Related Links

• Tariffs: Estimating the Economic Impact of the 2025 Measures and Proposals
• Tariffs Update: Potential Effects of the April 2 Announcements
• Tariff Update: Incorporating the April 9 Announcements
• Tariff Update: Incorporating Recent Decisions and Deals

Transcript

Tim Sablik: My guest today is Marina Azzimonti, a senior economist and research advisor at the Richmond Fed. Marina, welcome back to the show.

Marina Azzimonti: Thanks, Tim. Always a pleasure to be here.

Sablik: Today, we're going to be discussing the potential economic impacts of recent tariff changes, which is something that you've been following very closely. Over the past few months, you've written several Economic Briefs estimating the effective rate of new U.S. tariffs, and we'll include links to all of those pieces in the show notes for anyone who's interested.

To start our conversation, could you explain why calculating the ultimate economic impact of a tariff isn't as straightforward as just looking at the announced rate?

Azzimonti: The headline tariff rates are just the tip of the iceberg. What really matters is how they play out across countries and products. For example, as of May 2025, goods coming from China face average effective rates above 55 percent, while those from Mexico are closer to 13 percent. But even within Mexico, it depends on what you're importing. Auto parts that meet USMCA rules — this is the North American trade agreement between the U.S., Canada, and Mexico — are generally exempt from tariffs. But aluminum, that's a different story. If it wasn't melted and poured in North America, it doesn't qualify [for an exemption] and it can still be hit with a 25 percent tariff. So, the tariff landscape isn't just country by country, it's product by product. That complexity is what makes it so hard for businesses to navigate.

Second, it matters how much we actually import from each country. A high tariff might sound dramatic, like a 49 percent tariff on imports from, say, Cambodia. But when we look at the data, Cambodia only accounts for 0.4 percent of total U.S. imports, so it's definitely not going to move the needle much. In fact, that's exactly what we saw in the data we analyzed in our April Economic Brief. Despite the sharp increase in announced "Liberation Day" tariffs on countries like South Africa or Cambodia, their small share in the total trade basket meant their effect on overall average tariff rate was relatively minimal. It's a reminder that both the rate and the share of imports matter when assessing the impact.

Third, not all imports are final goods, like sneakers or iPhones. Many are intermediate goods — these are parts, components, or raw materials that firms use in production. This is where things get really tricky. The U.S. economy is deeply integrated into global supply chains. Firms might import inputs from multiple countries, transform them domestically, and sell them either here or abroad. A tariff on one input can ripple through the entire supply chain.

Finally, different sectors of the economy have different levels of exposure. An industry like automotive manufacturing, which relies heavily on imported parts, is going to feel tariffs much more acutely than, say, health care or hospitality which use mostly domestic inputs.

So, when you try to link a tariff announcement to an economic outcome — whether it's prices, employment, investment — it's not going to be straightforward. You have to account for the mix of trading partners, the types of goods involved, supply chain structures, and industry-specific vulnerabilities. That's why we go beyond the headline announced rates and compute measures like the average effective tariff rate to get a clearer picture.

Sablik: Thanks, that's a great overview and definitely illustrates how complex and complicated this is to analyze.

You mentioned the average effective tariff rate that you're calculating. What does that measure and how exactly do you go about calculating that?

Azzimonti: The average effective tariff rate, or AETR, is a simple but powerful way to measure how costly it is to import goods into the U.S. You can think of it like a sales tax, but instead of being added at the cash register, it's applied at the border. If the AETR is, say, 10 percent, that means a product valued at $100 at the port will cost $110 once it crosses the border.

Measuring it is actually pretty straightforward. You just need to take the total tariff revenue collected by the U.S. at the border and divide it by the total value of imports. Both of these numbers are published monthly by the U.S. Census Bureau, so it's very easy to track. At the start of 2025, that number was about 2.3 percent. So, even though tariff policy itself can get pretty complicated, the AETR gives us a clear apples-to-apples way to see how trade costs are changing over time.

But here's the catch. There is usually a lag between when tariffs are announced and when they start showing up in the official data. For example, the latest data on tariff revenue we have today is from April and we're in June. During times like 2025 when trade policy is shifting rapidly, businesses and policymakers can't afford to wait months to find out what's happening. They need to anticipate what's coming.

That's what our work has been. We build what you might think of as a predicted or projected AETR, an estimate of what the average tariff burden would be under newly announced policies.

How do we do it? We dive into detailed trade data. We know exactly what goods the U.S. imports from which countries and in what quantities. For instance, we know precisely how much the U.S. imported of a category like meat — all the cuts with bone-in, fresh or chilled, from Australia in December 2024. Then, we can match that up with official tariff announcements — which goods are affected, from which countries, and at what rates. Once we've mapped that out, we apply the new tariff rates to the 2024 trade flows. That is key because it lets us isolate the pure effect of the policy without assuming firms or consumers are already changing their behavior. So, it's a way of answering the question, "If these tariffs were in place last year, what would the average cost of importing have looked like?"

Sablik: As you already alluded to, trade policy has been changing and evolving fairly rapidly in recent months, and I should have noted that we're recording this on June 18. Can you walk us through how the average effective tariff rate has changed during the first half of 2025, so from January to now, June 18?

Azzimonti: Things were relatively quiet at the start of 2025. In early January, the AETR was around 2.3 percent, which means the average cost of importing goods into the U.S. was pretty low by historical standards.

But that started to change in February, when the U.S. announced a wave of new tariffs on imports from Canada, Mexico, and China, citing a national emergency over issues like undocumented immigration and drug trafficking. Most of those measures were paused shortly after they were announced, except for the 10 percent tariff on Chinese-imported goods which went into effect on Feb. 4. Because China is such a major trading partner — over 17 percent of our imports come from there — one policy change notched the AETR up to about 4 percent.

Then, in early March, things started escalating again. The U.S. imposed an additional 10 percent tariff on all Chinese imports and a 25 percent tariff on goods from Mexico and Canada went into effect. That pushed the AETR sharply higher to over 13 percent almost overnight. Some of that was rolled back temporarily due to exemptions under the USMCA trade agreement with Mexico and Canada, which brought the projected AETR down to 7 percent. But then, just a few days later, we saw new tariffs on aluminum and steel at 25 percent, raising the AETR to almost 10 percent.

Then came a major escalation on Apr2. That was the Liberation Day where tariffs were expanded in scope and partners that had mostly avoided earlier rounds — like the European Union, Japan, and Vietnam — were now included. This drove the AETR to about 22 percent.

But the story didn't end there. A week later, on April 9, some of these reciprocal tariffs were paused and a new structure was introduced: a uniform 10 percent tariff on all imports, except those from China which were increased to 145 percent tariff rate. Because China is such a large trading partner, as I mentioned before, that pushed up the overall AETR to over 27 percent.

In May, the landscape shifted again. The U.S. signed a trade agreement with the U.K. and, more importantly, reached a truce with China. Both countries agreed to reduce tariffs that had been imposed in April, settling on a 10 percent baseline plus the earlier 20 percent tariffs from March and rolling back all the most extreme measures. Based on [this] policy announcement, we estimate that the AETR fell to about 17 percent in May, still high relative to historical standards but down from the April peak. So, while tensions have cooled, tariffs remain elevated by historical standards — remember that we started the year at 2.3 percent — and the landscape is still very much in flux.

Now, as I mentioned earlier, all of these estimates gave us a projected AETR, assuming that businesses kept importing the same [amount] of goods from the same countries regardless of the new tariffs. But, of course, that's not how tariffs behave in the real world. When prices go up, demand tends to fall.

That's exactly what we saw. According to the latest trade data, total imports fell by 16 percent in April relative to March, with the steepest drops in imports from China, Mexico, and Canada, the countries that were most affected by tariffs. Because of this drop in trade volumes, the realized AETR — which is based on actual tariff collections — came in closer to 7 percent. This is significantly lower than the projected peak, but still is three times higher than how we started the year.

Sablik: Thanks so much for that. I'm sure we could say a lot more.

You mentioned how the effective rate in April, the realized effective rate, compared to where we started in 2025. How does the current effective tariff rate compare to prior periods in U.S. history?

Azzimonti: The rates today are remarkably higher than they have been in the past. To find anything comparable, you have to go back all the way to the 1930s to the era of the Smoot-Hawley Tariff Act. That was passed in 1930 during the early stages of the Great Depression and it raised average duties to about 22 percent. At the time, it was a massive spike in protectionism that is widely seen as having deepened the global downturn. Fast forward to today, the average effective tariff rate in 2025 peaked around 27 percent. So, in terms of scale, we are talking about a level of trade protection that we haven't seen in nearly a century.

What makes this episode very different from, say, the 2018-19 tariffs — which only pushed the AETR to about 2.5 percent — is that we are now seeing a broader structural shift. These new measures go beyond targeting specific sectors like steel or solar panels. They apply to entire countries, multiple industries, and cover a much wider range of goods.

Sablik: That's a good segue. How does the effective tariff rate vary across industries, and which industries are most exposed to the recent tariff changes?

Azzimonti: To understand how tariffs affect different parts of the economy, we look at something called the industry-level AETR. What this is telling us is not how much it costs to operate an industry, but rather how expensive it is to import goods associated with an industry after tariffs are imposed.

When we look at the results, there's a lot of variation across industries. Some are clearly feeling the heat more than others.

Take apparel, textiles, and leather goods. They're facing an AETR of over 30 percent. That's largely because these industries rely heavily on imports from China, which has been subject to some of the steepest tariff hikes this year.

But they are not alone. We are also seeing fabricated metal products, electrical equipment, and electronics face AETRs of about 20 percent. In those sectors, a big part of the story is the targeted tariffs on steel and aluminum, which are core inputs in everything from wirings and motors to appliances and machinery. So, even though these industries may not be importing finished goods directly from the highest tariff countries, they're still getting hit through higher input costs due to these goods-specific tariffs.

Another important case I want to highlight is transportation equipment, which includes vehicles and parts. It's facing an AETR of almost 15 percent. That may sound more moderate, but it's hugely relevant because this sector accounts for 13 percent of all U.S. imports — one of the largest import shares across U.S. industries — and it employs more Americans than any other manufacturing sector. So, even relatively small tariff increases here can have a big footprint in terms of both trade volume and labor market exposure.

On the other end, industries like chemicals and petroleum refining are facing much lower AETRs, often under 10 percent. That's because their imports are more diversified across trading partners or because they include products that were exempted under the most recent tariff wave, such as those under the USMCA.

So, the effective tariff burden is far from uniform. It really depends on who you trade with, what kinds of goods you bring in. That's why looking at industry-level AETRs is so valuable. It helps policymakers and businesses get a clearer picture of where the real pressure points are in the U.S. economy.

Sablik: Yeah, that's very helpful.

As we've been talking throughout this whole conversation, there's a lot of nuance into understanding how these tariffs trickle through different parts of the economy. In your most recent Economic Brief, you created a measure for the indirect impact of tariffs. Could you explain a bit about the intuition behind that measure?

Azzimonti: When our regional economists were talking to local producers, some of them were stating, "Well, I am in this industry, but I'm mostly sourcing domestically. So why should I expect to be hit by this?" That got us thinking, is just thinking of industry-level AETRs relevant for producers? Is this the end of the story?

So, what we try to understand with this new measure — which we call the impact factor — is how much tariffs raise the cost of producing goods in a given industry, not how much goods are going to cost if you import them and they belong to an industry. The answer to that is very non-trivial because you need to understand how production actually works. Most industries don't operate in isolation. They are part of a complex network of suppliers and buyers. Firms source imports from lots of other sectors — some domestic, some imported. So, even if a business doesn't directly import a tariffed good, it might still feel the impact through higher prices passed on by its suppliers.

Let's take a simple example: soda. You may think a soda company wouldn't be affected by aluminum tariffs if it's not directly importing aluminum. But most other producers buy pre-made cans from can manufacturers and those firms imported aluminum. So, when a tariff drives up the price of imported aluminum, that cost ripples through the supply chain and eventually hits the soda market. That's what we mean by spillover effect.

To capture these indirect effects, we built a measure called the impact factor. It uses detailed input/output data from the Bureau of Economic Analysis to map how much each industry depends on inputs from every other sector and how much those other inputs are imported. Then, we layer on the announced 2025 tariffs to estimate how much each industry's production costs are likely to rise, not just from the direct tariffs they face but from the ripple effect across the supply chain.

What we find is that industries like apparel, petroleum refining, machinery, and chemicals are among the most exposed, but, interestingly, not because they are directly targeted by tariffs. It's often because they rely on other sectors that were hit.

Here's another key insight. The impact factor values we estimate are generally lower than the AETRs discussed earlier. That said, even modest increases in production cost can add up, especially in industries with tight margins or complex global supply chains. So, while we are not necessarily looking at massive price spikes across the board, these cost pressures can squeeze profits, they can slow investment and, in some sectors, even show up in higher prices for consumers. That's why we're so worried and paying so much attention to them in the Fed.

Sablik: Thanks, that's a great overview. Even domestic suppliers can end up being affected by these tariffs that then pass on to other domestic firms.

Taking these spillover effects into consideration, what is the expected impact of tariffs across industries and geography? You gave a few examples, but maybe you can go into that a little bit more.

Azzimonti: When we look at the industry-level results, the biggest takeaway is that the most affected sector isn't necessarily the one you would expect. We saw the highest AETR were textile products, but it is motor vehicles and parts, for example, that is receiving the highest impact factor, according to our calculations.

First, motor vehicles and parts is a highly globalized industry that relies heavily on imported components. Second, it's been hit by a combination of overlapping tariff regimes — the U.S. currently has a 25 percent tariff on imported passenger vehicles, light trucks, and certain automotive parts. Those were imposed under Section 232 in April and May of this year. There are some exceptions if enough of the car is made in the U.S. or if some of the parts are under the USMCA treaty. But there are a lot of other cars that are sourcing parts from the European Union, for example. Third, the industry can also be indirectly affected by the steel and aluminum tariffs because those are very important inputs used in their production. So, when you stack all of them up — high input intensity, new part-specific tariffs, costly raw materials — the auto sector ends up being one of the most exposed in the economy.

Other industries that show high exposure are apparel, textile mills products, furniture, machinery, and electrical equipment. Again, part of this reason is that they rely on imports from sectors that were targeted. For example, a furniture manufacturer may source equipment or base materials from tariff-affected industries, even if it doesn't import a lot from countries that are being hit by tariffs.

One thing that's easy to overlook is that even services aren't totally insulated. Sectors like food services and entertainment might not import much themselves and they are mostly using labor inputs. But they still rely on goods that do, like packaging, kitchen equipment, or electronics.

Sablik: Thinking about the impact factor, which states or regions in the Richmond Fed's district seem to be the most exposed to the tariffs? To remind listeners, our district includes Maryland, West Virginia, Virginia, North Carolina, and South Carolina.

Azzimonti: Counties that have large manufacturing footprints — especially in the Midwest and parts of the South, in our Fifth District — tend to have the highest impact factor because they are home to the most global, integrated sector. Meanwhile, places with more service-heavy economies — like parts of the Northeast, the West Coast, D.C. area — are relatively shielded.

When we look at which counties are most exposed to the recent tariffs, a few clearly rise to the top. Smith County in Virginia, Scotland County in North Carolina, and Spartanburg County in South Carolina are among the top ones. What these counties have in common is a strong concentration in transportation equipment manufacturing, which was one of the sectors with the highest impact factors. Spartanburg, for example, is home to BMW's largest North American plant, which makes it a hub for auto manufacturing and global supply chains. That kind of footprint means they are particularly sensitive to rising input costs, not just from direct tariffs on auto parts, but also from spillover effects due to steel and aluminum tariffs.

Another one to mention is Alexander in North Carolina, where furniture manufacturing plays a big role. That sector is also exposed partly, because it imports material and certain components from counties facing stiff tariffs, especially China.

Sablik: Well, we've covered a lot. It's definitely clear that this research has been keeping you plenty busy, tracking all these various changes. Do you have any plans already for next steps in your research on this topic?

Azzimonti: Yes. This research has kept me awake several nights.

It's not only something I did on my own in a vacuum. Acacia Wyckoff, she was a [research associate] at the Fed. She was instrumental in this. We work very closely with Sonia Waddell; with Renee Haltom. We learned a lot from the regional group that are talking day to day to producers and giving us feedback as of how to think of this.

When we talk to our other policy economists, I realize that the million-dollar question going forward is, "How is this going to end up showing up in consumer prices, if at all, or firms down the supply chain?" The key concept is pass-through, meaning how much of this tariff cost will actually be passed along to consumers. So, if a company faces higher costs at the border, do they absorb it? Do they squeeze their margin? Do they bargain with their suppliers abroad? Or are they going to raise prices? Or are they just going to adjust their sourcing, sourcing from countries that have lower rates or sourcing domestically. So, we are just starting to understand how this is going to work.

What we are hearing from our ongoing surveys is that many firms are trying to shield consumers, at least for now. That could mean that pass-through is lower this time around than what we've seen in the past.

As new price data comes in, especially broken down by industry, we'll be able to compare it against our impact factor estimates, which essentially try to predict how much production cost should increase based on supply chain exposure and tariffs. Do we see a correlation between what [industries] we thought were going to experience a higher hit and how prices actually change in those industries. That comparison will help us to understand which sectors are absorbing costs, which are passing them on, and where the inflationary pressures might be building under the surface. It's a crucial next step in figuring out how trade policy is translated into real economic outcomes.

Sablik: Yeah, absolutely.

Listeners who are interested in following all that work, I definitely encourage you to go to our website, richmondfed.org, where you can keep track of everything that we're writing, researching, and learning about this topic.