GlobalGiving has hosted about 7500 projects that describe the work of organizations in 15 categories. I recently noticed that you can use this huge set of data to deduce which types of community work are clearly defined (by the implementers) and which are quite diversely defined. First, take a look:
Projects are the number projects in that category. Each project is described by the implementing organization, and each organization chooses some keywords that best describe that main focus of that project, as well as choosing a category that the project falls under.
Phrases are the number of unique phrases that can be extracted from the project keywords, with some filtering involved to that only legitimate keywords and two-word phrases remain. (Legitimate is defined by multiple projects using the same words).
The relative diversity of each of these categories can easily be seen by taking the ratio of the phrases to the projects:
Gender (projects focusing on women), Climate change, and human rights are the fuzziest and most diverse categories.
Does that surprise you? Probably not. But it should be food for thought that the reason these three categories appear at the top of the diversity list is that the nature of the problems involved are complex, and thus the scope of possible interventions organizations are currently implementing are correspondingly diverse.
Art is not complex. This surprised me. Part of the lack of diversity in how art projects are described is due to the small number of total art projects, but look again at the table. Hunger and Democracy projects are also quite rare on GlobalGiving, but both of these have quite a bit more diversity and fuzziness in how their implementers describe what they are doing.
The sum of all 7,469 projects is more diverse than some categories, and less than other categories. It falls right in the middle. I consider this a good reference point for interpreting the relative diversity of all categories.
Health projects are much more diverse than education projects, in terms of how they describe what they are doing.
Fuzzy concepts and fuzzy outcomes
A while back I was consulting with two different international organizations whose missions were to promote human rights. “Well, how do you define human rights?” I remember asking at the start of each conversation. And in both cases, the organization lacked a particularly clear definition. One of them said, “It is whatever our grantee organizations are doing with our human rights grants.”
Back then, that was a useless, circular definition. But today, after pulling a dictionary of phrases from 327 human rights projects, weighting each phrase according to how often self-described human rights projects use it, I can define it. Here is a weighted definition of human rights that can serve to replace the operational fuzziness that I once encountered. This is word crunching, and this is going to provide international development with structure for their ambitions in the future. (The numbers are normalized weights for each phrase (1-100 scale)):