An Electoral System You Can Count On
By Gary Krasner (01/12/06)
January 1st was the deadline for states to enact voting-related legislation that's in accord with the 2002 federal Help America Vote Act (HAVA). By so doing, states would be eligible for federal grants to finance wide-ranging measures to improve voting systems, making polling places and voting devices handicapped-accessible, ensuring that only eligible voters are registered to vote, and stuff like that.
The closeness of the election in 2000 brought attention to the flaws and sloppiness in state electoral systems. Two percent of the votes cast could not be counted for a variety of reasons. Votes from ineligible voters or nonexistant people could be just as massive. We now face the prospect of judges deciding all close elections, instead of voters.
For the most technologically advanced nation in the world, I feel certain I'm not alone in wondering why we can't accomplish so simple a task as accurately tallying votes. At first glance, it would seem like a rudimentary undertaking for a nation that landed a man on the moon 35 years ago. (Pardon the cliché.) But not all the flaws with the system are based on technology. And each of the two political parties have their own agendas for mending the system.
What Republicans Want
Each political party obviously wants what it thinks will benefit it the most. What Republicans want most is that votes which are cast are by eligible voters, and that local precinct workers have as little access as possible to completed ballots.
On the latter point, Republicans believe that to the extent that it's feasible, it may reduce ballot tampering. This concern can be traced to elections well before 2000. Ballot stuffing in Democrat-controlled inner cities is legendary. It can include scanning completed paper ballots more than once, or by padding extra votes into lever-type voting booths before the polls open or after they're closed. Or after an election is decided, election workers can "discover" completed ballots or provisional ballots that had not yet been tabulated. In the 2004 Washington state gubernatorial election, 573 such ballots were found, which miraculously made up the margin that the Democrat candidate needed to win. The opposite of ballot stuffing would include the intentional loss, damage or manipulation of completed ballots so that they cannot be tabulated, or the voter's intention be determined.
Democrats counter that ballot manipulation of these sorts can be prevented with better trained election workers or more poll watchers. But many counties cannot afford the expense, or find sufficient volunteers. This often leads to situations in which the minority party is unable to observe as many polling sites as the majority party. Under such conditions, it just takes a few hyper-partisan workers to do what they think is best for the nation—one way or the other.
With regards to the former point (voter fraud), polls show that four out of five Americans support the requirement of some form of photo ID to register to vote. Yet only 17 states require some form of documentation. Compare that to the ID requirements to rent a video or similar endeavors which we would agree holds less collective importance to our republic.
Republicans would note that in 2001, voters registered in several cities totaled more residents over the age of 18 that was indicated by the U.S. Census. Much to blame for this was President Clinton's "Motor Voter Law", which required state Departments of Motor Vehicles to register to vote all license holders, even those which offer mail-in registration which don't require identification. Board of Election officials were barred from challenging the legitimacy of an individual's voting status, and the law made it difficult to purge former voters who were deceased or moved away.
For example, voter registration rolls in Philadelphia spiked 24 percent since 1995, while during this time its population declined by 13 percent. In 1999, CBS's 60 Minutes found in California, completed mail-in registration forms for fictitious people, and the subsequent absentee ballots filed in those names. Authorities in Wisconsin convicted a heiress from New York working for Al Gore of bribing homeless people with cigarettes if they rode in a van to the polls and voted. In 1997, Almost five thousand homeless or other ineligible voters were paid $10 apiece and driven in vans to file absentee ballots for Miami mayoral candidate Xavier Suarez.
The provisions in HAVA which is hoped to meliorate registration abuses included incentives for states to improve screening of non-citizens, convicted felons (if the state bars them from voting), people registered with the names of deceased citizens, and the legally incompetent. Republicans support such measures, realizing that they can't compete with Democrats in reaching into the pool of poor, immigrant, or otherwise disenfranchised unregistered and/or ineligible voters (With the intent to defraud.)
Democrats might privately justify violating or stretching voter registration laws on the grounds that they need to level the playing field against the (supposed) well-funded Republicans. But Democrats obviously can't publicly argue in favor of allowing ineligible or unregistered voters vote. So instead, they counterattack with the claim that residency screening and requiring proper IDs (etc.) are racially-motivated voter intimidation to suppress voter turnout, which denies minorities and the poor their rightful franchise. Yet in reality, wide-scale, intentional voter intimidation is either a relic of the past, or negligible in terms of the percentage of the vote affected. Part of the reason is that it's the only form of voting fraud that's prosecuted to any extent. By contrast, racially-charged allegations of voter suppression have been enough to dissuade federal prosecutors from pursuing incidents of voter fraud. Yet permitting vote fraud to continue through lax registration requirements denies legitimate voters their rightful franchise just as much as unwarranted restrictions on voting might.
What Democrats Want
For obvious reasons, Democrats—and most prominently, the left-wing of that party—have been the most vocal in challenging the efficacy of the last two national elections. They claim there needs to be more accurate vote counts. Though one must wonder about the veracity of that claim, given that historians and pollsters think electorate has shifted more conservative over the last 3 decades. Jimmy Carter was the last Democrat that won the presidency with a majority of the vote.
Nevertheless, the type of election fraud Democrats are most concerned about is vote count, and their catch-phrase since 2000 has been, "let all votes count." Even if the intention of the voter can't be discerned, or his eligibility to vote is in question. Of course, they always leave out that second clause, but they'll gladly settle for those "flaws".
Fortunately, Congress and the Election Assistance Commission (EAC), set up by HAVA, looked beyond bumper-sticker slogans. One of them pertained to voter errors, such as overvoting and undervoting the ballot—which is most prevalent among the elderly, minorities, or poorly educated voters using punch card or optical scan voting systems. Studies show that loss of votes from "spoilage" (either the voter or the precinct worker unintentionally mishandles the ballot, rendering it unreadable) adversely affects minorities by a factor of 3 to 1. Essentially, it's the same socioeconomic group most prone to over- and under-voting ballots.
Curing this was one of the mandates of HAVA. The prevailing assessment is that to best assist the handicapped, elderly, non-English speaking or poorly educated people to vote accurately, calls for a technology that the Left has resolutely opposed: A digital ballot, or "DRE" ("Direct Recording Electronic"). Essentially, it's a computer that uses touch-screen activation for casting the vote. Since most households today have computers with mouse user-interfaces, I believe DREs should have that as an additional or backup option.
But Democrats on the left fear that this technology will provide Republicans with an edge in ballot tampering, essentially because most companies that produce DREs are run by Republicans. Their opposition to electronic voting builds on an existing bias: Corporations create DREs; corporations are evil; therefore, DREs are evil. Or at least, their vote tabulations cannot be trusted. But there are myths abound.
One is the notion that computer hackers can alter the vote totals via the internet. But DREs are stand-alone devices not connected to the internet. In the 20 years they've been used, there's no verified case of tampering with their vote totals, despite allegations that incidents where malfunctions occurred were instead intentional attempts at fraud.
Another example involves allegations that DREs produce a disproportionate number of undervoted ballots among minorities—the group that was supposed to benefit from DREs. But at least for the presidential elections in 1992, 1996, and 2000, the type of voting machine used accounted for a mere 2.5% of the difference in undervoted ballots. In other words, neither punchcards nor optical readers nor DREs with touch-screens actually produced significantly higher rates of undervoted ballots among Blacks or other minorities. The left might still allege fraud, yet the undervotes registered on DREs might have been either voter-intended, or device malfunctions.
The overall performance of DREs has not been perfect to be sure, any more than optical vote scanners have. There's lots of room for improvement for DREs. Touch-screens can be misaligned. Software can have bugs and produce glitches resulting in votes being lost or miscounted. There's also unconfirmed claims that localized electronic hacking is possible.
But many besides the left are attentive to problems with DREs. The GAO has called for "rigorous security and reliability standards", but it will not be ready by the next election. Such failings are intolerable. It undermines faith and integrity in the electoral process, and plays into the paranoia on the left—which can be summarized simply as, "rage against the (Diebold) machines."
The quality of the solutions thus far vary, and are of limited scope. Senators Clinton and Graham filed legislation that would require computerized voting devices to produce paper records of the vote. But as a remedy to prevent fraud, it's also necessary to direct what must be done with that paper after it's produced. I think my plan in the next section tries to address that.
Personally, I would require there be entries on each line for "no choice", for which the voter must select, to indicate an intentional non-vote for any candidate. Also, for the sake of perception if nothing else, companies that produce DREs should be restricted to far less access to the hardware and software than they appear to have now. My proposed plan addresses that as well.
The plan is necessarily bold and comprehensive, which is required to satisfy all the transparency issues. Because even with the remedies that GAO and others have urged, computers are by their nature complex devices, and who's inner workings are less visible than paper or punch ballots, which people on the left still prefer. The voting system I outline offers the efficiency and accommodation to the handicapped (etc.) that only computer interfaces can offer, but with the transparency and confidence that paper ballots are perceived to offer.
If DREs continue to be employed absent of a comprehensive plan like mine, then I fear that confidence in election results will wane. Computer glitches will never be proven to the losing side to be accidental. Opponents of electronic ballots will continue to argue that it's impossible to verify that the vote which was cast was the vote that got tabulated. And the other side will continue to allege that the vote that was cast wasn't from an eligible voter. And of course, whenever exit polls don't resemble the subsequent actual vote, the left will challenge the accuracy of the actual vote.
[Note that when this occurred in 2004, not only had left-wing Democrats dismiss the admissions from liberal mainstream media that their polling consortium had skewered their test samplings more toward Kerry, they also wouldn't accept the logical sequelae that it had the effect of discouraging Republican voters and depressing turnout for Bush. And all this occurred BEFORE their rage against the DRE machine had taken hold the way it has now. Today, the paranoia and distrust is at it's zenith.]
One can only appreciate the juxtaposition from the usual political stances: The Republicans favor the HAVA voting reforms that call for modernization, while Democrats prefer the status quo and urged states not to enact laws to qualify for the HAVA grants. They oppose robust voter eligibility screening and want states to keep using paper ballots or the old lever-operated mechanical machines.
Preventing Voter Fraud
The ideal remedy to all forms of voting frauds must be comprehensive, else those who feel cheated from ballot tampering will feel justified in perpetrating voter fraud, and vice versa. Thus, we cannot merely seek the ideal voting device, for example, without trying to get voter fraud under control. As John Fund notes in "Stealing Elections", prosecuting election fraud is almost nonexistent for a variety of both political and practical reasons. So we need a systemic, uniform and comprehensive solution.
Almost every type of flaw in our voting system which I mentioned in the previous sections can be traced to one factor common in all balloting systems used in the U.S.: A ballot once cast cannot be traced to the voter who cast it. It might seem obvious that this prevents us from discovering what Republicans seem most concerned about: ineligible voters. But it also prevents what Democrats claim to want most—that all ballots are counted, and counted accurately.
My proposed solution might also seem to violate an axiom in this country: the privacy of our vote. But with encryption technology, that should not be a concern. After I run through my proposed ballot system, you will see how that's accomplished:
The state supplies a registered voter with a unique registration number. That number will never have other data linked to it, other than the person's name, date of birth, current residential address, political party registration, and a face photograph of the person. As it's always been, voter registration information will be in the hands of local election officials, to allow them to properly identify people as eligible to vote or not. Election officials will never have access to how people voted. As you'll read later, such information might only need to be "unblinded" to authorized investigators to look into indications of fraud or tabulation errors. I suggest that it may be the state's attorney general, or some nonpartisan entity that the state may designate.
The sole reason for the ID number is that numbers are unique identifiers that cannot be duplicated, and it's required for the computer tracking process described later. Voters must register as usual. They will receive their voter registration card in the mail—sent to the address they claimed as theirs. The card itself will contain only 3 pieces of the total information: The person's photograph, his registration number, and his election district. It will also contain a magnetic strip of the registration number, so that the voter can "swipe it" while in the voting booth, for ease and speed.
That's right, voting will be a little like a transaction at an ATM booth. People will go to central locations to cast their votes, just like they do now. Casting your ballot will also be done in a private booth, just like it's done now. The voting machine would essentially be a computer terminal. The user interface will employ touch-screen activation, with the option for mouse input and voice recognition (for the visually-impaired).
Voters will either swipe their registration number into the computer, or enter it manually. The method of entering your votes might be through screen activation, or whichever method Congress deems best. After the voter reviews and confirms the entries on his ballot, he must enter his PIN (personal identification number). This PIN is just like a PIN you use on an ATM machine. It's not known to anyone but the voter. Without the correct PIN for that registration number, the votes entered into the computer will not be recorded, transmitted or tabulated. It won't go anywhere.
The process to ensure proper identification is necessary to stem voter fraud, and as I'll describe later, ballot manipulation. The PIN number prevents anyone else from voting using that voter registration number. The computer will store votes cast as "write-once" (then read anytime) data, to be sure. But the unique PIN adds an extra layer of assurance that votes cannot be altered (for any given registration number) or multiplied after they're cast. Only one completed ballot per registration number will be allowed. The photo of the voter on the card will prevent various forms of fraud as well, but at the same time represent a far less intrusive measure than say, having your fingerprints registered, as is required to vote in Mexican elections. Indeed, voter ID cards with photo, thumbprint, voter ID numbers, and magnetic strips are commonplace in the very same countries that the left urges us to emulate with regards to paper ballots.
The Computer Hardware & Software
Algorithms necessary for tallying votes are not complex—no more so than those which tally votes from inputs via paper scanners. So the transparency issue just with regards to understanding the program software isn't a daunting problem. Rather, the concern among some is that the use of computers tends to involve greater time and distance between the vote when it's cast, and the (usually) centralized location that tabulates that vote. The greater that distance, the greater are the opportunities to defraud. The idea that fuels this attitude is that fraud is inevitable and can't be prevented. And when it occurs, we should have a system in place that ensures that the effects will be limited just to the machine or voting precinct where the perpetrators operated. In other words, the preference for paper ballots seems to be that the inevitable cheating will be decentralized—limited just to the location where the perpetrators have access to ballots.
This appeasement to unlawful activities which violates the very basis of our democracy is unacceptable. Nevertheless, I describe below and in the next section ("The Fail-safe Feature") how the mechanism of voting and its redundancy features under my system deals with the centralization issue.
The system software and voting program will be open source-code written by a federal agency, commission or single government contractor. Congress will mandate the required features, capabilities and requirements that must be written into the software—many of which are difficult or impossible to implement when using paper ballots. For example, you'd be able to select the language of your choice. It must display the text in user-selectable variable font sizes for the visually-impaired. As an option for the hearing-impaired, it must employ an audio selection system—also with a language-selectable option.
Each state will be responsible for supplying the computer data which would consist of all the federal and local races, referendums, etc. The software will display solely those candidates up for election in each specific election district. Computers allow for last minute changes to be made.
Private companies will manufacture the computer hardware, which will be in exact accord with narrow specifications (screen size, audio specs, user-interface, recording device for output data, etc.) authorized by a Congressionally designated authority, such as GAO or OTA. This authority will also establish a uniform criteria for state election workers to test and maintain the machines. Neither the manufacturer nor election officials would be able to tamper with the software because the software would not be loaded into the computer manually or on-site. Instead, each terminal will be connected to a closed network hub at the polling place, which in turn will be connected via broadband connection to mirrored webservers run by the federal government (for the system and program software) and by the state (for the ballot data).
About Data Security
The system I'm proposing will require security measures on the same level as ATM machines, or better. Data encryption will be required, and not optional.
The network hubs will employ hardware IDs, similar to network "dongles" that many businesses use for security. This ensures that each polling place is legitimate, and will be authorized to receive the software and ballot data for each location.
Since the system software and voting program are downloaded into RAM, the computer hardware manufacturers will not install firmware in ROM that could contain suspicious or hackable code. In a couple of years, perhaps even the downloading of any and all code would be unnecessary, when the technology for web-based software becomes sufficiently capable. (web-based software is essentially software that can run on a computer without having to install on the computer itself, the code which runs the program.)
For hackers to successfully tamper with software, they must have two things: an exact copy of the code, and sufficient time to analyze what each line of code does. Hackers are denied both under this system. Even if they get past the other security measures, the state and federal governments would have 12 months between elections to modify the source code ever so slightly to stymie would-be hackers.
Besides, as you'll see, the system I'm proposing denies the opportunity for people at the local level to engage in vote manipulation. Most of the opportunity for fraud defaults to more centralized and upper level points. This reduces the pool of suspects—and together with stiffer penalties for vote fraud—would discourage even the most partisan from breaking the law. Yet the one vote per registrant limit renders hacking a daunting task on any level of system access. One would have to create entire identities, using residential addresses that will either be detected as nonexistent, or else a duplicate of another voter's.
In short, successful vote tampering would be virtually impossible. When the polls open, the precinct worker will flip the "on" switch on the computers. The operating system and voting software will be download from the federal webservers. Then the ballot data will be downloaded from each state's web servers. Before the polling places close, the precinct worker will just have to remember to turn the computers off and shut the lights before they leave.
The Fail-Safe Feature
The part of the plan that ensures error-free and tamper-free results involves a process that's best achieved through computer voting. It also employs the features that those on the Left seem insistent upon having: a confirmation system fed back to the voter, as opposed to a central monitor. Because according to the caviling as I understand it, no one can be trusted to catch fraud or errors, better than the voter himself. (Having partisans of equal strength present to monitor proceedings is an ideal that's not possible to produce in many counties and precincts.)
After the voter completes his ballot selections, the screen displays a summary screen of how you voted. If you made errors, you can go back and correct them. When you finally approve your ballot and enter your PIN, the system records your vote simultaneously to 3 separate entities, and all while you're still inside the voting booth:
First, the ballot you completed is transmitted directly to a central location where state election officials electronically tabulate the votes. All votes are tagged with the voter's registration number, and no votes can be tabulated without them. The registration number remains with the vote even after tabulation, but it's encrypted and therefore indiscernible to people at both ends of this transmission (i.e. both locations). This whole process is computerized anyway, with the electronic vote data remaining untouched and unseen by humans. The transmission and encryption encoding reinforces this blindness further, and would be similar to the security involved in ATM transactions.
Since voter registration numbers are still digitally linked to the votes cast at this stage, the vote data for that number can be printed and mailed to the registered voter in the same manner that people receive their bank statement. Along with the information on how you voted would be printed your registration number and address information, all on water-marked paper. It would be very hard to mass produce such printed materials to defraud the government and sway the election. Besides, what person would present a counterfeit or doctored printout when his name is on the printout, and would implicate him in the fraud?
But let's not get ahead of ourselves. All we've done thus far is blind any election officials to the identity of the voter, and speed up the transfer of the ballot from the voter to the final vote tabulator before unlawful ballot manipulation can occur locally, because we cut out the "middle man". But we're not done yet.
The second place to which your vote is instantaneously recorded is to a "write-once" (then read anytime) media installed and physically locked inside the computer terminal. The media might be an optical or laser disk, or perhaps a memory card. This data is likewise tagged with the voter's encrypted registration number. After the polling site closes, precinct workers manually remove the sealed box holding the media, which is then delivered to a second separate state agency, be it the Secretary or State, or Attorney General, or whichever entity the state so designates. If the seals are broken, there will be an investigation.
Regardless of which two state entities are the recipients of the raw vote data, federal law will mandate that they be separate entities independent of each other, with technical experts from all political parties present to monitor the entire proceedings. Each would be equipped with computers and software (software via the federal government) necessary to process the votes. The vote counts from each agency would have to match. If they don't, then the election cannot be certified until an investigation discovers why the tallies don't match. The third place where the vote is recorded may help uncover the reason for possible discrepancies.
The third "place" where the vote is recorded is to each individual voter: Before the voter leaves the voting booth, he takes with him a paper printout of his completed ballot. The voter is advised to review the printout before leaving the polling place. If there's an error, then the complaint must be registered right then and there. If the printout is correct, the voter is advised to retain this printout.
This printout represents the most accurate record of that voter's intent, because it's not printed out in the same way as the aforementioned ballot summary that state election officials print it and mail to the voter. As I explained above, the first state entity that obtains a record of your vote immediately tabulates it, then spits out the text data (comprising your ballot choices) onto a water-marked paper, formatted to appear like the ballot summary that was displayed on the computer screen that you approved when you were done voting. In other words, it's an indirect representation of your ballot in the same manner that a voice recording is transcribed to paper.
What is different about the printout that the voter obtains directly from the computer (while in the voting booth) is that this printout is as close to an optical photograph of the screen as you can get. In computer jargon, it's a dump of the screen buffer (i.e.: whatever is being displayed on the screen) directly to an output device, such as the printer port. "Screen captures" or "snapshots" is an old feature of computers. It can be triggered via hardware or software. When I owned an Apple II in 1980, hackers instructed users on how to solder a switch to the motherboard that would trigger a direct dump of the screen buffer. Hard-wired in such a manner, it would be difficult to shield from view any unlawful tampering. The bottom line is that the printout that the voter takes from the booth is the closest representation of the ballot that he saw on the screen and approved.
Assuming the printout accurately reflects the way the person voted, what happens next is the point in the process that will satisfy the wingnuts on the far left who think conservatives have been rigging the elections (when conservative win, that is):
When you get that printout of your vote in the mail, you can compare it to the printout you took home with you on election day. Alas, that's the final check in the process. The "snapshot" of the summary page (or pages) that indicated how you voted must match the page (or pages) printed out by the state agency at the point in time when that agency received and tabulated the votes. If you fail to get it in the mail, you can go to the Board of Elections to get it.
If the printouts don't match, you can place the ballots in a furnished post-paid pre-addressed envelop to the state attorney general, or some other designated nonpartisan or bipartisan committee responsible for investigating vote fraud. I would suggest people first make photocopies and perhaps send copies to their political parties.
In the final analysis, the success of this plan is based on the various hardware and data security measures, along with the decentralized two-tier parity checks: One between separate agencies, and the other between agency and voter.
Can This Plan Work Politically?
Implementation of this plan is technologically feasible and affordable. However, it breaks with some traditions, diminishes state authority to a degree, and would require Congress to amend federal election law to permit greater oversight and involvement by federal entities in state-run elections.
The part of this plan that makes it effective is a revolutionary idea, at least in American elections: For the first time, voter anonymity AFTER the vote is cast will be partially removed. Partially; not totally. The ballot the voter completes and is sent to the three locations, remains electronically tagged with the voter's registration number, even after the votes are tabulated (though not indefinitely thereafter). Computer encryption can be used to obscure a voter's name from his registration number. At most, all that a casual viewer might see is an encrypted registration number on a recorded ballot.
After an election is certified free of errors and fraud, there wouldn't be any need to retain that part of the digitized information that holds the personal identifiers. The software would permit state officials to easily strip off the names, street addresses, photographs, and voter registration numbers from the part that holds the election results. Once this information is purged from the data, journalists, historians, pollsters and political parties can then be given access to it to perform the usual statistical analysis of voting patterns measured against party affiliation and zipcode area.
Thus, the chances of a breach in the privacy of ones' vote is small, and far outweighed by the advantage in restoring faith and integrity to our elections.
Remember, the sole purpose of temporarily linking the vote to the voter is to (1) prevent voter registration fraud, (2) prevent errors or tampering of the votes just cast, and (3) catch possible errors and tampering of the tabulated votes later on.
At no point will local election workers handle ballots. Ballots will not be stolen or mysteriously lost or damaged. Stacks of completed ballots will not mysteriously materialize days after an election. Indeed, all forms of ballot tampering will be impossible at that level. At the very top level in which tampering might conceivably be attempted, it's discouraged by a daunting array of redundancy checks and features outside the control of one person or agency.
Voter error will be a thing of the past. Overvoting (selecting more than one candidate in a category) will be impossible. Undervoting—through an oversight—will also be impossible: The voter that accidentally misses a category will be alerted by the user friendly, multilingual interface which could also adequately alert the hearing- or visually-impaired voters.
Additional controls can be built into the process that would be impossible if paper ballots were used. Computers can do parity checks all along the way—and in real time. Attempting to cast votes from duplicate registration numbers will be blocked at the input stage, or will certainly be detected by later surveillance of the data. Nonexistent or duplicate addresses will be caught in the registration process, but also can be electronically searched for after the vote to investigate voting abnormalities. Party affiliations that don't match the vote, or cross-ballot voting patterns which drastically depart from prior election patterns or the current norm will trigger red flags before the day ends. None of this is possible under any other system. Certainly not with paper ballots.
Conclusion
Since the federal government will have a greater role in state elections under my plan, Congress will have to enact amendments to address the Constitutional issues. But a computer balloting system holds unparalleled advantages. For the individual voter, the technology will streamline the process on election day. There will be no justification for minorities to wait on long lines in the rain in order to vote. Handicapped individuals will be able to vote with dignity. The current worker-intensive system that requires large numbers of volunteers to police it will be history. Beyond the initial investment—which will be mitigated by the uniform standards imposed by the Feds—holding elections will be less expensive and there will be little of no voter fraud, vote manipulation, and tabulation errors.
The only thing wrong with computer ballots (or DREs) today is that the standards they must comply with vary greatly between counties and states. The hardware and software systems need to be more uniform, or preferably (as my proposal recommends), virtually identical. Voting machines must operate without errors. Yet bugs are likely to occur anytime you need to get a new program to run in a new or adapted operating system, which is running off a computer who's user interface is handled by thousands of voters who are complete computer illiterates. It's hard enough for a behemoth like Microsoft, let alone small computer companies. My ballot proposal mandates uniformity across states and counties. Private enterprise conservatives will just have to realize that when it comes to elections, there's no time to wait until for a corporate shakeout of poorly performing companies. National elections are too important for that.
Politically, it would be an uphill battle to implement my system. State and local election officials want to control "their" turf.
Democrats (in the majority) in large cities will suddenly become stalwart states rights' advocates. Their comrades on the far left have already become complete Luddites on this issue. Since 2001, they've become obsessed over alleged voting fraud. They've donned their tin-foil hats and won't listen to any proposals that involve computers. They see Republican conspiracies and black helicopters everywhere. (They actually believe Senator Paul Wellstone was assassinated by President Bush.) Their only solution is to use paper ballots.
But a return to the Middle Ages isn't a solution. Today, we place our very lives in the hands of computers. Computers are not only used to design planes, trains, automobiles, office towers, and nuclear power plants, they operate them as well. Commerce and banking are all computerized. System security, redundancy checks and backup provisions assure near perfect performance and reliability. When computers fail, others take over. Computers have improved our way of life, and they're the only means to resolve the problems which—commissions that have studied the problem have found—are inherent with paper-based ballots.
So it would be foolish to allow left-wing paranoia and distrust of Republicans in power to sway us from using a rational system that holds a cornucopia of advantages. As a side benefit, under my system, left-wing political rants, boorish allegations and attempts to undermine the legitimacy of Republican office holders would come to a crashing end.
(Printer friendly version) Email: Gary Krasner
