Emory Law Journal

Keynote Address, 2009 Randolph W. Thrower Symposium: The Intra-Executive Separation of Powers
Paul D. Clement 43rd Solicitor General of the United States and head of the national appellate practice at King & Spalding LLP.

What I thought I would do today is to talk a little bit about an aspect of the separation of powers that in my own view is a little under-discussed and a bit under-analyzed, and that is the internal separation of powers. This is not the well-documented tension among the three branches of government, but this is how to go about dividing authority and dividing responsibilities within the Executive Branch itself to provide a better result for governance. We’ll hear much about this, I think, in some of the later panels.

Certainly everybody as a starting point for talking about the Executive and executive power would point to the Vesting Clause in Article II. And the Vesting Clause in Article II says roughly that the power and the authority of the Executive Branch is vested in a President. Now there can be much dispute about exactly to what extent there is a unitary executive, and exactly how much should be read into the Vesting Clause. But I think one thing that everyone will agree upon—what even the most ardent advocate of the unitary executive would surely concede—is that the President cannot be a one-man band. The President cannot discharge all the various functions of the Executive Branch by himself, or someday, herself. So there has to be more than one person discharging the functions of the Executive Branch, and once you concede that essential fact, that obvious fact, then some thought should be given to how one divides authority within the Executive Branch.

Now my focus will be essentially two-fold. One, my focus will be on how that power is best allocated and divided in discharging one very important executive function, and that is the function dealing with the law and taking care that the law be faithfully executed, principally through the Justice Department. So that will be my focus. Second, my focus will be practical. There are others on some of the panels later who are better equipped and better able to discuss this as a purely theoretical matter. So one of the things I hope to add is a little bit of a practical perspective based on the last seven and a half years that I spent in the Justice Department, all of which I spent in the Office of the Solicitor General. So let me start, as a point of departure, to talk about the work of the Office of the Solicitor General and use that as a jumping-off point to discuss some broader principles about the different functions that are discharged by the Justice Department.

I. The Work of the Solicitor General’s Office

I am sure this is a very sophisticated audience, and there are varying degrees to which people are familiar with the day-to-day operations of the Solicitor General’s office and the work of the SG. Sometimes when I discuss the SG in a group of laypeople, I’ll ask people what’s the one thing that they know that the SG does. Sometimes I’ll get the answer: “SG, isn’t that the person who puts the warning labels on cigarettes?” And then I have to confess, “No, I was thinking of the other SG, the Solicitor General.” If I can get people to focus on the right SG, then if there’s one thing that people know about the SG’s office, it’s that the Solicitor General represents the United States, principally the Executive Branch of the United States, before the Supreme Court of the United States at oral argument. And I think that is, of course, both the most public role of the Solicitor General and also what for every person in the office who gets the opportunity to do so, the highest professional honor one can have: to step up in front of the United States Supreme Court and offer oral argument on behalf of the United States is truly a high professional honor.

But the important point for today’s purposes is although that is a high professional honor, it is just the tip of the iceberg in terms of the day-to-day work of the office. And in talking about the other aspects of that work, one of the very most important aspects of the office’s function is to determine which cases the United States will take up to the United States Supreme Court. And in thinking about this decision about which cases to take up to the Supreme Court, I find it helpful to contrast the process that is used in private practice—where I am now—and the process that is used in the government.

In private practice, the process is really quite straightforward. You’ll get a call from a general counsel of a corporation. The general counsel will tell you about this outrageous opinion from a court of appeals—it’s a terrible decision, it’s gotten everything wrong, it’s a horrible result, and we have to take it up to the Supreme Court of the United States. Now in private practice, the response—not every time, but virtually every time—is, “That’s absolutely right. This is a terrible decision and we have to take it up.” Now why is that? As a Supreme Court practitioner one should know that the chances of the Supreme Court taking the case are quite remote. The court right now, for the last couple of terms of anyways, has been taking a mere seventy cases a year out of a universe of over seven thousand certiorari petitions, so there is literally less than a one-percent chance that the average petition will get granted.

But nonetheless in private practice you know at least two relevant facts: one, although there’s less than a one-percent chance that certiorari will be granted by the Court, there’s a substantially higher percentage chance that your bill will get paid, and Supreme Court work is, after all, pretty good work, so that’s an important consideration. The second important consideration, though, which is related to the first, is that you know that if you simply say “no” to the client, and not “no, because this is just frivolous,” but “no, because you have a claim to make, you have a case to make to the Supreme Court, but it’s overwhelmingly likely to be rejected”—if you make that case to the client, chances are they’re just going to go to the next number on the speed dial and call another lawyer to file the petition in the Supreme Court. So the dynamic is quite different in that sense, and it makes it quite easy for the private practitioner to say, “That’s right. Let’s file a cert. petition.”

The process is quite different in the federal government, although it starts exactly the same way. The process starts with a call from a general counsel of one of the agencies—say, the Defense Department—and he or she will tell the Solicitor General that the Department has just gotten this horrible, awful result in the court of appeals, it’s terrible, and we must file a petition with the Supreme Court. That, however, is where the similarity ends, because at that point the Solicitor General will tell the general counsel not, “Absolutely, that’s right, we have to file.” The Solicitor General will tell the general counsel, “Well, that’s interesting. You may be right, you may be wrong. The decision may be right, it may be a case that merits Supreme Court review. I don’t know. Why don’t you put your thoughts on the matter in writing, and we will take that under consideration?” And that will happen; the general counsel of the agency will make the case for certiorari. That recommendation—that the United States file a cert. petition—will then go to the appropriate litigating division at the Justice Department. They will do their own independent analysis, and at that point the memos will go to the Solicitor General’s office. An Assistant to the Solicitor General will prepare his or her own thorough examination of whether the case is cert.-worthy. At that point a Deputy will annotate the various memos. That packet will finally go to the Solicitor General, and a decision will at that point be made either to take the case up to the Supreme Court or not to take the case up to the Supreme Court.

Without getting into the specific numbers as to how often based on a recommendation from an agency general counsel the United States ultimately files a petition, what I can tell you is that in the overwhelming majority of cases, the United States does not file a petition. The SG’s office says no and a petition is not filed, even though the general counsel wants to file a petition in the Supreme Court.

Now what explains that difference in practice? What explains that difference in result? I think the most straightforward explanation is good, old-fashioned monopoly power. The Solicitor General, within the Executive Branch of the United States, enjoys monopoly power essentially over whether to file a petition for certiorari and how to litigate on behalf of the United States in the Supreme Court. So unlike general counsel in private practice, who have a variety of counsel to choose from, and if they don’t like the answer the first person gives them they can move right down the speed dial, in the case of a general counsel at one of the agencies, if the Solicitor General is not inclined to file a petition for certiorari, the options are much more limited. It is possible to try to go over the Solicitor General’s head and appeal to the Attorney General or, in the most extreme case, to the President himself, but as a practical matter the simple expedient of calling another lawyer is not available.

Like all monopoly power, the Solicitor General’s responsibility to litigate on behalf of the entirety of the Executive Branch could be abused. You could have a Solicitor General who essentially took the fact that there is this monopoly power, there’s a bottleneck in the system, and used it as a way to put his or her policy views into place, and decided not to appeal particular cases not because of the law, but because of the underlying policy. And I should say based on personal experience that this abuse—this potential for abuse, anyways—is not entirely theoretical in the sense that you are given actual invitations for this kind of abuse.

And so in my experience, there could be a case where, say, the D.C. Circuit has struck down a rule that one of the agencies has promulgated. And in that kind of case you will often have some of the private counsel involved in the case make a pitch to the Solicitor General’s office as to why the United States should or should not file a cert. petition in a particular case.

And in some of these cases somebody affected by the underlying rule will come in and make the case against the United States filing a petition for certiorari, and it will often become quite clear quite early in the process that the argument this person is making for the United States not to file a cert. petition is not really a legal argument directed, for example, at whether there’s a circuit split or whether this is an issue that the agency can fix in other ways, but is indeed essentially a backdoor assault on the rule itself. The argument is, “Well, the D.C. Circuit struck down this rule. We never liked this rule. We lobbied against it in the first instance, but the agency promulgated it anyways. And now here’s our chance finally to vindicate our policy view and have you, the Solicitor General, strike the death knell for this rule, by not petitioning to the Supreme Court.”

And in that kind of situation, the reaction that I had and others in the Solicitor General’s office would have is to essentially stop the conversation at that point, jot down the address of the agency that had promulgated the rule, and hand it to the lawyer and say, “Look, if you want to get this rule taken off the books, it’s fine by me. We have plenty of other work to do in the office. So it’s not going to be any great hardship for the Solicitor General’s office if the agency that originally promulgated this rule pursuant to notice-and-comment rule-making decides to withdraw the rule and this case is essentially mooted and goes away. That’s fine by us. But the argument you’re making is really one you should direct to the agency that promulgated the rule and the policy makers in that agency. It’s not an argument that you should be directing at those of us in the Solicitor General’s office, because we have a different function within the Executive Branch. Ours is not to second guess the underlying policy. What we are focused on is to try to make legal judgments about whether a case is cert.-worthy.” And of course I’m not suggesting that there aren’t fine lines. I’m not suggesting that there aren’t areas in which the policy and the legal judgment become difficult to separate. But I do think understanding that fundamental difference, that the proper role of the Solicitor General’s office in this kind of context is a legal one, not a policy one, I think is a very important aspect of the work of the Solicitor General’s office.

II. Separation of Powers Within the Executive Branch

A. Policy Making Versus Legal Decision Making

And more broadly, I think it’s a very important aspect of the separation of powers internally within the Executive Branch. And so to me, perhaps the first and foremost important aspect of this internal division is the separation of the policymaking function from the legal decision-making function. As I say, they can blur in application in many cases, but I do think there are fundamental differences, and there is a temptation inherent in the Executive Branch and inherent in the way that the Justice Department is structured for people who are dissatisfied with the result they get from the policy makers in an agency to try to get a second bite at the apple, as it were, before the lawyers at the Justice Department. Certainly my view is that if somebody doesn’t like, say, transportation policy as reflected in a rule, they should talk to those responsible for making transportation policy in the Executive Branch over at the Department of Transportation. They should not try to get their favored policy result through the back door at the Justice Department.

And I think this basic insight, that there is an important difference between the policy positions on the one hand and the legal positions on the other, applies with particular force during a time like this, where there is a transition between administrations. There is always when there is a transition, even a so-called friendly transition, but particularly when there’s a transition in the presidency between a President of one party and the other party—there is always a profound temptation to switch positions in some cases that are pending before the Supreme Court.

The work of the Solicitor General’s office, at one level, is remarkably unaffected by an upcoming presidential election. There are other aspects, principally the policy-making aspects, of the Executive Branch that, as a presidential election looms on the horizon, actual policy making slows down substantially. It may pick up after the election with some regulations issued before January 20th, but the actual policy-making work in the October before the election can very much slow down.

Not so for the Solicitor General’s office. October of an election year is as busy as any other October, and it’s very busy because the Supreme Court term is starting out. And right through October, right through November, right through December, right through January, up until the 20th, the work of the Solicitor General’s office continues and briefs are filed, and the briefs are filed consistent with the positions articulated by the administration that’s in power until January 20th. Then on January 20th at noon, another President comes in from a different party with presumably different views, certainly on some of the policy matters and maybe even on some of the legal questions involved in these cases. And so as I say, there is a profound temptation to change legal positions.

I am certainly not one to say that that is a temptation that should always be resisted. Elections certainly have consequences. But there is a right way in my view and a wrong way to alter the position in these cases. To simply change a legal position without any alteration in the underlying policy comes at a great cost to the credibility of the Solicitor General and the Office of the Solicitor General before the Court.

On the other hand, if the policy-making agency responsible for a particular area changes the underlying policy as reflected in some sort of authoritative legal source, like deciding to withdraw a regulation or promulgate a new regulation, and the Solicitor General’s office then informs the Supreme Court of that new legal development, that’s certainly a completely appropriate way to proceed. And as I say, elections do have consequences, so one would expect some policies to change.

And so, for example, the United States recently withdrew a cert. petition it filed in the environmental area, but that withdrawal of the cert. petition followed a formal change in position by the EPA. And so in that sense, there’s a degree of accountability with the EPA for the change of position, and it’s not a situation where the only thing that’s changed is the legal position of the Solicitor General’s office because of the change of administration.

Just to give another example, the Supreme Court recently granted cert. for quite an interesting case in the area of preemption in the context of the National Banking Act, a case called Cuomo vs. Clearinghouse from the Second Circuit. This case involves an unusual provision of law because it’s a preemption provision that doesn’t address what I would call substantive preemption, i.e., whether or not federal law crowds out state law. Really, the question is, given that this is an area where there’s some role for state law and federal law, who will do the regulating? Is it the federal regulators or the state regulators?

In a brief filed with the Supreme Court, a brief in opposition to granting certiorari, the Solicitor General near the end of the Bush Administration took the position, consistent with the longstanding position of the Office of the Comptroller of the Currency as embodied in a regulation, that only federal banking regulators had the authority to enforce both federal and state law when it comes to national banks. Given all the concern that perhaps the problem with national banks is not that they’ve been over-regulated, it’s certainly conceivable that the administration will want to take a different view in this case. And I guess the point I would simply submit is that there is a right way and a wrong way to do that. To simply leave the regulation on the books from the OCC and have the OCC position not change, and have the United States through the Solicitor General’s office take the position that, “Well, as a legal matter, the regulation that we thought we could defend before we no longer think we can defend under step two of Chevron,” would come at a high cost to the credibility of the Office of the Solicitor General. On the other hand, if the Office of the Comptroller of the Currency changes position and withdraws the regulation and promulgates at least a notice of proposed rule making for a new regulation that takes a different view, then that seems to me to be quite a different matter.

And I will talk in a little bit about one of the various values of maintaining this internal separation of powers, but to foreshadow a bit, the single most important benefit of maintaining this internal separation of powers is accountability. And in a situation where the OCC clearly and formally changes positions, then there’s accountability for that change of position where it belongs, with the people responsible for regulating our national banks, not with the Solicitor General. And whatever talents the Solicitor General may or may not have, he or she ought not to be the person who essentially takes the hit for a major change in posture in the banking regulatory area.

B. Counseling Versus Litigating

Although this division between the policy making and the legal decision making is one of the most important aspects of this internal separation of powers as it applies to the Justice Department and its functions, I certainly don’t want to leave you with the thought that it is the only aspect in which there is a division of authority. Let me talk about at least two other important divisions of authority within the realm of the Justice Department.

One is the difference between the counseling function and the litigation function, and these are even more closely-related, so the lines can be even more blurred. But nonetheless, I think it is worth thinking about them separately and trying to distinguish between the function that lawyers in the Executive Branch play when they are counseling the President and others in the Executive Branch, as opposed to the position that the Justice Department takes when the counseling has already taken place, Executive Branch action has occurred, and the government is in the posture of defending an executed Executive Branch policy in litigation.

These are closely related functions, and the closeness of the functions is illustrated by the fact that back before 1933, both of these essential functions were housed within the Office of the Solicitor General. There was a specific Assistant to the Solicitor General who was responsible until 1933 for providing this counseling function on an ongoing basis. About 1933—perhaps not entirely unrelated to some of the great expansions of Executive Branch authority during President Roosevelt’s time—there was a decision that the counseling function was simply taking too much time out of the work of the Solicitor General’s office. And so like Adam’s rib, the Office of the Legal Counsel was created out of the Office of the Solicitor General. That Assistant was taken, given a separate office, and eventually elevated to the position of Assistant Attorney General with a separate staff. So we’re now at the point where OLC is not only separate, but is about the same size as the original OSG.

Although they are closely related, the counseling function is indeed different from the litigation function. And I think that there are at least two important differences. One is that the counseling function often involves issues that, by their nature, are unlikely—not impossible, but unlikely—to result in litigation. So some of the most difficult questions that had to be asked and answered in the context of the global war on terror were issues that might be litigated, but there were substantial obstacles to the litigation. And so they might be issues where litigation was very unlikely, and consequently there weren’t a tremendous number of judicial opinions that were directly on point. And even in a slightly less combustible context, like the Ineligibility and Incompatibility Clause, counseling often involves issues where you often don’t get litigation.

The other difference, and this may be slightly more controversial, but I think at the end of the day this is correct, is that there is a different standard that is applied in the counseling function from that applied in the litigation function. In the counseling function it seems to me that the proper question to be asked is really a yes or no, right or wrong, sort of 50/50 question, if you will; i.e., is it lawful for the Executive Branch to engage in particular proposed conduct? If the answer is yes, then a legal opinion can be written and the conduct can take place. If the answer is no, it’s not lawful, a legal opinion can or cannot be written, but importantly, the conduct should not take place.

The question that litigators ask themselves within the Executive Branch is often different. The standard, for example, when the constitutionality of an act of Congress has been challenged, traditionally—and uniformly in and out of administrations of different stripes and different flavors—has been to say that the Department will defend the constitutionality of the act of Congress as long as a good faith or reasonable argument can be made in defense of the statute. And so you have situations where a lawyer in the Solicitor General’s office, perhaps if he or she abstractly considered the questions, “What’s the best view? What’s the right answer? What’s the best answer? Is this statute constitutional?,” might say, “No, the statute is probably unconstitutional. But are we going to go ahead and defend it anyway? Absolutely, because we have a different standard to apply.”

I think that’s a pretty well-established principle. What is a little less well-established, but also something that does govern the way that the SG’s office has conducted itself, at least in practice, is that there’s also a more deferential standard, even when it comes to defending an Executive Branch action: Suppose an agency has, perhaps after consultation with OLC, decided to promulgate a particular rule that raises a potential legal question. At that point it becomes challenged in court and there is litigation. Now typically I don’t think that the question that the Solicitor General asks himself—or, soon, herself—is, “Well, you know, did OLC get this right? I mean, as a straight-up legal matter, is the better view that this action is lawful?” Once it gets into the litigation context, there is more deference given both to the decision of OLC that’s already been made, and the decision of the agency to take this position and put it into positive law. Thus there is a degree of deference. I don’t think it goes all the way to a reasonable good faith standard as in the statutory context, but there is deference to the decision already made.

These two differences in the functions—one, the standard being different, and two, the idea that in the counseling function you’re often in a position where you don’t have litigated cases and judicial opinions to rely on—point out an interesting difference between the role of the head of the Office of Legal Counsel and the role of the SG. Both of these officers of the Executive Branch are put in the somewhat awkward position of occasionally saying no to the President, and that is, of course, a very difficult thing. There are circumstances in which the President, wanting to conduct a certain policy, will be told by the Office of Legal Counsel that it is simply not a lawful option. There are also, on occasion, situations where there is a statute or a policy that the Executive Branch favors, and the Solicitor General will make the judgment that it’s just not defensible; we’re not going to defend the constitutionality of the policy or the statute.

Because of the two differences that I’ve highlighted between the two functions, the job of the head of OLC, in saying no, is much harder. The head of OLC, first of all, often doesn’t have the ability to say no backed up by a sheath of legal opinions from the Justices of the Supreme Court that explain exactly why the head of the Office of Legal Counsel is correct to say no. Likewise, the head of the Office of Legal Counsel, at least if my conception of the job is right, is really asking a straight-up 50/50 question. And so there are likely to be more occasions to say no. If you’re applying a standard that says, “Is it lawful?,” you’re going to say no a fair amount. If you’re applying a standard, “Is it obviously wrong, or is it unreasonable?,” or whatever the standard might be in a particular context for the Solicitor General, you’re going to be saying no less often.

If this difference is true, there is an anomaly that’s worth pointing out. And let me be very careful in offering two caveats—I am not trying to comment on any of the people who are nominees for these positions right now, currently, or saying anything about the qualifications of people who have had these jobs in the immediate past. But I think if you look at it historically, the Solicitor General position is one that has often been filled by people who come to the position with a great deal of stature from prior experiences—present company excluded—but, often lawyers who come off the federal bench, from the deanships of major law schools, or who otherwise come to the Solicitor General’s office with a great deal of gravitas.

With the Office of Legal Counsel, the typical head of the Office has been somebody who is much earlier in their career, is somebody who is perhaps coming from a professorship at a law school but not the deanship of a law school. And so we have something of an anomaly: the person filling the position who has the stature to more easily say no to the President of the United States occupies the job where it’s easier to say no and you have to do so less often. And the person who comes into the job with a little more difficulty in saying no—just because of the gap in stature and where you are in the hierarchy of the Justice Department—is the one who by virtue of the structure of the office and the nature of the counseling function should be expected to say no more often.

Now is there anything to be done about this? I think it’s at least worth asking whether or not the Assistant Attorney General level is really the right level for the Office of Legal Counsel. One thing that has been a byproduct of some of the controversy over some of the advice that the Office of Legal Counsel has given over the past couple of years is that it’s allowed a greater number of people—not a large number, but a greater number of people—to actually know what the Office of Legal Counsel is. I think very few people had even heard of the Office of Legal Counsel until recently.

And if you hear about the Office of Legal Counsel and you understand its important counseling role within the Executive Branch, I think you have two reactions. One is, “Well, why haven’t I heard of this before? This is a very important function.” And the second reaction upon reflection, once you understand how important and potentially difficult a role the OLC plays, you might wonder whether an Assistant Attorney General is the right level in the hierarchy within the Executive Branch for this very difficult and important function.

There are two ways to address this situation. One, and the more difficult in the sense that it would require legislation, which is always difficult, would be to elevate—essentially raise up in the hierarchy—the position, perhaps make it an Associate Attorney General for the Office of Legal Counsel. It might not sound like much of a difference to people who haven’t been in the Justice Department, but raising the position up to that higher level actually would make, I think, a significant difference in terms of the stature, both within the Executive Branch and perhaps in terms of the ability perhaps to attract candidates of even greater stature than we’ve already had.

Another, perhaps less demanding, solution—and one that would not require legislation—would be simply to have more of the opinions of the Office of Legal Counsel become, once again, opinions of the Attorney General. There was a longstanding tradition for the Office of Legal Counsel to really have almost all of its work be opinions of the Attorney General. Before the office existed, all the work product that embodied these kind of counseling decisions were opinions of the Attorney General. It has now become much more the tradition of the office to issue opinions under the signature of the head of the Office of Legal Counsel, or even a Deputy Assistant Attorney General, and they are now accumulated in bound volumes of the Opinions of the Office of Legal Counsel, no longer the Opinions of the Attorney General. I wonder whether there isn’t something to be said for getting back to the tradition of making these opinions of the Attorney General, since the Attorney General presumably is somebody who clearly has the stature to be in a position to say no to the President on important matters.

C. Prosecuting Versus Appellate Decision Making

Let me talk about one last division within the functions of the overall Justice Department function, and that is the distinction—and this is the most subtle of all—between the prosecutorial decision-making process and the appellate litigation decisions that the Solicitor General’s office is principally responsible for.

The SG’s office makes decisions about two thousand times a year as to whether or not to appeal an adverse decision for the United States. So every time there’s an adverse decision in the district court involving a case that the Justice Department litigated, in order to take that case up to the court of appeals, the lawyers directly responsible for the case need the authorization of the Solicitor General. So too if there’s a loss in the court of appeals and the lawyers want to file an en banc petition, they need the authorization of the Solicitor General.

In considering whether to appeal, the general tradition in the office has been to consider the merits of the legal arguments, and to figure out whether or not the legal position embodied in a case is correct. In addition, for en banc review, the office considers whether the case satisfies the criteria for en banc review under Rule 35 of the Federal Rules of Appellate Procedure. But the tendency has not been, in doing that appellate review function, to consider directly whether this is a prosecution we should have brought in the first instance.

There certainly are situations where, looking at the exercise of prosecutorial discretion in the first instance, you may wonder: “Why was this prosecution brought?” There’s obviously a temptation in the Solicitor General’s office, if you think it’s a case where the prosecution should not have been brought, to “no appeal” the case, not because the legal position that’s embodied in the case is incorrect, but just because it seems like a prosecution that should not have been brought. To accede to that temptation is a mistake, because it is important to keep those lines of authority and decision making separate.

Now, fortunately, that doesn’t mean that the Solicitor General has to be the proverbial potted plant when he or she comes across a case where it looks like the case should not have been brought in the first instance. Sometimes it is a simple matter of injustice; it seems like a very unjust prosecution to have been brought. Sometimes it really is more a matter of strategy in terms of the development of the law in a particular area. Perhaps that’s where the Solicitor General has some expertise to bring to bear.

But while the Solicitor General can and should raise questions about whether the prosecution should have been brought, these questions should not necessarily dictate the distinct decision whether to appeal. The practice that I followed in that kind of situation was to pick up the phone and call either the U.S. Attorney or the Deputy Attorney General, who is the supervisor of the U.S. Attorneys for purposes of these kinds of prosecutorial decisions, and raise the question, not try to provide the answer—because I didn’t think that was my role—but to raise the question of whether or not this was a prosecution that we actually should have brought in the first instance. That phone call allowed the machinery that’s directed at that slightly different question to be brought to bear on whether or not we should withdraw the prosecution at that time.

There are obviously equities and decisions in that process that are delicate. One doesn’t just say, “Well we shouldn’t have brought the prosecution so we’ll drop it without considering any other factors.” But there certainly were situations during my tenure where we looked at a case in our office, thought there was really no problem with the legal issue, but also thought that the reason the district court rejected the government’s position on a legal issue where we clearly were correct, was precisely because the judge was responding to the intuition that this prosecution should not have been brought. So rather than blur the distinct questions between whether the case should have been brought and whether it should be appealed, we tried to respect the fact that those are different functions with different channels of authority.

III. The Virtues of the Internal Separation of Powers

I do think that maintaining and honoring those lines has some very important benefits. So let me get to that. I’ve talked a lot about respecting these lines of authority, this internal separation of powers in the justice department, separating out the legal from the policy, the litigation from the counseling, and even the appellate consideration from the prosecutorial function.

So why is it worth doing? What is the value of maintaining this internal separation of powers? Let me try to identify four benefits. One is good old-fashioned efficiency. Adam Smith would be proud: division of labor promotes efficiency. The President of any party presumably wants people in the policy-making agencies who are at some level or another policy-making experts. Every Solicitor General I’ve ever met is fond of reminding people that the Solicitor General is the only Executive Branch officer required by statute to be learned in the law. The Attorneys General are no longer required to be learned in the law (Solicitor Generals love that). And none of the general counsel is required to be learned in the law; it is only the Solicitor General that is required to be learned in law.

I point that out because, of course, the Solicitor General is not required by statute to be learned in policy, let alone learned in policy of all the myriad of different policies that are addressed by executive agencies.

So if the SG redirects inquiries to the policy makers at all the various different agencies, then there ought to be advantages in having the policy makers making the policy decisions for which they are appointed and well-qualified, and then to have the litigators make the decisions that they are, by statute, supposed to be learned about making.

There is really in practice no way that the Solicitor General could be informed and conversant in all the different policy issues implicated by the cases that come across his or her desk. So the challenge is to focus on the legal issues at which the Solicitor General is supposed to be quite conversant.

A second important advantage, even above and beyond mere efficiency, is that maintaining these lines of authority promotes good working relationships within the Executive Branch. That’s very important because, as I said at the outset, in the modern regulatory state, the President could hardly be a one-man band. More to the point, there’s so many people that need to work together in the Executive Branch to promote and effectuate the President’s agenda that promoting good working relationships among those people is something that’s extremely valuable.

Let me just use, as an illustration of this, the relationship between the Solicitor General’s office and line prosecutors and U.S. Attorneys Much, of course, has been written lately about the relationship between main Justice and the U.S. Attorney’s offices. The last couple of years were a very difficult period for those relationships, but the difficulties in the relationship really didn’t extend to the Office of the Solicitor General.

If you think about that, that’s remarkable because the Solicitor General’s office is in the constant business of saying “no” to U.S. Attorneys in cases they care about. There are all sorts of cases where the U.S. Attorney wants to appeal a case to the court of appeals and the SG ends up saying “no.” These are cases that are often very high-profile cases in the district; these are cases where if the prosecution was brought and all the resources of a prosecution brought to bear, a U.S. Attorney feels like he or she is pretty far out on the limb and feels that when the decision to not appeal that case is made, essentially you’re sawing off the limb behind them.

Notwithstanding the fact that there is this constant source of potential tension between the U.S. Attorney’s office and the Solicitor General’s office, that really has not manifested itself as a difficulty. The reason for that is not any particular magic; it is because at various points in the process there has been a conscious effort to maintain these various differences in lines of authority. There was an understanding that the SG’s office was not second-guessing the decision to bring the prosecution in the first instance, that the SGs were always focusing on the legal issues and whether they were appealable and not asking the question directly whether or not the prosecution vel non should have been brought.

If the SG blurred those lines and when there were questions did not refer them to the U.S. Attorneys and the Deputy Attorney General and allow that separate process to take its course, the sources for tension would have much more of a tendency to bubble over. More broadly I think this is a particularly good illustration of how separating out the various distinct functions and maintaining lines of authority can promote good working relationships, but that’s certainly true in other contexts as well. For example the relationship between OSG and OLC is much more difficult if OLC is of the view that OSG is constantly second-guessing the decisions that OLC has made and vice versa.

A third advantage is that maintaining these different lines of authority and separation of these different functions helps provide a framework for making decisions within the Executive Branch. If you at least try to separate out the policy making from the litigation, the counseling from the litigation, the appellate decision from the prosecutorial decision, you provide a basic framework for who has the authority to make the final decision within the Executive Branch, subject to whatever possibilities for appeal there may be.

So, for example, one role that is among the most difficult roles for the Solicitor General to perform is a situation where you have conflicting views within the Executive Branch on a matter where only one view can be presented to the Supreme Court of the United States. This happens quite frequently, for example, in the context of decisions by the Supreme Court to grant cert. in a private party case under an anti-discrimination statute: An employee sues an employer under Title VII, the ADA, or one of these other anti-discrimination statutes; it’s a private party case. It gets to the Supreme Court, the United States government has not been monitoring this case, it knows nothing about this case, the court grants it, and the United States has about forty days to decide whether or not it wants to file a brief in support of one party or the other.

In a typical case, the Civil Division will tell the Solicitor General: “There is a public sector analog to this anti-discrimination statute that applies to the federal government as employer. The federal government happens to be the nation’s largest employer, so we clearly have an interest in this issue, and we clearly need to file a brief on behalf of the employer.”

The difficulty comes in that typically the head of the EEOC or the head of the Civil Rights Division will send a memo around that says: “This is a very interesting and important case. This provision also applies to local public employers.” In that context the Civil Rights Division is the principal enforcement agency for this statute, so of course we should file an amicus brief in this case and of course it should support the employee.

So what’s a Solicitor General to do in these contexts? You have to sort out a position that ultimately reflects the best views of the Executive Branch as a whole. And in some of these cases, it is very helpful to isolate what is policy and what is just pure legal judgment. Sometimes that will help you in identifying something of a hierarchy of the positions of the various interested parties within the government.

If one entity within the Executive Branch has taken a position that reflects a well thought-out policy, that has been promulgated into positive law as a regulation, and for the other interested party this is simply a position they came up with recently, or maybe they’ve litigated it once in a brief but there’s nothing more formal than that, that difference does help provide a basis for making the decision.

The other thing, though, that’s quite helpful is to discern whether the dispute is essentially legal or essentially a policy dispute. If it’s an essentially legal dispute about how to interpret an earlier Supreme Court case, for example, then ultimately the view of the Solicitor General is, “That’s what I get paid for. I will make that decision, and in most cases I will have the final word in making that decision.” But if it really is a policy dispute, then it’s not necessarily a dispute for the Solicitor General to resolve.

The maintenance of these internal divisions in the various functions of the Executive Branch also provides a framework for dealing with the White House. And, obviously, relationships between the Justice Department and the White House are another very important issue that has been raised in recent months.

There’s always something of a conundrum in thinking about the Office of the Solicitor General: the Solicitor General is well-known as having a very close relationship with the Supreme Court of the United States. It’s manifested itself in things like the Supreme Court, roughly twenty times a year, calling for the views of the Solicitor General in a pending case at the cert. stage. The Court doesn’t call for the views of anyone else—only the Solicitor General. It is quite a close relationship, so close that some have referred to the Solicitor General as the “tenth Justice.”

Now I’m always quick to add that I never heard any of the real Justices refer to the Solicitor General as the tenth Justice, and I think Drew Days has said that it probably gets it closer to the truth in referring to the Solicitor General as the thirty-sixth law clerk. But whatever the precise relationship, there definitely is a close relationship there, and also a tradition of a degree of independence of the Solicitor General from the rest of the Administration.

But if you look at an organizational chart, you would wonder why in the world is that the case? The organizational chart makes perfectly clear that the Solicitor General works for the Attorney General, who in turn is responsible to the President. What that means is that in any case as a matter of executive power and constitutional theory the Attorney General and the President could overrule the Solicitor General even on a purely legal determination.

Although that’s possible as a matter of constitutional theory, I think if it happened in practice the Solicitor General would be well-advised to leave office quite soon. Although there is always the possibility of the Attorney General and the White House second-guessing, superintending, overruling the legal decisions of the Solicitor General, if that happened frequently in practice that would be an indication that the system was not working well.

On the other hand if the Solicitor General were to try to arrogate to himself or herself policy-making functions, in that case I think the overruling probably would be forthcoming, and in that case would be healthy. This brings to light the relationship between the SG’s office and the White House. In general, the concern among those who value the “independence” of the Solicitor General is that the White House is going to be calling up the Solicitor General and weighing in on pending cases. But, in my experience, when the system is actually working the way it should, the calls will be infrequent and they will actually go the other direction; I, from time-to-time, would identify cases where it seemed to me quite clear that what was going on was that two parts of the Executive Branch had diametrically opposite policy views. In at least a couple of cases, I thought both policy views were defensible in court. One might be a little easier to defend than the other, so if it were up to me I’d probably pick on that basis. But both were defensible positions. The difficulty was just that two parts of the Executive Branch were not singing remotely from the same sheet of music when it came to policy.

In that context I thought the appropriate thing to do was to call up the White House Counsel’s office and basically say: “Look, there’s a policy disagreement within the Executive Branch. I could resolve it—I have views on these matters; I’m not a potted plant but I don’t think that’s my role. If it were a legal dispute I’d be happy to resolve it. But since it is a policy dispute could the President, or whomever he delegates the function to, resolve the policy position of the United States on this matter and get back to me? Once I have that input, then I can make a determination about how to defend that policy position in court.” If you think about and are able to separate out those functions, it makes a very difficult job of OSG a little bit more manageable.

I also think that considering these various functions also provides some insights into that very interesting and important relationship between the White House and the Department of Justice, more broadly. As I’ve already noted, when you’re talking about pure litigation functions, I think that the interaction ought to be relatively infrequent; I don’t think there’s anything per se impermissible about it, but it ought to be relatively infrequent just because the Justice Department, after all, are the litigation experts.

When you come to the counseling function I don’t think there’s any way to avoid consistent interaction between the White House and the Justice Department. After all, it’s the attorney-client privilege in action, and so you need that interaction.

When you get to the prosecutorial function, the situation is very different. That’s an area where the interaction, subject to limited exceptions, ought to be essentially verboten, and the Justice Department really ought to be left, perhaps not as a matter of constitutional theory, but as a matter of good government, to its own devices in the prosecutorial realm.

I would say there is the potential for some interesting hybrid situations. To give one instance, I’ll point out the experience I had in the context of the Moussaoui prosecution. In the Moussaoui prosecution there were of course some extremely difficult legal issues that the Justice Department had to confront in the course of that prosecution, including the ones that I was the most directly involved in—some very difficult Sixth Amendment issues involving confrontation rights and compulsory process rights, specifically the right to an accused to get access to evidence from other people being held as enemy combatants by other parts of the Executive Branch, and perhaps in foreign countries.

The Justice Department was not looking for insight from others in the Executive Branch on these Sixth Amendment issues. Those seemed like the kind of issues that were in the core of the Justice Department’s competence. On the other hand, though, there was always the possibility in the Moussaoui case that he would be taken from the criminal justice system and designated an enemy combatant. That determination was not a determination that was within the bailiwick of the Justice Department; that was a policy decision that would be made by the President acting through the Defense Department.

There was therefore a need in that case for a degree of interaction between that policy question and the legal issues that we were dealing with in the Justice Department that you wouldn’t have in the average case; the Moussaoui case was anything but average, of course.

Let me just finish, then, by highlighting what I think is perhaps the most important advantage of this internal separation of powers, and that is accountability. Accountability is always an important value in separation of powers discussions, and is particularly important with respect to this internal division. If these basic divisions of function and responsibility are honored, then decision makers can be held accountable for the result of the decision making. The policy makers can be held responsible for their policy-making decisions, good or bad. The litigators can be responsible for their litigation judgments, good or bad, but you at least know who to blame. Counselors can be held responsible for the counseling judgments that they made in advising the President, and prosecutors can be responsible for their exercise of prosecutorial discretion.

Just to give one illustration of this from my time in government is a situation we had involving a potential appeal of an adverse decision by the Fourth Circuit in a case involving the constitutionality of a campaign finance statute. In the particular case in question, the Fourth Circuit had struck down the statute as applied to the contributions made by non-profit groups. This seemed to us in the Solicitor General’s office to be a particularly significant extension of the Supreme Court’s earlier decision, which said that these same non-profit entities had to be exempted from the expenditure limits. But to except them from the contribution limits seemed to be quite a different matter.

Quite remarkably, in our view, the Federal Election Commission made a recommendation that we not petition the Supreme Court of the United States in that case. It wasn’t clear why that made sense as a legal matter. There was a circuit split, and generally the finding by a lower court that a federal statute is unconstitutional is a clear case for the case to be taken up to the Supreme Court even in the absence of a circuit split. Nonetheless we got a recommendation not to take the case up.

In certain circumstances one would give a great deal of deference to an agency like the FEC that has independent litigating authority in the lower courts. But in this case it was pretty clear that the recommendation that the decision not be appealed was really a policy-making judgment disguised as a litigation judgment. The only reason not to petition for certiorari in that case was because you didn’t really like the statute that much in the first instance, and so you were perfectly happy for it not to be applied to these particular entities.

That was a case where the SG’s office took the very unusual step of petitioning in the case, in the name of the FEC, notwithstanding the fact that the FEC had voted not to take the case up. The only ending to the story that’s relevant is that the FEC, for better or worse, and against its will, prevailed 7-to-2 in the Supreme Court of the United States.

So that’s really what I had to say this morning. As I say, others in some of the panels later this morning and this afternoon can do a better job than I could discussing some of the theoretical underpinnings of both the internal separation of powers and the external separation of powers. But from my practical experience of seven years in the Justice Department, I do think there are tremendous virtues in trying to identify internal separation of powers, internal differences in the functions of the Justice Department, and honoring and maintaining those different internal lines of separation of powers.

Thank you very much.


43rd Solicitor General of the United States and head of the national appellate practice at King & Spalding LLP.