Cato Papers on Public Policy, Volume 1: 2011-2012

CATO

PAPERS ON PUBLIC POLICY

Volume 1, 2011

CATO

PAPERS ON PUBLIC POLICY

Volume 1, 2011

JEFFREY MIRON

Editor

EDWARD H. CRANE

Publisher

THOMAS A. FIREY

Managing Editor

PETER VAN DOREN

Associate Editor

SALLIE JAMES

Associate Editor

Cato Papers on Public Policy, ISBN-10: 1-935308-48-3; ISBN-13: 978-1-935308-48-5. CPPP is published annually by the Cato Institute, a nonprofit, nonpartisan 501(c) (3) research organization based in Washington, D.C.

Correspondence regarding subscriptions, changes, of address, procurement of back issues, advertising and marketing matters, and so forth, should be addressed to:

Publications Department

Cato Institute

1000 Massachusetts Ave., N.W.

Washington, D.C. 20001

All other correspondence, including requests to quote or reproduce material, should be addressed to the editor.

© 2011 by the Cato Institute.

To subscribe to CPPP, visit www.cato.org/store or telephone (800) 767-1241.

Printed in the United States of America.

Cato Institute

1000 Massachusetts Ave., N.W.

Washington, D.C. 20001

www.cato.org

Publications Director: David Lampo

Marketing Director: Robert Garber

Circulation Manager: Alan Peterson

Cover: Jon Meyers

Contents

INTRODUCTION

Jeffrey Miron

ARTICLES

CAN THE TREASURY EXEMPT ITS OWN COMPANIES FROM TAX? THE $45 BILLION GM NOL CARRYFORWARD

J. Mark Ramseyer and Eric B. Rasmusen

Comment, Efraim Benmelech

Comment, F. H. Buckley

FREE TO PUNISH? THE AMERICAN DREAM AND THE HARSH TREATMENT OF CRIMINALS

Rafael Di Tella and Juan Dubra

Comment, Justin McCrary

COMPETITION AND INNOVATION

Michele Boldrin, Juan Correa Allamand, David K. Levine, and Carmine Ornaghi

Comment, Samuel Kortum

Comment, Andrew Atkeson

LABOR MARKET DYSFUNCTION DURING THE GREAT RECESSION

Kyle F. Herkenhoff and Lee E. Ohanian

Comment, Robert E. Hall

Comment, John V. Leahy

An Introduction to the Cato Papers on Public Policy and Annual Conference

This is the first volume of what will be an annual Cato Papers on Public Policy. The goal of the Papers and the associated annual conference is to produce new, high-quality research on public policy and to make this research available to a broad audience consisting of academics, policymakers, and journalists.

The research intends to fill a gap in the work that addresses the pros and cons of government policies. Academics produce significant research that analyzes public policies, but much of that work is abstract, technical, and not immediately relevant to real-world policy debates. The Cato Papers will evaluate significant economic or social policies using the techniques of modern economics. The papers will be less technical, on average, than a standard journal article, but more technical than a typical policy analysis. In a nutshell, the papers will aim to produce research that employs modern economic methodology but that is firmly focused on what policies are beneficial for the economy and society.

Jeffrey Miron

Editor, Cato Papers on Public Policy

Senior Fellow, Cato Institute

Senior Lecturer and Director of Undergraduate Studies, Harvard University

Can the Treasury Exempt Its Own Companies from Tax? The $45 Billion GM NOL Carryforward

J. Mark Ramseyer

*

Eric B. Rasmusen

ABSTRACT

To discourage firms from buying and selling tax deductions, Section 382 of the tax code limits the ability of one firm to use the net operating losses (NOLs) of another firm that it acquires. Under the Troubled Asset Relief Program, the U.S. Treasury lent a large amount of money to General Motors. In bankruptcy, it then transformed the debt into stock. GM did not make many cars anyone wanted to buy, but it did have $45 billion in NOLs. Unfortunately for the Treasury, if it now sold the stock it acquired in bankruptcy, it would trigger Sec. 382. Foreseeing this, the market would pay much less for its stock in GM.

Treasury solved this problem by issuing a series of notices in which it announced that the law did not apply to itself. Sec. 382 says that the NOL limits apply when a firm’s ownership changes. That rule would not apply to any firm bought with TARP funds, declared Treasury. Notwithstanding the straightforward and all-inclusive statutory language, GM could use its NOLs in full after Treasury sold out. The Treasury issued similar notices about Citigroup and AIG.

Treasury had no legal or economic justification for any of these notices, but the press did not notice. Precisely because they involved such arcane provisions of the corporate tax code, they largely escaped public attention. The losses to the public fisc were not minor—they cost the country billions of dollars in tax revenue. That the effect could be so large and yet so hidden illustrates the risk involved in this kind of tax manipulation. The more difficult the tax rule, the more easily the government can use it to hide the cost of its policies and subsidize favored groups. We suggest that Congress give its members standing to challenge unlegislated tax law changes in court.

* J. Mark Ramseyer is the Mitsubishi Professor of Japanese Legal Studies at Harvard Law School. Eric B. Rasmusen is the Dan R. and Catherine M. Dalton Professor in the Department of Business Economics and Public Policy of the Kelley School of Business at Indiana University.

We thank William Allen, Andrew Atkeson, Frank Buckley, Michael Doran, Sally James, Victor Fleischer, Michael Schler, and participants in seminars at the online Cyprosia, the Cato Institute, and the Harvard Law School for their many comments, whether positive or negative. We do not imply that any of these generous readers agree with our conclusions.

Can the Treasury Exempt Its Own Companies from Tax? The $45 Billion GM NOL Carryforward

Dona clandestina sunt semper suspiciosa.¹

1.  INTRODUCTION

Year after year, General Motors lost money—enormous sums of money. It designed cars. It built cars. But no one wanted to buy the cars. Over time, it accumulated huge operating losses (net operating losses, or NOLs). The tax code let GM carry forward these NOLs into the future. It let the firm save the losses for that day in the future when it would once again sell cars that people wanted.

The day never came. Instead, in June 2009 GM (call it Old GM) declared bankruptcy. It filed under Chapter 11 of the Bankruptcy Code and sold its assets to a new shell (New GM) in a transaction governed by Section 363 of the Code. Old GM’s shareholders lost their investment. They did not receive stock in New GM. Instead, Old GM’s creditors became New GM’s stockholders: the U.S. Treasury (with 61 percent), the auto unions, and Canada swapped debt claims against Old GM for equity stakes in New GM. Other Old GM creditors acquired a 10 percent stake in New GM as well. In the fall of 2010, the Treasury re-sold a large amount of its New GM shares to the public, cutting its share to 26 percent.

New GM has the factories, offices, designs, and some of the workers that Old GM had. It also acquired some $18 billion worth of Old GM’s NOLs.² New GM could not use them to reduce its tax liability immediately, since it was losing money. But in 2010, New GM did turn a profit and presumably will use its NOLs to avoid corporate income tax on that profit (Bunkley 2011).

Ordinarily, when one company buys another’s assets, it does not acquire its tax losses too. But the sale from Old GM to New GM qualified as a tax-free reorganization under Sec. 368 of the tax code: neither Old GM nor New GM incurred a tax liability, New GM entered Old GM’s assets on its books with Old GM’s adjusted basis, and New GM acquired Old GM’s NOLs.

The problem involved Treasury’s plans to sell the shares it took in New GM. If the combined equity stake of any group of shareholders in a loss corporation like New GM climbs by more than 50 percentage points, Sec. 382 of the tax code limits the firm’s ability to use those accumulated NOLs. Given Treasury’s large stake in New GM, if it sold its entire stake to the public, those new owners would raise their combined interest by 50 points. New GM would then lose its ability to avoid taxes on future income.

To solve this problem, the Treasury issued a series of notices. The Sec. 382 rules, it declared, would not apply to itself. When it sold its shares in New GM, the new owners might increase their ownership stake by 50 percentage points, but they would not trigger the Sec. 382 limits. The tax code offered no exception for government owned shares, and the Treasury did not purport to find one. Instead, it just declared that the law did not apply.³

The notices also apply to two other companies, AIG and Citigroup. Both of these companies had ownership changes over 50 percent as a result of the Troubled Asset Relief Program and would ordinarily, as in bankruptcy, lose their NOLs. If they retain them, that reduces the apparent (but not real) cost of the bailout because the government can resell its shares at a higher price.

Through these notices, Treasury accomplished two highly political goals:

•  It disguised (by billions of dollars) the true cost of the bailouts of GM and other firms.

•  It routed funds (again, several billion dollars) to the administration’s supporters at the UAW.

Ordinarily, if an administration wildly misstates the cost of its policies or routes public funds to its friends, the press notices and complains. In this case, it did not. The press missed the manipulation precisely because it involved such a complex and highly arcane provision of the tax code. The more obscure the law, in other words, the greater the risk of political manipulation: precisely because its strategy involved such an abstruse corner of the law, the administration was able to hide its politicized policies from the public.

We do not address the wisdom of the bailouts themselves. Neither do we ask whether firms should be able to carry forward operating losses, whether they should be able to reorganize tax-free, or why the United States has a corporate income tax at all.⁴ These are all interesting questions, but we have quite enough to do addressing the topic of selective tax relief through executive decree. Rather than explore these larger questions, we focus on the propriety of the Treasury’s manufacturing a tax break to distribute and hide government largesse. More generally, we focus on the wisdom of giving a president the ability to invent a tax deduction for his political supporters without a need to answer to the courts or Congress.

1.1  The Bad Man and the Law

Recall Justice Holmes’s description of the law as being the prediction of the Bad Man about whether a judge would stop him:

If you want to know the law and nothing else, you must look at it as a bad man . . . who cares only for the material consequences which such knowledge enables him to predict, not as a good one, who finds his reasons for conduct, whether inside the law or outside of it, in the vaguer sanctions of conscience. . . . If we take the view of our friend the bad man, we shall find that he does not care two straws for the axioms or deductions, but that he does want to know what the Massachusetts or English courts are likely to do in fact. I am much of this mind. The prophecies of what the courts will do in fact, and nothing more pretentious, are what I mean by the law. (Holmes 1897)

If a president is Holmes’s Good Man, he will obey the Constitution because it is the Constitution. The Treasury gave General Motors an illegal tax break. As a Good Man, he will read our article, feel remorse, and fire everyone involved.

If a president is Holmes’s Bad Man, on the other hand—and public choice theory suggests that it is Bad Men who have the best chance of being elected—he will obey the Constitution only when a court can make him obey it.⁵ If he hears of our article, he will ignore it. As a lawyer, he knows that nobody has standing to challenge someone else’s tax benefits in court. Thus, his prophecy about what a court will do is easy: nothing. The courts will reject any challenge for lack of standing, whatever the merits of a claim might be.

Only potential bad publicity would worry a Bad Man president. But publicity he can skirt by giving the funds through opaque provisions of the tax code. Publicity he can skirt by (take a deep breath) declaring an exemption from the application of Sec. 382 of the tax code to limits on carryforwards of NOLs following a sale under Sec. 363 of the Bankruptcy Code that uses preferred stock, credit bidding, and warrants by one company named GM to a different company also named GM. If the administration gave a billion dollars in cash to its supporters, the press would notice. If it gives it through the obscure details of the corporate tax code, the press will fall asleep.

In the article that follows, we explain the intricacies of the tax break (Section 2). We discuss the law involved (Section 3). If you think all presidents are Good Men, you may stop reading at that point. After all, following the Constitution is just a matter of understanding it. We explain it, you understand it, end of story. Lest some presidents be Bad Men, however, we conclude by exploring procedural reforms Congress might adopt to prevent a recurrence of what happened with GM.

2.  WHAT HAPPENED

General Motors was a public corporation with much unsecured debt, including $21 billion owed to the UAW Trust on behalf of retired workers and $27 billion owed to bondholders. None of these stakeholders was senior enough to see much return if the company liquidated in pieces. Probably, none would see much return even if the firm found a buyer for the whole company.

The senior creditors were a diverse lot. The U.S. Treasury had a secured interest in $19.4 billion from TARP loans and $30.1 billion in other loans. The Canadian government held secured claims of $9.2 billion. Government senior debt thus totaled $58.7 billion. Private creditors held another $5.9 billion in secured loans.

GM filed for bankruptcy under Chapter 11 of the Bankruptcy Code. To restructure its finances, it then negotiated a sale under Sec. 363 of the Code. For this transaction, it formed a new shell, New GM. Old GM then sold its assets to New GM. In exchange for its $21 billion unsecured debt to Old GM, the UAW Trust received 17.5 percent of the common stock of New GM, $6.5 billion in preferred stock, and $2.5 billion in debt. In exchange for their $27 billion of unsecured debt, the other junior creditors received 10 percent of the common stock of New GM and warrants for another 15 percent. The private secured creditors (the $5.9 billion claim) were paid in full. The Canadian government received 12 percent of the New GM common stock, and the U.S. Treasury received interests detailed shortly below.

To consider the stakes involved, note that in December 2010, New GM had stock worth $54.4 billion and liabilities of $12.9 billion (Ceraso, Moffatt, and Pati 2010), for a total asset value of $67.3 billion. In effect, the sale price in the 363 offer was:

•  $58.7 billion in senior credit claims,

•  $5.9 billion paid to private secured creditors,

•  $5.4 billion in stock (10 percent of $54.4 billion), and

•  a portfolio of harder-to-value warrants.

This yields a total of $67 billion plus warrants (Warburton 2010, p. 536).

Apparently, the 363-sale buyers paid $67 billion plus the warrant value for assets worth $67.3 billion. That seems a remarkably high price considering that no other bidder loomed on the horizon. The bankruptcy judge deserves praise for extracting so much value for Old GM’s creditors.

This $67.3 billion in asset value is not the net benefit to the 363-sale buyers or the senior creditors, however. That benefit depends on who owns the new GM equity and debt. Old GM’s private secured creditors received $5.9 billion in cash for their $5.9 billion in debt. The Canadian government gave up its $9.2 billion in Old GM debt but took a 12 percent stake in the common stock (worth 0.12 × $54.4 billion = $6.5 billion) plus $0.4 billion in preferred stock and $1.3 billion in debt in New GM—for a total value of $8.2 billion.

The most glaring anomaly involved the UAW. The union’s trust gave up unsecured claims of $21 billion and received:

•  17.5 percent of the stock of New GM worth (0.175 × $54.4 billion = ) $9.5 billion,

•  $6.5 billion in preferred stock, and

•  $2.5 billion in debt,

for a total of $18.5 billion. Given that the UAW Trust had been a junior creditor, this was a very good deal. By contrast, the other unsecured creditors gave up claims of $27 billion and received only 10 percent of the common stock and warrants.

Recall that the U.S. Treasury held secured debt totaling $49.5 billion. In exchange for its claims, it took 61 percent of the stock in New GM (stock worth 0.61 × $54.4 billion = $33.2 billion), $2.1 billion in preferred stock, and a $6.7 billion debt claim against New GM. All told, it received compensation of $42 billion.

Focus on the U.S. government. Through the Sec. 363 sale, it—apparently—lost ($49.5 billion − $42 billion = ) $7.5 billion. Anyone who loses only ($7.5 billion ÷ $49.5 billion = ) 15 percent on a $49.5 billion loan to a failing firm does well indeed. Yet appearances deceive. The government also gave GM investors $45 billion in NOLs. If the 363 sale had not gone through or the sale had been made to some outside buyer, these NOLs would have disappeared. The book value of these NOLs is $18 billion.

To be sure, Treasury was giving tax breaks partly to itself, and the book value of the NOLs exceeds their market value since it would take some years before GM could exhaust them. If the market value of the NOLs were, say, $12 billion (a little under the estimate of the stock analysts that we cite in Section 2.1 below), then that $12 billion was incorporated into the $54.4 billion equity value of the New GM, and we have overestimated the overall value of the deal for the Treasury. Of its $33.2 billion in stock, $7.32 billion ( = 0.61 × $12 billion) was a tax gift to itself.

More simply, consider the $12 billion worth of NOLs an additional loss to the Treasury. In effect, the Treasury lent GM $49.5 billion and lost ([$7.5 billion + $12 billion] ÷ $49.5 billion = ) 39 percent. If only Treasury could have inserted a further secret $20 billion of assets into New GM, New GM’s stock price would have been so high that Treasury would have appeared to make a profit from the entire affair.

2.1  As GM Told It

Here is how GM describes its tax situation:

We recorded valuation allowances against certain of our deferred tax assets, which under ASC 852 also resulted in goodwill. (General Motors 2010, p. 82)

In July 2009 with U.S. parent company liquidity concerns resolved in connection with the Chapter 11 Proceedings and the 363 Sale, to the extent there was no other significant negative evidence, we concluded that it is more likely than not that we would realize the deferred tax assets in jurisdictions not in three-year adjusted cumulative loss positions.

Refer to Note 22 to our audited consolidated financial statements for additional information on the recording of valuation allowances. (General Motors 2010, p. 138)

Table 1 from New GM’s securities filings (p. F-121 of its Form 8-K) shows that New GM claimed to inherit over $18 billion in tax carryforwards from Old GM.⁶ Stock analysts wrote:

We calculate an NPV of GM’s deferred tax assets at $17.2bn of which $4bn is related to pension contributions and more than $13bn related to accumulated NOLs and tax credits including R&D credits. (Morgan Stanley 2010)

and

Via a special regulation, GM’s highly valuable US tax assets (worth $18.9B in the US at 09-end) were left intact. . . . Our