Search

Since 2016, the photographs of officers being considered for selection have been removed and reinstated twice in the Officer Promotion Selection Boards (PSB), making PSBs an ideal setting to detect whether implicit or explicit bias occurs in the U.S. Navy. Using data from Defense Manpower Data Center and PSBs, I specify linear probability models to estimate the promotion outcomes of minority and female officers before and after the policy change of masking officer photos from the board. This study is similar to civilian studies on taste-based and statistical discrimination, such as blind auditions. I find that promotion outcomes are not statistically significantly changing in relation to the policy change. However, when controlling for the interaction of the photo masking policy and an indicator for minority, minorities are significantly less likely to promote to commander. I also utilize the racial, ethnic, and gender composition of the board to analyze its effect on promotion of minorities and women separately. I find that board composition does have a statistically significant impact on minorities’ promotion to lieutenant commander. On the other hand, I do not find that board composition has a significant effect on minorities’ promotions to commander and captain. Ultimately, however, due to the limitations of zone status and FITREP information in the data, this study is unable to confirm or disprove whether masking photos hurts diversity and inclusion. Lieutenant Commander, United States Navy Approved for public release. Distribution is unlimited.

Although females represent almost half of the U.S. civilian labor force, they account for less than 15 percent of the officers in the U.S. military. To account for this discrepancy, this thesis tests for gender bias within the U.S. military by analyzing unique datasets derived from Naval Postgraduate School. We first conduct a randomized control trial by means of a survey (n=234). One group responds to scenarios relating to one gender; the second group responds to the same scenarios but relating to the opposite gender. We then use statistical analysis and ordinary least squares models to compare responses between genders. Second, using NPS student evaluations of teaching (n=175,093), we conduct t tests, examine the correlation of evaluation questions on instructor effectiveness, and employ ordinary least squares models using student and course fixed effects, and instructor and course fixed effects while controlling for student, instructor, class and school characteristics to analyze how gender influences evaluations. Our results identify that students favor matched gender pairs, with the effect largest among male pairs. We found this effect to be of marginal economic significance. These findings may indicate the effectiveness of gender equality training, or may reflect the current social climate concerning gender bias. http://archive.org/details/theidentificatio1094558369 Outstanding Thesis Major, Australian Army Lieutenant Commander, United States Navy Approved for public release; distribution is unlimited.