December 1992 witnessed what some have called the “seige of Imbaba,” or the Egyptian government's effort to expel the Islamic militant group al-Gama'a al-Islamiya (henceforth known as the Islamic Group [IG]) from a Giza slum where the group had supplanted other actors as the dominant political presence. It took a force of at least 10,000 police and army troops weeks to root out the IG in a confrontation that was characterized by mass arrests and widespread human rights violations. Not long before, a local IG leader had held a press conference for Western journalists declaring the establishment of the “Islamic Republic of Imbaba.”3 How did it come to be that mere miles from the city center of Cairo, an Islamic militant group was able to create a state-within-a-state that operated largely beyond the reach of the Egyptian government?
The IG in Egypt
Since the founding of the IG in the 1970s, IG activists advocated that Egypt be ruled according to the dictates of Islamic law and eventually came to call for armed confrontation with the Egyptian government for its failure to do so (Sullivan and Abed Kotob 1999). Between 1981 and 1984, planning for the group's future activities largely took place within Egypt's prisons as a result of the group's reported complicity in the assassination of President Sadat (Mubarak 1995). It was decided during this time that the IG would seek to expand its activities to Cairo, seeking a broader constituency than it had already established in its historic strongholds of upper Egypt (Ashour 2009). As IG members began to be released in the mid-1980s, the group initiated the process of both increasing its activities and expanding its geographic scope of influence (Mubarak 1995). Communities in Greater Cairo populated by new migrants from upper Egypt would seem to offer a good opportunity for expansion and cultivation of new members.
The desire to increase the geographic scope of the IG's activities to Greater Cairo came at an auspicious time. The city's growth over the previous 30 years had been fueled by the development of what have been called “haphazard” or “informal” communities—‘ashwa'iyaat—populated primarily by migrants from rural areas (Oldham, El Hadidi, and Tamaa 1987). These areas were known as “informal” because they were developed on agricultural or government-owned desert land in violation of existing laws governing land use and “haphazard” because they did not emerge from an urban planning effort (Bayat and Denis 2000; Oldham, El Hadidi, and Tamaa 1987). Dwellers of these shantytowns, particularly in the 1980s, did not enjoy access to government infrastructure and typically occupied the lower rungs of Egypt's socioeconomic ladder. It seemed that the state had, to some extent, lost control of its capital city, raising issues about both the capacity and the competence of the Egyptian state (Dorman 2007, 24).4
IG Activities in Imbaba
Despite efforts to set up operations in a variety of areas (Bayat 2007a, 39), the IG established its most successful stronghold in the Giza neighborhood of Imbaba. Informal areas of Imbaba—like the poverty-stricken neighborhood known as Western Mounira—had no public schools, hospitals, clubs, sewerage, public transportation, or police station in the mid-1980s, leading Egyptian sociologist Saad Eddin Ibrahim (2002, 75) to describe the area as “Hobbesian” before the Islamists stepped in.5 According to one journalist, the residents of this neighborhood had been “orphaned by the central government” (Abdo 2000).
The IG sought to provide residents with direct assistance for their most pressing economic concerns. According to Ismail (2003, 100) the group provided services through a social work committee with a mandate of helping the poor with health and educational services.6 This included the distribution of food, books, and school supplies for the poor (Hafez and Wiktorowicz 2004; Mubarak 1995, 261). These services also frequently included the payment of school fees, the supplementing of marriage trousseaus, the distribution of meat on Islamic holidays, and support for orphans. The IG would provide money to families upon the death of a breadwinner and free medication.7 Physicians have tended to be very well represented in the IG,8 particularly doctors with hospital privileges who extended their services to local residents.9 As a result, the IG was very effective in its provision of health services to the local population through both primary care clinics and more specialized services, like drug rehabilitation programs.10
The group's activities reached beyond the provision of social services, however, to also include arbitrating conflicts and meting out the group's interpretation of justice. IG activists sought “the control of public space and the imposition of the moral code,” engaging in activities like sex-segregating public gatherings, forbidding music and dance, and encouraging women to veil (Ismail 2003, 80). In some cases, women who refused to veil were threatened and even had acid thrown in their faces (Mubarak 1995, 265). The spiritual directives of the IG were discussed in regularly held meetings and religious training workshops (Hassan 2000). Interestingly, the IG was unique within Egypt as a group that could hold mass meetings and conferences without a state security permit as a result of the autonomy it enjoyed within Imbaba (Hafez and Wiktorowicz 2004, 76). The type and scope of IG activity in Imbaba have ultimately led some to call what emerged a society existing in parallel to the one being administered by the state (Murphy 2002; Singerman 2009). A great deal has been written about the nature of IG activity in Imbaba, but the question of why the IG succeeded in Imbaba, rather than some other area in Greater Cairo, has not been addressed.
Why Imbaba and not Bulaq al-Dakrur?
Over 100 informal communities in Greater Cairo house around 60% of the city's 17 million residents (Kipper and Fischer 2009). Given the size of the city and the large number of ‘ashwa'iyaat as locations of housing for poor Cairenes, how can we understand the success of the IG in some areas and not others? As part of the group's desire to expand its influence from upper Egypt to Cairo, the IG sent group leaders to a number of Cairo neighborhoods (Bayat 2007a, 39; Mubarak 1995, 237–238). This included upmarket areas as well as poor but well-established “popular” neighborhoods (Bayat 2007a, 139).11 The IG enjoyed its greatest success, however, in the crowded ‘ashwa'iyaat.
The informal neighborhoods of Cairo provided a favorable location for IG penetration for a number of reasons. First, the “chronic absence of the state” (Denis 1996) meant that the group would not need to contend with local police and political authority as it sought to extend its influence. In addition, the lack of state-provided services offered an opportunity for the group to win over support from locals who felt neglected by the central government. The informal neighborhoods of Giza provided the additional benefit of being heavily populated by upper Egyptian migrants who may have been familiar with the IG's activities in southern Egypt.
Demographic and socioeconomic conditions alone, however, cannot predict the IG's successful establishment of local hegemony in Imbaba as a number of areas would have met the criteria outlined above. In particular, the informal neighborhood of Bulaq al-Dakrur resembled Imbaba on many of these dimensions. Bulaq al-Dakrur—like Imbaba—is an informal area located geographically close to the city center but peripheral to the city in terms of “social distance” (Ismail 2006, 7). Ismail argues that residents of Bulaq al-Dakrur have historically believed either that there is no state presence in their area or that they live in a state-within-a-state, though not one run by Islamists (2006, xxxiii).12 She contends that Bulaq al-Dakrur and Imbaba were highly comparable in socioeconomic terms, even quoting one informant who argues that “Bulaq can only be compared to Imbaba. The two areas compete in negatives” (Ismail 2006, 21).
Information collected about the two neighborhoods as part of the 1986 Egyptian census also suggests a number of demographic similarities between Bulaq al-Dakrur and Imbaba. The neighborhoods had almost identical percentages of young and old residents, students and self-employed, illiterate and more educated individuals, for example. Bulaq al-Dakrur and Imbaba were also much more similar to each other on social and demographic indicators than to neighboring areas like Agouza and Dokki. Despite these important points of similarity between the two neighborhoods, the IG only emerged as the key local arbiter and authority in Imbaba, whereas in Bulaq al-Dakrur governance was managed by local clan leaders and strongmen who were not explicitly religious in their orientation (Ismail 2006, 56).13
This raises the question of whether underlying levels of religiosity made Imbaba more susceptible to Islamist influence than Bulaq. Yet there is no historical evidence to suggest that the residents of Imbaba were more religious than individuals in other informal areas surrounding Cairo. Bulaq al-Dakrur was known to have housed one of the first cells of the IG, with cells only later spreading to areas like Imbaba and Haram in the Giza governorate (Morour 1990). Assuming that the group would first settle in areas where it was best connected and most likely to succeed, this suggests that Bulaq al-Dakrur was at least as religious as Imbaba. In addition, both Imbaba and Bulaq al-Dakrur were home neighborhoods to more than 10 defendants in the Sadat assassination trial (Kepel 1993, 222). If the number of Sadat assassination defendants might be taken as a proxy for the underlying religiosity of an area, then Imbaba and Bulaq al-Dakrur, again, are highly comparable. Finally, socioeconomic factors have long been thought to be correlated with but not deterministic of support for Islamist groups and practices. Literacy, for example, has been shown to be a robust determinant of support for fundamentalism among Muslim women (Blaydes and Linzer 2008). Yet Imbaba and Bulaq al-Dakrur had virtually identical rates of female literacy leading up to the late 1980s.
Individuals well positioned to speak about the social and political development of Giza governorate have argued that although both Bulaq al-Dakrur and Imbaba were characterized by comparable levels of physical and social inaccessibility for the areas' citizens, the neighborhood of Imbaba posed a unique geographic challenge for the state and security services. The vast majority of informal areas in Greater Cairo are characterized by narrow streets, minimal public space, and extremely high population density (Ismail 2003, 92). The frequent lack of established street names, house numbers, and accurate maps of these areas offer possibilities for “safe haven” from state authorities (Bayat 2007b). Yet Imbaba posed a particular challenge. According to Fouad Allam, the Egyptian official tasked with cracking down on Islamist organizations in Egypt during the 1980s and 1990s, the physical space of Imbaba was harder to control than other informal areas because of the many dead-end and one-way streets found there.14 Suhair Lotfi, a long-time researcher at the National Center for Criminological and Sociological Research located in Imbaba, concurs arguing that the physical geography of Imbaba was particularly inaccessible for state security and that the IG eventually invested more resources in Imbaba because it would be easier to defend against state authorities.15 According to Lotfi, the geography of Imbaba resembles a “citadel” with few points for vehicle entry and exit and many one-way roads and narrow alleys.16
This view is echoed in both scholarly and journalistic accounts. Imbaba came to be favored by Islamists because of its “invisibility” (Ismail 2003, 99) where the possibility of police interference would be minimized (Mubarak 1995). It was this ability to expand without fear of security interference, according to one IG representative, that provided the group with the opportunity and incentive to grow its activities in Imbaba (Mubarak 1995, 247). Streets in Imbaba were also very narrow and unpaved,17 not serviceable by police vehicles,18 and sometimes maze-like where a single narrow alley might run for hundreds of meters with dozens of tributary alleys.19 Satellite images of Imbaba from 1976 show a densely populated quarter with very few cross-cutting roads, making it virtually impossible for emergency vehicles to access (El-Sioufi 1981).20
The difficulty of police penetration of Imbaba became particularly clear after the December 1992 raids that swept the IG from its hegemonic position. According to one expert, many areas of Imbaba were highly inaccessible to the armored and other vehicles of state security.21 Although the security officials were eventually able to root out IG activists, the process took longer than anticipated because of the geographic conditions of Imbaba.22 Is it possible that the geographic characteristics of Imbaba impacted the ability of women from the area to receive health, educational, and other services? One-way streets and narrow roads would not have impacted the ability of pedestrians to move both within and to areas outside of the slum neighborhood. Indeed, residents of either of the two informal areas travel on foot to leave their neighborhoods and then by city bus, metro, or perhaps most frequently privately owned microbus to their destination. To move from the informal areas of Imbaba and Bulaq al-Dakrur to areas with more social services, individuals would need to cross railroad tracks and traverse other types of obstacles in both cases.23 As a result, it is does not appear that the types of geographic impediments to state sector intervention would impact ordinary citizens in the same way. Exploiting this difference between Imbaba and Bulaq al-Dakrur is part of the identification strategy I employ.