1. Keep the mandatory start point fixed (so it looks like mandatory is *growing* rather than wobbling everywhere. It’s pacman. It’s going to consume discretionary. ).
2. Make the discretionary part an even more muted color (try grey).
3. The interest should be a part of mandatory — it’s an interesting data point, but it’s not a separate category (purple). If you really need it to be part of the graphic (and I think it should be) try very light labeling and a divider (so it kinda looks like it’s own slice for those who are studying the chart rather than glancing at it).
4. Time. You’re doing a small multiples comparison (three pie charts over a time series). Make the years text bigger.
5. Also call out that the 2040 is a projection based on _____ — it hasn’t happened yet.
6. Consistent units. Don’t switch from billions to trillions.

My first instinct here is, how important is the larger magnitude? You normalized dollar amounts for constant 2009 dollars — great, already ahead of a lot of visualizations.

However, even in constant 2009 dollars, is there a missing “global normalization”? The US budget went from $0.9T to $3.5T in 40 years, a 2.8x increase. However, the global GWP went from $12T to $41T during that same time period, a 2.4x increase (per wikipedia, in 1990 dollars).

So dollars involved are much larger in magnitude, but that’s true in the average global economy as well. I’m not an economist so forgive me if that’s not a common normalization, but my gut instinct is there’s not much of a story in “the dollars involved are much larger in magnitude”, because that’s true of the average country as well.

As for the share-of-the-budget, pie charts are terrible for making that kind of comparison. Stick with Jon’s three %-of-whole stacked columns. There is a story in the makeup of the budget, and pie charts do not get that across.

Now you want to express the idea of two growing expenses “squeezing out” discretionary spending. We can do that by placing discretionary spending on the bottom of the stack, reinforcing the idea of something being crushed. We can reinforce this idea by using two shades of black for the crushing debt, and green for discretionary spending (the hero of this story).

Summarize the point in the graph title and you’re done.

This way we can easily see:
- Discretionary Programs decreasing each year
- Mandatory Programs increasing from 1970 to 2010, and then down for 2040
- Net Interest stable from 1970 to 2010, and then drastically increasing for 2040
- Total Mandatory increasing each year

I also added a horizontal stacked bar for comparing the actual values, for an additional perspective.

This alternative allows you to visually compare:
Relative contributions of each component for a given year (compare blue parts reading down a column);
Component to total for year (comparing the length of blue area to complete rectangle);
Spending for each year (compare lengths of boxes in any row);
Spending for any component for each year (compare blue parts in any row).

The first two tasks are easier because they use an aligned scale. Things could be rotated 90 degrees (you’d end up with lots of whitespace to the left) to make the second two tasks easier instead.

I tried 3 graphs. The first is a simple stacked column chart (excuse the lack of formatting), which is nice because the 100% aggregation is evident. Unfortunately, that approach makes it difficult to compare the specific categories across years; for example, the 57% to the 47% in mandatory spending is a bit hard to see, so I tried another approach where I broke it apart.

The third approach uses the dollar amounts as in Tim’s graph, but uses a stacked column chart. Here, I think, the visual focus switches to the aggregate amount and not the distribution, which I’m guessing is the main point of this slide.