The Constitution is a Disaster, But Designing a Sound Impeachment Rule for a Polarized Country Would be a Challenge Anyway
by Michael C. Dorf
After an assault on the democratic order, it is understandable to glorify the legal basis for that order, especially in the United States, which tends to Constitution-worship even in the best of times. Yet a moment's reflection reveals that the despicable attack on democracy by Donald J. Trump and those in the thrall of his personality cult was not merely an attack against the Constitution; it was facilitated by the Constitution's flaws.
Here are some features of our Constitution that combined to enable Trump's assault: (1) The Electoral College--our byzantine system for choosing a chief executive whose current powers are better suited to one elected by plebiscite---not only distorts campaigns to focus on a handful of swing states at the expense of voters in all the others but periodically turns what should be easy-to-call election-night outcomes into nail-biters that then make contests over a relatively small number of votes consequential. Elections that are not close thus fall within what we might call the "margin of propaganda." (2) The Electoral College also enables the loser of the majority vote to win the presidency "legitimately." (3) The decentralized nature of our election system (enshrined by Article II) creates multiple pressure points for threats of primary challenges or physical violence against elected officials if they do not "find" enough votes for the would-be autocrat. (4) The dispute-resolution mechanism set forth in Article II for a contested election disproportionately favors rural voters who are vastly over-represented in the Senate and in House delegations when counted (as they are under that mechanism) on a one-vote-per-state-delegation basis. (5) The super-majority requirement for a successful impeachment combines with party politics to prevent accountability for even the most outrageous conduct, so long as the demagogue retains support of about 30% of the population, which suffices to win primary elections. (6) The super-majority and state-centered nature of the Article V amendment procedure make it nearly impossible to change any of the foregoing.
To recognize that the Constitution is profoundly flawed is not to attribute all the blame to its framers, deeply flawed though they were in their desire to protect or acquiesce in the protection of the institution of human bondage. For despite the framers' moral blindness, they designed a system that has many admirable features. The core issues that continue to bedevil us might be thought to be problems of under-correction: (1) The election of 1800 revealed that the Constitution the framers designed for a nation without political parties fit very awkwardly with a nation with such parties, but the Twelfth Amendment corrected only one small aspect of that mismatch. And (2) In the wake of the Civil War, the Reconstruction Amendments left far too much of the ancien regime in place and, in one crucial respect (repeal of the 3/5 clause) actually enhanced the long-term political power of the most racist elements of the country (by increasing the representation in the House of the states of the former Confederacy).
If I had a magic wand, I would wave it to replace our barely-amended 18th-century slavery-preserving constitution with something much better suited to a multicultural 21st-century polity. Of course, neither I nor anyone else has such a wand. In future essays, I hope to provide some suggestions for measures that can achieve some tangible progress despite our profoundly flawed legal order. For today I want to turn my attention now to a genuinely difficult question of constitutional design that we would face even if we could amend our Constitution: what should the impeachment mechanism look like?
During the second impeachment trial of Donald Trump, his defenders and apologists warned that his conviction would set the bar too low for future impeachments. The warning was preposterous, of course. If seeking to extort a foreign power to fabricate dirt on a political opponent (Trump's first impeachment) and inciting a mob to attack the legislative branch of government in a barely failed coup attempt (his second impeachment) aren't impeachable offenses, then as Rep. Raskin rightly contended, nothing is.
And as a practical matter, it does appear as though nothing is an impeachable offense. Partisan polarization and party loyalty make the 2/3 threshold for conviction in a Senate trial impossibly high. Yes, it was impressive that seven Republicans nonetheless joined all fifty Democrats to recognize Trump's offense, but given the magnitude of Trump's high crime, the fact that 43 Republicans nonetheless voted to acquit him indicates that impeachment cannot be an effective tool pretty much ever, so long as a president or ex-president remains popular enough in their own party to swing primary elections.
To be sure, one might say that actually Trump would have lost the support of enough Republicans to result in conviction had he still been in office. Perhaps another ten such votes would have then materialized? After all, Senators McConnell and Portman pretty clearly indicated that they would have voted to convict if Trump were still in office.
Hah! It's hard to take that claim seriously coming from McConnell, given that he chose not to call the Senate back into session during the final days of Trump's term. And it's blindingly obvious that even if we give credit to McConnell and/or Portman on this score, the rest of the Republicans who voted to acquit were using the very bad argument that an ex-President can't be convicted in an impeachment trial as a makeweight to avoid having to take a position on Trump's conduct. But if Trump were still in office those same Senators would have relied on some other bogus explanation for their vote--perhaps the First Amendment or some extravagant claim about due process or whatever. Only a naif could possibly think that Republican Senators who voted to acquit Trump were guided by anything other than a political calculation about the relative risks of a primary challenge and the general election.
If we had that magic wand, should we wave it to lower the impeachment conviction threshold to a mere majority of the Senate? Maybe, but doing so would create a genuine risk of the sort that Trump's apologists disingenuously warned this time around. Whenever one party controlled the presidency and the other party controlled both chambers of Congress--which not infrequently occurs after a midterm election--there would be a risk that Congress would impeach the president based on mere policy differences.
History might be read to suggest that we need not worry about too-easy impeachment. The impeachment of Andrew Johnson, the aborted impeachment of Richard Nixon, and the two impeachments of Donald Trump were all sensible. The impeachment of Bill Clinton was questionable though within the realm of reasonable disagreement about what constitutes an impeachable offense. The Johnson impeachment was arguably over a policy difference, but it was such a momentous one--whether to pursue Reconstruction with vigor--as to rise to the level of a neglect of duty on Johnson's part. So is the worry unfounded?
I don't see how we can rely on our history to rule out the possibility of impeachments rooted in mere policy differences, because our history has unfolded in a world in which the operative rule was a requirement of 2/3 to convict. And the Johnson impeachment occurred at a time when Republicans had a supermajority in the Senate. I noted a moment ago that impeaching Johnson was acceptable given the nature of the policy difference, but it is possible to read the episode as showing that a dominant party will attempt impeachment when it can succeed based on its votes alone. If that's true when the threshold is 2/3, it will also be true when the threshold is a simple majority. Thus, I think it a fair inference that lowering the threshold for impeachment to a simple majority would greatly increase the risk of an impeachment over mere policy differences.
One response to that risk might be to accept it. Permitting a simple majority in Congress to remove the president would be the rough equivalent of a parliamentary vote of no confidence in a prime minister, which can occur in other constitutional systems that are broadly democratic. What would be so bad about that?
Maybe nothing, but the U.S. does not have a parliamentary system in general. And while it is possible to have a "mixed" presidential/parliamentary system, creating one seems like a task to undertake from the ground up, rather than as a somewhat accidental consequence of changing the impeachment rule.
We might alternatively think that pure policy differences won't give rise to impeachment because a President's Vice President will simply continue the same policies after taking office following a successful impeachment. That wouldn't have been true had the Johnson impeachment succeeded because Johnson had no Vice President, but in the wake of the 25th Amendment's provision for the appointment of a Vice President should the office become vacant (as it did for Ford and Rockefeller), there is little to be gained from removing a President.
Yet that argument hardly seems persuasive. Republicans impeached Bill Clinton in 1998, even though there was scarcely any policy daylight between him and Al Gore. Indeed, they did so even though Clinton's removal would have enabled Gore to run for election in 2000 as an incumbent, which arguably would have made him a stronger candidate. And of course, there will be circumstances in which a President and Vice President from different wings of the same party could be predicted to pursue different policy programs. So I don't think we can count on the Vice Presidency to eliminate or even greatly reduce the risk of a simple-majority rule for Senate conviction making impeachment into a tool used for policy-based impeachments.
I'll therefore conclude by suggesting two other directions in which we might go if we conclude that impeachment is effectively unavailable but that making it more available (if that were even possible, which it isn't, due to the very high threshold for constitutional amendment) would bring its own problems.
One approach would be to reinvigorate other mechanisms for holding presidents accountable. The Independent Counsel Act that was adopted after Watergate and that the Supreme Court upheld in Morrison v. Olson was permitted to sunset after Ken Starr abused it. An internal executive-branch mechanism sufficed to produce the Mueller investigation, but Mueller's acceptance of the Justice Department view that a sitting president cannot be indicted meant that his findings were first distorted by Bill Barr and then ignored by a too-Trump-friendly Congress. The Supreme Court might or might not allow a more robust Independent Counsel Act, but we won't know if Congress doesn't try to create one (which would likely require changing the cloture rule, itself a huge problem).
Another approach would be to see the ineffectiveness of impeachment as a product of polarization and to tackle that problem directly. Polarization is partly a result of gerrymandering, but while that explains some of what we see in state legislatures and the House, it doesn't explain the Senate. A comprehensive account of the causes, consequences, and means of reducing the ill effects of polarization is obviously beyond the scope of a blog post. For a good introduction, I recommend a 2015 volume edited by Prof Nate Persily. At least some of the proposals set forth in that book appear to be achievable without a constitutional amendment but would require substantial political will.