This paper presents a “broken stick” method to test for structural breaks in a regression model. The method is illustrated using output data across the EU and the results are bootstrapped to identify statistical significance.