What is the strongest evidence for the reduced-form model that
links money supply growth to inflation?
What does the AS-AD model, a structural model, say about money
supply growth and the price level?
Milton Friedman claimed that “inflation is always and everywhere
a monetary phenomenon.”A Monetary History of the United States,
1867–1960. We know this isn’t true if one takes a loose view
of inflation because negative aggregate supply shocks and increases
in aggregate demand due to fiscal stimulus can also cause the price
level to increase. Large, sustained increases in the price level,
however, are indeed proximately caused by increases in the money
supply and only by increases in the money supply. The evidence
for this is overwhelming: all periods of hyperinflation from the
American and French Revolutions to the German hyperinflation
following World War I, to more recent episodes in Latin America and
Zimbabwe, have been accompanied by high rates of money supply (MS)
growth.In most of those instances, the government printed
money in order to finance large budget deficits. The rebel
American, French, and Confederate (Southern) governments could not
raise enough in taxes or by borrowing to fund their wars, the
Germans could not pay off the heavy reparations imposed on them
after World War I, and so forth. We know that the deficits
themselves did not cause inflation, however, because in some
instances governments have dealt with their budget problems in
other ways without sparking inflation, and in some instances rapid
money creation was not due to seriously unbalanced budgets. So the
proximate cause of inflation is rapid money growth, which often,
but not always, is caused by budget deficits. Moreover, the MS
increases in some circumstances were exogenous, so those episodes
were natural experiments that give us confidence that the
reduced-form model correctly considers money supply as the causal
agent and that reverse causation or omitted variables are
unlikely.
Stop and Think Box
During the American Civil War, the
Confederate States of America (CSA, or the South) issued more than
$1 billion of fiat paper currency similar to today’s Federal
Reserve notes, far more than the economy could support at the
prewar price level. Confederate dollars fell in value from 82.7
cents in specie in 1862 to 29.0 cents in 1863, to 1.7 cents in
1865, a level of currency depreciation (inflation) that some
economists think was simply too high to be accounted for by
Confederate money supply growth alone. What other factors may have
been at play? (Hint: Over the course of the war, the Union
[the North] imposed a blockade of southern trade that increased in
efficiency during the course of the war, especially as major
Confederate seaports like New Orleans and Norfolk fell under
northern control.)
A negative supply shock, the almost
complete cutoff of foreign trade, could well have hit poor Johnny
Reb (the South) as well. That would have decreased output and
driven prices higher, prices already raised to lofty heights by
continual emissions of too much money.
Economists also have a structural model showing a causal link
between money supply growth and inflation at their disposal, the
AS-AD model. Recall that an increase in MS causes the AD curve
to shift right. That, in turn, causes the short-term AS curve to
shift left, leading to a return to Ynrl but higher
prices. If the MS grows and grows, prices will go up and up,
as in Figure 25.1 .
Nothing else, it turns out, can keep prices rising, rising,
ever rising like that because other variables are bounded. An
increase in government expenditure G will also cause AD to shift
right and AS to shift left, leaving the economy with the same
output but higher prices in the long run (whatever that is). But if
G stops growing, as it must, then P* stops rising and inflation
(the change in P*) goes to zero. Ditto with tax cuts, which can’t
fall below zero (or even get close to it). So fiscal policy alone
can’t create a sustained rise in prices. (Or a sustained decrease
either.)
Negative supply shocks are also one-off events, not the
stuff of sustained increases in prices. An oil embargo or a
wage push will cause the price level to increase (and output to
fall, ouch!) and negative shocks may even follow each other in
rapid succession. But once the AS curve is done shifting, that’s
it—P* stays put. Moreover, if Y* falls below Ynrl, in
the long run (again, whatever that is), increased unemployment and
other slack in the economy will cause AS to shift back to the
right, restoring both output and the former price level!
So, again, Friedman was right: inflation, in the sense of
continual increases in prices, is always a monetary phenomenon and
only a monetary phenomenon.This is not to say, however, that
negative demand shocks might not contribute to a general monetary
inflation.
Stop and Think Box
Figure 25.2 compares inflation with M1
growth lagged two years. What does the data tell
you? Now look at Figure 25.3 and Figure 25.4 . What caused M1 to
grow during the 1960s?
The data clearly show that M1 was
growing over the period and likely causing inflation with a
two-year lag. M1 grew partly because federal deficits increased
faster than the economy, increasing the debt-to-GDP ratio,
eventually leading to some debt monetization on the part of the
Fed. Also, unemployment rates fell considerably below the natural
rate of unemployment, suggesting that demand-pull inflation was
taking place as well.
key takeaways
Throughout history, exogenous increases in MS have led to
increases in P*. Every hyperinflation has been preceded by rapid
increases in money supply growth.
The AS-AD model shows that money supply growth is the only
thing that can lead to inflation, that is, sustained increases in
the price level.
This happens because monetary stimulus in the short term shifts
the AD curve to the right, increasing prices but also rendering Y*
> Ynrl.
Unemployment drops, driving up wages, which shifts the AS curve
to the left, Y* back to Ynrl, and P* yet higher.
Unlike other variables, the MS can continue to grow, initiating
round after round of this dynamic.
Other variables are bounded and produce only one-off changes in
P*.
A negative supply shock or wage push, for instance, increases
the price level once, but then price increases stop.
Similarly, increases in government expenditures can cause P* to
rise by shifting AD to the right, but unlike increases in the MS,
government expenditures can increase only so far politically and
practically (to 100 percent of GDP).