This is the ninth in is year’s series of posts by PhD students on the job market.
Firms in developing countries report rates of worker absence of between 5 and 20%. According to managers, the volatility of labor, driven by worker absenteeism, is one of their chief concerns. How bad is this for firms?
In my job market paper, I first examine the impact of worker absence on firm revenue and profit. I use detailed panel data on worker absence, worker turnover, and firm outcomes, for formal Indian manufacturing firms. I find that these firms face significant costs due to absence – comparable to the costs imposed by lack of access to credit and price shocks to other inputs.
In the policy space, labor regulations that protect workers are common (Botero et al. 2004; Micco & Pages 2006). As an unintended consequence of these regulations, firms may struggle to respond optimally to worker absence: firms state that they are constrained in their ability to hire and fire workers, for example.
In the second part of the paper, I exploit the variation in labor regulations across Indian states to study how regulatory constraints can influence the way firms cope with worker absence. Firms in states with less stringent labor regulations are able to mitigate between a third and half of the negative impact of worker absence. They do this by increasing worker separation and hiring more permanent workers as replacements (rather than temporary workers).
How bad is worker absence for firms?
To estimate the impact of worker absence on these firms, I use two strategies. First, regressions with factory fixed effects allow me to compare a firm’s outcomes and worker turnover in a year in which worker absenteeism is high with other years in which it is low. Second, I instrument for worker absence using an exogenous shift in workers’ cost of coming to work - the number of extreme rainfall days they experience – and estimate the resulting change in firm outcomes. This is likely to affect a broader swathe of workers, which helps account for the fact that the average absent worker may also be less productive. The magnitudes of these estimates are likely closer to the true cost of absence for firms. In the paper, I also use various methods to account for the direct impact of extreme rainfall days on firms.
Firms cannot find enough replacement workers to take the place of those who are absent: firms work fewer mandays in years in which worker absence is high (0.14% fewer worker-days for a 1% increase in worker absence). This results in a significant decrease in firm revenue and profits (0.21% and 0.3% respectively, for a 1% increase in worker absence). For a sense of magnitude, I compare my results to extensive work on other obstacles to firm growth: access to credit (Banerjee and Duflo 2014; De Mel et al. 2008), demand (Adhvaryu et al. 2013), and other input shocks (Abeberese 2017). My estimates of the effect of worker absence on firm output and revenue are comparable (and in some cases, even higher). Worker absence is at least as important to consider, both for firms and for public policy, as these other concerns.
How does less restrictive labor regulation influence how firms manage absenteeism?
In India, the Industrial Disputes Resolution Act (IDRA) of 1947 governs the interaction between firms and workers on a federal level. It encompasses a broad range of laws regulating the interaction between employers and employees, including the ability to settle disputes in labor courts, the hiring and firing of workers, the ability of workers to create and join unions, and workers’ rights in the event of firm closure. States have made various amendments to the law, resulting in state-level variation in cost of hiring and firing workers. I refer to states with fewer state-level restrictions (and therefore lower firing costs) as proemployer states.
Worker absence is 31% lower in proemployer states, relative to an 8.8% rate of absence in other states – a motivating fact that suggests that looser regulations may give firms mechanisms to deter workers from being absent.
More hiring and firing (of permanent workers)
Firms in proemployer states adjust to worker absence by firing and hiring more permanent workers relative to firms in other states. These firms experience 27% more worker separation and 58% more hiring of permanent workers in the month following an increase in worker absence.
This makes intuitive sense for three reasons. First, it is cheaper to fire absent workers in proemployer states. Second, the replacement permanent workers are of unknown worker-type to the firm, and are more easily let go if they turn out to be less productive. Third, the new permanent workers are likely to be absent less often than their counterparts in other states (because of the deterrent effect of higher rates of firing by the firm).
The pattern of the separations (a spike after one month following the increase in worker absence) aligns closely with the 30-day notice period all employers are required to give permanent workers.
The higher rates of separation and hiring hold only for firms with more than 100 employees, that are bound by the relevant sections of the IDRA, and not for smaller firms. This suggests that the pattern of results a) is not driven by voluntary separations, and b) is likely attributable to the IDRA, rather than to other fundamental differences between proemployer and other states.
Less reliance on temporary workers
Firms rely heavily on temporary (or contract) workers: these workers contribute 42% of total worker-days for the median firm. Firms’ agreements with labor contractors are set up so that temporary workers are considered entirely interchangeable (e.g. “I will need X workers for 5 days next week”) – absence is therefore not relevant for this group of workers. They are also not subject to the labor regulation codified in the IDRA. On the other hand, these temporary workers are less productive, on average (Bertrand et al. 2017).
We know from the previous result that firms in proemployer states are more willing to replace workers who are fired after periods of higher absence with more productive permanent workers.
Firms in other states rely more on new temporary workers instead, and hire 35% more temporary workers (relative to firms in proemployer states) in response to higher absence among permanent workers.
And better firm outcomes
What, then, is the overall benefit to firms of lower regulatory constraints?
Firms in proemployer states have smaller decreases in days worked (only 46% of the effect for firms in other states), revenue (67% of the effect for other firms), and profits (57% of the effect for other firms), in response to increases in worker absence. This suggests that less restrictive labor regulation that allows for higher worker turnover can mitigate between a third and a half of the costs of worker absence.
The finding that worker turnover is associated with better firm outcomes in the context of high worker absenteeism stands in contrast to an older literature that finds that worker turnover, in general, is bad for firms (Hancock et al. 2013 and Hausknecht and Trevor 2011 have meta-analyses of such studies).
What then?
While this paper focuses only on the constraints imposed by labor regulations on firms, these policies could have large welfare benefits for workers. These laws are meant to protect workers’ rights, and may well increase worker surplus by codifying how economic rents are shared between firms and employees.
However, my findings suggest that the consequences of worker absenteeism for firms can be significant, and that such policies may hinder firms’ ability to respond optimally to worker behavior. This results in worse firm outcomes. Arguably, firms’ greater reliance on temporary (contract) work arrangements driven by greater labor regulation may also reduce workers’ welfare.
More broadly, these findings are especially important when considering the growing impact of aging populations on worker absence in developing countries (Rasmussen et al. 2016). Cross-country projections up to 2030 suggest that between 0.8 and 1.6% of developing countries’ GDP will be lost to absence. As this becomes an increasingly serious problem for many countries, policy-makers will have to pay careful attention to the potential unintended consequences to firms of additional labor protections.
This paper also raises questions about the root causes of worker absence. This is the focus of related ongoing work with coauthors Suanna Oh and Yogita Shamdasani, where we are examining how workers’ need for flexibility in work arrangements affects their labor supply decisions, including absence.
Nandita Krishnaswamy is a post-doc in the Department of Economics at the University of Southern California.
Join the Conversation