Week 1 – The Base

Welcome to week 1!
I am excited to announce that the base simulator for our taxation simulations has (more-or-less) been created, with significant help from various publicly available GitHub projects. It’s by no means perfect, but it doesn’t have to be–it’s workable for now.

While significant progress has been made, we still have some crucial steps to take to ensure that our simulations are accurate and provide actual value. One of the most critical steps is determining the logistics of the taxes (e.g. rates, frequency) we will be simulating. More-likely-than not, this will be taken directly from OECD averages in real life for income taxes, seeing as that methodology was used by Oskarssen et. al and makes logical sense. However, it’s more complicated for LVTs and Wealth Taxes–are they event-triggered, or every (few?) turn(s)? What’s an appropriate rate? If they’re regular, how frequent should they be? What should the cutoff be to be eligible for the wealth tax? At this stage, I don’t have the answers, and I need to do more research before I’m prepared to definitively settle anything.

Another crucial detail is determining how to calculate GDP and Gini-coefficients (a measure of inequality; essentially, we will be comparing how much inequality there is pre- and post-tax to see how effective it is) within the context of the game. We need to find a way to calculate GDP accurately to observe its changes due to different types of taxation. In Monopoly, GDP can probably be considered the sum of the values of the properties, assets, and money held by the players (not the bank), and the Gini-coefficient problem has already been solved in another paper. In other words, this part shouldn’t be too hard to think-through or implement, but it’s another task on the to-do list for me to blog about.

Lastly, we need to adequately simulate player behavior and decision-making. This is a crossroads for my project–I need to be realistic with my practical abilities given the time-constraints (I can’t just make an A.I. that perfectly replicates human behavior!), so significant generalizations will likely be made. This is integral for the applicability of my eventual results–for example, income taxes may discourage real players from purchasing more properties or taking risks, but my simulations may miss that if the player behavior is too oversimplified.

Overall, I’ve got a lot of work ahead of me, but I’m excited and in a good place. Hopefully next-week I can at least report results on income taxes, and I’ll have more detailed answers to the rest.