Home  Profile  Partners  Projects  Team GINI  Contact Us 

Project Document
Goals and Objectives
Implementation Work Plan
M&E Plan
Picture Gallery

Database / Analysis

Scroll of News



Project Research & Dissemination Methodology

The project involves two main components: research and dissemination.


Literature Review: At the international level, research on the link between misgovernance and radicalization yields ambiguous findings. While some studies reveal a terrorism-enhancing effect of economic underdevelopment , others find that development and poverty do not constitute strong determinants of terrorism . Most studies do not find income –inequality to be significantly linked to terrorism, contrasted by the findings of Li and Schaub (2004) who find strong correlations between income distribution patterns and terrorism. While some researchers have linked economic downturns to increased terrorist activity , others have found no significant relationship between terrorism and economic performance in terms of growth, inflation or unemployment rates. While most studies suggest a strong negative or positive interrelationship between democracy and terrorism, (depending on how recently the transition was made) , Drakos and Gofas (2006a) find no significant link between the two. Higher degrees of political freedom and protection of civil liberties are found to dampen terrorist activity, while state repression is found to encourage terrorism and radicalization. In contrast, a few studies find only weak associations between political freedoms , while others characterize this relationship as non-linear. While some researchers find that governmental capacity and strength, (operationalized through proxies such as military manpower and government size) correlates negatively with terrorist activity , other studies contradict these findings. Researchers who control for regime stability find that more stable regimes tend to experience less terrorist violence , but not in the case of suicide terrorism. Adequate social welfare policies are found to reduce both the generation and risk of exposure to terrorist attacks. The link between education and terrorism is similarly weak and ambiguous. While certain studies have uncovered at best a weak link , others have linked higher illiteracy among males to higher terrorist incidence. Sound performance against indicators of governance, especially ‘rule of law’, bureaucratic quality’ and ‘government corruption’ are found to reduce ethnic tensions and discourage radicalization and extremism by reinforcing market principles, contract rights, and economic opportunity.

 At the national level, research is mostly limited to newspaper articles and opinion pieces. Academic research that does exist lacks methodological rigor, policy relevance or both. Robert Kemp (2008) attributes the rise of radicalism in Pakistan to the disintegration of state structures, which creates a vacuum filled by the growing influence of religious orthodoxy, compounded by poverty, unemployment and other factors. He does not, however, reveal the extent to which these factors enhanced the influence of orthodox “foreign elements”. Sohail Abbas (2007) surveyed prisoners held in Khyber Pukhtunkhwa who returned to Pakistan in 2001, having fought against US forces until the fall of the Taleban regime. The respondents were mostly employed, literate, and educated in public schools rather than madrassas. However, the small sample size (517 men), and the lack of evidence linking these men to terrorist acts in Afghanistan or Pakistan, undermines the validity and reliability of these findings. Christine Fair (2008) conducted a similar study based on surveys of the families of militants slain in conflict in Kashmir and Afghanistan. Again, a small sample size (141 families), and the use of ‘convenience’ sampling techniques undermine the veracity of findings. A sociological study by Shinwari (2008) posits poverty and lack of opportunity as drivers of growing militancy in FATA. However, as Safiya Aftab (2008) finds in her analysis of spatial distribution of poverty incidence across Pakistan, poverty levels are not nationally exceptional in FATA or Khyber-Pukhtunkhwa which are hotbeds of militant activity.

Research Hypothesis: Factors of misgovernance either cause or create an enabling environment for radicalization and the growth of extremist behavior. These factors involve the inadequate delivery of public goods and services to local citizens, in areas of public safety and law enforcement, dispensation of justice, land management, revenue administration, and social services such as health and education, among others. Targeted policy interventions can successfully address these misgovernance factors, thereby constricting the growth of radicalization and extremist behavior.

Research Questions:

1. Is there a relationship between misgovernance and radicalization? If so, what is the direction, strength, and linearity of causality?

2. Which misgovernance factors are most significant for pushing citizens toward radicalization and further into extremist behavior? What is the tipping point where radicalization assumes behavioral proportions for violence and extremism and for which individuals?

3. What are the mediating or intervening variables that condition the inclination or support for terrorism, for instance, tribal/family affiliations? What are the individual and composite effects of these variables in creating an enabling environment for radicalization and extremism? What are the reasons for differences (if any) in results for different regions or respondent profiles?

4. What are the policy interventions necessary to address these misgovernance factors, through improvements in the financial, administrative and performance management aspects of local governance, so that the growth of radicalization can be stemmed or reversed?

Analytic Framework: The purpose is to determine as accurately as possible, whether exposure to state or institutional failure (possibly manifesting as corruption) leads to militancy or support for militancy, and to what extent. We would explore the impact of poor service delivery (citizens’ perceptions of police, judiciary, land administration, education department, revenue, and other state institutions) governance mismanagement and inefficiency on inclination or support for terrorism. Since the research would provide access to highly robust data on the state of corruption (or perceptions of corruption), service delivery, and inefficiency vis-à-vis state institutions such as police, judiciary, education department, revenue, utilities related departments like WAPDA (the government run electricity- power company), we can also explore the impact of exposure to corruption on regime legitimacy and hence the indirect or direct effect of regime legitimacy, political support or system support on radicalization.

 The above hypotheses will be refined and various indicators developed to construct support for the terrorism and militancy dimension. Once data has been captured and processed, it will be analyzed using support for the militancy dimension (created through various indicators loaded together through factor analysis) as the dependent variable and various other variables like indicators of corruption, service delivery, socio-political efficacy etc. as independent variables and control variables using multiple regression as an analytical tool. The results will then be shared with local service providers to gain insights into the supply-side dynamics of misgovernance, evolve policy recommendations and identify areas for further research.

 As borne out by the literature review, this framework differs from previous research in focus and methodology. While previous research has focused on the link between socio-economic anomalies (such as poverty and unemployment) and terrorism, this project focuses on essentially a sub-set of state-citizen relations, by exploring the relationship between inadequate service delivery and the inclination or support for militancy. Also, the research is solution-oriented, probing the causal factors behind misgovernance in its supply-side dynamics. This means that research carries far greater policy relevance by framing analysis in the policy and institutional context with clearly defined policy dimensions. Methodologically, this research project is far more sophisticated as it will cover a much larger sample size, a randomized sample design, and inferential statistics for analysis (rather than simple descriptive statistics) to establish causal relationships. Also, Focus Group Discussions with local service providers will allow findings to be triangulated, the conclusions thus drawn to be far more robust, and the recommendations to be more grounded, specific, and actionable.


Once the research is complete, the findings may be formatted and summarized into a Special Report, as commissioned by USIP, which will be uploaded onto their website to enhance outreach and visibility.

 GINI will conduct a series of one-day seminars at the provincial (1 each in provincial capitals of Karachi, Lahore, Peshawar and Quetta) and national (1 in Islamabad) levels to disseminate and gain feedback on the Final Report. These seminars will invite the participation of provincial government officials (politicians and civil servants), media professionals (representatives of Press Clubs and Press Associations including journalists and commentators from print and broadcast media), academics (Vice Chancellors, department heads and researchers from provincial and national universities), civil society organizations (NGOs and think-tanks conducting research, advocacy, and/or capacity building for governance, human rights, conflict prevention and peace-building, service delivery improvement, etc.), and international donors focusing on governance or counter-radicalization efforts (including multilaterals such as World Bank and ADB, as well as bilateral donors such as USAID and DfID). The selection criteria for invitees will include experience, qualifications, and level of influence as opinion formers and policy stakeholders within professional domains, as related to either governance or radicalization and terrorism.

 These seminars will present the Final Report to the invitees and allow them to discuss and debate the conclusions and recommendations in detail. These discussions will focus on the merits of the research from political, academic and policy standpoints; the relevance of findings to provinces other than the targeted areas; the feasibility of the recommendations given political, economic and social realities; the roles of provincial and national stakeholders in realizing the recommendations of the report; and other points emerging from the findings of the report. The proceedings of each seminar and the key policy implications and recommendations will be recorded in Seminar Reports which will be prepared by the Research Analyst following each event and shared with project stakeholders so that they may receive intermediate feedback from the seminars. The Preliminary and Final Reports, the Special Report, as well as the Seminar Reports will be uploaded on the GINI website where organizational partners and other interested parties will be encouraged to view and download their contents.

 In addition, the Final Report will also be presented to USIP by the Project Team Leader and the Research Analyst at USIP offices in Washington, USA.

 This section provides details on the methodology adopted for survey research, Focus Group Discussion, and overall administration.

Survey Research: The research design envisages a multistage probability survey of FATA, NWFP(Khyber-Pukhtunkhwa), and the former Malakand Division. These three regions have evolved governance institutions and cultures that are markedly different in terms of historical evolution, and current structure and function. FATA has been governed under the pre-independence Frontier Crimes Regulation (1901); the former Malakand Division includes the former princely states of Swat, Chitral and Dir; while the rest of (Khyber-Pukhtunkhwa) has been administered by the provincial and local governments. Data also signifies varying levels of radicalization across the 3 regions. FATA is described as being under “Taleban control”, (with about 80% of terrorist attacks in Pakistan planned and executed from South Waziristan), the former Malakand Division is described as being under “contested control”, while most other districts of Khyber-Pukhtunkhwa are described as being under “Taleban influence”. These three regions have been chosen purposively to adequately represent these differences in the sample frame.

 The sample would be approximately 1000 in sample size within NWFP (Khyber-Pukhtunkhwa), chosen through probability proportionate to the size, another 500 in Malakand division and approximately 500 in FATA. Within these regions, villages (in rural areas) and circle numbers (in urban areas) will form primary sampling units stratified by administrative units such as Districts, Tehsils and Urban/Rural locations.

 Within each primary sampling unit we would choose 10 respondents identified through selection of 10 households through systematic sampling after identifying the first household through a random walk method. Within each household we would interview a respondent aged 18 years and above, identified through Kish Grid after listing the eligible respondents in descending or ascending order according to their ages, stratified by gender. We plan to select equal numbers of male and female respondents from each primary sampling unit. This sample would be representative of the regions thus chosen. Nevertheless, the 3 regions, (FATA, NWFP and Malakand Division) would be weighted according to their adult population. The disproportionate multistage cluster sample has been designed to keep enough number of cases in each region for sub analysis of each region. This sample design also helps to control for contextual effects or cultural or ethnicity aspects which is presented as an alternative hypothesis to explain the support for radicalization.

 Regression analysis will give adequate controls and can link the main explanatory variable to the dependent variable as causal variables (having controlled for mediating and intervening variables). To overcome the weaknesses emerging from the lack of control and treatment groups, we will use aggregate district level data to form control and treatment groups through matching (propensity score matching, CEM, genetic matching etc.). Incase the number of cases are too few to constitute a treatment group (given the relatively few number of districts sampled), we can match individuals in the survey data, working within the framework of stata or R (statistical software packages) to form control and treatment groups in case of dichotomous treatment variables (for instance, exposure to corruption or not). Such treatment or pre- processing of data will help us to cater for Omitted Variable Bias in the regression analysis (if any). Aside from these pre-processing purposes, we plan to use regression mainly to control for intervening and mediating effects.

Focus Group Discussions: Focus Group Discussions (FGDs) would be conducted in each of the 3 regions selected for survey research (Malakand Division, NWFP, and FATA) inviting the participation of lower, medium, and higher tier public service providers, organized as:

  •  Police, Prosecution and Judiciary;
  •  Land Management and Revenue; and
  •  Social Services departments including Health, Education, & WAPDA, Civil Works, etc.

 These clusters are organized according to the inter-institutional dependencies that determine the process of delivering each public good and service to the consumer. Police, prosecution and judiciary must function in an inter-related fashion to play their roles in the provision of public safety, law enforcement, and justice to the citizens. Similarly, Land Management is closely linked to the Revenue Department, while social services cover various departments that are similar in function and structure.

 Each management tier administering these public service areas would be represented with 9 FGDs (3 tiers x 3 public service areas) in each of the 3 regions. The agenda of the FDGs would be to:

  • Assess the level of extremist behavior in targeted regions, which may be linked to factors of misgovernance

  • Identify the supply-side factors that are responsible for negative citizen perceptions related to each public service area (if any) emanating from the surveys and highlight the reasons that explain differences

  • Evolve policy recommendations that may mitigate these factors, to improve negative public perceptions of local governance, which may promote radicalization

  • Identify areas for further research to explore the misgovernance-terrorism nexus in light of FGD findings

Administration: The project team will consist of a Project Team Leader, 1 Research Analysts, 1 Statistical Expert and 9 Field Operations Teams (3 for each region). The Project Team Leader will provide substantive inputs into the development of the research design, methodology, and instruments. He/she will also lead the analysis of research findings and provide overall supervision and guidance for the research. The Research Analyst will be responsible for designing and executing research, undertaking on-site spot checks, managing contingencies, and ensuring overall monitoring and evaluation. GINI has evolved detailed methodologies for developing survey instruments, recruiting and training field teams, and conducting interviews and Focus Group Discussions that have been applied successfully in the past and will be utilized for the proposed project. In addition, the following safeguards will ensure the integrity of the research process:

  • Interim deliverables for early vetting and revision

  • Contractual safeguards and financial incentives tied to compliance and performance standards for outsourced activities

  • Assurance of researcher capacity and skills at the recruitment and training stages

  • Random on-site spot checks conducted by the Research Analyst

  • Built in reliability checks in the research design, that gauge inter-rater and test-retest reliability levels

  • Multi-stage data processing that detects and corrects errors at the data entry, verification and validation stage

  • Brief trip reports compiled through research journals kept by enumerators

  • Monitoring adherence to Research Action Plans

The Statistical Expert will be responsible for developing the detailed Research Design, (including research instruments), performing statistical analysis and providing inputs into development of findings, conclusions, and recommendations. The Field Operations Teams (FOTs) will be composed of a Field Supervisor and 4 Enumerators who will undertake the research in each of the target Agencies and Districts. A total of 9 Field Operations Teams will be contracted. A team of Data Entry Operators will also be hired, led by a Database Manager who will enter, verify and process the data.

(C) 2012 @ giniweb.net, Website Designed & Developed by IT Team at GINI