Are you tired of static visualizations that hide the real story? Compositional variables—those parts adding up to a whole—are common in political science, but they often mask important dynamics. This article provides fresh strategies for analyzing them using time-series cross-sectional (TSCS) methods.

We extend existing approaches beyond single-time snapshots and discuss how researchers can incorporate contextual factors and spatial relationships into TSCS models of compositional change across geographic areas. Our methodology helps identify meaningful patterns that standard techniques might miss.

To demonstrate our approach, we analyze budgetary expenditures in all 50 US states from 2010 to 2020. This case study shows how seemingly simple budget categories can reveal complex interplays between state policies and national trends when viewed through the lens of dynamic visualization methods.

Here's what you need to know:

• Composition analysis challenges

* Static visualizations often oversimplify compositional data

* Standard time-series approaches struggle with interdependent parts

• Our approach

* Builds on recent advances in TSCS modeling

* Explicitly addresses spatial relationships alongside temporal trends

• Budget example findings

* Education spending increased during economic downturns across most states 📈

* Spending disparities between urban and rural states persisted despite overall increases 💰