Fabio Canova, "Should We Trust Cross Sectional Multiplier Estimates",
Journal of Applied Econometrics, forthcoming.
All files are stored in the file fc-files.zip. It contains a mix of
ASCII files (in DOS format) and binary files.
MATLAB CODES
1) The file for the Monte Carlo exercise is: multiplier_note_prgmc.m
2) The file to compute CV critical values is: critvalue_hom_test.m.
3) The file to compute multipliers is: multiplier_note_prg.m. This
file reads stdata.mat and usdata.mat which contains state and
aggregate US data.
DATA
The data are provided both in .mat and .txt (ASCII) formats. The
sample runs from 1980:1-2017:1
The ASCII format data are organized as follows
1) US State data
• Real Income per capita ( ypcdata.txt)
• Real Government expenditure per capita (gpcdata.txt)
• Real Federal transfers per capita (fpcdata.txt)
• Real Taxes per capita (tpcdata.txt)
Each of the four files has 38 time periods (columns) and 51 rows
(states) (it is a 38x51 matrix).
2) Aggregate US data ( uspcdata.txt)
• Real income per capita
• Real government expenditure per- capita
• Real Taxes per capita
Each series has 38 time periods ( it is a 38x1 vector)
3) Additional data ( additionaldata.txt)
• CPI inflation
• Oil price
• T-bill rates
Each series has 38 time periods ( it is a 38 x 1 vector)