MILLENNIAL

PERCEPTIONS OF
SECURITY
RESULTS FROM A NATIONAL SURVEY OF AMERICANS

MAREK N. POSARD | JENNIFER KAVANAGH

KATHRYN EDWARDS | SONNI EFRON

C O R P O R AT I O N
 Limited Print and Electronic Distribution Rights
This document and trademark(s) contained herein are protected by
law. This representation of RAND intellectual property is provided for
noncommercial use only. Unauthorized posting of this publication online
is prohibited. Permission is given to duplicate this document for personal
use only, as long as it is unaltered and complete. Permission is required
from RAND to reproduce, or reuse in another form, any of our research
documents for commercial use. For information on reprint and linking
permissions, please visit
[Link]/pubs/[Link].
RAND’s publications do not necessarily reflect the opinions of its research
clients and sponsors. R® is a registered trademark.
For more information on this publication,
visit [Link]/t/RR2571.
© Copyright 2018 RAND Corporation
M
illennials, those born between 1982 By 2040, millennials will make up an even larger
and 2000, are the largest segment of segment of the population (see Figure 2). As these
the U.S. population, with 84.3 million young Americans become more prominent in all
people (U.S. Census Bureau, 2015; fields—from politics to government to media to aca-
2017a). In 2015, they began to out- demia to business—their attitudes, preferences, and
number baby boomers, the 75.4 million Americans beliefs will likely have increasing weight in public
born between 1946 and 1964. As shown in Figure 1, discourse and U.S. policy (Weinbaum, Girven, and
millennials also outnumber those over the age of 70 Oberholtzer, 2016).
and under the age of 15. Although the millennials’ effects on U.S. society
have been widely discussed, their attitudes toward
FIGURE 1 security have not been carefully studied. Do millen-
U.S. Population by Age Group, 2016 nials’ attitudes toward security differ from the views
of previous generations? And, if so, what do these
Pre-baby boom perceptions imply for U.S. security policy in 2040?
(born before 1946)
9% This report attempts to answer these questions
based on a nationally representative probability sam-
Post-millennial ple survey of 1,608 adult Americans conducted from
(born after 2000) June to July 2017. The survey was administered to a
20%
sample of respondents on the RAND Corporation’s
Baby boom
(1946–1964) American Life Panel (ALP), which is a group of about
23% 6,000 Americans who are 18 years of age or older.
Millennial Our sample included people ranging in age from
(1982–2000) 21 to 71 as of 2017. We separated them into three
26%
Generation X groups: baby boomers; members of Generation X, or
(1965–1981)
22% Generation X’ers; and millennials.
The survey data revealed some statistically
significant differences in the attitudes of younger
SOURCE: U.S. Census Bureau, 2015. Americans toward economic and national security,
NOTE: The Census Bureau does not define post–baby boom along with some notable continuities with the views
generations. The RAND survey looked at millennials born between
1982 and 1996 because the youngest person in our sample was born of older Americans. In general, younger people
in 1996.
reported being more worried about economic secu-
rity and less worried about national security than
older Americans. They were also less likely to report
that living in a democracy is very important to them,
while older Americans were more likely to report that
Although millennials’ this was very important to them. Millennials were

effects on U.S. also the least concerned about illegal immigration,

the federal budget deficit, or increasing the size of

society have been the U.S. Armed Forces as security issues. They were
also less concerned than their elders about protecting
widely discussed, the United States against terrorist attacks or halting
the proliferation of weapons of mass destruction
their attitudes toward (WMD).
However, on other security issues (i.e., protecting
security have not been the privacy of U.S. citizens, investing in worker train-
ing and education programs, protecting civil rights
carefully studied. for minorities, expanding public benefits for families

1
FIGURE 2
Millennials in the U.S. Population, 2014–2040

100,000,000

90,000,000

80,000,000

70,000,000

60,000,000

50,000,000

40,000,000

30,000,000

20,000,000 Millennials
Generation X’ers
10,000,000 Baby boomers

0
2014 2016 2018 2020 2022 2024 2026 2028 2030 2032 2034 2036 2038 2040

SOURCE: U.S. Census Bureau, 2017b, Table 9.

NOTE: These trend lines represent extrapolations based on Census Bureau groupings of ages, which may not align with other popular definitions
of generations.

in need, global climate change), millennials and older • Baby boomers are the estimated 75.4 million
respondents did not differ notably in their attitudes, people who were born between 1946 and 1964.
and differences among respondents more strongly At the time of the survey in June and July
patterned with factors other than age. 2017, they were ages 53 to 71.
This report provides detailed analysis of the • Generation X’ers are the estimated 70.1 mil-
survey data and what might explain these differences. lion people born between 1965 and 1981. As of
It examines the question of what defines a generation 2017, they were ages 36 to 52.
and whether millennials display generational differ- • Millennials are the estimated 84.3 million
ences in their views of security when compared with people born between 1982 and 2000. As of
older Americans. It also explores the implications for 2017, they were ages 17 to 35. For our sample,
policymakers looking ahead to 2040. the youngest person was born in 1996, making
The definition of a generation is murky, and the him or her 21 years old in 2017.
definition of a millennial even more so. As we will Millennials’ attitudes might be shaped by such
explain, there is little consensus about when some events as the September 11, 2001, (9/11) terrorist
generations begin and end, or what seminal events attacks and the 2008–2009 Great Recession, while
influence and unite the people born during those those coming of age in the 1960s might be influ-
years. To understand the implications of our find- enced by the Cuban Missile Crisis or the Vietnam
ings, therefore, requires a more nuanced analysis that War (Carlsson and Karlsson, 1970). Some research
includes precise definitions of such terms as security finds that critical periods in the past can affect one’s
and generation, labels that are widely used but often attitudes in the future (Schuman and Corning,
with vague or conflicting meanings. 2012). Thus, people’s recollections of past events
For the purposes of this study, we defined the (e.g., 9/11 terrorist attacks, the Vietnam War, or the
generations as follows (estimates are as of 2016): Cuban Missile Crisis) might affect their views today.

2
Nevertheless, our data did not support the popular who are born in the same year. The U.S. Census
characterization of millennials as a generation with Bureau defines a generation as a group of cohorts that
attitudes uniquely different from their predecessors. share a statistically meaningful demographic event.
Rather, on most issues, millennial attitudes fell along The baby boom generation is the only generation
a spectrum that aligns with broader trends of how that is officially defined, because the Census Bureau
Americans’ views tend to change with their age and classifies the increase in birth rates after World
stage of life. War II as a meaningful demographic event. An
We therefore hypothesize that “age” is a more estimated 16.1 million Americans served in World
meaningful category than “generation” and that War II between 1941 and 1945. When the war ended,
popular characterizations of “millennial” attitudes, many returned home and had children. Figure 3 dis-
at least toward security, should be viewed with some plays the birth rate per 1,000 Americans from 1909
skepticism. Where millennial attitudes differ from to 2015. The birth rate in 1945 was 20.4, increasing
those of other Americans, we must also consider to 24.1 births per 1,000 people in 1946. That was the
whether their views are inherently different because largest increase on record by the National Center for
of their specific shared life experiences (a genera- Health Statistics. U.S. birth rates remained above
tional effect), or whether millennials are different these 1945 levels until 1965, when they dropped to
from baby boomers because the young are different 19.4 births per 1,000 people.
from the old (a youth effect). Our data tend to suggest The Census Bureau defines the cohort born
the latter. between mid-1946 and mid-1964 as the baby boom
generation because it was a unique epoch of elevated
births, producing a birth cohort that was much
What Is in a Generation? larger than in the previous quarter-century and that
The word “generation” means one thing in popular remained large for another 30 years.
culture and something entirely different to demog- In popular culture, however, the definition of a
raphers, who prefer to use the term “birth cohort.” A generation is a group of cohorts who share a mean-
birth cohort, or cohort for short, is a group of people ingful life experience. For example, many men from
what is now dubbed the “Greatest Generation” (born

FIGURE 3
U.S. Birth Rates per 1,000 People, 1909–2015

35

30

25

20

15

10

0
1909 1919 1929 1939 1949 1959 1969 1979 1989 1999 2009

SOURCE: National Center for Health Statistics, 2018.

NOTE: Red bars approximately represent the approximate start and end of the baby boom, from 1946 to 1964.

3
in the period beginning around 1914 and ending point (51.2 percent of millennials compared with
around 1924) fought fascism together as young 33 percent of baby boomers). Furthermore, there is
soldiers during World War II. Baby boomers grew speculation that the Great Recession has led millen-
up in a country preoccupied by the threat of nuclear nials to become more conservative investors than
conflagration, and many practiced sheltering under older Americans (Winograd and Hais, 2014).
their desks in school. Yet individual and subgroup Generalizing about generations can obscure
experiences vary widely. For African-Americans, the individual differences in life experience that
the assassination of black leaders during the 1950s might be much more powerful than the year of one’s
and 1960s might have been more formative than the birth. Despite the widespread belief that genera-
threat of nuclear war. tional attitudes matter, there is limited evidence
There is some evidence that the 9/11 terrorist on which experiences are (or are not) meaningful
attacks and the Great Recession might have affected to what percentage of the population and why.
the attitudes, opinions, and behaviors of millennials. Moreover, with only gradual changes in the birth
For example, the percentage of young adults report- rate over the past 50 years, the definitions of which
ing that it was likely for them to serve in the military Americans belong to which generation are arbitrary.
in the next few years increased from 15 percent in Pollsters and others have used conflicting dates and
August 2001 to 23 percent in November 2001, a few names to define the generations, especially those that
weeks after the attacks (Stafford and Griffis, 2008). followed the baby boomers, creating more confusion.
Millennials are said to have been heavily influenced The different names given to various groupings by
by the Great Recession (Archer, 2017). More than half birth year are shown in Figure 4.
of the millennials in our survey sample reported that Many of the generation names and dates shown
one or both of their parents had lost a job at some in Figure 4 were given by marketers, pollsters, and

FIGURE 4
Various Names and Dates Attributed to Generations

1950 1960 1970 1980 1990 2000 2010

Baby boom

Baby Bust (Washington Post, 1984)

20-Nothings (Life, 1993)

Grunge Kids
Generation X
13th Gen
Gen X-2 (Foot, 1996)

Gen X, Generation Y (MTV, 1997)

Millennial (Howe, 1998)

Echo Boomers
(Allied Academies, 2010)
Digital Natives
Baby boomers
Net Generation
Generation X’ers Nexters
Millennials The Trophy Kids
Post-millennials Gen Z (USA Today, 2012)
iGen
Multi-Gen
Homeland Generation

Gen X, Millennials, Homeland Generation (White House, 2014)

SOURCE: Bump, 2015.

NOTE: This figure does not include the way that millennials are defined in some reports by the U.S. Census Bureau.

4
journalists, and some names have proved more This has led to an explosion of research about the
enduring than others. attitudes and opinions of millennials as employees
Historians William Strauss and Neil Howe (1991) (Ertas, 2015; Myers and Sadaghiani, 2010), con-
coined the term “millennials” to refer to those grad- sumers (Bakewell and Mitchell, 2003; Gurău, 2012;
uating from high school in the new millennium, the Noble, Haytko, and Phillips, 2009), and citizens
year 2000. Millennials have also been dubbed “Echo (Jennings and Stoker, 2004). In 2013, Time magazine
Boomers” or “New Boomers,” because many are put millennials on a cover that derided the “Me, Me,
children of baby boomers, or “Generation Y,” because Me Generation” as “lazy, entitled narcissists” (Stein,
they follow Generation X. The U.S. Census does 2013). Some argue that millennials were socialized
not offer an official definition of what it calls “those to have an inflated sense of self (Twenge, 2014), while
typically referred to as millennials,” but some of its more-flattering assessments suggest that millennials
data categorize them as being part of the cohort born have unique creative and entrepreneurial qualities
between 1982 and 2000 (U.S. Census Bureau, 2015; (Martin, 2005).
Smith via Levy, 2016). This analysis attempts to consider multiple
The definition of the generation born after explanations for the survey findings while avoiding
millennials (e.g., post-millennials, Generation Z) value judgment. It rests on survey data, which are
is even more imprecise. They have been called the highly sensitive to interpretation. For example, how
“iGen,” “multi-gen,” and “homeland generation.” do we distinguish between attitudes that are held
Statistics Canada defines those born since 1993 as by millennials and attitudes that are held by young
“the new Generation Z or the internet generation people in general (as opposed to older people in gen-
since they were born after the invention of the inter- eral)? When we see millennial attitudes diverging
net” (Statistics Canada, 2015). Ernst & Young defined from those of their elders, is it because millennials
Generation Z as people born between 1997 and 2003 are different? Or is it because the survey question
(Merriman, 2015), while the Association of National measures something beyond attitudes—for exam-
Advertisers (2015) defined them as the group born ple, whether their income comfortably meets their
after 1995. financial obligations to older parents, children, and
so on.
To shed light on whether the differences we mea-
Generalizing About Generations
sured were a “generational effect” or a “youth effect,”
There is no shortage of generalizations about the we disaggregated the generational categories and
values and behaviors of people based on when they redefined them as age categories. Across numerous
were born (Wechsler, 1937). Bemoaning the follies questions, we observed strong and steady age trends.
of younger generations is a tradition that dates back Furthermore, we found support for these trends
to ancient times. Hesiod is said to have opined in the on most questions when we replaced generational
eighth century BCE, “I see no hope for the future categories with age categories or age in models. We
of our people if they are dependent on frivolous discuss these trends in the next section.
youth of today, for certainly all youth are reckless
beyond words. When I was young, we were taught to
be discreet and respectful of elders, but the present Survey Methods and Results
youth are exceedingly disrespectful and impatient of The objective of this project was to study millennials’
restraint” (quoted in Kohn, 2016, p. 16). security perceptions and then assess how their views
As millennials have become adults and entered could affect U.S. policy in the year 2040. To do this,
the labor market, some analysts have proposed that we surveyed adult Americans across age groups about
they have unique values and perspectives (Lyons, a range of economic and national security issues and
Duxbury, and Higgins, 2007; Weinbaum, Girven, compared their responses.
and Oberholtzer, 2016; Westerman et al., 2012).

5
Survey Methods white, compared with 78.5 percent of baby boomers.
Fewer millennials were married, and they reported
We administered our survey in June and
lower household incomes. These differences are high-
July 2017 on RAND’s ALP, which is a panel of
lighted in Table 1 and discussed at greater length in
about 6,000 Americans who are 18 years of age or
the following sections.
older.1 The ALP recruits a nationally representative
probability sample of American adults to complete
online surveys each month. If panel members do not TABLE 1
have a computer or internet access, the ALP provides Demographic Characteristics of the
these resources. People who become respondents in Unweighted Sample, by Cohort
the ALP regularly receive emails with a request to
Generation
visit the ALP member portal to fill out questionnaires Characteristic Millennials X’ers Baby Boomers
on the internet. Typically, an interview will not take
30.23
more than 30 minutes. Respondents are paid an Agea 43.80 (4.94) 61.38 (5.20)
(3.50)
incentive based on survey length, at a rate of $40 per Women 64.78 61.77 53.34
hour (Pollard and Baird, 2017).
Married 51.16 57.90 63.37
We invited a sample of 2,062 respondents from
Race and ethnicity
the ALP to complete our survey. Of these, 452 people
(21.9 percent) did not respond, while fewer than White 43.19 54.68 78.49

1 percent began the survey but did not complete it. African-
13.95 14.84 9.88
American
Our final sample comprised 1,608 respondents who
completed all of the survey questions, represent- Asian 5.32 4.19 1.45

ing 78 percent of our selected sample. (Table B.1 in Other 3.32 2.74 2.18
Appendix B shows statistics about the response rate Hispanic 34.22 23.55 7.99
for our survey.)
Education
We attempted to control the sample for party
Did not
affiliation, despite missing data for 14.5 percent of our graduate 38.21 40.16 38.08
sample (Pollard and Mendelsohn, 2016). Controlling collegeb
for party affiliation did not substantively change College
61.79 59.84 61.92
most results unless otherwise noted. In this report, graduate

we define statistically significant to mean a p-value of Household income

0.05 or lower, using a two-tailed test. (A probability- Under
36.54 30.53 25.18
or p-value of 0.05 or lower represents a 5-percent or $35,000

lower chance that we erroneously found a statistically $35,000 to

35.88 31.66 35.66
$74,999
significant difference.)
$75,000+ 27.57 37.80 39.16

Region
Survey Results
Northeast 21.3 18.9 17.7
There were significant demographic differences
Midwest 15.26 16.5 18.1
between the baby boomer and millennial respon-
South 33.3 36.9 33.8
dents in the sample, as would be expected from U.S.
Census and other data about the changing makeup West 30.1 27.8 30.3

of the U.S. population. There were more women and N 301 620 688
more ethnic diversity in the millennial sample than SOURCE: RAND A LP, June 21 to July 17, 2017.
in the baby boomer sample. Millennials self-identi- a Means and standard deviations displayed for age. All other variables in
this table display percentages. Some percentages may not sum to 100
fied as Hispanic at more than four times the rate of due to rounding. Unweighted sample statistics displayed.
baby boomers. Only 43.2 percent self-identified as b
It is likely that a percentage of younger millennials in our sample were
still enrolled in college and reported that they had not yet graduated.

6
 There are two ways to interpret the differences TABLE 2
in race and ethnicity. First, the United States is Political Affiliation by Birth Cohort,
becoming more diverse, especially among younger by Percentage
cohorts (Colby and Ortman, 2015). Second, younger
Affiliation Millennials Generation X'ers Baby Boomers
Americans might be less consistent in how they
Democrat 36.61 43.50 38.26
define their race and ethnicity. One study estimated
that 9.8 million (6.1 percent of the U.S. population) Republican 19.20 22.41 27.65
changed their race or Hispanic-origin responses Independent 21.43 21.09 26.37
from the 2000 to 2010 census (Liebler et al., 2017).
Other/ 22.77 12.99 7.72
It is beyond the scope of our survey data to identify not sure
the stability over time in the self-reported race and N 224 531 622
ethnicity of respondents. However, we speculate NOTES: Data represent the 1,377 respondents who completed the
that age-related differences are likely a result of both first wave of the 2016 RAND Presidential Election Panel Survey from
December 12, 2015 to January 6, 2016 (Pollard and Mendelsohn, 2016)
increasing diversity and changes in the way people and this Security 2040 survey from June 21 to July 17, 2017. Some
percentages may not sum to 100 due to rounding.
report their ascribed characteristics.
Surveys have consistently found that millen-
As shown in Figure 5, a sizable number of
nials over the age of 24 are better educated than
millennials reported high levels of financial anxi-
baby boomers (Frey, 2018). Our data on educational
ety: 53 percent of millennials reported being wor-
attainment by age cohort differ from those of other
ried about making ends meet each month. Nearly
sources because the youngest person in the sample
81 percent worried about saving enough, 68 percent
was 21, and many younger respondents might not yet
about paying off debt, and 36 percent about job loss.
have graduated from college but will do so in future.
Furthermore, Table B.2 in Appendix B shows four
It is therefore likely that the data showing that baby
logistic regression models predicting the odds that
boomers and millennials have equal levels of edu-
people in our sample were worried about making
cation reflect that some younger people are still in
ends meet each month, saving enough money, paying
school.
off debt, or losing their job. For each of these models,
Respondents were not asked specifically about
we found that millennials and Generation X’ers were
party identification. However, previous ALP surveys
significantly more worried about their finances than
have asked that question (from December 12, 2015, to
baby boomers were (after controlling for socio-
January 6, 2016) and found that the largest group of
demographic background—i.e., gender, race, marital
millennials (44.2 percent) identified as independent,
status, education, household income, and region of
as favoring some other party that is not Republican
residence), but the two former groups were not statis-
or Democrat, or as unsure about their political affili-
tically different from each other. We also reran these
ation (Table 2).
models using age instead of birth cohort categories:
Results showed a significant age effect for each mea-
Perceptions of Economic Security sure. However, millennials look similar to Generation
In our survey, we asked about four specific types of X’ers in their financial concerns after controlling for
events and experiences related to economic well- these same factors.
being. Respondents were asked, “How worried are These findings are not surprising; we would
you, if at all, about the following?” and were offered expect younger workers with less experience to have
four prompts in random order: “making ends meet lower wages and lower household incomes than older,
each month,” “saving enough money,” “paying experienced workers. However, there is growing
off debt,” and “losing your job.” For each prompt, evidence that current generations of younger workers
respondents were given five choices: “no opinion,” are not earning as much as previous generations did
“very worried,” “somewhat worried,” “not too wor- at the same age, suggesting that people in this group
ried,” and “not worried at all.”

7
FIGURE 5
Percentage of Respondents Worried About Finances, by Birth Cohort

100

Baby boomers (1946–1964)

80.7 Generation X’ers (1965–1981)
78.4
80 Millennials (1982–1996)
Percentage of respondents

67.8
62.5
60.8
60
53.6 52.5

39.7
40 37.1 36.2
34.6

20.2
20

0
Making ends meet each month Saving enough money Paying off debt Losing your job
SOURCE: RAND ALP, June 21 to July 17, 2017.
NOTE: Unweighted sample of those who answered with an opinion and who reported being “very worried” or “somewhat worried” about finances.

will not earn as much as previous generations have were younger, or are likely to earn less over their
over the courses of their lifetimes. lifetimes.
For example, if we look at the share of Americans Across the board, these percentages are lower
ages 25–34 who earn less than $30,000 today (or, than those for baby boomers but similar to those of
adjusting for inflation, the equivalent of $30,000 in Generation X’ers, indicating that millennials are not
earlier time periods), 25 percent of American men distinct but reflect greater financial worries among
and 79.6 percent of women earned less than this younger Americans.
amount in 1975, compared with 41.4 percent of men Again, this finding is not surprising: Baby
and 58.1 percent of women in 2016 (Vespa, 2017). boomers are more likely to be retired or approaching
A study from the Pew Research Center reported retirement age while younger Americans are more
that households in 2009 headed by someone who likely to be paying off loans for education, housing, or
was 65 or older had 42 percent more median net cars and to rely on individual or employer-sponsored
worth (in 2010 dollars) than households from the savings plans for retirement (Brown et al., 2012). We
same age group in 1984 (Fry et al., 2011). In compar- cannot predict whether the higher rates of economic
ison, households headed by adults younger than 35 anxiety among millennials will persist until 2040 or
had 68 percent less wealth than similar households whether millennials will experience rising incomes
in 1984. and fewer worries as they age.
A separate analysis using data from the U.S. Generation X’ers expressed more economic
Social Security Administration found that lifetime anxiety than baby boomers. As shown in Figure 5,
earnings, not just a point-in-time comparison, will 53.6 percent reported worries about making ends
be lower, particularly for men (Guvenen et al., 2017). meet each month, 78.4 percent reported worries
Hence, these insecurities could reflect that young about saving enough money, 60.8 percent reported
workers are earning less in the current period, are worries about paying off debt, and 37.1 percent were
earning less relative to older adults did when they worried about losing their jobs. This might reflect the

8
financial responsibilities typical of their life stages. shows three logistic regression models predicting
For example, Generation X’ers range in age from 36 the odds that people in our sample were very or
to 52, might have more children, and have had more somewhat worried about these national security
time to save for retirement than millennials under issues. Specifically, Table B.3 shows that millennials
the age of 30. and Generation X’ers were significantly less likely to
report being worried about these issues than baby
Perceptions of National Security boomers were. These differences were statistically sig-
We asked respondents, “How worried are you, if at nificant after controlling for sociodemographic back-
all, about the following?” with the following state- ground (i.e., gender, race, marital status, education,
ments: “North Korea’s nuclear program,” “growing household income, and region of residence).2 We also
authoritarianism in Russia,” and “Islamic extremist reran these models using self-reported age instead of
groups, such as Al Qaeda or ISIS.” For each state- birth cohort categories: Results showed a significant
ment, people could select one of five answers: “no age effect for each measure displayed in Figure 6.3
opinion,” “very worried,” “somewhat worried,” “not Once again, the differences between Generation
too worried,” or “not worried at all.” Figure 6 displays X and millennials were much smaller. The difference
the percentage of people who reported that they were was statistically significant for North Korea’s nuclear
very or somewhat worried about each of these three program and Islamic extremist groups, but not for
topics by their birth cohort. growing authoritarianism in Russia, once these same
As shown in Figure 6, concern about national demographic factors were controlled for.
security threats tends to rise with age catego-
ries. Baby boomers worried most about North The Importance of Living in a Democracy
Korea, Russia, and Islamic extremism, followed by The survey asked a single question: “How import-
Generation X’ers and then millennials. Appendix B ant is it for you to live in a country that is governed

FIGURE 6
Percentage of Respondents Worried About National Security Issues, by Birth Cohort

100

Baby boomers (1946–1964)

82.7 Generation X’ers (1965–1981)
Millennials (1982–1996) 80.1
80
Percentage of respondents

68.6 69.7
64.5
63.5 61.9
60
52.9
50.2

40

20

0
North Korea’s nuclear program Growing authoritarianism in Russia Islamic extremist groups, such as
Al Qaeda or ISIS

SOURCE: RAND ALP, June 21 to July 17, 2017.

NOTE: Unweighted sample of those who answered with an opinion and who reported being either “very” or “somewhat” worried about national
security.

9
 democratically?” People could select one of four Results from a separate logistic regression model
answers: “very important,” “somewhat important,” showed no significant differences in the odds that
“not that important,” or “no opinion.” Figure 7 shows respondents selected “not that important” versus
the distribution of these response by birth cohort, “very” or “somewhat” important.)
excluding people who reported that they had no Appendix B shows the results from an ordered
opinion. logistic regression model predicting the importance
While all of the respondents reported that living of living in a democracy. Specifically, Table B.4 shows
in a democracy was important to them, the greatest that millennials and Generation X’ers were signifi-
number of baby boomers found it very important, cantly less likely to report that living in a democracy
followed by Generation X’ers. Millennials were the was important to them after controlling for vari-
least likely to report that living in a democracy was ous sociodemographic background characteristics
important to them. Only 1.8 percent of people from (i.e., gender, race, marital status, education, house-
the baby boom cohort said that living in a democ- hold income, and region of residence). Furthermore,
racy was not that important to them. These effects we found no statistically significant differences
were significant after controlling for party identifi- between millennials and Generation X’ers after con-
cation and demographic background—i.e., gender, trolling for other background characteristics.
race, household income, college education, region, We were not able to address the question of
and marital status. (The largest differences by age whether Americans in general become more con-
category on this democracy question were between cerned with democracy as they get older or whether
those selecting “very” versus “somewhat” important. the finding reflects a growing disillusionment with
democracy among younger Americans. Longitudinal
FIGURE 7 studies using more-detailed questions would be
Percentage of People Reporting How required to shed light on this finding.
Important It Is to Live in a Country That However, within our own survey data, when we
Is Governed Democratically, by Birth substitute age categories for generational categories,
Cohort we see the same age pattern. As shown in Figure 8,
enthusiasm for democracy increases steadily with
Very important Somewhat important Not that important
each category of age. Moreover, as shown by the
100
ordered logistic regression analyses in Table B.4 of
Appendix B, we find significant age effects when
dividing the sample by generation, by age categories,
80 and by self-reported age.
Percentage of respondents

66.2 Policy Priorities

60 80.8
89.4
This survey also asked people to prioritize a range
of domestic and foreign policy issues. One question
asked individuals, “How much priority do you think
40
the U.S. should give to the following policy goals?”
The survey then listed a set of 11 issues:
1. “preventing the spread of weapons of mass
20 29.5
destruction”
15.4
8.8 2. “protecting the U.S. from terrorist attacks”
1.8 3.8 4.4
0 3. “reducing illegal immigration”
Baby boomers Generation X’ers Millennials
(1946–1964) (1965–1981) (1982–1996) 4. “increasing the size of the U.S. armed
SOURCE: RAND ALP, June 21 to July 17, 2017. forces”
NOTES: Unweighted sample of those who answered with an opinion.
Percentages might not sum to 100 because of rounding.

10
 FIGURE 8
Percentage of People Reporting How In contrast to their elders, millennials cared
Important It Is to Live in a Country That slightly less than baby boomers or Generation
Is Governed Democratically, by Age X’ers about reducing the federal budget deficit
Category (42.5 percent named the issue a “top priority”) and
a great deal less about reducing illegal immigration
Very important Somewhat important Not that important (only 17.3 percent called it a “top priority”). Although
100 both issues have been considered partisan dividing
lines in U.S. politics, controlling for party affiliation
did not change these results.
80 While a higher percentage of baby boomers
Percentage of respondents

(38.5 percent) than Generation X’ers (30.0 percent)

64.9 said that reducing illegal immigration was a top pri-
71.7
60 81.7 ority, the difference was not statistically significant.
86.5
90.3 Younger Americans were also less likely to say
that increasing the size of the U.S. armed forces was a
40 priority. While 32.8 percent of baby boomers deemed
it a top priority, only 24.5 percent of Generation X’ers
and 17.9 percent of millennials did. Controlling for
20 32.4 party identification had no effect for millennials,
24.3
14.0
Generation X’ers, or baby boomers.
11.0
8.0 Large majorities in every age group said that pro-
2.7 4.0 4.4 2.5 1.7
0 tecting the United States from terrorist attacks and
21–29 30–39 40–49 50–59 60+
and preventing the spread of WMD were top prior-
Age cohort
ities, but millennials deemed these issues somewhat
SOURCE: RAND ALP, June 21 to July 17, 2017. less of a priority than their elders. Controlling for
NOTES: Unweighted sample of those who answered with an opinion.
Percentages might not sum to 100 because of rounding. party identification did not change these results.
Appendix B shows the details from separate
5. “protecting the privacy of U.S. citizens” logistic regression models predicting the odds that
people in our sample reported each issue a top policy
6. “dealing with climate change”
priority. Specifically, Table B.5 shows that millennials
7. “investing in worker training and education were significantly less likely than baby boomers to
programs” report each issue as a top policy priority after con-
8. “reducing the federal budget deficit” trolling for their sociodemographic characteristics.
9. “protecting civil rights for minority groups” The same was true for Generation X’ers except on the
issue of reducing illegal immigration. In general, we
10. “strengthening local and state law
found that millennials were more similar to those
enforcement”
from Generation X than baby boomers after con-
11. “expanding public benefits for families in
trolling for other background characteristics in our
need.”
models.
Respondents could select whether these policy
topics were a top priority, somewhat a priority,4 no Where Millennials Are Not Different
priority, or whether they had no opinion on the topic.
On five issues, our models showed no statisti-
The results displayed in Figure 9 highlight only
cally significant differences among baby boomers,
the issues on which millennials and baby boomers
Generation  X’ers, and millennials after controlling
significantly differed in what they considered a “top
for demographic factors. These issues were protect-
priority.” On these five issues, opinions varied pre-
ing the privacy of U.S. citizens, investing in worker
dictably by age group.

11
FIGURE 9
Largest Differences Among Generations on Top Policy Priorities

100

Baby boomers (1946–1964)

Generation X’ers (1965–1981)
79.1
80 Millennials (1982–1996)
Percentage of respondents

71.1
66.6
63.5
60 56.8
54.1
50.8
44.0
42.5
38.5
40
32.8
30.0
24.5
17.3 17.9
20

0
Reducing the federal Reducing illegal Increasing the size of Protecting the U.S. Preventing the spread
budget deficit immigration the U.S. Armed Forces from terrorist attacks of WMD

SOURCE: RAND ALP, June 21 to July 17, 2017.

NOTE: Unweighted sample of those who answered with an opinion and named the issue a top priority.

training and education programs, protecting civil Conclusions: Implications for

rights for minority groups, expanding public benefits 2040
for families in need, and global climate change.
Millennials’ reported that top domestic pri- In this section, we apply results from this research
orities were protecting the privacy of U.S. citi- to attempt to assess possible implications for U.S.
zens (55.1 percent), worker training programs decisionmakers in 2040.5 Such projections are inher-
(51.8 percent), protecting civil rights (47.5 percent), ently limited; even if the response rates were perfect
and expanding benefits for families in need and all respondents understood the questions in the
(46.5 percent). In addition, 49.5 percent named cli- same manner, people change over time. We do not
mate change as a top priority. know whether the differences between millennials
Appendix B shows the details from separate and older Americans will disappear as the millennials
logistic regression models predicting the odds that age, marry, have children, divorce, or experience other
people in our sample reported each issue a top policy seminal life events. Will their attitudes be more like
priority. Specifically, Table B.6 shows no significant those of their parents when they reach the same age?
differences between baby boomers, Generation X’ers, On most issues, respondents’ attitudes and
and millennials. These results suggest that there is lit- opinions tended to pattern with their age, not their
tle difference in perceptions for some policies among generation. For example, the older the respondent,
these groupings once we control for their sociodemo- the more she or he tended to express worry about
graphic background characteristics. nuclear weapons, terrorism, North Korea, authoritar-
ianism in Russia, and Islamic extremist groups (such
as Al Qaeda or ISIS); the younger the respondent, the
less he or she expressed worry about national security
issues. More baby boomers reported being worried

12
about national security, followed by Generation X’ers, results suggest that millennials will be more worried
then by millennials. about their economic security, less worried about
On issues of economic security, the reverse national security issues, and less likely to believe that
tended to be true. Millennials and Generation X’ers living in a democracy is important by 2040.
were most concerned about financial security— Finally, if there are age effects (which we suspect
making ends meet each month, saving enough is key), then the current observations about millen-
money, paying off debt, losing one’s job—whereas nials will turn out to be just another example of the
baby boomers, who tend to be wealthier and nearing perennial expressions of dismay about young people.
retirement age, worried much less. We suspect that all three explanations—age, period,
Millennials were also the least likely to report and cohort—are plausible, assume that it mostly is
that living in a democracy was very import- age, and propose that longitudinal studies are nec-
ant to them (66.2 percent). Generation X’ers and essary to give a more definitive answer than we give
baby boomers placed much higher importance on here.
living in a democracy (80.8 percent and 89.4 percent, We propose that these findings could affect
respectively). U.S. security policy in 2040 in two different ways. A
The differences among the generations were public that places higher priority on domestic-related
sharpest on five issues, with millennials placing far security (e.g., personal finances) than on foreign pol-
less priority than their elders on reducing the federal icy (e.g., national security) might allow policymakers
deficit, reducing illegal immigration, increasing the more freedom in their decisionmaking. However, the
size of the U.S. armed forces, protecting the United American public might prefer that more resources
States from terrorist attacks, and preventing the go toward domestic and economic security at the
spread of WMD. expense of U.S. foreign policy activities aimed at
How do we interpret these results to understand improving global security.
security in 2040? The answer will differ based on The finding that millennials are less attached
whether we think these results are the effects of age, to living in a democracy can also be interpreted
period, or cohort. While it is difficult to disentangle in three ways. Millennials who did not experience
the age-period-cohort effect (Danigelis, Hardy, and the Cold War clash with Communist countries
Cutler, 2007; Robinson and Jackson, 2001), there might assume democracy is a given. Or they might
is some evidence that cohort effects exist based on feel that U.S. democracy is not bringing them the
when someone was born (Davis, 2004; Pampel, 2016) desired benefits. Finally, they might be disillusioned
and experiences of past key events (Schuman and with democracy, viewing its current paralysis as
Corning, 2012) can influence one’s current attitudes evidence that, in its current form or given current
and opinions. levels of polarization, the process is “broken.” If
If our results are based on period effects, then any of these explanations are true, support for core
people from different age groups might interpret the U.S. institutions could decline below current levels.
same situations in different ways. We would expect However, it is also possible that millennials’ views
the attitudes and opinions of millennials to remain will change as they age.
the same (e.g., the 9/11 terrorist attacks’ effects on Given the significance of this question, we
millennials or the Vietnam War’s effects on baby suggest that more quantitative and qualitative
boomers would remain) unless another key event research is necessary to understand the specific
occurs (e.g., a future terrorist attack, another war, assumptions, motivations, attitudes, and opinions
economic recession). If our results are cohort effects, that underlie the seeming devaluation of democracy
then we expect that key experiences from people’s by millennials.
lives would stick with them as they age. Thus, our

13
Appendix A: Survey Instrument 12 $50,000 to $59,999
13 $60,000 to $74,999
Appendix A displays the survey instrument com-
14 $75,000 or more
pleted by people on RAND’s ALP. The variable is
listed first, followed by the question wording, and [If “14” was selected for previous question] You told
then the answer choices. People did not see the vari- us that the total combined income of all members
able names when answering each question.A of your family (living here) during the preceding
12 months was more than $75,000. Thinking about
DEMOGRAPHICS
the total combined income of your family from all
What is your gender?
sources, approximately how much did members of
1 Male
your family receive during the previous 12 months?
2 Female
1 $75,000 to $99,999
What is your age? 2 $100,000 to $124,999
[Integer] 3 $125,000 to $199,999
4 $200,000 or more
Could you tell us what your current living situation
is? Now, we would like to know about other members
1 Married or living with a partner of your household, if there are any. How many
2 Separated other people live with you? (enter 0 for no one else).
3 Divorced [Range: 0 to 10]
4 Widowed
Do you work for someone else, are you self-
5 Never married
employed, or what?
Were you born in the United States? 1 Work for someone else
1 Yes 2 Self-employed
2 No 3 Other
Which category represents the total combined What is the highest level of school you have com-
income of all members of your family (living here) pleted or the highest degree you have received?
during the past 12 months? This includes money 1 Less than first grade
from jobs; net income from business, farm or 2 First, second, third, or fourth grade
rent; pensions; dividends; interest; Social Security 3 Fifth or sixth grade
payments; and any other money income received by
4 Seventh or eighth grade
members of your family who are 15 years of age or
5 Ninth grade
older.
1 Less than $5,000 6 Tenth grade
2 $5,000 to $7,499 7 11th grade
3 $7,500 to $9,999 8 12th grade—no diploma
4 $10,000 to $12,499 9 High school graduate, high school diploma
5 $12,500 to $14,999 or the equivalent (for example: GED)
6 $15,000 to $19,999 10 Some college but no degree
7 $20,000 to $24,999 11 Associate degree in college occupational/
8 $25,000 to $29,999
vocational program
9 $30,000 to $34,999
12 Associate degree in college academic
10 $35,000 to $39,999
program
11 $40,000 to $49,999
13 Bachelor’s degree (for example: BA, AB, BS)
14 Master’s degree (for example: MA, MS,
A
For more information about this survey and the data, see
MEng, MEd, MSW, MBA)
RAND Corporation, undated.

14
 15 Professional school degree (for example: MD, Do you consider yourself primarily white or
DDS, DVM, LLB, JD) Caucasian, Black or African American, American
16 Doctorate degree (for example: PhD, EdD) Indian, or Asian?
1 White/Caucasian
What is your current employment situation? 2 Black/African American
1 Working now 3 American Indian or Alaska Native
2 Unemployed and looking for work 4 Asian or Pacific Islander
3 Temporarily laid off, on sick or other leave 5 Other
4 Disabled
5 Retired Do you consider yourself Hispanic or Latino?
6 Homemaker 1 Yes
7 Other 2 No

Which of the following categories best describes EVENTS

the type of work you do? Have you ever skipped meals because you were
1 Management occupations unable to afford food? Select one.
2 Business and financial operations 1 Yes, in the past year
occupations 2 Yes, more than a year ago, but less than five
3 Computer and mathematical occupations years ago
4 Architecture and engineering occupations 3 Yes, more than five years ago
5 Life, physical, and social science occupations 4 No
6 Community and social services occupations Have you ever been more than one month behind
7 Legal occupations on a bill payment, such as rent, mortgage, heating,
8 Education, training, and library occupations or electricity? Select one.
9 Arts, design, entertainment, sports, and 1 Yes, in the past year
media occupations 2 Yes, more than a year ago, but less than five
10 Health care practitioner and technical years ago
occupations 3 Yes, more than five years ago
11 Health care support occupations 4 No
12 Protective service occupations
13 Food preparation and serving related Have you ever had a debt go into third-party collec-
occupations tion? Select one.
14 Building and grounds cleaning and mainte- 1 Yes, in the past year
nance occupations 2 Yes, more than a year ago, but less than five
15 Personal care and service occupations years ago
16 Sales and related occupations 3 Yes, more than five years ago
17 Office and administrative support 4 No
occupations Have you ever lost your job (i.e., you were laid off or
18 Farming, fishing, and forestry occupations fired)? Select one.
19 Construction and extraction occupations 1 Yes, in the past year
20 Installation, maintenance, and repair 2 Yes, more than a year ago, but less than five
occupations
years ago
21 Production occupations
3 Yes, more than five years ago
22 Transportation and material moving
4 No
occupations

15
When you were growing up, did either of your par- There are many things that people value in a job.
ents lose their job at one point? Select one. How important is job security to you? Select one.
1 Yes, my father 1 Extremely important
2 Yes, my mother 2 Very important
3 Yes, both parents 3 Somewhat important
4 No 4 Not too important
5 No opinion
Do you feel that you have been discriminated
against because of your membership in a group? If How worried are you that there will soon be another
yes, select the group(s) that apply, if not, select no. terrorist attack in the United States? Select one.
1 Race 1 Very worried
2 Age 2 Somewhat worried
3 Sex 3 Not too worried
4 Religion 4 Not worried at all
5 National origin
Do you think a terrorist attack in the United States
6 Handicapping condition
is more likely from a U.S. citizen, an immigrant, a
7 Marital status
refugee, or a short-term visitor (tourist)? Select one.
8 Sexual orientation
1 U.S. citizen
9 Political beliefs
2 Immigrant
10 Income
3 Refugee
11 No
4 Short-term visitor
At any point in your life, has a family member or
[The following questions are displayed as a table]
close friend been deployed to an active conflict zone
Have you voted in U.S. elections over the past
while serving in the U.S. military? Select one.
10 years? Check yes or no.
1 No
1 Yes
2 Yes
2 No
[If “yes” was selected for previous question] How are
I vote in most presidential elections (2008, 2012,
you related to the person who deployed? Select one.
2016).
If more than one applies, please select the closest
1 Yes
family member.
2 No
1 N/A
2 Grandparent I vote in most midterm elections (2006, 2010, 2014).
3 Parent or stepparent 1 Yes
4 Aunt or uncle 2 No
5 Sibling I vote in most local elections.
6 Cousin 1 Yes
7 Friend 2 No
[End of table display]
When they were at their largest, how much, if any-
thing, did you owe in student loans? (The balance [The following questions are displayed as a table]
on student loans is typically largest when you left How worried are you, if at all, about the following?
school.) Select one. [Randomize Display]
1 I have never had a student loan Making ends meet each month
2 Less than $25,000 1 No opinion
3 $25,001 to $50,000 2 Very worried
4 $50,001 to $100,000 3 Somewhat worried
5 $100,001 to $200,000 4 Not too worried
6 More than $200,000 5 Not worried at all

16
Saving enough money 4 Not too worried
1 No opinion 5 Not worried at all
2 Very worried
Growing authoritarianism in Russia
3 Somewhat worried
1 No opinion
4 Not too worried
2 Very worried
5 Not worried at all
3 Somewhat worried
Paying off debt 4 Not too worried
1 No opinion 5 Not worried at all
2 Very worried
Islamic extremist groups, such as Al Qaeda or ISIS
3 Somewhat worried
1 No opinion
4 Not too worried
2 Very worried
5 Not worried at all
3 Somewhat worried
Losing your job 4 Not too worried
1 No opinion 5 Not worried at all
2 Very worried
Global climate change
3 Somewhat worried
1 No opinion
4 Not too worried
2 Very worried
5 Not worried at all
3 Somewhat worried
A health shock or diagnosis 4 Not too worried
1 No opinion 5 Not worried at all
2 Very worried [End of table display]
3 Somewhat worried
[The following questions are displayed as a table]
4 Not too worried
How much priority do you think the United States
5 Not worried at all
should give to the following policy goals?
[End of table display]
[Randomize Display]
[The following questions are displayed as a table]
Preventing the spread of weapons of mass
How worried are you, if at all, about the following?
destruction
[Randomize Display] 1 No opinion
The integrity of U.S. democratic institutions 2 Top priority
1 No opinion 3 Some priority
2 Very worried 4 No priority
3 Somewhat worried
Protecting the U.S. from terrorist attacks
4 Not too worried
1 No opinion
5 Not worried at all
2 Top priority
Threats to civil rights of minority groups 3 Some priority
1 No opinion 4 No priority
2 Very worried
Reducing illegal immigration
3 Somewhat worried
1 No opinion
4 Not too worried
2 Top priority
5 Not worried at all
3 Some priority
North Korea’s nuclear program 4 No priority
1 No opinion
2 Very worried
3 Somewhat worried

17
Increasing the size of the U.S. armed forces Strengthening local and state law enforcement
1 No opinion 1 No opinion
2 Top priority 2 Top priority
3 Some priority 3 Some priority
4 No priority 4 No priority
Protecting the privacy of U.S. citizens Expanding public benefits for families in need
1 No opinion 1 No opinion
2 Top priority 2 Top priority
3 Some priority 3 Some priority
4 No priority 4 No priority
[End of table display] [End of table display]
[The following questions are displayed as a table] How important is it for you to live in a country that
How much priority do you think the United States is governed democratically? Select one.
should give to the following policy goals? 1 Very important
2 Somewhat important
[Randomize Display]
3 Not that important
Dealing with climate change
4 No opinion
1 No opinion
2 Top priority This survey asked about several national policies
3 Some priority and policy concerns. Is there any policy (defense or
4 No priority domestic) that you think we left out?
[Blank field for written response]
Investing in worker training and education
programs Could you tell us how interesting or uninteresting
1 No opinion you found the questions in this interview?
2 Top priority 1 Very interesting
3 Some priority 2 Interesting
4 No priority 3 Neither interesting nor uninteresting
4 Uninteresting
Reducing the federal budget deficit
5 Very uninteresting
1 No opinion
2 Top priority Do you have any other comments on the interview?
3 Some priority Please type these in the box below. (If you have
4 No priority no comments, please click next to complete this
survey.)
Protecting civil rights for minority groups
[Blank field for written response]
1 No opinion
2 Top priority
3 Some priority
4 No priority

18
Appendix B: Details on the TABLE B.1

Statistical Models Response Rates for the Survey

Appendix B displays results from models that we ref- Sampling Frame Percentages of the
Characteristics Frequencies Selected Sample
erenced in the main body of this report, along with
Size of selected sample 2,062 100.00
related statistics.
Completed the survey 1,608 77.98

Started but did not 2 0.10

complete the survey

Did not start the survey 452 21.92

SOURCE: RAND ALP, June 21 to July 17, 2017.

TABLE B.2
Logistic Regression Models Predicting Responses of “Very” or “Somewhat” Worried
About Finances
Making Ends Meet Each Month Saving Money Paying Off Debt Losing Your Job

Women (versus men) 1.23 (0.21) 1.30 (0.23) 1.26 (0.20) 0.88 (0.16)

White (versus non-white) 0.96 (0.19) 1.11 (0.23) 0.95 (0.18) 0.87 (0.19)

Married 0.86 (0.17) 0.78 (0.17) 0.91 (0.18) 1.06 (0.21)

Highest level of education

Did not graduate college — — — —

College graduate 0.71 (0.13) 0.97 (0.17) 0.88 (0.15) 1.08 (0.20)

Household income

Less than $35,000 — — — —

$35,000 to $75,000 0.68 (0.15) 1.67 (0.40)* 1.00 (0.22) 0.70 (0.16)

$75,000 or more 0.29 (0.08)*** 1.34 (0.35) 0.84 (0.21) 0.48 (0.13)***

Regions

Northeast — — — —

Midwest 0.60 (0.18) 0.88 (0.26) 0.76 (0.21) 0.81 (0.24)

South 1.01 (0.27) 1.05 (0.29) 1.12 (0.28) 1.15 (0.31)

West 1.08 (0.29) 1.37 (0.38) 1.07 (0.27) 1.56 (0.41)

Birth cohorts

Millennials 1.56 (0.35)* 2.03 (0.46)** 2.95 (0.64)*** 1.60 (0.38)*

Generation X 2.16 (0.39)*** 2.15 (0.41)*** 2.33 (0.39)*** 2.06 (0.39)***

Baby boomers — — — —

Constant 1.03 (0.46) 0.88 (0.39) 0.59 (0.25) 0.44 (0.18)*

Number of observations 1,519 1,519 1,519 1,519

Wald Chi-squared 82.32*** 31.38*** 55.65*** 40.18***

Pseudo R-squared 0.10 0.03 0.05 0.04

SOURCE: RAND ALP, June 21 to July 17, 2017.
NOTES: Odds ratios reported. Sample weights used; robust standard errors reported; * p < 0.05, ** p < 0.01, *** p < 0.001.

19
TABLE B.3
Logistic Regression Models Predicting Responses of “Very” or “Somewhat” Worried
About National Security Issues
Growing Authoritarianism Islamic Extremist Groups,
North Korea’s Nuclear Program in Russia Such as Al Qaeda or ISIS

Women (versus men) 1.40 (0.25) 1.50 (0.23)** 1.43 (0.22)*

White (versus non-white) 1.23 (0.26) 0.87 (0.16) 0.96 (0.17)

Married 1.06 (0.21) 1.06 (0.19) 1.11 (0.19)

Highest level of education

Did not graduate college — — —

College graduate 1.19 (0.22) 1.22 (0.20)

Household income

Less than $35,000 — — —

$35,000 to $75,000 1.50 (0.34) 1.19 (0.24) 0.96 (0.20)

$75,000 or more 1.74 (0.47)* 1.49 (0.35) 1.27 (0.30)

Regions

Northeast — — —

Midwest 0.86 (0.25) 0.76 (0.19) 0.65 (0.16)

South 0.77 (0.20) 0.92 (0.21) 1.00 (0.22)

West 1.00 (0.28) 0.94 (0.22) 0.65 (0.15)

Birth cohorts

Millennials 0.27 (0.06)*** 0.39 (0.08)*** 0.22 (0.05)***

Generation X 0.49 (0.10)*** 0.57 (0.10)*** 0.49 (0.08)***

Baby boomers — — —

Constant 1.72 (0.70) 0.80 (0.29) 0.55 (0.20)

Number of observations 1,519 1,519 1,519

Wald Chi-squared 57.69*** 37.52*** 70.99***

Pseudo R-squared 0.08 0.04 0.07

SOURCE: RAND ALP, June 21 to July 17, 2017.
NOTES: Odds ratios reported. Sample weights used; Robust standard errors reported; * p < 0.05, ** p < 0.01, *** p < 0.001.

20
TABLE B.4
Ordered Logistic Regression Model Predicting the Importance of Living in a Country
That Is Governed Democratically
Model 1 Model 2 Model 3

Women (versus men) 0.63 (0.13)* 0.63 (0.13)* 0.62 (0.13)*

White (versus non-white) 1.55 (0.36)* 1.54 (0.36) 1.52 (0.36)

Married 1.23 (0.27) 1.24 (0.27) 1.21 (0.26)

Highest level of education

Did not graduate college — — —

College graduate 1.90 (0.41)** 1.85 (0.40)** 1.86 (0.40)**

Household income

Less than $35,000 — — —

$35,000 to $75,000 0.59 (0.15) 0.60 (0.15)* 0.58 (0.15)*

$75,000 or more 1.23 (0.41)** 1.25 (0.42) 1.24 (0.41)

Regions

Northeast — — —

Midwest 0.96 (0.33) 0.96 (0.34) 0.95 (0.33)

South 0.98 (0.28) 1.00 (0.28) 0.98 (0.28)

West 1.22 (0.36) 1.25 (0.36) 1.24 (0.36)

Birth cohorts

Millennials 0.33 (0.08)***

Generation X 0.51 (0.13)**

Baby boomers —

Age categories

21 to 29 —

30 to 39 1.19 (0.35)

40 to 49 1.57 (0.50)

50 to 59 2.43 (0.78)**

60+ 3.11 (1.11)**

Age 1.03 (0.01)***

Number of observations 1,439 1,439 1,439

Wald Chi-squared 66.56*** 64.92*** 64.74***

Pseudo R-squared 0.08 0.07 0.08

SOURCE: RAND ALP, June 21 to July 17, 2017.
NOTES: Odds ratios reported. Sample weights used; Robust standard errors reported; * p < 0.05, ** p < 0.01, *** p < 0.001. The empty spaces in the
models represent the different ways we measured cohorts. Model 1 uses our breakdown of millennials, Generation X'ers, and baby boomers; Model 2
uses age categories; and Model 3 uses age as a continuous variable.

21
TABLE B.5
Logistic Regression Models Predicting Top Policy Priorities, with Significant Differences
by Birth Cohort
Reducing the Increasing the Protecting the
Federal Budget Reducing Illegal Size of the U.S. U.S. from Terrorist Preventing the
Deficit Immigration Armed Forces Attacks Spread of WMD

Women (versus men) 0.85 (0.13) 0.87 (0.14) 1.16 (0.20) 1.24 (0.23) 1.09 (0.18)

White (versus non-white) 1.01 (0.19) 1.67 (0.33)** 0.76 (0.15) 0.93 (0.21) 1.02 (0.21)

Married 0.93 (0.16) 1.18 (0.22) 1.12 (0.23) 0.92 (0.19) 1.16 (0.21)

Highest level of education

Did not graduate college — — — — —

College graduate 0.90 (0.15) 0.69 (0.12)* 0.70 (0.13)* 0.93 (0.18) 1.02 (0.17)

Household income

Less than $35,000 — — — — —

$35,000 to $75,000 1.35 (0.28) 0.72 (0.16) 0.78 (0.18) 1.37 (0.32) 0.93 (0.20)

$75,000 or more 1.83 (0.44)** 0.97 (0.24) 0.74 (0.20) 1.36 (0.39) 0.85 (0.21)

Regions

Northeast — — — — —

Midwest 1.52 (0.38) 1.17 (0.33) 0.56 (0.17) 0.82 (0.24) 0.69 (0.18)

South 1.56 (0.35)* 1.75 (0.45)* 1.45 (0.36) 0.99 (0.27) 1.32 (0.32)

West 1.35 (0.31) 0.99 (0.26) 0.84 (0.22) 0.79 (0.21) 0.86 (0.20)

Birth cohorts

Millennials 0.63 (0.13)* 0.37 (0.09)*** 0.26 (0.07)*** 0.26 (0.06)*** 0.34 (0.07)***

Generation X 0.69 (0.11)* 0.89 (0.15) 0.59 (0.11)** 0.66 (0.13)* 0.60 (0.10)**

Baby boomers — — — — —

Constant 0.88 (0.31) 0.52 (0.19) 0.71 (0.28) 3.32 (10.39)** 2.09 (0.78)*

Number of observations 1,519 1,519 1,519 1,519 1,519

Wald Chi-squared 22.23* 48.36*** 62.74*** 43.62*** 41.90***

Pseudo R-squared 0.02 0.06 0.07 0.06 0.04

SOURCE: RAND ALP, June 21 to July 17, 2017.
NOTES: Odds ratios reported. Sample weights used; Robust standard errors reported; * p < 0.05, ** p < 0.01, *** p < 0.001.

22
TABLE B.6
Logistic Regression Models Predicting Similar Top Policy Priorities, by Birth Cohort
Investing in
Protecting the Worker Training Protecting Expanding Public
Privacy of U.S. and Education Civil Rights for Benefits for Global Climate
Citizens Programs Minority Groups Families in Need Change

Women (versus men) 0.73 (0.12)* 0.86 (0.15) 1.36 (0.22) 1.17 (0.20) 1.10 (0.17)

White (versus non-white) 0.82 (0.15) 0.65 (0.14)** 0.41 (0.08)*** 0.52 (0.10)*** 0.62 (0.11)**

Married 0.65 (0.12)* 0.87 (0.17) 0.86 (0.16) 0.92 (0.18) 0.84 (0.15)

Highest level of education

Did not graduate college — — — — —

College graduate 0.86 (0.14) 0.83 (0.15) 1.66 (0.29)** 1.52 (0.27)* 1.89 (0.32)***

Household income

Less than $35,000 — — — — —

$35,000 to $75,000 1.09 (0.23) 1.01 (0.24) 1.01 (0.21) 0.71 (0.16) 1.29 (0.27)

$75,000 or more 1.12 (0.26) 1.17 (0.32) 0.84 (0.22) 0.46 (0.12)** 1.12 (0.28)

Regions

Northeast — — — — —

Midwest 0.65 (0.17) 1.33 (0.35) 0.60 (0.16)* 0.60 (0.16)* 0.75 (0.19)

South 1.04 (0.24) 1.67 (0.41)* 0.58 (0.13)* 0.73 (0.18) 0.86 (0.18)

West 1.01 (0.24) 1.33 (0.33) 0.60 (0.14)* 0.93 (0.24) 1.09 (0.24)

Birth cohorts

Millennials 0.70 (0.14) 0.87 (0.21) 0.83 (0.18) 0.88 (0.20) 0.84 (0.17)

Generation X 0.79 (0.13) 0.80 (0.14) 0.99 (0.17) 1.16 (0.20) 0.81 (0.13)

Baby boomers — — — — —

Constant 3.67 (1.43)*** 2.51 (0.98) 1.03 (0.38) 1.14 (0.45) 0.65 (0.23)

Number of observations 1,519 1,311 1,519 1,519 1,519

Wald Chi-squared 23.88 49.66 44.98*** 49.57*** 29.45**

Pseudo R-squared 0.02 0.05 0.05 0.05 0.03

SOURCE: RAND ALP, June 21 to July 17, 2017.
NOTES: Odds ratios reported. Sample weights used; Robust standard errors reported; * p < 0.05, ** p < 0.01, *** p < 0.001.

23
Notes Danigelis, Nicholas L., Melissa Hardy, and Stephen J. Cutler,
“Population Aging, Intracohort Aging, and Sociopolitical
1
Panel members generally do not give notice about their intent to Attitudes,” American Sociological Review, Vol. 72, No. 5, 2007,
leave the panel; rather, they simply stop participating in surveys. pp. 812–830.
To avoid retention of disinterested panel members, RAND peri- Davis, James A., “Did Growing Up in the 1960s Leave a
odically attempts to contact members who have not been active Permanent Mark on Attitudes and Values? Evidence from the
for more than a year and removes from the panel those who no General Social Survey,” Public Opinion Quarterly, Vol. 68, No. 2,
longer wish to take part or cannot be contacted. For more details 2004, pp. 161–183.
on the ALP, see RAND Corporation (undated).
Ertas, Nevbahar, “Turnover Intentions and Work Motivations
2
The effect of birth cohort was statistically significant after con- of Millennial Employees in Federal Service,” Public Personnel
trolling for the party identification of people in our sample when Management, Vol. 44, No. 3, 2015, pp. 401–423.
it was available. Frey, William H., The Millennial Generation: A Demographic
Bridge to America’s Diverse Future, Washington, D.C.: Brookings
3
When we controlled for responses to questions about economic
Institution, January 2018. As of February 11, 2018:
worries, we still found the same pattern of results. Thus, we do [Link]
not believe that economic worries are replacing worries about 2018-jan_brookings-metro_millennials-a-demographic-bridge-
national security across birth cohorts. [Link]
4
The specific wording of this answer choice was “some priority.” Fry, Richard, D’Vera Cohn, Gretchen Livingston, and Paul
Taylor, The Rising Age Gap in Economic Well-Being: The Old
5
While these projections are based on cross-sectional survey Propser Relative to the Young, Washington, D.C.: Pew Research
data, the time horizon (i.e., two decades) means that our projec- Center, November 15, 2011. As of February 16, 2018:
tions represent informed speculation. [Link]
uploads/2013/11/2011_Briefing_Fry_The-rising-age-gap-in-
[Link]
Gurău, Călin, “A Life-Stage Analysis of Consumer
References Loyalty Profile: Comparing Generation X and Millennial
Consumers,” Journal of Consumer Marketing, Vol. 29, No. 2,
Archer, Seth, “Millennials Are Still Spooked by Great Recession,” 2012, pp. 103–113.
Markets Insider, June 1, 2017. As of February 13, 2018:
[Link] Guvenen, Fatih, Greg Kaplan, Jae Song, and Justin Weidner,
millennial-investing-habits-financial-crisis- Lifetime Incomes in the United States over Six Decades,
2017-6-1002059757-1002059757 Cambridge, Mass.: National Bureau of Economic Research,
No. w23371, 2017.
Association of National Advertisers, “Figuring Out Generation
Z,” ANA Magazine, September 10, 2015. As of September 26, 2017 Jennings, M. Kent, and Laura Stoker, “Social Trust and Civic
(account required): Engagement Across Time and Generations,” Acta Politica,
[Link] Vol. 39, No. 4, 2004, pp. 342–379.
ana-2015-sep-figuring-out-gen-z
Kohn, Alfie, The Myth of the Spoiled Child: Challenging the
Bakewell, Cathy, and Vincent-Wayne Mitchell, “Generation Conventional Wisdom About Children and Parenting, Boston,
Y Female Consumer Decision-Making Styles,” International Mass.: Beacon Press, 2016.
Journal of Retail & Distribution Management, Vol. 31, No. 2,
Liebler, Carolyn A., Sonya R. Porter, Leticia E. Fernandez, James
2003, pp. 95–106.
M. Noon, and Sharon R. Ennis, “America’s Churning Races: Race
Brown, Meta, Andrew Haughwout, Donghoon Lee, Maricar and Ethnicity Response Changes Between Census 2000 and the
Mabutas, and Wilbert van der Klaaw, “Grading Student Loans,” 2010 Census,” Demography, Vol. 54, No. 1, 2017, pp. 259–284.
Liberty Street Economics, blog post, March 5, 2012. As of May 21,
Lyons, Sean T., Linda Duxbury, and Christopher Higgins, “An
2018:
Empirical Assessment of Generational Differences in Basic
[Link]
Human Values,” Psychological Reports, Vol. 101, No. 2, 2007,
[Link]#.U0MJwvldV8E
pp. 339–352.
Bump, Philip, “Your Generational Identity Is a Lie,” Washington
Martin, Carolyn A., “From High Maintenance to High
Post, April 1, 2015. As of February 11, 2018:
Productivity: What Managers Need to Know About
[Link]
Generation Y,” Industrial and Commercial Training, Vol. 37,
your-generational-identity-is-a-lie/
No. 1, 2005, pp. 39–44.
Carlsson, Gosta, and Katarina Karlsson, “Age, Cohorts and the
Merriman, Marcie, A New Generation with Unique Experiences,
Generation of Generations,” American Sociological Review,
Beliefs and Behaviors. They are Gen Z, Washington, D.C.: Ernst &
Vol. 35, No. 4, 1970, pp. 710–718.
Young, 2015. As of September 26, 2017:
Colby, Sandra L., and Jennifer M. Ortman, Projections of the [Link]
Size and Composition of the U.S. Population: 2014 to 2060, EY-rise-of-gen-znew-challenge-for-retailers/
Washington, D.C., U.S. Census Bureau, March 2015. As of %24FILE/[Link]
September 25, 2017:
[Link]
publications/2015/demo/[Link]

24
Myers, Karen K., and Kamyab Sadaghiani, “Millennials in the Stein, Joel, “Millennials: The Me Me Me Generation,” Time,
Workplace: A Communication Perspective on Millennials’ May 9, 2013. As of February 15, 2018:
Organizational Relationships and Performance,” Journal of [Link]
Business and Psychology, Vol. 25, No. 2, 2010, pp. 225–238.
Strauss, William, and Neil Howe, Generations: The History of
National Center for Health Statistics, “Births and General America’s Future, 1584 to 2069, New York, William Morrow &
Fertility Rates: United States,” January 15, 2018. As of April 26, Co., 1991.
2018:
[Link] Twenge, Jeanne M., Generation Me-Revised and Updated:
NCHS-Births-and-General-Fertility-Rates-United-Sta/e6fc-ccez Why Today’s Young Americans Are More Confident, Assertive,
Entitled—and More Miserable Than Ever Before, New York:
Noble, Stephanie M., Diana L. Haytko, and Joanna Phillips, Simon and Schuster, 2014.
“What Drives College-Age Generation Y Consumers?” Journal of
Business Research, Vol. 62, No. 6, 2009, pp. 617–628. U.S. Census Bureau, “Millennials Outnumber Baby Boomers and
Are Far More Diverse, Census Bureau Reports,” press release,
Pampel, Fred C., “Cohort Changes in the Social Distribution of June 25, 2015. As of December 20, 2017:
Tolerant Sexual Attitudes,” Social Forces, Vol. 95, No. 2, 2016, [Link]
pp. 753–777. [Link]
Pollard, Michael, and Matthew D. Baird, The RAND American ———, “2014 National Population Projections Datasets,”
Life Panel: Technical Description, Santa Monica, Calif.: RAND database, April 6, 2017a. As of April 26, 2018:
Corporation, RR-1651, 2017. As of April 2, 2018: [Link]
[Link] [Link]
Pollard, Michael, and Joshua Mendelsohn, Methodology of the ———, “Annual Estimates of the Resident Population by Single
2016 RAND Presidential Election Panel Survey (PEPS), Santa Year of Age and Sex for the United States: April 1, 2010 to July 1,
Monica, Calif.: RAND Corporation, RR-1460-RC/UCLA, 2016. 2016,” June 2017b.
As of April 26, 2018:
[Link] Vespa, Jonathan, The Changing Economics and Demographics
of Young Adulthood: 1975–2016, Washington, D.C.: U.S. Census
RAND Corporation, RAND American Life Panel, website, Bureau, 2017. As of December 20, 2017:
undated. As of December 20, 2017: [Link]
[Link] publications/2017/demo/[Link]
Robinson, Robert V., and Elton F. Jackson, “Is Trust in Others Wechsler, James, “A New Picture of Youth, No Longer ‘Flaming,’”
Declining in America? An Age-Period-Cohort Analysis,” Social New York Times Magazine, December 26, 1937.
Science Research, Vol. 30, No. 1, 2001, pp. 117–145.
Weinbaum, Cortney, Richard S. Girven, and Jenny Oberholtzer,
Schuman, Howard, and Amy Corning, “The Critical Period The Millennial Generation: Implications for the Intelligence and
and Generational Memory: Evidence for National and World Policy Communities, Santa Monica, Calif.: RAND Corporation,
Events,” Public Opinion Quarterly, Vol. 76, No. 1, 2012, pp. 1–31. RR-1306-OSD, 2016. As of April 13, 2018:
[Link]
Smith, Amy, via Richard A. Levy, U.S. Census Bureau, email with
the author, March 8, 2016. Westerman, James W., Jacqueline Z. Bergman, Shawn M.
Bergman, and Joseph P. Daly, “Are Universities Creating
Stafford, Darlene E., and Henry S. Griffis, A Review of Millennial Millennial Narcissistic Employees? An Empirical Examination of
Generation Characteristics and Military Workforce Implications, Narcissism in Business Students and Its Implications,” Journal of
Alexandria, Va.: Center for Naval Analysis, May 2008. As of Management Education, Vol. 36, No. 1, 2012, pp. 5–32.
February 16, 2018:
[Link] Winograd, Morley, and Michael Hais, How Millennials Could
Upend Wall Street and Corporate America, Washington, D.C.:
Statistics Canada, “Generations in Canada,” December 21, 2015. Brookings Institution, May 2014. As of February 15, 2018:
As of September 26, 2017: [Link]
[Link] Brookings_Winogradfinal.pdf
98-311-x/98-311-x2011003_2-[Link]

25
Acknowledgments nonprofit, nonpartisan, and committed to the public
interest.
We would like to thank the RAND Center for Global
RAND Ventures is a vehicle for investing in
Risk and Security for their support of this research.
policy solutions. Philanthropic contributions support
Specifically, we thank Andrew Parasiliti, Robin Meili,
our ability to take the long view, tackle tough and
Stanley Presser, Krishna Kumar, and Charles Ries
often-controversial topics, and share our findings in
of RAND International Programs for their feedback
innovative and compelling ways. RAND’s research
during various steps of this research project. We also
findings and recommendations are based on data and
thank Gregory Bauman, Hosay Yaqub, Erin Smith,
evidence, and therefore do not necessarily reflect the
and Steve Davenport for their support in facilitating
policy preferences or interests of its clients, donors, or
this work. Finally, we thank Michael Pollard and
supporters.
Greg Treverton for serving as reviewers.
Funding for this venture was provided by gifts
from RAND supporters and income from operations.
Security 2040
This report is part of a broader effort, an initiative of About the Authors
RAND Ventures, to envision critical security chal- Marek N. Posard is an associate sociologist at RAND.
lenges in the world of 2040, considering the effects His primary area of research focuses on social problems in
of political, technological, social, and demographic military organizations. His work has covered a variety of
trends that will shape those security challenges in the topics, including diversity in the military, unit cohesion,
coming decades. The research was conducted within social media use in military contexts, military families, and
the RAND Center for Global Risk and Security. the recruitment and retention of personnel.

Jennifer Kavanagh is a political scientist at RAND and

associate director of the Strategy, Doctrine, and Resources
RAND Center for Global Risk Program in the RAND Arroyo Center. Her research focuses
and Security on U.S. political institutions and public opinion and their
implications for U.S. foreign and domestic policy.
The Center for Global Risk and Security (CGRS)
works across the RAND Corporation to develop Kathryn Edwards is an associate economist at RAND.
multidisciplinary research and policy analysis deal- Her research focuses on applied public and labor eco-
ing with systemic risks to global security. The center nomics, how households make ends meet in the wake of
draws on RAND’s expertise to complement and earnings shocks, and contributors to inequality.
expand RAND research in many fields, including Sonni Efron is writer-in-residence at RAND. A former
security, economics, health, and technology. A board foreign correspondent, editor, and speechwriter at the U.S.
of distinguished business leaders, philanthropists, Department of State, she focuses on communicating to the
and former policymakers advises and supports the public on national security, international relations, and
center activities, which are increasingly focused on emerging technologies.
global security trends and the impact of disruptive
technologies on risk and security. For more infor-
mation about the RAND Center for Global Risk and
Security, visit [Link]/international/cgrs.

RAND Ventures
RAND is a research organization that develops
solutions to public policy challenges to help make
communities throughout the world safer and more
secure, healthier and more prosperous. RAND is
About This Report
Millennials, those born between 1982 and 2000, are the largest segment of the U.S. population, with 84.3 million
people, and by 2040, they will account for an even larger segment of the population. As these young Americans
become more prominent in all professional fields—politics, government, media, academia, business—their
attitudes, preferences, and beliefs will have increasing weight in public discourse and U.S. policy toward security.
But the millennial outlook has not been carefully studied. Do their attitudes toward security differ from the views
of previous generations? And if so, what do these perceptions imply for U.S. security policy in 2040? This
report—part of a series examining critical security challenges in 2040—analyzes survey data from a nationally
representative sample of adults, examines perceptions of economic and national security, compares attitudes and
opinions of millennials with previous generations, and concludes by making inferences about potential millennial
concerns about security in the year 2040. The report reveals that attitudes and opinions of security tend to pattern
with age, not generation. Specifically, older people expressed more worry about national security topics than
younger people, while younger people expressed more worry about economic security. Younger people also were
less likely than older people to report that living in a democracy was important to them.

[Link] $16.95
ISBN-10 1-9774-0085-X
ISBN-13 978-1-9774-0085-7
51695

9 781977 400857
RR-2571-RC