Thinking in Systems, Ch. 2: Types of System Dynamics

i'm going through this book chapter by chapter and explaining the basics from the chapter this video is about chapter two from her book and i would like to actually approach this video in a slightly different way i'm not just going through the content i'm actually going to go through how i encountered the content as an economist and some confusions i had about the way she was presenting it and how i resolved those so first she opens chapter two by saying she's going to go through the zoo of different systems where she basically says there's different

categories of systems that oftentimes have things in common but uh they're all wild beasts and they each have unique features so she she kind of made this analogy where it's sort of like yeah there's the monkey section of the zoo but monkeys evolved to have different properties and then there's a lion section of the zoo so that was how she opened it and when she did that what i was expecting was a title of a section that said here is one type of system this is the lion type of system and here's how i would

describe it so i was actually looking for like five different categorizations of of systems and i did not find that actually and yet when you look at her her section titles um they're basically one stock systems and two stock systems which i do not think are her actual categories of systems um so i had to figure out what does she mean by these different classifications if she's not giving a name for each different type and the conclusion i came to was that she's categorizing different systems according to what this stock within a system would look

like over time so like a balancing system sort of stays pretty steady and occasionally bumps below or bumps above but there are natural forces that sort of bring this back to a steady level that's one broad category of a system another category of a system is systems that sort of have this exponential growth pattern but then they have a collapse moment where they collapse and then may return to some kind of equilibrium perhaps a different equilibrium than before and then there's of course oscillation is another pattern you might see of the stock within a system

okay so the next point of confusion for me was the fact that she's talking so much about stock in the systems and her depictions of these systems over time are depictions of the stock and yet in the previous chapter she had said actually the elements within the system are not where the real action and dynamics are at it's the connection between the elements that matters most for a system so i was confused that she was talking about stock when stuck in and of itself the things in a system are not what really matters what matters

are the dynamics and the conclusion i came to here was actually that stock is easy to see easy to visualize easy to measure so if you're someone who is analyzing systems the connections between the the different parts and the incentives at play all of that is going to be invisible and it's really hard to analyze invisible parts of a system so she's focusing on stock because that's essentially what we can measure and visualize and analyze and the dynamics the interconnections that lead to the different movements within a system uh those we're only going to see

through the way they impact the elements okay so with that in mind let me sort of walk through some of the key points she's making in this chapter and uh to give you a preview i think one of the key points is what are the limiting factors of a system because we can't have a system that just grows and grows and grows forever there have to be forces that sort of bring that system back to the ground uh there's limits in the universe which is something that economists are very in touch with and i'll connect

that to some of economists shapes of our graphs we look at in a minute but that's one of the key points in this chapter she also talks about renewable versus non-renewable elements of the stock in a system and i think that's really really helpful to think through okay so if we're going back to the one stock versus two stock system i think stock is really the thing you're you're looking at to measure the dynamics of the system and the thing that might matter in some way though it's not necessarily the purpose of the system examples

of one stock systems include the room temperature where the temperature of the room is the stock a used car lot where the cars on the lot are a stock the money supply in the economy capital for a particular firm how much capital does it have in stock populations are a version of stock that's easy to analyze and may sometimes have this trait so that's a one stock system and then a two stock system may be the same type of system as one of the one stocks it's just where you've added another type of stock that

you're looking at in tandem with the first type so if you had population as the stock you're looking at population plus food supply as two separate stocks that have an interaction between them that could be a two stock system so the the one stock versus two stock is actually not about an inherent property of the system it's more about what are you looking at when you analyze the system and you might ask the question why don't you look at 10 different stocks within a system to figure out how it works and you could definitely do

that but of course dealing with models part of what you're trying to do is you're trying to reduce the complexity of models to understand what might happen in the dynamics and the more stocks you add to the system that you're analyzing simultaneously the more complicated the model gets so there may be cases where you want to add 10 types of stock into your your model for sure but it's not necessarily inherent that a 10 stock model is any better than a one stock model it just depends what you're trying to do with your model okay

so let me go through the three of these examples and these are not titles that she gave to these systems i'm just sort of giving you one of the forces that's keeping this in place so this is a system such as the number of cars on a used car lot where there's a balancing feedback loop in place so the the stock is fairly consistent over time if they sell too many cars in a weekend or sell more than expected then they're going to order more cars to come in the next month or the next week

and that's going to bring the number of cars on the lot back up to equilibrium and if they have a bunch of shipments that all arrive the same week and you didn't expect that you might have a bump up in the stock in which case they might lay off ordering new cars for a while and that'll bring it back to equilibrium so one of the features of a system that has a balancing feedback loop which we talked about in chapter one is that there may be a somewhat steady stock over time even if we perturb

the system there are natural forces that will bring it back to equilibrium and of course economists love equilibrium we're always looking for equilibrium now the second situation here is where there's a reinforcing feedback loop so exponential growth such as exponential growth of a population and you might imagine this like a population of bunnies that's sort of introduced into a new environment the bunnies reproduce and there's exponential growth in the number of bunnies in the environment but sometimes what can happen that can cause a collapse in that situation is that the population of bunnies eats up

all of the food supply so all of a sudden at some point there's going to be no food left to support the exponential growth and at that point you could have a population collapse and the population collapse may end in a new equilibrium that's different from the old equilibrium of rabbits in the population which was perhaps zero in my example so so this dynamic with exponential growth and then a crash is just one phenomenon you see in dynamic systems now down here we have an oscillating system and the oscillating system may oscillate around some stable

equilibrium but it's constantly going back and forth up and down around that and she pointed out that situations that lead to oscillating systems tend to be situations where there's a delay in the feedback loop such that you've overshot where you should be there's a balancing feedback loop that's going to bring you back to normal but because it's delayed you sort of overshoot your correction and you end up going down and three examples of the oscillating feedback loop mechanism are going to be the temperature in a room the federal reserve lowering and raising interest rates and

um the perhaps the political rightness or leftiness of a a government in an elected economy so i used to work at the federal reserve bank of kansas city and we had this uh conference room where we would meet and the economists would talk about uh various things i don't need to go into that but but the temperature in that conference room was always too hot or too cold and almost always when we went into that room to sort of make one of these decisions or come up with to talk about what economists talk about somebody

would make the same joke every time which is that the temperature in this room is just like the federal reserve and interest rates it was just the running joke and different people made that joke at different times but it was almost like somebody has to make the joke if it's too hot or too cold so what was the situation there the situation was you go into the conference room it's too hot people are really uncomfortable someone turns the thermometer way down to fix this situation and a little bit later it's going to be way too

cold and people are upset about that so they're going to turn it way up and it's going to get really hot and the fact that we couldn't get the the system in balance because it always felt so urgent that we changed the temperature that we overshot that was just this oscillating cycle and of course that relates to the interest rates because the federal reserve in in theory like an oversimplification of what they're doing with the fed is you want to raise interest rates when the economy is booming and lower interest rates when the economy needs

a little bit of a nudge but the problem with that is that the information we have about whether the economy is booming or or in a recession that information is delayed so by the time the fed figures out it should act um it might act in a way that's an overreaction given where the economy actually is now i'm not saying this is a good analysis of the fed or a good analysis of what's going on um i i was at the fed back in 2003 so that was a totally different era but the concept here

is that systems tend to over correct if there's this delay and that can lead to this type of to systems that look like this okay i think this is probably enough for one video so i'll probably do another video on chapter 2 where we look at limiting factors and renewable and non-renewable resources expect that video soon