Introduction
The disaster in Haiti required an immediate, decisive response. Yet, it was clear that to successfully address the situation a more cautious, nuanced approach would be required. Haiti was treated to a mixture of both. This nexus - between the ideal and the practical - is where relief agents went to work. And so too did the United Nations, simulated by my colleague and I, Terry Gish.
Throughout the tenure of this research seminar our work on Haiti aimed to model the administration of aid in the most realistic fashion. Our goal: To provide relief under the guise of the United Nations without encroaching on the turf of the United States or the Haitian government. With this we set about only to find administering aid to be deceptively difficult. Questions ranging from what kind of aid, to the types of its disbursement, offered a realistic perspective on the practical challenges these disasters generate. Administering aid is anything but simple, immediate or comprehensive.
To navigate the response effort, iterative processes were key. Constant repetition of thought highlighted the most realistic starting point; priorities were questioned and rearranged incessantly; and modeling rendered the final plan comprehendible. All the while the Haitian people were left wondering where their aid was. This point of tension, between the amount of aid demanded and the ability to supply it, was the biggest hurdle of the semester. It was also the most realistic.
Setting Priorities
We began by brainstorming the tasks that needed to be accomplished in Haiti. Not surprisingly, our list included most everything but the kitchen sink. (See Appendix A for the complete inventory of ideas.) Through class discussions, and a careful reading of “Value-Focused Thinking”, we came to appreciate the necessity of winnowing the list down. Some tasks we had listed were best suited to specific NGOs - others were to be undertaken by our partners, the United States and Haiti. As we studied the various iterations we had put together it became clear we were off track. To do everything is to do nothing well. Had we moved forward with our first draft, Haiti would likely be in the same shape it was immediately following the quake. By the third iteration we had deciphered what started to look like our fundamental objectives: Reconstruction management, emergency relief, freedom of movement, security and restoration of government capacity.
With these objectives in mind it became clear that our ultimate goal – or our strategic objective – was to provide aid to Haitians and their government. It was not to do everything. It was simply to engage the situation by leveraging the United Nation’s resources. Though, even this was discovered to be far from the most prudent approach. The United Nations has enormous capacity in an enormous range of fields. But was it critical that we leverage their expertise in reconstruction management at this stage of the recovery effort? What made more sense was to pick and choose the competencies that would directly improve the lives of Haitians quickly. This required a quick rereading of Maslow. After which, it was concluded that the United Nations would provide: Security, emergency relief and freedom of movement. These became our fundamental objectives. They were collectively exhaustive (the goals covered everything the United Nations could and should do), and mutually exclusive (each one was independent of the others).
We Have Objectives… Now What?
The primary focus of this class was the use of modeling to inform a constructive relief effort. Our tool - Logical Decisions (LDW) - was critical in visualizing the thought process and guiding progress in the construction of the recovery plan. With LDW Terry Gish and I created what is known as a fundamental objectives hierarchy. (See Appendix B for model). This hierarchy equates to a flow chart, with the strategic objective for the whole operation at the top, leading down to the fundamental objectives and their respective metrics.
The fundamental objectives shown in the model were chosen for their critical nature and feasibility of immediacy. Nothing was deemed less than 100% integral to the strategic goal of providing aid. Security was determined to be a pre-requisite for any other activity to take root. Similarly, freedom of movement in Haiti – Port au Prince especially - was included to shuttle aid in quickly and prepare the country for medium and long-run recovery efforts. Without transportation moving freely the disbursement of aid would be inadequate at best. Without security, the ability for aid to reach those in need would disintegrate at worst. Thus, these objectives were treated as fundamental to the plan. Lastly, emergency relief was considered fundamental because speed of disbursement, and the types of supplies demanded for basic survival, was key to successfully providing aid.
Having the right objectives is half the battle, however. Measuring them in an accurate, methodical way is the true aim. Again, we chose the most feasibly immediate and critical metrics to inform the interactions inherent in the model. Choosing these was the subject of much back and forth during the middle of the semester. Class discussions underscored the value specific and easily identifiable metrics have in end decision-making. After many iterations and much scaling down, we selected two metrics for each fundamental objective. Too many, we learned, and decision-making using the model as a reference would be no more valuable than blind choice taking. (See Appendix B for metrics.)
Breaking Down the Metrics
The second half of the semester was spent using another software tool titled: Microsoft Belief Networks (MSBN). Its utility derived from the ability to expand individual components of the LDW model, adding another useful iterative layer to decision making. In our case we chose to expand on the “airport operability” metric. (See Appendix D for this sub-model.) A key insight of the course came as we explored this expansion of our metric. For example: My colleague and I thought that airport operability was self-explanatory. If we “had it” things were on track. If we didn’t, we should get it. Simple. Exploring what airport operability actually requires in order to exist in more than name demonstrated the complexity each portion of the model, from the LDW hierarchy to the belief network, could possess if thought out properly.
In our case, it was difficult not to assume certain aspects of airport operability would be in place without specifically outlining them in the new model. After all, runways are, of course, critical, yet we took them for granted. A missing critical link such as runway infrastructure damages the applicability the model can have in any final decision-making process – the entire point of modeling. As we became more discerning in our selection process we started to over think it. From one iteration to the next our model would shrink and grow with no rhyme or reason. Did we include too much this time? Are we thinking of everything? The biggest hurdle with this utility is finding the point best described as “just enough”. For the iterative process can continue ceaselessly.
MSBN also allowed us to assess how each node in our model affected the others. (Find sample assessments in Appendices E and F.) Each node can either feed into other nodes (thus one will be dependent on another), or have other nodes feed into it. In many cases both will prove to be necessary. In our model, “personnel” extended its influence to three other separate nodes. In the assessment of the other fundamental nodes surrounding it, personnel influenced directly the various measurements. Another example is found by looking at the airport operability node itself (in this model it served as the strategic objective). This node was influenced by security, personnel, infrastructure and the maintaining of the airport’s perimeter. Determining airport operability, therefore, was dependent on the functioning of these other four attributes. This interaction - determining if security is truly critical to usable runway infrastructure - guided our thinking on other, more nuanced interactions.
Conclusion
Value-focused thinking is really iterative thinking. If done correctly, this process will organically uncover a useful model and its metrics. These will hopefully be instructive in making decisions for complex operations. The key is to use the model as an assessment of the available alternatives. Fifty iterations may lead you to the same alternative as the first. But the value lies in being sure. Using those fifty iterations allows a comprehensive and exhaustive survey of the choice landscape. Achieving this perspective is always the goal of the policymaker. Unfortunately, it is too oft not attained. Too frequently choices are made among alternatives inherent in the presentation of the problem, limiting their usefulness. It is by the process of iterative thought that more valuable policy alternatives can be uncovered. For the United Nations in dealing with Haiti, iterative thought helped select more critical ways to help the people, not the most obvious. In a disaster relief scenario nothing is more valuable.
No comments:
Post a Comment