|
|
|
World Future Society The Dream System
This article makes a case that a dream system can be and ought to be created. We begin with the basic network for voting and vote continue. The Technological Base Two million PCs with capabilities far in excess of what is needed can be purchased today, at retail, for less than $1.6 billion. An agency buying in such quantity should pay less than half. Very sophisticated system software costs no more than a few tens of millions of dollars. Special software is needed but would not add significantly to the ballpark estimate. One printer for every four or so PCs adds less than 10% to the cost. Ninety-nine percent of the PCs would be in the precinct voting places. Beyond standard election costs – space rental, supplies, maintenance, staff and official pay -- there are no further costs. The total cost of the proposed voting system should be around a billion dollars, much less than many would expect. At the close of any election, all votes need to be counted. An average of ten PCs and two or three printers in a precinct need to be networked to produce a precinct tally record. The precinct vote counts pass up the voting chain to about 1% of the PCs distributed mostly at the municipal level, some at the county level, and 50 at the state level. The links of the voting chain are the precinct, municipal, county, state, and national levels. Some large cities have sub-municipal levels. House districts cut across county lines and are totaled at the state level. Media will obtain outcomes at all levels from municipal on up. National media will want tallies as soon as the polls are closed at the municipal level if that is earlier than the state poll closing and at the state level as soon as all polls in the state close. In the three evening hours or so, as polls close from east to west, state by state, the media will report on the congressional and gubernatorial results and announce the running tallies both electoral and popular votes for president and vice-president. Exit polls, which contributed to the 2000 fiasco, will be irrelevant. Counting votes is simple addition for each candidate aggregated at most five times as the vote counts go up the voting chain. This can be done accurately and reproducibly with today's technology in a few minutes, each uplink taking as little as a fraction of a second. Each PC has only a few kilobytes of data that needs to be passed up the voting chain. After the last precinct in the nation closes, rapid tallying can be done for the final count or for a recount with complete accuracy. These processes are similar to the way that a large financial company with hundreds of branches can determine and confirm overnight cash balances accurate to the penny. Communication channels are required for at most only a few minutes at the close of the election and they need carry only a tiny bandwidth. There is no need to use the Internet, avoiding exposure to bugs, denial of service, encryption/ decryption and viruses. You can bet that thousands of hackers would lick their chops for a shot at what would be a biennial jackpot, corrupting a US national election through the Internet. Yes there are special processes required for properly handling absentee ballots, write-in votes, and dubious registration categories (ex-felons, address changes, etc.). These are discussed in the sequel. One basic idea is that no voter need be turned away from a voting place because s/he appears not to be registered. Many could vote in a category whose tallies are separately and coherently handled as they are carried up the voting chain. Category vote legitimacy need not be resolved if the counts are too small to affect the outcome. Two major components of the best-practices system which help handle registration problems are the resolution of the "too-close-to-call" situation at the state level and a special process for remedying what is called the "sanctity of the vote", that makes fraud by any voting officials from the precinct level up to and including top state officials, very unlikely and rather easily spotted. Channel Capacity Utilization Technology in a larger context So while today we could put in place a physical system for about a billion dollars, it should be even much less in the future. Would this country, the world's most technologically advanced nation and arguably the most advanced democracy, not look stupid if it did not try to resolve the tough non-high-tech practices and procedures needed to produce the dream system? How can we turn our backs on this dream? There is an intriguing historical context to this reluctance to pursue such a promising system. There was a time when the rising capability of new technology inspired a confident and growing country to undertake a series of highly successful projects like the Erie Canal in the early 1800s, Boulder Dam in the 1930s, and space exploration culminating in the moon landings in the 1960s. Information technology (IT) offers that potential today but so far seems focused on a single approach. The unique and evolving Internet is a complex, far-reaching development that will profoundly affect the lives of billions of people. It is too vast to call a "project" and too diverse in purpose to call a "system". But the product capabilities of IT are enormous and have led me to explore how to make a voting and vote counting system into a complete, best practices system for national elections, that could rightly be called a "dream system". Let us think about such a system not with the point of view that many futurists usually adopt, what will be or might be, but with the viewpoint of what ought to be. For this thought I commend Dee Hock (author of "The Birth of the Chaordic Age" and CEO Emeritus of VISA, a financial giant with $2 trillion annual revenue) who believes all organizations must continually re-engineer themselves from this point of view. Too-Close-to-Call The Electoral College Gridlock. The "Close-Election" Problem Resolution. There are a few technicalities that are readily handled that add a little complexity to this kind of resolution of the close-election problem. First it is theoretically possible, and so someday it may happen, that there are three or more candidates with vote counts within the break-even range. Also there is the possibility of a dead heat in the electoral college count. A further complexity is that if the state with the close-election had an odd number of electoral votes, an even split of its electoral votes would produce ½ votes. A resolution of these details could be drafted along the following lines. When the state with a close election has "broken even" (the two highest ranking candidates differ by less than the break even number) and has an odd number of electoral votes, the candidate who scored the highest in the count or final recount, if there are recounts, would get the extra ½ vote and the second highest candidate would loose ½ vote. The candidate with the highest scoring vote count would thus receive one more electoral vote than the second highest. If this produced a total electoral vote count that was a tie, then the break-even range would be enlarged to apply to the state that had the next largest break-even range. Similar extensions could be applied to handle the outcome in the case of a three-way "tie" or more. If the electoral college were to be upgraded and treated seriously, resolution and possible revision of these details for subsequent elections (never ex post facto) would give its members something potentially useful to do rather than, as in the past, utilizing people to do mindless tasks. The proposed resolution of the close election problem is desirable and equitable. The idea was mentioned and rapidly dismissed during the long Florida count, because it was tantamount to giving Gore the victory. Looking to the future, the proposed resolution would treat both parties fairly. Would the republicans oppose this idea because it would have meant a Gore victory if it were in place in 2000? A lesson of history going back to the Truman era suggests that they should not. Pained by five consecutive terms of democratic rule, republicans pushed through a two-term limit for presidents. For the next fifty years the only presidents who were prevented from a third term were two republicans, Reagan and Eisenhower, and both were so popular they might well have won again. Will the republicans turn down this idea for handling the close election problem and again make a mistake that will haunt them for fifty years? Practical politics with its perennial short-range view, suggests that they well might. Let us consider further ways to improve the election process. Potential Conflicts of Voting Place Activities. The Florida 2000 national election made it painfully clear that there are weaknesses and errors in both functions undermining the basis of our democracy. Major reforms are needed -- as is widely understood by state Secretaries of State, Election Commissioners and their staffs at the state and local level. There is a potential conflict among those workers, including both officials and volunteers, who staff the voting place on election day. The conflict comes from two facts: (1) while the workers have various official duties, they generally also all want to be helpful in explaining and guiding voters through the process. At the precinct level, many of the voters may know many of the workers and each other, so that the community feeling that everyone is working to help everyone else may be pervasive, (2) at the same time, the privacy of each person's voting intentions and filled out ballots must be respected. The two functions A and B at the human level are sometimes at odds. This problem is not new and has been understood and pretty well handled in the past. Voting places are laid out with privacy booths, ballot boxes, registration table, all somewhat separated from each other so that the workers handle their official functions, guide voters and respect voter privacy most of the time very well. There are relatively few errors. But that is the problem. This is just not good enough in very close elections, where one error in ten thousand can swing the election from one candidate to another. Reducing the Conflict. Fair, Accurate, Verifiable Registration. This can be accomplished by a special voice or data system controlled at the county or state level, which should conform to, respectively, state or national standards, backed by criminal statutes. These systems need not be very high-tech themselves. The important idea is that there be a central location at either the county or state level, through which all matters pertaining to registration could be cleared. It could be a telephone, voice based system using toll-free numbers to reach the central office staffed by an election official. As necessary, it could become a fully automated network system. If the card-holder challenges a removal, but is not given satisfaction, the special voice or data system automatically gives the card-holder a confirmation number which is a serial number related to the time and date of contact and the contact address (phone number, e-mail name, or whatever the special system uses for public access). All card-holders denied registration use a DR (Denied Registration) code, and along with the confirmation number or other contact address, then, on an "emergency, tentative access basis" must be allowed to vote at the precinct confirmed by the address on the card-holders card. The system is protected against fraud committed by the card-holder through the following measures. All such DR voters' ballots have a special DR designation added to the ballots they cast. When the votes are counted, the DR votes are kept segregated and are considered contested and tentative. If the total number of DR votes at the level where election contests are decided (three cases: city, county, and state), is too small to affect the election outcome, no resolution of the validity of these DR votes needs to be made. They are irrelevant to the outcome. This procedure will take care of the main problem of election officials at the state or county level improperly or illegally removing large numbers of people from registration rolls. Today when such behavior takes place, there are no audit trails whatsoever. The impropriety is unlikely to be discovered and if discovered it is difficult to prove such removals illegal and virtually impossible to do so in a timely way. After the polls close, officials can get away with a reasonable explanation or claim an excusable mistake. Once this special system has been adopted, officials will no longer care to risk exposure to what could result in a conviction for a felony that ruins their careers. Such risky behavior, likely to be significant only if thousands of names have been removed, would no longer be attempted when there would then be some fraction of those thousands with individual audit trails any one of which might establish that fraud was committed. Measures to Separate Voting Place Activities on Election Day. With chit in hand the voter passes through a door to a second room where voting takes place. At the door a clerk takes the chit and keeps count that s/he is getting 20 of one kind before the next kind comes in or, if not, reconciles the difference. The differences are quite easy to reconcile. For example, if one voter with a chit, decides not to vote at all and leaves the voting place, then only 19 ballots can be expected in that group. Other election day workers may be present in the voting room to assist in the use of the voting machines, but must not be close to the voter unless the voter asks for help. If there is overload – a need for a voter to wait because no voting machine is unoccupied -- then there should be some area, either just inside or just before entering the voting room (but not both) where voters can wait to vote. The single waiting area can eliminate possible confusion of who goes with what group marker. These low-tech procedures can be modified to some degree provided that one purpose is always met. The registration workers who know the voter by name, address, party affiliation, group code, or personal recognition, never know the sequential number assigned by the PC which will produce the voter's filled-out paper ballot, unless the voter chooses to divulge it at some point. The peer-to-peer network of precinct computers give sequential numbers to all voters in the precinct, skipping no possibilities, but not necessarily assigning them exactly sequentially, so that no one in the voting room or the registration room can know the number assigned to a particular voter whom they otherwise may know. The Voting Process. Throughout the voting process, the voter can switch from an item by item ballot display to a display showing the full ballot with the partially complete entries. From either display, the voter can change or correct any previous entry. When satisfied with all entries, the voter requests a CAST YOUR COMPLETED BALLOT button. After being prompted
The voter presses the key or touches or clicks on a button to respond. If the response is "yes", a screen display appears or lights-up, saying perhaps
and the voter confirms or cancels. The PC composes and stores the final ballot and prints one copy of the ballot for the voter who may choose to treat it as private or not. Either choice is acceptable and must be accommodated The PC has a coded number, discussed later, that allows no one to identify from any information stored in the PC or printed the individual voter's identity. The only one who knows is the voter him/herself. The computer keeps no running totals except on the number of ballots it has produced since the start of election day. The ballot has a stub that plays an important role in the process, described later. Absentee Ballots. Valid absentee and out-of-country ballots received in good order by election officials for each voting place can be entered, with a special group designation, so that the total votes for each candidate in this absentee and out-of-country group will be available along with grand totals almost simultaneously with the vote counting process. Every effort must be made to keep the percentage of absentee ballots down and to rule out absentee ballots that arrive after election day. Absentee ballots are paper ballots of fixed format that some PCs must be programmed to read by a component of systems software. After reading an absentee ballot and logging in its votes, the PC assigns it a serial number that includes an absentee code suffix. All absentee ballot votes become part of the precinct tally record. Some absentee voters may not be associated with a precinct and can vote only at the county or state level and only for county, state and national candidates. This should be discouraged but in some cases may be allowed. The absentee voting process should take place, as all regular voting does, on election day. Absentee ballots received after election day should be invalid and discarded. Tracking and Counting Ballots
The voting place is identified by code (d). The voting place print-out of all of its voting data, is called the precinct tally record. The first column of this record lists the up-to six characters, (e) and (f) above, in numerical order, identifying by number the individual voter so that the voter him/herself can find his/her cast votes – which no one else can identify. Each voter who cares to do so can confirm the correctness of the record with the voter's own PC produced ballot received when s/he completed voting. The precinct tally record could be a single large sheet. If a precinct had accommodated 200 voters and its ballots had a total of 50 candidates, the print-out would be a matrix of 200 rows and 50 columns, where all entries in the matrix would be checkmarks and, very important, all columns would be footed so that the number of votes received by each candidate at that voting place would be along the bottom line, simply the count of all checkmarks in the column. Tally records at the city, county, state and national levels would drop all information on individual ballots, showing only the total votes for each candidate when aggregated from the candidate's total votes at the next lower level. In the example above, at all levels above the voting place level, the matrix would have a single row and up to 50 or so columns. Sanctity of the Vote The precinct tally record
In the matrix of rows of voters and columns of candidates, entries are only (a) check-marks when the voter voted for that candidate and (b) blanks when s/he did not. All columns would be footed so that the number of votes received by each candidate at that precinct would appear along the bottom line, and each number is simply the count of all checkmarks in the column above it. With a copy of the precinct tally record any voter, using his/her serial number printed on the paper copy of the ballot received at the time of voting, can quickly find her/his row and confirm that the ballot choices agree with the checkmarks. The record has been set up to make it as simple as possible for any individual to confirm its accuracy with respect to that individual's ballot. Though few voters can be expected to go to this trouble, a few is enough to have a profound effect on the honesty of the election, as we will see. If, in fact, all 200,000 precinct tally records were available in one large office after all polls were closed, a battery of clerks with adding machines could, in principle, determine the exact vote for every candidate for every office at every level. It would likely be fraught with errors, a very bad way to do vote counts, Fortunately it is rendered unnecessary by a very accurate, quick and good way – a high-speed network linking the PC voting chain, as described previously. Write-ins Voter Empowerment to Confirm Recorded Ballot Choices Consequence of Voter Empowerment Eliminating Errors Along the Voting Chain Curing Corruption Vulnerability of PC Based Systems Increasing Voter Turn-out Acceptable include paid or volunteer get-out-the-vote drives providing rides to polls. A rough cash-value cost estimate of this activity is in the range of a dollar or two per vote. (Assumptions: volunteer's time is valued at minimum wage and full cost of vehicle per mile is about 50cents for a car with five people and a dollar for a bus with 30 people). Also acceptable is the cost of a presidential campaign, which cynically can be considered as vote buying by advertising. A major presidential candidate's campaign expenses are now just over a dollar per voter. IRS requests for voluntary presidential campaign financing have been set at a dollar on the voters tax return form. Another acceptable is precinct volunteers or officials providing a lapel clip, saying "I voted" (or similar) to every voter at no charge. This typically has a cash value of well under a dollar. Unacceptable include paying vagrants (or those who otherwise would probably not vote) to vote as directed. Cost can be a few dollars for drinks or a few cigarettes, or if cash, probably about ten dollars per vote would be needed for the voters rounded up and for the hustler who makes it happen. Such activities are associated with old Tammany Hall in NYC and the Chicago voting "machines" at the turn of the 19th century. Separated on a price/value basis, it is interesting to note that methods that cost about a dollar per vote are acceptable but ten dollars a vote are not. What drives the turn-out of 46 to 52 percent of adults who have voted in recent national elections? We do live in one of the most, if not the most, established democracy in the world and voting is an important citizen activity to maintain a democracy. Most people find that in the long interval after public school and before receiving social security checks, the significant contacts they have with government are tax payments and, in the past, conscription for war. Not very pleasant. The unpleasantness becomes especially irritating because the affected individual has little control. People often perceive this as lost freedom. Voting every year or two, or voting for president very four years, are the only actions most people can pursue to try to gain control over the forces that so frequently threaten and often do curtail their freedom. The reaction can be characterized by "throw the rascals out." Never mind that few rascals (ie, incumbents), especially at the national level, ever get thrown out. Today's successful politicians only have to convince 51% that they are not the rascals. Their ability to convince people that they are good guys has come to be their chief talent. Opinion surveys show a large percentage of voters vote for reasons like: "It's part of the American tradition that made this country great". "My parents voted religiously", "It's my civic duty", etc. People who think like this are not looking to be paid. They are going to vote anyway. In fact, the worse the country seems to be doing, the more likely they would vote. Many feel that our democracy does or says little to make them feel that the country really cares whether they vote. Candidates care, mainly for personal reasons. People would appreciate some official, meaningful, personal recognition for the act of voting. I left public high school at 16. In the intervening 49 years, until social security payments started coming in, I felt the country paid no attention to me as an individual beyond two activities, collecting taxes and conscription during WWII. Both these activities applied to large numbers en masse and are impersonal. This changed in a remarkable way ten years ago, the day I went to register at a Social Security office. An agent put my record on a monitor, showing my income by year over a forty-year period, a display new to me. I was enchanted. My life's work was condensed onto one screen. My own records were far less complete. I recognized some of my best income years and some not so great. For the first time in my adult life I felt that my country did care about me in a benevolent way. After 40 years of hard work, I would be getting monthly benefits. I was so moved, I almost cried. How can the country give us personal recognition for voting? Envision the next presidential election this way. At the polling place you have received a ballot with a high quality, adhesive-backed, peel-away-protected stub (like some postage stamp rolls). The stub appears valuable. You save it and so does your spouse. Let's say you are married and you file taxe Estimate of Costs and Turn-outs for Presidential Elections Case 1. Value of a ballot stub set at $1 and a lottery with $10million prize money: (1) 140 million vote, (2) 130 million aim to put the stub on their tax return, (2) 90 million of those are organized enough so that when they are filling out the return, they have the stub available or can find it. Total cost to US Treasury: $100 million, $90 million reduction in total revenue and $10 million in lottery prizes. But people are so excited about possibly winning the lottery, not to mention feeling virtuous for being good voters, that they generously report on average about $20 more of income leading on average to an additional $3 of taxes, or a total of $$$270million more of revenue. IRS has a net GAIN of $170 million. I ignore the increase in numbers filing due to the lure of the lottery above those who would have filed if there were no lottery, likely to be a significant number but difficult to estimate. Base case: no lottery, no tax deduction for affixing stub: (1) 100 million vote (average in recent presidential elections). This estimate shows a 40% increase in turn-out from 50% to 70% plus a gain to the IRS of $170 million over the base case. Case 2. No lottery but still $1 per stub return: Estimates of previous paragraph become (1) 120 million vote, (2) 80 million are taxpayers and aim to put the stub on their tax return, (3) 20 million actually do so. Cost to the IRS, $20 million and a 20% increase in turn-out from 50% to 60% Case 3. No lottery but now $10 per stub return: Estimates now become: (1) 130 million vote, (2) 90 million are taxpayers who aim to put the stub on their return, (3) 70 million actually do so. Cost to the IRS, $700 million with a 30% increase in turn-out rising from 50% to 65% The recommended case (1) not only puts $170 million into the US Treasury, it produces 40 million more voters, an increase of 40% over the base case. The US is transformed from a country with low election turn-out to one of the highest among democracies. Tax Credits at the State and Local Level Further Benefits To satisfy all the dream system requirements the system needs to operate remarkably only a few days a year. It can be available for other revenue producing activities over 95% of the time. It could be rented by large organizations that occasionally need to take quick, accurate votes, such as unions and corporate shareholder meetings. It could be used for very accurate opinion surveys. These revenues might cover much of the costs of handling elections. If some of the 200,000 precinct voting places were suitable as theaters or auditoriums for entertainment or educational functions, another possibility opens up: broadband transmission that goes to those theaters or auditoriums and does not have to go "the last mile" to each home. People go to the theater or the auditorium instead, and someday perhaps to gaming rooms or sporting events operating with systems based on the dream system approach. A built-from-scratch system for a much smaller number of locations, such as 2000, not necessarily all in the US, with a chain of theaters, auditoriums, and/or athletic stadiums would cost about $30 million, excluding the cost of the buildings themselves. Revenues downstream might well amortize such an investment. The dream system is essentially unconnected to the Internet and so can avoid the hackers, sociopaths, and criminals whose operations in every corner of the world cost legitimate users of the Internet over a hundred billion dollars a year, a hundred times the cost of the dream system. I would enjoy talking to others who are intrigued by the potential of a dream system and are interested in commenting or discussing any aspect of it. My preferred work mode with other people is collaborative, not competitive. As the nation decides election reforms, isn't this "dream election system" worth a closer look? APPENDIX – Excerpt from my article in World Times, Jan/Feb 2001, p. 2. Vulnerability of Electronic Voting Machines Without Dream System "But hold on. After showing these cool features to the election commissioner or to a party official or party fund-raiser, a salesman might add, 'Now on the back here is a little inconspicuous switch, which if you push to the left, regardless of what this machine tells the voter, will count the vote for the democrat. If you push it to the right, the republican gets the vote. Of course, we oppose the use of this switch. I know you’d never touch it. You’re ethical, but there are some buyers out there who consider that feature a plus and it might be a factor in their purchase order decisions. If my company X didn’t have this feature, our unethical competitors might get the order. What you do is up to you. We don’t want to know about it.' "Then his competitor might whisper, 'X’s feature is OK but my company Y has something better. Election commissions need to be allowed to make distinctions between candidates for different offices, not just for president. An on-off toggle is crude. It’s easily spotted. Someone might wind up in jail. My machine has more than a toggle. It’s controlled by a statistically biased random variable. It works like this. Suppose you decide your candidate in one contest needs a five-point advantage, another one needs only a three point advantage. The machine introduces random changes that favorably bias the tallies by the amounts you have instructed. No one can ever find out that any vote was changed. The machine eliminates all traces.' "Now we don’t have to pick on the poor salesmen trying to make a living in a tough game. Similar offers might be mentioned to the programmer who loads the candidate data, or by the engineer who trains the programmers. Or it could be accomplished by a party hack or a foreign agent who hands an envelope to one of these people. "Even if all of these shenanigans were outlawed, severely punished, and never took place, another weakness of the electronic voting machine is this. If a recount is required because of a close vote, then printed copies of each ballot is desirable, both by the election commission and each voter, so that both can confirm that the voter’s intentions were followed. If machines failed because of power failure or whatever, this would be absolutely necessary." About the Author
Send comments to: webmaster@wfs.org |
