Learning how to cooperate

Mycorrhizal fungi expand the plant root zone network. Photo credit NC State

Growing socioeconomic inequality and intensifying concerns about environmental sustainability both share in common a concern of an erosion of the social contract and popular mandate for the status quo economic system. Our interconnected globalized economy of a population of 7 billion people is one large complex collective action problem. Globally five transformations are identified which could shift the economy to become more equitable and environmentally sustainable

  1. Wealthy households should expect to invest their surplus resources in human capital of those on the margins of society to address inequality
  2. Families should number of children per household to be near replacement so that the scarce common resources can be shared most efficiently
  3. Environmental footprint consumption of individuals and firms should be contained within planetary boundaries limits
  4. Limited land, water and biodiversity resources must be conserved
  5. Energy, transportation, industrial systems must switch to become less dependent on releasing greenhouse gasses into the atmosphere from combustion of coal, oil and natural gas.

Underperformance of that system of providing equitable, sustainable prosperity is not only a concern that the outcomes fall short of expectations, but that those who are concerned that study the systems do not understand how to close these gaps. While different academic scholars from sociology, economics, non-governmental institutions and policy makers may claim to have found their own unique working solution, it is still an open collective knowledge gap if there remains sufficient uncertainty that it is a barrier for large scale implementation. It is an open control problem in the field of social sciences.

A control problem in this circumstance is considered ‘solved’ when a model of the system can be constructed with high prediction accuracy and contains some parameters which can be directly manipulated that can shift the outcomes. If either the model accuracy remains high or it does not contain any directly manipulated parameters, the control problem is still ‘unsolved’. The challenge of social systems is that they are non-deterministic, complex and thus predictions will always be limited to a cone of uncertainty which increases with the time horizon. What may work and be well predicted on small scales may lose predictive certainty when scaled to larger systems.

While there are precision limits for how much can be predicted in any complex system and time horizon, different fields have made advances in closing the uncertainty ranges. For example the cone of uncertainty in the 3-day forecasts for hurricanes has reduced from 300 nm to <80 nm from 1990 to 2018 (Masters, 2020). In contrast, the uncertainty of macroeconomic forecasts such as interest rates by the Bank of England are large — up to 2x the standard deviation for 2 year horizon and the accuracy has decreased in recent years (Bank of England, 2015). Modelling performance in political science has been similarly unhelpful. For example, fundamentals-based models do not perform much better than a coin toss for predicting an upcoming presidential election outcome (Silver, 2012).

Within this learning problem for large scale collective action are specific sub-problems that have been identified in the different fields. Sub-disciplines in social sciences of sociology, political science and economics often overlap with one another. Three well known problems in economics which can trace at least to the time of Adam Smith in the 18th century and still remain an active area of debate are

  1. Prisoner’s dilemma
  2. Relative performance of the state and market
  3. Human behavioral model.

Prisoner’s dilemma : conditions for cooperation

The field of game theory deals with multiple agent decisions where one player’s decision is contingent on how other agents will behave. The Prisoner’s dilemma is a general class of 2x2 games in which players are faced with a situation where they receive the best payout when they both cooperate, but receive a loss if the other player defects and a gain if they defect, or “free ride” (Ostrom, 1990). For a single round game, if there is no enforcement authority, players have no ability to communicate with one another and under certain conditions of the relative payoffs, each player has a dominant self-interested rational strategy to defect. The best solution overall for everyone — “Pareto optimal” is not the best solution for each acting in their own self-interest. For each individual they have the most to gain if they free ride and defect while others cooperate.

In such games the belief about how others influence each agent’s decision whether to cooperate or defect (Ostrom, 1990). One metaphor in particular uses a common grazing area described by Garret Hardin in 1968, the “Tragedy of the commons” (Hardin, 1968). If anyone can graze on the field then according to Hardin, the collective resources are doomed to over-grazing as each individual acting in their self-interest would take more than a sustainable share for themselves and leave a smaller amount for the collective. Over time eventually everyone is worse off and the resources are exhausted (Hardin, 1968).

Therein is the tragedy. Each man is locked into a system that compels him to increase his herd without limit — in a world that is limited. Ruin is the destination towards which all men rush, each pursuing his own best interest in a society that believes in the freedom of the commons.

- Garret Hardin, 1968 Tragedy of the Commons

Real world situations of common pool resources have evidence of both collective action failure, and successful cooperation (Ostrom, 1990). Where multiple rounds of interaction open the potential for reciprocity and tit-for-tat strategies which improve the potential for cooperation (Axelrod, 1984). Other researchers have found both theoretical and empirical evidence that communication, group size can influence the level of cooperation (Miller, 1998 ; Balliet, 2010 ; Ostrom, 1990). While the basic setup of the problem is simple to describe, its analysis and theoretical predictions of cooperation with only minor modifications to the setup are considered too complex to be able to generally predict cooperation consistently with certainty.

Enclosure, private and public goods

Where self-interest of all of the players is taken as a given presumably the only remedy for such a situation is through some form of enclosure and rights allocation either to one central authority or to multiple property owners. The model of private property rights can be implemented practically in situations where the product is excludable from others. Generally though enclosure is not always practical or obvious. The situation of the grazing commons can be generalized to other types of goods and services based on how easily that the product can be excluded for someone to use the product (excludable) and whether their use depletes from someone else’s opportunity to consume the good (rival). In total there are four possible combinations — private (rival, excludable), clubs (non-rival, excludable), commons (rival, non-excludable) and public (non-rival, non-excludable). The real world examples of Ostrom ultimately relied on different creative forms of enclosure with a hybrid ownership between the collective governance and the private members (Ostrom, 1990). Non-excludable, public goods however were not included in the scope of Ostrom’s study (Ostrom, 1990). An economy is made of a mix of these four different types of goods and services. While food which comes from land and real estate may fit well as private goods, others are either not completely rival or do not have a practical or cost effective means of enclosure.

Some goods such as education have properties of both private and public goods. Some of the benefits of education have positive leak beyond the direct parties involved to the rest of society, and similarly for healthcare. The demand for these public goods (or prevention of bads) is one of the arguments for the model of collection of taxes by a central state authority.

The role of the state and the markets

The prisoner’s dilemma tragedy for common resources has been scaled up metaphorically to the design of policy and institutions at the national level (Miller, 2020). Two general types of enclosure strategies have emerged in another open question, whether to appoint an authority (the state model) or to divide up the commons into private rights (the market model). Within this discussion is also three sub-topics — urbanization, corruption and technology advancement.

For a comparison of country-year dataset panels, urbanization as a phenomenon is positively associated with a number of transformations of daily life resulting in build up of physical urban infrastructure, improvements in health, education, increasing density and greater prevalence of market activity and institutions. The strong association between each of these trends of “development” is used as evidence that these market institutions are instrumental for social advancement and quality of life improvement (Miller, 2020). The greater density, opportunity for social exchange, of the urban environment may alter the mix of types of services by expanding the opportunity for public spaces for either public goods or public bads (West, 2017). Another trend associated with urbanization is Wagner’s law and Baumol’s cost disease are two theories that describe respectively, the general tendency for the demand and costs of public service to increase as the average income of an economy increases (Nordhaus, 2006 ; Nyasha, 2019 ; IMF, 2014).

The principal-agent problem is the dilemma of how to ensure that some agent executes his responsibilities in the interest of the principal, and not his own interests. While the case for more public services may increase with urbanization, how those goods are provided and who is the manager is a focal question. There are concerns of potential for corruption for both the firm and the state. For-profit firms may feel compelled to compromise ethics either for their own performance and pay, or in the interest of making a profit for the firm. A concern for democratically elected officials from the state is the potential for perverse rent-seeking pattern where a minority group uses their vote for personal gain rather than the interest of the group as a whole (Shughart, 2020).

The association of a healthy economy with an abundance of job opportunities is strong in the public conscience. While Keynesian economic policy instruments enable the state to correct for temporary fluctuations in employment demand, what is less clear is whether society would be better if this marginal employment was provided by a private firm. So the argument goes, if the state is less competent as an investor at generating a return, then state programs which achieve full employment but are invested poorly. Over time a pattern of under-performing investments could eventually lead to at a minimum an undermining of the state’s autonomy in a competitive environment, or worse a lower quality of life if assets are not replenished and maintained.

For a national economy, a return on investment can either come from accumulation of factors — energy, physical material stock, labor force, or productivity improvement through technology advancement. Two related questions are whether greater public spending tends to stimulate greater private sector investments — crowding-in, or whether it deters them — crowding-out, and whether the state or the private firms are more qualified at producing technology advancements. The question as to whether the state has either a net crowding-in or crowding-out effect on private investments appears inconclusive with a mix of evidence for both (Alfonso, 2008 ; Deleidi, Mazzucato Semieniuk 2019).

While not all new technologies are considered beneficial for society, a number of social improvements can be attributed to technology advancement in fields such as medicine, transportation, communications. Another concern of the manager’s competency is whether a democratically elected state is capable of replicating the creativity and risk appetite from private, profit seeking entrepreneurs in advancing technology. The argument of Economist Mariana Mazzucato is that the mission-oriented US defense research agency DARPA and US federal programs for pharmaceuticals research were instrumental for providing the basic research which led to later profitable industries in the private sector (Mazzucato, 2011). A potential advantage of a large, secure state over private firms is both the potential for a longer term investment horizon and access to larger capital. Both of these factors correspond to greater capacity for high risk ventures such as creating new markets, something that may be out of reach for individual firms. There are a range of outcomes. The US is one data point and there are counter-examples of states which have not produced similar results with the same level of R&D funding (Mazzucato, 2011).

Human behavioral model

The environmental conditions in which Homo Sapiens evolved their social behaviors may be different in important ways from the modern urban environment. Humans are a specimen of biology. They are complex biological specimens flawed, biased, behavior guided both by irrational emotions, heuristics and rational logic. Decision making is bounded by limited attention and information processing capacity. Another question related to cooperation in prisoner’s dilemma is how to model the behavior of the agents who are participating in the collective action scenarios.

How is complex problem solving efficient decision making organized from a limited, imperfect individual human species?

Mainstream neoclassical economic models used for analysis in game theory and to describe macroeconomic activity simplify the way that humans and firms interact with one another as conforming to an ideal form — Homo Economicus who is (1) self-interested (2) profit-maximizing and (3) perfectly informed. This model simplifies and excludes important features of Homo Sapiens. A more accurate, complete description of humans from the perspective of biology, anthropology is a social being which values the fate of their community, family and safeguard their own interests. They are also capable of cheating, corruption and this is no less true for private as it is for public institutions. Humans exhibit cooperative and competitive (selfish) dynamics, information is anything but perfect as bandwidth limits, heuristics, asymmetry are the rule rather than the norm (Schwartz, 2012 ; Sapolsky, 2017).

One example of the departure from pure self-interest, rationality is the social phenomenon of “altruistic punishment”. Lab experiments studying the effect reveal results that challenge the assumption of self-interest (Boyd, 2003). Evidence of pure altruism has been observed in economic game theory when a participant makes a move of either cooperation or punishment at a cost to themself even in one-off interactions with unaffiliated strangers (Boyd, 2003). In one 2004 experiment by De Quervian, participants are informed that one of the other players has committed an offense. They are offered the choice of either (A) doing nothing (B) enforcing a small fine at no cost to themself or © paying to have a larger fine levied on the other player. A non-negligible level of the participants opted for options (B) and (C ). Furthermore measurement of their dopamine levels during the action indicated that they enjoyed paying to punish the other player (Sapolsky, 2017). While the prevalence of such actions may vary, these observations offer evidence of potential larger scale cooperation networks beyond small groups (Boyd, 2003).

The field of Behavioral Science revises this model with the concept of “bounded rationality” to describe a more realistic description of decision making process which can be corrupted by conformity, cognitive biases and loss aversion (Kahneman, Tversky, 1979 ; Thaler, Sunstein, 2009). While it is acknowledged by Economists that use neoclassical assumptions they are a poor description of real human behavior, what is less conclusive is how these imperfections from the ideal model interact with institutions and scale up in complex adaptive systems for either positive or negative effect on general welfare.

Critical role of trust, cooperation

While there is plenty of knowledge of the solutions for what it takes to provide good quality healthcare, education and environmentally sustainable production methods, there is comparatively less understanding of how the human social systems succeed or fail in adopting these good habits and technologies. One part of that answer may relate to these three questions —

  1. Under what conditions do humans voluntarily cooperate to solve collective action problems?
  2. How do those lessons scale to the national economy?
  3. What form of agent model accurately represents the full range of human behaviors while still being flexible enough to scale those effects into larger systems?

Some insights can be inferred by examining system behavior in past case studies and comparing those that worked well from those that didn’t. As a complex adaptive social-ecological system, growing environmental unsustainability, socioeconomic inequality and slowdown of productivity growth can be thought of as symptoms of deeper flaws — loss of trust and cooperation in large scale collective action problems. These symptoms may also be associated with rising prevalence of short-term decision making, risk aversion, under-investment of human capital, higher share of competitive vs cooperative exchange, erosion of social cohesion from socioeconomic inequality and non-transparent (private, proprietary, classified) decision making.

Network diagram of feature relationships in an economy

The importance of trust and cooperation for maintaining a healthy, sustainable economy is recognized from a wide range of political and economic schools of thought (Farrell, 1993). The time horizons of investment payback for natural resources such as fisheries, forests can reach beyond a single lifetime, and this is even more dramatic for larger planetary dynamics such as the dynamics of greenhouse gases in the earth’s atmosphere in stabilizing the planet’s energy balance. Not only natural resources but also human capital investments have long payoff time horizons and only modest returns but are critical for technological advancement (World Bank Group, 2019). Trust and cooperation are critical prerequisites for the character of decision making to be willing to take risks and patience to commit to long term strategies.

The conditions for long term decision making for managing common pool natural resources are trust in the institutions and the community members, security that short run needs are taken care of and some personal stake in the long term outcomes such as the belief that one’s children will be exposed to the consequences (Ostrom, 1990). Trust is usually not established a-priori, but rather is built over time through a history of successful experiences of past cooperation (Farrel, 1993 ; Ostrom, 1990). Connectivity, transparency are instrumental in creating the conditions for building a collective history of shared meaning and past experiences. The opportunity for communication, in particular face-to-face interactions and ability for mutual monitoring and transparent, common shared libraries of information of past interactions are institutional factors which help to facilitate building confidence (Ostrom, 1990).

A problem that is well defined is half-solved.
- Charles Kettering, General Motors (1876–1958)

While the problem of how to move towards a more equitable, sustainable economy remains unsolved, it helps to have a concise description of the problem.


Alfonso, 2008 ECB working paper No 864 Macroeconomic rates of return of public and private investments. crowding-in and crowding-out effects

Axelrod, 1984 the evolution of cooperation

Balliet, 2010 Communication and cooperation in social dilemmas : a meta-analysis review

Bank of England, 2015 Evaluating forecast performance

Boyd, 2003 the evolution of altruistic punishment

Deleidi, Mazzucato Semieniuk 2019 SOAS Dept of Economics Working paper 226 : Neither crowding in nor out: Public direct investment mobilising private investment into renewable electricity projects

Farell, George Washington University, 1993, Trust and political economy : institutions and the sources of interfirm cooperation

Hardin, Garret, 1968 Tragedy of the Commons

International Monetary Fund (IMF), 2014 Public Expenditure Reform Making Difficult Choices

Kahneman, Tversky, 1979 Prospect theory : an analysis of decisions under risk

Masters, Yale Climate Connections 2020 The most reliable hurricane models, according to their 2019 performance

Mazzucato, 2011 The entrepreneurial state

Miller, 1998 Communication and cooperation


Nyasha, 2019 Government Size and Economic Growth: A Review of International Literature

Ostrom, Elinor, 1990 Governing the commons

Sapolsky, 2017 Behave : The biology of humans at our best and our worst

Schwartz, 2012 Refining the theory of basic individual values

Shughart, The library of economics and liberty last accessed 2020 Public Choice

Silver, Nate, 2012 Models based on fundamentals have failed at predicting presidential elections

Thaler, Sunstein 2009 Nudge : Improving decisions about health and happiness

West, 2017 Scale: The Universal Laws of Growth, Innovation, Sustainability, and the Pace of Life in Organisms, Cities, Economies, and Companies

Applied research, engineering, and projects for solutions to sustainable cities. SG Green New Deal https://aseangreennewdeal.wixsite.com/home