Religious Education in the Age of Mass Schooling Appendix A (Figures)

A Further Empirical Results Figures Figure A.1: INPRES Exposure and Islamic Schooling | Years of Schooling (b) Junior Secondary Madrasa (a) Elementary Madrasa .12 Islamic junior secondary (conditional) Islamic elementary (conditional) .03 .025 .02 .1 .08 .06 .015 1950 1960 1970 year of birth 1980 1990 1950 (c) Senior Secondary Madrasa 1960 1970 year of birth 1980 1990 (d) Any Level Madrasa .08 Highest education Islamic (conditional) Islamic senior secondary (conditional) .08 .06 .04 .02 .06 .04 .02 1950 1960 1970 year of birth 1980 1990 1950 1960 1970 year of birth 1980 1990 Notes: This figure reports analogous results to those in Figure 1, but here we restrict to individuals that completed the given years of education corresponding to the level at hand: 6 for elementary, 9 for junior secondary, and 12 for senior secondary. The outcomes in panels (a)–(c) are the same as those in panel (b) of Table 2, and panel (d) is the same as the outcome in columns 5–6 of Table 3. Figure A.2: INPRES Intensity and Entry of Private and Public Islamic Schools Elementary Madrasa (a) Private (b) Public .002 .004 INPRES x semi-decade: β 90% confidence interval INPRES x semi-decade: β 90% confidence interval .006 .002 0 .001 0 -.001 -.002 1950 1955 1960 1965 1970 1975 1980 1985 1990 1995 1950 1955 1960 1965 1970 1975 1980 1985 1990 1995 1980 1985 1990 1995 1980 1985 1990 1995 Junior Secondary Madrasa (d) Public .006 .0015 .004 .001 INPRES x semi-decade: β 90% confidence interval INPRES x semi-decade: β 90% confidence interval (c) Private .002 0 .0005 0 -.0005 -.002 1950 1955 1960 1965 1970 1975 1980 1985 1990 1995 1950 1955 1960 1965 1970 1975 Senior Secondary Madrasa (f) Public .003 .0015 .002 .001 INPRES x semi-decade: β 90% confidence interval INPRES x semi-decade: β 90% confidence interval (e) Private .001 0 -.001 .0005 0 -.0005 1950 1955 1960 1965 1970 1975 1980 1985 1990 1995 1950 1955 1960 1965 1970 1975 Notes: This figure disaggregates the Islamic school entry outcomes in Figure 4 (panels a–c) into private and public Islamic schools. The latter comprise 8% of all Islamic schools. All dependent variables are normalized by the 1971 child population. Figure A.3: INPRES Intensity and Entry of Private non-Islamic Schools (b) Primary (a) All .01 .005 INPRES x semi-decade: β 90% confidence interval INPRES x semi-decade: β 90% confidence interval .01 0 -.005 .005 0 -.005 -.01 -.01 1950 1955 1960 1965 1970 1975 1980 1985 1990 1995 1950 (c) Junior Secondary 1955 1960 1965 1970 1975 1980 1985 1990 1995 1985 1990 1995 (d) Senior Secondary .004 .003 INPRES x semi-decade: β 90% confidence interval INPRES x semi-decade: β 90% confidence interval .003 .002 .001 .002 .001 0 0 -.001 -.001 1950 1955 1960 1965 1970 1975 1980 1985 1990 1995 1950 1955 1960 1965 1970 1975 1980 Notes: This figure reports semi-decade-specific estimates of β in equation (2) on a balanced district–year panel. The dependent variable measures: the number of private non-Islamic schools across all levels (panel a), elementary (b), junior secondary (c), and senior secondary (d). Appendix C describes how we isolate secular schools among all private schools in the MEC registry. Figure A.4: Islamic School Attendance vs. Completion in Repeated Cohorts Highest Schooling Islamic, share of cohort .15 Enrolled in 2012 Completed in 2018 .1 .05 0 1994 1995 1996 1997 1998 1999 2000 Notes: This figure uses repeated observations of identical cohorts in the 2012 and 2018 Susenas rounds. We focus on cohorts born between 1994 and 1999 which were young enough to have been enrolled in school in 2012 but old enough to have completed high school by 2018. Black bars show Islamic attendance rates measured in 2012 while grey bars indicate Islamic completion rates in 2018.