https://www.nber.org/system/files/working_papers/w33739/w33739.pdf
This hot off the press paper, “Tariffs and Retaliation: A Brief Macroeconomic Analysis” examines the effects of a 10 pp rise in the US tariff rate. It has two unsurprising results – a reduction in US GDP, and a smaller reduction in rest of World (ROW) GDP and that ROW retaliation leads to even greater US losses – but one surprising result – an increase in US welfare in the no retaliation case based on a terms-of trade effect (the reduced US demand for imports reduces import prices).
Someone with actual experience with New Keynesian models – SOS! SOS! – ought to examine these results. As a critic of Old Keynesian models, I can only point to areas of concern which may not be valid criticism of this model.
1) I wonder if a two -country model can capture the elasticities of supply of US imports and demand for US exports that drives the fall in US import prices and rise in export prices.
2) I wonder if a two-good (US and ROW) model[1] of the production function may not over estimate the substitutability of domestic and foreign inputs into the production and thus underestimate the fall in output due to lower imports.
3) “US households reduce their labor supply – employment falls – in response to the terms-of-trade appreciation, which has positive welfare implications.” This is certainly not the usual upward sloping labor supply curve and the reduction in labor supply/positive welfare implication is jarring.
4) The model assumes that the Central bank _tightens_ monetary policy in response to the increase in tariffs. While explicitly modeling central Bank behavior is welcome, the assumption is not consistent with FAIT which (at least in a multi good model) woud increase inflation to facilitate adjustment to the relative price tariff shock.
5) “[I]n our model the US trade balance deteriorates.” If this is anything but an implication of the questionable monetary policy assumption, I don’t get it.
Maybe all this is just a preference for a different kind of model, nominal general equilibrium instead of New Keynesian model and different elasticities of substitution among goods and a different sign on the supply of labor WRT the real wage, so I hope others more technically qualified than I will have a go.
Image prompt: Men on both sides of a wall labeled “Tariffs” hurling rocks at other. (Why can’t image generation do words?).
[Standard bleg: Although my style is know-it-all-ism, I am aware that I could be mistaken or overstate my points. I would, therefore, welcome comments on these views (especially on this one).]
[1] Logically one needs at least imports, import substitutes, exports, and non-tradeable goods
Somehow, the image is even better with "tarfrs", but maybe that's just me!
Hmm, humans accidentally invented dyslexic AIs. Maybe someday we will discover a deep relationship between these AIs and dyslexic humans.