ltc-code 0.1.98__tar.gz → 0.2.0__tar.gz

{ltc_code-0.1.98 → ltc_code-0.2.0}/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.3
 Name: ltc-code
-Version: 0.1.98
+Version: 0.2.0
 Summary: Add your description here
 Requires-Python: >=3.9
 Description-Content-Type: text/markdown

{ltc_code-0.1.98 → ltc_code-0.2.0}/pyproject.toml

@@ -1,6 +1,6 @@
 [project]
 name = "ltc-code"
-version = "0.1.98"
+version = "0.2.0"
 description = "Add your description here"
 readme = "README.md"
 requires-python = ">=3.9"

ltc_code-0.2.0/src/ltc_code/20260614_new_build_scripts_update/check_cmo_apps.do

@@ -0,0 +1,374 @@
+* 2.1_nsc_prep_for_analysis
+* Sarah Cohodes
+* Prepare NSC data for analysis
+** transform long file to student level file
+** based on MA NSC programs
+
+clear all
+set more off
+cap log close _all
+set seed 123256789
+
+
+/* paths */
+global basepath "P:\Long_Term_Charter"      
+
+cd "$basepath"
+
+// raw data
+global raw "$basepath\data\raw"
+global lotto "$basepath\data\charter_lottery_data"
+
+// intermediate data
+global saves "$basepath\data\interim"
+global data "$basepath\data"  
+global data_clean "$basepath\data\clean" 
+
+// data  
+global data_cmo "$basepath\data\clean\new_build\cmo_apps" 
+global data_wide  "$basepath\data\clean\new_build\cmo_apps\stata_collapsed" 
+global data_long  "$basepath\data\clean\new_build\cmo_apps\stata" 
+
+// code
+global do "$basepath\programs\explore_sarah\check_cmo_apps"       
+
+// output
+global output "$do\tables_figures"   
+global log "$do\log_files" 
+
+
+**who is in sample
+global cmo 			aspire democracy_prep ilt christel_house rocketship yes_prep   
+global cmo_minus 	 christel_house democracy_prep ilt christel_house rocketship yes_prep  //makes combined file (don't need first alpha aspire)
+global cmo_plus 	combined aspire christel_house yes_prep  democracy_prep ilt  // rocketship 
+
+*rocketship built but not clear enough nondegenerate lottos to include, plus unbalanced.
+
+*just add new cmo here if one at a time
+*global cmo democracy_prep
+
+ 
+*************
+*SWITCHES* 
+*************
+
+global analysis_file		1 //assemble analysis file
+global analysis				1 //basic analysis
+*make version that is initialL_offer only
+
+
+if $analysis_file==1{
+
+foreach cmo of global cmo {	
+foreach type in wide long {	
+	use "${data_`type'}/`cmo'.dta", clear 
+	*CMO SPECIFIC FIXES
+	
+	if "`cmo'" == "yes_prep" {
+		cap g enroll=offer_accepted
+	}
+	
+	if "`cmo'" == "aspire" |  "`cmo'" == "rocketship" {
+		 g late_app=0
+	}
+	
+	if  "`cmo'" == "rocketship" {
+		destring school_year, replace
+		destring sid_cepr, replace
+		keep if school_year>=2016
+	}
+
+	*bring in cohort
+		merge m:1 sid_cepr using "$saves/IDS_ltc.dta", keep(1 3)
+		g nsc_sample = 1 if cohort>=2010&cohort<=2023 //update here when have new NSC data
+	
+	
+	*nsc degrees
+		merge m:1 sid_cepr using "$saves/flat_degrees_NSC_ltc.dta" , keep(1 3) nogen //NSC last updated 2024
+		
+		*merge nsc enrollment
+		merge m:1 sid_cepr using  "$saves/flat_enrollment_NSC_ltc.dta" , keep(1 3) nogen 
+
+		g opeid = ID_FSC_firstinst  
+		destring opeid, replace
+	
+		merge m:1 sid_cepr using "$data_clean/sent_to_nsc.dta", nogen keep(1 3)
+		replace sent_to_nsc=0 if sent_to_nsc==.
+		
+		
+		*College characteristics from Chetty paper
+		merge m:1 opeid using "$data_clean\chetty_college.dta", nogen  force keep(1 3) ///
+		keepusing(k_mean k_mean_m k_mean_f opeid superopeid institution_name superopeid_name ///
+		*mean*insuff* *mean*never* tier grad_rate_150_p_2013 exp_instr_pc_2013 flagship ipeds_enrollment_2013) 
+		ren institution_name institution_name
+			ren superopeid_name superopeid_name
+	
+	*spread insufficient and never attend outcomes
+	foreach v of varlist *_insuff* *_never* {
+		egen max=max(`v') 
+		replace `v'=max
+		drop max
+	}
+	
+
+
+			merge m:1 ID_FSC_firstinst using  "${saves}/temp2.dta", keep(1 3) nogen
+
+		* merging base file with new IPEDS data 
+			merge m:1 ID_FSC_firstinst using  "${saves}/temp0.dta", keep(1 3) nogen
+
+			*will have to come back to this it uses 1.12_ipeds educationdata
+			*merge m:1 unitid using "$data_clean\ipeds_data.dta", nogen  force keep(1 3)
+			
+			compress
+	
+	*** replace k_mean for those who don't attend college
+	***** note there are a few colleges that don't match and those get a different number	  
+	replace k_mean = k_mean_insuff if k_mean==.&opeid!=. // insufficient data college
+
+		replace k_mean= k_mean_never if k_mean==. 
+	
+		*outcomes are processed more below
+		
+			
+		foreach v of varlist att*Y1* {
+			replace `v'=0 if `v'==. & cohort<=2023
+				replace `v'=. if  cohort>2023
+		}		
+		foreach v of varlist att*Y2* {
+			replace `v'=0 if `v'==. & cohort<=2022
+				replace `v'=. if  cohort>2022
+		}
+		foreach v of varlist att*Y3* cmp*Y3*{
+			replace `v'=0 if `v'==. & cohort<=2021
+				replace `v'=. if  cohort>2021
+		}
+		foreach v of varlist att*Y4* cmp*Y4*{
+			replace `v'=0 if `v'==. & cohort<=2020
+				replace `v'=. if  cohort>2020
+		}
+		foreach v of varlist att*Y5* cmp*Y5* {
+			replace `v'=0 if `v'==. & cohort<=2019
+					replace `v'=. if  cohort>2019
+		}
+		foreach v of varlist att*Y6* cmp*Y6*{
+			replace `v'=0 if `v'==. & cohort<=2018
+				replace `v'=. if  cohort>2018
+		}
+		foreach v of varlist att*Y7* cmp*Y7* {
+			replace `v'=0 if `v'==. & cohort<=2017
+				replace `v'=. if  cohort>2017
+		}
+		foreach v of varlist att*Y8* cmp*Y8*{
+			replace `v'=0 if `v'==. & cohort<=2016
+				replace `v'=. if  cohort>2016
+		}
+		
+		
+	*this should be by first institution
+	tab college_name_first if tier == . & (cmp_BA_byY6 == 1 | att_4yr_byY2 == 1), mis	
+		
+	egen group = group(college_name_first)	
+		bys group: egen mode_tier = mode(tier)
+		bys group: egen median_tier = median(tier)
+
+	replace tier = mode_tier if tier == . & (cmp_BA_byY6 == 1  | att_4yr_byY2 == 1)
+	replace tier = median_tier if tier == . & (cmp_BA_byY6 == 1  | att_4yr_byY2 == 1)
+	replace tier = 8 if college_name_first == "BENJAMIN FRANKLIN INSTITUTE OF TECHNOL" & tier == . 
+	replace tier = 8 if college_name_first == "BENJAMIN FRANKLIN INSTITUTE OF TECHNOLOGY" & tier == . 	
+	replace tier = 8 if college_name_first == "BENJAMIN FRANKLIN INSTITUTE OF TECHNOLOG" & tier == . 	
+	replace tier = 7  if college_name_first == "DELAWARE STATE UNIVERSITY" & tier == . 	
+	replace tier = 5  if college_name_first == "FRAMINGHAM STATE UNIVERSITY" & tier == . 	
+	replace tier = 11 if college_name_first == "FLORIDA CAREER COLLEGE - MIAMI" & tier == .	
+	replace tier = 7  if college_name_first == "NEW COLLEGE OF FLORIDA" & tier == . 	
+	replace tier = 8 if college_name_first == "NEWBURY COLLEGE" & tier == . 	
+	replace tier = 4 if college_name_first == "NORTHWESTERN UNIVERSITY" & tier == . 	
+	replace tier = 7 if college_name_first == "THE OHIO STATE UNIVERSITY" & tier == . 	
+	replace tier = 7  if college_name_first == "TROY UNIVERSITY" & tier == . 	
+	replace tier = 8 if college_name_first == "UNITY COLLEGE" & tier == . 	
+	replace tier = 8 if college_name_first == "UNIVERSITY OF NORTHWESTERN OHIO - COLL OF TECH" & tier == . 	
+
+	* generating aggregate categories for each tier
+
+	g hselective = 0
+		replace hselective = 1 if tier == 1 | tier == 2 | tier == 3 | tier == 4
+
+	g selective = 0
+		replace selective = 1 if tier == 5 | tier == 6
+
+	g nonselective = 0
+		replace nonselective = 1 if tier == 7 | tier == 8 | tier == 9 | tier == 10 | tier == 11
+
+	g insufficient_dta = 0
+		replace insufficient_dta = 1 if tier == 13 
+		
+	g BA_byY6_hselective = cmp_BA_byY6*hselective
+	g BA_byY6_selective = cmp_BA_byY6*selective
+	g BA_byY6_nonselective = cmp_BA_byY6*nonselective
+	g BA_byY6_insufficient = cmp_BA_byY6*insufficient_dta
+
+	* generating aggregate categories for Barrons ranks
+
+	* fixing missing values of the Barrons rank variable
+
+	tab college_name_first if barrons_rank == "" & (cmp_BA_byY6==1 | att_4yr_inY2 == 1), mis
+
+	bys group: egen mode_rank = mode(barrons_rank)
+
+	replace barrons_rank = mode_rank if barrons_rank == "" & (cmp_BA_byY6 == 1 | att_4yr_inY2 == 1)
+
+	replace barrons_rank = "Highly Competitive" if (college_name_first == "NORTHWESTERN UNIVERSITY" & barrons_rank == "") | (college_name_first == "CASE WESTERN RESERVE UNIVERSITY" & barrons_rank == "")
+
+	g hcompetitive = 0
+		replace hcompetitive = 1 if barrons_rank == "Highly Competitive" | barrons_rank == "Most Competitive" | barrons_rank == "Very Competitive"
+
+	g competitive = 0
+		replace competitive = 1 if barrons_rank == "Competitive" |  barrons_rank == "Special" | (college_name_first == "LIBERTY UNIVERSITY" & barrons_rank == "")
+
+	g noncompetitive = 0
+		replace noncompetitive = 1 if barrons_rank == "Non Competitive" | barrons_rank == "Unranked"  | barrons_rank == "Less Competitive" | (college_name_first == "LANDMARK COLLEGE" & barrons_rank == "")
+
+	* generating graduation variables by aggregate ranks
+		
+	g BA_byY6_hcompetitive = cmp_BA_byY6*hcompetitive
+	g BA_byY6_competitive = cmp_BA_byY6*competitive
+	g BA_byY6_noncompetitive = cmp_BA_byY6*noncompetitive
+drop _m
+
+foreach n of numlist 1/6 {
+	label var att_any_byY`n' "Attend Any by Y`n'"
+	label var att_2yr_byY`n' "Attend 2-Year by Y`n'"
+	label var att_4yr_byY`n' "Attend 4-Year by Y`n'"
+}
+foreach n of numlist 4/6 {
+	label var cmp_any_byY`n' "Complete BA or AA by Y`n'"
+	label var cmp_AA_byY`n' "Complete AA by Y`n'"
+	label var cmp_BA_byY`n' "Complete BA by Y`n'"
+}
+	
+	sort _all
+	compress
+	save "$data_clean\baseanalysisfile_ltc_lotto_`cmo'_`type'.dta", replace
+} //end type loop
+} // end cmo loop
+
+* create combined file
+/*
+use "$data_clean\baseanalysisfile_ltc_lotto.dta", clear
+		drop if cmo_name=="ROCKETSHIP"
+		drop if cmo_name=="YES PREP"
+		drop if cmo_name=="ASPIRE PUBLIC SCHOOLS"
+		drop if cmo_name=="DEMOCRACY PREP"
+		drop if cmo_name=="INTERNATIONAL LEADERSHIP OF TEXAS"
+		drop if cmo_name=="GREEN DOT PUBLIC SCHOOLS"
+		drop if cmo_name=="HARMONY PUBLIC SCHOOLS"
+
+save "$data_clean\old_wide.dta", replace
+*/
+foreach type in wide long {	 
+	use "$data_clean\baseanalysisfile_ltc_lotto_aspire_`type'.dta", replace
+		foreach cmo_minus of global cmo_minus {	
+			append using "$data_clean\baseanalysisfile_ltc_lotto_`cmo_minus'_`type'.dta", force
+}
+sort sid_cepr
+compress
+	save "$data_clean\baseanalysisfile_ltc_lotto_combined_`type'.dta", replace
+
+}	
+
+} // end analysis_file loop
+
+if $analysis==1{
+foreach cmo of global cmo_plus {	
+	
+	use "$data_clean\baseanalysisfile_ltc_lotto_`cmo'_wide.dta", clear
+	
+	*set up offer/outcome types
+	if "`cmo'" == "ilt" {
+		global offer initial_offer
+		global outcomes att_any_byY2 att_4yr_byY2 att_2yr_byY2 
+	}
+	else {
+		global offer initial_offer offer
+		global outcomes att_any_byY2 att_4yr_byY2 att_2yr_byY2 cmp_any_byY6 cmp_BA_byY6 cmp_AA_byY6
+	}
+	
+	cap g late_app = 0
+	cap g application_cancel = 0 
+	cap  replace initial_offer = 0 if initial_offer ==.
+	
+	*basic setup
+	drop if application_cancel==1
+	drop if late_app==1
+		
+	bys sid_cepr: gen ww = _N 
+	gen w = 1/ww
+	
+	*identify degenerate lotteries
+	bys risk_set: gen risk_set_count = _N
+	bys risk_set: egen risk_set_offer = sum(offer)
+	g degenerate = risk_set_count == risk_set_offer
+	g nondegenerate = 1-degenerate
+	g all = 1
+
+	tempfile results_`cmo'
+	postfile t str32( model offertype lottotype outcome)  mean_control beta se n  using `results_`cmo'', replace
+	
+	foreach offertype in $offer {
+	foreach lottotype in all nondegen{
+		reghdfe p_his `offertype' [aw = w] if cohort<=2025&`lottotype'==1, absorb(risk_set) cluster(sid_cepr)
+				local beta = _b[`offertype']
+				local se  = _se[`offertype']
+				local n  = e(N)
+			qui margins , at(`offertype'=0)
+			local mean = r(b)[1,1]	
+				post t ("Balance") ("`offertype'") ("`lottotype'")  ("Percent Hispanic") (`mean') (`beta') (`se') (`n')
+		}
+	}
+
+	foreach offertype in $offer {
+	foreach lottotype in all nondegen{
+		reghdfe enroll `offertype' [aw = w] if cohort<=2025&`lottotype'==1, absorb(risk_set) cluster(sid_cepr)
+				local beta = _b[`offertype']
+				local se  = _se[`offertype']
+				local n  = e(N)
+			qui margins , at(`offertype'=0)
+			local mean = r(b)[1,1]	
+				post t ("First Stage") ("`offertype'") ("`lottotype'")  ("Enroll") (`mean') (`beta') (`se') (`n')	
+		}
+	}
+	
+	foreach var in $outcomes {
+		local label : variable label `var'
+	foreach offertype in $offer {
+	foreach lottotype in all nondegen{	
+		
+		reghdfe `var' `offertype' [aw = w] if   nsc_sample==1 & cohort<=2025& `lottotype'==1, absorb(risk_set) cluster(sid_cepr)
+				local beta = _b[`offertype']
+				local se  = _se[`offertype']
+				local n  = e(N)
+			qui margins , at(`offertype'=0)
+			local mean = r(b)[1,1]	
+				post t ("Reduced Form") ("`offertype'") ("`lottotype'")  ("`label'") (`mean') (`beta') (`se') (`n')	
+		
+		ivreghdfe `var' (enroll = `offertype') if nsc_sample==1 & cohort<=2025& `lottotype'==1, ///
+			absorb(risk_set) cluster(sid_cepr) 		
+				local beta = _b[enroll]
+				local se  = _se[enroll]
+				local n  = e(N)
+		*retain mean from above		
+				post t ("2SLS") ("`offertype'") ("`lottotype'")  ("`label'") (`mean') (`beta') (`se') (`n')	
+
+			}
+		}
+		}
+	
+	postclose t
+
+	preserve
+		use  `results_`cmo'', clear
+		export excel using "$output\basic_results_`cmo'_20260608.xlsx", replace firstrow(variables)
+	restore
+	
+} //end cmo loop
+} //end analysis loop