Automation, Taxation, and Social Security Sustainability

This paper examines the tax policy adjustments required to maintain a constant ratio of social security contributions to GDP within a pay-as-you-go system, depending on the degree of automation. We explore three alternative approaches to taxing autonomous capital, defined as a combination of robots and artificial intelligence: i) an income tax on autonomous capital, ii) an ad-valorem tax on investment in autonomous capital, and iii) a social security tax on robots, paid by employers. The analysis is conducted under two technological scenarios: one in which autonomous capital substitutes for both traditional capital and labor, and another in which it substitutes for labor while complementing traditional capital. For each scenario and tax scheme, we calculate the tax rates necessary to preserve the relative size of social security contributions. Our results indicate that, in the long run, the most efficient approach to taxing autonomous capital is a social security tax on autonomous robots paid by employers using such technology. Finally, we examine how these specific taxation schemes affect the functional distribution of income.