Pakistan Monthly Index

Download Data                                   Annotated Chart

We are pleased to host two new EPU indices for Pakistan developed by M. Ali Choudhary, Farooq Pasha and Mohsin Waheed in "Measuring Economic Policy Uncertainty in Pakistan".

Their indices reflect frequency counts of articles published in four leading English language Pakistani newspapers, Business Recorder, Express Tribune, Dawn and The News, from January 2015 to the present, while a longer version from August 2010 to present involves two of the four newspaper outlets (Business Recorder and Express Tribune). They follow the index construction methods in "Measuring Economic Policy Uncertainty" by Scott R. Baker, Nicholas Bloom and Steven J. Davis.

Pakistan is a small open economy exposed to a large number of economic-, non-economic- and political shocks. Thus, it important to study sources of economic policy uncertainty for the country as far back in time as possible. As a result, Choudhary et al. construct two indices: one that considers articles pertaining to economic policy uncertainty with two newspapers whereby web-harvesting is available as early as August 2010, and another that considers two additional newspapers with data from January 2015 onwards. They label the first index as EPU-2 and the second as EPU-4 in their annotated chart

To construct their EPU index, Choudhary et al. first obtain monthly counts of articles that contain terms pertaining to uncertainty (or uncertain, unpredictable, unclear, unstable), and economics (or economy) and one or more of the following policy-relevant terms: regulation, monetary policy, fiscal policy, central bank (or SBP), FBR (or tax authorities), policymakers, parliament, deficit, government, reserves, taxes, tariffs, legislation. After obtaining these raw counts, they scale by the number of articles published in the same newspaper and month. They standardize each newspaper's scaled frequency counts to have a unit standard deviation from August 2010 to April 2020 (for EPU-2) and from January 2015 to April 2020 (EPU-4), and then compute the simple average across newspapers by month. Finally, they multiplicatively normalize the series to have a mean of 100 for respective time periods of EPU-2 and EPU-4.