NYT Calls for Open Source e-voting machines
Here is the original article . The author pays the Open Source community quite a complement in their brief summary of the process and its potential merits as applied to e-voting. I am happy to see this in the mainstream, both Open Source and the snags surrounding e-voting today are key issues that deserve a lot more attention than they are currently getting.
Having said that, I fear the article is missing some key points necessary for a complete discussion of the problems we are facing today.
How will the user know the software running on the machine they are using is the same software posted openly for all to see?
To me, this is a pretty big deal. No matter how secure the system ends up being, the user cannot ever really know for sure that system is the one handling their vote. I have the highest respect for the Open Source process and values, but without means for user verification, the efforts would be largely useless.
How does the user know the vote cast is the vote recorded?
Again, this is a matter of representation. We can mitigate this problem with a paper record of the votes cast that users can examine before they are stored for potential recounts later. Please do not call these reciepts because they are not. They are records, a receipt is something the user would keep. We cannot have an anonymous vote if the user keeps their voting records, so we need a record of the vote cast, not a reciept.
A vote cast could be recorded for accurate count later, yet still be misrepresented in ways subtle enough to throw an election without ever triggering the recount. Even though an accurate record is kept, the lack of visibility in the electronic collection and tally processes, makes them subject to manupulation. Such manupulation could easily lie below levels that would trigger audits, yet still have an impact on close elections.
Voting with a machine is like voting by proxy.
When we use a proxy for something, two things happen:
- we are placed in a position where we must trust the proxy,
- others are in a position to manupulate said proxy without our knowledge.
This is where the electronic methods break down. The voting process is a fragile one. We need to be aware of the simple truth about voting, largely missed by the media so far; namely,
there is only one chance to capture the voters intent and that is at the moment they cast their vote. Unless that record is directly used for the tally, corruption and error can and will occur.
Voters that cast their vote using traditional human readable media, such as optical paper and punch cards, for example, see their vote recorded, moved, counted and reported to others for the final count. All of these actions can, and should mind you, be observed and verified using simple human means of communication and observation. This is democracy in its purest form. We should respect this above all else because our very nation is founded on its principles.
Performing our civic duty is a very American thing to do, why are we working so hard to replace that process with one that breaks the very line of trust our founding fathers worked so hard to secure in the first place?
The problem with e-voting is in the tally of the votes cast. Electronic media is not human readable; therefore, we cannot ever be sure the tally reflects the vote cast, unless we place our trust in another party to collect and tally the vote. How do we trust these third parties? Who shall choose? Why are we even following a path laced with doubt at its very core?
Here in Oregon, we decided to improve the voting process using the mail in ballot system. We still have paper, and the exact record of the vote cast is used for the tally. We can have recounts if we want to and people can still cast votes on election day and watch another person put their vote into a ballot box, if they so choose. What's wrong with that?
We should work hard to ensure our future voting systems feature at least these attributes:
- the record of the vote cast needs to be human readable on media that can survive a number of audits and that is difficult to change,
- the record of the vote process needs to be complete enough to know who voted and insure no over voting happens, but no vote shall be personally identifiable,
- those putting together the final tally should do so in a public manner, using human readable intermediate records for later observation and trust,
- the record of the vote cast must be directly used for the tally,
- aggragation of the vote, for the tally should be public as well.
None of this fits any electronic voting system we have on the table today, all of it is necessary to preserve the line of trust so important to the voting process.
I embrace technology in all parts of my life, with mostly good results I might add. No matter how hard I try to get around the matter of trust in this process, I cannot.
Is it really so hard to gather and tally a human readable record of the vote a few times a year? Is it hard enough to erode the core values our nation was built on in order avoid doing so?
I will gladly step up and do my part because my freedom and trust in my government demand it, shouldn't you?
Having said that, I fear the article is missing some key points necessary for a complete discussion of the problems we are facing today.
How will the user know the software running on the machine they are using is the same software posted openly for all to see?
To me, this is a pretty big deal. No matter how secure the system ends up being, the user cannot ever really know for sure that system is the one handling their vote. I have the highest respect for the Open Source process and values, but without means for user verification, the efforts would be largely useless.
How does the user know the vote cast is the vote recorded?
Again, this is a matter of representation. We can mitigate this problem with a paper record of the votes cast that users can examine before they are stored for potential recounts later. Please do not call these reciepts because they are not. They are records, a receipt is something the user would keep. We cannot have an anonymous vote if the user keeps their voting records, so we need a record of the vote cast, not a reciept.
A vote cast could be recorded for accurate count later, yet still be misrepresented in ways subtle enough to throw an election without ever triggering the recount. Even though an accurate record is kept, the lack of visibility in the electronic collection and tally processes, makes them subject to manupulation. Such manupulation could easily lie below levels that would trigger audits, yet still have an impact on close elections.
Voting with a machine is like voting by proxy.
When we use a proxy for something, two things happen:
- we are placed in a position where we must trust the proxy,
- others are in a position to manupulate said proxy without our knowledge.
This is where the electronic methods break down. The voting process is a fragile one. We need to be aware of the simple truth about voting, largely missed by the media so far; namely,
there is only one chance to capture the voters intent and that is at the moment they cast their vote. Unless that record is directly used for the tally, corruption and error can and will occur.
Voters that cast their vote using traditional human readable media, such as optical paper and punch cards, for example, see their vote recorded, moved, counted and reported to others for the final count. All of these actions can, and should mind you, be observed and verified using simple human means of communication and observation. This is democracy in its purest form. We should respect this above all else because our very nation is founded on its principles.
Performing our civic duty is a very American thing to do, why are we working so hard to replace that process with one that breaks the very line of trust our founding fathers worked so hard to secure in the first place?
The problem with e-voting is in the tally of the votes cast. Electronic media is not human readable; therefore, we cannot ever be sure the tally reflects the vote cast, unless we place our trust in another party to collect and tally the vote. How do we trust these third parties? Who shall choose? Why are we even following a path laced with doubt at its very core?
Here in Oregon, we decided to improve the voting process using the mail in ballot system. We still have paper, and the exact record of the vote cast is used for the tally. We can have recounts if we want to and people can still cast votes on election day and watch another person put their vote into a ballot box, if they so choose. What's wrong with that?
We should work hard to ensure our future voting systems feature at least these attributes:
- the record of the vote cast needs to be human readable on media that can survive a number of audits and that is difficult to change,
- the record of the vote process needs to be complete enough to know who voted and insure no over voting happens, but no vote shall be personally identifiable,
- those putting together the final tally should do so in a public manner, using human readable intermediate records for later observation and trust,
- the record of the vote cast must be directly used for the tally,
- aggragation of the vote, for the tally should be public as well.
None of this fits any electronic voting system we have on the table today, all of it is necessary to preserve the line of trust so important to the voting process.
I embrace technology in all parts of my life, with mostly good results I might add. No matter how hard I try to get around the matter of trust in this process, I cannot.
Is it really so hard to gather and tally a human readable record of the vote a few times a year? Is it hard enough to erode the core values our nation was built on in order avoid doing so?
I will gladly step up and do my part because my freedom and trust in my government demand it, shouldn't you?
posted by Doug Dingus at 10:04 PM
0 Comments:
Post a Comment
<< Home