Juan E. Saavedra

Are Educational Vouchers Only Redistributive

It is unclear if vouchers increase educational productivity or are purely redistributive, benefiting recipients by giving them access to more desirable peers at others’ expense. To examine this, we study an educational voucher programme in Colombia which allocated vouchers by lottery. Among voucher applicants to vocational schools, lottery winners were less likely to attend academic secondary schools and thus had peers with less desirable observable characteristics. Despite this, lottery winners had better educational outcomes. In this population, vouchers improved educational outcomes through channels beyond redistribution of desirable peers. We discuss potential channels which may explain the observed effects.

Educational Impacts and Cost-Effectiveness of Conditional Cash Transfer Programs in Developing Countries: A Meta-Analysis

We meta-analyze for impact and cost-effectiveness 94 studies from 47 conditional cash transfer (CCT) programs in low- and middle-income countries worldwide, focusing on educational outcomes that include enrollment, attendance, dropout and school completion. To conceptually guide and interpret the empirical findings of our meta-analysis, we present a simple economic framework of household decision-making that generates predictions, all else constant, for the association between certain program context and design characteristics, and impact estimates. We also present a simple model for the analysis of program costs, using it to compute cost-effectiveness estimates for a subsample of CCT programs. For all schooling outcomes, we find strong support for heterogeneity in impact, transfer-effectiveness, and cost-effectiveness estimates. Our meta-analytic results of impact and transfer-effectiveness estimates provide support to some – but not all – of the predictions from the household decision-making model.

Nudging Youth to Develop Savings Habits: Experimental Evidence Using SMS Messages

School financial education programs are often ineffective because the information they deliver is not relevant to imminent decisions, not understood and not reinforced. We present results from a field experiment in Colombia that articulates financial information to actual financial decisions. We randomly allocate 10,000 youth accountholders to four experimental conditions: i) twelve monthly messages with educational content aimed at helping youth determine spending priorities and using savings heuristics to achieve savings goals, ii) twelve monthly savings reminders which make savings goals salient, iii) twenty-four semi-monthly savings reminders and iv) control. Simple financial information delivered through text messages improves savings outcomes among youth. Consistent with a limited attention hypothesis, youth accountholders who received reminders almost doubled their total savings in their bank account relative to control accountholders during the twelve-month period during which we sent the SMS. The financial education treatment, in contrast, did not increase savings relative to control, suggesting that limited information may not be the most binding barrier that prevents youth from saving. Savings effects of reminders last even six months after youth stop receiving messages, the latest we can observe. JEL classification: D91 (Intertemporal Consumer Choice; Lifecycle Models and Saving), E21 (Consumption; Savings), I29 (Education, others)

Disentangling Disadvantage: Can We Distinguish Good Teaching from Classroom Composition?

This paper focuses on the use of teacher value-added estimates to assess the distribution of effective teaching across students of varying socioeconomic disadvantage. We use simulation methods to examine the extent to which different commonly used teacher-value added estimators accurately capture both the rank correlation between true and estimated teacher effects and the distribution of effective teaching across student characteristics in the presence of classroom composition effects. Varying the amount of teacher sorting by student characteristics, the within-teacher variability in classroom composition, and the amount of student learning decay, we compare aggregated residuals, teacher random effects, and teacher fixed effects models estimated in both levels and gains, with and without controls for classroom composition. We find that models estimated in levels more accurately capture the rank correlation between true and estimated teacher effects than models estimated in gains, but levels are not always preferable for recovering the correlation between teacher value-added and student achievement. For recovering that correlation, aggregated residuals models appear preferable when sorting is not present, though fixed effects models perform better in the presence of sorting. Because the true amount of sorting is never known, we recommend that analysts incorporate contextual information into their decisions about model choice.

Targeted Remedial Education: Experimental Evidence from Peru

An outstanding challenge in education is improving learning among low-achieving students. We present results from the first randomized experiment of an inquiry-based remedial science-education program for low-performing elementary students in the setting of a developing country. At 48 low-income public elementary schools in Lima, Peru and surrounding areas, third-grade students scoring in the bottom half of their science classes were selected at random to receive up to 16 remedial sessions of 90 minutes each during the school year. Control-group compliance with assignment (no extra tutoring) was close to perfect. Treatment-group compliance was roughly 40 percent, or five to six remedial sessions—a 4 to 5 percent increase in total science instruction time over the school year. Despite the low-intensity treatment, students assigned to the remedial sessions scored 0.12 standard deviations higher on a science endline test. But all improvements were concentrated among boys, for whom gains were 0.22 standard deviations. Remedial education does not produce within-student spillovers to math, or spillovers on other students.

Medium- and Long-Term Educational Consequences of Alternative Conditional Cash Transfer Designs: Experimental Evidence from Colombia

We show that three Colombian conditional cash transfer (CCT) programs for secondary school improve educational outcomes after eight years, depending on the stipend structure. Forcing families to save a portion of the transfers until they make enrollment decisions for the next year increases on-time enrollment in secondary school, reduces dropout rates, and promotes tertiary enrollment. Traditional stipends improve on-time enrollment and high school exit exam completion rates. These differences between stipends are statistically significant due to the effects on older students. Finally, a stipend that directly incentivizes tertiary enrollment promotes on-time enrollment in secondary school and in lower quality tertiary institutions.