Why Mail Ballots Are A Bad Idea by Charles E. Corry, Ph.D.
You can have an honest election, or you can have a mail in/absentee ballot election, but you can't have both at the same time.
By November 2006 the Help America Vote Act of 2002 (HAVA) (aka Leave No Voting Company Behind) tidal wave had washed almost completely across America destroying election integrity and trust in its wake, and brought to us by the same "leaders" who brought us war in Iraq; a war on drugs; the wholesale destruction of children, families, and marriage; torture; gulags; reinstituted indentured servitude and debtors prisons; and incurred a national debt of nearly $9 trillion dollars that is increasing by $1.7 billion a day.
One need not read far into the tabulation of problems by VotersUnite, or here, to realize electronic voting has been a massive failure. The innumerable problems, in many cases initiated by requirements of HAVA, has led to often extreme distrust of voting machines in polling places. Nowhere is this more apparent than in November 2006 election in Riverside County, California, one of the first counties to switch to electronic voting.
In many cases, dysfunctional voting machines and incompetent or dishonest election officials have led to outrageous waits for voters at polling places, e.g., in several Colorado counties in November 2006 the last voters were not able to cast a ballot until 1:30 AM the next morning at voting centers . In some cases, notably Ohio, election officials apparently deliberately put too few electronic voting machines in minority or Democratic neighborhoods, forcing many potential voters to turn away rather than wait in line for many hours to vote. And, if citizens are able to vote at a precinct, the innumerable problems with electronic voting machines documented here and on many other web sites leaves voters justifiably uncertain if their vote was counted and, if counted, counted correctly?
In order to avoid the lines at polling places, and with well-founded mistrust in touchscreen (DRE) voting machines, an ever-increasing number of voters have taken to using absentee ballots in the correct belief that a hand-marked paper ballot is more durable and accurate than an ephemeral entry on a computer screen.
But absentee mail ballots are still counted by electronic voting machines, only now it is done in the proverbial "back room" largely out of public view, which suits embattled election officials and voting machine manufacturers just fine.
While election officials are being pummeled by public distrust of electronic voting, they are, as in the past, going in the wrong direction. Despite an unbroken record over the past decade of making elections worse, the apparent stampede of these simpletons is to propose all-mail ballot elections, as has been done in Oregon. The Oregon experiment is reviewed, and not favorably, by Prof. Melody Rose and Thomas Hargrove. But far be it from our apparently retarded election officials to be deterred by failures elsewhere. Besides, the increased use of absentee ballots makes it appear to them that it is "the will of the people" to have mail ballot elections.
And we certainly wouldn't want to return to the old-fashioned method of hand marking and hand counting paper ballots at our local precinct that worked so well for so many years. That would make the waste of public money and distrust in electronic voting machines too painfully obvious.
In early 2006 I was asked by a local election official to tabulate the problems I'd seen with mail ballot elections and absentee balloting. Obviously, conscientious election officials do their best to minimize these problems. However, the "less conscientious" do their best to simply hide "mistakes" and all too often we've encountered, and document in this chapter, incompetent or corrupt election officials who ignore or are ignorant of the problems listed here.
It is also impossible for election officials to defend against and prevent all the problems listed here in a given election using mail ballots. Thus, while the limited use of absentee ballots may be necessary, their usage should be strictly controlled and the closest possible scrutiny applied to all ballots sent and received by mail.
As noted in this chapter, all-mail ballot elections have been widely touted, but have not been as successful as politicians would like us to believe. In no case should all-mail elections be used, especially in special district elections involving developers, or other elections where large dollar or tax issues are at stake.
With the present trend to allow voter registration forms by mail, fake or invalid voter registrations are not limited to mail ballots. But precinct voting puts an additional check as the individual must appear in person at each polling place. Since no personal appearance is ever required with mail-in registration and a mail-in ballot, it becomes trivial to vote under a fake or assumed name.
This problem is so widespread that the states of Alaska and Montana, and 261 counties in other states certified more voters than the actual adult population during the 2000 general election. Phantom voters are also widespread in Colorado and California.
While precinct voting requires the voter to appear in person and present some identification that matches the information in the precinct poll book, anyone can mail in a ballot for a name that appears on the statewide or county voter registration rolls. And with mail in elections there are thousands of ballots floating around.
While "inactive" voters, i.e., anyone who didn't vote in the last federal election can obtain a ballot by contacting the county clerk, most voters are unaware of that requirement and are never given a chance to vote.
Mailing ballots only to "active" voters allows county clerks to play games with costs as well. Obviously if you don't have to print and mail ballots to one-third of the registered voters the cost will be less. Is saving money a primary objective of an election?
One also hears claims that voter turnout is greater with mail ballot elections. The way that game works is that clerks divide the number of ballots returned (dividend) by the number they mail out (divisor) rather than the total number of registered voters as they should. Since ballots are only mailed to "active" voters, who are more likely to vote anyway, the dividend is increased, and the divisor is decreased by one third, the apparent percentage will obviously appear to increase. In fact, no significant difference in voter turnout has been established between precinct and mail ballot elections as shown below.
Games are also played with mailing ballots to other voters than those marked "active" in the poll books. Valid reasons for doing this include ballots mailed to voters who registered after the last general election and voters who request a ballot. However, there are no constraints or checks on who these supernumeraries are and it gives the unscrupulous plenty of room to send ballots to parties otherwise unqualified to vote or who deliberately send in falsified registration forms in order to obtain a mail ballot. And it should always be recalled that the first step toward stuffing the ballot box is obtaining a ballot.
Experience has shown time and again that if there is a loophole in the way an election is run that someone will take advantage of it. There is simply too much money and power traded in an election for fraud not to be an issue.
The pitfalls of voter registration are covered in a separate chapter but make it relatively easy to ensure that ballots are only sent to "selected" voters in mail ballot elections. That already occurs as noted above but can easily be fine tuned further.
Other tricks involve putting one group of voter's ballots in a different type of envelope that is easily recognized and intercepted. That method is most effective in small, special district elections. Or the voter's party affiliation may be printed on the envelopes, as done in Florida, making it easy to "lose" ballots from the "wrong" party.
Election officials are either unaware of, or choose to ignore the exploding problems with identity theft. In February 2006 150,000 voter records went missing in Denver, Colorado. In an ill-fated experiment with voting centers during the November 2006 all voter registration records for Denver City and County were put on dozens, perhaps hundreds of laptops distributed to all the voting centers. Theft of those records would have been, and probably was, trivial.
Voter registration records contain voters' names, Social Security numbers, birth dates, signatures, and addresses. Only the most naive or incompetent could claim there is no risk of election fraud and identity theft using these records.
Many times the address isn't valid, as anyone who has ever walked a precinct for a candidate will say. Or a dummy address was given, the address given was simply wrong, the address given was unreadable by the election staff, or they copied the address wrong, e.g., mistaking a 1 for a 7. In all such cases the voter doesn't receive the ballot requested and may not realize it until too late to obtain a replacement and is disenfranchised as a result.
Unless the mailed ballot is received back by the election officials there is really no way they can tell whether or not the ballot reached its destination. There is also the open question of whether the destination the ballot reached was the registered voter for whom it was intended, or was it intercepted (ballots are frequently stolen from mailboxes).
If the wrong writing instrument is used to mark a ballot the optical scanners that count the ballot may not read it correctly. Optical scanners used for counting votes may only read pencil, may be sensitive to type or color of ink used in a pen, or voter may use a colored pen or pencil, or even crayon that doesn't scan correctly.
For additional information on this problem see the report by Prof. Douglas Jones (PDF) on tests conducted in Maricopa County (Phoenix), Arizona. An example of this problem in an election is described for the November 2003 election in Garfield County, Colorado.
Improperly marked ballots may or may not be counted by the voting machines. In one documented case voters were instructed to use pen in the instructions on the envelope and to use a pencil on the ballot.
No method exists of checking for duplicate handwriting if an individual has registered under multiple names at multiple addresses by mailing in multiple registration forms. The chance of detecting such fraud is very low and the probability that the perpetrator will be successfully prosecuted is near zero.
Signature verification is used to reduce this but many spouses forge their partners signature. Digital signatures can be purchased or stolen. Scanning and printing someone else's signature is easily done with today's technology.
This problem can be avoided by using an extra secrecy envelope inside the mailed envelope. But an extra envelope adds to costs of printing, collating, and postage so many jurisdictions don't use them.
Careful ballot control and inventory is one of the most basic precautions inherent in an election to forestall ballot box stuffing. However, with mail elections tens of thousands of ballots simply "disappear," or are found in dumpsters, garbage cans, etc. Unscrupulous individuals then go "dumpster diving" at appropriate times after ballots are mailed and attempt to vote the discarded ballots.
America is a nation of transients and roughly one quarter of the population moves every year. However, citizens rarely tell the county clerk at their old address that they've moved. So mail ballots are received by the new tenants, owners, college kids, motels, realtors, developers, etc., and these ballots may be forwarded as well.
There are tens, and sometimes over a hundred different ballot styles in an election. Each ballot style contains only those candidates and issues a voter is eligible to vote on based on their physical address. In a precinct election it is fairly clear what candidates and issues are unique to that precinct and what ballot styles are given to voters by the election judges.
A fundamental protection provided by a voting booth in a polling place is privacy and the ability to vote in secret as one sees fit for the candidates and issues of the voter's choice, free of intimidation and outside influence (electioneering is usually forbidden within 100 feet of a polling place).
That fundamental protection is entirely lost with mail ballots. How you vote is open to inspection, influence, and review by everyone the voter interacts with. Some may find it acceptable, and even desirable, to sit around the kitchen table, or in a neighborhood or union meeting and fill out their ballots. Usually, however, such individuals have something to gain by influencing how people vote, e.g. the voter's employer.
In a precinct election a fundamental protection for voters is the prohibition of candidates, representatives, or any election materials, e.g. posters, signs, handouts, etc., typically within 100 feet (33 meters) of the polling place so that voters are undisturbed and unsolicited while they vote in secret.
Just the opposite occurs with mail ballots. Generally candidates or supporters can get a list of who has requested or been sent a mail ballot from the county clerk and then call the voter, send mail to them, even visit them in order to influence and pressure the voter into voting for them or the candidates and issues they support.
The major problem with buying and selling votes is verifying how the voter actually marked their ballot. In precinct elections the voter only receives a ballot after identifying themselves to the election judges. The voter then takes the ballot to a booth where they vote in private and no one else can see how they've voted. Once the ballot is marked any identifying information is removed and the ballot dropped in a sealed box.
At no time is a voter allowed to leave the polling place with their ballot, marked or unmarked, to avoid any possibility the voter might be selling their vote, formerly a very common practice in American elections. Nor are cameras or cell phones allowed in polling places.
As noted above, by design verifying how the voter marked their ballot is very difficult to do in a precinct election. However, with a mail ballot there is no problem at all verifying how the voter marked their ballot before they put it back in the mail to be counted. In fact, the buyer will often insist on mailing, or delivering the ballot themselves just to be sure.
In Texas this process is so ingrained that a cadre of vote whores are regularly employed to collect ballots and ensure they are "properly" marked. In Colorado this process has been especially pronounced in special district elections where real estate developers are involved.
In some locations, e.g., Florida, the voter's party is stamped on the envelope by the county clerk before mailing. Thus, activists can often ensure voters from the opposing party do not receive mail ballots. In other jurisdictions, or in special elections, other forms of marking, e.g., different colored envelopes have been used to distinguish between ballots for opposing camps.
In the November 2003 election in Garfield County, Colorado, the counting was done by the county clerk's son and, although the scanner stopped some 1,700 times, the problems were ignored. Eventually that election had to be hand counted by the Colorado Secretary of State's office and the outcome of one candidate race and one tax issue were reversed by the hand count.
In an April 1, 2003, city election in Colorado Springs poll watchers were restricted to a nine square foot area and could not directly observe any counting procedures or ask questions of anyone but the city clerk. These are but two examples of hundreds.
If there is a problem with a voter's ballot in precinct voting they are given an immediate opportunity to correct it. While in some isolated cases election officials may call a voter if there is a problem with their mail ballot, typically that doesn't happen. Even if a voter is notified of problems with a mailed ballot it leaves the door wide open to vote manipulation.
As a result, many valid mail ballots are disqualified, rejected, lost, stolen, changed, or delayed beyond the election deadline. Usually there is no easy way for a voter to find out if their mail ballot was counted. And there have been repeated incidents of uncounted mail-in ballots turning up in the clerk's office weeks or months after the election.
Again, the November 2003 election in Garfield County, Colorado, is a perfect example of this problem. However, the same problem was apparent when we looked at the logs from the Sequoia optical scanner in the November 2003 mail in election in Denver.
Problems with optical scanners misreading folds in mail ballots have also been discovered. Also, ink or toner may be transferred on a ballot in the press of mailing and the smudge read by the scanner as a vote or overvote.
In a mail in election, or with absentee ballots, it is impossible to know how often election officials turn off the switch on the optical scanners that rejects mismarked ballots. The problem is particularly acute where election officials use autofeeders to push ballots through the scanner. This problem can be somewhat alleviated by hand feeding the ballots but that leaves the process wide open to manipulation as to which ballots are counted and which rejected.
When a ballot is rejected for whatever reason by an optical scanner in a precinct election the voter is immediately given a replacement ballot and help, if needed, on how to properly mark their ballot before they return to the voting booth.
With mail ballots, when the scanner rejects the ballot election judges, or the clerk's crony or relative, fills out a new ballot "interpreting" the voter's intent as best they can. As in Colorado Springs, this is often done outside the view of poll watchers.
When the optical scanner rejects a ballot it is necessary to rescan the ballot either after the ballot has been redone by election officials, or by simply reinserting the ballot into the scanner in a different orientation. Thus, the ballot count may easily become confused and the election arithmetic uninterpretable. This is particularly true in small jurisdictions, special elections, or with inexperienced, corrupt, or incompetent election officials. Thus, the door to manipulating the vote count is left wide open.
Unless the voter calls the county clerk after the election they have no idea whether their ballot was received in time to be counted in the election. Even if their ballot was received they have no idea whether it was counted as marked.
Mail ballot envelopes are often opened and enclosed ballots reviewed by the same person, or by election judges sitting at the same table. Thus, it is relatively easy to determine who voted a particular ballot and how they voted.
In addition to the issues outlined above, counting mail ballots with electronic voting machines using optical scan methods leads to another set of problems. The more common ones known are listed below.
Note that there is a considerable difference in the requirements for a machine to scan a few hundred ballots, with just one or two ballot styles, in a precinct compared with the necessity of accurately and reliably scanning hundreds of thousands of ballots, with tens or hundreds of ballot styles, in mail in elections. An optical scanner that performs flawlessly in a precinct is much more likely to fail, or produce incorrect results when used to count ballots in the more demanding production environment of a mail in/absentee election.
However, as the ballots are counted in the "back room" at the county or city clerk's office, a concerted effort is often made to cover up problems and only the most obvious and egregious errors become public.
• Marking devices, i.e., pencil or pen, wrong ink color, mark intensity, etc., on paper ballots not recognized and votes are not counted by scanner. For an excellent review of such problems see the Statement regarding the optical mark-sense tabulators in Maricopa County, Arizona by Prof. Douglas Jones.
• Scanner heads become dirty or scratched and introduce reading errors. For example, voters at home may use correction fluid on their ballot that may wipe off on the read head of the scanner, or food gets spilled on the ballot that transfers to the scanner. This is a particular problem with mail elections where tens or hundreds of thousands of ballots may be scanned with a single machine.
Note: This tabulation is not exhaustive and many more scams are possible, and commonly used with mail ballots. That is why they are currently the method of choice for election fraud.
Since 2002 another requirement has been added to the voting process by the Help America Vote Act. That federal law requires that provision be made for handicapped voters to be able to vote unassisted.
The most common reason given to justify the necessity for an all-mail ballot election is lower cost. Aside from the point that the primary requirement for election officials is to provide an open, honest, and secure election, there is little evidence that an all-mail ballot election costs anymore than a well-run election done in polling places.
However, the typical cost comparison between a polling place election and a mail ballot election ignores the facts that a polling place election today is burdened with the HAVA mandates that every polling place have at least one expensive electronic voting machine that allows handicapped voters to cast a ballot unassisted, as well as provisional balloting that commonly requires at least one extra election judge in each polling place. These federal mandates are the reason a polling place election now appears to cost more. Table 1 shows a comparison of costs typically incurred in polling place and all-mail ballot elections.
Most election districts now also have early voting in polling-place elections, an additional expense. Further, most election districts also allow "no excuse" absentee balloting as well and roughly a third of voters now use that method even when polling places are available.
If the extremely expensive, and demonstrably unreliable and untrustworthy electronic voting machines were eliminated, and hand-counted paper ballots used, then the cost differences would disappear. Even with all the current impedimenta the cost difference is only about 20-25% more for a polling place + absentee + early voting + provisional ballot election, depending on whose numbers are used.
There appears to be little reason for early voting when voters can vote absentee, and that expense could easily be eliminated without significant inconvenience to voters and at considerable savings for the cost of an election.
I would also advocate returning to a requirement that a voter appear in person to obtain an absentee ballot. The current "no excuse" method of writing or calling to get a ballot is alarmingly insecure, particularly coupled with mail-in voter registration, as it currently is. Having the voter appear in person is infinitely more secure and saves the costs of mailing as well.
The other chestnut that is inevitably trotted out to justify, and ignore the obvious dangers of mail ballot elections, is the claim that they increase voter turnout. I have looked at that claim using election statistics from 1992 through 2006 provided by the El Paso County, Colorado, county clerk and Colorado Springs city election results.
The results from 21 elections are summarized in Table 2. There are 4 presidential races, 3 off-year congressional races, 5 coordinated county elections, 7 primary races, and 2 municipal elections included in the summary. Colorado law currently prohibits mail ballot elections in federal races so the three mail ballot elections included in Table 2 involve county and municipal elections only. But there are enough of each type of election to draw some preliminary conclusions.
By Mail 2
District 11 recall 5
3. Precinct elections include voting at a polling place or precinct, early voting, and absentee voting by mail. All mail in elections are done exclusively by mail and ballots are only sent to "active" voters, i.e., those who voted in the last federal election.
4. Registered voters in the City of Colorado Springs, which comprises roughly two thirds of the residents of El Paso County. Note that the city clerk only lists "active" voters as "registered" voters (in parentheses) in her tabulation when the election is by mail balloting.
• Presidential elections — These elections are usually of most interest to voters and the turnout, ranging from a low of 59% to a high of 81% in the four presidential races in Table 2 reflects that interest. Voter turnout in presidential elections is always higher than in any other type of election and these are always polling place elections under current Colorado law. Percent of voters who used absentee ballots ranges from a low of 9% before "no excuse" absentee voting began circa 1993, to 25% in the 2004 election and 34% in the 2000 election after "no excuse" absentee balloting was introduced.
• Congressional elections — Off-year congressional elections combined with state legislator races are also of great interest to voters. In the three races in Table 2 voter turnout ranges from a low of 46% to a high of 51%. Again, these are polling place elections under Colorado law. When "no excuse" absentee voting was initially permitted only 14% of voters used that method. As "no excuse" absentee voting was implemented, and distrust of electronic voting machines grew, 30-31% of voters now use an absentee ballot.
• Coordinated elections — These involve election of county officials and often tax or local issues. Five coordinated elections are included in Table 2 including one all-mail election in 2001. Voter turnout in coordinated elections ranges from a low of 19% to a high of 36%. The all-mail ballot election, with a turnout of 31%, falls in the middle of this range. The use of absentee ballots in these elections is consistently rather low, ranging from 8-12% when "no excuse" absentee voting first began, increasing to only 14-21% after it became common.
• Municipal and special district elections — It is in these types of elections that use of all-mail balloting is most dangerous. The numbers of eligible voters is smaller, special interests are likely involved, e.g. developers, and the outcomes can usually be more readily changed by differences of a few hundred votes. Four of these elections, three of which are mail ballot elections, are listed in Table 2. Numerous problems with the April 2003 Colorado Springs mail-ballot election are detailed elsewhere in this chapter. Turnout was a modest 37% for that election, which had a large number of candidates for city offices and a tax issue on the ballot. The December 2006 school district mail ballot election had only a 20% turnout and the April 2007 Colorado Springs mail ballot election had a turnout of only 27%. Note also that in the April 2005 Colorado Springs election turnout was only 15% but 50% of those who did vote, voted by mail (absentee). Thus, if mail balloting were to be given credit for increasing turnout, the April 2005 election should have had a much higher number of voters.
• Primary elections — These occur in August and voter apathy is apparent in the five primary elections in Table 2, which seldom have any real contests in heavily-conservative El Paso County. Turnout ranges from a pathetic 7% to 21%. Even a hotly contested Republican primary in 2006 for a congressional seat only provoked a turnout of barely 17%. When voters do participate in primary elections they now tend to vote more often by absentee ballot than in other types of elections. After "no excuse" absentee voting was introduced 22% of votes were mail ballots in the 2000 primary, increasing to 40% in 2006. It is also worth noting that electioneering was quite successfully used to influence those who voted absentee in the August 2006 primary election.
Clearly, voter turnout is influenced more strongly by the type of election, with presidential and congressional elections always drawing the largest turnouts, than by whether voting is based on polling place elections or mail ballot elections are used. Allowing "no excuse" absentee voting after 1993 has clearly increased the use of that method, with concomitant increases in the ease of casting fraudulent ballots and the virtual certainty that is now being done in many elections.
One can't logically ascribe any voter's desire to avoid standing in line for hours at a polling place and, thus, choosing to use an absentee ballot as a matter of "convenience." Instead, the increased use of absentee ballots appears to be the direct result of malfeasance by election officials and voting machine manufacturers whose expensive, unreliable, and untrustworthy voting machines have made polling place voting more difficult and time consuming, and deservedly increased public distrust and ire over election misconduct by orders of magnitude.
"Convenience" is a justification only the morally bankrupt would use for promoting the use of mail ballots.
It seems obvious that, at best, misguided efforts to increase voter turnout by making it more "convenient" to cast a ballot by mail have radically increased the probability and ease of election fraud. That contravenes the entire basis for an election and undermines the foundations of our democracy.
With the legions of scandals surrounding electronic voting machines, which are also used in a "backroom" to count mail ballots, public trust in fair and honest elections is plummeting. All-mail ballot elections have the appearance of simply one more effort by election officials to hide their mistakes and the problems with electronic voting machines, further increasing public distrust.
Another malodorous effect appears to be emerging from efforts to increase voter convenience and turnout. From the Colorado Springs mail ballot elections tabulated in Table 3 for 2003 and 2007 that were supposed to increase voter turnout and convenience, as well as encourage citizens to register, one finds instead that turnout dropped from 37% in 2003 to 27% in 2007. Further, despite an increase of 32,000 (8.6%) in the city's population, the number of registered voters is essentially unchanged at 222,600 ±100 in 2003 and 2007. And the number of registered voters in the city was 7,200 higher for the 2005 polling place election (Table 3).
By Mail 2
3. Precinct elections include voting at a polling place or precinct, early voting, and absentee voting by mail. All mail in elections are done exclusively by mail and ballots are only sent to "active" voters, i.e., those who voted in the previous federal election.
The data in Table 2, as discussed above, also fail to show any increased turnout that correlates with the "convenience" of a mail ballot election.
While it is premature to form any conclusions, the evidence to date suggests that making elections more convenient for voters and easier for citizens to register may actually have the opposite effects.