Lost Causal: Debunking Myths About Causal Analysis in Philanthropy
Why this work is in the frame
A frame that forgets how it found something cannot be audited. These are the routes that admitted this work.
Bibliographic record
Abstract
What if philanthropic evaluations told us that changes in the world had occurred, as well as how and why they occurred, including in what ways foundation funding and grantees contributed to those changes? What if evaluations made change pathways more visible, tested hypotheses and assumptions, and generated new insights based on what happened in the “black box” of systems change strategies? This type of learning comes from causal analysis — inquiry that explores cause-andeffect relationships. Yet currently in philanthropy, particularly for strategies and initiatives that feature high complexity, few evaluations use robust techniques for understanding causality. Instead, philanthropic evaluation tends to rely on descriptive measurement and analysis. These descriptions often are rich, meaningful, and in-depth, but they remain merely descriptions nonetheless. This article challenges the myths that hold us back from causal inquiry, allowing us to embrace curiosity, inquiry, and better knowing, even (or especially) if it means learning that our assumptions and theories do not hold up. We argue that philanthropy more frequently needs to examine causal relationships, using a growing suite of methodological approaches that make this possible in complex systems. Causal methodologies can challenge and strengthen the often uncontested beliefs that underlie philanthropic interventions, while offering evidence about enabling contexts and system drivers. Strong causal analysis considers not only the funder’s model and assumptions, but also the beliefs others hold about how and why change occurs, opening the door to more equitable and less biased ways of understanding change.
Fetched live from OpenAlex and de-inverted. Abstracts are not stored in this database: the inverted indexes are 8.6 GB of the frame’s 9.3 GB of text, and the host has 13 GB free.
Full frame distilled prediction
Teacher imitationNot calibrated prevalence, not ground truth. Human validation pending. Learned from the 10,348 direct Codex labels and 10,348 direct Gemma labels. Candidate is the union of thresholded teacher heads; consensus is their intersection. These outputs are machine_predicted_unvalidated and are not human labels or direct frontier model labels.
Codex and Gemma teacher scores by category
| Category | Codex | Gemma |
|---|---|---|
| Metaresearch | 0.007 | 0.001 |
| Meta-epidemiology (narrow) | 0.000 | 0.000 |
| Meta-epidemiology (broad) | 0.000 | 0.000 |
| Bibliometrics | 0.000 | 0.004 |
| Science and technology studies | 0.000 | 0.000 |
| Scholarly communication | 0.000 | 0.000 |
| Open science | 0.000 | 0.000 |
| Research integrity | 0.000 | 0.000 |
| Insufficient payload (model declined to judge) | 0.025 | 0.003 |
Machine scores (provisional)
The two teacher heads of the student model, read on this work. A score orders the frame for review; it never asserts a category, and the validation status ships verbatim with every row.
Baseline scores from an immature model (maturity gate not passed, 7 training rounds). Scores rank; they never assert a category.
score_only:v0-immature-baseline · verbatim from the scoring run: score_only means the number may rank works, and no category label ships from it