Summary of Thinking in Systems by Donella Meadows: Part 4

A System with Delays

• Picture a stock of inventory at a car dealership. There’s an inflow of deliveries from a factory, and an outflow of new car sales. By itself, this stock of cars at the dealership would behave like the water in a bathtub.
• Now, picture a feedback system that’s designed to keep the inventory high enough so that it can cover ten days’ worth of sales. The dealer needs to keep some inventory because deliveries and purchases don’t match perfectly every day. Customers make purchases that are unpredictable on a day-to-day basis. And deliveries from suppliers are also delayed at times.
• The dealer monitors sales. If they seem to be rising, she adjusts orders to bring the inventory up to her new desired level that provides ten days’ coverage at the higher sales rate. So, higher sales mean a discrepancy between current and desired inventory. This, in turn, means higher orders which will bring in more deliveries. And more deliveries will raise inventory so that the dealer can comfortably supply the higher rate of sales.
• This system is a version of the thermostat system. It has one balancing loop of sales draining the inventory stock. And there’s also a competing balancing loop that maintains the inventory by supplying what is lost in sales.

3 Types of Delays

• In addition to this, there’s something else that needs to be put into this simple model. To be more specific, there are three delays that we typically experience in the real world:
  • First, there’s a perception delay. The car dealer doesn’t react to just any change in sales. Before making ordering decisions, she averages sales over the past five days to sort out real trends from short-term dips and spikes.
  • Second, there’s a response delay. Even when it’s clear that orders need to be placed, she doesn’t try to make up the whole adjustment in a single order. Instead, she makes up a third of any shortfall with each order. In other words, she makes partial adjustments over three days in order to be sure that the trend is real.
  • Third, there’s a delivery delay. It takes five days for the supplier to receive an order, process it, and deliver it to the dealership.
• So, this system is still like the simplified thermostat system in that it consists of just two balancing loops. However, it doesn’t behave like the thermostat system.

Side Effects From Delays

• For instance, if the dealership experiences a 10% increase in sales, there are fluctuations between too little and too much inventory. A single increase in sales causes inventory to drop. But, the dealer watches long enough to be sure the higher sales rate will last.
• Then she starts ordering more cars to both cover the new rate of sales and bring the inventory up. But, it takes time for the orders to come in. And during that time, inventory drops even further. So, orders have to go up even more to bring inventory back up to ten days’ coverage.
• But eventually, the orders start to arrive, and inventory recovers. In fact, it more than recovers, since during the time of uncertainty about the actual trend, the dealer ordered too much.
• She now sees her mistake and cuts back. But, there are still high past orders coming in, so she orders even less. In fact, since she still isn’t sure of what will happen next, she orders too little. Now, inventory gets too low again. And so on, through a series of fluctuations around the new desired inventory level.

Why Fluctuations Occur

• Why do these fluctuations in inventory occur? It’s not because the car dealer is stupid.
• Instead, it’s because she’s operating in a system in which she doesn’t have timely information. On top of this, physical delays prevent her actions from having an immediate impact on inventory.
• Information insufficiency, physical delays, and fluctuations are common in inventories and other systems. For instance, if you try to take a shower, and there’s a very long pipe between the hot- and cold-water mixer and the showerhead, you’ll experience the hot and cold fluctuations because of a long response delay.
• Let’s say the dealer wants to change the behavior of the system. Since there’s not much she can do about the delivery delay from the factory, she decides to react faster. To do this, she averages sales trends over only two days instead of five before making order adjustments. But not much happens as a result. If anything, the fluctuations in the inventory of cars are a bit worse.
• Now, rather than shortening her perception time, let’s say she shortens her response time by making up perceived shortfalls in two days instead of three. In this case, things get even worse.
• This kind of contradictory result that you get when trying to fix a system can be seen quite often. In other words, we can be surprised by the counterintuitive behavior of systems when we try to change them.
• Part of the problem is that the dealer has been reacting not too slowly, but too quickly. Given the design of this system, she has been overreacting.
• For instance, things would go better if, instead of decreasing her response delay, she would increase it from three days to six. The swings back and forth between too much and too little inventory are greatly restrained with this change, and the system finds it new equilibrium fairly efficiently.
• Changing the delays in a system can make it either easier or harder to manage.

What System Thinkers Look For

• System thinkers are somewhat obsessed with delays. They want to know where they occur, how long they are, and whether they’re delays in information flows or physical processes.
• Some delays can be powerful policy levers. Lengthening or shortening them can produce major changes in the behavior of systems.
• In the big picture, one dealer’s inventory problem may seem small and fixable. But imagine that the inventory is that of all the unsold cars in America. Orders for more or fewer cars affect production not only at assembly plants and parts factories, but also at steel mills, rubber and glass plants, and textile producers.
• Everywhere in this system are perception delays, production delays, delivery delays, and construction delays.
• Next, consider the link between car production and jobs. Increased production increases the number of jobs. This allows more people to buy cars. That’s a reinforcing loop, which also works in the opposite direction. Less car production means fewer jobs, fewer car sales, and less production.
• Again, put in another reinforcing loop, as speculators buy and sell shares in the auto and auto-supply companies based on their recent performance. In other words, an increase in production produces an increase in stock price, and vice versa.
• That very large system, with interconnected industries responding to each other through delays, carrying each other along in their fluctuations, and being amplified by speculators, is the primary cause of business cycles.
• Economies are complex systems. They’re full of balancing feedback loops with delays, and they can swing back and forth between periods of decline and growth.

Click here to review part one of this summary, here to review part two, and here to review part three.

To get your own physical copy of Thinking in Systems, click here. For the Kindle version, click here. Or, to get a free copy of the audiobook with a 30-day standard trial, click here.