--- Oct CPS original +++ Oct CPS modified by Freddie&Jieyi @@ -7,41 +7,42 @@ from the October CPS . 2. CREATE AN ANALYSIS FILE age_profile_enrolled_students_cps_oct COLLAPSED BY SURVEY YEAR AND AGE GROUP TO GENERATE TABLES AND FIGURES DESCRIBING AGGREGATE TRENDS IN SCHOOL ATTENDANCE 3. CREATE AN ANALYSIS FILE enrollment_analysis_cps_oct COLLAPSED BY BIRTH COHORT AND 5-year age group TO DESCRIBE AGGREGATE TRENDS IN SCHOOL ATTENDANCE 4. CREATE AN ANALYSIS FILE age_profile_enrolled_CA_students_cps_oct that includes only respondents residing in the state of California to benchmark age enrollment figures from the CSU AND CCCC systems with survey data from the October CPS. ******NOTE: You will need to change the directory to your own working directory.***/ /* 1. CREATES AN ANALYSIS FILE main_analysis_cps_october_76_19 */ /***********************************************************************************************************/ /***********************************************************************************************************/ /***************************** Read in October Education Supplement CPS Data from Multiple Years ******************/ /***********************************************************************************************************/ /***********************************************************************************************************/ # delimit cr clear all -cd "C:\Users\cmslo\Box Sync\race project" +cd "/Users/chenjieyi/Downloads" +*cd "C:\Users\cmslo\Box Sync\race project" *cd "D:\Box Sync\race project" set more off clear quietly infix /// int year 1-4 /// long serial 5-9 /// byte month 10-11 /// double hwtfinl 12-21 /// double cpsid 22-35 /// byte asecflag 36-36 /// byte hflag 37-37 /// double asecwth 38-47 /// byte region 48-49 /// byte statefip 50-51 /// byte pernum 52-53 /// double wtfinl 54-67 /// double cpsidp 68-81 /// @@ -3489,271 +3490,349 @@ label define edstudy_lbl 18 `"No major"', add label define edstudy_lbl 19 `"Undecided"', add label define edstudy_lbl 96 `"Refusal"', add label define edstudy_lbl 97 `"Don't know"', add label define edstudy_lbl 98 `"No response"', add label define edstudy_lbl 99 `"Not in Universe"', add label values edstudy edstudy_lbl label define qedschus_lbl 1 `"Allocated"' label define qedschus_lbl 0 `"Not Allocated"', add label values qedschus qedschus_lbl label define qedschusyrs_lbl 0 `"Not Allocated"' label define qedschusyrs_lbl 1 `"Allocated"', add label values qedschusyrs qedschusyrs_lbl label define qedstudy_lbl 1 `"Allocated"' label define qedstudy_lbl 0 `"Not Allocated"', add label values qedstudy qedstudy_lbl - +/*Drop 2020 observations to keep away from covid shock*/ drop if year == 2020 + keep if month == 10 -keep if bpl == 09900 /*keep US-born people only */ -/*keep people who live in 50 states: need to go back and grab state and region vars*/ -keep if statefip >= 1 & statefip <= 56 - - - /*Data Imputation Flags*/ + +/*Keep US-born people only*/ +keep if bpl == 09900 + +/*Drop if U.S. state of residence is missing*/ +keep if statefip >= 1 & statefip <= 56 + +/*Data Imputation Flags: +For CPS supplements other than the ASEC, values are imputed for all years 1982 forward. +Keep track of imputation status. +*/ #delimit cr gen imputed_flag = 0 foreach var of varlist qage qsex qrace { replace imputed_flag = 1 if `var' != 0 } -replace imputed_flag = . if year <1982 /*for CPS supplements other than the ASEC, values are imputed for all years 1982 forward*/ - -/*note about IPW for this sample: - we preserve sex, race, and age joint distribution. because we are focusing on child outcomes here, we have to contend with differing sampling frames in the CPS with respect to age and attainment. -attainment questions were only asked of children over 15 in the early years of the cps. thus, unlike other supplements in doing IPW in this supplement, we do not use attainment in the IPW. This is a non-issue in other samples where we focus on adults*/ - +replace imputed_flag = . if year <1982 +/* +JY: After bpl==09900, all the years before 1994 are dropped. +Do we really need to consider 1982 since this step only have year 1994-2019? +We think the above line of code is not useful. +*/ + +/*Note about IPW for This Sample: +We preserve sex, race, and age joint distribution. +Because we are focusing on child outcomes here, we have to contend with differing sampling frames in the CPS with respect to age and attainment. +Attainment questions were only asked of children over 15 in the early years of the CPS. +CPS across years have differnt smapling frames, so we need to use IPW to adjust them. +Thus, unlike other supplements in doing IPW in this supplement, we do not use attainment in the IPW. +This is a non-issue in other samples where we focus on adults*/ egen trq = group(age sex race) egen num = sum(imputed_flag*edsuppwt), by(trq) egen den = sum(edsuppwt), by(trq) gen phat = num/den drop if imputed_flag == 1 gen weight = edsuppwt/(1-phat) replace weight = edsuppwt if year <1982 +/*FYQ: why edsuppwt can be used in October CPS (though only available in October), will that affect the comparison with March CPS?*/ /*check sample before drop*/ +/*FYQ: in what cases are weights less than zero? (JY: missing value?)*/ drop if weight < 0 +/*FYQ+JY: why we name "cps_october_76_19" but not 94_19 given the years available*/ save cps_october_76_19, replace +/*************Prepare for main_analysis file**************/ # delimit cr clear all -cd "C:\Users\cmslo\Box Sync\race project" +cd "/Users/chenjieyi/Downloads" +*cd "C:\Users\cmslo\Box Sync\race project" *cd "D:\Box Sync\race project" use "cps_october_76_19" - /*Identify if someone is currently enrolled in school: universe for adults starts in 1978*/ +/*FYQ: why are we comparing against 1987 in some places while against 1978 elsewhere*/ + +/*Generate a currently in school indicator +EDATT indicates whether a person is attending or enrolled in regular school, 1 means yes*/ gen inschool = 0 replace inschool = 1 if edatt == 1 replace inschool = . if year < 1978 +/*Generate a last year in school indicator +EDATTLY indicates whether the person was enrolled in regular school the previous October, and 1 means yes*/ gen inschool_ly = 0 replace inschool_ly = 1 if edattly == 1 -replace inschool_ly = . if year <1987 - +replace inschool_ly = . if year < 1987 + +/*Generate a currently in high school indicator +EDGRADE identifies the current grade or level of schooling for people currently enrolled in regular school, and 201 to 204 means 9th to 12th grade*/ gen inhighschool = 0 -replace inhighschool = 1 if edgrade>= 201 & edgrade <=204 +replace inhighschool = 1 if edgrade>= 201 & edgrade <= 204 replace inhighschool = . if year < 1978 +/*Generate a currently in college indicator +EDGRADE 301 to 304 means college year1 to year4*/ gen incollege = 0 -replace incollege = 1 if edgrade>= 301 & edgrade <=304 +replace incollege = 1 if edgrade>= 301 & edgrade <= 304 replace incollege = . if year < 1978 +/*Generate a currently in college year1 indicator +EDGRADE 301 means college year1*/ gen inyr1college = 0 replace inyr1college = 1 if edgrade== 301 replace inyr1college = . if year < 1978 +/*Generate a currently in graduate school indicator +EDGRADE 401 to 402 means graduate school year1 to year2+*/ gen ingradschool = 0 replace ingradschool = 1 if edgrade>= 401 & edgrade <= 402 replace ingradschool = . if year < 1978 -/*make a postsecondary indicator that includes undergrad or grad school to have something to compare to IPEDS enrollments*/ +/*Generate a currently postsecondary indicator that includes undergrad or grad school to have something to compare to IPEDS enrollments*/ gen inpostsec = 0 replace inpostsec = 1 if incollege == 1 | ingradschool == 1 replace inpostsec = . if year < 1978 +/*Generate a currently in non degree program indicator +EDVOCA indicates whether or not a person is currently enrolled in a training program other than regular school, and 2 means yes*/ gen innondegree = 0 -replace innondegree = 1 if edvoca == 2 /*check the crosstab of this one with college*/ +replace innondegree = 1 if edvoca == 2 replace innondegree = . if year < 1987 replace innondegree = . if incollege == 1 replace innondegree = . if inhighschool == 1 replace innondegree = . if ingradschool == 1 - +/*FYQ: for those who are enrolled in both vocational school and a regular high school/college/grad, why are we setting innondegree to "." instead of "zero"*/ + +/*Generate a last year in high school indicator +EDGRDLY identifies which grade the focal person was enrolled in the previous October, and 201 to 204 means 9th to 12th grade*/ gen inhighschool_ly = 0 replace inhighschool_ly = 1 if edgrdly>= 201 & edgrdly <=204 replace inhighschool_ly = . if year < 1987 +/*Generate a last year in college indicator +EDGRDLY 301 to 304 means college year1 to year4*/ gen incollege_ly = 0 replace incollege_ly = 1 if edgrdly>= 301 & edgrdly <=304 replace incollege_ly = . if year < 1987 +/*Generate a last year in college year1 indicator +EDGRDLY 301 means college year1*/ gen inyr1college_ly = 0 replace inyr1college_ly = 1 if edgrdly== 301 replace inyr1college_ly = . if year < 1987 +/*Generate a last year in graduate school indicator +EDGRDLY 401 to 402 means graduate school year1 to year2+*/ gen ingradschool_ly = 0 replace ingradschool_ly = 1 if edgrdly>= 401 & edgrdly <= 402 replace ingradschool_ly = . if year < 1988 +/*Generate a currently in public school indicator +EDPUPR identifies whether a person was in public school or private school enrollment, and 10 is public*/ gen public = 0 replace public = 1 if edpupr == 10 replace public = . if year <1978 +/*Generate a currently in private school indicator +EDPUPR 11 is private*/ gen private = 0 replace private = 1 if edpupr == 11 replace private = . if year <1978 +/*Generate a currently in public college indicator*/ gen incollege_public = incollege * public replace incollege_public = . if year < 1978 +/*Generate a currently in private college indicator*/ gen incollege_private = incollege * private replace incollege_private = . if year < 1978 +/*Generate a fulltime student indicator +EDFULL identifies the part-time or full-time status of people currently attending college or graduate school, and 1 is fulltime*/ gen fulltime = 0 replace fulltime = 1 if edfull == 1 replace fulltime = . if year < 1978 +/*Generate a parttime student indicator +EDFULL 2 is parttime*/ gen parttime = 0 replace parttime = 1 if edfull == 2 replace parttime = . if year < 1978 +/*Generate a currently fulltime in college indicator*/ gen incollege_fulltime = incollege * fulltime replace incollege_fulltime = . if year < 1978 +/*Generate a currently parttime in college indicator*/ gen incollege_parttime = incollege * parttime replace incollege_parttime = . if year < 1978 +/*Generate a 2-year college indicator +EDTYPE indicates whether persons who are enrolled in college attend a 2-year or 4-year institution, and 1 is 2-year college*/ gen two_yr = 0 replace two_yr = 1 if edtype == 1 replace two_yr = . if year <1978 +/*Generate a 4-year college indicator +EDTYPE 2 is 4-year college*/ gen four_yr = 0 replace four_yr = 1 if edtype == 2 replace four_yr = . if year <1978 +/*Generate a currently in 2-year college indicator*/ gen incollege_two = incollege * two_yr replace incollege_two = . if year < 1978 +/*Generate a currently in 4-year college indicator*/ gen incollege_four = incollege * four_yr replace incollege_four = . if year < 1978 +/*Generate a not in labor force indicator +EMPSTAT indicates whether persons were part of the labor force, and >= 30 means not in labor force*/ gen nilf = 0 replace nilf= 1 if empstat >= 30 +/*Generate a currently in college and not in labor force indicator*/ gen incollege_nilf = incollege * nilf + +/*Generate a currently full-time in college and not in labor force indicator*/ gen incollege_nilf_full = incollege * nilf * fulltime + +/*Generate a currently part-time in college and not in labor force indicator*/ +/*FYQ: In what cases would both nilf and parttime both be 1? Isn't that a data error if that occurred?*/ gen incollege_nilf_part = incollege * nilf * parttime +/*Generate a working for wages indicator +CLASSWKR 20-28 means works for wages or salary*/ +/*FYQ: Why do we need it? Is it because the data were not perfect so that some people say that they are wage_worker yet they are not in the labor force?*/ gen wage_worker = 0 replace wage_worker = 1 if classwkr >=20 & classwkr <= 28 replace wage_worker = 0 if nilf == 1 +/*Generate a in college and working for wages indicator*/ gen incollege_wageworker = incollege * wage_worker - /*make birth cohort vars*/ - gen byear = year - age - gen birth_cohort = 1915 if byear >= 1913 & byear <=1917 replace birth_cohort = 1920 if byear >= 1918 & byear <=1922 replace birth_cohort = 1925 if byear >= 1923 & byear <=1927 replace birth_cohort = 1930 if byear >= 1928 & byear <=1932 replace birth_cohort = 1935 if byear >= 1933 & byear <=1937 replace birth_cohort = 1940 if byear >= 1938 & byear <=1942 replace birth_cohort = 1945 if byear >= 1943 & byear <=1947 replace birth_cohort = 1950 if byear >= 1948 & byear <=1952 replace birth_cohort = 1955 if byear >= 1953 & byear <=1957 replace birth_cohort = 1960 if byear >= 1958 & byear <=1962 replace birth_cohort = 1965 if byear >= 1963 & byear <=1967 replace birth_cohort = 1970 if byear >= 1968 & byear <=1972 replace birth_cohort = 1975 if byear >= 1973 & byear <=1977 replace birth_cohort = 1980 if byear >= 1978 & byear <=1982 replace birth_cohort = 1985 if byear >= 1983 & byear <=1987 replace birth_cohort = 1990 if byear >= 1988 & byear <=1992 +/*Generate some age group variables*/ gen a20 = 0 replace a20 = 1 if age >= 18 & age <= 24 gen a25 = 0 replace a25 = 1 if age >= 25 & age <= 29 gen a30 = 0 replace a30 = 1 if age >= 30 & age <= 34 gen a35 = 0 replace a35 = 1 if age >= 35 & age <= 39 gen a40 = 0 replace a40 = 1 if age >= 40 & age <= 44 gen a45 = 0 replace a45 = 1 if age >= 45 & age <= 49 gen a50 = 0 replace a50 = 1 if age >= 50 & age <= 54 gen a55 = 0 replace a55 = 1 if age >= 55 & age <= 59 gen a60 = 0 replace a60 = 1 if age >= 60 & age <= 64 gen a65 = 0 replace a65 = 1 if age >= 65 & age <= 69 +/*FYQ + JY: Why we need a45_55 and why it is from 43 to 57*/ gen a45_55 = 0 replace a45_55 = 1 if age>=43 & age <=57 +/*Generate a 5-year age group variable */ gen age_5 = 0 -replace age_5 = 20 if a20 == 1 +replace age_5 = 20 if a20 == 1 /*Be careful that a20 = [18, 24] */ replace age_5 = 25 if a25 == 1 replace age_5 = 30 if a30 == 1 replace age_5 = 35 if a35 == 1 replace age_5 = 40 if a40 == 1 replace age_5 = 45 if a45 == 1 replace age_5 = 50 if a50 == 1 replace age_5 = 55 if a55 == 1 replace age_5 = 60 if a60 == 1 replace age_5 = 65 if a65 == 1 +/* Generate some binary gender vairables*/ gen male = 0 replace male = 1 if sex == 1 gen female = 0 replace female = 1 if sex == 2 +/*Generate some binary race variables +Our focus is on black and white who are not hispanic*/ gen white = 0 replace white = 1 if race == 100 & hispan == 0 gen black = 0 replace black = 1 if race == 200 & hispan == 0 +/*Create an alternative geography (super_state). +The 12th largest states get assigned their state.The rest get assigned to their census division.*/ gen super_state = 0 replace super_state = 32 if statefip == 1 replace super_state = 42 if statefip == 2 replace super_state = 41 if statefip == 4 replace super_state = 33 if statefip == 5 replace super_state = 1 if statefip == 6 replace super_state = 41 if statefip == 8 replace super_state = 81 if statefip == 9 replace super_state = 31 if statefip == 10 replace super_state = 31 if statefip == 11 replace super_state = 3 if statefip == 12 replace super_state = 8 if statefip == 13 replace super_state = 42 if statefip == 15 replace super_state = 41 if statefip == 16 replace super_state = 6 if statefip == 17 replace super_state = 21 if statefip == 18 replace super_state = 22 if statefip == 19 replace super_state = 22 if statefip == 20 replace super_state = 32 if statefip == 21 replace super_state = 33 if statefip == 22 @@ -3773,92 +3852,95 @@ replace super_state = 4 if statefip == 36 replace super_state = 9 if statefip == 37 replace super_state = 22 if statefip == 38 replace super_state = 7 if statefip == 39 replace super_state = 33 if statefip == 40 replace super_state = 42 if statefip == 41 replace super_state = 5 if statefip == 42 replace super_state = 81 if statefip == 44 replace super_state = 31 if statefip == 45 replace super_state = 22 if statefip == 46 replace super_state = 32 if statefip == 47 replace super_state = 2 if statefip == 48 replace super_state = 41 if statefip == 49 replace super_state = 81 if statefip == 50 replace super_state = 12 if statefip == 51 replace super_state = 42 if statefip == 53 replace super_state = 31 if statefip == 54 replace super_state = 31 if statefip == 55 replace super_state = 21 if statefip == 56 assert super_state > 0 -lab def ss 1 "CA" 2 "TX" 3 "FL" 4 "NY" 5 "PA" 6 "IL" 7 "OH" 8 "GA" 9 "NC" 10 "MI" 11 "NJ" 12 "VA" 21 "East North Central" 22 "West North Central" /* -*/ 31 "South Atlantic" 32 "East South Central" 33 "West South Central" 41 "Mountain" 42 "Pacific" 81 "New England" +lab def ss 1 "CA" 2 "TX" 3 "FL" 4 "NY" 5 "PA" 6 "IL" 7 "OH" 8 "GA" 9 "NC" 10 "MI" 11 "NJ" 12 "VA" 21 "East North Central" 22 "West North Central" 31 "South Atlantic" 32 "East South Central" 33 "West South Central" 41 "Mountain" 42 "Pacific" 81 "New England" lab val super_state ss - - +/*FYQ+JY: Why we name "main_analysis_cps_october_76_19" but not 94_19 given the years available?*/ save main_analysis_cps_october_76_19, replace - /*2. CREATE AN ANALYSIS FILE age_profile_enrolled_students_cps_oct COLLAPSED BY SURVEY YEAR AND AGE GROUP TO GENERATE TABLES AND FIGURES DESCRIBING AGGREGATE TRENDS IN SCHOOL ATTENDANCE*/ /*First, we want to describe within a survey year what share of students are adult students-- this is to make analagous pictures to the IPEDS pictures. This file creates variables that decompose school attendance within a survey year by age group to answer questions such as: "in a given year, what share of all students attending school were between ages 18 and 24?"*/ # delimit cr clear all -cd "C:\Users\cmslo\Box Sync\race project" +cd "/Users/chenjieyi/Downloads" +*cd "C:\Users\cmslo\Box Sync\race project" *cd "D:\Box Sync\race project" use "main_analysis_cps_october_76_19" -keep if inrange(age, 18, 69) - +/*Limit age between 18 and 69*/ +keep if inrange(age, 18, 69) + +/*We assign count = 1 to each observation. +Within each age-year group, the sum of the count variable tells us the number of observations in that group.*/ gen count = 1 -collapse (sum) count inschool inhighschool incollege inyr1college ingradschool inpostsec innondegree incollege_public incollege_private incollege_fulltime incollege_parttime incollege_two incollege_four incollege_nilf incollege_nilf_full incollege_nilf_part incollege_wageworker [pw=weight], by(year age_5) - +/*Count the number of realized (value = 1) obseravtions for each indicators, grouping by survey year and 5-year age group*/ +collapse (sum) count inschool inhighschool incollege inyr1college ingradschool inpostsec innondegree incollege_public incollege_private incollege_fulltime incollege_parttime incollege_two incollege_four incollege_nilf incollege_nilf_full incollege_nilf_part incollege_wageworker [pw=weight], by(year age_5) + +/*FYQ: Is this necessary? In what cases would we have missing values here?*/ drop if age_5 == . -/*make variables that tells us the total number of people in school, in high school, in college, and in grad school in our sample */ +/*Make variables that tell us the total number of people in school, in high school, in college, and in grad school in our sample for each survey year*/ sort year by year: egen inschool_sample = total(inschool) by year: egen inhighschool_sample = total(inhighschool) by year: egen incollege_sample = total(incollege) by year: egen ingradschool_sample = total(ingradschool) by year: egen inpostsec_sample = total(inpostsec) by year: egen inpublic_sample = total(incollege_public) by year: egen inprivate_sample = total(incollege_private) by year: egen inyr1_sample = total(inyr1college) by year: egen in4yr_sample = total(incollege_four) -/*make vars that tell us the age profile of all adult students */ +/*Make variables that tell us the age profile of all adult students for each survey year*/ gen age_sh_school = inschool / inschool_sample format age_sh_school %6.4f gen age_sh_highschool = inhighschool / inhighschool_sample format age_sh_highschool %6.4f gen age_sh_college = incollege / incollege_sample format age_sh_college %6.4f gen age_sh_public = incollege_public / inpublic_sample format age_sh_public %6.4f gen age_sh_private = incollege_private / inprivate_sample format age_sh_private %6.4f gen age_sh_gradschool = ingradschool / ingradschool_sample format age_sh_gradschool %6.4f gen age_sh_postsec = inpostsec / inpostsec_sample @@ -3870,57 +3952,56 @@ gen age_sh_4yr = incollege_four / in4yr_sample format age_sh_4yr %6.4f save age_profile_enrolled_students_cps_oct, replace /*3. CREATE AN ANALYSIS FILE enrollment_analysis_cps_oct COLLAPSED BY BIRTH COHORT AND 5-year age group TO DESCRIBE AGGREGATE TRENDS IN SCHOOL ATTENDANCE*/ /*The previous analysis file decomposes all school enrollment in a survey year by age group to answer questions like: "in a given year, what share of all students attending school were between ages 18 and 24?" In contrast, this analysis file describes school attendance patterns within a 5-year birth cohort at a specific age to answer questions like: "what share of the 1970 birth cohort attended school between ages 18 and 24?" The "share variables" answer those types of questions. Additionally, we also create share variables that are conditioned on attempt, and variables to describe the population of enrolled students. */ # delimit cr clear all -cd "C:\Users\cmslo\Box Sync\race project" +cd "/Users/chenjieyi/Downloads" +*cd "C:\Users\cmslo\Box Sync\race project" *cd "D:\Box Sync\race project" use "main_analysis_cps_october_76_19" -* Limiting age between 23 and 69 -* +/*Limit age between 23 and 69*/ +/*JY+FYQ: why do we start from 23 here? */ keep if inrange(age, 23, 69) gen count = 1 - -collapse (sum) count inschool inhighschool incollege inyr1college ingradschool innondegree incollege_public incollege_private incollege_fulltime incollege_parttime incollege_two incollege_four incollege_nilf incollege_nilf_full incollege_nilf_part incollege_wageworker [pw=weight], by(birth_cohort age_5) +collapse (sum) count inschool inhighschool incollege inyr1college ingradschool innondegree incollege_public incollege_private incollege_fulltime incollege_parttime incollege_two incollege_four incollege_nilf incollege_nilf_full incollege_nilf_part incollege_wageworker [pw=weight], by(birth_cohort age_5) drop if birth_cohort == . -/*make some share variables to describe schooling patterns within age/ birth-cohort*/ - +/*Make some share variables to describe schooling patterns within age/ birth-cohort*/ gen share_inschool = . replace share_inschool = inschool / count if count != 0 format share_inschool %6.4f gen share_inhighschool = . replace share_inhighschool = inhighschool / count if count != 0 format share_inhighschool %6.4f gen share_incollege = . replace share_incollege = incollege / count if count != 0 format share_incollege %6.4f gen share_ingradschool = . replace share_ingradschool = ingradschool / count if count != 0 format share_ingradschool %6.4f gen share_innondegree = . replace share_innondegree = innondegree / count if count != 0 format share_innondegree %6.4f @@ -3978,50 +4059,52 @@ gen cond_incollege_nilf_part = . replace cond_incollege_nilf_part = incollege_nilf_part / incollege if incollege != 0 format cond_incollege_nilf_part %6.4f gen cond_incollege_wagewkr = . replace cond_incollege_wagewkr = incollege_wageworker/ incollege if incollege != 0 format cond_incollege_wagewkr %6.4f save enrollment_analysis_cps_oct, replace /*4. CREATE AN ANALYSIS FILE age_profile_enrolled_CA_students_cps_oct that includes only observations from the state of California COLLAPSED BY SURVEY YEAR AND AGE GROUP TO GENERATE TABLES AND FIGURES DESCRIBING AGGREGATE TRENDS IN SCHOOL ATTENDANCE*/ /*We want to benchmark age enrollment figures from the CSU AND CCCC systems with survey data from the October CPS. This file creates variables that decompose school attendance within a survey year for respondents residing in the state of California by age group to answer questions such as: "in a given year, what share of all students attending school were between ages 18 and 24?" We focus on attendance in public universities. */ # delimit cr clear all -cd "C:\Users\cmslo\Box Sync\race project" +cd "/Users/chenjieyi/Downloads" +*cd "C:\Users\cmslo\Box Sync\race project" *cd "D:\Box Sync\race project" use "main_analysis_cps_october_76_19" + +/*Keep California sample only*/ keep if super_state == 1 + +/*Limit age between 18 and 69*/ keep if inrange(age, 18, 69) gen count = 1 - -collapse (sum) count inschool inhighschool incollege inyr1college ingradschool inpostsec innondegree incollege_public incollege_private incollege_fulltime incollege_parttime incollege_two incollege_four incollege_nilf incollege_nilf_full incollege_nilf_part incollege_wageworker [pw=weight], by(year age_5) +collapse (sum) count inschool inhighschool incollege inyr1college ingradschool inpostsec innondegree incollege_public incollege_private incollege_fulltime incollege_parttime incollege_two incollege_four incollege_nilf incollege_nilf_full incollege_nilf_part incollege_wageworker [pw=weight], by(year age_5) drop if age_5 == . -/*make variables that tells us the total number of people in school, in high school, in college, and in grad school in our sample */ - +/*Make variables that tells us the total number of people in school, in high school, in college, and in grad school in our sample for each survey year*/ sort year by year: egen incollege_sample = total(incollege) by year: egen incollege_public_sample = total(incollege_public) -/*make vars that tell us the age profile of all adult students */ - +/*Make vars that tell us the age profile of all adult students for each survey year*/ gen age_sh_college = incollege / incollege_sample format age_sh_college %6.4f gen age_sh_college_public = incollege_public / incollege_public_sample format age_sh_college_public %6.4f save age_profile_enrolled_CA_students_cps_oct, replace