Vidhya Soundararajan is a PhD candidate in Cornell’s Dyson School and is currently on the job market.
Firms are hiring more ‘contract workers’ by means of temporary fixed-term contracts mediated by third-party staffing agencies. With few binding regulations protecting contract workers, savings on labor costs (including wages, benefits, firing costs and time spent in disputes) remains the top reason for hiring them (Abraham and Taylor 1996). Researchers and policy makers recognize the inequitable terms dictating contract work (e.g., contract workers take a wage penalty among other poor terms) and investigate whether, at the very least, such work could improve the future earnings and employability of the workers. However, evidence here is inconclusive, varying across countries, sectors, and type of worker (Autor and Houseman 2010; Jahn 2013).
Notably missing from this discourse, particularly for developing countries, is how firms themselves are potentially adversely affected by excessive dependence on poorly paid and discontent contract workers, who have little incentive to devote themselves and contribute productively to the firms where they work (Panagariya 2004; Basu, Chau, Soundararajan 2015; Hirsch and Mueller 2010).
My research provides the first set of empirical estimates of the effects of contract work on firm productivity in any developing country. Strikingly, results indicate negative productivity effects from the excessive usage of contract workers.
Why India?
India, between the years 1998-99 and 2010-11, offers an interesting opportunity to study the effects of contract work on firm productivity for the following reasons: (1) The country witnessed an increase in the popularity of contract workers in the manufacturing sector, evident from the increase in their share of employment from 13% to 33%; (2) India’s manufacturing sector Gross Domestic Product attained new highs; (3) the wage ratio between regular and contract workers performing core manufacturing jobs within the same factory stayed around 1.4, indicating persistent wage penalty for contract workers; and (4) simultaneously, this period also saw a rise in strikes and protests by discontent contract workers.
Methodology
Productivity impacts of contract work are obtained by estimating a production function and a productivity evolution equation. I adopt Olley and Pakes (1996) (OP)’s structural method for estimations. Unlike OP’s exogenously evolving productivity growth process, my model allows productivity to grow endogenously (Doraszelski and Jomandreu 2013) based on firms’ own choices of current and lagged contract labor share. My Generalized Method of Moments estimator uses production inputs and/or their lags as instruments based on the timing restrictions that underlie the firm’s dynamic profit maximization model. Exogenous variation in current and lagged contract labor share is obtained using rainfall shocks and employment protection indices as instruments. Estimations use a 13-year firm level panel data set covering the years from 1998-99 to 2010-11 from India’s Annual Survey of Industries.
Productivity costs of contract workers
Results indicate that, on average, contemporaneous productivity effects of contract work are positive, suggesting that the gain in flexibility achieved by hiring contract workers contributes positively to productivity. However, in all industries, a section of firms face negative contemporary effects in employing contract workers, signifying that the low effort exerted by these workers could potentially lead to low productivity outcomes.
In contrast to contemporaneous effects, average productivity effects after one period (dynamic effect) are negative. These adverse dynamics indicate that contractual work arrangements that typically last only for a fixed term potentially hinder the accumulation of firm specific human capital, consequently affecting productivity. Further, most industries witness potential undesirable dynamic effects in employing excessive contract workers: that is, productivity increases with an increase in lagged contract share up to a threshold share, but declines with an increase in share beyond the threshold.
Implications for policy and future agenda
My research, along with other parallel bodies of work, shows that high workplace inequality and poor industrial relations contribute to a decline in productivity as well as product quality. Recognizing this, core industrial and labor policies in emerging economies (example: Make In India), should not only be tuned to attract investment (as they often do), but also focus on augmenting labor productivity in a step to enhance the living standards of poor workers.
What options do policy makers have in achieving this goal? If productivity losses are channeled through high workplace inequality, does setting a wage floor for contract workers improve productivity outcomes? On the other extreme, should we altogether prohibit contract workers from doing manufacturing jobs and restrict them to jobs peripheral to the establishment, such as cleaning and catering, like the amendment by the Indian state of Andhra Pradesh in 2003? My future research will follow up on these questions using reduced-form policy evaluation techniques.
