The Simplest Prediction in Economics
Supply curves slope upward, which is the first lesson of introductory microeconomics and one of the most reliably confirmed relationships in all of social science: offer more money for something and more people will provide it, whether the commodity is wheat or labor or taxi rides at two in the morning. Blood donation is one of the places where this iron law falls apart.
In 1970, British social scientist Richard Titmuss published The Gift Relationship, arguing that paying blood donors would not increase supply but destroy it. The claim was radical and rested on no controlled experiment, only a comparison between the unpaid British blood system and the partly compensated American one that showed higher rates of contaminated blood in the paid supply. Kenneth Arrow reviewed the book sympathetically but called the crowding-out hypothesis "really an empirical question," and for 35 years nobody ran the experiment that could settle it.
What Happened When Someone Finally Tested It
In 2008, economists Carl Mellström of the University of Gothenburg and Magnus Johannesson of the Stockholm School of Economics published the first randomized field experiment designed to test Titmuss's claim directly. They recruited approximately 262 Swedish university students and randomly assigned them to one of three conditions: an unpaid group invited to undergo a health screening as a prerequisite for blood donation, a paid group that received SEK 50 (roughly $7) for agreeing to donate, and a third group offered the same SEK 50 but with the option to redirect it to a children's cancer charity instead of keeping it.
Among men, the payment made no statistically significant difference across conditions. Among women, it was devastating: in the unpaid group 52% agreed to become blood donors, compared to just 30% in the paid group, which means the payment didn't merely fail to increase supply but cut it nearly in half. The charity-option group provided the most revealing data point of all, because when women could redirect the $7 to charity rather than pocket it themselves, 53% agreed to donate, virtually identical to the unpaid baseline. The money itself was not the problem; what mattered was what accepting cash for blood signaled about the kind of person you were, and whether the act remained a gift or became a transaction.
A Pattern That Repeats Across Fifty Experiments
The blood donation result does not stand alone in the literature. In a now-famous 2000 field experiment at ten daycare centers in Haifa, Israel, Uri Gneezy and Aldo Rustichini introduced a monetary fine for parents who picked up their children late, and standard deterrence theory predicted fewer late arrivals; instead, late pickups roughly doubled because parents treated the fine as a price that purchased the right to be late without guilt, and when the fine was removed twelve weeks later lateness never returned to its original baseline because the social norm governing punctuality had been permanently damaged by its conversion into a market transaction.
By 2012, Samuel Bowles and Sandra Polania-Reyes had cataloged over 50 experimental studies in the Journal of Economic Literature documenting the same fundamental pattern: explicit economic incentives reduced the very behavior they were designed to encourage, with fines crowding out compliance, bonuses crowding out effort, and payments crowding out generosity across cultures and across domains ranging from charitable giving to environmental conservation to workplace productivity. When Bruno Frey and Felix Oberholzer-Gee surveyed Swiss citizens in 1997 about siting a nuclear waste repository near their town, about 50% accepted without compensation out of civic duty, but offering substantial financial compensation dropped acceptance to 25% because the money reframed a question of shared obligation as a commercial bargain.
The Strongest Counterargument
The most serious challenge to the crowding-out thesis comes from within behavioral economics itself. Nicola Lacetera and Mario Macis, using both observational data from the American Red Cross and a controlled field experiment, showed that material rewards, particularly in-kind gifts like gift cards and T-shirts rather than cash, can increase blood donation rates without crowding out intrinsic motivation. Their study of nearly 100,000 blood drives found a 15% increase in donations at drives that offered economic incentives worth between $5 and $15, suggesting that it may be specifically cash, not all incentives, that corrodes motivation. A $10 gift card and a $10 bill carry identical purchasing power, but they send fundamentally different social signals about why the donor showed up. The Mellström experiment itself hinted at exactly this distinction: the charity option proved that the incentive's framing, not its dollar value, determined whether crowding out occurred.
What We Didn't Prove
The experiment used a specific population of Swedish university students in a cultural context where unpaid blood donation carries strong norms of civic duty and where Sweden maintains one of the world's highest voluntary donation rates. The crowding-out effect may be weaker in cultures where transactional relationships carry less stigma or where baseline donation rates are already low, leaving less intrinsic motivation for a payment to displace. The sample was modest at roughly 262 participants, and the gender asymmetry (significant crowding out among women but not men) has not received a fully satisfying explanation in subsequent work. The experiment also measured willingness to donate rather than completed donations, so whether the effect survives the full donation process remains unknown. Perhaps most critically, the payment was trivially small at about $7, and a sufficiently large payment might overwhelm the identity cost and restore the supply-curve prediction, consistent with Gneezy and Rustichini's separate finding that large financial rewards increased effort among charity solicitors while small ones reduced it below the unpaid baseline.
The Bottom Line
When people perform an act because they believe it is right, offering them money can remove the very reason they do it. A Swedish field experiment showed that paying potential blood donors $7 cut female participation nearly in half — not because the money was too little, but because it transformed an act of generosity into a purchase. Over 50 experiments across multiple countries and contexts confirm the pattern: human decency is not a commodity, and pricing it as one can permanently damage the supply.
What You Can Do
If you manage volunteers, donors, or any group that operates partly on intrinsic motivation, audit your incentive structure for the signals it sends rather than the dollar amounts it delivers, because a gift given as recognition functions differently from a payment given as compensation even when both are worth the same amount. Consider offering a charity-redirect option alongside any monetary reward, which the Mellström experiment showed can fully neutralize crowding out. And if you are ever tempted to introduce fines for unwanted behavior in a context governed by social norms (late daycare pickups, missed medical appointments, community violations), remember that the Haifa experiment showed the damage is not just immediate but permanent, because once a social obligation becomes a market price, removing the price does not restore the obligation.