Working papers

1. Causal Inference under Endogenous Network Interference: A Design-Based Mediation Framework

   Eaman Jahani

   Abs

   Standard methodologies for causal inference under network interference typically rely on the assumption that the social network mediating peer influence is measured pre-treatment and remains static. However, in many practical settings, ranging from public health interventions to digital platform optimizations, the treatment itself endogenously alters the network topology. In such scenarios, using the pre-treatment network snapshot to define exposure conditions leads to severe exposure misspecification and biased causal estimates. This paper investigates the inferential challenges posed by endogenous, treatment-induced network evolution and introduces a design-based mediation framework to recover meaningful causal estimands under network interference. We first formally characterize the bias in standard Horvitz-Thompson estimators that arises when researchers utilize pre-treatment networks to define target exposures. We prove that this bias is driven by the probability of misclassification and the outcome discrepancy between valid and invalid exposure states. To resolve this, we introduce a design-based mediation perspective that conceptualizes the post-treatment network as a high-dimensional mediator between the randomization design and observed outcomes. This framework orthogonally decomposes the global Average Treatment Effect (ATE) into two distinct causal pathways: the Natural Direct Effect (NDE), which captures effects operating through the treatment while holding the network distribution fixed, and the Natural Indirect Effect (NIE), which captures effects operating strictly through treatment-induced shifts in the network structure. We establish the conditions for valid inference on these mediated estimands and develop unbiased inverse-probability weighted (IPW) estimators. We further prove their consistency and asymptotic normality under sparse dependence and derive conservative variance estimators to facilitate valid inference despite the structural unobservability of joint counterfactual propensities. Analytic evaluations and Monte Carlo simulations demonstrate that our proposed estimators effectively eliminate finite-sample bias and maintain valid nominal coverage, providing the methodological framework for disentangling the direct efficacy of an intervention from indirect and structural channels in dynamic environments.

2. Celebrity messages reduce online hate and limit its spread

   Eaman Jahani, Blas Kolic, Manuel Tonneau, Hause Lin, Daniel Barkoczi, Edwin Ikhuoria, Victor Orozco, and Samuel Fraiberger

   Abs HTML

   Online hate spreads rapidly, yet little is known about whether preventive and scalable strategies can curb it. We conducted the largest randomized controlled trial of hate speech prevention to date: a 20-week messaging campaign on X in Nigeria targeting ethnic hate. 73,136 users who had previously engaged with hate speech were randomly assigned to receive prosocial video messages from Nigerian celebrities. The campaign reduced hate content by 2.5% to 5.5% during treatment, with about 75% of the reduction persisting over the following four months. Reaching a larger share of a user’s audience reduced amplification of that user’s hate posts among both treated and untreated users, cutting hate reposts by over 50% for the most exposed accounts. Scalable messaging can limit online hate without removing content.

3. Network Structure of Unequal Diffusion

   Eaman Jahani, Dean Eckles, and Alex Pentland

   Abs HTML PDF

   We develop a theoretical network model that captures the network structure of unequal diffusion. First, we argue that homophily alone does not fully explain the extent of unequal diffusion in the network. We show that any departure from the uniform distribution of links to information sources among members of a group limits the diffusion of information to the group as a whole. The variation from equal opportunity baseline not only has first order effects since some individuals will have fewer direct links to information sources, but also second order effects as the whole group will have fewer diffusion paths to the information sources. Second, we show empirically that most networks exhibit higher susceptibility to unequal diffusion across various grouping variables (e.g. gender or race) than expected simply by measures of homophily. Finally, we develop a variant of the Stochastic Block Model that parameterizes the cross-group linking of each node and by doing so fits not only assortativity on paths of length 1, but also diffusion assortativity on paths of length 2 and 3.

Refereed

1. Prompt Adaptation as a Dynamic Complement in Generative AI Systems

   Eaman Jahani, Benjamin Manning, Joe Zhang, Hong-Yi TuYe, Mohammed Alsobay, Christos Nicolaides, Siddharth Suri, and David Holtz

   Information Systems Research 2026

   Abs HTML

   Do users automatically benefit when generative AI models improve, or does realizing those gains require changing how they interact with the technology? We ran two preregistered experiments with 3,750 participants submitting nearly 37,000 prompts across two versions of OpenAI’s DALL-E to find out. The answer depends critically on task structure. In a bounded task with a clear objective, replicating a specific target image, roughly half of the performance gain came from the model itself, and half came from users naturally adjusting their prompts. In an open-ended creative task, designing logos for hypothetical organizations, nearly all of the improvement was attributable to the model alone. We then asked whether automated prompt rewriting, a feature embedded in many commercial AI products, could substitute for this user adaptation. It could not: rewriting modestly helped in the creative task but substantially hurt performance in the replication task, where precise user control mattered most. For organizations deploying AI in precision-oriented workflows, our results point to an underappreciated priority: investing in how people learn to interact with a model may matter as much as the model itself.

2. Long ties, disruptive life events, and economic prosperity

   Eaman Jahani, Samuel P. Fraiberger, Michael Bailey, and Dean Eckles

   Proceedings of the National Academy of Sciences 2023

   Abs HTML

   Social networks shape and reflect economic life. Prior studies have identified long ties, which connect people who lack mutual contacts, as a correlate of individuals’ success within firms and places’ economic prosperity. However, we lack population-scale evidence of the individual-level link between long ties and economic prosperity, and why some people have more long ties remains obscure. Here, using a social network constructed from interactions on Facebook, we establish a robust association between long ties and economic outcomes and study disruptive life events hypothesized to cause formation of long ties. Consistent with prior aggregated results, administrative units with a higher fraction of long ties tend to have higher-income and economic mobility. Individuals with more long ties live in higher-income places and have higher values of proxies for economic prosperity (e.g., using more Internet-connected devices and making more donations). Furthermore, having stronger long ties (i.e., with higher intensity of interaction) is associated with better outcomes, consistent with an advantage from the structural diversity constituted by long ties, rather than them being weak ties per se. We then study the role of disruptive life events in the formation of long ties. Individuals who have migrated between US states, have transferred between high schools, or have attended college out-of-state have a higher fraction of long ties among their contacts many years after the event. Overall, these results suggest that long ties are robustly associated with economic prosperity and highlight roles for important life experiences in developing and maintaining long ties.

3. Implications of COVID-19 vaccination heterogeneity in mobility networks

   Yuan Yuan,  Eaman Jahani, Shengjia Zhao, Yong-Yeol Ahn, and Alex Pentland

   Communications Physics 2023

   Abs HTML

   Our study utilizes network science to examine how uneven vaccine distribution affects mass vaccination strategies in the United States. Using mobility network data and epidemiological models, we find that distributing a fixed quantity of additional vaccines across Census Block Groups (CBGs) can vary case count reductions by up to 200l%. This highlights the impact of vaccination heterogeneity in mobility networks on epidemic outcomes. Our efficient algorithm identifies optimal vaccine distribution for maximum case reduction. Simulations show a possible 9.5% decrease in case numbers with just a 1% increase in the national vaccination rate if vaccines are optimally distributed. This result surpasses those from other vaccine distribution models. Our findings underline the need for policymakers to understand the interaction between vaccination patterns and mobility networks, suggesting that grasping geographical vaccine uptake variations could be as crucial as raising the overall vaccination rate.

4. An Online experiment during the 2020 US–Iran crisis shows that exposure to common enemies can increase political polarization

   Eaman Jahani, Natalie M. Gallagher, Friedolin Merhout, Nicolo Cavalli, Douglas Guilbeault, Yan Leng, and Christopher A. Bail

   Scientific Reports Jul 2022

   Abs HTML PDF

   Longstanding theory indicates a common enemy can mitigate conflict between members of rival groups. We tested this hypothesis in a pre-registered experiment where 1,670 Republicans and Democrats in the United States were asked to complete a collaborative online task with an automated agent or “bot" that was labelled as a member of the opposing party. Prior to this task, we exposed respondents to primes about a) a common enemy (involving threats from Iran, China, and Russia); b) a patriotic event; or c) a neutral, apolitical prime. Though we observed no significant differences in the behavior of Democrats as a result of these primes, we found that Republicans—and particularly those with very strong conservative views—were significantly less likely to cooperate with Democrats when primed about a common enemy. We also observed lower rates of cooperation among Republicans who participated in our study during the 2020 Iran crisis, which occurred in the middle of our data collection. These findings indicate common enemies may not reduce inter-group conflict in highly polarized societies, and contribute to a growing number of studies that find evidence of asymmetric political polarization. We conclude by discussing the implications of these findings for research in social psychology, political conflict, and the rapidly expanding field of computational social science.

5. Social Debunking of Misinformation on WhatsApp: The Case for Strong and In-Group Ties

   Irene V. Pasquetto,  Eaman Jahani, Shubham Atreja, and Matthew Baum

   Proc. ACM Hum.-Comput. Interact. Apr 2022

   Abs HTML PDF

   In this paper, we argue that WhatsApp can play an important role in correcting misinformation. We show how specific WhatsApp affordances (flexibility in format and audience selection) and existing social capital (prevalence of strong ties; homophily in political groups) can be leveraged to maximize the re-sharing of debunking messages, such as those accessed by WhatsApp users via ChatBots and Tip-Lines. Debunking messages received in the format of audio files generated more interest and were more effective in correcting beliefs than text- or image-based messages. In addition, we found clear evidence that users re-share debunks at higher rates when they received them from people close to them (strong ties), from individuals who generally agree with them politically (in-group members), or when both conditions are met. We suggest that WhatsApp leverages our findings to maximize the re-share of those fact-checks that are already circulating on the platform by using the existing social capital in the network, unlocking the potential for such debunks to reach a larger audience on WhatsApp.

6. ScamCoins, S*** Posters, and the Search for the Next Bitcoin^TM: Collective Sensemaking in Cryptocurrency Discussions

   Eaman Jahani,  Eaman, Peter Krafft, Yoshiko Suhara, Esteban Moro, and Alex Pentland

   Proc. ACM Hum.-Comput. Interact. Nov 2018

   Abs HTML PDF

   Participants in cryptocurrency markets are in constant communication with each other about the latest coins and news releases. Do these conversations build hype through the contagiousness of excitement, help the community process information, or play some other role? Using a novel dataset from a major cryptocurrency forum, we conduct an exploratory study of the characteristics of online discussion around cryptocurrencies. Through a regression analysis, we find that coins with more information available and higher levels of technical innovation are associated with higher quality discussion. People who talk about "serious" coins tend to participate in discussion displaying signatures of collective intelligence and information processing, while people who talk about "less serious" coins tend to display signatures of hype and naı̈vety. Interviews with experienced forum members also confirm these quantitative findings. These results highlight the varied roles of discussion in the cryptocurrency ecosystem and suggest that discussion of serious coins may be oriented towards earnest, perhaps more accurate, attempts at discovering which coins are likely to succeed.

7. Improving official statistics in emerging markets using machine learning and mobile phone data

   Eaman Jahani, Pal Sundsøy, Johannes Bjelland, Linus Bengtsson, Alex Pentland, and Yves Alexandre Montjoye

   EPJ Data Science Nov 2017

   Abs HTML PDF

   Mobile phones are one of the fastest growing technologies in the developing world with global penetration rates reaching 90%. Mobile phone data, also called CDR, are generated everytime phones are used and recorded by carriers at scale. CDR have generated groundbreaking insights in public health, official statistics, and logistics. However, the fact that most phones in developing countries are prepaid means that the data lacks key information about the user, including gender and other demographic variables. This precludes numerous uses of this data in social science and development economic research. It furthermore severely prevents the development of humanitarian applications such as the use of mobile phone data to target aid towards the most vulnerable groups during crisis. We developed a framework to extract more than 1400 features from standard mobile phone data and used them to predict useful individual characteristics and group estimates. We here present a systematic cross-country study of the applicability of machine learning for dataset augmentation at low cost. We validate our framework by showing how it can be used to reliably predict gender and other information for more than half a million people in two countries. We show how standard machine learning algorithms trained on only 10,000 users are sufficient to predict individual’s gender with an accuracy ranging from 74.3 to 88.4% in a developed country and from 74.5 to 79.7% in a developing country using only metadata. This is significantly higher than previous approaches and, once calibrated, gives highly accurate estimates of gender balance in groups. Performance suffers only marginally if we reduce the training size to 5,000, but significantly decreases in a smaller training set. We finally show that our indicators capture a large range of behavioral traits using factor analysis and that the framework can be used to predict other indicators of vulnerability such as age or socio-economic status. Mobile phone data has a great potential for good and our framework allows this data to be augmented with vulnerability and other information at a fraction of the cost.

Lightly refereed

1. Differential Network Effects on Economic Outcomes: A Structural Perspective

   Eaman Jahani, Guillaume Saint-Jacques, Pal Sundsøy, Johannes Bjelland, Esteban Moro, and Alex Pentland

   In Social Informatics. Nov 2017

   Abs HTML PDF

   In a study of about 33,000 individuals in a south Asian country, we find that structural diversity, measured as the fraction of open triads in an ego-network, shows a relatively strong association with individual income. After including all the relevant control variables, the effect of structural diversity becomes exclusive to the highly educated individuals. We hypothesize these results are due to concentrated distribution of economic opportunities among the highly educated social strata combined with homophily among members of the same group. This process leads to two important societal consequences: extra network advantages for the highly educated, similar to the rich club effect, and inadequate diffusion of economic opportunities to the low educated social strata.