This book explores how to set up an empirical model that helps with forecasting long-term economic growth in a large number of countries. It offers a systematic approach to models of potential GDP that can also be used for forecasts of more than a decade. It is an attempt to fill the wide gap between the high demand for such models by commercial banks, international organizations, central banks and governments on the one hand and the limited supply on the other hand. Frequent forecast failures in the past (e.g. Japan 1990, Asia 1997) and the heavy economic losses they produced motivated the work. The book assesses the large number of different theories of economic growth, the drivers of economic growth, the available datasets and the empirical methods on offer. A preference is shown for evolutionary models and an augmented Kaldor model. The book uses non-stationary panel techniques to find pair-wise cointegration among GDP per capita and its main correlates such as physical capital, human capital and openness. GDP forecasts for the years 2006 to 2020 for 40 countries are derived in a transparent way.

The author works for a commercial bank and has been the lead researcher in the bank's project called "Global Growth Centres 2020".

Auteur

The author works for a commercial bank and has been the lead researcher in the bank's project called "Global Growth Centres 2020".

Contenu
The importance of long-run growth analysis.- Assessment of growth theories.- The dependent variable: GDP growth.- Labor input.- Physical capital.- Human capital.- Openness.- Spatial linkages.- Other determinants of GDP.- The theory of forecasting.- The evolution of growth empirics.- Estimation results.- Forecast competitions and 2006-2020 forecasts.- Conclusion and outlook.