ltc-paper-scripts 0.1.0__py3-none-any.whl

This diff represents the content of publicly available package versions that have been released to one of the supported registries. The information contained in this diff is provided for informational purposes only and reflects changes between package versions as they appear in their respective public registries.
Files changed (35) hide show
  1. ltc_paper_scripts/README.md +87 -0
  2. ltc_paper_scripts/TICK_ALGORITHM_LICENSE.txt +26 -0
  3. ltc_paper_scripts/build_latex.py +293 -0
  4. ltc_paper_scripts/data/codex_synthetic/new_release_preview.xlsx +0 -0
  5. ltc_paper_scripts/exhibit_selection.py +52 -0
  6. ltc_paper_scripts/figure_ticks.py +51 -0
  7. ltc_paper_scripts/fonts/DejaVuSans-Bold.ttf +0 -0
  8. ltc_paper_scripts/fonts/DejaVuSans.ttf +0 -0
  9. ltc_paper_scripts/fonts/LICENSE.txt +96 -0
  10. ltc_paper_scripts/latex_figures.py +250 -0
  11. ltc_paper_scripts/latex_requirements.txt +4 -0
  12. ltc_paper_scripts/latex_source/outputs/tables/output/age25_earnings_table.tex +38 -0
  13. ltc_paper_scripts/latex_source/outputs/tables/output/balance_table.tex +54 -0
  14. ltc_paper_scripts/latex_source/outputs/tables/output/college_data_comparison_table.tex +20 -0
  15. ltc_paper_scripts/latex_source/outputs/tables/output/college_enrollment_completion_table.tex +31 -0
  16. ltc_paper_scripts/latex_source/outputs/tables/output/college_heterogeneity_table.tex +36 -0
  17. ltc_paper_scripts/latex_source/outputs/tables/output/college_mediation_additional_specs_table.tex +30 -0
  18. ltc_paper_scripts/latex_source/outputs/tables/output/college_test_score_mediation_table.tex +18 -0
  19. ltc_paper_scripts/latex_source/outputs/tables/output/earnings_heterogeneity_table.tex +54 -0
  20. ltc_paper_scripts/latex_source/outputs/tables/output/earnings_robustness_table.tex +29 -0
  21. ltc_paper_scripts/latex_source/outputs/tables/output/earnings_threshold_table.tex +27 -0
  22. ltc_paper_scripts/latex_source/outputs/tables/output/other_mediators_table.tex +31 -0
  23. ltc_paper_scripts/latex_source/outputs/tables/output/peer_family_comparisons_table.tex +20 -0
  24. ltc_paper_scripts/latex_source/outputs/tables/output/summary_statistics_table.tex +60 -0
  25. ltc_paper_scripts/latex_source/paper/appendix.tex +138 -0
  26. ltc_paper_scripts/latex_source/paper/preamble.tex +502 -0
  27. ltc_paper_scripts/latex_source/paper/tables-and-figures.tex +147 -0
  28. ltc_paper_scripts/mappings.py +328 -0
  29. ltc_paper_scripts/paths.py +23 -0
  30. ltc_paper_scripts/run_latex.sh +4 -0
  31. ltc_paper_scripts/split_inputs.py +163 -0
  32. ltc_paper_scripts/table_bindings.json +4084 -0
  33. ltc_paper_scripts-0.1.0.dist-info/METADATA +94 -0
  34. ltc_paper_scripts-0.1.0.dist-info/RECORD +35 -0
  35. ltc_paper_scripts-0.1.0.dist-info/WHEEL +4 -0
@@ -0,0 +1,20 @@
1
+
2
+ \setlength{\tabcolsep}{8pt}
3
+ \begin{tabular}{L{8cm} C{4cm}}
4
+ \toprule
5
+ \multicolumn{2}{l}{\textit{A. Parental-Income Rank: Descriptive Comparison}} \\
6
+ Charter-enrollee comparison mean
7
+ & @@peer_family_comparisons_table.0.0@@ \\
8
+ Childhood-tract comparison mean
9
+ & @@peer_family_comparisons_table.1.0@@ \\
10
+ Difference (percentile points)
11
+ & @@peer_family_comparisons_table.2.0@@ \\
12
+ \addlinespace
13
+ \multicolumn{2}{l}{\textit{B. Teenage Childbearing: Female, Age 20+ Sample}} \\
14
+ Enrolled vs. Control (pp)
15
+ & @@peer_family_comparisons_table.3.0@@ \\
16
+ & @@peer_family_comparisons_table.4.0@@ \\
17
+ Control Complier Mean (\%)
18
+ & @@peer_family_comparisons_table.5.0@@ \\
19
+ \bottomrule
20
+ \end{tabular}
@@ -0,0 +1,60 @@
1
+ % Workbook-driven table; source variable mappings are in build_latex.py.
2
+ % Values for binary variables are expressed in percentage points.
3
+
4
+
5
+ \setlength{\tabcolsep}{2pt}
6
+ \begin{tabular}{L{4.7cm} *{4}{C{1.3cm} C{1.1cm}}}
7
+ \toprule
8
+ & \multicolumn{2}{C{2.54cm}}{Age 20+}
9
+ & \multicolumn{2}{C{2.54cm}}{Age 25+}
10
+ & \multicolumn{2}{C{2.54cm}}{Tract Population 25+}
11
+ & \multicolumn{2}{C{2.54cm}}{National Population 25+} \\
12
+ \cmidrule(lr){2-3}\cmidrule(lr){4-5}\cmidrule(lr){6-7}\cmidrule(lr){8-9}
13
+ & Mean & SD & Mean & SD & Mean & SD & Mean & SD \\
14
+ & (1) & (2) & (3) & (4) & (5) & (6) & (7) & (8) \\
15
+ \midrule
16
+
17
+ \multicolumn{9}{l}{\textit{A. Family Background}} \\
18
+ \hspace{0.1cm}Parental income (\$)
19
+ & @@summary_statistics_table.0.0@@ & @@summary_statistics_table.0.1@@ & @@summary_statistics_table.0.2@@ & @@summary_statistics_table.0.3@@ & @@summary_statistics_table.0.4@@ & @@summary_statistics_table.0.5@@ & @@summary_statistics_table.0.6@@ & @@summary_statistics_table.0.7@@ \\
20
+ \hspace{0.1cm}Parental income rank (percentile)
21
+ & @@summary_statistics_table.1.0@@ & @@summary_statistics_table.1.1@@ & @@summary_statistics_table.1.2@@ & @@summary_statistics_table.1.3@@ & @@summary_statistics_table.1.4@@ & @@summary_statistics_table.1.5@@ & @@summary_statistics_table.1.6@@ & @@summary_statistics_table.1.7@@ \\
22
+
23
+ \addlinespace
24
+ \multicolumn{9}{l}{\textit{B. Demographics}} \\
25
+ \hspace{0.1cm}Female (\%)
26
+ & @@summary_statistics_table.2.0@@ & @@summary_statistics_table.2.1@@ & @@summary_statistics_table.2.2@@ & @@summary_statistics_table.2.3@@ & @@summary_statistics_table.2.4@@ & @@summary_statistics_table.2.5@@ & @@summary_statistics_table.2.6@@ & @@summary_statistics_table.2.7@@ \\
27
+ \hspace{0.1cm}Hispanic (\%)
28
+ & @@summary_statistics_table.3.0@@ & @@summary_statistics_table.3.1@@ & @@summary_statistics_table.3.2@@ & @@summary_statistics_table.3.3@@ & @@summary_statistics_table.3.4@@ & @@summary_statistics_table.3.5@@ & @@summary_statistics_table.3.6@@ & @@summary_statistics_table.3.7@@ \\
29
+ \hspace{0.1cm}Black (\%)
30
+ & @@summary_statistics_table.4.0@@ & @@summary_statistics_table.4.1@@ & @@summary_statistics_table.4.2@@ & @@summary_statistics_table.4.3@@ & @@summary_statistics_table.4.4@@ & @@summary_statistics_table.4.5@@ & @@summary_statistics_table.4.6@@ & @@summary_statistics_table.4.7@@ \\
31
+ \hspace{0.1cm}White (\%)
32
+ & @@summary_statistics_table.5.0@@ & @@summary_statistics_table.5.1@@ & @@summary_statistics_table.5.2@@ & @@summary_statistics_table.5.3@@ & @@summary_statistics_table.5.4@@ & @@summary_statistics_table.5.5@@ & @@summary_statistics_table.5.6@@ & @@summary_statistics_table.5.7@@ \\
33
+
34
+ \addlinespace
35
+ \multicolumn{9}{l}{\textit{C. Charter Enrollment (Enrolled Students)}} \\
36
+ \hspace{0.1cm}Age at enrollment (years)
37
+ & @@summary_statistics_table.6.0@@ & @@summary_statistics_table.6.1@@ & @@summary_statistics_table.6.2@@ & @@summary_statistics_table.6.3@@ & & & & \\
38
+ \hspace{0.1cm}Year of enrollment
39
+ & @@summary_statistics_table.7.0@@ & @@summary_statistics_table.7.1@@ & @@summary_statistics_table.7.2@@ & @@summary_statistics_table.7.3@@ & & & & \\
40
+ \hspace{0.1cm}Years enrolled in charter
41
+ & @@summary_statistics_table.8.0@@ & @@summary_statistics_table.8.1@@ & @@summary_statistics_table.8.2@@ & @@summary_statistics_table.8.3@@ & & & & \\
42
+
43
+ \addlinespace
44
+ \multicolumn{9}{l}{\textit{D. Student Outcomes}} \\
45
+ \hspace{0.1cm}Earnings at age 25 (\$)
46
+ & & & @@summary_statistics_table.9.2@@ & @@summary_statistics_table.9.3@@ & @@summary_statistics_table.9.4@@ & @@summary_statistics_table.9.5@@ & @@summary_statistics_table.9.6@@ & @@summary_statistics_table.9.7@@ \\
47
+ \hspace{0.1cm}College by age 20 1098T (\%)
48
+ & @@summary_statistics_table.10.0@@ & @@summary_statistics_table.10.1@@ & @@summary_statistics_table.10.2@@ & @@summary_statistics_table.10.3@@ & @@summary_statistics_table.10.4@@ & @@summary_statistics_table.10.5@@ & @@summary_statistics_table.10.6@@ & @@summary_statistics_table.10.7@@ \\
49
+ \hspace{0.1cm}College by age 20 NSC (\%)
50
+ & @@summary_statistics_table.11.0@@ & @@summary_statistics_table.11.1@@ & @@summary_statistics_table.11.2@@ & @@summary_statistics_table.11.3@@ & & & & \\
51
+ \hspace{0.1cm}Employed at age 25 (\%)
52
+ & & & @@summary_statistics_table.12.2@@ & @@summary_statistics_table.12.3@@ & @@summary_statistics_table.12.4@@ & @@summary_statistics_table.12.5@@ & @@summary_statistics_table.12.6@@ & @@summary_statistics_table.12.7@@ \\
53
+
54
+ \addlinespace
55
+ Number of people
56
+ & @@summary_statistics_table.13.0@@ & & @@summary_statistics_table.13.2@@ & & @@summary_statistics_table.13.4@@ & & @@summary_statistics_table.13.6@@ & \\
57
+
58
+
59
+ \bottomrule
60
+ \end{tabular}
@@ -0,0 +1,138 @@
1
+
2
+ \clearpage
3
+ \begin{figure}[!ht]
4
+ \centering
5
+ \caption{Distribution of CREDO-Estimated Test-Score Effects Across Charter Management Organizations}
6
+ \label{fig:appendix-credo-distributions}
7
+ \begin{subfigure}[b]{\textwidth}
8
+ \centering
9
+ \caption{All Charter Management Organizations}
10
+ \label{fig:appendix-credo-test-effective}
11
+ \includegraphics[width=0.92\textwidth]{outputs/figures/f_distribution_slide4.png}
12
+ \end{subfigure}
13
+
14
+ \begin{subfigure}[b]{\textwidth}
15
+ \centering
16
+ \caption{Charter Management Organizations in the Study Sample}
17
+ \label{fig:appendix-credo-study-cmos}
18
+ \includegraphics[width=0.92\textwidth]{outputs/figures/f_distribution_slide5.png}
19
+ \end{subfigure}
20
+ \fignotes{This figure plots the distribution of CREDO-estimated test-score effects across charter management organizations in the 2017 CREDO report. Panel A plots the distribution for all 277 organizations with a reported math or reading estimate, distinguishing estimates below zero from nonnegative estimates; the vertical line reports the mean among organizations with nonnegative estimates. Panel B plots the distribution for all organizations and overlays the distribution for the 15 organizations in the study sample that can be mapped to the CREDO data; the vertical line reports the mean for the mapped study organizations. For each organization, the plotted effect is the mean of its math and reading estimates, using the available subject when only one estimate is reported. Effects are measured in standard deviations relative to neighborhood public schools. The curves are kernel density estimates using Silverman's rule-of-thumb bandwidth, and each density integrates to one. The source is the 2017 CREDO CMO report \citep{credoCMO}.}
21
+ \end{figure}
22
+
23
+ \clearpage
24
+ \begin{figure}[!ht]
25
+ \centering
26
+ \caption{Effects on College Attendance at Each Age}
27
+ \label{fig:appendix-college-by-age-placeholder}
28
+ \includegraphics[width=\textwidth]{outputs/figures/college_attendance_by_age.pdf}
29
+ \fignotes{Each point illustrates the effect of charter enrollment on NSC college attendance at the indicated age. Attendance is measured separately at each age, rather than cumulatively. Effects are in percentage points; dotted lines indicate 95\% confidence intervals. The intended specification instruments charter enrollment with an admission offer and controls for lottery risk-set fixed effects.}
30
+ \end{figure}
31
+
32
+ \clearpage
33
+ \begin{figure}[!ht]
34
+ \centering
35
+ \caption{Leave-One-Network-Out Estimates of Effects on Earnings at Age 25}
36
+ \label{fig:appendix-cmo-leave-one-out}
37
+ \includegraphics[width=\textwidth]{outputs/figures/network_leave_one_out.pdf}
38
+ \fignotes{This figure reports leave-one-network-out estimates of the effect of enrollment in the charter network to which a student applied on W-2 earnings at age 25. Each bar reports the inverse-variance-weighted average of the network-specific local average treatment effects after omitting one charter network. Bars are sorted from smallest to largest, and network identities are suppressed. The teal annotation reports the baseline pooled estimate using all networks. Network-specific effects use admission offers as instruments for applied-to-network enrollment. Earnings effects are reported in thousands of 2025 dollars and include zeros for individuals without W-2 earnings.}
39
+ \end{figure}
40
+
41
+ \clearpage
42
+ \begin{figure}[!ht]
43
+ \centering
44
+ \caption{Effects on Earnings by Age Using the Firm-Characteristics Surrogate Index}
45
+ \label{fig:appendix-earnings-by-age-firm-chars}
46
+ \includegraphics[width=\textwidth]{outputs/figures/firm_surrogate_by_age.pdf}
47
+ \fignotes{The blue series illustrates effects on observed W-2 earnings at ages 18--26; the orange series illustrates firm-characteristics surrogate predictions at ages 25--31. Dotted lines indicate 95\% confidence intervals. The intended specification instruments charter enrollment with an admission offer and controls for lottery risk-set fixed effects. Earnings are measured in 2025 dollars and include zeros for individuals without W-2 earnings.}
48
+ \end{figure}
49
+
50
+ \clearpage
51
+ \begin{table}[!ht]
52
+ \centering
53
+ \caption{Robustness to Alternative Specifications and Outcomes}
54
+ \label{tab:appendix-earnings-robustness}
55
+ \footnotesize
56
+ \begin{threeparttable}
57
+ \input{outputs/tables/output/earnings_robustness_table}
58
+ \tabnotes{Columns 1--3 reserve robustness checks for W-2 earnings at age 25: using only the initial admission offer as an instrument, using initial and waitlist offers as instruments, and implementing the pre-analysis-plan specification in long-format data, respectively. The precise pre-analysis-plan specification remains to be finalized. Column 4 reports the OLS coefficient on charter enrollment with tract fixed effects, comparing charter enrollees with the broader tract population, including non-applicants; it does not use lottery assignment. Columns 5--7 reserve estimates using the baseline instrumental-variables specification for adjusted gross income, spouse earnings, and marriage at age 25. Columns 1--3 and 5--7 report local average treatment effects and control complier means. The Control Mean row is reserved for the non-enrollee comparison-group mean in Column 4, rather than a lottery control mean or control complier mean. Income measures are in 2025 dollars; marriage is measured in percentage points. Standard errors appear below estimates in parentheses. *, **, and *** denote statistical significance at the 10\%, 5\%, and 1\% levels. Dashes indicate pending entries.}
59
+ \end{threeparttable}
60
+ \end{table}
61
+
62
+ \clearpage
63
+ \begin{table}[!ht]
64
+ \centering
65
+ \caption{Impacts of Charter Attendance on Distribution of Earnings: Actual vs. Predicted Impacts Based on Log-Normal Distribution}
66
+ \label{tab:appendix-earnings-thresholds}
67
+ \footnotesize
68
+ \begin{threeparttable}
69
+ \input{outputs/tables/output/earnings_threshold_table}
70
+ \tabnotes{This table compares actual and predicted effects of enrollment in the charter network to which a student applied on the probability of having positive W-2 earnings and earning more than \$20,000, \$50,000, and \$100,000 at age 25. Columns 1, 2, 4, and 6 report actual local average treatment effects, using an admission offer as an instrument and controlling for lottery risk-set fixed effects. Columns 3, 5, and 7 are reserved for the predicted effects at the corresponding earnings thresholds; employment has no predicted column. The intended predictions are calculated under a log-normal earnings distribution using the constant-log-shift prediction. The Control Complier Mean row reports mean untreated outcomes among applicants whose charter enrollment is changed by the admission offer in the actual-effect columns. Effects and means are reported in percentage points, and earnings thresholds are measured in 2025 dollars. Heteroskedasticity-robust standard errors for the actual effects are reported in parentheses. *, **, and *** denote statistical significance at the 10\%, 5\%, and 1\% levels, respectively. Displayed values are rounded.}
71
+ \end{threeparttable}
72
+ \end{table}
73
+
74
+ \clearpage
75
+ \begin{table}[!ht]
76
+ \centering
77
+ \caption{College-Attendance Estimates Using Form 1098-T and National Student Clearinghouse Data}
78
+ \label{tab:appendix-college-data-comparison}
79
+ \scriptsize
80
+ \begin{threeparttable}
81
+ \input{outputs/tables/output/college_data_comparison_table}
82
+ \tabnotes{This table compares estimates of the effect of enrollment in the charter network to which a student applied on college attendance by age 20 using Form 1098-T and National Student Clearinghouse data. Column 1 reports the Form 1098-T benchmark once. Column 2 reports the original pooled NSC estimate. Column 3 is reserved for the NSC estimate after aligning its sample with the Form 1098-T sample; only the NSC sample changes, and the Form 1098-T benchmark is held fixed. Attendance is cumulative through age 20. We estimate local average treatment effects of applied-to-network enrollment using admission offers as instruments and controlling for lottery risk-set fixed effects. The Control Complier Mean row reports mean untreated outcomes among applicants whose charter enrollment is changed by the admission offer. Coefficients and means are reported in percentage points. Heteroskedasticity-robust standard errors are shown below the corresponding coefficients in parentheses. *, **, and *** denote statistical significance at the 10\%, 5\%, and 1\% levels, respectively. Displayed values are rounded.}
83
+ \end{threeparttable}
84
+ \end{table}
85
+
86
+ \clearpage
87
+
88
+
89
+ \clearpage
90
+ \begin{table}[!ht]
91
+ \centering
92
+ \caption{College Mediation Analysis: Additional Specifications}
93
+ \label{tab:appendix-college-mediation-additional-specs}
94
+ \scriptsize
95
+ \begin{threeparttable}
96
+ \input{outputs/tables/output/college_mediation_additional_specs_table}
97
+ \tabnotes{This table will report additional specifications for the college mediation analysis. Columns 1 and 2 will replicate the baseline two-stage least squares specification for W-2 earnings at age 25 (\texttt{wages25}) separately for the ever-college and never-college samples, respectively. Columns 3--5 will report treatment effects on any-college completion, bachelor's degree completion, and two-year completion in the 25+ sample, using exactly the cohorts used in the wage analysis so that the college estimates used in mediation align with the earnings sample. Columns 6--8 will report treatment effects on predicted completion for any college, four-year college, and two-year college, respectively, using the detailed completion rate of the first qualifying college attended from January of the age-18 year through December of the age-21 year, rather than the coarsened rate. Observed nonattenders receive zero; incomplete follow-up remains missing. The baseline specification instruments enrollment in the charter network to which a student applied with an admission offer and controls for lottery risk-set fixed effects. The Control Complier Mean row will report the corresponding mean untreated outcomes among applicants whose charter enrollment is changed by the admission offer. Earnings effects will be measured in 2025 dollars, and college effects will be reported in percentage points. Standard errors will appear below the estimates. }
98
+ \end{threeparttable}
99
+ \end{table}
100
+
101
+ \clearpage
102
+ \begin{table}[!ht]
103
+ \centering
104
+ \caption{Lottery Sample Selection}
105
+ \label{tab:appendix-lottery-sample-selection}
106
+ \footnotesize
107
+ \begin{threeparttable}
108
+ \input{outputs/tables/output/lottery_sample_selection_table}
109
+ \tabnotes{This placeholder will report the sample size remaining after each restriction, from the overall lottery sample through the final sample in Table~\ref{tab:balance}, excluding degenerate risk sets. The restriction sequence and any additional exclusions will be confirmed against the analysis code. Dashes indicate counts to be added.}
110
+ \end{threeparttable}
111
+ \end{table}
112
+
113
+ \clearpage
114
+ \begin{table}[!ht]
115
+ \centering
116
+ \caption{CMO Details}
117
+ \label{tab:appendix-cmo-details}
118
+ \footnotesize
119
+ \begin{threeparttable}
120
+ \input{outputs/tables/output/cmo_details_table}
121
+ \tabnotes{This placeholder will list the geographic area served, lottery years, and grades of contributed lottery data for each charter management organization in the sample. CMO labels follow the existing study roster. Dashes indicate details to be added.}
122
+ \end{threeparttable}
123
+ \end{table}
124
+
125
+ \clearpage
126
+ \begin{table}[!ht]
127
+ \centering
128
+ \caption{Childhood Program Magnitude Calculations}
129
+ \label{tab:appendix-intervention-calculations}
130
+ \scriptsize
131
+ \begin{threeparttable}
132
+ \input{outputs/tables/output/intervention_calculation_table}
133
+ \tabnotes{This table reports the assumptions and key inputs used to construct the comparison-program estimates in Figure~\ref{fig:childhood-program-earnings}. The estimated-effect column reports annual earnings gains in 2025 dollars. Intervention labels and parenthetical citations correspond to the labels shown in the figure. The charter-school estimate is constructed directly from the lottery analysis and is therefore not included here.}
134
+ \end{threeparttable}
135
+ \end{table}
136
+
137
+
138
+ \clearpage