Is Pakistan’s Army as Islamist as We Think?
New data suggest it may be even more liberal than Pakistani society as a whole.
The preeminent concern among Americans — and increasingly among many Pakistanis too — is that some personnel may support Islamist terrorism in the region and beyond or that perhaps a radical, rogue Islamist column may split off within the Army, endangering Pakistan’s stability and the security of its nuclear weapons. Others fear that radical personnel might even give nuclear devices or technology to terrorists. Equally important is that some U.S. observers equate greater Islamization with deepening anti-Americanism within the Army.
These apprehensions have been galvanized by the numerous terrorist attacks on military and intelligence institutions and personnel that have involved assistance from within the armed forces. Most recently, the May attack on a Karachi naval facility was likely facilitated by an al Qaeda cell within the Navy itself. Chief of Army Staff Ashfaq Parvez Kayani and other Army brass are worried about the enemy within. In recent months, Pakistani Army personnel have been arrested for ties with Hizb ut-Tahrir, an outlawed extremist organization in Pakistan.
But these fears may be overblown. Despite the importance of this issue, few data sources have permitted observers to evaluate these claims with any precision.
Using Army recruitment data from Pakistani analyst Shuja Nawaz, with whom I co-authored a paper in the Journal of Strategic Studies this year describing these data and the changes they evidence, I find no systematic evidence that conservative areas are producing more officers than other areas as late as 2002.
Islam’s Long History in the Pakistani Army
The Pakistani Army has long relied on Islam within the institution. The faith has long served — with varying degrees of success — as a unifying force to supersede ethnic, sectarian, and communal fissures that have long cut through Pakistan’s polity. During the tenure of Pakistan’s second military leader, Gen. Agha Mohammad Yahya Khan, the Army assumed a new role of defending not only Pakistan’s territorial sovereignty but also Pakistan’s "ideological frontiers." To defend Pakistan is to defend Islam.
Islam has also served to motivate the Army to fight an enemy that has always been conventionally superior: India. The Army cultivates deep respect for the military value of jihad, which is evident in its professional literature. The Army uses Islam to bolster its will to fight by debasing the enemy. During the 1971 civil war, Yahya Khan motivated his soldiers by declaring the Mukti Bahini (the Bengali guerrillas) to be a kafir army against which the Pakistani Army was waging a legitimate jihad. Brig. Javed Hassan (who retired a major general), while a faculty member at the Command and Staff College in Quetta, authored a study titled India: A Study in Profile. It is required reading at Pakistan’s National Defense University. Hassan argues that India is "less warlike" than Pakistan and attributes India’s military failures to its Hindu characteristics.
Pakistan’s Army likely needs such motivation against a larger, existential nemesis because — though it has started every war with India — it has never won any of them.
With little hope of defeating India on the battlefield, Pakistan has pursued an asymmetric form of warfare consisting of guerrilla and terrorist attacks under the security of its creeping nuclear umbrella. The Pakistani Army and the intelligence agency it controls (the Inter-Services Intelligence directorate, or ISI) have long instrumentalized Islam to prosecute Pakistan’s interests in Afghanistan and India, using a bevy of Sunni Islamist militant groups and other means to keep the country’s enemies off balance.
Gen. Muhammad Zia ul-Haq, who seized Pakistan’s government in a coup in 1977 and ruled until his mysterious death in a plane crash in 1988, was most notorious in efforts to Islamize the Army. He permitted religious groups to distribute their materials to the rank and file and officers alike. Under Zia, Islamic training was introduced in the curriculum of the Command and Staff College, which provides important training for promising officers of the captain or major rank. South Asia scholar Stephen P. Cohen found long ago that the Army’s professional journals contained numerous essays that studied the question of Islamization of the military and the degree to which the Pakistani Army should achieve greater adherence to Islamic principles. Pakistani military analyst Hasan Askari Rizvi has noted that Zia’s orthodoxy changed recruitment and retentions patterns by ensuring that piety was a part of an officer’s evaluation. Despite Zia’s efforts to Islamize the Army and the state, however, Cohen found that the changes in the officer corps, while important, were in fact quite modest at the time of his research.
Unfortunately, obtaining granular data on this issue is daunting. In principle, one could do surveys, but the sample size would have to be enormous (perhaps as high as 1 million) to identify enough military personnel to say anything meaningful about their views. The Pakistani Army has strictly sought to limit visibility of the organization, and U.S. defense attachés may interact with about 100 officers from an army of 520,000, mostly at the ranks of colonel and above. Attachés may even attend graduation at the National Defense University and do "beard counts," which is an extremely crude measure of Islamism in the military. The few U.S. officers attending the Command and Staff College (one attends each year) have the uncommon experience of meeting those with ranks of majors and captains, but the ISI scrupulously oversees foreign interactions and briefs officers for such meetings on the key positions to maintain. As U.S.-Pakistan relations have been fraught for decades, too little is known about this important institution. Yet understanding it is critical to the bilateral relationship.
What the Data Tell Us
In the absence of ideal data on officers, I did the next best thing: provide insights into the kind of areas that produce officers. To do so, I used the data obtained by Nawaz, which provided the numbers of officers recruited and retired from specific districts for several years between 1992 and 2002. To characterize these districts that produce officers, I used data from Pakistan’s Federal Bureau of Statistics.
Working with several quantitative analysts, I developed a panel dataset of household characteristics (such as employment, educational background, and demographic information) as well as district characteristics (number of hospitals, family planning clinics, and so on in a district). Specifically, this analysis uses data from 1991, 1995, 1998, and 2001 from Pakistan’s Federal Bureau of Statistics and district-level Army data from 1992, 1996, 1999, and 2002. The survey data and the Army data were merged on the one thing they have in common: the district. (The "lag" between recruitment and survey data is deliberate because socioeconomic and other effects should only influence recruitment outcomes in subsequent years.)
To understand how these district characteristics explain differences in recruitment outcomes, we used an analytical technique known as panel regression analysis. This kind of analysis allows us to discern the impact of key variables on recruitment outcomes for any given district and year.
In particular, we examined the effects of demographic, economic, military and "social liberalism" variables. (There is no direct measure for religiosity in the data.) Variables indicating some measure of social liberalism, which are at the core of this study, include the average age of female at first marriage; the numbers of sons in a household in private schools divided by all sons in the household; the average number of family planning clinics in a district; and the difference in educational differences of males and females.
Four key findings emerge from this analysis. First, among the demographic variables, districts with more people who could do basic math were more likely to produce officers. This suggests that foundational human capital matters to the Army, as expected.
Second, districts with more private high schools in a district are less likely to produce officers than those with fewer private high schools. This is consistent with the conventional wisdom that the Army no longer recruits from Pakistan’s elite families.
Third, as in the United States, the presence of retired officers is a strong and significant predictor of recruitment outcomes. This suggests that retirees create positive, pro-military environments conducive to recruitment. Retired officers also can offer helpful advice to aspirants about how best to impress the examiners and otherwise prepare for the rigorous admission process.
Fourth, our analysis of social liberalism characteristics suggests that in many ways more liberal districts are producing officers. One of the most important indicators of social liberalism, the propensity of households to send more sons to private schools, was strongly associated with recruitment in a district. Private-school utilization correlates with socially liberal views in Pakistan. (This appears to conflict with the finding above, but it does not. The numbers of private high schools reflect an aggregated demand for elite education — few Pakistanis attend any high school, much less private high schools — and signifies financial wherewithal to pay for such schools. Taste for private schools at a lower level signifies social liberality, holding district economic factors constant.) Similarly, in districts where men and women have roughly equal educational attainment or in districts where women are more educated, recruitments are more likely than in districts where men’s education exceeds that of women. The presence of family-planning clinics is also positively and significantly correlated with recruitment.
These findings suggest that more socially liberal districts are more likely to produce officers than are socially conservative ones.
The Limits of the Data
The Pakistani Army has used Islam for a variety of institutional and national goals. Elements of the Army have indeed radicalized as evidenced by the rare — but increasing and devastating — terrorist attacks in Pakistan that have involved military personnel. Our findings, however, suggest the Pakistani Army, at least until 2002, was no more likely to recruit from conservative areas of the country, suggesting in turn that perhaps — perhaps — there is less radicalization than is commonly assumed.
Admittedly, these conclusions are tentative, and these measures of social liberalism are no doubt imperfect. This study, moreover, cannot be conclusive as it can only speak to the districts that produce officers, not the worldview of officers themselves. Given the high stakes involved, this subject requires more thorough data collection and analysis. Understanding these dynamics is vital for the United States, but it’s perhaps even more important for Pakistan and Pakistanis who rely on their military to protect their country.