Cost overruns and delays in infrastructure projects: the case of Stuttgart 21
| Published date | 10 November 2020 |
| DOI | https://doi.org/10.1108/JPIF-11-2019-0144 |
| Date | 10 November 2020 |
| Pages | 256-282 |
| Subject Matter | Property management & built environment,Real estate & property,Property valuation & finance |
| Author | Bertram I. Steininger,Martin Groth,Brigitte L. Weber |
Cost overruns and delays in
infrastructure projects: the case of
Stuttgart 21
Bertram I. Steininger
Real Estate Economics and Finance,
School of Architecture and the Built Environment,
KTH Royal Institute of Technology, Stockholm, Sweden, and
Martin Groth and Brigitte L. Weber
RWTH Aachen University, Aachen, Germany
Abstract
Purpose–We investigate causes for the cost overrun and delay of the railway project Stuttgart 21. Besides, we
try to forecast the actual costs and completion date at an early stage.
Design/methodology/approach –The results of exploratory research show the causes for the costoverrun
and delay of Stuttgart21; we compare ourfindings with other railway projects. To estimate the costs at an early
stage, the reference class forecasting (RCF) model is applied; to estimate the time, we apply an OLS regression.
Findings –We find thatthe following causes arerelevant for the cost overrunand delay of Stuttgart 21:scope
changes,geological conditions,high risk-takingpropensity,extended implementation,price overshoot,conflict of
interestsand lack of citizens’participation.The current estimated costsare within our 95% confidenceinterval
based on RCF; our timeforecast underestimates or substantially overestimatesthe duration actually required.
Research limitations/implications –A limitation of our approach is the low number of comparable
projects which are available.
Practical implications –The use of hyperbolic function or stepwise exponential discount function can help
to give a clearer picture of thecosts and benefits. The straightforward use of the RFC for costs and OLS for time
should motivate more decision-makers to estimate the actual costs and time which are necessary in the light of
the rising demand for democratic participation amongst citizens.
Social implications –More realistic estimates can help to reduce the significant distortionat the beginningof
infrastructure projects.
Originality/value –We are among the first who use the RCF to estimate the costsin Germany. Furthermore,
the hyperbolic discounting function is added as a further theoretical explanation for cost underestimation.
Keywords Cost overrun, Time overrun, Infrastructure, Reference class forecasting, Hyperbolic discounting,
Principal-agent theory
Paper type Research paper
1. Introduction
In our globalised world, there is a significant need to travel more frequently and move goods
easily and fast. Also in regional terms, the travel and commuting pattern has changed in
recent years. More people commute over a longer distance and more often. This trend has led
to more and larger infrastructure projects in order to increase mobility and stimulate the
economy. The complexity of such mega infrastructure projects increases more than
proportionally with their size. At the same time, citizens demand greater participation in the
decision-making process of such projects –due both to their part as financial backers with
JPIF
39,3
256
JEL Classification —H54, O18, R42
The authors thank Nick French (editor), Seow Eng Ong (guest editor) and anonymous referees for
their comments.
This paper forms part of a special section “Industrial Infrastructure and Real Estate”, guest edited by
Professor Seow Eng Ong, Associate Professor Chyi Lin Lee.
The current issue and full text archive of this journal is available on Emerald Insight at:
https://www.emerald.com/insight/1463-578X.htm
Received 17 November 2019
Revised 14 June 2020
22 August 2020
Accepted 24 September 2020
Journal of Property Investment &
Finance
Vol. 39 No. 3, 2021
pp. 256-282
© Emerald Publishing Limited
1463-578X
DOI 10.1108/JPIF-11-2019-0144
their tax-money and also to their involvement as residents who want to live in a quiet, healthy,
and sustainable environment.
In their book “Megaprojects and Risk: an Anatomy of Ambition”Flyvbjerg et al. (2003a)
show that more large-scale infrastructure projects [1] have been implemented over the last
decades but they still suffer from delays, cost overruns, and market shortfalls. They call this
“megaprojects paradox”and instance the UK–France Channel Tunnel project in Europe with
an 80% cost overrun, the Denver International Airport in the USA with a 200% cost overrun,
or the Korean high-speed railway project in Asia with a 217% cost overrun. They conclude
that the reasons for the paradox are that the stakeholders in such projects are incentivised to
underestimate costs and environmental impact but overestimate revenues and economic
benefits. This is mainly driven by a lack of accountability and inappropriate risk sharing and
Flyvbjerg et al. (2003a) recommend reforms in the institutional setting during the decision-
making process and in institutional accountability.
In this research,we investigate the extent and causes for the costoverrun and delay of the
railwayproject Stuttgart 21; it wasthe megaproject in Germanywhich initiated the demandfor
more democratic participation of the citizens concerned in the decision-making process of
infrastructure projects. The European Central Bank (ECB) headquarters in Frankfurt, the
Elbphilharmonie Concert Hall in Hamburg, and the Berlin–Brandenburg (BER) Airport are
additionalexamples for cost and time overrunsof infrastructure projects whichhave received
extensivemedia coverage in Germany. Among the finished projects, the ECBwas constructed
three years later than planned and with a cost overrun of 48%; the Elbphilharmonie was
finished seven years later and the cost mushroomed with an overrun of 1,025%. Berlin–
Brandenburg (BER)and Stuttgart 21 are still unfinished. Since Stuttgart 21 was launched in
April 1994, it has repeatedly been subject to changes regarding the total costs and the
completion date.At the moment, the opening is planned for 2025–6 years later than planned,
and the cost overrun so far is around 228%.
These overruns are not recent phenomena and occur cross time and geographical
locations. Based on different infrastructure projects (roads, rail, bridges and tunnels) in
different regions, Flyvbjerg (2009) and Cantarelli et al. (2012a) detect a systematic
underestimation of costs at the beginning of project planning and no improvements in the
estimation process over the last 70–80 years. This result is surprising, as it implies that the
usual learning curve effects do not apply. The literature provides a range of different causes
and theories for the underlying reasons and suggestions that could help to arrive at more
realistic estimates, supporting a more efficient decision-making process at the beginning of a
project. We add to this literature by arguing that the short-term orientation of traditional
discount functions (exponential function) misrepresents the true contribution of the long-
term benefits of infrastructure projects which occur during the later operation period. The
hyperbolic discounting function could mitigate this trade-off. All the theoretical explanations
from the previous literature are discussed in the context of Stuttgart 21 to identify the main
causes of its cost overrun and delay and to be able compare them with other projects. In
addition, we use the reference class forecasting (RCF) model in order to see whether it
provides a more realistic estimate of the total costs at an early stage. A different forecast
model is applied for the time until completion.
The remainder of this paper is organised as follows: We give an overview of the project
Stuttgart 21 in Section 2. Next, we summarise the theoretical explanations for cost and time
overruns of railway projects in particular (Section 3) and of infrastructure projects in general
(Section 4). These theoretical explanations are discussed in the context of Stuttgart 21 to
identify the main causes of its cost overrun and delay (Section5). To estimate the costs at an
early stage, the RCFmodel is applied to Stuttgart 21 to see whether it leadsto a more realistic
estimate(Section 6). The final section givesconclusions and offers recommendationsas to how
costs and construction times might be estimated more precisely at an early stage.
Cost overruns
and delays in
projects
257
To continue reading
Request your trial