Some Ways to Think About Tariffs – and US Trade Policy

Overview

Tariffs are better than income taxes?
Large countries can make exporters bear the costs of tariffs?
Tariffs are small today
Tariffs cannot cover income tax revenue
Whether tariffs “bring back” jobs depends on why the jobs “left”
Lutnick’s Trifecta (Quadfecta?) – tariffs, equity, Ireland and gold cards.

In the Appendix – Tariff Economics – I go through some of the supply-demand theory of tariffs.

Side-note: I think LLMs are going to help policy-makers A LOT because they allow for much faster answers to precise policy questions.

Tariffs are better than income taxes?

The most common stated reason to dislike tariffs is that they increase costs. This stated reason to dislike tariffs applies also to corporate tax or to income tax. Those taxes also increase costs and reduce both the supply of goods or services and the benefits derived from goods and services below what they otherwise would be.

Moreover, the inefficiency caused by taxes – whether income taxes or tariffs – is non-linear with the rate of tax applied. If you double a tax rate, the economic costs – in terms of lost production and reduced social benefits – will more than double. Given income taxes are in the mid-double digits, corporate taxes in the low double digits and tariffs are in the very low single digits, income taxes (and corporate taxes) are likely doing more economic damage than tariffs – even if tariffs were to be doubled or quadrupled from their current very low baseline.

I’m not in favour of increasing tariffs. However, if the argument against tariffs is one of imposing costs on consumers, then similar arguments apply to income tax and to corporation tax [1].

Import taxes impose costs on exports, and on importers/consumers.
Corporate taxes impose costs on corporation owners, and on consumers.
Income taxes impose costs on employees, employers and on consumers. It’s true that the cost imposed on employees can be done so in a progressive fashion, but there is a real (and not progressive) cost that hits consumers – loosely analogous to the cost that tariffs impose on consumers.

If the argument against tariffs is one of economic inefficiency, then the same argument applies to a greater degree against income and corporation taxes – simply because those tax rates are much higher than tariffs, and economic costs are amplified at higher rates of taxation.

[1] Tyler Cowen recently raised a similar point on corporate tax – on the marginal revolution blog.

Large countries can make exporters bear the costs of tariffs

If you are the only seller of eggs, you have a monopoly, and you can charge more for those eggs than in a competitive market. Conversely, if you are the only buyer of eggs, you are a monopsony, and you can pay less than you would in a competitive market.

If you are a country that accounts for a large portion of importing done in the world, you are a monopsony. If you impose tariffs on your imports, yes, costs will increase somewhat for imported goods BUT a significant chunk of the tariff may be borne by the exporters. Yes, total imports will decrease, but, if the tariff rate remains low, you can have a situation where:

a) the cost of imported goods is a little bit higher

b) the price of exported goods (exported from other countries) is pushed lower [due to your country’s market power]

c) you the importing country government earn tariffs that exceed the increase in costs suffered by your importers.

Whether this works or not depends on i) whether there is retaliation (almost guaranteed? although still may not entirely cancel net effects), and ii) tariff rates being modest (as tariff rates are increased, the increased costs of imported goods start to outweigh the tariffs collected).

This is a big stretch but – in theory it’s even possible that tariffs can increase market efficiency if there is a situation where there is a company or country in a supply chain that is already exerting monopoly or monopsony power. For example, a Brussels tariff on big tech is a tariff in a market that has some monopoly characteristics. I’m not saying “tariffs are a win”, but it can be hard to predict the net effects of tariffs. Economic theory tends to focus on competitive markets, whereas real world markets are often dominated by quite few players, and these effects can be strong.

Tariffs are small today

I got ChatGPT to build up a list of tariffs (have a look here), by country, as a percentage of GDP and as a percentage of government spending.

Many EU countries have effective rates below 0.5% . The US has an average tariff of about 2.5-3%, with China in a similar range (2.2% of imports for 2023).

So, while there may be product-specific exceptions, if the US decides to impose much higher tariffs, it does not “level the playing field”. Rather, it would represent the US leading a solo push away from a field of low tariffs.

For completeness, one should look not just at tariffs but also non-tariff import costs including documentation and regulations [3] – called Non Trade Barriers (NTBs) or Non Trade Measures (NTMs). These barriers are difficult to measure, but real. ChatGPT’s deep research on this topic finds NTM tariff equivalents of roughly 5% for the US and Europe, and 8% for China [2]. I am uncertain on those numbers because the cited literature is sparse and the range of estimates within is large. At the same time, it appears that i) non-tariff barriers are more important than tariff barriers, ii) major trading regions (e.g. US, EU, China) have meaningful non-tariff barriers to trade.

To be even more complete in assessing trade barriers, one might also consider costs that fall neither under tariffs or non-tariff-measures, such as patent infringements or requirements for foreign companies to establish joint ventures with local businesses to operate within that market. Without a more complete quantification of these – and of non-tariff barriers – it is hard to entirely counter the arguments for tariffs as a remedy or response to an “un-level playing field”.

[2] Looking at trade-weighted import tariffs also hides that there are products and services with substantially higher tariffs or non-tariff barriers.

[3] technically, net costs after allowing for consumer benefits of regulations

Tariffs cannot cover income tax revenue

In 2023, US revenue was about $4.4T, while spending was $6.1T – a $1.7 Trillion deficit.

Of the $4.4T in revenue, about half – $2.2T – came from income tax.

Meanwhile, total imports in 2022 or 2023 were about $4T (~2022 figure).

So, for imports taxes to cover income tax revenue, this would require a 50% tax rate. This level of tariffs – or anything close – appears unrealistic.

I talked earlier about how a big country can partly impose the costs of tariffs on exporters – unless the tax rate is too high. It’s hard to know where too high is. My guess is that single digit rates keep the importing government in net benefit territory (although there are real costs – economic and political – that will be felt by importers). Meanwhile, anything close to a 50% tariff rate would seem high enough that it would crush import volume, and with it government tariff revenue.

Whether tariffs bring back jobs depends on why the jobs “left”

If the claim is that tariffs can help to re-shore or create new jobs, the question is why those jobs are not there in the first place. Is it because of a lack of import tariffs, or because of something else?

Perhaps many jobs are not in the US or Ireland because of:

environmental standards being upheld
restrictive planning laws and slow planning permission processes
labour practices being upheld [e.g. no child labour, health and safety at work] – although much Chinese manufacturing leadership today is driven by capabilities in robotics and automation?
cultures that frown more on excellence and achievement than in previous decades?

An aggregate increase in tariffs from 2.5% to 7.5% doesn’t mechanistically change any of those things. It just shifts incentives – but – if the question at hand is improving economic incentives for employment, there is perhaps more room to play with a 40% income tax rate than a 2.5% tariff – especially if looking at things on aggregate rather than sector-specific basis.

Howard Lutnick’s Trifecta (Quadfecta?)

Howard Lutnick is the US secretary for commerce under Donald Trump. Here he is on an insightful 1.5 hour long podcast talking about tariffs and more.

The US deficit is close to ~$2T per year. Apparently the DOGE/Musk goal is to cut spending by $1T and Lutnick’s goal is to find $1T of revenue. Lutnick has three (or maybe four) ideas for finding that revenue. With three ideas, he needs $330B per idea average. With four he needs $250B per idea (or $333B per idea from the best three!). He didn’t explain the exact numbers on getting there, but he gave his strategy and I’m adding some back-of-the-envelope numbers.

Tariffs

In 2022, US tariff income was ~$70B. That’s at a ~2.5% average tariff rate. With some growth/inflation, getting to $330B would mean bringing tariffs up to around ~10% (assuming little reduction in imports).

2. Government Equity in Purchase Deals

When the government makes a big order of goods (aircraft, vaccines) it often causes that company’s stock price to shoot up. Lutnick’s argument (makes sense from his finance background) is that the US taxpayer should get some of that unexpected upside (beyond what the US government gets in corporate taxes).

This seems challenging/problematic in that it pushes the responsibility for fairly negotiating arms-length deals with companies to give the US government a stake in them. It also requires the US government to decide what to do with that stake overtime. One counterargument is that the US government anyway has a difficult negotiation to manage when doing large purchases – whether there is equity ownership involved or not – and there are countries that seem to manage equity ownership well (e.g. Norway, although those are direct investments, not assets gained as part of purchase deals).

Can the US government find $330B in such deals each year? Pfizer is worth ~$127B in market cap today. Maybe the US government could have had ~$30B of that. It’s a good chunk of money, but lots more chunks would be required – per year – to hit that $330B. Is that a good thing for the government to be trying to find – and the incentives that may create? Is that better than not paying pensioners their social security payments (which will go underwater at some point in the next decade, apparently)? Is it good for the US government to own a large portion of key companies (nationalisation in another form)?

3. Ireland – and countries like Ireland with low corporate tax

Lutnick plans to investigate how foreign companies can be incentivised to relocate to the US – potentially by imposing tariffs on sectors like pharmaceuticals.

Meanwhile:

The US corporate tax rate is already lower than in a lot of other countries.
The corporate tax rate differences are much narrower than before. Ireland is now at 15% versus the US at 21% (much more narrow than the 12.5% for Ireland and 35% for the US previously).
Many of the loopholes allowing companies like Apple to avoid 15% tax in Ireland have been narrowed.

Ireland’s total corporate tax is about $28B per year. If all Irish companies moved to the US and started paying 21% instead of 15%, that would be $40B incremental per year for the US. Applying this upper bound calculation across all allied low tax countries, it is unclear whether one can get to $330B in incremental annual revenue. And that is before considering that the US is an important – but shrinking – minority of world trade.

4. The gold card

The gold card allows foreigners to pay $5M for a “green card” where you only pay tax on US income (not foreign income, which must be paid for typical green cards and by US citizens).

One philosophical problem is that this gold card encourages tax arbitrage, which is being denounced in strategy 3 above.

One practical problem is that the US may lose some revenue. The wealthiest foreigners who are on track to become a green card holder or citizen, may instead opt for a gold card.

Note that (with help from chatgpt for research):

There is already the EB-5 program in the US that allows for a visa and path to residency with about a $1-2M investment. It has a take-up of a few thousand per year, although it does not include this tax benefit of avoiding taxes on foreign income
Italy had a program allowing for a flat $100k tax on foreign income. It has a few thousand people availing of it per year and was discontinued.
The UK, until very recently, had non-domiciled status allowing no tax to be paid on foreign income. The UK also had an investment program that cost between 1-2 million pounds and provided a visa that could be combined with non-dom status. It’s uptake was just a few hundred people per year.
Programs allowing for flat or no tax on foreign income in Italy (discontinued) and the UK (discontinuing) had roughly thousands of people sign-up per year.

Notwithstanding the above, $330B of annual revenue would require 33,000 of gold cards to be sold per year. It seems on the high side relative to foreign programs, but I would not underestimate the demand for US visas – especially with this kind of tax benefit.

APPENDIX – Tariff Economics

A. Why tariffs are bad from a classical economics standpoint (because it reduces the benefits accruing to producers and consumers)

In classical economics, trade is good because everyone is drawing from a larger pool of capabilities. This increases specialisation and efficiency, which allows for increased benefits (producer and consumer surplus) for producers and consumers.

If a country decides to put tariffs on imports, that raises the price of those goods relative to other goods. In doing so, less of those goods will be imported. Overall, three things happen:

a. The importing country will import less goods – reducing the benefits they got from those goods (consumer surplus).

b. The exporting countries will export less goods – reducing the profits they got from those goods (supplier surplus).

c. The country imposing a tariff earns revenue (in the form of those tariffs).

An important question to ask… – what is the relative size of these three effects? The answer depends on the importing country’s market power, i.e. how much of those goods they import relative to others.

In the left hand figure below, you can see an example for a small country (like Ireland). The demand curve is shown in blue and slopes downwards – meaning that Ireland’s demand for this product increases at lower prices. The supply curve, by contrast, is flat. This is because – whether Ireland doubles or triples almost anything it buys – that will not greatly affect the price of buying that product on the world market.

And so, if Ireland imposes tariffs, then:

a) Ireland will import less of that good (and lose some of the benefits from importing those goods – the red triangle + the grey rectangle)

b) Ireland will earn some revenue from tariffs, shown in the grey rectangle

c) global suppliers will largely be indifferent and not suffer much loss, because Ireland is a small buyer.

Now, consider the case of a larger country – like the US or China. The demand curve still slopes down (the country will import more if prices are lower). However, as a large importer, as a larger quantity is imported, this will tend to drive up global prices. For this reason, the supply curve slopes upwards.

Now, if China or the US, or the EU for that matter, applies a tariff, there are three effects:

a) the US will import less of that good (and lose some of the benefits from importing those goods – the upper portion of the red triangle + the upper portion of the grey rectangle)

b) the US will earn some revenue from tariffs, shown in the grey rectangle (some of this will come from exporters, and some from importers).

c) foreign suppliers will export less of that good overall, and lose some of the profits (the lower portion of the red triangle and the lower portion of the grey rectangle) they were earning from exporting those goods.

Said differently, by imposing tariffs, importers (consumers/domestic manufacturers importing raw materials) are losing out BUT – for sufficiently small tariffs – most of that is captured by the government in the form of tariffs. At the same time, the other portion of tariffs (lower grey triangle) is essentially coming from exporters. The red triangle is “dead weight” loss – output that cannot be reclaimed by anybody.

Now, for small levels of tariffs, net government revenue can outweigh importer losses. But, for higher levels of tariffs (imagine the left side of the triangle moving to the left), the area of the grey rectangle will eventually get smaller than that of the red triangle, at which point tariff revenue is less than the loss suffered by importers.

As such, a country as a whole can “benefit” from tariffs, provided they are not too high. Furthermore, larger importers can benefit to a greater degree (notice how the ratio of the grey rectangle to the top half of the red triangle is much larger for large-countries imposing tariffs). This is because large countries affect global prices, and imposing tariffs forces exporters to lower their prices, which partially offsets the effect of the tariff for importers.

A crude assessment is that:

A small country that imposes import tariffs is penalising its own importers (consumers, manufacturers using raw goods) to collect revenue.
A large country that imposes import tariffs is penalising its own importers AND foreign exporters to collect revenue (which may, but certainly not guaranteed, be enough to outweigh – economically, but maybe not politically – how much it burdens its own importers).

This is an overall macro analysis though, and is challenged on two fronts:

Even though a large country can more easily achieve a “net benefit” by imposing tariffs, what’s really happening is that importers are suffering a loss, while the government is benefiting from a revenue gain. Yes, the government’s revenue gain is larger than the importer (consumer) losses, but – unless those importers/consumers are somehow compensated (unlikely, and complicated), there can be substantial pain. Sadly, pain on the part of end consumers is harder to identify as there are many factors influencing consumer prices. Pain on the part of importers of raw materials for manufacturing can target a smaller number of large businesses and be more apparent or politically felt.
Exporting countries may retaliate.

2. Retaliation

If one country imposes import tariffs on a large country, that large country (the exporter in this case) will suffer because its exporters will lose profits.

Assuming there is bilateral trade between the two countries, the exporting country can choose to retaliate by imposing tariffs on its imports. At a high level, each country is now collecting import tariffs at the expense of its own importers and its exporters.

From a game theory standpoint, yes, retaliating with tariffs (at a low level) can possibly improve a country’s net economic position, it will still a) hurt that country’s importers/consumers and b) not fix the problem of it’s exporters being hurt in the first place.

As such, retaliating – in the short term – typically puts the country in a worse political position in the short term BUT if countries know that tariffs will lead to retaliation, that can discourage countries from applying tariffs in the first place.

In deciding whether retaliation makes sense, some relevant questions include:

Is the government up for re-election soon, in which case retaliating is harder as it will hurt importers (and possibly further hurt exporters if there is escalation).
To what degree is the government reliant on the political support of specific business/trade union groups that are affected by tariffs.

In the specific case of the EU’s tariff response to the US – while the game theory of retaliating makes sense, it is risky for EU country leaders from a local politics standpoint. Donald Trump won’t be president for a third term (according to bookmakers), but tariffs are probably risky (in significant part, owing to retaliation risks) for JD Vance if he plans on running for president. So ultimately, the threat of retaliation is probably real and important to maintain.

As a side-note, US tariffs threats are both a blessing and a curse for the EU and Ireland:

They are a blessing because they get the EU/Ireland to focus more on the importance of defence, energy independence and economic growth.
They are a curse because (legitimate) complaints about tariffs dominate media and distract from focusing more on defence, energy and economic growth.

3. Uncertainty

Aside from any drawbacks in tariffs, the polarising and shifting nature of views on tariffs means that imposing tariffs is guaranteed to make it hard for businesses to plan. Any tariff that is imposed could very well be later removed. This slows investment and hiring.