Posts

Keynote Address at the 2010 Computational Social Science Society Conference: From Hunches to Evidence Driven Policy Design

I will be giving a keynote address at the 2010 Computational Social Science Society Conference (CSSS). CSSS 2010 is hosted by the Center for Social Dynamics and Complexity and the Consortium for Biosocial Complex Systems at. For more information on the conference, please click here [LINK].

From Hunches to Evidence Driven Policy Design: Leveraging Information through Simulation

Constructing public policy, whether at the national or local level, is a complex undertaking. Complexity arises from the number of stakeholders involved, varying agendas and incentives, resource constraints, a multitude of interacting variables, multiple time horizons, and even political climates. Due to these complexities, we too often categorize political and social problems as ‘wicked’ and unanalyzable. The default option is to take a haphazard approach to policy design, most often the outcome of the ego-based debates and negotiations of the decision-makers. In this keynote address, I will argue for a move from hunches (or intuition) to evidence driven policy construction. Today, due to the advancement of computational power and modeling techniques, we can simulate complex scenarios. Simulation gives us an ability to move policy construction from an activity primarily driven by historic case analysis and intuitions, to more of an applied science, where we can actually predict and control phenomenon. Through simulation we can, with reasonable certainty, ascertain the cost, benefit, risk, impact, and value proposition of a given policy. Using examples from simulation projects, such as a project that examined strategic options for dismantling terrorist networks, I will demonstrate how we can move policy design from being an ‘art’ to more of a ‘science.’

Conference on Intelligence and Nuclear Proliferation: Threat Identification, Policy Formulation and Decision Making, June 3-5, 2010

I will be speaking at the Conference on Intelligence and Nuclear Proliferation hosted by the Centre for Science and Security Studies (CSSS) at King’s College London in June. Kristen Lau and I have authored a paper that examines how information management failures led to an inability to adequately assess and detect nuclear threats in recent times. Lack of adequate information management capabilities have led to numerous international crises surrounding nuclear non-proliferation. For example, the inability to predict nuclear tests by India in 1998, the colossal failures surrounding assessments of Iraq’s WMD capabilities in early 2000, and today, the challenge of addressing Iran and North Korea.

Intelligence and Nuclear Non-Proliferation Programs: The Achilles Heel?
Intelligence is a critical component of all counter-proliferation activities. It allows us to assess and determine what makes up the current threat environment in terms of the proliferation of nuclear weapons and technology. When informed with an accurate assessment of the situation, policy makers are better suited to counter the proliferation threat. However, success and failure hinge upon how well information is managed during the intelligence process. The intelligence process as it relates to estimating nuclear capabilities or intentions is wrought with many challenges and complications. The denial and deception techniques employed by states running covert weapons programs and the dual-use nature of many weapons components create many difficulties for intelligence organizations. Additionally, illicit transnational networks obscure the situation further by serving as a source, for both nation states and non-state actors, for acquiring dual-use commodities and technologies. These challenges can lead to the miscalculation of a state’s capabilities or intentions. As was seen with the case of Iraq in 2003, western intelligence services grossly overestimated the capabilities of Saddam’s regime. This paper presents a comparative analysis of three cases of nuclear proliferation: India, Pakistan and Iran. Drawing from the analysis, the authors examine the lessons learned and propose recommendations for future counter proliferation policy and strategy.

To read prior papers published on this topic, please see:
• Desouza, K.C., and Lau, K.A.* “Managing the Proliferation of Weapons of Mass Destruction: An Information Management Perspective,” International Journal of Public Administration, 31 (13), 2008, 1457–1512. [LINK]
• Desouza, K.C. “Information and Knowledge Management in Public Sector Networks: The Case of the US Intelligence Community,” International Journal of Public Administration, 32 (14), 2009, 1219–1267. [LINK]

Measuring Agility of Networked Organizational structures via Network Entropy and Mutual Information

Yuan Lin, Sumit Roy, and I have authored a paper that examines the use of network entropy and mutual information to measure the agility of networked organizational structures. The paper will appear in Applied Mathematics and Computation.

Abstract
While the agility of networked organizational structures is important for organizational performance, studies on how to evaluate it remain scant, probably because the difficulty in measuring network evolution. In this conceptual paper, we propose two measures – network entropy and mutual information – to characterize the agility of networked organizational structure. Rooted in graph theory and information theory, these two measures capture network evolution in a comprehensive and parsimonious way. They indicate the uncertainty (or disorder) at the network level as well as the degree distribution at the individual level. We also propose an algorithm for applying them in the scenario of adding links to a network while holding the number of nodes fixed. Both simulated and real networks are used for demonstration. Implications and areas for future research are discussed in the end.

Lin, Y., Desouza, K.C., and Roy, S. “Measuring Agility of Networked Organizational structures via Network Entropy and Mutual Information,” Applied Mathematics and Computation, Forthcoming.