Les Hatton and Gregory Warr in Times Higher Education:
Given the entirely appropriate degree of respect that science has for data, the ongoing discussion of peer review is often surprisingly data-free and underlain by the implicit assumption that peer review – although in need of improvement – is indispensable.
The thing is, the peer review of scientific reports is not only without documented value in advancing the scientific enterprise but, in a manner that few care to acknowledge openly, primarily serves ends that are less than noble. Peer review is widely assumed to provide an imprimatur of scientific quality (and significance) for a publication, but this is clearly not the case.
While the many flaws of peer review are clearly laid out in the literature, its failure to protect the integrity of the scientific enterprise is notable. An estimated cost of irreproducible biomedical research is $28 billion (£20 billion) a year and“currently, many published research findings are false or exaggerated, and an estimated 85 per cent of research resources are wasted”, one paper found.
A prime example of the failure of peer review is the tainting of a significant segment of the biomedical literature by the use of misidentified and contaminated cell lines pointing, at best, to a culture of carelessness in cell biology research and the clear failure of peer review to discover and correct erroneous research.
There are many reasons why scientific peer review is ineffective.
More here.