The Default Rule of Prosecutorial Discretion
by Michael Dorf
On Friday I participated in a terrific one-day conference on presidential powers, organized and MC'd by Prof. Mark Tushnet at Harvard Law School. The conference was recorded; when the recording is available, I'll post a link, perhaps along with some further commentary. For now, I want to focus on a question that arose a number of times during the day: When, if ever, do laws governing the primary conduct of private actors include an implicit instruction to the executive branch to enforce those laws to the full extent? I'll then tie that question to my past work with Prof. Buchanan on the debt ceiling.
The conference overall considered presidential powers in a variety of contexts, while the panel I was on considered issues of executive non-enforcement. DC Circuit Court of Appeals Judge Brett Kavanaugh chaired, with comments from Prof. Tushnet, UC-Hastings Law Prof. Zach Price, and me, followed by discussion. Prof. Price has written three important articles (this one, this one, and this one) staking out the positions that: the Take Care Clause imposes a general duty on the president to enforce the law; case-by-case exercises of prosecutorial discretion are consistent with that duty; so is internal executive branch policy guidance on enforcement priorities; but public categorical programs like the Obama Administration's approach to immigration (DAPA & DACA) transgress a constitutional principle holding that the president--by contrast with English kings in the years leading up to the Revolution--lacks a power to suspend the law; and finally, most challenges to presidential non-enforcement are non-justiciable.
I don't entirely agree with Prof. Price. In particular, I am less troubled by the generalization, formalization, and publicity for the exercise of otherwise permissible prosecutorial discretion than he is, because generalization, formalization, and publicity are virtues that counteract a vice of case-by-case prosecutorial discretion: its tendency to permit arbitrary or discriminatory enforcement. Nonetheless, as this blog post of mine from 2014 indicates, I share Prof. Price's broader view that there are limits to a president's ability simply to choose not to enforce the law. I also share his view that these limits are not generally suited to adjudication: the chief and perhaps only check on presidential non-enforcement is politics--at least absent congressional indication that it means to curtail prosecutorial discretion. (More about that last point in a moment.)
The rough consensus* of the participants in the conference was that the pending case of United States v. Texas will not test these fundamental principles, despite the fact that the Court's cert grant specifically added a question for the parties to address: whether the Obama administration's immigration policy violates the Take Care Clause? The Take-Care issue is effectively off the table, because the challenge is not to the withholding of deportation itself, but to other benefits that come with it, especially work authorization. The government brief argues that such work authorization is permissible because the relevant immigration statute (in provision (h)(3)) specifically empowers the executive to grant work authorization. Texas counters that the provision cited by the government does not in fact delegate any power to grant work authorizations to otherwise removable undocumented immigrants; rather, it refers to other provisions of law that allow for work authorization.
I don't have a view about which side is right in this argument, but I share the consensus* of the conferees that answering the statutory question will not require the Court to address fundamental questions about the duties imposed by the Take Care Clause or whether those duties are judicially enforceable against the executive. As Georgetown Law Prof. Marty Lederman most forcefully argued at the conference (as he has argued elsewhere), United States v. Texas is just a case of statutory construction.
This was an academic conference, however, and so, naturally, we addressed the fundamental questions, regardless of their irrelevance to any pending Supreme Court case. Much of the discussion either implicitly or explicitly posed the question whether precedents (even if only executive-branch precedents) set by President Obama with respect to policies the conferees approved would come back to bite us in a future administration. With respect to foreign affairs and war powers, Yale Law Prof. Bruce Ackerman was most worried about a President Cruz or Trump using the authority claimed by President Obama, while Harvard Law Profs. Charles Fried and Jack Goldsmith were most strongly skeptical of Prof. Ackerman's worries. As Prof. Fried noted pointedly, the biggest foreign policy disaster of our time--the Iraq War--occurred with congressional approval.
With respect to domestic affairs, Prof. Price sounded the strongest cautionary note--with me not far behind. As I have argued before, over the long run, progressives have more to fear from a conservative president's non-enforcement of laws we want to see enforced (e.g., environmental protection, gun control, health insurance) than we have to gain from a progressive president's non-enforcement of laws we would prefer to see repealed. In general, progressives want the government to act; conservatives don't. Further, even apart from that long-term ideological cost-benefit analysis, I share Prof. Price's view that giving the president a blanket non-enforcement power is fundamentally inconsistent with the constitutional design. (The president's pardon power is sometimes offered as the basis for a counter-argument, but there are limits to the pardon power: It only applies in criminal cases, and it only applies retrospectively.)
To be sure, the executive branch cannot possibly enforce all of the laws to their full extent. Accordingly, as Harvard Law Prof. Laurence Tribe explained in his keynote, the most logical inference is that Congress does not intend for all of its laws to be enforced all of the time. If it did--and if there were a presidential duty to fully enforce--then the sheer volume of law would put the president to the sort of choice that Prof. Buchanan and I say congressional failure to raise the debt ceiling would: the president would have to choose among unconstitutional options. (Prof. Tribe reiterated his disagreement with our particular resolution of the debt ceiling "trilemma," even as he commended our analysis of the general phenomenon.)
Prof. Buchanan and I have argued that the combination of the Impoundment Control Act of 1974, the Supreme Court's decision in the Line Item Veto Case, the non-delegation principle, and the failure of Congress to enact any guidelines for prioritization in the event that the debt ceiling constrains the government's ability to pay its bills, collectively lead to the conclusion that when Congress enacts laws saying the executive should spend $X, it means the executive should spend $X--not that the executive should spend as much money as it has on hand according to whatever prioritization scheme the executive invents. Thus, a debt-ceiling crisis would present a genuine trilemma.
By contrast, Prof. Tribe, Prof. Lederman, and some others at the conference thought that the opposite is true with respect to laws regulating private actors. The background assumption there is prosecutorial discretion, and so the enactment of new laws does not implicitly carry the message that the executive should enforce those laws to the full extent possible. One could take the position that an express directive from Congress to the president to enforce some law to the full extent (or to some specified extent) is an unconstitutional interference with prosecutorial discretion, but no one at the conference took that position. Instead, it was common ground that Congress could issue such a directive; it simply hasn't done so with respect to immigration, marijuana, and various other laws. And therefore, given resource constraints, prosecutorial discretion survives.
To say that, however, is not to end the discussion. It is one thing to say that resource constraints and other factors--including the failure of Congress expressly to limit prosecutorial discretion--leave the executive branch with substantial prosecutorial discretion. It is quite another thing to say that this discretion is unbounded. Everyone would agree that the discretion cannot be exercised in a way that violates equal protection, for example.
But even absent the use of some illicit criterion for the exercise of prosecutorial discretion, I remain troubled by the possibility that a president could simply choose not to enforce some law on the grounds that: (1) he lacks the resources to fully enforce all of the laws, which is always true; and (2) he especially dislikes the policy of the law. Maybe in theory it is a sufficient check on that kind of presidential power that Congress can, if it chooses, expressly command the president to enforce certain laws, but if so, that check is entirely theoretical. So far as I am aware, there is no such express command in any of the laws that a President Cruz or Trump would likely choose not to enforce. And in the event that we have a President Cruz or Trump, you can be sure we will not have super-majorities of Democrats in both houses of Congress who could override the president's veto in order to enact legislation commanding the president to enforce the existing laws.
So my message to my fellow liberal/progressive constitutional scholars who defend a broad conception of prosecutorial discretion is: Be careful what you wish for.
----------
*When I refer to a "consensus," I exclude the two judges who participated in the conference: In addition to Judge Kavanaugh, First Circuit Court of Appeals Judge David Barron was another conference participant. Both were careful to avoid taking positions on questions that they might face in their judicial capacity. In addition to the people mentioned in above, other panelists were Princeton political scientist Nolan McCarty, University of Minnesota Law Professor Heidi Kitrosser, and Harvard Law Professor Richard Lazarus. Attorney and Harvard Law Lecturer Susan Davies was scheduled to attend but was called away.
On Friday I participated in a terrific one-day conference on presidential powers, organized and MC'd by Prof. Mark Tushnet at Harvard Law School. The conference was recorded; when the recording is available, I'll post a link, perhaps along with some further commentary. For now, I want to focus on a question that arose a number of times during the day: When, if ever, do laws governing the primary conduct of private actors include an implicit instruction to the executive branch to enforce those laws to the full extent? I'll then tie that question to my past work with Prof. Buchanan on the debt ceiling.
The conference overall considered presidential powers in a variety of contexts, while the panel I was on considered issues of executive non-enforcement. DC Circuit Court of Appeals Judge Brett Kavanaugh chaired, with comments from Prof. Tushnet, UC-Hastings Law Prof. Zach Price, and me, followed by discussion. Prof. Price has written three important articles (this one, this one, and this one) staking out the positions that: the Take Care Clause imposes a general duty on the president to enforce the law; case-by-case exercises of prosecutorial discretion are consistent with that duty; so is internal executive branch policy guidance on enforcement priorities; but public categorical programs like the Obama Administration's approach to immigration (DAPA & DACA) transgress a constitutional principle holding that the president--by contrast with English kings in the years leading up to the Revolution--lacks a power to suspend the law; and finally, most challenges to presidential non-enforcement are non-justiciable.
I don't entirely agree with Prof. Price. In particular, I am less troubled by the generalization, formalization, and publicity for the exercise of otherwise permissible prosecutorial discretion than he is, because generalization, formalization, and publicity are virtues that counteract a vice of case-by-case prosecutorial discretion: its tendency to permit arbitrary or discriminatory enforcement. Nonetheless, as this blog post of mine from 2014 indicates, I share Prof. Price's broader view that there are limits to a president's ability simply to choose not to enforce the law. I also share his view that these limits are not generally suited to adjudication: the chief and perhaps only check on presidential non-enforcement is politics--at least absent congressional indication that it means to curtail prosecutorial discretion. (More about that last point in a moment.)
The rough consensus* of the participants in the conference was that the pending case of United States v. Texas will not test these fundamental principles, despite the fact that the Court's cert grant specifically added a question for the parties to address: whether the Obama administration's immigration policy violates the Take Care Clause? The Take-Care issue is effectively off the table, because the challenge is not to the withholding of deportation itself, but to other benefits that come with it, especially work authorization. The government brief argues that such work authorization is permissible because the relevant immigration statute (in provision (h)(3)) specifically empowers the executive to grant work authorization. Texas counters that the provision cited by the government does not in fact delegate any power to grant work authorizations to otherwise removable undocumented immigrants; rather, it refers to other provisions of law that allow for work authorization.
I don't have a view about which side is right in this argument, but I share the consensus* of the conferees that answering the statutory question will not require the Court to address fundamental questions about the duties imposed by the Take Care Clause or whether those duties are judicially enforceable against the executive. As Georgetown Law Prof. Marty Lederman most forcefully argued at the conference (as he has argued elsewhere), United States v. Texas is just a case of statutory construction.
This was an academic conference, however, and so, naturally, we addressed the fundamental questions, regardless of their irrelevance to any pending Supreme Court case. Much of the discussion either implicitly or explicitly posed the question whether precedents (even if only executive-branch precedents) set by President Obama with respect to policies the conferees approved would come back to bite us in a future administration. With respect to foreign affairs and war powers, Yale Law Prof. Bruce Ackerman was most worried about a President Cruz or Trump using the authority claimed by President Obama, while Harvard Law Profs. Charles Fried and Jack Goldsmith were most strongly skeptical of Prof. Ackerman's worries. As Prof. Fried noted pointedly, the biggest foreign policy disaster of our time--the Iraq War--occurred with congressional approval.
With respect to domestic affairs, Prof. Price sounded the strongest cautionary note--with me not far behind. As I have argued before, over the long run, progressives have more to fear from a conservative president's non-enforcement of laws we want to see enforced (e.g., environmental protection, gun control, health insurance) than we have to gain from a progressive president's non-enforcement of laws we would prefer to see repealed. In general, progressives want the government to act; conservatives don't. Further, even apart from that long-term ideological cost-benefit analysis, I share Prof. Price's view that giving the president a blanket non-enforcement power is fundamentally inconsistent with the constitutional design. (The president's pardon power is sometimes offered as the basis for a counter-argument, but there are limits to the pardon power: It only applies in criminal cases, and it only applies retrospectively.)
To be sure, the executive branch cannot possibly enforce all of the laws to their full extent. Accordingly, as Harvard Law Prof. Laurence Tribe explained in his keynote, the most logical inference is that Congress does not intend for all of its laws to be enforced all of the time. If it did--and if there were a presidential duty to fully enforce--then the sheer volume of law would put the president to the sort of choice that Prof. Buchanan and I say congressional failure to raise the debt ceiling would: the president would have to choose among unconstitutional options. (Prof. Tribe reiterated his disagreement with our particular resolution of the debt ceiling "trilemma," even as he commended our analysis of the general phenomenon.)
Prof. Buchanan and I have argued that the combination of the Impoundment Control Act of 1974, the Supreme Court's decision in the Line Item Veto Case, the non-delegation principle, and the failure of Congress to enact any guidelines for prioritization in the event that the debt ceiling constrains the government's ability to pay its bills, collectively lead to the conclusion that when Congress enacts laws saying the executive should spend $X, it means the executive should spend $X--not that the executive should spend as much money as it has on hand according to whatever prioritization scheme the executive invents. Thus, a debt-ceiling crisis would present a genuine trilemma.
By contrast, Prof. Tribe, Prof. Lederman, and some others at the conference thought that the opposite is true with respect to laws regulating private actors. The background assumption there is prosecutorial discretion, and so the enactment of new laws does not implicitly carry the message that the executive should enforce those laws to the full extent possible. One could take the position that an express directive from Congress to the president to enforce some law to the full extent (or to some specified extent) is an unconstitutional interference with prosecutorial discretion, but no one at the conference took that position. Instead, it was common ground that Congress could issue such a directive; it simply hasn't done so with respect to immigration, marijuana, and various other laws. And therefore, given resource constraints, prosecutorial discretion survives.
To say that, however, is not to end the discussion. It is one thing to say that resource constraints and other factors--including the failure of Congress expressly to limit prosecutorial discretion--leave the executive branch with substantial prosecutorial discretion. It is quite another thing to say that this discretion is unbounded. Everyone would agree that the discretion cannot be exercised in a way that violates equal protection, for example.
But even absent the use of some illicit criterion for the exercise of prosecutorial discretion, I remain troubled by the possibility that a president could simply choose not to enforce some law on the grounds that: (1) he lacks the resources to fully enforce all of the laws, which is always true; and (2) he especially dislikes the policy of the law. Maybe in theory it is a sufficient check on that kind of presidential power that Congress can, if it chooses, expressly command the president to enforce certain laws, but if so, that check is entirely theoretical. So far as I am aware, there is no such express command in any of the laws that a President Cruz or Trump would likely choose not to enforce. And in the event that we have a President Cruz or Trump, you can be sure we will not have super-majorities of Democrats in both houses of Congress who could override the president's veto in order to enact legislation commanding the president to enforce the existing laws.
So my message to my fellow liberal/progressive constitutional scholars who defend a broad conception of prosecutorial discretion is: Be careful what you wish for.
----------
*When I refer to a "consensus," I exclude the two judges who participated in the conference: In addition to Judge Kavanaugh, First Circuit Court of Appeals Judge David Barron was another conference participant. Both were careful to avoid taking positions on questions that they might face in their judicial capacity. In addition to the people mentioned in above, other panelists were Princeton political scientist Nolan McCarty, University of Minnesota Law Professor Heidi Kitrosser, and Harvard Law Professor Richard Lazarus. Attorney and Harvard Law Lecturer Susan Davies was scheduled to attend but was called away.