This README file contains a short discription of the files in this folder and how these files work together to generate the result of the paper

The Data documentation for the Manuscript "A Test of the Permanent Income Hypothesis when Households are less Constrained" by EMMA AISBETT, MARKUS BRUECKNER, RALF STEINHAUSER, AND RHETT WILCOX contains the following files:

"Dictionary.dct" - Contains the original variable list from the proprietary Australian Scanner data from AC Nielsen for the years 2008-2009.

"Merge_data_gen.do" - This file contains the Stata code to merge the ATO (Australian Taxation Office) data and AC Nielsen data.

"tables and regression.do" - This file contains the regression results for all the tables presented in the main paper and the code for generating the two figures. Note that we have sent the code and data for these figures separately with the manuscript to allow for figure editing.

"appendix tables and figures.do" - This file contains the regression results for all the tables presented in the Appendix.

"ABS 8501.0 Retail Trade Australia Table1.xlsx" - This Excel file contains ABS data (Australian Bureau of Statistics) from Table 1 of "8501.0 Retail Trade, Australia," which is used in Figure 1. We used data from January 2008 to December 2009.

We utilized two proprietary data sets: the Australian Retail Scanner data from AC Nielsen (weekly data for 2008-2009) and ATO randomized payout data by postcode (daily).

From the ATO, we received an in-confident data file that includes the following variables: POSTCODE, STATE/TERRITORY, number [of payments], Distribution Daycode [numbered days], Distribution Day [date], Natural Disaster areas [1,0].

From AC Nielsen, we received a file containing the variables listed in "Dictionary.dct."

The AC Nielsen data is merged with the ATO stimulus payment date using the "Merge_data_gen.do" file to generate a merged dataset with AC Nielsen data and ATO payment schedules matched with households by postcode. All necessary variables are generated in this process.

The "tables and regression Do-file" is used to generate the two figures in the manuscript, along with all regression results that appear in the main document. Note that we have sent the code and data for these figures separately to allow for figure editing.

The "appendix tables and figures do-file" is used to generate all regression results for the tables presented in the Appendix.