A few days ago, Bill Easterly argued on Aidwatch that development comes not from solutions, but from functional problem-solving systems that motivate and facilitate solutions. In many ways, this is an argument for bottom-up, decentralized development, where people with local knowledge and first-hand experience of outcomes are the ones who determine what needs to be done.

On a deep level, Easterly is right, and not just because identifying the right solutions requires local knowledge. It’s also about sustainability, in a literal sense. In order for actual development to occur, solutions have to be imbedded in a local system that drives and sustains them without constant flows of money from NGOs or donor governments. Otherwise, it’s not development, but rather a permanent system of redistribution from wealthy countries to poor ones. It’s like claiming to cure an ill person by keeping her on life support. A person who is on dialysis and pain medication is alive and comfortable – and that is almost certainly better than the alternative. But, real healing would imply that we have figured out what’s wrong with the liver and fixed it, so that the body’s system is doing its own miraculous thing without mechanical intervention. Forcibly simulating the outcomes of good system can temporarily get you better health and education and housing, but it only goes so far.

A common response to Easterly’s blog was, “Great – but what exactly can we do to create better systems?” Good question. His response was that that was the subject for a whole separate book. That’s certainly true, but we’d like to start the conversation here with an overextended analogy/cautionary note, and a couple of ideas.

First and foremost: The way to build a better problem-solving system is NOT to simply take the body off life support. Now, that may sound obvious, and we won’t accuse Easterly of suggesting that all aid flows be brought to a screeching halt just because they’re not imbedded in good local systems. However, it does suggest two types of action that ought to be taken. First, use the time you are buying with life support to diagnose the problem. Second, look for solutions that provide the spark needed to jump-start the system.

Diagnosis is critical, because as all good economists know, the system isn’t failing for no reason. People act in accordance with the incentives and resources they face. If the result is a bad system – one in which the liver fails, education and health are poor, police and politicians are corrupt, and no one sticks their necks out to innovate or reform – then there is some reason that the bad system is prevailing. The best development research is diagnostic. What is the market failure? What is the cultural hold-up? These blockages have to be identified before a better system can take hold. At IPA, we always strive to conduct research in a way that identifies not just whether an isolated solution works, but why it works, so that we gain information about what was causing the problem in the first place.

Once problems are identified, the key is to find temporary solutions that provide the bridge from the bad system to a better one. These solutions may be the sort of narrow, non-generalizable, non-systemic projects that Easterly doesn’t like, but that can be okay under the right circumstances. Sometimes a one-off anti-corruption campaign, public education program, or infrastructure investment might be the blast that is needed to bump society out of a bad cycle and into a good one. The key is that if these solutions work well, they put themselves out of business. We should be looking for the antibiotics and defibrillators of development.