Whether you think of the program as a stimulus for the economy or the environment, it cannot continue to operate without funds.  Congress had to quickly approve another $2 billion for the program when the initial $1 billion appropriation was exhausted in less than a week.  The rebate fund lacks an inflow of cash to offset the outflow of payments.

I wrote about a STELLA model of a self-financing program to promote cleaner vehicle sales in May.  The “Feebates” model uses fees charged on sales of vehicles that have high pollution rates to finance rebates for cleaner vehicles.  I asked the Feebates model author, Andrew Ford what he thought about “Cash for Clunkers”.  He sent me an article he co-authored with Todd BenDor titled “Simulating a combination of feebates and scrappage incentives to reduce automobile emissions”.  Basically, the article describes a feebates model with a “cash for clunkers” type of rebate added to promote scrapping older vehicles.  The article was published in the journal Energy, 31(2006), 1197-1214.

Mr. BenDor was kind enough to send me the STELLA model so we could publish it to the web with isee Netsim.  We thought this would be a great way to start some discussion around the Clunkers program, public policy and using simulations to facilitate decision making.

Click here to run the simulation online.

• Can you design a policy that stimulates the economy and reduces emissions?
• This simulation uses fees to fund the program instead of deficit spending.  What other sources of revenue could we consider?
• Many states levy a “sin tax” on tobacco and alcohol.  How would the public react to fees on high-pollution vehicles?

Post your answers in the comment section below.

Image credit: http://www.flickr.com/photos/threadedthoughts/3808715790/

The most striking thing about this model to me is that it was created completely hierarchically.  I have seen many large models broken into sectors which are conceptually all at the same level.  I have seen other large models that are organized by feedback loops, which can at some times be large and unwieldy.  But I had yet to see an example that is truly hierarchical, with an appropriate dynamic hypothesis at each level.

The model is three-levels deep.  At the lowest level are models with very simple dynamic hypotheses.  At the next level up, groups of these smaller models are tied together to form more complex feedback loops, or loop sets, comprising a higher-level dynamic hypothesis (this is the complexity most of the models I develop have).  At the top-level, they are tied together to form a very high-level dynamic hypothesis.  One of the very nice things about this is that each part of the model, which was built bottom-up, has already been tested in isolation or within its group before the whole model is tied together.  All parts are in steady-state.  In this way, we have built confidence in all of the parts of the model and now are only testing the broadest feedbacks.

Critics of this approach insist that delaying the connection of the broadest feedbacks until this late in the development of the model hides important dynamics that affect all parts.  Not only is this considered risky, but the model does not generate results until the end.  After this experience, I can’t agree with this point-of-view.  There was very little risk in tying the pieces together at the end because they were well-formed pieces already rich in feedback, and (very important!) initialized in steady-state.  While the model did not address some of the overarching issues until the end, careful testing of the pieces of the model added insight at many steps along the way and even gave hints about what might happen when those final feedbacks were put into place.

One of the key questions this model attempts to address is how to combat the rising costs of health care.  The highest level dynamic hypothesis shows this to be exceedingly difficult because the processes in place at this level are all reinforcing loops:

Top-level of model (This is only a portion of the model. The full model includes three more modules for hospital services, and one for general accounting.)

The outer loop, R1, goes like this:  As population grows, there is more demand for health services (thru physicians).  As these services are used, more money is spent on technology (including pharmaceuticals), more money is devoted to research, and more causes of death are held in abeyance.  This increases expected longevity, which increases the population.  Of course, there must be physical limits to this (mustn’t there?), but it comes down to how much people are willing to pay to live a little longer.

The smaller loop, R2, goes like this:  Better technology (new or improved treatments) itself drives more demand for health care services, which creates more research dollars, and thus more technology.

In other words, desire for longer lives and better treatments itself serves to drive up health care costs.  This is a tough lesson.

The details are, of course, much more complicated than this.  For example, at some point, and some would say we have already reached that point, we will hit diminishing returns.  That is, we will have to spend far more money for a smaller incremental improvement in life expectancy.  Will the market bear the costs of this research?

Moving down a level into the Physicians module, we find another dynamic hypothesis:

Second level: Inside the Physicians module

As the number of physician visits increase, the physician capacity in hours increases (note the number of physicians itself does not typically respond as limits are set by AMA policy).  As the number of hours worked increases, physician prices increase.  Higher prices lead to fewer visits.  This is a simple balancing loop.  When the visit rate falls sharply, this balancing loop operates slightly differently:  As number of physician visits decrease, physicians have more free time to attend to each patient, lower their prices, and drive the visit rate back up both through lower prices and an increased frequency of visits.

Drilling down to the third and final level, we find this structure within Physician Visits:

Third level: Inside the Physician Visits module

This includes the effects of a managed care organization (MCO) trying to manage utilization rates of physician services (UM is utilization management).  As the visit rate increases, the MCO works to reduce the visit rate.  However, as the visit rate falls, patient satisfaction also falls and patients work to circumvent the MCO initiatives, thus bringing the visit rates up.  There are thus two balancing loops working in opposition to each other.  As you can easily imagine, they arrive at a compromise somewhere in the middle.

I have built a number of smaller models using modules and I have built them all from the top down.  I draw out my dynamic hypothesis using modules and then drill-down into each module to flesh out its dynamics.  Using Run by Module, I can test these modules in isolation and then in combination with others as they become available.  I start with one model file and I end with one model file.  I show this workflow below.

This model of the health care system was built from the bottom up.  This process was therefore quite different.  Each model at the lowest level was built and tested independently as its own model file.  Once I was convinced a model was working properly, I would move onto the next.  Once I had the pieces completed for a loop set of interacting models (for example, the contents of the Physicians module), I would create a new model file, create a sketch of the dynamic hypothesis at this higher level and then import the already working models into those modules.  After connecting the variables across those modules, I could test different combinations of modules in that loop set.

Finally, when I had finished all loop sets, I created one more model file for the entire model.  I drew the dynamic hypothesis for this model using modules and then imported the already working loop sets into the modules.  I then only had to connect the variables across this level and test the various combinations of modules until I was satisfied the entire model was doing what it was supposed to do.  I show this workflow below, assuming three submodels at the lowest level and a total of three levels.

Of course, it was a great advantage that the model had already been designed and implemented once before by Jim Thompson.  For such a large model, it would most often be disastrous to just start building at the bottom without knowing how it was going to tie together at the top.  Because he did such an excellent job in partitioning this model, it worked out very well.  I suggest, therefore, that a hybrid approach – both top-down and bottom-up – is more likely to succeed when starting a new project.  In this case, the architecture gets worked out ahead of time in the form of dynamic hypotheses at each level.  Then the smaller models can be built.  Indeed, this really is the approach I followed because Jim had worked out the dynamic hypotheses at each level ahead of time for me.  I show this workflow below, assuming three levels.

It occurred to us that the blog would make a good home for the list because others could help keep it current by posting comments with links to newly published papers or papers we may have missed.

To get the ball rolling, we compiled a list from journals we are familiar with – System Dynamics Review and Journal of the Operational Research Society. The result is by no means exhaustive. It is simply a way to get things started. We need your help to expand this list and to include other journals that you’re reading.

If you know of a published paper about a STELLA or iThink application that does not appear on this list, please let us know about it by submitting a comment. It would be helpful if you included citation information so that we can quickly verify it and update our list.

Thanks for your help!

System Dynamics Review

Journal of the Operational Research Society

Medici's Lever consists of educational games and a freestyle policy laboratory

A couple of months ago, Steve Peterson came by the office and showed me an online game created with iThink and isee NetSim called Medici’s Lever.  I was really impressed.

The game begins in San Jose, California.  Dr. Lilia Maria Delgado, a legend of Silicon Valley’s rise to high-tech super power, has died and left a half-billion dollars to establish the “Delgado Arts Endowment”.  In her will, Dr. Delgado simply states that the mission of the Endowment is to advance the cultural life of San Jose.

Your job is to fulfill the mission by experimenting with the various approaches of five different CEO candidates over a 40 year period.  Each candidate personifies a different theory of arts and cultural development and provides you with different investment levers that impact overall “cultural vitality”.

The San Jose Rennaissance is one of two educational games that are a part of Medici’s Lever.  The other game is very similar but takes place in a different region — this time a fictitious European city dubbed the Capital of Culture.

The final module of Medici’s Lever is an interactive Policy Laboratory that lets the user “dial in” initial parameters to define the region in which they want to experiment.

The 3 elements of the "logic model" are displayed as a pyramid in the games

Medici’s Lever is powered by an underlying iThink model (although users are unaware of this).  The games inform the player that it is driven by a “logic model” they must keep in mind when experimenting.

The premise of the logic model is: In order to maintain a healthy cultural ecology a region must strike a balance between three elements — cultural literacy, participation in cultural expression, and consumption of professional cultural goods and services.

In the games, the three elements are visualized as a pyramid.  What I found out after a couple of runs is that investing in the bottom of the pyramid pays off in the long-term.  The insight I gained was that a community needs to have people who appreciate art and culture and actively participate in it from a young age to be successful supporting the professional artists. In other words, it does not make sense to invest in a state-of-the-art performance hall if no one wants to attend the performances.

I think Medici’s Lever is an excellent example of how you can get folks thinking about a systemic issue in an easy and engaging way.

Background and Development

Medici’s Lever is the final project of Cultural Initiatives Silicon Valley, an organization that worked to implement a ten-year cultural plan for Silicon Valley from 1996 to 2006.  John Kreidler, Cultural Initiatives’ retired Executive Director, is responsible for the model concept and design.  The iThink model and web interfaces were developed by Steve and published online with isee NetSim.

It is important to note that at its inception, Cultural Initiatives planned to operate for one decade only.  As the end of that decade approached, John and Steve began working on bringing all of the ideas of the foundation together in a model that could be played as a game.  It was key that the game live online so others could benefit from the insights of the foundation long after it had sunset.

When Steve started developing the iThink model for the game, isee NetSim was in its beta cycle.  We worked together to see if at this beta stage isee NetSim was up to snuff for creating the type of online game that he envisioned. During the process, Steve gave me a lot of valuable feedback on isee NetSim (and rooted out quite a few bugs too).

isee NetSim turned out to be a good fit.  Steve has been using iThink for a long time so he knows it well.  isee NetSim enabled Steve to do all of the interface design inside iThink itself and let isee NetSim take care of converting it to run over the web.  Since the games could now be run in the browser, players would not need to download and install software to play the game.

Steve told me that the game is intended for policy makers, community leaders, educators, artists, business leaders — anyone who should be thinking about the role of arts and culture within a community.  I asked him how he thought Medici’s Lever would be used.

From my perspective, a game like this provides an excellent vehicle for focused discussion about arts & culture policy.  Medici’s Lever surfaces a host of questions about what is necessary to build a rich, sustainable cultural landscape.  It provides a rich context for informed discussion. Models like this are not “true” in an objective sense, but they have great potential utility.”

Freestyle allows the user to "dial up" conditions for real metropolitan regions anywhere in the world.

This is an interesting point about models in general.  This game is fictitious, but the dynamics are useful.  By allowing players to dial in different initial parameters (population size, immigration, birth and death rates, social cohesiveness and several cultural factors), users can create conditions for real metropolitan regions anywhere in the world.

Another important point is that the model runs over the web.  This makes it easy for people who are thinking about these issues to interact with the games and lab and have an informed discussion.  Players do not need a background in modeling and system dynamics. And they don’t need to download and install software.

If you want to experience Medici’s Lever for yourself, click here to play the game now.

Before branching out on his own, Chris worked for isee as a lead consultant, trainer, and learning environment developer. He co-developed the interactive teaching tool Systems Thinking: Taking the Next Step and has used our software to generate insight and understanding in others for many years. He is a real Systems Thinker and an excellent teacher. We think he is so good at teaching this systems stuff, we rely on him to facilitate our online training and to help us out at conferences and workshops.

Chris likes to tackle timely issues with Systems Thinking and is interested in how sharing models and videos online can help frame the discussion around them. Last year’s spike in fuel prices and public concern about future oil reserves inspired Chris to create a YouTube video and online Learning Labs to explore the supply and demand dynamics of the oil market.

YouTube video: “How long will oil last?”

www.youtube.com/watch?v=aTyrnQj03s4

Online Learning Labs

The following learning labs were developed with iThink and then published to the web using isee NetSim

How much oil is left?

Oil market dynamics

What do you think?

Chris is leading the effort to define an online Learning Lab format. We’d love to hear your reactions. Are there other topics you think this approach would be good for?