Yesterday, the OECD hosted a roundtable on “Conflict and Fragility -- International Engagement in Fragile States -- Can’t We Do Better?” to accompany their report on the issues.
Ten principles were defined, along with a brief assessment of each:
- Take context as the starting point -- Partly off-track
- Do no harm -- Off track
- Focus on state-building as the central objective -- Partly off-track
- Prioritize prevention -- Partly off-track
- Recognize the links between political, security, and development objectives -- Partly off-track
- Promote non-discrimination as a basis for inclusive and stable societies -- Broadly on Track
- Align with local priorities in different ways in different contexts -- Partly on track
- Agree on practical co-ordination mechanisms -- Off-track
- Act fast… but stay engaged long enough to give success a chance -- Off-track
- Avoid pockets of exclusion -- Off-track
So, the results are emphatically not encouraging: 8 out of the 10 of the principles are basically off-track! In fact, that would fall under “quite grim” in my book. If 80% of the cookies I baked were burnt, I’d be concerned about my skills as a cookie-baker.
I’ve had the pleasure of meeting some remarkable people in the development and humanitarian arena. In unison, they speak about the organizational dysfunction of the particular institutions in which they work, and a general frustration with the way the “whole system” works. On the whole, they are dedicated, ethically sensitive, jaded, overeducated.
I’ve also had the complacent pleasure also of standing aside, taking a sip of my martini, and watching a small slice of the action from the sidelines. There’s a bit of the theatre of the absurd to be confronted with the acronyms, jargon, the name-dropping, and the self-contained caste system in any specialized domain, whether Development or Management Consulting or Architecture.
I won’t comment here on what I think of the results of the report. ...Except to say that when they ask “Can’t we do better?” -- my intuition tells me that even the authors of the report and the wielders of the framework know that it's a rhetorical question. So that’s interest of the report itself, in a nutshell.
And it doesn’t seem agonizing enough. Why should we begin and end with rhetorical questions? The ten “principles” make sense. When I listen to the panelists, it makes sense. When I talk to the individuals dedicating themselves in the field or at their offices, they make sense and their individual “good intentions” emanate. The strategies are often thoughtful, coherent, and even compelling. Why is it so hard to make an impact then?
In the post-panel mingling, we stand in small circles, eat our burnt cookies, and muse about how inadequate all this is. How “technical solutions to political problems” are insufficient. How the “the priorities of donors” and the “needs on the ground” interact in some weird alchemy of dysfunction to make the execution of the best-laid (and financed) plans ineffective. How the incentives are skewed. And on and on. So the discussion itself becomes rhetorical, ineffectual, water swirling in a beach-side bucket while the waves comes crashing just outside it.
From an outsider’s perspective, it seems to me the the whole make-up of the people, institutions, and approaches that have evolved in Development have taken little notice of some larger paradigm shifts that are occurring that put in question some of the assumptions about how organizations should be managed, projects executed, people educated, and so forth. What needs review is not just the individual policies and programs and actions and principles, but the entire domain culture of the development and aid.
What does a domain culture entail? Well, for example:
- The internal organization of the institutions. Example: Walmart is organized differently from a law firm. They have different cultures. The way in which an institution works, defines its functions, and incentivizes its individuals comprises the most immediate aspect of the domain culture for most people.
- Inter-organization structure. Example: the fierce competition among leading technology companies is different from the way in which the members of OPEC operate. The structure of the way in which institutions relate to each other also is a manifestation of a modus operandi that characterizes the domain culture.
- Education, ideology, and theoretical underpinnings. It comes down to the people and what they’ve internalized as values and viewpoints. The various ideologies and underlying theories are formalized, transmitted via education and immersion in the milieu and accepting the narratives that dominate the particular field. Examples: Marxism vs. Capitalism. Secularism vs. Religious fundamentalism. Apple computer vs. Microsoft.
- Accumulated habits and tendencies. As opposed to formalized theories and articulated ideologies, there seem to be certain ticks and traditions which also constitute the culture of a particular domain. These get institutionalized over time, but there seems to be little theoretical underpinning for this. Why do we need a Ph.D. to work somewhere-- is that really a valid filter for performance? Why do we need to go through some painful procurement process? Why can’t we change these things? Well there’s no reason why we can’t, but we don’t.
There are others, but I hope the idea of a domain culture is clear from the above.
And my point is that the Development community seems held to a culture that seems from the 50’s (Ok I’m making a caricature) and incrementally updated, but only incrementally. It’s like we live in the world of quantum mechanics, but some of us still operate solely based on the rules of Newtonian mechanics.
Some of the changes that may serve to “update” the culture in the development field include varied tidbits like:
- The realization that uncertainty is a non-negligible effect. This should have impact on how we design actions, how we measure impact, and how we deal with when the uncertain rears its ugly head (often). Current programs seems to be based on a simple cause-and-effect type approach. If I do X, then we should good impact Y. But what we’re learning in the broader world is that causal relationships are not only non-linear, but that uncertainty plays a determining role in the outcomes of the best-laid plans. Dealing with tail risk is a central theme in finance today -- something that was unheard of just 15 years ago. Unless the program designs build-in a level of stochastic tolerance, the outcomes will continue to be derailed by uncertainty.
- Management structures within many of the organizations (from what’s been recounted to me mostly by people working in the UN system) are from the days of GM and (old) IBM. Hierarchy, slow approval processes, torturous vetting. This is necessary because of course we’re touching people’s lives, not selling gadgets. It is true that the lack of market pressure and the particular supply-demand curves of development actors as well as the high stakes make for a different set of rules in the this field. But the Development world might take some inspiration in understanding how companies today are trying to be nimbler, flatter in structure, and more innovative. The notion of disruption as something positive is coming of age. Fifty years ago, the idea in a typical corporation was to optimize the feedback loop between planning > control & action > measurement of results. This approach is being reconsidered for alternate models that place importance on agility, making small mistakes to avoid the big ones, and to accept that fact that “control” in the traditional management sense is more an illusion, and that the results are less definitive than one could measure. Actually these are just an analogous evolution from the “command-control-understand” model originating in enlightenment Europe to the more modern perspective of “measure, account for the uncertain, and take agile action” approach.
- Taking account of the granularity. Ultimately what we are trying to impact is people’s lives. These are affected in broad strokes by things like policies and programs and infrastructure, but the final unit of impact is the individual person. Thanks to the data collection and management capabilities we have now, we can at once treat, measure, and manage at once the macro- and micro- without giving up on the descriptive richness. There is less need for the reductionist approach for finding Key Performance Indicators and taking averages when the distribution is highly skewed (not that these should be thrown away). The traditional point measures and moments of a distribution can be reinforced with a handling of the entire dataset as such… This point is quite abstract and it would take a while to expound. As a simple example, imagine we have a certain distribution of income. Then traditional approaches would collapse that distribution into a single number like the ‘average,’ or a set of numbers that might include, say, the median and the mean. So we can speak of the “rise in average income” … while safely ignoring the fact that the income inequality has actually increased. (Now, no one these days is naive enough to report that, so it’s an extreme example.) However, my point here is that are ability and willingness to take in the whole distribution as such is made possible by advances of statistical techniques, computing power, and a realization that details are actually important. Noise might be signal disguised. This is a crucial shift from “let’s simplify the world and make a model or a framework” to “let’s see the messy world as it is, and see how action X might affect the entire distribution”.
Other areas which have bearing include the insights brought to us from behavioral economics, social media, big data, as well as alternative organizational and process models (such as “lean production / lean start-up”) can give us some hints into how to begin considering new ways to rejuvenate the practices and frameworks of the development field.
I am not close enough to really have insight into more precise or systematic approaches for how this cultural evolution can take place. Nor do I have a sense of how some of these shifts can be introduced and propagated. Because a shift needs to first impact people, who then design the institutions that then create projects and initiatives, it will no doubt be a long process. The institutions exist already -- they won’t change their charters easily. The temples are full of believers, and heretics will not be welcome. Meanwhile, for many who work in the system, there will continue to be a sense that something isn’t quite right; that we could in fact do better… and such reports will continue to be published.
While the development world has taken these changes up to a certain extent, the penetration is superficial and formalistic. Organizations use social media and tweet (but to what end?). There are papers on behavioral development economics. Frameworks acknowledge the complexity and multidimensionality of issues. However, the core culture underlying Development as a field remains stultified. It’s time perhaps to step back and open the windows; bring in a breath of wind.
Unless the participants in the "system" at the institutional and individual levels begin to entertain an evolution of the domain culture and possibly embrace a new paradigm, we will be seeing a lot more of these “Can’t we do better” reports -- the actors themselves being aware of the issues and yet somehow helpless in the echo-chamber of their familiar worldviews.