All posts by Eric Posner

Is Trump normalizing?

Seems like it. Since Charlottesville (almost a month!) or maybe the Phoenix rally (two weeks!), Trump has not made any big blunders. No super-embarrassing leaks. No obvious efforts to obstruct justice. Trump’s lawyer, Ty Cobb, says all the right things (well, almost), giving a strong impression that the administration has taken a cooperative attitude toward Mueller’s investigation. The DACA revocation seems to have been rolled out in a professional way, in contrast to the travel ban blunder at the start of the term. Am I wrong to think that Trump’s tweets have mellowed. Fewer personal attacks? Threats against judges? Attacks on the press? Or is it that since Trump has followed through on none of this threats, his words no longer seem ominous, but just background noise? Can only a law professor appreciate the irony that the Trump administration believes, or at least says it believes, that President Obama lacked the power to issue DACA. Is Trump, the supposed dictator, actually trying to weaken the presidency?

Suppose I am right (and I don’t know that I am)—what could explain this turn of events? The answer seems obvious. Trump has experienced the most spectacularly unsuccessful start of a presidency since I don’t know when. He has managed to provoke nearly everyone who might help him or at least not stand in his way—Republican senators, his cabinet officials, the intelligence community, the FBI, the press, the courts—and unsurprisingly has accomplished almost nothing. It seems possible that he has learned something, and that the staff changes in the White House reflect what he has learned. It is predictable that as the amateurs in the White House give way to experienced professionals, the administration will begin to resemble something like a professional organization. Trump has acquiesced in all this; does he deserve credit for doing so?

A parable for Google

There was once a village nestled in a small valley where the people lived long and prosperous lives. The local deity, unlike so many of the other gods in that land, smiled on the people and demanded only that they remember him in their prayers. But then one day a change came about. The people were hit by wars, plagues, and famines that nearly destroyed the community. At long last, the village, sensing it had displeased its god, sent its leading citizen up the mountain to seek enlightenment from the god and a path forward.

When he returned, he told the assembled elders that the god appeared to him in a fevered vision and explained that he was angered because the people had started using the god’s name in vain. This came as news; no one knew the god had a name. But it turned out that the god did, and that some subtle change in the evolution of the language had introduced the god’s name into everyday communication without anyone realizing it. And he forbade the people to use it.

“Well, what is his name?”, the people asked. And the man who returned from the mountain smiled sorrowfully and said, “I cannot say it lest I anger him.” But through gestures and circumlocutions, the man made the people understand that the god’s name coincided with the word for a farming implement, or newly imported type of jewelry, or some such thing—the identity of the actual item, and the word itself, is lost to time. The people duly avoided uttering the sacred word upon pain of execution.

Prosperity returned, and continued for generations. But with the passage of time, disagreement arose as to which item it was that corresponded to the god’s name. Yet the memory of the travails that beset the community remained powerful, and the fear that someone would inadvertently (or perhaps mischievously) break the taboo remained ever present. The elders consulted the sacred texts, and various theories arose as to the identity of the forbidden word. The very oldest villager recalled that the word began with an “a” or possibly an “s,” and so it was decreed that no words beginning with those letters may ever be uttered. A prominent theologian insisted that the word must have had three syllables, and so three-syllable words were struck as well. As theories multiplied, the governing council, out of prudence more than conviction, struck more and more words from the language. But because of the difficulty of eliminating words without mentioning them, the villagers were often confused, and as a result tried to use as limited a vocabulary as possible, so greatly did they fear the god’s wrath.

The people of the village seem to have satisfied their god, as they remain healthy and even prosperous. By disgorging the forbidden fruit, they obtained readmission to the Garden of Eden. But they also live in fear of causing offense to the god, and so watch carefully the few words that remain to them. Under the pall of the proscription, many treasured community activities stopped long ago. To tell the truth, a group of people more prosperous and stunted is hard to imagine.

Probability that Trump will be impeached next year

1. Adverse public opinion

2. Bad economy

3. Hostile media

4. “Incontrovertible evidence” of illegal activity

In his book about the Clinton impeachment, Bob Woodward quotes Senator Fred Thompson, a Republican who voted against removal, who cited the four factors above as necessary, and possibly sufficient, for impeachment and removal of a president. Thompson was generalizing from one (Nixon) or possibly two (Nixon and Johnson) data points, but his theory turned out to hold for Clinton as well.

The relationship between these factors is complicated; they are surely not independent of each other. Nixon enjoyed extremely high standing in the public (>60%) coming off his reelection, bolstered by significant foreign policy accomplishments. It was the gradual disclosure of increasingly incontrovertible evidence of illegal activity that destroyed his standing among the public. The bad economy and hostile media surely did not help. And while the media had always been hostile to Nixon, as it has been with Trump, the disclosure of evidence made it more hostile—as stalwart Republican papers like the Chicago Tribune eventually turned against him as a result of the disclosures.

Clinton, by contrast, remained popular throughout the revelations that led to his impeachment, and throughout the impeachment proceedings as well (#1). The economy was strong (#2). And while there was incontrovertible evidence of perjury and obstruction of justice (#4), the underlying behavior—adultery—was not considered very serious by the public. People disapproved of Clinton’s behavior but did not think a president should be impeached over it. We should amend #4 to state “illegal activity beyond the cover-up itself”—burglary, espionage, violation of people’s rights, rather than adultery. Clinton also faced a hostile media, though maybe not as hostile as the media faced by Nixon and Trump.

Trump is the least popular president of the modern era. While other presidents have done worse in the polls from time to time, no president has done so badly in the honeymoon period at the start of the presidency and during a period of economic prosperity and international peace. Trump’s approval rating is worse than Nixon’s even after the Senate Watergate hearings began in May 1973. At that time, less than 20% of Americans supported impeachment of Nixon. That number would not rise above 40% until almost a year later, after the Saturday Night Massacre and additional revelations. More than 40% of the public already wants Trump to be impeached.

So much for #1. We can also pass over #3. If Thompson is right, the only question is if (and when) the economy tanks, and if (and when) the “incontrovertible evidence of illegal activity” emerges. Economists estimate a 10-20% probability of a recession in 2018. If you think that the probability of incontrovertible evidence coming to light is 100%, then we can understand why prediction markets suggest a 20% probability of impeachment in 2018. The probability of impeachment is just the probability of recession (to a first approximation).

But that leaves the question of #4. At what point is evidence “incontrovertible” (and does it really need to be?), and how serious does the illegal activity need to be (and does it even need to be illegal?)? For both Nixon and Clinton, the evidence really was incontrovertible. Will we ever reach that stage for Trump? If Trump or top aides actually colluded with the Russians, and aides testify under oath, will that be enough? Do we need a memo? Audio? Video? What if Trump merely engaged in shady financial transactions with Russians long before he ran for election? Hard to say.

One thing that places Trump in a category different from Nixon and Clinton is that no one thought Nixon or Clinton was incompetent. Amoral, yes; impulsive, in Clinton’s case; but not incompetent. This is an additional factor to take into account, one that did not even occur to Thompson. If Trump convinces us all that he welcomes the prospect of a nuclear war with North Korea, then one way or another he’s going to get pushed out.

A better way to protect Robert Mueller

Daniel Hemel and I discuss the Tillis-Coons and the Graham-Booker bills, which seek to protect Mueller from being fired. We argue that the bills need to be strengthened. Trump can harass Mueller in many ways short of firing him (or getting him fired).

It might seem that things have calmed down since Kelly was put into place, and maybe we no longer need to worry about Mueller. But in fact we’re in the eye of the storm. When Mueller digs into Trump’s alleged financial relationships with Russian criminals, Trump may well risk the political blowback to stop Mueller in his tracks.

Time to talk about impeachment?

Until now, impeachment was off the table for three reasons. First, Republicans control the House and Senate and as long as they could put up with Trump, they would never start impeachment proceedings let alone convict. Second, in the immediate aftermath of the election, impeachment would have been seen as a repudiation of democracy, and hence as a kind of coup, even if the formalities of the procedure had been complied with. Third, there were no “crimes or misdemeanors” that could be pinned on Trump, at least none that were sufficiently clear to warrant the disruption of impeachment.

All this is changing before our eyes. Republicans in Congress are gradually distancing themselves from Trump. As long as he can’t deliver political victories like health care reform, they gain nothing from keeping him in office. And as Trump continually proves himself erratic, untrustworthy, and incompetent, the prospect of a long string of political defeats is becoming inescapable. We are a long way from this point of time (perhaps Kelly will delay it for a few days), but an inflexion point has been reached.

Moreover, as the election recedes farther into the past, impeachment can be based on Trump’s post-election failures rather than on disagreement with the people who voted for him. While the voters may have acquiesced in Trump’s conflicts of interest and lack of experience, the election cannot absolve him of his post-election conduct, including the continuing pattern of obstruction of justice, the chaotic administration of the White House, and the weird and possibly criminal involvement with Russia. If an economic downturn occurs or a crisis is mishandled, then the impeachment drumbeat will begin.

And finally, obstruction of justice provides the “crime or misdemeanor,” sanctified by the Nixon and Clinton precedents. The pattern of obstruction is by now so overwhelming that it can hardly be denied. That said, the House and Senate will need to decide whether obstruction alone warrants impeachment and removal, or some additional crime is necessary. The Clinton impeachment failed, in part, I think, because the cover-up was the only crime; the underlying behavior was never found to be illegal. Unless evidence of Russia collusion or other illegal or deeply immoral behavior is found and proven in a public way, I suspect that impeachment will not take place, even if Democrats obtain a majority in the House in 2018.

The probability of impeachment remains low, but not as low as it used to be.

The obstruction of justice statute as a check on presidential power

Almost without anyone noticing, a group of obscure statutes, mostly from the nineteenth century, have become, in the hands of the FBI and Justice Department, a significant check on presidential power. Don’t believe me? Consider our last nine presidents.

1. Nixon. Resigned after articles of impeachment drafted by House Judiciary Committee allege obstruction of justice.

2. Ford. Pure as the driven snow.

3. Carter. As innocent as the morning dew.

4. Reagan. Allegations (never proven but investigated) that he engaged in obstruction of justice in connection with the Iran-Contra scandal. Aides investigated, charged, and (in a few cases) found guilty (though rescued in various ways).

5. H.W. Bush. No investigation; but a plausible claim has been made that by pardoning Iran-Contra defendants, Bush blocked the investigation into his own role.

6. Clinton. Impeached for obstruction of justice; later sanctioned for contempt of court, based on what was effectively obstruction of justice.

7. W. Bush. FBI investigated Bush administration’s firing of U.S. attorney in New Mexico, based on obstruction-of-justice argument that it retaliated against him for failing to indict a Democratic state elected official before an election.

8. Obama. Clean as a whistle. (Or was he? Did he obstruct justice by publicly declaring that Hillary Clinton’s use of private email servers did not endanger national security. You be the judge. But my criterion is whether there was an official investigation.)

9. Trump. I would crash my server if I tried to describe all of his actions that would qualify as obstruction of justice based on these precedents. But it is worth mentioning that the public humiliation of Sessions counts as obstruction of justice, to the extent they are intended to spur him to get rid of or constrain Mueller, and also leaks, reported just today, suggesting that the president plans to fire Sessions. Also, if Trump pushes the Justice Department to investigate Comey and Clinton, yes, that too, to the extent that Trump hopes to block investigations into his Russia ties, or to retaliate against political opponents, as Daniel Hemel and I argue in Slate.

That’s 5 to 7 of 9 presidents (or their aides) embroiled in an obstruction of justice scandal, depending on how you define “scandal” (I’m inclined to include H.W. Bush and exclude Obama, but won’t insist on this.)

And also see our latest paper, which goes into more detail.

Is there a coming constitutional crisis?, #2

In my previous post, I suggested there isn’t, but maybe I wasn’t being sufficiently imaginative. It is possible that a constitutional crisis would erupt if Trump fires Sessions, or forces Sessions to resign (p=0.9). But how, exactly? A political crisis, yes. Many die-hard Trump supporters appear to be die-harder Sessions supporters. Trump’s already low popularity could plunge, putting his agenda into even greater peril than it is already in. But nothing stops Trump from appointing a new attorney general, and then tweeting and playing golf until 2021.

I want to maintain the notion that a constitutional crisis takes place when the government can no longer operate or faces significant disruptions because of disagreements among key public officials about what the constitution requires; broader definitions just seem to be gestures at drama. What does seem possible is that the firing of Sessions would create a breach between Trump and Republican Senators, many of whom are strong Sessions supporters. They could retaliate by refusing to confirm a new attorney general, or other nominees, for good, or (more likely) unless Trump appoints to the position someone who has a reputation for independence. This would mean that the standoff between Trump and the Justice Department would continue for the foreseeable future.

The underlying constitutional norm at stake is the independence of the Justice Department. This might seem like an odd claim: isn’t the Justice Department in the executive branch, which the president heads? But it seems increasingly clear that a new constitutional norm has arisen, possibly traceable to Watergate, according to which the Justice Department is semi- or maybe completely autonomous, much like the judiciary. You might think of Trump as a kind of unlikely James I to Sessions’ even unlikelier Lord Coke, who resisted the king’s efforts to subordinate the judiciary to the executive, and helped establish the principle of judicial independence. But for Trump, the relevant norm is not judicial independence but prosecutorial independence. Trump is trying to reverse the norm, but it has many defenders in both parties.

There is logic to all this. The principle of judicial independence is entrenched; not even Trump can touch it. But our system has maneuvered around this principle by granting increasing power to prosecutors, which they can use to harass people with investigations and selective prosecutions based on general laws that are otherwise not enforced. This power has grown to such an extent that it can easily be abused by a powerful president if he were really to exercise control over prosecutors, as the traditional legal sources maintain he has. The norm of prosecutorial independence, if one can call it that, has evolved over the last forty years to cabin the president’s discretion and prevent this kind of abuse. Trump is the first president to challenge it.

Is there a coming constitutional crisis?

Keith Whittington thinks there would be a crisis if Trump fires Mueller and then the Republicans in Congress refuse to hold Trump to “account,” which I assume means impeachment or something like it. I don’t see it. If Trump fires Mueller, and Congress doesn’t impeach Trump, then the outcome might be unfortunate but it is not a crisis since government would continue to operate as it always has. A crisis requires more than a constitutional violation (if that is what the firing of Mueller would be, but that is hardly clear). It requires an impasse of the sort that would arise when government officials receive conflicting orders from different sources claiming constitutional authority and are unable to resolve these conflicts in a consistent and predictable way, based on legal materials.

Such an outcome is certainly imaginable, but is it likely? If Trump fires Mueller, Mueller will step down. He won’t claim that the firing is null and void. Nor will anyone else. It is remotely possible that Congress would impeach Trump but I doubt it, and even then, there is no crisis unless Trump is actually convicted and then refuses to leave office.

We can imagine a more plausible path to a constitutional crisis, one that does not depend on the Republicans in Congress to abandon their political interests. Let’s map it out, with (conditional) probabilities.

1. Trump is found to have obstructed justice or committed some other serious crime (p = 0.5).

2. Special counsel Mueller brings charges against Trump in violation of apparent Justice Department policy (p = 0.2).

3. A court agrees to hold a trial rather than issue continuances until Trump leaves office (p = 0.3).

4. Trump is found guilty by a jury (p = 0.5).

5. Trump is sentenced to jail rather than required to pay a fine (p = 0.1).

6. Trump loses on his various emergency appeals (p = 0.5).

7. Trump refuses to report to jail and orders the Secret Service to turn away federal marshals who come to arrest him (p = 0.9).

8. The Secret Service obeys Trump rather than their boss, the Secretary of the Department of Homeland Security (p = 0.1).

I say at this point a genuine constitutional crisis has begun. But p = 0.000675. Does anyone think my probability estimates are too low, enough to make a difference? Or that there is a shorter path to crisis?

If Trump pardons, it could be a crime; with further meta-legal implications

Daniel Hemel and I argue that if Trump pardons family members and aides who committed crimes in connection with Russia, then Trump could be guilty of obstruction of justice. But if Trump is guilty of obstruction of justice, can’t he also pardon himself of the obstruction crime? Some legal scholars think that a president can self-pardon. Nixon apparently considered it at his lawyers’ suggestion. The argument, boiled down to its essentials, is that the Constitution doesn’t say he can’t.

I am skeptical, but even if the president can self-pardon, Trump faces a further problem that someone like Kurt Gödel would appreciate. The problem for Trump is that if he pardons himself, and our theory is correct, then the self-pardon is itself an obstruction of justice (an obstruction of the investigation and possible prosecution of Trump himself for pardoning his family members and aides). Trump would need to issue yet another self-pardon to relieve himself of criminal liability for the first self-pardon. He could do that, of course, but then there is a second- (or is it third-?) order crime—the obstruction of the investigation into whether Trump has committed a crime by pardoning himself.

Trump can continue self-pardoning himself at higher orders until the moment he leaves office. But the final self-pardon itself will be a potential crime like all the others, and there is nothing Trump can do about that. Whatever the case, he should be prepared to sign a lot of documents.

Über die Kraft der 140 Zeichen

Auf Deutsch. Ich denke, dass meine Twitterthesen ein Nerve gestrucken haben. How do Germans even fit their language into what they call a Kurznachrichtendienst (“short-news-service”)?

In the interest of fairness and balance, I quote below a comment from the Frankfurter Allgemeine website, translation brought to you by Google:

Professor obviously follows Twitter's boring and humorless normals. Twitter is incomprehensibly funny outside these areas. For example, when once again a hashtag is brought to the head, which in itself simply represents a parody and a troll of the Zeitgeist and the normal news, as for example the last-week #HeterosexualPrideDay - gorgeous. Or the political equivalent of the flying spaghettimonster and the memes made from it. Better than anything on TV and in newspapers where "wit" is offered.

Twitter is a slot machine composed of humans

Engineers design slot machines to be addictive. They exploit the phenomenon, first identified by B.F. Skinner, known as the “variable schedule of rewards.” It turns out that behavior is most effectively reinforced not when it is consistently rewarded but when there is a random element. When you pull the lever of a slot machine, you might win but you might not. Engineers figure out just how many times you need to win to keep you going, and build that into the algorithm. The algorithm will give you a large reward (the jackpot) in order to keep you going. But small rewards at more frequent intervals are necessary because too many jackpots would degrade their value (as well as bankrupt the casino). “Near misses” (the jackpot symbol is visible but not on the “payline”) can be constructed on the slot machine through algorithmic manipulation—they fool people though we all know abstractly that from the algorithm’s perspective a near win is no different from any other loss. But they are psychologically effective; that is why they appear more often than random.

Twitter works on the same principle. When you send out a tweet, the reward is variable. Depending on the number of followers you have, your tweet will normally not be retweeted or retweeted only occasionally. From time to time, you will hit the jackpot—a media personality (or his or her factotum) retweets it to thousands of followers, generating a cascade of retweets and likes. At other times, you receive a small reward of a few or dozens of retweets or likes. I suspect the power curve strongly resembles the slot-machine version.

Twitter builds addictive loyalty in other ways. The more followers you have, the less you want to lose them. As you scroll down through the increasingly boring tweets, you’re informed that new ones await you at the top of the screen. And then the tweets you receive also follow the variable reward system. Most are boring, some are interesting, a few are great.

Slot machines are entirely self-contained; that is why they are so effective at ruining people. Twitter is a decentralized form of slot machine, one that uses nominally autonomous human beings as its vectors. That is also why Twitter is not as addictive as slot machines. People aren’t as reliable as algorithms are. But I suspect Twitter’s engineers have caught on to this problem. Twitter itself tries to encourage people to write good tweets rather than bad tweets. I have often wondered whether Twitter creates artificial followers for people, or randomly gives prominence to their tweets. If not, it should (that is, from the standpoint of maximizing addiction). Twitter engineers should figure out a way to construct a “near miss” as well.

I sense that people are catching on that social media is a con. Maybe not all of it. Instagram—definitely yes. There, the addictive quality is reduced to its purest essence, with people simply posting to get likes, and liking to get more likes, and everyone caught in a Red Queen-style competition to nowhere. Twitter has much to learn from Instagram. Facebook, it’s not so clear. But the important point is this: a social medium does best by addicting its users, and so there are immense incentives for it to do that. Backed by vast sums of money, limitless data, and huge engineering talent, sooner or later these companies will figure it out. If you are not yet addicted, just wait for the next update.

Some people have said to me that they find value in Twitter. The major claim is that they receive useful information from reliable sources very quickly. Of course, if every element of Twitter were literally valueless, it would not have gotten off the ground. A slot machine, too, pays out once in a while. But unless you are a journalist with a breaking-news beat, ask yourself how important is it that you get information two minutes after an event rather than a day. Many other services exist to supply people with specialized knowledge that they need for their work. Bloomberg BNA’s Law Week emails tell me what I need to know about the latest law news. They couldn’t be less addictive.

I suspect the real reason that the chattering class—the academics and journalists who filled my feed—uses Twitter is that they are afraid of falling behind. If everyone tweets but me, am I no longer important? Does my opinion no longer matter? In this sense, Twitter is the Instagram for intellectuals—a negative-sum game that users play purely because everyone else does.

Twenty Theses about Twitter

1. People sign up for Twitter for two reasons: to obtain information and to exert influence.

2. Twitter serves these functions poorly. If you want information about a specific topic, a Google search is a more efficient way to obtain it. If you want information about current events, you do better by reading a newspaper.

3. Twitter provides information poorly because tweets are mostly driven by the latest outrage and are hence redundant. The rare tweet that contains an interesting or unusual idea is lost in the cataract.

4. Twitter is a poor device for exerting influence because of #5.

5. No tweet has ever persuaded anyone of anything.

6. Twitter’s real function is not to help people obtain information or exert influence.

7. Twitter’s real function is to enable people to obtain validation for their beliefs.

8. People send tweets with a single overriding purpose: to get the tweet “liked” or retweeted.

9. When your tweet is liked or retweeted, you enjoy a dopamine surge.

10. It doesn’t matter why the tweet is liked or retweeted, or even if the person on the other side read your tweet. You enjoy a fleeting illusion of mastery.

11. People retweet tweets that validate their own beliefs.

12. For this reason, the most effective tweet is a clever formulation of a view that everyone already believes. If one lacks cleverness, forcefulness provides a second best.

13. Tweets are either snide or outraged.

14. The effortlessness with which one obtains a dopamine response results in excessive use and a weakening of the response over time. Hence Twitter’s addictive quality. People increase Twitter usage in order to maintain a constant dopamine response.

15. Unfortunately, people might respond negatively to your tweet. When that happens, the self feels threatened, stress levels rise, and the organism engages in fight-or-block, resulting in either case in a form of infantile regression.

16. In the non-virtual world, successful people take care to keep up impressions, for example, they avoid making controversial statements to friends, colleagues, and strangers except when unavoidable, and even then do so in a carefully respectful way.

17. In Twitter, the same people act as if their audience consisted of a few like-minded friends and forget that it actually consists of a diverse group of people who may not agree with them in every particular on politics, religion, morality, metaphysics, and personal hygiene. Hence tweeting becomes a source of misunderstanding and mutual hostility. The Twitter paradox is that one seeks solidarity but is constantly reminded of one’s solitude. Fortunately, there is always the mute button.

18. Without realizing it, people who use Twitter damage the image of themselves that they cultivate in the non-virtual world.

19. The sense of validation that Twitter provides is as a potato chip is to a meal. A Frankfurt school theorist would say that the tweet is a commodified form of social engagement in Late Capitalism. Its effect is to alienate its users while immersing them in advertisements.

20. But Twitter doesn’t even make money for the capitalist class. It’s a black hole of value-destroying technology for all concerned.

Will the investigation of Trump backfire politically?

While I have argued that the investigation might succeed in a legal sense, it could backfire in a larger political sense. Imagine the following scenario. Mueller finds that some of Trump’s aides engaged in hanky-panky—talking to the Russians when they shouldn’t have, profiting from dubious financial transactions involving the same. But Trump himself did not call up Putin and offer to lift sanctions in exchange for help defeating Hillary Clinton in the election. The extent to which Trump benefited, financially or politically, from his aides’ illegal or dubious behavior is hopelessly ambiguous. A few of the aides go to jail or pay fines or walk the plank of community service.

Then Trump would declare victory, and politically he would be right. The contrast between the vast amount of media attention and public expense, and the smallness of the results, will give an impression of bureaucratic overreach. It will feed all kinds of narratives about the impotent fury of the deep state, the hostility of the media, and the sore loserness of the Democrats who cheered on the investigation. Remember the impeachment of Bill Clinton?

What if Mueller reports that Trump “obstructed justice” by firing Comey or lying to investigators? An obstruction of justice charge against Trump, if not accompanied by persuasive allegations that he engaged in an underlying crime of serious nature, will compound the damage to Mueller and the Justice Department.

The government takes obstruction of justice very seriously, even when the target is innocent of the underlying crime, because obstruction wastes government resources. But the public can be ambivalent.

The reason is simple. Whenever the government charges a person innocent of an underlying crime with obstruction of justice, there are two possible interpretations of events. First, the government’s preferred interpretation—that the person acted irresponsibly by choosing to conceal facts that could have helped the government find the actual culprit or close an expensive investigation. Second, that the government acted irresponsibly by harassing an innocent person who hid facts merely because they were private and embarrassing, or understandably sought to avoid squealing on friends or associates.

I suspect for the middling sorts who have not made up their mind about Trump and could possibly be swayed, the second interpretation will be more plausible than the first. A president will often have even better reason to keep secrets than ordinary people do—reasons of state, or even political reasons if embarrassing information may be exploited by opponents but are of little importance for the country (think of Clinton again).

Trump and his surrogates have laid the ground for this interpretation by accusing Mueller of bias. Mueller, because of his position, can’t defend himself. When he issues a report, if he does, or brings charges, he will need to let his findings speak for themselves. If the facts are not overwhelmingly damning, they will be ground to dust by the Trump obfuscation machine. Ken Starr was more damaged by the Whitewater/Lewinksy investigation than Clinton was. Will history repeat itself, the first time as farce and the second time as tragedy?

Trump and the Department of Justice

Russia Probe Causes Tension Among Top Officials at Justice Department, blares CNN. I’ll say. Trump is furious with Sessions for recusing himself from Russia-related investigations, and Rosenstein for appointing Mueller. Sessions is also mad at Rosenstein for appointing Mueller. Rosenstein, who is beginning to behave a touch erratically, might recuse himself from supervision of Mueller, who Trump hates and thinks is biased, because Rosenstein was involved with Trump’s dismissal of former FBI director Comey, which Mueller is investigating. Next in line for this snake pit is Associate Attorney General Rachel Brand. Brand is well-regarded but lacks the experience and reputation of Rosenstein, Mueller, and Comey, who have all been or are being ground into tiny bits by the giant Trump food processor, and is already suspect among Trump opponents because of her connections to the right. If she steps into Rosenstein’s shoes, expect no one on either side to trust her decisions—not Trump, because of her connection to the Republican establishment, not the Democrats, also because of her connection to the Republican establishment.

Some observations:

1. Because the DOJ works closely with the president, its top officials can be implicated in the president’s illegal behavior in complex ways—worst of course if they are complicit (Sessions?), but it could also be that they are witnesses (Rosenstein) or even victims (Comey).

2. These are lawyers, and to some extent will be more loyal to the law than to the president who appointed them (but see Mitchell, Kleindienst), but they are under great pressure and must make subtle judgments that are unavoidably political even if they will always say they are guided by the (hopelessly ambiguous) law.

3. Remember Kenneth Starr? Or Lawrence Walsh? A special counsel really can be biased, and can go too far, and make bad judgments. Gingrich, Trump, & Co. really are being clever in preparing the way for such a view. The self-defensive leaks from DOJ about the Trump obstruction of justice investigation, if that is where they came from, have already damaged the integrity of the Mueller investigation. Mueller is now in a very difficult position: he must be fair, appear fair (not the same thing), maintain staff morale, and so on.

4. What we have then is a dogfight pitting Trump & Co. versus an already compromised DOJ that nonetheless possesses vastly more bureaucratic experience, expertise, and savvy but mostly lacking in political instincts. Who will win? Expect mutual demolition.

Meta-Obstruction of Justice

(Written with Daniel Hemel.)

Can a President be guilty of obstruction of justice for firing a special counsel who is investigating whether the President committed obstruction of justice?

The question might not be hypothetical. A friend of the President, Newsmax Media CEO Christopher Ruddy, told PBS NewsHour Monday evening that Trump is “perhaps considering terminating the special counsel.” Ruddy reportedly reiterated that view in a text message to the Washington Post.

As a constitutional matter, Trump would probably have to direct the Deputy Attorney General, Rod Rosenstein, to do the firing. If Rosenstein refused, then President Trump could fire Rosenstein. Ultimately, he could almost certainly find someone at the Justice Department (or elsewhere in his Cabinet) who would issue the order to terminate Mueller, though this might require a few more firings along the way.

But that still leaves the question: Would firing the special counsel amount to criminal obstruction? (Or perhaps “meta-obstruction”: the obstruction of an obstruction investigation?) A similar question arose during Watergate, when the House charged Nixon with obstruction of justice because (among other misdeeds) he fired special prosecutor Archibald Cox, who was initially retained to investigate “offenses arising out of the 1972 election,” which ultimately would include obstruction of justice.

The legal analysis of meta-obstruction isn’t all that different from the analysis of obstruction plain vanilla. Firing a special counsel who is himself investigating obstruction allegations would likely satisfy the actus reus element: “influenc[ing], obstruct[ing], or imped[ing] . . . the due and proper administration of the law under which a[] pending proceeding is being had before a[] department or agency of the United States.” There might be some question as to whether a special counsel investigation counts as a pending proceeding before a federal department or agency, though at least for impeachability purposes, the precedent from both the Nixon and Clinton episodes indicates that the House considers the obstruction of a special prosecutor or independent counsel inquiry to be a high crime or misdemeanor.

But as we have argued in the context of plain-vanilla obstruction, obstruction’s mens rea element is much harder to nail down. The statutory question is whether the President acted “corruptly,” i.e., “with an improper purpose.” We have also argued that a President acts with an improper purpose if he seeks to shut down an investigation that he believes might bring to light evidence of criminal conduct, or possibly merely embarrassing behavior, by himself, his family members, or his close political associates, unless he can show that he acted consistently with his constitutional role as the faithful executor of federal laws and as commander-in-chief.

How does this apply to meta-obstruction? A first question would be whether the President believes that the special counsel’s investigation might turn up evidence of criminal conduct. So to get even more meta: If the President believes that he himself might have committed obstruction, then intervening in the special counsel’s investigation of his own obstruction would also probably be obstruction. But whether the President committed plain-vanilla obstruction depends on his motives for interfering with the initial investigation into ties between his aides and the Russian government. So we arrive at a question of meta-mens rea: if the President intervened in the initial FBI probe for corrupt reasons, then his intervention in the obstruction investigation would likely be corrupt as well.

The President might defend himself from allegations of meta-obstruction by arguing that he fired the special counsel for entirely proper reasons. He might argue that Mueller is compromised by his reported friendship with former FBI Director James Comey, or motivated by partisan interest. He could also possibly argue that the investigation has undermined his ability to govern—which sounds self-serving but remember that this concern motivated abandonment of the independent counsel statute. One can imagine an argument that the president should not be allowed to shut down an investigation of a primary crime (hotel burglary, electoral manipulation) but should be allowed to shut down an investigation into obstruction of justice, just because the latter law is dangerously ambiguous and gives zealous prosecutors too much discretion.

When Does the President Commit Obstruction of Justice?, III

(Written with Daniel Hemel.)

We argued in a prior post that a President who fires the FBI Director in order to halt the investigation of the President’s former aide can—under certain circumstances—be guilty of obstruction of justice. Motive matters. If the President wants to halt the investigation because the former aide has offered him a bribe, the President would be acting “corruptly”—and thus criminally. But we can also imagine circumstances in which the President’s intervention in an investigation would be a legitimate exercise of executive power. Drawing the line between criminal obstruction and prosecutorial discretion is not an easy task. But it is also clear that criminal obstruction statutes can apply to executive branch officials who wield prosecutorial power—as illustrated by the indictment and conviction of Nixon Attorney General John Mitchell on obstruction charges.

Imagine that the President wants to halt the investigation because he knows that the former aide, while working on behalf of the President’s campaign, helped a hostile power influence the outcome of the U.S. general election. The President seeks to stop the FBI’s probe in order to prevent this information from coming to light. If the information does come to light, the aide and possibly the President himself could be found guilty of a serious crime. This seems like textbook obstruction of justice: intervening in an investigation—and thereby impeding the “due and proper administration of the law”—based on a motive that is fundamentally corrupt.

Now imagine instead that the President wants to halt the investigation because he knows that during the interregnum between the general election and the President’s inauguration, the former aide—acting on the President-elect’s behalf—sought to open a secret backchannel with a historical adversary. The negotiations were aimed at achieving a comprehensive Middle Eastern peace accord, or a new nuclear nonproliferation pact, or the release of U.S. hostages. Moreover, the President believes that it would diminish the prospects for diplomatic success if the existence of these talks came to light prematurely. Here, we think it’s clear that the President’s intervention does not constitute obstruction: the President should have the power to suspend an investigation when he believes the national interest demands it.

A natural test for obstruction might thus be a case in which the President seeks to immunize himself from conviction of some (underlying) crime. But this seems too narrow. Consider some intermediate cases, where the president halts an investigation into possible collusion between his former aide and a foreign power because:

— (1) The investigation might reveal unflattering facts about the President, top aides, or his family;

— (2) The investigation might distract attention from the President’s domestic policy agenda; and/or

— (3) The investigation might impose financial, professional, and emotional costs on the former aide, a man whom the President regards highly.

These are all harder cases. We think a reasonable approach would be to say that a President commits obstruction of justice when he seeks to stop an investigation that he believes may bring to light evidence of criminal activity by himself, his family members, or his top aides, and possibly merely embarrassing information as well (scenario 1, and, we think, scenario 3). But it should be a defense if the President can show that he acted consistently with his constitutional role as the faithful executor of federal laws and as commander in chief. A President who stops an investigation because of urgent national security priorities would not be criminally liable.

Consider now scenario (2): the President stops the investigation because it might prove to be a distraction that prevents him from passing health care legislation or comprehensive tax reform. We think this would indeed amount to obstruction. The President’s constitutional role is to “take Care that the Laws be faithfully executed,” not to enforce the laws selectively so as to maximize his probability of political victory.

This analysis assumes that the President acts on a single motive for intervening in an investigation. But what if the President acts for multiple reasons—one that would constitute obstruction and one that would not? Imagine, for example, that the President fires the FBI Director because (1) he wants to cover up evidence of wrongdoing by his own campaign, but also (2) thinks that the FBI Director mishandled a high-profile probe of the President’s election opponent and sullied the bureau’s reputation along the way. Does the fact that the President had at least one “good” motive inoculate him from criminal liability?

The problem of mixed motives in obstruction-of-justice cases is not unique to the presidential context. Courts have long wrestled with the same question as applied to non-presidential defendants. We know of no case holding that any ounce of proper purpose requires acquittal (though the Trump administration takes the analogous position that the existence of any legitimate reason for the President’s travel ban excuses concurrent discriminatory motives for the policy). A “primary purpose” test, which can be found in other areas of the criminal law, might serve here.

Can the president commit the crime of obstruction?, II

(Written with Daniel Hemel.)

The President orders the FBI Director to halt an investigation of the President’s former aide. The Director refuses. The President responds by:

A1. Firing the Director.

A2. Taking a fifty out of his billfold and offering it to the Director as a bribe.

A3. Pulling out a pearl-handled Smith & Wesson and shooting the Director in the forehead.

Put aside the question of whether the President can be indicted while in office. To fix intuitions, imagine that the actions come to light the day after he leaves the White House: the question is whether to charge him with a crime now that he is a private citizen.

We think almost everyone will agree that A2 and A3 are crimes. Alan Dershowitz argues that the President has the constitutional authority to order the FBI Director to end an investigation, but we are nearly certain that even Dershowitz would agree that the President must not violate the statutory prohibitions against bribery and murder along the way. It is worth noting, however, that these statutes constrain the president’s enforcement discretion—a fact that does not render them invalid.

Now consider a second scenario. The President’s former aide, who is under FBI investigation, offers the President $100 if the President stops the probe. The President then orders the FBI Director to halt the investigation. The Director refuses. The President responds by:

B1. Firing the Director.

B2. Taking a fifty out of his billfold and offering it to the Director as a bribe.

B3. Pulling out a pearl-handled Smith & Wesson and shooting the Director in the forehead.

We think almost everyone will agree that B1, B2, and B3 are all crimes. Again, the President has the constitutional authority to order the FBI Director to end an investigation, but he commits a felony if he exercises that authority in order to procure a payment for himself.

Now consider a third scenario. The President orders the Attorney General to stop prosecuting nonviolent drug offenses. The AG refuses. Again, the President responds by:

C1. Firing the AG.

C2. Taking a fifty out of his billfold and offering it to the AG as a bribe.

C3. Pulling out a pearl-handled Smith & Wesson and shooting the AG in the forehead.

C2 and C3 are again clearly crimes. But we think that now, almost everyone would agree that C1 is not.

What makes B1 and C1 different?

The distinction lies in the motive. Congress has enacted several obstruction of justice statutes that make it a crime to “corruptly” influence, obstruct, or impede a proceeding, or to attempt to do so. In B1, the President acts corruptly out of a desire to procure a bribe. In C1, the President presumably acts because he thinks that the prosecution of nonviolent drug offenders is a misallocation of law enforcement resources. Whether one agrees or disagrees with that view, it is difficult to argue that the President in C1 is corrupt.

A1 is the harder case—indeed, the only one of these scenarios that strikes us as difficult. It is also the case closest to the one at hand. Is it more analogous to B1 (clearly a crime) or to C1 (clearly not)? Again, the fact that the President has the constitutional authority to order the FBI Director to end an investigation does not help us answer the question, because firing the FBI Director in order to end an investigation can be a crime nonetheless (B1).

Much depends on why the President wants to end the investigation of his former aide. If he thinks that the investigation is a gross misallocation of FBI resources that threatens to divert the bureau from its crime-fighting and counterterrorism responsibilities, then A1 begins to look like C1 (not a crime). But what if he wants to end the investigation because he is worried it might bring to light embarrassing information about himself, his family members, or his political associates? Then, A1 begins to look more like B1 (criminal indeed).

The apparent fact that President Trump fired the FBI Director in order to halt an investigation of the President’s former aide does not tell us whether the President has committed a crime. Clearly, some actions that the President might take in order to halt an FBI investigation would indeed be criminal (bribery, murder). And almost as clearly, there are some circumstances in which a President who fires the FBI Director in order to halt an investigation would be guilty of a crime (B1). But there are also circumstances in which a President who fires a law enforcement official in order to stop an investigation or a prosecution (or thousands of the same) would not be a criminal. Senator Howard Baker’s famous question during the Watergate hearings—“What did the President know and when did he know it”?—might be repurposed as: “What did the President do and why did he do it.” At this point, the “why” matters just as much as the “what.”

Can the president commit the crime of obstruction of justice?

Alan Dershowitz says no:

The president can, as a matter of constitutional law, direct the attorney general, and his subordinate, the director of the FBI, tell them what to do, whom to prosecute and whom not to prosecute. Indeed, the president has the constitutional authority to stop the investigation of any person by simply pardoning that person.

His argument is that the obstruction of justice statutes can’t apply to the president because the president possesses the constitutional authority to stop an investigation—by issuing an order to the attorney general or FBI director, by firing them if they do not obey the order, or by pardoning the person under investigation. Any interpretation of the obstruction of justice statute that blocked this authority would violate the Constitution. (Dershowitz agrees that Congress can impeach the president for obstruction of justice but only because impeachment is a political process; on this, see below.)

This argument is superficially appealing but I don’t think it’s right. To see why, imagine that Congress passes a statute that says (for example): “the president may not order the FBI director to stop an investigation of the president or any other executive branch official.” If Dershowitz is right, this statute would violate the Constitution and be struck down by a court.

Would it? The Supreme Court hasn’t told us, but it rejected the broad description of presidential authority that Dershowitz relies on. In Morrison v. Olson, it upheld the (now defunct) independent counsel statute even though the law created a roving prosecutor, gave him the power to bring charges against executive branch officials including the president, and prohibited the president from removing him. (The Attorney General retained the power to remove the independent counsel for “good cause.”)

In the Starr report, independent counsel Kenneth Starr had no trouble accusing Bill Clinton of committing the crime of obstruction of justice:

An effort to obstruct justice by withholding the truth from the legal process — whether by lying under oath, concealing documents, or improperly influencing a witness’s testimony — is a federal crime. There is substantial and credible information that President Clinton engaged in such efforts to prevent the truth of his relationship with Monica Lewinsky from being revealed in the Jones case.

Dershowitz would no doubt argue that Starr was mistaken, and that if Clinton had been prosecuted rather than impeached, a court would have thrown out this charge. I think it more likely that a court would seize on the “corrupt” intent language in the statutes, and hold that the president can do what he wants as long as he does not act “corruptly.” He can fire the FBI director for mismanaging the institution, or failing to comply with the president’s priorities, or even for being disagreeable or untrustworthy, or for standing in the way of another appointment, but not for investigating the president or his top aides. Trump could have lawfully fired Comey for mishandling the (Hillary) Clinton investigation or simply for failing to inspire confidence but not for investigating Trump’s pal Flynn.

Why does this matter? No prosecutor will bring charges against Trump. But if an impeachment occurs, then the allegation that the president has committed a crime lends gravity to the proceedings, and fulfills the constitutional command that the impeachment be based on “high crimes or misdemeanors.” That’s why it matters whether Trump committed a crime or not, as Dershowitz clearly understands.

Dershowitz says that Trump could have pardoned Flynn, and that is surely right. I’m not sure, however, that Dershowitz is right that misuse of the pardon power—for example, to thwart an investigation of the president himself by depriving investigators of bargaining power over an aide—cannot be an obstruction of justice. Suppose that the president pardoned someone in return for a bribe. The pardon may be valid, but the president committed a crime. A similar argument could be made if the president pardoned someone in order to obstruct justice. In any event, unless Trump pardons Flynn, the argument is of no importance.

Response to Matt Klein’s post on Alphaville on Harberger taxation

(Written with Glen Weyl.)

Matt Klein’s post on Alphaville about a recent paper of ours made a number of errors. In that paper, we propose a “Harberger tax” on wealth: people would be required to declare their self-assessed value of their property and pay a tax on that valuation; anyone else could force a sale of the property at the declared valuation. Individuals could deduct the value of any liabilities (like debt).

Klein’s central claim is that our proposal would benefit the rich at the expense of ordinary Americans. “A purer variant of plutocracy is hard to imagine.” However, a wealth tax would fall more heavily on rich people than on middle-class or poor people. The reason is simple: rich people by definition own more property than the rest of us do. Because the distribution of wealth is highly skewed in the United States, the tax would fall far more heavily on the rich than on the poor.

Klein’s claim seems to be based on two misunderstandings of our paper. First, he appears to think that the Harberger tax is based on the nominal value of assets rather than on net worth (equity) (see section 2.7). The homes of middle class people are typically encumbered with mortgages, leaving relatively little equity to be subject to the tax. Second, Klein neglects a classic result in the theory of property taxes: imposing a tax lowers asset values. A tax of, say, 2.5% expropriates 1/3 of capital rents (if the interest rate is 5%), which would reduce the value of assets by a third.

Keeping these facts in mind, let’s consider an example. A typical American family has about $50k of income and about $80k of net worth, encompassing about $50k of home equity and $30k of other assets. A Harberger tax of 2.5% would cause this net worth to fall by about 1/3 to roughly $55k, while the annual tax payment would be roughly $1400. The tax would also generate revenue, which, according to our calculations, would generate about $14,000 per family as a social dividend. The typical family would gain on net more than $12,000 annually.

Klein also claims that wealthy people would predatorily take assets from poor people they want to harm. This is silly, as it assumes that individuals would price their assets below what they would be willing to accept for those assets, which is never true at the optimal tax rate under our system. Rich people can, at present, “predatorily” buy up assets by offering lots of money to make poor people rich. Most people don’t view this as a major threat to the well-being of the poor.

Elsewhere, Matt Levine offers some helpful criticisms about the administrative workings of our system, which we will address in due course.