Track 1 · Machinery · lesson 7
Monetary and fiscal policy: the mechanism, not the argument
13 min
Track 1 · Machinery · lesson 7
13 min
There are two large levers. One changes the price and quantity of credit. The other changes how much the state itself spends and takes.
This lesson describes what the levers physically do. It does not tell you when to pull them, because that question is not settled and pretending otherwise is how a course on money becomes a pamphlet.
Monetary policy operates on credit: what it costs to borrow, and how much of it there is. It works indirectly, through millions of private decisions about whether a project still clears the bar.
Fiscal policy operates on spending and taxation: the state buying things, paying people, and removing money from households and firms. It works directly, because the state's own purchases are part of the output being measured.
They are frequently discussed as two settings on one dial. They are two different machines with different operators, different lags and different failure modes.
A central bank sets the rate at which institutions borrow from it and from each other overnight. That is a very short rate on a very specific transaction, and by itself it is nearly irrelevant to anybody's life.
It matters because of what it propagates into. Longer rates are set in markets, and a long rate is roughly what people expect short rates to average over that length, plus compensation for being locked in and for risk. Move the short rate, and — to the extent that people believe you will keep it there — the long rates move with it.
From there the previous lesson takes over. Every project whose return sat between the old cost of money and the new one stops being worth doing. Every asset priced off a discount rate gets repriced. Every household with a loan that resets finds a different number on the statement. Nobody experiences any of this as an instruction.
Predict
The state spends money into the economy and takes money out of it through taxation. The gap between the two is the deficit, and it is financed by borrowing.
Three things follow mechanically, and none of them is controversial.
Spending buys real output. A road built is concrete poured and people employed. Unlike the monetary lever, this does not depend on persuading anyone to change their mind — the purchase is the effect.
Taxes remove spending power. Money taken from a household that would have spent it reduces demand by roughly that amount. Money taken from a household that would have saved it reduces demand by much less, which is why who is taxed changes what the tax does.
Borrowing competes for savings. The state issuing debt is a borrower alongside everyone else, and it is a very large one with an unusually good credit standing.
Here is where the honest version of the argument starts, because both readings below use the same mechanisms and disagree about one input.
The mechanisms say: extra demand from anywhere bids for real output. The question is what happens when it does. If there are idle factories and people who want work and cannot find it, the extra demand can be met by producing more, and output rises. If everything and everyone is already busy, the extra demand cannot be met by producing more, so it bids up prices and pulls resources away from whatever they were doing.
Everything turns on how much slack there is. That is a fact about the world at a particular moment, it is genuinely difficult to measure, and it is not directly observable — which is why two careful people looking at the same economy can reach opposite conclusions without either being unreasonable.
Two readings
Extra state spending largely displaces private activity rather than adding to it, and shows up in prices and rates.
Where resources are idle, extra spending pulls them into use and adds to output rather than displacing anything.
Both sides accept
Both accept that real resources are finite, that spending has to be financed one way or another, and that the answer depends on how much unused capacity exists at the time. Neither claims its own answer holds in every state of the world. The dispute is about how often each state obtains, and about how reliably slack can be measured before the decision has to be made.
Which is which? Put each one on a side.
A programme announced in the second year of a deep downturn with high involuntary unemployment
A programme requiring a specialist skill that takes six years to train and is already fully employed
A programme funded by taxing households who would have saved most of the money
Four things survive both readings, which makes them worth more than either.
Timing dominates. The same action taken at two different points in a cycle has opposite effects. Most disagreements about whether something "worked" are disagreements about what the state of the world was when it was done.
Composition dominates too. Spending on something with idle supply behaves differently from spending on something with fixed supply. An aggregate number hides that completely, which is why aggregate arguments are so unsatisfying.
Both levers act with a lag, and the lags differ. Fiscal spending can be slow to start and fast to bite. Monetary policy is fast to start and slow to bite. Using them together does not add up in any straightforward way.
Expectations are part of the machine. A rate move that nobody believes will last barely moves long rates. A spending commitment nobody believes will be honoured barely changes anyone's plans. This is the part of the mechanism that behaves least like a lever and most like a negotiation.
Check
In most countries the monetary lever is held by an institution deliberately insulated from the electoral cycle, and the fiscal lever is held by people facing one. That arrangement is itself an argument someone won.
The case for it is a time-consistency problem: the short-run reward for cheap credit arrives before the election and the cost arrives after, which is a structure that reliably produces a particular decision. Handing the lever to someone with a long term and a narrow mandate removes that temptation.
The case against it is that the choice between, for instance, tolerating higher unemployment and tolerating faster price rises is a distributional judgement about who bears a cost, and there is a reasonable objection to unelected officials making it. Both cases are serious. The arrangement is a compromise between them rather than a technical necessity, and it differs by country.