The choice of decomposition methodology has important implications for how we view complex phenomenon. In this talk, Christopher Magee will first examine different modes for decomposing technological change (design viewpoint, types of technology, types of R&D, essential layer elements) discussing how each mode of decomposition influences our understanding of the phenomenon. His overall viewpoint on technological change will then be discussed as an introduction to a model of technological change that is consistent with this viewpoint and with known empirical facts about technological change. Implications and uses of the model will conclude the seminar.

They document asymmetry of the predictive distribution of GDP growth as a robust feature of the data, with negative skewness significantly increasing during recessions. When they include financial condition indices as additional drivers in the updating process of the time-varying parameters, they observe substantial improvements in the out of sample predictive ability for different horizons. This finding reflects the model's ability to pick up changes in the shape of the forecast densities in a timely manner. They find that episodes of financial tightening are robust drivers of the left skewness of the predictive distribution, ultimately sharpening economic growth predictions at the onset of recessions.

About the speaker

Professor Christopher Magee (PhD, MS, MBA, BS) has been with MIT for 18 years but earlier spent twice that time at Ford Motor Company first doing fundamental scientific research, and then doing and managing various other aspects of the spectrum of activities involved in technological change (applied research, product development, systems architecting, merged applied/fundamental research, computational engineering, manufacturing development, etc.). Most of his research at MIT has involved measuring, modeling and predicting technological change and its social impact.

All welcome. Please register on Eventbrite or contact for more information.


Research Themes

Research Programmes