Reason.com reports that this consensus is cracking [1]
And the correction at the end begs a fascinating question about
whether the media should present both sides of an issue. If
99% of research says A and 1% says not-A, is the media being
misleading by reporting both views? If the media doesn't
present not-A, then is it censoring opposing views?
Actually, they published an article highlighting the different view of a single researcher (ok - he had a co-author in one paper). Note that the paper showing the decrease in violence makes no claims for the reasons behind it.
I find his argument compelling, but that doesn't mean the consensus is any different.
Needless to say, this is very widely disputed and fought over. It's possibly more controversial than climate change.
However, it's quite widely studied[3], and the consensus position is fairly settled.
[1] http://www.jaacap.com/article/S0890-8567%2809%2960387-7/abst... (This is a meta-study)
[2] http://psycnet.apa.org/psycinfo/2014-41977-001/
[3] http://scholar.google.com.au/scholar?hl=en&as_sdt=0,5&q=impa...