Religious Education in the Age of Mass Schooling (Figures)

Figures Figure 1: INPRES Exposure and Islamic Schooling – Raw Summary (a) Elementary Madrasa .03 1968 birth cohort: first fully exposed to SD INPRES .015 .01 1962 cohort last unexposed to SD INPRES Islamic junior secondary (overall) high-INPRES districts low-INPRES districts .02 Islamic elementary (overall) (b) Junior Secondary Madrasa .005 high-INPRES districts low-INPRES districts .025 .02 .015 .01 .005 1950 1960 1970 year of birth 1980 1990 1950 (c) Senior Secondary Madrasa .07 Highest education Islamic (overall) Islamic senior secondary (overall) 1970 year of birth 1980 1990 (d) Any Level Madrasa high-INPRES districts low-INPRES districts .03 1960 .02 .01 high-INPRES districts low-INPRES districts .06 .05 .04 .03 .02 0 1950 1960 1970 year of birth 1980 1990 1950 1960 1970 year of birth 1980 1990 Notes: This figure reports mean Islamic school completion rates over time for districts with above-median (high) and below-median (low) INPRES intensity from 1973–1978. INPRES intensity is defined as the number of SD INPRES schools constructed from 1973-78 per 1,000 children in 1971. The rates are computed for cohorts from 1950 to 1994, pooling across annual Susenas data from 2012 to 2018, and they indicate whether the final level of education is elementary Islamic in panel (a), junior secondary Islamic in panel (b), senior secondary Islamic in panel (c), and any level Islamic in panel (d). These measures are computed over all individuals, while the corresponding Appendix Figure A.1 reports results conditional on individuals having completed the given level of education. The outcomes in panels (a)–(c) are the same as those in panel (a) of Table 2, and panel (d) is the same as the outcome in columns 3–4 of Table 3. The cohorts in gray are those that would have fully completed primary schooling before the SD INPRES program rolled out in 1973. The vertical dotted line captures the first cohort, born in 1968, that would have been fully exposed to SD INPRES given that they would have been 6 years old just prior to school construction ensuing. The cohorts born between 1963 to 1967 correspond to the partially-exposed cohorts. See Section 4.1 for further discussion of these distinctions across cohorts. Figure 2: INPRES Exposure and Islamic Schooling – Effects by Cohort Dep. Var.: Elementary Madrasa (a) Short-Run (b) Long-Run .002 .002 INPRES x birth year: β 90% confidence interval INPRES x birth year: β 90% confidence interval .001 0 -.001 0 -.002 -.002 -.003 -.004 1958 1960 1962 1964 1966 year of birth 1968 1970 1972 1950 1955 1960 1965 1970 1975 1980 1985 1990 1995 year of birth Dep. Var.: Secondary Madrasa (c) Short-Run (d) Long-Run .006 INPRES x birth year: β 90% confidence interval INPRES x birth year: β 90% confidence interval .02 .004 .002 0 .015 .01 .005 0 -.002 -.005 1958 1960 1962 1964 1966 year of birth 1968 1970 1972 1950 1955 1960 1965 1970 1975 1980 1985 1990 1995 year of birth Notes: This figure reports age-specific estimates of β in equation (1) based on annual Susenas data from 2012 to 2018. INPRES intensity is defined as the number of SD INPRES schools constructed from 1973-78 per 1,000 children in 1971. The dependent variable in panels (a) and (b) is an indicator equal to one if the individual’s final year of schooling was completed in an Islamic elementary school. Panels (c) and (d) are for an Islamic secondary school. Panels (a) and (c) correspond to the original cohort specification capturing variation in exposure to SD INPRES: fully-exposed born 1968– 1972 (dark gray), partially-exposed born 1963–1967 (white), and unexposed born 1957–1962 (light gray). Panels (b) and (d) expand exposed and unexposed windows to 1950 and 2000, respectively, though we only include among later cohorts those with completed schooling. The 1957 cohort serves as the reference age, given age fixed effects, in both the shortand long-run specifications. All specifications include survey year dummies, district of birth dummies and year of birth dummies interacted with the 1971 children population, the 1971 enrollment rate, and exposure to the water and sanitation program in the district of birth. The dashed lines correspond to 90% confidence intervals with standard errors clustered by district of birth. Figure 3: INPRES Exposure, Total Schooling and Islamic Education – Effects by Cohort Dep. Var.: Total Years of Schooling (a) Short-Run (b) Long-Run .25 .6 INPRES x birth year: β 90% confidence interval INPRES x birth year: β 90% confidence interval .2 .15 .1 .05 0 .4 .2 0 -.05 -.2 1958 1960 1962 1964 1966 year of birth 1968 1970 1972 1950 1955 1960 1965 1970 1975 1980 year of birth 1985 1990 1995 Dep. Var.: Any Level Madrasa (c) Short-Run (d) Long-Run .02 INPRES x birth year: β 90% confidence interval INPRES x birth year: β 90% confidence interval .004 .002 0 .015 .01 .005 0 -.002 -.005 1958 1960 1962 1964 1966 year of birth 1968 1970 1972 1950 1955 1960 1965 1970 1975 1980 1985 1990 1995 year of birth Notes: This figure reports results from the same specification as in Figure 2, looking here at total years of schooling in panels (a) and (b) and any Islamic schooling in panels (c) and (d). See the notes to Figure 2 for further details. Figure 4: INPRES Intensity and Entry of Islamic Schools New schools per 1,000 children (a) Elementary Madrasa (b) Junior Secondary Madrasa .008 .006 INPRES x semi-decade: β 90% confidence interval INPRES x semi-decade: β 90% confidence interval .006 .004 .002 .004 .002 0 0 -.002 -.002 1950 1955 1960 1965 1970 1975 1980 1985 1990 1995 1950 1955 (c) Senior Secondary Madrasa 1965 1970 1975 1980 1985 1990 1995 1980 1985 1990 1995 1990 1995 (d) Pesantren .008 INPRES x semi-decade: β 90% confidence interval .003 INPRES x semi-decade: β 90% confidence interval 1960 .002 .001 0 .006 .004 .002 0 -.001 1950 1955 1960 1965 1970 1975 1980 1985 1990 1995 1950 1960 1965 1970 1975 (f) Madrasa as % of new schools .02 .08 .015 .06 INPRES x semi-decade: β 90% confidence interval INPRES x semi-decade: β 90% confidence interval (e) Madrasa Diniyah 1955 .01 .005 .04 .02 0 0 -.02 -.005 1950 1955 1960 1965 1970 1975 1980 1985 1990 1995 1950 1955 1960 1965 1970 1975 1980 1985 Notes: This figure reports semi-decade-specific estimates of β in equation (2) on a balanced district–year panel. INPRES intensity is defined as the number of SD INPRES schools constructed from 1973-78 per 1,000 children in 1971. The dependent variable measures the number of elementary madrasa (panel a), junior secondary madrasa (b), senior secondary madrasa (c), madrasa diniyah (Islamic afternoon schools) (d), pesantren (Islamic boarding schools across all levels) (e), and madrasa in a–c as a share of total schools (Islamic, non-Islamic private, and secular public) (f) established by semi-decade and by district per 1,000 children in 1971. The 1965-1969 period is the reference period given district fixed effects. The gray shading captures the INPRES construction period from 1973-78. The dot corresponds to the period-specific β, and the bars correspond to 90% confidence intervals with standard errors clustered by district. All specifications include district fixed effects and year fixed effects interacted with the 1971 children population, the 1971 enrollment rate, and exposure to the water and sanitation program. Figure 5: INPRES Intensity and Electoral Support for Islam vs. the State (b) Islam (United Development Party, PPP) (a) State (Golkar, Suharto’s Party) 1973: beginning of SD INPRES First exposed cohorts vote in 1987 .08 Democratic era .06 .06 .04 .04 INPRES x election year: β 90% confidence interval INPRES x election year: β 90% confidence interval .08 .02 0 -.02 -.04 -.06 .02 0 -.02 -.04 -.06 -.08 -.08 1971 1977 1982 1987 1992 1999 2004 2009 1971 1977 1982 1987 1992 1999 2004 2009 (c) Islam vs. State (PPP minus Golkar) .08 INPRES x election year: β 90% confidence interval .06 .04 .02 0 -.02 -.04 -.06 -.08 1971 1977 1982 1987 1992 1999 2004 2009 Notes: This figure reports legislative-election-year-specific estimates of β in equation (2) on a balanced district–year panel. INPRES intensity is defined as the number of SD INPRES schools constructed from 1973-78 per 1,000 children in 1971. The dependent variable measures vote shares for Golkar, the party of Suharto and the New Order regime (panel a), the Islamic opposition party/ies (panel b), and the difference in vote shares between the two (panel c). In 1971, there were four Islamic parties that we group together, but from 1973 onward, the regime only allowed a single umbrella Islamic party, the United Development Party or PPP. The 1971 election was the last just prior to SD INPRES and serves as the reference election given district fixed effects. The gray area captures elections conducted under the New Order regime. The elections in 1987 and 1992 are the first in which INPRES-exposed cohorts would have been eligible to vote. The elections from 1999 onward took place after the fall of Suharto when the country democratized and both secular and Islamic parties proliferated. The bars correspond to 90% confidence intervals with standard errors clustered at the district level. All specifications include district fixed effects and year fixed effects interacted with the 1971 children population, the 1971 enrollment rate, and exposure to the water and sanitation program.