Published papers
Klinenberg, D. (2023). Does Deplatforming Work? Journal of Conflict Resolution, https://doi.org/10.1177/00220027231188909
Link to replication files
Presented at WEAI 2022 - Cryptocurrency and Social Media Group
Feature in The Forward
Danny Klinenberg (2022) Synthetic Control with Time Varying Coefficients: A State Space Approach with Bayesian Shrinkage, Journal of Business and Economic Statistics, DOI: 10.1080/07350015.2022.2102025
UCSB outstanding second- year paper 2020
Presented at the IAAE machine learning seminar 2020
Danny Klinenberg & Richard Startz (2022) Covid, colleges, and classes, Applied Economics, DOI: 10.1080/00036846.2022.2091110
Featured on the Brookings Institute Brown Center Chalkboard (link)
Working papers
Selling violent extremism (Job Market Paper)
Abstract:
Domestic violent extremist organizations have become a significant threat to Western democracies, with some groups attracting tens of thousands of members. How did these groups amass such large followings? Surprisingly, some of these organizations behave similarly to firms. I study the effect of three firm tactics – membership discounts, armed events, and advertisements – on the demand for membership in the Oath Keepers, America’s largest paramilitary organization. Using a synthetic control-like strategy, I find that discounts cause new member sign-ups to increase by over 60% and cause-related marketing by 150% respectively; however, sports sponsorships decrease it. Demand for membership is less responsive to price changes in counties with more income inequality, but more so in heavily politically conservative counties.
Link to most recent version
Presented at ESOC annual conference, 2022, Cal Poly- San Luis Obispo economics' department seminar series, UCSB Applied Microeconomics Lunch
Nonacademic coverage: Boston Globe, Wikipedia, Marginal Revolution
Deterrence using Reaction Curves: An empirical analysis of the Israeli-Gaza Conflict
with Eli Berman, Prabin Khadka, and Esteban Klor
Abstract:
We use response curves in a repeated game to formalize key aspects of integrated deterrence: escalation, de-escalation, incomplete deterrence, and deterrence by denial. In our approach, episodes of violence are due to the interaction of response curves, which follow a strategic logic of disincentivizing opponents from attacking, through both deterrence and compellence. To maintain credibility both sides punish attacks, disincentivizing larger attacks and yielding nonviolent lulls. We empirically estimate those curves using detailed incident data from the Israel-Gaza conflict between 2007 and 2018. Our estimates match the dynamics of the raw data: very frequent episodes of low lethality violent exchange. Response curves are stable and exhibit a posture consistent with incomplete deterrence: i.e., episodes de-escalate, but not to complete nonviolence in equilibrium. Major Israeli military operations shift the Gazan reaction curve inwards, yielding a less violent equilibrium.
DRC 29Jan2024 West Point.pdfWork-in-progress
Timing is Everything: Estimating strategic response in sequential data (with Eli Berman, Prabin Khadka, and Esteban Klor)
Many applied economic studies aim to estimate strategic behavior through reaction curves. Examples include two-sided conflicts, or economic trade wars, and algorithmic pricing between firms. Analysis is usually performed at a pre-specified time interval, such as days, weeks, months, years, etc, using a Vector Autoregression (VAR). Yet sides may respond within a day to one action, but wait a month after another. If data is recorded in arbitrary time intervals, then the researcher may mistake waiting to act for inaction. We analytically show that VAR analyses do not recover true reaction curves if the timing of reaction is not accurately recorded. We discuss alternative approaches to estimate reaction curves and investigate their usefulness in a Monte Carlo simulation.
Content Differentiation across Platforms (with Nathan Roll)
Online content creators differentiate their products across platforms, according to our analysis of over 421,000 videos from over 3,800 content creators on YouTube and competitor Bitchute from June 2019 to January 2022. Exploiting a feature of Bitchute, we identify which videos were directly uploaded to Bitchute and which were first uploaded to YouTube and then back downloaded onto Bitchute. We find that content creators produce videos that discuss ethnicity, death, conflict, swearing, and pose "us vs them" scenarios more when uploading directly to Bitchute compared to YouTube.