I first reported on this issue a couple of years ago. The trend continues:
As RePEc membership and the number of items online continues to grow, the number of abstract views and downloads per person continues to fall. The chart shows abstract views and downloads each month per RePEc member. The number of downloadable or viewable items per member is actually constant or rising. Possible explanations are:
1. The average quality of papers is declining - assuming better economists joined earlier on average.
2. Too much supply relative to demand. The total community of economists doesn't expand as fast as the number of RePEc members.
3. Of course, the average paper is getting older and older papers will be downloaded less.
4. A lot of people's papers were already on RePEc as journal articles and also working papers but they weren't registered. They are not adding papers as they register and so the number of downloads per person registered is declining.
5. A new phenomenon is that total downloads and abstract views have begun to fall:
Maybe this is attributable to improvements in Google Scholar and other means of finding papers that bypass RePEc?
Whatever the cause, a given number of abstract views or downloads is getting a higher RePEc ranking over time.