Not all trade wars are created equal

Abstract

▀ When interpreting estimates of the economic impact of trade wars, the devil is in the detail. Our review of more than 30 institutions' trade war projections finds huge variation – much of which can be explained by differences in assumptions rather than differences in models' structures or specifications. Relative to estimates based on extreme financial or tariff assumptions, our main trade war scenario results are moderate and comparable with the IMF's latest analysis. ▀ In our baseline forecast, trade policy measures have a relatively limited overall impact on global growth. This is in line with recent experience: to date, direct effects from higher trade costs have been small, and policy action has acted to offset adverse impacts on confidence. But the risk of further escalation in US‐China trade tensions remains, even following the (fragile) Trump‐Xi truce agreed at the recent G20 summit. ▀ Our review finds that estimates of the impact of such an escalation range from negligible impacts to a deterioration in some countries’ economic conditions approaching that seen during the global financial crisis. Results in Oxford Economics’ “rising protectionism” scenario – from our latest Global Scenarios Service – are moderate in comparison. ▀ We also find that all estimates showing substantial economic effects from a trade war in our sample are predicated on extreme asset price or tariff moves. In some cases, tariffs are assumed to rise more than at the time of the Great Depression in the 1930s. ▀ Such differences in assumptions can largely account for the divergence in trade war impact estimates. This is illustrated in our very low probability “full‐blown global trade war” scenario – incorporating severe tariff and financial assumptions more in line with institutions like the Bank of England – which generates a deterioration in economic conditions that broadly matches extreme estimates in our sample.