Most political analysts have already started to dissect the just-concluded Philippine 2013 elections—many of them focusing on the fate of individual senatorial candidates. Understandably, they pose such questions as why Grace Poe took the top spot, why Nancy Binay remained on 5th as predicted (despite the many brickbats thrown her way), or why Risa Hontiveros or Teddy Casino for that matter failed to land into the Magic 12 despite the all-out efforts and formidable strengths of their respective camps.
Insights emerging from these analyses are interesting and useful. But if I had the time (which I don’t), the resources (which I don’t), and the technical skills (which I have), I will conduct another kind of post-mortem analysis of the 2013 elections. I will use Big Data techniques to analyze the electoral patterns at various levels and in different dimensions. The AES Watch and Kontra Daya people should be able to conduct this type of electoral autopsy—not simply because they want to expose anomalies (although that’s an important aspect), but to understand and perhaps even reconstruct the entire process as it actually unfolded in practice.
The raw data is all there, although I’m not sure if the Comelec, PPCRV, or their media partners are willing to allow outsiders to access it at the lowest possible level, say, as files in CSV format. The database and data visualization tools for analyzing this huge but formless mass of data and resolving it into clear patterns are readily available, needing only a powerful computer and a database team who know their stuff. I’m sure the AES Watch can mobilize a lot—and I really mean a lot—of really intelligent computer scientists out there to at least initiate such a project.
Think of it this way: This raw data represents a live representation of the Philippine electoral terrain as it moved on May 13. You could sort of “Google-Earth” this terrain to zoom in and see precinct-level details, to zoom out and see big geological forms not so obvious in close-up view, or even to cross-section the terrain in so many ways. There are so many data patterns and trends to discover and visualize into all kinds of charts and maps, and incredibly insightful conclusions to draw, from such an investigation. And yes, including digital anomalies and statistical improbabilities to unearth.
Just to show you a tantalizing glimpse of the power of Big Data (which I had the rare opportunity to enjoy during my stint at GMA News particularly for the 2010 elections), here is an interesting tidbit: In certain cities, in certain precinct clusters (sometimes even adjacent clusters amounting to a whole barangay), the election returns show 100% or near-100% similarity of votes for a certain ticket or exact combination of candidates. Certain combination of candidates would get zero or near-zero votes in areas that were expected to be their political bulwarks. In some cases, the percentage share of votes for contending candidates would show up as fixed, unchanging figures (say, 44.5% for candidate A and 32.5% for candidate B) across many precincts.
Drill down further, or make cross-regional or cross-provincial comparisons, and you’ll be surprised at interesting patterns that you never suspected, electoral demons that you never imagined anyone would conjure. Do cross-references over time, and news clippings, and background checks, and you can actually write a voluminous report entitled, “The Astonishing Truth Behind the 2013 Elections”—all backed up by Big Data analysis.
I’m not saying anything more, at this point. It’s all up to you. Act now while the data is still there. Otherwise, what happened in past generations, in 2010, then in 2013, will happen again. And again, and again, and again, and again… till the entire citizenry realizes that there’s more to democracy than otherwise intelligent voters hypnotically following the motions of a quiet, well-oiled machinery—a machinery that we don’t quite understand except to note that it’s fueled by oodles of money that go into guns, logistics, media ads, and (now digitized) votes.