Grievances,Mobilization,and Mass Opposition to Authoritarian Regimes: A Subnational Analysis of East Germany’s 1953 Abbreviated Revolution

Abstract

Mass opposition to authoritarian governments is caused by economic grievances and factors which facilitate mobilization. In this article, I explore these competing explanations of revolution with a county-level analysis of the June 17, 1953, uprising against the socialist dictatorship in East Germany. I argue that grievances can drive unrest, but only when they are disproportionately large and clearly attributable to a regime. Mobilization capacity is the primary driver of unrest outcomes, but depends on group structure and communications networks which are difficult to capture using cross-national indicators. Independent farmers with intense grievances attributable to the East German regime’s agricultural collectivization policies were associated with unrest despite significant obstacles to mobilization. Construction workers with strong mobilization structures and dense communications networks were significant instigators of unrest despite small numbers and moderate grievances. These findings raise important questions for both theoretical and empirical treatments of revolutionary threats to autocratic regimes.

Keywords

democratization and regime change nondemocratic regimes social movements East European politics political economy

Whether on the streets of Vienna in 1848, in East Germany in 1953, or in Kiev in 2013, mass unrest is a real threat to the stability of undemocratic governments. This threat plays a central role in theories of authoritarian politics and democratization. On the one hand, economic grievances and the resulting danger of revolution are argued to have significant effects on the behavior of authoritarian governments, forcing dictators to make concessions in the form of democratizing reforms or redistribution.¹ Unrest has been linked to autocratic policies such as redistribution, food subsidies, and the provision of public services.² On the other hand, an important body of literature argues that the predominant cause of revolutionary threats is not the severity of economic grievances, but the capacity of groups to solve collective action problems and mobilize in opposition to a regime.³ Questions of mobilization and revolutionary collective action have come to occupy a key place in contemporary work on authoritarian politics.⁴ However, scarce data on mobilization capacity, grievances, and protests under authoritarian regimes mean that relatively few studies have been able to explore the competing contributions of both grievance structures and mobilization capacity to unrest.⁵

Here, I fill this gap with a county-level analysis of an historic uprising against the dictatorship of the Socialist Unity Party (SED) in the German Democratic Republic (GDR) on June 17, 1953. The June 17 uprising makes for an informative case study of mass opposition to authoritarian rule, because it appears unlikely to be explained by grievances and mobilization factors (Gerring, 2007). As in many authoritarian regimes, economic hardship was long-standing and ubiquitous in the GDR. Living standards languished well below prewar levels and investment focused on heavy industry rather than agriculture and consumer goods production.⁶ Under such conditions, we would expect it to be very difficult to find evidence that additional grievances among particular groups and related to specific government policies were a significant contributor to the uprising. Furthermore, the repressive socialist regime in East Germany included the nascent Ministry of State Security (Stasi) which suppressed dissent and the ability of the population to agitate against the government, while the presence of Soviet forces in the country had a chilling effect on the prospects for successful mobilization.⁷ In a context where personal networks and political organizations were potentially infiltrated by the Stasi, and the probability of a revolt leading to regime change was low, we would expect it to be very difficult to find evidence supporting the hypothesis that mobilization capacity played a significant role in propagating the unrest of June 17.

Unique data collected from declassified archives allow me to explore the competing contributions of economic grievances and mobilization factors in promoting unrest on June 17. I isolate groups with varying grievance structures and capacities for collective action and estimate how concentrations of these groups were correlated with collective opposition to the regime at the county level. I also examine the effects of bomb damage and housing shortages on unrest, these being proximate causes of economic grievances which varied in their attributability to the SED. Given the limitations of this historical data, my empirical findings are associational, not definitive proof of causality. My measures of grievances and mobilization capacity are indirect, relating the size of groups or levels of destruction and housing shortages to unrest outcomes at the county level. Nonetheless, the main statistical results hold under different model specifications and robustness tests. A careful examination of the quantitative findings in their historical context allows me to draw conclusions on the relative importance of grievances and mobilization capacity in determining unrest outcomes on June 17.

Informed by the experience of East Germany in 1953, I argue that our theoretical understanding of grievances needs to be expanded beyond the sort of broad economic hardship which affected most of the GDR’s population throughout the early 1950s. Such structural grievances are not sufficient to cause the revolutionary overthrow of a regime, even when they are acute. The sorts of grievances which provoke mass unrest are unexpected or incidental, and clearly attributable to the actions of government. I find that economic grievances which were difficult to attribute to the regime were not significantly correlated with unrest on June 17. Independent farmers, on the other hand, were thinly dispersed across the countryside, facing repression and considerable obstacles to collective mobilization. Faced with an increasingly harsh agricultural collectivization policy, and therefore harboring intense grievances directly attributable the regime, they were nonetheless significantly associated with unrest and opposition to the SED.

The sorts of intense, directly attributable grievances present among independent farmers in the GDR should be expected to be relatively rare under authoritarian regimes seeking to avoid instability. When grievances do not fulfill these conditions, unrest outcomes are primarily influenced by groups’ capacity to organize, reveal their preferences for regime change, and mobilize collectively in opposition to the government. This mobilization capacity depends on group-level decision-making structures which facilitate collective action by increasing the probability of a successful revolt and spreading the potential costs of repression across a broad membership. Relatively small numbers of individuals can play a crucial role in propagating unrest if they are mobilized in this way and have the technology to communicate their intention to rebel. Workers’ meetings on construction sites in East Berlin organized the first mass demonstrations against the regime in 1953. Construction workers were also relatively mobile and had dense communication networks across the country. They played a central role in propagating news of the Berlin demonstrations and calling for further mobilization among their colleagues. Despite their relatively low numbers, there is a robust correlation between concentrations of construction workers and unrest on June 17.

These findings are an important contribution to our understanding of mass opposition to authoritarian governments. They suggest that economic grievances can be a cause of opposition, despite repression and barriers to mobilization, when they are severe and clearly attributable to a regime. Given regimes’ efforts to prevent such grievances emerging, mobilization capacity should be expected to be the proximate cause of mass threats to undemocratic governments. However, my analysis shows that mobilization capacity depends on intragroup organizational structures and communications networks, suggesting that such tools of collective action should play a greater role in accounts of revolutionary threats to authoritarian regimes. This article also contributes to a growing body of research in political science and history which explores the East German dictatorship and its lessons for the study of authoritarian politics.⁸ Previous studies have posited many contributing factors to the uprising of June 1953, but this is the first quantitative analysis which allows their contributions to be systematically assessed and compared across the entire country. Moreover, my findings will be of interest to political scientists and historians who are analyzing postwar Europe with new interest. The origins of the Cold War political order on the continent are increasingly being sought in this initial phase of institution-building on both sides of the iron curtain.⁹ Revealing which economic and social factors led to the uprising in the GDR holds lessons for explaining the wave of unrest which spread across East-Central Europe in the mid-1950s and had profound consequences for the broader trajectory of regimes in the region (Ekiert, 1996).

Grievances, Mobilization Capacity, and Opposition to Authoritarian Rule

Economic grievances have long been considered a driver of political instability and revolutionary threats to autocratic governments. Gurr (1968) famously argued that individuals react with anger to discrepancies between their desired and achieved levels of wealth. In a similar vein, the seminal theories of authoritarian rule and democratization by Boix (2003) and Acemoglu and Robinson (2006) state that poor citizens harbor grievances against the rich and mobilize in revolution to redistribute wealth under an alternative regime. Empirically, cross-national studies have shown a link between structural measures of grievances such as poverty, inequality, or recessions and different kinds of civil strife, from urban unrest to civil war.¹⁰ However, the role of grievances in provoking rebellion and collective action is hotly contested. Canonical accounts dispute their role in the onset of civil conflict and revolutionary collective action, arguing instead that opportunities for rebellion and dynamics of mobilization have a decisive impact on the process of contention.¹¹ Particularly important in the context of repressive authoritarian regimes are social networks and organizations which facilitate preference revelation among individuals. The costs of collective opposition to any one individual are prohibitively high when they are unsure of others’ preferences and actions, and the likelihood of a successful uprising therefore appears low.¹² This is one reason why authoritarian regimes act decisively to suppress public expressions of willingness to protest or encouragement to mobilize.¹³ However, even relatively small increases in the observable size of the opposition can create band-wagon effects and lead to large increases in the size of a revolutionary movement (Kuran, 1989). These dynamics of diffusion and escalation of protest through preference revelation and band wagoning, whether via personal networks or demonstration effects, made a decisive contribution to the successful East German revolution of 1989, for example.¹⁴

I argue that the conception of grievances in the current literature on authoritarianism and democratization is too narrow, and must be broadened to account for grievance type and attribution to capture revolutionary threats. Authors who aim to conceptualize or measure mass threats to a regime rely predominantly on structural grievances, that is, those derived from an individual or group’s disadvantaged position in society.¹⁵ Thus, Boix (2003), Acemoglu and Robinson (2006), and Svolik (2012a, 2012b) argue that economic inequality captures the threat of mass opposition to authoritarian regimes, while Gandhi and Przeworski (2006) suggest further measures such as religious fractionalization and Kim and Gandhi (2010) use the proportion of manufacturing workers in the total labor force. However, structural grievances tend to be static or slow moving, conform with the expectations of citizens, and are relatively unlikely to cause rebellion. More important for conditioning the likelihood of unrest are incidental grievances, or those which are unexpected, violate some established norm of acceptability, and enhance a group’s capacity to mobilize. Moreover, grievances need to be disaggregated into those which are attributable to either government action or neglect and those which are caused by exogenous factors for which the regime cannot be blamed. Clear attribution of blame for deprivation to an authoritarian regime is likely to provoke unrest, because it provides a concrete target for mobilization and increases the likely payoff to collective action against the government. Javeline (2002), for example, shows that Russian citizens who attributed blame for wage arrears to the government were significantly more likely to protest in the 1990s. Thomson (2017b) shows that government food policy makes an independent contribution to urban unrest over and above that of market price fluctuations. De Juan and Wegner (2017) show that grievances around state service provision provoke protest in South Africa. As the GDR case at hand illustrates, structural grievances experienced by East Germans during the entire postwar period were not sufficient to cause unrest on their own. I will show that only disproportionately large, escalating grievances which were clearly attributable to the government caused collective opposition to the SED on June 17.

Hypothesis 1 (H1): Groups harboring severe, attributable economic grievances will be associated with greater unrest.

Economic grievances’ contribution to an unrest event must be considered alongside factors which solve collective action problems and facilitate mobilization. Individuals are unlikely to protest against an authoritarian regime, holding all else equal, because the potential costs of repression commonly outweigh any prospective benefits to opposition. However, when large numbers of individuals take part in collective action the regime might be toppled, make concessions to protesters, or at least spread the costs of repression over the entire group. This makes mechanisms which generate a shared belief in successful rebellion crucial for facilitating collective action under autocracy. Kuran (1989) and Bueno de Mesquita (2010) argue that relatively small numbers of rebels can embolden a much larger group to mobilize against a regime. Other scholars focus on sources of information, such as social networks, government transparency, or technology, which allow for the formation of shared expectations of protest.¹⁶ Social movement theory takes a broader view of mobilization resources which includes intragroup structures facilitating collective decision-making (Tilly, 1978). In what follows, I will show how construction workers, a relatively small group within the GDR, took the decision to oppose the SED in Berlin and subsequently played a decisive role in propagating unrest across the country. I will emphasize two factors, intragroup decision-making structures and communications networks. Workers’ meetings which allowed employees on large building sites in Berlin to coordinate in strikes and opposition to the regime played a vital role in initiating unrest on June 16. Construction workers were mobile and had communications links which then allowed them to spread word of their strike and propagate mobilization across the entire country the next day, so they played a central role in the uprising across the GDR.

Hypothesis 2 (H2): Groups with strong mechanisms promoting collective mobilization will be associated with greater unrest.

Grievances and Mobilization on June 17

On June 17, 1953, in the context of general economic hardship, growing repression, and regional political uncertainty following Stalin’s death, hundreds of thousands of citizens of the GDR mobilized in an attempt to overthrow the ruling party.¹⁷ The immediate cause of the uprising was a change in government policy in late May which increased output quotas for many workers and exacerbated already grave levels of economic grievances across the country. Labor unrest occurred sporadically before June 17, but organized mobilization against the regime happened first on the 16th among construction workers in East Berlin. Employed in large collectives which had strong hierarchical structures of worker representation, these builders decided to refuse the SED’s demands for increased output without increases in pay. They marched through the city to central government buildings demanding reform and, eventually, regime change (Wolle, 2013b, pp. 254-263). As word spread of this protest in the capital, it prompted a wave of unrest which broke out in the early hours of the 17th on construction sites and in industrial workplaces across the country. Joining workers, citizens in both urban and rural areas engaged in strikes, protests, and demonstrations with the explicit goal of removing the SED from power. In many locations, citizens stormed party buildings, prisons, and even offices of the secret police. Out of control of the East German authorities, the uprising was put down by Soviet army, which declared a state of emergency in almost all areas by midday, using troops and armored vehicles to break up demonstrations and occupy striking workplaces. By the evening protests in most areas of East Germany were extinguished, though a significant number of strikes and demonstrations continued for several days thereafter.

In this section, I describe occupational groups with clear grievance structures and mobilization capacities on June 17. I predict concentrations of these groups will be correlated with unrest. I also outline further indicators of economic grievances at the local level. I associate each group or grievance indicator either with a greater-than-average (upward arrow) or lower-than-average (downward arrow) level of economic grievance, attribution of grievance, or mobilization capacity as depicted in Table 1. For each of these variables, I test the null hypothesis (HX) that concentrations of the group or greater levels of the grievance is not associated with greater unrest in a county on June 17.

Table 1.

Grievances and Mobilization on June 17, Key Independent Variables.

	Economic grievance		Mobilization capacity	HX (unrest)
	Level	Attributability	Level	HX (unrest)
Group
Independent farmers	∧	∧	∨	$H 1.1$
Construction workers			∧	H2.1
Blue-collar workers			∧	H2.2
Economic condition
Living space	∧	∨		H1.2
Bomb destruction	∧	∨		H1.3

I will first describe the variables capturing grievances to test H1. Independent farmers suffered from acute economic hardship which was easily attributable to SED policy in the early 1950s. Despite structural barriers to mobilization due to their position in the countryside, they played a key role in the 1953 uprising. Most economic grievances in East Germany were attributable to government policy in the early 1950s. Unlike in West Germany, the recovery from the war was slow, and the SED placed burdensome demands on workers in the form of output quotas and strict food rationing.¹⁸ However, grievances could also be deflected by the government due to its youth, as the GDR had only existed for 4 years; the lasting effects of the war, which could not be blamed on the SED and its policies; reparations demanded by the Soviet Union; and efforts to construct a socialist economy which were viewed by at least some East German citizens as legitimate. Such heterogeneity in the attribution of blame for economic outcomes makes the use of any general measures of output as indicators of attributable grievances problematic.¹⁹

However, the SED’s agricultural policies stand out as unreasonably burdensome, unpredictable, and clearly attributable to the party and its agents. These policies had proceeded unevenly in several stages, each involving escalating levels of economic hardship and repression (Schöne, 2005). In the direct aftermath of the war, large farms and property belonging to Nazis were expropriated and distributed to owners of smaller holdings, workers, and party members. Subsequently, the SED pursued a policy by which farmers would combine their land, buildings, equipment, and livestock into agricultural collectives. Farmers who did not join were pressured to collectivize by draconian measures. These included being forced to sell produce at artificially low prices to the state, and being forced into debt or bankruptcy if they could not fulfill these compulsory supply quotas. Finally, in 1953, the SED had recently begun a still more radical strategy of collectivization involving overt repression, as party functionaries were encouraged to bully independent farmers into joining collectives, for example, by leading demonstrations to their homes and harassing them. Each stage of the collectivization process involved coercion and met with considerable resistance, not only from large farmers facing expropriation but also from others who were skeptical of the SED’s land reform policies, their real goals, and their prospects for success. Many farmers reacted with hostility to the prospect of combining their operations with neighbors, and they resented the impositions of party agitators demanding they join the collectives. (Port, 2007, pp. 87-89).

Although rural populations are notoriously difficult to mobilize in opposition to economic grievances or authoritarian governments (Scott, 1985), independent farmers played a key role in the unrest of June 17, giving it a truly nationwide character rather than being restricted only to cities. Opposition to collectivization was strong in the countryside, and by mid-1953, the situation in the villages was explosive. Unrest in rural areas had already begun in the weeks before June 17, and typically took the form of intimidation and attacks against collective farmers and party representatives by independent farmers. The nationwide events of June 17 provided the rural population with the opportunity to more forcefully voice their opposition to the SED and its agricultural policies. In more than half of rural municipalities, demonstrations, marches, strikes, and attacks on SED representatives took place.²⁰ Indeed, in many towns and smaller regional cities, independent farmers from surrounding areas played a key role in starting and organizing demonstrations and protests. The role played by independent farmers in the events was so important that in its aftermath, the Stasi concluded that “the Fascist putsch attempt on 17 June, 1953 showed that the class enemy has concentrated his strength in the countryside” (Schöne, 2005, p. 148). Because the SED’s agricultural policies made independent farmers a particularly disadvantaged group within the GDR, who directly blamed the regime for their grievances, I test the hypothesis that concentrations of independent farmers were correlated with unrest on June 17:

Hypothesis 1.1 (H1.1): Counties with greater concentrations of independent farmers experienced greater unrest.

Housing shortages and wartime bomb destruction were further causes of economic grievances in postwar East Germany. Unlike the condition of independent farmers struggling against collectivization, however, neither of these grievances could be clearly attributed to the SED. Housing shortages were one of the most prominent components of economic grievances in postwar Germany, where little had been invested in civilian accommodation before 1945 because the war effort took priority over all other economic activity (Port, 2007, pp. 33-38). For many years after the end of hostilities, bomb damage to cities’ housing stocks caused very significant hardship for the East German population, as families in towns and cities lived without essential infrastructure among the rubble and struggled to find suitable accommodation when the government failed to build sufficient replacement housing. Families were forced to crowd into apartments with strangers, live in basements of bombed buildings, and many had to move to the outskirts of damaged city centers. Factory workers were in many cases housed in cramped barracks near their workplaces, in primitive and unsanitary conditions.²¹ Housing shortages in the early years of the GDR were primarily attributable to factors exogenous to SED policy, and were presented as such by the government. The GDR was only founded in 1949, and thereafter rebuilding efforts were severely hampered by scarcity and focused on the construction of entirely new suburbs and towns centered around industrial plants, rather than the reconstruction of city centers where the majority of destruction had taken place. Therefore, in 1953, the regime was only 4 years old and deflected blame for grievances over housing shortages “by reminding housing applicants that the current situation was an unfortunate remnant of capitalism and the war” (Port, 2007, pp. 257-270). It could point also to investment in brand new housing stock outside city centers, although areas destroyed during the war were not yet rebuilt.²²

Bomb damage was not only nonattributable to government policy, it was also randomly distributed across cities, weakly correlated with urbanization and other socioeconomic variables, and not caused by a third factor which also determined levels of agricultural collectivization.²³ Although it was concentrated in West Germany, towns in what became the GDR also experienced significant damage. In total, around 400,000 Germans lost their lives due to air raids during the war and seven million lost their homes.²⁴ The large cities of East Germany, including Berlin, were almost all damaged by air raids. However, the degree of destruction caused by raids was determined by the accuracy and efficacy of attacks. These were in turn dependent on weather conditions, bomb loadings and performance, and city geography. Some smaller towns had very high levels of destruction, such as Zerbst, a town of only around 40,000 which had 56% of its housing stock destroyed. Nearby Potsdam, on the other hand, with a population three times larger and undoubtedly a more valuable strategic target for bombing, suffered only 20% destruction. The highest levels of destruction seen in German cities were caused by a small number of extremely destructive raids, which were contingent on a coincidence of random factors. Dresden, for example, was famously struck by an incendiary bomb attack close to the conclusion of the war. Sixty percent of the city’s housing stock was permanently destroyed, predominantly in that single raid, while nearby large cities such as Leipzig and Halle suffered much lower levels of destruction, at 25% and 5% of their housing stocks, respectively.

I argue that housing shortages and bomb damage are indicators of economic grievances which were only weakly attributable to East German government policy. I test the hypothesis that they were positively correlated with unrest on June 17, 1953:

Hypothesis 1.2 (H1.2): Counties with greater housing shortages experienced greater unrest.

Hypothesis 1.3 (H1.3): Counties with greater levels of housing destroyed by bombing during World War Two experienced greater unrest.

I now move on to discuss indicators of mobilization capacity on June 17. Construction workers played a central role in the uprising against the SED due to their unique ability to mobilize and spread news of their protest against the government. Although their output quotas had been increased in late May, these measures affected swathes of the economy and builders did not harbor greater economic grievances than other citizens or more clearly attribute their grievances to the SED. However, by organizing in independent workers’ meetings on large building sites in East Berlin, thousands of construction workers were able to collectively mobilize in a strike on the 16th. After the SED announced its policy of escalating austerity in May, discontent on construction sites in the capital increased. On June 14 and 15, organized meetings of hundreds of laborers were held on sites in East Berlin to discuss their grievances, prepare demands for delivery to the government, and agree to strike.²⁵ Importantly, these meetings were not organized or controlled by the state-sponsored workers’ unions, which actively suppressed mobilization, but by autonomous groups formed at each workplace. Although members of unions were present at these meetings, high-level functionaries were not, and discontent at government policy could not be silenced or adequately addressed before strikes were organized.²⁶ Able to decide collectively to strike and to coordinate across many construction sites, thousands of laborers marched through East Berlin to the government headquarters in the center of the city on 16 June and, in a heated exchange with high-ranking politicians, demanded that the SED allow for free elections and the reunification of the GDR with western Germany. These demands were rejected, and workers decided on a national strike and protest for the following day. Nationwide, construction workers had greater mobilization capacity than other citizens, because they had access to a communications network which allowed them to disseminate news of their strike, facilitating preference revelation and propagating unrest. Construction workers in the GDR were on average younger, more mobile, and difficult to control than workers in other sectors, who were usually tied to a single workplace. They were often part of large state-run cooperatives which had strong national ties as members were regularly sent to work on sites across the country.²⁷ This facilitated unrest on June 17, as construction workers who had been in Berlin heard of rising discontent on construction sites there, spread the news, and encouraged their colleagues elsewhere to engage in strikes and protests when these took place in the capital on the 16th. One example of this occurred in Cottbus, where a construction worker in a coking plant had witnessed the Berlin strike on June 16 and on his return on the 17th helped instigate a strike and demonstration which eventually needed to be put down by Soviet tanks (Diedrich & Hertle, 2003, p. 97). Cooperatives also had access to a nationwide telegraph network via which they spread the news of strikes by their colleagues in East Berlin and other cities, mobilizing construction workers across the GDR in opposition to the regime. A group of construction workers in Frankfurt on the Oder, for example, sent a telegram to their colleagues in East Berlin on June 18 expressing their solidarity and demanding that the government step down; if they did not receive a reply, they intended to strike (Diedrich & Hertle, 2003, p. 210). Compared to other East Germans who heard of the protests spreading across the country either through word of mouth or—if they were close to the capital—from reports on West Berlin radio stations, construction workers were well informed of the events of June 17 and were able to more effectively coordinate in opposition to the regime. I therefore hypothesize,

Hypothesis 2.1 (H2.1): Counties with greater concentrations of construction workers experienced greater unrest.

Blue-collar workers also did not have greater economic grievances than the rest of GDR population, but were concentrated in large workplaces which had the potential to facilitate mobilization. Following the Soviet development model of the time, labor and investment were directed overwhelmingly toward heavy industries such as steel, chemicals, and energy (Weber, 2012, pp. 198-199). Workers were prominent participants in the events of June 17, suggesting that class structure was a key determinant of mobilization and unrest outcomes. The size of the working class is commonly argued to facilitate collective action under authoritarianism (Kim & Gandhi, 2010) and a class-based explanation of the June 17 uprising resonates with conclusions drawn by West German historians who saw it as a “workers’ uprising.”²⁸ On the other hand, workers were organized in the corporatist Alliance of Free Unions (FDGB) which was created by the regime to help it prevent labor unrest, maximize worker output, and provide information on discontent among the working class. For example, in the late 1940s, the FDBG was expected not to help articulate workers demands for higher wages and living standards to the regime, but to convince workers that the government’s Stalinist austerity policies were to be followed and applauded (Hübner, 1995, pp. 36-39). Furthermore, the increasing presence of the Stasi in workplaces and the FDGB’s lack of engagement for workers’ rights undermined the positive effects of worker concentration and organization on mobilization. I therefore test the following hypothesis,

Hypothesis 2.2 (H2.2): Counties with a greater concentrations of blue-collar workers experienced greater unrest.

Data

I collected unrest data for 212 counties (Kreise) of the GDR, excluding East Berlin, from a volume by the German historians Diedrich and Hertle (2003). The volume is a collection of responses to a questionnaire sent to every district police administration by the national police headquarters on the evening of June 21. The questions which were of primary importance for the construction of the data set were “When and where did provocations begin in the district?” and “Where were the focal points [of the uprising] in the district?”²⁹ Each of the 14 district reports is reproduced in full, with explanatory and biographical notes.

In most areas, a pattern of escalation is described in the reports, with unrest beginning on the morning of June 17 as stop-work meetings within larger workplaces and growing into strikes, demonstrations, and in some cases attacks on state-run institutions such as party offices or prisons. Where unrest reached this latter stage and the East German police were unable to control the situation, Soviet troops were mobilized and in some cases shots were fired and civilians were killed. I confine my analysis to the events of June 17 only, and initially coded unrest following the dynamic of escalation outlined above, on a scale from 0, indicating complete calm, to 9, indicating at least one civilian death in a county:

0. Calm, with little or no disruption;

1. Assemblies or meetings within workplaces;

2. Strikes within workplaces;

3. Demonstrations which move outside workplaces and into a public area;

4. Looting or storming of a building associated with the state, such as a supermarket;

5. Storming of a party building;

6. Storming of a police station, Ministry of State Security building, or prison;

7. Mobilization of Soviet troops;

8. Shooting, including warning shots;

9. Deaths of civilians.

The dependent variables used in the following empirical analysis were constructed from the above coding:

Unrest (0/1) is a binary variable coded 1 if a strike or greater level of unrest occurred in a county; otherwise it is coded 0.

Mobilization is a four-value ordinal variable which measures the mobilization of a county’s population to open opposition to the regime. It is coded from 0-3 according to the scale above, with values greater than 3 coded as 3.

Unrest measures mobilization, normative protest against the regime, and repression and is the full 10-value variable coded according to the scale above.

The modal East German county was calm on June 17 (92 counties), while the mean level of Unrest across all East German counties was 1.82, a strike. Values exceeding those captured by the four-value ordinal Mobilization variable make up only around 10% of observations. Soviet troops were mobilized against protesters in 10 counties, shooting also took place in 10 counties, while protesters were killed in only three counties. The distribution of Unrest, along with independent farmers, bomb damage, and construction workers is depicted in the maps in Figure 1. Summary statistics of all variables and correlations between key variables are given in Tables 4 and 5 in the Online Appendix.

Figure 1.

Maps of unrest, independent farmers, bomb damage, and construction workers data.

Data measuring the socioeconomic characteristics of East German counties were collected from a detailed 1950 demographic, occupational, and housing census. The East German regime did not allow its statistical office to publish these reports and they have never been analyzed by researchers. After 1989, they were transferred to the federal archives in Berlin, where I collected them.³⁰ The key explanatory variables, as related to the hypotheses outlined above, are as follows:

H1.1: Ind Farmers. The proportion of independent farmers in total employment;

H1.2: Living Space. Square meters of living space per person;

H2.1: Construction. The proportion of construction workers in total employment;

H2.2: B.C. Workers. The proportion of blue-collar workers in total employment;

To test H1.3, I use the percentage of a town’s housing stock destroyed by bombing during World War Two. The variable Bomb Dest is collected from data published by the geographic historian Uta Hohn (1991), which covers all German towns with a population of 20,000 or more in 1939 and smaller towns which suffered considerable bombing damage. These data give observations for 72 of the GDR’s 214 total counties, and in my main models I assume that rural counties without a large town and excluded from Hohn’s data experienced no bomb damage during the war, giving them a value of zero on this variable.³¹ The average town or city in the original data set experienced destruction of 12.5% of its housing stock during the war, with Greifswald experiencing only 1.2% destruction and Dresden 60%. Across all 212 counties included in my analysis, the average level of destruction was 4%.

All models also include several control variables: Ln Pop Density, the natural log of the population per square kilometer; Pop 15-25, to control for Urdal’s (2006) finding that youth bulges cause rioting and civil unrest; the binary variable W Berlin Border for all counties bordering on West Berlin, to control for the fact that the protesters around the city were claimed by the regime to be either West Berliners, Western agents, or encouraged by Radio in the American Sector; the Border Dist indicator for all counties bordering West Germany; and the binary variable E Berlin Border to control for the fact that the uprising began in the eastern part of the capital.

Estimation and Results

I estimate logistic and spatial regression models to test the hypotheses laid out above. The spatial models include a lag $W y$ which is a weighted sum of unrest events in neighboring counties, normalized to range from 0 to 1.³² The average extent of the clustering effects in the model is measured by the spatial-autoregressive parameter λ. Results are presented in three separate tables by model type for ease of presentation and comparison. Table 2 presents the results of seven logistic regression models which take the binary Unrest (0/1) as dependent variable. These are graphed in Figure 2. Table 3 presents five models which take the 10-value Unrest as dependent variable and are estimated using linear spatial models. These are graphed in Figure 4. I present one more set of results in the Online Appendix. Table 6 and Figure 5 in Online Appendix show the results of four ordered logistic regression models which take the four-value Mobilization as dependent variable.

Table 2.

Results of Logistic Unrest Models.

	(2.1)	(2.2)	(2.3)	(2.4)	(2.5)	(2.6)	(2.7)
	Farmers	Constr	Liv Sp	Bomb Dest	BD (cities)	Working Cl	Unemp
Ind Farmers	18.1** (8.33)	23.2** (9.14)	30.0** (13.0)	22.7** (9.18)	22.6 (20.6)	27.3** (12.6)	22.7** (9.41)
Constr		44.2*** (13.7)	42.8** (17.1)	39.8*** (14.0)	6.44 (30.7)	45.0*** (14.0)	44.1*** (14.5)
Living Space			0.33 (0.31)
Bomb Dest				2.14 (1.98)
Bomb Dest					3.39 (2.55)
B.C. Workers						2.01 (2.85)
Unemp (M)							−7.74 (12.8)
Ag Workers							0.56 (3.05)
Ln Pop Dens	1.17*** (0.39)	1.30*** (0.42)	1.59*** (0.55)	1.26*** (0.43)	2.29*** (0.83)	1.32*** (0.44)	1.27*** (0.41)
Pop 15-25	−25.8*** (9.59)	−31.6** (12.6)	−11.9 (17.1)	−28.6** (12.8)	−15.5 (15.0)	−34.5** (14.1)	−34.6** (14.6)
Border Dist	−2.09*** (0.77)	−2.06*** (0.77)	−1.80** (0.79)	−2.21*** (0.74)	−0.88 (0.93)	−2.03*** (0.77)	−2.07*** (0.78)
W Berlin Border	1.96* (1.13)	0.92 (1.12)	0.53 (1.26)	1.04 (1.11)		0.96 (1.11)	0.95 (1.12)
Constant	−3.13 (2.56)	−5.07* (2.68)	−13.6** (5.88)	−5.20** (2.65)	−11.6** (5.27)	−5.77* (3.13)	−3.79 (3.36)
Observations	211	211	135	211	72	211	210
AIC	256.5	247.4	158.3	247.8	84.2	248.9	250.8

Standard errors in parentheses. AIC = Akaike information criterion.

p < .10. **p < .05. ***p < .01.

Figure 2.

Results of logit unrest models with 95% confidence intervals.

Table 3.

Results of Spatial Unrest Models.

	(3.1)	(3.2)	(3.3)	(3.4)	(3.5)
	Farmers	Constr	Liv Sp	Bomb Dest	Working Cl
Ind Farmers	24.1*** (6.03)	25.0*** (5.94)	30.3*** (9.58)	23.3*** (5.95)	24.2*** (8.34)
Constr		38.3*** (12.3)	31.8* (18.4)	30.1** (12.8)	37.9*** (12.6)
Living Space			0.37 (0.28)
Bomb Dest				3.16** (1.44)
B.C. Workers					−0.35 (2.56)
Ln Pop Dens	1.55*** (0.29)	1.50*** (0.29)	1.58*** (0.40)	1.36*** (0.29)	1.50*** (0.29)
Pop 15-25	−14.7* (7.67)	−15.7** (7.56)	−2.77 (15.3)	−11.5 (7.77)	−15.0* (9.03)
Border Dist	−0.70 (0.49)	−0.72 (0.47)	−1.39** (0.65)	−0.80* (0.46)	−0.71 (0.47)
W Berlin Border	0.66 (0.85)	0.053 (0.86)	−1.68 (1.37)	0.24 (0.86)	0.047 (0.85)
Constant	−6.26*** (2.36)	−7.43*** (2.31)	−12.6** (5.27)	−7.03*** (2.30)	−7.34*** (2.41)
E Berlin Border	1.92** (0.95)	1.16 (0.97)	2.60* (1.49)	1.15 (0.96)	1.15 (0.97)
$λ$	1.12*** (0.28)	0.97*** (0.27)	−0.041 (0.49)	0.95*** (0.27)	0.98*** (0.27)
Observations	210	210	134	210	210

Standard errors in parentheses.

p < .10. **p < .05. ***p < .01.

I first address the three hypotheses on the relationship between economic grievances and unrest. Testing H1.1, I find robust evidence that concentrations of independent farmers were associated with unrest on June 17. Because unrest was predominantly located in East German cities, the simple correlation between independent farmers and unrest was slightly negative, $r = - . 14$ . However, as the scatter-plot in the upper-left panel of Figure 3 shows, several counties with moderate to high shares of independent farmers experienced very intense unrest on June 17. In addition, as the lower-right panel shows, the correlation between population density and unrest is not perfect and many rural areas were very restive. It was the presence of independent farmers which explains this lack of a perfect correlation and the outbreak of unrest in the countryside. In Model 2.1, Table 2, I estimate a logistic regression model taking the binary Unrest (0/1) indicator as dependent variable, including independent farmers while controlling for population density. This model shows a significant positive relationship between the Ind Farmers variable and unrest. In fact, in all the binary logistic models presented in Table 2, the coefficient on the Ind Farmers variable is positive and significant at the $p < . 05$ level, except Model 2.5 which includes only cities and therefore misses the impact of independent farmers on unrest in the countryside. I graph the substantive significance of the results of Model 2.1 in the upper-left panel of Figure 2. The share of independent farmers in total employment varied from zero to a maximum of 0.17. Across the range of the Ind Farmers variable, and holding all else constant, the probability of a strike or more significant unrest increased by more than 40 percentage points, from .23 to .69. An increase in the variable from one standard deviation below the mean to one standard deviation above is associated with an increase in the likelihood of unrest from 0.27 to 0.51. This corresponds to the difference in the number of independent farmers observed between a municipal county like Halle or Zwickau and a rural county like Bad Freienwalde, north-east of Berlin. Considering that unrest was predominantly concentrated in urban areas, the ability of independent farmers to cause demonstrations in rural counties like Bad Freienwalde is remarkable.

Figure 3.

Scatterplots of independent variables and continuous unrest indicator.

In Table 3, I present the results of six spatial regression models taking the continuous Unrest indicator as dependent variable. These models also confirm the positive relationship between independent farmers and unrest. Model 3.1 includes Ind Farmers alongside a control for population density, and the coefficient on Ind Farmers is positive and significant at the $p < . 01$ level. This result is replicated across all further models in Table 3. I graph the results of Model 3.1 in the upper-left panel of Figure 4. Increases in the share of independent farmers in county employment from 2% to 16% are associated with an increase in the predicted level of unrest from complete calm to the mobilization of Soviet troops. Similarly, in spatial models reported in Table 7 in the Online Appendix, independent farmers are significantly correlated with the four-value Mobilization variable, and an identical increase in Ind Farmers is associated with a shift from complete calm to open resistance to the East German regime. Ordered logistic models in Table 6 in Online Appendix, also taking Mobilization as the dependent variable, show a positive relationship between independent farmers and unrest, although only at the $p < . 10$ significance level. The probability that a county was calm on June 17 declined from .60 to .35, and the probability of a strike increased from .17 to .32, over a two-standard-deviation increase in the Ind Farmers variable, as I show in Figure 5 in the Online Appendix. In contrast to independent farmers who were under significant pressure to collectivize, rural residents already in cooperatives, or “agricultural workers” to use the regime’s jargon, were not associated with unrest. In Model 2.7, I include the share of agricultural workers in total employment. The coefficient on this variable is positive, but small and has a very large standard error, while the coefficient on the Ind Farmers variable remains positive and highly significant.

Figure 4.

Results of spatial unrest models with 95% confidence intervals.

Moving on to the effects of housing shortages and bomb damage, I find no support for H1.2 or H1.3, that these variables were associated with unrest on June 17. As shown in the upper-right panel of Figure 3, the correlation between living space per capita and unrest is positive. This suggests, counterintuitively, that greater provision of living space was correlated with greater unrest on June 17. Housing shortages therefore would not appear to be an indicator of economic grievances causing unrest. However, there is a lot of noise in this relationship, with very different levels of unrest occurring around the middle of this variable’s distribution. The positive association between living space and unrest can be partly explained by living space per capita being strongly positively correlated with population density, and cities being particularly restive.³³ However, in the binary and ordered logistic regression models controlling for population density, the coefficient on Living Space remains positive and estimated with a lot of uncertainty. Similarly, the spatial Model 3.3 in Table 3 does not find a significant correlation between living space and unrest.

There are several potential reasons why living space is such a poor predictor of unrest. Average levels of living space per capita were extremely low in the 1950s GDR, and varied little among urban areas. Space per capita ranged from 7.1 square meters in rural counties like Grimmen in the north of the country to over 11 square meters in larger cities such as Leipzig, Jena, and Potsdam. Most urban areas had around 10 square meters of living space per capita. By comparison, in the early 1990s only developing countries had cities with less than 14 square meters of living space per person.³⁴ The universally dire housing situation in East German towns and cities did not vary enough across counties to produce significant differences in real living standards. In addition, the data on living space are from a census in 1950, and housing provision could have changed much faster than patterns of employment and bomb destruction between 1950 and June 17, 1953. Assessing another potential indicator of economic grievances, I find that counties characterized by greater unemployment also did not experience more unrest on June 17. In Model 2.7, Table 2, I include the unemployment rate among males, which ranged from almost zero to 12.5% across the GDR. The coefficient on the Unemp (M) variable is negative in this model, but has a very large standard error and is statistically indistinguishable from zero.

I find evidence of a weak positive association between bomb damage and unrest which is more supportive of H1.3. In both binary and ordinal logistic regression models, the Housing Dest variable is positively associated with unrest. However, in neither model does the coefficient approach statistical significance. As illustrated in the middle-left panel of Figure 2, an increase across the range of the Housing Dest variable is associated with a .25 increase in the probability of unrest in Model 2.4, but the standard errors around these point predictions are very large and overlap. There are, however, relatively few counties with a larger degree of bomb damage and many observations with zero damage. Therefore, in Model 2.5, I estimate a binary model of unrest including only the 72 observations where some bomb destruction was observed. In this model, the coefficient on the bomb destruction variable is larger and comes closer to statistical significance ( $p < . 18$ versus $p < . 28$ in Model 1.4) but from these logistic regression models I cannot reject the null hypothesis that there is no relationship between bomb destruction and unrest on June 17.

In the linear spatial regression Model Table 3.4 in Table 3, which takes the 10-value Unrest as dependent variable, I do find a positive relationship between bomb destruction and unrest which is significant at the $p < . 05$ level. The results of this model are graphed in the center-left panel of Figure 4 and show that an increase across the range of the Housing Dest variable from 0 to 0.6 is associated with an increase in the predicted level of unrest from 1.6, a strike, to 5.4, the storming of a party building. Taking these results, the small town of Zerbst, which lost 56% of its housing stock to air raids during the war, is predicted to experience significant unrest including the storming of state-run stores or party buildings, while nearby Potsdam which suffered only 20% destruction is predicted to experience only a peaceful demonstration. Unlike the logistic regression models, the spatial regressions take into account the full scale of escalation of the June 17 insurrection, as they measure violence and repression as the highest levels of the Unrest variable. This seems to be driving the significance of these results versus the logistic regressions. A spatial model taking the four-value Mobilization as dependent variable, presented in Table 7 in the Online Appendix, gives results more similar to the logistic regressions, with a positive but insignificant coefficient on the Housing Dest variable. Bomb destruction should therefore be seen as a weak correlate of unrest with, at best, some explanatory power but certainly less than concentrations of construction workers or independent farmers.

Turning now to hypotheses addressing mobilization factors, I test H2.1, that concentrations of construction workers were associated with unrest. I find robust evidence of a positive relationship between employment in the sector and unrest on June 17. As illustrated in the middle-right panel of Figure 3, the simple correlation between construction workers and unrest is positive and moderately large, at $r = . 26$ . This positive relationship is borne out in models controlling for confounding factors. Across all binary and ordered logistic regression models in Table 2 and Table 6 in the Online Appendix, the coefficient on the Constr variable is positive and significant at the $p < . 01$ level, except Model 2.5 which excludes areas without bomb destruction and therefore removes a lot of variation in the Constr variable. In the middle-right panel of Figure 2, I graph the substantive effects of employment in construction on unrest, from the binary Model 2.2 in Table 2. Across the range of the variable, holding all else constant, the likelihood of a county experiencing a strike or greater level of unrest increases by 0.61 from 0.16 to 0.77. Construction workers were able to propagate unrest despite making up a very small proportion of total employment, supporting my hypothesis that they acted as the leaders of the June 17 insurrection after hearing of the strikes in East Berlin on June 16. The mean level of employment in construction was only 0.045 across all counties, with the smallest numbers of construction workers found in remote rural areas such as on the border to Czechoslovakia and the greatest number in large cities, particularly those close to Berlin. At one standard deviation below the mean of Constr, with 0.03 of the workforce employed in the sector, the likelihood of unrest is 0.29 while at one standard deviation above the mean, with 0.06 of the workforce in construction the probability of unrest was .55.

The spatial regression results reported in Table 3 confirm the positive relationship between employment in construction and unrest on June 17. Across all models the coefficient on the Const variable is positive and statistically significant. In Model 3.2, an increase across the range of the variable is associated with a 5-point increase in the Unrest measure, from complete calm to the storming of a party building. In the ordered logistic Model 1, Table 6 in the Online Appendix, which takes Mobilization as dependent variable, the same shift in the Constr variable is associated with a reduction in the likelihood of a county being calm from 0.69 to 0.15. At the same time, the likelihood of a strike decreases from 0.13 to 0.11 and the likelihood of a demonstration increases from 0.09 to 0.55. Construction workers therefore were associated with the most radical forms of resistance to the SED regime on June 17, provoking outright insurrection rather than more moderate forms of opposition such as assemblies or strikes.

I find no support for H2.2, that concentrations of blue-collar workers were associated with the June 17 uprising. There is only a very slight positive correlation ( $r = . 05$ ) between working-class employment and unrest, as shown in the lower-left panel of Figure 3. Although the coefficient on the B.C. Workers variable is positive in both the binary and ordinal logistic regression models presented in Table 2 and Table 6 in the Online Appendix, it is not statistically distinguishable from zero in either model. I graph the effects of working-class employment on unrest from Model 2.6 in in the lower-left panel of Figure 2, showing that although the probability of unrest does increase from .29 to .60 across the range of the variable, these changes are estimated with a very large degree of uncertainty. In the spatial Model 3.4 in Table 3, the coefficient on the B.C. Workers variable is small, negative, and statistically insignificant and results from further spatial models are similar.

All models find that population density contributed to unrest. As illustrated in the lower-right panel of Figure 3, the simple correlation between population density and the Unrest variable is positive and moderately strong, at r = . 25. Of the 10 most restive counties on June 17, only three did not contain a large city, and only a few smaller and geographically isolated municipal areas such as Schneeberg in Saxony remained calm. In all the binary logistic models presented in Table 2, the coefficients on the Ln Pop Dens variable are positive and significant at the p < .01 level, and results are very similar in the ordinal models in Table 5 in the Online Appendix. As illustrated in the lower-right panel of Figure 2, an increase across the range of the population density indicator is associated with an increase in the probability of unrest from .27 to .71 in Model 1.2. Population density is also associated with unrest in all the spatial models reported in Table 3. The spatial clustering of unrest across the country also emerges from the models as an important feature of the June 17 insurrection. In the binary logistic models reported in Table 2, the coefficient on an indicator of counties bordering West Berlin is positive but not significant. However, areas on the West German border were significantly less likely to host unrest on June 17 because they were furthest from the epicenter of the revolt and had a large presence of military and security personnel. Border counties are also more likely to be calm in the ordered logistic regressions reported in Table 6 in Online Appendix. These models include an indicator of counties neighboring East Berlin, which is positively and significantly associated with greater unrest. Interestingly, the linear spatial models of the 10-value Unrest variable reported in Table 3 find that unrest in neighboring counties is a significant predictor of unrest, but areas near Berlin did not experience greater revolutionary activity on June 17. These results all shed light on the localized nature of the revolt and closely follow previous historical accounts. Unrest spread from its epicenter in East Berlin and local hotbeds of insurrection, most commonly larger towns and cities where construction workers instigated strikes.

Conclusion

The last decade has seen considerable scholarly interest in the politics of authoritarian regimes. Central to this research agenda is the question of revolutionary threats, or the likelihood of mass opposition to autocratic governments. In this article, I argue that previous approaches to theorizing and measuring revolutionary threats are underdeveloped. The predominant conception of economic grievances, which revolves around structural poverty or inequality across whole societies, does not capture the sorts of grievances which are most likely to cause collective opposition. These grievances are dynamic and incidental, breaking some convention of acceptability and clearly attributable to the regime. As I show, East Germans lived under conditions of acute scarcity for years in the early postwar era and did not attempt to remove the SED from power. The independent farmers who were associated with unrest in June 1953, on the other hand, were subject to harsh, unpredictable collectivization policies which were clearly attributable to the party and its agents. Regarding mobilization capacity, there has been significant theoretical attention devoted to collective action problems and revolutionary threats to authoritarian regimes. However, I argue that solutions to collective action problems are most likely to be found within groups or networks, rather than at the level of class or state where most previous analysis has been located. The size of the working class was not a predictor of unrest on June 17. Construction workers, on the other hand, who had independent collective decision-making structures and nationwide communications networks, were able to incite intense unrest even when present in small numbers in a county.

Policies which provoke acute, attributable grievances are unlikely to be actively implemented by authoritarian regimes seeking to avoid unrest. Opposition mobilization capacity, on the other hand, is ubiquitous and the predominant cause of revolutionary unrest. It also occurs at the group level and is very difficult for regimes to observe. This suggests that future studies of authoritarian politics should devote significant attention to the repressive and coercive agencies which detect and suppress groups’ mobilization capacity. Although they play a central role in autocratic regimes, secret police and intelligence agencies, presidential guards, the military, and other coercive institutions have been underexplored in the contemporary literature on authoritarianism and democratization. Here, the East German experience is informative. After June 17, reforms of the GDR’s coercive apparatus led to the Stasi growing from a relatively small institution primarily focused on counterespionage to a sprawling mechanism of surveillance and repression targeting the country’s entire population. Understanding mass opposition to authoritarian regimes requires exploring how they use such institutions to detect and suppress the capacity for mobilization in threatening groups.

Footnotes

Acknowledgements

I would like to thank Ben Ansell, Michael Biggs, Alexander Fouirnaies, Kevin Mazur, Katerina Tertytchnaya, three anonymous reviewers, one anonymous editorial board member, and participants at the American Political Science Association meeting in 2013, the Midwest Political Science Association meeting in 2014, International Studies Association meeting in 2016, workshops at the University of Oxford, the Peace Research Institute Oslo and Yale University. Holger Kern generously assisted with data.

Author’s Note

All errors and omissions are my own.

Declaration of Conflicting Interests

The author declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding

The author received no financial support for the research, authorship, and/or publication of this article.

Supplemental Material

Supplementary material for this article is available online at the CPS website .

Notes

Author Biography

Henry Thomson is an assistant professor in the School of Politics and Global Studies at Arizona State University. His research focuses on the political economy of authoritarianism and democratization, with a specific focus on agricultural and food policy.

References

Acemoglu

Robinson

(2006). Economic origins of democracy and dictatorship. New York, NY: Cambridge University Press.

Art

(2016). Archivists and adventurers: Research strategies for authoritarian regimes past and present. Social Science Quarterly, 97, 974-990.

Bates

R. H.

(1981). Markets and states in tropical Africa: The political basis of agricultural policies. Berkeley: University of California Press.

Beier

(1993). Wir wollen freie Menschen sein. Der 17. Juni 1953: Bauleute gingen voran. Köln, Germany: Bund-Verlag.

Boix

(2003). Democracy and redistribution. New York, NY: Cambridge University Press.

Boix

Svolik

M. W.

(2013). The foundations of limited authoritarian government: Institutions, commitment and power-sharing in dictatorships. Journal of Politics, 75, 300-316.

Brakman

Gerretsen

Schramm

(2004). The strategic bombing of German cities during World War II and its impact on city growth. Journal of Economic Geography, 4, 201-218.

Bruce

(2010). The firm: The inside story of the Stasi. New York, NY: Oxford University Press.

Bueno de Mesquita

(2010). Regime change and revolutionary entrepreneurs. American Political Science Review, 104, 446-466.

10.

Buhaug

Urdal

(2013). An urbanization bomb? Population growth and social disorder in cities. Global Environmental Change, 23, 1-10.

11.

Bundesarchiv. (2013). Der 17 Juni 1953: Auswahl aus den Beständen (Bibliotheksbrief).

12.

Casper

B. A.

Tyson

S. A.

(2014). Popular protest and elite coordination in a coup d’etat. Journal of Politics, 76, 548-564.

13.

Collier

Hoeffler

(2004). Greed and grievance in civil war. Oxford Economic Papers, 56, 563-595.

14.

Costalli

Ruggeri

(2016). The long-term legacies of civil war in young democracies: Evidence from Italy. Paper presented at the Order, Conflict and Violence Workshop, Yale University, New Haven, CT.

15.

Crabtree

Darmofal

Kern

H. L.

(2015). A spatial analysis of the impact of West German television on protest mobilization during the East German revolution. Journal of Peace Research, 52, 269-284.

16.

Crabtree

Kern

H. L.

Pfaff

(2018). Mass media and the diffusion of collective action in authoritarian regimes: The June 1953 East German uprising. International Studies Quarterly, 62(2), 301-314.

17.

De Juan

Bank

(2015). The Ba’athist blackout? Selective goods provision and political violence in the Syrian civil war. Journal of Peace Research, 52, 91-104.

18.

De Juan

Wegner

(2017). Social inequality, state-centered grievances, and protest: Evidence from South Africa. Journal of Conflict Resolution. Advance online publication. doi:10.1177/0022002717723136

19.

Diamond

L. J.

Plattner

M. F.

(Eds.). (2012). Liberation technology: Social media and the struggle for democracy. Baltimore, MD: Johns Hopkins University Press.

20.

Diedrich

Hertle

H.-H.

(Eds.). (2003). Alarmstufe Hornisse. Die geheimen Chef-Berichte der Volkspolizei über den 17. Juni 1953. Berlin, Germany: Metropol-Verlag.

21.

Ekiert

(1996). The state against society: Political crises and their aftermath in East Central Europe. Princeton, NJ: Princeton University Press.

22.

Fearon

J. D.

Laitin

D. D.

(2003). Ethnicity, insurgency and civil war. American Political Science Review, 97, 75-90.

23.

Finkel

Gehlbach

Olsen

T. D.

(2015). Does reform prevent rebellion? Evidence from Russia’s emancipation of the serfs. Comparative Political Studies, 48, 984-1019.

24.

Fricke

K. W.

Engelmann

(2003). Der “Tag X" und die Staatssicherheit: 17. Juni 1953—Reaktionen und Konsequenzen im DDR-Machtapparat. Bremen, Germany: Edition Temmen.

25.

Fulbrook

(2005). The people’s state: East German society from Hitler to Honecker. New York, NY: Yale University Press.

26.

Gandhi

Przeworski

(2006). Cooperation, cooptation, and rebellion under dictatorships. Economics & Politics, 18, 1-26.

27.

Gerring

(2007). Is there a (viable) crucial-case method? Comparative Political Studies, 40, 231-253.

28.

Goldstone

J. A.

Bates

R. H.

Epstein

D. L.

Gurr

T. R.

Lustik

M. B.

Marshall

M. G.

Woodward

(2010). A global model for forecasting political instability. American Journal of Political Science, 54, 190-208.

29.

Gurr

T. R.

(1968). A causal model of civil strife: A comparative analysis using new indices. American Political Science Review, 62, 1104-1124.

30.

Hechter

Pfaff

Underwood

(2016). Grievances and the genesis of rebellion: Mutiny in the royal navy, 1740 to 1820. American Sociological Review, 81, 165-189.

31.

Hendrix

C. S.

Haggard

(2015). Global food prices, regime type, and urban unrest in the developing world. Journal of Peace Research, 52, 143-157.

32.

Hohn

(1991). Die Zerstörung deutscher Städte im zweiten Weltkrieg: Regionale Unterschiede in der Bilanz der Wohnungstotalschäden und Folgen des Luftkrieges unter bevölkerungsgeographischem Aspekt. Dortmund, Germany: Dortmunder Vertrieb für Bau- und Planungsliteratur.

33.

Hollyer

J. R.

Rosendorff

B. P.

Vreeland

J. R.

(2015). Transparency, protest and autocratic instability. American Political Science Review, 109, 764-784.

34.

Hübner

(1995). Konsens, Konflikt und Kompromiss: Soziale Arbeiterinteressen und Sozialpolitik in der SBZ/DDR 1945-1970. Berlin, Germany: Akademie Verlag.

35.

Javeline

(2002). The role of blame in collective action: Evidence from Russia. American Political Science Review, 97, 107-121.

36.

Judt

(2005). Postwar: A history of Europe since 1945. New York, NY: Penguin Books.

37.

Kern

H. L.

(2011). Foreign media and protest diffusion in authoritarian regimes: The case of the 1989 East German revolution. Comparative Political Studies, 44, 1179-1205.

38.

Kern

H. L.

Hainmueller

(2009). Opium for the masses: How foreign media can stabilize authoritarian regimes. Political Analysis, 17, 377-399.

39.

Kim

Gandhi

(2010). Coopting workers under dictatorship. Journal of Politics, 72, 646-648.

40.

King

Pan

Roberts

M. E.

(2013). How censorship in China allows government criticism but silences collective action. American Political Science Review, 107, 326-343.

41.

Kleßmann

(2007). Arbeiter im “Arbeiterstaat” DDR: deutsche Traditionen, sowjetisches Modell, westdeutsches Magnetfeld (1945 bis 1971). Bonn, Germany: Dietz.

42.

Kowalczuk

I.-S.

(2013). 17. Juni 1953. München, Germany: C.H. Beck.

43.

Kramer

(1999). The early post-Stalin succession struggle and upheavals in east-central Europe: Internal-external linkages in Soviet policy making (part I). Journal of Cold War Studies, 1, 3-55.

44.

Kramer

Smetana

(Eds.). (2014). Imposing, maintaining and tearing open the iron curtain: The cold war and East-Central Europe, 1945-1989. Lanham, MD: Lexington Books.

45.

Kuran

(1989). Sparks and prairie fires: A theory of unanticipated political revolution. Public Choice, 61, 41-74.

46.

Kuran

(1991). Now out of never: The element of surprise in the East European revolution of 1989. World Politics, 44, 7-48.

47.

Landsman

(2005). Dictatorship and demand: The politics of consumerism in East Germany. Cambridge, MA: Harvard University Press.

48.

Lohmann

(1994). The dynamics of informational cascades: The Monday demonstrations in Leipzig, East Germany, 1989–91. World Politics, 47, 42-101.

49.

McAdam

Tarrow

Tilly

(2001). Dynamics of contention. New York, NY: Cambridge University Press.

50.

Opp

K.-D.

Gern

(1993). Dissident groups, personal networks, and spontaneous cooperation: The East German revolution of 1989. American Sociological Review, 58, 659-680.

51.

Pfaff

Kim

(2003). Exit-voice dynamics in collective action: An analysis of emigration and protest in the East German revolution. American Journal of Sociology, 109, 401-444.

52.

Plümper

Neumayer

(2010). Model specification in the analysis of spatial dependence. European Journal of Political Research, 49, 418-442.

53.

Port

A. I.

(2007). Conflict and stability in the German Democratic Republic. New York, NY: Cambridge University Press.

54.

Rozenas

(2016). Non-competitive elections and incentives for insurgency: Micro-level evidence from the post-war Soviet union. Paper presented at the Order, Conflict and Violence workshop, Yale University, New Haven, CT.

55.

Schöne

(2005). Frühling auf dem Lande? Die Kollektivierung der DDR-Landwirtschaft. Berlin, Germany: Ch. Links Verlag.

56.

Schroeder

(2013). Der SED-Staat: Geschichte und Strukturen der DDR 1949-1990. Köln, Germany: Böhlau Verlag.

57.

Scott

J. C.

(1985). Weapons of the weak: Everyday forms of peasant resistance. New Haven, CT: Yale University Press.

58.

Sperber

(2004). 17 June 1953: Revisiting a German revolution. German History, 22, 619-643.

59.

Spittmann

Fricke

K. W.

(Eds.). (1988). 17 Juni 1953: Arbeiteraufstand in der DDR. Köln, Germany: Verlag Wissenschaft und Politik.

60.

Svolik

M. W.

(2012a). Contracting on violence: The moral hazard in authoritarian repression and military intervention in politics. Journal of Conflict Resolution, 57, 765-794.

61.

Svolik

M. W.

(2012b). The politics of authoritarian rule. New York, NY: Cambridge University Press.

62.

Tarrow

(1994). Power in movement: Social movements, collective action and politics. New York, NY: Cambridge University Press.

63.

Thomson

(2013). Of sticks and carrots: Repression and food policy as cause and consequence of the June 1953 failed revolution in East Germany. APSA 2013 Annual Meeting Paper.

64.

Thomson

(2016). Rural grievances, landholding inequality and civil conflict. International Studies Quarterly, 60, 511-519.

65.

Thomson

(2017a). Food and power: Agricultural policy under democracy and dictatorship. Comparative Politics, 49, 273-293.

66.

Thomson

(2017b). Food and power: Regime type, agricultural policy and political stability. Unpublished book manuscript.

67.

Thomson

(2017c). Repression, redistribution and the problem of authoritarian control: Responses to the 17 June uprising in socialist East Germany. East European Politics and Societies, 31, 68-92.

68.

Tilly

(1978). From mobilization to revolution. New York, NY: McGraw-Hill.

69.

United Nations. (2000). Charting the progress of populations. New York, NY: Department of Economic and Social Affairs, Population Division.

70.

United States Strategic Boming Survey. (1945). The United States Strategic Bombing Survey: Over-all report (European War). Washington, DC: U.S. Government Printing Office.

71.

Wallace

(2013). Cities, redistribution, and authoritarian regime survival. Journal of Politics, 75, 632-645.

72.

Wallace

(2016). Juking the stats? Authoritarian information problems in China. British Journal of Political Science, 46, 11-29.

73.

Weber

(2012). Die DDR 1945-1990. München, Germany: Oldenbourg.

74.

Weyland

(2012). The Arab spring: Why the surprising similarities with the revolutionary wave of 1848? Perspectives on Politics, 10, 917-934.

75.

Wolle

(2013a). Aufbruch nach Utopia: Alltag und Herrschaft in der DDR 1961-1971. Berlin, Germany: Ch. Links Verlag.

76.

Wolle

(2013b). Der Grosse Plan: Alltag und Herrschaft in der DDR 1949-1961. Berlin, Germany: Ch. Links Verlag.