Maybe there's selection bias, where the companies who downsized were doing worse than the companies who didn't. So if both sets of companies ended up the same,

Maybe there's selection bias, where the companies who downsized were doing worse than the companies who didn't. So if both sets of companies ended up the same, and one set was initially doing worse than the other, then that'd mean layoffs worked