Income Taxes And Entrepreneurship

 

- Go ahead and get started. We are delighted to have Ellen McGrain here talking about income taxes and entrepreneurship. And Ellen is visiting, has been visiting here. And when do you leave? So how much longer? Friday. All right, so, so last chances to meet with Ellen.

- Well, I'm not on the moon.

- What? No, but person. I actually

- Am reachable

- Anytime, right? Anyway, so, so I think this is probably part of an ongoing project. 'cause we saw a great paper from you last year and then some previous stuff too on sweat equity stuff. So thank you

- As well. Always said this is, we've been involved in a joint growth project with the people at the Internal Revenue Service. They have a, a big mission, which is to think about, I use the term entrepreneurship, but to really think about business owners. So I want you to be thinking about, especially during my talk as I start to model them, I want you to be thinking about us business owners that actively manage a business. They might be a sole proprietor, a partner, or own an S corporation. They're, we call them to pastor entities. They're of great interest to the Internal Revenue Service because as many of you know, they're a big part of the tax gap. So part of the project is to think not only about redesign of, of tax policy, but ultimately about tax administration. So we're inching our way towards that as we, as we do this project earlier installments, were to, to get data out that specifically look at this set of people that I'm gonna be talking about today. So that was a big part of it is, and all of that stuff is up on, up on the website in downloadable forms. So for those of you who teach, it's a great resource and we tried our best to pull out as much as we could. So pulling out the data was a big part of it. And also thinking about, you know, what these business owners do day to day. So that's been a big part of thinking about theory development and how to model them. This latest installment is gonna focus particularly on income taxes. What I had presented, what Valerie was alluding to last year, was thinking about both income taxes and capital gains and capital taxes. And, you know, that work with Amal and Paolo Marini has really, I mean, our big conclusion is that the income tax taxation is much less distortionary than capital taxes. So that will be my focus today. And, you know, given Eric is here, he always likes to say, okay, what is the game that's, you know, what game are we playing? I'd like to say, you know, what's the want operator from, from this installment? What we've been working on in the past year is to really up our game in thinking about methods that can be applied to, you know, how should we tax business owners. So there's, we're, we're going to be, you know, instead of showing up to fight a fire with a squirt gun, we're gonna try to use actual hoses. So we're gonna bring, you know, better methods to think about transitional dynamics. The model I'm gonna write down today, you know, the transition occurs over 450 years, which sounds ridiculous since we live in a country that's only been around 250. So the idea that we're gonna do steady state to steady state is absurd. So we wanna be thinking about transitional dynamics. And now that we have the data, the, the other part of the, the hose to put out this fire is to think about models that actually fit both the macro data and the micro data. Okay? So if you walk, and I should say just disclaimer, if I walk down the hall in Minnesota and knock on the doors and say, how should we tax business owners, I'm going to get every possible answer I'm going to get. We should be taxing them, you know, very high rates to, we should be subsidizing that. Now, the people who are thinking about the very, very high rates, they're thinking about gains from redistribution. And you have these business owners that might in elastically supply effort to build their businesses and we should tax them. And that's gonna bring welfare gains on the other end. There are people who worry about financial, you know, constraints on these business owners. They worry about misallocation of capital in the hand of these not in the right hands. And so they're gonna get, you know, be more in line with, we should be subsidizing them today. Wait, today I'm gonna be really focusing not on financial. I know this will disappoint the guys from the finance side, but that those are, those are numbers being crunched right now. I'm gonna sound like Chris Sims, they're in the computer today. I'll be focusing on a model that without them, but there's gonna be enough to say to the people who work with igar like models that are, that are still working with the squirt guns. Hopefully. So in answer to him, I wanna set a new bar. I wanna, I wanna up the game both in terms of the method and the data that we're, we're working with John.

- When, when you knock on the doors at Minnesota, does anybody say we shouldn't be tax income, we should tax consumption business owners like everybody else?

- Not so much. I mean they, they talk about consumption on the job and that's part of the tax compliance compliance stuff.

- Is that, I mean, I know there's even stuff introduced in Congress just does IRS are you allowed to ask questions like that? It's just, we got an income tax, let's go see who pays it.

- Yeah, I mean that obviously when we're dealing with the IRS ar thinking about income taxes, if that's what they oversee, really where this is headed is thinking about both income taxation and enforcement. And it could be we end up with consumption because it's, you know, a losing game. Get them at the

- Porsche dealer.

- Get them at the Porsche. Alright, so let me go on now just a warning, because you guys are such a wonderful crowd and curious about everything. I put little blue buttons all over the place that have, you know, backup pictures. I told Dan, don't unfurl those because it's like putting catnip in front of cats. So I'm gonna kind of assert some things that you've already seen in previous talks. Obviously I'm happy to talk about some of the evidence that's in behind the blue buttons at the end. Okay, so here's where I start to assert some things and the blue buttons have evidence, as I kind of already alluded, the, the central characters in what I want to talk about today are the, are these business owners in the work we've done with the IRS, we have, in the data we've collected, we've learned that the surveys that we usually use to study entrepreneurs do a very poor job in reaching the very people we need reaching the, the dollars in entrepreneurship. Okay? And if we're thinking about the dollars in entrepreneurship, the reason we're missing them when we look at survey data is because we miss kind of the very right tail and we miss the left tail. We miss, you know, the high earners, but also those people who make losses. We also, I wanna point out, 'cause you know, I'm a macro economist, a lot of times I'm talking to macro people and they don't realize that the pasture businesses that I'll be talking about are actually most of the economy. They're most of business income in the US economy. People think of Jeff Bezos and Amazon, but that's just kind of a small part of our, our of our overall business income. Again, I won't unfurl that, but they're more than 50%. And if we really start imputing the misreporting that it's even bigger. So they're obviously a central group for thinking about income tax debates. So that's the motivation. Now there's a lot of challenge. The, as I already noted, the IRS cares about them because they report about half of their income, either through underreporting or over expensing. There's just limited informational requirements of business owners. So how do, how do we know that have, it's not reporting. We have estimates through audits and that's gonna be part of, part of the project. The next step of the project is gonna actually be using the national research program audits. So we'll have a better sense of of, of that. Right now this comes the, the reporting of half comes from kind of imputations based on what they do get glean from the NRB.

- Just define exactly what you mean by business in this context.

- Yeah. So every time I say business, I want you to count all the, so predators, all the, the partners, all of the, do they have to have employee scorp owners or you? No. Okay. Most businesses, by the way, just FYI, most businesses do not have employees. Yes, that's what I'm trying to Yeah, employees are kind of a pain. Who, who wants them? Okay. So, but the cha one challenge is of course that, that, that we're seeing happening

- Anyone who has non wage income,

- Say it again.

- Anyone who has non wage income

- Is not anyone. So I, I'm gonna, I'm gonna put off talking when I get to data, I will get more into it, but you know, when we go and think about owners, I don't wanna be including the person who gets a hundred dollars for reading the a ER report. And so there are thresholds when we do these definitions, it's, it's laid out in the things I'm not unfurling, but, but happy to talk about it.

- But do they have to file one of those in one of those ways rather than as a,

- These are gonna be people that file either on the schedule C 10 40, they file the 10 65 or they file the 1120 s. Think about those. I ideally we would also include smaller, you know, people actively managing C corporations. They're just much harder to see the owners. Okay. Now, something that Valerie mentioned is that one thing that has been of interest to us has been or kind of been central in our theorizing is that there's a lot of investments made by these owners in their business that we only see when they sell the business or, you know, put it on the books. So that can make it a little bit difficult because that could be where a lot of the value in these businesses is. Again, I won't unfurl the thing, but in the stuff we're doing with Paolo, we're looking at filings that actually let us see

- That that's what you showed at the CBO conference Les Summer, right? Didn't you present this, this

- As the Bob Hall?

- The Bob Hall thing?

- Yeah. Yeah,

- Thanks.

- Yeah, that's right. Okay. So that kind of makes it a little bit challenging is that it's harder to see. So hence we're going to be using theory to kind of see these latent things. Okay, so what specifically, here's the game, Eric. We're gonna develop a theory with occupational choice and business sweat capital in some sense. We've already developed that. We're gonna be stripping some things out so we can do a better job. I hope of estimating key elasticities. So we're gonna actually be looking at a little bit of a paired down thing relative to other things I've presented to you. But here's the key fire hose. Now we want to match everything up to nepa, nationally income in product accounts and this tax administrative data. Okay? So this is, now we're, we're trying to get serious about the data and we're going to be using some methods we recently have been working with to study the optimal tax rates and the transitional dynamics. So those are kind of the new, the new pieces

- When you think about the optimal tax rates and the elasticities, the elasticities are endogenous to the existing tax code. And it gets back to the question of whether you're just changing the rates or changing the code. One has a smaller one, has a larger potential effect on the elasticity you might expect to observe in that alternative scenario.

- Yeah. So they don't, these people don't follow the code. So I'm gonna be starting them at what they effectively are doing and then show you transitional dynamics plus and, and welfare gains and, and information if we were to change it today so that they're, you know, facing something new from now on.

- But if they weren't acting consistent with this one, why are we going to assume they're gonna act consistent with that one?

- I'm imagining that just, this is the hypothetical. Imagine we could perfectly enforce higher tax rates on them. So it's just, it's a, it's a counterfactual or lower. Okay, so let me start with the theory. There are going to be two sectors. This is, I want you to be thinking of these pass through businesses where owners are are involved because I wanna match everything up to national accounts. And there are workers in, well currently maybe not for long workers in the federal government and state government. I'm gonna include them with, you know, kind of the other sector. So think of that as all other businesses, including government plus C corp. And then there are gonna be households that are endowed with stochastic abilities to either run a business or to to do paid employment. Okay? So here you could think of a dentist and that'll be on one of my slides will be either setting up their own practice or working as an employee for another dentist.

- Yes. I presumably gonna leave out something Chad Jones pointed to as being big human capital investment. So you're going to school, if you live in a very high market tax rate country, say I'll just be a social worker. Whereas if you live in a low market tax rate company, you say, let's go get that computer science

- Degree. Yes. So we're, we're in the us Oh, for sure. And you are going to, so here there's going to be that stochastic ability is going to be you, you could think of it as the ability of the person, but it's fixed. Imagine, I want you to imagine that you're kind of coming in as a 25-year-old after college. Yeah. You could imagine doing the whole, let's start the world off. I'm five and you make some decisions, there's going to be another that's orthogonal to the tax policy. Yeah. That's gonna, yeah, well

- Look, part of the concern

- Is business formation. But, but can I, can I, can I hold you off? Can I hold you off? Because there's an investment. There's no, I can't It's impossible. It's impossible. There's going to be investment in your business and one, so gimme two seconds then there's government revenues on taxes and incomes, plus expenditures on goods and services. So I'm coming to the investment, John. So there's a value over the following states of the world. You have financial assets, think bank account, you have capa. CAPA for us is going to be the, think of it as the customer basis, trade names and other business capital from your sweat that's in your business. S is gonna be the shocks to productivity and self paid employment. S could be endogenize and, and be things that are specific to you, but they're not transferable. Okay? We want to distinguish always things that are transferable from things that are transferable because it's the things that are transferable that we can ultimately see the things that are inherent to you. They're just inherent to you.

- Ellen, how is, how is sweat capital different from human capital? Like, you know, the, the,

- It's it's

- The dentist, the lawyer, the, you know, these people

- Are highly educated. It's the patients being passed on. Okay. You have a nice demeanor and a nice smile. That's when you, okay, maybe not, maybe not, maybe you'll be a lousy dentist, but those things are not transferrable. So we're thinking of it'll be a letter Z and in fact let me, let me, let me pop up the model. So there's going to be the value of BUS being the business owner, okay? You get utility from consumption, you have dis utility from putting effort into the business to build List the yeah. Right. For these things, right? Right Now if you look at the profits, you see the third line John Pie, it is that there is some exogenous Z affecting your productivity, but that's not the patient list that can ultimately be passed on. One could endogenize Z and add another letter H to, there's human capital that's specific to you. That helps. And that can be built while a business owner identifying the dis, you know, the two will be a little tricky, but it, that could be in Dutch. But let me continue. So we've got utility over consumption, dis utility over effort. That dis utility over effort is going to be, what I'm gonna be very interested in, obviously, because I'm going to think about two versions of this model, especially with regards to income taxation. One where you just have fixed, you just have a, a fixed factor of production kappa bar if you like. And another where you accumulate those. Okay? Most of the profession just whether if they work with a decreasing returns to scale, you could think of them as having a fixed factor payments to the owner who's the residual claimant. But in this case, in our case, there's an elasticity that's critical. Okay? And I'm gonna show you that it's critical, otherwise it, you know, we've got, you know, that you get a return, an after tax return are on your assets. You pay tau B on your tax. That's gonna be the income tax of interest today. You get transfers. That's that little tr you pay some, there's, we're gonna add in all the other taxes. John, they're all in there. 'cause I gotta make everything add up.

- You hear that questions coming?

- Yeah. But, and it will match the us It's just what we're gonna play with is TBI and then in the profits we've got again the capa, which is the customer base. But they can also scale the business by renting external capital, think physical capital office space that for the dentist and maybe acqui you can even rent equipment. And then hiring workers. Their question

- Just quick, yes, this is kind of following up on John's comment, but think of it in the broader context of the longer run transition. I, I don't know the data, but I suspect that it's easier for somebody to, to become an entrepreneur if their parent, one of their parents was an entrepreneur. And so this is the case where if you had low tax, low tax rate environment, you could also get more entrepreneurship just from the learning by doing from the parent they going through,

- You could think of a parent passing on Kappa to the kid. Yeah. Now in this version, in this model, we, because we're not thinking about capital taxation and capital gains, we're going to focus on, you know, a dynasty. There'll be, it's not gonna have a life cycle. But you could think of that that CAPA being passed down. It's

- A really dumb question. What's that last term? On the first as a constraint, the one minus, one minus tau C time C. What is that? What consumption?

- Tax

- Consumption.

- Because you have a consumption, not an income tax. There is income. Tau B is an income tax. Then workers will pay an income tax on their problem. I'll show you that in a middle in a minute. So those are the income taxes. Then we have, you know, all the usual, we have a profit tax for corporations, et cetera. And any taxes on things you're holding in your mutual fund, like dividend taxes kinds of things would show up our, think of our as after all those payments. Okay. So there's a mutual fund, okay. They do all the investing,

- They pay my taxes for me and they give, they they

- They pay your tax. Yeah.

- Yeah. You answered a question to from Lori earlier about observability or at least I thought so. So here we're assuming there's no problem with the other tax authorities observing these profits, right? We're just abstracting from that.

- The way it's gonna show up is TB when I go to calibrate the model is 20%. TB statutorily is much higher.

- I see

- It's going to show up as their effective tax when I calibrate.

- See - So I'm taking is given that there is cheat

- Cheating going on. Micro found that in incentives you have to assume that there's a half probability that you'll get taxed or something like that.

- No, I'm just assuming everybody's cheating

- And and they're effectively paying half the tax rate. But

- Yeah, they're paying up. No, your micro, now you should ask, well Ellen, I thought you're gonna make it all add up. Wanna ask that? Okay, please. Good question Carol. Because the NEPA adds that in. They assume cheating. Do you know that half of the NEPA is imputed and it comes to business income? Wow. A lot of macro accounts don't know that.

- And again, because it seems ludicrous, they are, they, they're saying people are cheating, therefore we're, we know how much they're cheating. So the number actually is this

- National income and product accountants wanna give you a good assessment of but is truly going on. So what they do is they, from the audit data impute an estimate of the business income. That's what we have. And we are going to use that. We know there's reporting and we know there's IMP computing. So we know it. And our model is going to be designed to match those things.

- There's a certain circularity, we impute it and you're gonna make it add up to the imputed value. Which assumes that tax gap is basically based on the

- No, no, it's a matter of they don't fill in taxes, there's no tax collected. It's just, there's money. So it's, it's not circular. I

- Thought you said you were gonna make everything match up and

- I'll show you, you hold onto that question. Yes. The effective tax rate as well. Yeah. The effective tax rate guys. Yeah. This is not the truth if only half the people pay. Now of course when we go in our next installment, when we start thinking about enforcement and all those things, we're going to endogenize those choices. That's what where we're headed.

- Yes. That was the question. Name is the Z And then their choices, the business owner will

- Have, business owner has a, everything has an i A I, capital I. So the only

- Exogenous thing that,

- Yeah, the only exogenous thing here is

- In the w - They take is given the wages and the rents.

- Yeah.

- They take us, given the

- Wage doesn't depend on like who they hire. 'cause there's heterogeneity there too. I'm just trying to get a sense of where

- The these are that they're hiring.

- Okay.

- Yeah.

- Ellen, the part of the confusion here is that in reality what you have are numbers for our income. One is the income you have and then the one is your report.

- Yeah. - And those are two different numbers. Yeah. And, but you're just sort of swapping together and saying half Yeah, I'm gonna say

- Your your your

- Your two different numbers. Agreed. And that's where they the adding in Yeah.

- Everything me very aggressive tech.

- Yeah. No,

- We don't.

- I'll argue that later with you. We'll go, we'll go out for coffee. I'm gonna move on. Oh, oh, sure. Does it matter if there's cheating? Like people say

- Different.

- There will be an obviously when we go to think about doing that. There, there is. And I'm, I'm sloughing over that. Okay. That'll come. Yes. Oh, here was my dentist. So you gotta put your time in to build your patient list. You're gonna wrench your office space and your equipment and the NB are like your hygienist. So here's the worker problem. So now there's no effort to build Kappa, but kappa is a state for you. 'cause you could, we're gonna allow you to switch back and forth. So there's going to be a lambda that, that determines how much is lost. If you switch, like if you are a dentist working in your own practice and you now say, oh, such a pain in the butt, now I'm gonna go work for this other guy. Well, they're only gonna put up with you so long being away. Right? 'cause they'll go to the other guy.

- Okay. Just where was the other dentist they hired in the dentist problem. That was part of w So the dentist and the hygienist are both part of n

- Yeah.

- I mean

- I'm,

- I was worried about this composition part.

- No. Yes, yes, yes, yes. The other dentist would be a worker. Here's the work bubble, which where I have a

- Dentist, which is different than the hy hygienist. I'm just trying to see the hygienist and the dentist are gonna have different, I forget your things for skill. Yes. That were gonna be in the background.

- I understand. If you wanted a really, really to focus on the labor force, you would get into that. These guys are really not central for us. Yeah. But if you wanted to really think about the labor within these businesses, and you could talk about high skilled, those field and things like that. But they're not gonna be central in our taxing business. Because what's gonna be central is really the effort the guy puts in or gal puts in to building this that's gonna be central.

- Where does the worker's sweat capital does the, the worker has his own sweat capital, which he is using when he is working or not using when he's working.

- It's just decaying.

- So he is not putting any, he's not building meaning he goes to work with the other dentist list. He doesn't work on it anymore getting patients.

- So he, he's, those were his old patients in his old

- Dentist. So, so there's another margin. I'm not quite sure what in all this. Maybe you've already explained it and I missed it. So in addition to whether you're gonna be in a, an employee in another dental, a big dental practice, I work for Kaiser as a doctor or you're gonna be your own person, et cetera. You also have the, the decision on how much to reinvest in your business and how much to take out and get taxed. Where does that come

- In? Right? So think of, I want you to think of back when we're, we're the doctor, if we're the doctor that's working at Kaiser Permanente, right? And we're now thinking about setting up our own thing. We're gonna be starting basically with no kappa. We, we had nothing from ki we can't take it from Kaiser

- Mente. I know this 'cause my daughter-in-law is a dentist and she worked for a while for another dentist. But she always wanted to be have her own business. Yes. And then so she bought both the building and the patient list from a retiring dentist and then brought that up. It was a small building. And then she got so successful, she has now bought the much bigger building and the extra patient list. Yes. From another And then she even hires other dent, another dentist plus hygienists and stuff like that.

- Yes. So in the model that we had with that, that we discussed at the conference for Bob, we had the buying and selling of these businesses to put it in here. When we look at Tby, the commodity space blows. So it is actually, it's, it's on a to-do list that we really need to seriously up our game because that you start to, it's very difficult to to, to include all of that. That's fine. I'm just giving you anecdotes. Oh yeah, yeah. But I mean, yes, but that's exactly right. And, and when she started, she's going in and then building her thing or buying.

- It's no difference between income and capital gain here for example.

- Yeah. So again, these are, these are dynastic do not have the capital gains or selling in this version of the month. It's another level up. It's another, we need a bigger fire hose.

- What's the next, what's the next Just go one more.

- Okay. Sorry.

- The worker doesn't exert effort

- Here. We're going to assume that they inelastic supplied labor hours a week. Okay. They're all working 40 hours a week. The only elasticity in terms of labor is really the effort of the business owner to build the patient list.

- Isn't that what he's, he was worrying about that. The marginal tax rate effects efforts.

- I was worried about investment skills to become

- A Okay. But tax rate effects efforts,

- We are Okay. So you're just worried about it. One could add those additional margins. We're going to try to understand this model first because it actually, the getting at the elasticity is gonna be critical. So I want you to hold onto those great ideas. We purposely left out any kind of margin on the, the length of your day. It was a purposeful, thoughtful endeavor to leave that out.

- But, but you know, you can go from being a worker to an entrepreneur

- Yes.

- In the future.

- Yes.

- Knowing that you might want to have some effort put into maintaining your kappa. I mean if Kappa's a Kappa's a broad concept, it's not, you know, it's, it's, it's not, we

- Don't where you can build Kappa in your other,

- You don't have them. Right. And that's what you want, Jonathan.

- Okay. So the corporate sector's boring for anybody who's used to igar models. This is just like the usual corporate sector where they're going to, there's a technology a is a constant number and they pass dividends to the mutual fund. And the mutual fund passes a return to the household. Very boring. Here's our mutual fund. I won't go through this again. Just think there's a bank account and they pass it to the household. Competitive equilibrium. Yeah. Everything, every price taking here are all the choices that have to be made. All the fiscal variables, everything has to add up and clear. Usual. I'm gonna say something briefly about the computation. I don't wanna get too into it, but I wanna say in terms of that model, one thing that was a little bit, like I alluded to at the beginning, the transitional dynamics because of the sweat capital are slow. And, and we've got the two, you know, we've got the financial assets and the, and the sweat capital. So we've got the double, you know, the two, two state, two endogenous states. We've got the discreet choice and we've got these long transition periods, length 450. So to deal with those, we've kind of been working, this is, you know, the first part of upping the game. We've been working on improving certain algorithms for solving these models and for dealing with these transition dynamics mainly on methods, perturbation methods that give us a good guess for using global methods. So we get a very good guess and then we can kind of get to the answer fast.

- I missed something on the general equilibrium concept. So if I'm a dentist and become a worker or vice versa.

- Yeah.

- There's gonna be a, a displacement in the employment market for workers or dentists. Is there a cross-sectional distribution of types of people that are slotting into these positions based on the need for those positions? Or is there just one person that might be a dentist or he or she might be a, a worker that doesn't need a dentist?

- Are you asking if there's imperfect substitutability between,

- There's a continuum of, right, as a continuum of agents. Yeah. Is that, is that how

- It works? Yeah. So you're, you're an agent is defined as a CAPA z, which is the Right, your stochastic ability in or in self-employment and epsilon, which is your stochastic ability in pay that, that defines you.

- And when I, but am I infinitesimal so I don't need

- Yeah.

- And and there's a no feedback.

- Single consumption Good. I think. Yes.

- Yeah. And so, and it's sort of IID migration. So there's steady, we're always in steady state population of each type.

- We're not always in Stacey. So that's what we're going to do a tax reform and then

- Do a transition. We've

- Got a path of,

- I see

- Interest rates, we've got a path of wage rates. We've got the whole general a Librium 450 periods.

- And is there something that slows down the migration between types once the transition begins? Like, or do we instant

- Taxes.

- Taxes are gonna,

- I mean taxes will mean you're going to change the fraction of people who choose to run a business.

- You'll change that. They're gonna change instantaneously or is it?

- I'm gonna show you.

- Okay, great. Thank you. But

- The, and because of the accumulation of this CAPA capital, right?

- Yes. Turn off the capital. Well the only thing that kind of slows you down a little bit would be the other assets, but, but those don't slow you slow much. It's the cap that's really the reason. Got it. We're gonna have a slow transition. That's a big deal. Even if you stop like kind of quickly putting in effort, you still, this thing has got a decay over time.

- Now Ellen, how do you handle the fact that the state, the state of the distribution, state of CAPA is a continuum? How do you approximate that?

- How do I approximate the,

- The state Very one, one of the state variables is the distribution of Kappa.

- Yes. So a a kappa and Kappa are 2D. There's a 2D distribution over them.

- Yes.

- And then we're going to, for the stochastic shocks, those will be a large Yeah. So markup

- State variable is really an infinite dimensional state variable that at every point in time you have a distribution of what, two dimensions. So therefore that distribution is an element of an confidential space. How do you approximate that?

- Yeah. So that'll be approximated, you know, over, we're going to ha assume a certain size of the a grid or a domain and a certain length of the capa domain democratize that. And that is going to be, yeah, we're gonna have a, a distribution over every, think of it as over every point in that domain.

- So every every point is a continuum.

- Agreed. But every point on a grid over those.

- Every point on a grid.

- Yeah,

- I agree. So it's, you're, you're discre the,

- There's gonna be, think of the a a a grid where there's flying over the grid and we're going to be, so that's a flying

- Approximation of the, it's not a disco discretization of the state. The spine approximation

- On the 2D and then a markoff chain or the stochastic. We track that over time. Correct. We track that over time.

- So the but the, because there's no a there, there are no aggregate shocks here, right?

- There's no aggregate shocks.

- So the, basically the distribution of the book and the assets are themselves just follow a deterministic path

- Basically. Essentially. That's right.

- So - This is just, and you're finding that distribution that you're finding that path. Yeah. Yeah.

- But z and you, you do have these exogenous processes on Z and

- Absolutely.

- That are auto aggressive.

- Well, no.

- Okay. Hold on in the paper. Hold on because I need in the paper you did, but now that you've got something else,

- I need something else. Yeah.

- Do workers work for another company that pays profits tax? Who do the workers work for?

- The workers work either for entrepreneurs or for the corporations. Corporations slash government. And that has to add up. And so then somebody else pays the profits

- Tax. Okay.

- The corporations pay the profits tax. That's right. Okay, so let's get to, how am I going to pick some numbers and what the want operator is here? The want operator is, I want to match the NEPA and I wanna match micro moments that we have from the IRS and some auxiliary moments that we, we pulled from the census survey of business owners and I'll show you where we needed something. Partly because the IRS data was that the universe of owners and not owners slash business. And I'll show you where that comes in. This is the first part is very easy. The second part is where, where work is needed. Okay. What are the key

- Parameters? Say it again. One thing that I'm not, I can't quite figure out. So you have somebody who works for a corporation, but it's become very common in recent years for people working in high skill skilled people working in those businesses to form teams that get equity in the company. So how is that that's a way for them to realize their entrepreneurial potential inside.

- Are you talking about like a a we at Microsoft and they get stock

- Options. Yeah. For example

- Or, so those will, that'll

- Show up in how about just open AI

- Or that that'll show up in their W2,

- But how are you thinking of those people? I mean

- They're, it's they're, it's

- Entrepreneurial activity in

- Inside those in this world since it it would be, I mean they literally are in with workers.

- Yeah. It's ordinary workers. They're,

- They're just ordinary workers. Yeah, that's right. Okay. So what are the key parameters here? They're the taxes obviously. And this, I alluded to earlier that the TAL B is going to be about 20%, whereas workers are paying roughly 37% the productivity processes. We're going to, Ellen can I stop?

- I mean, yeah, so you said progressivity doesn't matter but some workers, federal, state, and local, some workers pay like 65% and I underst understand lots of workers pay 0%

- Or 10%

- Who do

- Have a very, not lots. That's where I'm gonna, I'm going to, you know, go to the board later. Not the what? Not lots.

- Most workers are about

- 50 people who don't work don't pay anything. So there's those people

- 50 percentage but people who pay or 50 to 80 grand a year, don

- X rates. I'm gonna hold off on this 'cause I can show you other data. This I won't, I don't even have a blue button to, to do it. So you think that a flat 37%? I think it's a beautiful approximation. It's, it is like the best one to do. Alright guys, I'm gonna move on a

- Major result then. I thought I was like him. I thought that everybody paid. You're saying that isn't true.

- Don't listen to him.

- I know, but I mean, no, hold on, I'm gonna put you

- Guys off. I'll run outta time. Okay, so the product productivity process process, this is gonna be kind of critical or we won't match our, our IRS data. So the AR ones won't, won't cut it. So I'll show you what we have to do. And then a key thing for elasticities is this sweat capital, the deterioration, like how much do you lose if you switch and the actual, you know, distil utility of effort, those are gonna be things I'm going to show you how it affects,

- I hate to tell you, but doesn't that mean that 37% of the economy is the government? If what you're saying is true 46% federal, state, and local. Oh got it.

- I I'm gonna table all these discussions until I can get my computer and pop it open and show you guys stuff.

- Modeling - Question. Does this have to do with this slide

- Or it's a modeling question when you take to the date. Yeah. So what in the model allows things like the labor share or the profit share to change over time Or there's, is it only taxes or is double count production? I'm just trying to get a sense like on the transition dynamics we see huge movements in the NPA data of profit shares moving and labor shares moving. I just wanna know which is it the Z or is it something that's gonna be able to move the profit share and labor share?

- So this right now is going to be take United States currently

- What year

- And I'm going, okay, so we're going to be using data like over the past couple decades because that's what our IRS

- Stations.

- So when I think about

- That transition dynamics, I don't want to think about transition dynamics in the existing data. I want to think, 'cause there's a huge amount of movement from 2000 to 2020 that we're in the labor share, the profit share in NIPPO data.

- You gotta be really careful when you say labor share to me.

- I'm gonna say whatever.

- I just gonna say that whenever Lucas

- Said

- Yeah.

- Data, - He also has to be careful when he says labor

- Share using nip a data there wasn't moving

- For Exactly. We can go over that.

- Okay, that's fine. So you're telling me it's not interesting to think about?

- No, it's very interesting. It's just, you gotta be careful when you say those words. But what I do wanna point to you is, is thinking about this estimating this, this is where there's some interesting stuff. So for the income mo business income moments, we start with this AR process, but the innovation Epsilon Z is going to be not just a Gaussian we're going to do because we want to get some key statistics from the IRS data. We want that the top 10% of our owners have 70% of the business income. We wanna get the distribution of income growth. Okay. What that distribution would look like, if I unfurled it would be that income changes at the 10th percentile are about minus point, minus 50% and at the 90th percentile are a hundred percent. So these are guys who are going minus 50 plus a hundred that spread for the paid employees is about a third to a half of what I've just described. So we want that, we wanna get that volatility and then we wanna get the right persistence. So we wanna match rank auto correlations and we do rank because we're dealing with IRS data where you've got potential problems, you know, in the very tails.

- And what causes that variation in your model To get those, to get that thing It's going to be what?

- Yeah,

- Born capital sweat capital or something.

- Yeah. So we need to get the, the, the high, the right skewness.

- Yes.

- We're going to model the innovations to that stochastic process to be this which is kind of, well actually lemme do the, I'll do the picture this a mixture so we can get the right tail. If we did the blue, we can't fit our IRS data.

- This is the innovation to to, or what do you call that

- Production inno It's, it's, it's the innovation

- But it's an explain. You're not gonna be able to say what causes this at all. Right? It's it's not, it's not gotta do the sweat capital part.

- No, this is the sweat capital is endogenous. Yeah. Yeah. So the sweat capital you, you, you are going to accumulate based on this. Okay. So that the blue will be fine for, for doing the paid employees. In other words, having a Gaussian just ar one process fine for the paid, but you'll miss the right tail for the other. Okay, so now I'm gonna tell show you some results. I'll first do the macro data, then I'll do the micro data.

- So lemme, so, so ZZ is, is exogenous productivity for the entrepreneur. So you, you made a choice to put that, put this unusual distribution there. It could have just as easily been put into the accumulation of sweat capital. Well the accumulation of sweat capital. Yeah. You do some advertising and suddenly a whole bunch of people show up and you know, and that's a big jump in your,

- So think of it that, that it is affecting like you're getting shocks. You could lose a, a client.

- Yeah.

- Okay. That would show up as, or there's a, there's a shock to demand for your stuff and you could lose some clients, but you're, you're end ly choosing effort.

- What's the e shock? Is that just normal or is that expon? Expon. So that's the, the E

- Is effort.

- Oh, E is effort. Yeah. Right.

- E is endogenous. So that's what you're choosing. Yeah. That's your time into building the business. But you are facing,

- Yeah. A so process

- You are facing these.

- Yeah. And is there any way to make that real? What what you back to have something in the economy that you can relate these shocks to you,

- Their incomes. So we are going to look to see what incomes come out of our model and match 'em up to the incomes we see in the data. That's what we're gonna do. Lemme start with the,

- You just look states out of this

- Income

- Tax, state expenditures.

- No, it's all in, everything's in. This is combined

- With that. Those numbers are way low compared to,

- Okay, so let me explain one thing for people who do the national accounts differently than me. We, I have taken all non-defense for public. So consumption is public and private, including non-defense. 'cause I think of health and education as consumption. IG the stuff gets tossed in the ocean. That's defense. That's that, that's not substitutable with, with private consumption. So that gets treated separately. Notice transfers here are high because I'm going to assume that again, your consumption includes the non-defense spending. The other thing I wanna point out, which is also in red, we wanna match the income to the sweat. Okay. How do we come up with that? We're going to take all of the earnings, including the, the imputed earnings, IE what they say in the audits and include it with the sweat income. Okay. Like the national accountants, we're going to get the total sweat income, not, not that reported on the tax.

- So Ellen,

- Yeah,

- She did that 15, 20 years ago is numbers almost radically d plus a decline in labor share.

- We'll talk later about labor share. You got it. It's a whole beehive.

- Okay, well,

- Okay. Hold, hold, hold up that thought.

- Gimme a stuff. I don't wanna stir the beehive. You, you do. I

- Agree. I agree. Okay, so the micro moments, here's where we're matching the top, the, the profit shares for the top. We actually, given our current parameterization we're actually overstating a little bit or too much the profit growth so that we have to kind of a adjust some things. So profit growth is just from the stochastic element of the pro pro plus the other factors. The the sweat will profit is Yeah, but mostly the, the most of the, the, you know, why we're going up and down. We, we will need those shocks. Yeah. The business entry and a business entry is 2% in the model. And the data exit rates for us are very useful in identifying our key parameters, the movement in and out of ownership and paid work self and paid work. And also the shape IE the elasticities Of the cost function of the effort. Okay. So we use two sets of data. The SBO, which gives us business age, it's crude because they've been up a lot of things. So it's a little bit crude, but you can see we're a little bit too flat in terms of our exit hazard relative to that data. But if we compare, if we look at low capa a proxy, young people by low capa people in the model, we can see that the stuff in the lower panel, which we match up to the owners in our IRS data kind of fit nicely. So what we need to do is include in the model. So we have to extend the model to include age of the business and age of the owner, which is not in there right now we have only really the business age. So we're still struggling with how to, how, how to fit these data. But I wanna show you why I wanna struggle with it. And it has to do with pinning down kind of the, the, the key things that are, that are determining the switching. So lay how much things deteriorate when you leave the business and go into work. That's going to change just the level of exit. If, if it's costlessly costless for you to go back and forth. Never no cost. So that's gonna change the exit rates. Also the, the a and the curvature of our distil are gonna determine the scale determines, you know, just the entry entrance because they do most of the investment. And more importantly, the the curvature on this determines the elasticity. Okay. And this is gonna be critical for the answer we get. I'm gonna skip over the numbers because I'm running out of time and I'm gonna give the predictions of the model. We give two sets of results transitions. What if you kind of move them towards their statutory level, how does it look and what is the optimal to how b what are the consumption equivalent welfare gains? And I'll also show you the revenue maximizing like the Laffer curves. Because in practice most people who study these things, that's all they have. Yeah. Yeah.

- Just to be clear, I'm not sure I missed this. When the guy moves to, from a worker to an entrepreneur.

- Yeah.

- He hires somebody,

- He can hire people and he can hire capital

- And and that's that response. And then those people will pay taxes. Right? That's in the model or that's not in the model.

- They're all paying, everybody's paying

- Taxes. But is it the case that when you change the tax rates, the entrepreneur now hires people, but of course he gave up a job. Is is that generally

- I'm believe leaving how w I'm leaving the worker fixed at 37%. Okay. Okay. And then I'm going to increase from 20% the effective rate on the business.

- Okay. I'm I was gonna much more simple question. You know, there's this intuition that we need on more entrepreneurs because they grow the economy. Yes. Because they have good ideas.

- Yes.

- That's in here or not

- It is potentially in here. 'cause there's gonna be welfare, welfare gains or

- Losses. First of your questions is yes, people when the tax rate changes, some of those people are gonna be workers and pay taxes. That's endogenous in there. The

- Workers have to work for other companies.

- The workers are always paying taxes. Tax. Exactly. So when everybody pays taxes, you pay business staff when you're owning. And then if I, if I, I stop being a business owner and go work for you, I stop paying my business taxes and I pay worker taxes under you. So that that's in the that's

- Right. That's in the mom

- Back on human ca you know, kind of innovation and those kind of things. That's John's question. She's taken the distribution of the epsilons as being fixed. Right? So that unless

- We change the distribution of these that are, it's just an enforcement question. Basically by getting rid of businesses, you are enforcing tax.

- You are, think of this experiment as you are for going from 20% to 40% as you are enforcing taxes. That's the transition. Then the question is, well what do you, what rate would you set if you could, but

- Ellen, what about the possibility that, the reason we have the taxes set up this way is we kind of know they're not gonna pay, but because we have all this benefit. So then we, and and

- One benefit, wait, lemme repeat what I said at the beginning. One benefit here is that they're very productive. They will do it anyway. And there is some redistributive

- Okay,

- Okay.

- Reception there. But that you made that extreme assumption

- And, and when I do sweat capital fixed versus variable, it's going to be making it extreme. Okay. Because when it's fixed, it's like having, it's kind of like being back in the igar model where there's inelasticity here. When I make it variable, now we start talking about elasticities. We're gonna go from, we're gonna enforce the taxes. I'm just gonna show you not the optimum yet, but how does it look with, there's just a fixed factor. You don't vary it. And but think Lucas 1978 almost, versus they can build their businesses. I don't show 450 years because you won't then see the beginning part, which is important. You can see with the fixed, and in fact I'm gonna go to the really critical thing, which is the owners you see with the fixed inel as hell, you get a little drop in the business owners, a little selection of the better guys stay, but kind of your, your no no impact variable. Very different. Okay. And if I looked at elasticities, which I do, I wouldn't look at the steady state to steady, is that a I've got five or you have a question? I have a question. Okay. Hold on one second. If we were looking at the steady state to steady state, the elasticities we would imagine here are enormous, but the critical ones are along the path. So for example, it was a big change in the TCJA. People have been studying what happened when the effective rate of tax on these very people was changed through the section 1 99 A for the nerds. They get a particular elasticity. We can compare to that elasticity and see kind of like how different is it? And I'm gonna show you after I answer these questions. Yes.

- There, there's a, if I understand correctly, there's a very strong assumption here. I'd like to hear your defense of it, which is you can't hire somebody else to produce sweat capital for you. That's right. But why? I mean, it's not obvious that, that, that's a sensible assumption. 'cause all the,

- If they were building it for you, it and they're actually, it's their thing then they're a co-owner. No,

- No. We have theories of organizational capital that you know well where you

- Hire. Oh, no, no. But if it's just, you're working with it. So, okay. It depends on whether or not you are building it. And it's part, it's your investment. I

- Own the firm. I hire Eric. He comes and develops my customer list. He's in, he signs an agreement

- With me that he can't take that customer a hundred percent. He's envy. He got w you paid him for exactly what he did. Guys, guys, guys, I get, I get a couple more minutes. You take it,

- Who owns it? Business. Who

- Gets to have it when you sell the business? Who gets to have it? Yes. That's the point. Not him.

- Okay. No, but you got Steve sector over here that doesn't produce sweat capital. As I understand it,

- They have a different kind of sweat capitals. They have a different kind of sweat capital and, and I and I missed that then when you're laying out the model. Yeah. But for the C corporations, they c corporations that KC, I'm assuming is r and d. You know, all the things that are, are being used in the capital stock of the Yeah. Also

- When we're driving people into the C sector by increasing the enforcement of taxes on the, on what you call the business owners. Where's

- That positive effect showing up? But hold on. If I go into r and d in that corporation, I am not an owner unless I hold stuff. But,

- But I trying, unless I hold stuff. I'm trying to understand the efficiency consequences of driving people away from the sector that produces sweat capital to the sector that produces r and d. And I'm, I'm missing that. Where is that

- And

- Why are they

- Going? And it's likely you would've,

- They're getting a higher tax. Yeah. Don't understand why it doesn't show up in these outputs.

- And it's likely you would've given Eric equity anyway and he would've taken a lower wage.

- I think, Steve, it's in their assumption that your, your business owning productivity and your wage productivity are completely orthogonal. Where the story, I, I have the same thing in your mind into your head, but there's gonna be the selection effect on wages. And I think that assumption is that when I become a worker now I'm just like a Yeah,

- You're like, you're like a little drone. A drone agree. You can still produce.

- That's an assumption I have.

- You're a drone for them. Yeah, but why can't you be, you gonna always do your own business

- Component that comes through. They think you'll be reallocating some of these skills

- Into this. It's not at all obvious that the best place to produce this intangible form of capital, whatever you wanna call it, is in these managerial firms rather than the C corporations.

- I'm not saying that that's not going on, I just, I did not focus on it here, but I'm gonna table this just so I get the last five minutes. Okay. 'cause I do wanna give kind of the punchlines. So if we do the same calculations as the people who watch section 1 99 A and then look two years, what we find, they find elasticity like, you know, percent change in in profit relative to the net tax rate, net of tax rate of 0.75. We find 0.62 here. So we're not, you know, obviously if we did study to state, to study state, we would say it's huge. But that's a very long-term thing. What you wanna be doing is kind of these shorter term things and what, what we can do with the model is also see where are those, where is it coming from? And in our model we can look at timet equals zero and say, okay, on the extensive margin, on the intensive margin, who is generating the, the, the changes in those profits. And we find that it's, it's mostly coming from existing firms, either entrants and their intense, their important on the intensive margin. They're, they're we're better selecting and then the extensive margin the mature firms are exiting. Okay. I'm going to, in my last minutes, I've already given, you know, what are the primary concerns here? We have not added any, you know, financial frictions or anything like that. Really what the forces right now are really thinking about kind of in many of the igar like models, people find very high tax rates because there are, there's a lot of inelastic rich people and there are big gains from, from redistribution. We have a, a fixed sweat case where that mo force will be most important and we have the variable case and the question is where it's elastic, how different do these look So we can get similar kinds of results like a tau B of 62% in the fixed sweat case. So this is just, there's decreasing returns to scale. There's fixed capital stock.

- Yeah. It tell us the objective. Is redistribution part of the objective or just

- This is maximize a utilitarian welfare function? What rate would you say?

- So it's re part of the objective or not? Well yeah,

- Because you're by shocks there's lots of shocks. Okay, so insurance. Insurance is the main modus. Yeah. You also have incomplete market. Yeah, there's incomplete markets, right.

- I also have a degree of inequality version somewhere.

- Now I'm gonna do the two together. So this is one of my main ahas. So you see in the fixed sweat case we've got 62% close to 1% gains, which is really in a macro sense large as opposed to 33% with 0.27 much lower. It's like a present value with short term benefits and long term costs. As people move out, the, the calculation is you can either stay status quo with, at, with the old tax system

- Below it and in the present value of the,

- Yeah, then, then you're doing the present

- Value. The short run

- By the way, I did also, and I'll show you that next and then I'll stop. We also do, oh, I should say one thing though. If I did steady state a, a sequence of steady states, we would've seen the optimum gain instead of 0.27% in the variable sweat, we would've found a gain of 1.5%. In other words, the steady state is extremely non reliable and a tax rate of 50%. So it's, it's making a big deal. The final thing is the Laffer curve. Hold on a sec. If we talk to the JCT, they're gonna be doing, you know, let's do the revenue maximizing rates which don't necessarily align with the welfare rates, but that's what they can do. So here, if we look at the change in present value of the, or just the Labrador, the number, I believe it's so flat, I forget even what the number was, maybe 35% can't even tell. And roughly 76% for the fixed sweat. So again, similar kind of result, although a slightly different, no, you know, higher numbers than what we see in the welfare case. And if we did the business tax revenue, oh and by the way, you know, a lot of revenue, here's total revenue, a lot of revenue gained in from these inelastic guys. But if there's any elasticity at all, not much. Okay. So we're trying to contrast the literature, which has really kind of focused on the red line and the blue, which is allowing for this endogenous sweat. So that's kind of the punchline. And the next step is to bring in the collateral constraints to start thinking about the financing part in how that that force, which is there could be some misallocation of who's running the business. How important is that? This was all revenue. So

- The guys who leave the small business are

- Now, yeah, that was total revenue. All the revenue. Now these rates don't differ a whole lot from Merley style. Your low, your lower rates don't differ a whole lot.

- Well I've never seen a dynamically fully matched to the IRS data melees. Love to see that. Love to see that.

- So Ellen, you're holding constant, the o the other taxes. Right

- On. Yes. I was holding constant the other

- Time. Yeah. So wouldn't you want to charge optimal tax? Wouldn't you want to do all three?

- My changes here is the transfers.

- Yeah, that - How the budget added up. Thank you. Alright.