Political analysts, pundits, and pretty much everyone to be honest frequently rely on Twitter
chirps tweets and Facebook posts to gauge aspects of many, many social issues, most prominently the upcoming May 2016 elections. With lots of Filipinos of all stripes making social media their new tambayan where they can share tsismis memes and social commentary, the Internet seems to be fertile ground for taking the Filipino social pulse like never before. Examples here, here and here (this last one was really fun).
We are, after all, the “social media capital” of the world, despite high prices and abysmal speeds, boosted by Free FB promos from telecom providers and the like. Compared to what we can get from Twitter and Facebook, pen-and-paper surveys and other old-fashioned methods of data gathering seem woefully out of touch.
However, estimates of Internet penetration in the Philippines compiled from various sources place the number of active Internet users at 44.2 million as of one year ago (January 2015), or slightly less than 44% of the Philippine population. That’s quite high – according to the same article, Internet access the world over is around 42%. But the natural question to ask is whether or not this 44% of the population is evenly spread across demographics – sex, age, income group, education level, region, etc – which would make it representative of the Philippines as a whole. There is definitely reason to think that the Internet is skewed towards younger people, or perhaps towards richer people, though I don’t know any way to estimate this more precisely without fine-grained data on individual Internet use. And those ads with Lola using FB beg to differ.
One thing I do look at here is where in the Philippines our “netizens” live. As it turns out, the 2010 Census of Population and Housing asked whether or not households had Internet access at home, and whether or not households had Internet access elsewhere. (I can only hope the 2015 Census asked this too so we can get some newer data). Aggregated results at the municipal and city level are available on the website of the Philippine Statistics Authority, here for example. That one has statistics for Batangas’s cities and towns, and there’s a separate spreadsheet for every province.
This is a map of the proportion of households in each city and town who report having Internet access at home on the left and Internet access elsewhere on the right. (EDIT: The map originally used a white to blue color ramp, but people found it difficult to see so I’ve changed it to white to red.) Unfortunately, the data make it impossible for me to gauge total Internet access, because a household could very well have answered “yes” to both, and I have no way to address double-counting, so separate maps for “at home” and “elsewhere” will have to do.
Internet Access at Home
Less than 1% of households in 525 cities and towns have Internet access at home. Between 1% and 10% of households in 949 cities and towns have Internet access at home, while 156 cities and towns have between 10% and 50% of households connected, and only 3 places – all in Metro Manila – have between 50% and 65% of households with Internet.
Evident in the map is a dark cluster of red representing a (relatively) high proportion of Internet access in Metro Manila and the surrounding provinces. There are also isolated patches of darker red clusters spread out across the country, which more often than not are provincial capitals. Clusters of white, representing places where less than 1% of households have home Internet access, can be seen in the Cordillera Administrative Region, in parts of the east coast of northern Luzon, in Masbate, in Agusan del Sur, in the Autonomous Region in Muslim Mindanao, and around southern Mindanao (Sultan Kudarat, South Cotabato, Sarangani, Davao del Sur).
Of course, population is not evenly spread out across Philippine cities and towns. As of the 2010 census, there were 31,813,129 people living in the 159 cities and towns where more than 10% of households had home Internet access.
Internet Access from Elsewhere
257 cities and towns have less than 1% of households with Internet access from some place other than home, while 791 have between 1% and 10%, 555 have between 10% and 50%, and 30 have between 50% and 96.4%.
One of the biggest differences between the two maps, aside from the fact that many more people access the Internet from “elsewhere” (whatever that meant to them) rather than at home, is the massive change in the numbers for the Cordillera Administrative Region. Among the 30 towns where more than half of households report Internet access from elsewhere are Bontoc (the capital of Mountain Province), and surrounding towns Sabangan and Sadanga, which report 70.9%, 82.7% and 55.3% respectively. (Favorite tourist destination Sagada isn’t included though). Also included are parts of Cebu, including Cebu City; the towns of Dauin, Sibaton and Sibulan in Negros Oriental, Bunawan in Agusan del Sur, Penarrubia in Abra, and Bagabag in Nueva Vizcaya.
Meanwhile, pockets of small white clusters can still be seen, such as in Lanao del Sur and Maguindanao. 54,054,760 people lived in the 585 cities and towns with more than 10% of households having Internet access “from elsewhere”.
Lastly, I check for whether clusters of high Internet access and low Internet access exist, using a measure of the correlation between the value for a place and the value for that place’s neighbors. There are many ways to specify neighbors, the most popular being “contiguity” (a place’s neighbors are the places that are next to it), but that doesn’t quite work for an archipelagic nation like the Philippines. I instead use an approach where a given city or town’s neighbors are the 8 other cities and towns nearest to it. The resulting structure of neighbors looks like this:
Red clusters, or clusters of high Internet access, can be found around Metro Manila and Cebu. For Internet access from home specifically, Ilocos has some high clusters as well, and for Internet access from elsewhere, the Cordilleras have that in spades (look at the splotch of red in northern Luzon). Satellites must be a thing or something. I don’t know.
Blue clusters, or clusters of low Internet access can be seen in the Cagayan area in northeastern Luzon, in northern Samar, and in western Mindanao. For Internet access from home specifically, Quirino, Aurora, and hotspot area Davao don’t have a lot of home connections, but those clusters disappear when considering Internet access from elsewhere.
Other colors (purple and pink) are much rarer, and depict clusters of high Internet access surrounded by low Internet access, and vice-versa.
Internet access is not evenly distributed across the Philippines. Duh, yeah, but it helps to see it on a bunch of maps. It’s a caveat among many against trying to chart trends in Philippine public opinion and the like using what you can find on the Internet alone. If I had access to the census data at the individual or household level (which is some pretty sensitive stuff though), we could have a clearer picture of Internet access as well as the relationship between Internet and other things like poverty rates, urbanization, education, etc. (I actually did try to do this using data at the municipal and city level, but I wouldn’t put too much stock in the results of just running a bunch of regressions on aggregated data.) We could also have a much better sense of the demographic makeup of Internet users in the Philippines. Please let me know if any readers have any ideas.
The data’s also from 2010, so you could say it’s outdated by now, but I doubt the picture of geographic inequity has changed much between then and now. More updated data of this type won’t be around until 2018 at the earliest.
QGIS, GeoDa and R were used for descriptive analysis.