And that's in countries with relatively reasonable hiring practices, like the US.
In Spain it's even worse: for positions at universities there are so-called "objective" scales of assessment, so that a paper scores a number of points depending on the position of the journal it's published in in the ISI JCR journal ranking. Points are added and the top scoring candidate wins.
For hiring postdocs in projects, the PI typically has some more leeway to set the assessment scales, but not too much. As a PI who is hiring right now, I can use an interview or personal assessment of candidates' work to tip the scales a bit (at my place up to 15% of the grade or so, little more than a tiebreaker) but if I get a candidate with significantly more JCR-indexed papers than another, I basically have to hire them even if I think their track record is worse (which is often the case as one can get lots of "points" by submitting tons of very similar papers on an easy-to-publish area). Because doing otherwise is, you know, not "objective".
So here, basically, if you publish in a discipline where JCR-indexed journals are dominated by the likes of Elsevier, you absolutely need to publish there or not get hired, period. When people in my field from more fortunate countries say things like "let's just ditch the journals and publish everything in arXiv", I would agree with them in an ideal world, but in my world they would push my whole country out of the field...
Well, you're talking of a publicly-owned university, and the perception of "mamoneo" and "amiguismo" funded by taxpayer money is, in my opinion, already high enough that I welcome anything that tries to make the hiring system more objective.
Mind you: The hiring policy at Google is that managers don't get to interview the potential hires they will manage, nor have any say in the hiring process. For technical positions, five engineers interview the candidate independently, and submit their feedback and recommendation to a hiring committee that makes the decision. The process tries to be as objective as possible, and the manager has no leeway at all.
That doesn't mean the Spanish system is any good, just that I don't think giving the PI more say is a great solution. The points assigned to "easy-to-publish areas" can be reduced if PIs give that as feedback. A way to evaluate a candidate without the numerical scale silliness could be devised too.
What I am against is the rigid numerical scales. There are always ways to game them (even if easy-to-publish areas can be flagged, which would be politically difficult to do, there is the problem of people who manage to publish almost identical work in several journals, the problem of ghost authorship, etc.) I believe it's basically impossible to judge a researcher's work without actually reading it.
That said, indeed many PIs can't be trusted to make decisions based on meritocracy, as there is the perception you mention and it's often justified. Probably the CVs should be evaluated or at least checked by someone different from the PI (for example by experts in the field from a different country, I think in Italy they do that kind of evaluation for some projects and grants). I just expressed some frustration as PI because I may be forced to hire a suboptimal candidate due to the rigidness of the scales, but what I want is not necessarily more autonomy for hiring, but simply any solution (like external expert evaluation) that will (1) give me the best candidate, and (2) not support the oligopoly of the likes of Elsevier like the Spanish system does.
In Spain it's even worse: for positions at universities there are so-called "objective" scales of assessment, so that a paper scores a number of points depending on the position of the journal it's published in in the ISI JCR journal ranking. Points are added and the top scoring candidate wins.
For hiring postdocs in projects, the PI typically has some more leeway to set the assessment scales, but not too much. As a PI who is hiring right now, I can use an interview or personal assessment of candidates' work to tip the scales a bit (at my place up to 15% of the grade or so, little more than a tiebreaker) but if I get a candidate with significantly more JCR-indexed papers than another, I basically have to hire them even if I think their track record is worse (which is often the case as one can get lots of "points" by submitting tons of very similar papers on an easy-to-publish area). Because doing otherwise is, you know, not "objective".
So here, basically, if you publish in a discipline where JCR-indexed journals are dominated by the likes of Elsevier, you absolutely need to publish there or not get hired, period. When people in my field from more fortunate countries say things like "let's just ditch the journals and publish everything in arXiv", I would agree with them in an ideal world, but in my world they would push my whole country out of the field...