How are categorical values adjusted for inflation?
We adjust categorical values for inflation using the Pareto method and then re-bracket the values. The goal is to preserve the original brackets as they are and simply reallocate people according to the bracket they would be in if the dollars were adjusted for inflation.
This is done as follows:
- First, we adjust the brackets using the consumer price index (CPI) using the Research Series. For example, $10,000 in 2000 was approximately $13,100 in 2014. We repeat this for the entire table.
While this looks nice, the problem is that the table is no longer compatible with brackets ranges in other years, so we have to move people around to get to the original bracket composition.
For each bracket, we create an array of buckets each at $100 intervals. For example, $0 to $13,100 is the first bracket and it contains 4,000 people. We create 131 buckets of $100.
- Then, using the Pareto distribution, place those 4,000 people into these buckets.
We use Pareto as more people tend to make amounts closer to the bottom part of the bracket rather than to the top, making Pareto an optimal distribution to use in these situations.
We do this for each income bracket, and end up with a long array of inflation-adjusted $100 buckets with people in these buckets. Then, it is a simple procedure to add these up back to our original $0 to $10,000, $10,000 to $15,000, etc.