Do Temporary Workers Face Higher Wage Markdowns? Evidence from India’s Automotive Sector
Davide Luparello · 2025
Read Do Temporary Workers Face Higher Wage Markdowns? Evidence from India’s Automotive Sector by Davide Luparello (2025) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/arxiv-2507-08222.txt.
Introduction
Luparello wrote a solo study in twenty twenty-five on if short-term workers in India's car trade face a worse wage squeeze than permanent staff. Short-term staff are half the trade's work force and earn far less than permanent staff. He builds a tool from plant data that spans near two decades to split that gap.
His model has choice-based work supply by worker type and plant-union talks for permanent staff. The whole pay gap comes from higher yield of permanent staff, he finds, not from a worse squeeze. A flat pay-back for the squeeze would raise both groups' gains but widen the gap too. That trade-off shapes any fix.
Catalog record
| CREATOR | Davide Luparello |
|---|---|
| YEAR | 2025 |
| COLLECTION | PAPERS |
| FORMAT | TXT |
| FILE SIZE | 117 KB |
| VERSION | 1.0 |
| LICENSE | http://creativecommons.org/licenses/by/4.0/ |
| RIGHTS | CC BY |
| SOURCE | arXiv |
Source: https://arxiv.org/abs/2507-08222
Browse the full library · About the Tower of Babel
Full text
The opening of the work, from the library's hosted text:
Do Temporary Workers Face Higher Wage Markdowns? Evidence from India’s Automotive Sector Davide Luparello Abstract Are temporary workers subject to different wage markdowns than permanent workers? This paper examines productivity, output markups, and wage markdowns in India’s automotive sector during 2000–2020. I develop a model integrating CES production, nested logit labor supply, and differentiated labor market conduct: Nash-Bertrand wage setting for temporary workers versus Nash bargaining for unionized permanent workers. Results reveal declining output markups as marginal costs outpace prices through productivity deceleration. Rising labor-augmenting productivity cannot offset declining Hicks-neutral productivity, reducing overall TFP. Labor market power substantially compresses worker compensation: wage markdowns persist at 40% for temporary workers and 10% for permanent workers. KEYWORDS: Markdowns, Markups, Productivity, India JEL CODES: L11, L13, L62 Introduction India’s automotive industry constitutes a cornerstone of the nation’s manufacturing ecosystem 1 1 1 Manufacturing’s share of GDP remains notably below that of comparable economies at similar development stages, a phenomenon that Lamba and Subramanian 2020 characterizes as ”premature deindustrialization” and ”precocious servicification.” . As of 2025, India ranks as the world’s fourth-largest automobile producer, with the sector contributing 7.1% to GDP and accounting for 49% of total manufacturing output. The industry relies heavily on low-cost temporary contract workers who, despite possessing similar qualifications and performing comparable tasks to permanent staff, earn substantially less and lack unionization rights ( Duvisac 2019 ) . This pronounced duality in employment arrangements raises a fundamental question: do temporary workers in India’s automotive plants experience larger wage markdowns than their permanent counterparts? To address this question, I compare marginal revenue products with observed wages across worker types. I develop a structural model that integrates production, labor supply, and labor-market conduct differentiated by worker type. Production follows a nested constant elasticity of substitution (CES) technology, with the outermost nest aggregating capital, intermediate inputs, and total labor. Labor enters as the composition of two variable inputs—white-collar and blue-collar workers—with blue-collar labor further divided into permanent and temporary employees. Output incorporates both Hicks-neutral and labor-augmenting productivity components, each evolving according to an independent Markov process. Consistent with established practice 2 2 2 Tracing back to the treatment of the transmission bias in Marschak and Andrews 1944 and Klette and Griliches 1996 . , firms choose input quantities with complete information about current productivity. Production data and optimality conditions alone prove insufficient to separate unobservable coordination costs from wage markdowns for blue-collar workers. I therefore build a labor supply framework that allows to identify elasticities for each worker type. Blue-collar workers, given their exogenously determined employment status—temporary or permanent—select among available workplaces to maximize expected utility, which comprises wages, workplace attributes, and type 1 extreme value taste shocks correlated at the state level that capture localized market conditions. This discrete choice specification yields standard nested logit choice probabilities 3 3 3 The approach of using Berry 1994 methodology to specify labor supply equations appears in Azar, Berry, and Marinescu 2022 and receives recent advancement in Rubens, Wu, and Xu 2024 . . Finally, I link labor demand and supply through conduct specifications: temporary workers receive compensation via Nash-Bertrand wage setting, while permanent workers negotiate wages through Nash bargaining between unions and plants. Knowledge of inverse labor supply elasticities and bargaining parameters enables separation of wage markdowns from coordination costs. My empirical strategy proceeds in distinct phases. First, I estimate CES substitution parameters and labor-augmenting productivity using GMM, leveraging the ratio of inputs’ first-order conditions and labor-augmenting productivity shocks for identification. Second, I estimate CES share parameters and the Solow residual 4 4 4 The composite measure of Hicks-neutral productivity and output measurement error. through GMM, utilizing the production function and Hicks-neutral productivity shocks for identification. Third, I apply Kalman filtering and smoothing techniques to decompose the Solow residual into Hicks-neutral productivity and output measurement error ( Hong and Luparello 2024 ) . Fourth, I derive marginal costs from the estimated parameters and compute output markups. These four steps yield all production parameters and productivity variables along with output markups. To separate wage markdowns from unobservable coordination costs, first I estimate blue-collar labor supply functions across worker categories using 2SLS techniques that instrument both endogenous wages and nested shares. I then estimate the permanent workers’ bargaining parameter using 2SLS that instruments the total surplus ratio. Results reveal declining output price markups—the Lerner index 5 5 5 The share of output price paid that does not remunerate marginal costs of production. falls from 60% to 40% over the period 2000--2020, mirroring broader profitability decreases. This decline stems from marginal costs outpacing output prices due to productivity deceleration: despite rising labor-augmenting productivity, declining Hicks-neutral productivity reduces overall TFP. Blue-collar wage markdowns 6 6 6 The share of workers’ marginal revenue product captured by the employer through market power. remain substantial despite rising wages: consistently 40% for temporary workers and stable at 10% under Nash bargaining for permanent workers 7 7 7 Or declining from 30% to 20% under limit Nash-Bertrand conduct. . Finally, the fall in labor share exhibits four times (temporary workers) or eight times (permanent workers) greater sensitivity to increases in labor market power than to output market power. This paper makes three principal contributions to the literature. First, the analysis advances the domestic outsourcing literature by demonstrating that plants’ incentives to hire temporary workers through contractors rather than permanent employees stem from higher market power facilitated by reduced union protection. These findings align with Autor 2003 and Estefan et al. 2024 , who demonstrate that enhanced worker protection may simultaneously reduce exploitation—the intended effect—while fostering employment of unprotected worker categories—the unintended consequence. My results also complement previous research documenting associations between domestic outsourcing and lower wages ( Dube and Kaplan 2010 for US janitors and guards, Drenik et al. 2023 for Argentina, Weil 2014 , Felix and Wong 2024 for security guards in Brazil), increased firm rent capture ( Appelbaum 2017 ), and heightened wage inequality ( Bilal and Lhuillier 2021 for France, Goldschmidt and Schmieder 2017 for food, cleaning, security and logistics services in Germany). Second, the paper introduces an estimator that integrates imperfect output market competition, factor-biased technological change, and heterogeneous factor market conduct across different inputs. This advancement extends the production approach to markup estimation literature 8 8 8 While this paper analyzes a specific industry, a substantial complementary macroeconomic literature examines labor market power within general equilibrium frameworks ( Berger, Herkenhoff, and Mongey 2022 ; Berger et al. 2024 ; Berger, Herkenhoff, and Mongey 2025 ) . —which typically operates under Hicks-neutrality assumptions ( De Loecker and Warzynski 2012 ; Morlacco 2019 ; Brooks et al. 2021 ; Mertens 2022 ; Rubens 2023 ) —while complementing research on directed technical change under competitive factor markets ( Doraszelski and Jaumandreu 2018 ; Zhang 2019 ; Demirer 2020 ; Raval 2023 ; Hong and Luparello 2024 ) and studies employing Berry 1994 share inversion to measure labor monopsony power ( Azar, Berry, and Marinescu 2022 ; Rubens, Wu, and Xu 2024 ) . Unlike Rubens, Wu, and Xu 2024 , my framework permits different conduct assumptions across production inputs, which is a substantial feature of the Indian automotive industry, and incorporates output measurement error. I disentangle this error from Hicks-neutral productivity