This dissertation is a collection of three self-contained essays that analyse the effects of monetary and fiscal policy using novel time series econometric methods. The research questions become important both academically and for practical policy-making after the global financial crisis of the early 2000s. The main contribution of the dissertation is in applying recent time series econometric methods to highly topical policy questions. Structural vector autoregressive (SVAR) models are an important tool in the empirical analysis of monetary and fiscal policy. The difficulty with conventional SVARs is the identification of structural shocks of interest needed for meaningful impulse response analysis. In this thesis I opt for a fairly novel approach to identify economically interpretable shocks which are then used to assess the macroeconomic effects of various economic policies. The so-called statistical identification consists of exploiting the statistical properties of the error processes, such as non-normality, to identify the structural model. In the three research chapters statistical, data-based information is combined with information from other sources. The assessment of economic policy can then be based on impulse response functions that are both economically meaningful and compatible with the sample data. In Chapter 2 I study the macroeconomic effects of the risk taking channel of monetary policy. The methodological improvement makes previously used identifying restrictions statistically testable and confirms that the balance sheet management of financial intermediaries led to a lower price of risk and higher real activity in the US before the financial crisis. Chapter 3 addresses the question whether increasing government spending stimulates real activity in the US. Unlike previous empirical research using SVARs I estimate a vector error correction model (VEC) that takes into account cointegration between the variables and use non-normality of the error processes for identification. The results show that when the empirical literature does not seem to reach a conclusion – in this case with respect to the sign or size of the fiscal multiplier – the identification strategy could play a role. In Chapter 4 I analyse the macroeconomic effects of the Bank of Japan’s, the Federal Reserve’s and the European Central Bank’s unconventional monetary policies. The use of a novel Bayesian SVAR method allows basing the whole analysis on the data and provides a formal way to assess the plausibility of given sign restrictions against the data. The analysis reveals differences in the output and price effects of the three central banks’ balance sheet operations.
|Award date||25 Aug 2017|
|Place of Publication||Helsinki|
|Publication status||Published - 25 Aug 2017|
|MoE publication type||G5 Doctoral dissertation (article)|
Fields of Science
- 511 Economics