In complex situations, an opportunity often avails itself in totally unexpected places, directions and forms that are not possible to discover or predict by logical computation (whatever artificial intelligence or big data employed), historic cases and experiences or trend extrapolation; this is where professional intuition of managers grounded in talent and years of experience shall drive the decision-making by using decision-making methods that use clues, heuristics and other tools of cognition. More formalised those methods are, more adept are managers at all levels in applying them higher are chances the organisation will be successful in navigating through surprises of unpredictability and attain its goals even under constraints of limited knowledge and time.
In this post I am presenting innovative features of decision making process and methods the Iraqi 3D Strategy’s employs, in order to flexibly navigate the local change agents and other partners toward building lasting peace and sustainable democratic institutions in Iraq.
Part II(b): Decision-making Methods
Strategy Review and Adjustment framework
As discussed in Part II(a), Strategy Review and Adjustment (SRA) is an iterative framework consisting of regularly conducted evaluations. It is a time and effort consuming exercise to be conducted by well-prepared managers. However, as practice has shown in various settings, the benefits of the process outweigh the invested resources. It is also highly recommended to formalise the process and its decision-making instruments from the outset of the Strategy implementation, with written guidance and resource materials provided.
The principal role of the review participants (decision-makers) is to make judgments and choices to bridge the gap between the ideas expressed in Theory of Change (TOC) and planning documents (PD) and reality, and to eventually choose the course of strategic action with highest potential for producing impacts. From assessment of recent developments to action point, each review comprises three blocks with their own management questions and decision making points, to allow arriving at optimal solution. Evaluations occurring at any point greatly influence what will happen in the next stage. A flowchart of generic SRA for the Strategy is represented in Figure 2; below I will describe each block’s decision making and illustrate it through a set of charts.
First block: Initial recommendation
Discussions of first block shall result in an initial recommendation. It is done through five steps: (a) Taking stock (What?); (b) Assessment (Why?); (c) Choice (What is next?); (d) Recommendation to implement (How?); (e) Monitoring arrangement (How we will know?). This is also a critical decision-making point. Figure 5 shows basic elements and flowchart of this block.
There are five option categories with regards to any level, dimension or component of the Strategy being reviewed: (a) Continue as it is; (b) Adjust and go; (c) Hold on; (d) Drop altogether/Terminate; (e) Replace. The latter option requires a new (or being held in reserve) course of action, with justification.
Second block: Testing
From this point there follows the next block of testing or, if deemed unnecessary, then the recommendation goes to the last block of seeking approval prior to implementation (justification for skipping the testing shall be provided then, along with the recommendation). Testing is an important and highly recommended for those choices which propose a considerable change in TOC, a totally new direction or set of activities (or of local change agents involved). There are essentially three groups of available methods which reviewers will use to test the recommended action, conditionally named as: (a) formal experimental methods (randomised experiments and quasi-experimental designs); (b) formal discursive methods (surveys, interviews, panels, and focus group discussions); and (c) informal methods (such as desk tests, simulations, role plays and thought experiments). Second block’s flowchart is represented in Figure 6 below.
The outcome of testing is one of three conclusions: (a) Affirmative (confirmed, may go); (b) Negative (hypothesis didn’t prove, no go); (c) Undecided (no clear result). Respectively, first conclusion brings the review process into last step of approval and related communication with ultimate decision-makers. Second and third conclusions return the process back to either considering second-best from the options short-listed or redesign it to be better fit for purpose. The new (renewed) recommended option then goes into testing, and so on until the review participants come up with recommendable solution. This is second critical decision-making point in the review process.
Third block: Approval
At this stage, the job of review panel members is to (a) document and communicate their decision through formal channels; (b) make presentations and address questions and concerns raised, in order to convince the Strategy’s stakeholders in (primarily) political advantages and validity of expected impacts, as well as financial, administrative/logistical, and knowledge-specific technical feasibility of the recommended TOC/PD adjustments and the course of action; and (c) get eventual approval, funding (additional, if needed), and commence the implementation.
This seemingly simple process most probably itself will be iterative and even chaotic, and its outcome may depend on a number of external factors not necessarily directly related to the quality and desirability of the recommended adjustments and measures. The outcome here comes along three choices: (a) Affirmative (proceed to implementation immediately upon the agreed plan, which might be slightly or significantly different from the one proposed but consensus between stakeholders and Strategy top managers has been reached); (b) Affirmative but delayed (do not proceed immediately, until further notice or prior agreed date); (c) Negative (adjustments rejected, reasons communicated and accepted by all parties).
The former two outcomes are rather straightforward and resembling the features of somewhat linear decision-making business process. Exception is situations when the delayed choice (for whatever reason) may become outdated and irrelevant over time—which is the case in many situations, especially in international strategies and programmes with high political sensitivity. In turn, in case of a negative outcome the review either goes back to square one or continues the implementation as before (without dramatic changes) until next review. Thus, the third and final critical decision of the Strategy review is made at this point and effectively concludes it.
SRA decision-making toolkit
The Strategy’s operational environment requires management at all levels to be aware of different decision techniques that have been proved successful in complex settings. Some of these methods are better suited to tactical level or to emergency situations under constraints; others work well in solving ‘large world’ problems. What is important: to ensure that guidance on these methods is readily available (including training), that all managers can effectively use them, and that the methods are sufficiently formalised to be employed Strategy-wise.
In the framework of the Iraqi 3D Strategy’s TOC/PD review process, each step employs a number of task-specific (sometimes the same) analytical tools and decision models—some of them are well-known (such as SWOT, PESTLE, Diagnostic trees, Force Field analysis, simplified SCBA, etc.) and broadly practiced and some are rather novel to international development programming (although widely practiced in business and, especially those dealing with urgent decision-making, in security sector).
In this section, I will briefly discuss some of that set of decision-making tools which I think are highly relevant to the Iraqi 3D Strategy (and any other international strategy/programme working in post-conflict, volatile environments) and shall be formally introduced into management’s decision-making, along with traditional methods. I will illustrate them within the TOC review framework, as relevant to the subject of this series of posts.
Multi-criteria decision models
Analytic Hierarchy Process (AHP) method
Multiple Criteria Decision Models (MCDM) are widely used in business and in public sector, for they can be flexibly adapted to decision makers’ objectives and preferences and help to keep the judgment objectives clear. I will illustrate this model on the example of Analytic Hierarchy Process (AHP), one of techniques to tackle complex problems of choice and prioritisation. What makes this method well suited to the Strategy’s approach are the following: (a) evaluation doesn’t require exhaustive historical statistical data but makes comparative judgments for multiple choices; (b) allows managers to examine ‘what-if’ scenarios and alter their own initial decision by reassigning the weight given to criteria (or factors, depending on the case considered); and (c) gives the team enough flexibility to choose and modify evaluation criteria, depending on situation (i.e. decision-makers may opt to replace one or more criteria of their metrics based on experience gained, and so arrive at better decision).
Decision scenario: The goal of the Strategy’s review team is to select the most reliable civil society partner to lead/coordinate a citizen platform partnering with provincial governorate of Wasit in participatory budgeting initiative sponsored by the Strategy.
As agreed by the team, the factors to be considered are: experience (in similar activities plus reputation); capacity (membership in terms of numbers and skills); inclusiveness (of members–gender, youth, minority); and outreach (localities and their representative populations covered). Figure 7 gives a graphic representation of decision model’s core elements.
Upon completing a series of steps when judgements are scored and weighed, review participants will arrive at decision which is based on scores earned by each candidate against all the (weighed) criteria. What is interesting is that the team will weigh the criteria as they think relevant to local circumstances and thus the same candidates would score differently in another province (say, in Babil) with different context and thus requirements. Matrix in Figure 8 represents the final decision based on a composite weight of all choices against all criteria for this illustrative case.
Probabilistic decision tree
This kind of decision tree illustrates the probabilities of alternative choices and is a proven method for choosing the most feasible course of action, when there is enough data (for computation) to support the decision-making process. I will demonstrate it on the following illustrative case.
Decision scenario: The Strategy’s managers/review team in Diyala province considers working on local economic development initiative, to assist the provincial authorities and their public-private partnership (PPP) platform supporting youth entrepreneurship and small business. The Strategy’s decision-makers have to decide whether to help setting technology business incubator (TBI) or craftsmanship workshop-incubator (PBI).
The decision is taken in the face of uncertainty about (a) whether competition will be undermining the start-ups (the rival products of foreign producers will be imported and imitation will be rapid thus reducing the profitability of pioneers); and (b) whether the local and regional customer markets will be booming or in recession for the initial period of five years.
However, having enough data it is possible to calculate economic and social returns under all of possible outcomes: ROC (return on the investment made by the PPP which is a social enterprise with matched public-private funding); Social effects (wellbeing of the community); Economic effect (jobs, public revenue).
The decision tree for this illustrative case is represented in Figure 9.
Advance knowledge as reference
Analogy-based models evaluate the decision object (whether initiatives, projects or strategic directions) by comparing them to known (well-documented) historical precedents of the same class. In development practice, typically managers are using case studies of ‘good practice’ projects in order to learn from them and try make a projection of some successful past experiences elsewhere (with certain, at times significant, conditionality attached due to diverging local circumstances and contexts) to their initiative under consideration.
As discussed in part one of this series, although informative this method alone is not sufficient (mostly due to limits of replicability), especially in such politically complex environments as Iraq today, where developments may take any turn and any point in time thus forcing the managers keep exploring, learning from local developments and adapting to them constantly.
There is another method of this class, which may prove very useful, especially when managers lack information about the context. This is known as recognition-primed method which works under constraint of limited information and thus relies on the decision-makers’ professional experience and intuition. Research findings over decades of observing real-life situations where this method is used (mostly in critical situations, emergency—by military and generally security, fire brigades; but also increasingly by civil agencies and businesses working in environments of high uncertainty and facing decision choices that do not fit into business-as-usual toolbox) prove that this method shall be among formal decision making tools of Strategy management. Conceptual flowchart of this method is represented in Figure 10.
Heuristic decision models
Heuristics are decision-making methods (simply put, rules of thumb) used to solve problems that traditional methods of logic and probability fail to handle. In management practice they have proven to work very well in uncertain settings and are relying on a minimum of time, knowledge, and computation to make choices adaptive to operational environment. In addition to these advantages, these methods are applicable at both strategic and operational levels of decision making. All this qualities make heuristics methods an excellent addition to the Strategy’s decision-making toolkit.
There are various methods (and various designs for some of them) in the class of heuristics driven decision models. Two of them are highly recommended for the Strategy. One is known as categorisation-by-elimination (CBE).
How CBE works? First, a manager/decision-maker or a management’s TOC/PD review team defines a full set of possible choices to be considered. Next, they establish cues (criteria, features, factors) most relevant to the choice decision on a given problem and rank them in terms of their importance. Third, they start evaluating the choices against cues, one at a time, starting with the most prominent cue. There is only one exit per cue, meaning only two choices: yes or no (either choice meets the requirement or it does not). If there is only one candidate choice meeting the criterion at this stage, then the search is stopped and the decision is to go with it as a recommended option. If there are more than one choices meeting the cue requirements, then the set is evaluated against the next cue, and so forth until there is one at a certain stage which is recommended. If all cues are exhausted or no choice meets the cue requirements (meaning that no single choice evolved as winner) then either choice is selected randomly from the remaining sub-set or new cues are added into the search. This means that last cue has two exits, to ensure that a decision will be made in the end.
Another decision tool which is used at individual and group level (especially in emergency situations but increasingly within broader set of settings) and is very effective in complex situations is fast-and-frugal tree (FFT). Its difference from CBE is that this method assesses one object against a set of cues organised by importance. Again, there is only one exit condition from any cue. Therefore it is more suited to situations where there are no alternatives and decision has to be made about that particular object.
FFT Decision scenario: The Strategy’s program in Anbar province supports the Government’s ex-fighter (militia) reintegration programme. At the first stage of the programme, assistance is provided to address the most immediate needs of ex-combatants. Cash payments were opted for at the initiation of programme by the implementing partners, as most effective albeit temporary instrument. However, the Strategy’s Early Warning and Response System (using survey findings) and field monitoring reports indicate that this approach is not as effective as planned and in fact even counterproductive. Therefore the decision shall be taken by Anbar team’s review panel, whether to terminate the assistance in this form or to continue. This illustrative case of decision making using FFT is presented in Figure 11.
What is distinct about this method, when compared to traditional decision trees, is that: (a) it gives a firm decision (that is why called deterministic method) and not a comparative probability between choices; and (2) it allows for reiteration, creativity and alteration (for example, considering the same set of cues arranged in different order according to their significance—thus offering two decision perspectives to compare).
To be continued. Next post II(c): Management Modalities and Operational Environment.
Previous parts of Diplomacy, Development, Defence in Action: 3D Emergency Strategy for Iraq posted on PolicyLabs:
Part I: The Strategy’s Foundations
Part II(a): Decision-making Model and Process
About the Author: Dr. Elbay Alibayov is an international development professional specialising in state-building and political processes in post-conflict countries. Most recently, he has worked in Baghdad assisting the Iraqi Government on a range of administrative initiatives and policy reforms. Before that, he helped building local governance structures and capacity through community-based natural resource management and local economic development initiatives in rural Afghanistan. In the course of eight years he has worked in Bosnia and Herzegovina, where he held various positions in the field (starting as head of field office in Srebrenica) and headquarters; have designed, implemented and overseen a broad range of reform strategies and demand-driven local and nation-wide initiatives; and have chaired and participated in the work of civil-military situation awareness groups, political coordination boards at all levels.