Michael D. Hurd, Maarten van Rooij, and Joachim Winter, "Stock Market Expectations of Dutch Households", Journal of Applied Econometrics, Vol. 26, No. 3, 2011, pp. 416-436. The data used in this paper come from two surveys conducted in the Dutch CentERpanel in April 2004 and April 2006. The fieldwork procedures are described in the paper. The data distribution contains the following files: Questionnaires ============== * StockExp_questionnaires.pdf This is a PDF file that contains the two questionnaires that were fielded in the Dutch CentERpanel in April 2004 and April 2006. Data file formats ================= [*.dta] These files are in Stata 11 format, with full variable and value labels. These can be reviewed by using the "describe" and "codebook" commands in Stata. [*.asc] These are the same data in ASCII fixed-width format with one observation per line. [*.txt] These are the corresponding codebooks produced by Stata. Raw data files ============== * StockExp_2004_raw.dta * StockExp_2004_raw.asc * StockExp_2004_raw.txt * StockExp_2006_raw.dta * StockExp_2006_raw.asc * StockExp_2006_raw.txt These are the raw data files as provided by CentERdata. The only changes relative to the files we obtained are: - In 2004, we obtained two separate SPSS files for weeks 17 and 21. We have combined these two datafiles into one Stata datafile. - We have added labels for a few variables that were unlabeled in the original files we obtained. - We have dropped the free-text answers to the feedback questions that were asked at the end of each survey. Note that the raw datafiles contain some background demographic variables provided by CentERdata that were not asked in our questionnaires. These are also labeled. Analysis data file ================== * StockExp_JAE.dta * StockExp_JAE.asc * StockExp_JAE.txt This is the data file that contains the processed data as analyzed and reported in the paper. Relative to the raw data files, the main changes are: - Added parametric and nonparametric estimates of the mean and variance of the individual expected return distribution. - Dropped a few observations with inconsistent responses, as described in the paper. - This dataset is organized as a "wide" panel, i.e., each row contains a person observation. Periods are identified by variable name suffix (_04 for 2004, _06 for 2006). All the *.dta files are zipped in the file dta-files.zip. These are binary files. All *.asc and *.txt files are zipped in the file text-files.zip. Since the *.asc and *.txt files are ASCII files in DOS format, Unix/Linux users should use "unzip -a". Please send questions to Joachim Winter (winter [AT] lmu.de).