Analysing Donald Trump’s tariff policies is akin to shooting at a moving target. Having confidently said at the start of the weekend there was “nothing” Mexico, Canada and China could do to prevent tariffs being imposed, Mexico and Canada managed to negotiate a reprieve on Monday.

This is a rare time to pity economists. It is hard enough to predict the effects of the US president’s policies but much harder when far too much is surprising everyone simultaneously, including Trump himself.

We don’t know the following: whether the US tariffs will happen; whether they are temporary or permanent; whether they are a bargaining chip or a punishment. We do not know what retaliation other countries will impose. All of these things are crucial, even before you get into the deep uncertainties that lie within economists’ models.

Economists should not feel singled out for especially bad treatment, however. Announcing tariffs at the weekend, the US president humiliated his Treasury secretary who had been touting a milder version of tariffs only a couple of days earlier. Scott Bessent had no idea either.

By Sunday, Trump had justified tariffs on the basis of the three countries’ culpability in the US addiction to Fentanyl, lax border security, trade deficits and his territorial ambitions over Canadian soil. Of course, his estimate of a Canadian trade deficit of $200bn was roughly four times too high. Trump is clueless too.

In a world where no one knows anything for certain, here is an attempt to give you some context. One thing I will not dwell on is the tiny price Trump extracted for withdrawing the immediate tariff threat. Mexico appears to have promised to send troops to its northern border where they are already stationed and Canada has promised to create a fentanyl tsar.

How big were the proposed tariffs?

Following Trump’s announcements, there were quickly some fantastic trade numbers flying around, such as the fact that 90 per cent of avocados consumed in the US came from Mexico. But what we really want to know is the proportion of US imports that will face new tariffs.

Obviously that number will change over time because US importers will seek other suppliers, but it is a good starting point.

The chart below uses the latest US Census Bureau numbers for the 12 months up to November 2024, showing that roughly 40 per cent of total US imports came from Mexico, Canada and China. With crude oil imports from Canada over the same period of about $100bn (assuming a Western Canada Select price of $60 a barrel), a little over half of these imports would attract a new 25 per cent tariff and the rest 10 per cent.

These figures allow another back-of-the-envelope calculation. With approximately 40 per cent of imports facing a roughly 20 per cent additional tariff on average, we can work out overall tariff rates for the US. If fully imposed, and with trade patterns constant, the US average weighted tariff rate would rise a little under 8 percentage points.

Putting this into historical context below, Trump’s weekend announcement was therefore huge.

The tariffs were bigger than the decline in duties associated with signing the original General Agreement on Tariffs and Trade (Gatt) after the second world war and the Smoot-Hawley tariffs that came with the Great Depression in the 1930s.

Trump likes to compare his tariffs with those of the 25th US president, fellow Republican William McKinley, who imposed a 49.5 per cent tariff in 1890. Comparisons with the effects of the McKinley tariffs also demonstrate important truths.

US International Trade Commission data shows the weighted average US tariff fell after their imposition in October 1890 because the proportion of dutiable US goods imports fell from 55 per cent in 1891 to 41 per cent by 1894.

Do not tell the president, who sees tariffs as a cash cow, but overall tariff revenue also fell because the trade in tariffed goods fell so sharply.

The fact that trade patterns are sensitive to large tariffs demonstrates a key problem with the dotted line in the chart below, which assumes no change in trade patterns. It is therefore wrong if used as a prediction, but it is still a useful number purely as an illustration of scale. These tariffs are large.

Back in the 1890s, what McKinley also soon found was that support for his Republican party declined almost as quickly as tariff revenues. Surprise, surprise, people disliked higher prices.

McKinley’s party lost nearly half their seats in the 1890 election for the House of Representatives just a month after the tariffs were imposed. Does Trump care? Who knows. Does he know? It does not appear like he does.

How big are the economic effects?

Now we know that assuming constant trade patterns is a terrible basis for a forecast, I am nevertheless going to stick with it because it is useful to demonstrate scale. Goods imports are roughly 10 per cent of US GDP, and if the price of these goods rose by 8 per cent (the size of the overall tariff increase), the US price level would rise about 0.8 per cent. This is why economists are suggesting an effect on inflation of between 0.5 per cent and 1 per cent.

This is a sensible number, but again a terrible forecast. Why? Because we do not know how much companies will use the tariffs as an excuse to push through other price rises or whether the tariffs will be absorbed into margins or by some exporters to the US.

The wider economic impact of Trump’s tariffs broadly represents a supply shock to the US and a demand shock to other economies. My colleagues at the FT’s Monetary Policy Radar have written a great piece on this.

That said, scaling economic effects is extremely difficult. We know tariffs and supply chain shocks are potentially very important for growth and inflation and also that the effects are almost always hard to see in global economic models which tend to assume away large disruptions because they were estimated on very small changes.

In other words, it is probably wise to listen more to Ngozi Okonjo-Iweala, director-general of the World Trade Organization, when she says that a global trade war with tit-for-tat tariffs could generate a “catastrophic” outcome. The IMF model shown in the chart below suggests trade wars do little damage.

Are US consumers really smart?

US consumer spending was notably strong in the fourth quarter national accounts published last week by the US Bureau of Economic Analysis. Even more interesting was that much of the real growth in spending came from durable goods. These contributed 0.85 percentage points of the 2.3 per cent annualised growth recorded in the fourth quarter.

Were US consumers anticipating tariffs and buying before the prices rose? Quite possibly. The data suggests that is a plausible story and one that would probably be followed by a buyers’ strike if steep tariffs persist.

What I’ve been reading and watching

• A former Fed official has been arrested and accused of passing economic secrets to China. As Liza Tobin says in a fabulous write-up: “If someone from a Chinese university offers you $450,000 for a part-time job, run away”

• Not that the FT’s stance on Trump’s trade war will surprise many people: the newspaper thinks it is absurd

• One group of Europeans demanding more tariffs are fertiliser producers. They want anti-dumping duties to be applied to Russia

• Martin Wolf is gloomy about long-term UK growth prospects

A chart that matters

Before the weekend’s excitement about tariffs, I had planned to delight you with an in-depth examination of housing and its contribution to UK inflation. This is a tale of fascinating data measurement issues, comparisons of rental price trends with the US and the important question of stocks and flows. Blame Trump that you have now missed out.

While the Bank of England’s Monetary Policy Committee is worried that the UK has more persistence in services inflation than other economies, much of the UK’s problem with services prices comes down to rental inflation. The chart below shows that UK services inflation excluding rent was lower in the latest data than either the US or the Eurozone.

It is difficult to see the UK as having a special inflation problem that others do not share. We shall see the BoE’s interpretation when it decides interest rates on Thursday.