ATM to machine the polls – the economics of automation

Frank & Ernest

There is much ado about how to computerize elections in the Philippines.  Here is my attempt at setting out the economics of the matter.

There are two goals: ensure integrity and security of the poll process, and minimize the cost of automation.  In economics, these are the “goods” we wish to buy or produce from our limited budget or resources.  The cost of automation includes not only the cost of training voters but also the foregone opportunity to use what people, such as poll watchers, already know about the election system in the Philippines.

A hybrid approach is likely the only route in the present environment, where perhaps only about half the voters are sufficiently tech savvy or tech “willing,” and it would be too expensive to implement full automation.  Thus, some remnant of the “paper” system needs to remain.  For the automated system, the least-cost route is an existing secure network that a voter already knows how to use.  I suggest that this is simply the ATM network of the local banks.  I estimate that half the voters know how to use ATMs.

How do we ensure integrity?

For the remnant paper system, voting is the same as before, and so is the “watching” of the vote count.  Candidates and other interested parties will still have a right and duty to “watch” each other in the counting.  What would be automated is the precinct-level canvass, since the authentication and input of the data here can be “watched.”  If done well, this eliminates the infamous “dagdag-bawas” in the post-precinct canvassing.  In recent poll contests, it seems that serious candidates have learned to protect themselves at the precinct watch stage.  Training will have to be conducted for the poll inspectors and watchers, but there will be no need to train the voters. 

The ATM is a ready-made vote-counting machine; all it needs is a common interface, set up at election time, which shows the ATM user his candidate choices.  The nice thing about the ATM is that it already has built-in security features: only the voter knows his ATM number, and he can vote in the secure environment of the ATM.  In order not to clog up the ATM system, voting can take place over, say, a two-week period, where the votes are not “final” until the date of the election (the voter can change his mind – which helps prevent vote-buying).  For after-election auditing, a voter can have his vote result emailed to him by the banks, and he will have a chance to compare that with the COMELEC database which will be open to him. An auditing firm can set up a random process to verify integrity, in the same way that such firms track bank transactions to uncover any bank frauds. 

A process will have to be devised to ensure that voters vote only once.  This can be done by assigning each voter a unique voter identification number (VIN).  At the precinct-level count, each ballot’s VIN must be entered into a precinct machine, connected on the web to the central COMELEC server.  The ballot will be rejected (or a blank one not even given to a voter) if the VIN has already been used at an ATM, or if the VIN is invalid.  All poll watchers will be trained to ensure this.  All the VINs actually allowed will have to be entered in the precinct-level canvass, so that an audit process can also randomly verify compliance with the one-vote-only rule. 

As to the cost, the usual answer is “It depends.”  But I would bet that the cost of this approach is smaller than the amounts currently budgeted for automation.  The ATM networks will be compensated for use of the ATMs, and for the special programming needed to convert (temporarily) an ATM machine into a poll machine.  The consortium of banks, in an act of civic-mindedness, can perhaps find a way to allow the security of their network to safeguard the security of the vote.

We often find informal polls offered in many websites, including personal blogs.  We don’t worry about security and integrity because the polls are mostly for entertainment.  Election of public officials is another matter, but it seems possible to have security, integrity, and low cost, with a little help from the banks!  A hybrid approach will also allow those who don’t trust machines to use the old paper ballot.


2 thoughts on “ATM to machine the polls – the economics of automation

  1. I originally posted this without checking on the state of the law and the politics. But the idea that we can multi-task the atm is good if we can solve the security issues. It also means we ought to be able to vote over the web.


Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s