The election results present a puzzle of sorts. On the one hand, voters expressed deep resentment towards inflation, under the belief that Biden contributed to rising prices, failed to address them, or both. On the other hand, Trump’s signature economic policy is tariffs—on imports from Mexico to Canada and now Israel—which most economists believe will raise prices. Why are voters, who are ostensibly so sensitive to high prices, willing to give Trump a pass on an obviously inflationary policy?
When I have posed this puzzle on Twitter, the standard neoliberal voices—from Jordan Weissmann to Eric Levitz to Matt Yglesias (aka “The Vox Boys”)—suggested that my brain is small. (Yes, the same Levitz who leaned entirely on an economist to interpret a contract for the counterintuitive proposition that insureds were immunized from Anthem’s proposed and now-retracted policy to restrict anesthesia coverage.) The Vox Boys reckon that voters put everyday low prices above all else. To believe this, however, you must also believe that voters don’t understand the implications of tariffs or don’t believe Trump will follow through with his threats. This neoliberal explainer is fairly unsatisfying, however, as it requires one to believe that voters are stupid.
An alternative explanation, which infuriates the Vox Boys, is that while voters care about low prices, they also care about other things like preserving blue-collar manufacturing jobs or supporting local businesses. To wit, voters tend to punish Democrats for removing trade barriers: A 2020 American Economic Review paper showed “trade-impacted commuting zones or districts saw an increasing market share for the Fox News channel (a rightward shift) … and a relative rise in the likelihood of electing a Republican to Congress (a rightward shift).” This desire to protect local businesses animates much of the New Brandeisian movement, which rejects the consumer welfare standard in antitrust, by among other things, recognizing harms to workers or small businesses.
Following the advice of her corporatist advisors like Tony West, Harris elected to attack Trump’s tariffs from the right, highlighting how the tariffs could raise prices. Indeed, the Harris campaign tweeted a video of Washington Post columnist Catherine Rampell bashing Trump’s tariffs, a few weeks after Rampell called Harris’s price gouging proposal “communism.” These attacks moved exactly no one in Harris’s direction. And no wonder: The Democrats are supposed to stand up for labor, who are the biggest beneficiaries of tariffs, especially those who work in the tariff-protected industries. Progressive advocates like Zephyr Teachout were calling for a recalibration on the anti-tariff message, but were ignored. Another victory for the Vox Boys and Girls!
When I pointed out that Trump managed to purge the neoliberal free-trading ideology from his party’s platform, appealing smartly to voters who care about jobs as well as low prices, Levitz quote-tweeted a screen shot of his summary of a 2019 study (and a link to his Vox article), purporting to show that that American exporters that were most exposed to Trump’s tariffs on their inputs—think steel, aluminum, solar panels, and various Chinese goods—experienced lower export growth in 2018 and 2019 than exporters who were unaffected by the duties. Per the Vox Boys, tariffs create harms beyond higher prices.
Before getting into the details of the study, let me note two obvious things. First, from a political perspective, the welfare of large traders engaged in importing and exporting (aka “trading firms”) doesn’t get much play in election conversations; so this anti-tariff argument will again fall on deaf ears. Second, one can’t evaluate a tariff from a cost-benefit perspective without also studying the beneficiaries of the tariffs. By focusing on the welfare of trading firms, however, this study implicitly downplays the welfare of workers whose jobs were protected by the tariffs.
Regarding the merits of the underlying study (available here), the focus on the impact of Trump’s tariffs on exporter growth is curious. If larger or faster growing exporters were more exposed to the “treatment,” then their growth would be expected to slow relative to the “control” group (smaller exporters not exposed to Trump’s tariffs); it’s easier to “grow” from a smaller base. Indeed, the authors acknowledge the difference in the size of the two study groups at page 2:
We find that U.S. importers facing import tariff increases employed twice as many workers compared to the average importing firm and about nine times as many workers as the average firm. Similarly, we find that U.S. exporting firms facing retaliatory tariffs were more than three times larger than the average exporting firm. Thus, the tariff increases hit the very largest trading firms in the U.S. economy.” (emphasis added).
Figure 3 of the study shows that cumulative growth rate in exports for the treatment group exceeded the control group in the two years leading up to the tariffs, with the gap between the two shrinking in each month. It stands to reason that, even absent the tariffs, the growth rates of the two groups would have naturally converged. In fact, the two trendlines differed by approximately 10 log points in early 2016, with most exposed export sectors exceeding all other export sectors by a comparison of 3 log points to -14 log points, respectively. This difference shrank to nearly zero by the beginning of 2018. The reversal that occurred after January 2018 reflects a continuation of the opposing pre-tariff growth directions. Yet the use difference-in-differences (DID) estimation to recover a causal effect, as the authors of this study intended, critically rests upon the “parallel trend” assumption—namely, that had the treatment never occurred (i.e., tariffs had never been imposed), the relationship between the treated and control groups would have remained constant over time. But the authors casually mention parallel trends just once in a footnote, claiming “Figure 3 suggests parallel trends in the months prior to the trade war.” While that statement might be true for the few months right before the Trump tariffs, it ignores the plainly obvious longer trend of convergence. Violation of the parallel trends assumption can bias the estimated effect, undermining the researcher’s ability to ascribe a causal interpretation to the treatment.
Finally, the magnitude of the effect, assuming it’s properly measured, doesn’t sound debilitating for large exports. The authors find a decrease in “log points” of around one (slightly smaller in 2018, slightly larger in 2019), which can be interpreted as a percent change for small differences. By comparison, exports were growing by between four and six percent in the year leading up to Trump’s tariffs, per Figure 1. A decline of one percent in the growth rate of exports for the largest trading firms that import tariff-affected inputs might be a small price to pay for protecting jobs and domestic industries.
Focus on the jobs
Levitz also points readers to a 2019 staff working paper at the Federal Reserve as evidence that Trump’s tariffs harmed workers. Setting aside any infirmities in the estimation or interpretation of results, at least this study focuses on a meaningful outcome variable. The staff working paper purports to show that U.S. manufacturing industries more “exposed” to tariffs lose more jobs from rising input costs (channel one) and retaliatory tariffs (channel two) than jobs gained or preserved from import protection (channel three). Exposure to import protection for a given industry is measured as the share of domestic absorption of that industry affected by newly imposed tariffs; exposure to the other two channels is measured similarly. It follows that for any given industry, exposure along these three channels could vary dramatically.
This study also uses a DID method to uncover the effects of the tariffs. The authors note the “issue of differing trends across industries prior to the implementation of new tariffs”—an admission that parallel trends may not be satisfied—and seek to address it by (1) removing industry-specific trends in 2017, or (2) differencing out the pre-trend path for each coefficient. After these various contortions, they find that “shifting an industry from the 25th percentile to the 75th percentile in terms of exposure to each of these channels of tariffs is associated with a reduction in manufacturing employment of 1.4 percent, with the positive contribution from the import protection effects of tariffs (0.3 percent) more than offset by the negative effects associated with rising input costs (-1.1 percent) and retaliatory tariffs (-0.7 percent).” (emphasis added). But this begs the question: What single industry would make such an equivalent move on each of these channels? If China is expected of dumping (say) solar panels, and Trump slaps a tariff on solar panels from China, why would the solar panel industry (now exposed to the import protection channel) be equally exposed to (say) rising input costs?
It would have been helpful for the authors to identify the aggregate employment effect across the three channels for any given industry. Were there industries with net job gains resulting from Trump’s tariffs? To wit, if an industry was only exposed to the import protection channel—that is, no input costs were increased by other tariffs and there was no retaliation for the industry in question—the best estimate of the jobs effect would be positive! By showing the size of the coefficients of the three channels for equal shifts in channel exposure, however, the authors have made it difficult to assess the economy-wide effects as well. We only know (assuming the specification is proper) of the relative magnitudes of the employment effects given a one percentage point exposure to each of the three channels. Tariff bashers will interpret the coefficients as if they can be summed up, but that is only for a hypothetical industry that experienced the same increase in exposure across all three channels.
This is not meant to impugn the integrity of either study. All empirical studies can be criticized. Rather, it is meant to suggest that the Vox Boys have found two studies that tell their story of tariff-induced harms and have decided to pump them up. But neither study materially advances the economic argument against tariffs.
In summary, the neoliberal critique of Trump’s tariffs finds little support in economics or among voters. The Vox Boys and Girls, who myopically focus on low prices over all other considerations, should be ignored. And Democratic Party should recalibrate their approach to tariffs, recognizing that, to be considered the party of labor once again, promoting labor interests should be their loadstar.
Washington Post columnist Eduardo Porter, in his recent piece, “Corporations are not destroying America,” seems to be taking cues on economic policy from his colleague Catherine Rampell. To Porter’s credit, he, contra Rampell, seems to actually read the materials he’s writing about. Yet the entire piece is emblematic of columns favored by the Post: ones that casually brush aside progressive policy ideas by dispatching with a straw man and infantilizing anyone to their left rather than engaging head on.
Porter takes issue with the “idea that American markets have become monopolized across the board, with dominant companies raising prices at will,” which he calls “ludicrous.” But that’s really not the point that progressives—or the Harris campaign, who Porter’s “memo” is addressed to—have been arguing. Could you find someone who thinks that every market in the country is overly concentrated with greedy fatcats leeching off the public? Absolutely. But it’s not the argument that progressives in the policy community and Harris have articulated.
The actual argument is not that every industry is overly concentrated, but that a number of key industries are, which enables opportunistic price gouging to ripple through key sectors of the economy and cause acute harms because of how critical those industries are. These are industries like meatpacking, airlines, credit-card issuing and processing, railroads, Internet search, ocean shipping, and baby formula. Porter can cite all the studies he wants that say overall concentration levels are not soaring or concerningly high and still not get to the substance of the argument. That said, there are numerous methodological and theoretical questions about measuring market concentration. But that’s hardly necessary here; the examples above are so obviously oligopolistic that there’s really no need for formal measurement.
Similarly, Porter simply misrepresents what opponents of corporate power claim big firms with pricing power are doing. He frames it as a matter of “dominant companies raising prices at will,” but the throughline in nearly all versions of progressives’ arguments is about companies leveraging particular disruptions, like inflation, as a smokescreen to exploit customers by raising prices in excess of the increased costs they’re incurring. Harris’s grocery price-gouging restriction has been much maligned by Rampell and other neoliberal pundits who usually either paper over that it only applies in emergencies (and only in the food industry) or elide it with a slippery slope “but what is an emergency?” distraction, preferring to hyper obsess over fringe cases when most of the time it will be pretty clear if we’re in an emergency; Harris’s plan specifically is inspired by the real-world exploitation of large groceries, as found by the FTC and revealed in the Kroger-Albertsons merger trial, during the last disaster.
Porter’s column also pointedly veers into an argument that narratives about “the monopolization of America often rests on evidence about corporate concentration at the national level. But the market relevant to consumers is, in many cases, local.” And that’s true to an extent, but Porter’s own example of hardware stores (directly following this) is a good demonstration of why this elides the harms to which progressives are trying to draw attention. Porter discusses “mom and pop” hardware stores vis-à-vis national chains Home Depot and Lowe’s (apparently he’s not a fan of ACE). But there are no mom-and-pop airlines or baby formula manufacturers or oil firms.
As far as the corporate big tech “Godzillas,” as Porter terms them, apparently they aren’t “squashing competition to reduce wages, keeping new rivals from entering their markets, and sticking it to consumers.” (Let’s ignore for a moment the very serious allegations that Uber and Amazon have suppressed the wages of their drivers. Or the recent verdict by a jury in Epic v. Google that Google overcharged app developers for transacting on the Play Store. Or the recent finding of a judge in U.S. v. Google that Google monopolized the search industry and the associated market for text advertising.) Rather, Big Tech’s fearsome power has come, per Porter, “first and foremost from deploying new technology and offering better value to customers.”
Presumably then, Facebook didn’t go buy out a bunch of other social media and networking websites and softwares. But it did. It bought ConnectU in June 2008, FriendFeed in August 2009, sharegrove in May 2010, Hot Potato in August 2010, Beluga in March 2011, Friend.ly in October 2011, and Instagram in April of 2012. Definitely not buying out competitors to sit on a huge pile of market share, just good old innovation!
And Amazon definitely got to where it is by besting its rivals through good old market competition! Except when it bought dozens of other online retailers and web service companies. Oh and when it artificially skewed its marketplace to disadvantage third party retailers in favor of Amazon’s own products, many of which were reproductions of other companies’ products.
Porter also seems to have overlooked the fact that Google acquired its way to power in the ad stack, gobbling up, among others, DoubleClick, Invite Media, and AdMeld.
The entire piece is vapid left-punching; Porter even actively agrees with the substance of the critiques against monopolists. For instance, this paragraph:
For sure, corporate consolidation has reduced competition in some markets. It was probably a bad idea to allow Whirlpool to take over Maytag in 2006, or to allow Miller to merge with Coors two years later. Hospital mergers deserve a much more skeptical view than they have received in the past. Let’s be careful about drug manufacturers buying out competitors for the purpose of killing a potential competing drug.
Yes, there is discourse about the American economy becoming more consolidated writ large, but the core of the debate is about intra-industry dynamics where market power creates unique opportunities for profit at the expense of consumers. Like when pharmaceutical companies buy out a firm working on a competing treatment. In sum, we have yet another piece to file away in the classic genre of “WaPo doesn’t have anything substantive to add, but feels the need to put down uppity leftists.” (It’s only gotten worse after new data revisions showed an even sharper uptick in corporate profits than earlier data had indicated.)
The last thought Porter offers is that “taking a wrecking ball to big business in the service of a rickety theory of harm will do everyday Americans no good.” What is the wrecking ball? Who is proposing to destroy major corporations?
The entire case laid out in this “memo” is that Harris, at progressives’ behest, is calling for leveling all large corporations. But Porter would do well to remember that the story of Rampell’s he links to in order to outline the consequences of this “wrecking ball” literally lies about what Harris’s pricing gouging law narrowly targeted at suppliers in the food industry (as opposed to all large companies) would look like. There are no “price controls” in the proposal, contrary to Rampell’s suggestion; food suppliers could justify price hikes during the next crisis so long as they were cost-justified. Rampell says Harris’s plan is likely to be modeled after a bill from Senator Warren (D-MA) that would give the FTC power to punish firms for price gouging.
As a justification for why such a law would be disastrous, Rampell says that “the legislation would ban companies from offering lower prices to a big customer such as Costco than to Joe’s Corner Store, which means quantity discounts are in trouble.” No, it doesn’t. There is absolutely no prohibition against quantity discounts. Rampell is, at best, warping this line stating that one standard that would give a firm “unfair leverage,” which would make it presumptively in violation of the statute, is when it “discriminates between otherwise equal trading partners in the same market by applying differential prices or conditions.”
If two firms are buying vastly different quantities of something, they are not “otherwise equal trading partners” of that supplier. What that line actually means is that if Joe’s Corner Store and Bob’s Corner Store both order identical shipments of widgets but Widgets Incorporated charges a much higher rate to Joe than Bob, then that is unfair pricing.
This is not the first time that Rampell has grossly misrepresented what Warren’s legislation says. Back in March, she said that it, and a similar bill from Senator Casey (D-PA), would be “[f]orbidding companies from changing the prices and sizes of everyday products without government say-so.” Neither bill said anything remotely like that. The fact of the matter is that Catherine Rampell is so committed to lashing out at the left that she either doesn’t bother to read the things she’s complaining about or is happy to just outright lie. Whatever the case may be, Porter should look elsewhere for insights about the economy.
Dylan Gyauch-Lewis is a Senior Researcher at the Revolving Door Project. She leads RDP’s Economic Media Project.
Did you ever notice that the same neoliberal economists are quoted routinely by economics reporters in the mainstream press? Take Ken Rogoff. He guest authors pieces on public policy at Brookings, is a professor at Harvard, semi-frequently authors op-eds, and is widely quoted in the media. While not quite as high profile as his colleagues Jason Furman and Larry Summers, Rogoff has been extremely impactful.
To give you an idea, I have compiled some statistics on the number of times famous economists have been quoted in the New York Times and Wall Street Journal since January 2020.
Economist | New York Times | Wall Street Journal |
Jason Furman | 187 | 156 |
Larry Summers | 153 | 164 |
Angus Deaton | 50 | 12 |
Joseph Stiglitz | 42 | 16 |
Kenneth Rogoff | 20 | 34 |
Isabella Weber | 8 | 6 |
As the table shows, Rogoff has been quoted 20 times in the New York Times since the start of the pandemic, trailing Nobel prize-winning progressives Angus Deaton and Joseph Stiglitz by only a small margin. Rogoff’s quotes in the more conservative Wall Street Journal exceeds these progressives, despite their international acclaim. (The table also shows the dependency of these papers of record on Furman and Summers, two Obama-appointed centrists—and disciples of deregulator extraordinaire Robert Rubin—who often reject the progressive policies of Biden.) In any event, the 34 quotes from the Wall Street Journal in a little over four years is an impressive display of influence.
In 2010, Rogoff co-wrote, with Carmen Reinhart, a paper titled “Growth in a Time of Debt” that came to define the acceptable boundaries of fiscal policy in the 2010s. And, while Rogoff has complained about being dismissed as an austerity-peddler, the fact remains that he and Reinhart became the go-to citation for governments when they slashed welfare spending and imposed sharp cost controls. The analysis that Rogoff and Reinhart (R&R) lean on was flawed from the start, however, and, for anyone without an Ivy League professorship, their oversight probably would have been career-ending. Despite efforts to substantiate his claims about the relationship between debt and growth rates in more recent work, huge methodological and theoretical issues remain. That Rogoff continues to be treated as a credible voice on economic issues is a striking indictment of our media ecosystem.
A Fundamentally Flawed Study
“Growth in a Time of Debt” was published to great fanfare when it came out with the financial crash of 2008 just barely in the rearview. The paper claimed to find a damning reason to pump the brakes on aggressive debt-financed government stimulus programs: When a country’s debt exceeds 90 percent of GDP, R&R asserted, its growth rate takes a massive hit, estimated at a drop of roughly three percentage points annually compared to countries below the cutoff—from 2.9 percent growth for countries with ratios between 30 and 90 percent to -0.1 percent growth for countries with ratios above 90 percent.
When a student and two professors at the University of Massachusetts—Thomas Herndon, Michael Ash, and Robert Pollin (HAP)—failed to reproduce those findings, they dug into the data and in 2013 found something else instead. R&R had made significant errors in their Excel sheet and sampling that inflated the number. R&R’s calculations excluded several years of data from New Zealand which, when included, lowered the difference in growth rates for countries above and below a 90 percent debt-to-GDP ratio from around three percentage points to just one percentage point. As noted above, R&R estimated real GDP growth to be -0.1 percent for countries with more than 90 percent debt-to-GDP; after correcting R&R’s inaccurate data, the UMass researchers found that the real figure was 2.2 percent. After corrections, the difference between real growth rates for countries above the 90 percent threshold, compared to countries with ratios below 90 percent, shrunk from three percentage points to one percentage point.
Rogoff and Reinhart did, in fairness, admit the error and correct it, making the same argument but with less dramatic figures. With other colleagues, they also produced more work that continued to show a similar trend. Even absent computational issues, however, there are still methodological issues and theoretical shortcomings that they never overcame.
For a start, R&R made a causal claim based on only correlational data, as several economists have pointed out. It could be the case that weaker growth leads to more government debt, rather than the reverse. Additionally, that R&R largely treated debt levels as a binary—either equal to or above 90 percent of GDP or below 90 percent of GDP—rather than a continuous variable could play a role. If they were to use debt levels as a continuous variable, they could model a relationship that reveals how each additional point of the debt-to-GDP ratio correlates to growth rates. Their method, however, merely sorted countries into two buckets: those with a “debt overhang” (their term, seemingly coined here, for when debt/GDP is greater than 90 percent) and those without. Then they more or less just took the averages (the averages were country-weighted in the original R&R).
R&R’s arbitrary 90 percent threshold is also worth discussing. For a start, the way that this threshold is determined is somewhat ambiguous, but it seems pertinent that when R&R published their most influential paper in 2010, that was a level that seemed to be fast approaching for many wealthy countries. Yet when HAP corrected the computational errors, the entire difference in growth rates was determined by the extreme outliers—that is, countries with debt-to-GDP ratios below 30 percent or above 120 percent. The UMass paper showed that, without those extreme outliers, there was no longer any strong correlation between debt and economic growth.
Another issue with the use of debt-to-GDP is that it does not account for government assets. Governments can raise revenue at any point by selling off ships, planes, tanks, land, buildings, and more or by selling intellectual property rights or exclusive leasing or permitting to companies. That there isn’t a serious effort to do so after countries hit that 90 percent threshold (or even go well past it) seems to indicate that in practice, the impacts of a debt overhang are preferable to taking extreme measures to stay below the red line.
Now R&R have insisted that they were never pushing austerity—and in fairness, they did include the caveat that fiscal stimulus should have been rolled back slowly. They never seemed to mind, however, that they were made famous by politicians like former Speaker of the House Paul Ryan and former British Chancellor George Osborne, who constantly cited R&R as a reason to impose austerity as rapidly as they could. In the wake of the controversy created by HAP’s 2013 debunking of “Growth in a Time of Debt,” journalist John Cassidy pointed out how R&R’s protestations in response to the austerity-pusher critiques completely clashed with the way R&R had marketed the paper when it was first published. Indeed, R&R were among 20 economists who publicly backed Osborne’s austerity policy in an open letter to The Sunday Times in February of 2010.
The revelation of terrible data management and the defensive response from R&R was enough for Cassidy to question whether listening to any economists at all was worthwhile. You would think, at the very least, economic reporters would discount R&R’s opinions on fiscal matters, but they haven’t even done that. For a mistake that would have likely derailed anyone with a less impressive pedigree, R&R have bounced back, still producing research, still pushing a (somewhat) toned down version of their argument from “Growth in the Time of Debt,” and still opining on economic policy in the media. Rogoff, in particular, is still quoted aggressively (as shown in the table above) and is using the new era of high interest rates to try and resurrect his old pet theory.
Old Wine in New Bottles
To be clear, unsustainably high levels of debt can be extremely problematic. But remember, that is not what R&R were saying. They were arguing that countries that experienced a debt overhang suffered long-term (economically significant) negative effects to their growth rates. Trying to use present circumstances to say, “See, we were right all along and all those people who decried our calls for fiscal responsibility were fools,” which is essentially what Rogoff has argued in a couple of op-eds this year, relies on a mischaracterization of both what R&R actually said and what their critics argued. In particular, Rogoff points to Adam Tooze using the word “austerity” over 100 times in his book Crashed. It is true that not all fiscal responsibility is austerity, and that not all austerity is fiscally responsible. But when Rogoff says things like Biden and Trump would both “blow up the debt,” he’s clearly hinting that mere “responsibility” is not all that he’s after.
After President Biden’s 2024 State of the Union address, Rogoff told Bloomberg that “Biden’s speech suggested blowing up the debt.” This is simply false, as Biden called for his policy proposals to be funded by higher taxes on the wealthy. Plus, this story ran after the president released his budget proposal, which includes cutting the deficit by $3 trillion annually. Maybe, Rogoff was interviewed before that, but anyone serious about advising policy as a neutral expert would have offered an updated statement. Would the debt continue to grow under Biden’s plan? Yes. Would it “blow up”? No.
And while Rogoff also asserts that Trump would likely blow up the budget, he includes the caveat that “we really have no idea what Donald Trump will do.” Anyone who creates this level of false equivalency between Biden and Trump on responsible budgeting is either oblivious or a total hack. Rogoff is known for Republican leanings; he advised John McCain in 2008 and reportedly “warmed up” to Trump after he took office.
Add it all up and we have a conservative economist who helped create a global push for austerity trying to resurrect that narrative, implying that Biden is no better than Trump on budget issues.
Rogoff’s legacy is one of creating cover for conservative governments to prematurely abandon fiscal stimulus, leaving millions of people out of work. What rocketed “Growth in a Time of Debt” to its high status among economists was how clear and dramatic it found the risk of high debt to be. That was proven to be bunk. But it was deeply rooted in the ethos of the austerity movement, so much so that the hawks at the Committee for a Responsible Federal Budget felt the need to defend their own position in the wake of the R&R controversy. Why is Rogoff still in reporters’ rolodexes?