4 ms·
It's rather funny how this article is framed. The unionization side lost the vote, which tends to imply that they are less popular. Despite that, all of the qu
by ufmace 6y ago
It's rather funny how this article is framed.
The unionization side lost the vote, which tends to imply that they are less popular. Despite that, all of the quotes are from people who supported the unionization. They have "a worker... who helped organize the union drive" and "the president of the Retail, Wholesale and Department Store Union that led the organizing effort". The article picture is of a pro-union protestor. They don't seem to have made any effort at all to find or quote people who were against it. They instead framed all opposition as obviously due to Amazon's "propaganda".
So what could be the reason for this - are their reporters just lazy, or are they hopelessly biased in favor of unionization? Well let's be honest, it's obviously both.
I'd also be interested to know here, what is the nature of the proposed unionization? Is this just Amazon workers, or would they become a small part of some massive national union organization? It quotes the "president of the Retail, Wholesale and Department Store Union", and their picture has some rather nicely printed signs by them, so I know which way I'd bet. If this unionization actually went through, would the individual workers have any say in what their contract would be, or would they just have to take whatever this massive national union thought is good for them? That might be a pretty solid reason to vote against it. Funny how it doesn't say anything about that.
- mimikatz 6y agoI tend to always go lazy over biased. More and more it is show reports follow narrative in many cases I am convinced it is because they can copy and paste about 2/3rds of the words in their stories that define the narrative. Also why Twitter gets cited so much you can easily find XXXX says XXXX on it.
- hackeraccount 6y agoThere's a third possibility. They know their audience. If the comments are anything to judge by I would say most of the readers would view the article as a fair representation of the event in question. I think the "problem" - and I can take the point that maybe it's no problem at all - is that the readers of arstechnica are people who are interested in technical issues and have very clear political views. The first is important but the second is critical.
- ufmace 6y agoThis sort of thing seems more like a self-reinforcing loop than something to be assumed innately. They write articles with a clear political slant that comes through strongly on some subjects. Potential readers who don't like that slant go elsewhere, and those who love the slant stick around, read everything, subscribe, and comment a lot. Perhaps more concerning are those in the middle, who would say they don't have a self-proclaimed position, but aren't sophisticated enough to recognize the slant. For those types reading media like that, they tend to absorb that slant as a law of nature instead of a political opinion. I remember once upon a time, it was considered to be a good thing for journalists to present both side of an issue. Maybe not every single issue, but at least anything important and contested enough to be worth holding a vote on. We were supposed to trust them, because they were supposed to know enough to do this. It was supposed to be dangerous and encourage extremism to present such issues as if there was only one legitimate side. I guess those days are long gone.