How can the quality of information flows be improved? 

Y finds that 60% of anonymous referees add value; 30% don't; 10% are entrant-rejecting incumbents: Neoclassical Expected Utility Maximisers (NEUM). 

Three anonymous NEUMs unintentionally inspired the 1998 HER Minisymposium on Editing and Refereeing (Goodwin, C., R. Leeson, W. Samuels, and V. Tarascio):

1. As journal editor, Warren Samuels gave a SHOE subscriber the BOOT (effectively out of the profession) after discovering that he or she had made fraudulent assertions in an email accompanying an anonymous self-serving rejection.

2. A prediction about 'The Rise and Fall and Fall of Keynesian Economics?' examined the almost non-existent epistemological foundations of the Phillips curve trade-off provided by the serving vice chair of the Federal Reserve. 

An anonymous referee insisted that all references to the serving vice chair of the Federal Reserve be deleted or condensed into a footnote and the article "rewritten along the following lines ..."

3. An examination of the Patinkin-Friedman correspondence was rejected by an HET journal on the basis of a two-sentence anonymous referees report: "One day Y might write something worth reading. That day has not yet arrived".

Proposed solution 

With the submitting author's permission, a referee should choose between having their report published on the journal's website 

a. immediately upon publication or rejection; or

b. two years later. 

Even with anonymity, this would change incentives: cheap shots would become more expensive (pattern recognition would allow names to be attached to NEUMs).

Sunlight illuminates monetary policy - shouldn't the Eleusinian Mysteries of the lumpen professoriate be restrained by a fundamental principle of law: procedural fairness?    

I authorise the publication of all my referee's reports.