Complexity and Poverty

The discussion on complexity and development continued on Twitter. I said that in my my previous post on complexity and development didn’t go far enough to raise what I think is a often a very overlooked point by complexity ‘theorist’ – which is that both good and bad things arise from emergence and self-organization. As it pertains to development, I think some of the most pressing problems like vulnerability, poverty, hunger, inequality, and corruption are persistent precisely because they’re deeply embedded inside complex systems. More specifically, they’re resilient – even the most powerful, well-funded interventions fail spectacularly.

As with the previous post, I think Owen is still feeling out the exact position he’s staking. I applaud him for being willing to get his hands dirty, as it were, by continuing the discussion and amending his argument as it unfolds. His response to the above idea was that, “poverty is the ‘no system’ situation.  Non-poverty requires a complex adaptive system…” He later backed down from the ‘no system’ claim.  He nuanced the claims later with, “many people live in systems which are not complex in a way which delivers them a good life,” and agreed with me that, “not all adaptation is positive. But you can’t end poverty without it. Complexity is necessary but not sufficient,” and that “rich countries and poor countries [are]… just a different kind of complex.”

This begs the question: what are the units of analysis? I agree that there are ‘healthy’ and ‘unhealthy’ socio-economic systems that either promote or hinder capabilities and positive freedoms. A baby born in Canada is likely going to have a longer, happier, wealthier, healthier life with a much wider array of opportunities than if it were born in rural Somalia. I agree with his main point, which has been raised by others, that healthier socio-economic systems are emergent properties of complex systems. There’s no one single key to having a society like that. To reverse engineer the emergence, I would say that it’s the result of having a society where desired action is encouraged with successive layers of positive feedbacks, and negative action discouraged with negative feedbacks. Social safety nets are in place to make sure individuals (or entire demographics, as it often happens) don’t become stuck in endless poverty or crime cycles.

But two points of disagreement remain. At the level of socio-economic systems, I disagree entirely that Canada has a complex adaptive system and leave pharmacy review Somalia doesn’t. They both do. The primary difference is that Canada has a system that rewards ‘good’ behaviour; Somalia has a system that rewards militancy, piracy, and corruption. It’s difficult to run a legitimate business when Al Shabaab might come to town any day. Even without them, it must be difficult to keep valuable goods in stock when your neighbours commit armed robbery for a living. I imagine that it’s also very difficult to keep a school up and running for the poor when there is no government to provide safety, tax revenue, or accreditation. This is all to say that militancy, piracy, and corruption are all adaptations inside an unhealthy system. People will do whatever they need to do to survive and thrive – regardless whether or not the system provides ways of doing it that don’t require violence. Put a different way, Canadians have a choice architecture for personal growth and feeding families that doesn’t require or reward violence.

My second disagreement is an implied disagreement, as I might be reading Owen’s argument incorrectly. My own experience and recent literature is showing that there’s a turn in global development, such that ‘middle income’ countries are coming to dominate the international system. Two thirds of the world’s poverty is in middle income countries like China, India, the Philippines, South Africa, Brazil, etc. These are systems where some people are doing fantastically well and others are doing terribly, though they’re – on average – not as absolutely poor as they were a few decades before

Further, there are fewer and fewer national socio-economic systems that look ‘healthy.’ The entire Eurozone is in crises. The UK is dealing with recession and austerity. America has a massive inequality issue, political deadlock, and a ‘jobless recovery.’ I used Canada as an example because it’s one of the only unambiguously ‘healthy’ countries I could think of. There are few systems where everyone is well-off or terrible, as is implied with the “rich country / poor country” dichotomy Owen uses. So I think we need to talk about poverty as an emergent structure inside the system, not the emergent structure. Some people are doing quite well inside the system while others fair much worse.

Which means we should lower the unit of analysis to either sub-groups (i.e., ‘socio-economic status’) or individuals for a more meaningful deconstruction of poverty. It’s at this level that I think complex adaptive systems analysis has the most to offer. Why do some people do well inside the same system that crushes the opportunities of others?

Which leads me to the example that brought me to complex systems analysis to begin with. I study farmer education in the rural South. With a few exceptions, the driving purpose of government’s supporting farmer education is to increase yield. This is, again, almost the exclusive thing groups like CGIAR work on. I bought into this paradigm at first, as it seems simple enough: the world needs more food. A farmer who grows more makes more, thus helps lift them out of poverty. It also provides more food for the urban poor.

But I learned that it’s far, far more complex than that. I began questioning whether it was really in a farmer’s self-interest to raise yields. Consider De Beers, the diamond maker, for a moment. They use their monopoly position to create scarcity by stockpiling their diamonds. Further, you see it in the discourse of CGIAR literature. Their logic isn’t that poor aren’t fed just because there’s more food, but because it’s not cheap enough. But why would any producer want to make less money for their product, especially if the margins are getting worse? 

I also began seeing that increasing yields works well for, and is often easier, for some farmers more than others. The only farmers who do well are, in fact, ‘growers.’ One can, and some do, become wealthy on such tiny profit margins by growing tons and tons of food, mechanizing production, buying chemical inputs in bulk, and insuring crops.

The opposite would seem true of smallholders: inputs cost more for them, whether it’s seeds, fertilizer, or pesticides. This is true both proportionally and absolutely, as smaller quantities are more expensive and profit margins are smaller. Smallholders are also more vulnerable to price swings, which are increasingly common (a feature of the complex global economy that they’re nested in). All of this means that smallholders are likely maximizing strategies to increase stability rather than production. This manifests itself in places like Kenya, where I was told “lazy” farmers don’t invest time and energy on fertile lands they own. But trying to ‘get big’ is a very risky strategy. Risk, when you already have very little to gamble with, can come at a very high price.

We’re left with a complex system where there’s almost nothing a smallholder can do but “get big or get out”, as Earl Buzz put it. The third or is to remain poor in a system that will likely not change in this lifetime. So smallholders adapt to this system by either trying to produce more (which is risky), picking up and moving to urban slums (where they now depend on the very same depressed food prices that drove them out to begin with), or using the low-risk, low-cost methods available to them to produce only as much as they need.

Smallholder poverty is the emergent structure of the system, as is rural-out migration and the slums that come with it. I believe most interventions aimed at helping them fail, and often cause more problems, because they fail to note the complexity of the situation. Increasingly yields would be an effective anti-poverty strategy if wasn’t nested in an interconnected adaptive system wherein increasing supply usually means a decrease in prices. The system isn’t static. It changes and moves with everything we do. I also think it wasn’t linear, lest-wise this analysis would be conventional wisdom. There’s clearly a point in which increased production tips from being helpful to farmers to being harmful. I think non-linearity is also found in that people don’t often look at slums and connect it any way to higher agricultural productivity. Something we generally just call “development.”

The complexity arises from the diversity of agents (buyers, growers, smallholders), and the interconnections and interdependency of the marketplace – what the large landowners are doing impacts the smallholders. But Owen says that perhaps I don’t focus enough on the linkages. I think the implied argument is that the more linkages the better, though I might be reading too much into a short text. If we use the above example, it’s all about the linkage. It’s how international institutes, governments, and and other farmers are (I feel) hurting smallholders while ostensibly trying to help them. Further, there’s also macro- linkages that I didn’t get into: like how European and American agricultural subsidies are awful for smallholders in Southern countries. By almost any objective measure, the less linked a Filipino or African maize farmer is to American farmers, the better.

Nearly every recommendation I would have to a government to strengthen the position of smallholders would involve de-linking them from the global economy. I don’t think agriculture is the only example. I think a lot of successful anti-poverty initiatives would require systematically mapping complex systems and removing or mitigating the links that produce fragility and vulnerability, but also doing them in ways that has the least negative externalities on other parts of the system. It’s not an easy task, but I believe it’s a more accurate approach than assuming that the world’s poor and vulnerable don’t already have a substantial amount of complexity dominating their lives.

Finally, I presented two situations (Somalia and smallholder farmers) in which I adaptations inside complex systems produce an emergent property of poverty. I would recommend, however, reading through Poor Economics with a theoretical lens of complexity. In it is a series of example of why under-education, healthcare choices, and malnutrition are all emergent properties arising from what I would call “local logics.”

One Comment

  1. Thank you for your analysis on this topic. It’s difficult to understand the complexities and the “interconnections and interdependencies” of the market system. It is true that de-connecting economies from global trade is what protects small-holders within the system. The United States has been doing this for decades with agricultural subsidies. Then these subsidies create producer adaptations that are based on phantom/false rewards. Unless government takes measures to “protect their own” it’s difficult to create conditions for these producers to succeed. I’m most familiar with this dynamic as it pertains to agriculture and other things like the textile or automobile industry, but how do you think it pertains to something like IT? I don’t think sub-saharan African countries are going to start producing computers, so how does what you are saying apply to this industry?

    September 29, 2012

Leave a Reply