Fiscal Scoring with Congressional Budget Office Director Phillip Swagel

>> Jon Hartley: This is the Capitalism and Freedom the Twenty-First Century podcast. An official podcast of the Hoover Institution Economic Policy Working Group where we talk about economics, markets and public policy. I'm Jon Hartley, your host. Today my guest is Phil Swagel, who is the Director of the Congressional Budget Office.

He previously was the Assistant Secretary for of the treasury for Economic Policy during the George W. Bush administration from 2006 to 2009. And he recently was a professor in International economics at the University of Maryland School of Public Policy. He was also a nonresident scholar at the American Enterprise Institute, a senior fellow at the Milken Institute, and co-chair of the Bipartisan Policy Center's Financial Regulatory Reform Initiative.

Welcome, Phil.
>> Phillip Swagel: Yeah, thanks so much. Thanks for having me.
>> Jon Hartley: Well, it's a real honor to have you on and to talk about CBO and how it works and its role. It's really important role. I want to first start with talking about where you grew up and how you first got interested in economics.

I mean, you did your undergrad at Princeton and PhD at Harvard studying economics. I mean, where did that interest begin for you?
>> Phillip Swagel: Okay, well, so I grew up in Southern California. My, my parents were from New York, both of them originally from the city. They met at, at Columbia, Columbia and Barnard.

But most of my life I grew up in California and I was kind of always interested in the news and current events. So I was the kid at summer camp in Ojai, California. I would fish the newspaper out of the trash bin by the counselors' dorms or bunks or whatever and read, it was the LA Times back then.

So I kind of always had that. So it's kind of natural that I'd be interested in something with, you know, with not politics, but current events, the real world. And then in college, I went to Princeton, as you said. You know, I just like that connection, the connection between kind of something analytic and in the real world.

>> Jon Hartley: That's fascinating, and I think, you know, many economists sort of, I guess first get interested in reading the newspaper and following what's going on. I mean, how did that interest sort of get fostered further when you were an economics major at Princeton, studying economics at Harvard? Who your advisors and sort of biggest mentors in that whole process?

>> Phillip Swagel: Yeah, for me, it was really important and I just lucked out, I had fantastic advisors and professors. I took microeconomics from Harvey Rosen, who, you know, then I worked with when I was at the White House and he was a member and then chair of the cea.

I took macro with the late Steve Goldfeld. Really phenomenal. Had been a CEA with President Carter. My senior thesis advisor was really a brilliant man who's now retired from Princeton, Avinash Dixit. And my PhD advisor was one of his students, Kalakrishna was my first advisor and then Alberto Alesina was my second.

And then Larry Katz was also on my committee, who also is just brilliant and amazing. I had actually, I was just thinking, as you asked, my first year of the Ph.D. program. So this would have been the fall of 1987. The macro class was taught by a mix of Larry Summers and Robert Barrow and they kind of alternated and which is great.

And Professor Barrow, Robert Barrow's brand new at Harvard, and he was just fantastic and brilliant and insightful, but also so was Larry Summers. And the day I remember I was going to mention was the day of the market crash in October of 1987. Larry was teaching that day, you know, again, first year PhD, macro whatever he was going to teach.

He said, well, today I'm just going to talk about bubbles instead. And just, you know, kind of off the top of his head did a whole class on financial markets and bubbles. And it's great. So yes, academically, yeah, I really, really lucked out. And I could talk more about after school some of the people I worked with also.

And of course some of your friends and co authors as well, like Kevin Hassett.
>> Jon Hartley: Well, that, that's terrific. And sorry, who ended up being the chair of your committee?
>> Phillip Swagel: So it was Kala Krishna, who's a trade economist and just brilliant, warm, funny, insightful, and she's now a professor at Penn State.

>> Jon Hartley: Fantastic, I mean, what an amazing group of people to have as mentors. And I can only imagine what it would be like to be in a class with Larry Summers as the October 1987, I always forget, the Black Monday or Black Tuesday crisis.
>> Phillip Swagel: Yeah, yeah, exactly, exactly.

I think it was, boy, I think it's Monday. But yeah, it's a good question, I'd have to go back and look.
>> Jon Hartley: I always get confused. But I know it's a 25% draw down in one day and it was one of the largest, I think still one-day draw down.

I know it came back later in the day, but one of the strangest market anomalies that the economists still haven't been able to fully explain yet. I guess I'm curious. You became a professor at University of Maryland. You served a Treasury during George W Bush administration, and now you're director of CBO.

I take it, economic policy has been something that you really enjoyed. What was your experience like during the Bush administration? You were there during, you know, the global financial crisis later in the George W. Bush administration. What was it like sort of working on things like TARP and so forth during that time?

Which feels like it was a while ago and we've had maybe a couple crises since, but I think it was definitely a turning point for economic policy. And you were there during that time. What was it like and what were you working on?
>> Phillip Swagel: Yeah, it's interesting. I mean, just like you said, sometimes it feels like a million years ago and sometimes it feels like yesterday and something will come up that reminds me and takes me back or I'll meet someone who I haven't talked to since then.

I mean, you know, it's a terrible time for the economy and the financial system, so there's a sense of urgency and there's a sense of mission as well, of, you know, we, what we were doing was important and, you know, obviously in. Well, I say in retrospect, rescuing the financial system and, you know, bailing out, I think, is the, the word people use.

And that seems like a perfectly fine word. Bailing out banks is controversial and unpleasant, but it was necessary. And that was something that I think, you know, even in real time we understood, you know, necessary, momentous, unpleasant, all those things. But, but overall, it's just the sense of mission, of working together and, you know, that the economy, you know, had to continue and the financial system had to be rescued as part of that.

>> Jon Hartley: That's terrific. And I'm sure, also, I think there's very few researchers who are able to stretch out both sort of the research world and the economic policy world. I mean, how, you know, after going back to academia, after having served at treasury during the Georgia W. Bush years, how did that inform your research tastes and things you're interested in writing about, doing research on?

>> Phillip Swagel: Yeah, I have to admit my attention span is shorter. You know, going back, going to academic conferences. I think this is pretty common for even the most accomplished people with much better CVs than mine. After policy world, it's hard to go back to academic conferences. I've kind of always been interested in things that were real world, and so that wasn't so hard, Thinking about financial regulation, that was pretty easy.

I had students who were interested in real-world things since especially as the policy school. So for me, Maryland's a very congenial place, and the policy school is the right place for me. So moving back and forth between academia and policy sort of policy world, it feels very natural.

And even at Maryland, of course, is inside the Beltway, in suburban Washington, I was able to be pretty involved in some policy things as well.
>> Jon Hartley: Well, it's terrific. And I know many great economists at the University of Maryland who've gotten involved in policy and are really distinguished.

I think Mike Faulkner right now, who's the Deputy Secretary of Treasury, want to talk about, CBO and really drill down into your current day job. I want to first get into why CBO matters as well as its history. So, the Congressional Budget Office, or CBO was created in 1974 by the budget Impalement Control Act, something we've been talking about a lot recently for other reasons, and things like surrounding Doge and so forth.

But, there's what is CBO in my mind. I see it as the official scorekeepers passing legislation and when there's some sort of a cost requirement that features, sort of features primarily in the Bird Rule and reconciliation process. Something that the Republicans and US Congress are going through a lot right now with passing their tax reform bill.

Can we maybe just take a minute to summarize how the budget process works? Sort starting from, sort of start to finish, starting from the budget resolution to the passage of the law and sort of where CBO fits into that when it comes up with the scores and where does it really matter?

>> Phillip Swagel: Okay, yeah, sure no, it's. No, thank you. I mean, CBO is an agency of Congress, and we're here to support Congress. So we get them the information they need when they need it. And that's both the kind of what they need and the time, when, and that's really our key role.

And the budget is obviously the budget process is a key part of that. Let's think about the reconciliation that they're working on now, right? Last week, Congress passed the budget resolution. That's a concurrent resolution between the two chambers, the House and the Senate. So there's no cost estimate to it.

That budget resolution sets out the broad parameters in a sense of the changes in the deficit that the overall, and then how each committee between the two chambers is going to contribute to that. And there's differences between the House and the Senate and there are different bodies. And so we are, we help them on a technical basis, right?

Coming up with the numbers involved in the resolution. And then we're helping the individual, individual committees that are working on the specific pieces of the legislation that eventually they will vote on. And so that, for example, health care, I mean, Medicaid is being considered. Obviously this is public record.

And so we work with committees of jurisdiction that work on Medicaid. In the House, it's E and C, Energy and Commerce is Chairman Guthrie. We work with his staff and we go back and forth as they refine ideas. We don't say do this, this is the right policy.

We, it's their policy. And we help them analyze what the budget implications are and what the other implications are. Eventually, the committees will come forward with legislation, and they'll have markups where the legislation is considered in public in the committee. And then we will come forward with a cost estimate.

And it's usually after the committee, after the committee process, we'll have cost estimates for each committee, so for each piece of what could be quite a large piece of legislation. And then at the end, the Budget Committee, first in the House, or probably first in the House, maybe then in the Senate.

It's up to them really, what order. The budget committees will stitch together the legislation coming out of the individual committees, and then we'll do another cost estimate for the entire thing. And we'll do that jointly with the Joint Committee on Taxation. And for anyone watching on video, well, you can see I'm pointing to the ceiling and I'm speaking from the fourth floor of the Ford House Office Building.

The JCT, the Joint Committee on Taxation, is on the fifth floor. And so all the parts of the legislation that involve changes to taxes to the Internal Revenue Code, they will estimate, but we'll work together with them. They're in charge of the tax, we're in charge of the non tax.

We'll do a grand estimate of it all. And then we'll do a dynamic estimate of the effect of the legislation on the economy. And then the feedback back to revenues and the deficit.
>> Jon Hartley: But you generally have your own models for analyzing, say, the effects of taxes as well.

And these are different from the JCT ones, I guess, so there's this whole budget process and I think a lot of people, it's obscure to a lot of people. In this idea that you what the House Budget and Senate Budget committees do, they come up with basically the top line numbers.

And then it comes down. Once those budget resolutions are passed, you think about passing a budget every year, then it comes into sort of lower committees or sort of downstream committees. The Appropriations committees or say House Ways and Means, Senate Finance Committee, and so forth. The appropriations committees and the House and Senate, and so forth.

They're the ones that are doing a lot of this work in terms of working within that top line number that's been outlined. And then it's usually, I think after they've written a bill or come up with a markup that you'll go out and score it. Typically, I guess in between that and passing the bill and you do lots of things.

There's a lot of things that go into these scores. When CBO says it'll cost, X trillion or billions of dollars for this particular piece of legislation. I want to sort of break down into, I guess a number of key sort of issues that go into each of these estimates.

One of them, which has become very topical right now is this whole idea of current policy versus current loss. CBO has done a lot of, traditionally done a lot of scoring under current law. I mean, some congressional Republicans right now would sort of like to switch to current policy to score.

So that the TICTO renewal bill and it's kind of a philosophical divide, I guess in terms of current policy says that we should have a baseline that's wherever the policy is today and use that as the baseline. And so, things like expirations and the law and so forth wouldn't be thought of as a change per se or part of the baseline.

For example, I think extending the taxes from Jobs Act looks inexpensive if we're, what we're doing is just keeping sort of the current provisions. If the current policy is the baseline, and the current law baseline is different in that it says sort of that the cost of expiring provisions does matter.

So it's really in my mind a philosophical divine. But I'm curious, how do you think about it and what should CBO and Senate Parliamentarian be doing? Is really your job just to sort of be agnostic to these sorts of issues? And just sort of come up with scores according to whatever sort of Congress is asking you to do, or where does sort of CBO fall into that in terms of, you know, current policy versus current law?

And how would you explain that issue? What do you think about it?
>> Phillip Swagel: Okay, yeah. It's an important question and something that the Congress is thinking about now. And the way you put it as a philosophical issue, I think is exactly right, that different people will look at it in different ways or look at it in both ways.

To say, like with the tax provisions, these are expiring. But a member of Congress might say, well, I think they should continue. And I just want to continue what we have now. And I feel like looking at it that way is the right way to look at it.

And another member might say, no, if the current law is this and the law has those provisions expiring, and that's the right way to look at it, we will provide the information that Congress wants. And that means we will provide comprehensive information this year. It means we'll provide both that if a member wants current policy basis, I mean, they'll have to tell us what is current policy, which provisions they think are current policy and are meant to be extended.

And, of course, I'll have to tell my colleagues upstairs at JCT what's current policy, since so much of that is the tax code, and we'll do that. And I mean, both agencies, CBO and GCP are here to support the Congress. And from the CBO perspective, we will give them the information they, you know, they want in the way they want it.

It'll be our analysis. Of course, we won't necessarily give them the numbers they want is whatever the numbers are, that's what we'll give them. On the other hand, we'll also provide the current law numbers as well. And that's, you know, first is by statute that the statutory baseline is on a current law basis with some exceptions for programs like Social Security and Medicare.

You know, it's our job to provide Congress with the information they want when they need it, is our analysis. But they're the ones who decide what information is useful to them. And different members with different philosophies, you know, again, as you put it, will vote differently based on different information.

And I just want to make sure that the information we provide is comprehensive. And that's really fulfilling the CBO mission.
>> Jon Hartley: Terrific, I wanna get into dynamic scoring because this is another one of these sort of, I think, hot topics. So dynamic scoring, CBO and JCT, they have sort of different types of models, and they have both, what's called a conventional static model, which is where a lot of these sorts of numbers often come from.

And then, they also have these general equilibrium models that do dynamic scoring that account for how GDP and the macroeconomy might be impacted by such policies. And they're accounting for things like spillovers and things like that and are a bit more complicated in that respect. And I think one area, I think of confusion, I think for a lot of people, is that a lot of advocates of dynamic scoring.

Often, I think sort of forget that in conventional estimates still include elasticity, taxable income, which includes sort of behavioral responses to taxation. I'm curious, like, what sorts of behavioral responses would the conventional static model include in CBO's case? I mean, does it include your revenue shortfalls from, say, tax avoidance minimization due to, and things like due to wanting to work less?

How does CBO sort of separate out these sort of conventional behavioral response responses, say, evasion avoidance, tax minimization from the labor response in dynamic scoring? And I mean, do you think there's a lot of value added and say, I guess we'll get to dynamic scoring in a second?

>> Phillip Swagel: Okay. Okay. No, no, absolutely. And so I start with where you started, that the conventional estimates have a lot of behavior in them. And you mentioned some of them, some of the dimensions of behavior. And we have that the JCT in their tax models, they have that.

One thing we analyze where this behavior is important is tariff policy. And there's that the president is limiting the use of the de minimis exception as a couple of large firms in China that ship, basically small shipments to the US and those essentially get around the tariffs. I don't know if it's evasion avoidance, right?

One's legal, one's illegal. I think it's what they were doing was legal and now it's going to be off, you know, out of bounds, and that's going to change the nature of commerce. And we have to figure out, well, what will it mean for the supply side, you know, for those firms, and then what will that mean for the customers?

And so there's a lot of behavior in there. It's probably not going to shift the overall size of the economy. A lot of dollars, probably not macro scale. But how would that adjust? And so we've got to think about that. There's other things like no tax on tips.

I mean, that would be jct, but they would have to think about, well, how will the nature of compensation change if, you know, could more activity shift over into tipped? You know, the kinds of things that have tips. And so that kind of behavior is important. I only just mentioned one more, which is, I mentioned before Medicaid policy.

We're doing a lot of work on that. A lot of the impact in the sort of the dollars, especially on the state side, and the coverage potential coverage changes comes about through decisions made by states. So the federal government will change the number of dollars, the allocation of dollars between the federal government and states.

And then, states have to decide how to respond. And different states will respond differently. They have different views toward Medicaid and uninsured and things like that. And we have to take into account of that behavior. So it's not of individuals, but it's of states. And we talk to the states, we analyze their situations.

But ultimately there's going to be our assumptions, our analysis built into those projections.
>> Jon Hartley: Fascinating. So I guess, like, okay, so I know there are tax models too. So like the things like evasion, avoidance and minimization. Break that out in a cbo, in the CBO tax model, or.

>> Phillip Swagel: It's hard. So for a cost estimate, JCT would do it. And they would have to think about that. About, you know, I guess, like the tax on tips, they would have to think about, right. How will the nature of work reconfigure? And what does it mean? So what does the behavior mean?

And we face the same thing with a tariff change. You know, how does the nature of commerce reconfigure itself? After legislation is enacted, then CBO is in charge of the baseline. And so, for example, the 2017 TCJA, the December 2017 Tax Act, JCT was responsible for the estimates.

But then once it was enacted, we put it into the baseline. And so in April 2018, CBO published kind of a long analysis of how we fit the newly enacted tax bill into the baseline and what did it mean for revenue and GDP and so on.
>> Jon Hartley: I'm familiar with some of that work because I know there's.

I have some kind of related work with my colleagues Josh Rao and Kevin Hassett on facts on investment. And I know there's sort of an elasticity that. Use from the literature plug into, I think your cap tax model, which is more about investment than say you're just thinking about the revenues in your baseline.

I guess sort of just and we'll get to that in a second. We'll talk a little bit more about models and transparency and things like that. I totally commend the CBO and being more transparent than the JCT and that sometimes, we don't even know what the JCT's elasticities of taxable income really are.

We think they're sort of less than one, but it's hard because at some level it's not quite an apples to apples comparison in terms of say, how a literature estimate compares in that it's not clear what sorts of things get counted. Hence through the question on evasion, avoidance, taxation.

That's more a JCT question. But I guess I'm just curious overall conventional scoring or conventional static models versus dynamic scoring. In thinking about how these policies, how they impact GDP, what do you think? Do you think there's value added there in dynamic scoring? I mean, I feel like a contentious issue for a very long time in terms of to what degree we should think about the estimates that come out of really general equilibrium models, which is where these sorts of dynamic scoring effects are being measured often.

I'm curious, what's your take on sort of dynamic scoring versus conventional stack models in general?
>> Phillip Swagel: Yeah, we want to do more of it and we've put a lot of work in over the last really couple of years. I'm trying to think of since the 2022 reconciliation legislation passed, right?

The reconciliation bills tend to be kind of all consuming. Then the Congress was divided in 23 and 24. So the PAISA legislation slowed down some and we had some time to prepare. And again, working with JCT and being ready for reconciliation, there are different roles. There's some things where legislation intrinsically has to have a dynamic score.

A big bill that affects immigration, you're changing the size of the labor force if the bill is large enough and intrinsically you're changing the size of the economy so that by its nature is dynamic. Something in tax policy where you have a really nice paper with Kevin Hasselt and Josh Rao that looks at the impact of change in tax policy on the economy or on investment, as you mentioned.

And in the economy, a big change in tax policy or even a small change that's focused on pro growth elements, that would change the size of the economy and then a dynamic estimate makes sense. And then there are members and many members who want us to do dynamic analysis on the spending side, sometimes they call at the investment side is a change in some policy that changes the size of the labor force, or maybe it's direct federal investment, building roads and bridges or highways or something else, that changes the productivity in the economy.

It changes the size of the capital stock. And members want that. And I want to give members of Congress the information that they think is valuable. And look, I think it's valuable too. I mean, the more the better, I can explain some of the challenges, but that's my overall philosophy is comprehensive, more information to the extent we can.

And I can go into some of the challenges.
>> Jon Hartley: I'm sure you have to have some sort of an estimate of what a fiscal, your fiscal multiplier would be, I mean, there's all sorts of varying estimates in terms of fiscal multipliers. I mean, there's certain micro estimates that I think are tend to be pretty high and then from maybe things like RCTs and things of that nature, quasi experiments of that kind of.

And that's from sort of more micro day. And then you have sort of the big macro sort of estimated elasticity. I think often of say, Valerie Ramey's work and the work of others, you know, John Sison, Emmy Nakamura and others, that you look at these like defense spending shocks.

And so there's, I think, a school of thought that says that those multiple multipliers are pretty low and maybe as low as, you know, 0.3 or something like that versus the micro multipliers might be, say, around one and a half or two or something like that. And then some people argue there's, I guess, tax fiscal multipliers that are very different from government spending ones.

I think Alberto Alsina found that the tax multipliers were often bigger than the government spending multipliers. I'm curious. I assume those would be something that you'd have to think about if a member comes to you and says, hey, incorporate the dynamic effects of building a bridge or some infrastructure project.

>> Phillip Swagel: No, that, that's right. That's exactly right. And we did one paper like that. It was requested by the Senator Rob Portman, who was leading group with Senator Kirsten Cinnamon and others that resulted in legislation on infrastructure investment. And that's what they wanted to know, that what's the dynamic feedback from infrastructure investment changes productivity, changes the size of the economy over a long period will feed back into Revenues and they had in mind really roads and bridges, which, you know, takes, well, it takes in the US years to build a road or a bridge, you know, sort of infamously sometimes.

And then we looked at the 30 year horizon of the, you know, productivity and revenue feedback. It did not pay for itself. But depending on how the road or bridge was financed, there could be a substantial offset that the added productivity would generate enough, enough revenue to pay for it was like as much as half of the cost of the road and bridge.

Leaving aside the crowding out, so leaving aside the debt that might be incurred to finance the road and bridge in the, in the first place. So there's still a sizable, a sizable impact that's on the spending side. And then, you know, like your work with Kevin and Josh on the tax side and the way I think of it and the way I explain it to members that different provisions have different pro growth impacts and the, you know, the sort of expensing, you know, think about like expensing versus a change in the corporate rate.

And the corporate rate is permanent, so maybe that's not on the table now. But think of like expensing versus the change in the personal rates. Those will have different impacts on growth. They're both pro growth, but different quantitative impact impacts.
>> Jon Hartley: Yeah, no, absolutely. And it's a huge, I guess, debate right now in Congress in terms of what to be focused more on sort of pro growth firm type tax adjustments versus, I guess, things are more, I guess on the individual side of things, things that were promised in the presidential campaign like tax on tips and things like that.

And doing questions I think right now, but maybe raising the person, right? It's a very interesting dynamic in terms of how these things are being considered. The salt deductions are a very big topic as well, but I wanna sort of get into this came back to sort of just CBO.

I wanna talk about just CBO scoring accuracy. So, you know, often our CBO scores wrong. I mean, I guess no score is ever perfect. But I'm curious, you know, we could think about, you know, the Taxes and jobs act of 2017 in the first Trump administration tax cuts which reduced the corporate rate and brought in expensing provisions for equipment.

And it also changed a lot in terms of how businesses pass through businesses are being taxed. And there were also changes to the individual code. There were opportunity zones. They capped the SALT deduction, all sorts of things in terms of your score or the CBO score for that.

And mind you, as a different CBO director who was there at the time, what did CBO get right and wrong there in your mind?
>> Phillip Swagel: Okay. And just to say, well, I started as CBO director in June 2019, but I own every mistake CBO has ever made and certainly members of Congress remind me of that.

And that's just, that's the reality. And you know, fortunately we get a lot right, but of course we get things wrong. And just as a general matter, the more familiar provision is and the more incremental legislation is, the easier it is for us to model it. And when something's entirely new or big, well, that's harder.

And you know, the ACA of course, was big and new. So just things like that are harder. The childcare provisions and the build back better legislation, those were big and new. You know, we never, in the end, that was never enacted, but that's the kind of thing that's harder.

The 2017 act is an interesting one. You know, of course, JCT did the original score, but we, what people really think about now, it seems like, is our estimates in the baseline is our projections, the technical word. So in April 2018, we published the first baseline update after enactment and that included the impacts of the, of the legislation.

So both we had stronger GDP growth, stronger business investment, stronger GDP growth, lower revenues, right? That stronger growth would improve revenues, but not so much to pay for itself and so on. And the beginning of 2018 bore that out, right? Investment was stronger until about the middle of 2018 when we saw some of the first wave of tariffs had a hit.

CBO is really actually very positive on the investment impacts of the provisions in tcg, tcja. The, you know, the lower corporate rate, the international provisions, the lower personal rates, our internal estimates, like the elasticity you mentioned, the change, essentially the change in the price of investment, you know, from the tax side, what does that do to the quantity of investment?

So we're not quite as high as what you and Josh and Kevin found, but we're within the range of your uncertainty and we're a bit above some of the other academic literature. So it's a bit of an irony that we are actually, we're very positive on the economics of the Act.

The revenue estimate that we did again back in this was published in 2018, April 2018. We are basically spot on for 2018 and 2019. That actual revenues came in actually very slightly below what we projected. So in some senses we were over optimistic, but really a tiny amount.

Now, of course, 2020 is a difficult year with the pandemic. There's so much going on, legislation. So let's leave that out because I don't think anyone could have projected what happened in 2020. 2021. After 21, 22, 23, revenues were quite a bit higher than CBO projected. And so that's a mistake and error.

But of course the next thing is to say, well, why what happened and why were revenues above the projection starting in 2021? And I would point to a couple of things, right? Inflation is one is in our estimation, is the biggest piece that inflation. There's high inflation starting in the spring of 2021.

We tax nominal income. And so that means that the nominal dollars of revenue were higher than before, you know, before the high inflation. And I don't think the enactment of TCGA is what caused the high inflation, right? So I wouldn't attribute the high inflation to tcga. And so it's not fair to say, you should have, you know, should have anticipated that.

The Fed, of course, did put in extraordinary monetary policy in response to the pandemic that increased asset values. And then we saw that capital gains realizations look to have been extraordinarily strong in 2022. That of course, TCGA would have also improved asset values. But the sudden spike in 2021 and revenue surge 22 looks like more the result of the pandemic and the Fed's response than tax policy.

And then the last thing I'd say is that of course, Starting in early 2021, there is a surge of immigration and that for the federal government, well, there's cost of immigration, of course, but those costs are much heavier on the state and local governments in the US and for the federal government, the revenue impact is larger.

So again, revenues were higher than CBO anticipated because of the immigration surge. And I just don't think CBO could have plausibly expected the immigration surge in 21 back in 2018. And I don't think TCGA is what caused the immigration surge either. So that, I mean, I'm going on long here, but you get the point that revenues were above what CBO projected, but only starting from the time of the high inflation, the Fed's qe, immigration and other things.

So once you take that into account, our accuracy looks, look, looks a whole lot better.
>> Jon Hartley: Got it. So it's the cap gains I Guess. And cap gains tax rates are not indexed to inflation.
>> Phillip Swagel: That's right.
>> Jon Hartley: The bracket creep there. And you're saying even, you know, undocumented immigrants are some degree paying taxes as well.

>> Phillip Swagel: That's right. That's right. And they're increasing the size of the economy. Well, many of them do, do, do pay taxes. Many of them in the immigration surge were brought in under parole, which essentially within six months. Many of them had work authorization, and then they had a big incentive to pay on the books because they'd have a date with a, you know, a judge and, you know, down the road.

So they had an incentive to pay on the books, to work on the books and pay taxes so that they can show that to the judge. Yeah. So it's the inflation, equity prices driven in part by QE and then the immigration surge. All of those meant revenues in 2021 and on were higher than could plausibly have been predicted back in 2018.

>> Jon Hartley: Got it. So I know many of the Texas and JOBS act provisions are set to expire. In fact, they were made temporary to begin with because of some of these things like the Berg Rule and things like that. But now there's a desire to make them permanent.

That's sort of the big feature of the current tax bill that's being pushed through Congress. How do you go about modeling the impact of ticta or taxation Jobs act expiration in your macro baseline? I know sometimes you ask me why there's no recession being shown under the 2026 year under current law given to these expire.

I'm curious how you think about those expirations from a modeling standpoint.
>> Phillip Swagel: Yeah, no, no, thanks. And, you know, first of all, to say, you know, there's no normative judgment here that when CBO says this, you know, this is what we think will happen, or this is the revenue or this is the GDP that doesn't say therefore it's the right thing to do or the wrong thing to do.

And members of Congress have many reasons and it's political philosophy and lots of things that are just beyond CBO and you know that. So that's where I start. We do have the expiration of the provisions in the 2017 act affecting the economy. And so we have GDP growth slowing in our baseline forecast.

We have GDP growth slowing from 2.1 in 2025. So 2%, 2.1% real GDP growth down to 1.8 in 2026, which you know, is a meaningful slowdown. But as you said, it's not a recession and it's kind of at, it is at odds with, I know with some people have in mind, this is quite a large increase in revenue under current law.

It's a bit less than 1% of GDP under current law. And you would think, well, that's going to slow down the economy by more than we have in our baseline. And there's a couple things going on that we have to, we basically have to model. Well, first of all, so much of the 2017 Act is permanent, including the most pro growth parts, right?

The corporate rate, the international provisions provide incentives for activity to be domiciled in the US and not to invert and not to go overseas. So that's, that's part of it is that a lot of the pro growth part is not expiring. The part that is expiring the most pro growth part of that is relatively small in terms of dollars.

This is some of the business side provisions essentially expensing and is the depreciation provisions. The bigger dollars are on the personal side. Now, every household faces, virtually every household faces a change in their taxes with the expiration of those provisions. But it's still also true that most of the dollars are associated with households with relatively high incomes and relatively high wealth.

And those households tend to have low marginal propensities to consume. So even though it's a lot of dollars, the MPC that comes with those dollars is relatively low, so the effect on the economy is attenuated by that. And so the effect on consumption and GDP is attenuated. And then there's the third and last part.

I guess there's two parts, 3 and 4. Well, one is we assume the Fed would respond in part to slowing economy under current law. The third that's interesting is that we talk to businesses about their view of the Tax act and there's a, a substantial number of businesses said to us, well, we kind of understood that even though the provisions, some of the provisions were permanent, well, of course there could be subsequent legislation.

And so there's a sense in which in modeling their investment decisions, businesses might not have included the full permanent effect of the provisions just because they know laws change and even a permanent law can change. And again, that would tend to attenuate the impact. It meant that the positive impact of the law was not full force.

And then the expirations in the same way, which actually last thought, which makes the econometrics hard. So what you and Josh and Kevin did and what other groups have done is hard because you're measuring the provisions of the act. But then businesses are acting according to their own interpretation of the provisions rather than the actual provisions.

So we face the same difficulty, but it's a combination of those things that the expiration has a negative effect on the economy in our estimation. But it's not, you know, not sort of full bore and not enough to bring the economy to, you know, certainly to a recession.

>> Jon Hartley: Well, that's fascinating how you think about scoring things like that, there're some of those provisions that are set to expire, I wanna just talk about, I guess taking a step back here. CBO has hundreds of models as far as I'm aware. Now some of them are being made public, including your tenure, I think like the cap tax model, for example, which measures the impacts of a tax change on capital and fixed investment spending.

I'm curious one, how do we get more of these public and what do you think about things like I guess the CBO show your work Act. This legislation has been sponsored by Senator Mike Lee, I'm curious. CBO is more transparent than JCT. I don't think any JCT models are really public at all, So I commend CBO for doing this.

But I'm curious, you know, there's, I think been a lot of interest in, in scoring and what goes into scoring in recent years. I think you see some of this with, you know, the popularity of, I guess, you know that you got the Yale budget lab now and the Penn Wharton budget initiative and all these things.

So, so I think there is demand for, for, for scores. But I guess, you know, there's always, I, I think what, what's funny about this sort of business is like there's always people who are upset about scores because too big or too small depending on I guess what, maybe what they hope to see.

But I'm curious, I think at some level, you know, putting more transparency into what's going into these models maybe could help alleviate some of that and thinking about how you change what exactly what assumptions are driving certain costs and so forth. I'm curious what is sort of the CBO plan for making models public and what do you think about this?

At some level I could see sort of negatives as well in the sense of why did you put that in there? And so forth, you know, so, so I, I don't think it's necessarily all peaches and cream, but I'm curious, what do you think about transparency? And how important is that to you as a CBO director?

>> Phillip Swagel: Okay, no, no, and let me start by saying I agree with your assertion that the more transparent we are, it will help us when there are disagreements, when someone's unhappy with an estimate or a number. And we can say, look, here's the assumption in there. And if you, whoever's unhappy, if you think we're wrong with that assumption, well, let's talk about that.

And then it's a substantive discussion, it's a focused discussion, and we do that. And that kind of transparency is helpful and sometimes we'll just have a different view. I mentioned earlier, build back better the childcare provisions, we actually thought that those provisions would have higher uptake than the, the Biden White House did.

And so our score on that little part of it was bigger than theirs, but that was because we actually thought the policy would be more effective, in a sense, than they did. So these, my predecessor did a lot on transparency, as Keith hall, and I'm continuing that. My colleagues are continuing that.

The Captax model you mentioned is a great one, it has a limited user base. I know you are busy, like, very highly knowledgeable people for whom it's really important. You people at Tax Foundation, Brookings, AI, a couple others and it's worth it, right? It's the transparency, it's puts a little bit of a tax on us, but it's worth it because the community who uses it will give us feedback and questions.

And sometimes if there's something they don't understand or you don't understand, and you ask us in our tax analysis, staff are amazing. You know, say, yeah, you're right, we could do that better. And that's, that's kind of feedback that's helpful. We get helpful, we get feedback like that from congressional staff.

I'm sorry, I'm pointing out my window toward the Capitol, so I'm like, at the Southwest corner of the Capitol complex, and so I'm pointing the northeast of the dome. You know, the health staff, where we work very closely with both, you know, both chambers, both sides on health legislation.

Sometimes, we'll give them a preliminary estimate and let's say, we expected, whatever, this to be different or that. And we'll talk about, we'll realize, well, no, they you have something, or they'll say, you need to talk to this outside expert and we will. So that's the kind of transparency that's useful.

I mean, it's all useful, but then. That's useful in a very mechanical, very immediate way. We put up other, as you said, we put up other modeling information, like from our healthcare health insurance coverage model. We recently, recently put up some information about health insurance premiums and our projections of private sector premiums.

I want to do more. The challenge is that it takes time, especially when legislation is moving quickly, we often just don't have time. And we all have time to talk about it with someone, but not to actually go and put it on our web, on our website, or in a way that's usable.

And we just can't stop and put up the models. And sometimes that's a time, sometimes the legislation is confidential, so we can't do it in advance. Sometimes it's judgment and I mentioned earlier, we talked about Medicaid and the state behavior. The state responses to federal changes in Medicaid is critical to understanding the effect of the policy.

And that's our judgment. There's no model in the mathematical sense, but there's our discussions, there's our judgment, there's our analysis. And so it's not easily encapsulated in something we post on the web. But I can sit with people, our experts can sit with people and talk about it.

And that's what I want us to do more of. And so the, Senator Lee's legislation, it's completely understandable and the kinds of things that he wants in that legislation, I agree with that. We need to put up more that be, make it so that other people can understand what we're doing.

It's hard for us to do it in a way that people can replicate, which is ultimately what, you know, what he would like. That's just hard. I'm not sure we're going to get there in a broad way, but I want to do as much as I can, you know, to satisfy the motivation behind that.

And so I'm. Yeah, that's it. It's. In a perfect world, we would be there. He wouldn't even need the legislation. We'd be there without it. But given the constraints, you know, we have, I want to do as much as we can. Actually, can I just say a word that anyone listening to the podcast, we have a Contact Us email on the CBO website, is communication.cbo.gov and those inbound emails matter.

Or if a listener or a viewer, you see a CBO report and something looks wrong, well, don't be shy. We do Want that feedback? We get plenty of feedback from academics, from members of Congress, from staff, but from anyone. If something looks wrong, I want to know about it.

And especially constructive criticism is always, is always helpful.
>> Jon Hartley: Well, that sound very kind you to be receptive to feedback in that way. I know many people have some very strong thoughts and feelings about various estimates. Actually, I asked ahead of doing this interview, I asked a few of my Hoover Policy, who are Fiscal Policy Initiative colleagues, Danny Hile, Josh Ryle, Tom Church and Ben Jarris, what they would ask you.

And so one of these questions that comes from Danny and a few others here is really, I'm curious what you think about sensitivity analysis and maybe doing some sort of CBO sensitivity analysis with an estimate. I mean, you tend to stick to point estimates, and I can understand why you might be reluctant to offer lawmakers, say, multiple figures.

It can also, I think you maybe leave you more vulnerable to getting things wrong and we could talk about the inflation Reduction Act, Affordable Care Act, Medicare, prescription drug costs, some of the newer things that you mentioned before. They're difficult to model because they haven't existed before but I'm curious how you might think about that.

Or, obviously sort of macro variables are different from things like the response, various response parameters, both these things are subject to some uncertainty. But do you have any thoughts on maybe something like a confidence rating, attaching a confidence rating to smear scores? The intelligence community, for example, uses an analytic confidence scale that allows analysts to indicate how confident they feel about their findings.

Uncertainty is always a difficult thing to measure. But I'm curious, would you ever think about, you know, sensitivity analysis and using sort of multiple estimates rather than just a single point estimate?
>> Phillip Swagel: Yeah. Yeah, and it's something that we think about internally. We've done some formal modeling, and some of the members of our panel of economic advisors have been helping, helpful in letting us do that in a kind of analytically sound way.

So the uncertainty around some of our numerical projections, budget deficits, growth, things like that, we compare ourselves to others, to the administration, to private forecasters. Of course, we could all be wrong. Many people were wrong about inflation in 2021, and we were one of them. So being with the herd isn't always the right thing.

But we do try to make those comparisons, we try to give people a sense of what the uncertainty is as best we understand it. And even in a cost estimate, we have a section on sources of uncertainty. We have a section on the Basis of the estimate, what information went in, and then what do we think are the sources of uncertainty for the congressional process?

They need a point estimate and ultimately we don't have a choice, right? There's, we have, we do a cost estimate for everything that gets voted out of committee that's going to the floor. We don't get to pick and choose what legislation we're doing cost estimates on. And yeah, so there's limitations.

But yeah, it would be nice if, if we could do more to indicate this uncertainty where you said the intelligence community is kind of an interesting one, like the, yeah, just to say, hey, this one, we did it really quickly. There's. If we had infinite time or even another three weeks, here's the things we do and be able to come back to it.

That's kind of, that's kind of interesting. We come back sometimes to estimates, but like the Affordable Care Act and the mandates, the mandate in the Affordable Care Act, we published something explaining why CBO thought what the agency thought and how it was different than what turned out. But yeah, we could definitely do more on that.

>> Jon Hartley: Well, it's fascinating because I think about standard air bars and how useful they would be. I remember several years ago, Loretta Master, when she was president of the Philly Fed, she recommended that FOMC policymakers should be allowed to include a standard error bar around their Dots in the SEP.

And the survey of economic projections that they put out, they put out these, both forecasts of say, things like GDP, inflation, unemployment, macro variables. But also they put out their forecast of where they think the policy rate, where the short term interest rate is going to be in the next several years in the long run.

So imagine if they could include a standard error bar as well to sort of put in some level of uncertainty, maybe the market would react to it in a different way or be an easier way to communicate both, I guess a point estimate and some degree of uncertainty as well.

I'm sure we could spend all day talking about all these sorts of. Phil, I really want to thank you for coming on. It's been an amazing conversation. I know you're in a very key spot right now given the tax bill that's going through Congress. And I know CBO is spending a lot of time coming up with cost estimates during the reconciliation process and working with members very closely.

So really want to thank you for coming on during this very busy time.
>> Phillip Swagel: Yeah, thank you. Thanks so much. Thanks for having me on. And thanks to Hoover for hosting the podcast. And, yeah, I look forward to continuing the conversation in the future.
>> Jon Hartley: This is the Capitalism and Freedom in the 21st Century podcast, an official podcast with the Hoover Economic Policy Working Group where we talk about economics, markets, and public policy.

I'm Jon Hartley, your host. Thanks so much for joining us.