This makes sense.. though I'm not sure it will do much in practice to improve political discourse. The combination of politicians having twitter accounts and twitter's content recommendation engine promoting politician's tweets make me think the perceived effect to the end user (twitter showing people messages from politicians or their supporters) will remain more or less the same.
But because twitter isn't taking money directly from those politicians, it perhaps reduces their perceived culpability. It seems like a good move for twitter.
It will remove the direct influence on who the ad gets sent to at least, a tweet will spread organically to a much wider audience than a targeted ad so messages cannot be as micro-targeted.
Good point. Now if only we could ban micro-targetting in general. It's a corrosive force to society in a wide array of scenarios, not just political advertising. (For instance, it's been known to facilitate discrimination in real estate and the job market.)
I think for things like jobs, housing, etc that have protected civil rights the only real way to ensure people's rights aren't being infringed is to make them completely open and post both the ad and the stats of where the ad was served/the settings for the targeted audience depending on the type of . Maybe even need to directly test that the bidder isn't racially targeting with random test offers if it's a google style bot bidding system. A similar system is probably needed for political ads because one insidious fact of web targeted ads is there's much less chance to respond to and potentially debunk them if they're highly targeted. With radio and television ads you couldn't do that they'd go to everyone watching so more people outside the targeted demo would see.
But because twitter isn't taking money directly from those politicians, it perhaps reduces their perceived culpability. It seems like a good move for twitter.