Awesome-Agent-Skills-for-Empirical-Research education-data-source-nacubo

install

source · Clone the upstream repo

git clone https://github.com/brycewang-stanford/Awesome-Agent-Skills-for-Empirical-Research

Claude Code · Install into ~/.claude/skills/

T=$(mktemp -d) && git clone --depth=1 https://github.com/brycewang-stanford/Awesome-Agent-Skills-for-Empirical-Research "$T" && mkdir -p ~/.claude/skills && cp -r "$T/skills/17-DAAF-Contribution-Community-daaf/dot-claude/skills/education-data-source-nacubo" ~/.claude/skills/brycewang-stanford-awesome-agent-skills-for-empirical-research-education-data-so-2aa833 && rm -rf "$T"

manifest: skills/17-DAAF-Contribution-Community-daaf/dot-claude/skills/education-data-source-nacubo/SKILL.md

source content

NACUBO Data Source Reference

NACUBO endowment data for U.S. college and university endowments (~650 institutions). Portal mirror provides only 7 market-value columns (total endowment, per-FTE endowment, year-over-year change; 2012-2022). Use when analyzing institution-level endowment size or trends. Full investment returns, asset allocations, spending rates, and governance data require separate NACUBO study access. For comprehensive endowment coverage across all institutions, use IPEDS finance data.

The NACUBO-Commonfund Study of Endowments (NCSE) is the most comprehensive annual survey of U.S. college and university endowments, covering ~650 institutions representing over $870 billion in assets. The Education Data Portal mirrors a limited subset (7 market-value columns); full investment, allocation, spending, and governance data requires direct NACUBO access.

CRITICAL: Portal Mirror Scope

The Education Data Portal mirrors contain only 7 columns of endowment data, focused on market values. Full investment data (returns, asset allocations, spending rates, governance) requires access to the complete NACUBO study.

Portal columns:
year
,
unitid
,
inst_name_nacubo
,
fips
,
endow_total
,
endow_per_fte
,
endow_chg_mktval
Value Encoding:
Variable Portal Format Notes
fips
Integer (
6
= California)
Not string abbreviations ("CA")
year
Integer (
2022
)
Fiscal year ending year
endow_total
Float64 (USD)
Full dollar amount, NOT thousands
endow_chg_mktval
Float64 (decimal fraction)
0.059
= 5.9% change, NOT
5.9
Missing data
null
NOT -1, -2, -3
See
./references/variable-definitions.md
for complete encoding tables.

What is NACUBO?

The National Association of College and University Business Officers (NACUBO) is a nonprofit professional organization representing chief administrative and financial officers at higher education institutions.

Collector: NACUBO with Commonfund Institute (currently); TIAA partnership FY2018-2022
Coverage: ~650 participating colleges, universities, and affiliated foundations (voluntary)
Scope: Investment returns, asset allocations, spending rates, governance practices, market values
Frequency: Annual survey (September-December collection, February publication)
Available years: 1974-present (50+ years of study); Portal mirror covers 2012-2022 (11 years)
Primary identifier:
```
unitid
```
(IPEDS 6-digit institution ID)
FY2024 total: 658 participants representing $873.7 billion in assets

Reference File Structure

File	Purpose	When to Read
`endowment-study.md`	NCSE methodology, participation, history	Understanding data source context
`endowment-metrics.md`	Market values, returns, spending rates	Interpreting performance data
`variable-definitions.md`	Key variables, size categories, institution types	Building queries, filtering data
`data-quality.md`	Coverage limitations, data caveats	Assessing data reliability
`asset-allocation.md`	Investment categories, alternative strategies	Analyzing portfolio composition

Decision Trees

What endowment data do I need?

Endowment research topic?
├─ Market values / fund size
│   ├─ Individual institution values → NCSE public tables
│   ├─ Size distribution → ./references/variable-definitions.md
│   └─ Historical trends → ./references/endowment-metrics.md
├─ Investment returns
│   ├─ One-year returns → ./references/endowment-metrics.md
│   ├─ Multi-year returns (3/5/10/25-year) → ./references/endowment-metrics.md
│   └─ Returns by size/type → ./references/variable-definitions.md
├─ Asset allocation
│   ├─ Traditional vs alternative → ./references/asset-allocation.md
│   ├─ Private equity/VC allocation → ./references/asset-allocation.md
│   └─ ESG/responsible investing → ./references/asset-allocation.md
├─ Spending and distributions
│   ├─ Spending rates → ./references/endowment-metrics.md
│   ├─ Spending purposes (aid, faculty, etc.) → ./references/endowment-metrics.md
│   └─ Budget contribution → ./references/endowment-metrics.md
└─ Data quality concerns
    ├─ Coverage/participation → ./references/data-quality.md
    ├─ Self-reported data issues → ./references/data-quality.md
    └─ Comparability across years → ./references/data-quality.md

Which data source should I use?

Comparing NACUBO vs IPEDS for endowment data?
├─ Need ALL institutions → Use IPEDS (mandatory reporting)
├─ Need investment returns → Use NACUBO (IPEDS doesn't have)
├─ Need asset allocations → Use NACUBO (IPEDS doesn't have)
├─ Need spending rates → Use NACUBO (IPEDS doesn't have)
├─ Need just market values → Either works
│   ├─ More institutions → IPEDS
│   └─ More detail → NACUBO
└─ Need governance/policies → Use NACUBO (IPEDS doesn't have)

Understanding study name changes

Which study by year?
├─ FY2023-present → NACUBO-Commonfund Study of Endowments (NCSE)
├─ FY2018-FY2022 → NACUBO-TIAA Study of Endowments (NTSE)
├─ FY2009-FY2017 → NACUBO-Commonfund Study of Endowments (NCSE)
└─ Before FY2009 → NACUBO Endowment Study (NES)

Quick Reference: NACUBO Data

Portal Coverage Note: The Education Data Portal mirrors contain only 7 columns focused on endowment market values. Full investment data (returns, asset allocations, spending rates, governance) requires access to the complete NACUBO study. See "Full Report Access" section below.

Portal Mirror Columns (Verified)

These are the only columns available in the Portal mirror (

nacubo/colleges_nacubo_endow

Column	Type	Description	Nulls
`year`	Int64	Fiscal year ending year (2012-2022)	0
`unitid`	Int64	IPEDS 6-digit institution ID	0
`inst_name_nacubo`	String	NACUBO institution name	0
`fips`	Int64	State FIPS code (1-56, territories: 72, 78)	1
`endow_total`	Float64	Total endowment market value in USD (full dollars, NOT thousands)	0
`endow_per_fte`	Float64	Endowment per FTE student in USD	1,260 (~15%)
`endow_chg_mktval`	Float64	Year-over-year market value change as decimal fraction (0.059 = 5.9%)	3,972 (~48%)

Important format notes:

```
endow_total
```
is in full USD (e.g.,
```
1.05e9
```
= $1.05 billion), not in thousands
```
endow_chg_mktval
```
is a decimal fraction, not a percentage (multiply by 100 for display)
Missing data uses null, not coded values (-1/-2/-3)

# Correct: convert decimal fraction to percentage for display
df = df.with_columns(
    (pl.col("endow_chg_mktval") * 100).alias("pct_change")
)

# Correct: endow_total is already in full USD
df.filter(pl.col("endow_total") > 1_000_000_000)  # Endowments over $1B

Key Metrics Available

Metric	Description	In Portal Mirror?	Granularity
Market Value	Total endowment assets (FMV)	Yes ( `endow_total` )	Institution-level
Market Value Per FTE	Endowment per FTE student	Yes ( `endow_per_fte` )	Institution-level
Market Value Change	Year-over-year change	Yes ( `endow_chg_mktval` )	Institution-level
Investment Return	Time-weighted return, net of fees	No (full study only)	1, 3, 5, 10, 25-year periods
Effective Spending Rate	Annual withdrawal as % of market value	No (full study only)	Institution type, size
Asset Allocation	Portfolio composition by asset class	No (full study only)	Size category, type
New Gifts	Donations added to endowment	No (full study only)	Aggregate by size/type
Budget Support	Endowment % of operating budget	No (full study only)	Size category, type

Key Identifiers

ID	Format	Level	Example	Notes
`unitid`	Integer (6-digit)	Institution	`166027`	IPEDS institution ID; primary join key
`inst_name_nacubo`	String	Institution	`Harvard University`	NACUBO version of name
`fips`	Integer (1-56)	State	`25` (Massachusetts)	State FIPS code

Endowment Size Categories

Category Code	Range	FY24 Count	% of Participants
1	Over $5 Billion	~25	~4%
2	$1 Billion to $5 Billion	~107	~16%
3	$501 Million to $1 Billion	~77	~12%
4	$251 Million to $500 Million	~97	~15%
5	$101 Million to $250 Million	~161	~24%
6	$51 Million to $100 Million	~111	~17%
7	Under $50 Million	~80	~12%

Institution Types

Type Code	Type Name	Description
1	Private	Private colleges and universities
2	Public	Public colleges and universities
3	IRF	Institutionally Related Foundations
4	Combined	Combined endowment/foundation

Typical Return Ranges (Historical)

Period	Average Range	Notes
1-year	-8% to +31%	High volatility
3-year	3% to 15%	Less volatile
5-year	5% to 12%	More stable
10-year	6% to 9%	Most commonly cited
25-year	6% to 9%	Long-term benchmark

Missing Data Codes

Code	Meaning	When Used
`null`	Not reported / missing	Standard for all NACUBO variables

Note: Unlike other Education Data Portal sources (CCD, CRDC, etc.), NACUBO data uses null values for missing data rather than coded values like -1, -2, -3. NACUBO is a voluntary survey with simpler missing data patterns; the Portal preserves null rather than applying coded values.

# Correct: Check for null
df.filter(pl.col("endow_per_fte").is_null())

# Incorrect: Checking for -1/-2/-3 (these don't exist in NACUBO)
# df.filter(pl.col("endow_per_fte") == -1)  # Won't find anything

Data Access

Datasets for NACUBO are available via the mirror system. See

datasets-reference.md

for canonical paths,

mirrors.yaml

for mirror configuration, and

fetch-patterns.md

for fetch code patterns.

Dataset	Type	Years	Path	Codebook
Endowments	Single	2012-2022	`nacubo/colleges_nacubo_endow`	`nacubo/codebook_colleges_nacubo_endowments`

Codebooks are

.xls

files co-located with data in all mirrors. Use

get_codebook_url()

from

fetch-patterns.md

to construct download URLs:

url = get_codebook_url("nacubo/codebook_colleges_nacubo_endowments")

Truth Hierarchy: When interpreting variable values, apply this priority:

Actual data file (what you observe in the parquet/CSV) -- this IS the truth

Live codebook (.xls in mirror) -- authoritative documentation, may lag

This skill documentation -- convenient summary, may drift from codebook

If this documentation contradicts the codebook, trust the codebook. If the codebook contradicts observed data, trust the data and investigate.

Public NACUBO Tables

Free tables available at: https://www.nacubo.org/Research/2024/Public-NCSE-Tables

Includes:

Participating institutions by market value
Average/median returns by size and type
Asset allocation summaries
Spending rate summaries

Full Report Access

Participants: Free access
NACUBO Members: $250
Non-members: $1,500
Academic researchers: Contact Commonfund Institute

Common Pitfalls

Pitfall	Issue	Solution
Using NCES-style codes	Portal uses integer FIPS, not string state abbreviations	Use integer fips codes (1-56)
Expecting coded missing values	NACUBO in Portal uses nulls, not -1/-2/-3	Check for null, not negative codes
Treating `endow_chg_mktval` as percentage	Column is a decimal fraction (0.059), not a percentage (5.9)	Multiply by 100 for display: `pl.col("endow_chg_mktval") * 100`
Treating `endow_total` as thousands	Column is in full USD (1.05e9 = $1.05B)	No division needed; value is already in dollars
Comparing to IPEDS values	NACUBO is voluntary (~650-813), IPEDS is mandatory (~6,000+)	Note sample differences in analysis
Year interpretation	FY2024 = July 2023 - June 2024	Align fiscal year definitions carefully
Assuming comprehensive coverage	Only ~650-813 of ~4,000+ institutions participate	Use IPEDS for population-level analysis
Ignoring partner changes	TIAA (2018-22) vs Commonfund (other years) may change methodology	Check methodology for specific years
Expecting investment returns in Portal	Portal only has market values, not returns/allocations/spending	Use full NCSE study for investment analysis

Key Caveats

Issue	Impact	Mitigation
Portal subset	Only 7 columns mirrored (market values only)	Full study required for returns, allocations, spending, governance
Voluntary participation	~650-813 of ~4,000+ institutions per year	Use IPEDS for comprehensive coverage
Self-reported data	Unverified accuracy	Cross-reference with IPEDS where possible
Selection bias	Larger, well-resourced institutions overrepresented	Analyze by size category
Partner changes	TIAA (2018-22), Commonfund (other years)	Check methodology changes
Fiscal year timing	July 1 - June 30	Match to other data sources carefully
Declining participation	813 in 2012 to 652 in 2021 in Portal data	Account for sample size changes in trend analysis

Common Use Cases

Research Question	Primary Variables	Reference File
How do endowment returns vary by size?	Return rates, size category	`endowment-metrics.md`
What drives endowment spending decisions?	Spending rate, spending policy	`endowment-metrics.md`
How are endowments invested?	Asset allocation percentages	`asset-allocation.md`
Which schools have the largest endowments?	Market value, institution name	Public tables
How has endowment performance changed?	Historical return series	`endowment-metrics.md`
What portion of budgets come from endowments?	Budget support percentage	`endowment-metrics.md`

Related Data Sources

Source	Relationship	When to Use
`education-data-source-ipeds`	Complementary; mandatory reporting covers all institutions	Need all institutions or just market values
`education-data-source-scorecard`	Complementary; student outcomes	Linking endowment size to student outcomes
`education-data-explorer`	Parent discovery skill	Finding available endpoints
`education-data-query`	Data fetching	Downloading parquet/CSV files

Topic Index

Topic	Reference File
Study methodology	`./references/endowment-study.md`
Partnership history	`./references/endowment-study.md`
Participation criteria	`./references/endowment-study.md`
Investment returns	`./references/endowment-metrics.md`
Spending rates	`./references/endowment-metrics.md`
Market values	`./references/endowment-metrics.md`
Size categories	`./references/variable-definitions.md`
Institution types	`./references/variable-definitions.md`
Variable list	`./references/variable-definitions.md`
Asset classes	`./references/asset-allocation.md`
Alternative investments	`./references/asset-allocation.md`
ESG investing	`./references/asset-allocation.md`
Selection bias	`./references/data-quality.md`
Coverage gaps	`./references/data-quality.md`
Comparability issues	`./references/data-quality.md`

Variable	Portal Format	Notes
`fips`	Integer ( `6` = California)	Not string abbreviations ("CA")
`year`	Integer ( `2022` )	Fiscal year ending year
`endow_total`	Float64 (USD)	Full dollar amount, NOT thousands
`endow_chg_mktval`	Float64 (decimal fraction)	`0.059` = 5.9% change, NOT `5.9`
Missing data	`null`	NOT -1, -2, -3