Technology and State Capacity: Experimental Evidence from Illegal Mining in Colombia

52 Pages Posted: 1 Oct 2021 Last revised: 28 Sep 2022

See all articles by Santiago Saavedra

Santiago Saavedra

Universidad del Rosario - Faculty of Economics

Date Written: September 19, 2022


New monitoring technologies can help curb illegal activities by reducing information asymmetries between enforcing and monitoring government agents. I created a novel technology using machine-learning on satellite imagery to detect illegal mining. Then I disclosed the technology and possible mine locations to government agents to study the impact on illegal activity. I randomly assigned municipalities to one of four groups: (1) information to the observer (local government) about the technology and potential mine locations in his jurisdiction; (2) information to the enforcer (National government); (3) information to both observer and enforcer, and (4) a control group, where I informed no one. I use an independent expert-validated dataset that measures gold mining to evaluate the effect of the intervention. I find that the treatment effect is relatively similar regardless of who is informed: in treated municipalities, illegal mining is reduced by 11% in the disclosed locations and surrounding areas. When accounting for negative spillovers --- increases in illegal mining in areas not targeted by the information --- the net reduction is still sizable (7%). These results illustrate the benefits of new technologies for building state capacity and reducing illegal activity.

Keywords: Illegal mining, Monitoring technology, Colombia

JEL Classification: H26, K42, O13, O17, Q53

Suggested Citation

Saavedra, Santiago, Technology and State Capacity: Experimental Evidence from Illegal Mining in Colombia (September 19, 2022). Available at SSRN: or

Santiago Saavedra (Contact Author)

Universidad del Rosario - Faculty of Economics ( email )

Casa Pedro Fermín
Calle 12C # 4-69

Do you have a job opening that you would like to promote on SSRN?

Paper statistics

Abstract Views
PlumX Metrics