An Advisory Recall of U.S. Senators?

There’s a move afoot to recall Senator Robert Menendez (D-NJ). New Jersey law authorizes such recalls, but it seems likely that the U.S. Constitution does not allow a state to force the removal of a senator based on a recall vote. The Seventeenth Amendment provides that the Senators shall be “elected by the people thereof, for six years,” and the Constitution provides for no means by which this term can be shortened by the voters of a state. A state can no more force the removal of a Senator by recall than it can force its Senators to leave office after four years instead of six. (For more, see this Congressional Research Service report on the subject.)

Because removal of a sitting Senator by the voters is likely unconstitutional, the then-New Jersey Secretary of State, Nina Mitchell Wells, refused to approve the proposed petition (something the Secretary of State is apparently required to do before the signatures can be gathered). The question is now being litigated, with a hearing set for Feb. 26.

Here’s my question: Accepting, as I do, that a state’s voters may not remove an elected Senator, would it be proper to still allow the recall to go forward, and then have the result have an advisory effect, even if it can’t have a legally binding effect? In this respect, a majority “yes” vote on a recall would literally be a “summon[ing] to return,” though not a legally binding summoning: It would express the voters’ desire that the Senator resign and return to private life, and the Senator would then decide what moral and political effect this should have.

The argument would be this: There should be nothing unconstitutional about the voters’ expressing their views that someone should resign, even if the expression is done through the formal process of a state-organized vote. The only unconstitutional thing would be to try to force the Senator to give up his office. So the signature gathering and election (if enough signatures are gathered) should be constitutionally permissible, and should have the maximal constitutionally permissible effect — which could indeed be a significant political effect, even if not a morally binding one.

Such advisory recalls don’t take place these days. But they are authorized expressly by at least one state — Arizona (see Title 19, chapter 2, of the Revised Statutes, especially article 2). And such advisory recalls seem consistent with the relationship between Senators and state legislatures in the 1800s, when Senators were elected by state legislatures. From Josh Chafetz’s Leaving the House: The Constitutional Status of Resignation from the House of Representatives, 58 Duke L.J. 177, 214 (2008):

As Professor Riker puts it, “[e]lection by state legislatures implied accountability to them,” and the legislatures demanded this accountability via instructions on how to vote, a practice carried over from the Continental Congress. When a legislature had strong opinions about an issue pending before Congress, it would pass a resolution of the following form: “Be it resolved that our Senators in Congress are hereby instructed, and our Representatives are requested, to vote for ….”

Without the recall provision of the Articles of Confederation, the only formal sanction the legislature had was a threat not to reelect — but because senators’ terms were much longer than those of state representatives, this was an uncertain threat: the legislative majority that the senators disobeyed might no longer be in power when the senators were up for reelection. Remarkably, however, senators’ sense of honor seems to have served to enforce instructions in the early Republic. Beginning with John Quincy Adams in 1808, senators who were unwilling to follow instructions from their state legislatures were expected to resign. Although the practice was not universally followed, Professor Riker counts fifteen such “forced resignations” (and another six “almost forced” resignations) between 1808 and 1854.

Senators were thus seen as being under continued authority of the legislature that elected them; and while they couldn’t be forced to do something, they could be advised to do it — and they were expected to follow the advice or resign. If that authorized instructions to vote, it seems to me it could likewise authorize instructions to resign. (To be sure, the instruction to resign might be less fair to the Senator, since it doesn’t give him an explicit roadmap for keeping his office; but the principle strikes me as similar.) The Seventeenth Amendment provided that the voters, rather than the state legislatures, should elect Senators. But that would simply mean that the voters could likewise have the power to issue advisory instructions to Senators, and if necessary advisory demands to resign.

The proper question, I think, is the standard “severability” question — whether the New Jersey voters who enacted the recall would have preferred, if the binding recall is held unconstitutional, (1) to allow the recall to go forward with merely advisory force, or (2) to not have any recall vote as to federal officeholders. This is of course a hard sort of counterfactual inquiry to answer. But that’s what pretty regularly done by courts that are trying to figure out whether the invalidity of one part of a statute or a state constitutional provision (whether enacted by the legislature or by the voters) should lead to the invalidation of the entire thing. And it’s largely inevitable if we want courts to try to salvage as much of the provision as possible.

Here, I think, there are two good arguments against the advisory recall. One is cost: It’s not clear that the voters who were willing to spend state money and voter time to authorize a binding recall would have been willing to do the same for a merely advisory recall.

But the other, more serious, argument is the likelihood that an advisory vote would be a poor measure of actual public sentiment. Precisely because the election is merely advisory, many voters might not be inclined to spend their time and effort casting a vote. As a result, the results might be skewed — even more than a typical election skews them — in favor of those who feel especially strongly about the issue. And a sitting Senator might face something that appears like a vote of no confidence by the state’s voters, when it’s really a vote of a very small and unrepresentative segment of the electorate. (Thanks to Scott Rafferty for pointing this out.)

In any case, that’s my tentative thinking on the matter; but I thought I’d raise the general issue, because it’s likely to get lost in the discussion of whether a binding recall is constitutionally permissible. As I said, I think it’s unconstitutional, but that shouldn’t settle the question of whether the recall vote is constitutionally permissible.