Dear Art,

Let's start with moral outrage and use it to energize a search for a 
"better way."  I of course agree that we use signals, and must, or else 
we would instantly drown.  Prudence is one of the virtues.  For example, 
I use the author signal, as we all do: anything by Art Diamond or Bruno 
Frey is worth reading.  And I do to some (i.e. the exactly optimal) 
degree use the journal test: I never, for instance, read anything in 
Econometrica, because I know from long experience that nothing published 
there makes a scientific contribution to economics--contributions to 
significance testing, yes, and to theorem proving, sure (you yourself 
have written most illuminatingly on the very point); but not to 
science.  I've had much better luck reading the Pakistan Development Review.

But the actual-reading choice procedure is to read the summary and the 
first page or so of an actual paper, glancing perhaps over the tables 
and the bibliography, and then  to read the whole thing only if I see 
signs of economic science going on there.  I admire the idea of 4 
submitted papers as in Britain.  We should do that for promotion and 
appointment in North America.  But what I'd want to have in a stack in 
front of me would be 100 such papers (that's about all I could handle if 
I were on such a board) of the typescripts, with names and journals 
removed.  Then I'd grade 'em like student scripts, and go into a meeting 
with colleagues who had read an overlapping sample as though into a 
degree meeting in Britain, ready to argue the case for this or that 
person in serious scientific detail: one or two Firsts; a select bunch 
of Upper Seconds; more Lower Seconds; a passel of Passes; and a number 
of Fails---such as the average paper in Econometrica.

What's the merit?   We would then write for each other, seriously 
persuasively, as historians on the whole do (because they on the whole 
read [quickly: see above] whole books of candidates), instead of for the 
depraved "standards" (as Bruno noted) of anonymous referees with no real 
stake in the scientific issue, except perhaps to stealthily kill off 
unusual ideas.

Warm regards,

Deirdre McCloskey