Are tariffs good?

“To me the most beautiful word in the dictionary is tariff” - Donald Trump

While his policy positions have often proved mercurial, Donald Trump has been notably consistent on one issue. The returning president is an ironclad skeptic of free trade, and has long called to impose import duties on key trade partners to protect US industry.

Under his first administration, he introduced global tariffs on steel, aluminium, washing machines, solar panels and targeted duties on goods from China, comprising a total of $380bn in imported goods. This time he has gone further, threatening a 10 - 20% flat tariff on all imports, up to 25% on Canada and Mexico and even 60% on all imports from China. 

Some of this may be Trump’s ‘Art of the Deal’, in which his threats to Canada and Mexico are a bluff to extract concessions from them and other trade partners. However, the core logic of Trumpian protectionism is that tariffs will reduce America’s trade deficit, stimulate investment in US industry, create more American jobs and boost US economic growth. 

Trump’s first round of tariffs achieved none of these things, instead accruing benefits to small but politically valuable constituencies at the expense of the broader economy. The second round will likely do the same. 

Let’s start with the US steel industry - a long-term cause célèbre for the MAGA base. In March 2018, Trump announced a 25% global tariff on imported steel. This should have been a boon for US steel workers. America is the world’s 4th largest steel producer, and with competitors now considerably more expensive, US smelters hoped to ramp up their production to fill the supply gap. 

And to some extent, they did. Shielded from foreign competition, US steel output increased by 5.1% between 2017 and 2021, while employment in the industry grew by around 6% from around 80,600 people to 86,000. 

The problem is that the 5,400 extra steel jobs created under the first Trump administration are barely a rounding error relative to the broader US economy. Tariffs raise prices: average steel prices increased by 2.4% over his first presidency, while prices for washing machines, another levied import, jumped by 12% in one year. They also trigger retaliatory duties from trade partners. Canada, China, the EU, India and Mexico all hit back with their own tariffs on US exports. 

This hurts the competitiveness of the downstream industry that employs a far larger share of the US workforce. Lydia Cox and Kaydee Russ, economists at Harvard and the University of California Davis, estimate that American jobs in industries that consume steel, such as the automotive sector, outnumber steel producing jobs 80 to 1.

The broader US manufacturing sector flatlined as a result, neither attracting significant new investment nor adding new employment. Foreign direct investment into US manufacturing - a prized photo-opportunity for any image-conscious politician - was actually on average 38% lower in the four years of Trump’s first presidency than in the preceding three years. 

Total employment in the manufacturing sector remains far below pre-GFC levels at around 12 million. 

The failure of the protectionist approach is most clear when viewed from the macro-level. The rise of China as an industrial and export superpower following its admission to the WTO in 2001 has reshaped the global trade and supply chains. Cheap Chinese goods have pushed down prices and led manufacturing capacity to move to China. As a result, Trump has long viewed China as the ultimate cause of American industrial decline, and accordingly slapped 25% tariffs on $50bn of Chinese-made goods during his first term. 

While the tariffs succeeded in reducing the size of the country’s trade deficit with China, the US global trade deficit still grew by 14% over the course of his first administration. Faced with higher prices on Chinese goods, US importers and consumers pivoted away from buying directly from China and towards third countries like Mexico and Vietnam. 

This is because trade imbalances clear globally rather than bilaterally, argues economist Michael Pettis. Modern supply chains are dynamic and follow the path of least resistance, so cheap Chinese goods will still wind up in the US if later stages of the production chain are moved abroad. This is why an increasing number of companies are “nearshoring” to Mexico, and why Chinese solar panel manufacturers have moved their assembly lines just south of their own border to Vietnam. 

Pettis’ key point is that tariffs do not address the root imbalance in the global economy: Chinese excess savings. China produces far more than it consumes because Chinese households save a very large share of their income, due to demographic factors such as the One Child Policy and a long-term development strategy that has prioritised investment in the supply side of the economy over raising household spending power. China’s household saving rate is around 30% - almost twice the roughly 16% seen in the Eurozone, and far above 4.5% in the US. China therefore has massive industrial capacity but weak internal demand, and shifts this excess capacity overseas in the form of cheap exports. 

The upshot is that US tariffs did not resolve this structural imbalance in the Chinese economy, and instead America’s trade deficits and China’s surpluses have continued to rise, but just not with each other. To change this, Beijing will need to undertake profound reforms of its economy and political system to increase household income shares and encourage households to spend more and save less.

The prognosis for the US economy with Trump's new proposed tariffs is poor. The Tax Foundation, a DC-based think tank, predicts that a 20% tariff on Canadian and Mexican exports and a 10% tariff on Chinese exports will reduce long-run GDP by 0.4% and cost 344,000 US jobs, even before considering inevitable retaliatory moves from these countries. 

While politically appealing, tariffs will not fix the US trade deficit, attract foreign investors, create jobs or make American manufacturing great again. Trump 2.0 needs a new answer to his oldest problem.