This paper by Danny Quah examines the empirical dynamics of economic growth across countries, regions, and states. The standard approach in these studies involves calculating average growth rates and regressing them on conditioning variables such as schooling, investment, government spending, and initial income levels. However, this approach assumes that each economy has a steady-state growth path, which is not supported by the data. The first part of the paper tests this assumption using cross-country income data from 118 countries. The results show that the underlying long-run growth patterns within each country are unstable, and income variability has increased over time. This suggests that significant economic disturbances are ongoing, and the traditional view of economies adjusting towards a steady state equilibrium is flawed. The second part of the paper introduces an alternative econometric strategy that does not rely on restrictive assumptions about long-run growth. It models the evolution of income distributions over time using a Markov chain approach. The analysis reveals that cross-country incomes tend towards extremes at both high and low ends, with a higher probability of convergence towards the richest countries. The paper also finds that growth rates show high cross-section mobility and little persistence. The conclusions highlight the need for more refined models to capture the dynamics of economic growth, particularly in the context of evolving income distributions. The paper suggests that future research should explore more interpretable structures and conditioning information, and develop models that can better handle data with both time-series and cross-sectional dimensions.This paper by Danny Quah examines the empirical dynamics of economic growth across countries, regions, and states. The standard approach in these studies involves calculating average growth rates and regressing them on conditioning variables such as schooling, investment, government spending, and initial income levels. However, this approach assumes that each economy has a steady-state growth path, which is not supported by the data. The first part of the paper tests this assumption using cross-country income data from 118 countries. The results show that the underlying long-run growth patterns within each country are unstable, and income variability has increased over time. This suggests that significant economic disturbances are ongoing, and the traditional view of economies adjusting towards a steady state equilibrium is flawed. The second part of the paper introduces an alternative econometric strategy that does not rely on restrictive assumptions about long-run growth. It models the evolution of income distributions over time using a Markov chain approach. The analysis reveals that cross-country incomes tend towards extremes at both high and low ends, with a higher probability of convergence towards the richest countries. The paper also finds that growth rates show high cross-section mobility and little persistence. The conclusions highlight the need for more refined models to capture the dynamics of economic growth, particularly in the context of evolving income distributions. The paper suggests that future research should explore more interpretable structures and conditioning information, and develop models that can better handle data with both time-series and cross-sectional dimensions.