-
I haven’t followed this debate in a field I know little about, and I don’t like the tone of this tweet. What I want to note is that if economics were used to publish meta-analyses, we could rely on them to assess the robustness of policy advising. Short thread ⤵️ @LHSummers/1145824471430905856
Mood +3 🙂
-
One paper (or even several papers) is only *1* data point. And there are so many published papers that one can always find one that goes in his.her direction. Meta-analyses agregate much more data points, and are a convenient way to assess how one claim fits with the literature.Permalink
Mood 0
-
Recently, France announced that homeopathy will not be reimbursed by the Social security anymore. This decision was made after reviewing a large corpus on its effectiveness – and the consensus is that there is no specific effect. Not just a couple of papers.Permalink
Mood +1 🙂
-
When we take decisions that will impact lives of millions, I’m not sure it’s overkill to have a look at the whole literature. The consequences are too large just to rely on a couple of papers.Permalink
Mood 0
-
What if instead on only relying on the IMF estimation of the Keynesian multiplier in Greece in 2010, we paid attention to several estimations instead of just one? IMF was wrong, and millions of people suffered the consequences.Permalink
Mood -4 🙁
-
And other fields already use meta-analyses. It’s not like a far reaching goal we may reach one day if we are lucky enough. The methodology is there, so we can build up on the work done in other fields to implement it "easily". But do we want to? That’s the whole question.Permalink
Mood +4 🙂
-
(Meta-analyses may also help to decrease the scientific influence based on affiliation instead of pure research, which would definitely improve economics too.)Permalink
Mood +5 🙂