Main content



Loading wiki pages...

Wiki Version:
### Overview This collection of Stata files is used to produce datasets, analyses, tables, and figures for the paper "Rising Intragenerational Mobility in the United States, 1969-2011". ### Scripts There are two types of scripts, the main script and sub-scripts #### Main - _main.do_ - sets project paths and calls on all sub-scripts and programs, in the appropriate order, to produce the main analysis, including all the main tables and figures featured in the paper. #### Sub-scripts These scripts process the raw PSID and Census Data - _igmnet_crdata_census_counts.do_ : creates (imputed) yearly counts of individuals in various age x gender x race x education groups for the purpose of re-weighting the PSID sample. - _: draws on datasets raw, downloaded PSID data to create person-year work histories for heads and wives in the PSID. ### Programs The rest of the data processing and analysis steps are implemented as Stata programs. There are programs for final data preparation, programs for doing analysis steps, and more general "utility" programs that help along the way. Note that all of the required programs are loaded into Stata's memory in the _main.do_ file. #### Data - _make_switch_ : using long-format panel data, creates a switching variable, in this case a variable identifying cases of occupational mobility at various levels of aggregation - _make_weights_ : weights data in memory (PSID data, in this case) based on marginal counts contained in an external file (Census derived). - _make_combine_categories_ : collapses some (small) occupational categories together and generates the period variable. - _make_mobility_table_ : collapses the sample to create a rectangularized mobility table of origins and destination by period - _make_table_covariates_ : generates the explanatory dummy variables for use in overlapping persistence log-linear models. - _make_table_increment_: increments the table based on either a model or a fixed amount. #### Analysis - _analyze_mobility_logit_ : takes person-biennial data, estimates logit models, and generates plots of trends in gross mobility at various levels of analysis. - _analyze_psid_mobility_tables_ : estimate models #### Utilities - _fromprevresults_ : a program from starting estimates using coefficients stored in Stata's memory or on disk - _nestest_ : a program from estimating a sequence of nested models, layering sets of explanatory variables, including support for factor variable - _spotodds_ : a helper program for calculating odds ratios in a two-way mobility table - _datestring_ : returns the date, time, and date+time as strings to be incorporated into, for example, log file names - _gof_ : calculate fit statistics for log-linear models estimated using glm. - _mstdizel_ : iterative proportional fitting command, this is a modification of mstdize that allows the two-way