NYU Tax Policy Colloquium, week 11: Nirupama Rao's "The Price of Liquor Is Too Damn High: State-Facilitated Collusion and the Implications for Taxes"

On Tuesday, we discussed the above paper, which was our second (the first being Saul Levmore, week 1) to discuss the choice and interplay between "tax" and "regulatory" instruments.  The paper concerns the measures states take to limit or control alcohol consumption, often with evident secondary (or should I say primary) aims of empowering cartels and favoring local over out-of-state producers.

Alcohol clearly is a good subject for Pigovian taxation.  Drunk driving accidents are certainly the clearest example of negative externalities resulting from alcohol consumption, but not necessarily exclusive.  One could also make internalities arguments, especially with regard to young drinkers.

As a theoretical matter, Pigovian taxes are easier to design in some cases, harder in others.  An article called "Taxation and the Financial Sector," which I coauthored with Douglas Shackelford and Joel Slemrod, makes the point that the negative externalities imposed by financial firms when they risk failure that may lead to bailout and/or horrendous macroeconomic consequences, are very context-specific and hard to capture accurately in a tax instrument.  One institution compared to another, or one state of the world compared to the other, may make all the difference in determining what the tax level ought to be.  On the other hand, for carbon taxation that is aimed at global warming, while the overall harm measure may be difficult to nail down, at least we know that all carbon molecules are relevantly the same.

Alcohol arguably is somewhere in the middle.  On the one hand, it's true that each unit of wine, beer, or spirits has a determinate amount of alcohol in it.  But even sticking to the externalities, it matters who is drinking, as well as how much that person is drinking at the same time.  One could imagine a sci fi dystopia in which everyone is equipped with monitors for alcohol in the blood that permits the imposition of personalized Pigovian taxes.  But since, in the real world, purchase not consumption of liquor is the occasion for levying the tax, we can't come close to doing that.  Thus, moderate drinking that may improve one's physical and mental state is hard to except from the regime that is aimed at drunk drivers and self-destructive alcoholics.

There are some fallback methods available.  For example, we (try to) ban under-age drinking.  And differential taxes on wine versus beer versus spirits can proxy for the things we'd like to measure directly.  For example, if we think that young drinkers tend both to prefer beer and to impose the highest drunk driving externalities, we might for that reason subject beer to a higher tax relative to alcohol content.  Not surprisingly, however, the states turn out to view things a bit more parochially.  California, with its big wine industry, is very nice to wine.  Pennsylvania, with its blue collar tradition that we hear about ad nauseum in every presidential election year, is nice to beer.

Explicit excise taxes on alcohol are mainly federal, and only secondarily (by dollar value) state and local-level, and most experts would probably say that, on balance, and despite the variability that we can't capture very accurately, they are set too low, rather than just right or too high.  Arguably this reflects the cultural difference in today's world between alcohol and, say, cigarettes.  I would speculate that "sin" taxes which are best rationalized on externalities grounds are highest when the sin is one that "we" don't indulge in - only "they" do.  By "we" I mean representative voters, with "they" being the Other in such voters' eyes.

Evidently, in the era that led to Prohibition, to some extent there was a view that only "they" drink - "we" don't.  Then if you go the Mad Men era, evidently "we" both drink and smoke.  Today, however, smoking is declasse while drinking is not.  So "they" smoke and let's tax it a lot, but "we" drink so let's not tax it so much.

OK, onto the state regulatory structures that the paper discusses.  The front part of its title, "The Price of Liquor is Too Damn High," might more accurately be stated as "The Pre-Tax Price of Liquor is Too Damn High, Albeit That the After-Tax Price Might Conceivably Be Too Low."  States do a number of cartelizing things in the market for alcohol sales that end up raising the price.  These measures may have Pigovian benefits if they suitably reduce alcohol consumption, and they apparently do reduce it.  But doing this through the creation of market inefficiencies means that the Pigovian revenues are either lost in pure waste or handed off to wholesalers who are aided by the state regulators in realizing monopoly profits.

A case in point, and the paper's main focus, is Pennsylvania's "post and hold" regime with lookback.  Here's how it works.  Pennsylvania has a lot of rules in the alcohol market that at least ostensibly are aimed at protecting small retailers, in particular against the likes of Costco.  (In an unregulated market, Costco would likely be able to get a better wholesale price, reflecting either cost-saving from the scale that it offers and/or its exercising monopsony power.)

To this end, Pennsylvania requires wholesalers to state in advance the uniform wholesale price that they will charge to all retailers for one month after the posting.  That is "post and hold."  It facilitates collusion in pricing between wholesalers, and apparently helps result in Pennsylvania's having significantly higher liquor prices than Massachusetts, an adjoining state without post and hold rules.

"Lookback" is the crown jewel of post and hold, so far as permitting wholesalers to reap cartel profits is concerned.  Under lookback, once you have posted your price for the month, you can lower it to match a competitor's posted price.  This makes it really attractive to aim high on your Stage 1 price, knowing that if someone undercuts you it won't be a problem, and that, for the same reason, you can't undercut them.  Especially with repeat players and monthly posting, this is simply a great device from the standpoint of wholesalers who want to collude in ways that would land them in jail for antitrust violations (or else make sufficient cooperation between would-be cartelizers unstable), but for the state's facilitating role.

Anyway, if the alcohol price would otherwise be too low from a Pigovian-plus-internalities standpoint, higher prices from post and hold plus lookback at least have the virtue of reducing alcohol consumption, conceivably closer to its optimal level.  But it's equivalent to the state's converting the Pigovian tax revenues into a combination of handouts to liquor wholesalers plus dissipation through waste.  The paper makes this case, in addition to showing (both theoretically and empirically) how post and hold plus lookback operate to raise retail prices for alcohol.

My article on Henry Simons has finally been published

My article on Henry Simons, which I presented at a Florida State University Law Review symposium a bit over a year ago, has finally officially appeared in print.  You can find it here.

The abstract goes something like this: "Surely just about everyone in the U.S. federal income tax field has heard of Henry Simons, if only for his famous definition of “personal income.”  Few may realize, however, that this proponent of 'drastic progression' in a broad-based income tax was also a self-described libertarian who generally denounced government economic regulation and was arguably the chief architect of the pro-free market law and economics movement at the University of Chicago.  This article provides a brief intellectual history of Simons’ work, aiming in particular to explain how and why he combined these seemingly disparate sets of beliefs, and what we may learn from them today."

I must confess I rather like, and enjoyed writing, this article, although it is not quite history and also, as a friend who prefers some of my other work told me, is not as "analytical" as I sometimes might be.  This reflects that it's partly a literary enterprise and character study.  

NYU book event regarding my international tax book

I'm very much looking forward to the following event, more fully described here.

"Fixing U.S. International Taxation by Daniel Shaviro – Monday, April 28th, 12:30 PM to 1:50 PM – Vanderbilt Hall Room 220, 40 Washington Square South:

"Please join us for a discussion of Professor Daniel Shaviro’s current book, Fixing U.S. International Taxation (Oxford University Press, 2014). 

"Following a presentation by Professor Shaviro of key themes from the book, leading experts in the field will offer commentary.  We are thrilled that Martin Sullivan, Chief Economist at Tax Analysts (publisher of Tax Notes), and Professor Itai Grinberg, Georgetown University Law Center, will join us.  We will also save time for a question-and-answer session with the audience."

At the event, the NYU Bookstore will be offering copies of the book for sale at a 20% discount.  I suppose inscriptions aren't out of the question either, should any demand for them materialize.

Later the same day, Sullivan will also be giving a talk on a topic of general interest to our audience, most likely pertaining to recent developments in the tax reform process.  I will post that here when it's up officially.

UPDATE: Sullivan's 6 pm talk at NYU Law School on Monday, April 28, will have the title: "Tax Reform 2017: Incremental or Fundamental?"

NYU Tax Policy Colloquium, week 10: Susannah Tahk's "The Tax War on Poverty"

Last Tuesday (April 8), Susannah Tahk of the University of Wisconsin Law School presented the above-named paper.  As noted in an earlier post, I was forced by external circumstances to fall short of my usual practice of promptly posting here a discussion of the issues that the paper raises.  But better late than never.

The paper documents and evaluates the rising use of tax provisions to do heavy lifting with respect to programs that are aimed at aiding the poor.  The most important example is the rise of the EITC, including its refundable element, and the relative decline of direct cash grants under TANF today as compared to AFDC twenty years ago.  There is also the fact that, say, the child tax credit, while aimed in large part at the “middle class,” is partly refundable, to the benefit of poor households with children.

It’s clear that use of the income tax system in lieu of direct transfers to pay money to poor people has potential optical advantages, compared to overtly using the transfer system – although it’s less clear to what extent these optical advantages are weakened, once one makes the payments refundable.  And of course 2012’s “47 percent” meme reflected a related part of the optical downside of formally netting transfers against income tax payments.

A broader issue – especially if refundability is inevitably limited and if there is some optical lean in the "income tax system" towards using exclusions and deductions rather than fixed-percentage, refundable credits – is that it may be hard to use this means when directing significant transfers to people on the very bottom.  Plus there is the vexed question – central to the anti-poverty debate – of what role work should play in the design of the programs, not to mention household structure such as number of kids.  These issues arise no matter what formal system one uses.

Suppose we had a uniform demogrant for all U.S. individuals, financed through explicit tax rate increases that were very limited at the bottom of the income scale.  In the typical language of public economics, this would not reduce “incentives to work,” which depend purely on marginal rates.  But in the typical language of anti-poverty discussion, the income effect of the demogrant would indeed reduce “incentives to work.”  That is, recipients would no longer face as dire a threat of privation (including starvation, assuming no other transfers) if they decided not to work.

This is not a case where one side or the other in this language mismatch is logically in error – it’s a question of how you want to use the term “incentive to work,” which  depends on what one cares about.  This in turn can depend not only on one's underlying normative views but also on empirical assumptions.  For example, a pure utilitarian would not care about "rewarding work" or 'helping the most deserving" as an end in itself, but might believe, depending on the empirical evidence, that in practice inducing work was either very important or not important at all (or anything in between).

The EITC is a sufficiently interesting and important provision to need substantial consideration all on its own.  In one sense, it’s “anti-insurance.”  Suppose that A and B, both poor, are both seeking scarce jobs.  A succeeds and B fails.  A is therefore better-off than B - not only does he have more income, but they both wanted the job.  With an EITC, A is then the one to reap further rewards from government transfer policy.  Purely from a static insurance standpoint, this makes no sense.

But on the other hand, suppose there are important reasons for encouraging work and “making it pay.”   Then the EITC’s anti-insurance structure can indeed be defended.

Making the picture more complicated still, one really wants to figure out (among other things) the overall marginal tax rates that actual or potential EITC recipients face.  At the same time that the EITC offers a large work subsidy, they may also be facing substantial positive implicit tax rates from the phase-out of various income-conditioned benefits.  Then later on, when the EITC begins to be phased out, the affected taxpayers either may or may not still be facing high implicit marginal rates from other phaseouts.  (These things vary with the taxpayer and by state or locality.)

Anyway, these are interesting and important issues that merit a lot more time and attention than I can offer here while catching up on my blog posts in the late stages of a busy semester.  But it was nice to have a session devoted to it - we hadn't done much distribution so far this year.

Tax day

As usual, I filed early via Turbo Tax.  I gave up doing it by hand some years ago, for reasons that include:

(1) Potential alternative minimum tax liability if you live in a high-tax jurisdiction such as New York, California, or Washington D.C.  Whether one ends up owing it or not, what a hassle to have to do the computation.

(2) The nightmare of, say, forgetting $100 of gross income until you are almost done.  With Turbo Tax, you just enter it and the computer does the rest.  If you're preparing your return by hand, any calculation that depended on adjusted gross income has to be redone.

(3) The last time I did it by hand, there was some insane rule for capital gains, such as trivial mutual fund passthroughs, that related to tax rate changes and special rates. The IRS (through Congress's fault, not its own) had to come up with a form that required something like 15 calculations that I recall as being on the order of, "Take 19% of the amount on line 5, and subtract it from 23% of the amount on line 7.  Then take the square root of the absolute value of the difference, round it to the nearest integer, and enter it on Line 12."

OK, I am admittedly exaggerating a bit on that one.  But only to convey a Deeper Truth.  I do indeed recall finding that, even though each calculation was mechanical, it was very hard to avoid an error (and thus to replicate the result) if one was doing very many of them in a row.

Turbo Tax, here I come, I decided - even though I have subsequently learned that this company may not be the world's best-faith political actor.

But not everyone files early, whether with Turbo Tax or not.  Yesterday (April 14), as I took a late night flight from California back to NYC, the man seated next to me appeared to be doing a federal income tax return on his laptop.  Good thing it wasn't April 15, as our plane was delayed and landed after midnight.

Activities on the side

Some tax shelter-related litigation in which I was an expert witness on behalf of the government has just apparently settled.  You can see an article discussing the case here.

In the words of the article, the case concerns “Texas billionaire banker Andrew Beal …. [and] relate[s] to Beal’s use of an abusive tax shelter the IRS calls Distressed Asset Debt, or DAD, to manufacture billions of bogus tax losses from junk Chinese debt…..

“In a DAD shelter, a U.S. taxpayer forms a partnership with a foreign owner of non-performing loans and then claims huge tax losses from the partnership—losses the foreign lenders, but not the U.S. taxpayer,  sustained.  The DAD technique spread among the wealthy in 2001, 2002 and 2003 after the Internal Revenue Service began a crack-down on better known abusive tax shelters, such as Son of Boss ….

“In 2004, after the IRS got wind of DAD, Congress changed the tax code to explicitly bar partnerships from being used to transfer foreign losses to U.S. taxpayers. Meanwhile, the IRS began auditing existing DAD partnerships, disallowing all their losses as shams and slapping on penalties…..

“Beal had created four separate DAD partnerships in 2002 and 2003 to hold non-performing loans made by the Chinese government, with the aim, the government says, of stockpiling $4 billion in artificial losses to shelter income.  (Even for Beal, whose net worth is an estimated $11.3 billion, $4 billion can offset income from more than a few years.)….

“[In an earlier case, the district court and Fifth Circuit]  nixed Beal’s attempt to claim $1.1 billion in tax losses from an investment of just $19 million in distressed Chinese debt, finding ... that the partnership was a sham that should be disregarded for tax purposes.  But the [courts] also ruled that Beal’s acquisition of junk Chinese debt had 'economic substance' because—and this sounds bizarre if you’re not a tax geek—he had a reasonable chance of making a real profit, even as he angled to claim big losses.  As a result …. both the district court judge and the appeals panel held Beal wasn’t liable for any penalties because he had opinions from both a  law and an accounting firm concluding that it was more likely than not that the partnership ploy would withstand IRS scrutiny. The appeals court called it a ‘close issue’ ….”

The issue this time around was simply whether Beal should face penalties for subsequent years’ disallowed deductions from DAD transactions.  Returning to the Forbes article:

"The government … has maintained that the circumstances of Beal’s three other DAD partnerships and in later tax years might be different—and more penalty worthy.

“Each side had already paid a big name tax professor to deliver an expert opinion on whether Beal could rely on those ‘more likely than not’ law and tax firm opinions to escape penalties.  New York University Law Professor Daniel N. Shaviro wrote for the government that the opinions in 2003 and beyond were weaker because they were based on 'factual premises that had been rapidly losing credibility.'  Specifically, they assumed that since Beal had made profits in distressed U.S. assets, he could reasonably expect to make money in the Chinese debt— which wasn’t turning out to be the case.  University of Chicago Law School Professor David A. Weisbach, whose expert opinion helped Beal escape penalties in the Southgate decision, weighed in again for his side.”

Anyway, this case has now been settled and there won't be a trial.

Back online soon

I haven't posted for nearly two weeks due to family issues that required attention but that are now pretty well stabilized.  Within the next couple of days, however, I plan to post concerning Susannah Tahk's paper at the NYU Tax Policy Colloquium last week, Nirupama Rao's paper tomorrow, perhaps the delights of April 15 (though in California at the moment, I e-filed early), some news concerning recently settled litigation in which I was an expert witness, and a book event at NYU Law School later this month.  Plus at least one other event at NYU that I'm looking forward to, along with a couple of upcoming appearances in Europe that I have on my calendar.

The problem with saying more at the moment is simply that I find my iPad a bit suboptimal for writing and editing posts, and left my laptop home in order to be more streamlined.

NYU Tax Policy Colloquium, week 9: Andrew Biggs' Public Employee Pensions: Investment Risk and Contribution Risk

Yesterday, Andrew Biggs of AEI presented the above-named paper.  (Title at page 1 of the link appears to be different, but it actually is indeed "Public Employee Pensions: Investment Risk and Contribution Risk.")

The paper addresses the funding issues posed by defined benefit (DB) plans for state and local government employees around the country.  This issue is a well-known time bomb that will either explode or else not, possibly in the near future, depending not only on what's done but also on the plans' good versus bad luck with their investments, many of which are in risky equity.

Currently the DB plans are estimated, if I recall correctly, to have about $1 trillion in present value of promised future benefits that remain unfunded.  However, this is based on discounting future benefits at a relatively high discount rate, which makes them look smaller.  The high discount rate (say, 8%, although it varies with the plan) is based on the expected rate of return from fund investments, which often are quite risky.

Given this actuarial convention in making the estimates - which isn't permitted for private sector DB plans that are subject to ERISA - suppose a DB plan is investing its assets at a very safe 2% rate.  It will appear to have very high unfunded liabilities, since they will be valued using this 2% discount rate.  The plan's managers decide, therefore, to bet the ranch.  They shift to very risky investments with an expected return of, say, 8% or 10%.  But this is not free money, of course.  The market values the underlying assets the same as the stodgy old 2% assets, because these items could hit a home run, on the one hand, or blow up, on the other.

Was this a wise investment choice by the plan?  Very possibly not, since the downside would leave it with gigantic unfunded liabilities.  But if I understand correctly how things are being done, the shift would cause the plan to appear as if it were much better-funded, because it could now use the 8% or 10% rate to discount its future liabilities.  This can both create dubious incentives, and mislead the audience for funding estimates to overlook the risk of the downside scenario in which the pension plan would end up being very short-handed.

The paper focuses on the question of how investment risk should lead us to think about the DB plans' future prospects - in particular, the risk that, if the plan investments do badly, there will either be defaults (or at least unexpected cuts) or a need for greatly increased contributions by workers or employers.  In particular, based on a Monte Carlo simulation where assets such as stock are assumed to have an 8% expected return (or lower in one of the scenarios), with a 12 percent standard deviation, it evaluates probabilities for insolvency, required contribution volatility, etc.  In particular, it offers parameters for the unsurprising, if disappointing, conclusion that you can't keep current contributions low and use risky assets to juice the expected return without creating significant contribution volatility and default risk.  In other words, using riskier assets fails to offer a free lunch in the paper's simulations (although it does of course turn out nicely for DB plans' stakeholders in the set of cases where the investments do well rather than poorly).

The quatrilemma whereby you can't have low contributions, plus long adjustment periods to make up shortfalls, plus low contribution volatility, plus low default risk, is based on presuming that a theory to the contrary that the DB funds' proponents and stakeholders sometimes advance is not in fact correct.  This is the theory that mean reversion in stock market returns means that long-term players, such as state and local governments if they can weather the interim storms, actually do get a free lunch in the form of higher expected returns without greater long-term risk.

To illustrate, suppose that we thought the stock market really was pretty much a lock to earn, say, 8% over time.  Then, if you can wait long enough, you can capture the higher return without bearing the variability that torments more short-term players.  For this to be true, however, periods of lower or higher returns must not be merely diluted, but positively offset.  E.g., if the market's been returning only 6% for a while, it's now likely to earn more than 8% for an offsetting period that gets it back on track.

Conceptually, it's as if an honest coin that yielded 5 straight heads is now likely to even things by running more tails for a while.  But admittedly the stock market is a sufficiently complicated beast (e.g., given the role played by market psychology) that we can't rule it out as decisively as we could in the coin example.  Still, it would require a rather odd market inefficiency that you would expect savvy investors to exploit.  (E.g., if the market is due to start earning an unusually high rate of return, the price level ought immediately be bid up to the point where we are back to the normal rate of return.)

I'm certainly open to theories of market inefficiency.  But, while this really is not my area, mean reversion does not appear to me to be among the more plausible candidates.  (And indeed I gather that Bob Shiller, who has gotten a Nobel Prize for writing about market inefficiencies, does not believe in it.)  For now it's just worth noting that, with mean reversion in stock prices or even a little bit of it, the dilemmas (or, rather, trilemmas and quadrilemmas) discussed in the Biggs paper might become at least slightly less binding.

Insofar as the existence of some or any mean reversion in relevant asset prices remains uncertain, I suppose you could say that supporters of the DB plans' current investment strategies and actuarial reporting conventions are taking a risk that they actually aren't taking on as much risk as they seem to be.  And even if the resolution of the second-order risk remains ambiguous, we will certainly learn at some point how their first-order risk-taking has come out.

Oops, one additional point on equal sacrifice theory and classical benefits-based taxation

I see that in my Part 1 post on the Weinzierl paper from yesterday's colloquium I promised to mention something at the end of my Part 2 post, and then I never did.  So I'll add that final point here.

Specifically, I noted that "Mill, in proposing equal sacrifice theory as a replacement for Smith's benefit theory, not only reached a very similar conclusion (flat tax above an exemption for necessities), but cited Smith as an influence and precursor rather than as the dead past that he wanted to bury."  I then offered to make an observation about this overlap.

One interpretation would be that equal sacrifice theory and benefits theory lead to a similar place, e.g., they might plausibly support a flat rate tax.  But, with all due respect for Smith and Mill, I question this because the underlying elasticities on which they turn are both (a) entirely distinct and (b) extremely hard to tease out - not just empirically but as a basic conceptual matter.

Again, under equal sacrifice the key elasticity that drives the optimal rate structure pertains to utility loss per dollar of tax paid as income rises.  This is against the background of a counter-factual hypothetical in which public goods were somehow provided for free.  

Under classical benefits-based taxation, the key elasticity pertains to income gain (reflecting ability enhancement) per dollar of public goods provided as income rises.  Here the counter-factual hypothetical is apparently a state of nature in which we have all the same people, with all the same innate talents, but none of the rudiments needed for a well-functioning economy.

Again, both models lead to a flat rate tax system with an assumed relevant elasticity of 1, based in each case on something that is unknowable.

Why would the relevant elasticities come out the same, causing the Smith benefit tax and the Mill equal sacrifice tax to look so similar?  This seems highly coincidental.  The main explanations that occur to me are:

1) Prominence and salience of 1 as a postulated elasticity.  Perhaps it feels safe and neutral to say that the posited trait grows at a steady rate with income, not faster or slower.

2) Prominence and salience of the flat tax result, which may look intuitively appealing (and also appear to have optical and political economy advantages), suggesting that the causal arrow ran in the other direction, i.e., from the intuitively favored tax rate to the mode of justification.

Either way, I draw an inference that we should not be all that confident that the equal sacrifice and benefit tax rationales are actually doing a whole lot of work.

Matthew Weinzierl's "Revisiting the Classical View of Benefit-Based Taxation," part 2

I broke off the prior post and am starting a new one here just for reasons of length.  This is a continuation that presupposes one's having looked at the prior one.

OK, on from equal sacrifice theory to benefit taxation. The paper that was posted for yesterday's session addresses “classical benefits-based taxation,” under which it is thought that people should pay a fair price for the benefits that they receive from government spending.  These include at a minimum specific public goods that particular programs address (e.g., national defense, the courts, and police protection), and perhaps more broadly the benefits of living in society that has a functioning economy, the rule of law, property rights, infrastructure, education, etcetera.  So once again, as I noted about equal sacrifice theory in the prior post, we are making policy assessments relative to a counterfactual – a hypothetical state of affairs that cannot in fact exist.  Only here, rather than being public goods supplied for free, it is no public goods, and presumably the existence of some version of a state of nature

 As Weinzierl notes, the classical benefits-based approach has been almost wholly eclipsed in the literature, due to challenges both normative (welfare economics and rejection of libertarian entitlement) and descriptive (how value public goods?  What’s the counterfactual from which benefit is measured?).  However, the paper develops a model in which something akin to Mirrlees' optimal income tax (OIT) model can be deployed to describe the basics of a benefits approach.

In Mirrlees, we tax income as a proxy for ability (assumed to be fixed and unalterable), which is itself a reverse proxy for marginal utility (i.e., when it is high, the marginal utility of a dollar is assumed to be low).  In the benefit tax world, we no longer care about utility.  All we care about is benefit from the provision of public goods.  In the paper's model however, income is a proxy for benefit, based on a modification of how ability is used.

In the paper's benefit tax model, ability, while it remains unalterable by the individual who possesses it, is assumed to be a joint product of (i) innate talent and (ii) public goods funded by tax revenue.  Innate talent and public goods are assumed to be complements.  At any talent level, more public goods leads to greater ability (i.e., one can earn more, e.g., by reason of infrastructure, a functioning exchange economy, the rule of law, the existence of educated consumers and workers, social peace enforced by the army and police, etc.).

Just as in Mirrlees, income is a proxy for ability.  But in the benefit tax model what's actually of interest is public goods' magnifying effect on innate talent to create higher ability.

For convenience, let’s assume that ability enhancement by public goods is constant as a percentage of income.  That is, assume an elasticity of 1 for income gain (reflecting higher ability) per dollar of public goods as income rises.  Thus, if I earned $1 million, I got ten times the benefit from public goods of someone who earned $100,000.

Just like under equal sacrifice theory as per the prior post, assuming a particular elasticity (but here a different one) of 1 leads to a flat rate system.  But the benefit tax system should be progressive if high-earners got more relative benefit from public goods than low-earners, and should have declining marginal rates if they got less relative benefit.

Leaving aside the question of normative motivation for embracing such a system, I think that some of the biggest questions that this model raises relate to (1) its dependence on comparisons to a hypothetical counterfactual state of affairs and (2) the concept of innate talent.  As to (1), even apart from the question of what motivates the counterfactual, I don't see we can define it usefully, or have a coherent debate on what it should be assumed to look like (and why).  As to innate talent, I don't think there is such a thing.  The value of any attribute depends on the environment in which it exists.  This is the lesson of evolution – fitness is relative to the environment, and changes when the environment does.  A trivial example is a star athlete with or without popular taste - and a global versus a purely local market - for the sport in which he or she can excel.  But consider as well having good math skills, or an aggressive personal style, or for that matter strong genetic resistance to particular diseases.  The same people or the same skills will do well in some environments and poorly in others.

This is not a problem for Mirrlees because we are assuming the existing environment, in which abilities have particular payoffs.  (Mirrlees does, however, ignore ongoing ability risk as the environment changes).  But it is a big problem for the concept of innate talent, if one is using it as part of the baseline from which to measure benefit from the existence of  a functioning exchange economy and all the rest.

That said, Weinzierl clearly has one of the more interesting and distinctive research agendas among active people in public economics these days.  I will look forward to continuing to follow his  work on non-OIT or at least non-welfarist normative approaches to tax policy.

NYU Tax Policy Colloquium, week 8: Matthew Weinzierl's "Revisiting the Classical View of Benefit-Based Taxation," part 1

Yesterday Matt Weinzierl of the Harvard Business School presented the above paper, as our colloquium crossed the halfway point for 2014.  It's useful to put this paper in the context of others that he has recently written, reflecting an ongoing project of engaging with some of the underlying philosophical issues in tax policy.  While the paper has a lot of math, along with shorthand references to economics ideas (e.g., Cobb-Douglas, Lindahl, and the Samuelson rule for public goods) that potentially make it tough sledding for a legal audience, in fact the set of underlying set of interests may well be a better fit in a law school environment such as ours than in many business schools.

A bit of background may be in order, and indeed is most of what I'll discuss in this post, with the aim of sketching out the underlying project for a broader audience.  Economists and some economist-fellow-traveling law profs, such as myself, regard optimal income taxation (OIT), derived from James Mirrlees' pioneering work in 1971, as a foundational approach for thinking about progressivity, redistribution, and tax rate  design.  OIT generally applies a utilitarian or other welfarist framework in which tax policy's aim is to increase social welfare, defined in terms of utility or subjective wellbeing.  Notions that people are "entitled" to the fruits of their own labor thus play no direct role, although one might view such notions as being transmuted into concern about the incentive effects of taxation.  OIT is all about trading off the adverse incentive effects against the redistributive benefits that are attributed to taking money from people to whom it is thought to have low marginal utility (e.g., because they are wealthy) and giving it those to whom we attribute high marginal utility (e.g., because they are poor).

OIT is thus potentially highly progressive and redistributive.  In practice, not necessarily - this depends on such factors as the elasticity of high-end taxable income, and indeed Mirrlees himself kept finding that a relatively flat rate structure was optimal in various specifications of his model.  But the underlying thought process is not one that, say, a U.S. presidential candidate would want to endorse too openly in the general election, at least judging from the Republicans' attacks and Obama's very cautious defenses of redistributive policy in the 2008 and 2012 presidential elections.

Weinzierl's work seeks to address this evident disconnect between academic and public discourse.  He wants to revive, and put coherent flesh on the bones of, traditional tax policy frameworks that might better capture people's underlying intuitions about tax policy.  The aim is to improve OIT as a positive theory, whether or not one actually finds these frameworks intellectually persuasive.

Perhaps his best-known paper in this vein is the height tax paper that he coauthored with Gregory Mankiw.  Here the argument is as follows.  From an OIT standpoint founded on utilitarianism, taxing tall people by reason of their height ought to be normatively appealing.  The reason is that height is a "tag," statistically correlated with high income and apparently high earning ability.  Insofar as it is fixed, however, one cannot respond to the height tax (unlike to a high rate of income taxation) through tax planning such as working less or sheltering more.  The fact that people might find this tax normatively unappealing, however, shows that they are not utilitarians, at least in a consistent and thoroughgoing sense.  Fair enough, although how one interprets this is open to further debate.  (E.g., what should be the impact on one's own assessment of the merits of utilitarianism?  Observing that we have conflicting normative instincts is just the start, not the end, of the analysis, since we need not view our own moral intuitions or otherwise as evidence of some underlying moral truth that we choose to embrace at the end of the day.

 In any event, Weinzierl's benefit tax paper for our session is more closely linked to a more recent paper of his tax discusses equal sacrifice theory.  Both the benefit tax paper and the equal sacrifice paper discuss normative theories of taxation that had their academic vogue in the past (e.g., Adam Smith advocated the former, and John Stuart Mill the latter), but that have largely been rejected academically despite arguably remaining important in how policymakers and the general public think about the distributional element in tax policy.

Mill’s equal sacrifice theory involves looking at the total utility loss that individuals suffer by reason of paying taxes, compared to a hypothetical counterfactual baseline in which all public goods from government spending could magically have been provided for free.  Against this background, suppose that it just happened to be the case that the total utility loss from one’s paying tax was constant as a percentage of income.  In other words, suppose there was an elasticity of 1 (or actually -1) for the utility loss per dollar of tax paid, as income rises.  Under this assumption taxpayers typically would experience the same loss of utility from paying (a) $2,500 of tax on $10,000 of income, (b) $25,000 of tax on $100,000 of income, and (c) $25 million of tax on $100 million of income.

Under this of course arbitrary assumption, equal sacrifice would call for adopting a flat rate tax, despite the fact that (under standard assumptions) the marginal utility of a dollar would remain far higher for poor than rich taxpayers.  Suppose that changing the above flat-rate system so that the rich would pay more while the poor would pay less would have negligible labor supply effects.  The change would cause poor individuals to gain more utility than rich individuals lost, indicating that a utilitarian ought to favor it. But equal sacrifice theory would say No: the point is to equalize sacrifice, not to maximize utility.

Equal sacrifice can, however, lead to a very progressive system - or for that matter, a very regressive one - depending on one's view as to how fast or slowly the marginal utility of a dollar declines as one's income rises.

Suppose we don’t worry about the motivation for favoring equal sacrifice as a normative theory.  (Weinzierl’s argument on this score is merely that it appears to have wide intuitive appeal, which I think is probably correct.)  It still faces a distinctive informational problem that utilitarianism does not face.  (Both face the problem of defining and measuring utility.)   Under utilitarianism, you merely need to look at the available policy choices, choosing whichever of them you deem to yield the highest level of social welfare.  Under equal sacrifice, you must evaluate all of these alternative choices relative to a counterfactual – a hypothetical state of affairs that cannot in fact exist.  This of course is the imaginary scenario in which public goods could magically have been provided for free.

Why have I spent so much time on this paper, before getting to the benefits tax paper that we actually discussed yesterday?  (I'll turn to that in my next post.)  Because - in common with Weinzierl - I regard the two topics as closely linked.  Indeed, he and I are not the only ones to think so.  Both equal sacrifice and benefit taxation look to some notion of treating distinct individuals in some sense comparably, as an alternative to the OIT / utilitarian approach in which tax policy seeks to improve general or collective welfare.  Mill, in proposing equal sacrifice theory as a replacement for Smith's benefit theory, not only reached a very similar conclusion (flat tax above an exemption for necessities), but cited Smith as an influence and precursor rather than as the dead past that he wanted to bury.  I'll return at the end of the next post to what I make of this Smith-Mill overlap alongside seeming theoretical distinctness.

It's always best to give as well as receive

Just back from a few nights at a lovely Punta Cana resort, recuperating over spring break from the arduous winter.  The highlights included, not just the beach, pool, and warm weather, but also this fella (a rhinoceros iguana, cousin to Central America's green iguanas) and two of his colleagues, who liked to sun themselves in a particular spot at the edge of some woods.

I got the distinct impression that he likes apple (also, banana).