New study screens charitable giving research for findings that actually replicate
Howard Lake | 11 July 2026 | News
A new working paper from the Center for Philanthropic Studies at VU Amsterdam sets out to do for the evidence base on charitable giving what Boyan Slat’s Ocean Cleanup set out to do for the seas: sift a vast and growing body of material for what is worth keeping.
Titled “The Science of Philanthropy Cleanup” and led by Professor Rene Bekkers, it asks a question fundraisers rarely get a clear answer to. Of the thousands of studies published on why people give, which findings actually hold up when other researchers test them again?
The scale of the problem is stark. In 2021 alone, Google Scholar indexed around 89,000 studies mentioning donations, and the volume of giving research has grown more than tenfold since 2000. No individual can keep track.
Advertisement
Fundraising meets bibliometrics
Bibliometrics is “the quantitative application of mathematics and statistics to study published scholarly literature”. It was part of my Masters in Information Science in 1992 when I was a fundraiser at Amnesty International UK. We focused then on the challenges of searching and current awareness in chemistry literature, as academic fundraising or philanthropy research was considerably more manageable by an individual. I’m pleased that the volume of fundraising research is now a challenge.
To find the reliable signal in the noise of all these studies, Bekkers and colleagues screened the literature in three ways:
- for retractions
- for statistical errors
- and for replications, meaning studies that repeat an earlier finding using fresh data or methods.
Retractions turn out to be rare in giving research. The authors found 36 retracted studies between 2007 and 2022, a rate of 0.069%, and the trend is downward. That could mean the field is trustworthy. It could equally mean giving research attracts less scrutiny than other disciplines, so fewer flaws are ever caught.
The statistical picture is less comforting. Of the papers the team could check automatically, 46% contained at least one statistical error, and in 13% the error changed whether a result counted as statistically significant. Both rates are higher than in psychology or communication science. More tellingly, 70% of those decision-changing errors happened to favour the study’s own hypothesis, a skew that points to researchers, consciously or not, nudging results toward the answer they wanted.
Replications are the heart of the paper, and they too are rare, fewer than 1% of all publications. About 60% reached the same conclusion as the study they repeated, which leaves a substantial minority of “known” findings that do not stand up.
What holds up
For fundraisers, the most useful section is the audit of which specific tactics survive independent replication.
The levers that hold up are, broadly, the practical ones. Directly asking people to give works: face-to-face and verbal solicitations reliably raise more money. Lowering the price of giving works, through matched funding and tax incentives, although the effect of tax reliefs is smaller than earlier studies claimed, including in the UK.
Framing a subsidy as a match rather than a rebate raises more than the equivalent rebate.
“Unit asking”, prompting donors to consider helping one person before deciding how many to help, holds up.
And efficacy matters: donors give more when overhead costs are lower, and when their gift is more likely to push a campaign over its target, which is the logic behind seed funding.
What doesn’t
Several widely repeated claims do not survive scrutiny, and some fundraising staples look distinctly unreliable.
The “watching eyes” effect, the idea that images of eyes make people more generous, does not reliably replicate.
Public recognition of donors does not dependably increase giving and in some studies reduces it, with donors sometimes preferring anonymity.
Thank-you gifts were found to reduce giving among existing donors rather than encourage it.
The oft-cited “identifiable victim effect“, that a single named beneficiary outpulls statistics, failed to hold up in field experiments and a registered report.
Social information nudges along the lines of “most people give X” produced far weaker or null effects on repeat testing. And moral appeals do not reliably shift behaviour: people with stronger prosocial values give more, but reminding people of those values does not make them give more.
Notably, the authors found no sign of publication bias within the replications themselves; null results were published at similar rates to positive ones. But they warn that meta-analyses built only on published studies are likely to overstate effects, because replications consistently find weaker results than the originals.
The practical takeaway is that the sector should treat headline findings with more caution before building campaigns on them. The authors call for more replication, for journals to adopt registered reports, accepting studies for publication before the results are known so that null findings are not buried, and for tougher statistical checks in peer review.
For a sector that leans on “what the evidence says” to justify creative and strategic choices, it is a timely reminder that not all published evidence is equal, and that some of the most quotable findings are among the least reliable. The paper is currently a preprint on OSF and has not yet been peer reviewed.
Sharing the research on LinkedIn, Professor Bekkers commented:
“We find that the more replicable mechanisms are material costs and solicitation, while replications indicate that material benefits and reputation do not work as we wrote previously. While replications and retractions are rare in the field of research on giving, we show how science can be self-correcting through replications.”
The Science of Philanthropy Cleanup was written by Rene Bekkers, Max Littel, Stephanie Koolen-Maas PhD, Dr Hoda Fahimpour, Arthur Gautier, Wahideh Achbaria, and published on 7 July 2026 by the Center for Philanthropic Studies, Department of Sociology, VU Amsterdam.
Related posts
UK slips down CAF World Giving Index but is third most generous country for giving money
UK no longer one of world’s most generous countries
Wealthy people in Britain could give £2.8bn more to charity a year, research suggests
€43 million lost to charitable causes over 10 years in Ireland
About Howard Lake
Howard Lake is consultant editor of UK Fundraising. He founded the site in 1994 and successfully sold it in 2022. As director of Giving X Ltd he is exploring growing giving on a massive scale.
He is the founder of Fundraising Camp and co-founder of GoodJobs.