Volume 23 – Number 4 October 2010
Session Reports (1)
SUN, PhD Colloquium
More than seventy participants of the ICSD 2010 came together on Sunday, July 25, to attend the 11th PhD Colloquium of the Student Chapter of the System Dynamics Society. The colloquium was organized by Katherine Dykes and Burcu Tan. Sponsorship this year was by the IBM Corporation that provided the event luncheon, refreshments and coffee breaks. The goal of the colloquium, as in previous years, was to provide a forum for students to present their work in progress at various stages. Four students presented work during plenary sessions and nine students presented work during the poster session at the end of the day. In addition, the colloquium featured three keynote presentations with a theme this year surrounding the topic of energy and sustainability modeling across continents. The featured speakers included Professor Erling Moxnes of the University of Bergen in Norway, Professor Namsung Ahn of Solbridge International School of Business in South Korea, and Professor Andrew Ford of Washington State University in the US. In addition, a luncheon talk was provided by Sponsor Representative, Dr. Ching-Hua Chen-Ritzo of IBM Corporation Research. All of the talks from students and invited speakers were well received. The link below will provide an overview and some of the highlights of the different presentations.
MON 11:00 AM Parallel, Model Building–Structure and Data
This Monday morning parallel session provided an in-depth view on current issues regarding the understanding of structure-behaviour linkages. It was chaired by Etiënne Rouwette and consisted of three thought-provoking papers: The Concerted Run on the DSB Bank: An Exploratory System Dynamics Approach by Pruyt and Hamarat (both presenting); Integrating System Dynamics and Enterprise Modeling to Address Dynamic and Structural Complexities of Choice Situations by Golnam, van Ackeren and Wegmann; and Sensitivity Analysis of Oscillatory System Dynamics Models by Hekimoglu. On average, more than thirty persons participated in the session that took place in the "Seoul" room on the second floor.
In the first presentation it was explained how "exploratory" system dynamics models can be used to address not only complex issues but deeply uncertain issues. The two authors from Delft University used a rapid-modeling approach to generate a system dynamics representation of the DSB bankruptcy that happened in the Netherlands last year. This "exploratory" model was subsequently analyzed using a methodology based on a large number of scenario runs. The second presentation described differences and similarities between enterprise modeling and system dynamics modeling; authors stem from different institutions in Lausanne, Switzerland. In particular, the comparison of the visual languages of the two modeling approaches highlighted some interesting points for cross-fertilization. The final paper of this session dealt with the topic of computer-based sensitivity analyses to determine parameterization of oscillating systems and is based on previous work at Bogazici University in Istanbul. In the presentation, it was described how important parameters of such a system can be approximated using regression analysis.
The Monday morning parallel session entitled "Planning for Energy and Business Opportunities" was chaired by Namsung Ahn. This session consisted of three presentations: 1) An Investigation on The Process of Modification of Subsidy Policy, by Reza Kazemi, Masoud Tavazoei, Nikrooz Nasr, Ali Mashayekhi 2) Policy Analysis for the Transformation of Switzerland's Stock of Buildings. A Small Model Approach, by Matthias Mueller, Silvia Ulli-Beer and 3) Inductive Modelling of an Entrepreneurial System, by Michael Yearworth.
The first fifteen minute talk started with Reza Kazemi discussing the dilemmas of subsidy in Iran. First, he briefly described different aspects of subsidy. Removing subsidy and its consequences was the next part of his speech. There are both long and short term effects on removing subsidy which make it a hard problem to solve. Next, he described a comprehensive model of the subsidy problem. At the end, he proposed some policies to remove subsidy in a way that causes the fewest difficulties to society. One question was asked about the oil price in Iran.
Matthias Mueller, as second presenter, talked about Policy Analysis for the Transformation of Switzerland's Stock of Buildings. He spoke on behalf on his professor who worked with him on this project. Around fifty persons took part in this session. Matthias walked around and described importance of buildings in Swiss climate and energy policy. He simulated three different scenarios in energy efficiency policies and analyzed different emission patterns. When Matthias finished there were two questions; one by Prof. Namsung Ahn and one from a member of the audience.
Finally, Michael Yearworth discussed Inductive Modelling of an Entrepreneurial System. European data was the source of his study. He started with lots of information about this issue to make it clear for the audience. He presented a very surprising model with lots of lines and relations. It was really amazing. Two questions were made about his work. They were mainly about the concept of the problem.
This issue was one of the most essential of the conference. All three papers were very well prepared.Mohammad Askari Tabatabaei
MON 11:00 AM Parallel, Public Debt Issues
Speakers: Domen Zavrl, University of Maribor, Slovenia
In light of the recent global financial crisis (and the use of public funds to bail out banks) I expected this session would be interesting. I was not disappointed. Of the three papers presented, the first dealt directly with the dynamics of the Euro Zone crisis. The other two papers took a rather different slant of public debt by examining poverty, debt accumulation, and public indebtedness in sub-Saharan Africa. A review of each paper follows.
An Institutional Dynamics Model of the Euro Zone Crisis, Doman Zavrl
Doman began by reviewing the puzzle of the financial crisis in Europe whose severity has surprised both political leaders and academics. Citing Greece as the epicenter, he presented time series data on the price of credit default swaps that told a compelling tale of the unfolding crisis.
The ambitious aim of this work was to question the relevance of present economic theory (which has so clearly failed to anticipate the looming debt problem) and to offer an alternative. Doman proposed a behavioral approach that combines system dynamics with institutional economics to shed more light on the 'animal spirits' that haunt market-based industrial economies. He presented a carefully crafted model, drawing on classic behavioral formulations in system dynamics.
The model has the capability to mimic aspects of human institutional decisionmaking that lead to asset bubbles and to trace the economic and political processes that transform mounting debt in the banking sector into public debt. He used the model to develop credible scenarios of the Euro Zone crisis and to test bailout options. In the brief time available an inviting glimpse was given of the underlying model and its dynamics. There was a lively audience discussion at the end that touched on policy conclusions and the challenges of communicating the model to policymakers and economists.
Modeling the Dynamics of Poverty Traps, John Ansah
John began by outlining a dynamic problem of chronic underperformance in sub-Saharan African economies and various theories (some exogenous and others endogenous) that scholars have advanced to explain the phenomenon. He then presented a concise system dynamics model linking five main stock accumulations: population, capital, labor productivity, public debt and accrued interest. He interpreted the conceptual model and considered how it might be related to existing theories of underperformance. I like such bold and concise models, and it seemed to me that this one was carefully formulated. John also mentioned a much larger disaggregated model from which the concise model had been distilled.
The presentation concluded with a review and discussion of policies to avoid poverty traps in African economies. Members of the audience also raised formulation questions about male/female disaggregation and productivity modeling.
The Dynamics of Public Indebtedness in Ghana, John Ansah
The third and final paper was also delivered by John Ansah. He continued his Africa modeling theme, but now switched his attention to debt relief policy for African economies using Ghana as a case. I found this topic to be an intriguing complement to Doman Zavrl's opening talk on the Euro Zone crisis; both involve debt transfer, though in much different socio-economic contexts.
John began by describing the so-called '2001 HIPC Initiative' intended to bring debt relief to highly indebted poor countries. In 2002 Ghana qualified. There has been much political debate about the merits of the initiative. Starting from two reference modes John developed a system dynamics model to test Ghana's program by examining the interplay between producers, households, public debt, taxation and spending. Among the model's findings was the insight that ineffective tax collection is a major contributor to rising debt. The presentation generated a useful discussion, including questions about the focus on fiscal policy and the apparent lack of monetary policy.
The dynamics of public debt affect all of us and have confounded policymakers and their advisors. The three papers in this session showed there is an important role for well-formulated system dynamics models in understanding debt accumulation and managing public debt–whether for advanced or developing economies. The session was expertly chaired by Carmine Bianchi.
MON 11:00 AM Parallel, Qualitative System Dynamics and Public Policy
The session began with a presentation by Joe Hsueh, What Can Formal Modeling Add to Qualitative Systems Thinking? A Simulation Model of the Shift the Burden System Archetype. The paper demonstrated why researchers should strive to build operational models and not rely solely on qualitative models. Formal modeling leads to deeper insights, and helps to avoid the simplistic and potentially incorrect assumption that when faced with a shift the burden situation, one should focus only on addressing the fundamental problems and not the symptoms. The lowest cost solution, which avoids an unfavorable tipping point, involves a mixed strategy that allocates some resources to address symptoms and rest to addressing fundamentals. This solution also reduces oscillations. [see full report]
The next paper, by Thanh Mai and Kambiz Maani, Systems thinking for sustainable tourism in the Cat Ba Biosphere Reserve of Viet Nam, uses a largely qualitative model to study a complex natural ecosystem where tourism is exploding but skilled resources, water, electricity, infrastructure, and recreational facilities are scarce. Systems thinking is used to determine key factors and the likely dynamics that would balance economics, ecology, and poverty relief. An iceberg analogy was used to help uncover mental models and better understand how the local people think. The team did twenty-six interviews involving over seventy participants to validate a preliminary model and to determine leverage points. [see full report]The final paper, by Jie Jiang, Jun Li, and Xu Honggang, System Dynamics Model for Transportation Infrastructure Investment and Cultural Heritage Tourism Development, answers the questions, how does transportation infrastructure affect tourist choice, and how do tourist choices affect tourism investment? Tourist groups decide based more on time, whereas individual tourists focus more on cultural variety. A case study applies a system dynamics model with an embedded logit model to analyze the impacts of transportation infrastructure investment on the tourism development for two cities in China. Insights include: more visits to both cities are likely, especially by group tourists, and the cities are likely to shift from competition to cooperation in the future. [see full report]
Motivation: Is it risky to work only with qualitative models/archetypes?
Situation: Started with the classic Shift the Burden CLD (ala Senge) and created an operational model where outsourcing is used to address "symptoms," and training is used to address fundamental problem.
Research questions: 1) Can formal modeling reproduce different shifting the burden behavior modes? [Answer: yes]; 2) What does formal modeling add? [Answer: deeper insights, and helps to avoid simplistic and potentially incorrect assumption that one should always and only address the fundamental problems and not the symptoms]
Test One: Neutralize all but the top loop.
Insight One: Without a fundamental solution option available, the "treat the symptom" solution is rational.
Test Two: Add simple cost structure to reflect the benefits of a speedy solution (via treating symptoms) versus a slower solution that addresses the fundamental problem.
Insight Two: Lowest cost solution involves mixed strategy that allocates some resources to address symptoms and rest to addressing fundamentals. A tipping point results with over-reliance on treating symptoms (higher costs in long run).
Insight Three: Use of fundamental solution can introduce oscillations!
Insight Four: Blending in symptomatic solution reduces both oscillation and cost; and there is a "sweet spot."
Audience Question One: Could you use with policy-makers? Good idea for future work!
Audience Question Two: Could model advise manager who starts with outsourcing to detect tipping point in time to avoid adverse behavior? Great idea; maybe should always have some training as a buffer.
Introduces the use of systems thinking (ST) and modeling [qualitatively, not empirically based] to anticipate the impact of tourism in Cat Ba (a designated island world biosphere reserve; not a park). Cat Ba is a very complex natural ecosystem where more than 2000 NGOs are operating to protect endangered species while tourism (mostly domestic) is exploding. Is this good or bad? It is certainly a severe challenge for sustainable tourism.
Research Question: Can ST help to determine factors and likely dynamics related to sustainable tourism that balances economics, ecology, and poverty relief?
Situation: Tourism is up, the environment is stressed, and poverty remains high. Skilled resources are limited. Water and electricity are scarce. Infrastructure and recreational facilities are limited. The problems are largely social, however, not engineering challenges. How to understand?
Solution: The team started with the iceberg analogy (from Senge) to help uncover mental models and better understand how the local people think. All current prevention effort is being applied at tip of iceberg.
The leader of the island is now using ST on his own, including giving lectures to over 600 people! The research team used CLDs and stock and flow diagrams (SFDs) to help understand the nonlinear feedback loops over a ten-twenty year time frame.
The team spent months conducting twenty-six interviews involving more than seventy participants to validate preliminary model and to determine leverage points. Dozens of issues (such as the need to educate women) were uncovered.
Presenter showed several SFDs and mentioned that a flight simulator was created; also noted a dilemma at the end regarding how to best model an attractiveness index. [out of time]
Questions: 1) How does transportation infrastructure affect tourist choice, 2) How do tourist choices affect tourism investment?
There are two types of tourists: group and individual. For a group, time affects choice, whereas for an individual, cultural value impacts choice.
A logit model was created for impact of touring time and cultural diversity on choice. Parameters estimated with MLE.
Case study applies the SD model with logit model embedded to analyze the impacts of transportation infrastructure investment on the tourism development of Xidi and Hongcun, World Heritage Villages in southern Anhui province, China.
One village had more visitors, but building a new road shifted many tourists to the other city, providing the data for estimating logit model parameters.
Then, as tourism shifted, investment strategy also changed, serving as an overall balancing feedback loop described by the SD model.
The SD model of time behavior matches the actual situation, thereby indicating that the model can be used to help plan future road-building investments.
Insights: More visits to both cities are likely, mostly by group tourists (as time to reach each city decreases). Cities are likely to shift from competition to cooperation for best results.
MON 2:00 PM Parallel–Fresh Water Management and Policies
This session with only one paper was chaired by Seetharam Kallidaikurichi. The system dynamic study of regional development of river basin under the constraints of water resources was presented by Shanshan Dai from Sun Yet-san University, China. The paper looked at the ever-worsening water scarcity issues in the Manas River Basin, an arid area undergoing rapid urbanization. According to the author, the problem of water stress within the region was mainly in a balancing feedback loop with exploitation (water supply) and a reinforcing feedback loop with allocation (water consumption). A hydrological-social-economic system was used to describe the development process. Policies based on both supply and demand side were tested to provide insights on future water resources management in arid regions. Discussions touched upon the special geographical and political situations related with water resources management in the Manas River Basin and prospects of this SD model's applications to other cases with different physical and socio-economic characteristics.
Fig.1 The Manas River Basin in Xinjiang, China
The author studied water stress/scarcity in two feedback loops: a balancing one with exploitation (water supply) and a reinforcing one with allocation (water consumption) (Fig.2). Based on this framework, a conceptual model was presented (Fig. 3). A hydrology-social-economic system was taken to describe and analyze the processes. Three sectors were considered on the supply side: natural surface water and groundwater, reservoirs and canals, and pumping capacity. The consumption side included three sectors: irrigation, industrial use and urban use. The SD models of groundwater, pumping water capacity and land sector were presented. An index called perceived water stress was defined by the author to quantify water sustainability of the region: Perceived water stress = water demand / available water. With a one-year time step, data from 1949 to 2005 were used for model calibration. A forty-four year simulation was conducted from 2007 to 2050.
Fig.2 Main Feedback Loops
Fig.3 Conceptual Model
Simulation results demonstrated that when technology enhancement reached its bottleneck, the increase of cultivated land would level off or decrease due to lack of water resources. Major policy implications proposed included: 1) industry adjustment was crucial, which would affect the availability of cultivated land; 2) feedback between groundwater level and pumping would decide the stability of water table and groundwater sustainability in the future.
Two questions were discussed after the presentation:
Q1: Can this model be successfully transferred to other studies on water resources management?
Q2: Who will decide water allocation among different sectors within the study area?
MON 2:00 PM Parallel, Pedagogy
Chaired by Michael Bean
In this session, the first speaker was Martin Schaffernicht. He presented a co-authored paper named What is learned in system dynamics education: a competency-based representation based upon Bloom's taxonomy. The second speaker was Erik Pruyt. He presented two papers; a single author paper entitled Making System Dynamics Cool II: New Hot Teaching and Testing Cases of Increasing Complexity and a co-authored paper named Essential Skills for System Dynamics Practitioners–A Delft University of Technology Perspective.
In his talk, Martin Schaffernicht emphasized "the need to integrate system dynamics into a competency-based curriculum." Martin and his colleague developed a tool, which would help to design a system dynamics (SD) course, based on Bloom's taxonomy. This work will be helpful integrating SD in the curricula of different professions. Nice work! Thanks Martin.
Erik Pruyt's first presentation was based on his single-author paper. In this work, he reported modeling cases developed for an introductory level SD course at Delft University of Technology. These cases serve as a bridge by helping students to overcome the gap between theory and actual model building. I believe Erik's approach will improve understanding of the students and their gain from introductory level SD courses. This is a critical work for all who teach system dynamics. Thanks Erik.
In his second talk, Erik Pruyt focused on the skill set necessary for SD practitioners. Erik and his colleagues made the initial step in defining SD modeling skills. Interesting work! Thanks Erik.
MON 2:00 PM Parallel–Understanding Model Behavior
Chaired by Imrana Umar, this session included three papers. The session was attended by nine conference delegates. Jinjing Huang made the first and third presentations, while Mustafa Hekimoglu made the second. All three presentations attracted probing and clarification questions from interested members of the audience.
An Eigenvector Approach for Analysing Linear Feedback System
by Jinjing Huang, Enda Howley, Jim Duggan
This presentation was impressive with its highly technical content. The speaker, Jinjing Huang, started off with a mathematical discussion of the problems associated with the conventional eigensolution. Jinjing then went on to explain that through their study of the eigenvector sensitivity, she and her team members had observed that the right and left eigenvector sensitivities associated with the same mode could not be evaluated separately. She then presented an analytical approach to the eigenvector-related sensitivity computation, i.e., a linear combination of the right and left eigenvector sensitivity. In response to a question from the audience, Jinjing confirmed that this can only be used to apply to linear systems, with only local sensitivities involved.
Sensitivity Analysis of System Dynamics Models by Behavior Pattern Measures
by Mustafa Hekimoglu, Yaman Barlas
In his well-paced and substantive presentation, Mustafa provided a definition of sensitivity analysis as an introduction to his presentation. He argued that since System Dynamics (SD) is a behavior-oriented simulation discipline, sensitivity analysis of SD should be evaluated in order to explore the effect of parameter uncertainty on the behavior pattern of the simulation.
He proceeded to highlight the procedure presented in his paper for a pattern sensitivity analysis of system dynamics models. He went on to describe the behavior pattern sensitivity methodology that was used. He then proceeded to justify his team's use of the regression method for statistical analysis (SA), and went on to show how the SA approach could be applied to: The Project Management Model (Taylor and Ford, 2006), Oscillatory SD models and the Simple Supply Line Model (Sterman, 2000). The stock-flow models were shown to the audience, and the parameters are listed. Output graphs were then shown and explained, and the results of regression and the sensitivity results revealed. The overall results for each of the models were then shared.
The results indicate that pattern measures of output behavior are appropriate in analyzing the parameter sensitivity of system dynamics models.
An Extension of Loop Deactivation in the Behavioral Method by Jinjing Huang, Enda Howley, Jim Duggan
Jinjing Huang shared the loop deactivation method presented in the paper that she and her co-authors wrote. She explained that the advantage of this method is that it can be used even where neither the control variable nor the unique edge is identified. This addresses the drawback in behavioral methods in which a loop would first have to be deactivated by fixing its control variable or a unique edge. In the loop deactivation method, a loop is deactivated by modifying its unique consecutive two edges which are able to distinguish this loop from other loops. The long wave model is used to demonstrate the loop deactivation approach and compare the analysis result with other dominant loop identification methods.Ong Ling Lee Elizabeth
Limits to Growth and Industrialisation–Insights from small and metaphorical system dynamics models and simulators by John Morecroft
Models may be used in a metaphorical sense. There is a continuum in model fidelity, ranging from low to high. An example of a model with low fidelity is the Romeo and Juliet model, which describes how the attraction of the two lovers for each other fluctuates over time. An example of a model with high fidelity is a validated and tested model of an airplane. Prof. Morecroft presented a model of a fishing industry. Model runs showed an increasing number of ships on sea which eventually reduced the fishing stock and thereby greatly diminished the rate of fish reproduction. Following the reductions in catch, the ships on sea fell again. The fishing model is used as a metaphor for industrial growth. Industrial growth and its relation to population growth are described by the World Dynamics Model developed by Forrester. The dynamics of the fishing and the World model are similar. The World Dynamics Model was not intended to be a high fidelity model. It intended to show general behaviour of an industrial society. The last run shown indicated what is needed to create a sustainable industrial society: reductions in capital investment, birth rate, pollution generation, natural resource usage and food production that lead to world equilibrium at a high quality of life.
A Simulation System for HPAI Diseases Control by Kwak, Kwak and Lee
HPAI stands for highly pathogenic avian influenza, commonly known as bird flu. If bird flu is found on a farm, all birds are killed. The impact on animals and on the economic and social circumstances in villages concerned is high. The HPAI project has three components: database, transfer to on-line tasks and simulation. Dr. Kwak showed a causal loop diagram explaining the number of birds infected. A second model showed the possibility of HPAI occurring. The model is used to track how HPAI came to a specific farm. Output screens showed a map of the region and the possible locations in which HPAI started. A diffusion model is used after this retrospective analysis, to estimate the spread speed of the disease. The model included a number of human and nonhuman factors that facilitate contamination and further spread. The models enabled the analysis of policies to slow or stop further spread, including circles around farms reported to be contaminated into which no transport is allowed. Finally, a social impact evaluation determined social costs depending on the previous diffusion data. Costs included, among others, farm impact, food consumption damage and retailer damage. Future improvements to the models include addition of missing data on farms and migration birds and improving the connection with weather systems.
This parallel session was dedicated to three different articles about agent-based modeling, dynamic control and system thinking in health care setting.
The first article, Using System Dynamics and Agent-based modeling to simulate a Proxy War for Resources, was presented by Ignacio Martinez-Moyano. This article was based on modeling the combat in Tajikistan by using agent based modeling. They modeled each of opponents as an agent, awarded by having access to the resources. They also provided a 2*2 matrix, to capture the probability of interactive actions. By modeling two stocks as capital (the amount of money dedicated to each party) and faction (the reservoir for arms) they modeled the combat as a path-dependent process.
The second presentation, Controlling the Direction of a Model Helicopter, was given by Oylum Seker. She started with a brief description of how a helicopter works. Then she presented the main model which simulates the motion of a helicopter based on decisions made by a naïve decision–maker, and then an expert. The decisions of the expert were provided by a heuristic model which was invented by Hakan Yasarcan, the second author. The main strength of this heuristic was its reliance on the past events, saved on a stock and used to make corrections in the future decisions. By comparing the results, this model aimed to prove the validity of the heuristic, and test it in a dynamic environment.
The third presentation was dedicated to a healthcare model for exhibiting the delayed discharges in UK hospitals. Sangeeta Sardiwal, presenting the main idea of her thesis on the article A Systems Thinking Approach To Investigating Delayed Discharges In The UK, insisted on the importance of systems thinking in approaching complex problems. She believed that more informed decision-making caused by integrating information systems about patient discharges could help to resolve problems, which are mainly unstructured and sophisticated. She proposed a combination of soft systems thinking methodology and system dynamics in order to approach this problem. The main part of her presentation was dedicated to the patients' flow in the hospital, which illustrated the mental model of decision-makers.
The session was opened 11:00 a.m. in the Berlin Room and was chaired by Elizabeth Ong. The first presenter was Brigit Kopainsky, the second speaker was Camilo Olaya, and the third was Eliot Rich.
Dynamics of Enforcement and Infringement of Intellectual Property Rights
This paper was about agricultural development in African countries. Issues such as food security and access as improving income and yield were discussed. Brigit said that one of the reasons why they were interested in Intellectual Property (IP) rights is that it is one of the big issues in agriculture in a developing country.
The presentation then progressed with a model which was like a scenario. First, Brigit showed the 'seed value chain and IP' in which 'plant breeding' and 'commercial seed production' interact with each other. She went on to explain how 'commercial seed production' interacts with 'seed distribution and sales.' These aspects put together with the Intellectual Property loop provided an overview of the presentation. Then there were explanations of how the enforcement loop of IP fosters innovation, propensity to file for protection, propensity to infringe on protection, and propensity to enforce protection. After that she suggested several models.
Theorizing about Crime: Elements for a Contribution of System Dynamics to Criminology
This paper was about crime in Colombia. The presentation was largely divided into three parts. The first part was concerned with Positivist theories or the objective causes of crime. The second part of the presentation concerned Rational Choice and last was the Social Learning theory of crime. After that, using system dynamics theory, the speaker, Camilo Olaya, deduced loops about the feedback loop. In addition, he said system dynamics made them think differently.
Refinement of A Supply and Demand Model for Vulnerability Black Markets
This paper presented an economic rationale for the existence and continuity of Vulnerability Black Markets (VBM). The presentation consisted of five parts. First, Eliot explained the background and purpose of this project and then he used a system dynamics model of markets for crime. As a result, they got two policies. One was market disruption that made the stock of exploits in black market to drop to zero and other one was enforcement. The conclusions were that unless the disruptions toward VBM forums were strong enough, sporadic market closures were followed by re-opening and that system dynamics modeling provides insight into behavior and apparent discontinuities in markets etc.
Chaired by Souleymane Bah and attended by twenty-five individuals, the Growth and Development parallel session examined ecological economics, fiscal policy, and the growth of economies of varying sizes.
Dolores Soto-Torres Dynamics in economic growth: A Perspective from System Dynamics
Dolores Soto-Torres presented the results of a simulation exercise of an economic system dynamics model to demonstrate the growth patterns of economies of varying sizes. The model that was presented considered the impact of the decisions of the consumers, firms, and government in an economy on the rate of growth of that economy. The simulation also took into consideration various governmental strategies as well as boom and bust cycles in order to illustrate the different paths of growth. It was suggested that simulations such as the one discussed during this presentation could be used to update future governmental policy and strategies that focus on the distribution of wealth, management of corruption, and educational policy.
All three presenters succinctly demonstrated the value of implementing a system dynamics model to further illustrate various aspects of economic growth and development.
Modelling the Effect of Allocation of Social Worker Duties on Delayed Discharges using System Dynamics in the U.K., by Sangeeta Sardiwal
This paper analyzed the effect of two different allocation policies for social workers and any impact an integrated information system between healthcare and social services may have on reducing delayed discharges in the U.K. Sangeeta used a combination of system dynamics and soft systems methodology to help create a better understanding of the qualitative data collected from interviews and conceptualize the problem at hand.
The simulation found that priority coping policies were effective in reducing delayed discharge in comparison to proportional coping policies when spare resource capacity was available, but became ineffective when there was no spare capacity. In addition, priority policies were found to be very disruptive to social workers' schedules. The alternative policy of using an integrated information system between health and social care to facilitate informed decision making seemed to be the most optimal solution of reducing delayed discharges.
Integrated Healthcare Delivery and Health Insurance Models for Studying Emergency Department Utilization, by Lianjun An, Ching-Hua Chen-Ritzo (Presenter)
In this paper, the authors analyzed the healthcare problem in the U.S. of emergency department overcrowding and the relationships between the demand for emergency department services, primary care services, the number of insured/uninsured people and the cost of health insurance premiums.
Ching-Hua described the three main components of the model which consisted of insurance choice, insurance costs and premiums and healthcare delivery. The three main leverage points which were used to analyze the model consisted of an insurance premium subsidy by the government, increasing the safety net primary care capacity for the uninsured and reducing the unemployment rate. All three levers had the effect of reducing the demand on emergency department services by affecting various factors in the model, ranging from increasing the number of insured people and decreasing the demand on emergency department services by a higher utilization of primary care.
Understanding the dynamics of planning and providing accommodation services for people with Intellectual Disabilities, by Lynette Lee, Mark Heffernan (Presenter), Geoff McDonnell
The objective of this paper was to analyze the demand in accommodation services for people with intellectual disability, based on several factors including an ageing and increasing population and potential shifts in policies determining how care is provided. The work in this paper took around four years to carry out, with some difficulties faced in gathering coherent data due to frequent shifts in the eligibility criteria and definition of what constituted an intellectual disability.
The model found that the number of people declared intellectually disabled will continue to rise for at least thirty years, with a considerable increase in the number of middle aged adults with mild intellectual disability who will require out of home accommodation support. The model predicts a need for a rise of 100% over the current provision of disability care by 2030. If those who do not need registered nurse care from aged care facilities are moved to disability care, the model predicts a rise of 120% over the current provision by 2030. With a ten year lead time for the set up of group homes, immediate policy settings will have considerable implications for the disability care system.
Chaired by Elisabeth Andersson
The first presentation of the session delivered by Shu-Jung Yang was on a piece of research work carried out by the presenter and his research student Yan Emma Liu. The presentation started with discussion on the craft, mass and Toyota production systems. The concept of paradox in production process was highlighted in terms of priority on "imitation" or "improvement" for maintaining competitiveness in the market. A system dynamics model was presented for sharing the understanding of competitive dynamics of a business firm. The model depicts the interaction between the domains of awareness, motivation and capability. The domain of awareness includes external competitive tension whereas the domain of motivation includes internal competitive tension. The capability domain on the other hand covers the dynamics of resource base to cater for the external and internal tensions. The insight gained with experimentation of the model indicated that in order to maintain a sustained competitive process, a firm should consistently invest in explorative activities for internal process improvements. External competitive tension would be vital in managing internal tension for learning and improvements.
The presentation took the given twenty minutes, and there was no question/answer session during the presentation.
The subject area of the second presentation was similar to the first in the sense that both the presentations focused into the process improvement dynamics at the firm level in manufacturing.
This second presentation however emphasized the issue of managing process improvement with the conflictive dual pressure of the need for production and need for improvement. General observation of premature demise of process improvement was one of the concerns, and investigation into how to increase throughput out of improved process design was the main focus. An important assumption was that there would be a single and constrained resource basket available for both the production and improvement activities. Another important aspect of the presented simulation model was that workers would learn during the improvement activities and it would increase their productivity as they gain experience in such activities. The simulation results revealed the existence of tipping point that distinguishes enduring high levels of production from outcomes with modest or no improvement. One of the important counter-intuitive findings was that slowing down the rate of starting improvement projects leads to better performance. It was also found that rather than focusing on achieving the highest possible output or rate of process improvement, managers should focus on building experience to get past the tipping point. It was emphasized that managers should develop and monitor signals or specific metrics that can bring better visibility of the state and rate of change of the important stock of experience. Overemphasis on the core improvement process ignoring the aspect of learning would develop poor result in managing the trade-off between production and improvement.
The presentation was well contained in the given fifteen minutes. Two major questions were asked.
Question 1: Are the goals of the new process improvement missing? The workers might not know why they are working differently. This might create the resistance to change.
Question 2: But why should the workers do different works? How one can know about the project start rate?
Better Healthcare through Better Design of Healthcare Research and Development, by Peter Hovmand
Peter Hovmand, as the third presenter of the session, shared the outcome of his research on R&D in healthcare sector. He focused into the management of R&D activities of a government funded research center. The complexity of the healthcare R&D organization was highlighted. The presented simulation model included interactions between R&D projects, project funding, human resources, research quality and space for research. The model depicts that adequate human resources and superior R&D quality would attract more project funding which in turn supports the human resources and R&D quality. So a well-managed R&D organization develops the virtuous effect that delivers high quality research output whereas a poorly managed organization might get trapped into vicious cycle that delivers inferior research output which would be much more critical in the health sector.
The presentation was completed in the given fifteen minutes. One main question was asked.
Question: Are the goals for the people who work in R&D projects included in the model?
The System Dynamics Newsletter is published four times a year by the System Dynamics Society.