This thesis examines four issues of relevance to the use of macroeconomic models at central banks. (1) Reconciling Microeconomic and Macroeconomic Estimates of Price Stickiness (With Adam Cagliarini and Allen Tran) We attempt to reconcile the high estimates of price stickiness from macroeconomic estimates of New-Keynesian Phillips Curves (NKPC) with the lower values obtained from surveys of firms' pricing behaviour, which also suggest that the frequency of price adjustment varies across sectors. Building on Carvalho (2006), we present Monte Carlo evidence showing that if heterogeneity exists estimates of the NKPC obtained using GMM overstate aggregate price stickiness and, if roundabout production is present, falsely find indexation of prices. (2) Estimating and Identifying Empirical BVAR-DSGE Models for Small-Open Economies Different modelling approaches place varying weight on theory and data. Del Negro and Schorfheide (2004) present a comprise between theory rich Dynamic Stochastic General Equilibrium (DSGE) models and Vector Autoregressions (VARs), by using the DSGE as a prior for a Bayesian VAR (BVAR). However, in their approach a small economy cannot be restricted from impacting on a large economy. I develop a method which can, and show how the VAR may be identified. (3) Empirical BVAR-DSGE Forecasts of the Australian Economy (With Sean Langcake) We use a BVAR-DSGE model to forecast the Australian economy, treating the world as observed and exogenous. Reflecting the importance of mining, we develop a DSGE model with a simple commodity sector. We find that the BVAR generally forecasts more accurately than the DSGE, but does not outperform a Minnesota VAR or simple univariate models. (4) Assessing Some Models of the Impact of Financial Stress Upon Business Cycles (With Adrian Pagan) After the global financial crisis much research has focussed on integrating financial factors into macroeconomic models. Two common methods include the financial accelerator and collateralized lending, for example Gilchrist et al (2009) and Iacoviello (2005). This paper proposes evaluating such methods by focussing on their implications for business cycle characteristics and whether stylised facts about the impact of financial conditions can be matched. We find that in the Gilchrist et al (2009) model financial factors may impact on particular cycles, but the average cycle is little changed. Some, but not all, of the stylised facts are captured.