Modelling open skies agr ts and air p ger competition
S.P. Shepherd*
A. Orta*
*Institute for Transport Studies, University of Leeds, Leeds LS2 9) T, UK.
S.P.Shepherd@ its.leeds.ac.uk
* Félix Parra 30, int 101, San osé Insurgentes, Mexico City, 03900, Mexico.
ortaarmando@ gmail.com
Abstract
This work aims to study competitive airline behavior, regarding pricing and supply decisions, and how
it adjusts to various structural assumptions, such as the existence of capacity expansion delays, by
employing system dynamics simulation techniques. Results show that, in a competitive environment,
the airline with the most aggressive market share expansion strategy would start transporting more
passengers and sustained an advantage for the first 15 years, of a 60-year horizon, only to be
undertaken by its competitor and end up sharing 50% of the market at the end. Furthermore, both
airlines undercut their fares to the point of reaching their operating unit costs, which goes in line with a
Bertrand competition behavior (Silva and Verhoef 2013), creating benefits to consumers.
1. Introduction
Given the promotion of a more liberal framework for international airline markets, it is
pertinent to assess the role of competition in the industry. Specifically, how the entry of new
carriers might modify fares and capacity in the medium and long term and how this then
impacts on travelers. For this case, system dynamics has proven to be a helpful tool for the
simulation and evaluation of the complexities of the airline business cycles, and the
effectiveness of competitive strategies in capital intensive industries (Lyneis 2000).
In the interest of this context, the purpose of this work is to study competitive airline behavior,
regarding pricing and supply decisions, and how it adjusts to different structural assumptions,
such as the existence of capacity delays and different costs structures, by employing system
dynamics simulation techniques. Section 2 gives an overview of the literature both from
traditional modeling and within the system dynamics field. Section 3 discusses the results
and policy implications. We conclude with section 4 covering conclusions and further
research.?
2. Literature review
2.1. Airline competition models
Given the liberalisation process and subsequent “Open Skies” agreements between the
United States and the European Union, member states of the latter started to pursue the
promotion of internal competition in the region. One of the main researchers that studied the
effect of liberal bilateral agreements in the EU is Marin (1995). The author analyses the
impacts on price competition and market structure of intra-European air traffic liberalisation
by proposing a theoretical model of firm behaviour in cooperative and non-cooperative
scenarios. Specifically, the cooperative scenario resembles a regulated market, where it is
assumed that companies behave in an oligopolistic framework with perfect collusion. In the
cooperative case, the market price equilibrium outcome will be equal to a monopoly setting
and a function of cost variables and market price elasticity of demand. Moreover, the non-
cooperative scenario simulates a market with free entry and price competition, resembling
the outcome of bilateral agreements. In that case, a Cournot-Nash behaviour is assumed
1A full version of the paper is available upon request.
which yields a competitive price equilibrium, as a function of own costs and firm’s own price
elasticity of demand.
Following the analysis of European air agreements, Schipper et al. (2002) built on previous
work and studied a dataset comprised of 34 routes that varied in liberalisation status
between 1988 and 1982. The authors use a similar theoretical approach as Marin (1995), but
assume that airlines make decisions on price and frequency of route flights. Using a Two-
Stage Least Squares technique, they found that on average, in fully liberalised routes,
economy fares decreased by 34 percent, and that frequencies increased by 36 percent.
However, this work only accounts for short-term effects, and did not consider the effect of
alliance formation in the airline industry, which is a global trend that aims to tackle soaring
costs and rampant competition (Button 2009). Additionally, another acknowledged limitation
is the role of capacity constraints at airports that could impede more frequencies and hence,
more competition.
Following on the effects of air travel liberalisation on network structure, Adler and S milowitz
(2007) study the global alliances and merger decisions under competition, given the location
of their network hubs, cost structures and revenues. The authors present a four-step game
theoretic competitive merger framework, where the examination yields a state where one US
airline allies with its European counterpart, and the remaining firms choose not to unite. This
outcome proves to be beneficial for both European agents, whereas the non-allied US carrier
is greatly affected. Furthermore, Adler and Smilowitz (2007) recognise that future research
could also contemplate a combination of non-stop and hub-and-spoke flights within a
network, frequency and aircraft size variables for the market share model, and an analysis of
the model over time periods, in a larger network setting.
Although research in airline competition is vast, the classical approaches might be limited in
their modeling capacity to reproduce the airline market system and its complexities. As Adler
and Smilowitz (2007) point out, there is a need to include dynamic simulation in competition
models to observe medium and long term effects of carrier decisions. Additionally, Silva et al.
(2014) recommend the implementation of different types of airlines, and Hansen (1990)
suggest the inclusion of other agents such as aircraft manufacturers, travelers, and unions in
the analysis.
2.2. System dynamics airline and competition models
One of the main studies that favor the use of system dynamics in forecasting and structural
analysis, with an application in the airline industry, comes from Lyneis (2000). The author
explains that the aircraft manufacturing market faces a highly cyclical behaviour, which
coincides with the rest of the airline industry (Vasigh et al. 2013). Because these cycles, such
as demand fluctuations over time, the decisions of an airline on capacity expansions are
challenging, as there is a risk of over (under) investing. Passenger demand is influenced by
external variables such as GDP, and price and frequency elasticities, and can also be
classified by region and type of traffic. On the airline side, endogenous carrier decisions such
as required fleet, frequency, costs, among others are influenced by airline demand and other
elements such as external costs (fuel price, inflation, regulations and congestions). These
items also affect decisions from manufacturers of aircraft which must invest in capacity to
reduce backlogs and to develop new technologies. According Lyneis this complex structure
and its interactions makes forecasting difficult, and hence ineffective for minimising
undesirable business cycles risks.
Similarly, Liehr et al. (2001) also acknowledge the highly cyclical nature of the airline industry
and analyse its composition to recommend “cycle management” measures to mitigate
shocks. As a recommendation to mitigate these effects, Liehr et al. suggest the creation of
an autonomous unit within the airline to ensure quasi-continuous capacity flow.
Furthermore, Pierson and Sterman (2013) build up from previous work and developed an
aviation industry behavioural dynamic model, which endogenously accounts for capacity
expansion, demand, pricing, wages, among other feedback elements. By employing
2
historical data, and estimating model parameters, they find that delays in aircraft
manufacturing are not a relevant feedback on the profit cycle, as pointed by Liehr et al.
(2001). The authors suggest that a less intense use of pricing management techniques might
increase profit returns for investors. However, a closer look ata more disaggregated airline
competition model is also advised to validate this recommendation.
Standard literature in airline competition, such as Marin (1995) and Schipper et al. (2002),
employs neoclassical economic theory, where it is assumed that carriers will only make
decisions on the production levels necessary to reach an equilibrium price where both firms
maximize their profit functions. However, these games are usually estimated in a static
setting, where firms’ objectives and market structures are assigned a set of linear equations
that are solved simultaneously, and yield a Nash equilibrium outcome (Viscusi et al. 2005).
An SD approach allows for the use of non-linear equations for the objective setting, which is
analyzed by using dynamic simulation. This situation creates and interesting opportunity to
build up from existing studies, theory and models to look closer at airline behavior.
3. Results?
With the model specification, we tested two structural scenarios, one where only one airline
would serve a route, and the other where there will be two airlines competing. In the
competitive scenario, we examined different strategies that two identical agents could
assume: conservative-conservative, where both airlines would seek a 50% of market share;
aggressive-conservative (also conservative-aggressive), where one airline aims to obtain an
80% of market share, whereas the other one plays a conservative strategy; and, aggressive-
aggressive, where both airlines compete for an 80% market share.
The model simulations were done with VENSIM PLE, where we assumed a 60-year horizon,
with a time step of 0.083 years, which is equivalent to a month. This means that every month
the airline will decide on changes to price, given the observed adjustments on demand and
supply. Additionally, Table 1 presents the different parameters that were considered in the
model simulations.
Table 1. Parameters employed in the model.
Parameter Description Unit Base value
Elasticity for intra-North America
Demand elasticity air travel at national level
of industry (InterVISTAS, 2007). Dimensionless -0.88
The equilibrium demand at the
reference price. It is represented
Initial reference per | as an average of 2 RT trips ina 400 | Seat*miles/period
capita demand mile route per person (illustrative). | /person 800
The population within the OD with
Reference a propensity to travel by air
population (illustrative). Person 1250
The strength of balance effect of
Sensitivity of changes in D/S in the price target
Demand/Supply (arbitrary, based on Pierson and
balance on price Sterman (2013). Dimensionless 0.4
2 The model structure and specificacion is available upon request.
Parameter Description Unit Base value
As the model period unit is one
Price adjustment year, it is assumed that the
time adjustment time is one month. Period 0.083
The price of a roundtrip seat in the
Reference price route (illustrative) S/seat 1000
Sensitivity of
attractiveness to Lower bound elasticity from Brons
price et al. (2002) Dimensionless 3
Depends on
Target market The desired proportion of market strategy (50%,
share (airline 1 & 2) | from the airline. Dimensionless 80%)
The strength of differences
Sensitivity of price between the current and the
to market share target market shares in the target
(airline 1 & 2) price of the airline (arbitrary). Dimensionless 0.25
The percentage of additional
Desired surplus desired capacity, compared to
(airline 1 & 2) demand (arbitrary). Dimensionless 0.15
Number of miles A trip of 400 miles RT * 50 times
flown per seat per period (illustrative). Miles/period 20000
Time to adjust The time required to fulfill the
order (airline 1 & 2) | order (arbitrary). Period 0.5
The time required to build the
Manufacturing lead | capacity (arbitrary, based on
time Vasigh et al., 2013). Period 2
Initial capacity The number of seats for the initial
(airline 1 & 2) period (illustrative) Seats 50
Retirements (airline | The time an aircraft is employed in
1&2) the airline (arbitrary). Period 7
Fuel costs per
available seat mile | Estimated cost for Delta Airlines.
(ASM) (airline 1 & (Vasigh et al., 2013, table 4.5, p. Cents*seat*miles/
2) 117) period 4.93
Estimated cost for Delta Airlines.
Maintenance costs | (Vasigh et al., 2013, table 4.5, p. Cents*seat*miles/
per ASM 117) period 1.09
Estimated cost for Delta Airlines.
Crew costs per ASM | (Vasigh et al., 2013, table 4.5, p. Cents*seat*miles/
(airline 1 & 2) 117) period 1.11
Other operating Estimated cost for Delta Airlines.
costs per ASM (Vasigh et al., 2013, table 4.5, p. Cents*seat*miles/
(airline 1 & 2) 117) period 0.67
Non-operating Estimated cost for Delta Airlines.
costs per ASM (Vasigh et al., 2013, table 4.5, p. Cents*seat*miles/
(airline 1 & 2) 117) period 8.48
Normal capacity
utilization (airline 1 | Estimated historical average load
& 2) factor (Pierson and Sterman, 2013) | Dimensionless 0.8
Normal profit
margin (airline 1 & | Estimated industry profit margin.
2) (Vasigh et al., 2013) Dimensionless 0.02
3.1.Monopoly scenario
When the model is adjusted to assume that there is only one airline in the market it is
possible to observe different, maybe unexpected behaviors. Figure 1 presents a selection of
the results from the simulation; in this case, we observe a decrease in price, which might
look counterintuitive. However, in this case, the specifications from the model assume that
capacity from the airline cannot be used in other routes, which translates into excess
capacity via a D/S balance below zero, at the beginning of the simulation period. This
situation forces the airline to lower its fares in the long run. Moreover, this decrease in prices
also occurs ata very slow pace, reaching its inferior limit at year 42 (compared to competitive
simulations) and might not even occur if we allow for capacity transfers.
Additionally, itis possible to observe that there exist cycles that affect capacity, demand, and
net income, just as observed in Liehr et al. (2001).
Price ($/se2%)
200 20%
0%
5 10 1 19 23 28 33 37 4 46 51 55 60 0 5 10 14 19 23 28 33 37 42 46 51 55 60
Time
(years)
2500
4 t
ts 4 000
i &
43 2500
$a 1500
y 1000
BZ 1000
£ soo 500
EY
0
0 5 4 19 2 5 60 5 10 14 19 23 28 33 37 «2 46 51 55.60
Time
(Years)
5
pe
S Gs
‘ Ye a:
a a 4
i i
H 2
5 10 4 19 23 28 33 37 42 46 51 55 60 05 wu 1 55 60
Time
(yeas) years)
Figure 1. Monopolistic scenario results. (Own work)
3.1.Competitive scenario
For the case of the competitive scenario, we assume that both carriers possess equal
characteristics but might follow different strategies. Moreover, the results presented in Figure
2 depict the aggressive-conservative scenario, which shows the most variability in market
share and other variables. In this case, the price level of airline 1 drops abruptly for the first
ten years, to later reach a floor which is its unit operating cost. This behavior can be
explained because of two effects: the initial spare capacity, compared to the airline demand,
which yields a D/S balance lower than one; and, the result of the difference between the
current and the target market share; both situations push prices down.
Because of this performance, airline 1 quickly starts gaining share, which translates into seat
capacity growth, and a higher number of desired demand and passengers transported.
However, because airline 2 holds a market share objective of 50%, it then reacts to the
strategy of its competitor by lowering its fares and catching up with its capacity, after 15
years. At the end, after 20 years, airline 1 cannot maintain its competitive advantage and
loses its share to airline 2. This reaction is explained by the impossibility of airline 1 to keep
pushing fares downwards in the long term due to its cost structure. This might not be
completely plausible, as for the last years airlines have been improving their cost structure
via technological improvements and business organization, which aids in keeping
competitiveness (Vasigh et al. 2013).
In this case, the first 15 years of the simulation period are similar to what Schipper et al.
(2002) presented. However, in the long run, if there are no improvements in technology or
business strategy, then airlines are not able to keep lowering fares, which was not
anticipated by the authors.
Interestingly, the supply for seat miles cycles of airline 2 bears lower peaks and tend to begin
shortly before airline 1 starts expanding its capacity. This might be because airline 1 reaches
its 15% surplus objective before its competitor. Nevertheless, this creates a situation where
due to fleet retirements and delays in the delivery of new seats, airline 2 might face a
temporary undersupply of seats which cause its fares to slightly increase, because of D/S
balance growth. This effect is later compensated with the provision of previously ordered
capacity. However, itis not possible to reach an equilibrium, as a consequence of these
cycles, which overshoots the model.
Furthermore, for the first years of the simulation, there is a considerable drop in net income
that is explained by the initial excess capacity from both airlines, and the lack of new orders
for the first five years. Nonetheless, as airlines compete for market share via price, they are
not able to transport many passengers until they receive the new seats ordered two periods
ago. Also, regardless the impossibility of maintaining market share in the model, airline 1
increases its net income considerably when it leads the route, which makes it attractive for
any airline to pursue an aggressive strategy.
Furthermore, regarding the NPV of net income payoffs for different strategies in the basic
scenario, Table 8, depicts how combinations of these schemes yield different results from the
last model. In this case, if both players go for a conservative strategy, there will be incentives
to deviate and be aggressive. Moreover, if players choose to be aggressive, then they are
attracted to follow a conservative strategy. For this set of combinations, a Nash equilibrium
will be either A-C or C-A, which is relatively consistent with the findings of Sterman et al.
(2007). Again, a sensitivity test for the attractiveness to price from the traveler could yield
more a precise set of results.
a)
Price (5/
0 5 10 14 19 23 28 33 37 42 46 51 55 60 0 5 10 14 19 23 28 33 37 42 46 51 55 60
Time Time
(Years) (Years)
i '
hy VAVAVIVVIVAV
$5 zo
of h
; Boo
Ho ; os
= 0 jp
H 200 H 200
0 ht 0
0 5 10 14 19 23 28 33 37 42 46 51 55 60 x 0 5 10 14 19 23 28 33 37 42 46 51 55 60
Tie Tie
(years) (years
- yt
is
ae BFS
3 3
2 2 2
i ua i 2
Ln
3 1
10
: 7
08 0
os 2 46 51 55 60 05 Wu 2 we BM 2 4 5
ime Time
(Yea) rears)
—Airline 1 Airline 2
Figure 2. Competitive scenario results, for the aggressive-conservative strategy. (Own work)
Furthermore, regarding the Net Present Value (NPV) of net income payoffs for different
strategies in the competitive scenario, for the first 20 years of the simulation, Table 2, depicts
how decisions from both agents can change the route outcomes, in terms of incomes. In this
case, if both players go for a conservative strategy, there will be incentives to deviate and be
aggressive. Moreover, if both players choose to be aggressive, then they are attracted to
follow a conservative strategy. For this set of combinations, a Nash equilibrium will be either
A-C, or C-A, which is relatively consistent with the findings of Sterman et al. (2007).
Table 2. NPV of net cumulative income payoffs for C - A strategies of the competition model, for the first 20 years
of the simulation period (Million $). (Own work)
A2
Conservative | Aggressive
Al (50%) (80%)
Conservative 15.73 |
15.3 | 16.12
(50%) 15.73 |
Aggressive / 4642 | 15.3 | 14.73 | 14.73
(80%)
In the case of the benefits for the consumer of having more competition in a route, Figure 3
depicts the total number of passengers transported per each scenario. It is possible to
observe that the monopoly arrangement is clearly the one where the less number of travelers
use the route, compared to the aggressive-aggressive scenario, where the output is slightly
larger that in the aggressive-conservative situation.
Thousands
wu
Passengers transported (Seats/Year)
BoM ow
0
0 5 10 14 19 23 28 33 37 42 46 51 55 60
Time
(Years)
Monopoly CC ——AC ——AA
Figure 3. Total passengers transported for each scenario. (Own work)
Moreover, Table 3 illustrates an estimated consumer savings for each competitive strategy.
This estimation is based on OFT (2010), where they build a methodology for estimating
savings from prosecuting non-competitive market arrangements such as cartels. In this case,
the base calculation is the monthly turnover from the monopolistic situation, which then is
multiplied by the difference in prices between the non-competitive scenario and the three
competitive cases. Afterward, total savings estimation is discounted through the evaluation
period and compared with the monopolistic turnover. In this case, itis possible to observe
how consumer savings increase if competition between airlines becomes more aggressive.
This goes in line with Marin (1995), where the conservative scenario is the closest to the
monopolistic situation (although it also offers some savings to consumers), and the most
aggressive case offers greater benefits to consumers.
The Future of Modeling and Simulation: Beyond
Dynamic Complexity and the Current State of
Science
August 15, 2014
Abstract
After a brief introduction to the state of the art of SD modeling, we
discuss recent and foreseeable innovations, and sketch a picture of what
the future field of (SD) modeling and simulation could, according to us,
look like. The pictures of the current state of the art, of the current state
of science, and of the foreseeable state of science, and three illustrations,
help us to sketch a functional road map from the current state towards
that future. Implementing this road map will require the field to voluntar-
ily reinvent itself. Since we do not know beforehand which new methods,
techniques and tools will be most useful, it is clear that the innovators will
have to experiment in a methodological sense. Without experimentation
and innovation, we could either stay on the aimless plateau or retreat into
a safe village. With experimentation and innovation, we may discover sev-
eral routes into the mountains, enjoy spectacular views, and reach many
high peaks.
1 Introduction
Many important issues within, or surpassing, the social sciences and humanities
show or may show intricate time evolutionary behavior, mostly on multiple
dimensions. Some of these dynamically complex issues are relatively well-know
and largely predictable, but have persisted for a long time due to the fact that
they are hard to understand or solve. Others —especially potential future issues
and grand challenges~ are largely unknown and unpredictable.
Most unaided human beings are notoriously bad at dealing with dynamically
complex issues. That is, without the help of computational approaches, human
beings are unable to assess potential dynamics of such complex issues and the
appropriateness of policy options to address them.
Modeling and simulation is a field that develops and applies computational
methods to study complex issues and solve problems in management science,
social science, environmental science, ete. Over the past half century, multiple
ing methods for simulating such issues and for advising decision-makers
facing them have emerged or have been further developed. Examples include
System Dynamics modeling and simulation (SD), Discrete Event Simulation
(DES), Agent-Based Modeling (ABM), Complex Adaptive Systems modeling
(CAS), Multi-Actor Systems modeling (MAS).
mode
Table 3. Estimated consumer savings for each competitive strategy, compared to monopoly scenario, for the first
20 years of the simulation period. (Own work)
Strategy Consumer benefit
Conservative-Conservative 23%
Aggressive-Conservative
(Conservative/Aggressive) 38%
Aggressive-Aggressive 43%
4. CONCLUSIONS AND FURTHER RESEARCH
This work has explored the possibilities of employing system dynamics to model competitive
airline behavior regarding pricing, capacity expansion decisions, and costs when carriers
pursue different strategies. Such analysis is motivated by the promotion of the liberalization
of the air markets around the world, and the need to estimate how the industry will behave,
considering its specific cyclical characteristics.
The simulations of the competition model with capacity delays and a cost structure presented
cycles on the supply side, as expected. These effects were transferred to the passenger
demand and profit aspects of the model. Moreover, the results showed that the airline with
the aggressive strategy would start earning market share and sustained an advantage for the
first 15 years, only to be undertaken by its competitor to end up sharing 50% of the market,
each. Furthermore, both airlines undercut their fares to the point of reaching their operating
unit costs, which goes in line with a Bertrand competition behavior.
Moreover, it was found that consumer savings increase as airlines pursue a more aggressive
goal in the market, compared to the monopoly situation.
As further research options, modeling dissimilarities concerning costs, resembling a legacy
carrier against a low-cost airline scenario, could modify these results, bringing differences to
the final shares of demand.
Regarding the airline strategies, it would be convenient to allow for the carrier to forecast
future demand for capacity expansions and attempt to anticipate and respond to strategies
from its competitors, as explored by Sterman et al. (2007). Furthermore, the model boundary
could be expanded to include route network structures, such as hubs or point-to-point
arrangements.
The potential of system dynamics in the simulation of competition is vast and could aid in the
understanding the medium and long-term outcomes of air liberalization.
10
REFERENCES
ADLER, N. and K. SMILOWITZ. 2007. Hub-and-spoke network alliances and mergers: Price-
location competition in the airline industry. Transportation Research Part B, 41(4), pp.394-
409.
AGUSDINATA, B. and W. DE KLEIN. 2002. The dynamics of airline alliances. J ournal of Air
Transport Management, 8(4), pp.201-211.
ALAMDARI, F. E. and |. G. BLACK. 1992. Passengers’ choice of airline under competition: the
use of the logit model. Transport Reviews, 12(2), pp.153-170.
BARBOT, C.,A. COSTA and E. SOCHIRCA. 2008. Airlines performance in the new market
context: A comparative productivity and efficiency analysis. J ournal of Air Transport
Management, 14(5), pp.270-274.
BELOBABA, P., A. R. ODONI and C. BARNHART. 2009. The global airline industry. Chichester,
West Sussex, U.K: Wiley.
BEN-AKIVA, M. E. and S.R. LERMAN. 1985. Discrete choice analysis: theory and application to
travel demand. MIT Press.
BRONS,M., PEELS, E., NIJKAMP, P. and RIETVELD, P., 2002. Price elasticities of demand for
passenger air travel: a meta-analysis. J ournal of Air Transport Management, 8(3), pp.165-
175:
BRUECKNER,J.K. and R. FLORES-FILLOL. 2007. Airline schedule competition. Review of
Industrial Organization, 30(3), pp.161-177.
BUTTON, K. 2009. The impact of US—EU “Open Skies” agreement on airline market structures
and airline networks. J ournal of Air Transport Management, 15(2), pp.59-71.
COLSON, B., P. MARCOTTE and G. SAVARD. 2007. An overview of bilevel optimization. Annals
of operations research, 153(1), pp.235-256.
CREEL, M. and M. FARELL. 2001. Economies of scale in the US airline industry after
deregulation: a Fourier series approximation. Transportation Research Part E: Logistics
and Transportation Review, 37(5), pp.321-336.
DOUGLAS, G. andJ. MILLER Ill. 1974. Economic regulation of domestic air transport. Theory
and practice.
FORRESTER, J. W. 1968. Principles of systems. Cambridge, Mass., Wright Allen.
GOMEZ-IBANEZ, J. A., W. B. TYE and C, WINSTON. 1999, Essays in transportation economics
and policy: a handbook i in honor of J ohn R. Meyer. Washington, D.C: Brookings Institution
Press.
HANSEN, M. 1990. Airline competition in a hub-dominated environment: An application of
noncooperative game theory. Transportation Research Part B: Methodological, 24(1),
pp.27-43.
HOLLOWAY, S. 2008. Straight and level: practical airline economics. Aldershot: Ashgate.
INTERVISTAS. 2007. Estimating Air Travel Demand elasticities Report for IATA.
LIEHR, M., A. GROBLER, M. KLEIN and P. M. MILLING. 2001. Cycles in the sky: understanding
and managing business cycles in the airline market. System Dynamics Review, 17(4),
311-332.
LYNEIS,J.M. 2000. System dynamics for market forecasting and structural analysis. System
Dynamics Review, 16(1), pp.3-25.
MANATAKI, I. E. and K.G. ZOGRAFOS. 2009. A generic system dynamics based tool for airport
terminal performance analysis. Transportation Research Part C: Emerging Technologies,
17(4), pp.428-443.
MANATAKI, I. E. and K. G. ZOGRAFOS. 2010. Assessing airport terminal performance using a
system dynamics model. J ournal of Air Transport Management, 16(2), pp.86-93.
MARIN, P. L. 1995. Competition in European aviation: Pricing policy and market structure. The
J ournal of Industrial Economics, pp.141-159.
MILLER, B. andj .-P. CLARKE. 2007. The hidden value of air transportation infrastructure.
Technological Forecasting & Social Change, 74(1), pp.18-35.
MILLER, B. andj. P. CLARKE. 2005. Investments under Uncertainty in Air Transportation: A
Real Options. Perspective. J ournal of the Transportation Research Forum, 44 (1), pp.61-
74.
11
OFFICE OF FAIR TRADING. 2010. A guide to OF T's impact estimation methods.London.
O'KELLY, M. E. 1987. A quadratic integer program for the location of interacting hub facilities.
European J ournal of Operational Research, 32(3), pp.393-404.
OUM, T.H. and Y. ZHANG. 1997. A Note on Scale Economies in Transport. J ournal of Transport
Economics and Policy, 31(3), pp.309-315.
PIERSON, K. andJ. D. STERMAN. 2013. Cyclical dynamics of airline industry earnings. System
Dynamics Review, 29(3), pp.129-156.
ROS, A.J. 2011. The determinants of pricing in the Mexican domestic airline sector: The impact
of competition and airport congestion. Review of Industrial Organization, 38(1), pp.43-60.
SCHIPPER, Y., P. RIETVELD and P. NIJ KAMP. 2002. European airline reform: an empirical
welfare analysis. J ournal of Transport Economics and Policy (J TEP), 36(2), pp.189-209.
SHEPHERD, S. 2013. Toll competition and dynamic toll-setting strategies. International J ournal
of Sustainable Transportation, 7(3), pp.186-203.
SICE, P., E. MOSEKILDE, A. MOSCARDINI, K. LAWLER and |. FRENCH. 2000. Using system
dynamics to analyse interactions in duopoly competition. System Dynamics Review,
16(2), p113.
SILVA, H. E. and E. T. VERHOEF. 2013. Optimal pricing of flights and passengers at congested
airports and the efficiency of atomistic charges. J ournal of Public Economics, 106, pp.1-
13.
SILVA, H.E., E. T. VERHOEF and V. A. VAN DEN BERG. 2014. Airline route structure
competition and network policy. Transportation Research Part B: Methodological, 67,
pp.320-343.
STERMAN, J. 2000. Business dynamics: systems thinking and modeling for a complex world.
Boston;London; Irwin/McGraw-Hill.
STERMAN,].D.,R. HENDERSON, E. D. BEINHOCKER and L. |. NEWMAN. 2007. Getting Big
Too Fast: Strategic Dynamics with Increasing Returns and Bounded Rationality.
Management Science, 53(4), pp.683-696.
SURYANI, E., S.-Y. CHOU and C.-H. CHEN. 2010. Air passenger demand forecasting and
passenger terminal capacity expansion: A system dynamics framework. Expert Systems
With Applications, 37(3), pp.2324-2339.
SURYANI, E., S.-Y. CHOU and C.-H. CHEN. 2012. Dynamic simulation model of air cargo
demand forecast and terminal capacity planning. Simulation Modelling Practice and
Theory, 28, pp.27-41.
U.S. DEPARTMENT OF TRANSPORT. 2005. Agreement between the Government of the United
States of America and the Government of the United Mexican States amending the Air
Transport Agreement of August 15, 1960, as amended and extended.Mexico City.
U.S. DEPARTMENT OF TRANSPORT. 2015. Air transport agreement between the Government
of the United States of America and the Government of the United Mexican States.
Washington, D.C.
VASIGH, B., K. FLEMING and T. TACKER. 2013. Introduction to air transport economics: from
theory to applications. Farnham: Ashgate.
VICKERS,].and G. YARROW. 1988. Privatization: an economic analysis. Cambridge, Mass;
London; MIT Press.
VISCUSI, W.K.,J.E. HARRINGTON andJ.M. VERNON. 2005. Economics of regulation and
antitrust. Cambridge, Mass; London; MIT Press.
VON STACKELBERG, H. 1952. The theory of the market economy. Oxford University Press.
ZITO, P., G. SALVO and L. LA FRANCA. 2011. Modelling Airlines Competition on Fares and
Frequencies of Service by Bi-level Optimization. P rocedia-S ocial and Behavioral
Sciences, 20, pp.1080-1089.
12
‘has
All too often, these developments have taken place in distinct field
the SD field or the ABM field, developing into separate ‘schools’, each ascribing
dynamic complexity to the complex underlying mechanisms they first and fore-
most focus on, such as feedback and accumulation effects in SD or heterogenous
:-specific networks (inter)actions in ABM. The isolated development within
eparate traditions has limited the potential to learn across fields and advance
faster and more effectively towards the shared goal of developing insights about
complex systems and supporting decision-makers facing complex
Today however, several initiatives are breaking through the silos opening
up new opportunities. Not only are different modeling traditions being used
in parallel and are hybrid methods emerging, modeling and simulation fields
also started to adopt, or accelerated their adoption of, useful methods and
techniques from data science, artificial intelligence, and recent developments in
operations research. Some of these innovations are already available while other
innovations still require a lot of experimentation. In this paper we will disc
that have recently been developed and are now being demonstrated
as well as innovations that are currently being developed and still require a lot
of experimentation.
The SD method is used here to illustrate these developments. Starting with
a short introduction to the traditional SD method in section 1, some recent
innovations are discussed in section 2, followed by some expected evolution:
esulting in a picture of the longer term future of social simulation
in section 4. Confronting the traditional method with the current state of the
art and with the expected future states of the art results in a list of recent and
ary future innovations, and, hence, in a functional road map for
software developers, and practitioners in section 5. Finally, conclusions are
drawn in section 6.
innovatioi
section 3,
neces enti
The remainder of this paper is available upon re-
quest.
References
Andersen DL, Radzicki M, Spencer RL, Trees $. 1997. The Dynamics of The
Field of System Dynamics. In: Proceedings of the 15th International Confer-
ence of the System Dynamics Society, August 19-22, 1997, Istanbul, Turkey.
Auping WL, Pruyt E, Kwakkel J. 2012. Analysing the uncertain future of copper
with three exploratory system dynamics models. In: Proceedings of the 30th
International Conference of the System Dynamics Society. St.-Gallen, CH
System Dynamics Society.
Moorlag, Auping, Pruyt. 2014. Exploring the effects of shale gas development on
natural gas markets: a multi-method approach. In: Proceedings of the 32nd
International Conference of the System Dynamics Society, Delft, NL, System
Dynamics Society.
Islam T, and Pruyt, E. 2014. An adaptive sampling method for examining the
behavioral spectrum of long-term metal scarcity. In: Proceedings of the 30th
International Conference of the System Dynamics Society. Delft, NL, System
Dynamics Society.
Kwakkel, Auping, Pruyt. 2014. Comparing Behavioral Dynamics Across Models:
the Case of Copper. In: Proceedings of the 30th International Conference of
em Dynamics Society. Delft, NL, System Dynamics Society.
Pruyt ct al. 2014. From data-poor to data-rich: System Dynamics in the era of
Big Data. In: Proceedings of the 30th International Conference of the System
Dynamics Society. Delft, NL, System Dynamics Society.
Kuipers J. 2014. Formal Behaviour Classification under Deep Uncertainty: Ap-
plying Formal Analysis to System Dynamics. In: Proceedings of the 30th
International Conference of the System Dynamics Society. Delft, NL, System
Dynamics Society.
Bankes SC. 1993. Exploratory modeling for policy analysis. Operations Research
41(3): 435-449.
Bankes SC. 2002. Tools and Techniques for Developing Policies for Complex
and Uncertain Systems. In Proceedings of the National Academy of Sciences
of the United States of America 99(3): 7263-7266.
Barlas, Y. 1996. Formal Aspects of Model Validity and Validation in System
Dynamics. System Dynamics Review 12(3): 183-210.
BenDor TK, Kaza N. 2012. A theory of spatial system archetypes. System Dy-
namics Review 28(): 109-130.
Bryant BP, Lempert RJ. 2010. Thinking inside the box: A. participatory,
computer-assisted approach to scenario discovery. Technological Forecasting
& Social Change 77(1): 34-49.
el AK. 2005. Simulation of common pool resource field experi-
a behavioral model of collective action. Ecological Economics 55(3):
120-436, http://dx.doi.org/10.1016/j.ecolecon.2004.12.014.
15
Clemson B, Tang Y, Pyne J, Unal R. 1995. Efficient methods for sensitivity
analysis. System Dynamics Review 11(1): 31-49.
Coyle GR. 1985. The use of optimization methods for policy design in a system
dynamics model. System Dynamics Review 1(1): 81-91.
Coyle GR. 1999. Simulation by repeated optimisation. Journal of the Opera-
tional Research Society 50: 429-438.
Diker VG, Allen RB. 2005. XMILE: towards an XML interchange language for
system dynamics models. System Dynamics Review 21(4): 351-359.
Dogan G. 2007. Bootstrapping for confidence interval estimation and hypothe:
testing for parameters of system dynamics models. System Dynamics Review
23(4): 415-436.
Graham AK, Ariza CA. 2003. Dynamic, hard and strategic questions: using
optimization to answer a marketing resource allocation question. System Dy-
namics Review 19(1): 27-46
Doyle JK, Ford DN. 1998. Mental models concepts for system dynamics research.
System Dynamics Review 14: 3-29.
Doyle JK, Ford DN. 1999. Mental models concepts revisited: some clarifications
and a reply to Lane. System Dynamics Review 15: 411-415.
Eberlein RL, Chichakly KJ. 2013. XMILE: a new standard for system dynamics.
System Dynamics Review 29(): 188-195.
Eker S, Slinger JH, Yiicel G. Investigating an automated method for the sensitiv-
ity analysis of functions. In P gs of the 30th International Conference
of the System Dynamics Society. St.-Gallen, CH, System Dynamics Society.
Feola G, Gallati JA, Binder CR. 2012. Exploring behavioural change through
an agent-oriented system dynamics model: the use of personal protective
equipment among pe: > applicators in Colombia. System Dynamics Re-
view 28(): 69-93.
Fallah-Fini S$, Rahmandad R, Chen HJ, Wang Y. in press. Connecting micro
dynamics and population distributions in system dynamics models. System
Dynamics Review.
Ford, A. and Flynn, H. (2005), Statistical screening of
System Dynamics Review 21(1): 273-303
ystem dynamics models.
Fiddaman TS. 2002. Exploring policy options with a behavioral climate—
economy model. System Dynamics Review 18(2): 243-267.
Ford A. 1990. Estimating the impact of efficiency standards on the uncertainty
of the northwest electric system. Operations Research 38(4):580-597
Forrester JW. 1961. Industrial Dynamics. Cambridge, MA: MIT Press. AND
Waltham, MA: Pegasus Communications. AND Productivity Press: Portland,
OR.
16
Forrester J. 1968. Principles of Systems. Wright-Allen Press: Cambridge, MA.
Forrester JW. 1969. Urban Dynamics. Cambridge, MA: MIT Press.
Forrester JW. 1991. The Beginning of System Dynamics. D-Note D-4165-1
rrester JW. 2007. $
Review 23(2-3):
stem dynami
9-370
~ the next fifty years. System Dynamics
Friedman JH, Fisher NI. 1999. Bump hunting in high-dimensional data. Statis-
tics and Computing 9(2): 123-143.
Groves DG, Lempert RJ. 2007. A new analytic method for finding policy-
relevant scenarios. Global Environmental Change 17(1): 73-85.
Hamarat C, Kwakkel JH, Pruyt E. 2013. Adaptive robust design under deep
uncertainty. Technological Forecasting € Social Change 80(3): 408-418.
Hamarat C, Kwakkel JH, Pruyt E, Loonen ET. 2014. An exploratory approach
for adaptive policymaking by using multi-objective robust optimization. Sim-
ulation Modelling Practice and Theory. Accepted for publication.
Hearne JW. 2010. An Automated Method for Extending Sensitivity Analysis to
Model Functions. Natural Resource Modeling 23(2): 107-120.
Homer JB. 2012. Models That Matter: Selected Writings on System Dynamics
1985-2010. Grapeseed Press, Barrytown, NY.
Homer JB. 2013. The Aimless Plateau, Revisited: Why the Field of System
Dynamics Needs to Establish a More Coherent Identity. System Dynamics
Review, 29(2):124-127.
Kampmann CE, Oliva R. 2008. Structural dominance analysis and theory build-
ing in system dynamics. Systems Research and Behavioral Science, 25(4):
505-519.
Kampmann CE, Oliva R. 2009. Analytical methods for structural dominance
analysis in system dynamics. In: R. Meyers (Ed.), Encyclopedia of Complesity
and Systems Science, Springer, New York, p8948-8967
Kwakkel JH, Auping WL, Pruyt E. 2013. Dynamic scenario discovery under
deep uncertainty: The future of copper. Technological Forecasting & Social
Change 80(4): 789-800.
Kwakkel JH, Pruyt E. 2013a. Using system dynamics for grand challenges: The
ESDMA approach. Systems Research & Behavioral Science. In Pre
‘SS.
Kwakkel JH, Pruyt E. 2013b. Exploratory modeling and analy:
for model-based foresight under deep uncertainty. Technologic
& Social Change 80(3): 419-431.
an approach
Forecasting
Lane DC. 2000. Diagramming conventions in system dynamics. J Opl Res Soc
51(2): 241-245.
17
Lane DC. 2010. Participative modelling and big issues: defining features of
system dynamics. Systems Research and Behavioral Science 27: 461-465.
Lempert RJ, Groves DG, Popper SW, Bankes SC. 2006. A general, analytic
method for generating robust strategies and narrative scenarios. Management
Science 52(4): 514-528.
Lempert RJ, Popper SW, Bankes $C. 2003. Shaping the next one hundred years:
New methods for quantitative, long-term policy analysis. RAND report MR-
1626, The RAND Pardee Center, Santa Monica, CA.
Lempert RJ, Schlesinger ME. 2000. Robust strategies for abating climate
change. Climatic Change 45(3-4): 387-401.
Logtens T, Pruyt E, Gijsbers G. 2012. Socictal aging in the Netherlands: Ex-
ploratory s ics modeling and anal lings of the 30th
International Conference of the System Dynamics Society. St.-Gallen, CH,
System Dynamics Society.
McKay MD, Beckman RJ, Conover WJ. 2000. A comparison of three methods
for selecting values of input variables in the analysis of output from a computer
code. Technometrics 42(1): 55-61.
Meadows DH. 1980. The Unavoidable A Priori. In Randers J. (ed.) Elements of
the System Dynamics Method, Productivity Press, Cambridge MA, 23-57.
Meadows DH, Richardson J, Bruckmann G. 1982. Groping in the Dark: The
first decade of global modelling. Chichester: John Wiley and Sons.
Meadows DH, Robinson JM. 1985. The Electronic Oracle. Computer Models
and Social Decisions. John Wiley & Sons, Chichester.
Meadows DH, Meadows DL, Randers J. 1992. Beyond the limits. Chelsea Green
Publishing Company, White River Junction, VT.
Miller J. 1998. Active nonlinear tests (ANTs) of complex simulation models.
Management Science 44: 820-830.
Oliva R. 2003. Model calibration as a testing strategy for system dynamics
models. European Journal of Operational Research 151(3): p552-568.
Oreskes
Confirmation of Numer:
641-647.
N, Shrader-Frechette K, Belitz K. 1994. Verification, Validation, and
al Models in the Earth Sciences. Science 263(5147):
Osgood N. 2009. Lightening the performance burden of individual-based models
through dimensional analysis and scale modeling. System Dynamics Review
25(): 101-134.
Petitjean FO, Ketterlin A, Gancarski P. 2011. A global averaging method for
dynamic time warping, with applications to clustering. Pattern Recognition
44(3): 678-693 doi:10.1016/j.patcog.2010.09.013
Pruyt E. 2013. Small system dynamics models for big issues: Triple jump to-
wards real-world complexity. TU Delft Library, Delft. Available at
|//simulation.tbm.tudelft.nl/|
Pruyt E. 2010. Scarcity of Minerals and Metals: A Generic Exploratory System
Dynamics Model. In Proceedings of the 28th International Conference of the
System Dynamics Society. St.-Gallen, CH, System Dynamies Society.
Pruyt E, Hamarat C. 2010. The Influenza A(HIN1)v Pandemic: An Exploratory
System Dynamics Approach. In Proceedings of the 28th International Confer-
ence of the System Dynamics Society. Scoul, Korea, System Dynamics Society.
Pruyt E, Hamarat C, Kwakkel JH. 2012. Integrated risk-capability analysis
under deep uncertainty: an integrated ESDMA approach. In Proceedings of
the 80th International Conference of the System Dynamics Society. St.-Gallen,
CH, System Dynamics Society.
Pruyt E, Kwakkel JH. (in press). Radicalization under Deep Uncertainty: A
Multi-Model Exploration of Activism, Extremism, and Terrorism. System Dy-
namics Review. In press.
Pruyt E, Kwakkel JH. 2012. A bright future for system dynamics: From art
to computational science and more. In Proceedings of the 30th International
Conference of the System Dynamics Society. St.-Gallen, CH, System Dynam-
ics Soc
Pruyt E, Kwakkel JH, Hamarat C. 2013. Doing more with models: Illustration of
a system dynamics approach for exploring deeply uncertain issues, analyzing
models, and designing adaptive robust policies. In Proceedings of the 31st
International Conference of the System Dynamics Society. Cambridge, MA,
em Dynamics Soc
Rakthanmanon T. 2013. Addressing Big Data Time : Mining Tril-
lions of Time Series Subsequences Under Dynamic Time Warping.
ACM Transactions on Knowledge Discovery from Data 7(3): 10:110:31.
doi:10.1145/2510000/2500489
Richardson GP. 2011. Reflections on the foundations of system dynamics. Sys-
tem Dynamics Review 27(3): 219-243.
Richardson GP, 2014. Drawing Insights from a Small Model of the Growth of an
‘Academic Field. Submitted for publication in: Proceedings of the 32nd Inter
national Conference of the System Dynamics Society, Delft, the Netherlands,
20-24 July 2014.
sis Perspective. Submitted for publication in: Proceedings of the 32nd
International Conference of the System Dynamics Society, Delft, the Nether-
lands, 20-24 July 2014.
Ruth M, Pieper F. 1994. Modeling spatial dynamics of sea-level rise in a coastal
area. System Dynamics Review 10(): 375-389.
Salch M, Oliva R, Kampmann CE, Davidsen PI. 2010. A comprehensive an-
alytical approach for policy analysis of system dynamics models. European
Journal of Operational Research 203(3): 673-683
19
Sterman JD. 1994. Learning in and about complex systems. System Dynamics
Review. 10(2-3): 291-330.
Sterman JD. 2000. Business Dynamics: Systems Thinking and Modeling for a
Complex World. Irwin/McGraw-Hill: New York.
Sterman JD. (ed.) 2007. Exploring the next great frontier: System dynamics at
fifty. System Dynamics Review 23(2-3): 89-93.
Struben J. 2005. Space matters too! Mutualistic dynamics between hydrogen
fuel cell vehicle demand and fucling infrastructure. In Proceedings of the 2005
nce of the System Dynamics Society. System Dynamics
Taylor TR Ford DN, Ford A. 2010. Improving model understanding using
statis ening. System Dynamics Review 26: 73-87.
Van Rossum G. 1995. Python Reference Manual. CWI, Amsterdam.
Vennix JAM. 1996. Group Model Building. Facilitating Team Learning bu Using
System Dynamics. John Wiley and Sons: Chichester.
Ventana Systems Inc. 2010. Vensim DSS Reference Supplement. Ventana System
Inc.
Ventana Systems Inc. 2011. Vensim Reference Manual. Ventana System Inc.
Walker WE, Marchau VAWJ, Kwakkel JH. 2013. Uncertainty in the frame-
work of policy analysis. In: Thissen WAH, Walker WE. (Eds.) Public policy
analysis: New developments. Springer Science + Business Media, New York.
215-261.
Walker WE, Rahman SA, Cave J. 2001. Adaptive policies, policy anal
policymaking. European Journal of Operational Research 128(2): 282~ a
Weaver CP, Lempert RJ, Brown C, Hall JA, Revell D, Sarewitz D. 2013. Im-
proving the contribution of climate model information to decision making:
the value and demands of robust decision frameworks. Wiley Interdisciplinary
Reviews: Climate Change 4(1): 39-60.
Wils A, Kamiya M, Choucri N. 1998. Threats to sustainability: simulating
conflict within and between nations. System Dynamics Review 14(2-3): 129-
162.
Yiicel G. 2012. A novel way to measure (dis)similarity between model behaviors
based on dynamic pattern features. In: Proceedings of the 30th International
Conference of the System Dynamics Society, St. Gallen, Switzerland, 22 July—
26 July 2012.
Yiicel G, Barlas Y. 2011. Automated parameter specification in dynamic feed-
back models based on behavior pattern features. System Dynamics Review
27(2): 195-215.
