What if the US imposes a tax on robots? This idea has been discussed by political analysts, experts, and Bill Gates (who favors the idea). Because robots can replace jobs, the argument goes, a stiff tax on them would give businesses an incentive to help retain workers, and offset the drop in payroll taxes when robots are used. So far, South Korea has reduced incentives for companies to install robots; EU politicians, on the other hand, considered a robot tax but did not.

A study by economists at MIT examines the existing evidence and suggests that the best policy in this situation would be to tax robots, but only slightly. Likewise, a tax on foreign trade would reduce jobs in the United States, the study found.

“Our research indicates that taxes on robots or imported goods should be very small,” said Arnaud Costinot, an MIT economist and co-author of a paper published on the research. “Although robots have an impact on income inequality … they still lead to an optimal tax that is modest.”

A separate study finds that taxes on robots should range from 1 percent to 3.7 percent of their value, while business taxes range from 0.03 percent to 0.03 percent. 0.11 percent, considering the current US income tax rate.

“We went into it not knowing what was going to happen,” said Iván Werning, an MIT economist and co-author of the study. “We had all the ingredients to be a big tax, so by stopping technology or trade, there would be less inequality, but … right now, we’re seeing a tax on one number, and it will be of trade, even less. tax.”

The paper, “Robots, Trade, and Luddism: A Statistically Sufficient Approach to Optimal Technology Regulation,” appears in advance online at Review of economic studies. Costinot is a professor of economics and associate head of the MIT Department of Economics; Werning is the department’s Robert M. Solow Professor of Economics.

Adequate statistics: Salary

The key to the study is that the experts did not start with an a priori idea about whether or not it is worth taxing robots and commerce. Instead, they used a “statistically adequate” method, examining scientific evidence on the subject.

For example, a study by MIT economist Daron Acemoglu and Boston University economist Pascual Restrepo found that in the United States from 1990 to 2007, one robot per 1,000 workers reduced the employment-to-population ratio by 0.2 percent; Each robot introduced into production replaced about 3.3 workers, while the increase in robots in the workplace lowered wages by about 0.4 percent.

In conducting their policy analysis, Costinot and Werning turn to this empirical study and others. They built a model to evaluate a few different scenarios, and introduced tools like income tax as another way to address income inequality.

“We have these other tools, though not perfect, to deal with inequality,” Werning said. “We think it’s wrong to talk about taxes on robots and commerce as if these are our only tools for redistribution.”

Specifically, the researchers used wage distribution data across five quintiles of income in the United States — the top 20 percent, the bottom 20 percent, and so on — to estimate the need for robot tax and sales. While empirical data indicates that technology and trade have changed the distribution of wages, the extent to which this change has contributed to the production of robots and the trade tax estimates of Costinot and Werning. This has the advantage of simplicity; the overall wage range helps economists avoid modeling by making too many assumptions about, for example, the role of automation in the workplace.

“I think where we are technically, we can link wages and taxes without making specific assumptions about technology and how production works,” Werning said. “It’s all included in that distribution product. We ask a lot of this empirical work. But we don’t make assumptions that we can’t test the rest of the economy. “

Costinot added: “If you are comfortable with high-level assumptions about how the market works, we can tell you that the only interesting thing that drives the best policy of robots or Chinese goods should be the answer “Wages in the amount of money. distribution of income, which, fortunately for us, people have tried to predict.”

In addition to robots, weather procedures and more

In addition to the minimum tax rate, the study includes additional conclusions about technology and revenue trends. Perhaps paradoxically, research concludes that after more robots are introduced into the economy, the impact of each additional robot on wages may decrease. In the future, the tax on robots may be reduced further.

“You can have a situation where we are deeply concerned about re-introduction, we have more robots, we have more sales, but taxes are going down,” said Costinot. If the economy is flooded with robots, he added, “That marginal robot you get in the economy is not that much of a difference.”

The learning process can also be applied to topics other than automation and marketing. There is a growing body of empirical work on, for example, the effects of climate change on income inequality, as well as similar studies on the effects of immigration, education and other things in the salary. Given the growing amount of empirical data in these areas, the type of model Costinot and Werning in this paper can be applied to determine, say, the level of a carbon tax, if the goal is to to support fair income distribution.

“There are many other applications,” Werning said. “There is a common logic to these problems, which this approach will lead to.” This suggests another avenue for future research related to the current journal.

In the meantime, for those who thought about overtaxing the robots, however, “they are qualitatively correct, but meaningless,” Werning concluded.



Source link