Is British output growth related to its uncertainty? Evidence using eight centuries of data

Abstract

We examine the empirical relationship between output variability and output growth for Britain using data for eight centuries covering the 1270 to 2014 period. Drawing on the economic history literature, we split the full sample period in four subperiods and use GARCH models to measure output growth uncertainty and estimate its effect on average growth. Within each sub-sample we allow output growth to depend on the state of the system, e.g. 2-regime switching model would switch between high-growth and low-growth regimes. We find that the effect of uncertainty on growth differs depending on the existing growth regime. Low-growth regimes are associated with a negative effect of uncertainty on growth, and medium or high-growth regimes are associated with a positive effect. These findings are consistent across the four states of economic development. Our results indicate why the empirical literature to date has found mixed results when examining the effect of uncertainty on growth.

Explaining the cross-country differences in the economic fallout during the COVID-19 pandemic crisis

“…Using 50 years of data on the G7, Neanidis and Savva (2013) find that uncertainty regarding output growth is related with a higher average growth rate mostly in a lowgrowth regime. Conversely, Bredin et al (2021) exploit a centuries-long sample of UK data and report that low-growth regimes are associated with a negative effect of uncertainty on growth, while medium-or high-growth regimes are associated with a positive effect. Christou et al (2020) also study 150 years of UK data and find a negative effect, the strength of which exhibits substantial time-variation.…”

Uncertainty and forecastability of regional output growth in the UK: Evidence from machine learning

“…First, random forests can accurately and flexibly analyze the links between regional GVA growth and a large number of predictors in a full-fledged data-driven manner. Second, random forests automatically capture potential nonlinear links between output growth and its predictors, including uncertainty, as shown to exist historically for the UK by Christou et al (2020b) and Bredin et al (2021), 2 as well as any interaction effects between the predictors.…”

“…To sum up, there exist competing effects of uncertainty on economic activity, and the overall effect would be contingent on the relative strength of these effects, though it is generally expected that the real option theory will dominate, and so uncertainty, on balance, has a negative impact on output, as observed during the global financial crisis and the recent outbreak of the COVID-19 pandemic. Christou et al (2020b) and Bredin et al (2021), 2 as well as any interaction effects between the predictors.…”