Economic policy uncertainty (EPU) has increased markedly in recent years in the U.S. and Europe, and some have posited a link between this phenomenon and subpar economic growth in advanced economies (see Baker, Bloom, and Davis, 2015). But methodological and data concerns have thus far raised doubts about whether EPU contains marginal and exogenous information about other economic phenomena. Our work analyzes the impact on EPU of several possibly endogenous variables, such as inflation and unemployment rates in countries where EPU is measured. We also consider longer-term technological factors, such as media fragmentation, which by undermining political consensus may indirectly elevate EPU. We find that about 40 percent of EPU movements can be explained by long-and short-run movements in these determinants, which is consistent with limited evidence that de-trended movements in EPU may contain marginal information about GDP growth and other macro variables.