Toward Fundamental Tax Reform

June 6, 2005

Unedited transcript prepared from a tape recording.

11:45 a.m.	Registration
Noon	Luncheon
12:30 p.m.	Keynote Speaker:	Douglas Holtz-Eakin, Congressional Budget Office
1:00	Panel I
	Presenters:	Michael J. Graetz, Yale Law School
		Robert Hall, Hoover Institution
	Moderator:	Alan J. Auerbach, University of California–Berkeley
1:45	Panel II
	Presenters:	Casey B. Mulligan, University of Chicago
		Kevin A. Hassett, AEI
	Moderator:	Daniel Shaviro, AEI and New York University Law School
2:30	Panel III
	Presenters:	William Gale, Brookings Institution
		Ronald Pearlman, Georgetown University Law Center
	Moderator:	Kevin A. Hassett, AEI
3:15	Adjournment

Proceedings:
MR. HASSETT:  Can I have your attention, please?  Attention, please.  I'd like to welcome everybody to our conference here today.  It's amazing how the clinking glass works better than any word.  Thanks, Max.

So I'd like to welcome everybody here to the American Enterprise Institute.  I'm Kevin Hassett, Director of Economic Policy Studies, and we're here today to discuss the new book that AEI just released that's edited by myself and Alan Auerbach called Toward Fundamental Tax Reform.

We're going to talk a lot more about the contents of the book in a bit, but without further ado I'd like to introduce our luncheon speaker. Douglas Holtz-Eakin is the Director of the Congressional Budget Office.  He's well-known to all of the tax geeks assembled in this room, and I've asked Doug to come here and talk to us about kind of the mechanics of fundamental tax reform from the CBO perspective.

And so there may or may not be a lot of proposals coming into the political procedure in the next few months, and exactly how will they be discussed, what kind of information will we be able to acquire from our professional staff about what might happen or not if the tax proposal is adopted?

Doug is going to provide a brief overview of this information for us for about 20 minutes and then we're going to have about 10 minutes for questions and then we're going to move directly to starting our first panel, which will be moderated by my friend and colleague, Alan Auerbach and he will at that point talk more about the book as well.  But now Douglas Holtz-Eakin.

Doug.

MR. HOLTZ-EAKIN:  Well, thank you.  It's great to be here and I want to thank Kevin for the invitation to come and talk.  I take it that despite the ostensible instructions Kevin gave me, my real job is to praise this event and this particular book.

The good news is I'm happy to do that.  Kevin was nice enough to share an electronic copy of this a while back, and I took the liberty of actually reading that and this is, I think, a wonderful addition to the collection of works on tax reform.  And I commend all the authors for their contributions and I think it's going to be a wonderful event that they put on today.

Now, Kevin has given me a little bit of cover on my remarks because I was trying to figure out what to say today and I ran into the following problem.  Traditionally, there are two kinds of talks on tax reform, at least in my eyes.

Type number one is the one that says what's wrong with the current tax system and why we need tax reform, and it goes through a litany of the failings of the current tax code, talks about the inefficiencies that it generates, describes in gory detail the fact that despite the fact that we have two individual income taxes, a regular and an alternative, and two corporation income taxes, we can't seem to collect the revenue we'd like in an efficient fashion.  The compliance costs strike everyone as a bit exorbitant.  The administrative challenge is a bit overwhelming and no one is happy with the fairness of it.  And so talk number one just sort of goes through Mao's long march through the failings of the current tax code.

Talk number two is essentially a chance to promote a specific way to fix these problems and to talk about your plan for tax reform.  My personal plan is to tax the people in this room and evade it, but I'm not going to be allowed to talk about that.

Or you talk a little bit about how to get there.  Should you have a big tax reform, one time, like 1986, or is there a sequence of modest legislative steps that cumulatively would yield you with a tax code that's way better?

Well, given those two alternatives, I realize that as the CBO Director I'm not allowed to promote a specific tax plan, so talk number two is off the table.  And given what I believe to be what Thomas Jefferson held as "We hold these truths to be self-evident," I shouldn't have to give the first talk.

And so that left me in a quandary.  I've got a big room full of all sorts of smart people in a prestigious location and a topic that I really like, which is tax policy, and I have no material.  So the solution that Kevin figured out for me was to drive straight into the weeds and talk a little bit about doing tax reform, talk a bit about the mechanics of it and the capacity to analyze reforms.

I wanted to talk a little bit about how ax reform does not exist in isolation; that, in fact, there's a complexity to tax reform that comes from the fact that its tentacles extend more broadly into the policy and legislative arena.  You can think of this as the collateral damage of the current tax code or the kinds of things you have to think about going forward.  And then I wanted to close with some challenges to the group for things that you have to figure out before we can actually get this done.

So let me talk a little bit about doing tax reform and some of the issues that will arise, and the first I want to discuss is really the capacity to analyze reforms.  And I hope this is a good-news story because I believe it to be the case that at the Congressional Budget Office, at the Joint Committee on Taxation and in the Department of the Treasury, as well as in a variety of think tanks and other research organizations, our capacity to understand the implications of tax reforms is much greater than it has been at comparable points in the past.

You can think of at least three, and there are probably more, kinds of formal modeling attempts on which there have been great advances in recent years and which would come into play in doing an analysis of tax reform.

The first would be growth models, and the broad macroeconomic implications for the long-term growth path of the U.S. economy of having a better set of tax incentives and a different fiscal policy more generally.

The second would be business cycle models and our ability to understand the near-term consequences of shifting from one tax code to another, perhaps in the form of a large reform, and understanding some of the transition issues which in the end are so important to thinking about a tax reform, how do we get from here to there, and what are the economic implications.

And the third would be the distributional analysis of a tax reform and the capacity to understand who pays in a real economic sense the burden of the current tax code, and as a result what would be the new set of burdens distributed across households of different incomes, households of different consumption, households of different ages, as we undertook a tax reform.

And in those three areas, I think there's been a tremendous amount of progress, and part of that progress has come from the fact that many of these groups are now doing this on a regular basis.  In my time at CBO, for example, the CBO has now undertaken each year a macroeconomic analysis of the President's budgetary proposals.

This is often labeled a dynamic scoring of the President's budget, but it is at some level an annual exercise in understanding the implications of a policy reform, and those implications can be drawn out from all three models.

We have looked--for example, when the President proposed the elimination of the double tax on dividends, part of the President's budget included that proposal and the CBO had to come to terms with how to analyze the implications of that dividend proposal.  What would be the long-term growth consequences?  What would happen to national saving?  What would happen to wealth accumulation?  Where would that wealth accumulation be located in the form of physical capital accumulation in the United States and elsewhere?  And what would, as a result, the growth consequences for the United States?

Now, we certainly didn't come to terms with it in any really narrowing sense.  We ended up, I think, putting out something like 12 different scenarios, giving combinations of different kinds of models, models which involved very little foresight, a traditional    ?     growth model in which agents in a myopic fashion accumulate their savings and develop wealth for the future, models with an overlapping generation structure where the foresight is limited to their lifetime and no further, and then some other models which had infinite horizons, Ramsey [ph] growth models of those kinds of households that have entirely too precise a vision of their future over the infinite horizon.

Given the combination of models, the ability to analyze them in an open- and closed-economy context and a variety of other variations one might want to look at, we provided no firm conclusions, but I do think we provided the Congress with some guidance as to how to sketch out the boundaries of what that kind of proposal would look like.

Now, it's also the case that we did not restrict ourselves to that particular piece of analysis in isolation.  It was part of a larger set of budgetary proposals.  Another proposal that year which might rings some bells was something which became the Medicare Modernization Act, $400 billion on prescription drugs for seniors.  That also had growth consequences, and it was important to recognize that there are competing impacts from all different kinds of policies.

We didn't pursue, but we could have, distributional analyses on the tax code.  There's a long tradition at CBO of computing effective tax rates, the actual economic incidence of taxes by income class, and those distributional analyses are presumably central to any tax reform as well.

But I think the important good news is that at CBO and elsewhere, there's been continual practice at this kind of policy analysis.  It's the kind of practice that keeps the models continually improving as we interact with the research community and understand things better.

It allows us to calibrate those models more closely to the current state of the economy and to the actual policies that are in place elsewhere in the budget.  And because tax reform will not occur in isolation, that's central.

And there are also some lessons that come out of that.  Lesson number one is there are still some enormously tough conceptual issues which I will leave it to the authors today to solve for us.  For example, with dividend taxes, it's still not obvious what will be the marginal investment incentives from relieving the double taxation of dividend income.  And since that's central to the analysis, this is a bit troubling.  I expect that's solved by five.

It's still true that these models are too primitive.  They're useful guides.  They capture some of the important and central economic relationships.  They are often devoid of important characteristics like financial sectors that will be central to any tax reform, and as a result don't give us all the guidance that we might like to have.

And a third lesson that has come out of this is that expectations about the future are crucial to understanding the impact of tax reform not just in the sense of what will individuals expect about the new tax code, but what do they expect about the current tax code and what are the expectations that are resident out there in the private sector about the evolution of the current tax code in the United States.

And on that note, it's useful to think about how the tax code is related to some other issues.  We were asked to testify a little while back in front of the House Budget Committee and talk about, you know, what is the role of the tax code in budget projections at the CBO.  And the short answer is the tax code is a nightmare from the point of view of making baseline budget projections, and it's a nightmare from at least three perspectives.

Perspective number one is the current statutory tax code, which has a whole series of phase-ins and phase-outs and sunsets, makes it very difficult to project receipts under current law, which is the bread and butter of what the CBO does.

It's difficult, number one, because you have to decide how people will react to the kinds of knife-edge cliffs that the tax code presents them as you go forward in time between now and 2011.  How will they respond to a sharp increase in taxes under current law when everything sunsets?  That's not obvious.

It's not obvious because we don't really know what people expect about the durability of the current tax provisions.  Have they decided, yes, this is it, we believe the law as written, we're going to plan accordingly, including the estate tax provisions?  Or do they have some deep belief that all of this is essentially up for grabs every year and so they make no particular long-run bet?  Or more likely, where are they in between about the durability of the tax code on a provision-by-provision basis?  Will we get marginal rates going up or down, or are they going to stay about the same?  What will happen to child credits?  What will happen to the estate tax?  It's not obvious, and as a result it's difficult to get the timing right from the point of view of projecting baseline receipts.

It's also complicated because, in principle, you would like to build your model of baseline receipts on top of an economic projection that's internally consistent.  But to do that, you have to first understand how you're going to handle those same expectations.

I mean, one possibility is to sort of go out and figure out what the private sector expects, make a projection of the economy based on that, and then layer on top of it a different set of rules for the tax code, which is the law as it evolves year by year.  That would be internally inconsistent, but might give you the best economic forecast.

Alternatively, and what we've tried to do is we have taken the tax code at face value and simply pretended that the private sector believes that the tax code will evolve as written under current law, built an economic projection where that assumption influences how people behave.

We see a sharp decline in labor supply with the rise in tax rates in 2011, for example, and that's internally consistent, but unlikely to be the best projection that one would make in any unconditional sense.

So the sunsets and phase-outs under current law are an important interaction between the tax code and the larger budget process.  Any reform would also have implications for the budget process in the ability to make those projections and in the transition to the new tax code.

Making good budget forecasts will be harder during the transition, and to the extent that the new tax code is stable and is levied on a base that is less volatile, it will make the revenue projection process more easy.  To the extent that we rely more heavily on the taxation of capital gains, a volatile source of income, it'll be harder.  To the extent that bonus income remains in the tax base and it remains as volatile as it has been over the past couple years, the problem will remain unchanged.

But the tax code will interact and any tax reform will interact with the larger budget process and the way that, as a result, the baseline is constructed and the scoring that will prevail for the entire policy objective.

It also affects other policy goals.  It doesn't take too much time to start thinking of places where tax policy and other policy objectives start running into each other.  The exclusion of the taxation for employer-sponsored health insurance jumps right out as a piece of tax policy that should be in the consumption base or in an income base.  But health policy objectives have kept that out.  How will that conflict be adjudicated?

Pension policies, the same thing.  Your view of the appropriate tax incentives for funding either defined benefit pension plans, defined contribution pension plans, will perhaps dictate how you feel about the structure of any tax reform going forward.  And those kinds of interactions, I think, are very important.

There are other places where I think it's less obvious, but equally important that this will show up.  For example, one could imagine the President's health insurance tax credit as embodied in his proposals got adopted into current law.  That would lower taxes and that would, by definition, contribute to the benchmark which would constitute a revenue-neutral tax reform.

However, you could equally well make the case that this is the kind of budgetary activity which could be put on the spending side as a voucher to purchase health insurance, and as a result would not lower revenues, but instead increase spending.

If you did the budgetary treatment that way, you'd have a different benchmark for what constituted a revenue-neutral tax reform.  There will be interactions in both directions.  What is the impact of tax reform on budgetary treatments?  And budgetary treatment will have a lot of impact on what constitutes revenue neutrality and what is or is not in tax reform.

More generally, the breadth of tax reform raises lots of implications for how you think about the budget.  A tax reform that was broad enough to include not just the current income taxes, but perhaps the payroll taxes or gas taxes, would have implications for the kinds of trust funds--and I want to use that term extremely loosely--that are currently in the budget.

Gas taxes are an integral part of an accounting mechanism that dictates how much goes on in the way of highway spending in the budget.  Altering the use of that as a revenue source would alter the way that business gets done and have big implications for such easy-to-pass legislation as highway bills.  And I don't have to raise your imaginations on what would happen if Social Security got rolled into any tax reform proposal in thinking about interactions with other policy objectives.

So from the point of view of thinking about tax reform, it has become pretty clear that there's a lot of work that needs to be done to support any kind of tax reform.  That work is conventional policy analysis, and there are now, I think, far superior, although far from perfect modeling capabilities housed within both the Congressional and the Executive Branches that can contribute to the understanding of the implications of tax reform for the economy.

It is also the case that there is, I think, a better recognition, but I think a far less well thought-out plan for looking at how tax reform interacts with other aspects of the policy and budget process.  I've sketched a couple of those.  I think it's very important to think hard about what expectations would be like.

I think it's important to note that there will be interactions with the budget projections, and to the extent that increased volatility of budget projections, increased difficulty in making good projects is part of the transition to a new tax code, that should be recognized up front.  In choosing the tax code, you can influence the long-run difficulty of that task.  Having the tax code settle down strikes me as one of the key aspects of that.

And that brings me to my last point, which is I think it's fair that if I have to come and somehow think of something interesting to say on tax reform, when I'm no allowed to promote my favorite plan or bewail the awful state of the current tax code, then I get to return the favor by asking you some hard questions.

And I think the part of the tax reform discussion that needs to happen going forward is how can we talk about having a tax reform that doesn't come undone.  I think it's great that there's now the capacity to better analyze tax reforms.  I think it would be awful if this turned into lifetime employment for folks in this room.

To the extent that we are going to have tax reform, one of the great mysteries is how we can make that tax reform more permanent and what are the aspects of structuring a tax reform.  What are the aspects of filing.  Do you get a permanent reform if you get some people's returns computed automatically so they don't visibly see tax preferences that are entered into the code?

What are the aspects of compliance that allow the code to be more durable, and what are the structures of rates and bases that will leave the tax code less--and let's face it, it's unlikely to be completely immune to continuous change by legislators, and a result the kind of evolution we've seen between 1986 and now, 20 years later, a tax code that looks very different than what I think the dream was at the time of that last big tax reform?

So that, I think, is an important topic.  It's a topic that ranks, in my view, as comparable in importance between picking between income and consumption bases, comparable in importance to trying to get the distribution of tax burdens correct.  Making sure that the code itself is something that is a more stable element of the policy process will improve not just tax policy, but it will improve the remainder of the budget and policy process, as well, and I think that would be something that it would be nice to get some insight into as well.

So I managed to fill 20 minutes by diving into the weeds on some things that could happen on tax reform, and as Kevin promised, I'd be happy to spend the last 10 answering any questions you might have.

Thank you.

[Applause.]

MR.          :  Steve?

MR.          :  Hi.  Steve Entenar [ph].  Doug, thanks for that on the modeling improvements we've had.  Isn't there a little bit of interaction among the modeling efforts in the following sense: If you really want to know the distribution of the tax burden or the gains and losers from a tax change, you do have to know the macro consequences because in some sense the incidence involves the change in incomes pre-tax, as well as what you end up having to pay?  Is that effort to integrate these models in that sense very far along?

MR. HOLTZ-EAKIN:  It's either very far along or stalled, depending on how you want to look at it.  If you accept--for example, imagine the question what is the incidence of the corporation income tax or tax on capital.  There's a reasonably good question.  If you accept a set of assumptions about the incidence, in which case let's say it falls entirely on capital, then there's an internally consistent set of macroeconomic assumptions--the savings incentives, investment incentives, distributional consequences.  Tremendous progress has been made.

If you don't think that's right, then, you know, we're still at ground zero.  We need to resolve who pays the tax and go back and redo the modeling in a second consistent fashion.  So that's in the eye of the beholder, quite frankly.

There are some questions that have remained for a long, long time.  You know, I started getting my Ph.D. in economics in 1980, and David Bradford, God bless him, was instrumental in my education on tax policy and he told me that in my lifetime I would never understand who paid the corporate income tax, nor would we get rid of the double tax on dividends.  And in my lifetime, I got to hear the President say we'll get rid of the double tax on dividends, but I'm sure I won't find out about capital taxation.

So I don't know if that's an answer or just--it tells you the lay of the land.  I mean, these things are hard.  I think it's entirely possible to get the models to marry up in a sensible fashion, but I can't tell you the answer.  And I actually think it's more useful to make sure that policymakers are cognizant of the key dimensions.

You know, they have the capacity to worry about everything.  So let's find the key assumptions, the ones that do have profound implications for distributional burdens, say, or growth consequences and highlight them so that they can go look past the models for what they'd like to do, because certainly you don't want to have models driving the policy.  You want it to inform it.

MR.          :  Jim Klumpner [ph], Senate Budget.  Doug, I very much appreciated your comments at the end about the durability of any tax code changes.  And I was wondering if CBO or anyone else had done estimates of the economic costs of keeping people constantly on the steep part of the learning curve.

MR. HOLTZ-EAKIN:  There's probably a long version of this.  No, we haven't.  I think there are papers in the research literature on this issue of code uncertainty and volatility of marginal tax rates, things like that.  We haven't done anything particular on that.

I think it's--I just happen--my instinct is that this is a really important issue that, given the comparative allocation of research energy, has been underserved; that the political economy of tax reform is just as important as the pure economics of tax reform.  And thinking about that strikes me as a good place to go.

MR. HASSETT:  Kevin Hassett.  Doug, it seems like if we're going to be trying to study the impact of tax reform on the economy and then provide papers that indicate what might happen if you do this or that that we also need to have a mechanism for following up and seeing if that paper we wrote three years ago was correct or not.  And I seem to recall that you guys did something on EGTRA [ph], not when you were there.  I wonder, has anyone ever gone back to see where it got things right and where it got things wrong, and so on?

MR. HOLTZ-EAKIN:  Well, there's no real need, Kevin.

[Laughter.]

MR. HOLTZ-EAKIN:  There has not been a particular follow-up to anything of that type.  There's certainly the case that you can go back and try to decompose the gap between any particular projection and what evolved.  That's standard business at CBO.  That's an often frustrating exercise because the order in which you do the compositions often matters and there are so many things going on, you don't get to isolate it too much.

But I do think that it's an important exercise.  I think it's crucial to doing work in the future to go back and always check to the extent you can were the misses due to unforeseen economic events, were they due to legislative changes, or were they do to aspects of the modeling, you know, hidden under the label "technicals" that were central.

And to the extent that you can find technicals dominating something, understanding why is important as well.  And that's true not just in this area, but many.  And so, yes, we've done a lot of work in that area.  But, no, we haven't looked at the particulars of the 2001 tax cut or 2003, for that matter.

MR.          :  Bill    ?    with American Enterprise Magazine.  If the '86 tax reform is taken as a model of something of an academic ideal and things that happened since then were a form of this back-sliding--as you said, you know, how can you keep things from becoming undone--what kind of mechanisms would you put in place to prevent things from becoming undone?

MR. HOLTZ-EAKIN:  Well, I think I wanted you to answer that question, so--

[Laughter.]

MR. HOLTZ-EAKIN:  But here's what I had in mind.

MR.          :  That's why I asked.

MR. HOLTZ-EAKIN:  You know, one possibility is you could view '86 as the perfect reform.  Not everyone shares that view.  That's for sure, and so it could be that '86 came undone because it was the wrong tax code.  We moved toward something that was very close--something that was an attempt to get as close as the political process could allow to a Haig-Simons [ph] income tax of the type that scholars in the 1950s espoused as the right tax base.

And that might have been just the wrong tax code for the wrong time.  I have a lot of sympathy for that view, and so it unraveled because it was the wrong reform.

Alternatively, I've heard people make a very disciplined case that, you know, it was the right reform and it unraveled because its big-fix nature made it very visible.  That was by design, so that the strategy of gore everyone's ox also means that you build everyone's grudge.  And what follows thereafter is a steady attempt to get my piece undone because you put it on the radar screen.  How do you insulate it from that?

Or perhaps it's the visibility of the tax code, which has become far more important in other aspects of policy--you know, health, as I mentioned, whatever; you know, the impact of the tax code as a policy instrument, the more heavy use of credits as an instrument of achieving policy objectives.  Maybe that's what does it.  So how do you wall it off from other pieces of policymaking and is it desirable to?

Those are the issues.  I don't have a concrete list, but I do think it's something we need to think hard about.

MR.          :  Howard Katz [ph], Cleveland Marshall Law School.  There are those who would argue in the area of reapportionment that the power of the computer and the increased knowledge it gives us has altered the process.

Is there a parallel in tax reform?  Are the winners and losers even more knowledgeable about the fact that there are winners and losers, and does that therefore make '86-type reforms even more difficult?

MR. HOLTZ-EAKIN:  I guess because I cling to the pollyannish belief that better-informed public policies are better public policies that I'd hesitate to go there.  I think what I believe, at least at this moment, is that the winners and losers have always been willing to identify themselves.

What we now have is a better capacity to adjudicate their claims and let the policymakers decide whether these guys who claim they're getting hammered by this reform really are or are not, these guys who others claim are unfairly benefitting really are or are not.  And I don't think as a result you're going to get more claimants showing up, more aggrieved parties on one side or the other.  I think you'll simply have better information about what you want to think about those claims in the policy process and be able to weigh them against one another in a more meaningful fashion, on a standardized set of estimates that have comparability in whatever those flaws may be.  But they'll be comparable and you can decide going from this candidate reform to this one what are the implications for the distribution, and does that appeal to you as a policymaker in terms of making your judgment.

So I don't think the total number of people showing up with claims is going to change much.  I don't think the process is going to be all that dramatically different.  You know, we might get showdown at Gutchie [ph] Gulch 2.  We'll see, but I think the kind of information that undergirds that and is in the hands of the policymakers will be different, and that could affect the dynamics going forward.

We're done?  Thank you.

[Applause.]

MR. HASSETT:  [IN PROGRESS] for an excellent talk, and I don't think those were weeds at all.

And so we'll now begin right away with the first panel.  I'd ask my colleague, Alan Auerbach, to come up.  He'll be the moderator.  Mike Graetz should be arriving any moment.  He got stuck behind a truck crash on I-95 on his way to the airport, but is apparently on a flight that landed half an hour ago.  So, hopefully, he'll be here in time for his remarks.  If not, Bill Gale has kindly agreed to bring his third panel performance into the first panel, if need be.  But what we're going to try to do is proceed with the first speaker and then see if Mike gets here on time in order to join the first panel.

Alan Auerbach is the Robert D. Birch [ph] Professor of Economics and Law, and Director of the Birch Center for Tax Policy and Public Finance, and former Chairman of the Economics Department at the University of California-Berkeley.  I think Alan has a long and distinguished career in tax policy, and this book that we're celebrating today would not have been possible without all of his hard work.

Alan.

MR. AUERBACH:  Thanks very much, Kevin.

Several months ago, Kevin called me up and said, you know, with the tax reform panel being appointed and President Bush showing an interest in tax reform, wouldn't it be a good idea to gather together people with different perspectives on tax reform to write succinct contributions with the focus of their own choice to give us an idea--give people not that familiar with the process access to what current thought is on tax reform?

And the extent of our input was choosing the authors and then letting them run from there, and we're very happy with the outcome.  The authors, nine in total, have different perspectives on tax reform.  Some, such as the two people on this panel, have very specific proposals that they've worked on over the years and written about quite extensively.

Others are very experienced in the tax policy process itself and have wisdom to provide regarding how the tax reform process might play out, what we should keep in mind as we try to adopt or think about adopting different proposals.  And still others are less involved in the tax policy process and are not pushing a particular well-fleshed-out proposal, but have insights to share based on their academic research.

And taken together, I think we feel that the papers in the book, which is available on the AEI website and also from your favorite bookseller, are very helpful not only for people less familiar with the tax policy process, but frankly for anyone, including those who are very familiar with what has been written over the years, to bring together in one place the current thinking on what tax reform might look like and what it should look like.

So without further delay, let me turn the floor over to Bob Hall, who has been an important contributor to thinking about tax reform for many years now, and ask him to give you a summary of his paper.

MR. HALL:  Thank you, Alan.  It's a pleasure to be here.  I guess the first thing I want to say is that I'm not involved in the tax reform process.  That takes place in this city exclusively, and those of us who live 2,600 miles away necessarily have a rather different role.

And my role now, and always, has been to talk about some structural ideas that are presented in the form of a very strict, politically utterly impractical set of ideas.  But they nonetheless provide a framework for thinking about tax reform that I believe is helpful in the tax reform process, and I think we've actually made some progress.

For example, the reduction in personal taxation of business income, the reduction in the dividend tax rate and the capital gains rate that took place in '03, is definitely a step in the direction that I believe we should move.  So, again, I'm sort of the utopian, and therefore not involved in the process.

So people always ask me, well, you know, what are the chances that we'll achieve the flat tax?  You know, to me, that's not a question I'm remotely capable of answering.  I have no domain knowledge that would help answer that question.  That knowledge resides here in our Nation's Capital and not in my head.

But what I do understand is some things I learned a long, long time ago about how to execute a consumption tax most efficiently and how to achieve simplicity and practicality in creating a consumption tax.

I take for granted our belief in the benefits of a consumption tax.  I'm not somebody whose beliefs about how the economy operates says that a consumption is a miracle, in contrast to an income tax.  I could probably live with an income tax, but I see the benefits.

The reduction in the tax wedges as they accumulate forward in time that exist in a consumption relative to an income tax, I think, is impressive.  And so I've always been part of the consumption tax.  I've been a consumption tax guy since I was 22 years old, when I first took graduate public finance.  And it has been gratifying to see--I didn't know and I wasn't taught all the deep reasons why it's a good idea, but it has been gratifying to see that happen.

Okay, so I'm going to take you through a little bit.  This is actually not emphasized in the paper, but I think it's probably a useful thing to be here to talk about how this system works.

It's basically a value-added tax.  The change--first of all, many of you have probably heard me say this before, but what's the history of the flat tax?  Alvin Robushka [ph], my co-author in this, co-conspirator, was involved as a political scientist in observing and even participating in the tax reform--or tax reaction movement of the 1970s.

And he came back from the field and said what the people want is the flat tax, and he said but I don't know quite what that is.  What is the flat tax?  So I said, oh, I know what that is; I learned that in graduate school.  And so I spent some months in the summer of 1981 sort of putting this together, and I had been taught the value-added tax is the way to implement a consumption tax.

But I also knew--and there was a lot of interest in this subject at that time in the 1960s in the question of how to make the tax system genuinely progressive, especially at the lower end.  And that's one thing it's hard to do, it seems, in a value-added tax framework where there's no personal part of the tax.

Okay, so the idea of this tax system is that there's an individual wage tax form, something not usually part of a value-added tax.  Now, subsequently, some countries have adopted systems.  Canada, in particular, when it implemented a national value-added tax, or what they call a sales tax, they actually did something that's kind of equivalent to this.  I think it's not suitable administratively for what the U.S. does, but there has now been a certain amount of field experience, I think generally favorable, in this area.

The basic idea is that you want to have an exemption, but if you create the exemption by some kind of a rebate that people can apply for independently of the tax they pay, there's a huge administrative problem in linking the two together.

So the idea here was to separate the value-added tax into two pieces, and the form one shown on the screen here is the personal piece and it applies to earnings, broadly conceived.  So that would be on line one wages and salary, and on line two pension and retirement benefits.  So line three gives total compensation, which is the personal tax base.

Now, the key to observe there is that it excludes any other form of income.  It's not an income tax.  It is, as it says at the top, a wage tax.  Now, of course, if you've studied the consumption tax, you know that a consumption tax is a one-time lump-sum tax, plus a wage tax.

So I'm going to talk about a business tax, but sort of the purest, simple modeling of this, a business tax has a present discounted value of zero, apart from its lump-sum element.  So the guts of any tax system--and, you know, even in income tax this is going to be true in terms of dollars--is going to be this compensation tax, and that's certainly the case for the flat tax.

All right, so the reason for this--the justification for it starts on line four, where you see the personal allowances.   And it turns out--I haven't tried to do detailed math--these are conservative in terms of revenue neutrality.  These were extrapolating numbers that were revenue-neutral before recent tax reductions.  So this is actually a revenue-gaining set of numbers at the 19-percent rate.

All right, so you get personal allowances which amount to a substantial $41,000 exemption for a family of four.  So this overcomes the single biggest deficit of the value-added tax, which is its inability to shield the lower-income families from paying any of the value-added tax, and then putting a progressive, rising fraction of the burden on higher and higher incomes.

Okay, so then the other piece of it is the business tax, which is just the other half.  If you say, well, how do you figure out what the business tax is?  Simple.  Take a value-added tax, back out the fact that individuals are paying a piece of that, the wage piece.  So in other words, that gives line 2b, which shows the deduction for wages and salaries which would not be present in a value-added tax--this, together with form one, together form a value-added tax.  So it's an air-tight, simple, administratively-proven consumption tax at a 19-percent rate, but it's progressive, unlike a value-added tax.

So, you know, we thought this was a pretty neat idea, and we published it in the Wall Street Journal and it created a little bit of a sensation at the end of 1981.  And, interestingly, it has really not changed.   The only thing that has changed have been the numbers.  The idea of progressive value-added tax was durable.  It has not been subject to fine-tuning, other than changing the numbers over time.

Okay, so what are its benefits?  Well, it's a consumption tax, so it provides the right incentive for capital formation because, going back for a second, like a value-added tax, you can see on line 2c that purchases of capital equipment and other capital are deductible.  So that makes it a consumption tax.  You take total GDP, subtract investment, you've got consumption, and that's just standard value-added tax principles.  So it's really simple.  It administrative benefits are proven.

Taxing business income at the source, which is, you know, the genius of a value-added tax, avoids more than a billion in Form 1099s and all the difficulties of administering an income tax which taxes business income at the destination.  So this idea of source taxation is central.

Of course, it's also the reason that it's politically a little hard to sell.  It looks like the rich are being excused.  They're pre-paying--they're receiving business income upon which the business tax has been pre-paid.  So it's as if there was a withholding tax.

You can say that 10,000 times even to an intelligent member of Congress and you get this blank stare at the end, and then the question is, well, why aren't the rich paying any tax on their business income or on their interest, capital gains and dividends?  So it's very hard to get that point across.  That's someone else's responsibility, not me.

Okay, so David Bradford, who is, you know, a huge contributor to this, pushed a very similar idea and a very good idea called the ex [ph] tax, same structure as the flat tax, but going back to the personal tax, there's more than one bracket.  And I've always felt--and Alvin doesn't let me say this in public, but I will here anyway--that there's some merit to that idea.

So you can make the tax more progressive, and probably in a desirable way.  For example--yes, here we go.  So the flat tax is the blue line, and so you can see that relative to an ex tax--and this is just an illustrative ex tax, the red line--you can see that there's a higher burden in the incomes in the $50, $60,000 range, and then correspondingly less burden above about $80,000, $85,000.  And that's illustrative of the difference in the distribution of the burden by earnings levels that could be achieved by adding just one more bracket.

In order to avoid some serious administrative problems, the top earnings bracket needs to be the same as the business tax rate.  And I see that as a constraint [TAPE CHANGE.]

So the example here is with a 25-percent business and upper personal tax rate.  So if you're concerned, as many people would be--and this is especially true, since the distribution of well-being in the U.S. since 1981 has dramatically widened, that a better distributional tax might be obtained through the ex tax, since as you can see from this picture, the flat tax was never flat.

So many people came to me in the old days and said, you know, I just figured out the flat tax is not flat; it actually has this angle in it.  It's that horizontal red line and common red and blue line, and plus the upward-sloping blue line, it's got an angle in it; it's not flat.  I said that's good; you understand it.

It's Robushka's idea to call it flat.  Of course, it was a genius name and I have to credit him for it, but it's not a very good description.  So the ex tax again is very close, and I'm very happy to embrace the concept because the administrative idea is very much the same.

Okay, so how could we get from here there?  I'm not master of how you do that politically.  I'm just thinking about what steps you could take in terms of sequential tax reform.  The first one is to push further in the direction that the President pushed through successfully in '03, and that is this system collects business taxes at the business level and does not further tax business at the personal level.  So all personal taxation of dividends and capital gains should be eliminated.  And, you know, we got pretty close to that in '03, so it seems like that's doable, push those rates to zero.

Then you get to the harder part of the business tax part of the story, which is interest.  To make this whole system work, you have to change something, and this is so, so important in terms of getting rid of leakages that exist in the tax system today.  You must remove the business deduction for interest and then remove the personal taxation of interest income.  That's by far the most important next step in tax reform.

Now, I don't want to suggest for a moment that that's easy, but that's what needs to be done.  That would take us most of the way.  We're getting into a very awkward place in tax reform because we haven't decided which way we're going.  We're giving personal incentives, more and more personal incentives for shielded income.  So creating a consumption tax on a cash flow basis by having qualified accounts, so if you save on those accounts you're excused from current taxation--so that's one way to get a consumption tax.

The other way is to give better incentives for investment to business.  If you do the two, through the interest deduction you get a subsidy to capital formation, which is as bad as excess taxation.  We're moving very strongly in the direction of debt-financed investment being subsidized, and that's a big mistake.  The way to get around that is to remove the deduction for interest at the business level and remove the taxation of interest at the personal level.  It would take care of the problem immediately.

Okay, so that's the big thing.  The rest of it is just some further fine-tuning at that point.  You want to simplify the personal tax, either adopt the AMT as the only tax or equally convert the income tax to a wage tax to phase other complications.

Also, the taxation of business income at the source pretty makes the whole idea of the AMT unnecessary, and so I think we can ditch the AMT at some point along this, and then we've pretty accomplished it, okay?

Thanks.

MR. AUERBACH:  Thank you very much, Bob.

Ready to go?

MR. GRAETZ:  I apologize for my late entry.  I had a four-hour train ride because the trucks on the highway on I-95 managed to collide and close the highway, but I'm here now.  Good.

So I want to talk about a proposal that I have been talking about for some time to change the tax system of the U.S.  I begin with the failings of the income tax, but these are extremely well-known, so I won't take much time with them.

Obviously, the tax code is too complicated.  The IRS has great difficulty administering the code and other programs within the code, so it can't do its job.  We all know about the AMT growth which will hit 30 million taxpayers if we don't change it.

And then the fourth bullet there is what I call chicken soup, which is that for any social problem--education, child care affordability, health insurance, retirement security, or long-term care financing--an income tax deduction or credit is the answer.  I'm fond of saying that the Congress uses the income tax the way my mother used chicken soup as a cure-all for everything.

Finally, the American public has become cynical about the income tax.  Individuals believe that the tax system is unfair, and young people complain particularly about the tax system and indicate that they have very little compunction about dishonestly completing their tax forms.

I believe as a result of this that the income tax cannot be fixed, as it was in 1986, without major structural revision.  And I also believe that keeping everyone in the system filing returns even if they start off only as a postcard will continue to create incentives for Congress to produce chicken soup.

The next slide is just my favorite slide, which is the number of thousands of words in the Internal Revenue Code and regulations.  The Internal Revenue Code itself is now at least four times longer than War and Peace, and considerably harder to understand.

Now, I want to turn to my proposal for making things better which is described in the book.  First, I would repeal the income tax, and I indicate here that instead of heeding the frequent calls to repeal the AMT, repeal the regular tax instead and modify the AMT.

The main modification is to increase the AMT exemption for married couples to $100,000 and to index the exemption to inflation as a way of keeping the income tax in check.   This means obviously that couples with less than $100,000 of income would pay no income tax.  I would lower the AMT rate or the income tax rate to a flat 25 percent, and you need to make some adjustments to the AMT to make sure that it allows deductions for income-producing activities.

The payoff from these changes is that you have 100 million fewer tax returns.  About 150 million people will file no returns.  For them, April 15th will be just another day.  The marriage penalty will be eliminated.  This retains deductions for home mortgage interest and charitable contributions, although obviously those could be modified.  And incentives for retirement savings and health insurance would be preserved.  This, I believe, is an important point, unless we come up with some alternative for the existing systems under the income tax.

The second piece is to clean up the corporate income tax.  Again, I would lower the corporate rate to 25 percent.  That would be the same rate that would apply to individuals and to flow-through entities.  I would more closely align book and tax accounting so that Congress would have to explicitly make any changes from book accounting for things such as depreciation, R and D, and foreign taxes.

I would consider at least replacing the foreign tax credit with an exemption for dividends paid.  This principally would be done to eliminate the barrier to repatriation of foreign earnings which we have under the existing system.

I argue that small business taxation ought to be greatly simplified probably through a single flow-through regime and much greater permission to   use cash accounting.  Book tax conformity will assist in the battle against tax shelters.  I believe this is actually the only solution to the tax shelter problem.

As a result of these two changes, what you see on this graph which I think is important is that the income tax would be about 4.1 percent of GDP, compared to the current level in the U.S. which is more than twice that of 9.2 percent, and its level of nearly 12 percent in 2002, much less than the European Union, which is 14 percent of GDP, or the OECD average which is 12.9 percent of GDP.  So the U.S. would be taking advantage of our status as a low-tax country much more so than we do now by having a much, much smaller income tax.

In order to make up the revenues that I've just lost, I would enact a value-added tax, a broad-based credit-method value-added tax in the range probably of 14 percent or so; 10 to 14 percent it says in the chart.  I would require that the amount of value-added tax be separately stated whenever goods and services are sold so that consumers are aware what they're paying in value-added tax.  So it's not a hidden tax as it is in some countries of the world.  I would exempt small businesses from collecting the VAT.  But, of course, they could elect into the system if they wanted to.

The payoff is that a value-added tax imposes compliance costs that are one-third to one-fourth as large as our income tax, according to Joel Slimdrod's [ph] estimates.  It is easily enforceable based on international experience and it minimizes economic distortions.

I believe the compliance costs may be even lower than that that Joel estimates.  A partner at Ernst and Young in Europe tells me that 5 percent of their partners work on value-added taxes, and that only 20 percent of the revenue personnel in Europe tend to work on value-added taxes, even though it collects around half of the revenue.

In the United States--and this is a common point that people are pretty well aware of by now--in the United States, the main distinction between our tax system and that of those used around the world is that we collect far less revenue from consumption taxes than other nations do, about 2.2 percent as a percentage of GDP, as compared to 6.6 percent in the OECD and 7.5 in Europe.

This value-added tax would essentially create a tax rate in the United States, combined with state sales taxes, of about 18.6 percent.  That would put us just between the OECD average and the EU average.  So essentially what we would be doing is raising our consumption taxes to be comparable with those around the world and lowering our income tax so that it would be far less than that elsewhere in the world.

The final piece of this idea is to provide a refundable payroll tax offset to replace the earned income tax credit and to protect low- and moderate-income taxpayers from increased tax burdens.  I believe that any movement away from an income tax to a consumption tax requires some adjustment to deal with the regressivity of a consumption tax, and this is how I propose to do it.

In the course of doing this, you eliminate the EITC's serious marriage penalties, and I would recognize the financial obligations of non-custodial parents to their children, while preserving the EITC's work incentives.  The IRS would simply tell employers what the offset was and this payroll tax offset would leave Social Security benefits unaffected.  Obviously, this would have to be larger if there are no value-added tax exemptions for food and other necessities, which I do not recommend.

The payoff from all of this is the number of returns that the IRS would have to deal with, in combination of both the individual and corporate income tax and the value-add tax, would be less than they were dealing with in 1946.  They would still have information returns and the like to deal with.  But this would shrink the income tax to the point where it would be possible for the IRS to do its job.

I think this proposal has a number of important advantages.  First is the elimination of the 100 million, of the 135 million tax returns.  This reduces compliance cost, it reduces administrative costs, and it will counteract the cynicism generated by the current system for middle-income taxpayers.

Secondly, the new tax system would encourage savings and investment in the United States.  It would eliminate the tax on savings for most Americans and reduce it for all Americans, thereby stimulating economic growth and creating additional jobs.  Unlike the current tax, this system eliminates all marriage penalties, which is something that the Congress has been unable to do under the current tax.

And this last point on this particular slide is one I want to stop and emphasize because I believe it's very important.  What I'm doing here is I am simply combining taxes that are commonly used throughout the world.  As a result, this system facilitates international coordination and fits well within existing international tax and trade agreements.

Most of the other proposals that have been advanced, including the flat tax and the ex tax and progressive consumption taxes and the like, will require a renegotiation at least of our income tax treaties and almost certainly of the GATT in terms of the way it deals with border adjustments.  This is going to be a very difficult and time-consuming process which is avoided by sticking with the tried and true.

This plan avoids the difficult issues of transition to an entirely new system that have haunted many other proposals that would eliminate the income tax completely.  As I've said, it would reduce the IRS workload so that it can do its job.  And with very few Americans filing tax returns, there will be much less temptation for political tinkering with the tax system; that is, it is important to realize that the provisions that have so complicated the code--the various education incentives and the like, to take one example--these are not special interest provisions; these are general interest provisions.  And by getting the vast majority of people out of the system, Congress would not have the same incentives to complicate the law.

With international capital mobility, starting with a relatively high level of value-added tax--and I have recommended a super-majority voting requirement for increasing tax rates--that  should reduces Congress's ability to reduce rates.

I have one minute.  There is a table there that cannot even be mentioned in one minute.  It simply shows that this proposal is estimated to be revenue-neutral, which is not the case for some of the proposals that have been advanced, most notably the so-called fair tax.

Let me just conclude.  I have a number of other slides here, but I will conclude with this one, which is to say that the President in his executive order to the tax reform panel insisted that one of their recommendations be reform of the existing income tax.

And I just want to emphasize that this is a proposal that can be combined with any reform of the income tax; that is, you can reform the income tax however you like.  And I have not emphasized--in trying to emphasize this structural set of issues, I have not emphasized all the changes that one might want to make to the income tax.

But after you've reformed the income tax, you can--by putting in place a 10- to 14-percent valued-added tax, you can still exempt income up to $100,000 and lower the income tax rates by about 10 percentage points, so that if, for example, the President's commission were to come up with a base-broadening proposal that lowered rates, you can lower them still further by enacting a value-added tax to replace most of the income tax.

This system, which its detractors like to describe as French, is remarkably American.  It is, in fact, the system that the United States had prior to the Second World War, which was a low-rate income tax on high-income Americans, combined with consumption taxes on everyone else.  At that time, the consumption tax was in the form of tariffs, which is a particularly economically inefficient form of consumption taxation.  A modern value-added tax, as we know, is a far more equitable and efficient tax, and so this is why this proposal is one to take us back to the future.

Thank you.

MR. AUERBACH:  Thank you, Michael.

We have time for a few questions.  If you'll raise your hand, I'll recognize you, and let me just ask if you wait for the microphone.  The first one is back there.  Please identify yourself, make your questions brief and indicate to which of the speakers the question is addressed.

MR.          :  My name is Dan Goldberg [ph], University of Maryland Law School.  A question to Robert Hall and to Michael, if I could; it's a related question.

Why do you use a subtraction method VAT instead of a credit invoice VAT in your flat tax?  And, Michael, I'd like to ask why do you leave the corporate income tax and business tax intact as long as you're suggesting a credit invoice VAT on the business level?

MR. HALL:  Okay.  Well, first of all, there's no particular reason, in principle.  I spend most of my time as a macroeconomist and I'm extremely conscious of the transitional macro costs of switching to a standard VAT because that transition requires that the price level rise by the amount of the tax rate, which would not be true under the flat tax.  So to my mind that, tilts it.

Now, of course, that means fighting every member of Congress on the border tax adjustment issue, which is a fight that we would inevitably lose.  So I could see it coming out the other way.  But, you know, Britain went through an extended period of macroeconomic grief after it adopted a standard VAT in 1979, and it was entirely related to the fact that it had to absorb an increase in the price level of something like 20 percent.

A 20-percent increase in the price level over a few years in an environment like we have today of 2-percent normal inflation is a huge macroeconomic hit and it's the one I think that has been--its importance has been underestimated by tax people who don't dabble in macro as I do.

MR.          :  If I could make my question a little more precise, under a credit method VAT you would view labor as another input, so that there would be a VAT credit for the cost of labor, the cost of employees, and the employees would then have to pay tax on their salaries, so that you wouldn't have the price effect.

MR. HALL:  What you've described is exactly what the flat tax does.

MR.          :  I know, exactly, except it's on the business level.  Instead of using a subtraction VAT, you'd use a credit invoice VAT, and I want to know why you've chosen not to use a credit invoice VAT, but instead have gone with a subtraction VAT.

MR. HALL:  Well, as you know, I'm only an amateur, probably, finance guy.  I'm a professional macro economist.  I thought the only difference was that the credit invoice dealt with the differences in tax rates, but otherwise they're perfectly equivalent.  If you have a uniform rate, they're perfectly equivalent, so I wouldn't see any difference.  It's a detail.  Is it anything more than a detail?

MR. AUERBACH:  Move on to--

MR. GRAETZ:  Well, I'll move on to my question, but let me just say one thing about this.  Charlie McClure's [ph] testimony before the President's tax reform panel, which is available on the Web, is very informative on this issue.  And at least according to Charlie, who I think has studied this more than anyone, it's not just a detail, and he regards the subtraction method VAT as having major disadvantages over a credit method.  But I'll leave that.  That was not a question to me.

The other question was why do you leave the corporate income tax in place, to which the answer is I have not eliminated taxes on capital income at the individual level.  And my concern is that if you do not eliminate taxes on capital income at the individual level, but you don't have a corporate income tax, then the corporation simply becomes a device for sheltering all capital income and so it'll end up on the corporate side.  So I don't think you can eliminate it completely.

MR. AUERBACH:  The next question is down here.

MR.          :  Will Almatruda [ph], Catholic U. Law School.  Neither of the speakers has said anything about estate or gift tax.  Could you remedy that for our benefit?

MR. HALL:  Well, in a consumption tax environment, there's no reason to distinguish how consumption is financed, so I would have zero gift and estate tax.

MR. GRAETZ:  Well, I particularly welcome this question, having just written a book which I want to now recommend called Death by A Thousand Cuts: The Fight Over Taxing Inherited Wealth, available on Amazon for $19.97.

The estate and gift tax is an issue that I think under either of these systems can be considered separately.  I would actually, given my druthers, go to taxing the recipient of inherited wealth rather than the decedent, and you could do that simply by including large bequests in income under the simplified income tax that is described.

And I suggested that in a version of this that I did in the Yale Law Journal, but it's not an essential part of the proposal.  You could have--if the Congress decides to repeal the tax, which, having studied the politics of it, I believe they're more likely than not to do, you could certainly have the system without an estate and gift--without a tax on inherited wealth.

MR.          :  Hank Goatman [ph].  Mike, why not use this as an opportunity to advance either the CBIT [ph] or some other form of integration so that we get all business income taxed the same way, no matter what form it takes place in?

MR. GRAETZ:  I would--you know, as you know, I have been associated with the CBIT proposal in an earlier time.   This is a proposal that was advanced--for those of you who are not into this lingo, this was a proposal that was advanced by the Treasury Department in a 1992 study on corporate integration which proposed that we first have an exclusion for dividends at the individual level to eliminate the double tax on dividends, and then would deny the interest deduction at the corporate level and eliminate the tax on interest at the individual level.

I've always thought CBIT was a good idea, and I agree with what I caught of Bob's comments as I came in about the difficulties of having the interest deduction continue.  But, again, the reason is--the reason I don't do it is that I really felt when we were at the Treasury writing that study that you could only eliminate the interest deduction if you did it in an international cooperative setting; that is, with other countries.

We would be the only country in the world to tax interest at source under a CBIT kind of proposal, because we would be denying the corporations a deduction even for debt which is supplied by foreigners.  And this is a huge change in the existing world regime of taxation and I simply don't believe it's the kind of change the U.S. can do alone.

And so while I regard the dividend exclusion as a step toward CBIT, I think you've got to take that step along with other industrialized countries, and that work hasn't been done yet.

MR. AUERBACH:  I think we have time for one more question.

Dan?

MR. SHAVIRO:  Dan Shaviro, of NYU Law School, and I have a question for Michael, which is both Bob and you have essentially VATs, but Bob has the level of a deduction for wages which are included by the workers, and this permits him to have a zero bracket.

He had showed an example while, I guess, you were still struggling with the trucks that had a zero bracket for a couple with no children of $27,000.  So for a 14-percent VAT, that would be a tax saving of $3,780 relative to their having to pay that.  That seems like a pretty big difference.

I'd also note that although Bob doesn't have payroll tax offsets, as you do, in his plan, it would certainly be reconcilable with it.  So anything you wanted to do on the payroll tax refundable credits, he could do, too.

My question is how would you really justify, for example, to these people who get a zero bracket worth $3,780 that you're not giving them that and instead you're giving them no need to file a fairly simple wage return?

MR. GRAETZ:  Two things.  One is the refundable credit that I've proposed--and if I went back to the chart, the refundable credit that I've proposed is designed not only to replace the EITC, but to offset the regressivity of the value-added tax.

So if you notice, on number two the current EITC is about $36 billion a year, and I've devoted what I estimated to be about 2 percent of the VAT as a way of essentially substituting for the zero percent bracket.  So I would assume that a wage-earner would get a credit of the $3,780, or whatever the right number is, to compensate for the VAT at the low-income levels, and it would be phased out as income increases.

So the idea of the payroll tax offset is to do two things.  One is to replace the EITC and the other is to mitigate against the regressivity that you would have with a value-added tax without the kind of zero bracket that the ex tax or the flat tax has.

You know, my problem with the ex tax--and, you know, it may be true of my proposal as well, for the reason mostly that Bob just suggested, which is that there will be price effects.  The CRS has estimated they're about three-quarters of the size of the tax.  So it would be about 10 percent, I suppose, on this tax.  But whatever they are, they're substantial.

And my problem is that having dealt in the 1990 budget negotiations and in the 1993 legislative period with non-border-adjustable taxes on American businesses, I simply do not believe that it is possible to get through Congress a plan which taxes imported goods only on the mark-up, so that we're taxing imported automobiles only on the mark-up and we're taxing U.S.-produced goods on the full value-added, on the claim which is the general claim--and I'm not denying it--that all economists always assert that this will balance out in currency adjustments over some period of time.

You know, I'm not sure.  I've seen the math enough.  I believe the math entirely.  And whether I believe it or not is unimportant.  I will tell you that American businessmen don't believe it and will not be convinced by the math, and they have a lot more sway than I do, or any of us, with Congress.

And so I just don't believe we're going to go to a consumption tax that's not border-adjustable.  So that's part of the reason for this, and I guess the other thing is that I don't--the ex tax taxes only wages at the individual level and this is, I guess, part of the problem with CBIT, is that I've always wondered whether we could have a tax that would be stable politically over time which taxed only wages at the individual level and didn't tax dividends or interest at all, or capital gains at all.

You know, that was the reason for the enactment of the 16th Amendment historically, and I don't believe that's going to be very stable either.  So it really relates to both questions.

MR. AUERBACH:  There will be a future seminar on the mathematics of border adjustments.

MR. GRAETZ:  I believe the math.  I've seen the math; I believe it.  They're absolutely right on that.

MR. HALL:  This is like the proposition that a feather and a canon ball fall at the same speed in a vacuum.  No amateur believes it.  You have to be a professional to believe it.

MR. AUERBACH:  Right.  I'd like to thank you for your questions and thank these two panelists for their presentations.

[Applause.]

MR. HASSETT:  We'll move directly to the next panel.

We're going to start the next panel right away.  We've got short panels and we're going to try to have a death march right through them all.  So Dan Shaviro has agreed to moderate this panel.  Casey, you're up here, too.

Dan?

MR. SHAVIRO:  Thanks, Kevin.  Hopefully, all of us are strong enough to survive without a bathroom break here.  I know I'm concerned myself.

I might have a couple of comments or questions to ask the panelists at the end, but I'm going to just let them start out right away.

By the way, Kevin, 15 minutes--is that what I should be timing you on?  Okay, thanks.

MR. HASSETT:  And is Casey going first?  Are you ready to go, Casey?  Okay.  Casey is first.

MR. SHAVIRO:  Okay.

MR. MULLIGAN:  In the keynote address, Doug pointed to the political economy of tax reform as a area that is in urgent need of more study if we want to do tax reform right.  And so what I try to do--I agree with that, so what I try to do in my chapter is to touch on a couple areas in economics that are starting--and I emphasize the word "starting"--to fill that void.

And at least one of those areas should be pretty palatable to this group because it really just takes the analysis that we've already used to analyze the right tax reform without regard to the politics and takes that one small, logical step in order to understand how the tax system might evolve after a reform.  Now, the result might be a little unpleasant, but we'll get to that.

So the idea that--let me give you the result first and then show you how I got to that idea by using the logic of optimal taxation, really.

The idea, seen in a number of places, some of Buchanan's [ph] work--I like the way Barrow [ph] put it a little more succinctly where he said that the general lesson for those who want a big government is to pay for government through a uniform tax on the broad middle class.  Call it the flat tax, call it the VAT tax, but that's the idea.  The idea is that you can feed a big government with a flat tax.

I guess Barrow doesn't like big government because then he concludes by saying, I believe that the implementation of a VAT would be a terrible idea.  You don't have to agree with the last part, but the analysis that I emphasize in my chapter is just the cause and effect from having a flat or VAT tax to having a larger government.  You may like that result, you may not like that result, but those would be the facts.

Now, this idea is old, and maybe the guys that started out here are right-wing wackos, so we don't have to pay attention to them.  But there are a bunch of others who work for the Democratic Party who have said the same type of thing, and others who haven't worked for any party saying the same thing that a better tax system may lead to larger government.

But what hasn't happened is that this old idea that's been around for a long time and seems to make sense isn't consistently applied.  In fact, Barrow when he was opining about the flat tax, which is, as we know, a uniform tax on the broad middle class, said a flat tax was a good idea and he ignored what he had said other places, namely that a flat tax would lead to a bigger government.  It's not a part of the core of public finance yet, so it sometimes gets forgotten.

Friedman [ph] is another example where he said that the flat tax can lead to big government, and on other occasions he has said the flat tax would be a good idea.  And most of the chapters in the book we produce--there's only two of them that even mention the possibility, and none of them really offer an administrative or other solution.  There's lot of analysis of how to administrate other details of the income tax, but no discussion of what can we do to control the size of government in one direction or another under a new and better tax system.

So that's the result, perhaps a scary result for some.  How did we get that result?  How did I get the result and how did the others who went before me?  Well, the idea is one we're familiar with from optimal taxation, which is that a dollar collected by the Treasury costs the taxpayers more than a dollar.  Taxpayers send a dollar to the Treasury, but it costs them more than a dollar.  Why?  Because they change their behavior in order not to pay two dollars, and that changed behavior has costs to it.

They're rational that they incurred those costs in the sense that they'd rather change their behavior somewhat and avoid the two-dollar tax in order to pay the one dollar.  But, still, in accounting for the full social costs of taxation, we've got to understand that a dollar in the Treasury costs more than a dollar.

How much more than a dollar?  A lot of people included in the book and in this room have worked on that question and there's some disagreement, although there's no disagreement that a dollar in the Treasury costs significantly more than a dollar.  Whether it's $1.40, which is a significant difference, 40 percent, or whether it's $3.00, a 200-percent extra, we can argue about.  But in any case, there is a significant access burden, as we call it, this cost that taxpayers pay that never makes it way to the Treasury.

So that's the economics, the very old economics of the excess burden.  But that same logic can tell us a little something about the politics, namely the same logic that leads us to believe that taxpayers will change their personal and business behavior in order to reduce their tax liability--they'll also change their political behavior in order to reduce their tax liability.

How much do they want to adjust their political behavior in order to reduce their tax liability?  It depends on how much taxes are costing them, which is the amount they're sending to the Treasury, plus the excess burden.  So the larger is the excess burden, the more taxes cost the taxpayers and the more they're going to fight back.  And that's how we take this basic economic public finance logic of excess burden and use it to say something about what happens to the size of government.

So maybe to give an example, if we thought--just to take a couple numbers off the screen, if we thought that the current tax system cost taxpayers three dollars per dollar sent to the Treasury, that's three dollars' worth of reason for a taxpayer to ask for government to hold back a dollar worth of tax collection.

If we reform the taxes and make it better--to use another number on the screen, make taxes only cost a $1.40, a dollar of taxes only cost $1.40, that's about half the reason to fight back against larger government; still, a reason to fight back, but half the reason, and you might think taxpayers would resist less.  And so that's the logic by which a better tax system can lead to a larger Treasury.

Now, you see as part of that logic, which wasn't in the optimal tax logic, that taxpayers are somehow part of the process of resisting the size of government.  So you have to buy that proposition, but let me try to convince you of it a little bit if you don't agree already.

If it's not taxpayers limiting the amount that goes to the Treasury, what is limiting it?  Well, I suppose technology could.  There's just some technological limit on how much the Treasury can collect.  But I don't think that's that pertinent in the United States, at least, because we can just look at other countries who are able to collect much more revenue per capita or as a percentage of GDP and they don't have dramatically better technologies than we do.  We have the best technologies in the world, so it's not technology that is stopping the tax collection

Something else is, and I would venture to say a big part of that something else are the taxpayers.

Another reason, it seems to me, that taxpayers are part of this process of resisting tax collections is a comparison of the payroll tax and the income tax.  I don't have the slide to show you today, but it's in my chapter.  I graphed over time since, I believe, the '30s the revenue collected by the payroll tax and the revenue collected by the personal income tax.

The payroll tax, as we know, is a flat tax, sometimes even better than flat because of the cap on it in a sense.  It's a very efficient tax, and if you read the Flat Tax book, in capital "f" and capital "t," they explain how the payroll tax is really a great tax from an efficiency point of view.

And it's the payroll tax that has been increased over and over and over again.  I think I counted 20 times the rate of payroll taxation was increased, as opposed to the income tax whose revenues as a fraction of GDP have been pretty stable.  There have been increases, as under Clinton, and there have been cuts, as under Reagan and under Kennedy and others.  So we see this idea shaping out over the history of taxation that the efficient tax, the payroll tax, is the one that's able to grow.

We can also talk about the foreign situations of raising tax revenue over the years, and VATs have been an important part of those revenue increases.

So we have this kind of paradox that the bad tax system might be better.  But you might say, and I would say, okay, a bad tax system has this silver lining that it creates resistance to tax collections, silver lining at least for those who think tax collections are big enough as they are.

But isn't there a better way to limit tax collections?  And that's really the second strand of research which I want to mention, which has been less productive, I'm afraid.  But there's been some research on the question of how different political rules, rules for making public policy, lead to different policy outcomes, in particular those that might limit public spending in a more direct way rather than this indirect way of having the bad tax that makes people mad and limits its collection.

So I'll mention a few of those ways, and a lot of our information here comes from looking across states of the United States, the 50 states, and how they collect revenue and the types of constitutional, electoral and other rules that they have.

Sometimes, it has been suggested that increasing the power of the executive, the President in the federal case, might limit the size of government.  The evidence there is pretty mixed across the states.

I might suggest capping the voting age at sixty-five but I wouldn't survive in this town, I suppose.  But I think that would limit the size of the government.  A super majority is a more serious thing that's been considered in several states and it seems like all the evidence there is not straightforward to interpret but it seems like super majorities help limit taxation.  Explicit caps--they're out there, limits on the rate of taxation.  It's not clear whether they a reaction to big government or something that can actually limit bit government but those are ideas as well.

So my first bullet pertains to some of the evidence across states.  We can also use theory here, although theory's not--I'm afraid it's not well-developed yet.  It's too bad cause I guess we want tax reform soon and we want to do it right.

But most of the theories behind these and other political rules for limiting tax collections revolve around voting, theories of how votes are cast and collected and how politicians campaign for those votes, et cetera.

The problem is what we're finding in the political economy field is that a lot of the public policies we see in democratic societies where there is voting and campaigns for the votes, et cetera, et cetera, we find those same policies, good or bad, in nondemocratic countries as well.

So it's a little bit hard for me to, with a straight face, use a theory based on voting and the rules of voting, to explain why the government  does what it does, good or bad, and I see those same outcomes happening in nondemocratic countries.

So we really need a theory, and maybe the theory of excess burden I sketched out a few minutes ago is such a theory, but we need a theory that's a little more general than that.

Okay.  I think the last point that I'd mention in my last 45 seconds is even if we could settle on using a theory or using data across states on a particular constitutional rule that might limit tax collections, still that constitutional rule has to be advanced and defended by somebody, and that somebody might be the taxpayers and maybe a set of mad taxpayers are better defenders of those things than taxpayers who aren't hurting that much.  So, thank you.

MR. HASSETT:  Thanks very much, Dan. As many of you know, just as this project was getting underway and we were scheduled to meet and go over our papers for the first time, our friend and colleague, David Bradford, had suffered an accident, and subsequently he passed away, and David had sent us his contribution to this book shortly before the accident, and we felt that it was extremely important to include David's work in the book because he brought so much to the table, and those of you who have read David's work before know that before you pick it up, you have to put your thinking cap on. 

Every sentence is going to be very carefully put together and very meaningful, and I'd have to say that the chapter that David provided for this conference was one of the best things that I'd ever seen him write, and I regret very much that he's not here to talk about the things that he discussed in his chapter, because they're so important.

When Alan and I talked about how to run a conference that turned into the book, we said, well, what we need is a bunch of folks who know a lot about this, who are willing to just sit down and kind a work out their differences, so we can try to figure out what we know and what we don't know, and the first person we thought of was David Bradford.  Those of you who know him know why.  Because he's a person, in addition to knowing more about it than any of us, who is also about the nicest person that you'll ever meet, and David had a long association with the American Enterprise Institute and with the tax community at large, which is evidenced in many different ways for the memorial activities that have been going on for David.

What I'd like to do briefly, in memory of David, is try to present what his chapter is about as this session right now.  I obviously won't be able to do David's ideas justice, but I can make an attempt to get his ideas across, as he would have if here were here.

David, if you look at his arguments for the X tax, tends to shy away from, or tended to shy away from arguments about how the X tax or consumption tax is going to make the economy boom.

And so David's argument for the X tax was not typically based on economic efficiency or the 7 to 10 percent higher GDP that Alan Auerbach and I mentioned in the introduction as our read, and something that maybe the literature says is a reasonable guesstimate for what might happen.

Indeed, in David's chapter, he mentions, as a theoretical matter, that it's not true that it's a general proposition that's generally true, that a consumption tax has to be more efficient and provide higher economic benefits or fewer economic distortions than an income tax, because a consumption tax necessarily doesn't tax some things, say capital, and therefore the tax rates on other things have to be higher, and so, as a theoretical matter, you could conceive of the tax rates on the other things causing enough damage, that a consumption tax is not better.

And so he says, well, it doesn't have to be true.  Then he seemed to have a little bit of disbelief or skepticism about estimates that suggested that there were big benefits.  It's not that he believes that they weren't true.  It's just that he felt that with regard to this tax reform debate, it was unnecessary to allude to these things because there are other reasons why you might want to have a consumption tax, that he thinks are convincing enough to decide the case before you even look at economic efficiency.

It's not to say that if one could show that a consumption tax harmed the economy, that he wouldn't have given this up.

I think that David, back when he first started forming the X tax, recognized that he thought that a consumption tax was the most desirable tax, and that he would like to think of a way to make it happen, and as Bob went through his slides, briefly, he mentioned the X tax, and the way to think about the X tax really is that it's a flat tax like Hora-Bushka [?], but it's got more rates, and the top rate is, at the individual side, on wages, is the same as the rate on businesses, and that is an important consideration because it makes it difficult to create tax scams by redefining what you are.

David lists a number of reasons, in his chapter, why he thinks that having such a structure is desirable.

First, he says it resolves the interest deduction mess, and I think that we could ask Dan to go into the interest deduction mess, if somebody would like to know the details, but defining exactly what is an interest deduction is the subject of many, many difficult and arcane rules.

In fact David said resolving the interest deduction mess, all by itself, is reason enough to adopt the X tax.

The second thing--he says that in his chapter.  The second thing is that it removes capital gains from the tax base, something that David views as useful simplification.  It will adjust the tax base for inflation.  It will create a level playing field for investment, and on this, David, perhaps like Bob, served the function of the thinker and the academic because he was unwilling to back down from the view that owner-occupied housing does not deserve special treatment, and that we should level the playing field for all types of capital, including housing.

David mentioned that the X tax would cut the complexity of retirement savings significantly, and also he mentions that it would facilitate individual filing, so that there wouldn't be a marriage penalty or subsidy anymore.

And so you can see where David's going about motivating the X tax.  He's listing all the things that are wrong with the code and the things that are addressed by the simple structure that he proposes.

The other thing that David does, that's a new contribution in this chapter, is he talks about the damage of changing rates over time, and as Alan and I were putting the conference together, we actually had some trouble with one of the funders of the conference, one of the persons who we wanted to approach to fund the conference, because we didn't want to actually say what the chapter's were going to be about, because we didn't want to guide anyone.

WE just wanted everyone, who we thought had really studied this a lot, to write down what they think the most important things are, and then we wanted to see what those were and then talk about them.

So that the application for funding was extremely short because we wanted to invite some guys and tell them to write what they think, more or less.

And one of the things we learned from this conference was that the unanimity of the view that the problem of fundamental reform is how do you make it stick, and there were two perspectives developed, one by Casey and others, Casey mentioned, the idea that if you have an efficient tax than maybe that's going to have an effect on government spending.

But David developed, in his chapter, an analysis of what happens if they mess around with the tax code again and how severe damage can be caused if you have a consumption tax but you change the rates.

So, for example, if you have a deductible IRA, is--that's how your qualifying account, that's how you get into a consumption tax, so people can define their consumption as their income minus what they put in some account somewhere.  Then if we increase the rate, then it's very easy for the tax rate on their capital income to go above a 100 percent, and David has an example like that in his chapter.

And so David actually modified in this chapter his X tax design, quite significantly I think in places, because of the danger of changing the tax code.

I think that one of the lessons from the conference really is that if this tax reform commission does propose something, that it's going to have to be accompanied by measures to address the things that David and Casey and others in the conference have brought up.

Finally, I think I would commend to you David's discussion of equity and the role of equity in defining a tax reform objective.  In particular, he mentioned that the primary attraction--one he said this in conversation to me and he almost says it exactly like this in the book--that one of the primary objections he has to income tax is that it's not horizontally equitable.  That if you have two people with the same lifetime income but they have it in different parts of their life, then you're taxing them differently and it doesn't strike him as equitable, whereas a consumption tax would.

And then he pivots from that and he says, okay, so if we want to have real horizontal equity, want to treat people the same, then that leaves the debate about vertical equity and that's really the beauty of David's design of the X tax, because what he says in the chapter is, well, I don't really know what the right answer is.  He doesn't say this explicitly but that's really the spirit of what he's saying, about how progressive the tax code needs to be.

But we should have a consumption tax that's flexible enough so that the politicians can work out something that pleases them, and there's no reason why a vertical equity argument should prohibit you from having a consumption tax.

And so on the grounds of equity, then, you'd have to say that you should have a consumption tax, if you believe both horizontal and vertical equity have meaning, cause a consumption tax will have more horizontal equity and it could have the same vertical equity.

That elegant argument was an example of the kind of argument that David would make all the time, and as Alan and I were concluding the book, one of the things that we thought was really a fitting justice to his memory, was that in the end it seemed like if there were a consensus--and I know Mike wouldn't be included in this part of the consensus.  But if there were a consensus of the group, it would be that if you're going to have a reform and you thought you could solve the political problems, and you thought you could solve the tinkering problems, and so on, then you would want to adopt something like the X tax.

As Bob Hall indicated in his remarks earlier, he even thought that there were arguments that seemed kind of attractive for the X tax as well, and I think that, in the end, almost a consensus of a large group of people of varying political affiliations, surrounding the advisability of David's X tax, is probably the best tribute that could happen for an academic like David.

And with that I'd like to thank you for your attention.

MR. SHAVIRO:  Thanks, Kevin, also for getting us back a couple of minutes.  I have a question or comment for each of you.  For Casey, I want to say this.  We all understand the idea that government activity can be associated with deadweight loss.  Also I think we all understand the idea that waste, although independently undesirable, if the things that are being funnelled to people are wasteful, then they kind of have less incentive to want them.

The problem I have with your analysis goes to the identification of the size of government with the amount of government spending, and one thing David Bradford always pushed was the idea that tax and spending are just labels for opposite direction cashflows, and in fact almost any provision of one kind can be relabeled as another kind, and I'm using an example of his, was to zero out defense spending and instead have a weapons supplier tax credit, which would be exactly the same program, and everyone have the same weapons for exactly the same money and everything would be the same.

Also likewise, when the government transfers things--this is the thing Gordon Telik [ph] has written about--it's clear that very often wasteful transfers are necessary, politically, if only to have a fig leaf.  I can't just have the technology hand me money to do what I want with.  Maybe I have to tell them I'm doing the Shaviro study that's of no value and that I'm going to waste some time and money on.

So I have trouble with getting from your basic analysis, which I think is powerful, to the idea that this means that spending is the thing that has to be cut, rather than looking directly at the allocative effects of government on the society.

I don't know if I should let you respond first, or just say something for Kevin too.

MR.      :  [inaudible].

MR. MULLIGAN:  I would agree with David as well, that measuring what government is doing is hard, which throws another monkey wrench in the idea that somehow we can put a line in the Constitution that keeps the amount of government interference in the right level, because there are so many ways to do the same thing.

The analysis still applies.  It applies to regulation for that matter.  Even if we had no Treasury.  If taking from one group and giving to another is a costly process, the group being taken from is going to be resisting more than if it's done in an efficient way.

Like the example of airline regulation which was benefiting some parties and hurting others.  But the efficiency of that regulation got worse and worse over time and that's what I think led to its demise.

So it's really across-the-board type of analysis.  Any area of government interference, whether it's called spending or taxes or regulation or something else, the less efficient it is, the more resistance there'll be to that interference.

MR. SHAVIRO:  Yeah.  I don't think I should follow up on this now, given that the audience will probably want to get in, but I'll just--for Kevin.  First of all, I want to say of course how sad I am, and I know all of us are, about the terrible, tragic thing that happened to David Bradford, who was not only a great economist, but really one of the most wonderful people I've certainly ever met.

One of the most interesting things about David's chapter is the part Kevin mentioned at the end, and actually David had a tradeoff, that not everyone totally agreed with him about, but it kind of, the tradeoff he made reflected kind of what his priorities were.

Which is he came to realize that a straight consumption tax, with expensing, has really big transition rate problems.  So he decided to fix that by trying to have some much more complicated-looking schemes, to have something a lot more like income tax accounting, that would really still be a consumption tax, economically, but it wouldn't be as sensitive to rate changes.

One problem with that, with some of the plans it was hard to know exactly how they'd work, but he certainly risked complicating the optics.  If one of the sales points of his plans, politically, if not analytically, was the simple optics of expensing, he certainly risked losing that benefit and that was, I think, you know, if David were here to answer the question, obviously not quite the same to have someone else, even someone as thoughtful as Kevin defending some other person's plan, that would really be a question I would have put to him, is, you know, are you losing some favorable optics of a really simple-looking plan?

Is that a problematic tradeoff for the lesser transition problems.

MR. MULLIGAN:  My read of this, and I did not speak with David about this, but let me, for those of you who haven't read his chapter yet, and I know there won't be many of those left after we're done today.  But David suggested that you could have the same economic effect of expensing if you just indexed the depreciation al