Firm decisions to use work sharing unemployment insurance

Although work sharing unemployment insurance (WSUI) allows firms to retain valuable employees while avoiding the other hiring and firing costs associated with layoffs, firms may not prefer WSUI over layoffs. First, the UI tax costs of WSUI should be at least as high as the UI tax costs of layoffs for a comparable labor input reduction, and these costs should vary significantly across industries and firms. Second, WSUI should only be effective in retaining valuable employees during temporary negative demand shocks, because during longer shocks affected workers have the option of finding new employment at fulltime hours with other firms. I build a simple theoretical labor demand model to formalize the firm's WSUI decision, and I then use administrative unemployment insurance data from the state of California to test the prediction that self-selection via the relative UI tax costs of work sharing versus layoffs affects firm decisions to use WSUI. Specifically, I formulate a simple simultaneous equations model with both qualitative and limited dependent variables, which can be estimated using the familiar two-step procedure popularized by Heckman (1979). The results suggest that selection based upon the relative tax cost of WSUI versus layoffs is an important determinate of a firm's decision to use the WSUI program. I also test whether firms with shorter expectations about the length of the demand shock are in fact more likely to use WSUI. These results show that once selection is accounted for, firms that expect demand shock to last from 1 to 3 months are more likely to use WSUI than firms with longer expectations. Finally, accounting for selection also highlights the fact that a significant portion of the reduced-form effect that some variables have on a firm's likelihood of work sharing is actually due to the relative UI tax costs of WSUI versus layoffs. For example, although reduced-form probit estimates suggest that unionized firms are significantly more likely than non-unionized firms to use WSUI, this effect is entirely due to the lower relative UI tax costs of WSUI versus layoffs faced by unionized firms.