Community Incentive and Distance Instrumental Variable Method

In Simple Words

Imagine trying to determine if eating at a local diner makes people happy, but happier people naturally choose to eat out more often. To isolate the real effect, a researcher gives randomly assigned discount coupons to various neighborhoods; comparing neighborhood happiness based solely on coupon density reveals the pure effect of diner visits. Similarly, offering random cash incentives and varying distance to testing centers across villages creates exogenous variation in community testing rates to measure genuine peer learning effects.

Research question

Measuring the causal effect of community-level health information or disease prevalence on individual behavior is confounded by endogeneity, as prevalence and testing decisions correlate with unobserved individual and village characteristics.

Participants

N = 2,654 individuals in 117 villages assigned incentives in 2004, with N = 1,995 followed up in 2006; 78.2% received non-zero monetary incentives (average $1.02), living an average of 1.99 km from results centers.

Study Design

Village-level averages of individual monetary incentives ($0 to $3) and distance to mobile counseling centers used as instrumental variables for the proportion of community members receiving test results.

Results

The first-stage regression predicting the proportion of village members learning HIV results yields a strong F-statistic of 23.98 without controls, with an average of 72.3% of village members learning test results (SD = 0.186).

Takeaway

Utilize village-level aggregated experimental treatments as instrumental variables when evaluating social spillover effects to eliminate endogenous peer selection bias.

Related findings

• Cross-Community Audit Methodology for Bias Detection →Economics (general)

  To test if store owners treat people differently based on heritage, researchers send out actor buyers from different ethnic backgrounds who follow the exact same script and purchase the exact same items. By comparing what happens to each buyer at the exact same stores, researchers can catch discrimination red-handed.

• Hawthorne Effect Control via Diagnostic-Only Counterfactual Design →Economics (general)

  Imagine a study testing whether a new study strategy helps students pass exams. If researchers stand inside the classroom watching the students every day, students might work harder simply because someone is watching them, not because the strategy works.

• Lottery-Based Instrumental Variable Identification Strategy →Economics (general)

  Imagine trying to test if wearing running shoes makes people faster, but fast people are naturally more likely to buy running shoes. Using a lottery to hand out running shoes completely breaks that confounding connection, ensuring the only difference between the two groups is pure chance.

• Instrumental Variables Correction for Management Measurement Error →Economics (general)

  Imagine trying to measure how much a medicine helps sick patients by looking at regular hospital records. Because doctors give the medicine primarily when patients get extremely ill, simple statistical comparisons make the medicine look weak or useless.