A Multivariate Time Series Analysis of Secular Stagnation: A Critical Realist Perspective

The study investigated whether excess saving over investment explains secular stagnation or is linked to structural break (hysteresis) in aggregate economic activity as measured by the changes in the level of GDP. The U.S. economy, including the advanced economies, is experiencing secular stagnation, potentially caused by excess savings over investment. This condition, if present, could have significant implications for how we can conduct economic policy. Instead of focusing on gross domestic product (GDP), a measure of economic activity, the study extends the existing literature by concentrating on aggregate private-sector output. The rationale is that if there is secular stagnation or hysteresis in GDP, we would expect similar observations in private-sector production, given its role as the main engine of aggregate economic activity. The methodology combines the critical realist retroduction mode of inference with time series econometric methods—the vector error correction model and autoregression distributed lag. In addition, the study examines whether there is evidence of structural break (hysteresis) in the level trends in some of the key macroeconomic variables linked to aggregate demand induced secular using the Clemente, Montanes, and Reyes' (1998) univariate method. The Gregory-Hansen structural break tests were used to examine whether the relationship had a structural break and to determine if the economy experienced hysteresis following the Great Recession. The econometric results do not show a systematic long-run relationship between excess saving, private-sector output, GDP, and the GDP gap. The relationship only holds in the short run. Furthermore, the observed structural break in private-sector production occurred earlier than 2007, the official date for the beginning of the recession.