By: Rajan Sehmi
"The Chancellor of the Exchequer is a man whose duties make him more or less of a taxing machine. He is entrusted with a certain amount of misery which it is his duty to distribute as fairly as he can." -Chancellor Lowe 1870
With all that 2020 hurled at us, Sunak's big plans for the UK economy had to be thrown out the window. Recovery policy after the next, including furlough and Eat Out to Help Out has meant his job has been demanding. All this investment has widened the deficit, just before we were about to balance the books. As a result of this increase in spending, Sunak had to devise a new tax policy to recuperate funds. And so, in the Budget 2021, Sunak announced a corporation tax rise from 19% to 25% for large companies in 2023, while government spending will be cut by a further £4 bn a year.
It is likely that the Treasury came to this conclusion through many mathematic models. However, imagine a system that simulates billions of iterations of different tax systems to find an optimal solution in just a few minutes.
Enter the AI Economist.
This Artificial Intelligence system seeks to address the goal of any economy: to find a balance between productivity and equality. 'Productivity' allows the economy as a whole to benefit through higher rates of GDP per capita. However, 'equality' addresses income inequality, ensuring that everyone has fair access to necessary resources and human rights. Too often, an economy may be hyper-productive, squeezing the most out of its workers, improving the economy on a face value GDP level, but remaining to keep its lowest-income members poor, the very individuals who spurred on this growth.
"Economic inequality is accelerating globally and is a growing concern due to its negative impact on economic opportunity, health and social welfare. Taxes are important tools for governments to reduce inequality. However, finding a tax policy that optimizes equality along with productivity is an unsolved problem. The AI Economist brings reinforcement learning (RL) to tax policy design for the first time to provide a purely simulation and data-driven solution. -Salesforce
The simulation uses a two-dimensional world to simulate an economy, though at a much more simplified level. There are only four types of agent, ranging from high to low skilled. Higher skilled workers earn income by building houses and then pay income tax to be redistributed amongst the other agents. Throughout these simulations, hoards of data are collected, thus we can use different tax models to monitor welfare and how agents react.
AI Economist also takes into account emotion. In high tax-high distribution models, workers may feel disincentivised to work, as they feel comfortable with increased welfare benefits. In the UK, welfare benefits were loathed in the 2010 Cameron government, with Chancellor George Osborne cutting £7 bn in the October 2010 spending review, £83bn of cuts in public spending until 2014.
"People who think it is a lifestyle to sit on out-of-work benefits … that lifestyle choice is going to come to an end. The money will not be there for that lifestyle choice." -George Osborne 2010
Perhaps welfare benefits do stunt productivity, but they also help those who need it most. This is where the AI Economist comes in to find an optimal balance.
As seen by the results, a free-market pushed productivity to its very limits. Great for the economic numbers, albeit at a massive loss to equality, ranking the lowest of models tested. The AI Economist model can be said to have reached a happy fiscal medium. But can this be applied to Sunak and the future of the British economy?
We are nearly there.
Although the AI Economist simulation could be regarded as a breakthrough, we are still many years away until we can use Artificial Intelligence and machine learning to formulate economic policy. Pragmatism when adjusting fiscal and monetary policy is essential. One wrong move and we may see the whole system collapsing. Thankfully, as technology gets better and better, we will undoubtedly see AI playing a more active role in our economy in the future.
For more information, you can watch these videos:
https://www.youtube.com/watch?v=4iQUcGyQhdA
https://www.youtube.com/watch?v=Sr2ga3BBMTc
https://www.youtube.com/watch?v=F5aaXrIMWyU
Or read the original paper here:
https://blog.einstein.ai/the-ai-economist/
Comments