Robot Tax

Definition

The robot tax is a legislative strategy aimed at discouraging the replacement of workers by machines and at strengthening the social safety net for those who are displaced. It covers both taxing robots themselves and reducing the tax incentives previously granted for automation investment.

Its fiscal rationale is that automation erodes government revenue. When robots replace workers, the state loses income tax revenue; in the United States, payroll and income taxes account for 81 percent of federal revenue. Whether robots actually create or destroy jobs on balance, however, is not settled: robots create and eliminate jobs at the same time.

History

Support for an automation tax among US politicians goes back to a 1940 bill introduced by Senator Joseph C. O'Mahoney. The start year of 2017 for this entry marks the point at which international debate intensified; earlier legislative attempts and tax credit reductions also existed.

The revival of the debate in the twenty-first century followed new technological advances such as machine learning. Estimates of automation risk diverge widely. One estimate put 47 percent of US employment at risk of automation, while an OECD-based estimate suggested around 9 percent.

In February 2017 the European Parliament discussed robot ethics standards in Mady Delvaux's proposal but rejected the robot tax portion. That same year Bill Gates endorsed the robot tax, explaining that if a worker doing 50,000 dollars' worth of factory work is replaced by a robot, the robot should be taxed the same way, and Mark Cuban also backed it, citing Quincy Larson's writing on the acceleration of technological unemployment. In 2019 the World Bank's World Development Report, written by Simeon Djankov and Federica Saliola, opposed a robot tax on the grounds that it would reduce productivity and increase tax avoidance by large firms and shareholders.

The South Korean case

South Korea recorded the world's highest robot density, concentrated in manufacturing: 1,000 robots per 10,000 workers, eight times the world average, amid a shrinking working-age population and the fastest ageing in the OECD.

In August 2017 the Moon Jae-in administration passed what was widely called the world's first robot tax. It did not tax robots but reduced the tax credit granted for robot investment. The tax revision was a measure to raise corporate taxation in response to the Trump administration's corporate tax cut, and as productivity-enhancing, safety and environmental protection facility credits were all lowered from 3·5·7 to 1·3·7 (environmental protection from 3·5·10 to 1·3·10), the productivity-enhancing share was interpreted as a robot tax. The scheme was due to sunset on 31 December 2017 but the government extended it to 2019, and according to a National Assembly Budget Office report the tax benefits amounted to 243.7 billion won.

Assessments of the measure diverged. A single senior official at the Ministry of Economy and Finance commented that reducing the productivity-enhancing facility investment tax credit could be called an early-stage Korean-style robot tax; this was not an official government classification. A National Assembly official called that an over-interpretation and placed the change in a continuing series of tax credit reductions, which had already occurred under the Park Geun-hye government: in 2015 Saenuri Party lawmaker Na Seong-lin introduced a bill amending the Restriction of Special Taxation Act to change 3·5·7 to 2·4·7, and a 2012 amendment excluded employment-replacing equipment such as automatic grinding and cutting machines from the credit.

Examples

Among politicians, San Francisco Supervisor Jane Kim set up a task force in 2017, UK Labour leader Jeremy Corbyn called for a robot tax the same year, and in 2019 New York Mayor Bill de Blasio proposed a robot tax charging large firms five years of corporate tax on jobs lost to automation, around his presidential run. Stephen Hawking also criticised machine owners in 2015 for lobbying against wealth redistribution.

On the academic side, the tax law scholar Xavier Oberson proposed giving robots taxpaying capacity so that government spending can be maintained even as workers' taxable income falls, and taxing robot owners until robots themselves have the ability to pay.

Distinctions

The robot tax aims to slow automation and reduce worker displacement in order to protect government revenue. Universal basic income, by contrast, is put forward as a social safety net response to robot-caused unemployment. Elon Musk proposed UBI in response to automation-driven unemployment, Gates backed the robot tax, and Yanis Varoufakis proposed a universal basic dividend instead of a robot tax.

The distinction between cutting tax credits and taxing robots is itself contested. South Korea's 2017 measure did not tax robots, so its classification as a robot tax set the Ministry of Economy and Finance official's remark about an early-stage Korean-style robot tax against the National Assembly official's objection that this was an over-interpretation.

Debates and relations

There are also restrained designs on the supporting side. Francisco Ossandón held that a limited robot tax is possible only when large firms pay it, when it is confined to specific activities (industry and finance), when the definition of a robot is narrow (physical smart machines or intangible intelligent software in financial activity) and when the rate is low, and he opposed a general robot tax.

Critics argue that the definition of a robot is fluid enough to make taxation meaningless. Jim Stanford and Tshilidzi Marwala pointed out that many twenty-first-century devices contain autonomous elements, blurring the definition. The robotics industry and policy authorities also objected: Savioke and the Association for Advancing Automation opposed the tax as a penalty on innovation, ABB chief executive Ulrich Spiesshofer likened it to taxing software and noted that countries with higher automation rates have lower unemployment, and EU Commissioner Andrus Ansip rejected it on the grounds that jurisdictions adopting it would lose competitiveness.