Skip to content

DAAF-Contribution-Community/daaf / .claude/skills/education-data-source-pseo

education-data-source-pseo

Verified

PSEO — Census data linking graduates to employment via LEHD wage records. Earnings percentiles at 1/5/10 years post-graduation by institution, degree, and CIP. Use for graduate earnings analysis. Coverage changes by release; Census V4.14.1 (2026Q2, posted 2026-08-25) lists 1,117 institution rows and partners in 37 states plus D.C. (plus online-only WGU).

references/industry-flows.md

View on GitHub
# Industry Flows

What industries graduates enter: employment by NAICS sector.

> **Portal Integer Encoding:** The `industry` column contains 2-digit NAICS integers (11-92, 99). Null values indicate no industry data for that row.

## Contents

- [Overview](#overview)
- [NAICS Sector Codes](#naics-sector-codes-portal-integer-encoding)
- [Industry Flow Variables](#industry-flow-variables)
- [Analysis Patterns](#analysis-patterns)
- [Matching Programs to Industries](#matching-programs-to-industries)
- [Limitations](#limitations)

## Overview

The Employment Flows tabulations include industry data showing which sectors employ graduates:

- **NAICS Sector**: 2-digit North American Industry Classification System
- **20 sectors** covering all economic activity
- Available at 1, 5, and 10 years post-graduation

Industry flows help answer:
- What industries employ graduates from a specific program?
- Are graduates working in fields related to their training?
- How do career patterns differ by major?

## NAICS Sector Codes (Portal Integer Encoding)

| Code | Sector |
|------|--------|
| `11` | Agriculture, Forestry, Fishing and Hunting |
| `21` | Mining, Quarrying, and Oil and Gas Extraction |
| `22` | Utilities |
| `23` | Construction |
| `31`-`33` | Manufacturing |
| `42` | Wholesale Trade |
| `44`-`45` | Retail Trade |
| `48`-`49` | Transportation and Warehousing |
| `51` | Information |
| `52` | Finance and Insurance |
| `53` | Real Estate and Rental and Leasing |
| `54` | Professional, Scientific, and Technical Services |
| `55` | Management of Companies and Enterprises |
| `56` | Administrative and Support and Waste Management |
| `61` | Educational Services |
| `62` | Health Care and Social Assistance |
| `71` | Arts, Entertainment, and Recreation |
| `72` | Accommodation and Food Services |
| `81` | Other Services (except Public Administration) |
| `92` | Public Administration |
| `99` | Unclassified/Federal (PSEO-specific) |

> **Data Note:** In Portal parquet files, the `industry` column is **String** type (not Int64) because some NAICS sectors span ranges (e.g., `"31-33"` for Manufacturing). All rows have a non-null `industry` value -- rows representing earnings-only data (without industry breakdown) appear with an aggregate industry value.

### Key Sectors by Common Graduate Fields

| Graduate Field | Primary Industries |
|----------------|-------------------|
| Engineering | 31-33 (Manufacturing), 54 (Prof Services), 23 (Construction) |
| Computer Science | 51 (Information), 54 (Prof Services), 52 (Finance) |
| Business | 52 (Finance), 54 (Prof Services), 44-45 (Retail) |
| Education | 61 (Educational Services), 92 (Public Admin) |
| Health/Nursing | 62 (Health Care) |
| Biology/Chemistry | 54 (Prof Services), 62 (Health Care), 31-33 (Manufacturing) |
| Arts/Humanities | 51 (Information), 61 (Education), 71 (Arts/Entertainment) |

## Industry Flow Variables

### Employment by Industry (Portal Schema)

In Portal data, employment counts are accessed via `employed_grads_count_f`, filtered by `years_after_grad` and `industry`:

| Portal Variable | Description |
|-----------------|-------------|
| `employed_grads_count_f` | Employed graduates count (filter by `years_after_grad` for Y1/Y5/Y10) |
| `industry` | NAICS sector (String type, e.g., `"54"`, `"31-33"`) |
| `years_after_grad` | `1`, `5`, or `10` |

### Querying by Industry

```python
import polars as pl

# Fetch PSEO data
df = fetch_from_mirrors("pseo/colleges_pseo_2020")

# Employment in Professional Services sector ("54"), 1 year post-graduation
prof_services = df.filter(
    (pl.col("industry") == "54")
    & (pl.col("unitid") == 228778)       # UT Austin
    & (pl.col("degree_level") == 5)       # Bachelor's
    & (pl.col("cipcode") == 11)           # Computer Science
    & (pl.col("years_after_grad") == 1)
    & (pl.col("employed_grads_count_f") > 0)
)
```
```r
# Fetch PSEO data.
# fetch_from_mirrors() is a Python helper; in R, build the URL from the mirror
# root in mirrors.yaml and read directly.
# Mirror failover: see `education-data-query/references/fetch-patterns.md` (R pattern).
config <- yaml::read_yaml("mirrors.yaml")
mirror <- config$mirrors[[1]]
url <- paste0(mirror$root_url, "/", mirror$vintage$hf_revision, "/", "pseo/colleges_pseo_2020", ".", mirror$format)
# NOTE: illustrative only — mirror parquet files are Polars-written and may
# declare string_view columns, so a plain read can fail under R arrow
# ("cannot handle Array of type <utf8_view>"). Real fetch scripts must use the
# view-safe parquet read from `education-data-query/references/fetch-patterns.md`.
df <- arrow::read_parquet(url)

# Employment in Professional Services sector ("54"), 1 year post-graduation
prof_services <- df |> filter(
  industry == "54",
  unitid == 228778,          # UT Austin
  degree_level == 5,          # Bachelor's
  cipcode == 11,              # Computer Science
  years_after_grad == 1,
  employed_grads_count_f > 0
)
```

### Multiple Industries

```python
# Filter to specific NAICS sectors
tech_sectors = df.filter(
    pl.col("industry").is_in(["51", "54", "52"])
    & (pl.col("years_after_grad") == 1)
    & (pl.col("employed_grads_count_f") > 0)
)

# Get distribution across all industries for a program
all_industries = df.filter(
    (pl.col("unitid") == 228778)
    & (pl.col("cipcode") == 11)
    & (pl.col("years_after_grad") == 1)
    & (pl.col("employed_grads_count_f") > 0)
).select("industry", "employed_grads_count_f")
```
```r
# Filter to specific NAICS sectors
tech_sectors <- df |> filter(
  industry %in% c("51", "54", "52"),
  years_after_grad == 1,
  employed_grads_count_f > 0
)

# Get distribution across all industries for a program
all_industries <- df |> filter(
  unitid == 228778,
  cipcode == 11,
  years_after_grad == 1,
  employed_grads_count_f > 0
) |> select(industry, employed_grads_count_f)
```

## Analysis Patterns

### Industry Distribution for a Program

Show where graduates from a specific program work:

```
Computer Science (CIP=11) Bachelor's from UT Austin, Y1 Employment:

┌─────────────────────────────────────────────┬──────────┬─────────┐
│ Industry (NAICS)                            │ Employed │ Share   │
├─────────────────────────────────────────────┼──────────┼─────────┤
│ 54 - Professional, Scientific, Technical   │ 280      │ 35%     │
│ 51 - Information                            │ 200      │ 25%     │
│ 52 - Finance and Insurance                  │ 120      │ 15%     │
│ 31-33 - Manufacturing                       │ 80       │ 10%     │
│ 61 - Educational Services                   │ 40       │ 5%      │
│ Other sectors                               │ 80       │ 10%     │
├─────────────────────────────────────────────┼──────────┼─────────┤
│ Total Employed                              │ 800      │ 100%    │
└─────────────────────────────────────────────┴──────────┴─────────┘
```

### Career Trajectory Analysis

Compare Y1 vs Y5 vs Y10 industry distribution:

```python
# Pseudo-code: How industry distribution changes over time
for year in [1, 5, 10]:
    industry_dist = get_employment_by_naics(institution, cip, year)
    print(f"Year {year}: {industry_dist}")
```
```r
# Pseudo-code: How industry distribution changes over time
for (yr in c(1, 5, 10)) {
  industry_dist <- get_employment_by_naics(institution, cip, yr)
  cat(sprintf("Year %d: %s\n", yr, industry_dist))
}
```

**Example finding**: Engineering graduates may start in manufacturing (31-33) but shift toward management (55) or professional services (54) by Year 10.

### Cross-Program Comparison

Compare industry destinations for different programs at same institution:

| Program (CIP) | Top Industry | % in Top Industry |
|---------------|--------------|-------------------|
| Computer Science (11) | 54 (Prof Services) | 35% |
| Nursing (51.38) | 62 (Health Care) | 92% |
| Business (52) | 52 (Finance) | 28% |
| Education (13) | 61 (Educational Services) | 75% |

## Matching Programs to Industries

### Expected vs. Actual Industry Match

Some programs have clear industry matches; others are more diffuse:

| Match Type | Example Programs | Characteristics |
|------------|------------------|-----------------|
| **Tight match** | Nursing, Education | 70%+ in single industry |
| **Moderate match** | Engineering, Accounting | 40-70% in 2-3 industries |
| **Diffuse** | Business, Communications | Spread across many industries |

### Field-Specific Expectations

**Health Sciences (CIP 51)**
- Expected: 62 (Health Care and Social Assistance)
- Typical match rate: 80-95%

**Engineering (CIP 14)**
- Expected: 31-33 (Manufacturing), 54 (Prof Services), 23 (Construction)
- Typical combined rate: 60-80%

**Business/Management (CIP 52)**
- Expected: 52 (Finance), 54 (Prof Services), 55 (Management)
- Distribution typically more varied

**Computer Science (CIP 11)**
- Expected: 51 (Information), 54 (Prof Services)
- Tech workers also common in 52 (Finance), 31-33 (Manufacturing)

## Limitations

### Granularity Constraints

| Limitation | Impact |
|------------|--------|
| 2-digit NAICS only | Cannot distinguish sub-industries |
| No occupation data | NAICS ≠ job function |
| Sector boundaries | Some industries span multiple codes |

### Interpretation Challenges

| Issue | Consideration |
|-------|---------------|
| Headquarters vs. work site | Large companies classified by HQ industry |
| Multi-industry firms | Conglomerates may be classified inconsistently |
| Federal employment | May appear in NAICS 99 (unclassified) |
| Contractor classification | IT workers at banks → Finance (52) not Information (51) |

### Missing Data

Flows counts are noise-infused (geometric mechanism) and post-processed rather than suppressed by a fixed <30 rule; Census documents the <30 suppression rule for earnings cells, and division-level flows may be suppressed when too much of the flow comes from states with missing data.

Industry data can also be missing when:
- Program or institution has insufficient coverage

### Example: Complete Industry Profile

For Business Administration Bachelor's graduates (hypothetical):

```
Institution: State University
CIP: 52 (Business, Management, Marketing)
Degree Level: 5 (Bachelor's)
Cohort: "2016-2018"

Industry Distribution Evolution:

Year 1:
┌─────────────────────────────────────┬────────┐
│ Industry                            │ Share  │
├─────────────────────────────────────┼────────┤
│ 52 - Finance and Insurance          │ 22%    │
│ 54 - Professional, Scientific       │ 18%    │
│ 44-45 - Retail Trade                │ 15%    │
│ 56 - Administrative Support         │ 12%    │
│ 72 - Accommodation and Food         │ 10%    │
│ Other                               │ 23%    │
└─────────────────────────────────────┴────────┘

Year 5:
┌─────────────────────────────────────┬────────┐
│ Industry                            │ Share  │
├─────────────────────────────────────┼────────┤
│ 52 - Finance and Insurance          │ 28%    │
│ 54 - Professional, Scientific       │ 22%    │
│ 55 - Management of Companies        │ 10%    │
│ 44-45 - Retail Trade                │ 8%     │
│ 56 - Administrative Support         │ 8%     │
│ Other                               │ 24%    │
└─────────────────────────────────────┴────────┘

Insight: Graduates shift from retail/food service toward 
finance, professional services, and management over time.
```

## Industry Code Reference

For complete NAICS definitions, see:
- [NAICS Code Labels (CSV)](https://lehd.ces.census.gov/data/schema/latest/label_industry.csv)
- [Census Bureau NAICS Reference](https://www.census.gov/naics/)
finishing-a-development-branchobra/superpowersUse when implementation is complete, all tests pass, and you need to decide how to integrate the workStars: 322kUpdated Claude CodeCodexVerifiedInstall
systematic-debuggingobra/superpowersUse when encountering any bug, test failure, or unexpected behavior, before proposing fixesStars: 322kUpdated Claude CodeCodexVerifiedInstall
test-driven-developmentobra/superpowersUse when implementing any feature or bugfix, before writing implementation codeStars: 322kUpdated Claude CodeCodexVerifiedInstall
skill-creatoranthropics/skillsCreate new skills, modify and improve existing skills, and measure skill performance. Use when users want to create a skill from scratch, edit, or optimize an existing skill, run evals to test a skill, benchmark skill performance with variance analysis, or optimize a skill's description for better triggering accuracy.Stars: 197kUpdated Claude CodeCodexVerifiedInstall