This paper analyzes the dependency of Government Spending Multipliers (GSM) for US economy on the Business Cycle, both empirically and theoretically. The empirical model based on Local Projection showed a non-monotonic relation between GSM and the Business Cylce, with the lowest multipliers occurring in periods of slack, and the highest occurring in normal times (whenever the GDP is around the Steady-State). Expansionary periods are related to intermediate multipliers. For understanding the transmission mechanisms behind the variability of GSM, a microeconomic model has been developed. Following the piioneristic contribution of the "Lambda-Model" of Cambpell and Mankiw, a "Smooth Transition Lambda-Model" has been implemented. From the model arised a State-Dependency for both the EIS (Elasticity of intertemporal Substitution) and the Propensity to Consume (PC), in particular showing a non zero EIS only in periods of real GDP contraction and a non-monotonic dynamic for the PC.
These last results have been integrated in a TANK (Two-Agents New-Keynesian) Model, in order to study the macroeconomic impact of the public spending innovation, and to reconcile theory and evidence.
The research is innovative for two class o motives, the one related to the spending multipliers literature, and the other one related to the more general consumption function literature.
The most innovative element of the research is the finding of a " State-Dependnet Consumption Funcion".
The traditional consumption function the I considered as the starting point of the research is the one proposed by Cambpell and Mankiw (CM) in the so called "Lambda-Model".
In the model of CM consumption function is twofold. The two componnt are the keynesian one and the ricardian one. The ricardian component is the one that links consumption variations to the real interest rate, and it is a consequence of the intertemporal maximization process (Consumption Smoothing). The keynesian component is the one that links consumption variation with income variations. What they found is a non statistical significance of the ricardian component and a strong significance for the keynesian one.
Starting from CM model, I developed a State-Dependent Consumption Function, conditioning both the components to the state of the business cycle.
The results can be summarized in the following points:
- Both the component of consumption changes across the business cylce
- The ricardian component is statistically differen from zero only during contraction periods
-The keynesian component has a non-monotonic dynamic with respect to the business cycle
These results are innovative because of the statistical significance of the model first: the inclusion of the business cycle as interaction term is important to improve the efficiency and significance of the Cambpell-Mankiw model.
The inclusion of the interaction term, on one hand stresses the importance of the keynesian component of consumption, while on the other hand underlines the role of the Elasticity of Intertemporal Substitution (ricardian component) during contractions.
These findings altogheter show how precautionary savings and the risk aversion of households increase during periods of contractions of the real GDP. It is a partial rehabilitation of the Consumption Smoothing Hypotesis, that can be resumed into a " Asymmetric Consumption Smoothing".
This result is important as previously introduced for two order of reasons: the first one is the implication for spending multiplier analysis, and the second one is more general and applies to both positive and normative aspects of the economic analysis.
About the spending multipliers, the Asymmetric Consumption Smoothing finding, can explain theoretically why multipliers differ across the business cycle, and particularly how the variablity of the Elasticity of Intertemporal Substitution and of the Propensity to Cosnumption can imply a lower multiplier during periods of GDP contraction and higher multipliers in normal times. The tranmission mechanims, arise by using a DSGE TANK model. The variabilty of the multipliers in the theoretical model are a consequence of three components interaction: the intertemporal IS function, the Marginal Rate of Substitution between consumption and labor, the proportion of Rule of Thumb Households.
The usefullness of the Asymmetric Consumption Smoothing in general can be summarized in the following points:
- a positive analysis of the Endogenous Business Cycle
- a positive analysis of the macroeconomic state dependent response of the economy to exogenous shocks
- a normative analysis of fiscal and monetary policy
-a normative analyis of the variability of optimal policies (monetary and fiscal) across the business cycle.