With hindsight it is possible to see in the course of the Ethiopian revolution a process of radicalization and post-revolutionary consolidation through which the Provisional Military Administration Committee (PMAC, or the Derg) established a stable new order on the ruins of the old. The direction of political and social change was by no means this clear during the first years after 1974. The pattern of revolutionary transformation involved a deep paradox, namely the conflict between the military leadership at the top and the various radical civilian forces below. At each stage of the revolution, the PMAC, aware of its own political weaknesses, sought to establish alliances with these civilian forces. Indeed it put into practice much of what the civilians were themselves demanding. But no stable alliance between them proved possible: the revolutionary transformation effected by the PMAC therefore involved both the implementation of a radical social program and the destruction or at least submission of other forces that had helped to bring these changes about.
Shaping the Derg’s Radicalism
There is little in the known intentions of the PMAC leadership that would explain why this military regime did implement a radical program. They were not members of a secret political party, or of some left-wing military conspiratorial group. The answer lies in the objective factors that shaped this malleable group of officers in the period after September 1974. The first was the pressure from the peasantry for land reform, and from the mass of urban poor, wage earners and semi-employed for an improvement in their living conditions. Without taking measures to meet these popular demands, the PMAC could never have survived, nor could it have begun that development of the economy upon which the longer-run survival of the regime still rests. Another factor was the pressure from the political parties themselves. The third factor that radicalized the Derg was the factionalism and the social conflicts reflected within it. They had been drawn from the intermediate sectors of Ethiopian society. A combination of a generic class hostility to the ancien regime and the particular radical program represented by Mengistu shaped the course of events within the PMAC. The fourth factor was international. While relations with the USSR had been improving slowly, the threat to the PMAC posed by the military situation accelerated the process of alignment with Moscow. The fifth reason was the most important: If it was to survive, the PMAC had to destroy the socioeconomic foundations of the old regime. This involved expropriation, a measure of mass mobilization, and the extension of state control throughout society.
The Major Reforms
The PMAC quickly instituted major reforms that established control of the main areas of the urban economy and destroyed the bases of the old regime in the countryside. On January 1, 1975, it nationalized all banks and 13 insurance companies. By late 1976, two thirds of all manufacturing was under the control of the Ministry of Industry. On March 4, 1975 all rural land was nationalized, tenancy was prohibited, and the peasantry limited to plots of only ten hectares. The regime dispatched up to 60,000 students to the countryside in a zemacha (campaign) to mobilize the peasantry. On July 26, it nationalized all urban land and all rentable houses and flats.
The proclamation of a republic, in March 1975, brought the many centuries of monarchy to an end. Three Islamic holidays were designated as national festivals on a par with their Christian counterparts. Combined with the land reform, these changes destroyed the position of the monarchy and disestablished the Coptic Church, depriving it of its material base, in land, and its ideological status. 
Of all the reforms enacted, the most important was the land reform. In the south of the country the subjugated Oromo peasantry seized the lands, ousting and in some cases killing the neftegna owners. Peasant associations were formed in many areas of the south. In the north, by contrast, where the prevailing system of rist land provided a form of ownership to the laborers, landowners were able to claim that the reform threatened the peasants’ rist rights and to present the reform as an Oromo-Muslim threat to the Amharic-Tigrean Christian order. Church lands not tilled directly by priests were also expropriated. In the Afar areas, the sultan was also able to mobilize resistance among his people, since in this Muslim area of the northeast, unlike the Oromo south, both landowner and peasant shared a single religious affiliation. Resistance broke out in many parts of the north, and effectively subtracted much of the provinces of Tigray, Gojjam and Begemdir, and parts of Wollo and Shoa, from government control during this period. Insofar as one can speak of a socially based counter-revolution, as distinct from resistance primarily on grounds of nationality, it was in these provinces.
The implementation of the reform transferred into social relations what had already been achieved politically by the deposition of the emperor. Rural incomes rose as the peasants were relieved of the landlords’ exactions, and the absolute amount grown also increased. Despite the lack of government cadres, peasant associations spread through much of the countryside: They had local judicial functions as well as the role of providing seeds and other inputs. By 1978 there were officially said to be 28,583 peasant associations in Ethiopia, with a total membership of 7.3 million households, backed by a system of peasant defense units encompassing 500,000 armed villagers. The number of service cooperatives had reached 343.
Apart from the activities of the counter-revolutionaries in the northwestern provinces, other substantial social problems have arisen. First, students sent out into the countryside to mobilize the peasantry around the reform soon clashed with the police and administrators already there. Landlords killed some zemacha cadres, and the PMAC itself was increasingly wary of this agitational force that lay outside its control. By the end of 1975, the government had recalled most of the students to the cities. The Peasant Association Decree of December 1975 was designed to establish a more permanent mobilizing structure in the rural areas, but the zemacha students brought back to the cities a deep suspicion of the PMAC’s policies. This fueled the conflict that had already broken out there — in the summer of 1975 between the Derg and its left opponents.
Second, the newly powerful peasantry withheld much of their produce from the urban market, and enjoyed their new freedom to consume more themselves. This, combined with war strains on the country’s transport system, led to severe shortages and inflation in the towns between 1976 and 1978. Most important of all, the new tenancies gave the peasants effective control over their plots. The central government was unable to supervise the local redistribution policies, and possession of means of production other than land gave the edge to the upper layer of peasants once redistribution took place.
Difficulties arose in the areas of coffee production, which accounts for up to 25 percent of the agricultural work force. The weakness of the state purchasing body, combined with the reliance of the peasantry on their former landlords for their subsistence goods, opened the door to new forms of exploitation of the peasantry by richer peasants and by merchants. Richer peasants were able to gain a disproportionate amount of land, and to ensure that it was they who controlled the new peasant associations, their credit, equipment and distribution.
As a result of these problems, and because of the impact of the Eritrean and Ogaden wars, food output did not even keep pace with population growth in the 1974-1978 period. Between 1977 and 1979 Ethiopia faced severe food shortages, both in the towns and in parts of the countryside, and even the reemergence of famine in the northern provinces.
The regime responded with the Economic Development Campaign. Agricultural output rose 2.4 percent in 1978-1979 and 4.8 percent in 1979-1980. A decree of June 1979 placed new upper limits on areas of tenancy and laid the basis for the transition to producers’ cooperatives. Official statements stressed the dangers of forcing the pace of collectivization, but the longer-run import of this decree was not lost on the peasantry. In early September 1979, there were clashes in Sidamo province in which an estimated 150 people were killed. By 1980 there were just 40 producer cooperatives in existence, mainly in the initial stage of organization.  State farms and cooperatives combined accounted for only 6 percent of agricultural output and 20 percent of marketed production.
The mobilization of rural resources for the Economic Development Campaign encountered substantial resistance from those who benefited most from the 1975 land reform. The problem is one that revolutionary regimes have had to face before, but none have found a balanced or democratic answer to it. The course of the land reform is indicative of the general character of the PMAC’s interventions. This very radical measure was carried out within six months of the PMAC’s coming to power, and just over a year after the first protest demonstrations in Addis Ababa. The officers who promulgated the reform had no political cadres with which to carry it out, and they lacked even a previous history of support for such social transformations. Their very weakness led them to try to win popular support by enacting such measures and by handing substantive power to the kebeles (neighborhood associations) in the towns and the peasant associations in the countryside. This delegation of power produced a new set of conflicts. Lacking the political mediations needed to resolve them in a non-coercive manner, the officers then sought to surmount them by administrative fiat and by direct coercion.
Similar problems arose in the context of the urban reforms. These were not, in the first place, as radical as the land reform. The state already controlled much of the banking sector and parts of industry, and there was no substantial private Ethiopian property to expropriate. Where a flourishing Ethiopian capitalist class did exist — in the retail trade — it was left relatively unmolested. Ethiopia’s largest export commodity, coffee, remained to a considerable extent in the hands of the same powerful merchants who had controlled it in the days of Haile Selassie. In addition, the private sector of the economy still included nearly all of the peasant sector as well as small-scale artisans.
The contradictory effects of the PMAC’s policies were clearest in the factories. Here, despite a relative quiescence compared to white-collar sectors in the upsurges of February to June 1974, there was growing militancy as a result both of the state’s “socialist” proclamations and of the left groups’ influence. By 1976, workers raised widespread demands for shop-floor control over production. The government repelled these with the appointment of powerful state managers, and with the imposition of a new trade union structure. Strikes, illegal since 1974, were prohibited in the 1975 Labor Code.
The housing law also encountered serious difficulties: the nationalization led to a net subtraction of the number of dwellings available for rental. Since many urban owners were in origin rural proprietors, the number of such people in the towns greatly increased as they were driven off the land by the 1975 land reform. The kebeles did carry out some housing redistribution and welfare programs, but they were unable to provide a sufficiently large set of retail outlets for food to offset the shortages and hoarding that began in 1976.
The kebeles became the site of an urban terror campaign by the Ethiopian Peoples’ Revolutionary Party (EPRP) against supporters of the government, and by ruthless state officials against their opponents. Each side sought to mobilize semi-employed youth against the other. The PMAC decision to establish and arm the kebeles, which numbered 1,800 by the end of 1977, nearly 300 of them in the capital — reflected the government’s weakness and their need to encourage popular action. Yet even more than in the countryside, the kebeles were taken over by political forces that the Derg had to disown. By the end of 1978, the kebeles were much more tightly controlled from above in financial and administrative matters, and they were progressively deprived of their security functions, losing their right to administer local jails and to carry arms.
Division and Consolidation
Apart from their conflicts with the civilian political forces, the Derg officers were themselves rent by serious divisions.  PMAC members who favored a rapid transition to civilian rule were in a minority, and the public history of the Derg exhibited a series of internal conflicts, in most of which the defeated officers were killed.
Between 60 and 80 of the original 126 members still eight remain within the Derg. The great majority of the Derg officers are graduates of the Holeta military academy. Almost all those from the senior academy at Harar were eliminated in the July 1976 and February 1977 purges. Amharas still seem to be a majority. The minority of Oromos are mainly Christian, and represent that section of Oromo society traditionally associated with the Ethiopian state. Of the five Eritreans believed to have been members at the beginning, all have either defected or been executed. Only one Derg member, the Oromo Ali Musa, is definitely a Muslim, despite the fact that around 40 percent of Ethiopia’s population adheres to Islam.
The ten-member Standing Committee is the core of the revolutionary government. Beyond it lies a Central Committee, with an estimated 32 members. The congress of all the surviving members is believed still to meet on an annual basis. While it has powers of discussion and ratification of some decisions, it no longer takes major initiatives against the will of the Standing Committee. Dozens of PMAC members not directly involved in the Standing Committee are distributed among civilian ministries or on the PMAC committees that oversee each ministry. Twelve have been appointed as provincial governors,  and since 1980 four have held positions in the Council of Ministers. The PMAC itself has ceased to be the major decision-making assembly. Its members have become more closely and more regularly integrated into the administration at both national and provincial level.
The original membership came about by a combination of direct election and accident. It was then frozen as the highest revolutionary authority. So far as is known, no Derg member has ever been subjected to new elections by the military who chose him, and none of the members who were excluded or who fled has ever been replaced. The surviving members have preserved both their clandestinity — no full list has ever been published — and their curious relation to the civilian bureaucracy.
The PMAC had the credit of having finally toppled the emperor, and it claimed to be only a provisional administration, so that some months passed before the full force of the civilian opposition gained momentum. In the first major outbreak of opposition, on May Day 1975, soldiers killed some demonstrators demanding an immediate return to civilian rule. During the year following the September 1974 seizure of power by the PMAC, the underground student-based nuclei, inspired by returning exiles, and the rather conventional leadership of the Confederation of Ethiopian Labor Unions (CELU) converged to form an alliance militantly opposed to the PMAC. The EPRP, which proclaimed its existence in August 1975 demanded an elected assembly and the immediate constitution of a people’s democratic republic. The CELU congress of September 1975 backed this program with general strikes that lasted for some weeks.
The PMAC crushed the CELU protests. In December it proclaimed a new restrictive labor law and officially dissolved CELU. But the initial instinct was not simply repressive: For another year the PMAC tried to win over its opponents in the EPRP and CELU. In early 1976 the press was opened to a debate between the All-Ethiopian Socialist Movement (MEISON) and the EPRP. The National Democratic Revolution, proclaimed in April 1976, was designed to meet some of the demands of the radical left, and for some time in 1976 CELU was permitted to operate again. Indeed, most of the EPRP program was represented in the NDR — with the exception of a commitment to a return to civilian rule.
By the end of 1976, the breach was all but final. In July 1976, the group who favored conciliation with the EPRP was eliminated from the Derg. In the same month the Derg introduced the death penalty for certain political crimes, and prolonged the state of emergency proclaimed in September 1975 during the general strike. The government blamed the EPRP for an attempt on Mengistu’s life on September 23, 1976, although a military faction opposed to him was probably responsible.  After September, the EPRP adopted a policy of assassination of PMAC supporters and between late 1976 and mid-1977 they killed several hundred of them, nearly all civilians. In response, the PMAC launched the campaign later known as the “red terror.”
The EPRP was active in a few cities — Addis Ababa, Dire Dawa, Jimma — and its military wing, the EPRA (Ethiopian People’s Revolutionary Army) carried out actions in Tigray province. The EPRP also made public in 1976 its alliance with the Eritrean People’s Liberation Front (EPLF). Many EPRP members were among the thousands who died in the “red terror,” and by mid-1978 the EPRP had ceased to operate in Addis Ababa and other towns. By 1981 the EPRA was reduced to operations in the mountains near Gondar.
The Derg simultaneously crushed the working-class opposition. The government established the All-Ethiopian Trade Union (AETU) to replace CELU in late 1976, but its first three presidents were assassinated by the EPRP. By December 1977, the AETU claimed to have a membership of 350,000.  Apart from the EPRP’s campaign of infiltration and assassination, the AETU was rent by conflicts between MEISON, which dominated its leadership, and the Derg. The PMAC dismissed the AETU leadership in May 1978.  After this, the loyalty of the mass of wage earners to this new imposed structure was bound to be suspect.
The EPRP and the “Red Terror”
The EPRP’s resort to urban assassination in the period from September 1976 to mid-1978 was a serious error.  In slaying leaders of the AETU or political instructors at the Yekatit 66 ideological school, as well as elected kebele officials, they were attacking people who, whatever their specific political positions, were trying to advance the Ethiopian revolution. At the same time, the EPRP’s decision to launch this campaign involved a fatal underestimation of their opponents’ capacity to survive and counterattack. It encouraged a situation in which the PMAC resorted to a violent counteroffensive which consumed most of the EPRP and many others.
The mistaken nature of the EPRP’s policies does not, however, justify the actins of the PMAC and their civilian allies. Executions as early as November 1974, dictated by the political situation at that time, were at least partly intended to instill fear in the population. The use of arrest, torture and execution of civilian opponents began with the clashes of May Day 1975, but this did not become a central part of PMAC policy until the end of 1976. Toward the end of April 1977, as the opposition were preparing for another May Day protest, up to 500 students were killed in Addis Ababa, and around the same time 22 people in the area of Berhanu Salem printing works were slain by a sadistic kebele leader, Girma Kebede. The campaign of “red terror” was officially launched in November 1977 and lasted until May 1978. In the 1976-1978 period, up to 30,000 people, mainly left-wingers or suspected left-wingers, were imprisoned, and several thousand killed in these campaigns. One POMOA spokesman made the following point about the summary justice of the “red terror.” “The only way to identify a counter-revolutionary is through the consciousness of the masses. The masses know who their friends and enemies are. After long observation they decide where someone belongs. They are not like the police, who have to arrest and then interrogate a person. The masses know who someone is already. They pick him up only after they have identified him.” 
No accounting or self-criticism was made by any PMAC member, and the members of the state repressive apparatus who practiced such actions remained in their posts. The PMAC’s claim that the EPRP-CELU opposition was counter-revolutionary in nature and reflected links to the USA and to Arab reaction is not sustainable. The EPRP-CELU alliance reflected the growth of a new consciousness among CELU members, both blue- and white-collar, after the PMAC came to power, a very real if delayed politicization that gave the EPRP its broader popular following.
On the other hand, the EPRP’s claim to have played a leading role in the February revolution of 1974 is baseless. It did not exist as a public organization until August 1975. Even the paper in which its ideas first appeared, Democracia, did not come out until June 1974, by which time the political initiative was with the radical army officers. There is also exaggeration in the self-image of the EPRP as “the party of the proletariat.” Its social base was always the students, and to a lesser extent intellectuals in state employment. It increasingly relied on unemployed urban teenagers in its own terror campaign, and recruited them on a militaristic basis. Its ties with the CELU leadership were ties between two distinct organizations, one student-based, the other a trade union body. This self-proclaimed vanguard ended up in a losing battle for survival, divorced from the popular following it claimed to have.
The difficulties faced by the EPRP from 1976 onwards were reflected in a number of organizational splits. In 1979, at the Fourth Plenum of its Central Committee, reportedly held inside Ethiopia, the EPRP argued: a) that it had misinterpreted the question of state power, by appearing to lay too much emphasis on the immediate establishment of a provisional people’s government; b) that it had been mistaken to give priority to the urban over the rural struggle; c) that it had failed to evolve an adequate form of united front work with other groups; d) that in dealing with the national question it had been overindulgent to “narrow nationalists” in Eritrea and to right-wing forces in Tigray (the TPLF); e) that it had taken too long to condemn “social-imperialism.” 
This self-criticism did not go to the root of the problem. The EPRP still characterizes the Derg as “fascist,” a term that indicates that they have not yet comprehended the nature of the state in contemporary Ethiopia and the changes it has instituted. The EPRP’s new rural strategy may well be able to survive because it is located in remote areas, but for this reason it will become even more marginal. The deeper problem, of relying on armed struggle as the main tactic deployed against the PMAC, remains unresolved.
The EPRP’s self-criticism for its indulgence towards the USSR is indicative of its general evolution. It began as a highly centralized conspiratorial group, given to workerist demagogy. It then became fully aligned with the policies of Peking. In 1979, it criticized China and transferred its loyalty to Albania. The EPRP’s claim that it always had a theory of Soviet social imperialism may be doubted. Prior to this self-criticism, the EPRP had held the view, shared by much of the Ethiopian left, that the PMAC was linked to US imperialism, and that Mengistu was part of a pro-American group within the government.
The EPRP’s account of its policies on the national question is also somewhat abbreviated. Although they argue in 1979 that they have always held the right of nationalities to self-determination “including secession,” scrutiny of the August 1975 statement reveals no such position. (Section IV of the EPRP program promises “to give full rights to the nationalities and peoples of Ethiopia to determine their own future” but talks only in terms of “a voluntary union based on equality and brotherhood” and of “internally autonomous regions.”) This reserve is all the more striking because, on available (albeit impressionistic) evidence, the EPRP had a disproportionately high percentage of Eritreans and Tigreans in its leadership. Material and political cooperation seems to have developed fully only in 1976 when the EPLF and EPRP issued a joint statement. Later divergences with the EPLF may have reflected the fact that the EPLF had undermined the EPRP’s position in Tigray province by supporting the rival TPLF. As with their policies on international issues, and indeed their whole relation to armed struggle, the EPRP’s policy on the national question would appear to have altered rather drastically in response to conjunctural factors. The real root of its erroneous approach lay in the kind of militaristic and highly dogmatic leftism its members had inherited from their exile student days.
The MEISON Clash
The PMAC’s ideological evolution was inseparable from its uneasy alliance with a number of other civilian forces. They consisted mainly of Addis Ababa students, dominated at the leadership level by returning exiles. The EPRP represented the section which had rejected the Derg outright, but five other groups allied with the regime for some time. Chief among these was MEISON. Its leaders were those who had first dominated the Ethiopian students in exile in the 1960s. MEISON, like the EPRP, claimed a clandestine pre-history that antedated its public emergence. According to this account, the organization was founded in 1968, and took up public positions with the paper Voice of the Masses in 1974. MEISON did not proclaim itself publicly until March 1976, seven months after the EPRP, and just prior to the NDR announcement in whose drafting it played a major role. Four other political groups collaborated with MEISON in supporting the PMAC. The Oppressed People’s Party of Ethiopia, or ECHAAT, was founded in 1976 as a breakaway from MEISON, whom it accused of neglecting the cause of the Oromo people. The WAS (Labor) League was established by Senaye Likke, himself a founding member, in his exile student days, of the Communist Labor Party of the USA. MALERED, or the Marxist-Leninist Organization of Ethiopia, was another small splinter from MEISON. All of these appear to have had some supporters or at least sympathetic protectors within the PMAC; they subscribed to the NDR, but they were in origin, products of the radical students and continued to recruit most of their members from this sector. The last group, Abyotawit Soded, or Revolutionary Flame, was formed by army officers at some point in late 1975 and was generally regarded as being the creation of certain PMAC members, among them Mengistu.
The alliance of these groups formed the basis for the establishment in late 1975 of what later became the Provisional Office for Mass Organizational Affairs (POMOA). In 1976 the five established an ideological training school, Yekatit 66 (February 24 in the Ethiopian calendar), which produced several thousand cadres in the next two years. In July 1977, the five established a common front known as EMALEDH (Union of Marxist-Leninist Organizations), designed as a further step on the way to founding a new party. But the level of factionalism proved to be such that no convergence was possible.
According to the NDR program, POMOA had four main functions: to prepare the basis for merging the five groups into a political party; to develop the mass organizations (trade unions, women’s organization, kebeles and so on); to run the Yekatit 66 ideological school; and to give ideological training to the militia. By the end of 1976, MEISON dominated the 15-member central council of POMOA. This posed a challenge to the PMAC by its pervasive extension through the civilian ministries, and through the kebeles, trade unions and parts of the peasant associations. The PMAC claims that MEISON took advantage of its influential position in POMOA, and of the crisis attendant upon the Somali and Eritrean wars, to attempt to seize state power. The other major issue raised by the PMAC was MEISON’s role in the terror. While the PMAC ordered and itself helped to prosecute the terror, and continued to carry it out after MEISON had disappeared, it is also apparent that the latter used its position of influence in government and in the kebeles to pursue old political feuds.
MEISON, for its part, charges that the PMAC was not carrying out the destruction of the old state machine but was retaining its “feudal-bureaucratic” structures.  It also charges that the PMAC blocked any restoration of political freedoms, apparently meaning freedoms for independent MEISON activity. It claimed that the PMAC was departing from the nationalities policy laid down in the NDR. The division in this matter was over Ogaden. While MEISON opposed the Somali invasion, it was an organization drawn in particular from Oromos and hence sensitive to the way in which the southern question was being handled. MEISON’s complaint was that in the face of the Somali invasion the Derg was rearming the old Amhara settlers in the south, the neftegnas. These landowners would then be in a better position to reestablish the power that the 1975 land reform had taken from them. MEISON saw this as a retreat on the nationalities issue, and as a “rehabilitation of reactionaries” that cast a slight on the mainly Oromo militia that MEISON had been helping to train. During 1977 MEISON also argued that pro-American forces remained strong within the PMAC, and at the same time criticized the Soviet Union for failing to provide adequate assistance.
The argument over rearming the neftegnas may well have a basis in fact, but it is phrased uniquely in class terms (“reactionaries”) and passes over in silence the national dimension of this issue. The criticism fails to take account of the gravity of the overall situation confronting the PMAC in the summer of 1977 in the face of a massive Somali onslaught. Precisely because of the national issues involved, and the possibility of Oromo sympathy for the Somalis, the PMAC knew that the local elements upon whom they could most rely would be the neftegnas.
MEISON’s move into clandestinity in the summer of 1977 reveals this underestimation of the gravity of the situation in an even starker form: Such a move could only weaken the resistance to the Somali invasion. Because of the invasion, the regime was gripped by a war fever that made oppositional political activity and the restoration of some political freedoms all the more unlikely. This disregard for reality would seem to be born out in their documents of this period: in the midst of a foreign invasion, they issued a call for the complete liquidation of the bureaucracy and criticized the Derg for using officers of the former imperial army.
But MEISON’s analysis is most flawed in its evaluation of its own role: Far from making self-criticism for its enthusiastic role in the “red terror,” it seeks to justify it. “The most important thing to realize is that this situation has been deliberately exaggerated in order to poison international opinion,” wrote party member Negede Gobeze in Le Monde on September 17, 1977. At the same time, MEISON overestimated its own ability to lead a clandestine struggle against the Derg. MEISON’s dramatic break was therefore based on an analysis which, like that of the rival EPRP, led many militants to imprisonment or death.
Prior to the 1974 revolution, there had been no group active within Ethiopia, including Eritrea, that professed adherence to socialism. By 1976, virtually all the contending groups had come to use a vocabulary influenced by Marxism and to propound socialist policies. After Mengistu’s accession to full power in February 1977, the ideological orientation became more explicitly Marxist-Leninist. The course of events in 1974 and 1975 must account for this change. As the histories of other revolutions show, sharp changes of political tempo can produce in weeks shifts of political consciousness, after years of apparent stasis. Given the fragility of the Ethiopian state, and the dramatic way in which it crumbled, the rapidity of this change does not in itself mean that the ideological shift within the PMAC was merely affectation. On the other hand, convenience and some opportunism must also have played a role. The regime was committed to carrying through radical reforms, it needed a new legitimating and nationally cohesive ideology, it felt challenged by civilian groups to the left, and it relied for essential support upon the USSR and its allies.
This pressure to the left was always controlled by the prime concern of the PMAC — to retain power. In adapting Marxism-Leninism, it selected those themes that served its purposes. If it began to inculcate ideas of class struggle and materialist analysis of society, it also drew on “red terror” and the need to fight all secessionists as “counter-revolutionaries.”
Eastern Europe was one source of new ideological influences that first reinforced but later competed with the civilian influence. Initially, in the absence of an Ethiopian Communist party, the influence of the left milieu in North America and Western Europe was much stronger. Until 1977, generically pro-Chinese views were more common than pro-Soviet ones among the Ethiopian left. The influx of large numbers of Soviet books and periodicals, linked to the strategic alliance with the USSR and to increasing numbers of Soviet-trained cadres, led by 1978 to the predominance of an orthodox Eastern European Marxism at the official level. State publications now condemned those whose views they opposed as Trotskyists, anarchists and Maoists. Even “Anglo-Saxon Marxism,” was anathematized. “One important factor that has inhibited the development of socialist theory in Ethiopia is the pseudo-Marxist Anglo-American literature which had been fora long time the mainstay of political thought for the radical Ethiopian intelligentsia. The theoretical distortion of Marxism a la Monthly Review, the so-called New Left Review and so on, have contributed their part to arrest the development and crystallization of serious socialist thought in Ethiopia.” 
This inflow of Marxist ideas required a massive cultural enterprise at both the educational and linguistic levels. While the level of mass “politicization” must remain suspect, the whole process has had some longer-term effects: very large numbers of people have now been mobilized in Ethiopia and thousands of new cadres have been formed as as a result. This is especially remarkable because of the absence of any particularist limiting phrase about the “Ethiopian” character of this socialism. There is no concession to supposed “third ways” of the Peronist or Baathist variety. While such a resolutely anti-parochial orientation does run the risk of obscuring the particularities of the local society, as has happened in Afghanistan, it may have prevented the Ethiopian revolution from making some of the cultural and class concessions that such qualified varieties of socialism tend to conceal.
Consolidating State Power
Some civilian groups have enjoyed temporary influence, but none has been able to play a leading role in the post-revolutionary regime. The refusal of the military to share real power, accentuated by the divisions within the PMAC over this question, prevented any consistent conciliation from being attempted. Yet this alone cannot explain the long delays involved in establishing a political party. Had the PMAC decided to set up a client political apparatus, it should have been able to do so. The Berg was, however, pushed along this path precisely by its inability to rule on its own, as reflected in its uneasy relationship with unions, kebeles, peasant associations and the Addis Ababa students. This weakness both impelled the PMAC to try to establish a working relationship with the civilian forces and at the same time constrained it from granting substantive independence to them.
The underlying cause of the Derg’s weakness was not just a shortage of competent political and administrative personnel, but a reflection of the balance of forces that existed after the revolution. Politically active civilian forces had emerged and, despite the military character of the regime, further politicization took place, in town and country. The civilian forces had a real political context from which to derive strength to resist the PMAC. This balance of forces had come about because of the way in which the revolution occurred: not through a straightforward military coup, but by a protracted challenge to the old regime in which both civilians and military had participated.
This combination made the revolution successful. At the same time it established the terms in which a debilitating post-revolutionary conflict was to take place, with each side claiming to be the true progenitor of the 1974 upheaval. A separate ingredient was the political culture of the civilian milieu, in particular its factionalism and tendency toward a rhetorical simplification of political reality. One factor was the detached status of much of the left, which operated through student activism and underground meetings, and lacked roots in broader social classes, urban or rural. These groups were divorced from those social contacts that could have exerted some strategic control upon them. A second factor was the refusal of the army to allow them to play an autonomous political role. A third factor was the pre-revolutionary ideological formation of these groups, heavily influenced by the trends prevalent in Europe and North America in the late 1960s, which displayed an impatience with the demands of careful political preparation. The 1974 revolution may in some measure have reinforced this early formation. The very advances made by a spontaneous popular upheaval and the apparent absence of political organization did little to instill the political sobriety and experience that leaders of other revolutions, in China or Vietnam, have had to acquire to negotiate the long years of pre-revolutionary struggle.
Underlying all these elements is the influence of Ethiopian society itself, where conspiratorial practices had historically prevailed over forms of collective political activity. The radicalism of the mid-1970s revitalized these traditions and gave them an appropriate disguise in the factional language of the international left. This reflected the manner in which the Ethiopian left had been formed at home and abroad, the lack of experience of practical mass work, and the failure to combine a socialist intellectual commitment with a concrete appreciation of the specific features of Ethiopian society or the political traditions of the country. Here the military were at an advantage. They were not just more determined and able to exercise repressive power when needed. By virtue of their formation within the imperial state apparatus, they were conscious of the political forces at work in their country and were endowed with some of the practical experience that would enable them to control a state machine and outmaneuver their more vocal but ineffective civilian rivals.
The Post-Revolutionary State
The state — the institutions through which political power is wielded and the means through which classes exercise their political domination over society — forms a central part of all social and political analysis. But the state cannot be seen as a mere instrument of class rule: It is able to act in some degree independently of, and even contrary to, the interests of those on whose support it rests. It is itself the forum or site of political conflicts between different social forces.
In revolutionary contexts, the normal controls of social class over the state may be attenuated, and the very conflicts of the revolutionary period are concentrated on gaining control of existing apparatuses or on establishing new institutions capable of replacing those through which the social relations of the old order are reproduced.  The autonomy of the state may also be greater in situations of transition from one mode of production to another: Social forces associated with both modes may compete for power, not just over the state but within it, in such a way that the institutions are released from the class controls imposed where one mode and one system of social relations are dominant. Finally, the autonomy of the state may be greater where class relations are themselves less developed, where the predominance of social relations are determined by factors other than class, such as ethnic and lineage bonds.
Ethiopia is a society undergoing a revolution, its social relations are transitional and heterogeneous, and its class forces are still partially developed. The state is therefore both a reflection of the conflicts within society, the object of that conflict, and a means by which those concerned to transform society can hope to achieve their aims. If the state is to a considerable extent limited by the objective structures of Ethiopian society, it is nonetheless an active agent in the process of consolidating a post-revolutionary order. It reflects to some degree the interests of class and social forces, but may also play an active role in class formation, in constructing new social relations of which its agents are a part.
The Ethiopian state was the site of the decisive conflict between supporters of the absolutist state and proponents of a new social order. The administrative expansion of the post-1941 period, combined with the growing agrarian crisis and international pressure, produced a revolution from above. Two other aspects of the imperial state were also important in subsequent events. First, unlike all other state apparatuses in Africa, this was not created by colonial rule. Linked to this was the role the state already played in the Ethiopian economy. The survival of a predominantly pre-capitalist order had precluded the emergence of an Ethiopian bourgeoisie capable of playing either the economic or the political role seen elsewhere in the Third World. Hence the state was already an important factor in economic activity. The social barriers represented by the nobility were swept away in the revolution. The unique character of Ethiopia therefore gave the state a particularly “autonomous” and potentially radical role.
Two main processes were dominant in the period after 1974: an expansion in the size and role of the military apparatuses, and a relative seclusion of the civilian apparatuses, coupled with their subordination to military control. At the time of the revolution, the Ethiopian armed forces totaled over 44,000 men, 41,000 of whom were in the army. Military expenditure in 1971 was an estimated $40 million. By 1980, the army had expanded to around 75,000 men, but was supplemented by a militia of relatively full-time soldiers totaling 150,000, or ten divisions. With the other paramilitary groups and the relatively static navy and air force, personnel as of 1980 total 230,000, and defense expenditure totals $385 million.  This numerical growth went together with an extension of military control over the civilian branches of state, and over the new mass institutions established after 1974.
The degree of transformation within the armed forces themselves is more difficult to establish. When constituted in 1974, the Derg was drawn from ranks below that of lieutenant colonel, with the result that none of the top commanders in 1976 were members.  It is possible to discern a certain social pattern in the outcome of the conflict within the Derg itself. The more senior officers were gradually ousted, only a few remaining in positions of power. The lowest ranks in the PMAC, the NCOs, were also generally excluded from the highest positions in the committee system that evolved. The group that emerged in control were the middle section of the Derg, those with some chances of promotion under the ancien regime, but who were blocked by the nobility and Harar graduates above them.
The personnel of the civilian apparatuses were reduced by the political turmoil of the revolutionary process. Many high-ranking members of the bureaucracy fled the country, either in 1974 or later. Yet the majority of the civil service apparently remained in place. The conflict with the EPRP in 1976-1978 led to further clashes within the administration — it was among graduates of universities and secondary school that the EPRP was able to recruit most successfully. In response, the PMAC imposed its own supervision on the ministries and instituted a process of political education for civil servants. Whole ministries participated in public discussions, work outings to the country and education programs designed to induce support for the regime’s goals. The success of these initiatives may be doubted. It was often the politically more courageous or aware civilians who were exposed and dismissed in such procedures, those who had favored the revolution. Those who remained were often those who were less politically active and found it easier to accommodate to the new situation. There was little attempt to transform the hierarchical organization of the civil administration itself, or to bring new personnel into the state machine. The response of the civilian administration was to bow to the inevitable, but to preserve relatively intact its personnel, institutional structure and, it would seem, its political preferences.
A much greater degree of transformation occurred in the provinces. The provincial administration had been less, modernized under the old regime and relied more on the aristocracy and local landholders. The process of social revolution removed many more of these personnel. One of the changes associated with the new regime was the imposition of much more control from the center: there was therefore a need for new systems of administration to be established. Here PMAC and other military personnel played a major role in a context more fluid and more transformed than that in the administrative structures at the center.
Those wielding state power in the post-revolutionary period were drawn almost exclusively from the lower sections of the imperial apparatus and were thus included in the petty bourgeoisie, a term with limited explanatory value given the variety of social categories it encompasses. More important is the fact that the reforms carried out did objectively benefit the social group from which the military leaders were drawn: the land reform strengthened the position of the richer peasants, while the growth of state control in general, and of the military with it, promoted the interests of those who had occupied intermediate positions under the ancien regime. However, given the freedom of decision of the military leadership, and the multiple pressures upon it, it would be mistaken to reduce its policies to mere expressions of class interests. The measures taken by the regime had, initially, contradictory effects and different potential outcomes. These depended both on the conflict within the state over what policies to implement and on the degree of resistance that further measures might encounter from those with greatest social power under the new regime.
The new regime’s relation to the population as a whole was marked by a number of distinctive characteristics. First, this was a far more centralized and interventionist state than the old. Over time the process of political change was evolving toward a new, administratively stronger state. Any solution to the regional and nationalities problem would have to accommodate this reality. Building on the policies of the old regime, the new state took the leading place in the economy, and by the end of the 1970s had erected a system of state controls and centralized planning. Whatever the final outcome, the state will be the main director of economic activity and investment in the coming period.
The second major change was in the relationship to the people. Once the regime was established, few could imagine that Ethiopia was a politically democratic country, in the sense of permitting those at the base to express their views openly or to take autonomous decisions. This was true neither for classes and nationalities nor for individuals. But this authoritarian structure should not obscure the fact that a new system of mobilization and communication had been established, through the political organizations of the regime. They were not based on the old recruiting or operating principles of ethnic particularism and heredity. The central criterion for recruitment was political loyalty — to the regime in general, and to its chairman in particular. Through these mechanisms, very large numbers of people were incorporated into the state system: over 7 million people in the peasant associations, up to half a million in the trade unions, several million in the women’s associations, and several hundred thousand at one time or another in the militia. The extension of social services also helped to strengthen the bonds between government and population: the literacy campaign in particular was designed to bring the people into political life. Politics, however much it was directed from above, was being brought to the population in a way not previously seen in Ethiopian history.
 To some extent, the Church divided like the army: An upper fraction, symbolized by the Abuna Teophilos, was associated with the ancien regime and removed from positions of influence; many of the lower clergy participated in the street demonstrations. According to one account (Süddeutsche Zeitung, April 24, 1980) there were 25,000 priests in Ethiopia, attached to about 20,000 churches. On available evidence, it would seem that clerical resistance has been far less than in comparable situations, and many priests have, as tillers of the land, joined the peasant associations. Religious institutions associated with foreign missionaries have, however, been involved in serious conflicts with the government.
 Hannelore Borel and others, Production, Marketing and Consumption of Potatoes in the Ethiopian Highlands, Institute of Socio-Economics of Agricultural Development, Technical University of Berlin, 1980, pp. 9-20. For details on the new system of producer cooperatives, see Ethiopian Herald, June 28, 1979.
 The following is based on the articles of Pliny the Middle-Aged (pseud.), “The PMAC, Origins and Structure, Part I,” Ethiopianist Notes 2/3 (1978), and “Part II,” Northeast African Studies 1/1 (1978).
 Africa Confidential, January 16, 1980.
 Marina and David Ottaway, Ethiopia, Empire in Revolution (New York: Africana Publishing, 1978), p. 107, and Rene Lefort, Ethiopie, la revolution heretique (Paris: Maspero, 1981), p. 185.
 Authors’ interview with AETU officials, Addis Ababa, December 1977.
 Guardian, May 29, 1978. Reports of further conflict inside the AETU appeared in Africa Confidential, March 11, 1981.
 EPRP policy can be traced through its initial statement issued in 1975, and through the pages of its journal Abyot (Revolution). A cogent statement of the EPRP position is given in Nega Ayele and John Markakis, Class and Revolution in Ethiopia (Nottingham, 1978). It would appear that Markakis later modified his views, to permit greater criticism of the EPRP. See “Garrison Socialism: The Case of Ethiopia,” MERIP Reports 79 (June 1979).
 For details see Amnesty International, Human Rights Violations in Ethiopia, December 1977, and ibid., November 1978. The killings did appear to have ceased by the end of 1978.
 The main document of this plenum is printed in Abyot, special issue, September 1979. Further details in Africa Confidential, January 14, 1981.
 MEISON’s analysis of the conflict is given in Voice of the Masses (August 1977), and in the interview given by Negede Gobeze to Le Monde, September 17, 1977. See also Le Monde, September 3. 1977. The PMAC’s reply is “The Ethiopian Revolution and the Right Opportunists: Stand of the PMAC,” Addis Ababa, May 1978.
 Ethiopian Revolution Information Center, Tasks, Achievements, Problems and Prospects of the Ethiopian Revolution (Addis Ababa, September 1977), pp. 27-28.
 Theda Skocpol, Stages and Social Revolutions (Cambridge, 1979). Chapter 4 elaborates an argument on the particular role of the state in revolutionary situations.
 International Institute for Strategic Studies (London). The Military Balance 1972-1973; and ibid., 1980-1981.
 For example, the Chief of Staff of the Armed Forces in June 1981, Brig. Gen. Haile-Giorgis Habte-Mariam, was not a member of the PMAC.