Web Voting Booth

It seems likely that the next U.S. Presidential election will be rigged just as badly as the 2004 election was, if not worse. I propose the creation of an alternative voting system. At a minimum, the system could serve mainly as a sort of "reality check" to the official voting records; with statistical adjustments, it could be more accurate than any "exit polls" or other such instruments currently employed. If the system were successful, it could eventually be a model for a replacement voting system to be used for the official vote count.

Why this hasn't been tried before

In ages past, the most any individual or small group could have hoped to do about this problem was to start an inquiry through the usual official channels. (Such inquiries have already been started, and they may well succeed in fixing the process in time, but there appears to be a substantial chance that they will not.) There was simply no way to collect votes from individuals across the entire country other than massive deployment of officials with the appropriate equipment to register votes, plus an additional massive number of individuals at various central locations to collate the results.

With the advent of the Internet age, however, it is theoretically possible for a single individual to collect votes from all across the country, tabulate them, and present the results in real-time.

With well-written software and adequate CPU power and bandwidth, this collection process could be more accurate, more secure, more flexible, more complete, than any of the methods which have been used in the past – in fact, it would be in pretty much every way superior to what we've had to put up with in the past.

With most libraries having at least a few networked computers, it should be possible to collect votes from absolutely anyone who wants to participate – regardless of income or education level. Public schools now teach basic computer usage, and the software can be designed to explain the process further if needed. As there will be no need to narrowly restrict the voting time-window, there should be little or no pressure to place your vote and make way for the next person.

Auditing

There must be verification at every level. An individual must be able to verify that her/his individual vote was counted correctly, and that the total of all votes on a particular issue has been counted correctly and includes her/his individual vote in the total. We also want the option of preserving anonymity, since this is traditionally an important part of the U.S. voting system.

The exact details remain to be worked out, but the basic idea goes something like this:

• Upon placing her/his vote, each individual receives a unique "ticket", i.e. a string of alphanumeric characters which is different for each voter. This string is confidential, i.e. there is no way for a non-privileged third party to determine which ticket was issued to a particular individual or which individual received a particular ticket. (The ticket should also come with a time-stamp, for second-level verification; there are additional safeguards which can be taken, such as including a digitally-signed checksum of the time-stamp in the ticket, making it very difficult to generate fake or duplicate tickets.)
• A complete list of all individual votes will be available online. Each vote will show the ticket number, the timestamp, and the value of the vote (i.e. how the individual voted). The data should be available in a format that is both easily translatable into spreadsheet format (for vote-counting) and legible to humans (for quickly checking that your vote was registered correctly).

Voter Verification

Obviously we need a way to prevent individuals from voting multiple times; there are a variety of ways of doing this and I don't really consider it an issue, but just to be thorough I'll mention a couple:

• As with the present system, each voter must register with the computer voting system no less than (say) 30 days before the next election in which the voter intends to vote. The voter enters her/his social security number and current address. The voting system sends a postcard to the address currently in their system for that SS number, asking "did you request this?" and giving a passcode for the recipient to use when responding "yes" or "no". This does leave room for potential troublemakers to abuse cards sent to old addresses and not forwarded properly, but no more so than the existing system of voter registration -- and there will at least be a record that the claimed address did not match the one on record, to help in resolving disputes.
• Voter registration cards are issued exactly as they are now; each card contains a unique voter ID. Each voter enters their voter ID into the system to allow creation of a user ID and password. This must take place at least (say) 30 days before the next election, to allow time for resolving disputes in case someone gets hold of someone else's voter ID. (Postcards should probably go out to the address-of-record too, to prevent abuse of non-voters' IDs.)

Hurdles

Obstacles common to any voting system:

• identification of qualified voters (see Voter Verification)
• prevention of duplicate votes (see Auditing)
• maintaining privacy of individual votes (see Auditing)
• (doing both of the above at the same time) (see Audiding)
• prevention of tampering (by officials or others)
• tabulating the results (relatively minor issue when the "voting machines" are all networked computers)

Obstacles particular to any privately-run (alternative) voting system:

• getting people to use it

Practical obstacles:

• software (there was much talk of developing open-source software for the digital voting machines after the last election – did anything actually get written?)
• making sure there is enough CPU power and bandwidth

Opportunities

We have the opportunity to condider some possible improvements to the basic structure of the voting system, and perhaps offer them as "extended voting options" to any voters who care to try them. If nothing else, the two-party system strikes me as unnecessarily divisive, and possibly nothing more than a remnant of the days when the system had to be designed to minimize the amount of data which was collected. Discuss.