Social Finance
Skoll ... ANDE Forum

SPEAKING OF SOCIAL IMPACT, WORDS MATTER

Burgess COMMENTARY

Peter Burgess

SPEAKING OF SOCIAL IMPACT, WORDS MATTER

ARTICLE HIGHLIGHTS:

• It’s hard to reduce a set of innovative initiatives that have affected lives, livelihoods, and living conditions in untold ways in a page, much less a spreadsheet.
• We need shorthand to describe and communicate the social goods that social entrepreneurs have created.
• We must develop social impact metrics and a system for governing them as if we were creating a language, not a set of statistics.

But try to summarize the impacts of even this small subset of social enterprises into a dataset. It’s hard to reduce a set of innovative initiatives that have affected lives, livelihoods, and living conditions in untold ways in a page, much less a spreadsheet. Social entrepreneurship is like pornography: we know it when we see it, but the parameters are hard to define.

But we must. We cannot build a global ecosystem that encourages and supports social entrepreneurs and their innovative approaches to pressing problems without being able to characterize an entrepreneur and their achievements with such summaries.

We need to be able to define goals and track progress toward them. We need to be able to aggregate experiences across social entrepreneurs in order to spot patterns in success and failures and adjust our efforts accordingly. We need to be able to succinctly describe the social impacts that we’d like to see and articulate the demand that attracts a supply of innovation.

In short, we need metrics for social impact. We need shorthand to describe and communicate the social goods that social entrepreneurs have created. However, we also need to be clear about the purposes that these metrics serve and the ways in which they shape the flows of financial resources, human effort, and creativity into meeting social need. Metrics ultimately create the map to move from finance through social entrepreneurship to development, and we need to make sure that this map remains accurate and inclusive.

The systems for describing social goods are starting to converge into a few approaches for representing social impact, with the help of initiatives including the Global Impact Investors Network’s Impact Reporting and Investment Standards, the Aspen Network of Development Entrepreneurs’ Metrics Conference, the Social Return on Investment Network and the Social Generally Accepted Accounting Principles, among others. It’s a historic moment for non-state development finance: akin to mid-twentieth century discussions in which nations agreed on frameworks for reporting economic growth and balance of payments as the basis for greater global collaboration in the international economy.

We must avoid converging on the wrong shorthand, and, even more importantly, learn from the past in governing the standards.

Metrics must remain languages and not become statistics. Not literally–this is a not an argument about qualitative versus quantitative data–but figuratively in terms of how we think about the shorthand and its limitations in capturing more complex underlying reality.

“Metrics must remain languages and not become statistics.”

The differences between language and statistics, as alternate sets of symbols, are subtle. Both attempt to describe an underlying reality to aid communication across distance, social networks, and common sensory experience. Both are meant to articulate abstract concepts. But there are differences, and they are important.

Statistics seek to distill; languages attempt to describe. Statistics measure particular attributes of a phenomenon that society deems important or representative. Sometimes these highlights turn out to miss the point–think of the difference between average (the arithmetic mean) and typical (the mode). Words, on the other hand, proliferate as new meanings become possible and relevant. We make up new terms when we need to, propagate them through social and professional networks, and someday the dictionary recognizes them.

We need this option to ensure that metrics for social impact keep up with rapid changes in the actual and known possibilities for impact. The advent of cheap mobile phones, for example, enabled new possibilities for building networks, matching workers to employers, disseminating information, and listening en masse to people who did not always have a voice. Climate science has highlighted the importance of reducing some kinds of air pollution to mitigate climate change, but the emissions’ climate impact cannot be compared to that of carbon dioxide without additional information about the context. Actions that respond to the science would not be recognized under a system that relies on the metric “carbon dioxide equivalent” as a proxy for climate impact.

Words’ meanings also evolve over time, while statistics’ metadata is designed to remain constant. Comparability over time is essential for tracking progress and the changing state of the world, but the social innovation ecosystem needs the ability to keep up with social context even more. Manifestations of workplace gender bias, potential remedies and perceptions of progress, for example, have changed substantially from the Mad Men era to the context described in Sheryl Sandberg’s Lean In and Anne-Marie Slaughter’s considered essay on the feasibility of “having it all.” So have the targets and priorities for social innovation in this area.

Languages also acknowledge variation across contexts; statistics claim universality. “Exclusion,” for example, has no meaning as a general term. The set of social occasions and economic and political processes that people may wish to be included in vary across contexts and over time, as do the specific barriers preventing them from attaining their goals. The signs of exclusion are often particular: one highly acclaimed research study used the fact of being forced into skinning dead animals and corralled into particular professions as indicators of Dalit marginalization from India’s mainstream economy, while food choice at weddings was considered a sign that social barriers were crumbling.

The “language” metaphor also highlights and draws attention to one of the main challenges for a global conversation about social impact: calibration of experiences, successes, and failures, across diverse contexts. Translation is recognized as an art as much as a technical matter.

Finally, language has the ability to handle communication in the first person as well as the third person. It has the ability to capture “I” statements–the direct voice of those whose lives social entrepreneurs seek to change–not just information about what “they” need. The conversation between the measured and the measurer is a fundamental tenet of social entrepreneurship.

Building a metrics language that recognizes innovation, unconventional methods, and diverse perspectives on what constitutes an improvement is first a matter of mindset. We need to restrain our comparative selves, the reductionist parts of our minds that seek comparability at ever-larger scale without thinking about what it costs in terms of information loss. Disability adjusted life years saved among women in Rwanda do not have the same social impact as disability adjusted life years saved among Peruvian men. This is true not just for the people whose life-years are saved, but also for the communities that they are part of, the social movements they may contribute to, the opportunities they may avail to change the world.

It also has two implications for the metrics governance. We are moving toward a global system for communicating about social impact that will, importantly, support common understanding of the meaning behind social impacts being discussed and studied. This dictionary for this language must be updated openly and often. The GIIN’s IRIS standards already have such a process in place, with a taxonomy committee to review new metrics, working groups to review developments that may require new terms to be added to the lexicon, and an a feedback form on the website to “suggest a new indicator.” We need to invest more in such efforts – expanding the taxonomy committee to include beneficiary and social innovator voices, for example; increasing the number of working groups; and increasing the outreach around “add an indicator.” Think of this kind of open process, but on the scale of the effort to update the Millennium Development Goals. The language would blossom.

The “add a new indicator” possibility also needs to become embedded as a standard practice throughout the ecosystem for assessing and tracking social impact. We need to leave room for all of the participants in the social innovation ecosystem to say something new. To not used the defined terms and format to describe the potential importance of their social innovation or their experience as a beneficiary of new responses to social need. To send a paragraph or a video instead of checking a box, giving a number, or selecting from a drop-down menu. Donors, impact investors, and incubators may have to spend additional effort reviewing reports that have a few such wild-card entries; managers may have to sit down with front-line employees who have new insights from their interaction with customers in the field, but these “wild cards” can be highly informative signals of the need for adding to the lexicon of social impact.

Nothing in treating metrics as language prevents rigor and precision. Nothing prevents accountability and clarity on agreements. Language does not stand in the way of learning. It’s more like working on grammar and vocabulary than accounting.