The tentative date for the resumption of registration is the 5th of November.
Come November the 5th, election offices nationwide will start accepting applications for:
.the registration of new voters;
.transfers of registration;
.changes of name; and
.corrections of entries.
There are also plans to conduct voter validation – that’s where voters come in to enroll their biometrics if they haven’t already done so in previous registration and validation exercises. Just to clarify – validation is NOT registration. When you come in to validate, it presupposes the existence of a valid registration. All you’re doing is confirming that you are, in fact, registered. You do this by submitting to the capture of your biometrics data – your image, your fingerprints, and your signature all electronically recorded.
Having announced all that …
Let’s talk a little about how cleaning up the list of voters with biometrics works.
First thing we need to wrap our brains around is the concept of a biometrics database. This means that a voter’s biometrics is stored in a huge computer, along with the biometrics of other voters. To date, the COMELEC has a biometrics database that includes approximately 50% of all registered voters. That’s about 20-plus million voters.
The second thing we need to be familiar with is the concept of an automated fingerprint identification system – AFIS (I think AFIS might be a trademark, but let’s just refer to it that way. Kinda like how colgate=toothpaste). An AFIS allows you to compare one set of biometrics data – fingerprints primarily – against all biometrics in the database. Those of you who’ve seen CSI will get this quite easily.
So, let’s say for instance, that I entered my biometrics into the AFIS and asked it to search through the database for any matches. The AFIS will take my biometrics and compare it with the other biometrics in the database. That’s one versus 19,999,999. If it finds a match, it’ll say so.
Now, if I’m a registered voter and the database is clean – and by extension, the list of voters – there should be only one match: James Jimenez. If I’m not a registered voter, then it’ll say “No Match.” But let’s say I registered in Manila and also in Taguig, then the AFIS will report two matches: one for James Jimenez in Manila, and another for James Jimenez (or whatever name) in Taguig. The difference in names won’t matter because the AFIS isn’t matching names, it’s matching fingerprints and fingerprints can generally not be faked. A result like that means that the list is inaccurate and steps should be taken to clean it up – and to file charges against me for having multiple registrations.
But here’s the catch. The system works best if you have a more or less comprehensive database. Which means that if there are 40-plus million voters, then you’d have to have at least (by my estimate anyway) 39.5 million sets of voter biometrics in the database.
Here’s why: Supposing in our example I had my biometrics captured in Manila but not in Taguig, even though I’m a registered voter in both places. When the AFIS searches the system, it’ll only find me in Manila but not in Taguig. Which means that the list will look clean, but it actually isn’t. That’s a false-positive. So, the further away we are from a 100% database, the more likely false-positives become, which defeats the clean up effort.
The third point to consider is this: as I said earlier, we have only about 50% enrollment. Which means that more than 20 million registered voters still don’t have their biometrics enrolled in the database. So, even if we did have an AFIS, there’s still too much room for false-positives, making clean-up still rather insignificant. What makes it worse is that we don’t even have an AFIS yet.
That’s right. No AFIS. It’s kinda like having a whole bunch of CDs, but no CD player.
We’ve asked for funding to allow us to procure an AFIS since 2004. Last I heard, it’s got cut from the budget for 2009. Again. Still, come November, we will once again gather biometrics; we’re gonna buy CDs but we still don’t have a CD player.
The big picture.
On the surface of things, this looks reasonable enough. But the surface of things rarely gives you the complete picture. And the complete picture is this:
Gathering biometrics requires the use of Data Capturing Machines (DCMs) of which we have a finite supply. Now this finite supply is – as all things are – vulnerable to a certain degree of attrition. Parts get worn out, computers crash, and so on. We replace them as fast as we lose ’em, but that’s like spinning your wheels: so much effort for little or no forward movement.
In the meantime, the registration process isn’t as efficient as it could be.
The law (RA 8189) says that applications for registration must be filed personally at the office of the Election Officer. Good. Now this law can be interpreted to allow the COMELEC to put up satellite offices apart from the main one. For instance, if we were to do this in Manila – there would be the EO’s office in Arroceros, but we could also put up satellite ‘offices’ in malls, schools, public markets, wherever. So, instead of having one office, we could have ten or fifteen or more. That would distribute the work load better and make things more convenient for people, i.e., they don’t have to know where Arroceros is, they can just go to school or the nearest mall.
Another advantage of this scheme has to do with the last minute rush to register. If you only have one point of registration, then EVERYone will go to that place at the last minute and inevitably overwhelm the system. This will lead to long long lines and much frustration – and ultimately, the non-registration of a lot of people. In contrast, if you had satellite offices, even the last-minute load will be spread out and so the problem will not be as bad. In fact, it might even turn out to be so manageable that it won’t be a problem at all.
But we can’t do that because, as I said, we have a finite number of DCMs – all of which are assigned to EO’s offices. And COMELEC policy as it stands is that the failure to have your biometrics captured means that your application will be deemed incomplete and not submitted. If your application isn’t submitted, then you won’t get registered as a voter. So, if we have just enough DCMs to serve EO’s offices, having satellite offices with no DCMs is pretty pointless: people will still have to go to Arroceros for data capture for their applications to be deemed submitted.
The solution most people advocate is the most unimaginative one: buy more DCMs so even satellite offices can have ’em. Easier said than done. Even assuming that we get the money to buy more DCMs, where would you store these babies? And like a car, you don’t stop spending for these DCMs after you’ve bought them because you still have to contend with things like spare parts and stuff.
The way I see it, throwing money at the problem won’t solve it.
Push back a little now, and try to gain a little perspective.
On the one hand, if we want as many people to register as possible, then we have to make it more convenient. Also, making registration possible in more places than just the EOs office means greater access to the process for everyone, and greater access + increased convenience = more registrants. On the downside, the clean-up process will be more tedious. So its really a case of more-registrants-more-tedious-clean-up.
On the other hand, we also want to build up the database because, with the AFIS, that’ll make clean up very easy. The downside is that, first, we don’t have an AFIS and we’re not getting it next year either; and second, the limited number of DCMs we do have prevent us from increasing access to the registration process through the use of satellite offices. It’s a case of less-registrants-more-biometrics.
Me, I see a value judgement here. Do we want more-registrants more than we want more-biometrics? If the answer is yes, then the obvious solution is to put the use of DCMs on hold. No DCMs frees us up to open satellite offices and increase access and so on. Better yet, the last-minute rush problem might even be eliminated.
If the answer is no, then we keep on using DCMs but the trade-off is that there there will be fewer registrants who will have access to the process. Worse, the last-minute rush of registrants will lead to a repeat of all deadline scenarios in the past: long lines, much frustration, lots of unregistered people.
Obviously, I’m in favor of shelving the DCM use, but not permanently. As inconvenient the use of DCMs is, it’ll all be worth it if we can actually use the data being collected. That is to say, if we had an AFIS. The problem is that we don’t have an AFIS, and so all this inconvenience is really being undergone for nothing. Well, nothing except for the expectation that an AFIS will be procured at some point.
But when will that happen? And in the meantime that we’re waiting for the AFIS to be funded, we’re experiencing so many difficulties getting people registered.
What about clean up?
I can hear you asking: what about clean-up? Well, first of all, even with DCMs, we still don’t get to use the biometrics data we’ve gathered because … yep, we have no AFIS. Second, clean-up is still possible because the registrations are computerized, i.e., the names of all registered voters are still entered into a database of names and other such details. Comparing names may not be as 100% efficient as comparing fingerprints, but a comparison of names will still identify multiple registrants. It’s just that we have to do a little extra work to verify if the James Jimenez from Manila is the same as the James Jimenez from Taguig. That’s not as high-tech as fingerprints, but it’s been proven to work.
I can also hear whining that goes like “we’re automating aren’t we? why go back to manual?” I have nothing but disdain for that kind of wrong-headed reasoning because it presupposes that we are automating for the sake of automation. WRONG. We are automating to improve the system. If one particular automation track is not doing that, then we have to reconsider doing things differently or – as I am saying – putting a different timetable on it.
Something to think about.