By Steve Kelman

Blog archive

Carrots or sticks?

benchmarks (Bakhtiar Zein/

In my blog last weekI discussed a new book, The Power of Bad, that presented evidence that we pay more attention to bad things that happen than to good ones, an underlying reason why the media emphasizes bad news about government. There is a chapter in the book, called "Heaven or Hell: Prizes vs. Penalties," that applies the argument to a completely different question -- whether rewards or punishments will influence people's behavior more. This topic too is an important one, both in our everyday lives and in the design of public policies, so I am talking about that chapter here as well.

I am also writing about this because it is one of the few examples I remember of something I've read that reversed my opinion about an important issue. I definitely see myself temperamentally as a softie, much more inclined to carrots than sticks. This is not just me -- the book's authors, John Tierney and Roy Baumeister, write that "Dangling a carrot is more pleasant than wielding a stick."

Yet I'm now reconsidering a very fundamental attitude I have.

The chapter starts with a broad-sweep account of the development of Christianity over the centuries in the U.S. New England was settled by Puritans, who morphed into Congregationalists (still big in my home town of Concord, Massachusetts). "Their clergymen were well-educated gentlemen, not charismatic rabble-rousers," Tierney and Roy Baumeister write. "They preached elegant, cerebral sermons based on the theology they had studied at Harvard and Yale, where rationalism was prized and emotionalism disdained. They had been taught to see God as distant and abstract, a vaguely benevolent deity nothing like the wrathful figure in the Old Testament who condemned sinners to perdition."

These clergy were challenged in the eighteenth century by the "fire and brimstone" preachers of the so-called Great Awakening, such as one who warned audiences not to be lulled by theologians who denied "the eternity of hell-torments" and stated that the denial of hell promoted "infidelity and profaneness." With the arrival of such preachers, Methodism grew in the U.S. from a tiny sect to, by 1850, the largest denomination in the country. The terrors of hell won out over the delights of heaven.

The book proceeds from this historical example to modern research and public policy practice. Psychologists have discovered that while rats had to be exposed to rewards several times to influence their behavior, and the reinforcement needed to be immediate, rats exposed to a punishment changed their behavior after a single exposure, and continued to do so even several hours later.

Dozens of studies show that people are much more likely to respond to public health messages (such as getting tested for cancer) if the message is based on fear. Schools provide fascinating public policy examples. Studies of students in Germany and Florida who have been held back and need to repeat a grade showed that, despite worries that this would destroy students' self-esteem, their grades rose the next year (as in fact did their self-esteem) and in subsequent years. Recent years have seen the development of the "no excuses" movement in school, which revolted against social promotion and "everyone gets a trophy."

"Instead of stressing collaborative projects, they regularly test each student's individual competence," the authors note. "Instead of encouraging 'mutually supportive' group discussions where students are called on only if they raise their hands, the teachers challenge students by 'cold-calling' them to answer a question."

At the Success Academy -- a non-profit chain of New York schools that uses these principles, enrolls 90% minority students and admits students by lottery -- 95% of students pass state proficiency exams, compared to one-third of minority students in ordinary schools. A study of which charter schools do the best showed that it was schools using a no-excuses approach.

So the book's argument is that just like bad things affect us more than good ones, so does punishment affect us more than rewards. Even if rewards are comforting and punishments seem cruel, they may typically not be the best way to influence behavior. This material was eye-opening for me. I need to think about it more to decide which of my own behaviors and attitudes I need to change.

Posted by Steve Kelman on Feb 03, 2020 at 12:40 PM


  • Government Innovation Awards
    Government Innovation Awards -

    Congratulations to the 2021 Rising Stars

    These early-career leaders already are having an outsized impact on government IT.

  • Acquisition
    Shutterstock ID 169474442 By Maxx-Studio

    The growing importance of GWACs

    One of the government's most popular methods for buying emerging technologies and critical IT services faces significant challenges in an ever-changing marketplace

Stay Connected