sent it along to some friends for their input, they’d be in the same bind. One after another, in just this fashion, users tumbled into the program. It became a standard, a “platform” in industryspeak. And Microsoft enjoyed a particularly appealing economic leverage: Developing Word may have cost millions, but once that work was done, each additional copy cost just cents to produce. This astonishing speed loop of profitability demanded a whole new economics. It also forced a reconsideration of what “competition” might really mean. Once Excel or Windows had settled into place, had become a standard, you couldn’t really compete with it. New, optimistic maybe even better rivals rushed into the marketplace, but they were all assaulting the impregnable wall of habit, of a locked-in technology. Should this be legal? Arthur wondered. Traditional economics said such monopolies were bad for everyone. (As did the Department of Justice and their global peers as they chased Microsoft for a decade.) But was that right? The “platform dividend” that accrued to Microsoft was surely large, but if you could somehow total up the benefit to the rest of us? The convenience, the efficiency, the benefits of Microsoft's billions of research spending might dwarf even Redmond’s massive profits. “Increasing returns,” Arthur wrote, “cause businesses to work differently and they stand many of our notions of how business operates on their head.” The essential phenomenon Arthur spotted at work two decades ago is something we now know as “network effects” - an idea that changed how we think about businesses, and particularly about the sticky and alluring power of gated, connected systems in nearly any setting. In the years after Arthur’s paper, billions of us ran madly along a course he had anticipated: We crashed our way as fast as possible into those single, winning businesses - rewarding them with near monopoly positions in exchange for the benefits of being “inside”. In the twenty years since A