COR648 8/25/97 This COR consists of four Excel spreadsheet files, and this document which describes each spreadsheet (5 files total). With these five documents, one can map 1960, 1980, and 1990 Census 3-digit occupation categories into the 1970 3-digit classification system. This mapping was used in Jennifer Sheridan's 1996 master's thesis (and CDE working paper 97-05 available at http://www.ssc.wisc.edu/cde/cdewp/97-05.pdf), and is used in the calculation of the percentage of workers who are female in the 'FEMJOBS' dataset available through the WLS webpage. Mapping 1960 Categories to 1970 Categories--COR648A.XLS This spreadsheet uses information from the 1970 Census, Volume 1, Part 1, Pages 1-718 - 1-724, Table 221, and also from Technical Paper 26, "1970 Occupation and Industry Classification Systems in Terms of Their 1960 Occupation and Industry Elements." Column 1 (A) - 1970 occupation code Column 2 (B) - Number of males in 1960 Census who would have been in 1970 occupation category (from Tech. Paper 26) Column 3 (C) - Number of females in 1960 Census who would have been in 1970 occupation category (from Tech. Paper 26) Column 4 (D) - Number of males in 1970 Census in 1970 occupation category (from 1970 Census, Table 221) Column 5 (E) - Number of females in 1970 Census in 1970 occupation category (from 1970 Census, Table 221) Column 6 (F) - Blank Column 7 (G) - Percent of all persons in 1960 who would have been in the 1970 occupation category who were female (C/(B+C)) Column 8 (H) - Percent of all persons in 1970 in the occupation category who were female (E/(D+E)) Column 9 (I) - Blank Column 10 (J) - Percentage of all males in Experienced Civilian Labor Force in 1960 who would have been in the 1970 occupation category (B/(total number males in 1960 l.f.)) Column 11 (K) - Percentage of all females in Experienced Civilian Labor Force in 1960 who would have been in the 1970 occupation category (C/(total number females in 1960 l.f.)) Column 12 (L) - J - K; used to calculate Index of Dissimilarity for 1960 Column 13 (M) - Percentage of all males in 1970 Census who worked in 1970 occupation category (D/(total number males in 1970 l.f.)) Column 14 (N) - Percentage of all females in 1970 Census who worked in 1970 occupation category (E/(total number females in 1970 l.f.)) Column 15 (O) - M-N; used to calculate Index of Dissimilarity for 1970 Mapping 1980 Categories to 1970 Categories--COR648B.XLS This spreadsheet uses information from the 1980 Census, Volume 1, Pages 1-166 - 1-175, Table 276, and also from Technical Paper 59, "The Relationship Between the 1970 and 1980 Industry and Occupation Classifications." The mapping of 1980 Census records to 1970 3-digit categories cannot be done directly (as it is for 1960 to 1970), because the Census Bureau chose a sample of 1970 Census records and recoded them to 1980 Census standards; to map 1980 census data back to 1970 categories, we would need to do the opposite (that is, take a sample of 1980 Census records and recode them to 1970 3-digit occupation categories). Thus, this mapping is an approximation based on 1970 data. Note: Row numbers 1 through 984 correspond to 1970 3-digit occupation categories Column 1 (A) - 1980 Census 3-digit occupation category Column 2 (B) - Number of males in the 1980 Census, who had occupations in the 1980 occupation category Column 3 (C) - Number of females in the 1980 Census, who had occupations in the 1980 occupation category Column 4 (D) - Total number of 1980 Census respondents who had occupations in the 1980 occupation category (B+C) Column 5 (E) - A summation of males from 1980 component categories who would have been in the 1970 categories. Formulas are derived from Table 2 in Tech. Paper 59. Example: Row 3 (column E) of COR648B.XLS corresponds to males in the 1970 occupation category "003-Computer programmers". 8.62% of 1970 Census male respondents who ended up in the 1980 category "064-Computer systems analysts and scientists" when recoded to the 1980 standards came from the category "003- Computer programmers" in 1970. Similarly, 100.00% of those 1970 male respondents recoded into the 1980 category "229- Computer programmers" came from the 1970 category "003". Taking 8.62% of the total number of males in 1980 category "064" (cell B64), plus 100% of the total number of males in the 1980 category "229" (cell B229) gives the number of 1980 respondents who would have been in category "003" in 1970. All occurances of 1980 categories scattered throughout the 1970 rows sum to 100%, but within a 1970 row, the sum of percentages can be more than 100%. Column 6 (F) - A summation of females from 1980 component categories who would have been in the 1970 categories. Formulas are derived from Table 2 in Tech. Paper 59. See example for Column 5 (E) above. Column 7 (G) - A summation of total respondents from 1980 component categories who would have been in the 1970 categories. Formulas are derived from Table 2 in Tech. Paper 59. See example for Column 5 (E) above. Totals in Columns 5, 6 and 7 are used to calculate the percentage of respondents in 1980 who would have been coded into a 1970 occupation category who are female. See COR648D.XLS below. Mapping 1990 Categories to 1970 Categories--COR648C.XLS This spreadsheet uses information from the 1990 Census, EEO Supplementary Report (CP-S-1-1), Pages 1 - 11, Table 1; from "Changes Between 1980 and 1990 Occupational Classification System" Census report; and also from the previous worksheet (COR648B.XLS). The 1990 occupation categories were mapped to the 1980 categories, then the same percentages used in the 1980-to-1970 mapping were used to map from 1990 to 1970. COR648C.XLS consists of two worksheets. Worksheet #2 maps the 1990 categories to the 1980 categories. Columns 1 and 2 are taken directly from the COR648B.XLS worksheet; Column 3 is blank; Columns 4 and 5 are taken from the EEO Supplementary Report and, categories which are different from the 1980 categories are combined directly (according to the "Changes Between 1980 and 1990 Occupational Classification System" Census report). These population figures from the 1990 Census, in terms of the 1980 categories, are then used as input to Worksheet #1: Note: Row numbers 1 through 984 correspond to 1970 3-digit occupation categories Column 1 (A) - 1980 Census 3-digit occupation category Column 2 (B) - Number of males in the 1990 Census, who had occupations in the 1980 occupation category (see Worksheet #2) Column 3 (C) - Number of females in the 1990 Census, who had occupations in the 1980 occupation category (see Worksheet #2) Column 4 (D) - Total number of 1990 Census respondents who had occupations in the 1980 occupation category (B+C) Column 5 (E) - A summation of males from 1980 component categories (1990 data) who would have been in the 1970 categories. Formulas are derived from Table 2 in Tech. Paper 59. See example for Column 5 (E) in COR648B.XLS Column 6 (F) - A summation of females from 1980 component categories (1990 data) who would have been in the 1970 categories. Formulas are derived from Table 2 in Tech. Paper 59. See example for Column 5 (E) in COR648B.XLS Column 7 (G) - A summation of total respondents from 1980 component categories (1990 data) who would have been in the 1970 categories. Formulas are derived from Table 2 in Tech. Paper 59. See example for Column 5 (E) in COR648B.XLS Totals in Columns 5, 6 and 7 are used to calculate the percentage of respondents in 1990 who would have been coded into a 1970 occupation category who are female. See COR648D.XLS below. All Population Figures, 1960 - 1990 in Terms of the 1970 Occupation Categories - COR648D.XLS Columns are labelled and are self-explanatory.