The magnitude and persistence of growth in gross domestic product are topics of intense scrutiny by economists. Although the existing techniques provide a range of tools to study the nature of growth and volatility time series, these usually come with shortcomings, including the need to arbitrarily define acceleration spells, and focus on a particular frequency at a time. This paper explores the application of "wavelet-based" techniques to study the time-varying nature of growth and volatility. These techniques lend themselves to a more robust analysis of short-term and long-term determinants of growth and volatility than the traditional decomposition techniques, as demonstrated on a small sample of countries. In addition to having desirable technical advantages, such as localization in time and frequency and the ability to work with non-stationary series, these techniques also make it possible to accurately decompose the association between growth trajectories of different countries over different time horizons. Such "co-movement" analysis can provide policy makers with important insights on regional integration, growth poles, and how short and long term developments in other countries affect their domestic economy.