9.15 Myths about Recounts

Criticism of the National Popular Vote plan in connection with recounts is an example of a criticism that actually applies more to the current state-by-state winner-take-all system than to the National Popular Vote plan. As explained below, recounts in presidential elections would be far less likely to occur under a national popular vote system than under the current state-by-state winner-take-all system.

Indeed, the question of recounts comes to mind in connection with presidential elections only because the current system so frequently creates artificial crises and unnecessary disputes. If we were debating the question of whether to elect state Governors by a popular vote, the issue of recounts would never even come to mind, because everyone knows that recounts rarely occur in elections in which there is a single pool of votes and in which the winner is the candidate who receives the most popular votes.

Tara Ross, an opponent of the National Popular Vote plan, has stated:

“The Electoral College typically produces quick and undisputed outcomes.”[376] [Emphasis added]

Ross has also said:

“The Electoral College encourages stability and certainty in our political system. Events such as those that occurred in 2000 are rare.”[377] [Emphasis added]

In testimony before the Alaska Senate, Ross stated:

“A direct election system … would result in … constant recounts.”[378] [Emphasis added]

Nothing could be further from the truth.

In fact, it is the current state-by-state winner-take-all system (i.e., awarding of all a state’s electoral votes to the candidate who receives the most popular votes in the state) that regularly produces artificial crises in the form of unnecessary recounts and disputes.

There have been five litigated state counts in the nation’s 57 presidential elections between 1789 and 2012 under the current state-by-state winner-take-all system. This rate is dramatically higher than the 1-in-185 chance (documented below) of a recount in which there is a single statewide pool of votes and in which the winner is the candidate who receives the most popular votes.

The current state-by-state winner-take-all system repeatedly creates artificial crises because every presidential election provides 51 separate opportunities for a dispute. This fact is illustrated by examining the five litigated state counts in the nation’s 57 presidential elections between 1789 and 2012. All five were artificial crises that would not have arisen if there had been a single large national pool of votes and if the winner had been the candidate who receives the most popular votes.

The 2000 presidential election was an artificial crisis created because of George W. Bush’s lead of 537 popular votes in the state of Florida. Gore’s nationwide lead was 537,179 popular votes—1,000 times larger than the disputed 537-vote margin in Florida. Given the miniscule number of votes that are changed by the typical statewide recount (about 294 votes), no one would have requested a recount or disputed the results in 2000 if the nationwide margin of 537,179 had controlled the outcome. In the absence of the state-level winner-take-all rule, no one (except perhaps almanac writers and trivia buffs) would have noticed that one particular candidate happened to have a 537-vote margin in one particular state.

In 1960, there was a recount and a court case that reversed the original outcome of the presidential race in Hawaii. Kennedy ended up with a 115-vote margin in Hawaii in an election in which his nationwide margin was 118,574.

Samuel Tilden’s 3% nationwide lead in 1876 was a solid victory in terms of the national popular vote (equal, for example, to George W. Bush’s nationwide percentage lead in the 2004 election). However, an artificial crisis was created because of the razor-thin margins of 889 votes in South Carolina, 922 in Florida, and 4,807 in Louisiana.[379] Few would have cared who received more popular votes in these three closely divided states if the President had been elected by a nationwide popular vote (which Tilden won by 254,694 votes). Again, the state-by-state winner-take-all system created an unnecessary artificial crisis.

Let us start with the facts about how rare recounts are in actual practice, how few votes are actually changed by recounts, and how few recounts actually change the outcome of an election.

FairVote has collected data on every statewide general election in the 13-year period from 2000 to 2012.[380]

There were 22 recounts in 4,072 statewide general elections between 2000 and 2012—that is, one recount for every 185 elections.

Table 9.15 shows, by year, the number of statewide general elections and recounts in the 13-year period from 2000 to 2012.

Table 9.15 The 4,072 statewide general elections 2000–2012 by year

Table 9.16 shows a breakdown according to the particular elective office or ballot proposition involved in the 4,072 statewide general elections and 22 recounts in the 13-year period between 2000 and 2012.

Table 9.16 The 4,072 statewide general elections 2000–2012 by type of election

Table 9.17 provides details about the 22 recounts of statewide general elections in the 13-year period between 2000 and 2012. The recounts in the table are arranged according to the absolute value of number of votes changed by the recount (shown in column 1). Columns 2, 3, and 4 identify the recount. Column 5 shows whether the original count was upheld or reversed. Column 6 shows the original margin, and column 7 shows the margin after the recount. Column 8 shows whether the recount was a candidate-requested or an automatic recount (explained below). Details concerning Wyoming’s constitutional amendments A and C of 2004 are explained in a footnote.[381]

Table 9.17 The 22 recounts of statewide general elections 2000–2012

The average change in the margin of victory as a result of a statewide recount was a mere 294 votes. This number is obtained by averaging the absolute value of the “change in vote” numbers found in column 1 of table 9.17.

As can be seen, the number of votes changed by a statewide recount between 2000 and 2012 ranges from 3 to 1,274.

All but two of the 22 recounts resulted in only a three-digit change in the original count, and the largest change was a change of 1,247 votes.[382]

As one would expect, half (11 of the 22) of recounts increased the apparent winner’s margin, and half decreased it.

The original outcome was reversed in only three of the 22 recounts—that is, about 1-in-7 recounts.

All of the recounts in which the original outcome was reversed had one thing in common, namely a low-three-digit original margin (specifically 137, 215, or 261 votes).

The three recounts that reversed the original outcome were:

• the 2004 Governor’s race in Washington state (where the original 261-vote lead became a 129-vote loss),
• the 2006 state auditor’s race in Vermont (where the original 137-vote lead became a 102-vote loss), and
• the 2008 U.S. Senate election in Minnesota (where the original 215-vote lead became a 225-vote loss).

The probability of a national recount can be estimated from the known probability of statewide recounts.

Using the 1-in-185 chance of a recount, and given that there is one presidential election every four years, one would expect a national recount about once every 740 years under a national popular vote (that is, four times 185).

In fact, the probability of a presidential recount under a national popular vote system would be even less than 1-in-185 (that is, even rarer than once in 740 years) because a close result is less likely to occur as the size of the voting pool increases.

This 1-in-185 frequency of problematic elections is dramatically lower than the five litigated state counts in 57 presidential elections up to 2012 that we have experienced under the current state-by-state winner-take-all system.

The reason there have been so many disputes in the mere 57 presidential elections is that there are 51 separate opportunities for recounts in every presidential election under the current state-by-state winner-take-all system. Our nation’s 57 presidential elections between 1789 and 2012 have really been 2,237 separate state-level elections. Thus, the current system repeatedly creates artificial crises in which the vote is extremely close in certain states, but not at all close on a nationwide basis.

One good way to visualize the difference between the two systems is to think of the chance of a recount as being a game of Russian Roulette in which there is one bullet in a 185-chamber gun. Under a national popular vote, the trigger is pulled once every four years. Based on history, we can reasonably expect the gun to fire once every 740 years (185 times 4). In contrast, under the current state-by-state winner-take-all system, the trigger is pulled 51 times every four years. Thus, we should not be surprised to have had so many litigated state counts in 57 presidential elections between 1789 and 2012. The trigger was pulled 2,237 separate times in 57 presidential elections under the current state-by-state winner-take-all system.

As previously mentioned, one would expect a national recount of a presidential election about once every 740 years under a national popular vote (based on the 1-in-185 chance of a recount and given that there is one presidential election every four years). When that exceedingly rare event occurs, it will also almost certainly be true that the results in one state (and probably several) would also be closely divided. Thus, if the nationwide count were extremely close, the current state-by-state winner-take-all system would very likely also produce a disputed count in one or more closely divided states.

Despite the fact that the average number of votes changed in a recount is very small (a mere 294 votes), it is common on Election Night for disappointed candidates who have lost by thousands of votes to bombastically announce that they are going to demand a recount. However, in the cool light of day, these candidates almost always realize that they have no realistic chance of reversing the outcome.

For example, in a race in North Carolina with a vote gap of 6,658 (0.15% of the 4,368,598 votes cast):

“The trailing Democratic Party candidate for [North Carolina] lieutenant governor said Monday she won’t seek a statewide recount, admitting that a new tally was unlikely to make up the nearly 6,900 votes she needs.…

“‘We face the reality that an extended battle would not alter the outcome of this race,’ Linda Coleman said at a news conference after conceding the outcome to Republican Dan Forest. ‘It was a hard-fought, spirited campaign and we have stark differences. But in the end, in a tight race, North Carolinians have chosen Mr. Forest as their next lieutenant governor.’

“Coleman had until Tuesday to demand a recount because her margin with Forest was less than 10,000 votes out of almost 4.4 million cast.…

“‘I don't think the money factor was an issue with her in making this decision,’ [Coleman spokesman] Beasley spoke said. ‘We just were faced with the reality of the numbers and that it’s hard to flip 6,000 votes in an extended recount battle. She [Coleman] doesn’t want to put the people of North Carolina and the state board of election employees through that.’”[383] [Emphasis added]

Similarly, Sandy Welsh decided not to pursue a recount in 2012 of her 2,231-vote loss in her race for Montana Superintendent of Public Instruction (a difference of 0.48% of the 468,563 votes cast).[384]

Not all recounts are conducted because the apparent losing candidate believes that he or she has any realistic probability of changing the result of the initial count. Nineteen states provide for “automatic” recounts of elections that are triggered because the original difference between the candidates is less than some pre-specified statutory percentage or numerical trigger. One reason that states conduct automatic recounts is to maximize public confidence in elections. Another reason is that recounts provide state officials and the public with the periodic opportunities to audit and evaluate the operation of the state’s election process. The government pays for automatic recounts. The percentage or numerical trigger for an automatic recount varies considerably among the 19 states. In many of the 19 states, an automatic recount will be conducted if the difference in the initial count between the first-place and second-place candidate (or ballot alternative, in the case of ballot propositions) is less than 0.5% of the votes cast.[385] Several states mandate automatic recounts with even larger differences. In many automatic recounts, no one (including the apparent losing candidate) realistically expects the outcome to change.

About two-thirds of the 22 recounts (14 of 22) in table 9.17 (column 8) were “automatic” recounts (as opposed to candidate-requested recounts).

Table 9.18 presents the percentage change in votes (column 1) that resulted from the 22 recounts of statewide general elections between 2000 and 2012 (table 9.17). Column 6 of the table here shows the total votes cast for the office or ballot proposition. Column 7 shows the percentage lead of the winner of the initial count (that is, the numerical lead shown in column 6 of table 9.17 divided by the total number of votes cast as shown in column 6 of this table). Many of the 19 states that conduct automatic recounts use this percentage as the criterion for deciding whether to conduct an automatic recount. Column 1 shows the percentage change resulting from the recount (that is, the number of votes changed in the recount as shown in column 1 of table 9.17 divided by the total number of votes cast as shown in column 6 of this table). The recounts in this table are arranged according to percentage in column 1.

Table 9.18 Percentage differences and changes resulting from the 22 recounts 2000–2012

As can be seen from column 1 of table 9.18, the percentage change that resulted from recounts is very small. Only one of the 22 recounts changed more than 0.1% of the original vote. The percentage change that resulted from 22 recounts ranged between:

• 0.0003%—that is, 3 votes in 1,380,750 (in Alabama) and
• 0.1073%—that is, 137 votes in 222,835 (in Vermont).

Recounts appear to be becoming rarer in recent years.

There were no recounts at all among the 419 statewide elections in November 2012.

The frequency of recounts has dropped by about half in recent years. As can be seen in table 9.15, there were twice as many recounts (15 of the 22) between 2000 and 2006 (the top half of the table) than in the six-year period represented by the bottom half of the table (7 of the 22). This apparent decline may be the result of the nationwide efforts made since 2000 to improve election administration and equipment, including enactment of the Help America Vote Act (HAVA) in 2002. One major improvement in election equipment is the drastic reduction since 2000 in the use of punched-card voting (with their hanging chads).

Improved technology can be expected to further reduce the need for recounts in coming years.

For decades, bank tellers, credit-card processors, and state lotteries have routinely and accurately handled and accounted for far more transactions than the mere 130 million votes cast in a presidential election on Election Day. Every day, approximately 145 million lottery bets,[386] 300 million bank-teller transactions, and 60 million credit card transactions[387] are accurately handled and accounted for in the United States. In particular, bets in state-administrated Lotto games resemble voting in that they require the accurate recording of the player’s selection of six (or so) numbers (typically on a paper slip). Unfortunately, the technology for administering elections and handling ballots in the United States is several decades behind that of the banks, credit-card processors, and state lotteries. No doubt, the technology for more accurately handling and accounting for large numbers of votes will catch up with that of these other industries during the next decade or so (and certainly sometime in the next 740 years) so that worries about recounts become a thing of the past.

Footnotes

Brendan Loomer Loy claims that the current state-by-state winner-take-all system acts as a helpful firewall that

“isolate[es] post-election disputes to individual close states.”[388]

Far from acting as a firewall that helpfully isolates recounts to particular states, the current state-by-state winner-take-all system is, instead, the repeated cause of unnecessary fires.

Under the current system, there are 51 separate statewide vote pools in every presidential election. Thus, our nation’s 57 presidential elections between 1789 and 2012 have really been 2,237 separate state-level elections. These 51 separate pools regularly generate 51 separate opportunities for artificial crises in elections in which the vote is not at all close on a nationwide basis (but close in particular states). This is why there have been five litigated state counts in the nation’s 57 presidential elections between 1789 and 2012. This rate is dramatically higher than the historical 1-in-185 rate for elections in which there is a single statewide pool of votes and in which the winner is the candidate who receives the most popular votes.

If anyone is genuinely concerned about minimizing the possibility of recounts, then a single national pool of votes provides the way to drastically reduce the likelihood of recounts and eliminate the artificial crises that are regularly produced by the current state-level winner-take-all system.

Trent England (a lobbyist opposing the National Popular Vote compact and Vice-President of the Evergreen Freedom Foundation of Olympia, Washington) has written:

“Containing elections within state lines also means containing election problems. The Electoral College turns the states into the equivalent of the watertight compartments on an ocean liner. Fraud or process failures can be isolated in the state where they occur and need not become national crises.[389] [Emphasis added]

The current system does not contain and isolate problems but instead creates artificial crises.

Footnotes

Brendan Loomer Loy warns that if we were to have a nationwide popular vote for President:

“Post-election uncertainty could stretch well into January, raising doubt about whether we would have a clear winner by inauguration day.…”

“With two centuries of legal precedent tossed aside, courts would have a very difficult time managing it all.”[390]

Loy’s scenario of a prolonged and unsettled election is based on the incorrect assumption that the existing U.S. Constitution, existing federal statutes, and existing state statutes would somehow be “tossed aside” under the National Popular Vote compact. In fact, the National Popular Vote compact was specifically drafted so as to operate within existing constitutional and statutory provisions in the same way that the current system does.

Finality of presidential elections would be ensured under the National Popular Vote compact by the same machinery that applies to the current system, namely the existing U.S. Constitution, existing federal statutes, and existing state statutes.

The U.S. Constitution establishes a strict overall national schedule for finalizing the results of a presidential election. These existing provisions would apply to elections conducted under the proposed National Popular Vote legislation in the same way that they apply to elections conducted under the current system. No prolonging of a U.S. presidential election until January is possible, thanks to these existing constitutional provisions and existing federal and state statutory provisions.

The U.S. Constitution provides:

“The Congress may determine the Time of chusing the Electors, and the Day on which they shall give their Votes; which Day shall be the same throughout the United States.”[391] [Spelling as per original] [Emphasis added]

Congress has exercised this constitutional power to set the uniform nationwide date for the meeting of the Electoral College by enacting the following statute:

“The electors of President and Vice President of each State shall meet and give their votes on the first Monday after the second Wednesday in December next following their appointment at such place in each State as the legislature of such State shall direct.”[392] [Emphasis added]

Under both the current system and the National Popular Vote approach, all counting, recounting, and judicial proceedings must be conducted so as to reach a “final determination” prior to the uniform nationwide date for the meeting of the Electoral College in mid-December.

The U.S. Supreme Court has made it clear that the states are expected to make their “final determination” six days before the Electoral College meets (the so-called “safe harbor” date established by section 5 of Title 3 of the United States Code).[393]

In addition, in most states, state statutes already impose independent earlier deadlines for finalizing the count for a presidential election. The U.S. Supreme Court has ruled that state election officials and the state judiciary must conduct counts and recounts in presidential elections within the confines of existing state election laws.

Note that the laws governing the finalization of the count (and completion of any recount) for a presidential election are entirely different from those governing, say, a disputed race for one of the 100 seats in the U.S. Senate (e.g., the 2008 Senate race in Minnesota).

It may be argued that the schedule established by the U.S. Constitution, existing federal statutes, and existing state statutes may sometimes rush the count, prevent recounts, and possibly even create injustice. However, there can be no argument that this schedule exists in the U.S. Constitution, federal statutes, and state statutes or that this existing schedule guarantees “finality” prior to the meeting of the Electoral College in mid-December. The existing constitutional and statutory schedule would govern the National Popular Vote compact in exactly the same way that it governs elections under the current system.

Footnotes

A recount is not a logistical impossibility or an unimaginable horror.

The task of recounting the votes cast for President in the nation’s 186,000 precincts is not a logistical impossibility, as evidenced by the fact that the original count is not a logistical impossibility.

A recount is a recognized ever-present contingency whenever a statewide election is conducted. There are about 400 statewide elective offices and statewide propositions on the ballot in a typical November general election in an even-numbered year. There is a probability of about 1-in-185 of a statewide recount (as discussed in section 9.15.1). As a matter of prudent planning, state election officials stand ready with contingency plans to carry out their duty to conduct a recount if one is required.

No state needs the assistance of any personnel or resources from any other state in order to conduct its recount. The personnel and resources necessary to conduct a recount are indigenous to each state. Thus, a state’s ability to handle the logistics of a recount within its own borders is unrelated to whether a recount is being conducted in any other state or all other states.

Under both the current system and the National Popular Vote approach, all counting, recounting, and judicial proceedings must be conducted so as to reach a “final determination” prior to the uniform nationwide date for the meeting of the Electoral College in mid-December.

The U.S. Supreme Court has made it clear that the states are expected to make their “final determination” six days before the Electoral College meets (the so-called “safe harbor” date established by section 5 of Title 3 of the United States Code).

Because all states must finalize their count (or finish their recount) by the “safe harbor” date in early December, and because the only remaining step required by the National Popular Vote bill is to add up the vote totals from all 50 states and the District of Columbia, the final national vote totals would be available before the Electoral College meets.

Even with a single pool of almost 130,000,000 votes, it is possible that the nationwide popular vote could be extremely close in some future presidential election (say, a few hundred votes or perhaps a few thousand votes). In that event, the initial vote count and the recount would be handled in the same way as they are currently handled—that is, under generally serviceable laws that govern all elections.

Any extremely close election will almost certainly engender controversy, and the eventual loser will often go away unhappy.

The guiding principle in such circumstances should be that all votes should be counted fairly and expeditiously.

Of course, if the popular vote count were extremely close on a nationwide basis, it would be very likely that the vote count would also be close in a number of states.

As U.S. Senator David Durenberger (R–Minnesota) said in the Senate in 1979:

“There is no reason to doubt the ability of the States and localities to manage a recount, and nothing to suggest that a candidate would frivolously incur the expense of requesting one. And even if this were not the case, the potential danger in selecting a President rejected by a majority of the voters far outweighs the potential inconvenience in administering a recount.”[394]

Footnotes

No state government has any obligation or power to judge the presidential election returns of any other state under either the current system or the National Popular Vote compact.

Existing federal law specifies that each state’s own “final determination” of its presidential election returns is “conclusive” (if done in a timely manner and in accordance with laws in existence prior to Election Day). The existing federal law was originally enacted in substantially the same form that it exists today shortly after the disputed Tilden–Hayes election of 1876.

In particular, the “safe harbor” provision of existing federal law specifies the conditions under which a state’s “final determination” is considered “conclusive.”

“If any State shall have provided, by laws enacted prior to the day fixed for the appointment of the electors, for its final determination of any controversy or contest concerning the appointment of all or any of the electors of such State, by judicial or other methods or procedures, and such determination shall have been made at least six days before the time fixed for the meeting of the electors, such determination made pursuant to such law so existing on said day, and made at least six days prior to said time of meeting of the electors, shall be conclusive, and shall govern in the counting of the electoral votes as provided in the Constitution, and as hereinafter regulated, so far as the ascertainment of the electors appointed by such State is concerned.”[395] [Emphasis added]

The nation’s long-standing policy of deferring to the states is echoed in the National Popular Vote compact. In particular, the compact’s wording is directly patterned after existing federal law. The compact requires each state to treat as “conclusive” each other state’s “final determination” of its presidential vote. The fifth clause of Article III of the National Popular Vote compact provides:

“The chief election official of each member state shall treat as conclusive an official statement containing the number of popular votes in a state for each presidential slate made by the day established by federal law for making a state’s final determination conclusive as to the counting of electoral votes by Congress.”

Thus, no state government has any obligation or power to judge the presidential election returns of any other state under either the current system or the National Popular Vote compact.

Footnotes

Tara Ross criticizes the National Popular Vote plan by saying:

“States have different criteria for what does (or does not) trigger recounts within their borders. These differences could cause a whole host of problems. What if the national total is close—close enough to warrant a recount—but a recount can’t be conducted because the margins in individual states were not close?”[396] [Emphasis added]

Of course, the ability to obtain a recount in situations “close enough to warrant a recount” is hardly ensured under the current state-by-state winner-take-all system, as demonstrated by the nation’s experience with Florida in 2000.

Moreover, there is no ability under the current system to obtain a recount in situations “close enough to warrant a recount” in the states that do not have recount laws (e.g., Mississippi). If, for the sake of argument, the 537-popular-vote margin that determined the 2000 presidential election had occurred in Mississippi (instead of Florida), there would have been no possibility of a recount. The initial count in Mississippi would have been the first, only, and final count.

In criticizing the National Popular Vote plan, Ross creates an additional misimpression by mentioning only automatic recounts (that is, recounts triggered merely because the original difference between the candidates is less than some fixed legislatively specified threshold). However, automatic recounts are not the only way to obtain a recount (or even the most usual way). In fact, only 19 states have such automatic recount laws.[397] In most states, there are numerous avenues available for obtaining a recount (with some state statutes providing as many as six ways). For example, 42 states allow candidates to petition for a recount. One of the most common forms of state recount laws is to permit a recount if the disgruntled candidate pays for all of the recount’s administrative costs in advance (with the candidate typically being reimbursed if he or she is vindicated).

Obtaining a recount under a national popular vote would not be as difficult as Ross suggests (even under the questionable assumption that no changes would be made to existing laws in response to enactment of the National Popular Vote compact).

As will be seen below:

• Under the current system, a timely recount in a presidential race may be warranted, but impossible to obtain in practice.
• A recount would be less likely to be needed under the National Popular Vote plan than under the current state-by-state winner-take-all system.
• Enactment of the National Popular Vote compact would provide impetus for the states to review and modify their existing laws to ensure timely recounts in presidential elections.
• Federal legislation would be an expeditious solution to the problem of guaranteeing a timely recount in a presidential election under the current system and any future system.

The reality today is that a timely recount in a presidential race under the current system may be warranted, but impossible to obtain in practice, in many states.

Presidential elections must be conducted within the constraints of the strict overall national schedule for finalizing the results of a presidential election established by the U.S. Constitution.

In particular, all counting, recounting, and judicial proceedings concerning presidential elections must be conducted so as to reach a “final determination” six days prior to the uniform nationwide date for the meeting of the Electoral College in mid-December. This principle applies equally to both the current state-by-state winner-take-all system of electing the President and the National Popular Vote compact.

The U.S. Constitution (Article II, section 1, clause 4) provides:

“The Congress may determine the Time of chusing the Electors, and the Day on which they shall give their Votes; which Day shall be the same throughout the United States.” [Spelling as per original] [Emphasis added]

Congress has exercised its constitutional power to set the uniform nationwide date for the meeting of the Electoral College. Title 3, chapter 1, section 7 of the United States Code states:

“The electors of President and Vice President of each State shall meet and give their votes on the first Monday after the second Wednesday in December next following their appointment at such place in each State as the legislature of such State shall direct.” [Emphasis added]

This statute was enacted in 1934, just after ratification of the 20^th Amendment in 1933. Prior to the 20^th Amendment, the President was inaugurated on March 4. The amendment advanced the inauguration to January 20.

Moreover, the U.S. Supreme Court has made it clear that the states are expected to make their “final determination” six days before the Electoral College meets (the so-called “safe harbor” date established by section 5, chapter 1 of Title 3 of the United States Code).

The important point is that there is not much time between Election Day in early November and the “safe harbor” date. For example, in 2008, Election Day was Tuesday, November 4, and the “safe harbor” date was 33 days later on Monday, December 8. In 2008, the Electoral College met on the following Monday, December 15 (the Monday after the second Wednesday in December). Congress met to count votes on January 6, 2009. The outgoing President’s term ended on January 20, 2009.

The schedule imposed by the “safe harbor” date was one of the major factors preventing a hand recount of presidential ballots in Florida in 2000.

In 1960, Hawaii conducted a recount under judicial supervision (which reversed Nixon’s original lead). However, the recount and judicial proceedings were not completed until after the Electoral College had met. As it happened, Hawaii’s three electoral votes did not affect the outcome of the presidential election in 1960. Congress met in joint session in January 1961 to count the electoral votes. The losing presidential candidate, Vice President Richard M. Nixon, presided over the joint session and graciously permitted Hawaii’s electoral votes to be counted for John F. Kennedy (while ruling that this action would not constitute a precedent). The reality, however, was that the recount was not timely.

In summary, no full ballot-by-ballot recount has ever been completed in a timely fashion for any U.S. presidential election.

The actual work of a recount does not itself consume a lot of time.

One of the major obstacles to obtaining a timely recount in a presidential race is that there cannot be a recount until there is a count.

Unofficial numbers are, of course, generally available from virtually every precinct and various units of local government on Election Night (or very soon thereafter). Although a candidate may know whether a recount is warranted in a particular election, a candidate’s request for a recount under existing laws is generally not legally “ripe” until after the official initial statewide count is complete.

The official initial statewide count typically takes several weeks. The reason for the delay is that official documents certifying the count from each county (or other unit of local government) must be sent to the state’s chief election official or state canvassing board. Then, the vote counts contained in these documents are added together to yield the official initial statewide count. The results are often not official until some board meets.

In Ohio in 2004, there were over 200,000 provisional ballots, and Bush’s final margin in the state was 118,601.[398] Senator John Kerry decided that the number of provisional ballots in relation to Bush’s statewide margin on Election Night (before counting the provisional ballots) did not warrant disputing the results and therefore conceded on the Wednesday after the election. However, had the margins been closer, a recount might have been warranted. A request for a recount is not ripe until the initial count is completed. However, the initial count was not certified in Ohio until the “safe harbor” day! Thus, a recount would have been impossible in the decisive state of the 2004 presidential election had a recount been warranted.

In 2012, eight of the 10 closely divided battleground states did not complete their initial count until November 29; five of the 10 did not complete their initial count until December 5; and two did not complete their initial count until just before the “safe harbor” date.[399]

In many other states, the initial count of a presidential election is generally not completed until a week or so before the “safe harbor” date (six days before the Electoral College meets).

The facts are that, under the current system, the possibility of conducting a timely recount of a presidential election is largely an illusion.

In most states, there is no deadline for completing the initial official count in sufficient time to permit the conducting of a recount (and likely post-recount litigation) that is consistent with the federal “safe harbor” date and uniform nationwide date for the meeting of the Electoral College.

Moreover, many state chief election officers can effectively preclude a recount merely by “slow-walking” the initial count so that it is not completed until just before the “safe harbor” date—thus preventing a candidate’s request for a recount from being legally “ripe” until it is too late to conduct the recount.

With the rising volume of absentee and mail-in voting in many states, provisional ballots, and military ballots, thousands of ballots are not counted until after Election Day, thereby further delaying completion of the initial count.

Thus, at the present time, a timely recount is impossible to obtain, in practice, under the current system in many states.

A recount would be less likely to be needed under the National Popular Vote plan than under the current state-by-state winner-take-all system.

There have been five litigated state counts in our nation’s 57 presidential elections between 1789 and 2012. All five of these disputed state counts were the result of the state-by-state winner-take-all rule in elections where there was a clear nationwide popular vote winner. Far from serving as a firewall that helpfully isolates problems to particular states, the current state-by-state winner-take-all system repeatedly creates artificial crises in situations where no problem existed in the first place.

Five litigated state counts in a mere 57 presidential elections between 1789 and 2012 is a dramatically higher rate than the historical 1-in-185 rate for recounts in statewide elections in which there is a single pool of votes and in which the winner is the candidate who receives the most popular votes.

Given that a recount has occurred only about once in 185 elections and that presidential elections occur once every four years, one could expect a recount about once in 740 years under the National Popular Vote plan.

The reason why there have been so many disputes in a mere 57 presidential elections between 1789 and 2012 is that each presidential election is really 51 separate state-level elections—with 51 separate opportunities for close results warranting a recount. Our 57 presidential elections between 1789 and 2012 have really been 2,237 separate state-level elections. The current state-by-state winner-take-all system has repeatedly created artificial crises in elections in which the vote was extremely close in certain states—but not close nationwide. This can be seen by reviewing the five litigated state counts in our nation’s 57 presidential elections between 1789 and 2012.

The 2000 presidential election was an artificial crisis created because George W. Bush’s total of 2,912,790 popular votes in Florida was a mere 537 more than Gore’s 2,912,353 votes. Under the statewide winner-take-all rule used in Florida, the 537-vote lead entitled Bush to all 25 of Florida’s electoral votes. There was, however, nothing particularly close about the 2000 presidential election on a nationwide basis. Al Gore’s nationwide lead was 537,179 popular votes (1,000 times larger) than the 537-vote margin that decided the 2000 presidential election. Given the miniscule number of votes that are changed by a typical recount (averaging only 294 votes), no one would even have considered a recount in 2000 if the nationwide popular vote had controlled the outcome. No one would have cared whether Bush did, or did not, carry Florida by 537 popular votes. There would not have been a dispute in an election where one candidate had a nationwide margin of over a half million votes.

A recount, court case, and reversal of the original outcome occurred in Hawaii in 1960. John F. Kennedy ended up with a 115-vote margin over Richard Nixon in Hawaii in an election in which his nationwide margin was 118,574. There would have been no recount in Hawaii in 1960 if the President had been elected by a nationwide popular vote.

In 1876, Democrat Samuel J. Tilden received 4,288,191 popular votes—254,694 more than the 4,033,497 popular votes received by Republican Rutherford B. Hayes. Tilden’s percentage lead of 3.05% was greater than George W. Bush’s 2004 lead of 2.8%—a margin generally regarded as “solid.” The 1876 election is remembered as having been close because Hayes had extremely narrow popular-vote margins in several states, namely:

• 889 votes in South Carolina,
• 922 votes in Florida,
• 4,807 in Louisiana,
• 1,075 votes in Nevada,
• 1,050 votes in Oregon,[400] and
• 2,798 votes in California.[401],[402],[403],

The closeness of the 1876 presidential election in the Electoral College was an artificial crisis created by the state-by-state winner-take-all system. The candidate who received more popular votes in these closely divided states would have been a mere footnote if the President had been elected by a nationwide popular vote (where Tilden had a 254,694-vote margin).

No presidential election since the 19th century has been won with a nationwide margin of fewer than 118,574 votes (Kennedy’s margin in 1960).

The closest presidential election since the 19th century (when the number of popular votes cast nationwide never exceeded 14,000,000) was the 1960 election in which John F. Kennedy led Richard M. Nixon by 118,574 popular votes (out of 68,838,219 votes cast nationwide). A margin of 118,574 popular votes is not particularly close on a nationwide basis. A six-digit nationwide margin would be unlikely to be challenged and even less likely to be reversed in a recount.

The 1960 presidential election is remembered as being close because a switch by 4,430 voters in Illinois and a switch by 4,782 voters in South Carolina would have given Nixon a majority of the electoral votes. If Nixon had just barely carried both Illinois and South Carolina, Kennedy still would have been ahead nationwide by almost 110,000 popular votes, but Nixon would have won the Presidency. In any case, the perceived closeness of the 1960 election was an illusion manufactured by the winner-take-all system used in Illinois and South Carolina—not because the nationwide margin of 118,574 was ever likely to be overturned by any recount. Indeed, the average change in the margin of victory as a result of a statewide recount was a mere 294 votes in a 13-year study of 4,072 statewide elections (section 9.15.1).

Table 9.19 shows the popular vote count for the Democratic and Republican presidential candidates in each presidential election between 1900 and 2008. In this two-party table, Theodore Roosevelt’s vote is shown in the Republican column for the 1912 election because he polled more votes as the nominee of the Progressive (Bull Moose) Party than did the Republican nominee, William Howard Taft. Column 5 shows the difference between the first- and second-place candidates. None of these elections was particularly close in terms of the nationwide popular vote. The closest election during this period was in 1960 when the popular-vote difference was 118,574.[404] This is true even though the number of votes cast nationwide in the first few elections of the 20^th century was only about 10% of present-day turnouts.

Table 9.19 Winning margins in presidential elections 1900–2008

In short, recounts would be far less likely under the National Popular Vote approach than under the current state-by-state winner-take-all system.

Enactment of the National Popular Vote compact would provide impetus for the states to review and modify their laws regarding timely recounts in presidential elections.

The observation that existing state recount laws are not based on national popular vote totals is something of a straw man in that it suggests that existing state recount laws are permanent and unchangeable.

When the U.S. House of Representatives passed a constitutional amendment in 1969 to establish a national popular vote for President, by a 338–70 margin, there were no detailed procedures for recounts in the amendment. The House did not pass accompanying recount legislation at the time it passed the amendment. Of course, it was generally understood that implementing legislation would have been enacted if the amendment had been ratified. The ratification of the amendment would have provided the impetus to update existing laws.

Similarly, the enactment of the National Popular Vote compact would provide impetus for the states to review their laws regarding timely recounts in presidential elections.

As Tara Ross says:

“To be fair, if NPV were implemented, then many state legislatures would probably work to make their recount statutes more lenient. Even if these states otherwise disagree with NPV, they would not want to be caught in a situation where they could not participate in a national recount. Moreover, as alluded to previously, many states already provide ‘optional recount’ statutes that allow recounts to be requested by candidates or voters even without a close margin.”[405]

Given that Congress has authority over the count and schedule for presidential elections, federal legislation is another way to solve the problem of guaranteeing a timely recount in a presidential election under the current system or the National Popular Vote plan.

Another way to remedy the existing practical difficulties of obtaining a timely recount in a presidential election would be for Congress to use its existing authority over the count and schedule for presidential elections to augment state elections with national recount legislation. This approach is discussed in section 9.15.7.

Footnotes

Although Congress does not control the manner of awarding a state’s electoral votes, Congress has constitutional authority over both the schedule of presidential elections and the counting of votes in presidential elections under Article II and the 12^th Amendment.

One example of the exercise of Congress’ authority over the count in presidential elections is the current federal law requiring each state to produce (prior to the meeting of the Electoral College in mid-December) a “Certificate of Ascertainment” containing the state’s “final determination” of the number of popular votes cast in the state for each individual presidential candidate. This procedure was first adopted after the disputed 1876 Tilden-Hays election and is now contained in section 6, chapter 1 of Title 3 of the United States Code.

Each state’s Certificate of Ascertainment provides the supporting evidence for the state’s “Certificate of Election” containing the names of the presidential electors who will cast the state’s votes in the Electoral College. In the case of a state using the statewide winner-take-all rule, the supporting evidence consists of the canvass of the statewide popular vote for President.[406] In the case of a state (such as Maine and Nebraska) using the congressional-district approach, the supporting evidence consists of the canvass of the district-wide popular vote for President.[407]

A joint session of Congress in early January reviews each state’s Certificate of Ascertainment and Certificate of Election as part of the constitutional process of counting the votes for the presidential election.

One way to solve the problem of guaranteeing a timely recount in a presidential election would be for Congress to pass a law guaranteeing presidential candidates the right to a timely recount.

A federal law would be beneficial to the operation of the current state-by-state winner-take-all method for awarding electoral votes because of the high frequency of disputes in presidential elections under the current system (five litigated state counts in the mere 57 presidential elections).

The high frequency of recounts under the current system results from the fact that each presidential election is really 51 separate state-level elections and that the nation’s 57 presidential elections have really been 2,237 separate state-level elections. Although the probability of a recount in any single statewide election is low (1-in-185 according to a study of the 4,072 statewide general elections in the 13-year period between 2000 and 2012[408]) and although recounts appear to be becoming rarer (there being no recounts among the 419 statewide elections in November 2012), the fact that each presidential election under the current state-by-state winner-take-all system is really 51 separate state-level elections means that there is a significant chance of future disputed presidential elections under the current system.

A federal law would also be potentially beneficial under the national popular vote approach, even though the probability of recounts would be much lower under a national popular vote because there would a single large national pool of votes (instead of 51 separate pools). One would expect an election close enough to warrant a recount under the National Popular Vote approach about once in 740 years (185 times four). In fact, the probability of a national recount would be even less than 1-in-185 because that rate is based on the history of statewide recounts, and recounts become less likely with larger pools of votes.

Time is of the essence in conducting a recount in a presidential election. The U.S. Constitution establishes a strict overall national schedule for finalizing the results of a presidential election. In particular, the Constitution requires the Electoral College to meet on the same day throughout the United States (currently the first Monday after the second Wednesday in December).

Because of this firm deadline, all counting, recounting, and judicial proceedings (state or federal) must be conducted so as to reach a “final determination” prior to the uniform nationwide date for the meeting of the Electoral College. The U.S. Supreme Court has made it clear that the states are expected to make their “final determination” six days before the Electoral College meets (the so-called “safe harbor” date established by section 5 of Title 3 of the United States Code).

A key consideration in constructing a practical schedule for recounts in presidential elections is the fact that there cannot be a recount until there is a count. That is, a recount cannot be conducted until the official initial count is completed.

Given the actual practices of many states (including many of the closely divided battleground states), there would be no time to conduct a recount under the current system of electing the President.

In Ohio in 2004, there were more than 150,000 provisional ballots. Bush’s final margin in the state was 118,601.[409] Senator John Kerry decided that the number of provisional ballots in relation to Bush’s apparent statewide margin on Election Night (before counting the provisional ballots) did not warrant disputing the results, and he therefore conceded on the Wednesday after the election. However, had the margins been closer, a recount might have been warranted. A request for a recount is not ripe until the initial count is completed. In 2004, the initial count in Ohio was not completed and certified until the “safe harbor” day (thereby precluding a recount).

Professor Danial Tokaji at the Michael E. Moritz College of Law at Ohio State University identifies the difficulties associated with a potential contest or recount in Ohio:

“There is no specific Ohio statute addressing a contest in a presidential election. Presumably, the generally applicable election contest procedure described above would apply. The Ohio statutory scheme, however, makes no reference to the federal statutes governing presidential election contests. This could prove problematic. Under the “safe harbor” provision of 3 U.S.C. § 5, Ohio must reach a final determination of election controversies within 35 days of the presidential election. A quick review of Ohio’s election contest procedure illustrates the problem. A contestor must file the petition within 15 days of the election results being certified (assuming no automatic or requested recount). R.C. 3515.09. Presumably, a contest concerning presidential electors involves a “statewide office” requiring the petition to be filed with the Chief Justice. See R.C. 3515.08. The court must then set the hearing within the 15-to-30-day window of R.C. 3515.10. Even without considering the time delay from election day to certification of results, meeting the 35-day safe harbor provision is doubtful. Add to this mix the uncertainty of the 40-day deposition period of R.C. 3515.16 if the contest is “in the supreme court.” Further consider the effect of an appeal—if possible—and the 20-day appellate filing window. Following the Ohio statutory scheme makes compliance with 3 U.S.C. § 5 unlikely.”[410] [Emphasis added]

In 2012, eight of the 10 closely divided battleground states did not complete their initial count until November 29; five of the 10 did not complete their initial count until December 5; and two did not complete their initial count until a day before the “safe harbor” date (which was December 11).[411]

In many other states, the initial count of a presidential election is generally not completed until a week or so before the “safe harbor” date (six days before the Electoral College meets).

The facts are that, under the current system, the possibility of conducting a timely recount of a presidential election is largely an illusion.

The precondition for conducting a full ballot-by-ballot recount of a presidential election is rapid completion of the initial count. Because there are only a few weeks between Election Day in early November and the meeting of the Electoral College, the initial count must be completed quickly enough to provide time for a full recount plus some additional time for post-recount litigation. Thus, an essential element of federal legislation giving presidential candidates the right to a recount must be acceleration of the initial count. Without acceleration of the initial count, there cannot be a timely recount.

Acceleration of the initial count costs money because of the overtime and additional staffing involved. There is, of course, no reason to spend the money necessary to accelerate the initial count unless there is good reason to believe that the presidential result (usually apparent to the candidates themselves on Election Night) is likely to be changed by a recount.

Taking all of the above considerations into account, we believe that an effective federal recount law for presidential elections should have the following features.

First, a federal recount law should move the uniform national day for the meeting of the Electoral College (currently established by federal law as the Monday after the second Wednesday in December) to December 30 (or to the previous Friday, if December 30 falls on a weekend). If there is no controversy over the outcome of a presidential election, the meeting of the Electoral College is ceremonial and irrelevant to implementation of the presidential transition. If there is a controversy, as much time as possible should be available to arrive at the most accurate possible determination of the presidential vote.

Second, a federal recount law should require that each state’s chief election official prepare and publish a recount plan 90 days before Election Day. This chief election official’s plan would provide for:

1. conducting an accelerated initial count of the presidential vote, if requested, that would be completed by November 30, and
2. conducting a recount of the presidential vote, if requested, that would be completed by December 14 involving a one-by-one examination of each ballot (to the extent possible given the state’s voting equipment and procedures).

In an undisputed presidential election, the only obligation imposed by the proposed federal count law on the state’s chief election official would be the preparation of this plan.

Most states would incur substantial incremental costs (notably in the form of overtime and additional staffing) in accelerating their initial count so that it could be completed quickly enough to provide time for a full recount as well as time for post-recount litigation.

States would also incur substantial costs if a full recount had to be conducted.

Thus, the chief election official’s plan would include a specification of all reasonable incremental costs for accelerating the initial count and all reasonable costs for conducting a recount.

Note that the federal right created by the proposed federal legislation is not an unfunded mandate on the states, because the requesting candidate would be required to pay, in advance, for all reasonable costs as computed by the state’s chief election official.

The chief election official’s plan would include standards for determining voter intent for all cases that may be reasonably anticipated given the state’s voting equipment and procedures. Most state-specific problems associated with counting votes are well known to state election officials as a result of their years of experience in conducting elections. However, these standards are, in many states, not clearly delineated. Instead, the standards are a mixture of various state statutes, case law, administrative procedures at the state and local level, and unwritten practices. A clear delineation of the rules for determining voter intent in the form of administrative standards would increase the efficiency of the initial count and recount and effectively reduce the number of issues that could be successfully raised in post-recount litigation.

Third, a federal recount law should give each presidential candidate on the ballot in a state the right to call for acceleration of the initial count of the presidential vote in the state, provided that the requesting candidate pays, in advance, for all reasonable incremental costs of that request.

In addition, a federal recount law should give each presidential candidate on the ballot in a state the right to call for a recount of the presidential vote after completion of the initial count, provided that the requesting candidate pays, in advance, for all reasonable costs of that request.

As a practical matter, a presidential candidate who has a realistic chance of overturning an apparent loss of the White House would have no difficulty quickly raising the money to pay for the requested actions.

These rights should be extended to the candidate’s successor if the candidate dies or resigns.

These rights should be given to the presidential candidates themselves (as opposed to the individual candidates for the position of presidential elector, political parties, or private citizens) because the candidates are in the best position to make a realistic and pragmatic political judgment, based on available information, as to whether the election involved is close enough to warrant a potential recount.

The fact that the candidate would have to pay the costs of a requested acceleration of the initial count and the costs of a requested recount would act as a disincentive against unrealistic requests.

Note that it is not desirable or possible to impose any preconditions on requests by the presidential candidates (e.g., closeness of the results). Such preconditions would necessarily have to be couched in terms of official election results which (if the initial count is not yet complete) would not be available at the moment when the candidate’s decision is needed.

The request for the acceleration of the initial count would have to be lodged quickly, say within six days after Election Day. In practice, it is usually clear by Election Night whether a particular election is close enough to warrant a dispute.

Television networks regularly make decisions to “call” an election on Election Night. Candidates for President (as well as candidates for Governor, U.S. Senate, and all other offices) regularly concede on Election Night. Both candidates and television networks make such decisions despite the fact that there are large numbers of in-person votes cast earlier in the day that are yet to be counted; large numbers of uncounted mail-in, military, and absentee ballots (which, in some states, need not even arrive at vote-counting centers until several days after the election); and large numbers of uncounted provisional ballots (for which voters have 6–10 days to provide evidence in support of their right to vote).

In practice, a presidential candidate’s decision to request an acceleration of the initial count would be made on the basis of the same mixture of political intelligence that candidates use in making their decision to publicly concede an election, namely available actual returns announced by election officials; exit polls; estimates of the number of uncounted absentee ballots, uncounted mail-in ballots, uncounted provisional ballots, and uncounted military ballots; and historical information and current polling indicating the likely breakdown of the absentee, mail-in, provisional, and military ballots.

Candidates make their decision to concede because the information at their disposal makes it clear that they have no realistic possibility of winning.[412]

Note that there is no practical way to refund the cost of accelerating the initial count to a “successful” candidate because no official count exists at the time that the candidate would make the request (and hence no benchmark for “success”).

Fourth, a federal recount law should make it clear that it is an option in addition to any procedure available under state law, state administrative procedures, or state case law. Thus, if the candidate fails to act by the deadlines contained in the federal recount law, the candidates would still be able to pursue whatever remedies may be available under existing state law.

Fifth, a federal recount law should clarify that the “safe harbor” date (defined in the existing section 5 of Title 3 of the United States Code) is the deadline for each state to complete its “final determination” of the presidential count in the state.

Sixth, moving the meeting of the Electoral College to December 30 (thus making December 24 the “safe harbor” day) enables a day such as December 14 to be a reasonable deadline for completing both the accelerated initial count and the recount. Such a deadline would leave 10 days for post-recount litigation. Note that five days would remain available for potential litigation after the “safe harbor” date (as is currently the case).

Seventh, a federal recount law should give each presidential candidate on the ballot (or his successor) the right to sue to enforce all the provisions of Title 3 of the U.S. Code concerning presidential elections. To avoid forum-shopping within a state, the action should be required to be brought in the federal district court located in the state capital (or the district court located nearest to the state capital if no federal district court is located in the capital).

Note that an additional advantage of a federal right to a recount in presidential elections is that it would preclude state officials from effectively precluding a recount by “slow walking” the state’s initial count. Slow-walking of the initial count effectively enables state officials (many elected on a partisan basis) to decide whether their own work product is subject to accountability.

Table 9.20 shows the schedule for 2016 under the proposed legislation.

Table 9.20 Schedule for 2016 under proposed federal recount legislation

The proposed federal recount bill (shown below) has the following elements:

• Section 1 of the proposed bill is the bill’s title.
• Section 2 of the proposed bill adds a new subsection (b) to section 5 of Title 3 of the United States Code containing five new parts.
  • Part (1) of the new subsection (b) of section 5 of Title 3 makes the “safe harbor” day (defined in the existing portion of section 5—now called subsection (a)) into an actual deadline for each state to complete its “final determination” of the presidential count in the state.
  • Part (2) of the new subsection (b) of section 5 of Title 3 requires the state’s chief election official to prepare and publish a plan (1) for conducting an accelerated initial count of the presidential vote that would be completed by November 30 and (2) for conducting a full recount of the presidential vote involving a one-by-one examination (to the extent possible, given the state’s voting equipment and procedures) of each ballot that would be completed by December 14—along with all reasonable incremental costs of conducting an accelerated initial count and all reasonable costs of conducting the recount. Such plan shall include standards for determining voter intent for all cases that may be reasonably anticipated given the state’s voting equipment and procedures.
  • Part (3) of the new subsection (b) of section 5 of Title 3 gives a presidential candidate on the ballot in the state the opportunity to call for an acceleration of the initial count in a state to be completed by November 30, provided the requesting candidate pays for all reasonable incremental costs of accelerating the initial count. The right to an accelerated initial count created by this subsection must be exercised within six days after Election Day. This right is extended to the presidential candidate’s successor if the candidate dies or resigns.
  • Part (4) of the new subsection (b) of section 5 of Title 3 gives a presidential candidate the opportunity to request a recount to be completed by December 14, provided the requesting candidate pays for all reasonable costs of the recount. The right to a recount created by this subsection must be exercised by December 2. This right is extended to the presidential candidate’s successor if the candidate dies or resigns.
  • Part (5) of the new subsection (b) of section 5 of Title 3 gives a presidential candidate on the ballot (or his successor) the right to sue to enforce all the requirements of Title 3. To avoid forum-shopping within the state, the action must be brought in the District Court located in the state capital (or the United States District Court located nearest to the state capital if no United States District Court is located in the capital).
  • Part (6) of the new subsection (b) of section 5 of Title 3 explicitly states that the recount made available under part (3) of the new subsection (b) of section 5 of Title 3 shall be an option available to presidential candidates in addition to any procedure available under state law, administrative procedures, or judicial determinations.
  • Section 3 of the proposed bill amends section 6 of Title 3 of the United States Code by requiring that the Certificates of Ascertainment be physically delivered to Washington, D.C., no later than the day after the “safe harbor” day. The current section 6 has an outdated time-consuming process involving registered mail and sets no particular deadline.
  • Section 4 of the proposed bill amends section 6 of Title 3 of the United States Code by moving the uniform national day for the meeting of the Electoral College (currently the Monday after the second Wednesday in December) to December 30 (or the previous Friday if December 30 falls on a weekend).
  • Section 5 of the proposed bill makes the bill effective for the 2016 elections.

Text of the Proposed Federal Recount Bill

To amend title 3, United States Code, to require a State to make available to a presidential candidate a timely initial count and a timely recount of the number of votes cast in the presidential election in the State, to change the date for a State to complete its final canvas or ascertainment of the number of votes cast for each candidate in a presidential election, to change the date of the meeting of presidential electors, and for other purposes.

Be it enacted by the Senate and House of Representatives of the United States of America in Congress assembled,

SECTION 1. SHORT TITLE.

This Act may be cited as the ‘‘Presidential Election Recount Act of ___’’.

SEC. 2. DEADLINE FOR FINAL CANVASS AND ASCERTAINMENT OF NUMBER OF VOTES CAST FOR PRESIDENT AND AVAILABILITY OF RECOUNT.

Section 5 of title 3, United States Code, is amended—

(1) by striking ‘‘If any State’’ and inserting ‘‘(a) IN GENERAL.—If any State’’ and by striking ‘‘concerning the appointment’’ and inserting ‘‘concerning the canvass or appointment’’.

(2) by adding at the end the following new subsection:

‘‘(b) DEADLINE FOR FINAL DETERMINATION OF CANVASS OR ASCERTAINMENT OF VOTES CAST FOR PRESIDENT AND AVAILABILITY OF ACCELERATED INITIAL COUNT AND AVAILABILITY OF RECOUNT—

‘‘(1) DEADLINE FOR FINAL DETERMINATION.—The canvass or ascertainment under the laws of each state of the number of votes given or cast for each candidate for President or presidential elector and the final determination of any controversy or contest concerning such canvass of ascertainment shall be made not later than 6 days before the time fixed for the meeting of the electors under section 7 of this title.

‘‘(2) PREPARATION OF PLAN FOR RECOUNT AND COSTS.—No later than 90 days before the time fixed for appointing electors under section 1 of this title, the state official or body that is authorized to conduct the canvass or ascertainment under the laws of each state of the number of votes given or cast for each candidate for President or presidential elector shall prepare, and make available to the public, a plan for accelerating the initial count of each ballot given or cast for each candidate for President or presidential elector in that state, with such accelerated initial count to be completed by November 30, and a plan for conducting a full recount involving a one-by-one examination (to the extent possible, given the state’s voting equipment and procedures) of each ballot given or cast for each candidate for President or presidential elector in that state, with such recount to be completed by December 10. Such plan shall include standards (not inconsistent with state law) for determining voter intent for all cases that may be reasonably anticipated, given the state’s voting equipment and procedures. Such plan shall include all reasonable incremental costs to the state associated with accelerating the initial count and all reasonable costs to the state for conducting the recount.

‘‘(3) ACCELERATION OF INITIAL COUNT.—If a candidate for the office of President appearing on the ballot in a given state (or a legal successor nominated in lieu of such candidate) shall, no later than 6 days after the time fixed for appointing electors under section 1 of this title, make a written request, accompanied by payment in full of the costs specified in the plan created under part (2) of this subsection, for accelerating the initial count, the state official or body that is authorized to conduct the canvass or ascertainment under the laws of each state of the number of votes given or cast for each candidate for President or elector shall conduct the initial count of the votes cast for each candidate for President or presidential elector, with such recount to be completed by November 30. If more than one candidate makes a request for accelerating the initial count in the state, the state shall divide the costs eqully among the requesting candidates and refund any excess payments received.

‘‘(4) AVAILABILITY OF RECOUNT.—If a candidate for the office of President appearing on the ballot in a given state (or a legal successor nominated in lieu of such candidate) shall by December 2 make a written request, accompanied by payment in full of the costs specified in the plan created under part (2) of this subsection, for conducting a full recount involving a one-by-one examination (to the extent possible given the State’s voting equipment and procedures) of each ballot given or cast for each candidate for President or elector in a given state, the state official or body that is authorized to conduct the canvass or ascertainment under the laws of each state of the number of votes given or cast for each candidate for President or elector shall then conduct a full recount of the votes cast for each candidate for President or presidential elector, with such recount to be completed by December 14. If more than one candidate makes a request for a recount in the state, the state shall divide the costs equally among the requesting candidates and refund any excess payments received.

‘‘(5) PRIVATE RIGHT OF ACTION.—An individual who is a citizen of the United States who is a resident of the State involved or a candidate for the office of President appearing on the ballot in a given state (or a legal successor nominated in lieu of such candidate) may bring an action against the State in the United States district court located in the capital of the State involved (or the United States district court located nearest to the state capital if no United States district court is located in the capital) for such declaratory or injunctive relief as may be necessary to ensure that the State is in compliance with this title.’’

“(6) NON-PRE-EMPTION.—The recount made available under part (3) of this subsection shall be an option available to presidential candidates (or a legal successor nominated in lieu of such candidate) in addition to any procedure available under applicable state laws, administrative procedures, or judicial decisions.

SEC. 3. REQUIRING PROMPT TRANSMISSION OF CERTIFICATIONS.

Section 6 of title 3, United States Code, is amended—

(1) by striking ‘‘as soon as practicable’’ each place it appears and inserting ‘‘immediately’’;

(2) by striking ‘‘to communicate by registered mail’’ and inserting ‘‘to communicate by overnight courier service’’; and

(3) by striking ‘‘to communicate under the seal of the State’’ and inserting ‘‘to communicate by overnight courier service under the seal of the State”.

SEC. 4. TIME FOR MEETING OF THE ELECTORAL COLLEGE.

Section 7 of title 3, United States Code, is amended by striking ‘‘first Monday after the second Wednesday in December next following their appointment’’ and inserting ‘‘30th day of December next following their appointment or the preceding Friday if December 30 is a Saturday or Sunday.’’

SEC. 5. EFFECTIVE DATE.

The amendments made by this Act shall apply to Presidential elections beginning with the elections held in November 2016.

Footnotes

A caller to a radio debate on the National Popular Vote plan asked:

“With a lot of voting machines without paper trails, there really isn’t a method of doing recounts. So, how would we do an effective recount if we need to?”[413]

Direct-recording electronic (DRE) voting machines are in widespread use in the United States.

Some DREs produce a voter-verifiable paper audit trail (VVPAT) that permits post-election ballot-by-ballot auditing or recounting of the results; however, others do not.

Computer World reported in October 2012:

“A total of 16 states use DREs that do not support a paper trail as their standard polling place equipment, according to Verified Voting. Of these, six states—New Jersey, Delaware, Maryland, South Carolina, Georgia, and Louisiana—will be completely paperless. All ballots that are cast in these states will be on DREs that support no paper trail whatsoever.

“The remaining states, which include Texas, Colorado, Florida, Virginia, and Pennsylvania, will use a mix of paper ballots and DRE voting systems that are paperless. But even here, the states of Virginia, Pennsylvania, and Tennessee will be almost completely reliant on paperless electronic voting systems. In Tennessee, for instance, all but two counties will use paperless DREs, while in Virginia all but seven of 134 countries will use paperless systems. Meanwhile, in a handful of states like Florida, only voters with physical disabilities will use paperless DREs.”[414]

The problem of DRE machines without a VVPAT is not uniquely associated with elections conducted under the current state-by-state winner-take-all system or those conducted under the National Popular Vote plan.

The deficiencies of DRE machines without a VVPAT are amplified by the current state-by-state winner-take-all system under which a small number of popular votes in a single battleground state can change the outcome of a national election. In 2000, for example, George W. Bush won Florida by a margin of 537 popular votes out of 5,963,110 votes cast, and those 537 votes determined the outcome of the national election.

Indeed, DRE machines without a VVPAT were used in battleground states such as Virginia, Colorado, Pennsylvania, and Florida.

In Virginia, 101 of 134 counties and independent cities in Virginia do not have a paper record of the vote. These places contain over four million voters.[415]

In Pennsylvania, 50 of the 67 counties do not keep a voter-verified paper record of voter choices.[416]

Jefferson County, Colorado, is the state’s fourth most populous county, and it uses iVotronic machines without VVPAT. There are 381,164 registered voters in the county.[417]

Florida uses DRE machines with a VVPAT for handicap-accessible voting.[418]

In terms of actual consequences in the real world, the (admittedly undesirable) use of DRE machines without a VVPAT is unlikely to have any effect on the outcome of any presidential election held under the National Popular Vote plan.

In the first place, the probability of a recount affecting the outcome of a particular presidential election in which there is a single pool of votes (i.e., a national popular vote) is low. As shown in table 9.16, there were 22 recounts in 4,072 statewide general elections in the 13-year period between 2000 and 2012—that is, one recount for every 185 elections. In terms of presidential elections, this probability indicates that there would be a recount in a nationwide popular vote for President only once in 740 years.

Recounts change only a small number of votes (an average of only 294 votes in statewide elections), and the probability of a recount reversing an election outcome is only one in seven.[419] Thus, the (admittedly undesirable) use of DRE machines without a VVPAT is, as a practical matter, unlikely to affect the outcome of any presidential election.

In the second half of the 20^th century, about two-thirds of all voting in the United States was done on lever-type voting machines. These machines (like DREs without a VVPAT) recorded the total count for each candidate, but did not keep a record of each individual ballot. Figure 2.13 shows the face of a lever-type voting machine used in 1960 in Alabama. Votes were recorded on mechanical counters on lever-type voting machines. After the polls closed, each voting machine was opened, and the vote count for each office was read from the mechanical counters. During the many decades when lever-type voting machines were in widespread use in the United States, there were no major adverse consequences to the nation because of the absence of the highest desirable degree of post-election verification. In particular, no presidential election was affected by the use of lever-type machines or the absence of a voter-verifiable paper audit trail.

The (unfortunate) inability to conduct a ballot-by-ballot post-election recount of votes cast on DRE machines without a VVPAT does not mean that an election would be thrown into chaos. It would not even mean that a recount could not be conducted. It would simply mean that the quality and thoroughness of the recount on those particular machines would be severely limited to that which lever-type voting machines provided in the second half of the 20^th century (e.g., to catching errors such as incorrectly recording the count from a machine, failing to include the count from a machine, or double-counting a machine).

Election authorities in the states have the ability to replace DRE machines without a VVPAT with machines with a VVPAT. Hopefully, they will exercise their power to do so.

Footnotes