Felix Reed-Tsochas

Human cooperation is dynamic and often declines even under favourable conditions ^1–4 . Many prevailing theories explain the decrease of cooperation in terms of strategic behaviour or learning, framed as evidence of rational behaviour or progression towards rationality ^5–9 . Here we show that a key source of long-term decline derives from deviations from rational behaviour that systematically vary over time. We analyse a natural social dilemma in the field—that is, group lending in Sierra Leone—tracking cooperative dynamics over a five-year period. Borrowers enter a joint-liability contract, structured so that if the group loan is not repaid in full, all members lose access to future credit ¹⁰ . This produces a threshold social dilemma with incentives to free-ride ^11,12 . The dataset includes 47,931 group payments made by 7,108 borrowers, augmented with a two-stage cluster sample of semi-structured interviews. We find a statistically robust pattern of punctuated decline driven by behavioural mechanisms ¹³ . Cooperation rates start out high but gradually decline due to decreases in group members’ cooperative motivation and effort. Sharp rebounds occur when loans are restarted and clients resensitized to their cooperative responsibilities, even though the group membership and dilemma structure are largely unchanged. This pattern persists over the five-year observation window, but with each successive restart the subsequent decline is more rapid. The findings have direct implications for preventing behavioural decline in cooperative programmes and institutions.

It is known that greater social cohesion increases a group’s ability to enforce cooperation. Despite this, defectors often go unpunished and groups with social structures that are a priori favorable often fail. A critical distinction is required between the structural effect on ability versus willingness to punish. We develop a theoretical framework in which variation in a group’s social structure generates a tension between ability and willingness to enforce cooperation. Structures that promote ability to punish also often reduce interest in carrying out sanctions, thus changing collective outcomes. Our empirical analysis involves a well-defined cooperative dilemma: group lending in Sierra Leone. We complement statistical modelling, based on a dataset containing 5,487 group repayments, with ethnographic analysis. We find: (1) Structural cohesion only increases economic cooperation between borrowers to a point, beyond which unwillingness outweighs increased ability to punish, reducing group repayments. (2) Groups with disconnected subgroups perform worse on average. Although borrowers are more willing to punish defectors in the out-subgroup, they are unable to do so effectively.

Introduction: Low-income and middle-income countries (LMICs) are crucial in the global response to antimicrobial resistance (AMR), but diverse health systems, healthcare practices and cultural conceptions of medicine can complicate global education and awareness-raising campaigns. Social research can help understand LMIC contexts but remains under-represented in AMR research. Objective: To (1) Describe antibiotic-related knowledge, attitudes and practices of the general population in two LMICs. (2) Assess the role of antibiotic-related knowledge and attitudes on antibiotic access from different types of healthcare providers. Design: Observational study: cross-sectional rural health behaviour survey, representative of the population level. Setting: General rural population in Chiang Rai (Thailand) and Salavan (Lao PDR), surveyed between November 2017 and May 2018. Participants: 2141 adult members (≥18 years) of the general rural population, representing 712 000 villagers. Outcome measures: Antibiotic-related knowledge, attitudes and practices across sites and healthcare access channels. Findings: Villagers were aware of antibiotics (Chiang Rai: 95.7%; Salavan: 86.4%; p‘ampi’ in Salavan (75.6%; 95% CI 71.4% to 79.4%). Multivariate linear regression suggested that attitudes against over-the-counter antibiotics were linked to 0.12 additional antibiotic use episodes from public healthcare providers in Chiang Rai (95% CI 0.01 to 0.23) and 0.53 in Salavan (95% CI 0.16 to 0.90). Conclusions: Locally specific conceptions and counterintuitive practices around antimicrobials can complicate AMR communication efforts and entail unforeseen consequences. Overcoming ‘knowledge deficits’ alone will therefore be insufficient for global AMR behaviour change. We call for an expansion of behavioural AMR strategies towards ‘AMR-sensitive interventions’ that address context-specific upstream drivers of antimicrobial use (eg, unemployment insurance) and complement education and awareness campaigns. Trial registration number: Clinicaltrials.gov identifier NCT03241316 .

Chemotherapy is often administered in openly designed hospital wards, where the possibilityof patient–patient social influence on health exists. Previous research found that socialrelationships influence cancer patient’s health; however, we have yet to understand socialinfluence among patients receiving chemotherapy in the hospital. We investigate the influenceof co-presence in a chemotherapy ward. We use data on 4,691 cancer patients undergoingchemotherapy in Oxfordshire, United Kingdom who average 59.8 years of age, and 44% areMale. We construct a network of patients where edges exist when patients are co-presentin the ward, weighted by both patients’ time in the ward. Social influence is based on totalweighted co-presence with focal patients’ immediate neighbors, considering neighbors’ 5-yearmortality. Generalized estimating equations evaluated the effect of neighbors’ 5-year mortalityon focal patient’s 5-year mortality. Each 1,000-unit increase in weighted co-presence with apatient who dies within 5 years increases a patient’s mortality odds by 42% (β = 0.357,CI:0.204,0.510). Each 1,000-unit increase in co-presence with a patient surviving 5 yearsreduces a patient’s odds of dying by 30% (β = −0.344, CI:−0.538,0.149). Our results suggestthat social influence occurs in chemotherapy wards, and thus may need to be considered inchemotherapy delivery.

Our work is motivated by an interest in constructing a protein-protein interaction network that captures key features associated with Parkinson's disease. While there is an abundance of subnetwork construction methods available, it is often far from obvious which subnetwork is the most suitable starting point for further investigation. We provide a method to assess whether a subnetwork constructed from a seed list (a list of nodes known to be important in the area of interest) differs significantly from a randomly generated subnetwork. The proposed method uses a Monte Carlo approach. As different seed lists can give rise to the same subnetwork, we control for redundancy by constructing a minimal seed list as the starting point for the significance test. The null model is based on random seed lists of the same length as a minimum seed list that generates the subnetwork, in this random seed list the nodes have (approximately) the same degree distribution as the nodes in the minimum seed list. We use this null model to select subnetworks which deviate significantly from random on an appropriate set of statistics and might capture useful information for a real world protein-protein interaction network.

Chemotherapy is often administered in openly designed hospital wards, where the possibility of patient–patient social influence on health exists. Previous research found that social relationships influence cancer patient's health; however, we have yet to understand social influence among patients receiving chemotherapy in the hospital. We investigate the influence of co-presence in a chemotherapy ward. We use data on 4,691 cancer patients undergoing chemotherapy in Oxfordshire, United Kingdom who average 59.8 years of age, and 44% are Male. We construct a network of patients where edges exist when patients are co-present in the ward, weighted by both patients' time in the ward. Social influence is based on total weighted co-presence with focal patients' immediate neighbors, considering neighbors' 5-year mortality. Generalized estimating equations evaluated the effect of neighbors' 5-year mortality on focal patient's 5-year mortality. Each 1,000-unit increase in weighted co-presence with a patient who dies within 5 years increases a patient's mortality odds by 42% (β = 0.357, CI:0.204,0.510). Each 1,000-unit increase in co-presence with a patient surviving 5 years reduces a patient's odds of dying by 30% (β =−0.344, CI:−0.538,0.149). Our results suggest that social influence occurs in chemotherapy wards, and thus may need to be considered in chemotherapy delivery.

Our work is motivated by an interest in constructing a protein–protein interaction network that captures key features associated with Parkinson’s disease. While there is an abundance of subnetwork construction methods available, it is often far from obvious which subnetwork is the most suitable starting point for further investigation. We provide a method to assess whether a subnetwork constructed from a seed list (a list of nodes known to be important in the area of interest) differs significantly from a randomly generated subnetwork. The proposed method uses a Monte Carlo approach. As different seed lists can give rise to the same subnetwork, we control for redundancy by constructing a minimal seed list as the starting point for the significance test. The null model is based on random seed lists of the same length as a minimum seed list that generates the subnetwork; in this random seed list the nodes have (approximately) the same degree distribution as the nodes in the minimum seed list. We use this null model to select subnetworks which deviate significantly from random on an appropriate set of statistics and might capture useful information for a real world protein–protein interaction network. The software used in this paper are available for download at https://sites.google.com/site/elliottande/ . The software is written in Python and uses the NetworkX library. Supplementary data are available at Bioinformatics online.

We examine the explicit use of social relationships as collateral against loan default, a core feature of modern microfinance. What structural factors determine

Human activities increasingly take place in online environments, providing novel opportunities for relating individual behaviors to population-level outcomes. I

Previous work has shown that species interacting in an ecosystem and actors transacting in an economic context may have notable similarities in behavior. However, the specific mechanism that may underlie similarities in nature and human systems has not been analyzed. Building on stochastic food-web models, we propose a parsimonious bipartite-cooperation model that reproduces the key features of mutualistic networks - degree distribution, nestedness and modularity -- for both ecological networks and socio-economic networks. Our analysis uses two diverse networks. Mutually-beneficial interactions between plants and their pollinators, and cooperative economic exchanges between designers and their contractors. We find that these mutualistic networks share a key hierarchical ordering of their members, along with an exponential constraint in the number and type of partners they can cooperate with. We use our model to show that slight changes in the interaction constraints can produce either extremely nested or random structures, revealing that these constraints play a key role in the evolution of mutualistic networks. This could also encourage a new systematic approach to study the functional and structural properties of networks. The surprising correspondence across mutualistic networks suggests their broadly representativeness and their potential role in the productive organization of exchange systems, both ecological and social.