A beautiful nugget aligned with the “find what works” ethos came across Thomas Barnett’s blog* earlier this week. Read the full post here. He describes a start-up consulting company which is devising a new approach to strategic planning. It’s a collaborative wiki-based approach (hence the name — Wikistrat), with a further twist: a distinction between vertical and horizontal effects to prompt the planners into fully exploring potential scenarios.

The vertical/horizontal distinction deserves further elaboration. Think of vertical effects as happening in very short succession, while horizontal effects ripple out from them over time. Barnett’s classic example, given his focus on security issues, is 9/11. The WTC attack had quick vertical effects on the US economy, security policies, and military activities, and each of those vertical effects had horizontal effects reaching out from them. For example, the US government response to the anthrax scare in a “post-9/11 world” was to threaten to break patent laws unless German drug manufacturer Bayer increased production of Cipro; this shifted the terms of the debate on the production of generic HIV antiretrovirals and other drugs for developing countries. This was a horizontal effect that would have been hard to predict from the original vertical effects. Another example Barnett uses: the US toppling of Saddam Hussein’s Sunni regime (vertical) resulted in the increasing influence of neighboring Shia Iran (horizontal).

In international development, as in most fields, failure to consider horizontal effects detracts from effective strategy. A good example is a social service program run by an NGO or central government that crowds out local government, lowering its capacity in the long-run. Similarly, health programs that target specific diseases may undermine health systems that could address multiple problems, as Bill Gates has recently figured out. (NB: “Vertical” and “horizontal” have different connotations on health issues.)

This description makes “horizontal effects” sound like “unintended consequences.” They’re similar, but I think there’s a practical difference for strategists. That difference lies in how you go about considering the effects. Vertical effects (or intended consequences) are easy for strategists to consider, as they follow a relatively straightforward causal chain. If we then try to predict the unintended consequences, we must engage in a creative, exploratory process that is susceptible to biases. Re-framing the search in terms of horizontal effects, however, gives strategists an causal framework for rigorously analyzing the effects. Using this distinction encourages strategists to think laterally, moving perpendicular from the expected effects to investigate what they mean in other arenas.

This is important because causality does not move linearly. It spreads out in ripples. Our brains are not well-suited for managing such complexity, and most strategic situations have too much ambiguity for a computer to handle. So the vertical/horizontal distinction gives us a grid for the world: we lay some effects along one axis, and then do separate investigations for the further effects on the other axis.

Wikistrat is attempting to operationalize this analytical framework. The knowledge and skills necessary to explore the effects are spread across disciplines throughout an organization. Attempts to bring interdisciplincary teams together can be effective, but are susceptible to inefficiencies. We also know that contextual knowledge is vitally important in development. A platform like Wikistrat may enable strategic planners to bring all this knowledge together in an effective way. If it does that, we may see much more effective interventions and better strategy coordination in the future.


* I recently resubscribed to the blog, having unsubscribed a few months back when I couldn’t handle the overwhelming content flow (links to and commentary on 8-10 articles a day). Barnett was the author of The Pentagon’s New Map (2004) and several other books that have been influential in foreign policy and military strategy circles — but oddly seem to get little play in the aid world. He also made a compelling and hilarious TED talk a few years back. I don’t agree with everything he says, but I’m impressed by his ability to move across disciplines (economics, politics, military strategy, diplomacy, technology, etc.) and present a coherent worldview of it all.

  1. You’re right on target with this post.

    As for the aid community: I spent years in the mid-90s working these sorts of ideas with USAID’s Africa Bureau during their re-engineering, and it was hard. The preference was for always tackling root issues in this very long-term effort, and my argument was more like, okay, go with your first year of this mega plan, but then recalibrate the whole thing after 12 months, ditching what’s not moving forward and running with those that are and reconfiguring the plan.

    But that was a hard sell.

    I’ve also sought to sell the Developing-in-a-Box idea at multiple aid conferences and to multiple aid contractors. That is, by design, more a purposeful vertical reach into a developing economy to do the minimum to trigger enabling connectivity (the vertical) and then simply standing back and watching what horizontals emerge (how the people surprise you with what they do with that suddenly available connectivity), and then looking for capability gaps on that and intervening again. it’s a very minimalistic effort that seeks to engage in planner behavior only sporadically, while being in searcher mode (Easterly) as much as possible. It’s also an attempt to make things a lot simpler for civil affairs officers in terms of commander’s intent: you want them to have a few simple principles like, When you hit town, first thing you do is set up security just outside of town for a market, invite everyone in, and then see what synergy results. Then try to figure out the next intervention that moves that process along. It’s an attempt to provide a doctrine of process versus a doctrine of established steps, because every situation really is different, and that’s hard to train a young soldier on, so you try to structure a 10 percent planner and 90 percent searcher step function.


  2. I read the original post and your comment and I couldn’t agree more with your first point, but am curious about the second.

    I have been living in Uganda for the past two years working for an international research organization. I partner with Ugandan NGOs, coordinate with the government and all of my employees are Ugandan. Maybe this is only to justify my presence/job but I firmly believe that an international presence is very useful in research and so would like you to expand on your point “we will get better answers if more of the research comes from the developing countries themselves, rather than being conducted by outsiders.”

    If there is anything I have learned from working on the ground in research, it’s that most data is HUGELY inaccurate. Everyone can improve. And I’ve found that an outsider’s presence can be very useful. For example, government schools get their funding based upon enrollment, so that’s a strong incentive to include ghost pupils on the registers collected by the government. Or, with very sensitive information, people are often very unlikely to disclose something like number of sexual partners to anyone in any way connected to their community.

    This is not to say that a perceived international presence can’t also be problematic. If people think they will receive assistance from some international NGO or donor they will definitely alter their responses.

    As with almost everything, involvement and key-decision making power should be defined by local people and those affected by such decisions.

    But research I think can usually benefit from an outsider perspective. This definitely goes both ways. International monitoring of US elections was a great idea. An outsider can have a unique take on a situation, be able to see it from different angles, have useful comparisons. This perspective is of course useless without the involvement of local people, but I think when taken together, it is definitely valuable.


Leave a Reply

This site uses Akismet to reduce spam. Learn how your comment data is processed.

%d bloggers like this: