README.txt

This code (Complexity Index.R) assess the sensitivity of the measure "complexity index" used in Caines, hoffmann and Kambourov (2017)

Data and code provided by Caines et al
- 'CHK_occupation_level.dta': Main results of Caines et al.
- 'occ1990dd_names.dta': Occupation titles
- 'occ1990dd_ONET_2.dta': Raw scores of descriptors
- 'ONET_pca_scores_produced.dta': Complexity index generated by code provided by Caines et al.
- 'data_build_occ_level.do': Constructs baseline occupation-level data included the main results
- 'tables_descriptive.do': Constructs Table A1

Data generated and code used by William, Victor, and Paul
- 'descriptor_overview_CMZ.csv': Overview of the descriptors for all occupations (e.g., sum/mean of scores, counts of 0 scores)
- 'factorloadings.csv': Descriptor factor loadings
- 'occ1990dd_ONET_2_CMZ': Cleaned and combined data on complexity index, complexity score, occupation weights, and raw descriptor scores
- 'pseudo_occupations_CMZ.csv': Simulation of pseudo occupations (see Section 3.5)
- 'sensitivity_CMZ.csv': Sensitivity check of the complexity index (see Section 3.4)
- 'Complexity Index.R': Code used to generate the above