Interview with Richard Cincotta on Does a Young Age Structure Thwart Democratic Governments?

This research is not about how you perceive relationships between young people (with whom you associate) and elders, and it does not try to determine who is responsible for actions occurring in political conflict. This research focuses on the political instability that is associated with youthful population age structures (a distribution in which young adults make up a large proportion of the adult population). To hear more about this relationship, I recommend that you listen to Eric Zuehlke’s excellent interview of Henrik Urdal of the International Peace Research Institute in Oslo (PRIO) on PRB’s website: http://www.prb.org/Journalists/Webcasts/2009/youthurbanizationconflict.aspx).
A great deal of research has been conducted on the political volatility of populations with a “youthful age structure” (called youth bulge populations by political scientists) using data from over 150 countries over the past 40 years. The conclusion of that research, by several authors, is that countries with a large proportion of young adults are more prone to a new outbreak of low-level political violence and instability than countries with more mature age structures. This result remains, even after filtering out the effects of income per person and different types of governance. The research on liberal democracy builds upon the work on civil conflict conducted by Henrik Urdal and colleagues at PRIO, Jack Goldstone (George Mason U.), Christian Mesquida (York U.), Elizabeth Leahy Madsen (Population Action International) and others.
My recent research (with John Doces, Bucknell U.) uses the category “FREE” in annual Freedom House assessments (www.freedomhouse.org) to indicate the presence of a liberal democracy, and we use data published by the UN Population Division to measure the youthfulness of national populations. We find that countries have a much smaller chance of being assessed as a liberal democracy when their age structures are young, than when those populations are more mature.
There are two effects. The first is the “advent effect”: countries with very youthful age structures are less likely to rise to the FREE category (liberal democracy) until fertility has declined significantly and, almost two decades later, the age structure has matured. This effect can be seen in the rise to high levels of democracy by countries in East and Southeast Asia (South Korea, Taiwan, Indonesia, and until recently, Thailand). The second, stronger effect is called the “maintenance effect”: countries that become liberal democracies when they have a very young age structure have trouble retaining that assessment.
In a demographic sense, “youth bulge countries” that become liberal democracies arrive at this regime type too early. These democracies tend to be fragile. Within 10 years, most shift to lower levels of democracy (“PARTLY FREE”) or become authoritarian (“NOT FREE”). Examples of countries that reached liberal democracy “too demographically early” include countries in South America (Venezuela, Colombia, Ecuador, Peru, Bolivia and Chile). Some became liberal democracies in the 1960s and 70s, but suffered through electoral violence, military take-overs, and insurgencies. Most of them retreated to more restrictive regime types. Chile is a good example. It rose to liberal democracy with a very young and radicalized population, then suffered a military coup and authoritarianism at the height of it youth bulge. As Chile’s age structure matured, however, its authoritarian regime became less popular—and ultimately it was replaced by a resilient liberal democracy.
Our theory predicts that, as their population age structure matures, the political environment of other Latin American countries will become less volatile and less fractionated, and will follow Chile’s lead. If governments take advantage of their “demographic bonus”, job growth is likely to occur, the pervasiveness of “youth culture” is likely to fade, and with fewer unemployed young people, political organizations (and government security services) will find it more difficult to recruit. Researchers who have studied how countries become democratic (including Samuel Huntington and Philippe Schmitter) have found that when politics become less volatile, when crime drops, and the economic climate improves, elites (big businessmen and property owners) are unwilling to support authoritarians. Why? Because authoritarians, their families, and cronies, are bad for business—they control commerce and trade, and extract wealth and favors.
Interestingly, our theory is consistent with the view of Thomas Hobbes (a 16th century political philosopher) who recognized that citizens tend to trade their political rights for greater security—a relationship that has come to be called the “Hobbessian bargain.” When there are external and domestic threats, citizens seem willing to give up their freedoms to someone who can guarantee their security. When threats disappear, citizens typically desire a more free society. Youth, who have not yet acquired material wealth, built a family, or organized a commercial network, are less interested in this tradeoff and can afford to be risk-takers.

Please see my response to your previous question.

Please see my response to Emeka Nwosu.

Mario,
This research, so far, does not find a clear relationship between population size and liberal democracy In fact, we do not find a relationship between population size and any type of regime—authoritarian, partial democracy, or liberal democracy.
The focus of the research is on the age structure of a county’s population. Several measures indicating the population’s degree of youthfulness provide a good indication of the likelihood that a country will be assessed as a liberal democracy, even when national income per capita, population size, and resource wealth are statistically taken into account. The measure that I have used is called the “youth bulge proportion”, which I measure as the proportion of young adults (15 to 29 years of age) in the working age population (15 to 64). I have tried other measures of age structural youthfulness, including the one that Henrik Urdal uses (15-24 year olds divided by all adults, 15 and older) and the one that Elizabeth Leahy has used (under 30 divided by the population). They produce similar results.
In order to create a “novel” way to test the hypothesis, I have used the ex-Soviet and Eastern European countries as a type of test of this hypothesis (see the paper linked to this website). It passes the test.

Anima and Meskerem:
This research often surprises readers. However, it doesn’t surprise demographers who are familiar with the way that development and political reform typically follow after progress through the demographic transition (declines in fertility and childhood mortality) and subsequent changes in age structure. The effect of age structure on regimes is a population-level effect, not one that is necessarily reflective of individual feelings, nor should it necessarily correspond to your perceptions of young adults whom you know. I recommend that you read my response to Emeka Nwosu and listen to PRB’s interview with Henrik Urdal.
Also, please try this thought experiment: Identify a country with a youthful age structure that underwent a “youth-led” democracy movement. Did it become a liberal democracy (with Western European style freedoms)? If the uprising succeeded, it probably became much less than liberal democracy—for example, Iran’s 1979 revolution (led to theocratic authoritarianism), the Russian and Chinese Revolutions (led to authoritarian communist states), the French Revolution (led to Napolean’s empire), the American Revolution (initially led to a partial democracy that restricted voting and tolerated slavery). Often youth-led movements are met with backlash that is meted out using equally idealistic youth recruited by the state—for example, the Basij militia in Iran’s suppression of the June demonstrations, and the PLA’s suppression of the Tiananmen
While idealism is, indeed, a quality associated with adolescents and young adults, we all need to remember that idealism takes many forms—many of which are not shared by our circle of friends and relatives. Popular notions of ideal communities—like those generated by religious fundamentalism, communism and ethnic nationalism—are not conducive to the rise and stability of a liberal democracy. Interviews of “suicide bombers” who survived because their vest-bomb did not explode, indicate that they are intensely idealistic rather than religiously suicidal (seeking the afterlife). Most young soldiers and police are also intensely idealistic.

Meskerem, I addressed your question in my response to Anima Sharma.

Robert,
I addressed your question in my response to M.E. Scherer.

Juanita,
You’ve brought up an interesting point. There are two types of comparable effects. You’ve mentioned the first: cohort effects, which arise from mass behaviors associated with a group in which individuals were born around the same years and went through similar experiences (such as, motorcycle riding by baby boomers). And then there are life cycle effects, which are patterns of changing behavior over the lifetime (for example, drivers’ risk-taking declines with age).

Geoff,
My answer to your question is somewhat controversial. I credit the unwillingness of senior demographers to promote or publish this work is a “cohort effect” (Clearly a senior demographer at PRB has broken this taboo!). Today’s senior demographers—most of them baby boomers—were trained by academics whose careers began shortly after World War II or even before. Before World War II, demographers ventured into topics in eugenics, which can be broadly defined as "the study of all agencies under human control which can improve or impair the racial quality of future generations." While this field included legitimate studies of heritability, it also harbored individuals who sought to explain the world in terms of genetic inferiors and superiors. This latter perspective was used in all corners of the globe as a justification for institutionalized racism, and became the ideological foundation of political fascism in Europe. To distance themselves from eugenics, post-war demographers distanced themselves from any type of demographic study of mass behaviors. Then along came another controversy: the resurgence of the Malthusian thesis in the 1960s and early ‘70s as an explanation for evolving world conditions and environmental degradation. As result, I believe, demography turned inward toward census and survey-related studies, and to influencing public health and sociology (and here demographers have succeeded).
Nonetheless, the pathway into demographic studies of state behavior had already been launched with Ansley Coale’s early work on age structure. His book (with Edgar M. Hoover) in 1958 predicted the demographic bonus (or demographic dividend). Economic demographers (including Ronald Lee, Andrew Mason, David Bloom, David Canning and others) have taken Coale’s thesis even further—and now the demographic bonus is accepted by international agencies as a product of fertility decline (see the recent article in the Economist, Oct. 31, 2009). But demographers, trained to stay away from topics that dealt with ‘mass behaviors” refused to venture into “Coalean” studies of conflict and governance, or to encourage its publication or promote it among students.
I believe this was a mistake. Demographic research on civil conflict suggests that a youthful age structure substantially increases the risk of an onset of civil conflict, but it does not find that age structure solely determines conflict or that it is powerful enough to “doom” states to failure. In addition, the conclusions of the research suggest that, after fertility declines and age structure matures, countries with youthful age structures will experience a reduction in “civil conflict risk” and an increase in the probability of being assessed as a stable liberal democracy. In other words, “the youth bulge thesis” does not assert that any country or region is necessarily war-like or undemocratic by nature—each country should experience some risk reduction as it goes through the age structural transition. There’s evidence that this is occurring: the geographic distribution of civil conflicts is shrinking as the number of “youth bulge” countries decreases (see the National Intelligence Council’s “Global Trends 2025” report, which illustrates this on page 20 as the “demographic arc of instability”; at: http://www.dni.gov/nic/PDF_2025/2025_Global_Trends_Final_Report.pdf .
Another fascinating aspect of this research: To a large degree, changes in women’s role in society, from child bearer and homemaker to participant in the workforce and public life, drives the age structural transition. This presents a powerfully optimistic image of a “demographic peace” that might emerge in a future time when the world’s women are fully participating in their society, and making their own choices, and the age structures of all the world’s countries have grown mature. It’s a nice image, but there are a few complications looming along the way (see Neil Howe and Richard Jackson’s article, “the Battle of the (Youth) Bulge”. See my comments on these issues in the Environmental Change and Security Project’s blog at: http://newsecuritybeat.blogspot.com/2008/10/guest-contributor-richard-cincotta-on.html

G.
Please see my response to Laura Underwood's question. It addresses some of the same issues.

Laura and G.,

The study does not focus on birth rates (the number of babies born for every 1000 people in the population), nor does it try to figure out the role of communications. It draws a simple conclusion—one that is obvious to those who understand how the demographic transition has spread around the world and how youthful age structures are currently distributed: countries with a mature age structure are more likely to become liberal democracies and stay liberal democracies than countries with youthful age structures.
What about the US? At the time of its founding, the US could not be considered a liberal democracy. Voting was originally restricted to land-owning white males, and one-third of the population of the southern region was enslaved. According to US Census Bureau, by 1800, half of the US population was under the age of 16 (comparable to Afghanistan and Democratic Republic of the Congo today). The two most recognized “regime data sets” (Polity IV and Freedom House assessments) many more years of political infighting, civil war, Indian wars, and racial and criminal violence passed before US citizens lived under a liberal democratic regime.
When did the US reach the liberal democracy mark? The Freedom House dataset begins in 1972, so it is impossible to determine when their assessment system would have deemed the US a liberal democracy from these data. Interestingly, when I’ve informally interviewed democracy researchers who use this scoring system, some argue that the US should not have made its present very high democracy rating until 1965—after the Voting Rights Act of 1965 was in place (which was used by the federal government to overturn local regulations (primarily in southern states) that restricted a significant proportion of African Americans from voting. I’m still not sure when this assessment would have first considered the US to have become “FREE”. My demographic model tells me there was a 50-50 chance in about 1920 (the year that women received the vote in federal elections).
In the late 1969s and early ‘70s, however, the US “youth bulge” was on the rise, driven by a surge in US fertility after the end of World War II, and the transition of those baby boomers from adolescence to young adulthood. Did the youthful US age structure contribute to the experimental youth culture and political unrest of the time, and ultimately to the conservative backlash that followed? Some researchers believe that it did. As the baby-boom generation entered the adult years, the US youth bulge proportion (the population of 15-to-29 year olds in the working age population, 15 to 64) rose to its highest level, 0.41—which is around the values that Malaysia has today; and it is higher than Indonesia’s current youth bulge proportion (0.39; Indonesia is now a liberal democracy). This value is still much lower than current values in Afghanistan (0.53), Iraq (0.50), or Nigeria (0.52). Despite that “border-line” sized bulge, this period was politically tumultuous and fractious. Public demonstrations by the Left were met by a backlash of voting that pushed the political system to the Right. However, the US has not declined from Freedom House’s category “FREE” from 1972 (when the Freedom House system was initiated), to today. Thus, age structure is not the only issue. It does, however, provide some insight into political dynamics in many countries.

Elizabeth,
As you well know, we can’t be sure that fertility will decline to much lower levels soon in all countries in the western, central and eastern regions of Africa, or in Afghanistan and parts of Pakistan and northern India. The UN Population Division and US Census Bureau projections assume slow rates of fertility decline in these countries, but these assumptions are based on the history of trends in other regions where, very often, women’s traditional and legal status was higher, where political systems were more stable, and where outlets for contraceptives and information where more accessible. Recent surveys show fertility trends in some of these countries to be “plateauing”—stalling after a short period of decline. And even if rapid fertility declined rapidly (as it did in Iran in the 1990s), these youthful countries would still experience at least another 20 years of “youth bulge”, until a younger “small family generation” came of age.
If fertility remains high in sub-Saharan Africa, the age-structural youthfulness that contributes to its countries’ political fragility will be extended further into the future. Without some slowdown in the ongoing surge of children and young adults, weak infrastructure-poor states will be continuously challenged to provide their rapidly growing numbers of young citizens with needed services and jobs. Failure to provide these will undercut the already tenuous legitimacy of the region’s weak states.
On the other hand, let’s not be too pessimistic. Three decades ago, few Western demographers would have predicted that the Middle East/North Africa Region (MENA) would now include countries near or below replacement fertility, such as Tunisia, Iran, Turkey, Lebanon, and Algeria. And others, including Morocco, Egypt, Syria, and Libya have experienced substantial fertility decline. According to my calculations, the age structures of Tunisia, Lebanon and Turkey are very close to the point where countries have had a 50-50 chance of being assessed as a liberal democracy (when the youth bulge proportion is about 0.39; the youth bulge proportion is measured as the population, aged 15 to 29, divided by the working-age population, 15 to 64).
Even with its very young age structures, sub-Saharan Africa has several liberal democracies—although the proportion of the total number of countries is relatively small (which is consistent with predictions). There are four liberal democracies in West Africa: Ghana, Benin, Mali and Cape Verde—Senegal recently dropped out of this category (this instability is also consistent with predictions). All four of the liberal democracies—South Africa, Botswana, Namibia, and Lesotho—are countries with a high prevalence of HIV and significant AIDS-related mortality.

M.E. and Robert,
As I’ve written to Emeka Nwosu and Meskerem Bekele, this research is about an age structural effect, not about individuals or groups within the population. As the age structure of a country becomes more mature (as it moves through th age-structural transition), its chance of becoming a liberal democracy, and remaining one, improves.
Personally, I believe that the “ease of recruitment” effect is the principal driver in this relationship. However, it is not clear that it is all that matters. Changes in age structure run in parallel with key societal changes that are difficult to define measure. These include: modernization (perhaps related to your observation), increases in parents’ perceived value of their children, and the improved status of women. Our research cannot demonstrate that these are influential in the “democratic transition”. However, it seems reasonable that each of these are associated with progress in the demographic transition and the shift from very young to more mature age structures.

Alberto,

My research only looks at very high levels of democracy, which researchers call liberal democracy (the Freedom House category "FREE"). However, by looking at the data, you do get a sense of how demography and economics influence the political stability of countries.
If you're thinking about Colombia, the data suggest a very positive future for your country. Colombia is going through dramatic changes in its age structure that are driven by a fairly rapid fertility decline. If age structure can be considered a clue to the future, I would suspect that the timeline for FARC is running out. With Colombia's youth bulge declining it will be more and more difficult for FARC to recruit. Improvements in educational attainment and job prospects should help speed this transition to a more peaceful and democratic state.
Here's a prediction from my work on civil conflict: the age-structural model suggests that insurgency (FARC) will be increasingly difficult to sustain. I believe FARC will be "out of action" in 5 to 10 years. When that happens, Colombia is likely to be assessed as a liberal democracy (it is very close now).
Actually, many poor countries are able to sustain low levels of democracy quite well. High levels of democracy (liberal democracy, with total freedom of the press and assembly, and organizational rights given to extremist parties) may be unadvisable.

Your question is well beyond the scope of my research. I will say, however, that it should help countries if all minority groups and regions were provided with educational opportunities (particularly for women), health services and economic activities. When large minority groups and regional groups lag in the demographic transition, they retain their "age-structural youthfulness" -- and they continue growing rapidly. This can produce ethnoreligious ... which exacerbate existing political tensions and produce new ones.
The most obvious example is Lebanon. While Christians and Sunni populations have passed rapidly through the demographic transition, Shiites have languished in poverty and been marginalized politically. The result has been a very youthful and politically active Shiia population capable of changing Lebanon's political balance.
Another example is India. UP and Bihar remain very youthful while South India and India's cities are passing through the age structural transition. If UP and Bihar continue to [be] left out of the educational, economic and fertility transitions, I think it will mean significant problems for the Indian state. Some say that this problem is already manifesting itself in migration to city slums and the re-emergence of the Naxalite Movement.

Usha,

The "youth bulge" research does not have answers for all your questions. Its scope is very limited. However, demographers note that India is a difficult case to categorize: Perhaps because of its size and heterogeneity, and perhaps because of its development policies, it seems to be going through several demographic transitions simultaneously.
The earlier and more rapid transition is occurring in the southern states and urban areas. Age structure is indeed maturing rapidly and education, mostly provided by the private sector, has produced an extraordinary generation of qualified workers and professionals in teh south and in the major cities.
On the other extreme is UP and Bihar. Fertility decline has stalled, and this should be of concern to the Indian government. Oddly, many Indian policymakers seem unconcerned; they see this youthful, poorly educated population as future workers. I believe this is a mistake. Instead, sustaining this youthfulness and population growth is likely to feed India's ongoing insurgencies and its slums.
India has an amazing record as a "liberal democracy", but it is one that is tarnished by several retreats from high levels (the Emergency period, 1977-78; and a few years during the BJP government). And no state in the world (except maybe Mayanmar) can compare with India's ability to fight off multiple ongoing insurgencies. I wish I could be more positive about India -- it's a very mixed picture.

Ghazy,

This is not about individuals. The relationship between age structure and liberal democracy is a population-level effect. The demographic transition, and the age structural transition that it drives, have large scale influences on countries. When the age structure is young, politics tend to be fractious and volatile; the risks are higher for civil conflict than at other times (80% to 85% of all civil conflicts in the last 40 years have begun in states where more than 60% of the population was under age 30).
When the age structure matures, it becomes easier for the state to provide services and jobs. It's a "demographic bonus" provided by fertility decline. With many workers, there are more opportunities to produce and tax.
When age structures amass large numbers of seniors --when age structures become very mature -- the state is again challenged to provide adequate services, particularly health services and pensions. In this stage of the age structural transition, workers may find themselves burdened by the demands of seniors. The challenges of this stage of the transition are yet to be realized. Japan and numerous European countries will encounter them in the coming decades.

Noel,

You may be correct. For most industrial countries, the last 60 years have, indeed, been the demographic good times. As Japan and Europe age, the needs of their senior citizens will grow and the size of their workforces will slowly decline. Can liberal democracies continue under these conditions?
On this, our research is completely mute. We don't know. One issue in Europe that may be significant is the influx of non-European workers. Some Europeans believe that these immigrants are creating divisive tensions that will eventually force some of Europe's liberal democracies to be "less liberal". On the other hand, many immigrants to Europe believe that they can be good citizens if [the] European notion of citizenship is expanded.
Good question.

Scott,

I'm guessing that your question is about "cohort effects" versus "lifecycle effects". In this research we have assumed that the lifecycle effects that make political mobilization/recruitment easy (such as risk-taking, sensitivity to peer pressure, idealism, naive acceptance of ideologies) dominate. The literature suggests that these "youth effects" decline rapidly during the 20s. While the amount of criminal violence varies greatly among countries, in nearly every country in the world, 15 to 30 year old young men account for 85 to 95 percent of all criminal violence. They are also the principal victims of that violence.
After that, they become far less important as perpetrators of violence. So, if you want to demonstrate future murder trends, just plot the proportion of young adults (15 to 29) in the adult population (15+). It works very well. There's a nice paper on this by Cohen & Land (1987) in American Sociological Review. I've followed the trend since then, and it fits very nicely. With all the talk about guns and drugs, few policymakers take the time to look at the demography of violence.

Hazel,

Some work has been done on this topic by Henrik Urdal (PRIO) and Bilal Barakat. It is now easier to include education in the mix because of a recently published dataset offered online by the International Institute of Applied Systems Analysis (IIASA) and the Vienna Demographic Institute (VDI). IIASA and VDI have generated age structures that show the educational attainment levels in each cohort. This is important because: (1) age structures vary significantly, even within some regions; and (2) age cohorts have varying impacts on production and politics, depending on how old they are. For example, a highly educated group of 70 year-olds is unlikely to contribute very much to workforce productivity, but they are likely to be considerably more healthy than a much less educated cohort of similar age.
In the Urdal-Barakat conference paper, I remember them finding that seconday school education was important, but it didn't take too much education to get them up to the level that reduced the proportion of conflict. You'll have to talk with Henrik on that one. It was not a finished paper and the authors may have done more work and reached other conclusions.
I think the big question was: Does education make people more likely than less educated people to be discontented in opportunity-sparse populations? Many political scientists think that educated people are more apt to be discontented and therefore politically extreme. I think the jury is still out on that one. And I'm not sure the issue will be easily resolved. As I've said before, there are "individual effects" and then "populaton-level effects", and those can be very different. Mixing them up creates the type of questions that you've seen on this particular forum ... and some people satisfied with this answer.

Ron,

Apparently, the age structure of the electorate is relevant to the ability of states to "consolidate" a liberal democracy ... in other words, to make high levels of democracy (and the civil liberties and political rights that compose it) the only choice that interests the electorate. The results of this research suggest that it helps to have a significant proportion of older adults to produce a political environment that a liberal democratic system can manage and remain in place.
This doesn't mean that leadership and institutions do not matter. They do. There are two countries that have overcome the "youth risk" and cruised through their youth bulge while maintaining Freedom House's "FREE" rating: Costa Rica (the most amazing outlier); and Jamaica (which experienced a considerable amount of political violence). They are both small countries, of course. But that's it. Even with India's dedication to liberal democracy, it has fallen twice from the FREE rating over the last 40 years (not bad for a huge, poor and heterogeneous country).
On the other hand, there are countries that this age-structural model suggests should be liberal democracies, but aren't: China, Russia, Cuba, and now Thailand (what happened there?). That's the beauty of this model; it produces statistical expectations about liberal democracy. Even when they are wrong, it makes you wonder why.
I believe that two types of countries can resist this demographic shift: (1) where a single party is synonymous wiht the state (China, and probably Iran); and (2) where a charismatic leader has been able to capture or inculcate popular support (Russia, Singapore, North Korea). Hugo Chavez is trying to follow this model, but he may not succeed. Venezuela, while behind other Latin American countries in the age structural transition, is maturing rapidly.

Ghazy,

There are differences between (1) the way that you observe the behavior [of a ] person or people, and (2) the influence of those people, in large numbers, on a population or system.
For example:

Observation of individuals: I notice that Americans buy fast cars. Individually, they love fast cars.

Observation of the population: The average speed of those car owners is about 50km/hr.

What happened? Lots of owners of fast cars create enormous traffic jams. There are population effects that do not mirror individual effects.

Henry,

I hope more research will be conducted on populaton effects of high educational attainment and the stability of high levels of democracy. However, we haven't tried that yet. Please look at my answer to Hazel Denton's question. I refer to a new dataset published by IIASA and Vienna Demographic Institute. They have generated data for educational attainment (4 categories) by sex and by age (5 year age groups). In the future, I hope I'll have a chance to add this to the model.
However, I have tried to see if it mattered in the transition to democracy among the ex-Soviet/Eastern European states, and it didn't seem to matter. I don't believe that result can be generalized to less developed regions.
You asked about education in democratic countries. One demographic point: in mature age structures, it is easier for governments to spend more on each child because the school age population is either growing slowly or declining. This is one of the rewards of the "demographic bonus" years. Thus, countries that advance through the age structural transition are likely to vastly increase their human capital stock. To find out more about that, see work by Wolfgang Lutz (IIASA), Anne Gujon (VDI) and colleagues.
You're right, of course, it takes more than demographic change to attain and maintain liberal democracy, but our research suggests that a mature age structure certainly helps.

Luz Maria,

A very nice quote. The research is solely concerned with the population-level effect, so it cannot validate hypotheses about individual and small group behavior within a population. However, Mr. Cicero touched on an interesting relationship that extends through human history. The young fight, the elders govern. Even in youth-led movements, this holds constant. In Cicero's model of a democracy, the young men provided the security and then were rewarded with political rights. The elders were then obliged to govern democratically. It was a "bargain".
About 1600 years later, Thomas Hobbes composed a more generalized (and somewhat more cynical) model of this bargain. He suggested that political rights and security were intimately related. Citizens desired political and civil rights -- their freedoms. But when citizens required a great deal of security, they were willing to forfeit their political rights to a sovereign. This is the general model that underlies the youth bulge theis on liberal democracy.
However, your quote brings up the subject of the importance of older adult generations. The results of the research suggests (to me, at least) that the older generation has to feel secure in order to opt for liberal democracy. If they do not feel secure, if they feel that a youth-led change risks their family and property, they will likely support "less-democratic" leadership. By nature, young people are excited by innovation and change. Ultimately it is the older generation that accepts change or decides not to stand in the way.