Social impact is any significant or positive change that solves or addresses social challenges (Business Impact, 2014). The essence of impact is simple – it entails identifying problems and implementing solutions. In this realm, human beings are especially useful, given our capacity to channel thought to solve practical problems. Even then, however, most social challenges are difficult to understand. They often manifest through systems (or networks) that are complicated, unpredictable, and equivocal. Given the human ability for overcoming challenges, tackling these problems requires people to brainstorm, implement, and test possible solutions. In this sense, the impact processes should be distributed over many agents, especially for the most pressing problems globally, “grand challenges”. Though distribution has many benefits, the actual extent a project is distributed and the amount it helps depends on a few factors. These include the problem’s difficulty and the resources available for solutions, namely, money, time, and capacity, especially leadership (Figure 1).
Origin & Usefulness
Distribution is the extent that tackling a social problem is decentralised, occurring through systems of persons and organisations, as opposed to single entities. Distribution occurs within and across three processes – design, implementation, and execution (Figure 2). In designing, distributed action involves engaging diverse stakeholders to analyse the problem, brainstorm interventions, and propose solutions. Distributed implementation uses many agents, including some from design (like individuals, NGOs, and ministries) to execute and test various solutions. Outcome distribution sees many persons, communities, countries, or other social groups benefit from solutions. In some sense, there is distribution in all impact projects, as even the most extreme case involves two entities, one to create and implement, and another to benefit.
Distribution helps solve more difficult social problems. The difficulty of solving a social problem corresponds to how complex, uncertain, and evaluative it is (Ferraro, et al., 2015). 3 A problem is complex if the system from which it originates has many agents whose interrelations and effects are non-linear. Uncertain problems are hard to forecast, due to the fickleness (or irrationality) of actors making the effects of change unclear. Evaluative problems have many interpretations. They transcend professional and geographical boundaries, changing meaning depending on the underlying beliefs of those reflecting. Grand challenges, such as combatting global mental illness, have all three characteristics (Collins, et al., 2011).
Grand challenges require more people along all stages of impact. For the design phase, the varied perspectives brought by multiple observers enable a better understanding of the problem so that solutions are better. As mentioned, the origins of grand challenges are hard to identify. They manifest in different ways by social groups and their interpretations are equivocal, often reflecting observers underlying opinions, beliefs, and biases. They require a multi-disciplinary approach as they manifest across professional boundaries. Having multiple perspectives in the design phase increases available information and fosters more targeted solutions. Distributed design also bringsthought diversity that pushes the innovative thinking useful for solving grand issues (Garud, et al., 2011). Moreover, incorporating people, especially stakeholders from the system, endows them with a sense of ownership, and incites them to facilitate change.
Even though having multiple stakeholders increases the chance of solutions working, the underlying unforecastable nature of systems means that effective solutions are not readily identifiable. Testing a range of solutions (prototyping) helps identify those that work. Distribution facilitates experimentation by creating groups that can each test their specific solutions. It also allows room to address the nuances of individual problems. Widespread testing enables more data collection to revise and improve these interventions. 4 For example, in a project to increase contraception usage among Ethiopian women, IDEO’s implementation had many teams for prototyping (Wyatt, 2019). Using teams improved the solutions, allowing IDEO to constantly refine them for the benefit of over 70,000 women (IDEO, 2020).
Social impact usually benefits many people instead of lone individuals. In this case, outcome distribution is across persons. The more widespread a problem, the greater the need for many to benefit. For instance, the Inter-American Development Bank’s (IDB’s) coronavirus relief project in Brazil in 2020 had to help millions of vulnerable people suffering from poverty due to the pandemic (IDB, 2021). 5 For the most part, projects have widespread benefits by scaling out, namely, increasing the number of persons that benefit from interventions. Nevertheless, targeting system leverage points can accomplish change more efficiently. Scaling up (changing policies or laws) or scaling deep (tackling cultural roots) can reduce the need for distribution. Scaling up fosters change as policies cover many persons within a jurisdiction while scaling deep helps future generations by inciting changes to norms and beliefs(Riddell & Moore, 2015).
However, not all social problems are grand. Some are more easily understood because they are simple, forecastable, and/or unequivocal. These problems may also be technical ones that have established solutions. Barbados fixing a sewage leakage on its South Coast in 2019 was one such smaller problem (King, 2019). The causes were well understood – the sewage pumps were in poor shape after not being maintained. The solutions were clear – repair the pumps and conduct more checks. In this case, a less distributed approach was better.
Prerequisites for Effective Distribution
Distribution is not effective if it involves the wrong agents. Ika et al. (2012) estimate a high proportion of World Bank projects fail due to poor stakeholder selection. It is therefore paramount to consider who is involved in impact. As it takes time to find the correct stakeholders, the process can be iterative. The best practices suggest choosing important actors from the underlying system during the design phase based on objective criteria (Chaudhury, et al., 2017). This happens after initial stakeholder mapping. The ones chosen can be those most interested in, affected by, or capable of addressing the problem (Green, 2002). At the same time, participant selection needs to consider the underlying power dynamics of the system. Chosen stakeholders may not depict the actual system but may instead represent the interests of elites or those facilitating participation (Taylor, 2001).
Nevertheless, even if the correct agents are chosen, distribution has diminishing returns for a problem of a given difficulty for three reasons – stakeholder competition, group inefficiencies, and stupefying bureaucracy – all of which increase with the distribution. Stakeholders in social impact often compete over resources and are motivated by self-interest. This results in the collective action problem, whereby individual group members, despite common goals, pursue self-interest, resulting in suboptimal results for everyone else (Ostrom, 2010). Significant research also shows that large groups are unproductive and suffer from cognitive biases. Groups of stakeholders, just like other groups, are prone to suffering from groupthink, riskier collective decisions, the Abilene paradox, and coercive persuasion (Cooke, 2001). Often, rules are introduced to control group members and prevent self-interest and bias. But, at times, these rules are too restrictive, reducing productivity and discouraging creativity.
To overcome these issues, productive distribution requires structures that encourage prolonged engagement among multiple stakeholders towards action. Hybrid forums are a good example of a working format (Callon, et al., 2009). These forums have strict standards for membership that result in better stakeholders. They also have distributed authority, lateral accountability, and mutual monitoring. Further structures appeal to individual members’ unique interests. This happens through multivocal inscriptions – routines, processes, and guidelines that are ambiguous enough to satisfy intrinsic motives. Using partially equivocal language that still guides enables different interpretations. The 17 Sustainable Development Goals (SDGs) are a great example. For instance, Goal 11 - “making cities inclusive, safe, resilient, and sustainable” - uses general language for unique adoption according to societal needs and individual interests.
Constraints: Money, Time, Capacity
Though additional distribution may improve a project, it is not always feasible as money, time, and capacity act as constraints (Figure 3). Engaging diverse individuals across a project, though it improves the understanding of the problem and the quality of solutions, is costly. These come in two forms, tangible costs (salaries, benefits, travel, accommodation, training etc.) and intangible costs (time, conflict, loss of autonomy, etc.) (Anggraeni, et al., 2012). Both costs increase with people’sinvolvement and, hence, with the distribution. As most impact initiatives have strict budgets, it’s important to optimise distribution within this limit.
There is also a significant intangible time cost due to deliberation, especially in design. In considering the views of diverse stakeholders, it is difficult to find common ground that satisfies all involved parties. This is especially true if stakeholders are engaged in more than one point in design (Cottrell, et al., 2015). Consensusis more difficult under greater distribution. As such, for more pressing issues, a less distributed design phase may be better. While more individuals for the implementation phase may enable a quick rollout, it requires careful oversight which most projects may lack, and which may be too time-consuming.
Many impact initiatives lack the capacity to create the environment that effective distribution requires. This occurs for many reasons, including overriding bureaucracy or rigidity, the need for control, and misaligned intentions. These are prevalent in larger international organisations like the United Nations (UN) (Weiss, 1982). In most impact projects, management serves two groups, those supplying resources and those benefitting from the project. Those supplying resources often have strict deliverables and deadlines that management must follow. This rigidity leads management to consolidate control centrally, as they are accountable for meeting these goals (Cooke & Kothari, 2001). Bureaucracy may also dictate which stakeholders are included in the impact process and prevent some organisation forms, such as hybrid models, from being used. Moreover, the goals of the project may contradict with management’sinterest. Those managing the project may simply want to accomplish the project goals, which are usually hard to measure, and not care much about creating the actual tangible impact that distribution facilitates (Bahadorestani, et al., 2020).
Need for Leadership
Functional distribution, therefore, requires adept leadership to foster healthy structures within these constraints. The persons responsible for leading change initiatives must be skilled systems/networks leaders who hone participatory, transformational, and facilitative leadership styles. Participatory leadership is characterised by leaders’ trust in subordinates and shared decision making through group participation (Sagnak, 2016). Transformational leadership entails creating a shared vision and empowering followers to help achieve it by connecting it to their deeper intrinsic motivations, such as their ideals and values (Burns, 1978). Facilitative leaders help groups and individuals become more effective, especially at self-organising, through building their capacity to reflect on and improve their work (Schwarz, 2002). A combination of all three styles is necessary for functional distributive impact.
Comentarios