A model you can recite is worth very little. A model you reach for automatically, on problems that didn’t announce themselves as stock-and-flow problems, is worth a great deal. This last teaching lesson is about that promotion — from “I understand the bathtub” to “I run the bathtub on everything.” We’ll boil the whole course into a five-question checklist, debug a few real problems with it, and tie the loose threads back to the other models in your latticework. By the end, “stock or flow?” should be the first cut you make on anything that builds up over time, as reflexive as checking your mirrors.
The five-question checklist
When any problem involves something that accumulates — money, people, carbon, trust, inventory, skill, anger — run these five questions, in order. Most reasoning errors die at one of them.
- What’s the stock? Name the level you actually care about — the thing you could photograph. Not the rate. (Debt, not deficit. Headcount, not hiring. Weight, not calories.)
- What are the flows? Name the inflow(s) filling it and the outflow(s) draining it, each with a “per time” unit. Miss a drain and your whole analysis tilts.
- What’s the net flow — and which way is the stock moving? Inflow minus outflow. Positive rises, negative falls, zero holds. This is the direction; nothing else sets it.
- Is someone quoting a flow but implying the stock? The deficit-fallacy check. If a rate is being celebrated for going down, ask: is the inflow now below the outflow? If not, the stock is still climbing.
- How long will this take? The inertia estimate: gap in the level ÷ net flow toward it. Stocks are slow; get a number before you promise a timetable.
That’s the entire course as a procedure. It looks almost too simple written down — but watch how often a confident plan or a scary headline collapses the instant you make it answer all five.
The checklist, pocket-sized
Stock? Flows? Net (which way)? Flow-dressed-as-stock? How long? Five questions. Run them on anything that accumulates and you’ll catch the level/rate confusion, the missing drain, the deficit fallacy, and the impossible timetable — the four ways accumulation reasoning usually fails.
Debugging a personal-finance problem
Take a common one: “I got a big raise but I’m not getting any richer.” Run the checklist.
- Stock? Wealth — net worth, a level in dollars. (Not income, which is a flow.)
- Flows? Inflow: salary, investment returns. Outflow: spending, debt payments, lifestyle creep.
- Net? Here’s the reveal: the raise lifted the inflow, but if spending rose to match — the classic lifestyle inflation — the net flow barely moved, so the wealth stock barely grows. A bigger faucet with an equally bigger drain is dynamic equilibrium: roaring flows, flat level.
- Flow dressed as stock? Exactly the error in the complaint. “A big raise” is a flow event; “getting richer” is a stock change. They feel like the same sentence and they are not. High income with high spending is a busy tub that never fills.
- How long? Even with a healthy net flow, wealth is a big stock to move — gap ÷ net flow says years, not months. Patience isn’t optional; it’s arithmetic.
Notice the checklist didn’t just describe the problem — it located the fix. You don’t get richer by raising the inflow if the outflow tracks it; you get richer by opening a gap between them and letting the accumulation rule do its slow, relentless work. Which, not coincidentally, is the whole secret of the next model.
Two friends both earn $90k. Ana saves $1,500/month and spends the rest; Ben got a raise to $120k and now spends all of it, saving nothing. Who is building wealth, and why — in stock-flow terms?
Debugging a policy problem
Now a public one: “We added 50,000 affordable homes this year, so the housing shortage should be easing.” Checklist.
- Stock? The shortage — the gap between homes needed and homes available, a level.
- Flows? Inflow to the housing stock: new construction (those 50,000). Outflows and competing inflows: demolitions, conversions, and — crucially — new household formation and in-migration, which raise the needed side.
- Net? If the city added 50,000 homes but gained 70,000 new households, the shortage stock got worse, even though construction was real and large. The inflow people quote (homes built) is only one of several flows moving the level.
- Flow dressed as stock? “We built 50,000 homes” is a flow; “the shortage is easing” is a stock claim. The leap skips the demand-side flows entirely.
- How long? Housing stock is enormous and construction is a trickle against it — gap ÷ net flow is measured in years even when every flow points the right way. One good year barely dents it.
The checklist exposes the favorite move in policy debates: announce the one flow you control (homes built, jobs created, troops sent) and let the audience hear it as a stock outcome (shortage solved, unemployment fixed, war won) — while quietly ignoring the other flows moving the same level. You won’t fall for it again.
A government wants to shrink a stock. Sort each proposed action by whether it works on the INFLOW or the OUTFLOW of that stock.
Place each item in the right group.
- Run a budget surplus to shrink the national debt
- Speed up case processing to shrink a court backlog
- Stop new borrowing to slow the growth of the national debt
- Reduce new filings to slow the growth of a court backlog
- Cut carbon emissions to shrink atmospheric CO2
- Plant forests and capture carbon to shrink atmospheric CO2
The three threads back to the latticework
Stocks and flows aren’t an island — they’re the floor several other models stand on. Three connections worth making explicit, because each one gets sharper once you see the bathtub underneath it.
- Compounding is a stock wired to feed its own inflow. Money earns interest that becomes more money that earns more interest — the faucet’s setting climbs with the level. That’s a reinforcing loop built on the accumulation rule, and it’s why patient saving (a steady positive net flow into a self-feeding stock) becomes a fortune. The slow inertia you cursed in Lesson 4 is the same patience that makes compounding pay.
- Bottlenecks are a statement about flows. The slowest stage caps the throughput, so a stock drains or fills no faster than its narrowest exit, and work piles up — a stock — in front of the constraint. Want a stock to move faster? You can’t push the level; you have to widen the bottleneck flow.
- Feedback loops are the reason we started here. A loop works by sensing a stock and adjusting a flow — a thermostat reads the temperature (stock) and toggles the heater (flow). Every loop in that course was a wire running from a stock back to the flows that feed it. You can now read any loop as plumbing.
The unifying punchline across all three: you can never change a stock directly — only its flows, or the loops that drive them. This is the deepest practical lesson of the model. Staring at the level, wishing it were different, bailing frantically — that’s flailing at the stock. Real leverage is always upstream, on a flow or on the loop structure setting that flow. Fix the plumbing, not the puddle.
Fill the tub
You can't push the level — only the taps
Set the faucet and the drain — two independent rates — then drag the time slider. Watch the level integrate the net flow: it ramps up when the faucet wins, drains when the drain wins, and holds perfectly steady when they match (at any level).
Inflow 6 beats outflow 4 L/min: a net of +2 L/min. The stock climbs by 2 L every single minute — a straight ramp — even though neither flow ever changes. By minute 12 the level has reached 74 L. The level is the running total of the net flow, not the flow itself.
Usually the one with a loop wrapped around it — a flow that feeds back on the stock — because changing that doesn’t just nudge the level once, it changes how the system responds forever after. Throttling a one-off inflow drains the tub a little; re-wiring the loop that controls the inflow changes the tub’s whole future behavior. That’s why the feedback-loops course put “loop structure” and “the system’s goal” near the top of its leverage ladder and “tweaking a number” near the bottom. In stock-flow terms: bailing water (fighting the stock) is weakest; turning a tap (changing a flow) is stronger; rebuilding the plumbing so the taps regulate themselves (changing the loop) is strongest of all. Find the flow that a reinforcing loop is amplifying or a balancing loop is defending, and you’ve found the place where a small, well-aimed push moves the whole system.
Recap
Big picture
Stocks & Flows: the whole course in one picture
- The grammar of accumulation
- Stock vs. flow
- Stock = a level you can photograph (liters, $, people, trust)
- Flow = a rate you can only film (per minute, per year)
- The test: count the "per times" in the unit
- The accumulation rule
- Change in stock = inflow − outflow
- A stock is the running total (integral) of its flows
- So it smooths, lags, and remembers
- Equal flows → flat stock (dynamic equilibrium)
- The deficit fallacy
- A falling inflow is NOT a falling stock
- Ladder: inflow falls → stock still rising → peaks → falls
- Reflex: is the inflow now BELOW the outflow?
- Inertia, lag & overshoot
- Stocks are slow; time ≈ gap ÷ net flow
- Delayed reading → over-correction → overshoot
- Inertia is also a buffer: stability vs. responsiveness
- Debugging tool
- Stock? Flows? Net? Flow-as-stock? How long?
- You change a stock only through its flows or loops
- Threads: compounding, bottlenecks, feedback loops
- Stock vs. flow
Put the whole model to work
You run the five-question checklist on 'crime fell because arrests are up.' Which question most directly exposes the weak reasoning?
Check your answer to continue.
Where this goes next
You’ve turned the bathtub into a tool. The five-question checklist — stock, flows, net, flow-dressed-as-stock, how long — is now something you can run on a household budget, a climate pledge, a hiring plan, or a housing debate, and it catches the four classic ways accumulation reasoning fails. You’ve seen that stocks and flows are the floor under compounding, bottlenecks, and feedback loops, and that the model’s deepest lesson is about leverage: you change a stock only through its flows or the loops that drive them — fix the plumbing, not the puddle.
One thing remains: proving to yourself that it’s stuck. The Final Exam pulls the whole course together — the two-second test, the accumulation rule, the deficit fallacy, inertia and overshoot, and the debugging checklist — into one graded, one-way run. Several questions look easy until you notice a flow wearing a stock’s clothing. Reason each one all the way through, the way you now know how.