• Fiscal Policy
• New Method
• Structural shock
• Uncertainty
• VAR

This thesis focuses on the estimation of uncertainty and its effects on macroeconomic activity. We use a new methodology which consists of three main passages. We firstly estimate a VAR finding out a structural shock. Using the just identified structural shock, we estimate the corresponding uncertainty. The latter is obtained by regressing the square of the structural shock on the same information set used in the previous VAR. Finally, we combine together the first two steps to get the uncertainty shock and its effects on the VAR variables. We show that this method is related to the instrumental variable
identification method proposed in Stock and Watson 2018 and Mertens and Ravn 2014. Precisely, the impulse response functions obtained with this method are asymptotically equivalent to the ones that can be obtained by using the IV method with the square of
the structural shock as the external instrument.
We use this method to study the effects of fiscal policy uncertainty. We identify a fiscal foresight shock using a Cholesky scheme and taking as a benchmark the survey of professional forecasters used in Forni and Gambetti 2016. Our sample includes US data spanning from January 1981 to July 2017.
It emerges the existence of a variable uncertainty. More specifically, when we regress the square of fiscal foresight shock on the past of VAR variables, in fact, the latter turns out to be linearly forecastable, even though - by construction- fiscal foresight shock is not.
We find that fiscal policy uncertainty has negative effects on economic activity and, more precisely, on real variables like GDP, consumption and investment.

This thesis focuses on the estimation of uncertainty and its effects on macroeconomic activity. We use a new methodology which consists of three main passages. We firstly estimate a VAR finding out a structural shock. Using the just identified structural shock, we estimate the corresponding uncertainty. The latter is obtained by regressing the square of the structural shock on the same information set used in the previous VAR. Finally, we combine together the first two steps to get the uncertainty shock and its effects on the VAR variables. We show that this method is related to the instrumental variable identification method proposed in Stock and Watson 2018 and Mertens and Ravn 2014. Precisely, the impulse response functions obtained with this method are asymptotically equivalent to the ones that can be obtained by using the IV method with the square of the structural shock as the external instrument. We use this method to study the effects of fiscal policy uncertainty. We identify a fiscal foresight shock using a Cholesky scheme and taking as a benchmark the survey of professional forecasters used in Forni and Gambetti 2016. Our sample includes US data spanning from January 1981 to July 2017. It emerges the existence of a variable uncertainty. More specifically, when we regress the square of fiscal foresight shock on the past of VAR variables, in fact, the latter turns out to be linearly forecastable, even though - by construction- fiscal foresight shock is not. We find that fiscal policy uncertainty has negative effects on economic activity and, more precisely, on real variables like GDP, consumption and investment.