Welcome to the 2008 Computational Economics Modelling Exhibit ...

Below are the final versions of projects submitted by students in ECC3855, October 2008. They were given only a few weeks to take an idea of their choosing to modelling reality, learning NetLogo along the way. If you have any questions about these models or wish to contact one of the authors, please contact Dr Simon Angus (simon.angus AT buseco.monash.edu.au).

NB: Code is made available for research and educational purposes. The authors of each project remain the sole copyright owners.

Quicklink to this page: http://tinyurl.com/ECC3855ProjectExhibit2008

The Models

Physico-Social Systems

Project Title

The Bit Torrent Network

Authors

Kashi Trathen and Rob Ball

About

This model simulates the dynamics of a peer-to-peer bit torrent network. Bit Torrent is a popular protocol for distributing large files amongst multiple users over the internet, and studies have shown it to account for approximately 70 percent of total IP traffic (Yinka, 2005). It works by breaking a file into many segments (usually 100 or more depending on the size of the file) and allows individual users who are downloading the file to simultaneously upload completed segments to other users. This means that users are downloading from each other as well as from the server or the initial distributors (Cohen, 2008). Compared to traditional web hosting where everyone downloads directly from the server, Bit Torrent significantly reduces bandwidth usage and hardware requirements for the initial distributor, thereby resulting in substantial cost savings. In some cases (as this model will demonstrate) it is possible for the initial distributors (seeds) to drop out completely without undermining the network, provided enough file segments are in distribution.

/~sangus/cgi-bin/mointeach.cgi/MonashU/ECC3855ComputationalEconomics/ProjectExhibit2008?action=AttachFile&do=get&target=Ball-Trathen-screen.png

Code

BallTrathen-BitTorrent.nlogo

Project Title

Highway Traffic Incident Reduction

Authors

Jacqui Zucco and Ravi Dutta

About

A primary consideration for the road traffic authorities when setting speed limits on roads is to reduce the likelihood of accidents occuring. Speed impacts on the likelihood of crashes being a constributing factor in one out of five accidents. Research suggests that the risk of involvement in a road traffic accident doubles with every 5km/h increase in speed over 60km/h. Whilst safety on roads is an important concern to the traffic authorities they must weigh this against the need for efficient movement along the roads. They balance these issues by attempting to maximising throuhput on the roads whilst minimising the likelihood of accidents. The road traffic authorities have even more recently been biased towards safety by setting lower speed limits in many areas. Is this approach justified? Many drivers on the other hand believe that driving at higher speeds will allow them to reach their destination faster. Is this an accurate belief?

/~sangus/cgi-bin/mointeach.cgi/MonashU/ECC3855ComputationalEconomics/ProjectExhibit2008?action=AttachFile&do=get&target=Zucco-Dutta-screen.png

Code

ZuccoDutta-HighwayIncidents.nlogo

Project Title

Do Differential Speed-limits Lead to Crash Reduction?

Authors

Dinendra Chandrasinghe and Yunhee Park

About

This project models the movement of cars, p-platers and trucks on a one way highway with two lanes. The underlying research question of the model is to find out the best policy that could be implemented on the speed limits of vehicles so that it leads to the least number of accidents on the highway. In order to investigate this policy, we carry out the following actions: 1. Placing the same speed limit on all three types of vehicles (base case scenario) 2. Placing differential speed limits on all three types of vehicles 3. Placing the same speed limit between 2 types of vehicles and a differential speed limit on the third type.

/~sangus/cgi-bin/mointeach.cgi/MonashU/ECC3855ComputationalEconomics/ProjectExhibit2008?action=AttachFile&do=get&target=Chandrasinghe-Park-screen.png

Code

Chandrasinghe-Park-TrucksOnHighways.nlogo

Project Title

Immigration Border Patrol

Authors

Craig Mawdsley and Daniel Rana

About

"What strategies used by custom officials are most effective in reducing the number of 'illegal immigrants' smuggled into Australia?" A study into people smuggling and trafficking by the Australian Institute of Criminology in May 2001 identified a growing trend in people smuggling over Australian waters. The reported number of arrivals by boat in 2001 was around 4,000, up from around 150 in 1997 Smuggling is carried out by amateur operators to trans-national organised crime syndicates. As the operations became more complex, a more organised response was required by border control authorities.

/~sangus/cgi-bin/mointeach.cgi/MonashU/ECC3855ComputationalEconomics/ProjectExhibit2008?action=AttachFile&do=get&target=MawdsleyRana-screen.png

Code

MawdsleyRana-ImmigrationPatrol.nlogo

Social Systems

Project Title

Melbourne Street Bashings

Authors

Peter Mutton and Phil Stefanovski

About

Given that there has been a growing number of street bashings in Melbourne this year the Victorian Government has been focusing on policies to reduce this number, such as the 2am lockout. We attempt to evaluate certain scenarios that lead to, or reduce the number of, bashings in a given time frame. Thus, our question is how does police presence and the 2am lockout affect the probability of street violence in the Melbourne CBD?

/~sangus/cgi-bin/mointeach.cgi/MonashU/ECC3855ComputationalEconomics/ProjectExhibit2008?action=AttachFile&do=get&target=MuttonStefanovski-screen.png

Code

MuttonStefanovski-StreetBashings.nlogo

Project Title

Bar Fights

Authors

Cheng Yang and Ronald Wong

About

This is a policy-driven model which seeks to explain the incidence of violent behaviour in nightclubs and bars depending on the level of security (number of bouncers), strength of alcohol served (assuming that all drinks are mixed and not bottled), and maximum admission rates. Although the obvious answer to reduce violence is to admit nobody and serve to alcohol, this model also seeks to maximize utility and club/bar profits, as reflected by the number of drinks sold. The aim of this model is to be simple enough to identify clear relationships to produce policy-relevant conclusions. At the same time however, we attempted to incorporate enough complexity into the model to make our results realistic. We spent considerable time in trying to accurately model how alcohol affects behaviour. Blood alcohol concentration levels and its effect on turtle behaviour is modelled based on data from Virginia Polytechnic Institute (2008). Alcohol elimination rate in the model is based on Reasor and Montgomery (1992).

/~sangus/cgi-bin/mointeach.cgi/MonashU/ECC3855ComputationalEconomics/ProjectExhibit2008?action=AttachFile&do=get&target=YangWong-screen.png

Code

YangWong-BarFights.nlogo

Project Title

Competition Amongst Franchises

Authors

Leigh Krafchek and Hayden Neeland

About

What impact do competing franchise outlets have on the profitability of the parent company? Is it more beneficial for the parent company if each individual outlet is in direct competition or is it imperative that each outlet is under strict instructions to adopt a similar overall business strategy in relation to pricing, marketing and expansion? How do consumers respond to these different scenarios?

/~sangus/cgi-bin/mointeach.cgi/MonashU/ECC3855ComputationalEconomics/ProjectExhibit2008?action=AttachFile&do=get&target=KrafcheckNeeland-screen.png

Code

KrafcheckNeeland-Franchises.nlogo

Project Title

Tackling the Crime Rate

Authors

Ruwangi Welikala and Johan Shah

About

One of the most important issues concerning countries is street crime. It is socially wasteful and can cause large costs to an economy. It is not surprising then that, policies that aim to reduce crime has been at the core of developed as well as less developed countries' policy agenda. The amount of crime that occurs in a given society, is not only the concern of the average citizen but businesses, politicians and the like. Thus it is crucial to determine the choices of policies and their implications in mitigating street crime. More specifically, countries must often make decisions regarding whether to adopt a top-down or a bottom-up solution to combating pervasive levels of crime. However it is often the case that governments give more emphasis to top-down approaches but in general they do use a mix of these two policies. Our focus is to determine which one of these policies, if any, is more effective. In essence, our research question is: What is the best way to reduce street crime? More specifically should we take a top-down, bottom-up or a combination of both approaches?

/~sangus/cgi-bin/mointeach.cgi/MonashU/ECC3855ComputationalEconomics/ProjectExhibit2008?action=AttachFile&do=get&target=WelikalaShah-screen.png

Code

WelikalaShah-Crimerate.nlogo

Natural Systems

Project Title

Overfishing

Authors

Andrew Johnson and Savannah Seixas

About

The NetLogo Overfishing model explores the real-life crisis of overfishing facing the world's fisheries industry and its effect on the ecosystem. Fishing is considered to be unsustainable if it tends to result in extinction of the fish population. Conversely, fishing is considered sustainable if it maintains the ecosystem over time, despite fluctuations in population sizes. To begin addressing this issue, users of the model are invited to consider a very simple ocean ecosystem containing only plankton (producer) and fish (consumer). Users are able to explore the effects on the population of each species by adjusting the available level of sunlight, initial populations, their reproduction rates and energy. The effects of predation are further extended by introducing predators (a consumer of fish) into the ecosystem. Finally, users can explore the effects of fishing through the introduction of fishing trawlers. Users are challenged to find a sustainable level of fishing by adjusting the initial number of trawlers in the ocean, changing the maximum legal catch of fish and varying the annual global change in demand for fish.

/~sangus/cgi-bin/mointeach.cgi/MonashU/ECC3855ComputationalEconomics/ProjectExhibit2008?action=AttachFile&do=get&target=Johnson-Seixas-screen.png

Code

JohnsonSeixas-Overfishing.nlogo

Project Title

Penguin Thermodynamics

Authors

Jing Leong and Jessica Berndt

About

The emperor penguin (Aptenodytes forsteri) breeds during the Antarctic winter, so far from their regular food source that male penguins fast for about 4 months whilst completing the breeding cycle. For successful incubation of their eggs, and for their own survival, male emporer penguins form huddles, allowing them to conserve their energy stores. Without huddling, male penguins would reach their critical body mass after 3 months, and would be forced to abandon the egg in order to return to the open sea to feed. Penguin huddles comprise hundreds of individuals, and occur more frequently during periods of high speed winds. Once in the huddle, the penguins move according to wind direction. Individuals exposed to the wind move slowly along the flanks of the group to reach the opposite, sheltered side. By doing so, the huddle moves collectively across the fast-ice. By forming huddles, male penguins minimise heat loss to reduce energy expenditure, whilst maintaining their body temperature at a near constant level. This is crucial for reproductive success, as full embryonic development requires a temperature of at least 35 degrees C. The normal core temperature at rest should be about 37.5 to 38 degrees C. During the breeding cycle, core temperature is reduced to 36.7 degrees C on average. Maintaining a lower core temperature requires less energy, also reducing energy expenditure during incubation. However, male penguins will not lower their body temperature below 35 degrees C, unless incubation fails.

/~sangus/cgi-bin/mointeach.cgi/MonashU/ECC3855ComputationalEconomics/ProjectExhibit2008?action=AttachFile&do=get&target=LeongBerndt-screen.png

Code

LeongBerndt-PenguinThermodynamics.nlogo

Project Title

Shark Attack!

Authors

Trent Wiltshire and Matt Eglezos

About

This model seeks to consider the factors that influence the possibility of a shark attack. It asks whether the risk of a shark attack varies as: the number of swimmers changes; the number of sharks changes; the beach is (not) patrolled by a lifeguard; the response rate of shark patrols change. Thus the research question posed is: Is the likelihood of shark attack affected by length of time in the water, the patrolling of beaches and the number of beachgoers? It is certainly true that the risk of a shark attack remains quite low. However, data from the International Shark Attack File (ISAF) (administered since 1958 by the Florida Museum of Natural History at the University of Florida) indicates that the number of unprovoked shark attacks has been gradually increasing, at least since 2003 (ISAF, 2008a). The ISAF suggests that the 'decadal number of unprovoked shark attacks [has grown] at a steady rate over the past century' (ISAF, 2008a). While the risk of drowning remains greater than the risk of a shark attack, '[s]hark attacks remain a genuine but unlikely danger for humans entering the water' (Australian Shark Attack File, 2008b). Of the shark encounters that have occurred, trend data suggests that an average of 30% of shark attacks are fatal.

/~sangus/cgi-bin/mointeach.cgi/MonashU/ECC3855ComputationalEconomics/ProjectExhibit2008?action=AttachFile&do=get&target=WiltshireEglezos-screen.png

Code

WiltshireEglezos-SharkAttack.nlogo