Video is going to use the more data than, say, search because that's the nature of video. After all, video is more-or-less why broadband exists in the first pla

Video is going to use the more data than, say, search because that's the nature of video. After all, video is more-or-less why broadband exists in the first place. And more to the point, the extreme variability in the amount of data used by various applications provides a pretty good argument as to why you really shouldn't use an undifferentiated lump-sum sum of bandwidth to rank people's priorities when it comes to internet use. I mean, I may use more bandwidth watching a two hour movie at the end of the day than I used in every other hour combined. That doesn't tell you anything about the relative value of what I did with the rest of it. And even if video does use more than half of people's bandwidth, that's still an awful lot of other stuff going on. And as far as connectivity in schools goes, the value of that is truly debatable. There is a growing body of evidence to suggest that academic performance tracks disturbingly closely to household income. In other words, failing schools are just one more symptom of a much larger problem with entrenched poverty. And when it comes to solving that problem, getting access to as many people as possible (as opposed to limiting it to their kids, and only when their kids are in school) seems like a fundamental pre-condition. In the absence of hugely comprehensive and cash-intensive approaches to schooling in economically devastated areas (e.g., the model developed by Geoffrey Canada), a limited focus on schools that avoids the myriad ways impoverished home life undermines academic study may end up doing more harm than good, in the same way junk food provides a lot of calories without providing much nutrition. On the other hand, if you make serious headway on the poverty problem, the schools will start to take care of themselves. Indeed, this is exactly what happens in suitably well-off areas. Recognizing that entrenched poverty is the product of structural issues that go way beyond education, dealing with those issues directly seems like the higher priority. To that end, subsidized access seems like a no-brainer. To the extent that a fully-functioning smartphone with an internet connection is becoming a passport to the modern world, we should be going out of our way to ensure that people who can't afford to buy them on their own are being given what they need. It doesn't have to be an iPhone 6, but it does need a working browser and access to communication apps.