¹⁸Guvenen (2007) refers to a process like this one as a ‘restricted income process’ (RIP) as distinguished from a process that he proposes which is similar but which allows each individual to have a distinct idiosyncratic mean growth rate. Guvenen’s argument that each household has its own growth rate is intuitively plausible (indeed, it occurred to earlier authors who tested and rejected it), but Hryshko (2010) argues that there is no evidence that the Guvenen income process describes the data better (in a quantitatively meaningful way) than the restricted income process. Since incorporation of Guvenen’s income process introduces serious modeling difficulties, it seems prudent to avoid using it unless the evidence for idiosyncratic growth factors becomes much more compelling.