People-Powered and Non-Violent Social Movements: Forcing Gradualist Democratic Reforms in Authoritarian Societies

Not all successful unarmed civil insurrections against dictatorships take place in a dramatic mass uprising with hundreds of thousands occupying central squares in the capital city. There have also been cases of nonviolent struggles against autocratic regimes that failed to topple the dictatorship in a revolutionary wave, but did succeed in forcing a series of legal, constitutional and institutional reforms over a period of several years that eventually evolved into a liberal democratic order. These more gradualist transitions have taken place across different regions and against different kinds of authoritarian systems. This webinar will tell the story of pro-democracy movements in three of these countries—Brazil, South Korea and Kenya—and how they were able to force, over time, autocratic governments to agree to substantive democratic reforms. By focusing on the role of civil society, this presentation challenges dominant, top-down, institution- and elite-based approaches to democratization.


INTRODUCTION
Recent decades have witnessed the power of unarmed civil resistance struggles in bringing down autocratic regimes and ushering in democratic governments along with a concomitant rise in the academic literature examining the phenomenon. Much of the attention, however, has gone to those uprisings which, following dramatic scenes of hundreds of thousands of people in a central square demanding the end of dictatorship, the regime collapses after only days or weeks of mass protests, as occurred in the Philippines in 1986, Czechoslovakia in 1989, Serbia in 2000, and Tunisia in 2011. Less understood have been the largely nonviolent struggles against autocratic regimes that failed to topple the dictatorship in a revolutionary wave, but did succeed in forcing a series of legal, constitutional, and institutional reforms over a period of several years which eventually evolved into a liberal democratic order. As with the more sudden and dramatic transitions, these gradualist transitions have taken place across different regions and against different kinds of authoritarian systems, including Brazil, Argentina, Uruguay, Guyana, South Korea, Taiwan, Kenya, Ghana, Hungary and Croatia. Each of these countries emerged from dictatorship as a result of protracted nonviolent struggles with no clear discernable date of transition from authoritarianism to democracy.
This article challenges the tendency by some analysts to mistakenly characterize these gradualist forms of democratization as primarily top-down transitions by enlightened government leaders or the outcome of a struggle between competing political elites rather than recognizing the critical role of civil society in forcing these reforms from below. This article examines pro-democracy movements in three of these countries-one each from Asia, Africa, and Latin America-and how they were able to force, over time, autocratic governments to agree to substantive democratization, including free competitive elections, freedom of assembly and speech, a free and independent press, a peaceful transition of power, and an independent judiciary. These three cases are particularly pertinent in that, in each of them, the first free elections resulted in the election of candidates identified with the old autocratic order, yet these elected presidents were nevertheless obliged to honor the democratic changes made possible by the resistance movements.
This study looks at pro-democracy movements engaged in strategic nonviolent action and their impact on the democratic transitions in Brazil, South Korea, and Kenya. As with the more abrupt democratic transformations, the most significant driving force was the application of strategic nonviolent action. These were situations in which the pro-democracy movement was not strong enough to bring down the regime, but the regime was not strong enough to defeat the pro-democracy movement. As result, it could be argued that, under such circumstances, a protracted struggle was inevitable. Over a period of several years, popular movements-in tandem with emerging political parties and elite opposition figures-were able to force increased political openings over time, even if the face of government attempts at retrenchment and repression.
The examination of these three cases illustrates that the application of strategic nonviolent action has served as a critical component in gradualist democratic transitions, complementing negotiations, international pressure, elections, the actions of opposition leaders and political parties, and other more conventional methods of political influence. It looks at the role of organization and leadership, strategy, coalition-building, and nonviolent discipline in determining the success of these popular struggles and its implications for future transitions from authoritarianism to liberal democracy. Particularly in light of recent cases in which military officers or civilian demagogues have successfully undermined democratic institutions in a number of countries, this study underscores how establishing and maintaining a democratic system often requires civil resistance by the populace to defend it and that it cannot be left to conventional political mechanisms and established political leaders.

THEORETICAL APPROACHES
The traditional theoretical approaches to democratization fall into a few broad perspectives, including modernization theory, class power theory, and elite choice theory.
The modernization approach focuses on social and economic characteristics of countries and political institutions (Lipset, 1959;Huntington, 1991;Boix and Stokes, 2003). Among the factors considered in this approach are economic development, cultural homogeneity or pluralism, and a legacy of British rule are among the factors thought to promote democratization. Although useful in identifying the correlates of institutionalized democratic structures, this approach does not address the actual process of democratization, or how and why democratization may occur in countries which do not fit their model.
The class power approach to democratization focuses on changing structures of class and state and types of political regimes, asserting that democratization results from a positive coalition of class forces and transnational relations, making the case that countries in which workers and capitalists have aligned themselves against historically anti-democratic large landowners and are not crushed by state repression, the prospects for democratization increase (Moore 1966;Rueschemeyer et al., 1992). As useful as this approach may be in terms of the impact of changing social, economic, and political structures regarding certain class coalitions and state relations and their impact on democratization, it also misses the more proximate agency-driven aspects of democratization.
The elite choice approach focuses on elite negotiation, compromise, and pacts that characterize many transitions (O'Donnell and Schmitter, 1986;Karl and Schmitter 1991;Higley and Gunther, 1992). This approach focuses on the need for agreement between regime reformers and moderate opponents and their ability to convince or neutralize hardline elements within their respective camps to accept compromise (Przeworski 1991). While this approach is useful in accounting for the role of political elites in democratic transitions and consolidation, it fails to account for the pressures put on the system through civil resistance and takes an inappropriately negative view towards nonviolent resistance and other forms of extra-institutional politics as being a threat to democratization.
Indeed, mass-based nonviolent resistance has often been viewed as an obstacle to democracy rather than as a driving force in democratic transitions. The historical record, however, shows that political elites rarely give up their privileges without sustained mass mobilization from below (Schock 2005). Karatnycky and Ackerman (2005) examine sixty-seven democratic transitions from the 1970s through the 1990s, finding that the occurrence of mass-based nonviolent resistance was a significant factor in most democratic transitions and countries with strong and cohesive nonviolent civic coalitions were more likely to be more democratic in the post-transition era. Similarly, Chenoweth and Stephan (2011) found that in countries that experienced successful massbased nonviolent resistance campaigns, there was a relatively higher level of democracy and lower level of the recurrence of violent civil conflict in the years following the struggle compared to the conditions in countries that experienced successful violent struggles.
Gleditsch and Celestino (2013) argue that nonviolent protest and direct action promote democracy, while violent revolution creates a series of events which furthers autocracies in the future. It challenges elite-driven perspectives that democracy emerges from elites' pacts while noting how non-elite actors are significant role players in regime changes. Their experiment spans roughly 100 years, going beyond the focus on civil resistance campaigns which force transitions but those which consolidate democratic governance, which they attribute to effects on dispersing power and increasing influences for comparison and concessions. Bethke and Pickney (2019) make the case that groups which participate in nonviolent resistance campaigns and then enter political office in democracy are the strongest for strengthening democratic quality. They also suggest that democratic regimes that transition through nonviolent resistance have a more developed civil society post transition. Furthermore, they argue that post-transitional democracies may also be better equipped to create democratic legacies and presence of nonviolent civil resistance prior to the -transition and increased freedom of expression post-transition were highly related.
Della Porta (2016) uses a contentious politics approach, focusing on the mechanisms between structure and action. She notes there are two major paths which arise from opportunities that come from alignment of elites. She observes how "participated pacts" could jeopardize a civil society's change at democratization in the event repressive regimes blocks the freedom of association limits important knowledge to insiders. Focusing on the actors' agency, giving them leverage to influence external events. Among the independent variables she suggests influences the democratic transition include military position, structural influence, and policy and that socioeconomic, political, and cultural conditions do in fact inform the choice of actors in protests.
The literature involving comparative case studies of social movements engaged in strategic nonviolent action in support of democracy is fairly limited but has been growing in recent years. With two exceptions, noted below, none of these include any of the three case studies covered below. Ackerman and Kruegler (1994) offer strategic principles that explain the trajectories and outcomes of nonviolent movements. They examine six nonviolent struggles (the first Russian Revolution of 1904-1906; the German resistance against the Franco-Belgian occupation of the Ruhr region in 1923; the Danish struggle against the Nazi occupation from 1940 to 1945; Gandhi's Indian national independence movement in 1930-1931and, the civic strike in El Salvador in 1944) and test their hypothesis that the chances of success for popular movements are enhanced if such movements adhere to twelve key strategic principles, which they placed into three categories: principles of development, principles of engagement, and principles of conception. Ackerman and Kruegler assert that the success or failure of nonviolent movements, as with many military campaigns, depends significantly on the movements' development of a wise strategy and appropriate tactics. Good strategy, these authors maintain, is based on the formulation of clear political goals followed by the development of specific campaigns and tactical maneuvers designed to achieve shortterm, medium term, and long-term goals.
Kurt Schock (2005) examines successful pro-democracy struggles in South Africa, the Philippines, Nepal, and Thailand as well as the failed pro-democracy movements in China and Burma, and how the use of strategic nonviolent action resulted in regime change in some cases, but not in others. He takes a relational approach to nonviolent action, emphasizing the importance of resilience ["the capacity of contentious actors to continue to mobilize collective action despite the actions of opponents aimed at constraining or inhibiting their activities," (2005:142)] leverage ["the ability of contentious actors to mobilize the withdrawal of support from opponents or invoke pressure against them through the networks upon which opponents depend for their power," (2005:143)] and third party intervention in the outcomes of nonviolent struggles. Schock's central conclusion is that the skills, strategies and attributes of a nonviolent movement can potentially overcome adverse structural conditions. Marchant and Puddington (2008) compare structural conditions in 64 democratic transitions between 1975 and 2006 in order to evaluate "whether one can identify underlying, preexisting conditions that favor the emergence, success, or failure of . . . civic movements." They found that "Neither the political nor environmental factors examined in the study had a statistically significant impact on the success or failures of civil resistance: had a statistically significant impact on the success or failures of civil resistance movements. Among the major implications of this finding is that civic movements are as likely to succeed in less developed, economically poor countries as in developed, affluent societies. The study also finds no significant evidence that ethnic or religious polarization has a major impact on the possibilities for the emergence of a cohesive civic opposition. Nor does regime type seem to have an important influence on the ability of civic movements to achieve broad support. (2008:1). McAdam et al. (2001) identified nonviolent action as a form of "contentious interaction" on the same spectrum with (though obviously different from) other forms of contention, including terrorism. The contentious politics approach emphasizes the interactive and relational aspects of nonviolent action, including the relationship between nonviolent challengers, local opponents, government officials, the media and external actors. Using past examples of popular nonviolent movements, McAdam et al. (2001) concluded that a combination of strategic decision making, mass nonviolent action, elite defections and external support lead to changes in power relationships between conflicting groups, which ultimately determines the outcome of the movement.
Sharon Erickson Nepstad (2012) considers not just the strategies of the resistance but the counterstrategies of the other side. It pairs what the author considers to be successful and unsuccessful struggles against Communist regimes (East Germany and China), military regimes (Chile and Panama), and personalist dictatorships (the Philippines and Kenya.) Among the variables determining the likelihood of success for a pro-democracy movement, according to Nepstad, are maintaining nonviolent discipline, thinking ahead to anticipate the regime's countermoves and ways of responding, and maintaining unity and cohesive leadership. There are serious questions regarding whether Kenya should be considered a failed case, for reasons described below. Rossi and Della Porta (2019) look at the relationship between social movements, trade unions, and transnational advocacy networks in resisting authoritarian regimes, bringing in overlapping perspectives from both the social movement and democratization literature. They look at the diverse roles such movements can play in democratization and the benefits from strikes and protest, political organizing in urban areas, and receptive elites open to democratization. One particularly relevant study is Jonathan Pickney (2020) which combines both a global statistical analysis of all such transitions between the end of World War II and 2011 along with case studies from Nepal, Zambia, and Brazil. He argues that the two most critical factors are sustained high levels of social mobilization and focus less on maximalist revolutionary goals to building new political and social institutions.
The vast majority of the cases studies in the literature have been movements against authoritarian regimes or occupying powers that were either suppressed, won through a dramatic mass uprising, or-even in cases where the nonviolent struggle was protracted-there was a clear discernable date where a country went from authoritarianism to democracy. None of these examine the question in this article regarding transitions from authoritarian rule to democratic governance in which the various significant legal, constitutional, and institutional reforms took place over several years.

METHODS
A comparative case study method was chosen in order to find similarities in cases of such gradualist democratization through the utilization of strategic nonviolent action and to better understand how and why they succeeded. Since the ten or more democratic transitions in recent decades which fit this category would be too much for this short qualitative study, I have chosen three cases which clearly follow this model and reflect a diversity in geography, culture, regime type, and political history. Most of the data for the Kenya case study was gathered through a series of sixteen detailed in-person interviews in Nairobi in 2013 and 2018 with activists, journalists, and academics familiar with these events through observation or participation. Most of the material for the South Korean and Brazilian cases was gathered subsequently through secondary sources along with three interviews in each case with expatriates currently living in the United States. In all the interviews, I took copious notes on a laptop computer. In each of the three cases below, I briefly examine the political history that resulted in authoritarian rule, early acts of resistance, the emergence of an organized opposition, government attempts to suppress the opposition, subsequent reforms and retrenchments, the eventual transition to a liberal democratic order, and subsequent political developments. Collectively, they underscore my contention that not only can popular movements engaged in nonviolent civil resistance bring about democratization as a result of a sudden and dramatic ouster of autocratic regimes, they can also bring about democratization by forcing gradualist reforms over time. Following the examination of these three cases, I put forward what appear to be common factors contributing to the successes of these struggles and the implication of these findings to the understanding of democratic transitions.

Kenya
The end of dictatorship in Kenya is thought of as the "second liberation" after the liberation from British colonialism. While there are still quite a number of remaining obstacles holding back the full consolidation of democracy in that country, Kenyan society is dramatically freer by any number of measures than it was 30 years ago. The largely nonviolent pro-democracy struggle which emerged in Kenya during the mid to late 1980s against the dictatorship of Daniel arap Moi failed in its goal to force his immediate ouster. Indeed, he was subsequently reelected twice and remained in office until 2002. However, the movement succeeded in building about a series of minor concessions by the regime which eventually added up to major changes in the country's political institutions and political culture.
There have been a number of countries where civil insurrections have succeeded in overthrowing a dictator while the regime remains intact. In Kenya, in many respects, it was the other way around. Due to a divided opposition, Moi was twice reelected by a narrow plurality, but the mobilization of popular forces resulted in less repression and more political openness and accountability, eventually forcing him to step down and pave the way for the country's current relatively democratic system and burgeoning civil society. These democratic institutions remain strong despite the subsequent elections which brought ideological successors of the old regime to office, ongoing threats and harassment of some dissident groups, and brief spasms of ethnic violence.
When Kenya's founding President Jomo Kenyatta died in August of 1978 and was succeeded by his vice-president Daniel arap Moi, there was hope for political liberalization. Political prisoners were released and other reforms were enacted. However, with the debt crisis prompting the International Monetary Fund and other international lenders to impose structural adjustment programs which led to widespread economic hardship, demands that Kenya abandon its neoliberal economic policies grew strong enough that the regime launched a wave of repression. This assertion of political power from the country's long-suppressed civil society underscored the potential for change, though the country remained a one-party state, with the Kenyan African National Union (KANU) the only legal political party. Despite growing authoritarianism, Joel Barkan (1988) noted how Kenya during this period "was nonetheless a relatively open and resilient system with multiple secondary centers of power and measure of real competition-and hence accountability-at the local and regional levels." It was this political space, however limited, through which pro-democracy activists emerged and tried to find means by which some level of resistance could emerge.
There had been some level of defiance to the regime previously, but it had been largely restricted to dissident elements within the ruling party. In the 1980s, this defiance was taken out of the parliament building and into the streets. This shift became increasingly necessary as Moi began cracking down on dissent within KANU: Raila Odinga, Kenneth Matiba, Charles Rubia, and other prominent politicians found themselves detained for the better part of a year.
The first major focus in this stage of the struggle involved the call to repeal amendment 2A of the Constitution, which imposed one-party rule under KANU. The government considered it treasonous to try to repeal the constitution, including this amendment, yet there was widespread recognition among the opposition that allowing for opposition political parties and the resulting political space for opponents of the regime was a critical first step in advancing democracy.
Among the first overt protests against the regime were a series of actions which took place at the University of Nairobi, long a center of debate and resistance among its highly politicized student body. In February 1985, students launched a boycott of classes in protest of the expulsion of students for alleged antigovernment activities. Refusing government demands to return to their classes, thousands of students converged in downtown Nairobi only to have their rally violently broken up by police. An opposition gathering on the university's athletic fields was also violently suppressed. One student was killed, hundreds were arrested, and a number were sentenced to long prison terms.
In 1986, there was a particularly severe wave of government repression against opponents, largely targeting members of Mwakenya-an underground movement advocating a return to the nationalist and socialist ideals rooted in the Mau Mau movement that had led the anti-colonigal struggle-which had particularly strong influence within educational institutions and trade unions. Though Mwakenya could be considered the ideological heirs of the anti-colonial struggle, Moi labeled them as unpatriotic, terrorists, and traitors. It was quickly infiltrated by government agents and the regime used guilt-byassociation to isolate activists. After the crackdown, some members left the country to receive military training in Libya in the hopes of launching an armed revolution, but it soon became clear that such a strategy was a non-starter. Indeed, such calls for armed resistance was used by the regime as an excuse to engage in even greater suppression of dissent.
The growing repression included a dramatic increase in torture, which was documented in a widely circulated report from Amnesty International. (New York Times, 1987) Moi's reaction was to threaten to arrest anyone affiliated with Amnesty International who tried to enter Kenya. Meanwhile, resistance spread to labor, as 10,000 textile workers defied the government's ban on strikes and walked off their jobs in August of 1987. This was the largest of more than one hundred strikes which had taken place during the previous year. Protests spread to the predominantly Muslim coastal region, as thousand demonstrated for their right to organize. In November, in response to the arrest of seven student leaders in Nairobi for "sedition," 3,000 of their fellow students came out in protests. Police brutally attacked the protesters and shut down the university, resulting in parental protests at the treatment of their children and the disruption of their education. Defiance against the Moi regime escalated further the following year following the death of Peter Karanja, a leading political dissident, in custody. More significantly, the decline of perceived legitimacy in the government became apparent in the March 1988 parliamentary elections, composed only of KANU candidates, in which 87% of voters boycotted the polls. Taken altogether, this unprecedented level of dissent left the regime in a quandary. Moi became both more extreme in his repression while at the same time becoming more accommodating in terms of reforms. Whatever hope the regime may have had that such a combination might lessen the appeal of popular defiance, it resulted in just the opposite: As with the many other cases in which a regimes have engaged in disproportionate repression against nonviolent opponents, it resulted in what is commonly referred to as "backfire" (Martin, 2015) or the "paradox of repression" (Kurtz and Smithey, 2018), whereby the legitimacy of the regime is diminished in the eyes of the public and spurs the growth of nonviolent resistance. Meanwhile, the reforms offered hope that the activism was having its desired effect, resulting in the resistance growing ever stronger.
With repression at the universities and elsewhere making it increasingly difficult to find political space for opposition activities, Kenya's churches became increasingly important centers of resistance. With virtually any other nongovernmental or non-KANU public gathering banned, Sunday morning services and other religious gatherings became virtually the only place where Kenyans could gather freely. Similarly, with increasing censorship of the media and repression of journalists, sermons became a rare vehicle through which Kenyans could hear critical perspectives of government. In January 1990, the prominent Presbyterian minister Rev. Timothy Njoya explicitly called for a nonviolent revolution comparable to the one 2 months earlier which had brought down the Communist dictatorship in East Germany. The mysterious deaths of Foreign Minister Robert Ouko in February and Anglican bishop Alexander Muge in August-both of which oppositionists blamed on the Moi regime-prompted major protests, as did the decision to evict 30,0000 residents of an urban shantytown to make way for business developments by wealthy regime supporters.
This period also witnessed the first signs of major divisions within the ruling elites, another phenomenon common amid the rise of civil resistance. (Schock, 2003). In May of 1990, former cabinet members Sharles Rubia and Kenneth Matiba publicly called for the repeal of Amendment 2A, multiparty elections, and an end to the endemic corruption. Matiba had resigned in protest in late 1988 in reaction to election fraud and Rubia had been forced out earlier that year for challenging Moi's increasingly dictatorial powers. They announced a major rally of regime opponents for 7 July, an event which became known as "saba saba," the Swahili word for the date 7/7. Saba saba was the first major opportunity for Kenyans in opposition to come together, interact, and express their opinions. After years under laws that prevented more than three unrelated people to gather without a permit, it was an opportunity to end their isolation, recognize that they are not alone in their opposition, and come forward and express themselves. Despite the risks, in an open act of rebellion against the system, many tens of thousands representing a broad cross-section of the nation came out to express their feelings in a way they had not been able to do previously. On 7 July, pro-democracy demonstrations indeed took place throughout the country, particularly in Nairobi. The regime considered them "illegal gatherings" and protesters were viciously attacked by paramilitary forces and some protesters fought back. In addition to teargas and truncheons, government forces also opened fire. Twenty-eight protesters were killed and more than a thousand were arrested.
By mid-1991, the regime found itself faced with a broad array of increasingly organized dissent, not just from lawyers and other professionals, but churches, former politicians who had broken with the regime, independent business leaders, and an increasingly angry and vocal, if disorganized, general public.
The tactic of holding political rallies only to have them violently broken up by authorities would not seem to be a sustainable strategy, yet it played an important role of graphically exposing to both Kenyan and international audiences the extent of human rights abuses in the country and contributed to the backfire against the regime. It also helped in raising Kenyans' awareness of rights that were being denied. Perhaps most importantly, it was an indication that the collective threshold of fear was overcome when the government would reiterate its ban on public rallies, but enough people decided that, despite the illegality of the action, that they would show up anyway. As with other successful civil resistance struggles (Williams, 2018), overcoming fear is a key factor in overcoming repression.
In 1991, Rubia, Matiba, Odinga and two other prominent opposition leaders formed the Forum for the Restoration of Democracy (FORD). They were soon arrested by the regime, but were released following strong domestic and international protests. More significantly, the protests forced Moi to finally legalized opposition political parties, a recognition that he could no longer maintain one-party rule in the face of growing popular demands. Robert Press (1996:71) noted "that concession followed domestic resistance that had grown from a few isolated acts by individuals to involve an emerging challenge by a wide array of opponents to the regime." FORD held their first rally in January 1992 in Nairobi, attracting a crowd of more than 100,000 people. The group represented a broad cross-section of Kenya's ethnic and regional mosaic. On January 25, there was a big rally in Mombasa, the coastal city that is the country's second largest. Despite being a KANU stronghold and despite threats to stop the rally by force, officials were too intimidated by the size of crowds to try. In both rallies, attendees ranged from those from the country's poor majority to business leaders and academics. Building on the growing grassroots movement for democracy, most observers expected that Odinga, the anticipated FORD nominee, would be able to easily defeat the increasingly unpopular Moi.
The regime and its supporters then began manipulating divisions within FORD, resulting in Matiba breaking off from FORD to form FORD-Asilia, based among the Kikuyu ethnic group. An additional Kikuyu-led faction, calling itself the Democratic Party, emerged as well. To exacerbate ethnic divisions further, Moi began a conscious effort of encouraging ethnic strife between Kalenjin and other ethnicities in the Rift Valley, claiming that Kikuyus and other more recent migrants were attempting to seize their land. The resulting violent ethnic clashes led to over 1,500 deaths and 300,000 people being displaced. Not only did the Moi regime fail to intervene to stop the fighting but actually supplied weapons and cash to Kalenjins engaged in attacks on unarmed villagers. The ethnic cleansing help solidify the region's support for his re-election and was used to bolster his claim that multiparty democracy encouraged ethnic hostilities and that Kenyans would be safer under his continued strong-armed rule.
Moi made other attempts to limit the fairness of the election, such as denying opposition groups rally permits, limiting their ability to campaign in certain outlying areas, stacking the electoral commission, refusing to register over a million young voters who had not yet received their official ID cards, allowing harassment of opposition activists, denying opposition parties access to stateowned media until just a few days before the election, prohibiting independent election monitoring, and blocking 45 opposition candidates for parliament from filing their papers, thereby allowing a number of KANU candidates to run unopposed. Despite all this, as well as a fair amount of fraud on Election Day itself, it was the division in the opposition that was most responsible for Moi's re-election. Though the 1992 election may have indeed been stolen, it would have been virtually impossible for the regime to have gotten away with it had the opposition been unified. Kenyans are divided as to what would have happened if Moi had refused to yield power in such a situation. Most knowledgeable observers believe that not only would the government been subjected to international sanctions, the fraud would have been so massive and obvious, there would have likely been a popular uprising similar to those that took place following stolen elections in the Philippines in 1986, Serbia in 2000, Ukraine in 2004, and Gambia in 2016. (Zunes, 2020) The combination of seasoned politicians with grassroots activists would have resulted in a major resistance effort. Despite large numbers of Kalenjins in army, there would likely have been widespread defections as well.
Despite this enormous setback, the elections opened up other avenues for change. In allowing for opposition parties, the regime also enabled the legal registration of a number of NGOs which ended up begin crucial in advancing the campaign for greater democracy. Civil society had been weak up until this time due to cooptation and suppression, but began to blossom in the aftermath of the reforms. In addition to underground organizations coming to the surface, new NGOS sprung up as well. Many of these civil society groups received international support.
Despite the emergence of NGOs and opposition parties, there was widespread recognition that there could be little substantive change if the same laws and constitution was in effect. If there could be such change, neither Moi nor any other president could get away with abuses. As a result, with a multiparty electoral system now in place, attention could be focused on protecting civil liberties, such as releasing political prisoners, curbing police abuses, ending detention without trial, and other restrictions limiting what Kenyans could do to build a more just and democratic society. Despite allowing for multiple parties and the emergence of NGOs, Kenyans still did not have the freedom to organize politically. Protests and other means of nonviolent resistance continued to be violently suppressed, but resistance continued, particularly in the capital of Nairobi, where mobilization was becoming increasingly routine. The regime then began deporting dissidents from Nairobi to their hometowns. Despite these efforts, it was clear that a culture of resistance had emerged, as Kenyans recognized that elections were not enough, but major legal and constitutional reforms were necessary and civil resistance was the means to move forward. A major campaign in support of political prisoners, led in large part by their mothers, was part of a larger struggle for constitutional changes guaranteeing civil liberties. Activist Tom Djeri (2013) noted how the campaign by the mothers typified the long pro-democracy struggle, noting that "At every stage of struggle there were people who wanted one specific thing, but their allies would use them to press for something bigger." It is not surprising, then, that some of the most overt use of force by the regime took place during constitutional reform struggle in the early 1990s. Moi was so scared of public defiance-manifested in rallies, marches, sit-ins, encampments, and other forms of nonviolent direction action-repression became more overt in many respects. Rather than use uniformed security services to disrupt events organized by legal organization, the regime began increasingly to rely on KANU thugs-often trucked in by the government and many of them visibly intoxicated-to attack the opposition and provoke street fights, with the police intervening only if KANU forces were losing.
During this period, resistance had become institutionalized through the NGOs, the churches, lawyers and other professionals, and opposition politicians. Robert Press notes how "activists were able to increase the pressure on the authoritarian regime to make concessions on human rights and democracy despite the relative weakness of their organization. Often the organizations were poorly staffed, poorly financed, and poorly equipped. More important was the accumulating effects of their resistance on both the public and the regime." He noted that these growing acts of resistance "encouraged public support for reforms as seen in the public's participation in mass demonstrations and other expressions of dissent" and that "it served as evidence to the regime that demands for reform had transformed from individual acts or resistance to an informal resistance movement that had wide appeal." (Press, 2006).
As a result of the reforms, opposition activists now had an inside track through their representatives in parliament and an outside track in terms of ongoing protests. Parliament became a public venue for publicizing injustices in the country as a number of activists became parliamentarians. Detention of dissidents became increasingly rare and the use of torture had dramatically subsided. Unfortunately, the focus on work within the parliament and elite politics led to a decline in protests and related nonviolent action campaigns for several years, not rebounding until the uptick to the 1997 elections. At the same time, the movement also moved beyond the capital. While protests had largely been centered in Nairobi in the earlier years, by the later part of the 1990s, there was growing resistance in other town and cities, and even some rural areas. Part of this came from the growth of NGOs and their encouragement of active participation in the political process. It was during this period when, as a result of lesser censorship, independent media emerged as major force for democratic change. If intellectuals were the leaders of the movement in the 1980s, by the 1990s it was opposition politicians and civil society leaders who were at the helm. Throughout both periods, however, while the leadership was from the middle class, activists came from a wide range of social backgrounds. Indeed, given that the poor had less to lose than the country's relatively small middle class, they were often more willing to take a chance. As Wachira Waruru (2013) put it, at a typical rally one would find "Middle class people in the leadership, but the crowd was mostly poor." A group calling itself the Citizens Coalition for Constitutional Change (CCCC) was formed which put together a draft constitution, met with politicians and civil society group, held meetings, and organized protests. They demanded changing the rules of the game and a people-centered process of drawing up a new constitution. A National Convention Executive Council was organized, made up of conveners drawn from among NGOs, youth, churches, and others. Beginning in the spring of 1997, they started organizing constituent assemblies. It was becoming increasingly difficult for Moi to dismiss the legitimacy of such a broad-based opposition and resist calls for change, forcing him to the negotiating table. Much to the disappointment of the broad opposition movement, however, he insisted on only meeting with leading politicians from the opposition parties, not civil society leaders. The resulting reforms agreed upon were much less substantive than what the grassroots movement had hoped. Still, it had become apparent that it was the movement, not the elite oppositionists, that forced the regime to come to the table in the first place. As activist Davinder Lamda (2013) noted, "What these regimes understand is mass action. It is the only time they will listen, only time they are willing to talk".
Despite ongoing repression, growing outrage at human rights abuses and rampant corruption led to an upsurge in protests in 1997. On May 31, protesters shut down Nairobi's central business district. Brazen acts of defiance even came to the floor of the parliament in the presence of Moi himself. One of the largest uprisings in the course of the struggle took place on July 7, 1997 with protests throughout in country and in virtually every major towns. Unlike the first saba saba, there were much better planned and organized. Despite killings and detentions, Moi was unable to stop this ever growing and increasingly well-organized movement. With elections scheduled that fall, it was widely assumed it would be his last year in office.
In August, however, the elite oppositionists once again betrayed the pro-democracy movement by refusing to unite behind a single candidate. The divided opposition split the votes of the pro-democratic majority in the December election, allowing Moi to once again win with a narrow plurality. Despite this, the pro-democracy struggled had succeeded in most of its demands, removing the authority of unelected chiefs, overturning prohibitions against public gatherings, revoking sedition laws, and ending detention without trial. While Moi maintained his residency in the State House, his autocratic grip on the country had been substantially weakened. He couldn't repress like he used to. As journalist Kevin Mchiro put it, "Moi remained, but the dictatorship went. He still inhabited the House, but he no longer had the same power." (Mchiro, 2013). No longer could Moi order people detained without trial. No longer could he charge opponents with sedition for criticizing his policies. Unlike many African leaders, he would not be "president for life; " thanks to new term limits, people knew he would be gone by 2002 at the latest. As Tom Djeri put it, "We did not win the struggle against Moi, but we won the struggle for the nation." (Djeri, 2013).

South Korea
Except for short periods, South Korea had been under autocratic rule since emerging from nearly a half century of Japanese occupation followed by effective partition in 1945. General Park Chung-hee had ruled the country since he seized power in a military coup in 1961. The threat from communist North Korea and the overriding task of rebuilding from the devastation from the Korean War of 1950-53 provided the excuse for the denial of democratic rights. Repression increased following his declaration of martial law in 1972.
In August of 1979, primarily female textile workers engage in a sit-in at the YH Trading Company in protest of large-scale layoffs. The government brought in riot police who attacked the women, beating them severely. The demonstrators scattered, many of them fleeing into the nearby headquarters of opposition leader Kim Young-sam, but the attacks continued, resulting in one death. Various protests continued into the fall, with increasing involvement by students, particularly at Pusan National University and Kyungnam University in Masan.
Increased unrest led to concerns of a full-scale pro-democracy rebellion. Park and other government leaders were bitterly divided on whether to compromise or engage in massive and violent repression. Apparently as a result of a heated argument during a dinner meeting, intelligence chief Kim Jae-kyu shot and killed Park. Prime Minster Choi Kyu-ja became interim president, who-following a brief state of emergency-enacted a series of reforms and released a number of dissidents, including students, religious leaders, and journalists, as well as Kim Daejung, the opposition leader who had become the country's most famous political prisoner after being kidnapped from Japan where he had been living in exile. As the pro-democracy movement took advantage of this slight opening and began gaining strength, however, anti-democratic elements in the military, led by military leader Chun Doo-wan, increased their pressure on Choi not to compromise further.
On May 14, a massive demonstration Seoul brought over 100,000 students and workers into the streets. General Chun successfully pressured Choi to declare full martial law to try to stop the protests, which gave the general effective control of government. Colleges and universities were closed and the National Assembly was shut down, along with the headquarters of most political parties. Labor strikes were outlawed and all political activity was banned. Student leaders were detained, and the country's three most prominent dissidents-Kim Dae-jung, Kim Young-sam, and Kim Jongpil-were arrested.
Despite the crackdown, in the city of Kwangju in the southwestern part of the country, student-led demonstrations continued. On May 19, Chun sent in paratroopers who brutally attacked the peaceful demonstrators with truncheons and fix bayonets, killing as many as 100 people. Students, now joined by thousands of townspeople, fought back. The nonviolent discipline which had characterized most of the pro-democracy protests of the previous years quickly evaporated, as growing numbers of students, backed by thousands of enraged city residents, raided armories and police stations to obtain weapons and drove the troops from the city. Kwangju became a liberated democratic zone within the authoritarian state. With no one to enforce the ban on demonstrations, over 300,000 people took to the streets on May 20 protesting the martial law regime. On May 27, however, with the apparent acquiescence of the U.S. government which released South Korean forces under U.S. command, Chun dispatched 20,000 elite troops to retake the city. In one of the most infamous massacres of the late 20th century, an estimated 2000 people were killed as the army retook the city.
In the aftermath of his crushing of the Kwangju uprising, Chun banned over 150 periodicals, dismissed a large number of journalists, and launched a "purification campaign," which-while including some corrupt officials and various criminals-was largely focused at purging dissidents. Universities found themselves under a heavy police presence. Within months, Chun formally declared himself president. There was a lull in pro-democracy activism over the next few years, but gradually some renewed organized dissent emerged, resulting in the suspension of certain draconian national security laws and the abolition of the unpopular midnight to 4:00 am nationwide curfew. By late 1983, Chun withdrew his forces from the universities, allowing for limited political activities, releasing student prisoners, and reinstating dismissed faculty.
The opposition took full advantage of this opening with a new wave of protests emerging in 1984. No longer limited to students and a few high-profile dissidents, other social movements-particularly labor unions-emerged in an organized fashion. Indeed, the student/worker alliance played an important role in the burgeoning pro-democracy struggle, as new organizations emerged that year, including the Council for the People's Democratic Movement in June and the National Council for Democratic Reunification, while politicians purged from the political arena by the Chun regime formed the Council for the Promotion of Democracy. (Jung and Kim, 1993) Significantly, the country's rapidly growing middle class increasingly sided with pro-democracy forces and joined in the demonstrations and other acts of resistance. The growing civil resistance campaign forced the regime to allow opposition parties-including the New Korean Democratic Party (NDKP), with its explicit opposition to the dictatorship and support for democracy-to participate in the February 1985 elections for the National Assembly.
In addition to Chun's inability to fully contain student dissidents despite police effectively occupying university campuses, he may have also hoped that he could win over some of the more conservative democratic elements to his side to ensure victory in upcoming election and there was also growing concern about international criticism in light of the country hosting the forthcoming Asian Games in 1986 and the Olympics in 1988. According to Korean social scientists Hyun Choe and Jiyoung Kim, "The Chun regime used a policy of appeasement in an attempt to shore up its political legitimacy and stability by weakening political motivation of anti-government forces. However, contrary to Chun's expectations, rather than being satisfied with the Chun government's "generous" gesture, democratic forces took this opportunity and threated the policy stability of the Chun regime." Choe and Kim (2012). NKDP candidates not only defeated the pro-government party, but the conservative Democratic Korea Party as well. This victory enabled the both the democratic movement and the public some realistic hope that democratization could be achieved in the relatively near future. The regime's declining political legitimacy had been demonstrated not just through large-scale civil resistance, but through the electoral process as well. The NKDP soon learned, however, that despite its electoral gains, it could do little in practice to advance democracy without concomitant pressure from the outside.
The failure of the NKDP to move the regime towards greater democratic reforms through negotiations and institutional strategies resulted in a return to a grassroots movement through a new coalition known as the National Liaison for Democratization (NLD), which organized large scale demonstrations which threatened the very stability of the regime. Under such pressure, Chun agreed in late April 1986 to begin talks for constitutional reform. In response, the NLD disbanded. Chun established the Special Constitution Revisions Committee to begin negotiations with opposition groups to allow for a more democratic electoral system and liberalize the constitution. However, without the NLD providing pressure, Chun continued periodic crackdowns on dissent, such as at Geonguk University in October of that year, where 20,000 students from twenty-nine universities had gathered for the opening ceremony of a new nationalist pro-democracy organization; the university was placed under siege for 4 days and nearly 1,300 students were arrested. Apparently overconfident, the regime continued its repression of student dissent, including the killing during interrogation of student activist Bak Jong-cheol. The backfire effect revived the nonviolent resistance campaign and created further demands for a new constitution which would allow for direct presidential elections.
Until this point, the regime had hoped to woo away conservative oppositionists to cooperate with them in maintaining the indirect presidential system with increased civilian input or through forming a broad but militarydominated coalition that could win a multi-party vote. Recognizing that neither scenario was likely in the thencurrent political climate, they announced in April of 1987 that the talks were ending and the current constitution and political system would remain in place for the indefinite future. While Chun announced he would not seek re-election, he named General Roh Tae-woo, a friend since they were students together, to be his successor.
This prompted additional protests, led by the growing minjing (common people) movement, which included socialist, nationalist, religious, economic justice, and peace activists. Demands were not just to bring down a president who had seized power illegitimately and to have free direct elections, but to punish those responsible of the Kwangju massacre. The intellectual base on the movement was still centered in the universities, but it had broad support from many sectors of society, challenging not just dictatorship, but militarism, big business, and foreign domination. This led to the most dramatic upsurge in protests to date: On June 10, with a massive protest planned, the regime mobilized 60,000 police to suppress it, but the crowds-which numbered over 400,000-were too large. By June 27, nearly one and a half million signed on to the call for direct presidential election. Three days later, General Roh-with Chun's approval --announced he would accept most of the opposition's demands, including the direct presidential election system, release of political prisoners, liberalizing the press, and allowing greater autonomy for colleges and universities, and other social reforms.
These dramatic concessions apparently came when the regime realized that the protests and various forms of noncooperation could not be halted, leading to fears of disruptions during the Olympic games scheduled for Seoul in September 1988 when the country would be flooded with international media. Furthermore, the growing diversity of the opposition was making it more difficult to suppress, not just in terms of numbers, but in the reaction of domestic and international opinion. It was harder to get away with attacking businesspeople, middle aged women, and clergy than it was to attack radical students. They also hoped, as had the Moi regime in Kenya, that the opposition would fail to unite, given that Kim-Dae-jun and Kim Young-sam both had strong ambitions and strong personalities, and Roh could win the election with a plurality. This was exactly what happened, with the general being elected president in an apparently fair election, but with only 36% of the vote.
Winning a plurality enabled the military to maintain the presidency, but it was more problematic by having only a plurality in the National Assembly. Increasing political freedoms made it more difficult for Roh to impose his agenda on the country as well. In a successful effort to retard the democratization process, General Roh in 1990 convinced two leading dissidents, Kim Young-sam and Kim Jong-pil and their respective parties to join a grand coalition they called the National Liberal Party, giving him a solid majority in the Assembly. This also made it possible to build a political machine powerful enough that-even in a free and fair presidential election-progressive dissident leader leading Kim Dae-jung could be defeated. The 1992 presidential election became a race between Kim Dae-jung, Kim Young-sam, and some minor party candidates. While there were fears of a military coup if Kim Dae-jung appeared on the verge of victory, the combination of his likely defeat and the solidification of democratic institutions prevented such a reversal. Indeed, Kim Young-sam won with a 41% plurality, becoming the first civilian president since the military coup of 1961, albeit with strong military backing. Though many Koreans feared that Kim Young-sam would be a puppet of the military as a result of the grand coalition, the nonviolent pro-democracy struggle which had emerged during the previous decade was strong enough that President Kim was emboldened to prosecute Chun, Roh and other generals involved in the Kwangju massacre of 1980. They were convicted and sentenced to penalties ranging from 3 years to life imprisonment. Indeed, while most other countries which experienced a more gradualistic democratic transformation effectively amnestied those responsible for human rights abuses during the authoritarian period, the nonviolent struggle in South Korea had helped create a strong enough civil society and democratic impulse that the civilian government could engage in successful investigations of the military government's human rights abuses and prosecution of its perpetrators. In the 1998 election, Kim Dae-jung was finally elected, serving for 5 years. He won the Nobel Peace Prize for his pro-democracy and human rights activism in 2000.
Despite the constant threat of war with North Korea, the high levels of militarization, and a powerful industrial oligarchy, South Korea has generally been seen as a strong democracy in subsequent years, with regular elections recognized as free and fair. A widespread civil resistance campaign in 2017 against the notoriously corrupt President Park Geun-hye, daughter of the former dictator, resulted in her impeachment, removal from office, and imprisonment. Civil society movements remain influential along with a wellfunctioning electoral system.

Brazil
A U.S.-backed military coup in 1964 ousted João Goulart, a leftleaning democratically-elected president, and brought to power the first in a series of right-wing military dictators in Brazil, marking the first in a sequence of right-wing coups that would plunge most of South America into years of severe repression. Eventually, Brazil and most other Latin American military dictatorships relinquished power by the 1980s as the economic situation deteriorated and domestic and international pressure for democratization increased. Brazil's democratic transition was more protracted, however.
A nascent pro-democracy movement had emerged in the years immediately following the coup, in the form of wildcat strikes and other nonviolent protests, but it was compromised by the emergence of urban guerillas whose efforts largely centered on bank robberies and kidnappings which received little support from the population and allowed the government to get away with massive repression. Torture and other severe human rights abuses became commonplace against both violent and nonviolent opponents of the regime. By 1974, the armed resistance movement had been completely destroyed.
Not long afterwards, however, dissent from within civil society began to grow, initially by academics, journalists, and other professionals. In 1975, dissident Sao Paolo journalist Vladimir Herzog was arrested by the regime and tortured to death. Though he was Jewish, an ecumenical funeral was held at the city's cathedral led by the Catholic cardinal, attracting tens of thousands of people. It marked a beginning of organized nonviolent protests against the dictatorship, which spread throughout the country.
In response, the new military president, General Ernesto Geisel, promised a "slow, gradual and safe" political opening. Though he commenced to lift some restrictions on freedom of assembly, which allowed for greater civil society mobilization, the goal of such limited liberalization was primarily to simply create a façade of popular participation in order to enhance the legitimacy of military rule while severely limiting the scope of opposition to the government (Lamounier, 1989).
The liberalization, known as abertura (opening) or the distensao (decompression), led to the gradual reemergence of social movements, often with the support of the Catholic Church. The church hierarchy, under the leadership of the National Conference of Brazilian Bishops, began speaking out against torture and some of other particularly egregious human rights abuses and sponsoring programs focused around a "centrist defense of civil liberties and human rights" (Della Cava, 1989, p. 146) More significant was the emergence of ecclesial base communities (CEBs), which grew from around 30,000 at the beginning of the decade to close to 100,000 by 1980. The CEBs provided opportunities for people from poor backgrounds to develop leadership skills and develop solidarity networks. This was particularly important since a number of the grassroots groups had a hard time mobilizing their constituencies, not just because of government repression, but in part due to the traditionally patrimonial Brazilian society which had not created a favorable cultural climate for grassroots struggle. Movements by women, peasants, and slum dwellers did gradually developed, however, with the Church helping the disparate struggles develop a strong solidarity network. Catholic clergy emerged as a central force in the movement against the dictatorship, including many in the hierarchy, such as those in the Archdiocese of Sao Paolo, which provided an umbrella for key opposition groups. Given the centrality of the Catholic Church in Brazilian society, there were limitations on how brutally the state could respond, at least to clergy in prominent posts.
While directly challenging military rule was still a risky endeavor, the dramatic growth of civil society organizations focusing on particular grievances-such as neighborhood women's groups, mothers' groups, and others challenging sexism at home and in society--were able to link their particular issues with the violence inherent in military rule. (Nelson, 1996) By 1977, press censorship was relaxed, habeus corpus for political detainees was resumed, and exiles were allowed to return. This limited liberalization, along with divisions between hard line and moderate elements in the regime, helped lead to a new wave of protests, which became particularly apparent in the labor sector.
A May 1978 sit-down strike of 2,500 metalworkers in Sao Bernardo do Campo, part of the industrial belt just south of Sao Paolo, was joined by other workers and neighborhood groups. By the end of year, over a half million workers from fourteen sectors of the country had gone on extended strikes, including workers in schools, hospitals, banks, and other public services. In 1979, there was an even larger wave of strikes, in which 3.2 million workers participated. Churches, even cathedrals, opened their doors for labor activists whose union halls had been raided. Going well beyond industrial workers, the walkouts included increasing numbers of professional workers as well: state employees, teachers, professional, bank employees, non-managerial professionals, and others. The strikes spread to the agricultural sector, with more than 100,000 sugar cane workers in the northeast refusing to show up for work in October. The following year, over a quarter million went on strike. The strikes, often initiated by unofficial younger leaders, challenged not just employers and the state, but the official pro-government union leaders as well.
Brazil had long been a country with a high level of unionization, but on a corporatist model which had easily been usurped by the military regime. By the end of the 1970s, however, more than onequarter of the unionized workers had become part of this new, more radical, and democratic syndicalism. One of the major waves of strikes took place just 2 days prior to the March 1979 inauguration of President João Figueiredo, the fifth general to serve in that office since the coup. Despite attempting to seize direct control of the unions and refusing to meet the workers' demands, the wave of strikes showed the new president-considered a compromise candidate between hardline and moderate officers-that force could not resolve problems in industrial relations. Following yet another new round of strikes in 1980, the workers again failed in their immediate demands but sent a clear signal that rising labor militancy that could not be subdued. The government realized that it could no longer control the country's workers. A key demand was that workers be able to negotiate with their employers rather than the state. Recognizing the impact the strikes were having on them economically, the leadership of the Sao Paolo business community agreed with the demand and also put forward a manifesto calling for the return to democracy. Meanwhile, in September of 1980, one million students and 40,000 professors went on strike, closing half of Brazil's universities.
For the Brazilian generals, the strikes and other disruptions served as a reminder, as it did with the Kenyan and South Korean regimes, that the power of government is ultimately dependent on the willingness of the population to obey. (Sharp, 1973) Recognizing the military could not hold onto power indefinitely, President Figueiredo promised to hand over to civilian rule by 1985. Protests continued, however. A general strike shut down Sao Paolo in July 1983, underscoring the evergrowing power of labor.
The Party of the Brazilian Democratic Movement (PMDB) had become the leading opposition party when political parties were legalized in 1979. In 1982, growing protests forced the government to allow for direct elections for all offices except the president. As with the previous elections, the system was stacked in favor of conservative pro-military candidates, yet opposition candidates won most of the important 1982 races. Recognizing that the majority of Brazilians opposed the military and their right-wing allies, a key demand of the opposition was the direct election of a president rather than through the militarydominated electoral college.
By 1984, demands for a direct election of the president grew dramatically, with millions of Brazilians wearing T-shirt saying, "I want to vote for President!" The Diretas Ja (Direct Elections Now!) movement emerged as the largest nonviolent resistance movement in Brazilian history. The PMDB initiated the campaign, but was quickly endorsed by Catholic clergy and other civil society actors. In an outpouring of civic spirit not seen since before the 1964 coup, regional rallies were organized bringing close to a million people to the streets in Rio de Janeiro and 1.5 million in Sao Paolo, featuring not just opposition political leaders, but pop stars, soccer announcers, and other entertainers. Despite the fervor, the vote on the amendment in Congress allowing for a direct presidential election fell short of the two-thirds majority required.
Despite failing in the short term, the massive outpouring made clear that abertura-the regime's efforts to permit a degree of political openness while perpetuating quasi-authoritarian rule under military leadership-was failing. That 1984 protests proved that such efforts from above could not contain the demands from below and that Diretas Ja "spelled the death of authoritarianism in Brazil" (Avritzer, 1995). Indicative of this growing realization were the large-scale defections of civilian legislators from the government party, forming a center-right party known as the Liberal Front to negotiate with PMDB moderates for a democratic transition.
The mass protests, combined with a decidedly unpopular right-wing civilian presidential candidate selected by the military to succeed General Figueiredo in 1985, resulted in enough defections in the electoral college for the surprise election of PMDB candidate Tancerdo Neves, a centrist governor of Minas Gerais who had served in several high positions in the federal government prior to the 1964 military takeover. Despite concerns of a possible coup, the strength of his support discouraged enough right-wing officers to prevent such an attempt. On the eve of his inauguration, however, he was stricken and died a month later. Jose Sarney, his vice-presidential running mate, who had been part of a pro-military party until he had formed a small conservative party which allied with the PMDB just prior to the election campaign, became president.
Though Sarney proved to be an unpopular president, the democratic transition had been complete. Direct elections were held at 4-year intervals subsequently, the military has been under firm civilian control and-despite great inequities and other problems-the country remained democratic. Luiz Inácio Lula da Silva ("Lula"), a leftist union leader who played a major role in the pro-democracy movement and had run unsuccessfully for president three times earlier, was finally elected in 2002 and served two terms. A series of controversial judicial and legislative actions removing his leftist successor in 2016 and banning both her and Lula from running resulted in the election of a far-right populist Jair Bolsonaro in 2018. His authoritarian leadership style has severely tested the country's democratic institutions, but the country has avoided sliding back into dictatorship and polls indicate he will not be re-elected. none of these cases, however, was there a clear demarcation from dictatorship to democracy. In each case, pressure from below enabled the regime to open a little, with the pro-democratic opposition immediately filling that space to successfully press for a greater opening which in turn would be filled with additional pressure for change. This continued, sometimes with temporary setbacks, until each country had achieved a level of political freedom, fair elections, and accountable government to be generally considered democratic. For example, in the South Korean case, Korean scholars Hyun Choe and Jiyoung Kim argue that there were structural opportunities which stimulated democratic movement forces and public participation. First of all, the opposition party and social movement groups started to transform the frame though framing (e.g. constitutional revision for direct presidential elections) and active struggle. Such efforts bore visible fruit as the opposition party won the 1985 elections. This victory further expanded the cognitive opportunity as well as the structural opportunity. In other words, the 1985 victory in the general election raised hopes for the public "expectation in the success of collective action." This stimulated popular participation in collective action, erode the regime's power to control democracy movements, and further expanded structural opportunity. (2012:6) Both the willingness of the opposition to return to large scale civil resistance and even simply the threat of a return to civil resistance provided the democratic opposition with leverage against the forces of the state. In each case, there was a dual track of pressure on the regime: from within such conventional political sources as legislative bodies, opposition political parties, and their leaders as well as from without through pressure from civil society groups engaging in various forms of civil resistance. It was the outside track engaged in nonviolent resistance which led the regimes to recognize that they could not simply outmaneuver the opposition political leaders while also pressuring the opposition leaders not to compromise. In each case, the prospects of continued economic disruptions from ongoing protests led increasing numbers in the business community to pressure the regime to compromise as did other elite elements concerned with how the ongoing repression and instability was harming the country's image internationally. When negotiations stalled, when promises remained unfulfilled, or when more repression was unleashed, the willingness of activists to again take to the streets, engage in noncooperation, and other tactics of civil resistance kept the pressure on the regime to compromise further.

FACTORS CONTRIBUTING TO SUCCESS
As described above, divisions within the opposition slowed the democratization process, yet having what amounted to a more radical flank ready to mobilize created a strong incentive to move the process forward. In each case, the use of strikes linked economic grievances with the autocratic system that was incapable of addressing them. Each case took advantage of the paradox of repression by using the regimes' heavy-handed response to peaceful protests to gain greater support domestically and internationally. The nonviolent discipline in most of these campaigns and their close ties with respected opposition political leaders made it difficult to depict the protests as the work of a marginal and dangerous malcontents.
It is not within the scope of this study to compare the difference in outcomes between the three countries. However, even in the cases of Brazil and Kenya, which have failed to consolidate democracy as thoroughly as South Korea, the civil society institutions and practices of popular resistance developed in the course of the struggle make it highly unlikely that they will slide back into dictatorship and give hope that these countries' current challenges with establishing a stronger liberal order can be overcome. Indeed, the agency and democratic processes inherent in nonviolent movements helps build the foundations of democratic society more so than guided democratization from above (or armed struggles, centered around a military hierarchy and vanguard mentality) (Zunes, Kurtz, and Asher, 1999).
In certain respects, it could be argued that having to engage in a protracted struggle in winning democratic rights was advantageous because it involved changing the mindset of the people instead of just forcing changes at the top. This is why, despite the return of Moi to power in the Kenya's 1992 elections, the election of a military president in South Korea's 1988 vote, and the coming to office in 1985 of a president who had until recently been allied to the military in Brazil, pro-democracy activists in those countries remained optimistic. Civil society had emerged strong enough to prevent more overt forms of government repression and--despite serious economic problems and political turmoil, particularly in Kenya and Brazil--the democratic institutions which emerged in those countries remain strong.
There were some distinct differences in the cases. For example, while Kenya allowed for multiparty elections before significant liberalization in terms of open dissent and political organizing, South Korea and Brazil had liberalized to some degree prior to allowing direct presidential elections. In terms of numbers, Brazil had the largest mass actions, though-despite weakening the regime's grip on power and eventually forcing the military out-many of those large protests focused on economic issues rather than directly challenging dictatorship. Kenya's movement was smaller, yet focused directly on issues of governance and over time was still able to chip away at the regime's legitimacy and force a democratic opening.
In all three cases, neither the state nor the movement could completely defeat the other. As a result, pro-democracy forces were forced to play a long game, pressing for a democratic transition over time.
Why these movements succeeded deserves further study. However, there are three major variables which appear to have enabled them to succeed: Effective organization and leadership: These movements appear to have been structured enough to provide the necessary leadership, discipline, organization, and strategic coherence to be effective in the face of repression but decentralized enough so that the arrest or killing of key leaders would not cripple the movement. In Kenya, it was respected clergy, lawyers, parliamentarians and other civic leaders who, through a number of ad hoc organizations, eventually formed FORD. In Brazil, it was trade unions and the Catholic Church. In South Korea, it was students, intellectuals, and ultimately trade unionists who coalesced into the National Liaison for Democratization (NLD) and later the Minjing movement.
Discernment and Strategy: In all three countries, there appears to have been an awareness of under what circumstances it would be beneficial to work within the limited legal framework allowed by the state, when and under what circumstances to negotiate, and under what circumstance to engage in open defiance. In Kenya, the opposition agreed to participate in elections despite ongoing limits to civil liberties, with the awareness that legalized political parties would not only be able to participate in elections, but be an organizational base for ongoing resistance. In South Korea, pro-democracy forces were willing to put their resources into support for political parties like the NKDP or into grassroots coalitions such as the NLD, depending on the circumstances. A similar balancing act took place in Brazil between the broader grassroots democratic movement and what became the Party of the Brazilian Democratic Movement. This appears to support the thesis by Ackerman and Kruegler (1994) regarding the importance of developing clear political goals followed by the development of specific campaigns and tactics designed to achieve short-term, medium term, and long-term goals.
Coalition-building: In each country, the pro-democracy resistance demonstrated an ability to build broad alliances among disparate sectors of the population united in a common goal, despite efforts by the government in the transition period to divide and co-opt certain segments of the movement, including a willingness for different segments of the movement to play complementary roles within and outside the established proscribed political order. While the Brazilian opposition maintained a stronger grassroots orientation, all three countries witnessed a close collaboration between opposition political figures and the base of the resistance movement, bringing in churches, workers, students, professionals, and other sectors.
Nonviolent discipline: Much of the success of these movements rested in the ability to maintain a nonviolent discipline in the sectors engaged in active resistance in order to broaden support for the movement, sow divisions within the regime and its security services, limit to some degree the level of regime violence, and deny the regime the excuse to slow or reverse reforms. With the exception the small urban guerrilla movement in the late 1960s and early 1970s in Brazil and the latter phase of the May 1980 Kwangju uprising in South Korea, the opposition struggles in all three countries rejected the use of arms. While rioting and other non-lethal episodes of violence took place in all three countries, the resistance was overwhelmingly nonviolent in orientation.

POLICY IMPLICATIONS
This phenomenon of protracted pro-democracy struggles has several implications for foreign governments, particularly in the advanced industrialized world, in finding a balance between a policy of unconditional support for regimes which continue to violate human rights and refuse to more fully open their political system and a policy of maintaining punitive across-the-board sanctions regardless of reforms, both of which provide little incentive for changing policies. What would be most helpful is some kind of "carrot and stick" strategy: maintaining pressure through carefully targeted sanctions in order to challenge ongoing repressive policies and anti-democratic structures combined with incentives-such as lifting certain sanctions-in return for specific verifiable democratic advances. Since fine tuning such policies to maximum effectiveness from the outside can sometimes be a challenge, it is imperative to solicit the thinking of democratic civil society organizations within the country as to what kind of balance would be most useful to their efforts.
Another potential policy implication for foreign governments is in regard to the question of direct support for pro-democracy groups. This is an area which requires great sensitivity, especially from countries with a history of colonialism and other forms of foreign military intervention, as well as governments which are apply a double-standard through their willingness to support other autocratic regimes which they consider "strategic allies" (Ibrahim and Zunes, 2009). At the same time, there is certainly a strong moral, strategic, and legal case for providing such support (Ackerman and Glennon, 2007). Given the protracted and grassroots nature of these struggles and the tendency of elite oppositionists to place their personal ambitions above the pro-democracy struggle (as witnessed in Kenya and South Korea), such support should be directed more towards grassroots organizations whose agenda is genuinely focused on democracy rather than the fortunes of a particular favored political figure or political party. Indeed, support for what may formally be considered non-political organizations could still be quite beneficial since, by empowering individuals to take positions of leadership and creating alternative institutions from those controlled by the state, they create conditions which advance the process of democratization.
Though far less endowed in terms of resources but with less political complications would be support from global civil society. International nongovernmental organizations with a record of consistent support for democracy regardless of the incumbent regime's geopolitical orientation have more credibility in offering such support. Capacity building organizations which can provide workshops, printed material and online resources on strategic nonviolent action can play a particularly useful role as well. Further research on the history and dynamics of strategic nonviolent action, particularly those of a protracted nature, would obviously be useful as well.