Skip to content

Latest commit

 

History

History
41 lines (33 loc) · 3.74 KB

README.md

File metadata and controls

41 lines (33 loc) · 3.74 KB

Equitable Taxation Model

Problem Statement: Inequitable taxation increases the burden on families that need help.

Primarily, there are three types of tax policies: progressive, regressive, and proportional, and most countries around the world adopt a progressive taxation system. A percentage of income forms the tax, the quantum of which may vary according to income thresholds and may also include a fixed component. The general principle is that those who earn more pay more. But this leads to economic inequality globally and leads to various concerns as it impacts health and social welfare. Finding an optimal tax policy that optimizes inequality and productivity is an unsolved problem.

Though this seems fair on the outset, such a system does not account for the differences in spending needs or habits of various households. Due to ailments, families with more dependents or higher medical costs would incur higher subsistence costs. However, they would be charged the same amount of tax as another household that earns the same amount but has fewer subsistence costs. It discourages individuals from working, which leads to lower productivity.

There are mechanisms for tax relief for cases where you have more dependents than other households, medical costs, donations to charity, and many more. However, the scope of these tax reliefs is generally limited and would not cover other circumstances which may be just as valid. Even when one is eligible for tax relief, obtaining the relief is generally not automated and tedious. It also opens the possibility of fraudulent claims or declarations.

We aim to address these issues by augmenting the existing taxation system with a predictive model. It would allow us to predict the expenditure of a particular household for the past year and automatically adjust the tax levied on them.

Conclusion

The Equitable Tax Model that we have created effectively addresses the issues mentioned in the problem statement. Based on the pre-collected information about each household, governments can accurately predict their necessary and unnecessary expenditure. The model has an R-score of 0.929 in predicting necessary expenses and 0.877 in predicting unnecessary expenses.

We can achieve more equitable taxation by using a formula that uses a ratio of unnecessary/necessary expenses to adjust the current tax. Naturally, having higher necessary expenses would therefore lead to lower taxes and vice versa.

The model we have created does not replace the existing tax system but instead augments it. The only modification to the actual tax amount is through the abovementioned ratio. Therefore, it would be easier for governments to understand and implement.

Our model also uses common census data to make its predictions, like the number of dependents/education levels. All countries usually collect this data about their citizens. As such, any country could apply our model.

There are a few enhancements that can still be made to the model. Before implementation, more expert opinions could be sought to advise on the features to select/additional features to gather, likely to have the highest impact on expenses. The tax adjustment ratio could also be customized to increase/decrease weightage on additional spending. It would help to further refine the proposed tax.

In conclusion, our model is based on sound socio-economic principles and addresses the problem statement effectively.

Results

  • Adjusted Vs Current Tax for unnecessary expenditure

alt text

  • Distribution of tax amount

alt text