where the t vectors represent i.i.d. random variables, and D denotes
dummy variables, which indicate the respective position
of revenues relative to their long-run equilibrium. These dummies
will take on a zero value if revenues are below their steady
state level and a value of one else. Additionally, we include several
covariates in the estimation of Eqns. (5)–(8) in order to isolate
a cleaner effect of income elasticity. These other control
variables, i.e., variables besides changes in tax rates and tax
policy dummies, are (i) contemporaneous public expenditures,
(ii) public expenditures lagged by one quarter, (iii) election year
dummies, indicating each quarter of a respective election
year, 13 and (iv) three dummy variables marking different inflation
regimes defined as intermediate inflation episodes for rates
between 5 and 10%, high inflation episodes for the range from
10% to 20%, and finally hyperinflation periods indicating quarters,
for which the inflation rate exceeds 20%. 14