Adrianto

Does Employee Ownership Hamper or Facilitate the Adoption of New Technologies?

with Avner Ben-Ner and Ainhoa Urtasun

Several studies have explored the determinants of firms’ decision to adopt robotics technology. For instance, firms are more likely to adopt robots in areas where there are more robot integrators  (Brynjolfsson et al. 2023) and robot users (Leigh et al. 2022). Moreover, labor cost, economies of scale, profitability, process complexity, and competition are other factors that significantly predict robot adoption; more profitable, less skill-intensive, and exporter firms are more likely to adopt robots in the future (Koch et al. 2021), and wages, level of routine work, the proportion of high education, the importance of scale of production, and international competition are all positively associated to high robot adoption (Fernández-Macías et al 2021).  

The identity of those who earn in a firm interacts with most of the factors that affect the benefits of robot adoption. However, the appears to be no evidence regarding the effect of ownership on the likelihood of adopting new technology, including robotics. There are reasons why ownership matters in affecting robot adoption. Robots are designed to substitute for human tasks in jobs that are primarily manual and routinized. They are also deployed in dangerous tasks to reduce work safety incidents or tasks requiring high precision. Nonetheless, they cannot fully replace human tasks (at least not yet), and thus, human involvement in nonautomated as well as new tasks—such as maintenance and quality control—is needed. Thus, substitution and complementarity work simultaneously, and the net effect depends on which one is stronger. This is of primary importance for workers as it directly impacts their employment prospects and well-being.

Employee well-being is pursued more vigorously by firms where employees have better representations to negotiate with the employer, such as in unionized firms or firms in which employees hold ownership rights through broad-based plans, such as employee stock ownership plans (ESOPs). Empirical evidence shows mixed findings regarding the former. For instance, a seminal work on union membership shows a negative effect on job satisfaction (Borjas 1979), but a recent study shows the opposite (Blanchflower et al 2022). In this case, workers and the employer sit on the opposing side of the table.

In ESOP firms, by granting ownership to workers, a broad-based ownership plan works through two different channels. First, it allows the potential to gain a slice of future profits and thus aligns the financial interests of both parties (workers and other owners). Second, employees are provided a representation of their voice through a trustee and ESOP committees. This is manifested in a wide range of employee satisfaction measures (Freeman et al. 2010, Adrianto et al. 2024).  The pursuit of employee well-being implies the adoption of technologies that advance worker safety, job satisfaction, employment stability, and financial results. These objectives may conflict with each other; for example, financial outcomes may be reduced, at least in the short run, by investment in technology that improves worker safety or by giving workers opportunities to advance their skills. Therefore, the adoption of technologies by such firms is likely to be different than that of firms without employee ownership that are likely to emphasize financial outcomes even if they come at the expense of others.

The first question addressed in this paper is whether the adoption of industrial robots is affected by employee ownership. This is what the Luddite movement did in England in the 19th century, during which textile workers opposed the use of cost-saving machinery capable of standardizing processes, which consequently reduced the need for and thus was detrimental to the existence of skilled artisans. Worker position to the introduction of automation technology continues in contemporary societies, although not uniformly so. Whereas in the United States, unions tend to be against new technologies, in Scandinavia, employers and unions appear to have forged agreements, whereby the gains from automation are shared, and union opposition to new automation has vanished. We investigate whether the better reflection of employee interests in ESOP firms and their choice of technology affect—delay or speed up—the decision to adopt robots.

In the second question, we ask whether employee ownership moderates the postadoption effects of robots on labor demand. Robots displace some tasks predominantly performed by lower-skill workers.  Robots also require additional human tasks of maintenance, programming and quality control, typically carried out by more skilled workers (the reinstatement effect). In addition, robots can produce the same output level more efficiently (the productivity effect), which can lead to business-stealing away from the competitors and generate labor demand to perform nonautomated tasks. With these opposing effects in mind, we ask whether hiring and employment differ between the two types of firms.

Sharing is Caring: Employee Stock Ownership Plans and Employee Satisfaction in U.S. Manufacturing

with Avner Ben-Ner, Jason Sockin, and Ainhoa Urtasun

Do employees fare better in firms they partly own? By examining workers' reviews of their employers on the website Glassdoor, we offer the first expansive comparison of employee satisfaction between firms in which workers own company shares through an employee stock ownership plan (ESOP) and conventional firms in which they do not. Focusing on production workers and managers in an industry-labor market matched sample within U.S. manufacturing, we find that employees report greater satisfaction in employee-owned firms overall and within specific aspects of jobs such as their firms' culture. Such differences in job quality cannot be rationalized by differences in skill demand and are greater when the ESOP is the product of collective bargaining. This work highlights how match quality can differ by ownership arrangement.

Robots and Work

with Avner Ben-Ner and Ainhoa Urtasun

This paper examines the effects of robot adoption on employment and skills in US manufacturing plants (2010-2022). Using a difference-in-differences method, we find a 150% increase in job postings and a 15% increase in employment in plants that adopt robots compared to non-adopters matched by industry and labor market. Increased competitiveness due to robots raises output with positive employment spillover effects in non-robotized stages of adopting plants and in upstream and downstream non-robotic plants within adopting firms. Skill composition remains largely unchanged, with more design, production, maintenance, repair, and programming skills required in robotized areas. Productivity and robot-human complementarity effects dominate displacement, with job losses limited to outcompeted non-adopters. Our multilevel approach reveals underestimation of robotization effects in prior firm-level studies. Industry-level employment effects are negligible due to counterbalancing gains and losses. Findings suggest the need for policies facilitating robot adoption while supporting displaced workers.

How things are made matters: The effects of technology on the organization of work

with Avner Ben-Ner and Ainhoa Urtasun

Given a particular product to produce, firms have several alternative production technologies from which to choose. This paper examines the effect of production technologies, directly and indirectly through complexity and task interdependence, on outcomes essential to the organization of work. Our study uses online job vacancy postings in the U.S. manufacturing sector during 2017-2021 to analyze technical occupations (i.e., engineers, technicians, and operators) in plants that implement one of six primary technologies: subtraction, forming, molding, additive manufacturing, chemical, and assembly. Controlling for different forms of automation, location, and other factors, we find that the differences in the division of labor, specialization, and span of control among technologies are driven by differences in complexity. Additive manufacturing, chemical, and assembly are technologically more complex than forming, molding, and subtraction, and, as a result, they need more jobs to be designed, more tasks and skills to be bundled into a job, and fewer employees to be overseen by a manager. Moreover, each technology exhibits a distinct pattern of two forms of task interdependence—reciprocal and sequential, and therefore the effects on the three outcomes are more nuanced.

• Best paper in the 84th Annual Meeting of the Academy of Management: "The Effects of Robots on the Workplace" (2024)
• PhD Student Teaching Award, Carlson School of Management (2023)
• Best Participant in the Organization Development training, Oranye Development (2017)
• Highest Distinction, Master of Science in Public Policy and Management, Carnegie Mellon University (2014)

• Summer Fellowship, Carlson School of Management (2024)
• Ph.D. Student Conference Travel Fellowship, Carlson School of Management (2024)
• The Employee Ownership Foundation/Louis O. Kelso Fellowship (2023-2024)
• Ph.D. Student Conference Travel Fellowship, Carlson School of Management (2023)
• Ph.D. Student Conference Travel Fellowship, Carlson School of Management (2022)
• Travel grant for the WAIM Conference, the Work in the Age of Intelligent Machine Research Coordination Network (2022)
• PhD Student Conference Travel Fellowship, Carlson School of Management (2022 & 2023)
• Financial Education and Training Agency (FETA) Scholarship (2020-2024)
• Scholarship Program for Strengthening the Reforming Institution (2012-2014)

• Ad-hoc reviewer, Organization Science (2024)
• Ad-hoc reviewer, the Annual Meeting of the Academy of Management (2024)
• Discussant in the Kelso Workshop, New Brunswick, NJ (2024)