Explaining interstate military friendly policy diffusion across U.S. universities: uncovering vertical-diagonal diffusion

ABSTRACT

This article examines the diffusion of diversity management policies across large state bureaucracies that operate in a national marketplace of ideas. We analyze organizational antecedents and horizontal, vertical, and vertical-diagonal diffusion mechanisms associated with the degree of diversity management convergence among research universities from 2008-2016. Findings suggest the long-theorized, but rarely measured, non-linear pattern of diffusion explains organizational adoptions and the degree of policy convergence across the system. We also show a vertical-diagonal process whereby policy spills over to influence policy choices in other lower-level governance systems. This spillover may explain how organizations make policy decisions in complex governance arrangements.

KEYWORDS:

• Military veterans
• Policy diffusion
• Education
• Diversity Management

Disclosure statement

No potential conflict of interest was reported by the author(s).

Notes

1. We rely on the term ‘service members’ throughout the article to include military veterans, active-duty personnel, National Guard, and Reservists.

2. For-profit institutions do not compete for research grants, and therefore may feel less compelled by government pressure to change practices.

3. We describe the most common policies in the online appendix.

4. We do not examine bottom-up diffusion, because there is little evidence these policies diffused in that pattern.

5. These data only include the number of students receiving GI-Bill education benefits or Department of Defence Tuition Assistance Program benefits and not the total number of student veterans enrolled or completion rates.

6. The data on time-to-completion and student debt was from a self-reported survey distributed and reported by Military Times each year. Their ‘Best for Vets’ reports from 2011–2018 show a marked difference in these two measures between proprietary (for-profit) organizations and public organizations. This data was not included in the analysis for this project because many from our sample are excluded in this report.

7. These policies include (1) a single point of contact for all student veteran affairs (2) student veteran liaisons in existing offices (3) Veteran specific orientation (4) Priority registration if available for other student groups (5) Fee waivers (6) Deferred tuition (7) Institution-sponsored events highlighting veterans or military service (8) Campus chapter of Student Veterans of America (9) Academic credit for military service (10) Student veteran specific housing (11) Student veteran centre and (12) Dedicated Student veteran curriculum/degree programs.

8. To preserve space, we do not present the empirical results from the in-state model, but they are available in the online appendix.

9. We estimate Poisson regressions because our dependent variable is a count measure of the number of policies one university emulates from another.

10. We include dummy variables from University_i as our panel fixed effect to control for any unobservable within the institution converging on the peer’s existing policies.

11. We estimated the models using standard robustness checks described in recent diffusion research (see Nicholson-Crotty & Carley, 2016 and Boehmke 2009a). In the interest of space, we only present the most restrictive estimations (institutional and year fixed effects conditioned on university years in which emulation was possible). Less restrictive estimations not conditioned on emulation possibility and cubic splines to control for duration dependence showed similar results and are available from the corresponding author.

12. The state and institutional policies are distinct in our estimations. This vertical diffusion is therefore not a coercive mechanism in which the state imposes policies onto institutions, but instead serves as a social or political symbol for the values of the policymakers.

13. Although there are many problems with six-year graduation rates as a proxy for student performance, it is the most widely used metric for student success in higher education for state performance measures (Ishitani 2003) and public management scholarship (Fay, Fryar, Meier, & Wilkins, 2021).

14. We present the external and internal antecedents in separate tables due to space limitations. All empirical models include both external and internal antecedents in the equation.

15. The dotted lines are the 95% confidence intervals.

Additional information

Notes on contributors

Daniel L. Fay

Daniel L. Fay is an associate professor in the Askew School of Public Administration and Policy at The Florida State University. His research interests include organizational theory, diversity issues in public management, veterans’ policy, policy diffusion and higher education policy and management. His work appears in Research in Higher Education, Social Science Quarterly, and The Journal of Technology Transfer, American Review of Public Administration, Policy Studies Journal International Journal of Organizational Theory & Behavior, and Public Administration Review.

Abby Kinch

Abby Kinch is the Vice President of Programs and Services at Student Veterans of America. Her research interests include strategic planning and management, intelligence analysis, intelligence operations, emergency management, training, and policy.

Frances S. Berry

Frances S. Berry is Reubin O’D. Askew Eminent Scholar and Frank Sherwood professor of Public Administration in the Askew School of Public Administration and Policy at Florida State University. Her research interests are innovation and diffusion, performance and strategic management, public policy, and networks and policy implementation. She is a fellow of the National Academy of Public Administration.