PA 30: Social Dilemmas, Commitment Problems, and the Origin of State Institutions

The Ransomware Dilemma and Commitment Problems

In the context of ransomware attacks, a fundamental dilemma arises regarding the exchange of cryptocurrency for decrypted files. A victim faces a significant commitment problem: even after paying the ransom, there is no inherent assurance that the hacker will actually provide the decryption key to restore access to the data. This skepticism on the part of the victim could potentially undermine the market for ransomware exchanges. If victims cannot be assured of recovery, they may choose never to pay, resulting in a loss of potential profit for the attacker. To solve this commitment problem, hackers have developed specific strategies to appear as "honest hackers." Many hacking groups strive to build a reputation for reliability, as this reputation increases the likelihood that future victims will pay. Additionally, hackers may provide proof of their ability to decrypt specific files as a demonstration of technical capability. More astonishingly, some groups run 24/7 help desks or customer support services to guide victims through the recovery process, ensuring that the hostile takeover of a computer is followed by an atypically high level of service. This investment in customer support and reputation-building is essential because, in the absence of such trust, the lack of confidence would cause the entire exchange market to fail.

Course Logistics: Midterm Exam and Writing Assignment

The first midterm exam is scheduled for next Wednesday, exactly one week from today. Students registered with the Center for Accessible Education (CAE) will take the exam at the testing center, while all other students must be present in the classroom to take the exam online. Despite being an online exam, attendance in the physical room is mandatory. The exam will consist of closed-ended questions, and students are required to bring a fully charged device. In the event of an unexpected IT failure, printed paper copies of the exam will be provided as a backup. The exam will be administered via a lockdown browser through CCLE (CCLE mentioned as "Pro and Learn" in context), which allows for nearly instantaneous grading and detailed diagnostics. If a question is missed by a large majority of students, it may be discarded before final grading. Students may use earplugs to block sound during the exam, provided they are not electronic devices like AirPods. Following the midterm, an in-class writing assignment will be held on the subsequent Monday. This assignment requires students to diagnose a social dilemma presented in a short, newspaper-style scenario. Students must identify the problem (e.g., positive or negative externalities), the resulting level of behavior (e.g., undersupplied according to the rule of undersupply when private benefits are lower than social benefits), and an appropriate policy response (e.g., a subsidy). The response is limited to 500 words and will be graded based on logical organization and the quality of the rationale provided rather than length.

Game Theory: Extensive Form and Sequential Moves

To analyze commitment problems and social dilemmas, we utilize complete information games where players can anticipate the incentives of others. Unlike simultaneous move games, these are modeled as dynamic games where players move in sequence. These games are represented using the extensive form, which is a decision tree structure starting at the top or left of a page. A node represents a point of choice, and the possible actions emanate from that node as branches (arrows). To solve these games, we employ backwards induction, a process of reasoning backward from the end of the game tree. This method is intuitive, much like a player in chess or checkers anticipating a respondent's reaction before making a move. In the case of a concert ticket exchange on the internet, the first player chooses whether to send money. If they send it, the second player (the seller) chooses whether to honor the agreement or cheat. In a hypothetical model where the seller gets a payoff of $2$ for cheating and $1$ for honoring, they will choose to cheat because 2 > 1. Recognizing this, the buyer looks down the tree and sees that sending money results in a payoff of $-1$ while keeping the money results in $0$. Because 0 > -1, the buyer refrains from buying, leading to a suboptimal equilibrium of $(0, 0)$ where a mutually beneficial transaction never occurs due to the lack of third-party enforcement.

The Dominant and Subordinate Pig Experiment

The commitment problem is further illustrated through the experiment involving a dominant pig and a subordinate pig sharing a pen. On one side of the large pen is a lever that must be pressed to release food into a trough on the opposite side. If the dominant pig presses the lever, it must run to the other side, but the subordinate pig is already there and can consume most of the food before the dominant pig arrives. If the subordinate pig presses the lever, the dominant pig simply waits at the trough and uses its physical strength to push the subordinate pig away, consuming all the food. The payoffs can be modeled as follows: if the subordinate pig presses the lever, it might end up with a payoff of $2$ while the dominant pig gets $3$. However, the subordinate pig anticipates that if it contributes, it will be pushed away and possibly left with less or nothing. In various scenarios, if the dominant pig cannot credibly commit to shared consumption, the subordinate pig simply refuses to press the lever. This leads to an equilibrium where the subordinate pig may be "bored" or hungry, but the dominant pig is also worse off. This demonstrates that dominance is not always beneficial, as the dominant player’s power to exploit the weaker player prevents the weaker player from cooperating in the first place. The dominant player would actually benefit from tying their own hands to make a promise of non-exploitation credible.

Thomas Hobbes and the State of Nature

Thomas Hobbes, in his work Leviathan, describes the "state of nature" as a baseline of human existence characterized by the absence of a central authority. In this state, there is no industry, culture, or navigation because the "fruit of investment" is entirely uncertain. If a farmer cannot be sure they will enjoy their harvest, they have no incentive to plant. Consequently, society in the state of nature lacks arts, letters, and security. Hobbes famously describes the life of man in this condition as "solitary, poor, nasty, brutish, and short." This is a social dilemma where even the strong are not safe, as they live in perpetual fear of violent death or misappropriation. Without institutions to constrain the powerful, there is no accumulation of wealth because there is nothing to steal from those who refuse to produce. Thus, the absence of a coercive third party leads to a massive social loss.

The Farm vs. Prey Model

The state of nature can be modeled through the choice of two actors, Emile and Bala, who must decide whether to "farm" or "prey." If both farm, they create a flourishing economy with a payoff of $10$ each. However, using backwards induction, we see the dilemma: if Emile farms, Bala chooses between farming ($10$) and preying ($20$). Bala will choose to prey (20 > 10). If Emile preys, Bala chooses between farming ($-5$) and preying ($1$). Bala will choose to prey (1 > -5). Emile, anticipating Bala’s responses, sees that if he farms, he gets $-5$, but if he preys, he gets $1$. Consequently, both choose to prey, leading to a payoff of $(1, 1)$. This is a social dilemma where both are significantly worse off than the $(10, 10)$ outcome they could have achieved if they could have committed to not stealing. This predatory behavior prevents any long-term investment in human capital or technology.

Mafias and the Provision of Protection

In environments where trust is fragile and state governments are absent, non-state institutions like the mafia can arise to solve commitment problems. As scholar Diego Gambetta notes, the mafia provides a valuable service by acting as an enforcer in transactions. In an illicit market, a seller might pay a mafioso (a third-party enforcer) to protect the buyer against the seller's own potential to cheat. By creating a credible threat of punishment for fraud, the mafia allows for trade to occur in markets that would otherwise collapse. The seller essentially "purchases protection against himself" to make his promises credible. For example, in drug markets, gangs try to maintain order because violence attracts police and deters customers. The goal is to keep the business profitable by reducing the risks associated with the transaction, effectively serving as a private protection racket that creates social value within its specific context.

The State as a Stationary Bandit

The state can be viewed as an evolution of the protection racket. Max Weber defined the state as an entity that holds a "monopoly on the legitimate use of physical force (coercion)" within a territory. Mancur Olson expanded on this by distinguishing between "roving bandits" and "stationary bandits." A roving bandit plunders a village and moves on, which is a self-defeating strategy because victims eventually stop producing anything worth stealing. In contrast, a stationary bandit settles in a territory, claims it as their domain, and provides security and public goods like roads and sanitation. They do this not out of benevolence, but because a secure, productive population provides more tax revenue. By punishing predation among its citizens, the state enables them to escape the state of nature and invest in farming or industry. This security generates a "peaceful order" that allows for the accumulation of wealth, which the state then taxes to fund its own operations. An example of this is the Islamic State (ISIS) in Iraq; despite their brutality, bureaucratic documents show they functioned as a stationary bandit by providing services like trash collection and taxing citizens to fund their governance, rather than relying solely on plundering resources.