![Sample our Education journals, sign in here to start your access, latest two full volumes FREE to you for 14 days]({"type":"image" , "src" : "/sda/5932/TOC-Subject-Sample-2021-Education.jpg"})
Abstract
This paper examines the child development outcomes of two cohorts of children who were exposed to the same intervention at different points in time. One cohort was eligible to access playgroups during the first year of a five-year project cycle, beginning at age four. The other cohort became eligible to access these services during the third year of a five-year project cycle, beginning at age three. The younger cohort was more likely to be exposed to playgroups for longer and at more age-appropriate times relative to the older cohort. The paper finds that enrollment rates and enrollment duration in preprimary education increased for both cohorts, but the enrollment effects were larger for the younger cohort. In terms of child development outcomes, there were short-term effects at age five that did not last until age eight, for both cohorts. Moreover, the younger cohort had substantially higher test scores during the early grades of primary school, relative to the older cohort. We document the extent to which program impacts can vary as a result of differences in project implementation.
Acknowledgments
We would like to thank Dedy Junaedi, Upik Sabainingrum, Anas Sutisna, Lulus Kusbudiharjo and Mulyana for managing the fieldwork. We are grateful to Sophie Naudeau, Alaka Holla and Halsey Rogers for comments on an earlier version. The findings, interpretations, and conclusions expressed in this paper are entirely those of the authors. They do not necessarily represent the views of the International Bank for Reconstruction and Development/World Bank and its affiliated organizations, or those of the Executive Directors of the World Bank or the governments they represent.
Notes
1 During this period, the project team also worked with the government to develop a number of policies and guidelines for ECED services at the central and district levels. However, most of these were not formally promulgated until the end of the project in 2013 and do not confound the analyses presented here.
2 The 60 priority villages per district selected to participate in the project were the ones with the highest score using this formula. Proposals were expected for each of the two centers being proposed in a given village. Funding was assured by virtue of inclusion of the village in the project. As per the Project Implementation Completion and Results Report, 99.8% of the expected proposals were received and approved.
3 The Indonesia Early Childhood Education and Development project was funded through a credit from the International Development Association in the amount of $67.5 million and a grant from the government of the Kingdom of the Netherlands in the amount of $25.3 million. In addition the government of Indonesia provided $34.94 million in funding for the project.
4 The exchange rate to the dollar was around 9000 IDR to 1 dollar at the time the block grants were given. So USD 18.000 was about IDR 162 million. The minimum wage around that period was about IDR 12 million per year (Online at https://www.bps.go.id/site/resultTab). GDP per capita was about IDR 32 million per year (Source: World Development Indicators). Civil servant teachers earned about IDR 32 million Rupiah per year (Source: Authors calculations based on survey data among 2,700 civil servant teachers collected in Nov 2009).
5 Teachers were paid IDR 250,000 per month during the life of the project.
6 See Hasan et al. (Citation2013) for the playgroup schedule for a typical day.
7 This is not the fee that households had to pay. The median monthly user fee was 5,000 IDR in 2010 and 10,000 IDR in 2013. Rural households at the time reported a monthly wage of 1.7 million rupiah as per Bureau of Labor Statistics (2016).
8 See for instance the estimates quoted in Barnett (Citation1997) and Evans et al. (Citation2000).
9 Districts included in the project were selected on the basis of a formula described in Appendix 3 of Hasan et al. (Citation2013). Districts included in the evaluation were willing to randomize participation. However, within districts villages were selected on the basis of a formula described in Appendix 4. Appendix 7, 8 and 9 test the evaluation design, the internal validity and external validity of the design. These show no differences that would confound the analysis.
10 Appendix 7–9 in Hasan et al. (Citation2013) document that these villages are well balanced on a range of observable characteristics.
11 Only 32 children from the two cohorts are siblings to each other.
12 Data were collected in 2013 and again in 2016.
13 We use the short-form to match previously published estimates in Brinkman et al. (Citation2017) with which we compare this paper’s findings. Results using the long-form of the EDI are qualitatively similar and are available upon request.
14 A comparison of assets ownership by households in the evaluation sample with that of the rural sub-sample of the SUSENAS (a nationally representative household survey) suggests average rates of asset ownership and education levels are by and large similar between the two. See also Hasan et al. (Citation2013).
15 Results by batch are reported in Appendix Tables 1–3.
16 In both cases—wealth and parenting practices—we split the sample into those above the mean and those below the mean.
17 Appendix Table 4 provides results which show that the differences between groups—either poor versus non-poor or low versus high parenting practices are not statistically different from each other.
18 Our experience in the field and the data on enrollment histories does indeed underscore the fact that parents view playgroups and kindergartens as substitutes and do not adhere to the ages of eligibility for the different services.
19 Since all of our impact estimates focus on intent-to-treat, these take-up figures do not affect the interpretation of our results. The small non-zero estimates of enrollment in project centers among children from non-project villages are possible in those few cases where households lived near enough to a treatment community and project playgroup. However, this was rare.
20 See Appendix Table 4.
21 See Appendix Table 5.
22 Appendix Table 6.
23 See Appendix Figure 2.
24 See Appendix B for the empirical strategy used to estimate the impacts for the 4 year-old cohort.
25 These are obtained using the approach described in Appendix B. The items used to construct scores for the 4 year-old cohort are identical to those used for the 1 year-old cohort.
26 Our test statistic to examine whether the estimates from the two cohorts are statistically different is assuming
Since
is typically positive, our assumption yields a conservative estimate of the test statistic.
27 As a robustness check, we also examine whether treatment effects on the EDI and test scores vary by exposure to playgroups at the appropriate ages among the younger cohort. Results are reported in Appendix Tables 5 and 6. Overall, we find larger treatment effects for social competence and emotional maturity among children enrolled in playgroups for one to two years at ages 3–4 relative to their peers enrolled in playgroup for less than one year at ages 3–4. These results are consistent with the argument we present here that longer exposure to playgroup at the appropriate age explains part of the different impacts we observe between the younger and older cohorts.
28 The 2012 GDP per capita in PPP terms was US$4,876. In our calculations of rate of return, we assume that rural wages are a third of this number.
29 See Appendix Table 7. The median monthly fee was 5,000 IDR in 2010 and 10,000 IDR in 2013. Rural households at the time reported a monthly wage of 1.7 million IDR (Bureau of Labor Statistics, 2016).
30 See Appendix Table 8.
31 By the time the younger cohort was old enough to enroll in project playgroups, many more playgroups were charging fees. The introduction of fees was a direct response to the project funding coming to an end and the centers needing to devise an alternative sustainability strategy (Hasan et al., Citation2019).