With varying motives, using the newest weapons, engulfing a wider area, and claiming more lives, terrorism in Pakistan has attracted considerable local and global attention. Existing literature covers many topics but no study specifically addresses temporal and spatial incident patterns, patterns in weapon use and target selection, and distribution of terrorist types. Lack of comprehensive analysis limits our ability to formulate and test hypotheses, to do operational planning, and to develop counterterrorism policy.
Using data from the GTD, this paper describes empirically the incident patterns of terrorism in Pakistan from 1974 to 2007 in a multidimensional way: temporal and spatial patterns, patterns in target types, weapon types, terrorist types and the patterns pre and post U.S-led invasions of Afghanistan and Iraq. The study offers new insights about the measurement of terrorism, the definition of the terrorists, the cyclical nature of terrorism, the role of conflict, the choice of weapon, the sponsorship of terrorism, the selection of targets, and the dynamic and reactionary nature of terrorism.
Theoretical, operational, and policy implications of the findings are discussed. KEYWORDS: Terrorism in Pakistan, spatial patterns in terrorism, temporal patterns in terrorism, terrorism and conflict, terrorism target selection, terrorism sponsorship, Syed Ejaz Hussain (Ejaz) is a deputy inspector general from Pakistan Police, currently a Ph.D. candidate at the University of Pennsylvania working with Professors Lawrence Sherman, Randall Collins, and John MacDonald. His research interests include terrorism, policing, and criminological theory.
The Review on Impact of Terrorism on Financial Markets of Pakistan
Since the emergence of 9/11 event in U.S.A., the world has witnessed significant increase in terrorist activities. Pakistan unfortunately has been the most vulnerable to the paradigm shift of terrorism. Its vulnerability is mainly subjected to its geographical location. Nonstop terrorist activities toppled over the country’s political, social and economic structure. Now the life of a normal ...
TERRORISM IN PAKISTAN: CHANGING INCIDENT PATTERNS
3
With varying motives, using the newest weapons, engulfing wider area, and claiming more lives, terrorism in Pakistan has attracted considerable local and global attention. Consequently, a large literature has been published on root causes of terrorism in Pakistan ; Pakistan’s Jihad Culture ); Madrassa Culture ; sectarian terrorism ; Pakistan’s counterterrorism efforts ; Pakistan as sponsor of terrorism ; and Pakistan’s performance as an ally of the U.S. in its war against terrorism ().
No paper specifically studies temporal and spatial patterns, patterns in weapon use and target selection, and distribution of terrorist types. If data were presented in publications (see Barshied, 2005: 21; Grare, 2007: 139), they typically just give a short-term view of annual comparisons or rates based on press or government analyses. This lack of comprehensive analysis of terrorism incidents limits our ability to formulate and test hypotheses, to do operational planning properly, and to develop counterterrorism policy.
Trying to fill in this gap, this paper describes empirically the changing incident patterns of terrorism in Pakistan from 1974 to 2007 and attempts to answer the following four questions: What is the temporal distribution of terrorist incidents in Pakistan? What is the spatial distribution of these incidents? What is the temporal and spatial distribution of target types, types of weapons used, terrorist groups involved, and the number of people killed and wounded? And what are the patterns of terrorism incidents pre and post the US-led invasions of Afghanistan and Iraq?
The rest of the paper proceeds as follows. It begins by describing the Global Terrorism Database (GTD) as its data source, defines the variables analyzed, and explains the analytical methods used. The paper then presents results on temporal patterns, spatial patterns, and distributions of victim types, weapon types, and terrorist group types. Results also compare and contrast these patterns, pre and post U.S-led invasions of Afghanistan and Iraq. Contrary to the
The Term Paper on The Crisis of Militant Groups and Terrorism in Pakistan
In recent years Pakistan’s security concerns have been rising as its major cities have been rocked by rising suicide attacks. It has emerged that several new terrorist groups have come up, with several other established groups reconstituting themselves. Further, a more violent crop of militants that is disinterested in political solutions has emerged. Experts also say that these groups have formed ...
4
usual format of an empirical article where the discussion section is separate from the results, here discussion is presented in the results section. There is a reason behind this choice. The results section is lengthy, comprising six sections; postponing discussion until the end would have required the readers to retain a lot of information waiting for explanation. The paper then concludes with a summary of the findings and their importance for theory, policy, and operations.
DATA AND METHODS
DATA
Initially, I proposed using data from three sources: the Global Terrorism Database (GTD), the South Asia Terrorism Portal (SATP)1, and the Pakistan Police forces’ records2. After reviewing the tradeoffs of the GTD, the SATP, and the Pakistan police, I used the GTD data only for the following reasons. First, the GTD is an open-source database maintained by the National Consortium for the Study of Terrorism and Responses to Terrorism (START) at the University of Maryland and is the world’s largest database on terrorism incidents. It covers period from 1970 to 2007 and includes 81,641,000 cases, globally.
Second, the GTD has a vast amount of information on 120 variables. This information enabled me to collapse some of the categories and verify information on the others. Third, the GTD is supervised by academicians trained in 1
South Asia Terrorism Portal (SATP) is the biggest open source web-based data available in the South Asian region maintained by the Institute for Conflict Management, India. Information is derived from newspapers and is presented
mainly in the form of descriptive news arranged chronologically. 2
Since 2006, Federal Investigation Agency (FIA) has started compiling data at national levels. FIA has formatted the data including variables of date and location of the incident, the weapons used and the target, the number of killed and wounded, and motive. FIA data would be the most reliable for the time period they are available.
The Essay on Business Intelligence And Data Warehouses
Difference between the structure of database and warehouse transaction Database is designed to make transactional systems that run efficiently. Characteristically, this is type of database that is an online transaction processing database. An electronic strength record system is a big example of a submission that runs on an OLTP database. An OLTP database is typically controlled to a single ...
5
the technicalities of research, and the SATP and FIA data are supervised by police officers—in the case of the SATP, a former police officer. Fourth, the GTD data are in a ready to use format and available on CD, while FIA data were procured through request to the senior officers and the SATP data are not formatted. Fifth, the main audience of this paper is the academic community and they recognize the GTD well and are confident of its veracity. Many articles have already been published using the GTD data (see, LaFree et al., 2009).
Despite all these advantages, the GTD has to be used with two warnings.
First, the data were lost for 1993. Although the GTD has reconstructed the total number of incidents, people killed and wounded, details of the incidents are not available below the yearly aggregates. Therefore, analyses below year levels encounter the problem of missing values. As the missing values are less than 2% in my analysis, I deleted them case-wise. Second, as the data were originally compiled from the newspapers, the possibility of media bias cannot be ruled out. Variables
Terrorist Incident. The GTD defines terrorism as the “threatened or actual use of illegal force and violence to attain a political, economic, religious or social goal through fear, coercion or intimidation.” Its application is confined to non-state actors. The word “incident” is defined in the American Heritage Dictionary as “a definite and separate occurrence; an event.” The GTD codebook (2009) considers an incident as single if occurring in both the same geographic and temporal point.
Time and Location. The variables of time include day of the week, day of the month, month, and year during 1974-2007. The choice of 1974 as the beginning year is deliberate. Pakistan in its current geographic form exists only since 1971, when East Pakistan separated as Bangladesh and 1974 is the first year in which a terrorist incident for Pakistan is recorded in the GTD. Pakistan, for historical and administrative reasons, is divided into four provinces and three federally administered areas as shown in Figure
1. The four provinces include the Punjab, Sindh, North West Frontier Province (NWFP), and Baluchistan, and the federal areas include Islamabad (capital city), Federally Administered Tribal Areas (FATA), Kashmir, and the Northern Areas. These provinces and federal areas are further subdivided into 121districts and seven agencies, the administrative equivalent of a district. Districts are the basic administrative units within the provinces. All federal areas are combined as Islamabad in Figure 1. Figure 1. The Administrative Divisions in Pakistan
The Term Paper on Extermism in Pakistan
Who was Shahzab and why he was murderd? Twenty-year old Shahzeb Khan was in dark grey shirt and trousers, the clothes he wore to the Valima of her sister when 9mm bullets pierced through his car and entered his body, which eventually killed him. He was the only son of his parents. That was December 25, 2012. The incident was a result of a small argument. Shahzeb Khan raised his voice when his ...
PAKISTAN
PUNJAB
SINDH
BALUCHISTAN
NWFP
FATA
ISLAMABAD
DISTRICTS
DISTRICTS
DISTRICTS
DISTRICTS
AGENCIES
DISTRICTS
36
23
27
24
7
15
The GTD provides names of the location either as name of the district, city, village, or province—whatever is available. By tracing district names from village and city names, I compiled information on the variable of the district name. To tabulate data on the province level, I extracted the province name from the district name or the province/state name variable already present in the GTD.
7
Target Types. The GTD provided 22 categories of target types. On the basis of their commonalities, I collapsed these categories into 11 categories. For example, categories of food and water supply, telecommunication, utilities were collapsed to get a category of infrastructure and facilities. The eleven categories are listed as: private citizens, property, and businesses; infrastructure and facilities; civil administration; military and paramilitary; police; sectarian; foreigners; educational institutions; political leaders, meetings, activists; civil society institutions; music and barber shops; and minorities.
Civil administration means government officials excluding officers in uniform; military includes army, air force, navy and paramilitary forces; infrastructure means gas pipelines, train tracks, and government buildings; sectarian targets include Shia or Sunni individuals, worship places, and congregations; minorities include Christian, Hindu, and Ahmaddiyya (a Muslim sect declared non-Muslim under Pakistani law in 1974).
Barber shops, music, and CD shops were combined to make category of music and barber shops. Civil society institutions include media and non-governmental organizations.
Weapon Types. By reorganizing some of the weapon categories provided in the GTD, I came up with the following types: Firearms; explosives, bombs, dynamite; incendiary; melee (hand to hand fights or with blunt weapons); projectiles (rockets, rocket propelled hand grenades, mortars); suicide bombers; and others. I made a separate category of suicide bombers and projectiles, taking them out of explosives in general to study them in detail because of their current usage in Pakistan.
The Essay on Terrorism Groups Government Bombing
28 October 2000 Domestic Terrorism On The Rise Domestic terrorism has risen sharply over the past 10 years. According to a recent Bureau of Alcohol Tobacco and Firearms report, bombings or attempted bombings increased from 2, 098 in 1990 to 3, 199 in 1994, a 52% increase. Property damage from bombs rose to $7. 5 million, with 308 people injured and 31 killed. The Oklahoma City bombing, which is ...
Group Types. The GTD shows 82 terrorist groups involved in terrorism in Pakistan but instead of using their names, I classified these groups on the basis of their motives: sectarian, ethnic, political, Muslim militants, foreigners, and Al-Qaeda. It simplified the analysis, and furthermore it was not the group name but the group motive which mattered. For example, sectarian
8
terrorists in Pakistan used different names—Sipah-e-Sahaba, Lashkar-e-Jhangvi, Lashkar-eOmar, Group of 313, Sipah-e-Muhammad—but their motive was to kill people on a sectarian basis. They used different names for personal, operational, or political reasons. Therefore, I combined groups with the same basic motive as one group. Al-Qaeda, because of its importance, needed a separate category. So I kept it separate from the categories of Muslim Militants and Foreigners, although by definition it could be included in either of them. Number of People Killed and Wounded.
The GTD provides the number killed and wounded in a terrorist incident including terrorists; I counted the victims only. To investigate the
severity levels of terrorist incidents, I computed the number of people killed and wounded per incident. ANALYSES
I analyzed the data by using descriptive statistical methods, univariate time series analysis, and the Geographical Information System (GIS).
Descriptive Statistics
Descriptive statistical methods used here include, mean, median, standard deviation, and graphs.
Univariate Time Series Analyses
Time series are sequences of observations ordered in time, and they have four components: trend, cycles, seasonality, and irregulars. This trend component is the long term movement in a series, whereas cycles are regular fluctuations which are not seasonal. Seasonality
9
is a component of a series which is dependent on time of the year and irregular is what is left over once the other three components of the series have been accounted for. Autoregression in time series occurs when observations in a time series at different points of time are correlated. Spatial Analyses
The Essay on International Terrorism America World Americans
After reading Martha Crenshaw's essay on The Causes of Terrorism, thoughts were provoked that allowed me to look at the causes of terrorism from an entirely new perspective. It took my preconceived notions and feelings towards terrorism out of the picture and really let me look at the subject from a balanced point of view. I found this appealing because I have never been able to comprehend how ...
For locating patterns in the geographical distribution of incidents, the best method was to use coordinates as the space point, which in this case were not available. The best alternative available in the present case was to use the district as the geographical unit to show the geographic spread of incidents. I used geographic information system (GIS) base layers for Pakistan and Asia from DIVA-GIS and used ArcGIS to make GIS maps.
RESULTS
According to the GTD, in 34 years of terrorism in Pakistan, 2590 incidents took place in which 5840 people were killed and 11597 wounded. As shown in Table 1, the average number of incidents remained at 76 per year, with an average of 173 people killed and 341 wounded. The people killed and wounded per incident were 2.25 and 4.48 respectively.
The median number of incident was 36, with a median number of 106 people killed and 261 wounded. The statistics of standard deviation, and the minimum and maximum in Table 1 show high dispersion in the distribution of the number of incidents, killed, and wounded. Table 1. Summary of Terrorism Incidents, People Killed and Wounded in Pakistan (1974-2007)
Variable
Sum
Mean
Median
Standard
Min
Max
10
Deviation
Incidents
2590
76
36
124
0
665
Killed
5840
172
106
251
0
1273
Wounded
11597
341
261
449
0
2385
TEMPORAL PATTERNS
Before presenting and discussing the time patterns in terrorism incidents, a brief review of the macro dynamics of terrorism in Pakistan would facilitate the readers’ understanding. At its independence from the British in 1947, Pakistan had diverse population in terms of ethnicity and religious sects. Ethnic groups included, Bengali, Punjabi, Sindhi, Baluchi, and Pathan and the Muslim sects comprised Shias and Sunnis, further subdivided into many groups based on their theological differences.
This diversity resulted into the secession of East Pakistan (1971), and the insurgency in Baluchistan (1970-1974).
Evolution of terrorism in its present form is the result of five internal and external factors intersecting together over the years. Please see Figure 2. First, General Zia-ul-Haque came to power in 1977 after toppling and arresting an elected Prime Minister Mr. Zulfiqar Ali Bhutto. Mr. Bhutto was later executed by General Zia in 1979. A direct result of this execution was the formation of a terrorist group named Al-Zulfikar. The group committed acts of terrorism, including hijackings.
Throughout his life, General Zia feared the resurgence of Mr. Bhutto’s Peoples Party. Sindh was Bhutto’s resident province, and hence their stronghold was a nightmare for Zia. To break Bhutto’s influence in Sindh, he helped form Muhajir Qaumi Movement (MQM) (Kukreja, 2003: 143), a language-based party in Karachi and the other urban areas of Sindh. Violence by and against the MQM, which passed through certain
11
evolutionary stages became responsible for almost 90% of terrorism in the cities of Karachi and Hyderabad and 40% in the country. Second, the Iranian Revolution (1979) established its influence in the Shia community of Pakistan. Consequently, Shias demanded exemption from the Zakat & Usher Ordinance (1980) which was based on the Sunni jurisprudence. They also required the government to enforce Shia jurisprudence in their personal matters.
To counter the Shia ascendance, General Zia helped Sipah-e-Sahaba Pakistan (SSP) (Haqqani, 2006), an anti-Shia Deobandi religious organization. Coincidentally, the SSP was further supported by the Iraq-Iran war of 1980’s, as the parties to the war made Pakistan a proxy theater (Fair, 2004: 104).
The SSP was funded by both Iraq and Saudi Arabia (Stern, 2000: 124) making it a “cash rich organization” (Kamran, 2008: 80).
The SSP, later on, gave birth to militant splinter groups, the most dangerous and notorious of them Lashkar-e-Jhangvi (LeJ).
Almost 30 % of terrorism in Pakistan is traced back to the SSP or its subgroups, and the Shia groups like Sipah-e-Muhammad (SMP).
Therefore, language and sectarian motivations explain about 70% of terrorism in Pakistan. Third, the Soviet Invasion of Afghanistan (1979) prompted the U.S. and Saudi Arabia to invest nearly six billion dollars (Weiner, 1998) in the region to organize, train, and arm fighters against the Soviets. Thousands of Muslim fighters were brought to Afghanistan and Pakistan in the name of jihad. American and Saudi money pushed the Mullahs in both Pakistan and Afghanistan to prominence3, established Madrassas as sanctuaries of jihad culture, and made carrying a Kalashnikov an honor in the society. In addition, it strengthened General Zia against the Peoples Party and Shias.
Figure 2. Terrorism Incidents in Pakistan (1974-2007)
3
($6-$8 billion were distributed to them: Bhatia and Sedara quoted in Shuja et al., 2009: 14)
800
12
Afghanistan Invasion
600
Soviet Invasion
Incidents
400
Soviet Withdrawal
200
Iraq Invasion
0
Iranian Revolution
1970
1980
1990
Year
2000
2010
Data from Global Terrorism Database, University of Maryland
Fourth, after the Afghan Jihad against Soviets ended in 1989, the U.S. left the region with thousands of battle-hardened, mobilized mujahedeen, 80, 000 of them in Pakistan (Rashid, ***).
This void created by the US leaving the region without demobilizing the Mujahedeen according to Shuja et al. (2009: 10) drew Inter Services Intelligence (ISI) into action. It tried to divert mujahedeen in two directions: to create strategic depth in Afghanistan (Rashid, 1999: 118) and to support the struggle in Indian-held Kashmir.
With the help of the ISI, mujahedeen (Taliban) came to power in Afghanistan, and they gave refuge to Arab mujahedeen like Osama bin Laden, their erstwhile companions in jihad. After the establishment of Taliban government in Afghanistan, two important factors led mujahedeen to challenge the United States. First, the U.S. created grievances in them by ignoring mujahedeen once the Soviets withdrew. Second, the success of the mujahedeen against the Soviets—a super power—boosted their confidence. Arab
13
mujahedeen had their own grievances especially the presence of the U.S army in Saudi Arabia. Mujahedeen, ultimately, challenged the U.S. in the form of 9/11 which brought the U.S. again to Afghanistan now not as an aide of mujahedeen but as an enemy. Fifth, the U.S. invaded Afghanistan in October 2001 and the Government of Pakistan assisted. Pakistan as a result bought enmity of Arab mujahedeen and the Taliban.
Terrorism nowa-days in NWFP, FATA, and some parts of the Punjab is a direct result of Pakistan’s confrontation with the Arab Mujahedeen and Taliban. To further aggravate the situation, drone attacks inside Pakistani territory have created “a backlash among the tribesmen and even among the general population of Pakistan (Shuja et al., 2009, p15).
Once the Northern Alliance with the U.S. help formed their government in Afghanistan, India was allowed to establish four consulates and one embassy in a small country like Afghanistan. Foreign Minister of Pakistan alleged India to be sponsoring race-based terrorism in Baluchistan4 .
These allegations against India are confirmed by C. Fair from Rand Corporation when she said: “Indian officials have told me privately that they are pumping money into Baluchistan” (2009).
The weapons used against the Pakistan army in Swat operation were coming from Afghanistan and these were U.S. made5 which U.S says were stolen from the arms supplied to Afghan forces. The U.S. is also concerned about China building Gwadar port in Baluchistan which China is believed to use this port “to project force and undermine U.S. and regional security.”6
4
http://www.nation.com.pk/pakistan-news-newspaper-daily-english-online/print/Politics/23-Nov-2009/Indiafuelling-terrorism-in-Pakistan-Qureshi accessed on January 4, at 12:06 p.m. 5
http://www.dawn.com/wps/wcm/connect/dawn-content-library/dawn/the-newspaper/front-page/stolen-us-armsbeing-used-in-swat-ispr-059 accessed on January 2, 2010 at 2:48 p.m. 6
http://www.vqronline.org/articles/2007/spring/schmidle-waiting-for-worst/ accessed on 01/05/10 at 6:46 p.m.
14
Now, I present results on time patterns of terrorism incidents in Pakistan at
three levels: year, month, and day of the week. Then I compare and contrast these results with the world data from the GTD except Pakistan.
Yearly Patterns
Figure 2 mapping the outbreaks of terrorism onto other international events depicts that Pakistan has seen five peaks of terrorist violence since 1974. The first peak is located in 1987 with 60 incidents out of which 22 occurred in NWFP alone and 11 in Islamabad. The GTD shows them as incidents committed by unknown terrorists. The location and the year of the incidents strongly suggest their link with Afghan secret agency or the Soviet secret agency— KGB to put pressure on Pakistan for its role in Afghan Jihad against the Soviets.
The second peak is visible in 1991 with 150 incidents out of which 77 incidents occurred in Sindh followed by the Punjab with 34 incidents. This peak was the result of two streaks in terrorism in Pakistan running together: language-based and Sectarian terrorism. The third and the highest peak in Figure 2 is seen at 1995 with 665 incidents out of which 630 incidents occurred in Sindh alone. It was mainly language-based terrorism confined to Karachi and nearby city Hyderabad. The fourth peak in occurred in 1997 with 206 terrorist incidents. These were mainly sectarian and mostly in the Punjab. However, there were some language based incidents in Sindh as well.
Presumably, because of strict government actions against sectarian organizations and language-based terrorists in Karachi, the situation improved and there was some peace time from 1998 to 2003. After six years of pause, terrorism resurfaced in 2004 with more power and revenge this time in two forms: Taliban in NWFP and FATA, and Ethnic (Race) groups in Baluchistan. Out of 256 incidents in 2007, 133 incidents occurred in NWFP and 64 in the adjacent FATA. Baluchistan also has 34 incidents.
15
There is a correlation of 0.6 between terrorism in Pakistan and the world, as shown in Figure 3, suggesting an association between the two. Cross correlogram of Pakistan and the world incidents shows a maximum correlation of 0.5 at Lag 3 indicating the lagged effect of world incidents on Pakistan after three years.
Incidents
2000
3000
4000
5000
Figure 3. Terrorism in Pakistan and the World7
0
1000
Spearman’s rho=.6
1970
1980
1990
yearly
World Minus Pakistan
2000
2010
Pakistan
Monthly Patterns
Table 2 giving a month wise break up of incidents over 34 years (1974-2007) demonstrates that the months of May, June, July, August, September, and October remained the top six months. In case of the World, May, March, June,
July, August, and October were the
7
Terrorism incidents in the world except Pakistan
16
leading months. Chi-square goodness of fit test indicates statistically significant differences in months. And all the summer month on the top give a clue to the presence of some seasonality. Table 2. Monthly Distribution of Terrorism Incidents in Pakistan (1974-2007)8 Month of Incident
Frequency (N=2489)
Percent (N=2489)
6
302
12.13
5
238
9.56
9
233
9.36
10
231
9.28
8
226
9.08
7
214
8.60
Scholars for a long time have been interested in the twin questions of whether there exists seasonality in crime and if yes why is it there—because of temperature influences on criminals or availability of more victims. Research generally supports routine activity theory—availability of more victims due to good weather and long hours—in explaining more crime in summer months (Morken and Linkar, 2000; Landau and Fridman, 1993; Hird and Ruparel, 2007).
Day of the Week
As shown in Table 3, Monday is the leading day of terrorism in Pakistan with 20% incidents followed by Sunday 15% and Saturday 14%. Friday remained the day with the minimum 11% of incidents. In case of the world, Monday is also the leading day of terrorism with 15% incidents. However, Saturday and Sunday are at the bottom with 13 and 12 percent
8
Only top six months are shown.
17
incidents respectively as shown in Table 3. Chi-square goodness of fit tests indicate statistically significant differences in days of the week for Pakistan and the world. Terrorism patterns on days of the week could also be explained with the routine activity theory—availability of more victims on Mondays and less on holidays. Friday has the minimum incidents in case of Pakistan and Saturday and Sunday in case of world. One commonality between Friday, Saturday, and Sunday is the concept of holiday. Friday is generally a day of peace, rest, and prayers for Muslims and remained a weekly holiday in Pakistan for many years. Saturday and Sunday are generally the weekly holidays all around the world.
Terrorist incidents are occurring less on Friday, Saturday, and Sunday because terrorism generally is against the private citizens and the government institutions as victims, and these are available less on holidays. Reemergence of terrorism on Monday makes sense because after holidays, it’s the first day of the week, more activity is expected.
Table 3. Incident Patterns on Day of the Week—Pakistan and World Day of Week
Monday
Sunday
Saturday
Wednesday
Thursday
Tuesday
Friday
Total
Pakistan %(N=2487)
19.90
15.36
14.52
14.11
12.67
12.10
11.34
100
World %(78447)
15.36
12.92
12.49
15.22
15.06
14.91
14.05
100
18
Trend, Seasons, Cycles, and Autoregression
Correlogram of daily data in Figure 4, shows that incidents on lag 1, lag 2, and lag 3 show a correlation of more than 0.5. Then again stronger correlation emerges after every 14 days. One explanation could be that the terrorists make their preparation, preempt, and commit acts of terrorism for some days. In the meantime, the law enforcement agencies react and the terrorists go into hideouts to come again after 14 days once the law enforcement activity slows down.
Autocorrelations of Daily Incidents (Moving Averages)
0.00
0.50
1.00
Figure 4. Autocorrelation of Daily Incidents
0
20
40
Lag
60
80
100
Bartlett’s formula for MA(q) 95% confidence bands
Figure 5, showing weekly data, indicates that significant strong pulses are felt at week 14 and 28, suggesting a more or less 14 week seasonal pattern. Autocorrelation of observations suggests the same boom and bust pattern of violence as discovered in daily data. The stronger return of
19
autoregression after 14 weeks suggests a seasonal pattern of about three months confirming the findings as suggested in tabulated data in Table 2.
-0.50
Autocorrelations of Weekly Incidents
0.00
0.50
1.00
Figure 5. Autocorrelation of Weekly Incidents
0
10
20
Lag
Bartlett’s formula for MA(q) 95% confidence bands
30
40
20
0
50
Incidents
100
150
200
250
Figure 6. Trend and Cycles in the Number of Incidents
1975q1
1980q1
1985q1
1990q1
1995q1
Quarterly
Incidents
Trend
2000q1
2005q1
Cycles
Figure 6 presents fluctuations of more than 3-month duration. Trend line over 34 years shows upward movement. Whereas polinomial of the order 11 shows 10 turning points and five cycles of varying duration. Figure 7 of number of killed and wounded shows a similar pattern in terms of rising trend but cycles show a difference—an extraordinary rise at the end of the graph indicating beginning of a very pwerful cycle.
Figure 7. Trend and Cycles in Number Killed and Wounded
0
Number Killed & Wounded
500
1000
1500
21
1975q1
1980q1
1985q1
1990q1
1995q1
quarterly
2000q1
Number Killed & Wounded
Trend
2005q1
Cycles
Temporal Patterns in Number of People Killed and Wounded
Figure 8 graphs quantum of terrorism in Pakistan in a different way—by counting the number of people killed and wounded in terrorist incidents excluding the terrorists themselves. Here the patterns of terrorism in Pakistan seem different; the graph instead of showing a peak in the middle at 1995, shows peaks at 1987, 1995, and 2007, a different pattern from the incident graph in Figure 2.
Figure 8. People Killed and Wounded in Terrorism in Pakistan (1974-2007)
0
Number Killed & Wounded
1000
2000
3000
4000
22
1970
1980
1990
Year
2000
2010
Depicting the distribution of severity levels—defined as the number of people killed and wounded per incident—Figure 9 makes it clear that more people were killed and wounded per incident in 1980 (35), 1984 (20), and 1987(20) than in 1995 (2), the peak year in respect of the number of incidents. Reaching at 2 persons wounded or killed per incident in 1995, the graph has taken an upward turn and is averaging about 15 persons killed and wounded per incident in 2007.
Figure 9. People Killed and Wounded Per Incident (1974-2007)
0
Killed and Wounded per Incident
10
20
30
40
23
1970
1980
1990
Year
2000
2010
The patterns based on the analysis of number killed and wounded in terrorism (magnitude) are different from patterns based on the analysis of frequency (number of incidents).
One possible explanation for this difference is the weapons’ different degrees of effectiveness. This finding urges us to measure terrorism not in terms of incidents only but in terms of number of people killed and wounded also. Might be there is a need to develop some index for measuring terrorism.
SPATIAL PATTERNS
In the following section, I display the spatial distribution of incidents at province levels, at district levels, and then at country level, the three important hierarchical administrative units in Pakistan.
24
Geographical Information System (GIS) map in Figure 10 presents the geographical spread of terrorism in Pakistan. Terrorism incidents mostly occurred in the four provincial Figure 10. Geographical Spread of Terrorism Incidents in Pakistan (1974-2007)
capitals—Karachi (1211, 49%), Peshawar (133, 5%), Lahore (109, 4%), and Quetta (82, 3%) and the country capital Islamabad (69, 2.7%).
Some of the rural districts also attracted more than twenty terrorism incidents as shown by the large number of dots at some places. In addition to the capitals, dots concentrate in FATA, Swat, Southern Punjab, and the central Baluchistan.
25
To show the geographical distribution of terrorism in different provinces and federally administered regions of Pakistan, I made bar graphs as presented in Figure 11. Figure 11. Spatial Distribution—Provinces and Districts (1974-2007)
The bars clearly show that the province of Sindh has the maximum number of
incidents (1338, 54%), followed by Punjab (347, 14%), NWFP (280, 11%), Baluchistan (242, 10%), and FATA (155, 6%).
Islamabad Capital Territory, Northern Areas, and Kashmir combined as Islamabad had the minimum incidents (87, 4%).
To look at the spatial distribution of incidents below country and province levels, I have shown the incidents at district levels in Figure 12. Top 20 districts are shown with Karachi at the
26
top with 1211 incidents—48% of the total incidents in the country and 90 % incidents in the province of Sindh. Karachi is followed by Peshawar, Lahore, Quetta, and Islamabad—all are capitals. Some other peripheral districts which also
Figure 12. Spatial Distribution—Top 20 Districts (1974-2007)
Spatial Distribution: Top 20 Districts
D.I. Khan
Kurram
Bannu
Sheikhupura
Gujranwala
Bolan
Bajaur
Jhang
Multan
Swat
Dera Bugti
S. Waziristan
Rawalpindi
Hyderabad
N. Waziristan
Islamabad
Quetta
Lahore
Peshawar
Karachi
15
15
17
17
17
19
19
21
26
33
36
44
46
48
67
69
82
109
133
1211
0
500
Incidents
1,000
1,500
attracted higher number of incidents included North Waziristan, Hyderabad, Rawalpindi, South Waziristan, Dera Bugti, Swat, Multan, Jhang, Bajaur, Bolan, Gujranwala, Sheikhupura, Bannu, Kurram, and D.I. Khan listed in order of their frequency.
Spatial patterns show that provincial capitals and some other districts attracted more terrorism than the others. Provincial capitals are likely to attract more terrorism because they are
27
the power base of the government. Terrorism there makes it more didactic, theatrical, and symbolic. Additional factor which may have increased terrorism in them are their bigger sizes allowing for availability of more targets, more media coverage, and more hiding places available. Other districts excluding capitals which have more terrorism in them are the districts which have some source of conflict in them. For example, the districts of Dera Bugti, Kohlu, and Sibi are where the gas fields are in Baluchistan and these are one source of grievance of the locals against the federal government.
In the same way, the districts of Jhang, Faisalabad, Multan, and Bahawalpur, attracted lot many terrorist incidents in them because these districts are source of sectarian conflict between the Sunnis and Shias. Districts of Swat, D.I. Khan, and agencies of North and South Waziristan, Kurram are having more terrorism because they have some localized conflict in them.
The Evolution of Terrorism in Karachi (1974-2007)
To further investigate the highest number of incidents in Karachi—49 % of the national total—I made a temporal distribution of the incidents to study their evolution over time. Figure 13 shows that Karachi started to have perceptible levels of terrorism in 1990 with a peak in 1995 with 616 incidents.
The percentage of terrorism in Karachi in the national total is far higher than what is due to Karachi in terms of its population 18m (10%) and area 3,530 square km (0.44%).
Factors which may explain frequency of terrorism in
Karachi include but are not limited to: being capital Figure 13. The Evolution of Terrorism in Karachi (1974-2007)
0
200
Incidents
400
600
28
1970
1980
1990
Year
2000
2010
of the Sindh province, source of the conflict, its demographic composition, and being a metropolitan. First, because it is capital of the province, terrorism here has more symbolic and theatrical value. Second, Karachi is unique; it is not only a capital but the sole source of conflict in the province of Sindh. The bases of this conflict are its socioeconomic conditions and demographic changes. Karachi had a population of 4, 00,000 in 1947 which swelled up to 18, 00, 00,000 in 2009—an increase of 450 times9. This exceptional increase in population was because of many reasons: Karachi became the first national capital; the only seaport; the first international
airport; major industrial city; financial center; and house of millions of Urdu-
9
It is instructive to look at Karachi’s population graph presented in http://en.wikipedia.org/wiki/Kharak accessed on 01/07/2010 at 4:54
29
speaking migrants from India, major chunk of Afghan refugees, Pathans from NWFP, and Punjabis from Punjab.
Karachi’s queer demographics serve as a perennial source of conflict here. Although, it is capital of Sindh province but the Sindhis are only 7.22% of the total population. Ninety three percent of the population comprises of immigrants: Urdu speaking (48.52%), Punjabi (13.94%), Pashto (11.42%), Balochi (4.34%), Saraiki (2.11%) and others (12.44%).
Karachi has become the largest Pashtun city in the country, there are more Baluchis in Karachi alone than in Baluchistan, and it is the sixth largest Punjabi town (Kureja, 2003, p 152).
Now the politics of Sindh revolves around ethnic conflicts: Sindhi v. Muhajirs, Muhajirs v. Punjabis, Muhajirs v. Muhajirs, and Muhajirs v.
Pathan—creating a situation of everybody against everybody else. Sectarian violence in the city has further aggravated the situation creating a new line of division as many of the MQM are Shias and most of the Pathans are Deobandi. It is not just the linguistic composition of the city but the prevalence of youth, which also mattered. In 1987, the year in which violence in Karachi actually started, almost 36% of Karachi’s population was between the ages of 14 and 30. 71% of them were literate, as compared to the overall Karachi literacy figure of 55% and the overall Pakistan figure of 26.17%. 22.4% of its members were graduates.
No wonder the infrastructure—housing, roads, transport—could not keep pace with the fast growth of the population. Originally housing conflicts turned into ethnic rivalries, and transport problem provided the immediate context. The first ethnic violence was between the Urdu-speaking Muhajirs and Pathans in 1987 on killing of a Muhajir female college student by a Pathan passenger
van driver in an accident. This clash was aggravated by the easy availability of
30
weapons at unusually cheap prices. Between 1986 and 1989, the prices of guns went down by 40 to 50% in Karachi. A pistol could easily be bought at Pakistani Rupees 3000 (US $40 now) and a Kalashnikov for 16,000 Pakistani Rupees (US $188).
The MQM with the active support of the government split in 1992 (Fair, 2004, p. 112).
Violence within the two MQM factions was most intense during 1993 and 1994.
THE DISTRIBUTION OF TARGET TYPES
Private Citizens, properties, and businesses remained at the top with 844 (39%) incidents against them. Infrastructure and facilities remained the second important target with 315 (13%) incidents. Police and military also remained under attack in large number of cases—251 (10%) against police and 151 (6%) against military and paramilitary forces. Sectarian targets attracted 221 (8%) incidents. Civil administration was attacked in 147 (6%) incidents, foreigners in 131 (5%) incidents, civil society institutions in 32 (1%), educational institutions 22 (0.9%), minorities in 19 (0.8%) and music shops and barber shops in 18 (0.8%) cases.
Private Citizens, property, and businesses remained at the top as victims probably because of three reasons. First, they are a soft target, undefended, and have no deterrence. Second, they are available in large numbers; all groups targeted them. Third, once attacked, they compel the government to yield to terrorist demands. Data show that police are the target in almost 10% incidents in Pakistan and the world both. This pattern is easy to understand because terrorism whether it is against the government or some other entity, police must come in as the first line of response. Therefore, they are targeted necessarily.
Figure 14. The Distribution of Target Types
31
Distribution of Target Types
Music & Barber Shops
18
Minorities
19
Educational Institutions
22
Civil Society Institutions
32
Foreigners
131
Civil Administration
147
Military & Paramilitary
151
Political Leaders & Meetings
186
Sectarian
221
Police
251
Infrastructure & Installations
315
Private Citizens, Properties, Businesses
844
0
200
400
600
Incidents
800
Figure 15. Evolution of Target Type-Current Trends for Select Targets
0
20
Incidents
40
60
80
100
32
1970
1980
1990
Year
Police
Civil Administration
Educational Institutions
2000
2010
Military & paramilitary
Music Shops & Barbers
Although the language-bases organization attacked the police in the 1990s, after 9/11, the trend generally is to target government agencies like police, military, schools, and barber and music shops as shown in Figure 15. Military is at the top followed by police, civil administration, educational institutions, and music and barber shops.
Targeting of military and paramilitary, police and the civil administration is interpreted as reaction against the government policies or a deliberate attempt by the vested interests to destabilize the state of Pakistan. Educational institutions (girls’ schools) are targeted because according to Taliban, their education is unIslamic. In the same way, Taliban target music and barbers’ shops because they consider listening music and shaving beards as against Islam.
THE DISTRIBUTION OF WEAPON TYPES
33
Figure 16. Distribution of Weapon Types
806
Incidents
400
600
800
1,000
Distribution of Weapon Types (1974-2007)
1003
200
264
75
Suicide
Melee
49
0
81
Firearms Explosives Incendiary
Projectile
34
Figure 16 indicates that firearms were the main weapon in 1003 cases—near 40% of the terrorist incidents, very closely followed by explosives in 806 (32%) incidents. Incendiary was used in 264 (11%) incidents, suicide attacks in 81 (3%) cases, melee in 75 (3%), and projectiles in 49 (2%).
Firearms remained on the top as weapons because these are easily available, less expensive, and do not need much of training to be fired. Added reason was the MQM’ liking for firearms as asserted by C. Fair (2003, p. 112).
She asserts that the MQM’s use of rocket launchers was inaccurate and ineffective.
Therefore, they did what they could do competently— acts of assassinations, terror, and street violence—and pistol was their favorite weapon. Another dimension of weapon type analysis could be the study of casualties per type of weapon. Such an analysis is presented in Table 4. Suicide attacks killed and wounded 42 people per attack, explosives 9.4 per incident, firearms 4.3 per incident, and projectiles 7.3. Table 4. Killed and Wounded per Weapon Type (1974-2007)
35
Weapon Type
Killed & Wounded Total
Killed & Wounded Per
Incident
Firearms
4354
4.3
Explosives
7543
9.4
Suicide Attack
3361
42
Projectiles
359
7.3
36
Figure 17 presenting the use of firearms and explosives is very revealing. Starting from 1987 till 1990, explosives dominated the terrorism scene in Pakistan. First time in 1991, firearms started to dominate and for seven years till 1997, firearms kept on dominating. Then from 1998 onward till 2007, explosives are the weapon of choice with the terrorists. It is interesting to note that when explosives are dominating the scene, they do it for quite some time, so do the firearms. Explosives are comparatively difficult to acquire and use and their use also need meticulous planning.
Therefore, my explanation of the observed pattern supported by the data is that when explosives are used, a foreign hand behind the terrorism in generally there, for example pre-1991 years, and post-1997. Secondly, explosives are more likely to be used directly against the government targets, or to destabilize the government, otherwise. Figure 17. Evolution of Firearms and Explosives (1987-2007)
0
Number of Incidents
100
200
300
37
’87 ’88 ’89 ’90 ’91 ’92 ’93 ’94 ’95 ’96 97 ’98 ’99 ’00 ’01 ’02 ’03 ’04 ’05 ’06 ’07 Explosives
Firearms
As we learned that suicide attack is the most damaging type of weapon used by the terrorists and its frequency is increasing day by day, its in-depth investigation is necessary. Therefore, I present the suicide attacks separately. Figure 18 shows that suicide attacks in Pakistan started in 1995 but were not so frequent until 2001. From 2001 onward the graph took a sharp upward turn and 56 suicide attacks resulted just in 2007 alone.
The seriousness of the issue can be gauged from the fact that worldwide from 1980-2001, 188 suicide attacks were reported (Pape, 2003, p. 343) but in Pakistan, in just one year, 56 suicide attacks took place. Suicide Figure 18. Trends in Suicide Attacks
0
10
20
Incidents
30
40
50
38
1970
1980
1990
Year
2000
2010
terrorism follows a strategic logic (Pape, 2003, p. 344).
The data here show that suicide attacks are used against hard targets, or where a lot of damage is required. In Collins’ opinion, growth of suicide attacks, using explosives, might be explained by the demonstration of their greater effectiveness, achieved through the delivery system. According to him, the humans operating covertly have the highest reliability rate of delivering the weapon close to its target.
(R. Collins, personal communication, November 18, 2009).
While explaining this sharp rise, Hassan (2004, p.42) wrote “foreign jihadis brought the technique; Arab clerics escaping from Afghanistan preached its virtues; Pakistani merchants and smugglers provided funding; and local zealots supplied bombers.”
THE DISTRIBUTION OF TERRORIST GROUPS
39
1878
500
Incidents
1,000
1,500
2,000
Figure 19. Distribution of Terrorist Groups (1974-2007)
206
103
47
27
26
13
0
121
Unknown Lang
Sect
Militants Race ForeignerPoliticalal Qaeda
Two hundred and six (8%) terrorist attacks are traced back to Ethnic (Language) groups, 121(5%) to sectarian groups, 103 (4%) to militant Muslims, 47 (2%) to Ethnic (Race), 21 (1%) to foreigners, 26 (1%) to political terrorists, and 13 (0.5%) to al-Qaeda as shown in Figure 19. A major portion of 1878 (75% of the total) incidents is attributed to unknown terrorists. Although the GTD shows them as untraced but majority of these incidents have been traced out and linked to language-based and sectarian terrorists.
Table 5 shows that number killed by al Qaeda per incident are 10.15, against 2.25 overall. Number of wounded per incident attributed to al Qaeda are 25.27, against the overall mean of 4.47. it shows that al Qaeda incidents are, comparatively, severer. Table 5. Severity of Al-Qaeda Incidents
40
Variable
Observations
Mean
Std. Dev.
Min
Max
Number Killed
13
10.15385
9.007831
0
24
Number Wounded
11
25.27273
24.02537
2
70
TERRORISM IN PAKISTAN PRE AND POST THE U.S-LED INVASIONS OF
AFGHANISTAN
The study compares and contrasts the two periods by using descriptive statistics, GIS map, and bar graphs. To make comparisons valid, I used data immediately preceding and following the Afghan invasion and immediately preceding and following the Iraq invasion for equal number of days. The U.S. invaded Afghanistan on October 7, 2001 and including the attack day, it makes 2277 days to December 31, 2007. To count 2277 days backward it takes us to July 14, 1995. Whereas the U.S. invaded Iraq on March 20, 2003 and it counts 1748 days from the invasion day to December 31, 2007. To count 1748 days backward, I took June 6, 1998 as the starting day.
Results of data analysis on the change in total number of incidents, severity levels, change in spatial distribution of terrorism, change in weapon types, target types, and weapon types are given in Table 6. Summary statistics show that the frequency of incidents went down but their severity levels went up.
Table 6. Comparison of Pre and post Statistics
Before
After
Incidents
827
619
Killed
1410
2277
Killed per Incident
1.92
3.68
41
Wounded
2675
4840
Wounded per Incident
3.70
7.82
Figure 20 and Figure 21 show the spatial distribution of terrorist incidents in Pakistan pre and post U.S-led invasion of Afghanistan. Figure 20 makes it vivid that terrorist incidents in Pakistan before the invasion were concentrated mostly in Punjab and Sindh—the two eastern provinces of Pakistan bordering India. Figure 21 showing the distribution of incidents after the Afghan invasion indicate a geographical shift from the eastern regions of Pakistan to western regions like Peshawar, Swat, Quetta, FATA, and Bolan and Dera Bugti areas of Baluchistan.
Table 7 portrays the changes in distribution of terrorism incidents in terms of number of incidents pre and post the U.S-led invasion of Afghanistan. Incidents in terms of percentages increased in NWFP, Baluchistan, FATA and incidents significantly decreased in Punjab and Sindh. Capital Islamabad has attracted almost the same percentage of incidents.
Figures 20 and 21. Spatial Distribution Before and After U.S-led Invasion of Afghanistan
42
43
Table 7. Pre and post Percentages of Incidents in Provinces and Regions Incidents
Before %
After %
Punjab
22.84
6.86
Sindh
71.32
10.20
NWFP
1.34
24.92
Baluchistan
1.70
31.44
FATA
0.49
23.08
Islamabad
2.31
3.51
Total
100.00
100.00
Table 8 below presents the percent changes in the incidents attributed to them and indicates that Muslim Militants, and Ethnic (Race) increased their incidents but language-based ethnic incidents, and sectarian decreased substantially from 5.19 % to 0.97%. Table 8. Percent Changes in Terrorist Types
Terrorist Type
Before%
After%
Ethnic (Language)
5.19
0.97
Sectarian
8.33
3.23
Muslim Militants
0.97
12.44
Ethnic (Race)
0.12
7.43
Al-Qaeda
0.00
2.10
Table 9 shows that military and paramilitary forces, infrastructure and facilities and civil administration attracted a higher percentage of incidents after the U.S-led invasion of Afghanistan.
44
Table 9. Percent Changes in Target Types
Target Type
Before%
After%
Private Citizens & Businesses &
Property
31.28
20.36
Infrastructure & facilities
11.11
15.51
Police
10.02
10.66
Religious Figures/Institutions
12.20
7.43
Military & paramilitary
1.57
19.06
Civil Administration
5.07
9.21
Educational Institutions
0.72
1.94
Music shops, Barber shops
0.12
2.75
Table 10 makes it clear that explosives and suicide attacks have increased substantially from 16 % to 58%, and 1 % to 12% respectively. The use of firearms has decreased from 49% to 23% and incendiary from 21% to 1%.
Table 10. Percent Change in Weapon Types
Weapon Type
Before%
After%
Firearms
49.03
22.94
Explosives/Bombs/Dynamite
16.06
57.67
Incendiary
20.65
0.65
Suicide Attack
0.72
12.12
Projectile (rockets, mortars, RPG)
1.69
2.75
TERRORISM IN PAKISTAN PRE AND POST THE U.S INVASION OF IRAQ
45
Table 11 indicates that terrorism incidents increased from 162 to 573 and people killed and wounded from 427 to 2160 and from 1208 to 4543, respectively. The number killed per incident and the number wounded per incident increased from 2.64 to 3.77 and 7.46 to 7.93. Table 11. Comparison of Pre and post Statistics
Before
After
Incidents
162
573
Killed
427
2160
Killed per Incident
2.64
3.77
Wounded
1208
4543
Wounded per Incident
7.46
7.93
Incidents in Punjab and Sindh decreased from 36% of the total to 6% and from 42 % to 7%, respectively. Incidents in NWFP increased from 3 % to 26%,
Baluchistan from 8 to 33%, and Table 12. Pre and post Percentages of Incidents in Provinces and Regions Region
Before
After
Punjab
35.63
5.98
Sindh
42.50
6.70
NWFP
3.13
26.45
Baluchistan
7.50
33.33
FATA
3.13
24.28
Islamabad
8.13
3.26
Total
100.00
100.00
FATA from 3 to 24%. As shown in Table 12, incidents in Islamabad also decreased from 8% to 3%. Table 13 presents that Muslim militants and Ethnic (Race) increased their share of incidents
46
while Ethnic (Language) and sectarian terrorists decreased their percentage share in the total incidents.
Table 13. Change in Terrorist Group Types
Terrorist Type
Before
After
Ethnic (Language)
4.29
0.35
Sectarian
20.25
2.27
Muslim Militants
4.91
12.06
Ethnic (Race)
0
8.04
Al Qa’eda
1.23
1.92
Table 14 while showing changes in target types tells us that terrorism against infrastructure increased by 7%, against military 20%, civil administration 4% but decreased against minorities by roughly 2%.
Table 14. Change in Terrorist Target Types
Target Type
Before
After
Private Citizens & Businesses &
Property
27.61
20.62
Infrastructure & facilities
9.20
16.08
Police
8.52
10.49
Religious Figures/Institutions
20.25
6.82
Military & paramilitary
0.61
20.45
Civil Administration
6.13
9.79
Foreigners
12.88
3.32
Educational Institutions
1.84
2.10
Minorities
3.68
0.87
47
Music shops, Barber shops
0.61
2.80
Table 15. Change in Weapon Types
Weapon Type
Before
After
Firearms
36.42
21.68
Explosives/Bombs/Dynamite
51.85
58.57
Suicide Attack
2.47
12.76
Projectile (rockets, mortars, RPG)
7.41
2.27
Changes in weapon types as brought out by Table 15 indicates that use of firearms decreased 14%, projectiles by 5% while suicide attacks increased by 10%, and explosives by 7%.
DISCUSSION AND CONCLUSION
The results from this study provide a story for the changing patterns of terrorism. Put succinctly, this is a story of seven C’s: conflicts, communism, capitalism, cycles, capitals, citizens, and Kalashnikov. These C’s combined provide us the gist of spatial and temporal patterns of frequency and magnitude of terrorism, and patterns in victim, weapon, and terrorist types.
While giving the gist, we can say that terrorism in Pakistan is an extreme form of reaction to inappropriately handled political and economic grievances turned into ethnic and religious dissensions. Internal and external vested interests provided the support, which aggravated the situation. Findings have theory, practice, and policy related implications. The U.S-led invasions
48
of Afghanistan and Iraq have brought some profound changes in terrorism patterns in Pakistan. Frequency of incidents have increased, more people are dying, weapons have changed from firearms to explosives especially noticeable is the increase in suicide attacks, and the government institutions especially army is the target. Terrorism has moved from east to west of the country with substantial increases in NWFP, Baluchistan, and FATA. Sectarian and language-based incidents have decreased but terrorism committed by Ethnic (race) and Muslim militants have increased.
The striking shift in the geographic concentration of terrorism post-USA invasion of Afghanistan is prima facie evidence of the dynamic nature of terrorism. Terrorism has moved from east to west because now area of conflict has shifted to the west of the country. The second explanation might be sponsorship of the post 9/11 terrorism in Pakistan which allegedly comes from India, the U.S. or Iran and they are sitting on the western border of Pakistan. Conflicts—ethnic, political, and religious—like many states are endemic to Pakistan, sometimes real, sometimes perceived, and sometimes concocted, even.
Despite these conflicts, life in Pakistan went as usual, unless the government agencies turned these natural conflicts into well-defined divisions and the outside state actors used these divisions to advance their geostrategic interests in the region. All of these conflicts have resulted from some genuine socioeconomic grievance, for example, lack of civic amenities, unfair distribution of resources among provinces, issues of provincial autonomy, and changes in demographics.
Conflagration caused by inappropriately handling these conflicts became the essence of terrorism in Pakistan. Ultimately, the political conflicts turned into sectarian, ethno-linguistic, ethno-secessionist, and religious motives for terrorism. Conflict not only explains varied motives for terrorism but spatial distribution of terrorism at some places too. Places with the source of conflict in them, for
49
example, the central districts of Baluchistan, the southern Punjab, Waziristan, and Karachi, attracted more terrorism as compared with the others.. Communism and capitalism—are the two essential part of the story of terrorism in Pakistan; no account of terrorism is complete without a discussion of the two. Since earlier days of Pakistan’s establishment and after after 1979, geostrategic politics of the region revolves around battles between communism and capitalism forms of governance.
Pakistan was affected by communism in two ways. First, a sponsored zeal for fighting communism in Afghanistan filled Pakistan with mullahs, madrassas, mosques, movements, bigots, narcotics, jihadists, and weapons. As a collateral damage of the Afghan war, Pakistan has to bear a big quantum of sectarian terrorism as well. Second, to avenge for grievances against Pakistan for its role in Afghan war, communist Afghanistan and the Soviet Union sponsored terrorism activities in Pakistan.
Kalashnikov is symbolic of the ‘Kalashnikov Culture’ developed in Pakistan after the Soviet invasion of Afghanistan. That’s one of the worst things ever happened to Pakistan. The U.S. bought weapons from China and supplied to Mujahedeen through Pakistani ISI. Half of the weapons never reached the mujahedeen or were sold in the local market by the mujahedeen themselves.
Arms became dirt cheap. Arming of the society made taking up arms on any issue, an easy option for the disgruntled. According to C. Fair ( 2004, p. 101), “as the rise of terrorist movements in Karachi were generally coincident with the spread of small arms in Pakistan, these outfits have had little problem acquiring light machineguns and rocket launchers” which explains so much violence and the resultant lethality.
50
Capitalism becomes more relevant to explain terrorism in Pakistan after 9/11, especially in FATA and Baluchistan. The U.S-led invasion of Afghanistan has three commonly known objectives: hunting out al Qaeda and to fight Taliban; pumping the oil reserves in the Central Asian States through Pakistan; and
encircling and containing China. Because of its geostrategic location in the region, Pakistan is the key to achieving all the three objectives. First, against al Qaeda and Taliban, Pakistan has already helped the U.S a lot except for a few instances where Pakistan’s national interest was severely damaged.
Second, pumping oil reserves is also possible and Pakistan has no objection if the U.S. succeeds in securing a corridor to Turkmenistan through Afghanistan. Third, issue of China’s containment is problematic for Pakistan and Pakistan’s reservations, most probably, have created problems for it in Baluchistan. Pakistan is neighbor of China and a China’s shortest possible access to the Indian Ocean near Persian Gulf. Pakistan has built Gwadar port near the Strait of Hormuz at the mouth of Persian Gulf with the Chinese help. The U.S. analysts believe that this port is likely to project China’s naval presence in the region near the Gulf where 55% of the oil reserves are.
It’s making China’s access to the world market easy. It will also make China’s relations with Pakistan stronger. India having enmity with China, also takes Chinese presence near its waters as something intimidating. There is likelihood of Gwadar port being offered as an alternative to Dubai and Iran’s newly built free trade port Chabahar. So, the U.S. India, Iran would never like this port to work10 and terrorism in Baluchistan is closely linked to the establishment of Gwadar port. Cyclical movement is the characteristic pattern of daily, weekly, monthly, and yearly incident data. Terrorism occurs in waves: terrorism today is correlated with what happened
10
For a detailed account of the U.S. interests in Baluchistan please read a U.S. government document “Baloch Nationalism and The Geopolitics of Energy Resources: The Changing Context of Separatism in Pakistan” by Robert G. Wirsing, April 2008
51
yesterday; terrorism this week is correlated with what happened last week; there is seasonality in monthly data; and yearly data show longer cycles every time. Cycles are reflective of the internal and external dynamics of terrorist organizations. Terrorists’ preparation, initiative, and the government’s indecisiveness take the violence to the peek. Terrorists’ exhaustion or the government’s action brings the violence down but for some time. As the conflicts, the basis for the violence, remain intact, the terrorism comes again with new agenda. Capitals—provincial and national—are the hardest hit places, as brought out by the spatial analysis in the paper.
Capitals are the hardest hit because they have symbolic value as the seat of the government, making the terrorist act more theatrical. Moreover, whatever the source of the grievance, the government serves the strategic logic of terrorism better—challenge the authority, attain theatrical value, and be more symbolic. The case of Karachi is unique; it’s not only the provincial capital but the source of conflict too. This leads us to the proposition that capital and conflict combined create more terrorism compared to either capital or place of conflict singly.
Citizens— private citizens, their property and businesses—remain on the top as victims of terrorism in Pakistan. Most likely, they are defenseless and hence soft targets; they are innocent, hence hitting them creates more alarm as compared to hitting the police or military personnel. Hitting the citizens is in line with the strategic logic of the terrorists; citizens are more likely to compel the government to accept terrorist’s demand. The paper has a value for theoretical, policy, and operational reasons. As far as its contribution to theory is concerned, there are some important observations to make. First, terrorism is still hard to define and measure and as it is true that one man’s terrorist is other
52
man’s hero, it’s is also true that one times hero may be other times villain. Second, no account of terrorism is complete without discussing the internal and external state actors and the geostrategic politics of the region; terrorism in Pakistan is never apolitical. Third, terrorism is a dynamic phenomenon changing patterns with time: temporal patterns, spatial patterns, victim type, weapon types, and motives all change with the situation. Fourth, unresolved conflicts are the essence of terrorism. Fifth, terrorists select targets because of some strategic logic.
Either the targets afford them maximum propaganda value, or they have symbolic value for them, or they are confronting them to counter their actions, or they are the direct enemy, or there are some ideological grounds for attack. Sixth, terrorists’ weapon choice may indicate their linkages. Generally speaking, locally sponsored terrorists would use firearms and the externally controlled terrorists would most probably use explosives. Frequency of events is the forte of the locally sponsored organizations and magnitude of the events strength of the externally sponsored.
Target selection also indicates the likely sponsorship of terrorism. Locally sponsored terrorists would attack their direct enemies most of the time while externally sponsored terrorists would attack infrastructure or the government agencies or the private citizens. The paper’s results from the analysis of time and spatial patterns of terrorism in Pakistan are of particular relevance for policy. Resources could be deployed keeping in view the time patterns and spatial patterns observed.
Additionally, conflicts generated by socioeconomic conditions have surfaced as the mainstay of terrorism. It makes it important for the policy makers to look into the prevailing conflicts. Conflicts, generally, have valid grounds and could be dealt with conflict management through constitutional measures. It is also important for policy makers to look at how they should place themselves within the changing geostrategic politics of the region. For the immediate
53
amelioration of the situation, an enhanced law enforcement infrastructure is required. Finally, every terrorism ends but end of terrorism does not mean beginning of peace (Cronin, 2009).
Terrorism cycle, which is going on in Pakistan now-a-days would end in near future, unless fuelled by the vested interests. Policy makers will have sufficient time to avert or lessen the severity of the next cycle and that’s what they should be preparing for now. The analysis of data raises a number of important theoretical questions, as it answers the others.
First, is there a regular pattern by which government-sponsored organizations lead to spin-offs (LeJ in case of SSP and MQM (H) in case of MQM), over what time period, and by what conditions? Second, lethality is increasing, what is it about the current flavor of state destabilizing terrorism that motivates such higher causalities? Is the planning better than in the past? Are the actors less controlled by political parties that in the end want some legitimacy with the polity?
Is there less evidence of specific deterrence with these groups after large sweeps and arrests are made? Third, why should the sectarian and language-based terrorism decrease, instead of continuing and adding to the overall amount of terrorism? Is it because of life cycle of growth and decline through which all terrorists organizations pass; or there is a structural limit to the amount of terrorism that can go on at one time?
In sum, the descriptive exploratory analysis in this paper reveals the patterns, producing new insights, prompting new thoughts, and generating new hypotheses contributing to make up for lack of ‘facts’ and lack of ‘theory’ on terrorism in Pakistan. However, there is one main study limitation. As compared with the FIA, the SATP, and other sources, the GTD data are lower in number. Therefore, absolute numbers should be interpreted keeping this thing mind. However, percentages are not significantly different except for the statistics on terrorist group types. But
54
we could rely on the relative position of each group according to the percentages of cases they were involved in.
References
Asal, V., Fair, C. C., & Shellman, S. (2008).
Consenting to a child’s decision to join a jihad: Insights from a survey of militant families in pakistan. Studies in Conflict and Terrorism, 31(11), 1001-1022.
Barshied, C. (2005).
The sipah-e-sahaba dossier, Center for Policing Terrorism, The Manhattan Institute.
Cohen, S., Fair C. C., Ganguly, S., Gregory, S., Shah, A., Tellis, A., What’s the Problem With Pakistan? Foreign Affairs, March 31, 2009
55
Cronin, A.K. (2009).
How Terrorism Ends: Understanding the Decline and Demise
of Terrorist Campaigns, (Hardcover – Aug 24, 2009).
Gayer, L., (2007).
Slums, guns, and yellow devils: a genealogy of urban conflicts in Karachi. Modern Asian Studies, Cambridge Univ Press
Gove, P. B., & Merriam-Webster, I. (2002).
Webster’s third new international dictionary of the english language, unabridged. Springfield, Mass.: Merriam-Webster. Grare, F. (2007).
The evolution of sectarian conflicts in pakistan and the ever-changing face of islamic violence. South Asia, XXX(1), 127-144.
Hassan, N. (2004).
Al-Qaeda’s Understudy, The Atlantic Monthly, June 2004. Hird, C. & Ruparel, C. (2007).
Seasonality in Recorded Crime: Preliminary Findings. [London, England]: Home Office Online Report, February 2007. Retrieved on January 14, 2010 from http://www.homeoffice.gov.uk/rds/pdfs07/rdsolr0207.pdf.
Haqqani, H. (2006).
Weeding Out the Heretics: Sectarianism in Pakistan. Current Trends in Islamic Ideology (Vo. 4) November 01, 2006. Hudson Institute’s Center for Islam, Democracy and the Future of the Muslim World.
Kamran, T. (2008).
Contextualizing sectarian militancy in Pakistan: A case study of Jhang, Journal of Islamic Studies.2009; 20: 55-85
Kukreja, V. (2003).
Contemporary Pakistan: political processes, conflicts, and crises. Sage Publications India Pvt. Ltd.
56
Lafree, Gary; Dugan, Laura; Korte, Raven (2009).
Criminology, Feb 2009, Vol. 47 Issue 1, p1745, 29p Landau, S. F. & Fridman, D. (1993).
The seasonality of violent crime: the case of robbery and homicide in Israel. Journal of Research in Crime and Delinquency, 30 (2): 163-191. Looney, R. (2004).
Failed economic take-offs and terrorism in pakistan. Asian Survey, 44(6), 771-793.
Morken, G. & Linaker, O. M. (2000).
Seasonal variation of violence in psychiatric patients and the general population. European Journal of
Psychiatry, 14 (3): 153-160. Nawaz, Shuja (2009).
FATA — A Most Dangerous Place. Center for Strategic and International Studies, January 2009.
Pape, R.A. (2003).
The strategic logic of suicide terrorism, American Political Science Review, August 2003; 97, 3
Pinault, D. (2003).
Shia-sunni relations in contemporary pakistan. Journal of South Asian and Middle Eastern Studies, Vol.26, no.Iii, 26(3), 62-84.
Rashid, A. (1999).
Taliban: Exporting Extremism. Foreign Affairs. 22 (1999) Riedel, B. (2008).
Pakistan and terror: The eye of the storm. Annals of the American Academy of Political and Social Science, 618, 31-45.
Schaffer, T. C. (2008).
The madrassah challenge: Militancy and religious education in pakistan. Survival, 50(5), 199-200.
Stern, J. (2000).
Pakistan’s jihad culture. Foreign Affairs, 79(6), 115 Tellis, A. J. (2008).
Pakistan’s record on terrorism: Conflicted goals, compromised performance. Washington Quarterly, 31(2), 7-32.
Weiner, T., (1998).
Afghan Taliban Camps Were Built by NATO, The New York Times, August 24, 1998, Washington.
Williams, B. G. (2008).
Talibanistan: History of a transnational terrorist sanctuary. Civil Wars, 10(1), 40-
