At the time of the founding of this nation, some voting was done with paper ballots and some was accomplished orally. Ballots are a type of artifact. In ancient times, artifacts used for voting might have been pebbles, beans, small balls, or pieces of pottery or natural shell. (The word "ballot" comes from the Italian for "small ball.") In the area that became the New England states, paper ballots were generally used from the very beginning of voting. In Virginia, voting was done orally before a clerk who would ask the voter his choices and write down the responses. Thus, originally, the US had both artifact and non-artifact voting systems.
Obviously, oral voting was not secret, but a positive attribute was that the results were indisputable. By the time of the Civil War, most states that had previously employed that technique had converted to voting by ballot. Often, a stated reason for elimination of oral voting was intimidation by creditors, landlords, and employers threatening respectively their debtors, tenants and employees. A second reason for the change was that counting ballots was a less time-consuming process when the turnout was large. A few states were holdouts, and in 1871, the federal government adopted a law requiring voting by ballot for members of the House of Representatives. The Congress has authority under the Constitution to dictate the "manner" of voting for federal elections but not for state elections. Nevertheless, the law made no mention of voting for Presidential Electors; US Senators were still being named by state legislatures at the time. The last state holding out against voting by ballot was Kentucky, which did not fully adopt it until 1891.
In the 19th century, up to its last decade, most ballot voting was not secret, even though secrecy was often a stated reason for replacement of oral voting. Ballots were printed and distributed by the political parties, and were of different sizes and colors. (A state law requiring that ballots be "white" was easily circumvented because different shades of white could be used.) Observers at polling stations could identify which party's ballot a voter inserted into a ballot box. Voters could vote a split ticket; small pieces of paper with candidate names on them (called "pasters"), as well as paste pots, were sometimes made available at polling stations. The pasters could be glued over printed names on the ballot. Thus, the activity of casting a split ticket took extra effort.
A second reaction was against non-secret voting. As the economy became industrialized, there began to be masses of workers who were intimidated by their employers. A significant interest group thus was formed. The result of the desire for greater integrity and secrecy was the implementation of two new voting methods.
One new voting method was the replacement of separate party-issued ballots by a single neutral ballot containing the names of candidates of all parties for all offices. This "Australian" ballot (named for the nation where the concept first originated) was issued only by official election authorities and was made available only at polling stations on election day. This procedure prevented the issuance of counterfeit ballots and eliminated the turmoil and violence which had occurred outside of polling stations as party activists attempted to force voters to accept their particular ballot. Voters' choices were now secret, as the same form of ballot was used by each voter, regardless of the varying selections. An example of this type of ballot is shown by Saltman, 2006, p. 101. The first state adopting the Australian ballot statewide was Massachusetts in 1888, and by 1896, about 40 states, one-by-one, began to use the process. There was never a federal law or requirement; adoption and implementation was by the states themselves.
The second new method was the invention and deployment of the mechanical lever voting machine. The seminal invention occurred in 1889 and the first use in a federal election was in Rochester, New York, in 1896. The advantage of the machine was that it did not employ paper ballots and therefore no paper-ballot frauds could be perpetrated. (It has its own serious internal design defect, described below, which never generated demands for "paper trails" or the machine's abandonment.) Furthermore, since voters did not fill out ballots, they could not make non-standard marks; unapproved marks had raised the issue of "intent of the voter" and there were heated debates and lawsuits as a result of close elections concerning the intent of voters who had cast the crucial ballots. Thus, the type of dispute that occurred in Florida in 2000 has a long history.
The voting process with lever machines was secret, also. In the initial implementation, the voter entered a private, enclosed area (a "booth") where voting was carried out but, in later designs, the voter was protected from being observed by a curtain. The machine's face had the appearance of an Australian ballot, with each party's candidates in a separate parallel column and the various offices presented in separate parallel rows. In the initial design, a voter selected a candidate by pushing a locking push-key next to the candidate's name and later, using a subsequent invention, by positioning a small lever pointing to the candidate's name. The use of the lever made it possible for the voter to change a selection, not possible with the locking push-key. The sum of the votes for each candidate was indicated on a separate counter hidden inside each machine. The values could be viewed after voting ended and the machines were opened.
Diffusion of this type of machine slowly occurred and, by 1964, almost two-thirds of all US voters were using it. In that year, almost all other voters were using paper ballots that were hand-counted, and the remaining few were employing newly deployed computer-readable paper ballots whose holes or marks, indicating votes, were automatically sensed. In the late1970s and early 1980s, mechanical lever machines began to be replaced in small percentages by their computer-based direct-recording electronic (DRE) equivalents, and that conversion has continued through the present. Some of the mechanical devices continued to be used in 2005 state elections, for example, throughout New York. (In 2006, the US Department of Justice filed a suit against that state. New York had accepted funds under the federal Help America Vote Act of 2002 but had failed to utilize the funds for agreed purposes, such as the replacement of its lever machines or development of a statewide computerized voter-registration file.)
Thus, at present, US voters may find artifact-based or non-artifact voting devices at their polling stations. Non-artifact voting, first orally, then with mechanical lever machines, and more recently with DRE machines, has been used somewhere in the US since the first votes were taken in the original states. Non-artifact voting systems require special techniques for design and implementation to enable independent verification to be carried out.
3. Concern over Software Fraud Begins
Soon after computerized voting with punched cards began to be used in 1968 in Los Angeles County, the question of the possibility of fraudulent software arose in that region. The issue was highlighted in a page-one story in The Los Angeles Times (Bergholz, 1969). The article described an experiment undertaken by a group of computer scientists. The group was divided into two. One sub-group secretly changed a computer program that was to count votes on ballots by adding a bias routine, and the second group was supposed to find the added code. In the experiment, the attacking sub-group seemed to have the edge; the defenders could not find the malicious routine because it had erased itself after performing its nefarious work. This fact highlighted the insidious nature of program manipulation.
The newspaper story created a stir in the Los Angeles area. The county government established an Election Security Committee to review the situation. Later, that government, as well as the California State Commission on Voting Machines and Vote-Tabulating Devices, let a number of contracts to specialists who further analyzed the issues. Additionally, articles by technical experts proposing remedies were written in journals for computer professionals. Recommendations from all of these efforts, identifying methods of preventing software fraud, were reported and categorized into differing subjects as follows (Saltman, 1975, pp. 35-38):
(1) require audit trails of computations;
(2) limit physical access to systems;
(3) allow observer teams to watch proceedings;
(4) undertake recounting;
(5) set requirements for design of computer programs;
(6) carry out testing of computer programs;
(7) improve security of computer systems and operator procedures;
(8) review and achieve better administrative management; and
(9) adopt and implement state-level regulations.
In 1969, when the concerns first arose, there were no national standards and no institutionalization of testing of computer programs used for vote-counting. A better institutional environment, suggested by (9), would arise slowly. California has taken the lead among the states with the development of significant state regulatory activities. A role for the federal government, an issue not raised in the Los Angeles recommendations, eventually was considered by Congress. The Clearinghouse on Election Administration was established in 1972 and moved into the Federal Election Commission (FEC) in 1975. The development of the first federal voluntary standards began in 1984, and their completion and issuance occurred in 1990. The process of certification of independent testing laboratories (ITAs) was started soon afterwards. Finally, federal institutions specifically focused on the effective administration of elections were created in 2004 under the Help America Vote Act of 2002 (HAVA). ITAs are now called Voting System Testing Laboratories (VSTLs).
Next Page 1 | 2 | 3 | 4 | 5 | 6 | 7 | 8 | 9
(Note: You can view every article as one long page if you sign up as an Advocate Member, or higher).
