In empirical terms, there appears to be some evidence that could be read as being for
and against the FTPL. For instance, the well known cases of hyperinflation in Europe in
the beginning of the 20th century, discussed by Cagan (1956) and by Sargent (1982),
seem to indicate that the inflationary periods of Germany, Austria, Hungary and Poland
were only surpassed after the problems in terms of fiscal deficits had been delt with.
When studying the monetary characteristics of hyperinflations in Austria, Hungary,
Poland and Russia in the 20s and in Greece, Hungary and Russia in the 40s, Cagan
(1956) used models for money demand based on the strong relation existing between
money and inflation.9
Loyo (2000) explains the hyperinflation of Brazil in 1975-1985 as a result a of a vicious
circle between the nominal growth of government debt and the corresponding rollover
of existing debt. Also, Woodford (1999, pp. 393) maintains that fiscal policy in the US
might have been non-Ricardian before 1951: “That period would represent a historical
example of the kind of interest-rate pegging regime for which the theory was
developed.”10
In the first half of the 70s, the idea of a regime of fiscal dominance, Ricardian in
Woodford’s terminology, may be relevant, since there was a decline of fiscal balances
and an increase of inflation, namely in Italy and in the UK, see figures 1, 2 and 3. Also,
and as mentioned by Sargent (1987), the fact that the government budget constraint is
pratically absent from macroeconomic models in the 60s and 70s, may be an indication
that the budget constraint was then seen more like an equilibrium condition and less as
constraint.
In the beginning of the 80s there was an increase of budget deficits in several western
economies. According to the assumptions of the FTPL, there should have been a price
level rise in order to reduce the government liabilities in real terms. For instance, in
Italy, see Figure 2, there was a decline of inflation throughout the 80s, what seems to
disagree with the FTPL. This is even more visible in the case of Germany, see Figure 1.
9 A posteriori, the econometric problems of those models were pointed out namely by McCallum
(1989).
10 It was in March 1951 that was made the accord between the US Treasury and the Central Bank in
order to give the Central Bank more autonomy to change interest rates, Hetzel and Leach (2001)
give a reading of the implications of that agreement.
15