Donald Trump has now unilaterally imposed huge global tariffs, upending the world economy. But we did get a preview of Trump’s trade approach in his first term, allowing researchers to analyze the political consequences. Thiemo Fetzer finds that China, the EU, Canada, and Mexico reacted to the first term tariffs strategically, trying to hurt Trump’s constituents. Omer Solodoch finds that the first term trade war announcement immediately hurt Trump politically, reducing approval and affecting voting intentions. They both say the new trade war is bigger, with political consequences likely to grow.
Guest: Thiemo Fetzer, University of Warwick; Omer Solodoch, Hebrew University
Studies: “Tariffs and Politics”; “Political Costs of Trade War Tariffs”
Transcript
Matt Grossmann: How the first-term trade war hurt Trump, this week on The Science of Politics. For the Niskanen Center, I’m Matt Grossman. Donald Trump has now unilaterally imposed the largest and most widespread global tariffs upending the world economy. But we did get a preview of Trump’s trade approach in his first term, and we might get a preview of the stakes for Trump now by looking at the political consequences from the first term.
This week, I talked to Thiemo Fetzer of the University of Warwick about his economic journal article with Carlos Schwarz, Tariffs and Politics. He finds that China, the EU, Canada, and Mexico reacted to the first-term tariffs strategically trying to hurt Trump’s constituents. They had to balance those strategies against their own consumers’ interests, but it was enough to affect Trump’s voters.
I also talked to Omer Solodoch of Hebrew University about his Journal of Politicxs article with Edward Mansfield, Political Costs of Trade War Tariffs. He finds that the first-term trade war announcement immediately hurt Trump politically reducing approval and affecting voting intentions all before any individuals had been hurt economically. They both say the new trade war is bigger with political consequences likely to grow. I ask what’s different about US policy and the international response this time and how that might affect voters in the midterm elections. I think you’ll enjoy these conversations starting with Thiemo.
So you studied retaliatory tariffs during Trump’s first term in an article in the Economic Journal. So remind us what the major findings and takeaways from that article were.
Thiemo Fetzer: Essentially, what we did in this paper was study retaliation feasibility trade-offs. So essentially we were looking at how countries retaliate to the aggressive trade tariffs stance that was imposed by the first Trump administration. And we looked at how China, Europe, Mexico, and Canada in particular, responded to these tariffs by designing retaliation and rebalancing measures to take out the economic blow, but also to essentially try to hit back at what was perceived as a very aggressive trade policy stance.
And in essence, what we find is that retaliation was carefully politically targeted, but there’s a broad range of subtle trade-offs that have been sort of traced out and that has been tried to be sort of traded off. The European Union was particularly sort of sharp in terms of picking goods that both maximized the degree of political targeting, not just hitting the average Trump voting area by retaliating towards goods that are produced in sort of the average Trump voting area, but by specifically focusing on the areas where there’s a lot of swing voters that have been moving around quite fluidly across elections.
So there was this clear political objective to targeting where it politically hurts, but at the same time there’s obviously offsetting incentives whereby the EU retaliation in particular is guided by a regulation that says, “Well, when a country attacks us, we try to inflict pain back, but we try to protect the industries that have been hit by the act of aggression. And at the same time, we want to provide a retaliation in a way that does not necessarily target goods where the US in this case is the sole supplier.” So that it doesn’t hurt essentially own economy. And that’s the careful trade-off that is navigated.
Countries achieve that better or worse. In particular, what we notice is that China was very aggressive in its retaliation, but is also very constrained in what it can retaliate against because the volume of the tariffs that the US has imposed was quite high. The Chinese imports from the US includes a lot of high value items such as airplanes and so on, but the bulk of it is actually a lot of quite agricultural commodities and very basic goods where there’s ready alternatives available.
And so ultimately what our observation from the first trade was, so to say, was that the Chinese retaliation seemed to have been very aggressive, but also in a way that could have induced a lot of pain on Chinese own consumers, which then the Chinese gradually rebalanced by essentially sourcing inputs from other suppliers and essentially inducing a round of trade rerouting. And that was quite a stock where essentially the US here has more leverage, so to say when it comes to goods trade because trade between the US and China in terms of goods is quite unbalanced.
And so the US can sort of hit more and the Chinese, if they retaliate more, it might lead to significant trade rerouting or potentially short-term economic pain.
Matt Grossmann: So you have a pretty fine-grained data here on the geographic and industry level pattern of the tariffs. So tell us a little bit about that and how you found these stark differences, especially in the geographic content of the Chinese tariffs versus the others.
Thiemo Fetzer: Sure. I mean, at the end of the day, what we do in this paper is we construct a measure that essentially looks at production data at the state level that’s broken down into counties based on very granular employment data that allows us to measure the incidence or the likely incidence of tariff retaliation across counties and county by sector, so to say. So quite granular cells. And that would then obviously pick up that if retaliation is skewed towards, let’s say agricultural commodities, let’s say soybeans, that this would hit some regions in the US that are soybean producing much more than other areas that are not primarily engaged in that sort of economic activity.
I think the key novelty or the key bit as an analytical layer that we add in this paper, which actually received not as much attention, is that we construct a broad set of counterfactual retaliation responses. So not just do we look at what is the degree of the political targeting of the actual implemented retaliation response by just measuring essentially what is the correlation between Trump voting in 2016 or the switch from Obama to Trump sort swing voting, what’s the correlation so to say with our exposure measure and the tariff exposure measure, retaliation exposure measure and the support for Trump.
We also do that for counterfactual retaliation responses. So retaliation responses that the other countries could conceivably have chosen but chose not to implement. And that sort of allows us, as I said, to trace out the retaliation possibilities frontier. I think that’s where we can then obviously characterize to what extent a given retaliation appears like an outlier relative to the counterfactual of possible chosen alternatives that could have possibly be chosen through the selection ultimately of different bundles of goods that are being retaliated against.
And again, this is where the EU is very clever so to say. I mean, I remember at the time I spoke with an unnamed official at the European Commission that sort of indicated, well, we have our tools to design retaliation focusing on iconic brands, very clear American products that however, have a very clear political signature in terms of the essentially constituencies or the areas so to say where these are traditionally being produced in a way that was very clearly mimicked and aimed at affecting the political economy and the underlying support in Congress for Trump’s trade policy.
And I think that’s where the granularity of the data allowed us to essentially provide a comment on, as I said, not just the actual choice and evaluating the actual choice, but evaluating the actual choice relative to a broad set of counterfactuals that makes it quite compelling to essentially understand what other countries think might hurt Trump the most relative to what they can choose in terms of their retaliation design.
Matt Grossmann: You also looked at some individual survey data and panel data. So tell us about that part of the paper and what you found.
Thiemo Fetzer: So in the work, we specifically… I mean we put a lot of care and attention to trace out the sharpness of the targeting that is really picking up the distinct electoral geography of essentially electoral support for Donald Trump. And that’s why we look at the correlations between the retaliation response and the support for Trump vis-a-vis other Republican candidates in the 2016, so before the actual election, before the 2016 election. So we document that it’s a very clear signature that hits areas that had a particular strong support for Trump as a candidate already which is again highlighting the granularity of the targeting.
We’re also looking at to some extent whether the retaliation had an impact on the electoral outcomes in particular 2018 midterm elections. And it did suggest that there was some sort of electoral cost of not the trade war, but the actual retaliation response and hurting the electoral base of Donald Trump in the midterms. So in some level it would suggest that retaliation can work.
We should not lose sight of the fact that a lot of this debate right now is actually focused on trade in goods. That is goods that physically crisscross the ocean that can be measured at the border, so to say, at the customs border as a good enters into, let’s say, the European single market or the Chinese market. But of course we are now potentially moving in terms of escalation in a direction that might be much more nuanced because there is now a discussion over whether a service sector trade might be a part of a potential retaliation response.
And service sector trade is very difficult to measure and it’s much more difficult to geographically allocate, so to say, where is that service being produced and in particular where the economic gains the ultimate residual economic gains from that service sector trade arising. This is the discussion about big tech firms.
Matt Grossmann: So beyond your own paper, what do you think of the current conventional wisdom on the potential for this to really matter in election outcomes? Obviously, trade outcomes have been shown to matter for many elections. On the other hand, US elections are pretty set in stone for most parts of the country. So I guess what’s the state of the art in terms of the potential impact as countries are considering at this time?
Thiemo Fetzer: I do think that we might see a very similar dynamic that could play out this time around that they might… If there’s retaliation. Again, if the retaliation is confined to goods trade, we might, again, see a similar dynamic whereby then the US might be seeing a need to provide and supporting measures, for example, for its farmers that might lose market access, which is exactly what happened in the first time around. I think around just the electoral timing that we see in the US.
The US because of the very tight electoral cycles means that essentially an executive administration has about two years essentially to get its business done. And there’s a big risk that obviously by the midterms, this could result in sort of an unraveling of the political economy whereby electoral support in the houses is being lost. Of course, this time around a lot of the trade measures are being motivated under national security prerogatives, invoking presidential powers that… Well, there’s an active debate about… for legal scholars around whether Donald Trump actually has these powers to wield.
So this could potentially extend the time window, but essentially the fact that the electoral cycles are so short means that essentially the US is going full broadside right now on the tariffs to try to use that limited time. Whereas obviously other partners might be just speculating for time and hoping that their respective retaliation responses might matter in rebalancing and sort of ultimately constraining tying the hand of the executive by the time that the midterms come about. That might be the hope that other countries that are engaged in this dispute or trading blocs might have.
But right now, since the escalation has been so aggressive from the start, this could go… Just again theoretically, this could go down a very rigid path of escalation whereby we need to think a little bit and discuss a little bit more of the broader strategy and the broader motivations behind this US aggressive stance, I think to help also other countries potentially engage with the US administration. Because I think right now there’s just a lot of confusion also which might be a negotiating tactic by just throwing a lot of yes maybe, yes no, maybe yes. But obviously the question is whether there’s a system behind that, whether that’s part of the negotiation or whether that’s essentially also pointing towards a degree of confusion, which obviously would potentially… It already is leaving its marks in financial markets and the role of the dollar, which, again, if we engage with some of the thinking and the work of some of the academics that are associated with the US stance in particular around this trade war, this trade escalation, suggests that maybe this is exactly what they want to achieve, make the US less investible, devalue the dollar, essentially weaken potentially the role of the dollar in the global economy, at least as a reserve asset, because there is some non-negligible arguments that would suggest that there is a potential dilemma that comes with a country being the reserve currency issuer, because it ties the country’s hands when it comes to monetary policy. And so, I do think that we also need to engage a little bit with that intellectual backdrop, because that seems to motivate a lot of the administration’s actions.
Matt Grossmann: So you’ve talked about the trade-offs that countries face in implementing retaliatory tariffs, I wonder if you could maybe give an example for today. So as the EU, for example, considers retaliation, what does it have to consider in terms of how things might be politically useful, but how they might impact the domestic economy?
Thiemo Fetzer: I can make it quite sharp vis-a-vis the measurements that we’ve used. Essentially, in designing retaliation, as I said, countries will look at the degree to which the retaliation is correlated with, for example, at the regional level, county level or so, in terms of electoral preferences of the US electorate, as measured through election data, so to say, and breaking down that retaliation basket down to essentially the areas and the sectors where this is being produced within the US. That’s one dimension, that political targeting, trying to induce pain, so to say, on the aggressor.
The other dimensions is, for example, are there already substitutes available? If there’s retaliation targeting towards soybeans, reducing imports of soybeans, soybeans are a non-complex global commodity that are readily traded, there’s ready alternatives available in the global market, and so, for example, retaliating against soybeans might be something that is less costly for the European Union to do, because actually, there’s alternative suppliers that could be tapped in to source soybeans, for example, Brazil and a few other countries. And so, that’s one consideration, you want to retaliate against something that induces pain, but at the same time, something that is not hurtful to your own consumers, and so the Europeans could potentially then source soybeans from another country.
The third dimension that is typically important is just looking at trade elasticities, which suggests to what extent the supply is elastic or inelastic globally. So if the supply is very elastic, it means that if the cost of exporting from the US to Europe for soybeans are going up because of the tariffs, it actually means that the trade flows could collapse substantively very quickly, again, simply from the perspective of this being a globally traded commodity. And so, that means that retaliation actually is effective, because the response is very elastic. And so, these are typically some of the metrics that will be looked at, but again, that is focusing on goods.
But now, in this particular escalation, we are already talking and there’s a debate around the potential invocation of an anti-coercion tool by the European Union, or expanding the retaliation response to the service sector side of the economy. And we have to bear in mind that the US, actually, when it comes to service sector exports, it’s looking very good. The US actually, through its tech companies, is exporting a lot of digital services. Europe, on average, doesn’t have any digital giants, a lot of the ICT infrastructure in Europe is provided by US tech companies, so the Magnificent Seven, so to say, or a big part of the Magnificent Seven. And there’s a question, of course, whether Europe could essentially retaliate against that part of the US export bundle in order to induce essentially compliant behavior to push back against the act of aggression, and I think that hasn’t happened before.
It would be a very big deal if this were to happen, and it would happen within the context of actually there being a major technology shock through artificial intelligence that actually makes it potentially much easier now for European alternative suppliers to Google, Microsoft and co to potentially emerge in the marketplace, especially if there is a strategic push to push out the American tech infrastructure as part of and as induced by the US aggression. And so, that could be a very big deal that, again, we haven’t seen before. And how would this be done? Well, one mechanism is through essentially the suspension of intellectual property rights that the US is holding in certain domains, which is exactly on a more subtle and problematic dimension when it comes to national security. That’s one of the dimensions that’s often invoked in the discussion between why is it that the US is taking on this aggressive stance towards China. China has market access barriers for its tech market, so American tech firms have very little penetration in China, or actually no market access, and that might be seen as a distortion of the level playing field globally.
Matt Grossmann: We talked about the politics being different on the side of US trading partners this time, but what about the potential for the politics to be different in the US? The initial model really depends on a more traditional economic voting, where people are going to react if they are financially harmed by policies or retaliatory policies, but some of the rhetoric surrounding the latest round is really a lot more warlike, so we might… And Trump has said this is short-term pain for long-term gain, that we’re in a battle, especially with China. So to what extent should we expect anything more like a rally round the flag war phenomenon surrounding these policies rather than the direct economic effects being the most politically important?
Thiemo Fetzer: I do think that, again, the statement, states make wars and wars make states, so to say, there is, in my understanding, in particular when it comes to Canada, very strong reaction, in essence, very patriotic, so to say, that Canada doesn’t want to be bullied by its big neighbor and perceives it as such. And so, there might be huge shifts in public opinion vis-a-vis the US essentially undermining US soft power, the same time we have the cuts to USAID, that probably is going to help contribute to eroding the US soft power.
Again, if you zoom out of this, which I sometimes do, I see actually this isn’t that bad in the long run, and so on that one, I might not be in disagreement. But again, as I said, the potential for things to go wrong is just incredibly high. So we do see a rally around the flag effect, so to say, in the US. In Europe, there’s now a big push towards actually deepening the integration of the single market, there’s a lot of initiatives, with particular focus on the service sector, which is the bit where the US tentatively has a significant edge. And so, it could well be that, in the end of the day, we might end up in an equilibrium with a much more fragmented global economy in a way where the US might have the losing hand, because already now we see that these tariffs have left their mark on the stock market, US consumers are heavily invested in the stock market, and so we see that this will have an impact on consumption.
Again, this might affect the trade balance, so to say, positively, but people typically do not like when their standard of living is declining, or their perceived standard of living is declining. And so, there could be a huge backlash effect, simply through just the major revaluation of the US dollar and US dollar denominated assets that could arise.
Matt Grossmann: So one of the reasons that experts have been skeptical of these policies is because they’re often ostensibly about bringing back US manufacturing, and it doesn’t look like they’re well-targeted to do so. I know that you’ve studied the sector before and that you’re familiar with that literature, so give us just a sense of the relative role of trade policy in domestic manufacturing and why people are skeptical that this will make a difference in a positive direction.
Thiemo Fetzer: Yeah. In some level, one could think of the tariffs as being a type of stick or carrot, depending on what your perspective is. Technically, the tariffs will make the imports of certain goods more expensive and it could induce some firms to potentially arise in the US to produce said goods domestically in the US. Technically, this can work. There’s evidence that trade policy-induced or geopolitical escalation-induced trade barriers can induce some form of industrialization. There’s work that looks, going back into economic history, at the Napoleonic Blockade, that highlighted then how when the UK was essentially blockading the European continent because of Napoleon wars, subsequently, the French got very good at producing essentially spinned cotton, which is something that was one of the main export commodities or goods that the British would have exported and dominated the market. So going back in econ history, there is some narrative in this.
There’s also recent evidence, I have a paper where we look at trade in production networks, and particularly onshoring, that takes advantage of a natural experiment that took place in 2017 in the small country of Qatar, which kind of is the ideal context to study such a mechanism. Qatar at the time was importing the bulk of its food and low-value processed goods from its big neighbors, in particular the UAE and Saudi Arabia, and then in 2017, in June, from one day to the other, a simmering escalation or dispute escalated, resulting in a full-on trade blockade, whereby from one day to the other, essentially Qatar lost access to the main supply for a lot of stable commodities and basic goods, for example, dairy, dairy products, poultry and so on. And the Qataris within a relatively short period of time then started importing goods that are used to make dairy, so there’s these pictures of planes full of dairy cows that have been flown in. And within a relatively short period of time, Qatar was onshoring its dairy industry, when before, it was reliant on its neighbors.
But if we look at the example of the cotton from the econ history side and if we look at the example from Qatar, these are goods that are relatively low-tech, and so, well, in the grand scheme of things, they can involve a high-degree of capital investment, but actually, a lot of this, when we think of the type of goods that the US might want to onshore, it’s high-tech goods. And the US is already good at relatively low-tech goods, if you look at the US export basket, well, on one side, we’ve got airplanes, defense and high-tech, but then on the other hand, the bulk is agricultural goods and commodities. And so, it’s not clear to me whether we should put too much hope in that incentive and this mechanism, because ultimately, a lot of the high-value-added goods trade is very much a function. It requires long, long processes and periods to plan factories, to build supply chains, to build supply networks. These are things that do not happen-
Supply networks. These are things that do not happen overnight. And so at some level, I do think that the previous administration with the CHIPS Act and so on, and the incentives, the positives incentives that were used to encourage onshoring probably proved much more effective. And we saw a lot of at least investment announcements and some electric vehicle factories being built. It could be that now the uncertainty that is being created by the tariff measures and of the aggressive stance vis-a-vis trading partners that might be technology partners instead of helping rebuild US manufacturing might be alienated to invest in the US, and so this could backfire massively. And that’s actually one of my main concerns. And then, of course, there’s also still the very fundamental challenge that a lot of these high-tech products require access and reliable supply chains to certain raw materials. And again, this is where the different planning horizon might come in.
The Chinese have pretty much monopolized a lot of the markets, a lot of the goods that are vital, for example, to the green transition, rare earths, copper refining and so on, rare earths refining. They have a lot of control in the supply chain of these goods. And that could result in essentially the US, even though when they have established a factory, they’re still very much reliant on imports from China of certain goods and commodities to make, let’s say, higher value-added goods. So it’s a very risky strategy and I think that at some level, the positive incentives usually I would like to think work better than negative incentives. And again, these responses, again, I’ve not evaluated them from the supply chain view, but it does strike me as these instruments being very blunt, disrupting existing budding partnerships and supply chains that have been built up over the last two years of the Biden administration, and thereby actually causing harm if the ultimate objective is to re-shore and to onshore some of the production.
Matt Grossmann: So we know the country’s acted to hurt Trump politically last time, but we can also look at the initial policy announcements from the first term to understand their immediate effect. That’s what Omer was able to do.
So you studied trade wars during Trump’s first term in your new article in the Journal of Politics. Tell us about the major findings and takeaways.
Omer Solodoch: All right, so first a little bit maybe on our motivation. When we started this research, I think it’s important, when several studies like Kim and Margalit and Blanchard [inaudible 00:33:04], they showed that voters living in heavily affected areas turned against Trump and the Republican Party in the 2018 midterm elections. And what was also clear is the fact that the vast majority of voters didn’t live in heavily affected areas or work in targeted industries. And we were interested in the reaction of the broader voting public or the average voter who doesn’t live in those areas heavily targeted by China’s retaliatory tariffs. So to do that, we basically, we would employ an event study, and we focus on a very narrow window of time around mid-June 2018 when the US-China trade war escalated dramatically, and both sides announced substantial tariffs imports from the other trade partner. And we tracked three public opinion surveys that coincidentally were in the field conducting interviews right before and immediately after the trade war escalation.
And all three surveys also used random sampling. So respondents have an equal probability to be sampled before or after the trade war escalation. And this random sampling together with a narrow window of time around the trade war kind of makes us more confident that any shift we detect in public opinion is caused by the trade war escalation rather than other events or developments. And overall, what we find is a significant six percentage point drop in Trump approval immediately following the trade war escalation. And we find this effect across all three surveys, separate surveys, and across a wide range of modifications and estimation strategies. We also find this effect as electoral implications. We can talk about data a little bit later.
And importantly, I would say that we also find that this drop in Trump approval is accompanied by a drop in public support for Trump’s protectionist trade policy, and using possible outcome tests, we find no similar effects on other policy areas that were salient back then, like voters didn’t become more opposed to Trump’s immigration policy or foreign trade, sorry, foreign policy towards North Korea. And these two things were on the news back then, and instead, those voters turned against Trump and also against his trade policy. And together, these findings really give us a confidence that the effect we find is attributable to the trade war escalation.
Matt Grossmann: So remind us of what this looked like back then. What were these announcements that occurred while the surveys were in the field? What did it look like in terms of the tariffs that were imposed and the retaliatory tariffs that Americans were facing?
Omer Solodoch: So Trump’s first term tariffs were very high compared to the pre-2018 trade war situation, but much, much lower than the current trade war in 2025. And in 2018, there were several rounds of tariffs imposed by Trump and very quickly retaliated by China. We are talking about the first round around March when Trump imposed the tariffs on washing machine and solar panels. And then in June, we’re talking about the first round of tariffs imposed directly on China and again in September and October. So yeah, that’s about it, and before the midterm elections in 2018.
Matt Grossmann: So you try to differentiate between whether people responding to personal economic losses or the overall losses of the community. And related to that is this idea about whether people are responding to direct effects of the policies or just kind of the announcement of the policies and in response to that. So what do you find there?
Omer Solodoch: Right. So I think that a well of the evidence shows that there was a really quick reaction before any economic effects were realized. And we see it in our study, which like I said, focuses on a really narrow window of time around the trade war escalation. So there couldn’t be any really economic damage already materializing realized by voters. So we see the boarders really respond to their expectations, their worries, their concerns about the future economic pain that they might experience, them personally, but even more so the nation as a whole.
Matt Grossmann: And the other interview for this episode is about how the retaliatory tariffs at the time were quite well targeted to try to hurt Trump politically. So on the one hand, that might match your story, but on the other, you sort of find that they might not have needed to be all that targeted because people were kind of responding to the general trend and not necessarily only in their specific industries.
Omer Solodoch: Right. So I think we can think of that like the tariffs and the trade war had two separate effects, that together they amount to the total effect, the total public opinion and electoral effect of this trade war escalation and trade war tariffs. So one of the clear effects that we see in other studies is voters living in heavily impacted areas were targeted strategically by China and other trade partners. They became more opposed to Republican candidates in the 2018 midterms. But we also see that regardless of living in those heavily targeted or impacted communities, you can see a turn against Republican candidates and against Trump right after the trade war escalation in June.
Matt Grossmann: So you do also study the 2018 election. So what do you find there? And kind of put it in context of the overall electoral shift against Trump. How much of a difference did it make, and where do you come down relative to other estimates?
Omer Solodoch: Right. So I think that the effect we find is pretty, not only statistically significant, but also substantively large. We see six percentage points drop in Trump approval and a seven percentage points drop in voters intention to vote for Republican candidates in 2018. And by the way, they switch from Republican candidates to Democratic candidates. And it’s not like the reduction in support for Republican candidates is translated to no turnout or something like that. And it’s really difficult to say what were the implications eventually in terms of seats in Congress, but we do know from other studies that the more targeted effects, so the electoral effects on those voters living in heavily impacted communities, this amounts to about between, it depends how you calculate it, but between three to eight more seats in Congress without tariffs. So that’s from the literature. In our case, we only see these seven percentage points drop in support for Republican candidates.
Matt Grossmann: And how would this fit in with the larger literature on the effects of trade on incumbents and elections? Is this relatively large compared to what others have found? Is this the kind of effect that you would generally expect this kind of sociotropic immediate effect, or was this sort of a weird one-off?
Omer Solodoch: So I think it makes sense, and it’s not weird in these terms, but it is unique and it’s not exactly what we would expect from previous literature because previous work showed the trade is a low salience issue, low priority for voters. And voters rarely hold politicians accountable for the trade policies. And they know very little about the economics of trade and so on. But at the same time, another trend in the literature shows that voters do respond to import shocks, to trade shocks, the China shock. So we have these two different sets of findings, but again, those reactions, those effects that we do see in the literature is in those cases of voters living in heavily impacted areas. And what we find bridges these two different findings because we show that when trade is a high-salience issue, when there is a trade war, for example, then also the average voter, and not only those voters living in the highly affected communities, also the average voter reacts and responds to the trade war escalation, to trade policy and so on.
Matt Grossmann: So this is timely now because we’re back in a trade war in Trump’s second term. So tell us how the current trade war differs from the one that you studied in ways that might change its effects on voters, and I guess give us the current context. Does it look like Trump learned anything from the first round or that we’re just doing round two about the same way?
Omer Solodoch: Yeah, that’s a good question. I think that there are at least three major differences between the 2018 and the 2025, let’s say trade wars. The first one is in terms of magnitude. So when you go back to 2018, it seemed like these tariffs are really, really substantial, and the war in many ways. But when you look at, for example, tariffs imposed on China back then, and today in 2025, you see that today it’s just much, much more substantial. So that’s one difference. Another big difference is in speed or the timeframe. So back then, it was very gradual in 2018. There were several rounds of tariffs from April to March to October 2018, and then later on also in 2019. And all of these tariffs together, they don’t amount to the single liberation day in 2025.
So it was much quicker, much bigger. And also because of the sheer size of the 2025 tariffs, I think that they would, countries like China that imposed really heavy tariffs right now in retaliation, they would be much less able to be precise and to target specific electoral bases and to harm, for example, Trump’s base like the data show that they probably did in 2018. So because of the size of these tariffs, they could be less precise. So these are three major differences, I think. And in terms of what Trump have learned, maybe one thing that could point to something that Trump and his administration have learned is maybe start early, much before the midterm elections, hoping that perhaps the economic damage-
The economic damage wouldn’t be so bad somehow, maybe because some of the declarations want materials to actual policy, I don’t know. And then having voters going to the polls where a lot of other events are happening and maybe the trade war is not the first thing on their mind. So I think that might be one thing that could not minimize, but make the electoral ramifications for Trump and the Republican Party smaller. That seems to be one thing.
Matt Grossmann: Yeah. One piece of the effects in the literature has been about the retaliatory tariffs and how they were imposed, but your findings seem to suggest that maybe just the tariffs themselves could have a negative effect. So how about the global context? How are countries responding this time, differently than they did last time? And I guess how much does that matter? Or once the stock market has gone down a lot in the US due to the tariffs themselves or what people pay higher prices for consumer goods from China, that will be enough, just our own tariffs to have that negative implication.
Omer Solodoch: Yeah, I wouldn’t say that our study shows that people react only to those tariffs imposed by the US and not to retaliatory tariffs. In the mid-June trade war escalation, both events were very close to each other. So on June 15, Trump announced tariffs on China, and very quickly afterwards, China threatened and declared that they’re going to impose their tariffs on the US. So what we see is actually a voter’s reaction to the trade war escalation to the beginning of the direct, I think, economic clash between the US and China. Because before that, the Trump administration imposed tariffs on products, not specifically on China or other countries. So I think voters reacted to that. So it was a combination of both things, and it was a situation where the trade war escalated.
And I think that retaliation actually should be effective or should affect voters and how they think of Trump, the extent to which they approve the president, the incumbent more generally, and their voting outcomes, their voting behavior. Because if the president is able to get concessions from those trade partners, that might indicate something about competence and voters might be happy with that. So actually when China retaliates, I think that beyond the economic damage that this retaliation might cause, there is also something about the perceived competence of their own presidents, of the voters leaders in getting concessions. So I think it is an influential factor.
Matt Grossmann: So you mentioned that the timing is different this time, but the tariffs are larger. They’re also extremely high salience this time, and at least the initial indications from public opinion polls on trade policy itself are that people are moving against protectionist policies again. So what I guess would cause us to hypothesize that we get even larger effects this time, and what might minimize those effects?
Omer Solodoch: So I don’t think that we’re going to see a one-to-one increase in voters reaction as a result of the size of these tariffs. So it’s more about perceptions. And back then in 2018, the media covered the trade war escalation as a really dramatic shift, and no one knew that in 2025, it’s going to be even larger. So the headlines were as alarming as to date. So voters reacted to that more, I think. So we should expect similar reactions, but a steeper drop in the stock market, for example, or more panic, more concerns about markets, this could absolutely have more important effects, bigger effects. So in that way, I think we might expect a deeper dive in Trump approval and voters intention to vote for the Republican Party. I think we should expect a larger decrease in these outcomes.
Matt Grossmann: Perhaps surprisingly at the time, the Trump tariffs against China ended up being extended into the Biden administration, and it seemed like maybe the Democrats were also saying that although trade wars are not that popular, maybe kind of opposition to China is popular. So I’m just wondering if we might ever see a rally around the flag effect if this is really considered more like a war-like posture between the United States and China. In other words, is it possible to get big enough on enough fronts that this just ends up being sort of support for the incumbent against the foreign enemy?
Omer Solodoch: I think that’s an interesting point, and I would say that we see a rally around the flag in another set of economic warfare tools, and that is in economic sanctions. So we see these rally round the flag effects when countries impose sanctions, economic sanctions from other countries. And usually the public rallies around the leader and the support for the incoming government just rises in response. We don’t see it in the trade war, at least not in the US-China trade war in 2018. And I think what might explain this difference is first, American voters I think attribute the blame for the trade war to Trump, more clearly because he initiated the first round of tariffs and so on. So that’s one difference.
And the other difference is, again, retaliation, right? So when there are economic sanctions, there is this dynamic of just one side getting the sanctions, and that’s it. Here we’re talking about the US for example, first, like I said, Trump initiates and imposes sanctions on countries like China, and then also appears incompetent when he doesn’t manage to get concessions from China. So this different dynamic here maybe doesn’t allow for a rally around the flag effect and is more consistent with an anti-incumbent effect.
Matt Grossmann: So Trump has said that we should expect short-term pain, but that it’s in the service of long-term gains. And this does fit a little bit with the partisan business cycle literature, which suggests that Republicans are less likely to sort of stimulate the economy right when they get in, but maybe actually have better electoral timing, including our four-year election cycles, if not the two-year ones. What do you make of that potential? Is there a chance that, yes, there’s an immediate negative effect, but the possible long-term effects are either small or reversed?
Omer Solodoch: That’s a difficult question. I think there is some evidence that voters, and especially those protected this time by US tariffs, they turned towards Trump. But I think the evidence is not so strong that this increase in their vultures in support of Republican candidates, for example, in 2018 midterms, they were quite weak and usually not statistically significant, at least in one study that was conducted on this issue. So overall, is it possible that some voters, especially those protected maybe by those tariffs imposed by the US on imports from China and other countries? Yes, it is possible. But again, I would just suggest that it’s not the majority of voters working in these industries, living in those communities that are heavily dependent on these industries and so on. So overall, I think that I would expect that the more major, the stronger force here would be economic concerns on the direction of the nation’s economy.
Matt Grossmann: So there’s not a lot of political upside we’re seeing here, and most of the economic literature suggests pretty negative implications of these policies. Even the people who are most sympathetic to reducing the value of the dollar or trying to increase US manufacturing don’t seem to like the way that these are actually being implemented. So no one’s claiming kind of ownership of this intellectually. Is that your same read of the literature, or is there something that might kind of explain or justify these moves in the economic literature?
Omer Solodoch: Yeah, that’s my interpretation as well. I think that there is no sense in imposing, no economic sense in imposing tariffs based on trade deficits for example. There is a case to be made about imposing limited and precise tariffs on strategic assets, on strategic industries, but this seems to be very far from the case of Trump’s trade wars. So overall, I think that’s my interpretation as well.
Matt Grossmann: So you have also been involved in this larger debate in academia about the relative role of economics and culture in changing voters preferences. And a lot of this occurred surrounding the 2016 election in the US where there was some evidence that trade-affected areas turned against Democrats, but also obviously some evidence of cultural view change as well. And I know you’ve studied this in Europe as well, so kind of fit what we’ve been talking about in this kind of broader literature question of not just kind of the relative importance of economics and culture, but in sort of how trade effects fit into this set of changes with the rise of populism and the growth of the education divide.
Omer Solodoch: Yeah, good question. So I would say that both economic and cultural factors matter, and in many cases, they interact as we know in ways that make it very difficult to identify and separate them. And even though I tend to assign more weight on, like you mentioned, on a set of cultural factors, I also think that economic factors matter a lot and that the balance between them, between the set two sets of factors, economic and cultural could change over time. And for example, in response to financial crises or trade wars. So we might see a greater impact of these set of economic factors. And overall, I think that, trying to think of the trade war and this question about economic and cultural origins of support for populist parties more broadly, I think that one of the most intriguing stories happens along the rural urban divide, I would say, and specifically the cultural resentment in rural communities is clear. We have clear evidence on that.
But also it seems that these rural communities, they respond to economic shocks, to trade shocks, to the trade war in ways that are very consistent with their economic interests. So it’s very interesting to see, to try and learn more about this interaction of culture and economic interest in the case of rural borders. I believe in the rural urban divide in the US trade war case and elsewhere.
Matt Grossmann: And anything that didn’t come up so far that you wanted to leave us with or tout about what’s next?
Omer Solodoch: Maybe just for a last point, I would say that in many cases we see that consumer interests almost never affect borders attitudes toward trade. So people consider the impact of trade on jobs and so on. Very rarely they really act consistent with their interests as consumers. And I think that right now, when we’re going to witness this huge shock in terms of inflation caused by tariffs, maybe it’s a perfect storm and the most likely case where voters should act consistent with their consumer interest. If it’s a force that sometimes affect trade references and voter behavior, then maybe it’s a perfect storm to test this hypothesis. That’s maybe my last point on this.
Matt Grossmann: There’s a lot more to learn. The Science of Politics is available bi-weekly from the Niskanen Center. I’m your host, Matt Grossmann. If you like this discussion, here are the episodes you should check out next, all linked on our website. Did Chinese Trade Competition Increase Nativism and Elect Trump? Why Foreign Policy is Still Bipartisan. Will a Good Economy Save Trump? Does the Biden Economy Have Bad Election Timing or an Unfair Fed? And Inflation Hurts Presidents, and it’s Not the Media’s Fault. Thanks to Thiemo Fetzer and Omer Solodoch for joining me. Please check out Tariffs and Politics and Political Costs of Trade War Tariffs, and then listen in next time.