There is growing consensus from experts and practitioners that international aid needs dramatic change in order to become more effective, more transparent, and less vulnerable to criticisms of waste, corruption, and even net harm.
A new school of thought from the emerging discipline of complexity research supports a major rethinking of aid and the relationship between “outsiders” and the communities outsiders hope to help. Despite this, reform momentum seems to remain concentrated around improving – as opposed to challenging the existence or role of – aid agencies and the government bureaucracies throughout the developing world with whom they partner.
An alternative interpretation of the complexity research suggests minimizing the economic role of aid agencies and governments in developing nations, thus maximizing economic freedom for individuals to steer development towards a more reliable path to prosperity.
The top-down, foreign-led design of international aid programs has been long criticized for its overreliance on simplistic, linear strategies and for being fundamentally at odds with the complex way nations actually develop.
Beginning with Nobel Prize development economist P.T. Bauer’s heretical 1971 book, “Dissent on Development,” a long list of intellectual heavyweights and experts have rigorously challenged the aid industry’s juggernaut-like way of trying to help the world’s poor. Nobel Prize economist Eleanor Ostrom dismissed the contribution of aid agencies complaining they may at times say the right things but they lack a deep understanding of the problem.
Even more damning, Richard Dowden, director of the Africa Institute, once suggested that modern aid agencies can be even less thoughtful about local people and cultures than were their colonial forbearers. What’s more, Angus Deaton, whose Nobel Prize was awarded in 2015, concluded in his 2013 book, “The Great Escape: Health, Wealth, and the Origins of Inequality” that aid agencies might be doing more harm than good in the communities they are meant to serve.
Encouragingly, several of today’s mainstream academics and even some aid practitioners sympathize with those criticisms and advocate energetically for reform. However, their proposed solutions might not go far enough in demonstrating they have truly learned the lessons of aid’s failures. Instead, what appears to be an unshakable commitment to the central roles of aid agencies and national governments in steering development might represent just another version of the same fatal conceit that plagued today’s aid advocates’ predecessors. If that’s true, the question then is how might we better use their insights to further our common goal of seeing the world’s poor prosper?
Confronting the Inefficacy of Aid So Far
Beginning in 2003, the Organisation for Economic Co-operation and Development (OECD) has spearheaded a global effort to improve the effectiveness of aid, specifically addressing aid’s historic paternalism, lack of transparency, and failure to demonstrate significant results. To date, more than 160 countries representing both donors and recipients have signed on to this effort and have agreed to a new set of principles that continue to be refined. The most recent version is enumerated below:
- Ownership of development priorities by developing countries: Partnerships for development can only succeed if they are led by developing countries, implementing approaches that are tailored to country-specific situations and needs.
- Focus on results: Our investments and efforts must have a lasting effect on eradicating poverty and reducing inequality, on sustainable development, and on enhancing developing countries’ capacities, aligned with the priorities and policies set out by developing countries.
- Inclusive development partnerships: Openness, trust, mutual respect, and learning are at the core of effective partnerships in support of development goals, recognizing the different and complementary goals of all actors.
- Transparency and accountability to each other: Mutual accountability and accountability to the intended beneficiaries of our cooperation, as well as to our respective citizens, organizations, constituents, and shareholders, is critical to delivering results. Transparent practices form the basis for enhanced accountability.
The observance of those principles is meant to subdue the influence of donor countries to make more room for recipient countries’ own knowledge about their specific and complex needs. The hope has been to start doing things better, to learn from past mistakes, and to introduce better listening and better measurement to focus on results.
Unfortunately, the pace of change has been disappointing. The 2016 report of The Global Partnership for Effective Development Co-operation, the organization tasked with monitoring progress on the OECD goals, offers mixed results and little cause for celebration. Despite its upbeat tone, the report underscores the persistent, and perhaps inevitable, challenges associated with outsiders trying to help without interfering. As just one example, progress toward the elimination of perverse practices such as conditional support (strings attached to aid that serve donor country interests even though they might undermine local priorities) remains well below targets.
While the principles set forth by the OECD do speak to important insights about the type of solution needed, some experts are starting to question whether we fully understand the complexity of the problem itself.
Aid and Dynamism
Oxford University Press, 2013.
One of the more helpful ways to think about why traditional aid has not worked in developing economies is explained in Ben Ramalingam’s 2013 book, “Aid on the Edge of Chaos: Rethinking International Cooperation in a Complex World.” He compares international aid’s approach to the misguided forest fire prevention policies that dominated late 19th-century national parks. He explains, “[f]ire policies have not protected the forests but in fact have placed them at considerably greater risk,” in part because the intervention to prevent fires itself disrupts the natural process of emergent forest health. The counterintuitive warning then is, “preventing small fires can lead to large fires.”
Like ecological systems, economies are also complex in the technical sense meaning a linear strategy employed to influence economic outcomes will fail to account for the near countless factors that threaten the outcomes such plans seek.
Those factors are not only exceedingly numerous, they are interdependent, so how they affect each other cannot be reliably predicted using traditional methods that tend to isolate variables for testing and prediction. As a result, interventions driven by linear strategies, even those supported by evidence, run afoul of the way complex systems actually function.
This confounding dynamism has inspired a new line of research under the broad banner of complexity science with an aim to discover how we can do development better. In one sense, this approach is a nod to Friedrich von Hayek’s 1974 Nobel Prize lecture, “The Pretense of Knowledge,” in which he calls on the scientific community to recognize its own limits in applying known scientific tools to complex problems. And yet, it’s also an attempt to transcend those limits by improving the scientific toolbox that experts employ in the face of that complexity.
This tension is nothing new to scientific thinkers. In the 2015 New York Times bestselling book “Super Forecasting: The Art and Science of Prediction,” authors Philip E. Tetlock and Dan Gardner write, “For centuries, scientists had supposed that growing knowledge must lead to greater predictability because reality was like a clock – an awesomely big and complicated clock but still a clock – and the more scientists learned about its innards, how the gears grind together, how the weights and springs function, the better they could capture its operations with deterministic equations and predict what it would do.”
Tetlock and Gardner explain how the work of meteorologist Edward Lorenz showed, as early as the 1970s, that “in nonlinear systems like the atmosphere, even small changes [or errors] in initial conditions can mushroom to enormous proportions.” For economic development, this means underpowered and simplistic strategies are more likely to lead to ruin than they are prosperity. To replace the clock comparison, Tetlock and Gardner offer the concept of a cloud as a more fitting representation of complex reality. While a lot might be known about clouds, we’re unable, for example, to predict the shape of one.
They conclude, “Unpredictability and predictability coexist uneasily in the intricately interlocking systems that make up our bodies, our societies, and the cosmos. How predictable something is depends on what we are trying to predict, how far into the future, and under what circumstances.” That line between unpredictability and predictability echoes the “edge of chaos” that Ramalingam suggests is the sandbox we must play in if we want to get better at economic development.
In practice, this means recognizing that complex systems analysis must eschew the search for a single solution, what Ramalingam calls “best practicitis.” He calls on aid agencies to first cure themselves of this ailment and, secondly, to adopt a new mindset that looks instead for self-organized, emergent solutions within specific contexts. Those solutions, when identified, cannot simply be bottled for redistribution. Instead, aid practitioners must seek ways to accelerate the iterative nature of innovation diffusion among networks to spread the solution’s impact.
For example, he tells the story of Jerry and Monique Sternin, who traveled to Vietnam to work on improving infant health. They discovered almost by accident that some poor families were bucking the trend of malnutrition through a series of unique habits. Those outliers, referred to in the literature as positive deviants, represented for their communities the best hope for change. The solution then, though unrecognized, already existed there, on the ground, inside the complex system. Importantly, as outsiders, the Sternins decided not to try to reinterpret those habits into a list of best practices to be taught paternalistically to others. Instead, they facilitated increased network engagement among the deviants with the broader population so that they could accelerate the diffusion of the relevant knowledge. It worked. Malnutrition rates went down 65% to 80% among a total population of 2.2 million.
The Sternins concluded, “‘Best practices’ evokes the immune system rejection response to a foreign body, and there is also a social immune system rejection response to outsiders coming in and saying: ‘Hey, look at the answer here. We have already solved the problem.’ With [positive deviance], the solution and the host in a sense share the same DNA so you don’t get that push back or rejection.”
The knowledge relevant for finding the solution in a complex system then is what the deviants discover, as evidenced by the outcomes they exhibit, combined with the knowledge others possess who share the same “DNA” and allows them to recognize, value, and adopt the innovation for themselves. It’s precisely that knowledge that makes it productive for each of the locals in the population to be the key decision makers when it comes to solving their own problems.
Productive Knowledge in Complex Systems
An example of product complexity mapping. "The Atlas of Economic Complexity," MIT, 2013.
Economist Ricardo Hausmann and his co-authors in “The Atlas of Economic Complexity: Mapping Paths to Prosperity” refer to this knowledge as productive knowledge. This is the knowledge that is critical for a solution to emerge in a complex system and, for Hausmann, it is the knowledge that produces saleable economic activity. This knowledge is distributed widely across individuals and discovered only through the dynamic process of those individuals engaging one another in various contexts, such as in a market, making decisions on their own behalf.
Of course, the word “distributed” is a misnomer since it implies some act is undertaken by some actor to do the distributing. In fact, the critical knowledge already exists within the individuals themselves as a result of their various experiences and cognition. Furthermore, much of that knowledge is tacit, meaning those who possess it might not even be conscious of it, might not be able to articulate it, nor would they necessarily recognize its influence on how they evaluate options in any decision set they face. Lastly, and very importantly, productive knowledge cannot be obtained by outsiders to exercise on others’ behalf.
Unlike Ramalingam who seeks to reform aid agencies so they can better manage complexity, Hausmann places his hopes for reform on countries’ ability to achieve product differentiation. He writes, “The policy message for most countries is clear: create an environment where a greater diversity of productive activities can thrive, paying particular attention to activities that are relatively more complex or that open up more opportunities.”
Using international trade data, Hausmann and his co-authors created what they call the Economic Complexity Index to measure, by proxy, the productive knowledge of a country by assessing the complexity of the products it exports relative to other countries. They claim that economic complexity drives long-run levels of income and growth. To provide some guidance on what to do with that insight, they also developed the Complexity Outlook Index which measure the potential set of products a country could be making, weighted by their complexity, and based on their own productive capabilities. Taken together, Hausmann et al. claim their measures predict future growth better than virtually any other variables tested in the growth literature.
While Hausmann et al. prescribe no role for aid agencies, per se, in using their tools, they do intend for their indices to be used widely by anyone invested in economic development, including firms looking to relocate or diversify. Despite that broad invitation, most of their suggestions for action lie within the purview of national governments and one hopes the biggest impact of their work won’t be the expansion of industrial policy throughout the developing world.
The last thing developing countries need their governments to do is to try to pick winners and losers in the marketplace. Development economist William Easterly’s research on international trade has revealed the “power law” of export success, which observes that for some phenomena large outcomes are more likely than would be suggested by normal distributions. As a result, the chance of picking a winner actually goes down exponentially the higher the threshold used to define success. For developing countries especially, this means the stakes for such a gamble are too high.
Improving Governments to Improve Aid
In 2017, Harvard University’s Matt Andrews, Lant Pritchett, and Michael J.V. Woolcock published, “Building State Capability: Evidence, Analysis, Action.” The book is both an argument for focusing on OECD’s second principle, capacity development, as the linchpin of success in economic development, as well as an improvement guide for government bureaucracies.
The authors argue that until governments in developing countries are more sophisticated and competent, they never will be able to deliver the kind of public services aid programs are meant to support. For them, this is the key to economic development. A central and persuasive tenet of their prescribed strategy is recognizing that institutions cannot simply be transplanted from foreign countries; a practice they call isomorphic mimicry. Because they are part of complex systems, institutions have to be developed locally, in context, so that they represent the needs, culture, and practices of local environments.
To help governments improve their capacity then, the authors have developed a program for bureaucrats called PDIA. Rooted in the insights of complexity science, PDIA stands for Problem-driven Iterative Adaptation and offers a series of exercises for government teams to work through to discover their own unique pathways to bureaucratic competence. The steps in the process are worth repeating here since they echo the broader traits of complex systems solutions:
- Focus on specific problems in a particular local context as nominated and prioritized by local actors.
- Identify motivational problems.
- Foster active, ongoing experimental iterations with new ideas gathering lessons from these iterations to allow solutions to emerge.
- Establish an authorizing environment for decision-making that encourages experimentation and positive deviance.
- Engages broad sets of agents with highly varied skills sets to ensure that reforms are viable, legitimate, and relevant.
Notably, each of those points can be restated in a way that illuminates core principles of the market economy, the complex system at the heart of successful economic development. For example, “prioritization by local actors” recognizes the value of what Hayek called the knowledge of time and place. It is the knowledge that individuals possess about their own preferences and circumstances that cannot be centralized or obtained for productive use by others and it echoes the productive knowledge of Hausmann.
Recognizing the importance of motivational problems mirrors the central insight of economics that incentives matter, particularly given the principal-agent problem many aid workers must try to navigate. In its broadest context, this problem refers to the complications that arise when someone is meant to act on someone else’s behalf, which is why outsiders in the context of aid are at a particular disadvantage.
“Iterative change” implies an unpredictable but productive evolution of change that learns along the way. That learning and change may implicate disruption of old models to get to a new solution. This concept is not unlike Joseph Schumpeter’s description of entrepreneurial change via creative destruction, the unpredictable but likely consequence of a complex system unfettered by centralized attempts to impose linear-based outcomes.
A key motivation for emphasizing the importance of the authorizing environment for decision-making in “Building State Capability” is the insight that decentralized decision-making is a necessary approach to achieving innovative solutions. Similar to the productive efficiency of markets that are rooted in decentralized rights to property, decisions are best made by those who bear the costs and, at the same time, stand to reap the benefits of those decisions. In this way, all decisions implicate tradeoffs. Central authorities cannot grasp all relevant tradeoffs in a complex system such as a developing economy.
Lastly, recognizing that complex problems are best tackled by a diversity of people calls to mind the powerful influence of specialization in achieving exponential gains. Furthermore, that which determines whether something has achieved vibrancy, legitimacy, and relevance in a complex system is that broad set of third parties whose individual decisions and actions either support or fail to support the success of any initiative or product. This type of consumer sovereignty does not imply an all-knowing or perfectly rational set of third parties. It simply means success is dependent on the actions of many independent actors and the special contributions they make that, when aggregated, support a growing economy.
It’s encouraging to see mainstream experts conclude that the answer to economic development has something to do with those core economic ideas. It seems unlikely, however, that local governments, following a foreign how-to guide on discovering emergent solutions for their own bureaucracies, will be an effective solution to the problem of underdevelopment. In fact, based on their own measures of state capability, the authors put the odds of success dismally low for most developing countries, with only eight on track to achieve their definition of strong capability this century.
Despite their intellectual contributions, the framework these thought leaders operate within is fatally flawed. It takes for granted that the solution to economic development will emanate from the technical knowledge of foreigners and that the solution will be worked out between those foreigners and the governments attempting to run the countries where development is needed most. Short of that grander ambition, however, their analysis of public policy may leave a clue for how best to apply the insights of complexity science to unlock the productive knowledge Hausmann talked about.
Public Policy and State Capability
In looking at a state’s capability to advance various public policies, Andrews et al. distinguish among the simple and the complex using four categories that represent the difficulty a bureaucracy would have implementing and maintaining them. First, is the policy transaction intensive? Second, does it require a lot of discretionary decisions be made? Third, does it serve the public or impose an obligation on them? Fourth, is it dependent on known or unknown technology?
Re-creation of "Figure 5.1. Four key analytic questions about an activity to classify the capability needed." found on p. 108 of "Building State Capability," Oxford University Press, 2017.
They then take it as their goal to help national and local governments around the world become better at carrying out the more complex policies, namely, those that are transaction intensive, require bureaucratic agents to make wise, discretionary decisions, impose unpleasant or unwanted obligations on the public, and that depend on the bureaucracy innovating successfully.
Undoubtedly they see the daunting nature of their chosen task. They concede, therefore, that there ought to be a “genuine debate about the tasks a government can realistically perform,” especially when other civil society actors including private organizations can perform those functions just as well or better. This observation begins to look like a case for limited government, but they also suggest there should be a plan to reintegrate those functions back into government in the future as state capability improves.
This begs the question, though, in a world of struggling state capability why not begin instead by emphasizing proven public policies that rely less on government for success? That means advocating for those policies that are less transaction intensive, less dependent on omniscient government agents, less burdensome on citizens, and those policies that depend only on basic competencies that are clearly understood by those charged with carrying them out.
It seems those are the types of policies that would go the furthest in demonstrating appreciation for the insights of complexity science, namely that the most productive knowledge critical to discovering solutions is possessed by the individuals themselves who are the ostensible beneficiaries of aid programming. The conclusion then must be to expand the decision set for individuals as much as possible to provide maximum flexibility for discovering productive change and growth.
Public Policy and Economic Freedom
Nobel laureate Amartya Sen’s 1999 book, “Development as Freedom” details the non-severable relationship between those two ideas. He writes, “The perspective of freedom has been used both in the evaluative analysis for assessing change and in the descriptive and predictive analysis in seeing freedom as a causally effective factor in generating rapid change.”
It is precisely the freedom within complex systems that allows solutions to emerge. If governments and aid agencies are to have a positive influence on economic development, they must be judged within the context of freedom. Sen explains, “A variety of social institutions relating to the operation of markets, administrations, legislatures, political parties, nongovernmental organizations, the judiciary, the media, and the community, in general, contribute to the process of development precisely through their effects on enhancing and sustaining individual freedoms.”
You can find an abundance of those types of policies to choose from if you look in places such as Fraser Institute’s Economic Freedom of the World Index or the World Bank’s Doing Business report, both of which assess countries based on a review of public policies that represent the expansion of individual economic choice and a reduction in government scope and functions.
There you will find policies that make it easier and cheaper to start a business, easier and cheaper to trade across borders, easier and cheaper to register property and get construction permits, and policies that implicate an overall smaller government size with fewer regulations to competently enforce.
Admittedly, many good policies, even some of those included in those global indices, require strong institutions supported by competent and capable governments. How much more likely though are governments to do those things well if they, as Andrews et al. suggest we debate, are limited in their tasks to only that which is truly necessary?
Yet, in the same way that outsiders blunder when they try to help other people achieve economic success, they also blunder when they try to get other governments to strengthen their institutions. In an article published by the American Journal of Economics and Sociology, the economists Peter Boettke, Christopher Coyne, and Peter Leeson use the term “sticky” to describe how successfully institutions persist within a country. In their analysis of what makes stickiness more likely, they conclude that chances are best when local actors lead the process of improvement.
For this reason, local civil society organizations focused, through research and advocacy, on expanding economic choices for all stand the best chance of leading effective change. It is their civic entrepreneurship that determines what reforms are needed, which of those reforms are possible, and what form they should take to achieve relevance and sustainability in the local context.
Unleashing Productive Knowledge in a Complex System
Improvements in Doing Business scores translate to reductions in poverty. Using new findings published by the World Bank, Atlas Network estimates that policy changes achieved by a sample of 10 local civil society organizations supported with project grants translate to the equivalent of 405,000 people lifting themselves out of poverty.
For example, Centre for Civil Society in India succeeded in convincing the Modi government to eliminate minimum capital requirements for new businesses, a policy that disproportionately burdens the poor and one that scores high on Andrews et al.’s complexity spectrum for bureaucratic capacity. Its elimination then is a win-win for economic development.
Changes like this align with the insight from complexity science that the knowledge needed for economic development is possessed by those in poverty since they are in the best position to assess whether they have enough capital to take on the risk of starting a business. It certainly makes more sense than concluding the best way forward is to make aid agencies more adaptive, or bureaucracies more capable, or national governments more calculating in picking winners and losers in the product marketplace. Those do nothing to expand economic choices for the true drivers of poverty alleviation: the poor themselves.
As outsiders, we can play our part by supporting independent research and advocacy civil society organizations led by local intellectual leaders who recognize that the future economic success of their country is a function of the expanded economic opportunities all people enjoy as their governments try to do less on their behalf. This solution shares in the spirit of the OECD principles in seeking solutions at the local level and takes them to their natural conclusion.
The teams associated with Andrews, Ramalingam, and Hausmann have made major contributions to our understanding of complexity and yet most of their hopes for using that increased knowledge are pinned on the centralized decision-making capabilities of aid agencies and national governments. Their solutions do take seriously the insight that within complex systems individuals freed to engage in dynamic processes will discover solutions that are more promising than their centralized, top-down counterparts. Yet they underappreciate the big opportunity this insight truly presents us. It’s an opportunity to transform the way we think about helping others around the world by first recognizing that, because productive knowledge is widely distributed in a developing economy, our first priority is ensuring that economy receives less interference from us. As outsiders, we will always lack the particular knowledge needed to solve their economic problem.
Hausmann says it well when he writes, “The secret to modernity is that we collectively use large volumes of knowledge, while each one of us holds only a few bits of it.” The key then is to stop trying to centralize that knowledge via outsiders or governments and to simply unleash it. The more we resist the temptation to solve economic problems for others the more we will improve on that score. Ramalingam offers us this final piece of encouragement, “Even if a narrow, simplistic, mechanistic, reductionist form of global altruism is our legacy, it needn’t be our fate.”