Versión de HTML Básico
within the period comprised Mexico faced various economic recessions
the
model performed satisfactorily without the need to introduce dummy variables.
Regarding the literature that uses the time series econometric analysis, King &
Plosser (1984) decompose M1 in high powered money and deposits. The effect of
deposits in real output ranges from 0.740 to 0.784, while high powered money
exposes a coefficient of 0.510. For this authors, the period involved is from 1953 to
1978 on an annual basis for United States. As far as Koustas (1998) is concerned,
he rejects long-term neutrality having obtained positive values greater than 0.20,
comprising a time period from the first quarter of 1953 to the fourth of 1993
regarding the Canadian economy.
In estimating the effect of money in real output, Noriega, Soria & Velázquez (2002)
consider an array of countries in their analysis of money neutrality. Regarding
Mexico, they examine broken trends between M2 and output from 1932 to 2000,
concluding that it would be difficult to argue that money is neutral.
Further, Shelley & Wallace (2003) examine the Mexican case between 1932 and
2001, using real GDP and M2. Neutrality is rejected for the whole period. However,
considering that there is a break in the Mexican economy at the beginning of the
1980’s, the authors adjust for a change in the mean growth as from 1981. This is
because the Mexican economy grew consecutively for the first 50 years, as a
structural change is perceived between 1981 and the next year. With this last
adjustment, neutrality is not rejected in the long-term. In the case of Guatemala
26
That is to say, 1995:1-1995:4 (four months); 2001:2-2002:1 (four months) and
2008:4 - 2009:4 (five months), -considering GDP growth with a year lag.