I'm not sure what you're saying. You say that unlimited users use significantly higher average data than the capped users but that is not what the study shows.

I'm not sure what you're saying. You say that unlimited users use significantly higher average data than the capped users but that is not what the study shows. Although that might have been true before the carriers introduced throttling, the actual article shows that the top 5% of unlimited users use about the same data as the the top 5% of limited users. That is what is surprising about the results. What you're saying -- which is that data hogs push the unlimited average much higher than the limited average -- is not actually what is happening. AT&T doesn't have to try to match the medium data useage (although that is indeed what is happening). I think what you should be saying is that AT&Ts throttling keeps both the medium and the mean about the same. This is because people on the unlimited plans don't necessarily use more data than the tiered people. If you look at the actual data (which, unfortunately wasn't in the NYT article) the top 5% of data users across all plans and carriers is about 3.7GB. This is already below the 5GB cap. I think what the article posits makes sense. The purpose of throttling unlimited users who are already below the 5GB cap is to get grandfathered people off unlimited. It's a rotten deal because they (and by "they", I mean mostly AT&T) want to make their unlimited plans undesirable so that you never use much more than 5GB, which is the cap for the tiered plan anyway. Sure, you get theoretically higher monthly bandwidth but the quality of service is actually worse than the tiered plans.