Why Public Policy Fails
by Bernardo Mueller, University of Brasilia
The views and opinions expressed in this article are those of the author and do not necessarily reflect the official policy or position of the Florence School of Regulation.
The public policy fails. It fails often and fails everywhere. From small tasks, such as keeping the streets free of potholes, to massive endeavors such as providing healthcare, it seems that public policy continually comes up short of expectations. In Why Public Policies Fail: Policymaking under Complexity I discuss the reasons why policies so systematically underperform. The main reasons are probably not those that you imagine.
Before getting into the determinants of underperformance the article provides three examples to make the point that it is not just the occasional ill-conceived, poorly-planned and shoddily implemented policy that disappoint, but rather that the problem goes beyond inexperience, incompetence and malfeasance. These are examples of policy approaches that have had wide acceptance by experts, academics and society in general, and have been accepted as innovative, smart and rigorous ways to solve problems and get things done. But that, nevertheless, have, more often than not, proved disappointing.
The first is the idea of microcredit to reduce poverty. It involves small loans to extremely poor people organized in tight-knit groups, who would not otherwise have access to credit. It is conceived to stimulate entrepreneurship by giving the poor a boost towards building human, social and financial capital. The idea was praised by academics, international organizations, governments, NGOs and others as an innovative way to get at one of mankind’s greatest problems. The United Nations declared 2005 as the International Year of Microcredit. It was widely adopted across the world. Yet, when the first policy evaluations came in, the impacts on individuals and communities proved disappointing. Banerjee et al (2015), for example, found very little evidence of transformative effects. Relative to expectations, microcredit has thus far failed.
The second example is the approach to managing fish stocks, as well as other resources, through individual transferable quotas (ITQs). Fish stocks across the world have been collapsing in the past decades and traditional policies have failed to avert the impending disaster. ITQs create a market for property rights over fish that would not otherwise exist due to the problem of open access. The right to fish can be traded, which leads to a better allocation of effort and resources, as the least efficient sell their rights to those with lower costs and greater skill. These schemes were enthusiastically embraced by economists as an idea well-rooted in economic theory that leverages the power of markets to provide good incentives and internalize externalities. But here, once again, there was disappointment. ITQ’s have been widely implemented across the world but have had little impact. Evaluations by Costello et al (2008) in Science and by Chu (2009) in Fish and Fisheries, while finding some modest impact in terms of reducing stock loss, still paint a disappointing picture relative to the expectations that the approach once elicited.
The third example refers to the surprisingly large share of global economic activity that is concentrated in a small number of mega-projects. These are “large-scale, complex ventures that typically cost US$1 billion or more, take many years to develop and build, involve multiple public and private stakeholders, are transformational, and impact millions of people” (Flyvbjerg, 2014 , p. 6). Examples of mega-projects are the International Space Station (estimated cost US $ 150 billion), Dubailand (an amusement park in Dubai, US $ 64 billion), and the Hong-Kong-Macau bridge (US $ 10.6 billion). Spending on mega-projects makes up 8% of world GDP and is growing. Flyvbjerg (2014: 11) analyzed all mega-project to date in order to ascertain how well such a large part of the world’s economic resources are being used:
Success in megaproject management is typically defined as projects being delivered on budget, time, and benefits. If, as the evidence indicates, approximately one out of ten megaprojects is on budget, one out of ten is on schedule, and one out of ten is on benefits, then approximately one in a thousand projects is a success, defined as on target for all three. Even if the numbers were wrong by a factor two — so that two, instead of one, out of ten projects were on target for cost, schedule, and benefits, respectively — the success rate would still be dismal, now eight in a thousand.
These are just three examples, but the claim that public policies fail is more general. A glance at today’s newspaper should make the point. The question is: why do public policies fail so ubiquitously?
Why do public policies fail?
Part of the explanation for why public polices so often, leading to inefficiency, disappointment and waste, is obvious. All phases of making public policy, conception, projection, implementation and operation, are permeated by incompetence, ignorance, corruption, excessive optimism, self-interest and transaction costs. They are hard to do and full of perverse incentives. However, the central argument here is that even if we manage to resolve or drastically mitigate all of these problems, it still remains the case that projects and policies will largely fail. More data and information, better governance, better experts, more transparency, nobler public spirit and political will, checks and balances, etc. can usually improve things. But these eventually run into decreasing returns and there always remains an irreducible level of failure that is impervious to even the best attempts at getting things right.
This irreducibility comes from the complex system nature of most public policies and the environments in which they operate. Complex systems are characterized by the interaction of many, diverse, agents, leading to the emergence of system-wide phenomena that are not designed or intended by any of the parts. The system is different from the sum of the parts. In these situations, the traditional approach by economists and public policy-makers to improve public policy is not appropriate. The traditional approach is linear, reductionist, Gaussian, static, ergodic (it assumes that the future will be the same as the past) and largely ignores the interactions that are the source of the problem. Decision theory, for example, which is one of the cornerstones of these approaches, requires comparing the costs and benefits of all situations that may occur, taking into account the probabilities of each state in the world. But in a complex system, not only is it not known what will happen, but it is not known what can possibly happen.
How to deal with public policies in complex domains?
Given the uselessness of much of our knowledge to improve the performance of public policies, what can be done? Could new information technologies like Artificial Intelligence, Big Data, Machine Learning, blockchain, neural networks, in addition to who knows what will emerge in the future, be the solution? Many of these techniques and processes have characteristics that make them good tools for dealing with complex systems. In “Why Public Policies Fail” I describe some modern approaches that seem promising to deal with this problem because they do not need to forecast or control the system. However, my conclusion there is that even if these techniques and instruments are able to improve our ability to create, implement and manage public policies, we never have the level of control and predictive ability that the traditional approach supposes to be possible. Even with more appropriate approaches to complex systems, there are limits to what can be achieved. In the end, it will be necessary to adopt a posture of greater epistemological modesty and recognize our limitations, admitting that the level of control and agency that we usually aim for cannot be achieved.
This does not mean that there is nothing that can be done. Complex systems can be influenced and swayed to avoid some types of results and induce others, even if it is not possible to impose finer tuning. The important thing is to recognize the nature of a complex system when dealing with one, and to use the interventions proper for a system with such characteristics, which sometimes can be simply doing nothing.
Leap in the dark or epistemological humility?
In the 1960s Albert Hirschman noticed a pattern in several projects that he visited in different countries on his travels as a development economist. Many countries embarked on large and ambitious projects, such as dams or new industries, with a naive optimism that did not see the complexity and difficulty inherent in the undertaking. He coined the expression “Hiding Hand Principal” to refer, in an allusion to Adam Smith, to this tendency for policymakers and project managers to underestimate the level of uncertainties and complications inherent in the project they had taken on (Hirschman, 1967). According to him, if this cold reality were not masked by the simplistic optimism of the actors, few projects would be attempted. And although the projects often did not achieve their goals, they did many times put the country to an unanticipated situation that allowed flashes of creativity to adapt the project for other purposes that proved viable along the journey and sometimes even turned out to be better than the original idea.
Hirschman’s attitude matches in part with the argument presented here. Although he did not use complex systems theory in his analysis, the idea that there is fundamental uncertainty behind public projects and policies is similar. However, the policy recommendation is quite different. While Hirschman reacts to the impossibility of predicting the future and controlling a complex system by suggesting a leap into the dark with the hope that in the end everything will work out, I am suggesting to be more realistic with what can be achieved and adapt the scope and nature of the intervention to this reality, even if it means that we cannot do everything we would like to do.
