ii

This work was funded by grants from Google, Inc., The Kresge
Foundation, and Lumina Foundation. The study is a public-private
partnership among Student Veterans of America, the National Student
Clearinghouse, and the U.S. Department of Veterans Affairs Veterans
Benefits Administration.

Million Records Project
Author: Chris Andrew Cate, Vice President of Research, Student Veterans of
America

Keywords: (1) Student veterans, (2) Completion Rates Student veterans, (3)
Time-to-completion Student veterans, (4) Level of education Veterans, (5)
Degree fields Student veterans

Student Veterans of America is a 501(c)(3) nonprofit organization that
provides military veterans with the resources, support, and advocacy
needed to succeed in higher education and following graduation.

Student Veterans of America is a registered trademark.

Cover design by Reingold
Cover photos: Student Veterans of America

Copyright 2014 Student Veterans of America

All rights reserved. No part of this report may be reproduced in any
form, to include electronic or mechanical means (e.g., photocopying,
recording, or information storage and retrieval), without the express
written consent of Student Veterans of America.

Published 2014 by Student Veterans of America
1625 K St. NW, Suite 320, Washington, DC 20006-1679
Student Veterans of America URL: http://www.studentveterans.org
Telephone: (202) 223-4710
Email: millionrecordsproject@studentveterans.org

Suggested citation: Cate, C.A. (2014). Million Records Project: Research from Student
Veterans of America. Student Veterans of America, Washington, DC.

iii
Acknowledgments
Student Veterans of America would like to thank our generous funders and partners for
supporting this historic project. Without their steadfast commitment to the success of student
veterans, this initiative would not have been possible.



iv
Executive Summary
Accurate data on the postsecondary academic outcomes of todays generation of student
veterans have been difficult to find. Inconsistent methods of collecting such information has led
to confusion about the completion rates of student veterans in higher education, and without
strong, empirical data, the uncertainty will persist.
The Million Records Project was envisioned, planned, and implemented by Student
Veterans of America to address these gaps in knowledge and, with additional research,
ultimately determine best practices and policies that promote student veteran success. The result
of a public-private partnership among Student Veterans of America, the National Student
Clearinghouse, and the U.S. Department of Veterans Affairs, the Million Records Project
explores the postsecondary academic outcomes of nearly 1 million student veterans who initially
used Montgomery and Post-9/11 GI Bill benefits between 2002 and 2010. The outcomes
explored in the first phase of the project include student veteran postsecondary completion rates,
time-to-completion, level of education, and degree fields.
The Million Records Projects initial results showed strong postsecondary outcomes for
the current generation of student veterans. A majority (51.7 percent) of student veterans in this
sample earned a postsecondary degree or certificate, despite the challenges they face. Like other
nontraditional student populations, they tend to be older, have families to support, and juggle
employment and school. Unlike their nontraditional peers, however, student veterans are unique
in that they may delay their enrollment or interrupt their progress in higher education due to
military obligations, and may have service-connected disabilities.
As Reservists and National Guard units perform a greater function in todays military, the
perception that most veterans join the military after high school, serve their country, separate
v
from service, and then enter higher education is changing. Consequently, veterans follow very
different paths to graduation compared with their traditional peers. Despite these and other
challenges common to nontraditional students, student veterans are persisting towards their
educational goals and succeeding. At both the two- and four-year degree level, the majority of
veterans who graduate do so within four and five years respectivelya rate similar to the
traditional student population.
The vast majority of student veterans are enrolling in public institutions (79.2 percent),
with a smaller percentage enrolling fairly evenly in private nonprofit (10.7 percent) and
proprietary schools (10.1 percent). Following a similar distribution, the large majority of student
veterans who graduate do so from public schools (71.7 percent), with the remaining population
graduating in much smaller but fairly even percentages from private nonprofit (15.5 percent) and
proprietary (12.9 percent) schools. Based on the sector of initial enrollment, the private
nonprofit sector had the highest student veteran graduation rate (63.8 percent), though a
significant number of those students (21.6 percent) actually completed their degree at a public or
proprietary institution.
Of those that completed, approximately 9 out of 10 (89.7 percent) initially earned degrees
at the associate level or higher. In addition, many student veterans achieve higher levels of
education: 31.3 percent of the sample who initially earned a vocational certificate, 35.8 percent
of the sample who initially earned an associate, and 20.8 percent of the sample who initially
earned a bachelors degree went on to also earn a degree at a higher level. Results also indicate
that a high percentage of student veterans are pursuing degrees in business, public service,
health, science, and engineering.
vi
The Million Records Project represents one of the most comprehensive examinations of
student veterans postsecondary academic success in decades. The initial results of the project
are critical first steps in filling research gaps regarding student veteran outcomes, but more is still
needed. This and future studies have the potential to aid policymakers, institutions of higher
education, and other stakeholders as they decide how best to allocate resources to better serve
our nations veterans. In addition, Veterans Service Organizations, such as Student Veterans of
America, The American Legion, and Veterans of Foreign Wars, can use this new, up-to-date
information to advocate more effectively on behalf of their constituents. It is Student Veterans
of Americas hope that the results of this project will demonstrate that Americas investment in
its nations veterans is paying offconfirming that todays veterans are earning college degrees,
entering the workforce armed with the skills and knowledge they need to succeed, and
continuing to contribute to society in meaningful ways.

vii
Table of Contents
Acknowledgments .......................................................................................................................... iii
Executive Summary ....................................................................................................................... iv
Table of Contents .......................................................................................................................... vii
Introduction ..................................................................................................................................... 1
Literature Review ............................................................................................................................ 4
Introduction ................................................................................................................................. 4
Storyline 1: Student Veterans Are Not Completing Postsecondary Programs ........................... 4
Nontraditional characteristics ................................................................................................. 4
Students with disabilities ........................................................................................................ 5
Use of GI Bill benefits ............................................................................................................ 8
Storyline 2: Student Veterans Are Completing Postsecondary Programs .................................. 9
Historical evidence. ................................................................................................................. 9
National surveys .................................................................................................................... 10
National Databases ................................................................................................................ 13
The Paradox .............................................................................................................................. 16
Million Records Project ............................................................................................................ 18
Data From the U.S. Department of Veterans Affairs. ........................................................... 19
Data From the National Student Clearinghouse ................................................................... 19
Utility of the Million Records Project ...................................................................................... 20
Methods......................................................................................................................................... 21
Design ....................................................................................................................................... 21
Sample ....................................................................................................................................... 21
viii
Variables ................................................................................................................................... 23
Outcomes of Interest ............................................................................................................. 23
Demographic Variables ........................................................................................................ 26
Procedure .................................................................................................................................. 28
Results ........................................................................................................................................... 31
Demographics ........................................................................................................................... 31
Postsecondary Completion Rates .............................................................................................. 33
Time to Completion for Initial Degree ..................................................................................... 34
Degree Fields ............................................................................................................................ 38
Science and Engineering degrees. ......................................................................................... 39
Postsecondary Academic Outcome Comparisons .................................................................... 42
Initial school sector. .............................................................................................................. 42
Branch of service. ................................................................................................................. 46
GI Bill usage. ........................................................................................................................ 48
Discussion ..................................................................................................................................... 50
Postsecondary Academic Outcomes ......................................................................................... 52
Policy Implications ................................................................................................................... 63
Research Implications ............................................................................................................... 65
Implications for Practice ........................................................................................................... 66
Future Research and Directions ................................................................................................ 67
Closing ...................................................................................................................................... 69
Works Cited .................................................................................................................................. 71
1
Introduction
For nearly 70 years, veterans have used the GI Bill to ease their transition from military
service to the civilian workforce. The benefit provides valuable financial support to help
veterans afford the cost of attending postsecondary educational and vocational training programs
and persist to degree attainment.
Historians and economists have well documented the positive, beneficial outcomes that
the original GI Bill produced for World War II veterans and the United States economy (Bound
& Turner, 2002; Greenberg, 1997; Stanley, 2003). However, these studies were historical in
natureconducted several decades after that eras GI Bill benefits concluded and once those
veterans entered the workforce. Historical research, while it provides useful insights, it is not as
valuable to policymakers and stakeholders who rely on current data to inform their decision-
making process.
In addition, this generation of student veterans differs from those of the Cold War Era
who differed from Vietnam Era student veterans who differed from those who emerged after
World War II. As the military continuously adapts to current circumstances, todays veterans
reflect such changes. For example, Operation Enduring Freedom (OEF) and Operation Iraqi
Freedom (OIF) mobilized more reservists and National Guard personnel than previous combat
eras. The effects of a mid-term mobilization on the academic outcomes of these service
members remain largely unknown. This lack of knowledge makes it difficult for college and
university administrators and faculty to implement proper policies that enable student veterans to
deal with such disruptions, withdraw appropriately, and then re-enroll in school.
With more than 1 million beneficiaries having used or currently using the Post-9/11 GI
Bill, the need to report on their postsecondary academic outcomes has significantly increased.
2
The Post-9/11 GI Bill is a massive investment of more than $30 billion in the success of our
nations veterans. However, a lack of data on their postsecondary outcomes, and the lack of an
established method to collect such data, make it difficult to accurately measure the return on the
GI Bill investment. National databases often fail to accurately identify student veterans or track
their postsecondary academic outcomes. National surveys are also ill-equipped to measure the
academic success of student veterans due to response bias and sampling error.
Lack of information on the postsecondary outcomes of student veterans also limits the
efficacy of those tasked with supporting them, such as higher education institutions,
policymakers, service providers, and other key stakeholders. Scarce resources are often
allocated to programs that have not been empirically evaluated to determine how they impact a
student veterans postsecondary academic outcomes. Without an empirical study to measure the
postsecondary success of a large sample of student veterans, the publics only source of
knowledge on these individuals comes from media portrayals, anecdotal stories, word of mouth,
or case studies, which are difficult to validate and generalize to the entire student veteran
population. Finally, if an unfounded and unsubstantiated perception develops that student
veterans are doing poorly in college, Congress may believe that taxpayer dollars are being
wasted on incomplete educations, and the GI Bill may be cut. Any reduction in benefits will
have a ripple effect on a separating veterans life trajectory. With little or no postsecondary
education or vocational training, he or she will be ill-prepared to enter todays credential-
dependent civilian workforce and compete for high-paying, sustainable careers.
The need for empirical data to drive policy and programmatic investments regarding
student veteran outcomes led to the development of the Million Records Project (MRP). The
MRP, a public-private partnership between the U.S. Department of Veterans Affairs (VA), the
3
National Student Clearinghouse (NSC), and Student Veterans of America (SVA), was designed
to address many of the weaknesses found in previously established national databases and
surveys. By matching VA and NSC data, SVA analyzed a national sample of 1 million student
veterans who first used their GI Bill benefits between 2002 and 2010 to accurately measure the
postsecondary academic outcomes of the current generation of GI Bill beneficiaries. The MRP,
for the first time in the GI Bills history, provides policymakers, stakeholders, and the American
public accurate, near real-time data on the completion rate, time to degree completion, level of
education, and degrees pursued for todays student veterans.
This report summarizes the results of the project and will enable policymakers and
stakeholders to make data-driven decisions that will affect millions of current and future student
veterans. The MRP also provides institutions of higher education a benchmark against which
they can compare their own student veteran completion rates, and it significantly adds to the
limited amount of research on the topic while also setting the stage for future studies on student
veterans and other nontraditional student populations.

4
Literature Review
Introduction
Two different storylines have developed in recent years regarding student veteran
postsecondary completion rates: one in which data appear to support the notion that student
veterans have a high postsecondary dropout rate and the other that has evidence supporting high
student veteran postsecondary completion rates. The review will first examine available
evidence regarding the potentially high postsecondary dropout rate of student veterans and the
risk factors associated with such results. It will then explore the contrary storyline of student
veterans succeeding in high numbers in higher education. The review will continue with a
discussion on the weaknesses found in national databases and how that contributes to the
confusion surrounding student veteran postsecondary academic outcomes. The literature review
concludes with an overall description of the Million Records Project and its potential
contribution to this field of research.
Storyline 1: Student Veterans Are Not Completing Postsecondary Programs
Nontraditional characteristics. Horn (1996) uses seven characteristics to classify
postsecondary students as nontraditional: delayed enrollment; part-time enrollment; financial
independence; full-time employment while enrolled; having dependents; single parent; and did
not receive standard high school diploma. Horn also establishes a scale of nontraditional status
based on the number of nontraditional characteristics the student met. A student having one
nontraditional characteristic ranked as minimally nontraditional; those with two or three
nontraditional characteristics are classified as moderately nontraditional; and having four or
more nontraditional characteristics means one is highly nontraditional. Using these
classifications and U.S. Department of Education data, Horn found that nontraditional students
5
were less likely than traditional students to have completed their degree after five years and were
more likely to have withdrawn from higher education. Horn also found that the fewer
nontraditional characteristics a student had, the more likely it was that he or she would have
earned a bachelors degree in five years.
By applying Horns characteristics to student veterans from the post-World War II era, it
becomes clear that they were the first nontraditional students to enroll in higher education in
large numbers. As Humes (2006) describes, many student veterans of the post-World War II era
were older, married, had children for at least part of their collegiate career, and were financially
independent. Most of todays student veterans mirror their predecessors and can also be
classified as nontraditional students using Horns characteristics. They are frequently older than
their traditional counterparts due to a multiyear break between high school and college; a high
percentage are married and have families of their own. Although not a characteristic listed by
Horn, student veterans also have a greater depth and breadth of world experiences compared
with traditional students, largely due to their military service and overseas experiences with other
cultures (Ackerman, DiRamio, & Garza, 2009; DiRamio, Ackerman, & Mitchell, 2008).
Students with disabilities. College students with physical and cognitive disabilities,
including mental health diagnoses, often have difficulty persisting in and completing
postsecondary educational and vocational programs. Hurst & Smerdon (2000) examined
National Center for Education Statistics (NCES) data on students who attended community
colleges and found differences in the persistence rates of students with disabilities (53 percent)
compared with those without disabilities (64 percent). Other research found that students with
disabilities might have lower academic results and lower confidence in their academic abilities
compared with students without disabilities (Cosden & McNamara, 1997). Research focusing on
6
students with mental health diagnoses also suggests a link between psychological symptoms and
academic performance, which subsequently leads to students withdrawing from postsecondary
institutions (Megivern, Pellerito, & Mowbray, 2003).
For service members and veterans, major technological and medical advances have
increased the likelihood of surviving combat-related injuries. It has been estimated that 90
percent of injuries sustained in Operation Iraqi Freedom (OIF) or Operation Enduring Freedom
(OEF) will be survived (Klocek, 2008). This is a striking increase when compared with 70
percent of battlefield injuries survived during the Vietnam War, and fewer than that in Korea and
World War II. The U.S. Department of Defense reported that as of Feb. 28, 2014, approximately
51,000 service members have been wounded during OIF, Operation New Dawn (OND), and
OEF, which indicates that a substantial portion of todays veterans may be enrolling in higher
education with minor to severe service-related disabilities.
While physical injuries associated with battle, such as loss of limb or severe burns, are
common among OIF and OEF service members, an increasing number of todays veterans have
also sustained a traumatic brain injury (Klocek, 2008). Traumatic brain injuries (TBIs),
sometimes known as the invisible wounds of war, are not nearly as recognizable as other
physical disabilities but can be just as challenging for veterans to cope with. Also called severe
concussions, close head wounds, and mild brain injuries, TBIs occur when a person
receives a severe blow to the head, either by a physical object or a concussion blast like that
produced by a detonated improvised explosive device (Klocek, 2008). The symptoms of a TBI
parallel those of a concussion. However, concussion symptoms typically dissipate after a day,
while TBI symptoms can last for years or a persons entire life.
7
TBIs can also cause changes in a persons behavior and cognition, specifically memory.
A person with a TBI may have difficulty focusing and recalling recent events. Hearing, vision,
balance, and spatial orientation may also be impacted in persons with TBIs. While the effect of
TBIs on student veteran academic persistence is unknown, it is clear that such cognitive injuries
negatively impact ones ability to remember, remain focused, and see and hear wellall of
which are vital to a college students success.
In addition to TBIs, veterans of the current wars are susceptible to other invisible
wounds, or mental health diagnoses, such as posttraumatic stress (PTS), depression, or
substance abuse. Hoge, Auchterlonie, and Milliken (2006) found that the prevalence rate for any
mental health disorder was 19.1 percent among OIF veterans and 11.3 percent among OEF
veterans.
Service-related injuries, whether visible or invisible, will likely impact all facets of a
veterans post-service life, including their academics. For example, research has chronicled
student veteran experiences in college (Ackerman, et al., 2009; DiRamio, et al., 2008) and found
that veterans who sustained lasting wounds in battle may have unpredictable attendance in class
due to pain and other symptoms associated with their injuries. In addition, the side effects from
medicines used to treat physical pain or mental health diagnoses could affect a veterans ability
to concentrate, take notes, and remain alert, thus negatively affecting his or her academic
performance (Church, 2009).
Given these findings, it is likely that a proportion of veterans from the recent military
operations that enroll in higher education will have combat wounds, mental health diagnoses, or
both.
8
Use of GI Bill benefits. Some media outlets have reported that 88 percent of student
veterans drop out of their postsecondary educational and vocational program within the first year
(Briggs, 2012; Wood, 2012). Veteran advocates, including SVA, The American Legion, and
Veterans of Foreign Wars believe that claim to be inaccurate for a number of reasons. However,
the statistic is discussed in this review because it is still commonly used as an indication of
student veteran postsecondary academic outcomes, both by the media and general populace.
What little is known about the 88 percent statistic, is that it is based on a weak methodology.
Use of VA education benefits, specifically enrollment in the GI Bill, was used as a proxy for
enrollment in a postsecondary institution. While use of VA education benefits is a strong
method for identifying student veterans, it is a weak measure of persistence and academic
outcomes. For example, student veterans may stop using VA education benefits but continue to
remain in school, and supplement costs with other sources of financial aid (e.g., scholarships,
grants, Title IV funds). In the process previously described, this student veteran would be
classified as a dropout or withdrawal. This scenario also indicates that student veterans may not
be using all of their GI Bill benefits to earn a degree, which can mistakenly be construed as
leaving school prior to completion.
In summary, student veterans share characteristics with both nontraditional students and
students with disabilities or mental health diagnoses. These groups of students typically have
high postsecondary dropout rates. In addition, student veterans who delay using VA education
benefits may be mistaken as college dropouts. Taken altogether, this evidence suggests that
student veterans may have poor postsecondary academic outcomes.
9
Storyline 2: Student Veterans Are Completing Postsecondary Programs
While the evidence above suggests that student veterans face challenges that would
typically lead to low postsecondary academic completion rates, evidence also exists that shows a
high percentage of student veterans may be earning postsecondary degrees.
Historical evidence. Historically, veterans have a demonstrated record of academic
success in postsecondary environments. Approximately 8 million of the nearly 16 million post-
WWII student veterans who were eligible for the GI Bill earned postsecondary degrees or
completed vocational training programs (Greenberg, 1997). Research from this era indicates
that, as a group, the academic performance of student veterans, measured by cumulative GPA,
was slightly higher than that of their nonveteran counterparts (Frederiksen & Schrader, 1952).
Several other studies from the post-WWII era generally corroborated these findings (Garmezy &
Crose, 1948; Gowan, 1949; Love & Hutchison, 1946). As for Vietnam War-era student
veterans, a report from VA in 1976 found that approximately two-thirds completed their
postsecondary programs (Department of Veterans Affairs, 1976). Joanning (1975) replicated
and extended this research, finding that the academic performance of post-Vietnam-era student
veterans, as measured by GPA, was equal to or better than their nonveteran counterparts.
It is also important to note several key environmental differences when comparing the
postsecondary academic outcomes of GI Bill beneficiaries from past generations with those of
today. The most notable difference is that most conflicts from World War II through the
Vietnam War relied on a conscripted military force. Todays wars are being fought by an all-
volunteer force. The modern military gives young adults a choice: volunteer for service or
remain a civilian, in which case higher education is a viable option. For young men and women
who lack the financial means or the skills to attend college, however, the choice to serve in the
10
military is more than a call to duty; it is a means of escaping environments that have little
potential for socioeconomic advancement and limited vocational options by which to learn
discipline, and potential trades (Teachman, 2007).
Along with the rise of an all-volunteer force, public policy changes have also enabled
more young adults to enroll in postsecondary institutions rather than join the military. For
example, the Higher Education Act of 1965 and other financial aid programs have made college
affordable for many young people. These new financial aid policies, in combination with
educational deferments for conscription, allowed more young men and women to go directly into
postsecondary education from high school, and have resulted in the military having to compete
with colleges and universities (Teachman, 2007).
Although veterans have historically performed well in higher education, todays military
and postsecondary environments are vastly different than those of the past, thus making it
difficult to conclude that student veterans will continue to replicate the successes of their
predecessors. Just as the military continues to innovate and change, so does each generation of
student veterans. The current generation of student veterans faces different challenges upon
separation than did their Cold War predecessors, who faced different challenges than Vietnam
War veterans. A historical argument alone is insufficient to prove that todays student veterans
will succeed academically.
National surveys. Recent national surveys provide further evidence of a high rate of
student veteran postsecondary completion. The 2010 National Survey of Veterans (2010 NSV;
Westat, 2010), conducted by Westat for VA, examined and elicited feedback from VA
beneficiaries on VA programs and services. In the section regarding education benefits, the 2010
NSV asked veterans about degree or program completion. According to the results, 63 percent
11
of survey respondents reported that they completed the postsecondary educational or vocational
program for which they used their VA educational benefits (Westat, 2010). When analyzing the
2010 NSV by service era, the 45-year period between the end of the Korean War and Sept. 11,
2001 shows a stable postsecondary completion rate between 66 and 68 percent. A majority of
the participants (51.1 percent) in the 2010 NSV who reported serving after Sept. 11, 2001 also
reported the completion of their postsecondary educational or vocational training program (Cate,
2014). It should be noted that this finding is preliminary and will most likely increase as time
passes and more student veterans are able to complete their postsecondary programs.
Although the 2010 NSV seems to indicate a high student veteran postsecondary
completion rate, the results become weaker when given a closer review, such as investigations
into the completion rates of veterans who separated from the military after Sept. 11, 2001. The
sample size for this group is extremely small, thus yielding weak conclusions. Another flaw is
the surveys reliance on self-reported data, which is a common survey weakness that can lead to
imprecise results. An individual might misunderstand a question or the directions and provide an
inaccurate response. In addition, the survey relies on a single question to measure completion. It
asks the respondents if they have completed the program for which they have used VA education
benefits. The broad question can be interpreted in a variety of ways that may lead to
misinterpretation, such as participants replying no if they finished their program after having
used all of their benefits. This question also does not include the possibility of student veterans
completing their vocational and educational programs without use of their VA education
benefits. This again may lead to a response error that affects the final results. Finally, the NSV
is typically conducted every 10 years, making the results less relevant and accurate and more
difficult to interpret as time passes from the initial collection of data.
12
Another national survey, the 2012 American Community Survey (ACS; U.S. Census
Bureau, 2012) conducted by the U.S. Census Bureau, collects information on a wide range of
demographics, including age, sex, income and benefits, education, and veteran status. Therefore,
it allows for comparisons between veterans and nonveterans within the U.S. population. While it
does not specifically ask about completion, the education level reported can serve as a proxy for
degree attainment. The 2012 ACS shows that a larger percentage of nonveterans self-reported
attainment of a bachelors degree or higher as compared with veterans (29.3 percent to 26.7
percent, respectively).
However, the ACS also has several flaws in tracking student veteran outcomes. First,
participants are asked to describe their highest level of education attained. Respondents have a
predetermined list of potential answers and the ACS combines some college and attainment
of an associates degree into one category, which makes it difficult to ascertain whether the
person attained a degree, withdrew, or is still enrolled in a postsecondary program. Second, the
ACS does not include a question regarding vocational or on-the-job training programs, which
veterans can attend using their education benefits. It is unclear whether student veterans would
equate these programs as some college or choose high school diploma only. Last, the ACS
does not have sufficient data on veterans military service. Without this information, it is not
possible to determine when a student veteran separated from the military and started his or her
postsecondary studies, making it difficult to conduct detailed analysis on student veteran
postsecondary academic outcomes. As a result, researchers cannot ascertain with certainty
whether or not the small difference between the education levels of veterans and nonveterans
will remain stable across all age groups.
13
National Databases. Contributing to the confusion regarding student veteran
postsecondary academic outcomes is the fact that national-level data on student veterans have
been difficult to find, analyze, and interpret due to poor collection methods, narrow inclusion
criteria, and mistakes in identifying student veterans. Most traditional national postsecondary
databases exclude a portion of the student veteran population while including other military
populations, which makes accurately analyzing student veteran postsecondary academic
outcomes difficult at best.
U.S. Department of Education. The U.S. Department of Education manages the National
Center of Education Statistics (NCES), which tracks postsecondary student academic outcomes.
The NCES maintains several databases, such as the Integrated Postsecondary Education Data
System (IPEDS), that contain data on postsecondary students, as reported by institutions of
higher education and financial aid records. However, many of the NCES databases do not
properly identify and track students who are veterans, making it difficult to interpret these results
due to weak sampling methods.
IPEDS, the database most frequently used to track postsecondary student outcomes, is a
collection of interrelated annual surveys sent to every college, university, technical, and
vocational institution that participates in federal student financial aid programs. These schools
are required to report data on enrollments, program completions, graduation rates, and
institutional data. However, IPEDS collects data only on first-time, full-time students entering in
the fall term. The database excludes many students who transfer schools, start at community
colleges then transfer to a four-year university, temporarily withdraw from school for personal or
military-related reasons, attend part-time at some point in their academic career, or those who
14
pursue on-the-job training or vocational certificates. Unfortunately, student veterans fit most of
these characteristics and are often mistakenly removed from the database and not counted.
The National Postsecondary Student Aid Study (NPSAS), a second NCES database that
contains information about student veterans, also does so using a weak method. The NPSAS is a
comprehensive research dataset on postsecondary student demographics, financial aid, and
enrollment. The primary weakness of the database is in the method used to identify student
veterans. NPSAS uses the Free Application for Federal Student Aid (FAFSA) in classifying
samples as active duty service members or veterans; student interviews and institutional records
supplement the data.
The FAFSA contains two questions about military service: 1) does the applicant currently
serve on active duty in the U.S. Armed Forces? and 2) is the applicant a veteran of the U.S.
Armed Forces?. These questions misclassify several categories of military-connected students.
For example, a reservist could be activated and deployed, return home to reserve status, and
retain that reservist status while attending school. The reservist would be eligible for VA
education benefits, but would not be classified as currently serving on active duty or as a
veteran of the Armed Forces on the FAFSA and would therefore not be flagged as a student
veteran in NPSAS. A second example is a service member in the Inactive Ready Reserve who is
in the process of separating from the military after serving on active duty. If enrolled in college,
this service member would be misidentified as a nonveteran in the NPSAS based on their likely
responses to the FAFSA indicating that they are neither a veteran nor serving on active duty.
Tracking student veteran outcomes using FAFSA-related identifiers can also exclude
student veterans because GI Bill benefits are not included in Title IV funding under the Higher
Education Opportunity Act (HEOA, 2007). As a result, student veterans are not required to
15
complete and submit a FAFSA to receive GI Bill benefits. Those that do not complete and
submit a FAFSA are excluded from the NPSAS database.
Furthermore, Department of Education databases use a very broad definition of veteran
education benefits (HEOA 420(c), 2007). Under Title IV of the HEOA, veteran education
benefits not only include GI Bill benefits, but also Reserve Officer Training Corps (ROTC)
scholarships, Department of Defense Tuition Assistance program funds, and Survivors and
Dependents Educational Assistance Program benefits. This makes it extremely difficult to
isolate the impact of specific benefits.
Department of Veterans Affairs. In contrast to the Department of Education, VA is able
to identify nearly every student veteran, but has only recently been instructed (through Executive
Order 13607) to track and collect information on student veteran outcomes. Historically, the
Veterans Benefits Administration (VBA) of VA is mainly responsible for disbursing tuition
payments to schools after a student veterans enrollment is verified. To accomplish this task, VA
collected information related only to the amount and destination of the benefit, such as the
student veterans institution, enrollment intensity (part-time or full-time), and the amount of the
disbursement. These limited data do not accurately translate into student veteran postsecondary
academic outcomes. For example, some student veterans who exhaust their education benefits
may go on to complete postsecondary degrees or credentials, and others may complete degrees
or credentials before using all of their benefits.
While VA has recently established a new system in its certification process to obtain
additional data on student veteran academic outcomes, a gap in the collection process still exists.
The VA will not have data on a student veterans enrollment and completion if the veteran is not
using his or her GI Bill. While use of VA education benefits is a strong method of identifying
16
student veterans, it is a weak stand-alone method for determining postsecondary academic
outcomes, as it does not take into account student veterans who complete certificates or degrees
using other means of financial aid.
In summary, compared with Department of Education databases, the VA education
benefits database, the 2010 NSV, and the ACS have better methods for identifying student
veterans, methods that allow for more accurate estimates of their postsecondary academic
outcomes. However, the VA benefits database, the 2010 NSV, and the ACS were not designed
to measure or track postsecondary academic outcomes, which is the strength of the Department
of Educations databases. These weaknesses have contributed to the current contradictory results
and inaccurate perceptions of student veteran postsecondary completion. The 2010 NSV notes
that the student veteran postsecondary completion rate is 68 percent; the ACS reports that 56
percent of veterans have completed at least some college or higher; and NCES indicates that the
six-year time-to-completion rate for baccalaureate-level student veterans starting in 2003 was 9.9
percent (95 percent CI [4.2, 15.6]) and the six-year completion rate for associate degrees was
11.2 percent (95 percent CI [5.6, 16.7]; U.S. Department of Education, 2009).
The variance in the reports makes it difficult to clearly understand the postsecondary
academic outcomes of student veterans. The ACS and 2010 NSV suggest that student veteran
completion rates are high, but the U.S. Department of Education data suggest the exact opposite.
Contradictory results make it nearly impossible for policymakers and stakeholders to make data-
driven decisions.
The Paradox
The lack of clarity on student veteran postsecondary academic outcomes has led to the
development of two contradicting storylines. One storyline suggests that student veteran
17
postsecondary completion rates are low, given that they are nontraditional students; many have
service-related injuries or mental health diagnoses; some do not exhaust their VA education
benefits. The second storyline suggests that student veteran completion rates are high, given that
student veterans have a long history of college success that originates with the post-WWII
generation of student veterans and two recent national surveys indicate that student veterans may
have high postsecondary completion rates and levels of education. These seemingly
contradictory stories form the paradox. How can both exist simultaneously?
One storyline of the paradox could be wrong. Perhaps veterans do not have high risk
characteristics or belong in these high risk groups. This is unlikely. A wealth of research has
established that veterans are nontraditional students who are generally older, have a break
between high school graduation and college enrollment, are more likely to be married with
families compared with traditional students, and some proportion of veterans have service-
related injuries or mental health diagnoses. This would suggest that student veterans would have
low postsecondary completion rates.
The survey results on the postsecondary success of student veterans could be wrong.
Research on student veterans in general is scarce and accurate data on their academic outcomes
are even more rare. While the two surveys mentioned, the 2010 NSV and the ACS, detail an
individuals veteran status and contain limited data on postsecondary information, they were not
designed to provide in-depth data or analysis on student veteran postsecondary academic
outcomes. In addition, the traditional national-level academic databases either do not accurately
identify and track student veterans or these databases do not connect completion data with an
individuals veteran status.
18
If these surveys are correct, though, that indicates that the paradox exists and although
student veterans share characteristics with students that historically have low rates of
postsecondary completion, student veterans are resiliently overcoming such obstacles to achieve
academic success. However, to fully investigate this concept, a new method of data collection
and tracking of student veterans must be established. The new method must address the
weaknesses found in the national-level surveys and databases previously described. First, it
would have to be able to accurately identify current student veterans, excluding veteran
dependents and active duty service members not using VA education benefits enrolled in
postsecondary programs. Second, it would need to be able to track student veteran degree
attainment at the individual level rather than the institutional level, so that student veterans are
not excluded when they transfer schools or withdraw from college due to military service or
personal reasons. Third, it would need to rely on objective data preferably collected directly
from VA and institutions of higher education.
These criteria served as the basis for the Million Records Project.
Million Records Project
It is evident that current systems were not designed to effectively or accurately measure
the postsecondary outcomes of student veterans. To better understand student veteran
postsecondary completion rates, SVA entered into a partnership with VA and the National
Student Clearinghouse (NSC) to create and develop the Million Records Project (MRP). The
MRP addresses several of the weaknesses inherent in current federal databases and national
surveys that track postsecondary academic outcomes, thus producing a more accurate estimate of
student veteran completion rates.
19
Data From the U.S. Department of Veterans Affairs. The Million Records Project
uses VA information to identify veterans who used education benefits. VA provided 500,000
records of veterans who initially used their Montgomery GI Bill Benefits (MGIB) between 2002
and 2010 and another 500,000 records of veterans who initially used their Post-9/11 GI Bill
benefits between 2009 and 2010 for a total of 1 million records. VA confirms a veterans status
with the Department of Defense prior to distributing benefits, thus it is certain that the sample
consists wholly of U.S. military veterans.
Data From the National Student Clearinghouse. The National Student Clearinghouse
(NSC) is a nonprofit organization that collects student enrollment and degree data at the
individual level directly from participating institutions of higher education that use the NSCs
services. The NSC provides individual student enrollment and program completion to the
education finance industry, the U.S. Department of Education, and to colleges and universities.
At the time of this reports completion, the NSC collected data on approximately 96 percent of
all postsecondary students in the United States.
In recent years, the NSC has done in-depth analysis of its own records to produce several
research reports called Signature Reports, which focus on the postsecondary completion rates of
U.S. college students. While similar in some respects to IPEDS, NSCs Signature Reports differ
in that they include many postsecondary students (e.g., those who change enrollment intensity or
transfer institutions) that are not part of IPEDS. The NSCs Signature Reports offer a more
comprehensive measure of postsecondary completion in the United States.
By using the NSCs database, the Million Records Project obtained accurate program
completion information based on institutional recordsnot self-reportswhich reduces
response error and survey-related biases and increases the validity of the data.
20
Utility of the Million Records Project
Establishing an accurate database by which to better measure student veteran
postsecondary academic outcomes is a necessary first step to better supporting this population of
students. Accurate data will allow student veterans to use their limited benefits more effectively
by being more informed consumers, thus increasing completion rates and improving the return
on the GI Bill investment. With a better understanding of national-level completion rates,
individual schools and programs will be able to benchmark the success of their own student
veterans. In addition, an accurate national rate of postsecondary completion allows Student
Veterans of America and others to analyze the programs, practices, and services that increase
such rates. Policymakers and stakeholders at all levels will begin to make data-driven decisions
that impact the lives of student veterans. Finally, it will allow researchers to begin to determine
whether student veterans are a unique group of nontraditional students or potentially require
greater assistance to aid in their postsecondary completion.
21
Methods
Design
This was a secondary data quantitative analysis study designed to report student veteran
postsecondary completion rates based on initial school enrollment cohorts, student veterans
time-to-completion, their highest level of education, and their majors or degree fields. A
secondary purpose of this study was to explore differences in the primary outcomes of student
veterans based on available demographic variables, such as branch of service and gender. Data
were obtained from VA through a Freedom of Information Act (FOIA) request. VA, the NSC,
and SVA collaborated to obtain completion data for 1 million student veterans who initially used
their GI Bill benefits between 2002 and 2010. Data analyses were conducted using descriptive
and inferential statistics including frequencies, means, and crosstabs.
Sample
The population of interest was United States military service members or veterans
enrolled in a postsecondary educational or vocational training program between 2002 and 2010.
The sample comprised 1 million veterans who either used the MGIB between 2002 and 2010 or
the Post-9/11 GI Bill between 2009 and 2010. Of the 1 million records, roughly 10.1 percent
(101,105) appeared in both the MGIB and the Post-9/11 GI Bill data files. The duplication
occurred because some student veterans were eligible for both benefits and exercised their option
of switching from the MGIB to the Post-9/11 GI Bill. These duplicates were removed, thus
yielding a total sample size of 898,895 veterans who initially used their GI Bill benefits between
2002 and 2010.
According to public VA data, VA processed education benefits claims for 4,067,476
MGIB and Post-9/11 GI Bill veterans between 2002 and 2010 (Department of Veterans Affairs,
22
2013). This cumulative statistic includes first-time as well as previous GI Bill uses; it does not
indicate that there were 4 million initial GI Bill users during that timeframe. However, the VA
beneficiary data (4,067,476) can serve as a proxy for the total population size of MGIB and Post-
9/11 beneficiaries between 2002 and 2010. As such, the Million Records Project sample covered
22.1 percent of all student veterans who used the MGIB and Post-9/11 GI Bill during the time
period and produced a margin of error of 0.09 percent.
While use of the VA dataset allowed for valid identification of student veterans, it did
exclude several potential groups of student veterans. Veterans who did not use their VA
education benefits to earn a postsecondary degree or certificate are excluded. Not only does this
specific subgroup represent a small minority of the student veteran population, but the process to
identify and include such individuals would have been extremely difficult and costly, and would
have had minimal impact on the results.
The sample also excluded dependents to whom veterans have transferred their VA
education benefits. For the most part, dependents have different obstacles to degree completion
than student veterans. The goal of this project was to better understand the academic outcomes
of student veterans. There is clearly a need to study dependents and they may be included in
future projects. In addition, the transfer of benefits is a relatively new option and presents a new
level of complexity to this field of research and practice. For example, one veteran can transfer
benefits to numerous dependents, each receiving a fraction of the benefit.
Student veterans who used VA education benefits other than the MGIB or Post-9/11 GI
Bill were also excluded. By focusing on the VA education benefits that are most utilized
currently, the MGIB and Post-9/11 GI Bill, the research captures a more comprehensive and
representative sample of todays student veterans. It is possible, though, that some individuals in
23
the sample had previously used the Department of Defenses (DOD) Tuition Assistance program
(TA) or a different VA education benefit program to earn postsecondary credit. However, there
is no method to match those in the MRP sample with information regarding TA usage, as it is a
DOD program. Parsimony was also a factor in this decision. Developing a sample of
beneficiaries using multiple VA education benefits programs and DOD data would have been
extremely difficult and, as this was essentially a pilot project, simplicity was of paramount
concern for SVA, VA, and the NSC. However, like veteran dependents, studies on other VA
education benefit programs as well as use of TA funds would add valuable information to this
field of research and can be included in future projects.
Students were defined as individuals registered for at least one course credit at a VA-
approved postsecondary institution or program. Verification of a student veterans enrollment
was conducted by VA. VA requires postsecondary institutions and programs to certify a student
veterans enrollment before it disburses funds. In addition, schools must notify VA if there are
any changes to a student veterans enrollment status, such as addition or subtraction of credits or
withdrawal from school.
Finally, the study defined a postsecondary institution or program as any school or
program that has been certified by VA to receive education benefit funds. This includes all types
of two-year and four-year institutions: public, private not-for-profit, and proprietary schools.
This also includes traditional brick-and-mortar institutions, online programs, vocational
certificate programs, and on-the-job training programs.
Variables
Outcomes of Interest. We focused on five indicators of student veteran postsecondary
academic performance, based on data obtained from the NSC.
24
Postsecondary Completion. This variable is an omnibus measure of student veteran
postsecondary completion. It includes completion from any postsecondary educational or
vocational programs that report such data to the NSC. Vocational certificates, associate-level
degrees, baccalaureate-level degrees, and postbaccalaureate degrees were thus analyzed.
Completion of Associate Degree. This variable focuses on the percentage of student
veterans whose first degree, as reported by the NSC, was an associate-level degree. It includes
students who attended public, private not-for-profit, and proprietary institutions. Student
veterans who initially enrolled in a postsecondary educational or vocational program after
December 31, 2010 were excluded from this analysis due to the two-year normative time-to-
completion for associate degrees; a student who initially enrolled after December 31, 2010 would
most likely not have had enough time to complete such a degree given the date of the data match.
Completion of Baccalaureate Degree! This variable focuses on the percentage of student
veterans who earned an initial baccalaureate level degree, as reported to the NSC. It includes
students who attended public, private not-for-profit, and proprietary institutions. Student
veterans who initially enrolled in a postsecondary educational or vocational program after 2008
were excluded from this analysis due to the four-year normative time-to-completion for
baccalaureate degrees; a student who initially enrolled after 2008 would have lacked sufficient
time to complete such a degree.
Initial Degree Level. This variable focuses on the individuals initial degree level. This
variable measures the level of the first degree an individual earned according to NSC records. It
provides a baseline for any future degrees the individual may have earned.
Highest Level of Education Completed. This variable focuses on each students highest
level of education as reported to the NSC. This variable measures the highest degree earned by
25
each veteran in the sample up until the date of the data match. The GI Bill may have an indirect
effect on a veterans overall level of education in that attainment of higher degrees would not be
possible if lower degrees were not earned with the help of the benefit. This variable expands
upon the previous variable by including any professional or graduate degrees a student veteran
earned, such as a masters or doctoral degree. However, it focuses only on the highest degree
earned, excluding all lower level degrees.
Time-to-completion! The time it takes a student veteran to complete a degree is perhaps
as important as earning the degree itself. Using data provided by the NSC, a time-to-completion
variable was constructed by subtracting the student veterans initial enrollment date from his or
her initial degree completion date (if available). Summary statistics, the average, the middle
response, and the most frequent responses, are reported for initial associate- and baccalaureate-
level degrees.
In addition to reporting the summary statistics for student veterans time-to-completion
for associate and bachelors degrees, a breakdown of student veterans time-to-completion based
on U.S. Department of Education guidelines is also reported. For associate level and other two-
year degrees and certificates, the percentage of student veterans in this sample who completed
their degree in two years (100 percent of time), three years (150 percent of time), and four years
(200 percent of time) is reported. For baccalaureate degrees, the percentage of student veterans
in this sample who completed their degree in four years (100 percent of time), five years (125
percent of time), and six years (150 percent of time) is reported.
Degree Field. In addition to reporting to the NSC whether students have completed their
program, schools also have the option of reporting the academic field of the degree using the
Department of Educations Classification of Instructional Programs (CIP) codes. While the full
26
CIP code is six digits, due to Personal Identifying Information (P.I.I.) concerns, only the first two
digits were released by VA to SVA via the FOIA request. However, the truncated CIP codes
allowed for broad comparisons among degree fields as well as discrete analyses.
The project analyzed and reported the top five most frequent degree fields for initial
associate and bachelors degrees from the sample. A secondary analysis and report on degree
fields measured the proportion of the sample that earned degrees in science and engineering
fields. The project used National Science Foundation guidelines for classifying CIP codes into
Science and Engineering categories and reported the percentage of the sample whose initial
degrees were in Science and Engineering fields at the associate and baccalaureate level.
Demographic Variables. Several independent variables were used in this study that will
provide descriptive information, but also will be used to explore differences between student
veteran subgroups.
GI Bill. Between 2002 and 2010, veterans were eligible for two GI Bill programs, the
Montgomery GI Bill (MGIB) and the Post-9/11 GI Bill. Due to eligibility periods and the
criteria for different VA education benefits programs, student veterans had the option of
switching from the MGIB to the Post-9/11 GI Bill, if they were eligible for the Post-9/11 GI Bill.
Therefore, three groups based on GI Bill eligibility were designated for this report: (1) MGIB
Exclusive, (2) Both MGIB and Post-9/11 GI Bill, and (3) Post-9/11 GI Bill Exclusive. However,
since the Post-9/11 GI Bill had been in existence for less than four years at the time of the data
match, some outcome results for those who used only the Post-9/11 GI Bill are not reported.
These student veterans have not had ample time to complete bachelors degrees and any results
would be considered premature and highly likely to change over the next few years.
27
Degree Attainment and Use of the GI Bill. Two variables were provided by VA. The
first variable indicated whether the individual earned a degree at any level prior to using their GI
Bill benefits. The second variable indicated whether the individual earned a degree after using
their GI Bill benefit. These variables provided insight into when and how veterans utilized their
GI Bill benefits to earn degrees.
Institutional Demographics. The data files analyzed for this project included limited
institutional nonidentifying demographic information, such as degree level offered (two-year or
four-year) and sector (public, private nonprofit, or proprietary). Comparisons were conducted
between the larger groups, such as two-year versus four-year schools and programs.
It is important to note the coverage rate of each sector in the NSC database when making
cross-sector comparisons. As of Dec. 30, 2013, NSCs coverage rate for the public and private
nonprofit sectors is high (99.2 percent and 93.1 percent, respectively). The coverage rate for the
proprietary sector is lower (67.5 percent). In addition, the coverage rates for the proprietary
sector increased between the years of the sampling from 48.2 percent in 2003 to 53.7 percent in
2010. The total effect the coverage rate may have on reporting results for the proprietary sector
is unknown, but the lower coverage most likely increases the potential error when reporting
results for this sector. Therefore, drawing conclusions based on the institutional sector analysis
should be done with caution.
Service Branch! A student veterans service branch was the final independent variable
used for comparisons. The VA collects a veterans previous service branch at the time of
application for VA benefits and then verifies it with the DOD. All five branches, Army, Air
Force, Navy, Marine Corps, and Coast Guard, were included in the dataset and used in the
comparisons. In addition to the five branches of the military, two agencies, the U.S. Public
28
Health Service and the National Oceanic and Atmospheric Administration (NOAA), are also
eligible for Post-9/11 GI Bill benefits. Although employees from these agencies represented a
small percentage of the overall sample and had little influence on the overall results, there was
no reason to exclude this group. They were included in the analyses.
Procedure
The sample was the result of a public-private partnership between VA, the NSC, and
SVA. The VA identified a sample of student veterans based on their use of VA educational
benefits (GI Bill usage). This sample was matched with NSC data that contains the degree
attainment records of approximately 96 percent of postsecondary students in the United States.
The NSC removed all institutional identifying information from the data prior to returning the
matched data to VA. SVA helped facilitate the partnership between VA and NSC.
Step one of the process required VA to properly identify student veterans for the sample.
VA created several filters to select appropriate individuals for inclusion. One filter excluded all
veteran dependents (spouses and children) that used the GI Bill, thus creating a list of only
veterans. Another filter excluded students who had received benefits while attending institutions
that were known not to report data to the NSC. This filter was established to increase the
samples coverage rate. However, it may have inadvertently created an oversampling of the
public and private nonprofit schools because of the low coverage rate among proprietary
institutions. Once the filters were in place, VA randomly selected 1 million veterans from its
education benefits databases who initially used their GI Bill benefits between 2002 and 2010.
Half of the sample focused on veterans who initially used MGIB benefits and the other half used,
at least in part, Post-9/11 GI Bill benefits.
29
Next, VA transferred the identifying information for the sample of student veterans, such
as name, date of birth, and social security number, to the NSC using secure file transfer
protocols. The VA followed current guidelines established by the Privacy Act of 1974, which
addresses the use of computerized databases that might affect the privacy rights of individuals,
and the NSC followed the Family Educational Rights and Privacy Act (FERPA) to ensure and
protect the privacy and anonymity of all individuals in the sample.
The NSC matched the data file provided by VA of veteran education benefit records with
its own records of degree attainment and other academic outcomes, where available. The data
match between the VA data file and the NSC database occurred on June 14, 2013. At that date
an estimated 75 percent of schools reported degree attainment data for the spring 2013
graduating class. Matching at this time excluded any summer degree completers that would
typically be included under the graduating class of 2013.
Of the 898,895 student veterans in the sample, the NSC found degree data for 859,297
individuals in its NSC data for a coverage rate of 95.6 percent. The 39,598 students without a
record in NSC data most likely attended an institution of higher education that does not report
postsecondary academic outcomes to the NSC. Missing academic data does not imply anything
about that students outcomes.
Once data from VA were matched with NSC data, NSC removed all institutional
identifying information, such as name of school, schools address, and the Office of
Postsecondary Education (OPE) code, and created a separate FERPA-compliant data file void of
all personal identifying information (e.g., name, date of birth, and social security number). NSC
returned both data files to VA.
30
SVA submitted a FOIA request to VA for access to the public version of the data. Once
the FOIA request was approved and the Veterans Benefits Administration ensured that any
personal identifying information had been removed based on guidelines established in the
Privacy Act of 1974, VA securely transferred the data using secure file transfer protocols to SVA
and it was then analyzed using SAS version 9.3.
31
Results
Demographics
Table 1 presents the demographic results of the Million Records Project alongside a U.S.
Department of Education report that profiles undergraduate students who used veteran education
benefits during the 2007-2008 academic year (U.S. Department of Education, 2009).
Approximately 1 in 5 (21.1 percent) individuals in the sample were female, which is not
unexpected considering that the DOD reports that 14.6 percent of current active duty personnel
are female (Department of Defense, 2013). As females generally have higher postsecondary
completion rates than men and the general student population has a more balanced ratio of men
to women than did the MRP sample, the MRPs results may not be directly comparable to
national completion rates. If the MRP sample included more women, it would have likely made
the samples overall completion rate rise and be more representative of the overall student
population. Over half of the sample (56.3 percent) were in their 20s when they first enrolled in a
postsecondary or vocational training program and nearly one-quarter (23.7 percent) were
between 25 and 29. This is similar to a Department of Education (2009) report that stated 45.9
percent of undergraduate students who received veteran education benefits were between 24 and
29. A direct age comparison with the MRP results, however, was not possible due to the
different age groups used in the Department of Education report.
Furthermore, the vast majority of the sample initially enrolled in either a public school
(79.2 percent) with the remainder split fairly evenly between private nonprofit institutions (10.7
percent) and proprietary schools (10.1 percent). Caution should be exercised when drawing
conclusions from this result; the NSCs coverage rate for proprietary schools is lower than it is
for the public and private nonprofit sectors, which may account for the higher proportion of
32
Table 1
Demographic Comparison between Million Records Project
1
and U.S. Department of
Education
2
Results
Demographic MRP Dept. of Ed

Female 21.1% 25.2%

Age at Initial Postsecondary Enrollment
Under 20 20.4% 0.2%*
20-24 32.6% 14.7%*
25-29 23.7% 45.9%*
30-39 17.4% 26.4%
40 or older 5.0% 12.9%

Initial School Sector Enrollment
Public School 79.2% 65.7%
Public, Non-Profit 10.7% 13.1%
Proprietary 10.1% 10.4%
Other or multiple schools ** 10.8%

Branch of Service
Air Force 18.1% **
Army 39.7%
**
Coast Guard 1.5%
**
Marine Corps 17.1%
**
Navy 23.5%
**
Other
3
<.1%
**
1
N = 859,297
2
Department of Education, 2009
3
NOAA and PHS both qualify for Post-9/11 GI Bill benefits
* U.S. Department of Education used different age groups: 18 or younger, 19-23, 24-29
** No Data
33
public school enrollment. As for branch of service, the Army comprised the largest segment of
the sample (39.7 percent), followed by the Navy (23.5 percent), Air Force (18.1 percent), Marine
Corps (17.1 percent), and the Coast Guard (1.5 percent).
Postsecondary Completion Rates
Of the entire Million Records Project sample (n = 859,297), 70,382 (8.2 percent) had an
earliest initial postsecondary enrollment date of Jan. 1, 2011 or later. These records were
excluded from all postsecondary academic outcome analysis because these veterans have likely
not had enough time to earn a postsecondary degree. Their inclusion would have artificially
skewed the samples overall completion rate. The remaining 788,915 records were analyzed. Of
this sample, a majority (407,483) indicated the attainment of a postsecondary degree, ranging
from a vocational certificate to a doctorate, for an overall student veteran completion rate of 51.7
percent.
Introducing another variablewhen the individual actually used the GI Bill benefit
depicts some interesting findings. Of those with a postsecondary completion (n = 407,483), 79.5
percent (n = 323,798) earned a certificate or degree after enrolling in VA benefits. However,
more than one-third (40.8 percent; n = 166,073) completed at least one postsecondary program
prior to using any GI Bill benefits. An example of this would be a service member who earns a
certificate through the DODs Tuition Assistance program while on active duty and then later
uses the GI Bill to earn another degree. Approximately two-thirds (64.6 percent) of these
166,073 veterans first earned an associate or lower-level degree and nearly half (49.6 percent)
then went on to use their GI Bill benefits to earn another postsecondary degree. When
evaluating these results, it is important to note that VA used an artificial cutoff date to determine
whether a degree was earned after or before enrolling in the benefit. This may have led to an
34
overestimate in the number of student veterans earning a degree after enrolling in the benefit and
an underestimate of those that completed a degree prior to using the GI Bill.
Time to Completion for Initial Degree
Of the sample reporting at least one postsecondary completion (n = 407,483), 18.9
percent of the records (76,859) did not include a valid earliest enrollment dateusually
because the students earliest enrollment occurred before the institution began submitting data to
the NSC. Therefore, 330,624 records were analyzed to determine time-to-completion for initial
degrees.
Table 2 displays the summary statistics (mean, median, and mode) for initial associate
and bachelors degree levels. The samples average time-to-completion for an associate-level
degree was 5.1 years. However, the middle response was four years and the most frequent
response was two years. The samples average time-to-completion for a baccalaureate-level
degree was 6.3 years, the middle response was five years, and the most frequent response was
four years. The large difference between the means and the other measures suggests that the
distribution of years is nonparametric; outliers may be influencing the overall mean. Discussion
of these potential outliers will be explored in the next section. These results also suggest that the
average is not the best measure time-to-completion for student veterans.
Table 2
Summary Results of Time-to-Completion for Initial Degree by Degree Level in Years
Degree Level Mean Median Mode
Associates Level 5.1 4.0 2.0
Bachelors Level 6.3 5.0 4.0
1
N = 144,343
2
N = 130,189
35
Initially, time-to-completion analysis followed the U.S. Department of Educations
standard reporting practices ending at the four-year mark for associate-level degrees and at the
six-year mark for bachelors degrees. However, the summary statistics for the samples time-to-
completion indicated that reporting beyond the traditional four-year and six-year benchmarks
was warranted as outliers may appear at higher intervals. Table 3 displays the samples time-to-
completion for those who earned an initial associate level and baccalaureate level degree;
certificate and graduate degrees were excluded due to the varying time rates within those degree
programs.

Approximately half (52.6 percent) of the sample who first earned an associate-level
degree did so within four years. For those in the sample who initially completed a bachelors
degree, a majority (50.5 percent) completed it within five years and a cumulative total of 59.4
Table 3
Cumulative Percentage of Time-to-Completion for Initial Degrees Earned
Completed In Associates Level
1
Bachelor Level
2

2 years or less 29.0% **
3 years 42.6% **
4 years 52.6% 40.0%
5 years 61.2% 50.5%
6 years 69.0% 59.4%
7 years ** 67.1%
8 years ** 74.2%
1
N = 144,343
2
N = 130,189
** No Data
36
percent completed it within six years. It is likely that individuals reporting extended times-to-
completion were not continuously enrolled and many factors, both personal and military-related,
may have contributed to their longer academic careers.
Highest Level of Education
Table 4 summarizes results for initial degree level earned and the highest level of
education completed regardless of GI Bill usage. Approximately 1 in 10 (10.3 percent) first
earned a certificate, nearly half of the sample (48.5 percent) first earned an associate-level
degree, and 38.6 percent first earned a baccalaureate-level degree. As for highest level of
education, however, nearly half of the sample (47.3 percent) earned a baccalaureate-level degree,
29.2 percent earned an earned an associate-level degree, and only 6.3 percent reported a
certificate as their highest degree. When comparing the initial degree levels with the highest
level of education a pattern appears: lower degrees (certificates and associate-level) decrease,
while higher degrees (baccalaureate-level and above) increase. Overall this suggests that student
veterans continue their education beyond the initial degree to earn higher degrees over the course
Table 4
Comparison of First Degree Earned and Veterans Highest Level of
Education
1

Level First Degree Highest Level of Education
Certificate 10.3% 6.3%
Associates 48.5% 29.2%
Bachelors 38.6% 47.3%
Masters 2.4% 15.7%
Doctorate ** 1.6%
1
N = 407,483
** No Data
37
of their lifetime.
Further evidence that veterans pursue higher levels of education over the course of their
lives is seen when comparing the first degree earned with future degrees earned. Nearly one-
third (31.3 percent) of those who initially earned a vocational certificate or diploma went on to
earn a higher degree. Similarly, 35.8 percent of those that initially earned an associate degree
later earned a higher degree. Finally, 20.8 percent of those who initially earned a bachelors
degree later earned a graduate level or doctoral degree.
Table 5 compares an individuals highest level of education with whether they earned a
degree before or after using GI Bill benefits. For those who earned any degree prior to using
their GI Bill benefits, 4.7 percent reported a certificate, 22.7 percent reported an associate-level
degree, 43.8 percent reported a baccalaureate-level degree, 25.8 percent reported a masters-level
degree, and 2.9 percent reported a doctorate as their highest level of education. For those who
earned a degree after enrolling in GI Bill benefits, 7.3 percent earned a certificate, 33.6 percent
earned an associate-level degree, 49.7 percent earned a baccalaureate-level degree, 8.8 percent
Table 5
Comparison of Highest Level of Education by Use of GI Bill Benefit
1

Degree Before Benefit Degree After Benefit Sample
Highest Level of Education
Certificate 4.7% 7.3% 6.3%
Associates 22.7% 33.6% 29.2%
Bachelors 43.8% 49.7% 47.3%
Masters 25.8% 8.8% 15.7%
Doctorate 2.9% 0.7% 1.6%
1
N = 407,483
38
earned a masters-level degree, and only 0.7 percent earned a doctorate as their highest level of
education. Those who earned a degree prior to using GI Bill benefits had higher levels of
education than those who relied solely on the benefit to earn a degree. Some caution should be
exercised in drawing conclusions from these results, as the actual date of enrollment in the
benefit was not available and VA used an artificial cutoff date to determine degree-attainment
prior to enrollment in the benefit. It is therefore difficult to conclusively identify individuals in
this sample who earned a degree after using their benefits.
Degree Fields
Tables 6 and 7 show the five most frequent degree fields, based on the CIP codes,
reported by institutions to the NSC, for those initially completing associate-level degrees and
baccalaureate-level degrees. While there is overlap in the degree fields between the associate
and baccalaureate levels, their order varies. At the associate-level, the most frequent degree field
was liberal arts and sciences (33.5 percent), followed by business (17.8 percent), homeland
security, law enforcement, and firefighting (9.7 percent), health professions (9.3 percent), and
engineering technologies (7.6 percent). By contrast, the most frequent degree field at the
Table 6
Percentage of Degree Field for Initial Associates Degree
1

Major Percentage
Liberal Arts and Sciences 33.5%
Business 17.8%
Homeland Security, Law Enforcement, Firefighting 9.7%
Health Professions 9.3%
Engineering Technologies 7.6%
1
N = 141,402
39
bachelors level is business (26.2 percent), followed by social sciences (10.8 percent), homeland
security, law enforcement, and firefighting (6.6 percent), computer and information sciences (5.8
percent), and health professions (4.9 percent).
Science and Engineering degrees. Using the National Science Foundations (NSF)
criteria for classifying U.S. Department of Education CIP codes to Science and Engineering
degrees, the MRP grouped the samples CIP codes into two categories: Science and Engineering
(S&E) degrees and non-Science and Engineering degrees (non-S&E). Table 8 displays the
Table 7
Percentage of Degree Field for Initial Baccalaureate Degree
1

Major Percentage
Business 26.2%
Social Sciences 10.8%
Homeland Security, Law Enforcement, Firefighting 6.6%
Computer and Information Sciences 5.8%
Health Professions 4.9%
1
N = 128,710
Table 8
Percentage of Science and Engineering Degrees by Degree Level
1

Degree Level N S&E Non-S&E
Total Sample
2
312,923 17.8% 82.2%
Associates level 141,402 7.5% 92.5%
Bachelors level 128,710 32.6% 67.4%
1
Science and Engineering as defined by NSF and based on initial degree obtained
2
Includes post-baccalaureate degrees as initial degrees in NSC records.
40
percentage of initial degrees earned in Science and Engineering fields by degree level. Analysis
shows that of the 312,923 records that contained a CIP code for the initial degree completion,
17.8 percent were Science and Engineering degrees. When accounting for degree level, the
number of Science and Engineering degrees awarded at the baccalaureate level (32.6 percent) is
drastically higher than at the associate level (7.5 percent).
Tables 9 and 10 compare the time-to-completion for Science and Engineering degrees,
non-Science and Engineering Degrees, and the entire sample based on initial degree level. At
the associate level, there was little variance in the times-to-completion between the three groups.
However, at the baccalaureate level, a higher percentage of the sample pursued bachelors
degrees in Science and Engineering completed their degrees faster than those in non-Science and
Engineering majors.
Table 11 compares the mean, median, and mode for the time-to-degree of Science and
Engineering degrees, non-Science and Engineering degrees, and the samples overall time-to-
completion. The median for non-Science and Engineering bachelors degrees (six years) is
Table 9
Cumulative Percentage of Time-to-Completion for S&E Associates Degrees
1

Completed In S&E Non-S&E Sample
2 years or less 28.5% 27.7% 29.0%
3 years 42.9% 41.3% 42.6%
4 years 52.1% 51.5% 52.6%
5 years 60.0% 60.2% 61.2%
6 years 67.5% 68.1% 69.0%
1
N = 141,402
41
higher than both the Science and Engineering bachelors degrees (five years) and the overall
sample (five years).

Table 10
Cumulative Percentage of Time-to-Completion for S&E Bachelors Degrees
1

Completed In S&E Non-S&E Sample
4 years or less 39.4% 37.1% 39.0%
5 years 52.5% 48.3% 50.5%
6 years 61.6% 57.3% 59.4%
7 years 69.0% 65.2% 67.1%
8 years 75.9% 72.6% 74.2%
1
N = 128,710
Table 11
Time-to-completion Summary Statistics for Science and Engineering Degrees by Degree
Level
Mean Median Mode
Associates Level Degree
Sample 5.1 4.0 2.0
S&E 5.3 4.0 2.0
Non-S&E 5.2 4.0 2.0

Bachelors Level Degree





Sample 6.3 5.0 4.0
S&E 6.2 5.0 4.0
Non-S&E 6.5 6.0 4.0
1
N = 312,923
42
Postsecondary Academic Outcome Comparisons
Limited comparisons across three demographic variables, initial school sector of
enrollment, branch of service, and type of GI Bill used, were also conducted.
Initial school sector. Figure 1 displays the completion rates for the entire sample
according to the school sector in which the veteran first enrolled. This is not necessarily the
sector in which they first used their GI Bill benefits or the sector in which they graduated from.
Of those in the sample that initially enrolled in a private nonprofit school, 63.8 percent earned a
postsecondary degree from any school sector; 50.8 percent of those that initially started at a
public school went on to earn a postsecondary degree; and 44.9 percent of the sample that
initially enrolled in a proprietary institution earned a postsecondary degree. As noted previously,
the NSC coverage rates for the three groups vary, with the proprietary sector having the lowest
coverage during the sampling timeframe (2002 to 2010). The exact effect this has on the results
is unknown, but most likely the margin of error for the proprietary sector results is higher than
for the other sectors and should be considered when interpreting the results.
50.8%
63.8%
44.9%
Public Private, Non-Profit Proprietary
Figure 1
Postsecondary Outcomes by Initial Enrollment Sector
(N = 788,915)
43
The proportion of postsecondary completions according to school sector also parallel the
enrollment percentages for each sector. The public school had the highest percentage of
completions with 71.7 percent. The private nonprofit sector had 15.5 percent of all sample
completions, and the proprietary sector had 12.9 percent. Again, the low coverage rate for the
proprietary sector during the sampling time frame should be considered when interpreting these
results.
Figure 2 displays the completion rates for those earning a degree within the same initial
sector of enrollment, the completion rates for those that earned a degree in a different sector than
their initial enrollment, and the percentage that did not complete. The public sector had the
highest within-sector retention rate (42.9 percent) for completions, followed by the private
nonprofit sector (42.2 percent), and the proprietary sector (36.8 percent). While private
nonprofit institutions had a high overall completion rate, it also had the highest migration rate
(21.6 percent). This suggests that approximately 1 in 5 student veterans that initially begin at a
two- or four-year private nonprofit institution eventually left the school and the sector entirely
42.9% 42.2%
36.8%
7.9%
21.6%
8.1%
49.2%
36.2%
55.1%
Public Private, Non-Profit Proprietary
Figure 2
Postsecondary Outcomes by Initial Enrollment Sector
(N = 788,915)
Completed: Same Sector Completed: Different Sector Non-Completion
44
before completing their first degree. It is also important to note that noncompletion does not
necessarily mean withdrawal from school. The individuals in this group may still be working
toward their degrees. The migration rates for the three sectors may also increase over time as
those currently classified as noncompleters earn a degree in a new sector.
Tables 12 and 13 display the time-to-completion for initial associate and baccalaureate
degrees for each sector of initial enrollment. While the proprietary sector had the lowest
completion rates of the three, they also had a larger proportion of students completing both
associate-level and baccalaureate-level degrees faster. This suggests that those who initially
enroll in a proprietary school and earn a degree do so more quickly than those who initially
enroll at a public or private nonprofit institution.
Figure 3 displays the highest level of education based on the initial school sector
enrollment. There was little difference across the sectors, with a bachelors degree being the
most frequent highest level of education. However, within each sector, private nonprofit schools
had higher rates of graduate degrees (28.5 percent) compared with public (15.9 percent) and
Table 12
Cumulative Percentages of Time-to-Completion for Initial Associates Degrees by Initial
School Sector of Enrollment
1

Completed in Public
Private
Non-Profit Proprietary Sample
2 years or less 25.8% 36.9% 59.3% 29.0%
3 years 39.5% 49.1% 72.5% 42.6%
4 years 49.8% 59.0% 80.1% 52.6%
5 years 58.6% 67.2% 85.5% 61.2%
6 years 66.8% 74.2% 89.7% 69.0%
1
N = 144,343
45
proprietary schools (11.3 percent). This could be due to the limited number of graduate
programs offered at proprietary schools compared with public and private nonprofit institutions.
Finally, an examination of the initial degree for Science and Engineering degrees by
Table 13
Cumulative Percentage Time-to-Completion for Initial Bachelors Degrees by Initial
School Sector of Enrollment
1

Completed in Public
Private
Non-Profit Proprietary Sample
4 years or less 32.3% 54.8% 68.7% 40.0%
5 years 44.4% 64.6% 78.3% 50.5%
6 years 53.8% 72.1% 85.0% 59.4%
7 years 62.3% 78.1% 89.5% 67.1%
8 years 70.2% 83.2% 93.1% 74.2%
1
N = 130,189
ubllc
rlvaLe non-
roL
roprleLary Sample
CraduaLe 13.94 28.81 11.69 17.31
8achelor's 46.00 32.31 31.60 47.29
AssoclaLe's 31.23 16.30 29.43 29.13
CerucaLe 7.23 2.18 7.23 6.23
0
10
20
30
40
30
60
70
80
90
100
!
"
#
$

&
#
'
(
)

Figure 3
Highest Level of Education by Initial Enrollment in School Sector
(N = 407,483)
46
initial school sector was conducted. Results found minimal variation between the sectors. The
rates for private, nonprofit, and proprietary schools were the same (18.8 percent) and slightly
above the samples average of 17.8 percent. The rate for the public sector (17.6 percent) was
slightly below the samples average.
Branch of service. Table 14 displays the completion rates for each branch of the
military. The analysis found that three of the five military branches had completion rates higher
than that of the overall sample (51.7 percent): Air Force (66.9 percent), Coast Guard (53.5
percent), and the Navy (51.9 percent). The Marine Corps had a completion rate of 44.9 percent
and the Army had a completion rate of 47.0 percent.
There was little variance in the highest level of education among military branches. The
highest degree most commonly earned across the five branches was a bachelors degree, with the
Marine Corps having the largest percentage (50.1 percent). The Air Force had the fewest
number of veterans reporting a bachelors degree as their highest degree (43.2 percent), but had
the highest proportion of post-baccalaureate degrees (20.8 percent).
Table 14
Branch of Military Service Completion Rates
Branch of Service Completion Rates Sample Size
Air Force 66.9% 147,403
Army 47.0% 312,561
Coast Guard 53.5% 12,299
Marine Corps 44.9% 130,958
Navy 51.9% 185,525
Other
1
91.1% 169
1
NOAA & PHS
47
Tables 15 and 16 compare the time-to-completion for initial associate-level and
baccalaureate-level degrees by branch of service. There was little variation in the time-to-
completion rates among the branches. A majority of veterans in nearly all five branches
completed their associate degree in four years, with the Army (49.8 percent) falling just below
50 percent. A similar result was found for bachelors degrees, with all five branches having at
least a majority completing their degree within six years.
Table 15
Cumulative Percentages Time-to-Completion for Initial Associate Degrees by Branch of
Service
1

Completed in Air Force Army Coast Guard Marine Corps Navy
2 years or less 30.8% 26.2% 29.7% 30.2% 30.1%
3 years 43.3% 39.8% 42.8% 45.1% 44.0%
4 years 53.7% 49.8% 52.4% 54.9% 54.0%
5 years 62.8% 58.5% 60.0% 63.1% 62.1%
6 years 71.3% 66.4% 66.7% 70.3% 69.4%
1
N = 144,343
Table 16
Time-to-Completion for Initial Bachelors Degrees by Branch of Service
1

Completed in Air Force Army Coast Guard Marine Corps Navy
4 years or less 41.7% 36.6% 37.3% 37.1% 42.4%
5 years 53.9% 48.1% 48.1% 49.1% 53.0%
6 years 62.6% 57.4% 56.2% 58.5% 61.2%
7 years 69.8% 65.5% 63.6% 66.7% 68.4%
8 years 76.5% 73.0% 70.6% 74.2% 74.9%
1
N = 130,189
48
Analysis of Science and Engineering degrees among the branches found that all five have
rates similar to the overall samples average. The Marine Corps and Navy had the highest rate of
degrees in the Science and Engineering fields (both at 19.3 percent), followed by the Army (18.6
percent). All three were above the sample average of 17.8 percent. The Coast Guard was
slightly below the sample average at 17.0 percent and the Air Force had the lowest rate of
degrees (14.1 percent) in Science and Engineering fields.
GI Bill usage. Although it is premature to draw any definitive conclusions about the
Post 9/11 GI Bill or its comparison with other VA education benefits, due to its recent
implementation, some comparisons were conducted to explore early trends between the MGIB
exclusive group and the hybrid MGIBPost-9/11 GI Bill group.
Table 17 displays the comparison of academic outcomes between the two groups of GI
Bill beneficiaries. A comparison of student veterans who used only Montgomery GI Bill
benefits versus student veterans who used both the MGIB and Post-9/11 GI Bill shows minor
differences between the two groups. Student veterans who used portions of both GI Bills had a
slightly higher completion rate (57.8 percent to 59.7 percent) and a slightly higher percentage of
Science and Engineering degrees (17.6 percent to 18.9 percent). These student veterans also had
a higher proportion of baccalaureate degrees as their highest level of education compared with
student veterans who used only the MGIB. However, student veterans using only the MGIB had
a greater proportion of masters and doctorate degrees, which may indicate that they have been
attending postsecondary education and vocational programs longer and that this difference will
narrow over time.

49
Table 17
Comparison of Postsecondary Outcomes by GI Bill Chapter Used

MGIB Both
Post 9/11
GI Bill Sample
Completion Rates
1
57.8% 59.7% ** 51.7%


Highest Level of Education
2

Certificate
6.8% 4.8%
**

6.3%
Associates
28.0% 26.8%
**

29.2%
Bachelors
46.9% 52.8%
**

47.3%
Masters
16.6% 14.3%
**

15.7%
Doctorate
1.7% 1.3%
**

1.6%


Science and Engineering Degrees
3

S&E
17.64% 18.20%
**

17.84%
Non-S&E
82.36% 81.80%
**

82.16%
1
N = 788,915
2
N = 407,483
3
N = 312,923
** Not enough data for comparison
50
Discussion
The Million Records Project is intended to increase the knowledge of stakeholders,
policymakers, and the general public with regards to the postsecondary academic outcomes of
this generations student veterans. For the first time in the history of the GI Bill, empirical
evidence has been obtained on the academic achievements of current student veterans, and it
offers a glimpse into potential future trends. It is also anticipated that the Million Records
Project will have substantial policy, practical, and research implications.
While this report provides an in-depth look at the postsecondary outcomes of student
veterans, some weaknesses in the study should be noted. The inclusion of only Montgomery GI
Bill (MGIB) and Post 9/11-GI Bill beneficiaries limits the reports findings to these specific VA
education benefits programs. Caution should be used when generalizing these findings to other
VA education programs, such as the Montgomery GI Bill Selective Reserve (MGIB-SR) benefit
or the Veterans Educational Assistance Program (VEAP). In addition, including other education
benefits programs, such as DODs Tuition Assistance, would show in greater detail the total
amount of financial assistance student veterans received due to their military service, which
would allow for a more robust discussion.
Another weakness of this report is the lack of data on the samples enrollment status.
The results are based on certificate or degree attainment. Including enrollment information will
allow for more accurate results when measuring persistence rates, sector migration rates, and the
number of non-completers who are, in fact, still enrolled in higher education programs.
The exclusion of enrollment intensitypart-time or full-time enrollmentis another
weakness. This data point, connected to enrollment, would provide additional information and
allow for stronger conclusions when evaluating the time-to-completion results. For example, this
51
variable would help researchers determine if a veterans long time-to-completion is a result of
military service, enrollment intensity, or some combination of the two.
An additional weakness is that some of the variables did not have complete, detailed
information. For example, the specific date when an individual enrolled in the GI Bill benefit
was not provided. The VA rather used a cutoff date of 2002. Anyone having a degree prior to
that date was assigned to the earned degree prior to benefit group and those not having a
degree prior to the cutoff point were assigned to the earned degree after benefit group. This
skews the results by potentially undercounting the earned degree prior to benefit group and
overestimating the earned degree after benefit group.
Finally, the low coverage rate for the proprietary sector in the NSC database makes
comparisons across sectors difficult, as the margin of error is higher for proprietary schools.
Therefore, conclusions and discussions based on sector comparisons should be done with caution
as a higher coverage rate could lead to different results.
The Million Records Project also has a number of key strengths that should be noted.
First, the large sample size of 898,895 veterans over a span of 10 years produces an extremely
low margin of error and strong population estimates, which allows for a robust discussion about
the data with little worry about misinterpreting the results due to sampling error. Second, VAs
random selection of the sample also reduced potential sampling error while increasing the
external validity of the results. Third, by using the NSCs database, the project obtained degree
data based on records from institutions of higher education, thus eliminating potential response
bias from individualsthough, as noted above, the low coverage rate for proprietary schools
does create a response bias based on school sector and should be considered when interpreting
the results.
52
Postsecondary Academic Outcomes
A majority of the sample earned a postsecondary degree or certificate, for a
postsecondary completion rate of 51.7 percent, which is comparable to the 2010 NSV results.
While it is atypical to use an average across several years of postsecondary completion data, the
purpose of this study was to gain a better measure of the postsecondary academic outcomes of
the Post-9/11 generation of student veterans, and a group average over the era is the most
appropriate statistic for that purpose.
The MRP design allows for analysis of the postsecondary academic outcomes for
individual academic year cohorts, but the results are likely to differ from other national-level
statistics on postsecondary completion rates. Methodological differences are a main cause of
such variance. Traditional databases, such as IPEDS, NCES, and NSC, track postsecondary
completion rates using a cohort model with specific enrollment and completion dates, a sample
limited to first-time students, and a time-to-completion capped at six years. By contrast, the
MRP sample studied student veterans over many years and included individuals with previous
degree attainments. If the same methods used by IPEDS and NCES were applied to this sample
of student veterans, a lower completion rate for student veterans would be expected, because
more of them would be excluded or listed as dropouts if their time-to-degree was longer than six
years.
However, the goal of the MRP was not to directly compare student veterans with other
populations but rather to begin to fill the void in research on veterans in higher education by
providing a more accurate measure of student veteran postsecondary outcomes. This report is a
necessary first step. A cohort-based design would require several years of examination to
determine trends and it may not fully capture external influences like increased deployments.
53
This overview of the postsecondary academic outcomes of student veterans sets the stage for
future studies to make comparisons using a more traditional cohort-based approach.
In addition, using a group average of postsecondary completion for the service era allows
for parallel comparisons with previous research on student veteran postsecondary completion
rates from earlier eras. The 2010 National Survey of Veterans, when selecting only student
veterans who served after September 2001, notes that 51.5 percent reported completing their
educational or vocational program for which they used GI Bill education benefits (Cate, 2014).
The completion rate reported in the MRP is similar to the 51.5 percent, but slightly lower than
the completions rates reported for other service eras in the 2010 NSV, which are approximately
65 percent. The difference could be a function of time; those service eras have had more time to
enroll and complete a degree than many of the veterans in the MRP sample. This gap is likely to
close over time. The MRPs findings not only provide updated and more comprehensive data,
but the results stand in stark contrast to 2012 media reports that claimed that 88 percent of Post-
9/11 student veterans were dropping out of college in their first year.
Although direct comparisons with other completion rate statistics are not possible due to
differing methodologies, the Million Records Projects completion rate is far higher than what
may be expected for nontraditional students. A 1996 study of nontraditional student completion
rates found that within five years only 44.3 percent of the sample earned any type of
postsecondary certificate or degree (Horn, 1996). In fact, student veterans are attaining degrees
at a rate similar to that of all studentstraditional and non-traditional. However, it is taking
longer to reach that mark. The National Student Clearinghouse reported a 56.1 percent national
completion rate for the AY 2007 cohort over a six-year period (Shapiro, Dundar, Zisken, Yuan,
& Harrell, 2013). The MRP sample required longer to reach comparable levels of completion,
54
which is not surprising considering the military deployments over the course of the sample. This
suggests that at a postsecondary academic outcome level, student veterans may resemble
traditional students rather than their nontraditional counterparts, and are very persistent. They
may withdraw from college for military or personal reasons, but will return. Again, caution
should be exercised in making direct comparisons between these cohort designed studies and the
MRP methodology.
In addition, this samples 51.7 percent completion rate will most likely continue to rise.
The MRP measured postsecondary academic outcomes as of June 14, 2013 with incomplete data
from the graduating class of 2013 and no data on summer completers. Student veterans enrolling
in school closer to 2010 or after, which is a large percentage of Post-9/11 GI Bill beneficiaries,
would not have had enough time to earn a degree or certificate and are not recorded in the NSCs
database as attaining a certificate or degree. Therefore, as student veterans who recently enrolled
in institutions of higher education have the opportunity to finish, this samples overall
completion rate should increase. The 51.7 percent completion rate is, however, a benchmark
based on empirical evidence.
The average time-to-completion for student veterans who first earned an associate degree
was 5.1 years. For those who first earned a bachelors degree, the average time-to-completion
was 6.3 years. Both data points indicate that some veterans may require slightly more time to
complete a degree than the traditional time-to-completion expectations established by the U.S.
Department of Education. However, it is important to remember that time-to-completion does
not mean that the student was continuously enrolled and it is more likely that the student veteran
had to withdraw and change from full-time to part-time.
55
It is also important to note that the averages are influenced by outliers, and the median
and mode are better indicators of a student veterans time-to-completion. For student veterans
who first completed an associate degree, a majority did so in four years or less with the most
frequent time-to-completion being two years. For those first completing a bachelors degree, a
majority did so in five years, but the most frequent time-to-completion reported was four years.
This indicates that a majority of student veterans are earning their degrees in the U.S.
Department of Educations established timeframe, while a smaller but important minority, as
noted by the longer average time-to-completion, requires a few extra years.
Based on the MRP time-to-completion results and other research (Ackerman, et al., 2009;
DiRamio, et al., 2008), it can be hypothesized that student veterans follow different paths to
degree attainment, which is evidenced by the varying times-to-completion for this sample.
Figure 4 visually represents at least three such paths. Recognizing these three pathways will
allow institutions of higher education to properly plan for changes in a student veterans
enrollment over the course of their academic career.
Path Alpha is the traditional path recognizable to most people. An individual enters the
military after high school, fulfills his or her service obligations, and upon separation enrolls in
higher education. With their military service complete, they are able to persist to graduation
relatively uninterrupted by military obligations. Student veterans on this path typically earn an
associate-level degree in two years and a baccalaureate-level degree in four. Active duty service
members are the largest segment of the military and therefore it is not surprising that most
student veterans probably follow this path as evidenced by the median and mode time-to-
completion.
56
Some evidence also suggests that being a reservist in college and being called up can
significantly disrupt a student veterans postsecondary progress (Ackerman, et al., 2009;
DiRamio, et al., 2008). Path Beta reflects the increased use and incorporation of reservist and
National Guard units in todays military. In this path, an individual enrolls in college after
joining the military as a reservist or National Guard member or vice versa. In both cases, the
student veteran is at risk of being activated during the school term, which is what occurs in this
path. The student veterans military obligation causes a temporary stop-out. Depending on
when the activation occurs and the specific institutions policies, the student veterans academic
progress for that term could be negated and listed as a withdrawal (shown as red and black lines
in Figure 4). The student veteran deploys for the designated amount of time, is deactivated,
returns home, and re-enrolls in school. Student veterans in some instances may have to wait
months before continuing their studies if their unit is deactivated in the middle of a semester
(shown as a gray block in Figure 4). Finally, the student veteran continues his or her studies, but
may face several more activations over the course of the academic career.
In the scenario just described, a total of 20 months may have been artificially added to the
students time-to-completion: four months of credits lost due to activation during the term, 12
months deployed, and then another four months waiting to re-enroll. What was a two-year
degree quickly becomes a four-year associate degree and the four-year normative time-to-
completion for a bachelors degree may take six years to completeprovided that the student
veteran is activated only once. Student veterans that follow Path Beta may account for some of
the high average time-to-completion found in this study. In addition, since several Army and
Marine Corps reservists and National Guard units were activated numerous times during
57


F
i
g
u
r
e

4

S
t
u
d
e
n
t

V
e
t
e
r
a
n

H
y
p
o
t
h
e
s
i
z
e
d

C
o
m
p
l
e
t
i
o
n

P
a
t
h
s

P
a
t
h

A
l
p
h
a
:


S
t
u
d
e
n
t

v
e
t
e
r
a
n

j
o
i
n
s

t
h
e

m
i
l
i
t
a
r
y
,

f
u
l
f
i
l
l
s

h
i
s

c
o
n
t
r
a
c
t

t
o

t
h
e

m
i
l
i
t
a
r
y

a
n
d

o
n
c
e

c
o
n
t
r
a
c
t

i
s

o
v
e
r

e
n
r
o
l
l
s

i
n

p
o
s
t
-
s
e
c
o
n
d
a
r
y

e
d
u
c
a
t
i
o
n

o
r

v
o
c
a
t
i
o
n
a
l

p
r
o
g
r
a
m
.


T
y
p
i
c
a
l
l
y

t
h
e

p
a
t
h

t
h
a
t

m
o
s
t

a
c
t
i
v
e

d
u
t
y

s
e
r
v
i
c
e

m
e
m
b
e
r
s

f
o
l
l
o
w
,

a
l
t
h
o
u
g
h

s
o
m
e

m
a
y

t
a
k
e

c
o
r
r
e
s
p
o
n
d
e
n
c
e

c
o
u
r
s
e
s

o
r

o
n
-
b
a
s
e

c
o
u
r
s
e
s
.

M
i
l
i
t
a
r
y

S
e
r
v
i
c
e

A
t
t
e
n
d
i
n
g

S
c
h
o
o
l

P
a
t
h

B
e
t
a
:


S
t
u
d
e
n
t

v
e
t
e
r
a
n

j
o
i
n
s

t
h
e

r
e
s
e
r
v
e
s

o
r

n
a
t
i
o
n
a
l

g
u
a
r
d
,

b
u
t

a
t
t
e
n
d
s

a

p
o
s
t
-
s
e
c
o
n
d
a
r
y

e
d
u
c
a
t
i
o
n

o
r

v
o
c
a
t
i
o
n
a
l

p
r
o
g
r
a
m

a
t

t
h
e

s
a
m
e

t
i
m
e
.


A
c
t
i
v
a
t
i
o
n

o
f

s
t
u
d
e
n
t

s

u
n
i
t

m
a
y

c
a
u
s
e

b
r
e
a
k
s

a
n
d

d
i
s
c
o
n
t
i
n
u
i
t
i
e
s

i
n

t
h
e

s
t
u
d
e
n
t

v
e
t
e
r
a
n

s

p
o
s
t
-
s
e
c
o
n
d
a
r
y

e
n
r
o
l
l
m
e
n
t
.


D
e
p
e
n
d
i
n
g

o
n

t
h
e

i
n
d
i
v
i
d
u
a
l

s
c
h
o
o
l

s

p
o
l
i
c
y

a
n
d

w
h
e
n

u
n
i
t

s

a
c
t
i
v
a
t
i
o
n

o
c
c
u
r
s
,

s
t
u
d
e
n
t

v
e
t
e
r
a
n
s

m
a
y

l
o
s
e

n
e
a
r

a

t
e
r
m

o
f

w
o
r
k
.


A
l
s
o
,

d
e
p
e
n
d
i
n
g

o
n

w
h
e
n

u
n
i
t

i
s

d
e
a
c
t
i
v
a
t
e
d
,

s
t
u
d
e
n
t

v
e
t
e
r
a
n

m
a
y

w
a
i
t

m
o
n
t
h
s

b
e
f
o
r
e

r
e
t
u
r
n
i
n
g

t
o

s
c
h
o
o
l
.











P
a
t
h

G
a
m
m
a
:


S
t
u
d
e
n
t

v
e
t
e
r
a
n

i
n
i
t
i
a
l
l
y

e
n
r
o
l
l
s

i
n

a

p
o
s
t
-
s
e
c
o
n
d
a
r
y

s
c
h
o
o
l
.


W
i
t
h
d
r
a
w
s

f
r
o
m

s
c
h
o
o
l

o
r

p
r
o
g
r
a
m

a
n
d

j
o
i
n
s

t
h
e

m
i
l
i
t
a
r
y
.


A
f
t
e
r

s
e
r
v
i
c
e

r
e
-
e
n
r
o
l
l
s

i
n

s
c
h
o
o
l
.





58
Operation Iraqi Freedom and Operation Enduring Freedom, this path may also explain their
currently lower rates of completion compared to the other service branches.
Path Gamma offers another explanation for the long average time-to-completion. In this
path, an individual enrolls in an institution of higher learning, but leaves to join the military as an
active duty service member. Once the service member fulfills his or her military contract and
separates, they re-enroll in higher education to earn a degree. Due to the increase in military
enlistments following the terrorist attacks of Sept. 11, 2001, student veterans following this path
may account for a larger percentage of the sample than in previous service eras. For those who
extended their length of service beyond their initial contract, their time-to-completion could
range from several years to more than a decade. Such cases account for some of the outliers
found in the data that also contributed to the higher average time-to-completion for student
veterans.
Aside from the pathways, the results also indicate that the GI Bill benefit may not be the
main motivation driving veterans to enroll in higher education. A high percentage (40.8 percent)
of the student veterans who earned a postsecondary degree or credential in the sample earned the
degree prior to using VA education benefits. These individuals may have completed a degree
prior to ever being eligible for VA benefits, which suggests that todays military is comprised of
well-educated individuals. In addition, two-thirds (64.5 percent) of those earning a degree prior
to accessing VA benefits completed a two-year associate or vocational certificate. They may
have done so using the DODs Tuition Assistance program to earn a degree while on active duty.
It could also indicate that student veterans are conscientious consumers of their benefits. If they
recognize that attaining a lower-level degree at a reduced cost can also fulfill general education
59
requirements at a four-year institution, student veterans may choose to save GI Bill benefits for
use at more costly schools and to pursue higher degrees.
The results further support this idea with nearly half (49.6 percent) of those who earned a
degree prior to accessing their GI Bill benefits also earning a degree after usage. For the entire
sample that completed a degree, either at the baccalaureate or associate level, 79.4 percent did so
after accessing their benefits. While the motivation for enrolling in higher education or the
motive behind using ones VA benefits may be unclear and outside the scope of this project, the
data suggest that the GI Bill incentivizes veterans to persist in earning postsecondary degrees and
leads to increases in a veterans overall level of education. Further research into when and why
veterans choose to use their VA education benefits would provide much needed clarity and
answer many of these questions.
One-third (33.4 percent) of student veterans in the sample earned a bachelors degree or
higher, which is slightly lower than the results of the 2012 American Community Survey (U.S.
Census, 2012), which reported 42 percent of the entire U.S. veteran population who attended
some college earned a bachelor degree or higher. However, as more time elapses and more
student veterans in the MRP sample progress through education, the percentage of baccalaureate-
level degrees and higher are expected to rise and this gap will most likely narrow. While the GI
Bill may not be capable of fully funding a student veterans postsecondary education, it does
provide the foundation, both in terms of knowledge acquisition and financial freedom, on which
veterans can pursue additional higher-level degrees like their post-WWII era counterparts who
went on to become doctors, engineers, and researchers (Humes, 2006). In doing so, they are
likely to be more competitive in the civilian workforce and earn higher wages over their lifetime
(Bureau of Labor Statistics, 2010).
60
An analysis of degrees pursued, with a focus on Science and Engineering fields, also
sheds light on the employability of student veterans. Todays veterans appear to be earning
degrees in a few specific fields, as evidenced by the overlap in the five most frequent degree
fields for associate- and bachelor-level degrees. Three degrees appear in the top five for both
degree levels, although in different order: business; homeland security, law enforcement, and
firefighting; and health professions. This suggests that a large segment of the sample will be
entering the business or service sectors of the workforce. Also, the high frequency of student
veterans pursuing liberal arts and sciences, as defined by the U.S. Department of Education, at
the associate level suggests that many attend a two-year institution to complete their general
education requirements and may later transfer to a four-year university to earn a higher degree.
Of the sample with a degree completion and a degree field record, 17.8 percent earned a
Science and Engineering degree, as defined by the National Science Foundation. At the
associate-level, the percentage of Science and Engineering degrees decreases to 7.4 percent.
However, it increases to 32.6 percent when looking only at those who first completed a
bachelors degree. According to NSF data, the average number of Science and Engineering
bachelors degrees conferred to all students between 2002 and 2010 was approximately 31.8
percent (National Science Foundation, 2013). The MRP results suggest that student veterans are
attaining bachelor-level Science and Engineering degrees at a rate similar to their civilian
counterparts.
In addition, the study examined differences in time-to-completion for baccalaureate
students in Science and Engineering programs with their peers pursuing other bachelors
degrees. The results indicate that those majoring in Science or Engineering completes their
degrees more quickly than student veterans in non-Science or Engineering fields. This may be
61
more a result of the structured course schedules of many Science and Engineering programs that
minimize the amount of electives students can take. However, the limited number of months that
the GI Bill will pay for classes is another incentive for this group to finish faster. Currently, the
GI Bill allocates funds for only 36 months of classes at the full-time enrollment level
approximately four years of college. MRP results show that only 39.4 percent of the sample
completed their Science and Engineering bachelors degree within four years. While a higher
rate than their non-Science and Engineering counterparts, approximately 20 percent still needed
an additional one to two years to finish their degree. If attending at a continuous, full-time rate
for their entire college career, these students are likely to have exhausted their benefits prior to
completion and paid for the additional cost out of their own pocket or utilized student loans.
The Million Records Project also compared student veteran postsecondary completion
rates based on the sector of higher education in which the student initially enrolled. The results,
while not definitive due to the varying coverage rates previously discussed in this report, show
that private nonprofit institutions, public schools, and proprietary schools can all improve their
postsecondary completion rates. However, some sectors may need to improve more than others.
Based on the sector of a student veterans initial enrollment, the private nonprofit sector has the
highest student veteran completion rate at 63.8 percent, followed by public schools at 50.1
percent, and the proprietary sector with the lowest at 44.9 percent. Essentially this indicates that
if student veterans initially enrolls at a public or private nonprofit institution of higher learning,
they are more than likely to earn a degree over the course of their lifetime from any school. The
converse is potentially true for those first enrolling in a proprietary school; they may be less
likely to complete a degree in their lifetime from any school.

62
These completion rates are based on the school sector in which the student veteran first
enrollednot where they first completed their degree (See Figure 3). When sector transfer rates
are applied to these results, the rates become nearly even across the three sectors. In addition,
the private nonprofit sector receives a large increase from students who completed their degrees
in a different school sector. Sector migration is a broad exploration and a poor indication of
student veteran transfer rates. More detailed initial enrollment level data and enrollment data in
general would allow for discrete analyses, allow researchers to note non-completer students that
are still enrolled and persisting, and generate stronger results and interpretations.
In evaluating sectors, it is important to recall that the coverage rates for the proprietary
sectors was lower compared with the other two sectors. This created a potential sample bias
skewing data towards the public and private nonprofit sectors and should be considered in the
discussions of sector enrollment. The MRP results show that most student veterans (79.2
percent) initially enrolled in public schools and roughly 1 in 10 initially enrolled in either a
private nonprofit (10.7 percent) or proprietary institution (10.1 percent). Such a large disparity
in enrollment means that not all increases in completion rates by sector are equal. For example,
a 1 percentage point increase in completion rates in the public sector is equal to at most eight
times as many student veterans graduating in the private or proprietary sector if those sectors
experience the same one point increase. This information should help guide the allocation of
scarce resources to maximize impact.
It is also critical to note that the completion rates are reported at the sector level and
detailed analysis by school was out of the scope and ability of this report. Therefore, individual
schools within a given sector may have significantly different postsecondary completion rates
than the sector completion rates reported here. Although the private nonprofit sector has the
63
highest completion rate, it is probable that there are private nonprofit schools with lower student
veteran postsecondary completion rates. Conversely, the proprietary sector has a lower
postsecondary completion rate, but it is likely that some proprietary institutions graduate
veterans at a high rate.
In summary, the analysis shows that student veterans graduate at significantly higher
rates than their nontraditional student peers and perform at a level comparable to traditional
students. Their time-to-completion varies and reflects the fact that some student veterans follow
different paths to degree attainment. The results also suggest that student veterans may strongly
consider when to use their limited GI Bill benefits to maximize impact, as evidenced by those
who delay usage until enrolled at a more costly institution or to pursue a higher-level degree.
The findings also suggest that GI Bill usage leads to higher levels of education over the course of
a veterans lifetime. Last, the largest percentage of student veterans graduating with bachelors
degrees are preparing for careers in business, public service, science, and engineering fields.
Policy Implications
The Million Records Project has numerous public policy implications. Most notably is
the fact that this report establishes a proven methodology to empirically and accurately measure
the impact of the GI Bill in near real time. Prior to the Million Records Project, policymakers
and stakeholders had to wait years or even decades to know the effects of a specific GI Bill
program. Not only has the Million Records Project drastically reduced that time to a fraction of
what it once was, but this process is replicable, valid, comprehensive, and cost-effective.
The MRP also empowers policymakers and stakeholders to make data-driven decisions
regarding the future of the Post-9/11 GI Bill, which is one of Americas most robust investments
in the post-service success of veterans in history. The MRP provides empirical evidence that
64
student veterans are using their GI Bill benefits to earn postsecondary certificates and degrees.
However, it is important to note that student veterans may need more time to complete their
certificates and degrees due to personal and military obligations. For some student populations,
breaks in enrollment or changes in enrollment intensity may results in permanent withdrawals
from school. Yet for veterans, the MRP results suggest that they are persistent and will continue
through to degree attainment.
As noted above, the Million Records Project also analyzed the performance of student
veterans enrolled in different sectors of higher education. The results mainly indicate that
greater research is needed to understand how and why student veterans persist in each sector, but
the level of student veteran enrollment in each has significant policy implications. Far more
student veterans are enrolling in public institutions than in private or proprietary schools. That
means that even a small increase in the completion rate of veterans at public schools will yield a
larger number of veterans actually earning degrees. Stakeholders may use this data to allocate
limited resources more effectively.
It is also well known that America is facing a serious shortage in the number of students
graduating with Science and Engineering degrees. Following WWII, thousands of student
veterans earned such degrees and then made numerous advances in medicine, technology,
mathematics, and other fields. Todays veterans are well-equipped to pursue such degrees to fill
the skills-gap plaguing the nation. They not only have experience using modern technology
from their time in the military, but many have security clearances, are well-versed in other
languages and cultures, and are now earning degrees and credentials in these critical fields. The
MRP found the percentage of student veterans earning bachelors degrees in Science and
Engineering is basically even with the general population. With this baseline established, work
65
can begin to increase the number of student veterans earning Science and Engineering degrees,
thus helping to close the skills gap.
The findings showed that nearly 1 in 5 student veterans took five to six years to complete
a bachelors degree. The GI Bill only allocates funds for 36 months or approximately four
academic years if the student veteran attends full-time for all months, meaning that many student
veterans may be at risk of dropping out, switching majors, or taking on student debt to complete
their degrees before their GI Bill is exhausted. If legislation were passed that granted Science
and Engineering students an extra year of GI Bill benefits, it would incentivize more veterans to
pursue such majors and increase the likelihood that they complete these high-demand degrees. It
would also potentially reduce the number of student veterans who may switch from a Science
and Engineering degree to another field because their GI Bill benefits will be depleted prior to
completion. More research, however, is needed to explore potential factors that increase the
number of student veterans earning Science and Engineering degrees.
Research Implications
The Million Records Project has implications for the research community as well. Those
examining veterans and military service members in higher education, as well as the larger
population of nontraditional students, now have a new method for analyzing postsecondary
outcomes. The NSC can provide robust data not only on degree completion, but also on
enrollment. Furthermore, the MRP results indicate that research on student veteran
postsecondary academic outcomes should utilize expanded timeframes for completion beyond
those used by the U.S. Department of Education. If researchers use the U.S. Department of
Educations timeframes, a significant proportion of student veterans may inadvertently be
missed.
66
The Million Records Project also adds a new dimension to the study of nontraditional
students and the classification of student veterans as nontraditional students. Even though they
share many of the same traits, as defined by Horn (1996), traits that typically result in few
postsecondary completions, student veterans are earning postsecondary degrees at rates higher
than their nontraditional counterparts. Contrary to theoretical predictions, student veterans are
actually succeeding in postsecondary environments at the same level as their traditional student
peers. This topic should be further explored to discover the factorsboth character traits of
student veterans and institutional supportsthat lead to positive academic outcomes and how to
replicate such practices for other nontraditional students, thereby increasing the overall
completion rate.
The results also empower researchers interested in exploring veterans in higher
education. Prior to the Million Records Project, a main weakness of many research articles
focusing on student veterans was the inability for a researcher to state if his or her sample was
representative of the entire student veteran population. The MRP provides researchers limited
national level statistics to use when developing studies, selecting samples, and comparing results.
This may lead to a snowball effect of stronger research on veterans in higher education and a
rapid expansion of the currently limited knowledge base on these students, which in turn can
improve and inform both policy and practice.
Implications for Practice
Last, the Million Records Project will have countless practical implications, especially
for institutions of higher education that enroll student veterans.
In addition to shedding light on possible paths to completion, the MRP is a blueprint that
schools can replicate to track the academic outcomes of their student veterans. They can then
67
compare their results with both a national completion rate and to the completion rate for their
sector of higher education. The results will empower administrators to smartly allocate scarce
resources. For example, if the school is graduating veterans at a high rate, it may have reached a
point of diminishing returns. The school should, most likely, sustain current levels of support. If
a school has a low completion rate, that school may need to increase support. Future work, as
described in the next section, will assist schools in determining exactly what programs and
service to invest in.
Furthermore, many schools face budget cuts and have had to prioritize the types of
programs and services they provide to students. Since there have been limited data on student
veteran completion rates, some schools may question the value of allocating resources for this
population at all. The fact that student veterans are completing at high rates, however, indicates
that an investment in supportive services is likely to attract more student veterans, increase
persistence, and thus increase the total amount of GI Bill payments being made to the school.
Future Research and Directions
The MRP was a necessary first step to vastly expanding the collective knowledge base
regarding veterans in higher education. However, it is only the first step.
The immediate next project will both broaden and deepen the scope of the MRP results.
Phase II will replicate the data-match conducted in the Million Records Project, but only for
veterans using the Post-9/11 GI Bill benefit. The sample will consist of 1 million student
veterans that enrolled in the benefit between 2009 and 2013. The Phase II data-match will also
include enrollment data, thus allowing researchers to calculate retention, persistence, and transfer
rates. It is also likely to determine key loss-points for student veteranstimes that they are most
likely to drop out. If successful, interventions can be designed and implemented strategically to
68
increase persistence. The Phase II project will also explore the number of student veterans that
stop out, but return to school, and to determine the average length of that hiatus.
Finally, Phase II will explore the campus policies, practices, and services that most
influence student veteran postsecondary outcomes at a sample of institutions on a subsample of
the population. The selected schools will be analyzed according to sector and level, thus better
preparing them to implement policies, practices, and services that have successful improved
academic outcomes in their sector.
Future studies should explore ways to further measure the overall return on the GI Bill
investment. Postsecondary completion rates are only one measure. Other research can focus on
how GI Bill beneficiaries impact the economy, such as job attainment, wages, amount of taxes or
social security paid, and number of months unemployed. Additional measures, such as savings
to VA and federal government programs, should also be considered, such as use of private health
insurance over VA medical insurance.
Last, the Million Records Project sets the stage for research related to the student loan
debt of military veterans. While the GI Bill is a robust financial benefit, the time-to-completion
rates noted in this study show that the benefit may not always cover a student veterans entire
academic career. There are also gaps in the benefit that can increase the cost of a degree. For
example, if a student veteran is classified as an out-of-state resident, his or her GI Bill benefits
are capped and will not cover the full cost of tuition at that institution. This forces student
veterans to make difficult choices about their academic future: either pay for college with student
loans, transfer to a more affordable institution, or drop out entirely. The amount and impact of
student debt on student veterans persistence in college and their long-term goals remains largely
unknown.
69
Closing
For nearly 70 years, the GI Bill has been an essential tool to help veterans transition from
the military to the civilian workforce. Over the decades, many researchers and historians have
examined and reported on the benefits of the GI Bill to veterans and the country. The research
has shown that student veterans have high levels of postsecondary academic outcomes.
The Million Records Project adds to this previous research on student veterans and the GI
Bill, with notable exceptions. This report has shown that, like all previous generations, a
majority of student veterans in the Post-9/11 GI Bill era are earning postsecondary degrees and
that this percentage is likely to rise over the foreseeable future. It refutes certain previously held
notions that a majority of student veterans use their educational benefits on vocational
certificates or on-the-job training programs. The MRP findings indicate that a majority of
student veterans are earning bachelors or graduate degrees. These levels of postsecondary
degree attainment, combined with the leadership skills developed in the military, position student
veterans to become tomorrows leaders in numerous fields, such as business, public service,
science, and engineering.
This report breaks new ground as well. Unlike previous research that waited several
years or decades to examine student veteran completion rates and the effects of the GI Bill, the
MRP utilized real-time degree attainment databases. This method gives policymakers,
stakeholders, Veterans Service Organizations, and the American public the opportunity to know
in near real-time the return on their GI Bill investment. Having real-time attainment data on
student veterans instead of anecdotal, outdated data will also allow for more effective allocations
of scarce resources; and it gives the American public the reassurance that tax dollars are being
used wisely by those who have defended our nation.
70
While the Million Records Project is a vital first step in understanding the current
generation of student veterans, it is not the last. This report provides as many answers regarding
the postsecondary outcomes of veterans as it poses questions for future research. The next phase
of analysis should focus on the drivers of academic success for student veterans so that effective
programs can be brought to scale, thus increasing the likelihood that all student veterans achieve
their academic goals and earn a postsecondary degree or credential.
71
Works Cited
Ackerman, R., DiRamio, D., & Garza , R. L. (2009). Transitions: Combat veterans as college
students. New Directions for Student Services, 2009(126), 5-14.
Bound, J., & Turner, S. (2002). Going to war and going to college: did World War II and the
G.I. Bill increase educational attainment for returning veterans? Journal of labor
economics, 20(4), 784-815.
Briggs, B. (2012, July 2). Thousands of veterans failing in latest battlefield: college. Retrieved
from http://usnews.nbcnews.com/_news/2012/07/02/12509343-thousands-of-veterans-
failing-in-latest-battlefield-college?lite
Cate, C. (2014). An Examination of Student Veteran Completion Rates over Service Eras: An
In-Depth Analysis of the 2010 National Survey of Veterans. 2(1). Student Veterans of
America.
Church, T. (2009). Returning veterans on campus with war related injuries and the long road
back home. Journal of Postsecondary Education and Disability, 22(1), 43-52.
Corder, G. (2009). Nonparametric statistics for non-statisticians: A step-by-step approach.
Hoboken, N.J.: Wiley.
Cosden, M., & McNamara, J. (1997). Self-concept and perceived social support among college
students with and without learning disabilities. Learning disability quarterly, 20(1), 2-
12.
Department of Defense. (2013). 2012 Demographics: Profile of the Military Community.
Washington, D.C.
Department of Veterans Affairs. (1976). Completion rates for education and training under the
Vietnam era GI bill: A study. Ann Arbor: University of Michigan Library.
72
Department of Veterans Affairs. (2013). Department of Veterans Affairs Education Program
Beneficiaries: FY 2000 to FY 2012. Retrieved from
http://www.va.gov/vetdata/docs/Utilization/EducNation_2012.xls
DiRamio, D., Ackerman, R., & Mitchell, R. L. (2008). From combat to campus: Voices of
student-veterans. Journal of Student Affairs Research and Practice, 45(1), 78-102.
Frederiksen, N., & Schrader, W. (1952). The academic performance of veteran and nonveteran
students. Psychological Monographs: General and Applied, 66(15).
Garmezy, N., & Crose, J. (1948). A comparison of the academic performance of matched
groups of veteran and nonveteran freshmen at the University of Iowa. The Journal of
Educational Research, 41(7)(7), 547-550.
Gowan, A. M. (1949). Characteristics of freshman veterans. . Journal of Higher Education,
20(4), 205-206.
Greenberg, M. (1997). The GI Bill: The Law That Changed America. Lickle Pub Inc.
Hoge, C., Auchterlonie, J., & Milliken, C. (2006). Mental health problems, use of mental health
services, and attrition from military service after returning from deployment to Iraq or
Afghanistan. JAMA: Journal of the American Medical Association, 295(9), 1023-1032.
Horn, L. (1996). Nontraditional Undergraduates, Trends in Enrollment From 1986 to 1992 and
Persistence and Attainment Among 198990 Beginning Postsecondary Students. U.S.
Department of Education, NCES. Washington, DC: U.S. Government Printing Office.
Humes, E. (2006). Over Here: How the GI Bill Transformed the American Dream. Orlando,
FL: Harcourt.
73
Hurst, D., & Smerdon, B. (2000). Postsecondary Students with Disabilities: Enrollment,
Services, and Persistence. (NCES 2000-092). Washington, D.C.: U.S. Department of
Education.
Joanning, H. (1975). The academic performance of Vietnam veteran college students. Journal
of College Student Personnel, 16(1), 10-13.
Klocek, J. (2008). The physical and psychological impact of your injury and disability. In N.
D. Ainspan, & W. E. Penk, Returning wars' wounded, injured, and ill: A reference
handbook (pp. 50-66). Westport, CT: Praeger Security International.
Love, L., & Hutchison, C. (1946). Academic progress of veterans. Educational Research
Bulletin, 25(8), 223-226.
Megivern, D., Pellerito, S., & Mowbray, C. (2003). Barriers to higher education for individuals
with psychiatric disabilities. Psychiatric Rehabilitation Journal, 26(3), 217-231.
National Science Foundation. (2013). Degrees awarded in all fields and in science and
engineering fields, by degree level: 1966-2010. Retrieved from
http://www.nsf.gov/statistics/nsf13327/tables/tab1.xls
Shapiro, D., Dundar, A., Ziskin, M., Yuan, X., & Harrell, A. (2013). Completing College: A
National View of Student Attainment Rates - Fall 2007 Cohort. Herndon: National
Student Clearinghouse.
Stanley, M. (2003). College Education and the Midcentury GI Bills. The Quarterly Journal of
Economics. 118(2), 671-708.
Teachman, J. (2007). Military service and educational attainment in the all-volunteer era.
Sociology of Education, 80(4), 359-374.
74
U.S. Census Bureau. (2010). Current Population Survey, 2010 Annual Social and Economic
Supplement. Washington, D.C.: U.S. Census.
U.S. Census Bureau. (2012). VETERAN STATUS BY EDUCATIONAL ATTAINMENT FOR
THE CIVILIAN POPULATION 25 YEARS AND OVER.
U.S. Department of Education. (2009). Issue Tables: A Profile of Military Service members and
Veterans Enrolled in Postsecondary Education in 2007-08. Washington, D.C.
U.S. Department of Education. (2009). National Center for Education Statistics, 2003-04
Beginning Postsecondary Students Longitudinal Study, Second Follow-up (BPS: 04/09).
Westat. (2010). National Survey of Veterans, Active Duty Service Members, Demobilized
National Guard and Reserve Members, Family Members, and Surviving Spouses. Final
Report, Deliverable 27.
Wood, D. (2012, Oct 25). Veterans' college drop-out rate soars. Retrieved from
http://www.huffingtonpost.com/2012/10/25/veterans-college-drop-out_n_2016926.html