brycewang-stanford/Auto-Empirical-Research-Skills / skills/17-DAAF-Contribution-Community-daaf/dot-claude/skills/education-data-source-meps
education-data-source-meps
VerifiedMEPS — Urban Institute modeled school-level poverty (% at 100% FPL), from CCD + SAIPE (public schools, 2009-2022, 2-3yr lag). Use when FRPL is unreliable due to CEP. Consistent cross-state measurement. Public schools only.
| File | Kind | Size |
|---|---|---|
| SKILL.md | skill md | 13.8 KB |
| references/comparison-to-frpl.md | markdown | 8.0 KB |
| references/data-quality.md | markdown | 10.8 KB |
| references/data-sources.md | markdown | 8.3 KB |
| references/methodology.md | markdown | 7.2 KB |
| references/variable-definitions.md | markdown | 10.7 KB |
references/data-sources.md
View on GitHub# MEPS Data Sources
Input data used to construct Model Estimates of Poverty in Schools, including CCD, SAIPE, and ISP data.
## Overview
MEPS combines multiple federal data sources to estimate school-level poverty:
```
MEPS = f(CCD school data, SAIPE district poverty, ISP direct certification)
```
Each source contributes different information to the model.
## Common Core of Data (CCD)
### What is CCD?
The Common Core of Data is the primary federal database on public elementary and secondary schools and districts in the United States.
- **Source**: National Center for Education Statistics (NCES)
- **Coverage**: All public schools and districts
- **Frequency**: Annual
- **Years**: 1986-present
### CCD Components Used in MEPS
| CCD Dataset | MEPS Usage |
|-------------|------------|
| School Directory | School identifiers, location, type |
| School Enrollment | Total enrollment, demographics |
| School Characteristics | Grade span, charter status, locale |
| District Directory | District identifiers, boundaries |
| FRPL Counts | Where reliable, as model input |
### Key CCD Variables for MEPS
| Variable | Description | Role in MEPS |
|----------|-------------|--------------|
| `ncessch` | 12-character school ID | Primary identifier |
| `leaid` | 7-character district ID | Aggregation unit |
| `enrollment` | Total student count | Weighting |
| `free_or_reduced_price_lunch` | FRPL eligible count | Model input (where reliable) |
| `fips` | State FIPS code | Geographic identifier |
| `urban_centric_locale` | Urban/suburban/rural | Model covariate |
| `charter` | Charter school indicator | Model covariate |
| `school_type` | Regular, special ed, vocational | Sample selection |
### CCD Data Quality Notes
- **FRPL in CEP schools**: Often reported as 100%, unreliable as poverty measure
- **Missing data**: Some schools have missing enrollment or FRPL counts
- **Definitional changes**: School IDs can change with restructuring
- **Lag time**: CCD data typically 1-2 years behind current year
## Small Area Income and Poverty Estimates (SAIPE)
### What is SAIPE?
SAIPE provides annual estimates of income and poverty for states, counties, and school districts produced by the U.S. Census Bureau.
- **Source**: U.S. Census Bureau
- **Coverage**: All states, counties, and school districts
- **Frequency**: Annual
- **Years**: 1995-present
### SAIPE Methodology
SAIPE combines multiple data sources:
1. American Community Survey (ACS) - primary survey data
2. Individual income tax returns (IRS)
3. Supplemental Nutrition Assistance Program (SNAP) data
4. Decennial Census data (for benchmarking)
Uses a regression-based model to produce estimates for small areas where direct survey estimates would be unreliable.
### Key SAIPE Variables for MEPS
| Variable | Description | Role in MEPS |
|----------|-------------|--------------|
| `population_5_17_poverty` | Children 5-17 in poverty (count) | District calibration |
| `population_5_17_poverty_pct` | Percent of 5-17 in poverty | Model target alignment |
| `population_5_17` | Total children 5-17 | Denominator |
| `median_household_income` | District median income | Model covariate |
### SAIPE in MEPS
SAIPE serves as the **calibration anchor** for MEPS:
1. MEPS model generates school-level estimates
2. School estimates are adjusted so district totals match SAIPE
3. This ensures MEPS aligns with Census poverty data at the district level
### SAIPE Data Quality Notes
- **Model-based**: SAIPE itself is modeled, not direct counts
- **Margin of error**: Estimates have uncertainty, especially for small districts
- **School-age specific**: Uses 5-17 population, not total population
- **Geographic boundaries**: Based on school district boundaries which can change
## Identified Student Percentage (ISP) Data
### What is ISP?
The Identified Student Percentage is the share of students directly certified for free meals through categorical eligibility programs.
- **Source**: State education agencies (via USDA/FNS)
- **Used in**: MEPS 2.0 (December 2025)
- **Coverage**: Schools participating in National School Lunch Program
### ISP Components
Students are identified through participation in:
| Program | Typical Income Threshold | Coverage |
|---------|-------------------------|----------|
| SNAP | 130% FPL | Nationwide |
| TANF | Varies by state | Nationwide |
| Medicaid | Varies by state | Select states |
| Foster Care | N/A (categorical) | Nationwide |
| Homeless | N/A (categorical) | Nationwide |
| Migrant | N/A (categorical) | Nationwide |
| Runaway | N/A (categorical) | Nationwide |
| Head Start | 100% FPL | Where applicable |
### ISP Advantages Over FRPL
| Characteristic | FRPL | ISP |
|----------------|------|-----|
| Requires family application | Often yes | No |
| Affected by form non-response | Yes | No |
| Directly tied to program data | No | Yes |
| Available in CEP schools | Distorted | Yes |
### ISP in MEPS 2.0
MEPS 2.0 incorporates ISP as a key predictor:
- Provides poverty signal even in CEP schools
- More directly measures program participation
- Less affected by form collection policies
### ISP Data Quality Notes
- **State variation**: Which programs are linked varies by state
- **Matching quality**: Student-level matching to programs varies
- **Timing**: ISP measured at a point in time, may not reflect full year
- **Not a poverty rate**: ISP is program participation, not income verification
## Data Linkage
### How Data Sources Connect
```
CCD Schools ─────┬──── ncessch (school ID)
│
├──── leaid (district ID) ──── SAIPE Districts
│
└──── School characteristics ──── Model covariates
ISP Data ────────┴──── Matched to schools via state/school identifiers
```
### Join Keys
| Join | Key Variable(s) | Notes |
|------|----------------|-------|
| CCD School ↔ MEPS | `ncessch`, `year` | Direct match |
| CCD School ↔ CCD District | `leaid` | Many-to-one |
| CCD District ↔ SAIPE | `leaid`, `year` | Direct match |
| CCD School ↔ ISP | State-specific matching | Complex matching |
## Data Availability
### Coverage by Year
| Data Source | Available Years | Update Frequency |
|-------------|-----------------|------------------|
| CCD | 1986-2023 | Annual |
| SAIPE | 1995-2023 | Annual (December) |
| ISP | 2011-present | Annual |
| MEPS 1.0 | 2006-2019 | One-time release |
| MEPS 2.0 | Extended range | December 2025 |
> **Portal Status (Feb 2026):** MEPS 2.0 has not yet been integrated into the Education Data Portal mirrors. Portal data still reflects MEPS 1.0 (2009-2022).
### Known Data Gaps
1. **Private schools**: Not included in CCD or MEPS
2. **New schools**: May be missing first year of operation
3. **Closed schools**: May have incomplete final year data
4. **Restructured districts**: ID changes can complicate longitudinal analysis
## Accessing Source Data
### CCD Data
Available via the mirror system. See `datasets-reference.md` for canonical paths and `fetch-patterns.md` for fetch code patterns.
Key CCD datasets:
- School Directory: `ccd/schools_ccd_directory`
- School Enrollment: `ccd/schools_ccd_enrollment_{year}` (yearly)
- District Directory: `ccd/school-districts_lea_directory`
### SAIPE Data
Available via the mirror system:
- District Poverty: `saipe/districts_saipe`
Also available directly from the Census Bureau:
- https://www.census.gov/programs-surveys/saipe.html
### ISP Data
- Not directly available via the Education Data Portal mirrors
- Incorporated into MEPS 2.0 estimates
- Raw data available through USDA Food and Nutrition Service
## Data Quality Assessment
### Reliability Hierarchy
| Data Element | Reliability | Notes |
|--------------|-------------|-------|
| Enrollment counts | High | Well-collected in CCD |
| School identifiers | High | NCES standards |
| SAIPE district poverty | Moderate-High | Model-based but Census quality |
| FRPL counts (non-CEP) | Moderate | Varies by state |
| FRPL counts (CEP schools) | Low | Often 100%, meaningless |
| ISP | Moderate | Varies by state matching |
### Recommendations
1. **Trust MEPS over raw FRPL** - MEPS accounts for known data quality issues
2. **Use enrollment for weighting** - Most reliable CCD variable
3. **Check for ID changes** - When doing longitudinal analysis
4. **Note CEP status** - Helps interpret FRPL data quality
Related skills
| Skill | Stars | Updated | Works with | Status | Install |
|---|---|---|---|---|---|
| slack-gif-creatoranthropics/skillsKnowledge and utilities for creating animated GIFs optimized for Slack. Provides constraints, validation tools, and animation concepts. Use when users request animated GIFs for Slack like "make me a GIF of X doing Y for Slack." | Stars: 198k | Updated | Claude CodeCodex | Verified | Install |
| ponytailDietrichGebert/ponytailLazy senior dev mode for any coding task: the smallest change that fully solves it, and a reply a busy human understands in one read. | Stars: 194k | Updated | Claude CodeCodex | Verified | Install |
| ponytail-debtDietrichGebert/ponytailHarvest every shortcut: comment into one debt ledger, so deferrals get tracked instead of forgotten. One-shot report. | Stars: 194k | Updated | Claude CodeCodex | Verified | Install |
| ponytail-debtDietrichGebert/ponytailList every `shortcut:` comment (and older `ponytail:` ones) as a debt ledger. One-shot report, changes nothing. Use for "ponytail debt", "what did ponytail defer", "list the shortcuts", /ponytail-debt. | Stars: 194k | Updated | Claude CodeCodex | Verified | Install |