Rudolf Kerschbamer and Regine Oexl: The effect of random shocks on reciprocal behavior in dynamic principal-agent settings, Experimental Economics
Date: July 21, 2022
ORCID: 
0000-0001-5596-8746REGINE OEXL
0000-0002-7666-7157RUDOLF KERSCHBAMER

DOI-number of this material: 10.48323/tc8j2-4c904

******************
The material includes
1. The raw data files (18 excel files, with the ending .xls; 3 sessions of observations for each of the 6 treatments)
2. Two do files (ending .do, to work with Stata 15). integrate_files.do constructs the raw_data.dta - file from the raw data files. Analysis.do file uses raw_data.dta to generate a file raw_data1, with which the user can generate all figures and tables in the main body and appendix of the article. The corresponding numbering of figures and table from the article is as indicated in the Analysis.do file.
3. The zTree programs used to collect experimental data (6 files; ending .ztt, one for each treatment).Data for this experiment was collected using Z-tree (Version3.3.12). One additional file (Boxes.jpg) is included since it is used by these .ztt programs.
4. Theoriginal experimental instructions (18 files, ending .doc, in German). A translated version is published in the appendix of the published paper.

*******************


Explanation of the main variables of interest, using the context of the experimental design; these are the variables that are then contained in the raw_data.dta  file. 

There are six treatments, a 3x2 design: 
* cohort 		= 1 == S treatment
  			= 2 == D treatment
 			= 3 == NRG treatment
* treatment 		= 1 == no shock
 			= 2 == shock
In each condition, we run 3 sessions with 24 subjects:
* session 		runs from 1 to 6 within each cohort; first three sessions are `no shock
* subject		runs from 1 to 24 within each session
* id			uniquely identifies each subject
* period 		runs from 1 to 10
* type 			is equal to 1 (principal) or 2 (agent)
* group 		is the group within which subjects were rematched
* age			is the age of the subject
* gender		is the gender of the subject (1 is female)
* brett_num 		is the risk aversion of the subject (runs from 1 to 100)
Variables of interest:
* adjustment_10 	is the adjustment
* average_y 		is the average output
* effort			is the effort
* wage			is the wage 
* desired_effort	is the desired effort
* effort_min_des_effort is the effort minus the desired effort
* shock 		is the shock in the treatments where there is a shock
* output 		is the sum of effort and shock
* Payoff 		is the payoff of the individual
* average_y		is the average output of the agent
