Microfinance institutions (MFIs) provide credit, savings, and other financial services to the poor and must successfully manage large volumes of small transactions. SymBanc is a system dynamics simulator designed to introduce students to the complexities of managing a Microfinance Institution (MFI) or to engage experienced practitioners in a discussion of the key determinants of success in such a dynamic environment. The simulator allows students and practitioners to grow an MFI from a single branch to a large network by making a variety of decisions about target market, staffing and facilities, loan and savings product design, and sources of external funding. Detailed feedback enables them to fine-tune their strategies during a simulation. This paper begins with some background on Microfinance Institutions and then presents the structure of the model underlying SymBanc and results of typical simulations. Initial experience using SymBanc and future enhancements contemplated for it are also described.
We were at the Lyons Pub. Peterson walked in with a presentable young man.
This is Randy, she said, pulling up a chair. Hes just back from Egypt. Ordering a beer, Peterson fished a photo from her pocket. Doesnt Randy look grand in front of the Sphinx?
Sedgewick turned to the young man Tell us about your trip.
Randy smiled in recollection. I stayed at Le Meridien in Giza. Costs a bloody fortune, but its worth it. Has a swim-in bar, dont you know.
And the Sphinx? Sedgewick prompted.
Randys expression turned weighty. Big. Damn big.
So youve never actually been to Egypt. Sedgewick said sadly. And, never seen the Sphinx.
Like Randy, many of us return from a model without true insight. In this workshop well deeply explore a model or two. You will see how eigenanalysis complements and speeds traditional approaches to understanding models. Math-phobics and math-lovers are equally welcome.
This report builds on a previous epidemiological model of a pneumonic plague outbreak that incorporated three behavioral responses as exogenous drivers and evaluated their importance in allowing us to replicate the actual outbreak (Heinbokel& Potash, ISDC-2003). The current paper describes our subsequent efforts to incorporate those critical and controlling behavioral dimensions into this model as critical feedback loops. We conceptually deconstructed the event into four segments: becoming aware of the outbreak, deciding to act in response, choosing a specific response, and returning to normal behavior. We utilized current psychological theories, such as the Psychometric Paradigm and Brunswiks Lens Model, to build small, conceptually clear, transferable, and combinable behavioral submodels to simulate the first three segments involving information and social networks, social trust, and risk perceptions. We believe these modeling efforts comprise first steps in a critical process of translating current, frequently static, risk theories to dynamically responsive vehicles that can be flexibly and quantitatively applied to reliably aid in understanding and influencing responses to such public health threats, other extreme events, and other dynamic risk scenarios in general.
The use of System Dynamic software tools are becoming a popular way of investigating complex problems. However, along with the use of these tools exists the risk of relying too heavily on the numerical part of the analysis and neglecting the preparation phase for analysis. Any modelling procedure in System Dynamic modelling goes through a conceptual phase that uses the Learning Loop approach. This phase is most often done unintentionally. Using the Learning Loop approach consciously facilitates the group modelling process to acquire four successive phases, i.e. Definition, Clarification, Confirmation and Implementation. This enables a clear structure in the process, from acquiring the task to documenting the results. Only by intentionally using the Learning Loop approach in a managed manner, can the full potential of the process be exploited. Qualitative analysis does not replace simulations with a computer model but simulations should serve as a continuation to reconfirm or refute qualitative hypothesis and a simulation should only occur when the mental model has been tested. Systems Analysis, including its thinking, analysis and dynamics, is not a method, but rather an adaptive learning behaviour. It is a behaviour that finds the optimally adapted method, applying at some times SD computer tools.
A simple approach was used to evaluate the potential human population that the pre-industrial Icelandic environment could sustain. A model was constructed that simulated the population size according to potential biological production available for livestock. Biological production was determined by the extent of the total poten-tial vegetation cover based on the Degree-day concept. Fluctuations in the mean an-nual temperature cause changes in the potential vegetation cover and as a conse-quence change the biological production sustaining livestock and ultimately human population. The simulation's results indicate that the potential population that the environment could sustain during the pre-industrial period fluctuated around 40-80 thousand. The results further indicate that the severe land degradation experienced after the settlement period had a marginal impact on the population size. The pre-historical population did however overshoot the natural sustainability on few occa-sions.
The workflow in System Dynamics may be seem to follow certain general flow patterns within the adaptive-iterative approach required. We constructed some diagrams to gain transparency and understanding of different tasks in the model-ling process. The modelling workflow involve systems analysis, group modelling and system dynamics. The systems analysis as executed during group modelling consists of three model building stages and one implementation task. The stages involves Definition, Clarification, Confirmation and Implementation. After defin-ing the issues and questions, the process evolves the Causal Loop Diagram (CLD) iteratively with the Stock and Flow Diagram (SFD), which form the construction drawings for the programming of the model incorporated as a System Dynamic Tool Diagram (SDTD). The third stage is the testing of the computerized model version created by the System Dynamic tool (SD-tool) and the fourth task is the implementation of the outputs into results and policies. This may be considered an iterative process in all the stages. Innovation is the emergent output from a proc-ess operating according to the learning loop: Finally the roadmap given to our students for going from question to model diagram is shown.
A new archetype, The Tyranny of Small Steps (TYST) has been observed. Explained through a system dynamics perspective, the archetypical behaviour TYST is an unwanted change to a system through a series of small activities that may be independent from one another. These activities are small enough not to be detected by the surveillance within the system, but significant enough to encroach upon the tolerance zone of the system and compromise the integrity of the system. TYST is an unintentional process that is experienced within the system and made possible by the lack of transparency between an overarching level and a local level where the encroachment is taking place. The Örby case study illustrates a real life manifestation of the TYST archetype in planning. The TYST illustrates the necessity for total transparency in any systems in order to avoid unintended consequence of the archetype. The TYST process may be regarded as a part of wide range of complex systems but depending on the conditions, it can remain dormant, and only become active when the conditions for lack of transparency are fulfilled.
Formal model analysis tools are essential elements in understanding how structure drives behavior. Conventional model analysis relies heavily on a time-consuming experimental iterative process. Current formal tools are not mature enough for application to most models. This paper presents a loop dominance analysis approach based on eigenvalue elasticity analysis (EEA). EEA, although a potentially strong formal model analysis tool, has drawn criticisms over the years for a number of reasons. The approach proposed in this study attempts to bring proper solutions to the issues raised by those criticisms. To this end, a ten-step procedure is proposed. Among the most prominent features of the proposed procedure is the ability to track the influences of feedback loops on a specific variable of interest. Others include the ability to track the loop dominance dynamics over time and an attempt to the codification of the proposed features of the EEA. The application of the proposed approach is demonstrated using a simple economic long wave model and two other models, all chosen from earlier methodological studies on formal loop dominance analysis. The results of these applications also facilitate the comparison of the proposed approach to other formal model analysis tools.
This paper investigates the dynamics of accumulation processes of strategic capabilities in manufacturing, i.e. cost, quality, time orientation and flexibility. The analysis is conducted with the help of an exploratory system dynamics model that represents a hierarchy of these accumulative capabilities. By applying a dynamic view, concepts from the operations management literature are tested and shortcomings are identified. In a further step, the exploratory model is parameterized with empirical data from a large international survey of manufacturing plants. Implications concern the distribution of managerial attention on the different capabilities and its dynamic consequences. The value of this paper lies in the insights gained by the transformation of a verbal model in a quantified simulation model and the learning resulting from simulation experiments.
The Methodology of System Dynamics claims to promote understanding of complex systems. Accepting this claim, the question Does experience or an education in System Dynamics help people to solve simple, dynamic problems? arises. It guides the conduction of our experiment. The first hypothesis about no influence of additional information for problem solving has to be accepted. The performances of two different information treatment groups are not significantly different. Our second hypothesis, that people with and without experience in System Dynamics will have the same performance, has to be rejected. A significant difference between the performances of experienced people and people with no or little experience exists. A possible reason for this circumstance is that an education in System Dynamics doesnt immediately, but over a longer time horizon, enables people to comprehend dynamic systems. At last, the experimental design will be discussed and several weaknesses will be pointed out.
Keywords: Experiment, Applicability of System Dynamics, Hypothesis Testing, Dynamic Problem, Education, Comprehension