users from:
Rolling Stock (RS) Other methods of tr a ns p o rt
Infrastructure mainte-
nance (IM) Cost reduction (CR) in:
Operation (O) Conventional railway
Airplane
Coaches
Operational car costs
Reduction of external costs (REC)
in:
Congestion
Accidents
Environment
Infrastructure maintenance
2According to the assessments given by users and expressed in surveys,
the generalized cost of alternative ways has been expanded, to intro-
duce other useful components such as comfort, security, etc. To this
end, these costs have been increased by 1.8 for car, 1.2 for rail, 1.15 fo
r
coach and 1.5 for airplane.
Source: Authors’ compilation based on [7,11,12].

P. COTO-MILLÁN ET AL.
110
Regarding prices, it has been assumed that all sections
are valued at constant prices of 2008. Thus, the method-
ology applied for quantifying each of those costs is as
follows:
3.1.1 Infrastructure Construction Costs
The AVE infrastructure includes the track as well as the
earthworks, signalling, stations, catenary, etc. The total
cost of the 855 kilometres amounts to 7,928 million Eu-
ros. Each kilometre of infrastructure built amounts to a
cost of 9.27 mill i on Eur os.
3.1.2. Infrastructure Maintenance Costs
The annual maintenance cost for the infrastructure3 has
been estimated at 11.75 thousand Euros per km. It is im-
portant to point out that in the long term this cost is
probably slightly sensitive to the levels of demand. This
effect has not been considered, as it has been assumed
that it would be offset by a probable downward trend of
the maintenance unit costs linked to the economies of
scale.
3.1.3. Rolling Stock Costs
The costs considered for the three different types of roll-
ing stock are as follows: 20.7 million Euros for long-
distance trains; 9.6 for regional shuttle trains; and 11.4
for variable gauge trains.
3.1.4. Costs for Obtaining The Necessary Rolling
Stock Units
The methodology used to determine the necessary units,
according to [10], is as follows. The average capacity of
Siemens and Talgo trains for long haul is 361 passengers.
A load factor of 0.63, which means 228 passengers car-
ried per unit, was considered. When multiplying this
value by the average mileage covered by each unit
(450,000 kilometers) per year, the result is that every
102.60 million passenger-kilometers per year are neces-
sary to incorporate a new train. For shuttle trains and
variable gauge trains a new train is necessary every
57.12 and 54.72 million passenger-kilometers.
It is necessary to should point ou t that the calculations
are based on ideal operating conditions or maximum
efficiency conditions in the incorporation of the new
rolling stock.
3.1.5. Oper ating Costs
This chapter includes all costs derived from operating the
AVE, which have been obtained separately for both
types of material. For long-haul material, the cost is 8.21
Euro cents in 2008 per passenger-kilometer, while for the
other segments of supply it is 6.98 cents.
3.1.6. Resid ual Value
For the rolling stock, the useful life considered is 20 years.
Together with this, a linear depreciation during this pe-
riod has been assumed. For the entire infrastructure, a
value of 45 years has been estimated [16]. Therefore, af-
ter 40 years of operation, and assuming a linear deprecia-
tion rate, the residual value would be minimal, approxi-
mately 10% of val ue of the investm e nt made.
3.2. Benefits
The benefits considered in the project and the methodol-
ogy used for their assessment are detailed below.
3.2.1. Time Savings
Part of the benefits related to the introduction of the
AVE consists of the time savings for passengers origin-
nating from other means of transport and also for new
passengers. For its assessment, it is necessary to know
the modal travel times from the origin to the destination
for each mean of transport (including the times to and
from the station or airport) and the monetary values as-
signed to the total travel time (see Table 3). Likewise, in
order to calculate the savings corresponding to the gen-
erated trips, it is necessary to have access to the mone-
tary costs of each means of transport that makes it possi-
ble to obtain the income (see Table 3).
Table 3. Model travel times, Modal monetary values of
travel time and monetary costs per transport mean. route
Madrid-Barcelona.
Model travel times
Car Airplane Coach Train AVE
Journey time5 h 45'55' 7 h 35' 6 h 35'2 h 35'
Access and
dispersion
time 1 h 35' 50' 1h 50'
Total time 5 h 45'2 h 25' 8 h 25' 7 h 35'3 h 25'
Modal monetary values of t r a v e l t i m e
(Euros of 2008 per p as se ng er/ ho ur )
Car Airplane Coach Train
6.04 25.56 3.25 13.01
Monetary costs per transp or t mean (Euros of 2008)
Airplane Coach Conventional train AVE
1ª Class 2ª Class Busi-
ness Tourist
122.05 22.84 58 44 128.6886.70
3This value includes other concepts such as station maintenance
(ADIF). Source: [7,16] and ow n el aboration.

111 P. COTO-MILLÁN ET AL.
The modal monetary values of travel time used in this
work have been provided by the Administrative Depart-
ment responsible for transport in Spain. Such values,
which are detailed in [16], have been updated. The same
values have been used in [10] for the Madrid-Sevilla
AVE and in [7] for the Madrid-Barcelona-French bor-
der.
3.2.2. Cost Reduction in Alternative Means of
Transport
Another type of benefits, inherent to the operation of the
new product, consists of the reduction of costs in the
alternative means of transport: conventional railway,
airplane, coach and car.
3.2.3. Redu ction of Costs in Co nventional Railway
The intermodal substitution produced by the AVE is es-
pecially pronounced for the conventional railway, which
practically becomes a marginal mean of transport on this
route in terms of its passenger transport function..
In order to determine the corresponding cost reduction,
the structure of the production4 cost of the average day-
time train has been used, it being the most affected by
the reduction of the offer. Of all its components, only
half of the cost associated to the chapter of stations has
not been included. However, the entire amortisation
chapter has been included, based on the fact that the train
could be used alt ernat i vely in another route.
As the offer along this route has practically disap-
peared, in order to determine the total cost it is valid to
use a unit cost ratio per passenger-km, which has been
obtained by applying to the national average value a co-
efficient representing the occupancy differential existing
in this route with regard to the na tional average.
This has been estimated at 4.63 euro cents per travel-
ler-kilometre originating from conventional train travel.
3.2.4. Reduction of Costs in Air Transport
The transfer of travellers from air transport to the AVE
brings about a reduction in costs for air transport opera-
tors in the Madrid-Barcelona route. In this mode of
transport, unlike the case of the conventional train, aside
from the cut in the number of fligh ts, the reduction of the
offer takes place through a reduction of the average oc-
cupancy.
For this reason, instead of basing itself on a unit cost
per traveller-km, the method used is based on establish-
ing the reduction of the number of flights in order to ob-
tain the cost savings for the trip by applying the average
cost of a flight.
The cost savings per unit per trip have been obtained
by subtracting from the total cost the fixed items such as
structural costs and part of the commercialisation costs,
as well as all taxes. The amortisation chapter has been
kept, given that an aircraft can be used on another line, as
it is clearly understood when the operator is renting the
aircraft.
The final result is 15.23 Euro cents for the year 2008
per passenger-km originating from air travel.
3.2.5. Red u ction of Costs in Coaches
In coaches, the effect of the intermodal substitution
brought about by the introduction of the AVE generates
cost savings for the operating co mpanies, essentially as a
result of the reduced number of trips. In order to calcu-
late these savings, the costs structure of a representative
coach gathered by the [16] has been used. It is worth
noting that in the cost saving calculation, all of its com-
ponents have been considered, including depreciation, as
there is a possibility, in the form of opportunity cost, to
use the vehicle for another line in the medium term.
Therefore only the costs of taxes have been excluded,
since it is assumed that the costs of structure or organiza-
tion are included in the discretionary coach data. The
savings determined amount to 3.92 Euro cents perpas-
senger-km.
3.2.6. Red u ction of Car Costs
In the case of private vehicles, it is not possible (except if
the vehicle is rented) to substitute the trip planned ini-
tially by another one in a different place during the same
period of time. Therefore cost chapters considered as
“fixed”, such as part of the amortisation and insurance,
should not be included in determining cost savings. As to
the part of the amortisation chapter included, [16] has
been used as starting point, where the authors estimate
that half of said chapter corresponds to the passing of
time, whereas the other half is related to the use of the
vehicle.
Continuing with this assumption, the structure of the
cost savings has been obtained, eliminating the fixed
chapters and the taxes, for each of the routes that make
up the corridor. The costs considered in each section and
for which specific methods have been used are: vehicle
amortisation, maintenance, fuel consumption, lubricant
consumption and tyre wear. The calculated savings per
unit amount to 52.87 euro cents per passenger-km.
3.2.7. Red u ction of External Costs
For the assessment of external costs, the results of [7],
which ar e shown in Table 4, have been considered.
4Therefore, organizational costs, known as structural costs have no
t
been included.

P. COTO-MILLÁN ET AL.
112
Table 4. External marginal social costs (euro cents perpas-
senger-km).
Environ-
mental Infrastructu re conservation Accidents Congestio - n
Car 1.89 0.84 3.05 2.14
Train 0.60 2.62 0.17 -
Coach 0.67 0.16 0.78 0.37
Airplane 2.46 - - -
AVE 0.49 - - -
Source: [7] and own elaboration.
4. Parameters
In the simulations conducted on the project type and in
order to bring this work to a rigorous sens itivity analysis,
several parameters have been considered. The empirical
evidence shows that the following parameters are the
ones that have greater impact on the profitability of the
project.
4.1. The Discount Rate Parameter
In order to determine the final net benefit of the project,
it is necessary to update the annual benefits and costs
with regard to the year considered as base, by means of a
real social discount rate r, in such a way that the present
value (NPV) of the investment is:
 
00 11
nn
t
t
tt
BC
NPV rr




t
t
(2)
Where; Bt, Ct and r represent for benefits, costs and real
social discount rate respectively.
The Manuals for the assessment of projects of the
Spanish Ministry of Development, responsible for trans-
port infrastructures, recommend the use of a social dis-
count rate of 6% in real terms. This value has also been
used in other infrastructure projects (especially roads)
carried out in Spain during the nineties. [17] uses even
higher rates (8% and 10%) for the economic evaluation
of the Barcelona rounds.
However, since the incorporation of the Spanish
Economy to the European Economic and Monetary Un-
ion, the consolidation of favourable economic expecta-
tions has been translated into a consolidation of low val-
ues of real interest rates below 3%. Therefore, it seems
logical to use as a basic alternative in our assessment of
the model project a magnitude of 4% for the real dis-
count rate in the medium term.
4.2. Project Duration Parameter
Two project’s life periods have been considered, 40 and
60 years, which seem particularly suitable given the pro-
ject’s size and the Madrid-Sevilla AVE experience, par-
ticularly bearing in mind that u sing a discount rate of 4 %
would penalize less net income in future years.
The evaluation is made with constant prices of 2008. It
involves the maintenance of relative prices of different
goods and services during the life of the project.
4.3. Economic Growth Rate Parameter
It has been assumed that the annual GDP growth rate is
3% during the entire project. This value would corre-
spond to the growth rate of the potential Spanish GDP
according to the many studies carried out on the Spanish
economy until the 2008 crisis and matches the average
growth rates over the past 10 years. For the sensitivity
analysis, annual economic growth rates of 2.5% and
3.5% have been con si dered.
5. The Demand
Given its importance in relation to the magnitude of
AVE net benefit, the main characteristics of the AVE
demand are analyzed, distinguishing its two components.
The introduction of the high-speed rail brings a sig-
nificant reduction in the generalized cost of rail transport,
this fact generates a significant volume of demand linked
to this new mode of transportation. Such reduction,
which occurs in the non-cash components of the cost
(time, comfort, etc.) produces two effects on the demand
which are called, induction and substitution. They corre-
spond to the trips that would not have been done if this
new service had not existed and to those that would have
been developed in another mode of transport.
5.1. Induction
The component of the AVE demand, usually termed in-
duced, is comprised by all the new trips. This “genera-
tion effect” must not only include passengers who have
never made such a trip, but also another component
formed by the increased frequency of trips by those who
were already travelling on that rou te before the existence
of the AVE. The average of annual trips made by users
in the Madrid-Seville route increases very significantly,
from 11.1 to 15.2 [18].
In addition, according to RENFE Operadora, in the
route Madrid-Barcelona, 40% of travelers move fre-

113 P. COTO-MILLÁN ET AL.
quently, ie at least once every two weeks.
5.2. Substitution
The introduction of high-speed train brings not only a
significant reduction in the travel time but also in other
components of the generalized cost, such as safety and
comfort. From a survey of users of the Madrid-Seville
AVE in [7] show the importance of such components
within the preferences of travelers. In this sense, it is
noteworthy that, within the main reasons for choosing
the AVE, comfort has a weight (29%) almost similar to
time (30%) and significantly higher than price (11%).
Due to the high magnitude of the substitution effect,
the introduction of the high-speed train produces very
significant effects on the demand for the modes of
transport which compete with it. Apart from the disap-
pearance of the conventional train, the introduction of the
AVE causes a significant decrease in air traffic. Accord-
ing to AENA, such decrease is of 40% in the case of
Madrid-Barcelona, for the first two years of the service,
which match the calculations of [7]. In the case of the
private car, the losses are lower than in the previous
cases, approaching 20%. Finally, in the case of the bus,
there appears to be a strong impact on long-haul trips
(11% loss), as both products are hardly substitutes.
The initial demand of the corridor before the intro-
ducetion of the AVE has been calculated using data from
transport operators and surveys of mobility. Factor re-
placement and the new generation of travel demand have
been determined from the data of the first two years of
online s ervice, prov ided by RENFE and AENA, and fol-
lowing [18], (s ee Table 5).
Table 5. Number of passengers on the Madrid-Barcelona
route, 2007-2009 (in millions).
Transport mode Before
AVE After AVE
2007 2008 2009
Airplane 4.81 4.01 3.06
Conventional railway 0.76 ------ ------
AVE ------ 2.14 2.70
TOTAL 5.57 6.15 5.76
Source: Own elaboration based on data from AENA and RENFE Opera-
dora.
In addition, to apply the tool of benefit-cost analysis
for the evaluation of the project, it is necessary to know
not only the initial demand to and from AVE Ma-
drid-Barcelona but also the demand of the intermediate
points of the corridor (Zaragoza, Tarragona, etc). To-
gether with this, it is necessary to know the evolution
during the holding period considered, broken down into
its two components, traffic generated and diverted traffic,
as well as the revenues needed to assess the benefits
produced by the generated trips.
In this sense, it has been estimated that the total num-
ber of travellers for the High Speed corridor as a whole
(Madrid, Zaragoza, Lleida, Tarragona and Barce-
lona),that is to say, considering not only travellers with
origin-destination Madrid-Barcelona, was 5.71 M trav-
ellers in 2009, according to the information provided by
RENFE Operadora. Besides, it has been estimated that if
this line is extended until the French Border and con-
nected to the French High Speed network, demand could
increase 15%. This would mean reaching 6.56 M travel-
lers. In [19] it is proved the existen ce of positive network
effects for the HSR Madrid-Barcelona with a possible
connection to the French high speed network.
It has been considered that the demand for the AVE
evolves from the first four years in which the evolution
of demand is similar to the logistic curve, with a GDP
elasticity of 1.4. This value is equivalent to that obtained
for air transport in Spain [20], and is consistent with
those obtained in other countries [21]. It is recalled that
several analysts as [16] argue that the AVE is more
similar to an airplane than to a conventional train.
The AVE social profitab ility is ob tained by comparing
the costs and benefits described above, updated to the
base year by the social discount rate chosen. A value of
demand from which the present value of the net social
benefits is positive is obtained for each chosen value of
the GDP growth rate and the other parameters.
6. Results
Figure 2 shows the values of the costs and benefits of a
project type, updated to the base year with a real social
discount rate of 4%, assu ming a duration of 40 year s and
a 3%, growth rate of the Spanish GDP, which are the
values chosen for the baseline scenario. Th e results show
that the costs and social ben efits of the AVE are sensitive
to the demand of the corridor.
In addition, the slope of the profit curve is signify-
cantly higher than that of the cost curv e. The c aus e of the
weak sensitivity of the costs to the level of demand lies
in the important weight of fixed costs, particularly those
relating to infrastructure [22]. In this sense, when com-
paring with other modes of transport, the profitability of

P. COTO-MILLÁN ET AL.
114
the AVE is much more dependent on traffic density in
the corridor as the supply of additional units of rail ser-
vice incorporates a much smaller additional cost due to
the intense effect of the economies of scale.
It is also shown that the level of demand for the inter-
section point of both curves, which represents the mini-
mum level of traffic that makes the project type socially
profitable, is 6.5 million of equivalent passengers.
Table 6 shows the results of the sensitivity analysis
performed with different parameter values. The values
obtained for the threshold of demand vary between 6.1
and 8.4 million passengers. Additionally, the introduce-
tion of higher values for human life and the duration of
the project (60 years) slightly reduced the magnitude
cited.
For values of the AVE demand lower than those ob-
tained previously, a first option to consider in order to
make the project profitable is reducing costs, particularly
those related to infrastru cture, which would move up the
cost curve.
Source: Own research.
Figure 2. Evolution of costs and profits in relation to de-
mand. (Assumptions: Rate of discount 4%; Growth rate
3% and duration 40 years).
Table 6. Minimum volume of demand for a profitable high
speed train (Millions of equivalent passengers), (sensitivity
analysis).
GDP growth rate
Rate of
discount 2% 2.5% 3% 3.5%
4% 7 6.7 6.5 6.1
6% 8.4 8 7.6 7.1
Source: Own research
Alternatively, the profit curve could be moved up-
wards, by decreasing the threshold of demand that lev-
erages the project. An example of this alternative is the
use of the AVE for the transport of goods, which would
generate increases in various chapters of the benefits:
time savings, reduced maintenance costs of conventional
road and rail infrastructure, etc.
7. Conclusions
The evaluation of any AVE project reveals the existence
of numerous factors that influence to a greater or lesser
extent its profitability. The demand is pointed out as the
most important factor, as this new mode of transportation
is characterized by its high speed, more than twice that of
the conventional rail, but also by high fixed costs of inf-
rastructure, virtually independent of the number of pas-
sengers, which make it necessary to hand le high volumes
of demand to achieve an acceptable level of profitability.
In addition, the AVE profitability is sensitive to other
parameters such as: social discount rate, economic
growth rate, project duration or value of human life. The
various simulations performed for the expected values of
4% for the real social discount rate, 3% for the GDP
growth rate and project duration of 40 years, makes it
possible to verify that the volume of traffic that marks
the threshold of social return is 6.5 million equivalent
passengers.
In the case of the Madrid-Barcelona AVE, the number
of travellers in the corridor as a whole was 5.71 M, ac-
cording to RENFE Operadora. If no other possible ef-
fects are considered, the project shows a negative bal-
ance in the present circumstances. These results can be
applied to other HSR projects with similar characteristics,
that is, covering similar distance trips, with the same
alternative transport modes (bus, private vehicle, con-
ventional train and plane) and which generalized cost is
similar too.
Failure to meet this demand threshold, it would be
necessary to consider alternative projects, associated
with lower costs or higher profit levels.
Furthermore, the development of a high performance
network on these corridors, taking advantage of the ex-
isting infrastructure, would generate new profits—net-
work effects—in the high-speed network service. Given
the benefits associated with these network eff ects, which
magnitude grows parallel to the extension of the high-
speed network, the threshold of demand obtained from
6.5 million passengers could be reduced significantly in
subsequent projects.
8. Acknowledgements
The authors gratefully acknowledge the suggestions and

P. COTO-MILLÁN ET AL.
115
criticisms from two anonymous rev iewers and the editor.
However, errors and omissions that may exist are the
sole responsibility of the authors.
9. References
[1] D. A. Aschauer, “Is Public Expenditure Productive?” Jo-
urnal of Monetary Economics, Vol. 23, 1989, pp.7- 7-100.
doi:10.1016/0304-3932(89)90047-0
[2] PEIT, Plan Estratégico de Infraestructuras y Transportes
2005-2020. Madrid, Department of Transportation, 2005.
[3] COM, 2001/370-European Commission, WHITE PAPER,
“European Transport Policy for 2010: Time to Decide,”
Brussels, COM (2001) 370 final.
[4] N. Adler, C. Nash and E. A. J. H. Pels, “High-Speed
Rail&Air Transport Competition: Game Engineering as
Tool for Cost-Benefit Analysis,” Tinbergen Institute
Discussion, , 2010, Paper No. 08/103/3.
[5] G. De Rus and C. Nash, “In what circumstances is in-
vestment in HSR worthwhile?”.Working paper 4th, An-
nual Conference on “Railroad Industry Structure, Com-
petition and Investment,” Universidad Carlos III de Ma-
drid, 2006.
[6] G. De Rus, “The BCA of HSR: Should the Government
Invest in High Speed Rail Infrastructure?” The Journal of
Benefit-Cost Analysis, Vol. 2, No. 1, 2011, pp.1-26.
[7] P. Coto-Millán and V. Inglada, “Social Benefits of In-
vestment Projects: the Case for High-Speed Rail,” In Es-
says on Microeconomics and Industrial Organization,
Second Edition, Physica-Verlag, Springer-Verlag Com-
pany, Germany, 2003, Chapter 23.
[8] J. Dodgson, “Railways Costs and Closures,” Journal of
Transports Economics and Policy, Vol. 18, No. 3, 1984,
pp. 219-235.
[9] G. De Rus and V. Inglada, “Análisis Coste-Beneficio del
Tren de Alta Velocidad en España,” Revista de Economía
Aplicada, Vol. 1, No 3, 1993, pp. 27-48.
[10] G. De Rus and V. Inglada, “Cost-Benefit Analysis of the
High-Speed Train in Spain,” The Annals of Regional
Science, Vol. 31, 1997, pp. 175-188.
doi:10.1007/s001680050044
[11] P. Coto-Millán and V. Inglada, “Market Failures: The
Case for Road Congestion Externalities,” In Essays on
Microeconomics and Industrial Organization, Second
Edition, Physica-Verlag, Springer-Verlag Company,
Germany, 2003, Chapter 22.
[12] P. Coto-Millán and V. Inglada, “Innovación en el
Transporte: El Tren de Alta Velocidad,” Economía
Industrial, No. 353, 2003, pp. 83-88.
[13] O. Álvarez and J. A. Herce, “Líneas Ferroviarias de Alta
Velocidad en España,” Economía Aplicada, Vol. 1, No. 1,
1993, pp. 5-32.
[14] V. Inglada, “El Papel de las Infraestructuras en la
Competitividad y el Desarrollo Económico,” Estudios
Territoriales, Vol. 1, No. 97, 1993, pp. 397-409.
[15] A. Bonnafous, “The Regional Impact of the TGV,”
Transportation, No. 14, 1987, pp. 127-137.
doi:10.1007/BF00837589
[16] MOPT, “Manual de Evaluación de Inversiones en
Ferrocarril de Vía Ancha,” Ministerio de Obras Públicas
y Transportes, Madrid, Department of Transportation,
1991.
[17] P. Riera, “Rentabilidad Social de las infraestructuras: Un
Análisis Coste-Beneficio,” Civitas, Madrid, 1993.
[18] V. Inglada, “Análisis Empírico del Impacto del AVE
Sobre la Demanda de Transporte en el Corredor
Madrid-Sevilla,” Revista de Estudios de Transportes y
Comunicaciones, No. 62, 1994, pp. 35-51.
[19] V. Inglada, P. Coto-Millán and B. Rey, “Effects of Net-
work Economics in High Speed Rail: The Spanish Case,”
Annals of Regional Science, No. 41, 2007, pp. 911-925.
doi:10.1007/s00168-007-0134-6
[20] P. Coto-Millán, V. Inglada and J. Baños-Pino, “Marshal-
lian Demands of Intercity Passenger Transport in Spain,
An Economic Analysis 1980-1992,” Transportation Re-
search E, Logistics and Transportation Review, Vol. 33,
No. 2, 1997, pp. 79-96.
[21] A. D. Owen and G. D. A. Phillips, “The Characteristics
of Railways Passenger Demand,” Journal of Transport
Economics and Policy, Vol. 21, No. 3, 1987, pp. 231-253.
[22] L. S. Thompson, “High-Speed Rail in the Unites States:
Why isn’t There More? ” Japan Railway & Transport Re-
views, October, 1994, pp. 32-39.