What Widens and What Closes the Divide

The digital divide is the gap between people who have reliable, fast internet access and the skills to use it, and people who don't. That gap already exists because of income, geography, and infrastructure. So when a service moves entirely online, it doesn't create a new problem from nothing; it takes an existing gap and makes it matter more, because now the service that used to be reachable in person is only reachable to people already on the right side of the divide.

Three kinds of choices push the divide either way. Governments decide whether to fund rural broadband, subsidize devices, or keep in-person options alongside online ones. Companies decide where to lay cable, which neighborhoods get fiber, and whether their apps work on cheap phones with limited data plans. Individuals decide whether to spend scarce money on a data plan, and whether they have time to learn new software. Trace one chain: a household has no home internet, so a student does homework on a library computer with a one-hour limit, falls behind on skills that employers now expect by default, and a decade later has a thinner resume and fewer job interviews than a peer who had broadband at home the whole time.

Mobile-first countries show the divide isn't one-size-fits-all. Where mobile networks arrive before wired broadband, people skip straight to smartphones for banking, school, and government forms, which closes the access gap faster but still leaves out anyone without a phone or affordable data. This connects to bias in data sets: training data scraped from online activity overrepresents people who already had reliable access, so the models built from it work less well for everyone else.