Professors Kathleen DeLaney Thomas and Erin Scharff discuss fake tax news and the effect it has on public policy.
This transcript has been edited for length and clarity.
David D. Stewart: Welcome to the podcast. I’m David Stewart, editor in chief of Tax Notes Today International. This week: unbelievable.
The tax field and tax system are often misunderstood by the general public. While admittedly it can be hard for anyone to wrap their head around the U.S. tax system, misunderstandings often color how people feel and think about tax policies or changes.
Add in intentional misinformation campaigns, often found on social media, and there’s a lot of false and fake tax news being spread around. Our guests this week took a close look at how fake tax news spreads and its effect on tax laws and policy.
Joining me now to talk more about this is Tax Notes contributing editor Robert Goulder.
Bob, welcome back to the podcast.
Robert Goulder: Thanks for having me, Dave.
David D. Stewart: Could you tell us about your guests?
Robert Goulder: Yes, two professors: Kathleen DeLaney Thomas with the University of North Carolina School of Law in Chapel Hill, and Erin Scharff with Arizona State University Law School in Phoenix. They have an article out called “Fake News and the Tax Law,” published in the Washington and Lee Law Review , and it’s fascinating stuff.
David D. Stewart: What sort of issues did you get into?
Robert Goulder: Well, what they did is interesting. They painstakingly compiled a data set of fake news, subjecting it to objective parameters to figure out what actually is “fake news” and filtering out things that are opinion, filtering out what you might call “false news,” distinguished from fake news.
They did all this, they had the status set, and then they analyzed it, searching for trends, commonalities, and recurring themes. Unsurprisingly, one of their main conclusions is just what you alluded to in your introduction — there’s so much complexity in our federal income tax system that it is perpetually ripe for misinformation.
David D. Stewart: All right, let’s go to that interview.
Robert Goulder: Kathleen and Erin, welcome to the podcast.
Kathleen DeLaney Thomas: Thank you so much.
Erin Scharff: It’s a pleasure to be here.
Robert Goulder: Now, one thing that occurs to me is right off the bat in your article you say taxation, this whole field, it is just ripe for misinformation, whether that is intentional or inadvertent. And it’s been that way for a long time, predating the onset of social media. Why is that? Why is taxation so vulnerable to misinformation?
Kathleen DeLaney Thomas: Tax is hard. As we all know, tax is an incredibly complex, complicated subject. Most people don’t have a very in-depth understanding of it, for very understandable reasons, because it’s so complicated. Because people don’t bring a lot of prior knowledge, it’s easy to mislead and misinform. As you said, this is an age-old problem that we’re looking at because that has essentially always been the case, although we argue — and maybe we can get to it a little bit later in the discussion — that we think it’s only gotten worse, or certainly hasn’t gotten better.
Erin Scharff: I would add that it’s a hard system that everyone has to use. Not everyone needs to understand how we get rockets to the moon; it’s confusing and complicated to do that physics, but we don’t have a sense that we understand it. We have a sense that that’s rocket science. Because taxpayers actually regularly interact with the code, they have a desire to understand it, and it is super complicated.
Kathleen DeLaney Thomas: The last thing I’ll just add to that is even the way tax information gets communicated by reputable sources sometimes is wrong. The popular press makes technical mistakes. Again, this all goes back to complexity. But even when there’s no malicious intent, this is just an area where — again, understandably — a lot of mistakes happen, and so it’s easy for false information to spread.
Robert Goulder: Your article identifies some classic examples of this. One concerns marginal rates, right? There’s just a lot of people who simply don’t get how marginal rates work. And there’s a lot of people who really fail to grasp what their average rate is. They might think they’re paying a much higher percentage than they actually are. Could you elaborate on some of those classic misperceptions?
Erin Scharff: I teach average versus marginal rates every year in my tax class when I teach basic introductory tax. Because if you’re going to leave knowing one thing, having taken a federal income class, you should leave being able to understand the difference in average and marginal rates.
I think that complexity is a whole bunch of things related to the tax code. A piece of it is Americans have a phobia of math, [at least] in my relationship with my students. Just the idea of taking an average is, by itself, I think, harder than we recognize for many people.
You add on top of that politicians constantly talking about a particular rate. If you go and look at the rate brackets themselves, you don’t see the zero rate that is the standard deduction if you’re going to look [at] it on a sheet of paper. So the idea that there’s this disjuncture already between taxable income and the rate at which it’s applied, those are complicated things.
On top of an America [where] I think there’s a reasonable number of people who are just like, “Averages? Yikes!” You can’t see my facial expression because this is a podcast, but I’m doing a facial expression that has that “yikes” look.
Kathleen DeLaney Thomas: A lot of this information, things like tax rates, there’s literature showing this is the case with the deductions as well. If something’s highly salient, it’s an idea that’s out there, people are used to hearing it a lot. That’s what they know about, and that’s what they often think tends to apply to them.
I might have an idea of what the top marginal rate is because that’s a number that I hear or that I see posted online or I read it in a newspaper, and that’s my only concept of what a tax rate is. As you alluded to, there’s studies on this showing that people don’t really have an understanding — again, I think this is very understandable — between the difference from a marginal rate, which is just that rate on the very last dollar of your income, and your average rate, which is essentially how much tax you pay divided by all your income. Erin mentioned this requires math, but even conceptually, the average person doesn’t necessarily have an appreciation for what these things mean.
This is one of the most basic parts of the tax law, right? In terms of our interactions with it every day, how much tax do I owe? How much do I pay in taxes? Studies show a lot of people really don’t have any idea. When they’re asked to estimate what their average tax rate is, it’s not in line with [what] their actual average tax rate is. They might have, again, these salient numbers in their mind they’re used to hearing, but they don’t really largely have an understanding of exactly what they’re paying on an average basis or what that even means.
We saw the same thing with deductions. Certain deductions people hear about a lot. People are highly familiar with things like charitable contribution deduction, mortgage interest deduction — those are very salient. There’s a perception that those deductions apply to lots of people when in reality they don’t because 90 percent, or close to 90 percent, of people take the standard deduction. But because people have heard of these deductions, they might mistakenly think these are deductions they can claim, which could be problematic because that could drive their behavior.
Erin Scharff: The average versus marginal rate confusion also drives confusion in economic decisions. There was a study that we cited in the paper that suggested there’s some misperception that overtime is always taxed at a higher rate, and folks are worried that if they’re going to get a raise, their after-tax income is going to be smaller because they don’t realize that it’s only their marginal income that’s going to be subject, potentially, to a higher tax rate. So it is not just a policy confusion, it is also a confusion that may lead people to make suboptimal economic decisions.
Robert Goulder: We’ve established that some misinformation is sort of inadvertent, or, as you say, “math is hard,” but there can be overt campaigns to influence people as well. The whole narrative surrounding the estate tax comes to mind. What does that episode tell you about tax and fake news?
Erin Scharff: Michael Graetz and Ian Shapiro offer, I think probably many of your listeners already know, a really eloquent encapsulation of this saga. But the story I always take away from this is this quote from Frank Luntz that “a compelling story, even [if] factually inaccurate, can be more emotionally compelling than a dry recitation of the truth.” A concerted campaign was waged by a very small portion of wealthy families to take on the estate tax and reframe it for folks. They highlighted stories — many of them demonstrably untrue — about people losing their family farms and people losing their small businesses.
Look, the estate tax as a tax on the wealthy was, if not resoundingly popular, not something that the average American was particularly worried about. But suddenly the estate tax is now a tax on hardworking, thrifty business owners, and that changed its salience.
So not only did Republicans want to repeal it, by the end of it a lot of moderate Democrats signed up for repeal, despite the evidence that these stories just weren’t true. They were compelling.
Robert Goulder: Now, your article distinguishes between fake news and false news. Can you help us understand the difference?
Kathleen DeLaney Thomas: Fake news — and I will say we draw our definitions from what’s really a growing body of literature by scholars from all disciplines on fake news as it’s becoming increasingly studied — but fake news, as we use the term, refers to news that’s designed specifically to mislead. So with malicious intent, intentionally created content meant to mislead people.
False news we define as information that’s incorrect, but with less malicious origins — perhaps a mistake or confusion by the author. They come from two different places in terms of intent. In the paper we lump them together. Even though we do discuss them separately, for purposes of our analysis, we put them together under one big basket, which we ultimately refer to as fake news really for simplicity.
The reason we lump them together in our analysis is because we argue the effect on the reader is often the same. These stories often start one place and get passed on so many times, through so many sources, it’s hard to point to, well, the original author made a mistake. It might be the case that the original author of a post made a mistake, but then it was shared many times by people who knew better. Or created by someone with malicious intent, but shared by somebody who thought it was valid.
So rather than having to trace through those, if the end result is information that misleads the reader, we characterize that as something we’re interested in studying, although we do recognize acts of misinformation come about in these separate ways in terms of origin.
Robert Goulder: I think one of the really compelling things about your article is the data set that you compiled. It’s really what makes your article so unique and forceful. You have these real-world examples of fake taxes. Can you tell us about your data set, how you selected the items, how far back you went in time to retrieve them, what filters did you use? What were your methods? Elaborate on all that, if you could.
Kathleen DeLaney Thomas: I’ll say, Erin and I spent a very long time thinking through this question of how we were even going to come up with the data set. I think we spent a lot of time on the internet and various websites figuring out what to do. We also surveyed literature of scholars and other disciplines who had looked at fake news data sets. Nobody had done it in tax, but just to see again what other practices were.
The approach we ultimately settled on was to use stories that had been fact-checked by third-party fact-checkers. These are websites your listeners would’ve heard of, like PolitiFact or Snopes or FactCheck.org. So third parties that fact-check stories that are already out there.
We did this for a few reasons, one of which is if you go to those third-party websites, they archive the stories, whereas if you try to go directly to fake news websites yourself, they’re often defunct because they’ve now moved on to a new URL and they’re hard to run down if a third-party source hasn’t archived them. It also basically created a resource for us where these third parties have made the determination, fake or false or not, which we thought took our own bias out of the process.
It was a more effective way — this is sort of the most important part — it was the most effective way of finding stories — and this is what we were most interested in — that had essentially gone viral. We were looking for stories that had been shared thousands of times, viewed by thousands of people.
Erin and I ourselves could go onto Twitter, for example, and easily find a tweet that we think is fake news, but likely it hasn’t been shared very many times or viewed very many times. But for a story to end up on FactCheck.org or PolitiFact, it’s on that website because it’s essentially gone viral at that point and gotten enough attention, it’s ended up on one of these websites. This created a place for us where we could limit ourselves to the fake stories that had been shared and viewed many times.
We essentially compiled a data set where we looked at every story that mentioned tax on these third-party fact-checking websites. You can do it all in one place by going to Google Fact Check. That was where we started. We ended up with 56 unique stories. Some stories popped up multiple times on different websites. It’s the same meme, you see it a few different places. So we only counted that once.
We ended up with 56. They go back to 2016, and that’s just as far back as the fact-checking goes. It’s really about 2016 where these third-party fact-checkers start fact-checking these social media memes. So it starts in 2016, and we looked at every story we could find up until May of 2022, which is when we really sat down and started writing. And so for that we end up with 56 stories.
Erin, I want you to say more, if you want, about how we narrowed it down more and what we excluded.
Erin Scharff: Well, I think I’m happy to say that — I was going to say a couple of things about Google Fact Check and the organizations that do fact-checking, because this is something that I actually — I knew Snopes existed before we did our project, but I had not known that there was actually an international fact-checking network that fact-checking organizations need to submit an application to that requires them to adhere to a standard of conduct, including transparency. And then this international organization makes an assessment.
So the pool of fact-check websites included some websites that I was familiar with and some websites that I was not. But Google Fact Check relies on this international code of principles signature organization to sort of make sure that — look, it’s hard sometimes to know whether something’s true or not in a complicated area. In our paper we discuss some examples of close calls, but at least all of the organizations we’re looking for had some indicia that they were doing all of this in good faith. They are the organizations that both Google Fact Check and Facebook rely on in their own fact-checking. I wanted to add that because that’s something that I learned in the context of this paper.
In terms of where Kathleen led us down, which was how we narrowed our data set from all of the stories on tax to the 56 that were ultimately in our study, we excluded stories that were about foreign taxes. There’s a lot of stories about United Kingdom taxes that end up in the fake news data sets that are checked. We excluded those.
We excluded a small number of stories that were about state and local fake tax news. On the one hand, I care a lot about state and local taxes; I think that’s important. On the other hand, they’re less likely to be fact-checked. We worried that we weren’t getting quite a representative sample of the kind of stories that are out there. They’re just less likely to go viral in ways that capture national fact-checkers.
We also excluded fact-checks of politician statements. We did that for a number of reasons. One, sometimes they fact-check politician statements that we think were just misstatements. President Biden once misstated the marginal tax rate. We don’t think that that was likely intentional, in part because it is just so easily verifiable that we think it’s unlikely he was trying to convince Americans that the tax rate was a different tax rate.
But another reason that we think it was appropriate to exclude politician statements is sort of the ways that fact-check organizations choose which stories to fact-check.
And Kathleen already alluded to one criteria that was important to us, which is these organizations fact-check stories that are viral, that have gone and are being shared widely. For us, that’s what we were particularly interested in. What are the stories that people are sharing?
But fact-check organizations often have a separate criteria of checking statements of political officials or candidates for elected office simply because they’re prominent political officials or candidates for elected office. That seems like an important role for fact-check organizations to hold those seeking and in public office accountable to tell the truth, but we were less interested in that piece of fact-checking work because we really were interested in, what are the stories that are capturing attention on social media, and why were they capturing attention on social media? And politician statements are fact-checked often without reference to how viral those statements have gone.
Robert Goulder: Yeah, thinking about it makes sense to filter those out. OK, so you did all this hard work, you compiled this data set. Tell us, what are the takeaways, what are the identifiable trends that you were able to observe?
Kathleen DeLaney Thomas: This was the fun part, because I don’t think we went in with any priors about what the trends were going to be. We just sort of pulled all the stories and read them to see what we could find.
I’ll mention one of the surprising threads which we discuss in the paper, which is a focus on new taxes. I think we say in the paper something like it borders on hysteria in terms of how upset and fearful people are that the government’s going to enact a new tax. So not just raising taxes, but new taxes.
Over 20 percent of the stories we found, the fake stories, were about new taxes being enacted. Examples include a livestock tax, a federal property tax on your home, a tax on your 401k, a national sales tax. I think there were at least a dozen unique stories about new taxes. The common thread there was that Congress had slipped a new tax in a bill behind closed doors and that if taxpayers didn’t watch out, they were going to be paying this tax soon and they didn’t realize it.
It reflected a couple things we thought were really interesting, which is, one, that people tend to be or seem to be very averse to new taxes. We talk about some literature supporting that idea. And two, that people really don’t trust the government.
Maybe not surprising, but it seemed to be particularly reflected in these stories about, again, widely shared viral stories that Congress buried in some big bill of many pages a new tax that no one was being told about.
Erin Scharff: I would say, you said this was all hard work, but it was hard work that was very rewarding. I’ll echo, because it was just fascinating to see what people were talking about in terms of fake taxes on social media, including all of these fake stories. I think a couple of things stood out to me.
One, just the persistence of what we call in the paper “the old chestnuts of tax denial,” the widely shared, “I’m giving you a gift that’s not taxable to the waitress” kind of memes that federal income tax is unconstitutional.
And just with all of the transparency that’s enabled by the internet, all of the sort of free tax information that’s now available, that hasn’t stopped this echo chamber of non-truth about taxes, I think maybe not surprising, but I definitely found salient in our data.
The other thing that I think I found interesting was just I think what you see as sort of the emotional salience of the ways taxes affect Americans. Trump’s tax policies themselves are rarely seen as subjects of fake tax news on either side, liberal or what you could call conservative fake tax stories. The fake tax news about Trump was all about his own personal taxes. To me that suggests it is not the continued story, that what is interesting to people is fascinating stories and emotions.
There’s a bunch of important — and I don’t want to dismiss it — policy implications about the ways President Trump did or did not pay his taxes. But at the end of the day for a lot of Americans, those aren’t the important stories affecting their own tax liability, and yet major changes happening to the tax code and the Trump administration weren’t being debated in these kinds of forums as regularly.
Kathleen DeLaney Thomas: There was also a separation that was clear to us between some stories that appeared to be very clearly intentionally fake, as the way we discussed is defined earlier — completely fabricated. Usually these stories would have some tiny snippet of truth and then [be] taken in a fabricated direction. When we talk about these stories, they’re not articles, they’re almost all a meme, a graphic with a line of text, something that would be easily shared on Instagram, Twitter.
We call them stories, but we’re really just talking about a really easy to read one or two sentences. A lot of them clearly seem to be the type of story that had some malicious intention. But then we also separately discussed in the paper stories that had been fact-checked and rated false that we think almost certainly stemmed from confusion, if not by the original author, the fact that they’ve been so shared we think reflected widespread confusion about particularly complicated parts of the tax law, the forms going on.
An example we give in the paper are the recovery rebates: the CARES [Coronavirus Aid, Relief, and Economic Security] Act, that $1,200-per-person check that went out as a recovery rebate. We all remember these were structured as advance refundable tax credits. The idea was, of course, to use the tax system to deliver this money because it was a built-in mechanism. It was very efficient, and it was effective at getting money out really quickly. But the way the legislation was written to the average reader would be confusing. What does it mean to have an advance refundable tax credit?
We found fake news widely shared about the fact that these rebates were going to have to be repaid with your tax return. It’s actually a really compelling example. There’s a TikTok video that went viral — we discussed it in the paper — where there’s a man, and he just reads directly from the legislation. It’s pretty stunning. He doesn’t even have to come up with something fake to say. He just reads the legislation directly into the camera and then looks up and says, “See, what that means is you’re going to have to repay this check on your tax return next year.” It’s easy to hear him read this legislation and think, “Yeah, I guess that’s what that means,” because the legislation is very confusing.
We cite this as an example of here’s something that a lot of people probably legitimately misunderstood because it was confusing, and it’s confusing to people to read and hear that they’re getting an advance refundable tax credit and that it’s a credit against their tax and not know what that means in terms of are they going to have to pay this money back.
Erin Scharff: The guy concluded the video with these lines: “Guys, stay safe out there. Media and public is not telling you exactly what this bill is.” He’s citing the language of the bill.
And it is not that I think the best way to write tax legislation is always to make it easily digestible to the public. I understand why that would be a challenge on the one hand. On the other hand, the easy access to the code means sometimes people are going to look at it and puzzle and be legitimately confused and concerned.
Robert Goulder: Well, that’s where we transfer from a discussion about media and fake tax news and start thinking about notions of what’s good for American democracy. Is there actually a case for changing how we draft statutory language so that episode won’t happen?
I mean, it seems kind of nuts to say, “Oh, we’re going to change how we draft legislation because we don’t want there to be fake news.” But I admit I would like for the fake news to go away. Is that fair? How do we stop it?
Kathleen DeLaney Thomas: That is fair. I think I walked away from this project with the feeling that the fake news was never going away, and we need to think about how we communicate with the public and with taxpayers in this environment of fake news.
We have to think about the fact that — and this is really what makes — we argue now very different than, for example, the time in history chronicled by Graetz and Shapiro during the estate tax. That was a slow, years-long campaign of misinformation.
Now people are barraged with fake, false, bad misinformation all the time, especially if you are on social media a fair amount, and especially if you get your news on social media, which studies show a lot of people do. We’re now in an environment where people are constantly subject to misinformation.
After doing this study, I have thought a lot about, “How do we communicate true information in this environment under the assumption that there’s probably very little we can do to make the fake news go away?”
Erin Scharff: I think that’s right. One of the things we opened the paper with is fake news has been around for as long as there has been media being shared, hoaxes and deliberate misinformation. What’s changed is the ease of sharing it.
I think one of the takeaways I have from the paper is, I live in Arizona. When I think about sort of the most immediate threats of fake news, lots of my fellow citizens of this state do not believe that we’ve had legitimate elections or do not believe the official results of those elections. Those are really dangerous forms of fake news in a democracy that I think are really first-order problems of shared governance that we’re going to have to address.
In some ways our paper is about a second-order level of problem, but a very, very serious second-order level of problem, which is, we did not scrape 4chan or Truth Social because that’s not the social media websites that our fact-checkers are using. They’re using what we could think of now, maybe, as mainstream media — social media, Twitter and Instagram and Facebook. These are people participating in what presumably is a bipartisan social media ecosphere where there’s still a lot of sharing of misinformation.
I think that thinking about it as both a policy problem and a taxpayer compliance problem, and thinking about those separately, is going to be really important to thinking about solutions. Because some of this stuff is purely a policy debate, and some of it is really, do people feel comfortable taking a COVID relief payment? Do people understand what is now going to be reported from their Venmo and Zelle accounts?
And what is the responsibility of tax administration to use the channels that people now use to get their information — social media — to provide accurate, clear, and concise information? I think we need to think seriously about the fact that people watch videos more than they read news articles, and how do we reach those people with accurate tax information?
Kathleen DeLaney Thomas: One thing we discussed in the paper is that we think a baseline response — and it’s not going to solve the problem, I think it’s more of, at a minimum, something Erin just alluded to — which is thinking about channels of communication of information. Thinking about communicating with taxpayers — and we mean from the government’s perspective, from the IRS — through things like social media.
We use, again, the CARES Act recovery rebates as an example. Because again, as I mentioned, you had TikTok videos, other easily shareable and digested memes essentially saying you’re going to have to pay your COVID rebate back, circulating viral on social media. The corrected information, if you wanted to find it, as far as we could tell, was on an FAQ page on the IRS website.
Now, I will say we were looking for this information a year later, so it’s possible something was posted online that we could no longer track down, but as best as we could find, there was information on an FAQ, on a text-heavy website. You’d have to go looking for it, and you’d have to scroll down and find it. It was hard to find and not something easily shareable.
We suggest, just as an example in the paper, couldn’t you have a meme with a graphic that just says in big, bold font, “Your recovery rebate does not have to be repaid. Beware of false information.” That could be tweeted and posted and hopefully shared by perhaps academics or concerned citizens.
The IRS can’t make its own information go viral. We understand that. We understand there’s a limit to how much you can combat fake news with true news simply by using social media. But again, and as Erin alluded to, we suggested at least at a minimum, it’s worth thinking about how good information gets communicated to taxpayers.
We also argue it’s not very expensive to do these things. When you look overall at the cost of different enforcement measures and levels of communication, social media is something that’s relatively cheap.
Robert Goulder: Erin, final thoughts?
Erin Scharff: Well, it’s just been a pleasure to be here would be my first final thought. Thank you for having me.
Fake news is going to be something that we are going to be grappling with for the foreseeable future, and thinking about how we communicate good tax information as folks who listen to this podcast that are tax professionals I think is going to be a critical part of our sort of moral obligations going forward. It is not enough just to think about what good policy is, but to make sure we’re able to communicate accurate, truthful information in ways the public can understand and take that part of our job seriously.
Robert Goulder: There you have it. It’s a fantastic article, and I do encourage you to read it. “Fake News and the Tax Law.” The authors, Kathleen Delaney Thomas with the University of North Carolina School of Law in Chapel Hill, and Erin Scharff with Arizona State University School of Law in Phoenix. Thank you for the article and thank you for being on the program.
Kathleen DeLaney Thomas: Thank you so much for having us.
Erin Scharff: Thank you.