By Zhu Zhimeng 朱知萌 Peking University School of Transnational Law
1. Introduction
The digitization and automation of the global economy have the potential to destabilize what is arguably the world’s most important tax tool–personal income taxes. Therefore it has wide-ranging economic and social implications.[1]
These effects include but are not limited to, an increase in the gap between rich and poor, an increase in unemployment, and a reduction in national tax revenues. After reviewing global literature, I found that the overall international consensus of experts is that it is imperative to tax robots.
In June 2017, Nobel laureate Robert Shiller said that to alleviate the income gap caused by the robot revolution, it is necessary to reconstruct the tax system and that the imposition of a “robot tax” may be politically more acceptable than simply taxing high-income people. Taxation as wage insurance helps people who have been robbed of their jobs by new technologies to complete the reemployment transition. This is a practice that coincides with humanity’s innate sense of justice.[2]
The McKinsey Global Institute has declared that AI has had a devastating effect on the achievements of the Industrial Revolution “at ten times the rate and at three hundred times the scale, or about three thousand times the rate.” The profound changes and disruptive impacts of technological advances in many fields have opened up a whole new era for us.[3]
Taxing robots has attracting Chinese officials’ attention. A 2019 article entitled “Is “Robot Tax” Feasible?”[4] by Shanghai Research Institute of the Shanghai Municipal People’s Government concludes that “it is particularly important to have good tax policies for intelligent robots” and vaguely suggests the designer of the Chinese robot tax should “combine the characteristics of China’s own tax system”. The scarce literature discussing taxing robots in China suggests the Chinese government is observing international developments rather than becoming involved as a rule-maker. Rather than taxing robots, promoting robots is now the focus of the Chinese government. While Chinese literature barely discusses whether robots should be taxed in China, the Shanghai government acknowledges in the article that robots are developing in China and that the discussion of a “robot tax” in China is also timely. This article will summarize the benefits and drawbacks of taxing robots in China by reviewing Chinese domestic and global literature and discuss whether the Chinese government will one day tax robots in light of China’s national conditions.
2. The benefits of taxing robots in China
Taxing robots is a new issue in tax law. Bill Gates popularized the notion in 2017 when he called for a tax on robots. Gates’ rationale, shared by many proponents, is simple. “If a human being was earning $50,000 a year, the government would collect tax on that income. If the human being had the misfortune of being replaced by a machine, the government would not collect the tax on income anymore. Plus, the poor worker is out of a job.”[5] While the Western world is heatedly discussing the reasons for a robot tax, Chinese scholars only did an incomplete discussion on this topic.[6] In summary, there are mainly four reasons leading global scholars to think that robots should be taxed.
First, taxing robots will narrow the gap between the rich and the poor. The 2021 OECD policy brief “Artificial Intelligence and Employment” concludes that AI is likely to further widen the gap in the labor market, especially between those who have the skills to use AI effectively and those who do not.[7] Ryan Abbott, professor of University of Surrey School of Law wrote in Harvard Law & Policy Review[8] 2018 that while automation creates wealth, the distribution of wealth is uneven. Since 1950, the CEO-to-worker pay ratio has increased a thousandfold, but overall wages have stagnated for thirty-five years. Increased automation is likely to accelerate these trends.[9] Guerreiro et al. from the National Bureau of Economic Research[10] recommend a robotics tax using a generational model that overlaps with conventional and unconventional skilled workers in the workforce.[11] A government’s goal is to redistribute income to displaced day-to-day workers. They found that “automation can destroy many of the jobs held by older generations and lead to a sharp rise in income inequality” and conclude that the best robot tax is not zero as long as there are skilled workers in the workforce.[12]
The Shanghai Research Institute of the Shanghai Municipal People’s Government has the same view—— Robots represent the progress of human society, but the key problem is that the benefits brought by new technologies are not equal to all people, and the newly created value and excess profits are mainly attributed to robot owners, that is, capitalists.[13]
Second, taxing robots can alleviate the risk of unemployment.[14] If a person makes $50,000 a year, the government taxes that income. If one is unfortunate to be replaced by machines, the government will no longer receive income taxes. In addition, the poor worker is unemployed.[15]
In 2013, Oxford Martin’s College released a study, “The Future of the Job Market,” showing that nearly half (47 percent) of U.S. jobs in 702 categories will be affected by automation technology over the next 20 years. In 2017, the MIT and Boston University study “Robots and Jobs” pointed out that for every 0.1 percentage point increase in the number of robots in the number of workers, the US employment rate will fall by 0.18-0.34 percentage points and workers’ wages will also be reduced by 0.25-0.5 percentage points.[16] At the same time, the South Korean government planned to implement a policy to reduce tax incentives for automation companies, aimed at collecting robot taxes in another way, with the aim of controlling the high unemployment rate. Some argue that state revenue and taxation should compensate people who have lost their jobs due to robots by providing them with skills training and employment opportunities. The Shanghai Research Institute of the Shanghai Municipal People’s Government also believes that low-skilled, labor-intensive industries and conventional repetitive jobs represented by migrant workers in China are being replaced by robots at an accelerated rate based on econometric model estimation.[17]
While some think creating more jobs is related to unemployment, some scholars argue that robots will create more jobs and actually boost employment.[18] Much of the stimulus driving these various calls for tax robots stems from society’s fear of unemployment and human intimacy with the workforce.[19] In history when concerns about the mass unemployment caused by automation arise, technology has created more jobs than it has destroyed.[20] After the technological revolution, society needs a period of “adjustment” to reorganize itself and then begins to benefit from new technologies. In other words, the employment rate will rise after the “adjustment” period.[21] Demographic observations made by Vikram Chand etc., an associate professor of the Faculty of Law of University of Lausanne suggest that a perspective concerning robots centered purely on economic employment is bound to lead to uncertain outcomes——“The question may not be how to tax technology, but how to tax it in a world where technology goes hand-in-hand with developments we haven’t seen in thousands of years (eg, potential migration of population, stopping for the first time in global population history, or climate change not seen since the end of the last ice age).” The authors summarize some of the measures taken by the government, as well as the various options in the academic literature, and still insist a tax on robots.[22]
Third, taxing robots will assure the tax revenues used in national infrastructure.[23] Without robot tax, countries using robots may lose a significant portion of the income they earn as a result of taxing formal employment income.[24]
Dissenting opinions found in various data that the collection of personal income tax remained relatively stable.[25] But the opposition data reflects only one aspect. At present, the property tax rate in most countries is much lower than the income tax rate. For example, China implements an excess progressive tax rate of 3% to 45% for the comprehensive income part of personal income tax, including wages, and the income tax rate for high-income people in the United States is 25% to 55%. According to the January 2017 study by the McKinsey Global Institute, even in a developed country like the United States, about half of its current work activities can be automated and intelligent through existing technologies, which corresponds to nearly $2.7 trillion in annual wages in the United States, in other words, intelligence and automation may lead to the loss of hundreds of billions or even trillions of dollars in tax revenues per year by all levels of government in the United States.[26]
Taxing robots will also impact tax allocation. Unlike humans, robots do not buy cars, clothes, food, electronic devices, or contract services. As a result, the production and consumption of goods and services are likely to decline. A decrease in consumption may result in a reduction in VAT collection.[27] Dissenting views are that there is ample evidence that the share of capital income in total income has been increasing for various reasons, including globalization. But this trend certainly cannot be extrapolated into the future, especially the adoption of smart web service robots and technological changes accelerating faster, and even leading to sudden structural changes in the economy.[28]
And, as the population ages, the public sector may need to increase social benefits and pensions through tax revenues.[29] This is also the problem the Chinese government faces.
Fourth, enterprises themselves should not have obstacles to tax payment. From the perspective of the enterprise itself, in the future, whether the company uses AI, or the company itself is an AI practitioner, the company’s efficiency will eventually increase. In general, the economic efficiency of the entire society will be improved, and the company should be willing to hand over a certain proportion to the state.[30]
3. The drawbacks of taxing robots in China
There are mainly four objections against taxing robots in global literature. However, in my view, the four reasons are all easily refuted or internally illogical.
The first objection says if the robot includes capital characterized by cross-border liquidity, the tax will result in relocation to jurisdictions with lower taxes. However, this argument is weak for several reasons:
“(i) The organic composition of financial capital moving in and out of global economies, against which the international community has earnestly raised the issue of base erosion and profit shifting, is quite different from the capital represented by robots or AI.
(ii) In whatever form the income from capital has been selected to be taxed in a particular tax code, the fact remains that capital income is the same as any other source of income and, therefore, should be taxed as all income. That capital income has turned out to be difficult to tax has not converted that income in form or substance to ownership of a unique property that forbids it from being taxed.
(iii) In the initial phase of robot technology, many countries offer tax incentives for their development; commensurately, there should be little fear of flight as in other forms of capital.
(iv) As the use of robots inevitably increases, raising productivity along with incomes, taxing robots has the potential of generating revenue for government expenditure.
(v) Empirical evidence does not reveal that low capital taxation is necessarily linked to high economic growth. In that case, tax havens would have exhibited high growth but, in reality, act more as conduits rather than destinations of growth. At the same time, areas with relatively high taxation such as the People’s Republic of China and India have tended to experience high growth.” [31]
Second, taxing robots does not hinder the pace of progress in robots. Proponents of taxing robots argue that there is a social cost to people’s displacement. Taxing machines would replace lost direct income and be used to train and educate unemployed workers. But the pace of progress is inevitable. No tax would significantly change that.[32] But taxes are not collected to discourage progress, but to distribute wealth more equitably.
Third, the robots to be taxed are hard to define. David Brunori, a partner at Quarles & Brady and a professor at The George Washington University, in “Taxing robots and some good tax policy” asks a series of questions: “How exactly do we define these taxable machines? Do we use some definition of “mechanical” or “programmatic”? Do we define taxable machinery through the replacement of workers? If that’s the case, aren’t all machines taxed? Isn’t it true that almost every machine has reduced working hours? Isn’t that what machines are all about? Thousands of auto workers have been replaced by bolted robotic arms. ATMs have replaced thousands of bank tellers. Self-driving trucks are replacing truck drivers. Email functionality basically ends the need for people to deliver documents. Nothing or nothing is taxed.” But this article also acknowledges that, in any case, all state tax systems need to be reformed.[33]
In addition, robots are technologies developed that enjoy Research and Development (R&D) discounts[34]. The adjustment of tax policy can provide guidance for the intelligent development of the industry. At present, most countries in the world encourage enterprises to carry out automated and intelligent research and development, which is reflected in the tax policy as direct tax incentives, as well as indirect preferential policies such as allowing enterprises to accelerate depreciation on the investment of automation equipment. In contrast, after the realization of intelligence and automation, the number of human employees required by enterprises will also be reduced, thereby reducing the social insurance and housing provident fund that enterprises need to bear for employees.[35];
In addition to the problem of defining machines, challenges are inevitable when it comes to how to tax robots. One of the challenges is that robots may be smart enough to avoid taxes on their own. As robots and artificial intelligence evolve further, it is expected that they may also begin to understand tax compliance norms and push tax avoidance to the limits. This can be achieved through an intensive and continuous search for global tax avoidance practices by advanced AI, and further, expand these practices through its own advanced technologies and processes. And only a very finely structured tax system can successfully bring robots into tax networks. [36] There are many theories about how to collect taxes. This article will not discuss the conundrum of how to collect taxes.
Fourth, taxing robots will influence innovation. There are renewed concerns about the potential for a regulatory tax on robots to slow the spread of robotics. The cost of a slowdown is broader economic inefficiencies and massive welfare losses.[37] Robots are more efficient, and taxing them reduces overall productivity.[38] If the cost of R&D and automation is less the tax benefits available to companies than they need to pay for “additional” employees, companies are likely to opt for an automated, intelligent path.[39]
4. Conclusion
I believe that the controversy on whether taxing robots or not is partly sustained by the immature development of robots. Because this industry is still in a preliminary stage and has not yet flourished to an advanced stage, we may not know whether we should suppress and inhibit our entire development a little bit by taxing robots. In my view, in the long run, robots are definitely going to be taxed, and the trend of social development is that which resource creates high wealth will definitely become the focus of taxation.[40] I suggest that the current stage is to cultivate the development of our artificial intelligence and improve efficiency. The robot industry is still in its infancy. Let it move forward because it is a good thing. Our work efficiency can be improved until robots bring universal value to everyone.
In view of the above, it is predicted that the Chinese government will ultimately tax robots based on weighing the factors for and against them. However, an irresponsible policy will result in disastrous impacts given such a density of population, like seriously affecting innovation or destroying China’s fledgling robotics industry. Therefore, it is understandable why the Chinese government is monitoring the situation.
- In A. Grau (ed), Interactive Robotics: Legal, Ethical, Social and Economic Aspects, Springer Nature, 2022, Ch 17. available at: https://link.springer.com/book/10.1007/978-3-031-04305-5 ↑
- Shanghai Research Institute of Shanghai Municipal People’s Government. “Is “Robot Tax” Feasible?”. China Development Observation Magazine, 8 Nov. 2019, pp. 1, available at http://www.chinado.cn/?p=8515. ↑
- Shi Junming, Chen Shuangxiong. “Discussion on the collection of ‘robot tax’ in the intelligent era”. Dacheng Shanghai Office, 2020, pp.1. https://mp.weixin.qq.com/s/Asm8cEowdzEyP18scV_nrQ. Accessed on 10 July 2022. ↑
- Shi Junming, Chen Shuangxiong. “Discussion on the collection of ‘robot tax’ in the intelligent era”. Dacheng Shanghai Office, 2020, pp.1. https://mp.weixin.qq.com/s/Asm8cEowdzEyP18scV_nrQ. Accessed on 10 July 2022. ↑
- David Brunori. “Taxing robots and some good tax policy.” State & Local Taxes Weekly, vol. 18, Art. 5, 2017, pp. 1. ↑
- The scarce discussions on this issue in China I can find are: A review of research on robot tax at home and abroad(国内外对征收“机器人税”的研究述评) written by staff from Fiscal and Tax Research Division, Research Bureau, People’s Bank of China(中国人民银行研究局财政税收研究处)(2020) http://www.cqvip.com/qk/81137x/202003/7101305559.html, which is a word to word copy of Legal issues with the robot tax(机器人税的法律问题) written by staff from Shanghai Pudong New Area State (Local) Taxation Bureau(上海市浦东新区国家(地方)税务局) in 2018 https://www.cnki.net/kcms/doi/10.19376/j.cnki.cn10-1142/f.2018.03.006.html and Shanghai Research Institute of Shanghai Municipal People’s Government(上海市人民政府上海研究院): Is “Robot Tax” Feasible?(上海市人民政府上海研究院) in 2019 https://mp.weixin.qq.com/s/eI0D5LPSLqbhezYpKo0iIw saved at: https://web.archive.org/web/20220711135150/https://mp.weixin.qq.com/s/eI0D5LPSLqbhezYpKo0iIw. ↑
- Stijn Broecke, Artificial Intelligence and Employment, OECD, December 2021 ↑
- The Harvard Law & Policy Review (HLPR) is published twice annually and serves as the official law review of the American Constitution Society for Law and Policy. For more information please see https://harvardlpr.com/about/. ↑
- Ryan Abbott, Bret Bogenschneider. “Should robots pay taxes? tax policy in the age of automation.” Harvard Law & Policy Review, 2018, pp. 155-156, available at https://openresearch.surrey.ac.uk/esploro/outputs/journalArticle/Should-Robots-Pay-Taxes-Tax-Policy-in-the-Age-of-Automation/99513953102346. ↑
- The National Bureau of Economic Research (NBER) is a private, nonpartisan organization that facilitates cutting-edge investigation and analysis of major economic issues. It disseminates research findings to academics, public and private-sector decision-makers, and the public by posting more than 1,200 working papers and convening more than 120 scholarly conferences, each year. https://www.nber.org/about-nber ↑
- Guerreiro, J., S. Rebelo, and P. Teles. 2020. Should Robots Be Taxed? Working Paper. 23806, National Bureau of Economic Research. https://www.nber.org/system/files/working_ papers/w23806/w23806.pdf. ↑
- Parthasarathi Shome. “Taxation of Robots.” The Governance Brief, ADB, Issue 44, 2022, p. 4. https://www.adb.org/publications/taxation-robots ↑
- Shanghai Research Institute of Shanghai Municipal People’s Government. “Is “Robot Tax” Feasible?”. China Development Observation Magazine, 8 Nov. 2019, pp. 1. ↑
- Shanghai Research Institute of Shanghai Municipal People’s Government. “Is “Robot Tax” Feasible?”. China Development Observation Magazine, 8 Nov. 2019, pp. 1. ↑
- David Brunori. “Taxing robots and some good tax policy.” State & Local Taxes Weekly, vol. 18, Art. 5, 2017, pp. 1. ↑
- Shanghai Research Institute of Shanghai Municipal People’s Government. “Is “Robot Tax” Feasible?”. China Development Observation Magazine, 8 Nov. 2019, pp. 1. ↑
- See more Shanghai Research Institute of Shanghai Municipal People’s Government. “Is “Robot Tax” Feasible?”. China Development Observation Magazine, 8 Nov. 2019, pp. 1. ↑
- Ryan Abbott, Bret Bogenschneider. “Should robots pay taxes? tax policy in the age of automation.” Harvard Law & Policy Review, 2018, pp. 158-159. ↑
- Kathryn Kisska-Schulze, Rodney P. Mock. “The robotic revolution: a tax policy collision course.” Temple Law Review, 2021, pp. 218, available at https://www.templelawreview.org/article/the-robotic-revolution-a-tax-policy-collision-course/. ↑
- Ryan Abbott, Bret Bogenschneider. “Should robots pay taxes? tax policy in the age of automation.” Harvard Law & Policy Review, 2018, pp. 158-159. ↑
- Vikram Chand, Svetislav Kostić and Ariene Reis. “Taxing Artificial Intelligence and Robots: Critical Assessment of Potential Policy Solutions and Recommendation for Alternative Approaches – Sovereign Measure: Education Taxes/Global Measure: Global Education Tax or Planetary Tax.” World Tax Journal, 2020, pp. 724-728, available at bfd.org/sites/default/files/2021-09/International%20-%20Taxing%20Artificial%20Intelligence%20and%20Robots%20Critical%20Assessment%20of%20Potential%20Policy%20Solutions%20and%20Recommendation%20for%20Alternative%20Approaches%20-%20IBFD.pdf. ↑
- Vikram Chand, Svetislav Kostić and Ariene Reis. “Taxing Artificial Intelligence and Robots: Critical Assessment of Potential Policy Solutions and Recommendation for Alternative Approaches – Sovereign Measure: Education Taxes/Global Measure: Global Education Tax or Planetary Tax.” World Tax Journal, 2020, pp. 728-729. ↑
- EBMA CENTER. Internet + Bootcamp: 10 Insights from Campers, Shanghai Jiaotong University Antai EMBA, 20 Nov 2019, https://mp.weixin.qq.com/s/Vd_0NLnsdiJD3seuU8E9cA. Accessed 10 July 2022. ↑
- Vikram Chand, Svetislav Kostić and Ariene Reis. “Taxing Artificial Intelligence and Robots: Critical Assessment of Potential Policy Solutions and Recommendation for Alternative Approaches – Sovereign Measure: Education Taxes/Global Measure: Global Education Tax or Planetary Tax.” World Tax Journal, 2020, pp. 756. ↑
- Shafik Hebous, Alexander Klemm. “Alexa and Siri Do Not Pay Taxes.” International Monetary Fund, 2022, pp. 4, available at https://iceanet.org/wp-content/uploads/2022/01/Klemm.pdf. ↑
- Shi Junming, Chen Shuangxiong. “Discussion on the collection of ‘robot tax’ in the intelligent era”. Dacheng Shanghai Office, 2020, pp.1. https://mp.weixin.qq.com/s/Asm8cEowdzEyP18scV_nrQ. Accessed on 10 July 2022. ↑
- Vikram Chand, Svetislav Kostić and Ariene Reis. “Taxing Artificial Intelligence and Robots: Critical Assessment of Potential Policy Solutions and Recommendation for Alternative Approaches – Sovereign Measure: Education Taxes/Global Measure: Global Education Tax or Planetary Tax.” World Tax Journal, 2020, pp. 714. ↑
- Shafik Hebous, Alexander Klemm. “Alexa and Siri Do Not Pay Taxes.” International Monetary Fund, 2022, pp. 4. ↑
- Parthasarathi Shome. “Taxation of Robots.” The Governance Brief, ADB, Issue 44, 2022, p. 4. ↑
- EBMA CENTER. Internet + Bootcamp: 10 Insights from Campers, Shanghai Jiaotong University Antai EMBA, 20 Nov 2019, https://mp.weixin.qq.com/s/Vd_0NLnsdiJD3seuU8E9cA. Accessed 10 July 2022. ↑
- Parthasarathi Shome. “Taxation of Robots.” The Governance Brief, ADB, Issue 44, 2022, p. 4. ↑
- David Brunori. “Taxing robots and some good tax policy.” State & Local Taxes Weekly, vol. 18, Art. 5, 2017, pp. 2. ↑
- David Brunori. “Taxing robots and some good tax policy.” State & Local Taxes Weekly, vol. 18, Art. 5, 2017, pp. 2. ↑
- Discounts for high technology industry. ↑
- Shi Junming, Chen Shuangxiong. “Discussion on the collection of ‘robot tax’ in the intelligent era”. Dacheng Shanghai Office, 2020, pp.1. https://mp.weixin.qq.com/s/Asm8cEowdzEyP18scV_nrQ. Accessed on 10 July 2022. ↑
- Parthasarathi Shome. “Taxation of Robots.” The Governance Brief, ADB, Issue 44, 2022, p. 4. ↑
- In A. Grau (ed), Interactive Robotics: Legal, Ethical, Social and Economic Aspects, Springer Nature, 2022, Ch 17. ↑
- Parthasarathi Shome. “Taxation of Robots.” The Governance Brief, ADB, Issue 44, 2022, p. 4. ↑
- Shi Junming, Chen Shuangxiong. “Discussion on the collection of ‘robot tax’ in the intelligent era”. Dacheng Shanghai Office, 2020, pp.1. https://mp.weixin.qq.com/s/Asm8cEowdzEyP18scV_nrQ. Accessed on 10 July 2022. ↑
- EBMA CENTER. Internet + Bootcamp: 10 Insights from Campers, Shanghai Jiaotong University Antai EMBA, 20 Nov 2019, https://mp.weixin.qq.com/s/Vd_0NLnsdiJD3seuU8E9cA. Accessed 10 July 2022. ↑