Economics - Theses
Permanent URI for this collection
Search Results
1 - 1 of 1
-
ItemStudies in business cycles and macroeconomics( 2018)This thesis comprises three self-contained chapters. Each chapter is linked through the common motivation of understanding fluctuations in aggregate economic activity. Sectors with higher separation rates have a larger response in their vacancy yield when aggregate vacancies vary. Chapter 2 constructs series of labour search models to investigate the link between separation rates and vacancy yield. The analysis shows that the separation rate does not affect the vacancy yield elasticity. In turn, productivity and vacancy creation costs, which are correlated with the separation rate, are considered. It is shown that these two variables can account for the observed relationship between the vacancy yield elasticity and separation rates. Aggregate macroeconomic series display asymmetries in different phases of the business cycle. Peaks are sharp and quick, while troughs are protracted and relatively flat. The past literature has attributed this asymmetry to firm level frictions. Chapter 3 takes a sectoral view, investigating sectoral turning points and sectoral growth rates around aggregate and sectoral turning point dates. It is shown that more sectors are contracting around the aggregate peak date than are expanding around the aggregate trough date. Furthermore, sectors do not display asymmetry in their growth rates around their own turning points dates. These two findings go some way to explaining the observed asymmetry around aggregate turning point dates. The Chapter also shows that input-output linkages play an important role in determining the timing of an industry's turning point dates. Gross Domestic Product (GDP) in Australia is published two months after the end of the reference quarter. This delay presents a significant challenge for policy makers who must make decisions in real-time. Chapter 4 assesses the ability of different data-sets to now-cast Australian GDP, and anticipate discrete economic events. Particular attention is paid to business and household surveys, and how these datasets fair against a benchmark auto-regressive (AR) model and a factor augmented vector-auto-regression (FVAR) model. The results show that the FVAR model generates the most accurate point now-casts, however, the survey models are still useful when they signal GDP growth rates that are less frequent.