Same, But Different: Testing Government Expenditure Shock Measures
Abstract
In this study we test whether different government expenditure innovation measures
proposed in the literature are good proxies for government spending shocks. We compare
five government spending shock measures: forecast error of survey of professional forecasters
(SPF), forecast error of forecast of Federal Reserve published in Greenbook (GB), military
news, Fisher-Peters measure, Ben Zeev-Pappa measure. To compare these measures, we
utilize the reliability estimator of a proxy, R², and F statistics which are proposed by [1].
We find that SPF and GB constitute the best proxy variable for government spending shocks
with regard to different reliability estimators. SPF and military news proxies have the highest
multipliers comparing to the others. Their both multipliers are greater than 1. Multiplier
for Ben Zeev-Pappa and Fisher-Peters measures are less than zero. GB also has a positive
multiplier but less than one.
Citation
Hassani, Ashkan (2021). Same, But Different: Testing Government Expenditure Shock Measures. Master's thesis, Texas A&M University. Available electronically from https : / /hdl .handle .net /1969 .1 /196422.