Rare events in the american GDP time series, 1790 -present : Factor artefact?

Abstract

This paper studies the cyclical behaviour of the annual American Gross Domestic Product (GDP) series. We  show that rare events have varied effects, which give useful information on the nature and the amplitude of economic shocks. Our methodology is, more precisely, an efficient testing procedure to control the robustness of historical time series constructions.

KEYWORDS: Cliometrics, GDP, Historical Time Series, Outliers, USA.