Racial Discrimination in Sourcing: Evidence from Controlled Experiments
Author(s):
Karca D. Aral (Whitman School of Management, Syracuse University) and Luk Van Wassenhove (INSEAD)
Journal:
Production and Operations Management, forthcoming.
Summarize the essence (findings or implications) of your paper:
Discriminatory racial biases can significantly impact sourcing decisions.
Research Questions:
1) Can racial discrimination affect sourcing decisions?
2) If so, what is the magnitude of its impact?
3) What can be done to address racial biases in a sourcing context?
What We Know:
The survival of small minority-owned businesses crucially depends on their ability to win contracts from buyer firms. Hence, the existence of racial discrimination in sourcing would indicate that entrepreneurship, which is widely deemed as an important tool for addressing racial wealth disparity (Bradford and Osborne, 1976; Borjas, 1999; Boston, 1999, 2006; Bradford, 2003; Fairly, Robb, and Robinson, 2022), cannot be an organic economic leveling mechanism left on its own devices. On the other side of the spectrum, the absence of discrimination in sourcing would mean that market forces may eventually remedy the economical dimensions of racial inequality. Clearly, answering whether discrimination exists in a sourcing setting is of extreme importance to help create and guide related policy decisions and managerial guidelines.
Arguably, a formal and quantified supplier selection process leaves no room for discrimination to affect sourcing decisions. In fact, supplier selection decisions for critical supplies at large-scale corporations are formalized requiring, for example, 3rd party supplier assessments and the use of scoring rules for total-cost ranking (Aral, Beil, and Van Wassenhove, 2021). However, many purchase decisions are less formal. In practice, procurement managers are typically responsible for a category of supplies managing hundreds of suppliers and placing a few thousand orders a year. Among these, generally, only few critical supply categories have formalized procurement processes tied to incentive schemes (if any). In fact, tracking procurement and setting up related incentive schemes in sourcing is notoriously difficult in practice (e.g., Mercker et al., 2014; Bader, 2023). This leaves procurement managers without supervision on supplier selection decisions for many supply sub-categories with little to no monetary incentives on potential cost savings. This suggests that procurement managers’ personal discriminatory biases may affect sourcing decisions – potentially through their brief interactions with suppliers’ sales managers. While research on labor market discrimination may offer some insights on how discrimination may be affecting this interaction, the salient features of the sourcing context requires focused attention.
Originality/Value - Novel Findings:
The predictions from the statistical and taste-based views in labor economics are opposing and inconclusive in the sourcing context. Whether discrimination exists in the context of sourcing decisions and its potential magnitude are open questions, which we address in this paper. We note that identifying and evaluating potential discrimination in sourcing is not possible using public data because buyer firms are not required to disclose their suppliers, information on the race of the sales managers of current, past, and potential suppliers, or their proposal specifications. Further, audit style field experiments in the sourcing context while potentially relevant and informative, would be prohibitively controversial. This is because sourcing decisions carry a high dollar value, and an audit study on sourcing would consume significant resources of targeted companies at the expense of shareholder value. Thus, we design and conduct controlled experiments to identify whether and to what extent discrimination exists in sourcing. To the best of our knowledge, our paper is the first to analyze the effect of discrimination on sourcing decisions.
The headline result of our study is that racial discrimination affects sourcing decisions. In particular, we find that when facing a supplier with a sales manager with a distinctively black name, buyers are 6.5% less likely (statistically significant at 1% level) to select that supplier compared to a supplier with a sales manager with a distinctively white name. Additional analysis confirms the robustness of our findings.
Implications for Practice:
Our results suggest that organizations, even when committed to responsible operations, may unknowingly be acting as discriminatory buyers through the actions of their procurement managers, hence perpetuating the racial divide -- identifying sourcing as another important discrimination channel ignored in the previous literature. At the micro level, discrimination can significantly hurt the firm’s profitability by limiting its supplier pool. Our findings suggest that bias training and supplier diversity programs (where firms set targets for purchases from minority-owned suppliers) not only serve corporate social responsibility goals but can also improve firms’ profitability by expanding their supplier pool.
Implications for Policy:
Our findings can also help inform the policy debate on the importance of corporate diversity legislation. In the U.S., there is limited regulation requiring firms to promote corporate diversity and the existing legislation is often challenged in court. Our findings suggest that equal-opportunity legislations similar to those already in place in the labor market may also be needed in the context of sourcing and may enhance targeted firms’ corporate social responsibility profile and profitability.
Implications for Society:
We identify that addressing racial discrimination in sourcing is a vital step in leveling the racial wealth gap.
Implications for Research (How might this research affect future research):
Our results open up potential opportunities for future research. First, future experimental research can identify other moderating factors (e.g., number of potential suppliers, existence of an incumbent supplier, level of uncertainty, etc.) that may contribute to discrimination in sourcing, and whether discrimination can impact other operational functions of the firm. For example, discriminatory biases may impact operational investment decisions in a principle-agent setting, and order quantities and willingness to share forecast information in a supply chain setting. Second, how discrimination in sourcing changes with geography is an open question. Our results are US-based and focused on discrimination against the black minority. Replications of our work in different countries and cultural settings would be informative. Third, further research employing a similar experimental design can address whether gender discrimination can affect sourcing decisions. Fourth, future empirical work could address the effectiveness and consequences of supplier diversity programs and legislation. Related empirical data (when available) could be used to study, for example, how the operational structure and financial conditions of the firms affect their likelihood and success in implementing supplier diversity programs. Fifth, theoretical and empirical OM research can address how to optimize various factors contributing to the success of supplier diversity programs. Finally, future research can investigate the degree with which supplier diversity programs can contribute to diverse suppliers’ and their buyers’ future success.
Full Citation:
Aral, K.D., and L. Van Wassenhove, 2023, Racial Discrimination in Sourcing: Evidence from Controlled Experiments, Production and Operations Management, Forthcoming.
Abstract:
Entrepreneurial activity is widely deemed as an important tool for addressing racial wealth disparity. However, the survival of small minority-owned businesses depends on their ability to win contracts from buyer firms. Whether discrimination can exist in sourcing and affect this ability is an unexplored question potentially due to an implicit assumption that business-to-business settings are immune to such biases. In this paper, we study whether racial discrimination can affect sourcing decisions in controlled experiments. We find that when facing a supplier with a sales manager with a distinctively black name, buyers are 6.5% less likely (statistically significant at 1% level) to select that supplier compared to a supplier with a sales manager with a distinctively white name. Our findings suggest that equal-opportunity legislations similar to those already in place in the labor market may also be needed in the context of sourcing. For executives, our findings suggest that supplier diversity programs and procurement bias training can boost corporate performance by enlarging the supplier pool, while also enhancing a firm’s corporate social responsibility profile.