This post is a late addition to the Constitution category. All the other ideas here have been ruminating in my mind for years, but this one is a fairly recent entrant. I am convinced, though, that it’s extremely important as well for USA version 2 (and perhaps for the survival of USA version 1).
The problem
In late 2017, the Washington Post/University of Maryland poll reported that 42% of Americans did not believe that Trump legitimately won the 2016 Presidential election. In late 2021, the same poll reported that 29% of Americans did not believe that Biden legitimately won the 2020 Presidential election. Assuming negligible overlap between these two sets of dubious citizens, this means that 71% of the American public only believe in the integrity of a close Presidential election when their preferred candidate won. This is consistent with this ABC/Ipsos poll, in which only 20% of Americans reported being “very confident” in the election system, and this international Gallup poll, in which 59% of Americans reported “no confidence” in the election system (among polled countries, only Chile and Mexico had lower reported confidence). Something must be done about this – it is not a sustainable situation in which a substantial supermajority (apparently, at least) links confidence in electoral integrity with the success of their particular candidate.*
This profound lack of confidence, however, only seems to apply to the Presidential election – the confidence results on any given state (at least among those I have examined) remain fairly high, regardless of what electoral integrity concerns may or may not be present in any given state.
Part of this, I think, is the result of civic ignorance. Media insistence on reporting “popular vote totals” (which are meaningless for a wide variety of reasons beyond just the fact that we don’t elect the President that way) as well as the rise of the execrable National Popular Vote Interstate Compact (which almost certainly violates at least the Compact Clause of the constitution). Every time the popular vote (as reported by the media) produces a different result from the actual (electoral college) vote, citizens are routinely seen to be surprised by this. They seemingly do not realize that the President is not elected by an election, but by an appropriate aggregation of the results of 50 separate elections. Furthermore, this is so only because all 50 states have chosen, through their state legislatures, to choose their electors by indirect popular election (as opposed to appointment, or games of chance, or choosing the oldest eligible electors in the state, or …)
The Constitution (Article II, Clause 2) specifies that:
Each State shall appoint, in such Manner as the Legislature thereof may direct, a Number of Electors, equal to the whole Number of Senators and Representatives to which the State may be entitled in the Congress: but no Senator or Representative, or Person holding an Office of Trust or Profit under the United States, shall be appointed an Elector.
Note that the state legislatures are given plenary power to appoint electors (more precisely, to direct the manner in which they shall be appointed). This power is not given to the governors, nor the Secretaries of State, nor any State Courts, but to the individual State Legislatures.
One reason for the problem
The 2020 Presidential election exhibited a number of strange features that hadn’t been observed before (or, at least, not in such a way that it might have affected the overall outcome). Principally, because of last-minute voting changes related (at least ostensibly) to the COVID-19 pandemic, at least six states** (Arizona, Georgia, Michigan, Nevada, Pennsylvania, and Wisconsin) held their Presidential elections in ways differing from the expressed directions of their legislatures. It is not my intent to argue that each (or any) of these changes affected the outcome of that state’s election (unknowable since the correct election wasn’t held); nor am I even asserting that there were more than the normal number of local irregularities in these elections than usual – those things always happen at some level. Rather, I am asserting that, in each of these cases, someone other than the State Legislature changed the “manner” of the appointment of electors in that state. Furthermore, all these changes appear to have moved each of these states’ vote totals in the same direction (toward the Democrat candidate).
This, frankly, doesn’t pass the smell test. These changes, together with the SCOTUS refusal to even listen to this argument, led (IMHO) to the increasing desperation of Trump supporters between the election and the electoral college vote. I would also argue that all of this could have been avoided if the State Legislatures in question had each passed a “close enough” resolution stating that the changes were acceptable minor ministerial alterations to the manner they had approved. Better yet (to avoid this in the future), each State Legislature should adopt, as part of their Presidential election code, a statement as to who, in the State government, is authorized to make an official pronouncement as to:
- whether or not the election was held in accordance with legislative directions, and
- what the result was (if a constitutionally valid election was held)
- what the backup procedure for appointing electors is, if the election was not valid
Note that a legislature may not wish to delegate this authority, and may require an explicit legislative stamp of approval before sending the slate of electors on to the Federal government.
At the moment, this procedure is governed by the Electoral Count Act of 1887 (which was designed to prevent a repeat of the 1876 election/debacle). The main problem with this status quo is that the statute in question is quite poorly drafted and may be unconstitutional for a variety of reasons. From my perspective, the most relevant problem is that, if there are multiple slates of electors from various entities in a state (but note that this is not what happened following the 2020 election – there were merely some contingent slates approved because of some pending litigation), it gives preference to the one certified by the state’s governor. This is, of course, constitutionally backwards, if the governor and the legislature are on opposite sides of a controversy. There will never be, however, an opportunity for judicial review of the Electoral Count Act because of the short time period between the election and the convening of the electoral college. There’s just not enough time to mount a constitutional challenge, with appropriate standing, where the issues are neither unripe nor moot (see here).
Suggestions (federal)
As a result of the 2020 election, there are a number of efforts at reforming the Electoral Count Act. While many of these are good approaches at eliminating the internal inconsistencies, none of them that I have seen address the fact that there is no requirement that a State Legislature (or proxy) certify that the election WAS conducted along the lines the legislature had specified. Proposals deal with requiring the legislatures to specify who is to certify the election count, but not the electoral procedures.
Given the time constraints in moving from state elections to the electoral college, and the rarity of events that truly invoke any of the electoral college machinery, the controlling language needs to be clear and indisputable. For this reason, I would suggest that USA version 2 incorporate into the Constitution itself, enough detail to resolve any disputes on the identity of electors, AND require (or at least incentivize) State Legislatures to predetermine a method for certifying both the method and result of any indirect election used for selecting electors.
As far as the actual election procedure goes, there should be very little mention of that in the Constitution or statute of the central government, since all elections are state elections (as they should be).
Suggestions (states)
On the topic of state procedures that support election integrity, I will point out two things. First, with regard to poll procedures – there are two fundamental ways in which a voter may be disenfranchised:
- A voter who should be allowed to vote is not allowed to vote for some reason
- A person who should not be allowed to vote is allowed to vote, thus effectively cancelling out some opposite (but otherwise valid) vote of another voter (note that this category includes multiple votes from an otherwise valid voter)
In designing a polling procedure, it is important to realize that Type 1 disenfranchisements are preferable to Type 2 disenfranchisements, for 3 reasons – transparency, short term cure, and long term cure.
- Transparency: when a Type 1 disenfranchisement occurs, the disenfranchised voter knows about it at the polls.
- Short term cure: normally when a voter is disallowed from voting, he is allowed to cast a provisional ballot, which may be included in the vote totals if the problem (lack of identity verification, for example) is cured in a short period of time.
- Long term cure: once a voter knows that he has been denied the franchise for a curable reason (perhaps lack of photo identification), this can be cured before the next election and no Type 1 disenfranchisement will happen again.
Type 1 disenfranchisements are thus easily detected and cured, while Type 2 disenfranchisements are difficult to detect and impossible to cure – the only solution is prevention by making requirements that risk a (detectable, curable) Type 1 disenfranchisement (ID requirements, etc.).
Second, with regard to vote tabulation: the counting of the votes should be done in a secure, transparent, auditable, accountable fashion. Paper ballots with routine machine counting (but possible to hand-count) and strict chain of custody requirements seem (to me and others – see here on mail-in ballots in particular) to provide the best combination of efficiency, audit trail and transparency.
Again, this should all be done at the state level. Any federalization of elections risks providing a single point of failure for election fraud.
* Following the original writing of this post, Lott and Smith (detailed study here and summary article here) produced a study showing just how shoddy the record-keeping was in at least the 2020 Presidential election: 14 swing states were identified, and the 100 largest counties in those states polled with regard to their compliance with record-keeping requirements in the Civil Rights Act of 1960 – 94% failed to follow the law, and in the other 6% who did follow the law there was an average discrepancy of nearly 3% between the number of reported votes in the election and the number of votes in the records. Not exactly a ringing endorsement for the integrity of our current election system!
** For completeness, here is a short summary of the facts – I’m not sure these are the best links for this purpose, but they are readily available:
- Arizona – 29-day registration deadline changed (Federal court)
- Georgia – signature verification requirements changed (Secretary of State)
- Michigan – absentee ballot applications mailed out to everyone (should be on request per statute)
- Nevada – machine inspection of signatures was allowed, instead of the human inspection required by statute
- Pennsylvania – signature verification omitted (Secretary of Commonwealth); elimination of candidate’s right to challenge illegal ballots during absentee ballot count (PA Supreme Court)
- Wisconsin – unmanned drop boxes used in contravention of statute (Secretary of State)
Comment on the 2020 Presidential Election
It appears to be nearly obligatory at this point to comment on the 2020 Presidential election and its aftermath. Since I mentioned it earlier, I will do so a bit more expansively here.
To get a sense of what the issues were surrounding the 2020 Presidential Election, I recommend reading the now-infamous Eastman memo (the full 6-page one, not the two-page summary), these two critiques of the legal reasoning of the memo, and Eastman’s response to one of the critiques (Bessette’s), together with an additional elucidation of the situation here.
My position is that none of these legal narratives is a locked-down certainty, and thus:
- The state legislatures in question should have clarified whether or not the elections in their states comported with their direction, or
- The murky situation between the poorly-drafted ECA and the Constitution (including the 12th Amendment) should have been decided by SCOTUS (I’m not a fan of judicial review, obviously, but it’s what we have at the moment).
Since there was neither the time (nor the courage on the part of the Robertson court) for approach 2, approach 1 was all that could have been done to handle things correctly.
Since that didn’t happen (i.e. a correct approach), the particular incorrect approach that was taken was that an all-hands-on-deck war against Trump and anyone who voted for him was declared by a coalition of most of the powerful institutions in American life (most of the government, most of the media, tech companies, etc.). One could argue (correctly) that Trump personally could have done much more to defuse the situation leading up to the January 6 violence at the Capitol, but I claim that, by that point, large swaths of the American populace believed that no one in their government was listening to them because, well, no one in their government was listening to them. To be completely clear, I am in no way saying that the violence was justifiable, only that it was understandable, foreseeable and (most of all) likely preventable.
The situation was that Trump supporters could “petition for redress of grievances” all they wanted, but no one in power was even going to listen to their arguments (various legal attempts through many different avenues were made that attempted to get a hearing at SCOTUS in time – all were either held up in lower courts until moot, or denied cert). To be sure, many arguments were based on attempting to show election fraud – this is, to my mind, a certain losing strategy: the election procedures (including any anti-fraud measures, or lack thereof) are a state matter, and fraud is extremely difficult to prove – the correct approach is to prevent it, not prosecute it. On the other hand, Texas v. Pennsylvania, et al. (though it was poorly drafted and contained tangential fraud allegations that diluted its main point) did address the central issue here – that some states did not follow the “manner” “directed” by their legislatures; it was denied cert and never heard on the merits. In my view, the best outcome would have been for this case to have resulted in a SCOTUS injunction against the four named states that their legislatures respond to the question of whether or not their elections were held in the “manner” they had “directed” (as the Constitution requires). Instead, none of the elite ruling class was interested in hearing what any of the unwashed deplorables had to say.
One is reminded of the line, “We have Petitioned for Redress in the most humble terms: Our repeated Petitions have been answered only by repeated injury.”