lacea.knit

class: center, middle
# "Full-Time Schools and Educational Trajectories: Evidence from High-Stakes Exams"
### Francisco Cabrera-Hernandez `$^a$`  Maria Padilla-Romo `$^b$` Cecilia Peluffo `$^c$`
#### `$^a$` CIDE and HSE `$^b$` University of Tennessee-IZA `$^c$` University of Florida 
November, 2022
###LACEA Meeting. Lima, Peru.

---

## Motivation
.center[Out of 100 children entering elementary school only 24 reach University.]

####.center[Figure: Educational Trajectory in Mexico 2003-2014]

.center[.italic[Source: Mexican Secretariat of Education, 2015.]]

---

## This Paper

- What are the effects of Full-Time Schools (FTS) on High-Stakes Test Scores, High School Placement, Preferences and Non-Cognitive Skills in the long-run?

- We use the Mexican FTS program and the Mexico City's metro area centralized high-school admission process.

- We exploit the staggered implementation of the programme to provide new evidence of FTS effects on educational trajectories.

- Focus on the differential effects in richer an poorer contexts.

---

## What do we learn?

###We provide the first long-term causal evidence of FTS:

- High-stakes **test scores increase 4.9%SD**

- Increases **On-time graduation** by 1.5pp

- Increase the probability of **applying to a top 5 high school** by 1.2pp.

- Non-Cognitive Skills are improved, **especially among women and poorer students**.

- 13 out of 100 students ever exposed to FTS are **placed in better quality high schools**

---

## What do we know?

- **Positive short-run academic achievement** *(Bellei, 2009; Almeida et al., 2016; Padilla-Romo, 2021; Cabrera-Hernandez, 2020; Aguero et al., 2021)*

- **Effects on mothers and grandmothers' labour supply** *(Padilla & Cabrera, 2019; Cabrera & Padilla 2020; Garganta and Zentner, 2020; Berthelon et. al. 2020)*

- **On child labour** *(Kozhaya and Martinez-Flores, 2022)*

- **On Mothers' fertility decisions, teenage pregnancy and juvenile crime** *(Berthelon & Kruger, 2011; Cabrera & Marquez, 2022)*

- **In the long-run delays childbearing, increase educational attainment and earnings** (Dominguez and Ruffini, 2021).

---

## The FTS programme

-   It started in 2007 in 500 schools, it reached 25,000 by 2018 or 60% of eligible.

-   For grades 1 to 9, it extends school-day from 4.5 hours to 8.

-   80% of municipalities are treated.

-   Increased inputs, money, staff and teachers, and infrastructure (e.g. kitchens).

---

## Time-Space Variation

---

## Mexico City's High School Enrollment

From 1996 there is a single application process through one high school admission exam.

####Process:
<img src="data:image/png;base64,#Image2a.png" width="100%" style="display: block; margin: auto;" />

---

## Data and Sample

-   Individual-level (IL) and school-level (SL) data from the SEP and COMIPEMS.

-   IL links information on low-stakes exams in elementary (ENLACE) and high-stakes in high school (COMIPEMS).

-   Includes preferences over high schools, demographic information, self-reported non-cognitive outcomes.

-   SL data includes participation in the FTS Program and the share of PROGRESA students in 2006.

-   Our sample covers universe of students who attended an elementary school and once between 2007 and 2013.

-   And applied to a public high school for the first time between 2010 and 2019.

---

## Descriptive Statistics

---

## Identification

-   Cohort-by-cohort variation in students' FTS enrollment during elementary school:

`$$HS_{iect}=\nu_e+\theta_{c}+\delta FTS_{iec}+X_{iect}\beta+u_{iect}$$`

-   `$HS_{iect}$` is the composite high school test score;

-   `$\nu_e$` are elementary-school fixed effects; `$\theta_{c}$` are cohort fixed effects;

-   `$FTS_{iec}$` is a variable equals to one if student `$i$` was ever exposed to full-time schooling;

-   `$X_{iect}$` are student characteristics, including gender and mothers' education;

---

## Dynamic Effects

-   We further allow our model to capture dynamic effects:

`$$HS_{iect}=\nu_e+\theta_{c}+\sum_{k\neq -1}\delta_kFTS_{ieck}+X_{iect}\beta+u_{iect}$$`

-   Our variable of interest, `$FTS_{ieckt}$`, indicates the degree of exposure to full-time schooling in elementary school.

-   For `$k\geq 0$`, this variable is equal to one if the student was enrolled in a full-time elementary school for `$k+1$` years.

-   **We compute estimators robust to dynamic and heterogeneous effects. Interaction-Weighted estimator developed by Sun and Abraham, 2021).**

---

## Results on High-Stakes High School Placement Exam

---

## Results High School Placement Exam

---

## Results in Context

-   Out of the 130,239 students ever enrolled in a full-time elementary school

-   17,306 students (13%) are placed in higher-ranked schools in their priority list vs. the absence of the FTS

-   This affects peer composition, graduation from HS, and college enrollment (e.g., Jackson, 2010; Pop-Eleches and Urquiola, 2013; Dustan et al., 2017; Estrada and Gignoux, 2017)

---

## Heterogenous effects

---

## Effects on Trajectory

---

## Effects on Non-Cognitive Outcomes

---

## Effects on Work Ethic

---

## Effects on Prefrences

---

## Effects on Test-Taking Probability

---

## Composition Effects

---

## Lee Bounds

---

## Conclusions

-   Full-time schools have long-lasting benefits for children.

-   Increase probability of graduating from middle school on time, high school admission test scores, subsequent placement, and preferences for high-quality high schools.

-   Non-Cognitive effects that concentrate on females and poorer children.

-   Effects on preferences for top high schools that concentrate in students from low-SES schools.