Richard Sever, [writing for the *Times Higher Education Supplement*](https://www.timeshighereducation.com/opinion/academics-should-stop-fixating-editorial-power-fallacy), with the argument that the tyranny of journal impact factors—with all their career-determining arbitrariness—is a symptom of a bigger problem: The small number of academic jobs for PhDs in the life sciences. He's skeptical of the alternatives to the impact-factor shorthand:
> It is right to lament the fact that where someone publishes has become more important than what they publish. But judging people on where or with whom they trained would arguably be worse. Article-level metrics are little better: they can be gamed and inevitably become targets. And anyone who thinks the assessment system should work more like Yelp or eBay ratings probably needs to use Yelp or eBay a bit more.
It's a good point, that other reputation proxies have their own problems--and that aping the app reputation economy, in particular, is a dangerous mistake. Of course the journal impact factor is uniquely unsuited to article-level assessment, but Sever's point is that the underlying problem is the need to sift out the vast reserve army of post-docs:
> Real change will come only with a restructuring of academic careers. Forward-thinking PhD programmes that prepare candidates for a variety of different potential careers are a good start. A bolder and perhaps more necessary step would be to admit the fiction that a postdoc is a training position for group leadership and to create more permanent staff positions below principal investigator level.
Sever's right here too, but the "bolder" label for a frank acceptance of the current system seems a tad defeatist. If the underlying academic system is the problem, then a "bolder" step would be to question *that*.