9.13 Myths about Election Administration

Under the National Popular Vote compact, a presidential election would be administered inside each state in the same way that it is now administered. The compact makes no changes in a state’s laws or procedures for preparing ballots, operating polling places, handling absentee ballots or early voting, or counting votes at the precinct, city, town, or county level. Local and county election officials would conduct elections exactly as they do now.

The National Popular Vote compact would make no change in the process of aggregating the vote counts from the local level in order to ascertain the total number of popular votes cast in the state for each presidential slate.

The only change introduced by the National Popular Vote compact occurs after a state has finished tallying the statewide total number of popular votes cast for each presidential slate.

At that point, the votes cast for each presidential slate in all 50 states and the District of Columbia would be added together to produce a national grand total for each presidential slate (section 6.3.3). This vote total would be, of course, the official version of the same adding process that the media, the political parties, and various watchdog groups already do on Election Night and in the days immediately following each presidential election.

Under the compact, the presidential slate with the largest national grand total from all 50 states and the District of Columbia would be designated as the “national popular vote winner.” The chief election official of each state belonging to the compact would then certify the election of the entire slate of presidential electors that is affiliated with the “national popular vote winner.” For example, if the Republican slate is the “national popular vote winner,” the state’s chief election official in every state belonging to the compact would certify the election of the entire slate of Republican presidential electors.

The effect of the National Popular Vote compact would be that all the presidential electors of all states belonging to the compact would be affiliated with the presidential slate that received the largest total number of popular votes in all 50 states and the District of Columbia. These presidential electors from the states belonging to the compact would collectively represent the nationwide will of the voters.

Under the compact, the presidential electors would meet in their states, as they do now, in mid-December and cast their electoral votes.

Because the compact would only go into effect when it has been enacted by states possessing a majority of the electoral votes, the presidential slate receiving the most popular votes from all 50 states and the District of Columbia would receive a majority of the electoral votes in the Electoral College.

The fiscal analysts associated with virtually every state legislature that has considered the National Popular Vote bill have concluded that there would be no significant additional administrative burden or financial cost associated with implementing the compact.

The mechanics for counting and tallying votes at the precinct, city, town, county, and state levels would be the same under the National Popular Vote compact as they are under the current system.

Neither the current system nor the National Popular Vote compact requires—or permits—any state election official to become involved in judging the election returns of other states.

Existing federal law (the “safe harbor” provision in section 5 of Title 3 of the United States Code) specifies that a state’s “final determination” of its presidential election returns is “conclusive” in the counting of votes by Congress (if done in a timely manner and in accordance with laws that existed prior to Election Day).

The wording of the National Popular Vote compact is patterned directly after the existing federal “safe harbor” provision. It would require each state to treat as “conclusive” every other state’s “final determination” of its vote for President. Clause 5 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.”

Accordingly, assuming each state complies with federal law, no state would have any power to examine or judge the presidential election returns of any other state under the National Popular Vote compact.

Under the National Popular Vote compact, the mechanics for counting votes for President at the precinct, city, town, county, and state levels would be the same as they are today.

The only administrative difference would be that, after counting all the votes in the state, each state’s chief election officer would add up the popular vote totals from all 50 states and the District of Columbia to determine which slate of presidential electors would be called upon to cast the state’s electoral votes.

When the National Popular Vote bill has been introduced in state legislatures, state fiscal officials have uniformly concluded that it has no significant fiscal impact on the state. In most states, this determination has been explicitly stated in the financial analysis that is routinely produced by the legislature’s professional staff prior to the time that the legislature considers the bill.

The arguments in favor of conducting audits apply to all elections, regardless of the office being filled. The statistical procedures for conducting audits are applicable to all elections.

Audits are conducted in some states today, thanks to statutory audit procedures and administratively established audit procedures.

Federal legislation has been proposed to require audits in all federal elections—including presidential elections. For example, the proposed Voter Confidence and Increased Accessibility Act of 2009 (H.R. 2894 of the 111^th Congress introduced by New Jersey Congressman Rush Holt and a considerable number of co-sponsors) would require audits for all federal elections, including presidential elections.

One important difference between presidential elections and elections for the U.S. House and U.S. Senate is that the U.S. Constitution establishes a strict overall national schedule for finalizing the results of a presidential election. The existing constitutional provisions (and existing supporting federal statutes) apply equally to elections conducted under both the National Popular Vote plan and the current system.

Specifically, the U.S. Constitution requires that the Electoral College meet on a uniform nationwide day in every state.[343] Congress has specified the Monday after the second Wednesday in December as the date for the meeting of the Electoral College.[344]

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” day established by section 5 of Title 3 of the United States Code).[345]

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

Many of the most important reasons for conducting an audit are lost if insufficient time remains available to conduct a full recount if the audit discovers a problem. Indeed, in the “Principles and Best Practices for Post-Election Audits” endorsed by numerous organizations involved in election-administration issues (including the Brennan Center for Justice, Common Cause, Verified Voting, and numerous state-level groups), one of the best practices is:

“Post-election audits must be completed prior to finalizing official election results and must either verify the outcome or, through a 100% recount, correct the outcome.”

Thus, in the case of presidential elections, a practical and realistic schedule for audits must allow time for a potential full recount (and also time for potential post-recount litigation) prior to the uniform nationwide day for meeting of the Electoral College. Thus, audits in presidential elections must be conducted in an expeditious and timely manner (soon after Election Day) so as to allow time for a potential full recount and potential post-recount litigation.

Fortunately, audits do not take long. Today, audits are routinely conducted within a couple of days by the states that have statutory audit procedures or administratively established audit procedures. There is thus no reason why audits cannot be conducted for presidential elections under either the current system or the National Popular Vote approach.

Proposed legislation such as H.R. 2894 provides for audits of presidential elections. This (generally excellent) proposal could be improved by amending the formula for determining the intensity of auditing that is required in presidential elections so that the level of intensity of the audit is determined by the apparent margin in the nationwide count (as opposed to the apparent statewide count) in case the appointment of presidential electors is based on the national popular vote. Alternatively, the highest level of intensity already provided for in H.R. 2894 for the audit might be automatically applied to presidential counts. Note that this suggested improvement concerning the issue of intensity does not relate to whether an audit will be conducted—but merely to the audit’s level of intensity.

In short, there is nothing in the National Popular Vote plan that would prevent a post-election audit.

Footnotes

The Help America Vote Act of 2002 (HAVA) permits a voter to cast a “provisional ballot” under certain circumstances, including (but not limited to) situations in which:

• the voter does not have the type of identification (if any) that may be required by state law;
• the voter is not listed on the election roll for a particular precinct (perhaps because the voter went to the wrong polling location or because the voter recently moved); and
• the voter arrives at the polling place on Election Day but previously requested an absentee ballot (thus raising the question of whether the voter has already voted).

A provisional ballot is typically inserted into a large envelope whose exterior contains an explanation as to why the ballot was cast on a provisional basis. The outside of the envelope contains the voter’s signature and often contains additional identifying information beyond the voter’s address (e.g., a driver’s license number).

Provisional ballots are usually counted within six to 10 days after the election (depending on state law).

Processing provisional ballots is a tedious administrative process. The specific processing required depends on the reason why the provisional ballot was cast in the first place. For example, if a ballot was cast provisionally because of lack of certain required identification documents, the signature on the outside of the envelope may be compared visually with registration records before the provisional ballot is approved. If a driver’s license number is used as part of the identification process, the number provided by the voter on the outside of the envelope may be compared with the state’s database of driver’s licenses. According to a Miami Herald story:

“Each provisional ballot takes about 30 minutes to review and inspect, said Ron Labasky, counsel for the state association of election supervisors.”[346]

According to the U.S. Election Assistance Commission (a body established by the Help America Vote Act of 2002), about two-thirds of all provisional ballots are found to have been cast by legitimate voters and, therefore, eventually counted.[347]

Hans von Spakovsky has stated that a nationwide election of the President

“would … lead to … contentious fights over provisional ballots.”[348]

Hans von Spakovsky has also stated:

“Every additional vote found anywhere in the country could make the difference to the losing candidate.”[349] [Emphasis added]

We agree with von Spakovsky that any vote “anywhere in the country could make the difference” in a nationwide vote for President. Indeed, the most important reason to adopt the National Popular Vote plan is to make every vote in every state politically relevant in every presidential election. We do not view the fact that every vote “could make the difference” as an evil.

Von Spakovsky continues:

“Provisional ballots may not affect the outcome of the majority vote within a state under the current system because the number of provisional ballots is less than the margin of victory. However, if the total number of provisional ballots issued in all of the states is greater than the margin of victory, a national battle over provisional ballots could ensue.

“Losing candidates would then have the incentive to hire lawyers to monitor (and litigate) the decision process of local election officials.…

“Lawyers contesting the legitimacy of the decisions made by local election officials on provisional ballots nationwide could significantly delay the outcome of a national election.”[350] [Emphasis added]

Our view is that ballots cast by legitimate voters should be counted. We also believe that a candidate who is slightly behind in a close election has every right to “monitor” the handling of provisional ballots and, if necessary, “litigate” the question of whether a particular voter is legally entitled to have his or her vote counted. A losing candidate is certainly entitled to present his or her case to the courts “if the total number of provisional ballots … is greater than the margin of victory” based on the non-provisional ballots.

We do not view the proper counting of all legitimate votes as an evil; however, if anyone entertains this viewpoint, provisional ballots are far more likely to create a problem under the current state-by-state winner-take-all system than under a nationwide vote.

Under the current state-by-state winner-take-all system, the outcome of the national election frequently depends on the outcome of one or more closely divided battleground states. The number of provisional ballots in closely divided states is typically larger than the initial margin of victory based on the non-provisional ballots. Thus, even when there is a clear winner of the national popular vote, the possibility exists, of a dispute involving provisional ballots in a closely divided battleground state that could, under the current system, determine the outcome of the national election.

For example, in 2004, George W. Bush had a nationwide lead of 3,012,171 popular votes—far greater than the number of provisional ballots nationwide. There has been an exceptionally high amount of provisional voting in Ohio in recent elections, including 2004. In 2004, there were more than 150,000 provisional ballots in Ohio, and Bush’s margin was 118,601 in Ohio in 2004.[351] The outcome of the 2004 election would have been reversed with a switch of 59,393 votes out of a total of 5,627,903 votes in Ohio. On the Wednesday after Election Day, Senator John Kerry decided that the provisional ballots were unlikely to reverse the apparent outcome in Ohio. If the number of provisional ballots had been somewhat higher or if Bush’s margin among the already counted regular ballots had been somewhat lower, the provisional ballots in Ohio would have decided the Presidency in 2004 (despite Bush’s already known nationwide lead of three million votes).

There has been about one such “near miss” election each decade under the state-by-state winner-take-all system. Table 1.23 shows there have been six presidential elections between the end of World War II and 2012 in which a shift of a relatively small number of votes in one or two states would have elected (and, of course, in 2000, did elect) a presidential candidate who lost the popular vote nationwide.

In 1976, for example, Jimmy Carter led Gerald Ford by 1,682,970 votes nationwide; however, a shift of 3,687 votes in Hawaii and 5,559 votes in Ohio would have elected Ford.

In 1968, a shift of 10,245 in Missouri and 67,481 in Illinois would have elected Hubert Humphrey as President despite Richard Nixon’s nationwide lead of 510,645.

The 2000 presidential election was decided by 537 votes out of a total of 5,963,110 votes in Florida—far greater than the number of provisional ballots that are currently cast in Florida.

Although the 2008 presidential election was not as close as 2000 or 2004, a relatively small number of votes determined the outcome in several states in which the number of provisional ballots exceeded the leading candidate’s margin in that state, including Missouri (McCain’s 3,903-vote margin out of 2,925,205 votes), North Carolina (Obama’s 14,177-vote margin out of 4,310,789 votes), and Indiana (Obama’s 28,391-vote margin out of 2,751,054).

There were nine closely divided battleground states in the 2012 election (section 1.3). Thus, there were nine states where provisional ballots could potentially have played a decisive role under the current state-by-state winner-take-all system.

Provisional ballots can be expected to produce disputes in future presidential elections because of the recent enactment of voter-identification laws in some closely divided battleground states. For example, although the voter-identification law enacted in Pennsylvania in 2012 did not take effect in time for the 2012 presidential election, it is expected to take effect in 2013.

The likelihood that provisional ballots might trigger a dispute in a presidential election is higher under the current state-by-state winner-take-all system than under a system in which there is a single national pool of votes.

A November 6, 2012, article in National Journal entitled “The Ohio Vote Count Could Be a Mess” stated:

“The Buckeye State has supplanted its Southern cousin Florida as the marquee battleground of the 2012 presidential election—the state most likely to tip the race to either President Obama or Mitt Romney.…

“Ohio also bears another, more ominous similarity to the 2000 Florida: If a close race demands a recount, conditions are ripe for a repeat of the delays, confusion, and chaos that racked the Sunshine State. And just like 12 years ago, the state’s ultimate winner could very well determine who is the next president.…

“The most obvious flash point involves provisional ballots, those cast if a voter’s eligibility is in question. Election officials don’t count provisional or absentee ballots until 10 days after Election Day. In case of a narrow margin and with hundreds of thousands of such votes still to be counted, neither candidate could claim victory. (Ohio recorded 200,000 provisional ballots in 2008, a number expected to rise this time.”[352] [Emphasis added]

A similar issue arises in connection with military and overseas absentee ballots. Under the Military and Overseas Voter Empowerment Act (MOVE), each state determines its deadline for receiving absentee ballots from military and overseas voters.

Although the process of properly counting all the legitimate votes may take some time, there should be no concern about the delay. Electing the right person to office is more important than a slight delay in ascertaining the outcome. As discussed in detail in section 9.15.3, the U.S. Constitution establishes a strict overall national schedule for finalizing the results of presidential elections. These existing provisions apply equally to elections conducted under the current state-by-state winner-take-all system as well as elections conducted under the National Popular Vote plan. 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). The nation knows, from experience in 2000, that the outcome of a presidential election must be resolved (one way or the other) in accordance with the schedule specified by the U.S. Constitution.

The possibility of disputes over provisional ballots is an example of a potential problem that is more likely to occur, and more likely to matter, under the current state-by-state winner-take-all system than the National Popular Vote plan.

Footnotes

At about 11:15 PM eastern time on Election Night in 2012 (shortly after the polls closed in California and other western states), the television networks called the 2012 presidential election in favor of President Barack Obama. Shortly thereafter, Governor Mitt Romney addressed the nation to concede that he had not won the election and congratulate the winner.

How is it possible for television networks to “call” elections and why do candidates concede on Election Night when there are:

• millions of votes cast on Election Day that are yet to be counted;
• millions of uncounted mail-in, absentee, and military ballots (which, in some states, need not even arrive at vote-counting centers until several days after the Election Day); and
• millions of uncounted provisional ballots (for which voters, in many cases, are not even required to step forward and provide evidence in support of their right to vote for 6–10 days)?

Both candidates and television networks routinely and confidently make decisions about the ultimate outcome of an election based on a combination of information sources, including:

• exit polls conducted outside polling places on Election Day,
• telephone and other types of polling indicating the likely breakdown of absentee, mail-in, provisional, and military ballots,
• estimates (obtained from both election officials and polling) of the number of uncounted absentee, mail-in, provisional, military, and regular ballots, and
• actual election returns (obtained from elections officials on Election Night).

Using these techniques, knowledge of the winner of the national popular vote for President has always been evident on Election Night.

In contrast, knowledge of the winner of the electoral vote has not always been evident on Election Night.

For example, because of the current state-by-state winner-take-all system, knowledge about the winner of the Presidency in 2000 was delayed until 34 days after Election Day (and six days before the meeting of the Electoral College on December 18, 2000). In contrast, the winner of the national popular vote in 2000 was evident shortly after the polls closed.

The 34-day delay in learning the identity of the President was an artificial crisis created by the current state-by-state winner-take-all system. The eventual deciding factor in the 2000 election was George W. Bush’s lead of 537 popular votes in Florida rather than Gore’s nationwide lead of 537,179 popular votes (1,000 times larger than the disputed 537-vote margin in Florida).

Notwithstanding these facts and history, it has been claimed that if the President were elected by a nationwide popular vote, knowledge of the winner would be delayed because votes from all 50 states (instead of just 10 or so battleground states) would matter in determining the winner.

On November 27, 2012 (three weeks after Election Day), the following complaint concerning the official count was posted on an election blog:

“Apparently only 17 states have completed their count of all ballots.…I think the implications for National Popular Vote are pretty obvious—had this been a closer election (say, Bush–Gore or Kennedy–Nixon close) we’d still not know who the president was.… The Electoral College seems to have provided conclusive clarity rather quickly.”[353] [Emphasis added]

Of course, in the very election that was “Bush–Gore close”—namely the Bush–Gore election in 2000—knowledge about the winner of the Presidency was delayed for 34 days because of the current state-by-state winner-take-all system.

In 2004, knowledge about the winner of the Electoral College was delayed until Wednesday morning even though it was clear on Election Night that President George W. Bush had won the national popular vote by about three million popular votes. If 59,393 Bush voters in Ohio had shifted to Kerry in 2004, Kerry would have ended up with 272 electoral votes (two more than the 270 necessary for election). The 59,393 voters in Ohio were decisive, whereas Bush’s nationwide lead of more than three million votes was irrelevant.[354]

Despite the complaint on the election blog concerning the 2012 election, the 2012 election was not close in terms of the national popular vote. President Obama’s multi-million-vote nationwide lead was evident on Election Night. However, the closeness of the race in numerous battleground states (e.g., Ohio, Virginia, Florida, Colorado, Nevada, Iowa, New Hampshire) suggests that if President Obama’s nationwide lead had been smaller than his actual nationwide lead of 4,966,945 votes (as discussed in section 9.31.9), knowledge of the winner of the 2012 election would likely have been significantly delayed because of the state-by-state winner-take-all system.

The complaint on this blog fails to distinguish between the two levels of “knowing” the outcome of an election.

The first level of “knowing” typically occurs on Election Night even though there are millions of uncounted ballots—regular, absentee, mail-in, provisional, and military. Nonetheless, sufficient information is available to enable television networks to reliably “call” the election and, more important, to compel losing candidates to concede defeat.

The second level of “knowing” the outcome of a presidential election comes later—namely the official count.

The official winner of the 10 closely divided battleground states was not known on Election Night. In fact, the official counts from eight of the 10 battleground states did not come in until after November 29—the day when the blogger complained that we might not “know who the president was” if the President were elected by a nationwide popular vote.

After Election Day in 2012, David Wasserman of the Cook Political Report monitored the official vote counts from each state and immediately posted each new result on the web.[355] Although procedures vary from state to state, the official count typically is certified by the Secretary of State or a board (e.g., Board of Canvassers, Board of Elections). Wasserman announced the completion of the official statewide count for almost all states with a Tweet.

Table 9.14 shows the approximate dates on which the 50 states and District of Columbia announced their official results of the presidential election (based on David Wasserman’s Tweets in most cases). The dates for five states are labeled “before”—indicating that the table contains the date on the state’s Certificate of Ascertainment. The Certificate of Ascertainment is typically created and signed (by the Governor) several days after the completion of certification of the official statewide count. Column 1 of the table indicates the order in which each state completed its official count. Column 5 flags the 10 states that many considered to be battleground states in 2012 (New Hampshire, Florida, Wisconsin, Nevada, Iowa, Ohio, Colorado, North Carolina, Virginia, and Pennsylvania).

Table 9.14 Approximate dates when states completed their presidential vote counts in 2012

 

As can be seen in table 9.14, eight of the 10 battleground states completed their official presidential count after November 29—the day when the blogger complained that we might not “know who the president was” if the President were elected by a nationwide popular vote. These eight states were:

• Wisconsin,
• Nevada,
• Iowa,
• Ohio,
• Colorado,
• North Carolina,
• Virginia, and
• Pennsylvania.

The blogger’s reference to the Kennedy–Nixon election in 1960 was also incorrect.

Kennedy was identified as the clear winner of national popular vote early in the morning after Election Day.

The New York Times’ front-page headline article on the day after the election (Wednesday, November 9, 1960) was “Kennedy is the Apparent Victor.”[356]

On Thursday November 10, 1960, the headline of the New York Times was “Kennedy’s Victory Won by Close Margin.”

“Fifty-two additional electoral votes, including California’s thirty-two, were still in doubt last night. But the popular vote was a different story.… Senator Kennedy's lead last night was little more than 300,000 in a total tabulated vote of about 66,000,000 cast in 165,826 precincts.”[357] [Emphasis added]

On Friday November 11, 1960, the headline of the New York Times was “Kennedy's Margin Is Under 300,000.”

Nonetheless, uncertainty about the electoral-vote continued. A front-page article in the New York Times on Saturday November 12, 1960, reported:

“The Republican National Chairman, Senator Thruston B. Morton … asked party officials in eleven states today to begin legal action to get recounts. The states were Delaware, Illinois, Michigan, Minnesota, Missouri, Nevada, New Mexico, New Jersey, Pennsylvania, South Carolina and Texas.”[358] [Emphasis added]

The electoral vote count remained unclear until Thursday November 17, 1960. The headline of a New York Times article on that day’s front page announced that “California Is Put in Nixon’s Column by Absentee Vote.”

“Senator Kennedy led in the tally of regular ballots with a majority of 34,568, but the absentee returns changed the picture. Mr. Nixon's lead rose to 13,160 with about 20,000 absentee ballots still to be counted. Most of these are in Republican areas.

“The absentee returns gave Mr. Nixon 132,168 to Mr. Kennedy's 84,458. State-wide, absentee and resident, the count was: Mr. Nixon, 3,219,211; Mr. Kennedy, 3,206,051. An official canvass, due by Nov. 28, will give the final result.”[359]

The final official count in California in 1960 was 3,259,722 for Nixon and 3,224,099 for Kennedy—a difference of 35,623 out of 6.5 million votes.

Under the current state-by-state winner-take-all system, the outcome of the national election frequently depends on the outcome of one or more closely divided battleground states.

For example, in 2004, George W. Bush had a nationwide lead of 3,012,171 popular votes. There has been an exceptionally high amount of provisional voting in Ohio in recent elections, including 2004. In 2004, there were more than 150,000 provisional ballots in Ohio, while Bush’s margin was 118,601 votes.[360] The outcome of the 2004 election would have been reversed with a switch of 59,393 votes out of a total of 5,627,903 votes in Ohio. On the Wednesday after Election Day, Senator John Kerry decided that the provisional ballots were unlikely to reverse the apparent outcome in Ohio. If the number of provisional ballots had been somewhat higher or if Bush’s margin among the already counted regular ballots had been somewhat lower, knowledge of the winner of the election in 2004 would have been delayed until the provisional ballots were counted (despite Bush’s already known nationwide lead of three million votes).

There is a far greater chance that knowledge of the winner of a presidential election will be delayed under the current state-by-state winner-take-all rule than under a system in which there is a single national pool of votes and in which the winner is the candidate receiving the most popular votes nationwide.

In any event, there should be no concern about the delay introduced by the official counting of ballots, because the U.S. Constitution establishes a strict overall national schedule for finalizing the results of presidential elections. 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. States are expected to make their “final determination” six days before the Electoral College meets (the so-called “safe harbor” date). The nation knows, from experience in 2000, that the outcome of the presidential election must be resolved (one way or the other) in accordance with the schedule specified by the U.S. Constitution.

An unusual situation developed in 2012 when Hurricane Sandy disrupted many parts of New York state a week before Election Day. On the day before Election Day, Governor Andrew Cuomo issued Executive Order No. 62, allowing any registered New York voter in the federally-declared disaster areas to cast a provisional ballot at any polling place in the state. The affected areas consisted of the five counties of New York City (Bronx, Kings, New York, Queens, and Richmond) and the counties of Nassau, Rockland, Suffolk, and Westchester. The Executive Order required every county in the state to transmit the resulting provisional ballots to the Board of Election in the county where the voter was registered.

The Executive Order resulted in 400,629 provisional ballots on November 6, 2012—about four times the number of provisional ballots handled in New York in 2008.

Counting provisional ballots is a time-consuming and labor-intensive task even under normal circumstances (see section 9.13.6). One reason that counting the provisional ballots resulting from the Governor’s Executive Order was unusually time-consuming is that a provisional ballot given to a voter outside his or her normal precinct would, almost always, contain some offices for which the voter was not entitled to vote. The detailed instructions accompanying the Executive Order illustrate the complexity of the situation:

“For example, a voter staying with family in Orange County who was displaced from Westchester, would be entitled to vote for statewide contests and Supreme Court (because those 2 counties share a judicial district) and possibly a congressional, state senate, or state assembly contest. A voter who sought refuge further upstate might only be eligible to vote in the statewide contests, as they would share no other offices/contests.”

Thus, when the provisional ballots resulting from the Executive Order arrived at each voter’s own local Board of Election, the receiving county had to determine whether that particular voter was entitled to vote for each separate contest that appeared on the sending county’s provisional ballot. A voter who was temporarily displaced to an adjacent county might, for example, still be in his or her own congressional district and state Senate district, but not his own Assembly district.

Obviously, if New York had been in the position of determining the national outcome of the presidential election (as Florida was in 2000 and as Ohio was in 2004), all of these provisional ballots would have been counted expeditiously—regardless of the cost of the overtime needed to complete the task.

In actual practice, the New York State Board of Elections certified a statewide count for President before the “safe harbor” day without considering the unexpected volume of provisional ballots. The state’s first certified count showed that the Obama-Biden slate had received 4,159,441 votes and that the Romney-Ryan slate had received 2,401,799 votes—a margin of 1,757,642 votes.[361]

Then, on December 31, 2012, the Board of Elections certified an amended statewide count showing that the Obama-Biden slate had received 4,471,871 votes and that the Romney-Ryan slate had received 2,485,432 votes—a margin of 1,986,439.

New York was not a closely divided battleground state in 2012, and therefore it was evident that its 400,629 provisional ballots could not have affected the nationwide outcome. Similarly, if the National Popular Vote compact had been in effect in 2012, it would have been evident that New York’s 400,629 provisional ballots could not have affected the nationwide outcome. Douglas A. Kellner, Co-Chair of the New York State Board of Elections has stated:

“If the final New York count had been required to determine the identity of the President, the New York State Board of Elections would have accelerated its official count—regardless of whether the outcome of the election was being determined by the state-level winner-take-all rule or the national popular vote.”

Footnotes

It is sometimes argued that the quality and trustworthiness of elections is so high in closely divided battleground states that the country should not risk an election in which the 40 or so non-battleground states might affect the outcome of a presidential election.

A small number of questionable votes in a single state is unlikely to change the outcome of a presidential election conducted on the basis of the national popular vote. It is, however, a historical fact that a small number of votes may affect the nationwide outcome of a presidential election under the current state-by-state winner-take-all system. For example, the 2000 presidential election was decided by 537 votes out of a total of 5,963,110 votes in Florida—one of the numerous battleground states that used direct-recording electronic voting machines in 2012.

The trustworthiness of elections has been questioned in numerous closely divided battleground states, including in Ohio, Florida, Colorado, and Pennsylvania.

In each of the states mentioned, proponents of various controversial measures argued that elections were insecure and unreliable.[362] In citing these examples, our purpose is not to agree or disagree with the rationale or propriety of these new measures, but to dispute the claim that elections in today’s closely divided battleground states are inherently more trustworthy than the rest of the country.

In Florida, for example, Governor Rick Scott (R) signed into law a controversial measure in 2011 that imposed more than 75 restrictions to combat voter fraud. The changes limited early voting, purged voter rolls of non-citizens, and made it more difficult for third-party organizations to register voters.[363] The article “The Battle over Election Reform in the Swing State of Florida” reviews numerous additional controversies concerning election law in Florida.[364]

In Colorado, Secretary of State Scott Gessler (R) launched efforts to remove certain ineligible registered voters from the voter rolls.[365]

In Pennsylvania, stringent voter identification legislation was enacted. Politics PA reported on June 25, 2012:

“House Majority Leader Mike Turzai (R-Allegheny) suggested that the House’s end game in passing the Voter ID law was to benefit the GOP politically.

“‘We are focused on making sure that we meet our obligations that we’ve talked about for years,’ said Turzai in a speech to [Republican State Committee] committee members Saturday. He mentioned the law among a laundry list of accomplishments made by the GOP-run legislature.

“‘Pro-Second Amendment? The Castle Doctrine, it’s done. First pro-life legislation—abortion facility regulations—in 22 years, done. Voter ID, which is gonna allow Governor Romney to win the state of Pennsylvania, done.’”[366] [Emphasis added]

Ohio was the key battleground state in both the 2004 election and the 2012 election. In 2012, for example, it accounted for 73 of the 253 post-convention campaign events in the 2012 election (table 9.3).

In Ohio, Secretary of State John Husted attempted to eliminate early voting during the weekend before Election Day; however, this change was rejected by federal courts.

A November 6, 2012, article in National Journal entitled “The Ohio Vote Count Could Be a Mess” stated:

“The Buckeye State has supplanted its Southern cousin Florida as the marquee battleground of the 2012 presidential election—the state most likely to tip the race to either President Obama or Mitt Romney.…

“Ohio also bears another, more ominous similarity to the 2000 Florida: If a close race demands a recount, conditions are ripe for a repeat of the delays, confusion, and chaos that racked the Sunshine State. And just like 12 years ago, the state’s ultimate winner could very well determine who is the next president. Part of the reason is that swing states such as Ohio haven’t adopted some of the reforms that Florida enacted after its infamous recount.…

“The most obvious flash point involves provisional ballots, those cast if a voter’s eligibility is in question. Election officials don’t count provisional or absentee ballots until 10 days after Election Day. In case of a narrow margin and with hundreds of thousands of such votes still to be counted, neither candidate could claim victory. (Ohio recorded 200,000 provisional ballots in 2008, a number expected to rise this time.).…

“The possibility of an outright recount further clouds Ohio’s outcome. The state will conduct an automatic recount if the difference between Obama’s and Romney’s tallies is less than one-quarter of 1 percentage point. But officials won’t begin that process until the election results are certified, which might not happen until early December. Each county has 21 days to certify its results before submitting them to the secretary of state.”[367]

We are not aware of any evidence that the trustworthiness of elections in closely divided battleground states is better than the rest of the country.

Footnotes