Making a big gift to draw other donors is only useful if the charity is highly effective

This article first published in the Financial Times in December 2017.

The practice of donors offering to “match” gifts made to charities they support has become common in the charity world. In the run-up to Christmas, many have been offering to double donations. In the recent Challenge Campaign, for instance, money given to selected charities was doubled by the Big Give, set up by Sir Alec Reed, founder of Reed, the recruitment firm.

Others schemes propose different deals: one donor was offering to give $2 for every $1 given on Friday to Charity Navigator, a charity rating agency; while PayPal will add 1 per cent to all donations made through its platform until New Year.

The UK’s Department for International Development does it too. Its UK Aid Match offered up to £5m from the UK aid budget to eight charities.

The theory is that the match attracts new donations. If your aim is helping your chosen charity to raise money, is it a good idea to offer to match other people’s contributions? The rigorous evidence suggests that it is. 

One experiment showing this, by economists Dean Karlan and John List, then at Yale and the University of Chicago respectively, ran in the US in 2005. Over 50,000 donors to a US civil liberties non-governmental organisation were randomly assigned to receive one of several versions of a fundraising letter. One was a standard solicitation with no match. Another version said — truthfully — that for every dollar that somebody gave, an anonymous donor would give another dollar: a 100 per cent match.

Offering a match increased the chance that each recipient would give by 19 per cent, and the size of the average gift was 22 per cent higher.

Steffen Huck and Imran Rasul of University College London found much the same in a 2006 experiment with the Bavarian State Opera House in Munich. There, donors who offered a 100 per cent match were 14 per cent more likely to give than those who were not offered any match, and the average gift was 24 per cent greater. A third study in the US in 2014 found the same: offering a match raised more than half as much again as a standard solicitation.

But what if you vary the terms of the match? Here, some really interesting differences emerge.

The US experiment also had groups of donors who were offered other inducements: for every dollar they gave, the anonymous donor would give $2 (like Charity Navigator’s donor); for another group, the match was $3 for every $1 that they gave. The level of the match turned out not to matter: the more generous matches did not entice more donors to give, nor to give more each.

By contrast, in the German experiment another group was offered a 50 per cent match. That deal had precisely the same effect on people’s likelihood of giving as the 100 per cent offer. But, surprisingly, each person gave more: €101 on average, versus €92 for the people offered the 100 per cent match. The researchers concluded that the large donation and match “crowded out” or discouraged other donors.

The German experiment also suggests a better route to increase the total raised. A fourth group in that study was told that the anonymous donor would match any donations above €50 (up to €60,000). That device raised the most — a little above the amount raised by the 100 per cent match letter. It achieved that mainly by increasing the number of donors who gave above the threshold to qualify for the match.

The 2014 US study found a method which is more effective still — although I hate it. Those researchers mailed various request letters to 40,000 US donors. Some were a standard “ask”, some offered a 100 per cent match and some said that an anonymous donor was covering “all the overhead costs associated with raising the needed donations”. Notice that this is not covering all the overheads to run the actual work — just to raise the funds. The “overhead” letter raised two and a half times as much as the standard ask, and nearly twice what the match offer raised — mainly by persuading nearly twice as many people to donate.

My concern about this last device is that it probably reinforces the fabulously unhelpful and dangerous notion that charities’ overheads are somehow separate from their “real work”, and optional. They aren’t, as we discussed last time. Overheads enable the “real work” and skimping on them is often a false economy.

Charitable giving is pretty flat as a proportion of GDP, at least in the US and UK. Fundraising seems to be virtually a zero-sum game. In other words, devices such as offering to match donations do not seem to increase the amount that people give overall: they just reallocate it. In terms of improving life for the intended beneficiaries, that’s fine only if the charities which can offer matches are better than those that don’t. If you’re going to offer a significant sum to a charity in order for it to attract other donors, superior effectiveness is the key criterion to check.