2026 Undergrad Econ Debate: Should firms that replace workers with AI be taxes on lost salaries?

Good afternoon everyone. Welcome to the UEC economics debate for this year. Today's important topic will be if companies should have replaced employees with AI and if they should be taxed on lost salaries. Yeah. So should companies that replace employees with AI be taxed on lost salaries? So please come to the podium one at a time and speak into the mic here. And now the affirmative side you have 15 minutes and you may begin. >> Every generation has said that new technology will destroy jobs and for the most part they are wrong until now. And here's why AI is different. Looking into some historical context, we can see many instances of new technology replacing jobs. One is a loom replacing the hand weaver. But while the invention of the loom replaced the hand weaver, it also created factory jobs, textile engineers and clothing designers. Similarly, the assembly line in the 1910s killed manual labor but created electricians, technicians and supervisors. And with computers, they replaced clerical work but also created jobs such as programming and IT. All of these technological advancements have something in common. It displaces humans upwards. Technological Technology automated the physical work which gave birth to cognitive and creative roles. But AI attacks this pattern. Goldman Sachs found that AI can match or outperform 47% of industry professionals on economically valuable tasks. Additionally, Anthropic's own CEO has warned that within 5 years AI could eliminate half of all entry-level white-collar jobs. AI doesn't just push people to a new tier. It removes a tier entirely. Focusing on taxation, a good historical contextual example is the Luddites of 1811. People had mocked them for fearing tech, but they do not fear tech and they do not fear innovation. They were skilled workers demanding that productivity gains be shared, something that this tax aims to do. Society has navigated new technologies and come up with new rules, such as child labor laws, the 40-hour work week, and unemployment insurance, which are all made to share the economic gains of new technology. An AI tax would be no different in principle. In looking towards the future, we can see that AI is not a distant threat. The World Economic Forum's 2025 Future of Jobs Report found that 41% of employers are already planning to downsize their workforce due to AI. AI is replacing jobs and that is in the company's budget. While new jobs may appear, workers cannot afford to wait 30 years for that to happen. This tax builds the bridge that we need now. >> In 2013, French economist Thomas Piketty released his magnum opus, Capital in the 21st Century. In his work, he uses historical data from from 20 countries to to model inequality over time. He finds two variables, R, the the rate of return on capital, and G, the rate of growth in an in an economy over years. Shown here. The crux of his argument is that when R is greater than G, inequality increases. We contend that AI as a as a technology will increase R and worsen wealth inequality. Wealth inequality on its own has horrible effects on our economy, includ- and our society, including loss of social mobility, pol- political po- polarization and social tension, psychological and health and health declines, as well as political influence on democracy. This we- this tax on lost salaries uh serves as a as an intervention into the economy. From 1913 to 1950, we can see inter we we can see intervention or a loss in R with respect to G, where G is greater than R. This comes at a time where wealth was taxed heavily. And this tax aims to do so. AI also has a peculiar effect on inflation and deflation. In the short run, it'll have inflation as it has an a large demand a large demand injection in the form of credit um due to create creating uh chip processors and bottle and inflation will also be caused through bottlenecks as water is scarce and chips are scarce. Both of these in the short run cause could potentially cause inflation. In the long run, however, AI could cause deflation. As AI as a technology will displace as we mentioned we mentioned before white-collar jobs and therefore um take money out of the pockets of of working-class families, lower the money supply, and lower profits for firms. Causing deflation in the long run. Our tax aims to to sedate this. In a sense, it'll slow down AI development allowing wealth inequality to to advance a little bit but not as strongly as it could while while stalling of development of this technology enough for us to create better reforms and advance on this technology at sustainable pace. >> If you look beyond just general unemployment, we see an underemployment rate. We see an underemployment rate around 23.6%. This is including people who are making less than $26,000 per year before taxes. And if we look even further into this, we see for college graduates the unemployment rate is already at 42.5%. This is a stark example of how AI will grow this unemployment rate as entry-level jobs are the most susceptible to AI replacement because of how easily they are automated. And we're already seeing this in some majors as criminal justice has an unemployment rate of 65% and it is one of the most susceptible jobs to AI replacement because it is the most susceptible to automation. And it makes it a direct example of how AI will grow this unemployment even further. And we already have seen 100,000 layoffs or hiring freezes due to AI in 2025, and that number is just going to grow as AI grows further and further. As the CEO of Anthropic said that AI could spike unemployment from 10 to 20% in the next 1 to 5 years. And aside from that general unemployment that would be catastrophic to our economy, we would also see major problems in our federal revenue as federal revenue is very tied to employment. And this is because it is 50% from income tax and 35% from social security tax, making it 85% from employment taxes. And we saw this decrease in federal revenue from loss in employment and loss in income in 2007 when GDP federal revenue relevant to GDP fell from 17.9% to 14.6% in 2009. So, AI is going to cause more unemployment, which is going to deplete our federal revenue, causing short-term problems in funding unemployment and welfare. And when it comes to unemployment, this cause a very vicious cycle where we're seeing unemployment on the rise due to AI, and then also funding for necessary unemployment benefits being lowered because of AI as well. And in the long term, this could cause even more problems if we see major depletions in federal revenue, making the government have to cut very necessary spending in programs and services. So, this AI tax is a solution to this problem. It would replace the loss in revenue from income tax and social security tax, and it would replace that and make and it would replace that tax and make sure that the that the income is still generated for the federal government. And in the long run, it would also be able to help that as well. It would also make sure AI development is slowed down, so necessary reforms can be made and and to keep unemployment and underemployment at manageable rates. >> [clears throat] >> As mentioned by Connor, federal funding could be at risk and could potentially even go down in the next few years because of disruptions of AI. And with many welfare systems dependent on federal funding, such as the social security program and Medicaid, which are both programs which millions of Americans rely on, we could see both of these programs along with many other welfare systems be put under extreme strain. As it is, it's expected that the social security trust fund could run out by 2032 at the earliest, and if not, 2067, and that too at a best-case scenario. In the case that AI uh replaces a large chunk of human labor and employees, a huge chunk of taxes which the government collects would be depleted, and uh the government would thus struggle to support these welfare programs, which again, millions of Americans rely on. And this is not just a American phenomena, but this would be a problem worldwide because uh program or sorry, countries like Korea and uh alliances like the EU also have strong social security nets which are at risk because of their heavy reliance on federal funding from tax. We can see here that in 2019 around 88% of the social security benefits in the US were collected from payroll tax. And this just reiterates that the loss on taxes in the case that human labor is replaced by AI on a large scale, which it is expected to be, it would prove to be a a big loss to the social security uh network. Mass AI implementation could also lead to a significant decline in quality of life. Uh even though it might lead to higher GDP outputs and productivity in the nation, it is expected that there would be significant unemployment as a byproduct of that, which could lead to a large deterioration in the standard of life conditions. And this com- this is a phenomenon which we have seen before in American towns which were in the Rust Belt. And once those towns shifted from manufacturing in the 1980s, quality of life uh went down severely once the economy shifted from manufacturing to more service-based industries. Unemployment could also then trickle down between generations, and that could create a vicious cycle of poverty amongst many American families. And that too is a byproduct of AI replacing human labor, and in the case that there is no appropriate tax uh introduced. We also know that tax funding will be required in the coming years for safety programs such as AI insurance, uh unemployment benefits, and other programs to prepare workers and employees for the incoming shift of AI into the workplace. And I had the privilege of asking Mr. Jared Bernstein, the economic advisor to the White House, and he too firmly believes that it would be crucial to introduce some form of tax or some some form of tax credit in order to support these programs in order to prepare the population for the incoming future, which is inevitable. We also can see uh in the latest White House AI action plan that the government is planning on introducing AI-scaled programs and has also introduced some programs with the Bureau of Labor Statistics to try to predict what sort of effects would happen on on the wage markets and in the labor markets, but it is crucial to have a rigorous system in place which includes features like AI insurance and unemployment benefits, and these programs should be able to sustain themselves in the coming years. Tax on AI would be difficult to implement, but it could be done like how South Korea did in 2017 by introducing a tax on robots, which essentially uh taxes companies based on automation which replaces a skilled worker. Other forms of tax could include increased corporate taxes to offset for the lost taxes due to the unemployment caused by AI, along with offering tax incentives and credits to companies which retain and retrain workers. There's no doubt in our minds that without any intervention, companies would be able to extract profits while shifting the long-term cost and burden of unemployment onto the existing public welfare systems, which are already strained as it is. Our group firmly believes in the need to introduce some form of tax on AI that replaces human labor as it is essential to prepare to hope for the best but prepare for the worst. And this is something that we do know that AI is unlike any other technological revolution in the past and it will have severe negative reverberations on our communities. Thus, the time to act is now and I would like to reiterate our claim and our group's claim on the urgent need to tax AI on uh its replaced human labor. >> Uh all right. Thank you, uh negative side. Uh we're going to now move on to the opening statements from the negative side and we'll So, we'll begin now. You guys also have 15 minutes. >> [snorts] >> Good afternoon, everyone. Thank you for coming today. Uh we're the negative side of the debate and we're arguing that companies uh who replace employees with AI should not be taxed on lost salaries. Uh our main arguments include that this tax would discourage the adoption of productivity-enhancing technology, incentivize firms to misclassify AI-related layoffs, harm [snorts] small businesses, uh create incentives for offshoring to avoid taxes, and cause inflation. First, industries that have adopted AI to save time has have also increased their overall productivity. An increased productivity decreases unemployment. As you can see from the red dashed line in the first graph, industries with 1% point higher time savings experienced 2.7% point higher productivity growth relative to their pre-pandemic trend. >> [snorts] >> In the second graph, you can see how historically increased productivity has resulted in lower unemployment. It may be argued that historical trends might not accurately predict the effects of uh that AI will have, but hiring for AI-related roles has increased even more. A study co-authored by MIT Sloan Associate Professor Lawrence Schmidt found that even workers in high-wage roles heavily exposed to AI, positions near the top of the pay scale, saw their share of total employment grow by about 3% over the last 5 years. That's because AI boosted firm productivity. Companies that use the technology grew faster, which [snorts] helped sustain or even expand headcount in high-exposure positions. >> If we introduce a tax on companies that that replace employees with AI, we have to assume that companies will be transparent about why the layoffs are happening, but that assumption doesn't hold up in the real world. Major businesses already have strong incentives to frame layoffs in ways that protect their reputation, avoid backlash, and limit regulatory scrutiny. If a new tax is tied specifically to AI-driven job loss, companies won't simply accept that cost. They'll adapt their messaging. Layoffs will be attributed to restructuring, market conditions, or efficiency improvements, even when AI is the underlying driver. In effect, the policy the policy risks creating a system where the most important information, what AI is actually doing to the workforce, becomes harder, not easier to track. The lack of transparency has broader consequences. Policymakers, economists, and the public depend on accurate data to understand how technical how technological changes affect on If companies systematically obscure the role of AI to avoid taxation, we lose the ability to measure its real impact. That makes it harder to design effective responses, whether that's workforce retaining retraining, education reform, or social safety nets. Instead of confronting the economic shifts caused by AI head-on, we'd be navigating a distorted picture shaped by the corporate incentives to minimize tax exposure. Ultimately, a tax like this doesn't just risk being ineffective, it may actively undermine our understanding of the problem it's meant to solve. Rather than encouraging honest reporting and responsible adoption of AI, it encourages concealment. If the goal is to address the risks AI poses to workers and the economy, we need policies that promote clarity and accountability, not ones that push companies to hide the truth. >> I also want to extend a bit on Logan's part on transparency against AI tax on lost employees based on budget cuts. Should a company be taxed over budget cuts, it could lead the business to bankruptcy or necessary setbacks to continue growing. Also, it could drive more competition to unbalance. Implementing a certain tax on AI uh specific tax then rolls to define a salary on how much was lost to a machine. Government budget cuts like the IRS weaken the ability to implement or enforce new taxes on companies that automate jobs. Meanwhile, companies are increasingly replacing workers to fund AI, which reduces tax revenue even further. An example of this is small businesses. The result of firing workers or changing workers could have have an effect if they're taxed with AI. They could use it as an helping tool, but still require workers but involve uh temporary pause. They adopt AI not for profit maximizing, but also for survival to keep them affiliated in uh graph over there. While they don't have HR team to manage tax in this scenario, also most small businesses don't have a need to replace full-time employees with AI at all. Only certain roles of scheduling, bookkeeping, or marketing. Putting a tax could punish efficiency and innovation. Making it harder for smaller companies to compete with large businesses and create uncertainty to discourage small businesses from growing. >> Building off of Donald's point, if a small business or medium-size business is already using AI just for survival, a tax doesn't create uncertainty, it creates an ultimatum. In today's globalized economy, that ultimatum will lead to companies offshoring to countries with more conducive taxation rates. We can We have to recognize AI is a weightless capital. And unlike a physical factory or coal mine, AI infrastructure is just code. If the United States makes it too expensive to innovate here, firms will just limit won't just limit or even stop the use of AI. They will simply move their digital headquarters, their data centers, and their high-value management roles to countries that welcome this technology instead of penalizing it. To understand the scale of what we risk losing, please look at this chart um that lists the highest-rated countries that privatize AI investments. As you can see, the United States is the global leader with over 471 billion invested in AI over the last decade, nearly four times our closest competitor, China. This investment won't isn't just a number, it is the drive behind our future productivity. By taxing lost salaries, we are effectively placing an innovation penalty on our own economy. This tax will force American firms to choose between staying staying domestically under technological stagnation or moving to countries like Ireland and Singapore where they have zero automation taxes and are aggressively trying to catch up with the US. When we hinder the adoption of new technologies through taxation, we don't have jobs. We invite economic paralysis, thus incentivizing companies to stick with slow and efficient processes just to avoid a tax bill or follow the alternative offshoring. In a global market, you cannot compete while holding yourself back. If we make progress more expensive within the US, we are no longer protecting workers. We are simply exporting our competitive advantage and the high volume jobs of the future to the rest of the world. >> To better understand this proposal to tax companies for AI replacing jobs, must acknowledge the potential benefits and consequences for the American people. To do this, it is essential to look at who is paying the tax and what will be done with said tax. The most recent example of a mass tax that comes to mind is President Trump's tariffs. On liberation day in April of 2025, Trump initiated a mass tariff of 10% on nearly all imports. This immediately dropped the S&P 500 by over 10% to which it did recover quickly, but this indicated a scare in the American consumer. This scare was further backed up by the rising consumer price index and inflation data in the following months. March of 2025 had CPI numbers of 319. Just 6 months later in September, that figure was nearly 325, marking a 1.6% increase. To further acknowledge that point, inflation rose from 2.4% in March to 3% in September of 2025. In words, the tariff was passed on to consumers. The correlation between Trump's tariffs and a potential tax on AI replaced jobs is clear. While the potential AI tax intends to punish corporations for replacing jobs by eating into profits, we all know that is an economic fantasy. In simple terms, a tax on AI-replaced jobs will raise prices for the American consumer, representing potential stagflation as jobs get replaced and unemployment rises. The velocity of money will slow down, and this recession indicator will be met with rising prices and a rising CPI, further hurting the American economy and the American consumer. When it comes to the opposite side of this tax, we must look at the receivers. This sparks the question, can Indo-Americans trust their government to use tax revenue for the betterment of society? The numbers say no. According to a 2024 Yahoo Finance poll, merely 18% of Americans believe their tax dollars are being spent effectively. From the same study, respondents were asked what they are not okay with their tax revenue going towards. The most common answer was that Americans were not supportive of their tax dollars being put towards war funding. Since February 28th, America has been in a complicated conflict with Iran. This war has cost $59 billion. While it has cooled off as of late, it's reasonable to assume that a surplus in tax revenue will go overseas to supporting defense. Serving as evidence for this point, Trump published a 2027 budget request earlier this month. He requested [snorts] $1.5 trillion allocated to defense spending, the most ever. With this came roughly a 10% cut on all domestic spending. Therefore, if a tax on jobs replaced with AI, the money would go straight overseas instead of being invested back home for those who have lost their jobs. In conclusion, a tax on businesses that replace workers with AI will inevitably fall on the consumers. This will raise inflation during times of rising unemployment. Subsequently, while the supposed goal of this tax is to support those who have lost jobs, there is no credible evidence that the surplus will be used to support them. Americans will be paying more into a government that they do not trust to spend responsibly. This tax will not support the American citizens. That is all for our opening. Thank you. >> Uh Here is my question. >> Yeah, for sure. >> All right. Thank you both sides for your opening statements. Uh we will now go into a 5-minute break, but once we pick back up, we will get into our rebuttal sessions where both teams will go back and forth in 4-minute blocks, starting with the affirmatives, but uh for now we'll go into a 5-minute break. Thank you. All right. Uh our 5-minute break is now over. Uh so we will start our first round of rebuttals, starting with the affirmative side again. Uh so please begin. You have 4 minutes. >> According to our colleagues, um AI AI development will move overseas, and it's a weightless capital. This simply is not true. AI infrastructure is heavily expensive and is one of the main bottlenecks going into the future. By 2030, electricity demand will rise 5.7%. 50% of that is due to AI. And um Open AI um has demanded that the US built 100 gigawatts of energy a year to accommodate this demand. According to Deloitte, by 20 um by 2027, um 92 gigawatts are needed, but all grid capacity is sold out. Um um Besides electricity and um and um and grids, we also need microchips. It's not It's not easy to move AI to an AI development to another country. We have physical things needed to create um to develop this technology. >> The US government should not be scared of development in other countries. As you showed, the US is already miles ahead of China and other countries in AI development, and the US is also not afraid to ban other technologies from other countries like we did with China and cars. This isn't So, AI is not something we should be scared of in other countries in their development. And also, even if they do develop in a faster rate than us, it is important to note that this tax is a protection of employment. It's That's the main reason for this entire tax, to make sure we focus on employment and the American people first. >> The other side says that the US citizens would be dubious of where their taxes would go, and therefore the tax would not be important and should not be implemented as it would not be substantial to the population, but this is the opposite as the tax would provide money to the government for services, which would help the population and would reduce losses of jobs to AI as taxes would incentivize companies to not replace labor with AI. And stated previously, within 5 years, um 50% of entry white-collar jobs will be taken. Um and employment or unemployment will then hurt the population. So, saying that the population will not care about this is misleading and harmful. >> As of 2024, it was cheaper to hire a human worker as opposed to AI, according to a MIT study. And this statement was said by Neil Thompson, who's the principal investigator at MIT CSAIL. I say this because you the opposing team raised the point of small businesses being at risk of going bankrupt. This honestly this refutes that. Uh and additionally, small businesses and business businesses in general receive a lot of incentives and benefits for hiring human labor, such as the work opportunity credit. Additionally, uh regarding the figure which showed the productivity growth rate versus unemployment, which uh compared the productivity growth rate and unemployment between 1960 to 2000. But between those four decades, most, if not all, technological additions and, you know, new new technologies were involved to assist human labor and not replace them. So, this uh this trend of productivity going up and causing unemployment to reduce with the addition of AI at a mass level is just something we can't predict. To close this, I would like to reiterate our team's statement and our belief in taxing AI and that it's not related to punishing innovation and halting growth, but it's to ensure that the benefits of AI are shared equally amongst our communities. >> Okay, thank you, affirmative side. Uh negative side, it's now your turn. You have 4 minutes. >> Um to start, to the point of AI taking away entry-level jobs and the point that you just made that protecting jobs should be the main focus of this tax, AI is a tool and it is what you make of it. And you still need people to analyze and control it. Companies cannot be fully founded on AI usage. And I think if we force companies to keep redundant positions, these entry-level jobs that we've proven are we no longer need because of the productivity of these AI machines, we're actually discouraging innovation towards new, more effective roles. If taxing these companies, like you say, will cause them to contain remain keeping most of these entry-level positions, then we're allowing them to stay stagnant and not find the actual jobs and create the actual careers that will be needed in this future as we move in this new time of AI. And again, to this point, AI technological advances required this transition. All technological advances required this transition. AI is not unique in its improvements. We must learn how to work with these tools instead of resisting change Uh to keep entry-level jobs the way that they have been in the present time. To the point that the tax is necessary for federal wealth, uh tax will mean nothing without a dedicated aid to improve the job market. And as we've kind of discussed, as David mentioned, it is true that we cannot guarantee a place for these taxes to go to, even if there was goals for some sort of a re-education uh program, which I don't know if it was actually talked about, but something like that is not guaranteed to actually work based on how our government works. But even if this tax was put in place, most companies would choose to pay this tax instead of keeping on redundant employees. Employees do end up being more costly in the long term as you incorporate all the employees that would be paid off, unlike um a similar AI machine. Not every AI machine requires all this large-scale uh infrastructure and money. A lot of people are dependent on tools in smaller ways that are still impacting salaries and workers. And a lot of companies would rather pay these costs than uh pay for employees, which are more costly with their training and the time off and the things that come with having real human capital, as opposed to AI and machines that are able to do more in a short amount of time. >> [snorts] >> And to the point that it is necessary to prevent long-term economic issues, it is true that AI takes away jobs, but a tax is not the effective solution. As we mentioned in our statement, it will only make the problem harder to track. Companies will find ways to make excuses for why these uh taxes or these people are being put out of their jobs or why they're moving overseas to make sure they're not paying for these taxes. If we force companies to pay these taxes, they're going to bury their leads, which means in a few years when we're continuing to have these problems with entry-level jobs vanishing and uh companies paying their employees less and less, it's going to be much more difficult to find an effective reality to actually target this tax. >> [snorts] >> Um in addition, tax um by nature is designed to discourage a business decision. Um like I'd said, in a free market, there many loopholes around this discouraging system. Instead, we should promote policies that establish better floor wages instead of capping something. Um Um the important thing to focus on here is innovation and how the US does in the real world. As you had mentioned in your rebuttal statement, thank you. As you had mentioned in your rebuttal statement, um that you think that the US does not need to compete against other countries. This is just frankly not true. We might be in the lead right now, but in order to keep that lead we need to continue innovating and continuing to encourage and incentivize our companies to do what they can to find new solutions for this new world in which we are using AI to improve on things instead of simply staying in a stagnant reality just so we can keep on entry-level jobs that are no longer needed. Um as more companies are able to use these productivity, we will fall behind. >> [snorts] >> Um So yeah, just to reiterate our final point. Um the point is not that AI is not taking jobs. This is definitely a true thing, especially with entry-level jobs, but what we need to focus on is finding a better solution that is not a tax to find a way to encourage companies to do so while not discouraging from federal gains and from business success. Um okay. >> Thank you. Thank you. Okay, thank you for round one. Uh we're going to go into round two of rebuttals starting with the affirmatives. Uh you have 4 minutes. Uh you bring up redundant jobs um and how they can be seen as useless um or just waste of money, but how will uh senior-level jobs be filled in the future if it is estimated that 50% of white-collar entry jobs are expected to be replaced? Um you cannot just become skilled in something. It takes years, it takes practice um to and dedication to the job um to become a senior in the in the profession and become talented. Um and this will result um in less skilled labor in the future, and um, job shortages for important roles that run our society. In addition to this, as um, as they've admitted themselves, jobs will be lost. And this is a serious macroeconomic issue. AI is not a person, it cannot consume, it cannot buy. As jobs are lost, aggregate demand will go down, and this is a large issue for our economy. >> And to the point that this tax will fully slow down development of AI and make it very slow, that's not true. This tax is made so that reform can be made. It's made to slow down the progression at a very minimal rate. That's why the tax only is for replaced employees. And that minimal rate will allow for more reform and AI growth to be made by the US government. >> In 2025 alone, 55,000 workers in the US were laid off. This is a number which economists, journalists, everyone is expecting to go up exponentially in the coming years in both the white-collar fields and the blue-collar fields. We know that AI is replacing human labor at an alarming late alarming rate. And this, as Sebastian said, will lead to extreme macroeconomic problems. I would also like to go over the fact of AI AI taxes potentially hindering development and causing the US's competitors to gain an edge advantage, which is not necessarily the case. Uh, from the start, we've reiterated that our stance on AI taxes would be would not necessarily be to harm innovation, but would be to find the sweet spot between innovation and growth and enforcing an AI tax does not mean that America's development would halt. It would just mean that America would continue to develop while considering the welfare of its own citizens. And um with regards to the points made regarding workers being or entry-level jobs being replaceable and in the coming years that um you know, human human uh that AI you know, it's easier to incorporate. That is not always necessarily true. AI when it when and if it if it should be used as a tool which would be helping human labor as opposed to replacing it, that could yield to extremely productive results. >> In order to offset this lost entry-level jobs, we propose along this along this uh tax a job guarantee. This will This will provide a um entry-level jobs to people at a minimum at a minimum wage and stabilize the economy macroeconomically. This is an effective use of this tax and will ensure prosperity for the future. >> Okay, uh thank you affirmative side. Uh negative side, uh you guys can begin for 4 more minutes. Just one more time. Thank you. >> Hey, uh I missed my homework uh because there's war going on in Iran. Um this would not make any sense to any professor as like a makeup uh excuse cuz those things have nothing in in relation. I feel like that's the exact same situation we're battling in here. Look, we have lost jobs uh since the pandemic. The job market has definitely slowed down. But then are we seeing a correlation? Is the 55,000 jobs that we have lost this year primarily due to AI? Is it so easy to quantify the jobs lost and how they're being lost to AI? Because I feel and I'm and our entire stance is basically AI helps us boost the productivity and it's really difficult to quantify an employee's productivity in terms of the output they take out. We need to really make sure the entire productivity boost that these companies are getting has to be passed down to the employees and does not end up becoming a profit on the company's books. We've seen as per Karl Marx, the profit for companies is definitely a theft. We are seeing this theft on an increased scale. The same employee is now able to make bigger and bigger numbers and bigger and bigger profit for the company and the company in return instead of passing down that profit is making sure that they are getting the money and instead make sure that people get laid. Instead of making sure that people get like taxed, companies get taxed for firing people, I think we should find out a way to make sure this productivity benefit gets passed down to people. We never really said that the companies would not care. They need they argue we need entry-level jobs to establish senior level roles in first. We don't really think the entry-level roles should completely be lost. Instead we argue that this tax will not incentivize innovation in these entry-level jobs. We need these jobs to be necessary because they are useful, not because people need entries into companies. [snorts] We need people to do something and provide value at at at their worst days. We must find a better solution that will not be found using taxing businesses. Meant to slow down AI growth to make room for reform was a pay was a point that was made by the uh the other side. Uh I don't think that's ideal. I need we need uncapped innovation. Uh we need to uh we we don't really need to slow down on innovation. We need to really promote and incentivize the growth the new technology and really embrace what's out there uh instead of becoming slowed and uh instead of becoming slowed and and and making our point. We need to ascend. We need to promote businesses to bring their own uh business to bring their in innovation all across, make it more accessible, and make it more uh powerful for everyone to get access to them. Anytime we spend uh slowing down is the time we are spending falling behind. We need to make sure these companies are not left unchecked on the profit they make and they are held responsible and they make sure there are better reforms for that money to flow back into the economy. Thank you. >> All right. Thank you both sides for your rebuttal session. Uh we're now going to move into a 10-minute segment of uh Q&A from the judges and then the affirmative side can begin with their answers. So, yeah, we'll move on to that now. >> So, thank you for your presentations. Some very interesting on both sides. Uh >> [clears throat] >> let me first ask you all the on the no side. You um you seem to argue different sides of this. On the one hand, you argued that putting on a tax uh would lead to inflation. And I I assume that would occur because firms would raise their prices. But on the other hand, you argue that firms would not raise their prices, they would move abroad. They would just do it someplace else. So, my first question to your side is which is it? Are they going to raise their prices and stay and continue employing, or are they going to pick up and leave? On your side, you argued that I couldn't I wasn't clear what was supposed to be done with the tax revenue that was raised. On the one hand, it seemed like what you were arguing is that that was simply replace lost revenue that would occur because there'd be fewer taxes for social security payments, etc. But on the other hand, you sometimes implied that this tax revenue would be used to help the workers themselves who were being laid off. So, if I would like you to clarify how the tax revenue is going to be used and why. On your side, you kept arguing that raising this tax was going to stifle innovation. But the question is you didn't specify anything about how high the tax would be, what the historical impact of taxes have been on innovation. I'm not sure that there's a lot of historical evidence that increasing taxes by a modest amount reduces innovation. It could be go the other way that it forces firms to be even more innovative or more efficient. So, I'd like you to clarify that a bit. On your side, you never really addressed their point that this could lead to firms um hiding their activities, miscalculating or misindicating who was laid off because of AI or who was laid off because of something else. They argued that there'd be a reduction of our uh transparency about how these firms are operating, and I'm not sure you all ever really addressed that. >> Okay, thank you. >> Firstly, Professor, thank you for your question. Uh with regards to the question, my point to that would be that essentially the revenue which would be collected by the government from companies that replace human labor with AI, it would essentially be going into both programs. Like I had said earlier, the social security program and many other welfare programs in the US are heavily dependent on uh you know, taxes collected by by humans and workers. And that in 2019, 88% of the total funding for the social security program was from human uh wages. Of course, in the case where in the future when AI does lead to mass and unemployment and a severe reduction in taxes collected by the government in the form of income tax, that tax which the government would then apply onto the companies could be used to fill that void which was created by the lack of taxes collected from human labor. Additionally, uh like many other welfare programs that exist, uh the government could support and expand upon them in order to prepare and brace the economy and population with regards to the incoming um market and employment changes. Uh there are systems in place such as unemployment insurance uh and uh unemployment insurance, but these systems will be put under extreme strain in the coming years due to the extreme magnitude of layoffs and job replacements which will be seen in the future. And thus it is crucial that the government does indeed support those programs as well. Just have a safety net for the the workers who end up losing their livelihoods. >> With regards to how layoffs would be would be measured, two methods come to mind. The first would be checking salaries across years with respect to output. If in one year a firm pays an X amount in salary and in the next year salaries decrease but output remains the same or increases, we can deduce that some of it could be as a result of AI integration. The same could be done with wages. If a certain amount of wages were were paid in a year, we can divide by the average weight wages and we can see how many hours workers worked that year. Subsequently, if in the next year hours decrease but output remains the same, we can deduce that AI was was used to make up for that loss in in worker out in worker power and keeping the same output. >> Thank you for your question, Professor. I'd like to tackle the first one you asked on if companies would move overseas or if it would cause inflation. And in short, my answer is I can't really tell you. I think it would cause both. I think that what we can all guarantee is that the tax would not come out of a firm's profits. What we know about big business in America is they're going to force any profits or any I'm sorry, any taxes they can onto their consumers. So I believe that in the short run it would cause inflation as companies hike their prices prices to maintain of a small tax as they begin to cut employees and in the long term as they cut more and more employees if we can assume that's the direction they're going in, then they will move overseas to avoid the tax altogether. So, in short, I believe that the answer is inflation in the short term and offshoring in the long run. >> On the innovation part, I feel like we have actually historically seen taxes take out innovation. We've seen the death of our shipbuilding industry. We've seen the chicken tax, which was implemented on Germany during the Civil War, and then how that led led to loss of domestic market. And then, in a series of change, though not directly proven, how the how how we lost the entire market for like automobile and everything on that front. >> So, again, thank you. Very interesting debate so far. I guess the thing that stands out for me is we're debating an issue about the future, about which we know very little, actually. So, I would like to ask both sides why you would be against the tax or for the tax, given that we are essentially heading into a state of complete uncertainty. That is, we're not sure what jobs will arise, what jobs are going to disappear. So, how does your arguments that you've made for or against fit in that framework of uncertainty? I would again, if you could address that issue. >> Uh thank you, Professor, for the question. Um paraphrasing John Maynard Keynes, uh with regard to the future, we simply do not know. And that is correct. And so, with this tax, we would like to take the somewhat more conservative approach in protecting workers' rights and protecting um their interests instead of um protecting the interests of innovation and of a over a corporation. We'd like to secure the workers' future. We'd like to secure employment and the domestic prosperity. That's the basis for our tax. >> And just to add on to Sebastian's point, AI is by far and I think all of our lifetimes one of if not the most transformative and one of the most you know, effective forms of technology and innovation that we've seen. And our group's stance from the start has been that the so-called AI tax which we want to implement in the future which no one knows where it's headed would not necessarily be to punish innovation. It would be to protect our communities, protect our welfare, our labor markets, and our economies while enjoying that growth at the same time. Yeah, that's >> Thank you for that question. I think it was definitely something that came up while we were preparing for the debate. It's different from I know some of the questions we've had in the past where it's easier to use research that's very concretely based on the issue itself. And kind of the answer that we came up with and I can speak to it first hand. I'm pre-law. I'm hoping hoping to be a lawyer and I think that's one of the fields that people are kind of scared is going to at least be partially taken over by AI and made a lot more difficult as a field. And I think our answer in the general is just we don't know the future. You're absolutely right. So why would we start calling innovation now when we never really have in the past especially as a country like America as cliche as it is to say, we are very focused on innovation and trying to be the best and trying to predict for the future to be as great as it can. So while there are fears of how it could impact us, I think it's better to see it as how we can chase this future and try to be as productive and good as we can be. Thank you. >> Okay, well, I'm Itai Sher from the economics department. I really enjoyed this this debate. So, I sort of have a related, you know, question and a similar premise for for both um sides. So, first for the pro side, so let's let's suppose that automation is a real problem at, you know, as you suggest. But still, why should we have a dedicated AI tax rather than more generally increasing capital taxes? Why specifically an AI tax? What could be accomplished by this that wouldn't be accomplished by increasing the effective capital tax rate? And wouldn't a broader tax be harder to game and avoid these kind of classification problems that we have been talking about? So, why are you proposing this kind of narrower tax? What what would that achieve? So, that's my my question to you. And, you know, to the con side, I think that most economists agree that the um you know, effective tax rate on labor is higher than on capital in the US. And so, in some sense, the government's already putting its thumb on the scale against labor and in in favor of automation. So, if that asymmetry distorts decisions toward automation, wouldn't the right response be to sort of equalize the tax treatment of labor and capital, which moves you, you know, somewhat towards the other side's proposal? So, are you open to that kind of compromise with the other side? >> You have 2 minutes. >> Just three? So, I'll try to keep this short cuz I only have 30 seconds. But like you mentioned, why do we propose AI tax instead of increasing capital taxes. Well, one of my suggestions during my talk was to potentially introduce higher capital taxes to offset for the lost taxes on labor along with other forms of taxes which could include taxes on robots which would be taxing you know, robots which replace human labor. So, the reason why we would prefer sticking to taxing AI in general would be because essentially to protect the labor and employment of the United States. We believe in the American dream that with hard work, you know, we can uplift ourselves, but if we're not employed and if AI replaces all of us, that's not really possible. >> Okay. I think you make a great point and I think we decided we would be willing to compromise on that. Our biggest argument here is against the implementation of a direct AI tax and something that is you know, harming innovation in such a direct way and not the most effective way. We're not against maybe coming up with some better way to even the scales out as you had said between workers and companies. We're more against something that harms businesses without actually improving the lives of any real workers. So, I don't think we'd be against that type of compromise at all. Thank you. >> All right, thank you for the Q&A segment. Sorry, we got like a little cramped on time, but we'll now move into our closing statements. So, the affirmative side can start and you will have 2 minutes. No new arguments at this point, please. Thank you. >> Firstly, I'd like to thank the judges for being here today and the opposing team for giving us a good fight. And we all know that the future is uncertain. No one knows where we're headed, but we do know that the future does look troubling, especially if we don't rectify and work on the system, the existing system which we have in place. We've seen uh you know, my peers have discussed the tax ramifications, the ramifications on the job market, the ramifications on the economy, and what sort of impacts those could have if we don't take action now or in the coming future. There's a famous saying which says that the best time to plant a sapling was 10 years ago, and the second best time is now. And so, it's it's of utmost importance that uh you know, the state uh and the the the US takes urgent steps into uh addressing this issue and protecting the welfare of its citizens while also uh you know, enjoying the fruits of productive growth. Uh without any intervention, like I had said earlier, companies which thrive and mainly rely on AI uh AI automation and AI usage as opposed to human labor would be enjoying the profits while shifting the burden of unemployment and all the byproducts of that onto the welfare system and the general population. So, our proposal of the so-called tax on AI is not to hurt innovation, it's just to protect our our welfare. And Sebastian. >> Oh, that's it. Thank you for your time. >> Okay, thank you, affirmative side, for your closing statements. And now, the negative side, if you could please give your closing statements in 2 minutes, and that will conclude the debate. >> We all share the same goal, an economy that works for workers. The disagreement today is about the tool, and not the goal. Our opponents have offered a tax that sound intuitive, make companies pay for the jobs they eliminate, but intuition isn't policy. We talked about this. We don't really know what the future is going to look like. Policy lives or dies on mainly three questions. Can it be defined? Can it be enforced? And does it actually help the people it names? And can it be quantified? >> [snorts] >> On definition, the US Census Bureau couldn't get firms to agree on what a robot is. On enforcement, the same companies that paid just 4.9% in federal tax last year would simply relabel AI layoff as a restructuring or a shift uh operations that they want to do like off offshore. Uh as Pfizer and Burger King have already demonstrated, and on outcomes, the federal government's own trade adjustment assistance program left displaced workers earning 3,300 less than those who never enrolled. We would tax productivity to fund the program with a negative track record, and I don't think any of us want to do that. >> To conclude, we claim that this AI tax will be ineffective in aiding workers the way that it claims to, because the tax amount was not defined. There are two ways this could end. If a small tax is placed, companies will simply pay the tax and be more willing to pay the small amount than to prevent their ability to innovate. If they do this, this means it will not benefit the employees who have been laid off. If it is a much larger tax, companies will either choose to move overseas or raise prices for consumers. In this case, the nation still fails to get that tax revenue that we are looking for. In our argument, we aren't defending companies. >> Okay, thank you. >> Uh all right, that uh concludes our debate. So, thank you to everyone that participated, and thank you to the judges for your thoughtful questions. Okay. >> [applause] >> Uh yeah, also thank you to everyone that came in in the audience that's watching, and we'll now give judges some time to do deliberations, and we'll go from there. >> So, congratulations to both groups. You've made a lot of good arguments, um and your presentations on the on the whole were very effective. So, congratulations to you both. Uh we all took very extensive notes. You can see here all these different notes that and we all did uh to try to um assess who won. Uh we thought that on the on the uh yes side, uh you made a number of very strong arguments uh that were logically connected. The main arguments were that uh AI may uh reduce tax revenue for the government, that that would have negative impacts on the ability to protect the social safety net, and to fund important government programs. Um number one, number two, that if a lot of people were laid off, that would reduce aggregate demand, and so the government would have to be able to sustain aggregate demand. That was another strong argument. Number three, that uh this um tax was not designed to stop innovation, but to slow it down, that everything was possibly happening too fast. So, those were strong arguments on your side. On your side, you made it we thought a number of strong arguments as well, that it would be difficult to enforce this tax, that uh companies would have an incentive to obscure what they were doing, which would make it more difficult to trace how AI was working, and so forth. Uh you also made a very strong argument about the importance of innovation, not uh only for firms, but also for workers, and that you were concerned about how all this would affect workers. Um and finally, that given the uncertainty, uh it's better to err on the side of promoting innovation, perhaps, than to than doing something that might hold it back. So, those were all strong arguments that we thought you made as well. In the end, weighing all this and all the notes and this and that, we thought the yes side won. So, congratulations to the yes side. Um >> [applause] >> But, nice job everybody.