DOCUMENTATION FOR TOTAL ANNUAL INDIVIDUAL INCOME FROM THE MARCH CPS This text file contains an explanation of the individual annual income files (income_smooth and income_raw). These data were calculated using the unicon-create variable "income" which measures all personal income from all sources in the previous year. The data are organized in state-year format. There are 40 observations (1963-2002) for each state, plus the District of Columbia and an observation with the total for the US (= 2,080 records). All earnings data have been adjusted to 2002 dollars using the CPI deflator for all urban consumers. The columns of the text file are separated by spaces and do not contain variable names; in these text-format files, state names are surrounded by quotation marks and two-word names contain spaces (e.g.,"New York"). The variables in the data appear in the following order: 1. statename 2. stcode - Two-letter state postal abbreviation 3. year 4. p10income - 10th percentile of total annual individual income (all) 5. p20income - 20th percentile of total annual individual income (all) 6. p30income - 30th percentile of total annual individual income (all) 7. p40income - 40th percentile of total annual individual income (all) 8. p50income - 50th percentile (median) of total annual individual income (all) 9. p60income - 60th percentile of total annual individual income (all) 10. p70income - 70th percentile of total annual individual income (all) 11. p80income - 80th percentile of total annual individual income (all) 12. p90income - 90th percentile of total annual individual income (all) 13. meanincome - average total annual individual income (all) 14. minincome - Minimum total annual individual income (all) 15. maxincome - Maximum total annual individual income (all) 16. sdincome - Standard deviation of total annual individual income (all) 17. ineqincome - Ratio of the 90th percentile of total annual individual income to the 10th percentile of total annual individual income (all) 18. gini_income - Gini coefficient for total annual individual income (all) 19. n_income - Number of non-missing values (all) 20. nwgt_income - Weighted number of non-missing values (all) 21. men_p10income - 10th percentile of total annual individual income (men) 22. men_p20income - 20th percentile of total annual individual income (men) 23. men_p30income - 30th percentile of total annual individual income (men) 24. men_p40income - 40th percentile of total annual individual income (men) 25. men_p50income - 50th percentile (median) of total annual individual income (men) 26. men_p60income - 60th percentile of total annual individual income (men) 27. men_p70income - 70th percentile of total annual individual income (men) 28. men_p80income - 80th percentile of total annual individual income (men) 29. men_p90income - 90th percentile of total annual individual income (men) 30. men_meanincome - average total annual individual income (men) 31. men_minincome - Minimum total annual individual income (men) 32. men_maxincome - Maximum total annual individual income (men) 33. men_sdincome - Standard deviation of total annual individual income (men) 34. men_ineqincome - Ratio of the 90th percentile of total annual individual income to the 10th percentile of total annual individual income (men) 35. men_gini_income - Gini coefficient for total annual individual income (men) 36. men_n_income - Number of non-missing values (men) 37. men_nwgt_income - Weighted number of non-missing values (men) 38. wom_p10income - 10th percentile of total annual individual income (women) 39. wom_p20income - 20th percentile of total annual individual income (women) 40. wom_p30income - 30th percentile of total annual individual income (women) 41. wom_p40income - 40th percentile of total annual individual income (women) 42. wom_p50income - 50th percentile (median) of total annual individual income (women) 43. wom_p60income - 60th percentile of total annual individual income (women) 44. wom_p70income - 70th percentile of total annual individual income (women) 45. wom_p80income - 80th percentile of total annual individual income (women) 46. wom_p90income - 90th percentile of total annual individual income (women) 47. wom_meanincome - average total annual individual income (women) 48. wom_minincome - Minimum total annual individual income (women) 49. wom_maxincome - Maximum total annual individual income (women) 50. wom_sdincome - Standard deviation of total annual individual income (women) 51. wom_ineqincome - Ratio of the 90th percentile of total annual individual income to the 10th percentile of total annual individual income (women) 52. wom_gini_income - Gini coefficient for total annual individual income (women) 53. wom_n_income - Number of non-missing values (women) 54. wom_nwgt_income - Weighted number of non-missing values (women) 55. cpideflator - Deflator used to adjust values for inflation (this is provided so that the user may calculate the unadjusted values). This deflator uses the consumer price index and is uniform for all states. 56. statecpi - This is a state-specific deflator from 1963 to 2000, except for Alaska and Hawaii. A description of how this deflator was created can be found in Berry, William D., Richard Fording and Russell Hanson, "An annual cost of living index for the American states, 1960-1995", Journal of Politics, 62:550-567. The updated deflator files can be found at http://www.icpsr.umich.edu:8080/ICPSR-STUDY/01275.xml UNIVERSE: The March CPS universe for total individual annual income includes private and government, as well as self-employed respondents. We excluded respondents who reported working 0 weeks in the previous year and those reporting income of $1 or less. For more information, see unicon documentation below or visit http://www.unicon.com. WEIGHTING: We used the "wgt" variable provided by unicon, divided by 100, to produce weighted estimates. IMPUTATION: Since the CPS grouped some states together from 1968-1976, the values for most individual states were imputed for these years. For a description of the methodology used for this imputation, see Appendix B of our report at: http://www.princeton.edu/~joshg/inequality/IneqDataReport.pdf. Maximum and minimum values, as well as N's, were not imputed from 1968-1976. They are recorded as missing values for those years. *********************************************************************************** UNICON DOCUMENTATION: income Person's total income _income Unicon recode - Person's total income Original location, length, and name of variable: 62 63-67 68-75 76-79 80-88 88B-99 P157 P171 P61 P247 P248 P440 6 10 6 7 7 8 TOTINC TOTINC TOTINC PINCTOT PINCTOT PTOTVAL Topic: income Related variable: aincome - allocation flag tpcdto - topcode flag Code: 62-67 68-75 76-80 81 82-84 85-88 88B-95 96-99 Btm Val -N -09999 -150000 -29997 -29997 -150000 -389961 -389961 Top Val +N +50000 500000 425996 654999 +N 599994 varies NIU 99999 Rec:NIU . Note: 1962 Universe - adults in the income sample (14<=age & mis^=4 and mis^=8) Construction - incwag + incse + incfrm + incuer 1963-1967 Universe - adults (14<=age) Construction - incwag + incse + incfrm + incuer 1968-1975 Universe - adults (1<=popstat<=2) Construction - incwag + incse + incfrm + incss + incint + incpa + incomp + incoth 1976-1999 Universe - adults (1<=popstat<=2) Construction - incern + incuer Although not specifically stated, our tests show that the variables are individually topcoded and then added. CAUTION - The 'select' and 'unselect' options should be used carefully with the UNICON RECODE, as selection occurs after recoding of the variable. (i.e., The options use the recoded values rather than the original raw values.) ** Survey Questionnaire Source Items - income ** 1962-1967 Derived from sum of incwag, incse, incfrm, incuer 1968-1975 Derived from incwag, incse, incfrm, incss, incint, incpa, incomp, incoth 1976-1999 Derived from incern + incuer