“Now Scarlet he thought the most beautiful of all Colours . . . whereas the first Time he saw Black, it gave him great Uneasiness”

“Now Scarlet he thought the most beautiful of all Colours . . . whereas the first Time he saw Black, it gave him great Uneasiness”

Four years ago, while I was casually researching Molyneux’s problem, I stumbled on one of the most fascinating texts I have ever read. It is an account, three centuries old, of the experiences of a boy cured of congenital blindness at the age of thirteen.[1] He was not, in the narrowest sense, actually blind. He could tell night from day, and “for the most Part in a strong Light, distinguish Black, White, and Scarlet”; but he had no perception of shape, and his sense even of those colors was imperfect enough that when he was later shown them, after gaining his sight, he did not recognize them.

His blindness was owing to cataracts, in both eyes; on some day before 1728, the surgeon William Cheselden took an instrument to one eye and dislocated the cataractous lens.[2] With the opacity removed, the boy could see:

Now Scarlet he thought the most beautiful of all Colours, and of others the most gay were the most pleasing, whereas the first Time he saw Black, it gave him great Uneasiness, yet after a little Time he was reconcil’d to it; but some Months after, seeing by Accident a Negroe Woman, he was struck with great Horror at the Sight.

When he first saw, he was so far from making any Judgment about Distances, that he thought all Objects whatever touch’d his Eyes, (as he express’d it) as what he felt, did his Skin; and thought no Objects so agreeable as those which were smooth and regular, tho’ he could form no Judgment of their Shape, or guess what it was in any Object that was pleasing to him: He knew not the Shape of any Thing, nor any one Thing from another, however different in Shape, or Magnitude; but upon being told what Things were, whose Form he before knew from feeling, he would carefully observe, that he might know them again; but having too many Objects to learn at once, he forgot many of them; and (as he said) at first he learn’d to know, and again forgot a thousand Things in a Day. One Particular only (tho’ it may appear trifling) I will relate; Having often forgot which was the Cat, and which the Dog, he was asham’d to ask; but catching the Cat (which he knew by feeling) he was observ’d to look at her stedfastly, and then setting her down said, So Puss! I shall know you another Time. He was very much surpriz’d, that those Things which he had liked best, did not appear most agreeable to his Eyes, expecting those Persons would appear most beautiful that he lov’d most, and such Things to be most agreeable to his Sight that were so to his Taste. We thought he soon knew what Pictures represented, which were shew’d to him, but we found afterwards we were mistaken; for about two Months after he was couch’d, he discovered at once, they represented solid Bodies; when to that Time he consider’d them only as Partly-colour’d Planes, or Surfaces diversified with Variety of Paint; but even then he was no less surpriz’d expecting the Pictures would feel like the Things they represented, and was amaz’d when he found those Parts, which by their Light and Shadow appear’d now round and uneven, felt only flat like the rest; and ask’d which was the lying Sense, Feeling, or Seeing?

Being shewn his Father’s Picture in a Locket at his Mother’s Watch, and told what it was, he acknowledged a Likeness, but was vastly surpriz’d; asking, how it could be, that a large Face could be express’d in so little Room, saying, It should have seemed as impossible to him, as to put a Bushel of any thing into a Pint.

At first, he could bear but very little Sight; and the Things he saw, he thought extreamly large; but upon feeling Things larger, those first seen he conceiv’d less, never being able to imagine any Lines beyond the Bounds he saw; the Room he was in he said, he knew to be but Part of the House, yet he could not conceive that the whole House could look bigger. Before he was couch’d, he expected little Advantage from Seeing, worth undergoing an Operation for, except reading and writing; for he said, He thought he could have no more Pleasure in walking abroad than he had in the Garden, which he could do safely and readily. And even Blindness he observ’d, had this Advantage, that he could go any where in the Dark much better than those who can see; and after he had seen, he did not soon lose this Quality, nor desire a Light to go about the House in the Night. He said, every new Object was a new Delight, and the Pleasure was so great, that he wanted Ways to express it; but his Gratitude to his Operator he could not conceal, never seeing him for some Time without Tears of Joy in his Eyes, and other Marks of Affection: And if he did not happen to come at any Time when he was expected, he would be so griev’d, that he could not forbear crying at his Disappointment. A Year after first Seeing, being carried upon Epsom Downs, and observing a large Prospect, he was exccedingly delighted with it, and call’d it a new Kind of Seeing. And now being lately couch’d of his other Eye, he says, that Objects at first appear’d large to this Eye, but not so large as they did at first to the other; and looking upon the same Object with both Eyes, he thought it look’d about twice as large as with the first couch’d Eye only, but not Double, that we can any Ways discover.

From William Chesselden, “An Account of Some Observations Made by a Young Gentleman, Who Was Born Blind, or Lost His Sight So Early, That He Had No Remembrance of Ever Having Seen, and Was Couch’d Between 13 and 14 Years of Age,” 402 Philosophical Transactions 447 (1728), 448-50.

I wondered in 2012, and I wonder now, what became of the boy. No one seems to know his name. Is it written in some lost paper of Cheselden’s?

1. He either was actually born blind, or went blind quickly: “He Had No Remembrance of Ever Having Seen.”

The boy was couched “Between 13 and 14 Years of Age.” I take that to mean that he was thirteen when the first eye was treated, and fourteen when the second eye was (see the quoted material below).

How (strong) judicial restraint would lead to an incoherent interpretation of the Constitution

How (strong) judicial restraint would lead to an incoherent interpretation of the Constitution

In a famous essay published in the Harvard Law Review in 1893, James Bradley Thayer argued that a court should invalidate a statute only if its unconstitutionality was plain and palpable—“not open to rational question.”[1] Like Thayer in 1893, I in 2016 think the courts are much too powerful; but I think Thayer’s rule leads to paradoxes. The purpose of this post is to explain why.

Thayer’s rule admittedly does not lead to paradoxes if one thinks only of cases of a certain sort. “No . . . ex post facto Law shall be passed,” says Article I, §9, cl. 3. Suppose Congress makes it a federal crime punishable by a year’s imprisonment to drive above 100 mph within the limits of the District of Columbia. Sometime thereafter, John Doe is caught by police driving 120 mph on a DC road. He is arrested, and is headed for the slammer—but only for a year. Before he is tried, however, Congress passes a second statute increasing the punishment for this crime from one year to two years, while specifying in the clearest English that the enhancement of the penalty is to apply retrospectively. When the government tries to have Doe put away for two years rather than one, does it act constitutionally? On this point, the Constitution’s text is ambiguous, and Thayer’s rule tells us how a court should confront the ambiguity. It should adopt the narrower of the two interpretations and uphold the new law, declaring it to be constitutional. That may be good advice, or it may be bad advice; what it certainly is not is paradoxical advice.

But now take another example, this one involving the Fourteenth Amendment’s requirement that “No State shall . . . deny to any person within its jurisdiction the equal protection of the laws.” Under one interpretation of the clause, it would be a simple prohibition on outlawry.[2] On another, it would be a prohibition on racially discriminatory laws.[3] Simplifying enormously, let us suppose that these are the only viable interpretations of this provision. What should a Thayerian court do if a state legislature should adopt two laws, one a racially non-discriminatory law prescribing outlawry, the other a racially discriminatory law not involving outlawry? It may hold that:

  1. Both laws are constitutional (the Thayerian option)
  2. The racially discriminatory law is constitutional, and the outlawry law is not
  3. The outlawry law is constitutional, and the racially discriminatory law is not
  4. Both laws are unconstitutional

In this case, it seems impossible for a court to remain faithful to Thayer’s rule while simultaneously adopting a consistent interpretation of the Constitution. There is no consistent interpretation such that both statutes are constitutional: at least one is invalid. There is also no consistent interpretation such that both statutes are unconstitutional: one, at least, is valid. But to strike down either and uphold the other is to strike down a statute whose unconstitutionality is clearly “open to rational question,” and thus to infringe Thayer’s rule. Every course is forbidden—unless incoherence is allowed.

The above admittedly assumes that courts are going to be in the business of answering directly whether a law is or is not constitutional. A Thayerian might want to abandon that part of our judicial-review tradition. Instead of asking whether a law is constitutional, a court might ask whether there is a consistent interpretation of the Constitution under which the law and others the legislature has adopted are all constitutional. Where the answer was yes, the court would be bound to uphold the law; where the answer was no, some other rule would determine which statutes were to be invalidated. One idea would be to give earlier statutes priority over later ones, but it does not really matter just what this other rule is; the important thing is to preserve coherence. Now a benighted legislature that had adopted a racially discriminatory policy could prescribe outlawry—but only if it became colorblind first. A cruel legislature that had prescribed outlawry could take up racism, but only if it was willing to let up on its vindictive streak.

Unfortunately, even this does not solve the problem—not entirely or enough. Take the same equal protection example, and tweak it ever so slightly: one statute is adopted by the legislature of Virginia, and the other by the legislature of New York. What, now, does a court do?

1. James Bradley Thayer, “The Origin and Scope of the American Doctrine of Constitutional Law,” 7.3 Harvard Law Review 129 (1893), 144. It is famous among law professors; I would not recommend working it into small talk.

Thayer actually limited his rule to acts of Congress. In referring to “Thayer’s rule,” I excise that limitation.

2. Herbert Wechsler, “Toward Neutral Principles of Constitutional Law,” 73.1 Harvard Law Review 1 (1959), 19.

3. Cf. Robert H. Bork, “Neutral Principles and Some First Amendment Problems,” 47.1 Indiana Law Journal 1 (1971), 11, passim.

How to pick the losing position on everything and still win an election

How to pick the losing position on everything and still win an election

Imagine an election in which only two issues are contested. The first (let us say) is whether the capital gains tax rate should be raised from 21.07 percent to 21.071 percent. The second is whether the hunting season should be extended by one day.

The candidate Alexander, whose only object is to win the election, consults with his pollster, who tells him everything he thinks he needs to know. By a margin of 60 percent to 40 percent, the public prefers the higher capital gains rate; and by the very same margin of 60 percent to 40 percent, they favor the extended hunting season. Alexander promptly adopts these positions and begins campaigning on them.

His opponent, Bruce, has no real idea what is going on. He looks at Alexander’s stances, and takes all of the opposite ones—not because he necessarily agrees with them, but because that sounds like the sort of thing an opposition candidate should do.

When the election is held, there is perfect turnout; not a single voter has changed his mind on either issue; and every voter votes exclusively on the issues. It may seem, then, that Alexander is headed to a safe victory. But when the votes are tallied, it turns out that Bruce has won in a landslide, 80 percent to 20 percent.

Alexander’s loss may be a surprise, but it is no mystery. Call the higher capital gains tax rate C, and the lower c; call the longer hunting season H, and the shorter h. There were, then, four policy possibilities: CH, Ch, cH, and ch. Suppose the electorate consists of three groups, each with its own preference ordering:

  • (20%) CH > Ch > cH > ch (or CH > cH > Ch > ch)
  • (40%) Ch > ch > CH > cH
  • (40%) cH > ch > CH > Ch

In that case, C and H have 60 percent support over c and h, respectively; but CH loses to ch, 20 percent to 80 percent.

The Alexander of this story has committed what I will call the “winning issues fallacy.” That fallacy supposes that if an issue would be a winning issue for a candidate in a single-issue election, it must be a “winning” or at least advantageous issue for him in a multi-issue election.[1] To think that way is to put the focus in the wrong place. Electorally speaking, the important question is not: How popular is this stance? It is: Holding all other stances on issues (by both candidates) constant,[2] how popular is this stance among people whose votes will be affected by it? And one can answer that question only by looking at where position bundles stand in individuals’ multi-issue preference orderings.

Suppose a poll shows support for limiting farm subsidies at 55 percent against 45 percent. The winning issues fallacy encourages us to say that a candidate is better off adopting this popular position. From a purely political standpoint (the argument will run) it is always better to side with 55 percent of the voting public than with 45. Seeing through the winning issues fallacy, perhaps the better bet is that making this popular proposal will marginally hurt a candidate. Very few people not on the receiving end of the subsidies are likely to have their votes “tipped” by this issue, and it is the tipped votes that matter.

Take any issue, big or small, and the same logic applies. It does not matter what the public thinks about the EU, or a border wall, or refugees, or paid maternity leave, or any other issue. It matters what a slice of the public thinks, the people whose votes are “tippable” given the other positions the candidates hold.

In extraordinary circumstances, one position (A) might outperform another (B) in the polls by a margin of 9:1, and yet it might be political suicide for either of two candidates to adopt the more popular position. How? Suppose that the B-favoring 10 percent will ordinarily split their votes about 50/50 between the two candidates, but that they care so deeply about their preference for B over A that every single one of them will mechanically vote against any candidate who comes out in favor of A. Among the A-favoring 90 percent, on the other hand, only a handful of people will ever change their votes in response to a candidate declaring for A. In that case, a candidate who adopts the position favored by nine-tenths of his electorate has no hope of being elected unless he would otherwise have won in a landslide. (The exception would be if both candidates came out in favor of A—but it is uncertain that that would be a stable equilibrium.)

Of course, it matters how other stances are held constant. Hold them constant with one set of values, and a position may be advantageous. Hold them constant with another, and the same position may be disadvantageous. In the Alexander-Bruce hypothetical, Alexander chose CH and Bruce ch. If (adopting Bruce’s perspective) we hold everything constant except Bruce’s stance on the capital gains rate, it is clear that h is his best play. If he chooses it, he wins in a landslide. If he instead chooses H, he loses 40 percent to 60 percent (as the election is then a contest between cH and CH). Things change if instead we keep other positions constant by leaving Alexander’s stances at CH while fixing Bruce’s stance on the capital gains rate at C. Now Bruce’s best play is H, which scores him a tie, as opposed to the 40 percent to 60 percent loss of putting Ch up against CH. The ideal, in any real-world situation, is obviously to understand all of the possibilities and pick the optimal bundle given the positions of one’s opponent.

The bottom line, however, is that the winning issues fallacy is a fallacy. Otherwise, it would not be possible for Bruce to lose an election by switching from ch to cH—that is, by substituting a more popular position for a less popular one, while holding all of his other positions constant.

1. To be sure, probabilistically speaking Alexander made the best gamble—assuming he had no information about voters’ two-issue preference orderings, and no way of obtaining any useful information on that point. (On the information he had, there were more possible “voter states” such that he would win with CH than there were those such that he would win with any one of ch, Ch, and cH.) But in a typical election this assumption will not hold.

2. I mean: (a) all of the positions of the other candidate, and (b) all of the other positions of this candidate.

Another side of George Washington

Another side of George Washington

In 1797, James Monroe published (brace yourself for an eighteenth century title) A View of the Conduct of the Executive, in the Foreign Affairs of the United States, Connected with the Mission to the French Republic, During the Years 1794, 5, & 6. In it the future president of the United States attempted to justify his behavior as minister to France, and attacked in heated terms the administration of George Washington, who had fired him.

It turns out Washington obtained a copy of Monroe’s work, and sprinkled it with annotations. Some of his more pointed comments put a smile on my face:

In the month of May, 1794, I was invited by the President of the United States, through the Secretary of State, to accept the office of minister plenipotentiary to the French Republic. (iii)

GW: “After several attempts had failed to obtain a more eligable character.”

It had been too my fortune, in the course of my service, to differ from the administration upon many of our most important public measures. (iii)

GW: “Is this adduced as conclusive evidence, that the administration was in an error?”

That the administration had injured me, was a point upon which I had no doubt; that it had likewise compromised its own credit, and with it that of the United States, was also a truth equally obvious to my mind. (xxx)

GW: “But not so in either case to an impartial and discriminating mind.”

Upon mature reflection, therefore, it appeared that I had but one alternative, which was to remain where I was, and proceed in the functions of my office, notwithstanding the embarrassments to which I might be personally subjected, or to retire, and in retiring, to do it tranquilly, without explaining my motives for it; or by explaining them, denounce the administration to the public. . . . Besides, it seemed probable that my retreat at that moment, in either mode, might have some influence in inducing the French government to adopt a system of policy toward us, which it was equally my duty and my wish to prevent. I resolved, therefore, to stand firm at my post. . . . (xxx) [1]

GW: “Curious and laughable to hear a man under his circumstances talking seriously in this stile, when his recall was a second death to him.”

But I own also that my reluctance was diminished by the knowledge that the administration possessed the treaty with England, whilst Colonel [David] Humphreys was in America, and the presumption thence arising, that this objection was weighed and overruled before his departure. (xxxii)

GW: “‘And he wrote to his father, ending with this line, I am my lovely Nevia ever thine.’” [2]

The appearance of the treaty excited the general disgust of France against the American government, which was now diminished by the opposition which the American people made to the treaty. (xxxv-xxxvi)

GW: “Who were the contrivers of this disgust, and for what purposes was it excited? Let the French party in the United States, and the British debtors therein, answer the question.”

But it was my duty to answer this letter, which I did without a comment; for it was improper for me to censure, and useless to advise. (xlvi)

GW: “When a rational answer and good reason cannot be given, it is not unusual to be silent.”

Much, too, was said in that address of the advantage of our accommodation with Britain, as likewise of the favorable disposition of that power towards us, without the slightest attention being shown to the French Republic. (xlviii)

GW: “To state facts for the information of Congress, and not to write eulogiums on the French nation and conduct, was the object of the then President. If Mr. Monroe should ever fill the chair of government, he may (and it is presumed he would be well enough disposed) let the French minister frame his speeches.”

But, by this attack on me, a new topic has been raised for discussion, which has drawn the public attention from the conduct of the administration itself; for, in consequence, the only question now before the public seems to be, whether I have merited the censure thus pronounced upon me by the administration, or have been dealt hardly by. But this was a mere political manoeuvre, intended doubtless to produce that effect. (liv)

GW: “Self-importance appears here.”

It being known that, with other members of the Senate, I had opposed in many instances the measures of the administration, particularly in that of the mission of Mr. Morris to France, and of Mr. Jay to London; from the apprehension those missions would produce, in our foreign relations, precisely the ill effect they did produce. (lx-lxi)

GW: “Unpardonable to appoint these men to office, although of acknowledged first-rate abilities, when they were of different political sentiments from Mr. Monroe, whose judgment, one would presume, must be infallible.”

Whether the nature of this crisis contributed in any degree to influence our measures, by repelling us from France and attracting us towards England, is submitted for others to determine. (lxiii)

GW: “As he has such a happy knack at determining, he ought not to have let this opportunity escape him.”

What would have been the condition of these States had France been conquered, and the coalesced powers triumphed, it is easy to perceive. (lxiii)

GW: “In turn, what will be the consequences of France overturning so many governments? and making partition of so many countries? One, it is supposed, is right—the other, wrong; from the actors in the Drama.”

And I now declare, that I am of opinion, if we stood firmly upon that ground, there is no service within the power of this republic to render, that it would not render us, and upon the slightest intimation. (123)

GW: “That is to say, if we would not press them to do us justice, but have yielded to their violations, they would have aided us in every measure which would have cost them—nothing.”

1. This quotation is not in Ford’s edition. It is taken from The Papers of George Washington: Retirement Series, vol. 2, ed. Dorothy Twohig (1998).

2. This annotation, and the quotation that precedes it, are absent from Ford; see the previous note for the source. The annotation is apparently an inexact quotation from an English translation of Martial, Epigrams, book I, 68 (or 69 in the system some older texts seem to have been using).

Why (national) immigrant crime rates are basically irrelevant

Why (national) immigrant crime rates are basically irrelevant

The most natural way of determining the impact of immigration on crime is to compare immigrants’ crime rates with natives’. Are the rates for immigrants lower? Then immigration can be said to make the country a more law-abiding place. Is it the reverse? Then the opposite can be said. This simple math is exactly correct, if what one is concerned with are national crime rates. But it is not clear that national crime rates matter.

Suppose there exists a nation of 100 million people. Each of them lives in one of ten cities, each of which has a population of 10 million people. Call the murder rate of this nation, whose population we will imagine to be entirely native born, X.

Now 10 million immigrants arrive.

Case 1. The immigrants found an entirely new city. If their murder rate is X, the national murder rate is unaffected. If it is greater than X, the national murder rate increases—but no pre-existing city is affected. If it is less than X, the national murder rate decreases—but again no pre-existing city is affected.

Case 2. The immigrants have a low murder rate of X/2, but they all move to a city with a still lower murder rate of X/10. The national murder rate falls: but the murder rate in each and every city is either the same or higher than before.

Case 3. The reverse. The immigrants have a high murder rate of 2X, but they all move to a city with a still higher murder rate of 10X. The national murder rate increases: but the murder rate in each and every city is either the same as before or lower.

These illustrate the general point, which is that it matters where the immigrants settle. If the murder rate changes in the United States, the murder rate changes for North America—but no Canadian will think himself safer or less safe on this account. The national murder rate is similarly artificial. It is the average number of murders committed by a set of human beings no one person interacts with. And as these cases show, the murder rate may do one thing nationally while in each locality it does nothing or just the opposite. (City rates are also imperfect, of course—and would be especially so with cities this large. I am using them as a better, not a perfect, metric.)

With a little more work, we can spin out starker hypotheticals. I close with Case 4, in which every city’s crime rate is made to rise while the national crime rate falls—and this though the nation is made up exclusively of cities.

Case 4. Suppose eight of our ten cities have a murder rate of X. One (Safe City) has a murder rate of X/10; another (Violent City) has a murder rate of 10X. The immigrants have a low murder rate of X/2. We will suppose that X is 6 murders per 100,000, and that each immigrant who commits a murder commits exactly one. We have, then, 300 immigrant murderers, each of whom will commit exactly one murder.

We divide up the immigrants thus:

  • Group A consists of the great bulk of them—9,700,000 people, 210 of whom commit a murder in the relevant year. The group’s murder rate is therefore roughly 2.16 murders per 100,000.
  • Group B consists of 200,000 people, 24 of whom commit a murder in the relevant year. Its murder rate is accordingly 12 murders per 100,000.
  • Group C consists of 100,000 people, 66 of whom commit a murder in the relevant year. Its murder rate is accordingly 66 murders per 100,000.

Group A moves to Safe City; Group C, to Violent City. Group B, meanwhile, splits into eight subgroups, each of which has the same murder rate; and each of these moves to a different one of the remaining cities. Since 2.16 > 0.6, 66 > 60, and 12 > 6, the murder rate ticks up in each and every one of the nation’s ten cities. Meanwhile, the national murder rate falls. The nation’s localities have all become statistically more dangerous as the nation has become less so. [1]

1. Of course, as with Case 2, so with Case 4. We can run this example in reverse, so that the nation’s localities all become statistically less dangerous as the nation becomes more so.
Why the American constitution should be more like the British

Why the American constitution should be more like the British

The American Constitution may be the most loved legal text in the world. I wonder whether that love serves us well at this point in our history, when our political system is increasingly dysfunctional. Stepping away from its status as a national symbol, and considering it objectively, the Constitution seems to contain two major defects. The bigger of the two is that it makes the president independent of the legislature. The second is that it gives the Senate an equal share of the legislative power. Books aside, the real effect of these two features of the American Constitution is to ensure that the executive and the judiciary can push aggressive policy agendas with little effective resistance from the legislature.

Suppose in some future amendment we were to up-end these features with five reforms:

  1. Members of the House of Representatives are to be chosen once every four years, i.e., in 2016, 2020, 2024, etc.
  2. The president is to be chosen by the House of Representatives, which may remove or replace him at any time. The vice presidency is abolished: When the president is removed, he continues to act as president until a successor is chosen. When he cannot or declines, the acting president is whomever may be specified by statute.
  3. Article I, §6, cl. 2 is rescinded as far as the House of Representatives is concerned. (This would allow a member of the House to simultaneously hold an executive office, including the office of president of the United States. [1])
  4. The presidential veto is abolished.
  5. The Senate may delay a bill for up to a year; after that, the House may enact it over the Senate’s objection. In the special case of a “money bill”—a bill dealing exclusively with taxation and/or spending—the Senate may delay the bill for no more than 30 days.

I think this change would be a major advance. By introducing into the American system the crucial feature of Britain’s parliamentary constitution—the accountability of the executive to a chamber of the legislature, and one which is overwhelmingly dominant in relation to the other—we would have closed much of the space for executive and judicial usurpation.

The reform would, in the first place, end the Imperial Presidency—sort of. It depends, admittedly, on what one means by that. Right now, the president has both enormous power and limited accountability. For all I know, Congress might choose to delegate more power to an executive it could control. But the beauty of making the president accountable to the legislature is that it makes the question of how much power he possesses far less important.

When Congress vests a power in the head of some department, everyone agrees that it almost vests it in the president. Why? Because the president has great control over who serves as the head of that department. If the president instructs (say) the attorney general to do X, it is true that the attorney general can decline. What keeps that from being terribly interesting is that the president can fire him at any moment. If the threat of being fired is not enough to keep him in line, the president can follow through on the threat and instruct the acting attorney general to do X (and under present law, the president can probably choose as the acting attorney general “a[ny] person who serves in an office for which appointment is required to be made by the President, by and with the advice and consent of the Senate”).[2] The attorney general’s powers, then, are only “kind of” his own.

A “parliamentary presidency” would place the president himself in an analogous position: his powers would only be kind of his own. That at least strips much of the danger out of presidential power. Take an extreme hypothetical: the hopefully never-to-be-enacted Presidential Taxation Act, which vests in the president the power to adjust marginal tax rates by executive order. Under the reformed (parliamentary) presidency, even this statute seems almost tolerable. The House would remain in ultimate control. If it opposed the president’s chosen rates strongly enough, it could pressure him into reversing course, or (barring that) remove him from office.

Of course, the reform would have another effect: it would make it much easier to adopt legislation. A party that won a House election decisively would generally speaking be able to enact its full agenda, though the Senate, if hostile, could slow things down and force the House to either accept or refuse to accept certain popular amendments. (That caveat is not trivial. The House might bundle together in a single bill an unpopular provision favored by its base with a large number of popular ones. The Senate could respond by passing an amended version of the bill that dropped the unpopular provision. The House would then be faced with a choice: It could fold, accepting the Senate’s amendment; or it could wait just under a year and enact its bolder version—while paying the political price.) It is hard to predict just what this would do to the quality of the United States Code. Wise legislation would be easier to adopt, and foolish legislation easier to repeal. Foolish legislation would be easier to adopt, and wise legislation easier to repeal. Who is to say which effect is dominant?

My own view is that policy considerations are probably a wash. By this I do not mean that legislation is unlikely to be better or worse. What I mean is that it is probably about as likely to be better as it is likely to be worse [3]—and about as likely to be significantly better as it is to be significantly worse, etc. Take any change in the quality of legislation and call it X. I suppose that X and -X are equally likely. If I am right, the situation is like a little game in which the player is given two options. If he chooses the “boring” option, he is gifted $1,000 and goes home happy. If he chooses the gambler’s option, he has to flip a fair coin. If the coin comes up tails, he gets nothing; but if it comes up heads, he receives some amount of money over $2,000—specifically, a quantity such that the amount of additional utility he derives from it versus the $1,000 is equal to the amount of additional utility he would derive from the $1,000 versus the $0. Which option should the player choose? There does not seem to be a right or wrong answer to this question; but most people will probably either be indifferent between the two options, or close to indifferent.

Meanwhile, the greater ease of adopting legislation would have two positive consequences for power dynamics under the Constitution:

1. The minor one. It would be clearer to voters who bore responsibility for what. The president could not blame Congress, nor Congress the president—nor the House the Senate, or the Senate the House—for a failure to deliver on campaign promises.

2. The really important one. There would be a more meaningful check on the power of the courts in the area of statutory interpretation. Suppose the Supreme Court of the United States interprets a statute in a manner that the governing party disapproves of. Under the reformed system, it would only take a majority of the House to push through legislation nullifying the decision by amending the statute. If it was clearly a sound decision—if the statute’s language required it—there is still nothing wrong with this. The legislation will not have retroactive effect: it will be just another case of a legislature substituting new law for old. If the decision was clearly unsound—if the statute’s language required another result—the court will have been blocked from effectively usurping legislative power. If (as is most likely) the decision was somewhere in between, the judges were exercising a quasi-legislative discretion. The supremacy of the House of Representatives would act as a curb on that discretion, limiting their ability to use it to advance politically controversial causes.

I will close with this: Small-d democrats denounce the electoral college and the malapportioned Senate. At the same time, there have been touches of Caesarism in recent presidents; for now, all is calm, but will that last? What would a statesman do, if he wanted to kill three birds with one stone?

1. It would also restrict the application of the “emoluments” language to the Senate.

2. 5 U.S.C. § 3345. See “Authority of the President to Name an Acting Attorney General,” 31 Op. O.L.C. 208 (September 17, 2007). A complication is that this statutory power is triggered only when an officer “dies, resigns, or is otherwise unable to perform the functions and duties of the office.” It is not certain that being fired would count as being “unable to perform the functions and duties of the office”—though if death and/or resignation qualifies (note the word “otherwise”), presumably termination does too.

3. Perhaps better is a little more probable than worse. Suppose there are two bills before Congress addressing some problem. Call them A and B, and call the status quo S. Suppose one party controls the House and thinks A > B > S; the other party controls the Senate and thinks B > A > S. Stipulate (as seems reasonable) that where both parties believe that a policy would be an improvement over the status quo, it would be. We assume, then, that either of A and B would be an improvement over the status quo. Under the reformed system, A or some compromise between A and B is almost certain to be adopted. Under the current system, the status quo may be perpetuated indefinitely as each side blocks the other’s proposal in the hope that the other side will fold and adopt its own first preference. (The same situation, of course, can exist between the Congress and the president.) A British-style system at least prevents these standoffs.