CROSSOVER International Conference on Policy Making 2.0 17-18 June, 2013 Lloyd Institute, Trinity College, Dublin, Ireland System Dynamics applications for policy modelling in the Defence sector Riccardo Onori, MBA, PhD (onori@dis.uniroma1.it) Dept. of Computer, Control, and Management Engineering
System Dynamics applications System Dynamics is basically an approach other than a computer simulation modeling methodology for studying and managing complex feedback systems, such as business and other social systems. System Dynamics, as a decision-making modeling approach, is easily used in contexts where standard analysis is made difficult by the wide range of available data, especially in those systems highly influenced by the so called soft variables, those variables, usually connected to human behavior, that are not directly measurable (i.e. Customer Satisfaction, Panic Level, and so on). Properties of dynamic problems: Contain quantities that vary over time Variability can be described causally Important causal influences can be contained within a closed system of feedback loops 2
Applications in the Defence sector There are various SD applications for policy modeling in the Defence sector. During the CROSSOVER project the Consortium built a state of the art KNOWLEDGE BASE, alive and dynamic, that exposes tools, models, key people, case studies, etc. Many applications can be founded in that repository. In order to show how the SD approach can support the policy modeling in the complex frameworks of the Defence sector, we will briefly revise three different recent applications on which we have been involved: The assessment of an international Crisis: the case of MALI The workforce planning in the italian Military sector The Critical Infrastructure resilience and protection 3
The assessment of an international Crisis: the case of MALI 4
General context International crisis situations are complex by nature, and the number of actors involved, the many relations and influences, the freedom of action enjoyed by an ever-growing number of individuals, can be seen as typical attributes of complex social systems. Many contemporary attempts to intervene in this kind of situations have been affected by the lack of understanding of the many interrelations that influence actors behaviors and ultimately the final outcome of the crisis situation. The contemporary crisis situations are typically characterized by: extensive violence and loss of life, massive displacements of people, widespread damage to societies and economies, need for large-scale, multi-faceted humanitarian assistance, prevention of humanitarian assistance by political and military constraints, significant security risks for humanitarian relief workers in some areas. 5
Situation Awareness process Perception Perceive status, attributes, and dynamics of relevant elements in the environment - events - people - environmental factors Comprehension The processes of pattern recognition, interpretation, and evaluation Projection Project the future actions of the elements in the environment 6
MALI, Step 0: Analyse the context Population Religions Ethnic Groups GDP (PPP) >15 millions 90% Muslims, 9% Animists, 1% Christians Mande 50%,Peul 17%, Voltaic 12%, Songhai 6%, Tuareg and Moor 10%, other 5% $ 18 billion GDP (per capita) $ 1.100 Export ($2,3 billion 2011 est.) Cotton, gold, livestock Import ($2.5 billion 2011 est.) Oil, machinery, building materials, food, textiles Population growth 2.6% (doubling time of about 30 years) Fertility rate 6,35 children per woman Average age 16,3 years Unemployment rate over 30% (with worsening projections) % of population below poverty threshold 47% (2006 UN est.) Children malnourished (< 5 years) 28% Infant mortality (< 5 years) 17,8% 7
MALI, Step 1 Define the main actors 8
MALI, Step 2 Analyze the history of a crisis In 2002, the United States started the so called Pan-Sahel Initiative (PSI) in Mali, Niger, Chad and Mauritania, to strengthen their borders in order to control illegal arms and drugs smuggling, as well as the circulation of transnational terrorists (from 2004, PSI became the Trans-Sahara Counterterrorism Partnership). The initiative was primarily based on military support (in terms of training and equipment). 9
MALI, Step 3 Explore conditions of perceived threat Elemental conditions (i.e.: survival, deep legitimate interests, ideologies, values, etc ):Tuareg ethnical identity, Islamic Fundamentalism Structural conditions (i.e.: illegitimate governments, ineffective governance, inequalities, forbidden access to the political life, social exclusion, access to natural resources, geographical issues / borders, etc.): Post-colonial borders, Corruption, Patronage, Mafia-like relationships, Perception of abandonment and discriminations, Access to (presumed) mineral deposits Immediate (proximate) conditions (i.e.: INTERNAL: Security-sector out of control, proliferation in personal arms possession, human rights abuses, flows of displaced persons and refugees; EXTERNAL: bad neighbours, geopolitical competition, refugees and runaways, terrorists): Impact of the Arabic Spring and the Crisis in Libya (militias returning, arms diffusing after October 2011), More than 200k IDP s and 250k refugees in neighbouring countries, Numerous violations of human rights (especially on women and children) 10
MALI, Step 3 Explore conditions of perceived threat Trigger conditions (events generally triggering and/or accelerating a violence escalation, i.e.: elections, arrests or assassinations of key people, environmental disasters, primary goods prices increase, coup d état, economical crisis, etc.): Rearing of food price, Coup d état, Alert for locusts invasion Crisis-induced conditions: Material: light arms proliferation Emotional: hates, revenges (i.e.: recent outcrop of numerous inter-tribal conflicts in North Mali) 11
MALI, Step 4 Identify the problem-defining variables Governance State Credibility and Political Stability Violence Separatist Cause Economic Welfare Jihad and Terror Causes Operational Capability of Armed Forces Humanitarian Conditions I slamized territor y Available Food I DPs and Refugees 12
MALI, Step 5 Build-up a causal loop diagram - - State Credibility and Political Stability - - Inclusive policies Lybian Crisis Military Influence on Politics - - Patronage and Mafia-like Relationships - Separatist Cause - Induction of Neo-colonialism perception Operational Capability of Armed Forces - Economic Welfare - External Military Intervention - - Available Food - Armed Rebellions - - - Quality of L ife - I DPs and Refugees - AQIM Hiring Humanitarian Violations - I slamized territory Jihad and Terror Causes Intl. Terroristic Acts 13
MALI, Evaluation of a military intervention policy It is a policy that, as a matter of fact, has been put into being, starting from January 13, and is still ongoing. Notwithstanding the relative quick success in freeing the towns captured by the Terrorist, the military effort, so far, have not been able to influence in a resolute manner the various issues at stake. In fact, if the external military intervention is generally triggered by the fact that the Islamized Territory percentage goes beyond a certain value, and if such intervention is generally aimed, on one hand, at reducing the Islamized territory and on the other at providing some help to the local Armed Forces in being more effective against Armed rebellions (even by direct intervention of the International Armies), this generally creates three side effects that are: an inurement effect for which the local armed forces rely on the international armies, thus getting further depressed; a surge in Jihadists terroristic acts due to the ignition of a sense of contrast to the foreign occupation ; an induction to the perception of a sort of Neocolonialism attitude that ultimately further deteriorates the Government Credibility in the eyes of the various involved parties. 14
Workforce planning in the Military sector: the italian case 15
Military Workforce Planning in Italy Transformation and modernization of Armed Forces Abolition of conscription -> professionalization -> reduction of personnel Military Rank Progression regulated by a State Law. 1955: old Angelini Law 1997: new Governmental Bill no. 490 named New Officers Advancement Law (Nuova Legge di Avanzamento degli Ufficiali NLAU) NLAU: attain a new, contained workforce balance by January 1st 2015 (end of transition phase ) Italian Air Force (ItAF) aircrews, from the initial total personnel strength of 1235 officers (such is the last documented number in 2009) to the 2015 goal of 1132 officers During the period from 1997 to today, addressed by the law as a transition phase to achieve the related reduction in numbers, the overall number of aircrews should have diminished (while it has grown...). It is thus evident that a more structured and effective approach to workforce planning is required. 16
Historical behaviour 17
Context some insights After the Academy (2 years training): permanent employment contract cannot be laid off by voluntary dismissal get dismissed only due to retirement age or loss of military status for physical or other reasons Minimum bound-to-serve time (16 years) imposed upon becoming officers When this time expires, Air Force pilots are free to leave the service (many in the past have chosen to join civilian Airlines) In the 90s, phenomenon of dramatic proportions and the Air Force introduced an incentive program to keep pilots from leaving Retention bonus to pilots willing to sign for additional 2yrs service (maximum of five bi-annual bonuses - ten years after which, 46yrs old pilots would supposedly no longer be a prime choice for civilian Airlines ) 18
Qualitative definition of system and boundaries 19
Causal-Loop Diagram: overall Financial§or& Paycheck& budget& R1& Ranks& Delay& Total&Workforce& Promo? ons& B3& Delay& Academy& graduates& Delay& Basic&training& B1& Cadets& Dismissals& B2& Cadets& to&enroll& Workforce& Gap& Recruitment& planning& Figure 4: M ain variables and feedback loops 20
Aging-chain characteristics/1 All ranks have been aggregated in three main groups: Junior ranks (from 2nd Lieutenant to Captain) Senior ranks (from Major to Colonel) General ranks (Brigadier, Major and Lieutenant General) The control variables of this process are: YEARS IN RANK, the number of years that need to be served in each rank; TARGET PROMOTIONS, the number of promotable officers set by the law; ADJUSTMENT VALUE, aimed at matching the number of promotions to the surplus/deficit condition of the next rank strength. 21
Aging-chain characteristics/2 Promotions from one aggregate rank (N-1) to the next (N) fall under the Promotion by selection criteria in addition to a minimum numbers of years to serve in the preceding rank, candidates for promotions are evaluated by a commission and, after being included in a graded list, only a portion of them (the highest ranking ones) is promoted in accordance with the available positions in the next rank (pull-flow). Figure 9: General SFD flows between generic ranks are represented (Senior Ranks Box, without 22
Aging-chain characteristics/3 Conversely, Figure 9: General within SFD each flows between aggregate generic rank ranks are pool represented (push-flow), (Senior Ranks there Box, are without some instances of Promotion by seniority, where the only pre-requisite to be details on the outflow already showed here -, is reported in Figure 10) promoted to the next rank is to have served a pre-determined number of years in the preceding one (applicable to 1st Lieutenant, 2nd Lieutenant and Majors). Figure 10: Senior Ranks detailed SFD 23
Conclusions: lines to take 1. RECRUITMENT A constant recruitment rate policy needs to be introduced. This is the only condition that may allow for further policy actions to be effective and their outcomes predictable. Trying to correct internal balance with a variable inflow rate of officers induces the risk of unmanageable oscillations 2. RETIREMENT The early retirement policy of the Air Force should be revised to achieve the retirement rates that will keep the level of Colonels at the target value. This measure is still within the array of available means of the existing regulation and should, unless things change in the next years, be implemented without problems. 3. PROMOTION There is the need to extend the flexibility to adjust the number of yearly promotions in certain ranks indefinitely. Corrections in the number of promoted officers have, in fact, an immediate impact on the adjacent ranks, thus allowing re-balancing of situations that have not been foreseen (such as an unexpected spike in the rate of voluntary early dismissal) 4. DISMISSAL/RETENTION The Air Force does not have any dismissal policy for the ranks up to Lieutenant Colonel. Conversely there is a retention policy for pilots. So it appears that such an incentive policy should come to an end. 24
Conclusions The Italian Government has passed a Law whose impacts have not been thoroughly understood nor assessed ex-ante. Our model has shown that, with the actual constraints given by the system s structure, it will not be possible to reach the New Law targets by the target date 2015. It might though be possible to reach those workforce number targets by implementing additional policy measures and by first assessing their potential impacts on a model like the one developed. Additionally, more dramatic changes still to come due to current budget spending review: further personnel reduction of 30.000 to 40.000 soldiers (a 21% reduction on the actual total strength across the whole Armed Forces), must be achieved over the next 10 to 20 years, which will make things even mode difficult (thus introducing additional constraints in the system). 25
Critical Infrastructure Protection 26
Critical Infrastructure Protection Several research and policy studies have been conducted to develop methods to improve protection of Critical Infrastructures, and to analyze their behavior both from an economical and social point of view. Utilities, and other organisations who operate a country s critical infrastructure (CI), historically have a high level of competence concerning the assessment and mitigation of threats to the security and integrity of infrastructure they manage. They also have well established and tested procedures for recovering the infrastructure in the aftermath of a disruptive event. However in recent years new threats have emerged and the probability of known threats occurring has increased. As a result, traditional approaches to risk assessment and recovery planning have started to be questioned as to their continued relevance and fitness for purpose. 27
US Dept. of Homeland Security CIP/DSS Under current and future cyber and physical threat environment, the United States in 2003 decided to undergo a comprehensive approach to security for its critical infrastructure using vulnerability, consequence, and risk analyses. The approach had to address uncertain and evolving threats, consider a wide variety of assets and infrastructures, and use consistent methodologies and criteria. The Critical Infrastructure Protection (CIP) Program sponsored by the U.S. Department of Homeland Security (DHS) had three primary goals: Develop, implement, and evolve a rational approach for prioritizing CIP strategies and resource allocations using modeling, simulation, and analyses to assess vulnerabilities, consequences, and risks; Propose and evaluate protection, mitigation, response, and recovery strategies and options; and Provide real-time support to decision makers during crises and emergencies. 28
US Dept. of Homeland Security CIP/DSS Relationship between CIP decision makers, decisions, and the CIP/DSS. Source: Bush et al. (2005) 29
US Dept. of Homeland Security CIP/DSS CIP/DSS architecture Source: Bush et al. (2005) 30
US Dept. of Homeland Security CIP/DSS Generic influences in CIP/DSS critical infrastructure models. Source: Bush et al. (2007) When applied to the Critical Infrastructure management, the System Dynamics approach allows to consider every component of the causal relations and loops between e.g.: the public s behavior, the level of availability of the CI and the public s safety. 31
Other SD works in the CIP field Modeling Economic Impacts to Critical Infrastructures in a System Dynamics Framework, Dauelsberg and Outkin (2005), studies on the consequences of infrastructure disruptions from an economic point of view, leaving room for development on the softer aspects of the system behavior. In their approach they considered a disruption as an event that topples the equilibrium of the system; such a definition can be certainly applied to such an event as a natural or man/made disaster. Understanding and Managing the Threat of Disruptive Events to the Critical National Infrastructure, Warren and Thurlby (2012); they further pointed out that System Dynamics can create tools that are both simpler and more accurate than those used to date in the managing of disruptive events to Critical Infrastructures, specifically in the decisionmaking support field. 32
Research challenges The part of the problem that is less well understood is the relationship between long-term, strategic choices and the ability of infrastructure networks to withstand disruptive events. Those choices concern investment in the assets themselves, in the IT infrastructure, especially the network control systems, and in the people managing the system. It is not so obvious how much impact any particular choice will have over long periods of time, nor how choices on different issues will interact. The issues that need to be better understood are therefore: how long-term choices on strategic issues make the network more resilient (less likely to be damaged by a disruptive event) how these and other choices can minimise the service loss when disruptive events do occur how strategic and operational choices can minimise the time for the network to recover, and thus the total cumulative loss of service 33
Resilience as a policy issue Any improved approach to risk and resilience must be systemic in nature (Pitt Review, HM Government, UK, 2007). The Pitt Review defined this requirement by several criteria: It should consider the system as a whole, not just individual components or subsets in isolation. The forces that cause the disruptive events and the events should themselves be part of the system. It must capture the feedback between parts of the system (i.e. the effects caused by actions on the system can themselves be the cause of further effects). It should recognise that actions can both improve the situation and make it worse (i.e. the feedback can be positive and negative). Actions taken in advance to create or eliminate an effect may have delays in achieving this. The effect of an action can change over time as the system changes. The process of development and adaption of the system is continuous. 34
Conclusions 35
Considerations In the Defence sector among the others anticipating future events, promptly detecting emergencies and evaluating the impact of different policy choices before they are implemented are necessary features of good governance. In particular, the anticipation of future trends and impacts of different policies through a systemic approach is fundamental for the most relevant policy decisions So far, we have been able to assess some possible uses of the System Dynamics / Systems Thinking approach to: the analysis of a Crisis Situation, the evaluation of a regulatory decision, the analysis of critical infrastructure dependencies. These by introducing a phased approach to the analysis and by identifying a some possible System Archetypes describing the dynamics of the systems. What we need now is a smart-decision and systemic approach for understanding system of systems behavior. 36
CROSSOVER International Conference on Policy Making 2.0 17-18 June, 2013 Lloyd Institute, Trinity College, Dublin, Ireland Thanks for your attention! Contacts: onori@dis.uniroma1.it