US robot impacts in developing countries: Evidence from Colombian workers

This paper examines how robotization in developed countries impacts workers’ earnings and labor dynamics in developing countries. Using nearly a decade of Colombian Social Security records and US robot adoption data, we provide robust evidence that workers in industries highly exposed to US robots experience lower cumulative earnings overall and lower employment for older workers and those employed in larger firms. Workers switching firms within the same industry avoid earnings losses, while those moving across industries face the largest losses, consistent with loss of industry-specific human capital. These effects are driven by reduced Colombian exports to the US, particularly in the primary sector, and are more pronounced among older, high-earning workers and those in larger firms, plausibly more integrated into global value chains and, therefore, more directly exposed to foreign automation.