Trade Policy Uncertainty

Here I am in silence
Looking 'round without a clue
--Information Society

Recently we discussed a measure of economic policy uncertainty developed by three high profile scholars. Now, a group of researchers at the Federal Reserve have developed a related measure more specific to regulatory issues related to trade called 'trade policy uncertainty (TPU).'

Their Trade Policy Uncertainty Index is derived from a textual analysis of seven large newspapers and is reported as the percentage of articles that contain particular key words proposed to reflect trade policy uncertainty. To complement their 'news-based' index, the researchers have also developed a smaller sample sized 'firm-based' TPU from textual analysis of quarterly earnings conference calls. Data for both can be found here and a .xls of new-based TPUcan be downloaded from this page.

To assess the validity of their new-based TPU, which they clearly prefer, the researchers compare it to their firm-level TPU:

and a categorical trade-related economic policy uncertainty measure from Baker et al:

These comparisons indicate reasonable relationship between the measures although the Fed's news-based research TPU measure is the most 'spikey' over the past three years of Trump policy. This is important because it is policy actions during the Trump administration that have clearly motivated this research.

The obvious limitation of the Fed's new-based TPU measure is that it sources data from seven large newspapers shown in other research to slant their content toward the political left. Couple this with the mainstream media's clear bias against Donald Trump, and it is easy to hypothesize that measures of trade policy uncertainty collected from large newspapers are likely to show exaggerated levels since Trump's election. Some evidence of precisely this appear in the second graph above. Note that the Fed's news-level TPU over the past three years is much higher than the Baker et al. measure (which employs a broader methodology) during the same period, and that the Baker et al measure experienced a larger jump back in trade policy uncertainty back in the mid-late 1990s than during the Trump years.

Nonetheless, these uncertainty measures are interesting and may have utility. The Fed researchers apply their new-based TPU measure in empirical analysis of various econometric data. Their results suggest that higher TPU creates a drag on economic activity by lowering trade and reducing fixed investment.

The later reflects what some scholars term the 'wait-and-see' approach to regulatory uncertainty--wait until the regulatory fog lifts before committing to significant investment.