A Dollar of Government Spending Does Not Produce a Dollar of Output
It produces more in some conditions and less in others. The fiscal multiplier is the ratio, and the fact that it varies so much is why fiscal policy arguments never resolve.
The Basic Idea
When a government spends, the money goes to someone who spends part of it, and the recipient of that spends part again. Each round adds to output, and the total effect can exceed the original amount.
The fiscal multiplier is the ratio of the change in output to the change in spending. A multiplier of 1.5 means a dollar of spending produced a dollar fifty of output.
Why It Is Not Always Above One
The circulation story is incomplete because it ignores what the spending displaces.
If the economy is at capacity, government spending competes for resources already in use. It bids up prices and wages rather than adding output, and it may push interest rates up, reducing private investment. That displacement is called crowding out, and where it is complete the multiplier approaches zero.
The multiplier is not a property of spending. It is a property of the conditions the spending happens in.
What Determines It
| Condition | Multiplier |
|---|---|
| Economy with substantial slack | Higher |
| Economy at capacity | Lower |
| Interest rates at their floor | Higher |
| Central bank offsetting the stimulus | Near zero |
| Closed or large economy | Higher |
| Small open economy | Lower, spending leaks to imports |
The central bank reaction is the most underappreciated. If a central bank is targeting inflation and raises rates in response to fiscal stimulus, it deliberately offsets the effect, and the multiplier can be close to zero regardless of anything else.
Conversely, when policy rates are at their lower bound and cannot fall further, the central bank does not offset, and multipliers are much larger. This is why the estimated multiplier varies so much across studies looking at different periods, and why both sides of the argument can cite credible evidence.
Composition Matters Too
Different kinds of fiscal action have different multipliers.
Spending directed at people who will spend it, such as unemployment benefits and support for low income households, has high multipliers because the recipients have little saving buffer and spend nearly all of it.
Tax cuts for higher earners have lower multipliers, because a larger share is saved rather than spent. Direct government purchases have relatively high multipliers because the entire amount enters the economy immediately, with no decision by a recipient about whether to spend it.
The Timing Problem
Multipliers are estimated over specific horizons and effects differ across them. Infrastructure spending has a low multiplier in the first year, because projects take time to start, and a higher one later, both from the activity and from the productive asset created.
That mismatch is a practical problem. Recessions are identified with a lag, legislation takes time, and spending takes longer still, so stimulus frequently arrives as the economy is already recovering, which is when the multiplier is lower.
Why Estimates Disagree So Much
Published estimates range widely, and the honest reading is that they are measuring different things. A study of an economy at the lower bound with substantial slack and one of an economy at capacity with an active central bank should produce different answers, because the underlying quantity genuinely differs.
Treating the multiplier as a single number to be discovered is the error. The useful question is always what it is likely to be under current conditions.
The Bottom Line
The fiscal multiplier depends on slack, on what the central bank does in response, on how open the economy is, and on who receives the money. It is well above one in a slump with rates at their floor and close to zero in a fully employed economy with a central bank determined to offset. Arguments citing a single figure are almost always describing conditions rather than a constant.