Recursively add missing dataframe columns for arbitrary number of dataframes

Question

I wish to add columns into a variable number of PySpark dataframes recursively until they all share the same columns. (The added columns will be populated with null values).