Hardfork 19 has been a great success, except for one potential issue. Anecdotally, I see lots of curators vote 10%-25% for others' post, but give themselves or their voting colluders a full 100% vote. Many do these for comments as well. This self-vote is 4 times as powerful as before, and thus, self-upvoting can potentially account for at least 4x as much of the reward pool as before. I say at least because they would vote others with a lower weight to save their much more costly VP, and thus more of the reward pool overall goes towards self-upvotes beyond the 4x greater strength. Many pointed out this abuse vector before HF19, but Steem developers and witnesses seem to have turned a blind eye to this potential issue. 

I'm no developer, I don't know how to access this data. What I'm sure all of us would like to see is how the self-voting behavior has changed post-HF19. 

So developers - we need some statistics. Basically, a chart that plots the self-upvotes potential payout as a proportion of all author rewards, daily, over time. If we can go back to earlier in the year pre-no-whale-experiment, that'd be ideal. 

I hope to be proven wrong, but we might have an issue here. Of course, voters may choose to do as they please, but if significant amount of the reward pool goes towards self-voting, it could be detrimental or at least sub-optimal to the network.

Addendum: 

Something else to think about. For a $100 upvote, it currently requires only a 125,000 SP account. That could easily reward themselves thousands of dollars per day, cashing in both curation and author rewards.

Edit - To clarify, I didn't intend this to be a discussion about the ethics of self-voting. We just want to see the effects of it. Most of the comments seem to be about how self-voting is acceptable or not. That's quite alright, but the end question here is - does this change in HF19 make Steem less attractive to new users?