The Multiplier — The Cycle, Plate iii

If Say's Law fails — if demand can fall short of what the economy produces — then the level of spending is not a passive consequence of production but an active force governing how much gets produced and how many are employed. The first tool for seeing how that force works is the multiplier, which Richard Kahn made precise in 1931 and which Keynes placed at the heart of his theory of the cycle. The multiplier reveals that a change in spending does not act on the economy once and stop; it ripples outward, multiplied, so that demand commands a leverage over output and employment that the classical economists, with their eyes fixed on supply, had never reckoned. A dollar of new spending is not a dollar of new income but several; and a dollar withdrawn is several lost.

The mechanism is a chain. Suppose some new spending enters the economy — a firm builds a factory, a government funds a public work. The money paid out becomes income in the hands of the workers and suppliers who receive it. They do not hoard it all; they spend a fraction on goods and services, saving the rest. That fraction becomes income to the shopkeepers and producers they buy from, who in turn spend a fraction of it, which becomes income to others, who spend a fraction again — a diminishing series, each round smaller than the last as a slice leaks into saving at every step. The total addition to the economy's income is the sum of this whole series, which is a multiple of the original injection. How large a multiple depends on how much of each round of new income is passed along in spending rather than held back in saving — what is called the marginal propensity to consume. Where people spend a large share of new income, the chain is long and the multiplier large; where they save much of it, the chain is short and the multiplier small. The single act of spending swells into a sum far greater than itself.¹

1. The mechanism is a chain. New spending becomes income to those who receive it; they spend a fraction (saving the rest), which becomes income to others, who spend a fraction, and so on — a diminishing series, a slice leaking into saving each round. The total addition to income is the sum of the series, a multiple of the original injection. The multiple depends on how much of each round is spent (the marginal propensity to consume): spend a large share and the chain is long, the multiplier large; save much and it is short, the multiplier small.

What multiplies upward multiplies downward, and this is half of why the cycle is so violent. The very same chain that amplifies an injection of spending amplifies a withdrawal. If spending falls — investment collapses, orders are cancelled, a financial fright makes everyone hold back — the spending that vanishes was someone's income, and that someone, poorer, now spends less, depriving another of income, who spends less still, the contraction rippling and multiplying exactly as the expansion did, in reverse. So a modest initial shock — a drop in investment, a panic in the markets — can produce a large fall in total income and employment, magnified through the economy by the multiplier running backward. The economy, in this respect, has no shock-absorber; it has an amplifier, which turns small disturbances into large swings in both directions. The boom feeds on itself as each dollar of new spending calls forth more; the slump feeds on itself as each dollar withdrawn calls forth less.²

2. What multiplies up multiplies down. The same chain amplifies a withdrawal: spending that vanishes was someone's income; poorer, they spend less, depriving another, who spends less still — the contraction rippling and multiplying in reverse. A modest shock (a drop in investment, a panic) can produce a large fall in income and employment. The economy has no shock-absorber but an amplifier, turning small disturbances into large swings both ways. The boom feeds on itself; so does the slump.

⁂

From the multiplier follows a result that overturns the plainest common sense: the paradox of thrift. For a single household, saving more is simple prudence — spend less than you earn, lay money by, and grow more secure and, in time, richer. But let every household try to save more at once, and the result is perverse. The extra saving is, by definition, less spending; the less spending is, by the multiplier, less income for others; the less income means those others can save less than they had hoped, and indeed must cut their own spending, costing income to still others. When the dust settles, the economy's income has fallen, employment with it, and the total amount saved may be no greater than before — may even be less — for saving comes out of income, and income has shrunk. The attempt of all to save more has made everyone poorer while adding nothing to the pool of saving. What is virtuous for the one is ruinous for the all. The household that tightens its belt in hard times is prudent; the nation that does the same, every household at once, deepens the very slump it fears.³

3. The paradox of thrift. For one household, saving more is prudence. But if every household saves more at once: the extra saving is less spending; less spending is (via the multiplier) less income; less income means others save less than hoped and cut their own spending. Income falls, employment with it, and total saving may be no greater — may be less — for saving comes from income, which has shrunk. The attempt of all to save more makes everyone poorer while adding nothing to saving. What is virtuous for one is ruinous for all.

The paradox of thrift is one instance of a deeper principle, and naming it explains why macroeconomics had to become a science of its own: the fallacy of composition, the error of supposing that what holds for a part must hold for the whole. In the economics of the individual and the single market, the counsels of prudence are sound — save, economize, cut your costs, accept a lower price to clear your stock. In the economics of the whole, these can reverse: if everyone economizes together, the demand that is everyone's income collapses, and all are made worse. The whole is not the part enlarged; the aggregate economy obeys laws the household does not, because in the aggregate one person's spending is another's income, so that the prudence of each subtracts from the income of all. This is the deepest reason that reasoning from the household to the nation — "a family must live within its means in hard times, and so must the country" — is not homely wisdom but precisely the fallacy that turns a downturn into a depression. The economy in a slump may need someone to spend more, not less; and that is something no prudent individual, looking only to his own budget, can be expected to do.⁴

4. The deeper principle: the fallacy of composition — assuming what holds for a part holds for the whole. For the individual, prudence is sound (save, economize, cut costs, accept a lower price). For the whole, these reverse: if all economize together, the demand that is everyone's income collapses. The aggregate obeys laws the household does not, because one person's spending is another's income. So "a family must live within its means in hard times, and so must the country" is not wisdom but the fallacy that turns a downturn into a depression. This is why macroeconomics is a separate science.

The multiplier and the paradox of thrift together recast what a slump is and how it might be cured. A slump is not mended by universal economy — that only deepens it; it is mended by someone spending enough to arrest the downward chain and set the upward one going, the restored income restoring the spending that restores employment. This points toward the role of a government able to spend when no one else will, a theme a later plate takes up; and it points back to a prior question the multiplier itself does not answer. For the multiplier explains how a shock, once it strikes, is amplified into a swing — but not what the shock is, nor why the most volatile part of spending, investment, lurches up and down to set the cycle in motion. Two answers follow. One locates the lurch in the financial system, which builds fragility in the boom until it cracks; the other locates it in money, whose quantity can collapse and drag spending down with it. The next plate is the first of these: the mania.⁵

5. The multiplier and paradox of thrift recast the slump: not mended by universal economy (which deepens it) but by someone spending enough to arrest the downward chain and start the upward one. This points to a government able to spend when no one else will (a later plate), and back to a question the multiplier does not answer: what is the shock, and why does investment lurch to set the cycle going? Two answers: the financial system building fragility in the boom (the mania), and money whose quantity can collapse (the contraction). Next: the mania.

The multiplier is the engine that gives demand its leverage over the economy. A change in spending ripples outward, each round becoming income that is partly spent again, so that the total change in income is a multiple of the first — amplifying booms as it runs forward and deepening slumps as it runs back, turning small shocks into large swings. From it comes the paradox of thrift: that if all try to save more at once, spending falls, income falls, and saving itself may not rise, so that what is prudent for the household is ruinous for the nation. This is the fallacy of composition, and it is the reason the economy as a whole is a study apart from its parts — and the reason the household's instinct to economize in hard times is exactly the wrong counsel for the economy entire. The multiplier shows how shocks are magnified; the next plates ask where they come from. The next plate is the mania.