Skill complementarity in production technology: new empirical evidence and implications
(Andrey Stoyanov and Nick Zubanov)
Abstract:Matched worker-firm data from Danish manufacturing reveal that 1) industries differ in within-firm worker skill dispersion, and 2) the correlation between within-firm skill dispersion and productivity is positive in industries with higher average skill dispersion. We argue that these patterns are a manifestation of technological differences across industries: firms in the ``skill complementarity'' industries profit from hiring workers of similar skill level, whereas firms in the ``skill substitutability’' industries benefit from hiring workers of different skill levels. An empirical method we devise produces a robust classification of industries into the distinct complementarity and substitutability groups. Our study unveils hitherto unnoticed technological heterogeneity between industries within the same economy, and demonstrates its importance. Specifically, we show through simulations on a simple general equilibrium model that failing to take technological heterogeneity into account results in large prediction errors.
What do employee referral programs do?
(Guido Friebel, Matthias Heinz, Mitchell Hoffmann, and Nick Zubanov)
Employee referral programs (ERPs) are randomly introduced in a grocery chain. Larger bonuses increase referrals and decrease referral quality, though the increase in referrals is modest. Still, ERPs are highly profitable, partly, because referrals stay longer than non-referrals, but, mainly, because non-referrals stay longer in treated stores than in control stores. In a post-RCT firmwide ERP rollout, referral rates remain low for grocery jobs, but are high for non-grocery jobs, which are perceived as more attractive. Our results (1) are consistent with referral-making being driven by money and altruism toward friends; (2) show that ERPs can have substantial benefits beyond generating referrals. The most-supported mechanism for (2) is that workers value being involved in hiring.
Downsizing announcements and workplace performance.
(Guido Friebel, Matthias Heinz, Ingo Weller, Nick Zubanov)
We estimate the effect of downsizing announcement on workplace performance using data from a German bakery chain of 193 shops. Faced with intensified competition, the firm decided to sell or close down 57 of its worst performing shops. We identify the effect of downsizing from a plausibly exogenous variation in the timing of the sale or closure announcement in the individual shops. We find that the announcements of the shop being sold to a new owner and being closed down reduce sales by six and 21 percent, respectively. The negative effect of downsizing increases with the share of workers with a permanent contract, even though permanent workers faced a much lower unemployment risk. We relate our findings to the literatures on downsizing and psychological contract.
Making managers matter
(Guido Friebel, Matthias Heinz, and Nick Zubanov)
Middle managers are the backbone of modern organizations. Can they be persuaded to improve performance outcomes without additional direct incentives? We address this question within a field experiment in a retail chain of 238 stores that was struggling with high store employee turnover. Our treatments were 1) to ask store managers on behalf of the CEO to do what they can to reduce turnover; 2) to raise employee awareness of career and development opportunities; and 3) the two combined. The first treatment and the combination of the two reduce turnover by a third - an effect stable for at least eight months after the treatments began. The second treatment produces a smaller and borderline-significant effect. From a subsequent telephone survey of store managers we find that managers in the treated stores increased their efforts to tackle turnover over the treatment period; in particular, they communicated more with their employees. It is noteworthy that making managers matter does not require additional incentives or any other organizational change.