Log of Changes First version April, 19, 2019 This version August, 20, 2023 Serdar Ozkan (serdar.ozkan@toronto.ca) Sergio Salgado (ssalgado@wharton.upenn.edu) Note: You might be getting the following error file xxx.dta cannot be modified or erased; likely cause is read-only directory or file This is generated by Dropbox (or equivalent). When a file is being uploaded to Dropbox, it locks the file and cannot be erased or replaced. If STATA creates and erases files faster than the speed at which Dropbox uploades files, STATA will give you that error. To solve it, simply pause your sync until the code is done. ---------------- August 20, 2023 - Added a new dta with exchange rates from WDI; Added corresponding changes in 0_Initialize.do - Created a wrapper that runs all codes from 0 to 5. The file is called runall.do ---------------- March 24, 2023 - Changes of earnings, Adding variable out_share that is the share of this below min income. - The variable used for the income levels is the same as the one used for log earnings (but not logs). ---------------- March 13, 2023 - All codes updated. Codes 1 to 5 now run directly from code 0_Initialize.do. The insheet code has also being update to account for the changes suggested by Luigi. - The sample creation code 1_GenSample.do was updated to include the log change; Statistics for log changes have been also included in 4_Volatility.do ---------------- April 25, 2022 Updated the following files -myprogs.do to fix a problem in the concentration measures: the shares of the q1 to q4 were not correctly calculated as they were using >= making some individuals be in the same quantiles. -1_Gen_Base_Sample.do. The calculation of arc-percent changes had an error in that it was based on the same sample used for the res earn. In practice, this meant that 0s were not taken into account. This was changed in two ways 1) separated the creation of researn1F and researn5F from the arc versions. The arc version are now calculated using labor income including individuals with income above the min income threshold in a new section between lines 220 and 260; 2) we have changed the arc-changed to the standard measure without adjusting by the within group means. -6_Insheeting_datasets.do. Change in final step of the construction of the Rank csv file. ---------------- March 24, 2022 Updated the 6_insheet code. The main changes are as follows - Added th gender-by-age groups in the concentration statistics - Added Kelley skewness, Crow-Siddiqui, and percentile ratios that were missing for some of the heterogeneity groups - Erases the averaging across years (before called years 9999). This reduces the # of moments that are disclosed February 24, 2022 Updated the insheet code to use standard merge command (1:1). Also changed the part of the outsheet command to avoid wrong prints into excel file (some ranks where printed as 99.00002, instead of 99) January 26, 2022 Added two new global variables to 0_initialize.do. These define the name of the country, the type of education variable available in each country (either numeric or string), and the minimum number of observations required to calculate different moments. Added a new file 6_Insheeting_datasets.do that takes the CSV results and creates a CVS file that will be uploaded to the website. Notice this file must satisfy the disclosure requirements for each team. This disclosure requirement is defined by the global minnumberobs (see 0_initialize.do). Erased 6_Core_Figs.do file. Renamed 6_Paper_Figs.do as the main file for plots, now called 7_Paper_Figs.do Fixed bug in 3_Inequality.do that saved incorrectly the gender as 1/3 indexes Erased the concentration measures calculated as in Gomez (2018) from 3_Inequality.do Added a new section in the inequality code that calculates the autocorrelation of residual earnings. These new results are important and will be added to the website. January 10, 2022 Updated figure codes 6_Core_Figs.do 7_Paper_Figs.do adding global Tcommon = ${yrlast} - 20 + 1 To define the maximum year used for the heterogeneity plots. December 18, 2020 The code 7_Paper_Figs.do has been updated to fix a problem in the calculation of the Kelley Skewness. We have also removed the mobility section from 6_Core_Figs.do. This code is obsolete--all core plots are done in 7_Paper_Figs.do. We will remove 6_Core_Figs.do from the GitHub page in a future update of the code. December 7, 2020 Updates the 1_Gen_Base_Sample.do, 3_Inequality.do, and 5_Mobility.do codes. In 1_Gen_Base_Sample.do we change the definition of alternative permanent income, created a new version of residual earnings controlling for education dummies and age, and fixed a bug for the calculation of arc-percent changes. In 3_Inequality.do we added time series statistics for the new residual earnings measure (with education and age). The code 5_Mobility.do has been streamlined and we changed the selection of the sample used to calculate the mobility plots. We also added an entire new file, 7_Paper_Figs.do, that produces the main core figures for the draft. See Figures_Paper.pdf for additional details and an example of each figure using the Norwegian data. November 6, 2020 Fixes the volatility code. The in previous version, the moments conditional on permanent earnings were calculated after dropping all observations that did not have researnXF and arcearnXF with X \in {1 5}. August 6, 2020 Fixes a bug in the construction of permalt income in 1_Gen_Sample.do (not it is constructed between years $yrfirst+2 and $yrlast and age > begin_age + 2 Fixed also a small big in the transition matrices in 5_Mobility.do July 21, 2020 Relative to March 03, 2020, this version updated all do-files, from sample creation to the generation of the plots. March 03, 2020 Updated 1_Gen_Sample.do, 6_Core_Figs.do, and (very minor) changes in myplots.do and 3_Inequality.do. Below the important changes - On 1_Gen line 158 that was incorrectly dropping observations with little earnings dropping all observations with 0 earnings. Dropped that line. We also added a condition on arc-percent that only kicks in if labor earnings in t and in t+k are below the min income. - On 6_Core, adjusted some of the figures to put the figures of 5-year changes centered in the moving window. That is, if the moment is calculated as t+5, the plot is centered in t+2. Also, added a new version of figure 5a to have two axis. January, 18, 2019 Updated the codes to the version 2.0; This new version of the code contains several major changes relative to the original version of April 2019. Among others - Adds a new set of results for the arc-percent change in income - Modifies the change income growth measure to allow for declines below the min value in t+k. - Adds new features to the plots such as recession bars and differential color schemes. - Adds several new plots: density plots, cohort plots, etc. August, 14, 2019 We have modified 1_Gen_Sample.do so now it records the gender for which the profiles are calculated. These changes are made between lines 210 and 236 July, 05, 2019 We have modified some of the calculations in the gen_base and mobility codes. In the first, we have added bys male: egen avgall = mean(totearn) gen permearnalt`yr' = avgall*totearn/avg // This is because we want to control for age effects In lines 388 to have permearnalt in the correct scale. This does not change the results. The mobility code has more changes. In particular, we have modified the transition matrix calculation to account for 0. Individuals with 0 permanent earnings are now grouped in one category only, whereas the rest of individuals (with positive permearnalt earnings, are separated in 10 groups, to a total of 11 rows in the transition matrix). See lines 240 to 266. We have also saved some summary stats within each cell (see line 248 for instance). April, 19, 2019 The folder contains the first version of the code for the Global Income Dynamics Database project. The code was developed in Stata 13 by Serdar Ozkan and Sergio Salgado. The original set of do files is the following 0_Initialize 1_Gen_Base_Sample.do 2_DescriptiveStats 3_Inequality 4_Volatility 5_Mobility 6_Core_Figs GenData myplots myprogs See the file Code_Guidelines_April2019.pdf for additional details.
higginsbrian/QE-MasterCode
This folder constains the latest version of the QE Master Code developed by Ozkan and Salgado
Stata