Science, technology, engineering and mathematics (STEM) fields are widely credited as the primary drivers of economic growth through innovation, with engineering universally identified as especially critical. Yet as other nations have strengthened their engineering talent pools, the United States has struggled to cultivate an engineering workforce that reflects its diversity and takes full advantage of its human capital. Reflecting this dilemma, African Americans have consistently posted the weakest persistence and bachelor's degree completion rates of all racial/ethnic groups in engineering, and by some indications, their postsecondary outcomes are worsening. The purpose of this study was to develop understanding about potential institutional levers for improving engineering bachelor's degree attainment both for underrepresented minorities (URMs) broadly and Black students specifically. Drawing on the higher education production function, I used multiple sources of institutional panel data for 324 engineering schools/colleges from 2005 to 2011 to uncover differential relationships between faculty predictors and engineering bachelor's degree production by student race/ethnicity and institutional context. I used multiple imputation to handle missing data and estimated fixed effects linear regression and dynamic panel models of engineering degree production, then I assessed institutions' degree production efficiencies using stochastic frontier analysis. The findings indicate that from 2005 to 2011, the number of engineering bachelor's degrees conferred to Black students declined 10%, with the smallest declines occurring at highly competitive institutions (2%) and the largest declines at HBCUs (30%). Results from the fixed effects models indicate that engineering faculty-to-student ratio was positively related and the proportion of research faculty negatively related to engineering bachelor's degree production for every student subgroup in at least one institutional setting. The share of URM faculty was positively related to degree production for URMs and Blacks in some settings. However, no faculty measure was predictive of degree output for every student subgroup across every institution type. And in every instance where a faculty variable was related to degree output for multiple student subgroups, the magnitude of the estimated effect was greatest for Black students, then URMs, then all students. Ultimately, the study suggests that leveraging institutional resources to improve student outcomes in STEM calls for targeted analyses to develop strategies that reflect the heterogeneity of STEM disciplines, STEM students, and educational settings. [The dissertation citations contained here are published with the permission of ProQuest LLC. Further reproduction is prohibited without permission. Copies of dissertations may be obtained by Telephone (800) 1-800-521-0600. Web page: http://bibliotheek.ehb.be:2222/en-US/products/dissertations/individuals.shtml.]