Psychology study discredits psychology studies; why you shouldn't be surprised

Last week, The Reproducibility Project published their study finding that only 36% of 100 psychology experiments could be reproduced.  “The field of psychology sustained a damaging blow,” the New York Times reported.  It was one of a string of scandals affecting the field in recent years.

Criticism has come from all sides.  Some blame the original researchers.  Others say the reproducers have overlooked nuances of the experiments, going so far as to call it a “vigilante exercise.”  One thing is not debated: extreme pressure exists for researchers to deliver new, publication-worthy results.

Intense, career-defining pressure is not unique to the research world.  High stakes incentives have led to perverse results not only for scientists but also for CEOs, factory workers, teachers, and more. 

Once bad behavior begins, new rules and regulations are put in place.  Sure enough, academic journals have started to implement stricter guidelines for submitted papers.

Yet rules and regulations alone can’t solve the problem.  Psychologists can turn to their peers for help find a longer-term solution.

As hundreds of psychology experiments have shown, when our reason for working – our motive – is disconnected from work itself, our performance suffers. No industry is immune from the cobra effect.



As the legend goes, during the days of British colonialism, the city of Delhi in India had a cobra infestation.  To combat the snakes, the colonial government implemented a logical new policy: a bounty for every dead snake brought in.  Dead snakes piled up.  But all was not as it seemed.

High stakes incentives have led to perverse results not only for scientists but also for CEOs, factory workers, teachers, and more

A few shrewd entrepreneurs realized there was good money to be made from dead cobras.  They opened cobra farms.  Once the government realized what had happened, they canceled the bounty. As the value of cobras plummeted, the snakes were released.  In the end, the city had more cobras than when the initiative began.

Because it was hard to measure the true goal—fewer live cobras—the government created a system to incentivize what was easy to measure—more dead cobras.  Unfortunately, the two measures were not equivalent.  In fact, the incentive created the exact opposite.

The cobra effect occurs when we use a reward or punishment to incentivize a behavior, and get an unintended consequence as a result.  The reward or punishment causes people to take the shortest possible path to success.

Unfortunately, cobra farms are not limited to colonial Delhi.



Scientists often describe the phenomenon of “publish or perish”.  The New York Times describes this:  "Scientists have pointed to a hyper-competitive culture across science that favors novel, sexy results, and provides little incentive for researchers to replicate the findings of others, or for journals to publish studies that fail to find a splashy result.”  If it is true that scientists fudged (or selectively reported) results, we have a cobra effect.  Rather than incentivizing good science, we got the opposite.

However, the issue is not limited to science at all.  If anything, cobra effects are far more common in industry.  

For example, in one company we worked with, customer service agents were incentivized to keep their customer calls below a certain length of time.  Agents were penalized and eventually fired for spending too much time on each phone call.  The intent of the punishment was positive.  Customers didn’t want to spend unnecessary time on calls.  Long calls cause costs to go up, which would either cause prices to go up or make the company less competitive. 

The cobra effect occurs when we use a reward or punishment to incentivize a behavior, and get an unintended consequence as a result

Have you ever been on hold at a customer service line, heard the sound of someone picking up your call, and then heard the line go dead?  You may have been experiencing a cobra effect.  At this company, when an agent’s average time on calls went above the company limits, they started picking up phone calls and then immediately hanging up.  Their average call time went down.  This was a cobra effect.

We’ve seen cobra effects across all types of organizations.  We’ve seen examples in school systems across the country. Teachers have cheated on standardized tests to protect their school’s funding or their jobs.  When faced with extreme pressure to produce predictable quarterly earnings, CEOs have been known to dabble in maladaptive behaviors from cutting much needed investments in growth, to entering the accounting grey area and cooking the books.  This is a cobra effect.



The natural reaction that organizations have to cobra effects is creating a never ending game of “whack-a-snake.”  Each time a cobra effect happens, our knee-jerk reaction is to create more policies, rules, punishments, and clever incentive schemes.  We create armies of overseers (who also need to be overseen for their own cobra effects).

Instead of spending so much time aligning incentives, we should instead spend the time aligning motives


Yet the cobra effect isn’t the by product of bad people, per se, but instead a predictable consequence of a system where the motive for doing the work is different from the work itself.  For example, in colonial Delhi, the motive for bringing in dead cobras was to get the reward.  The motive was separate from the activity.  If, on the other hand, the motive had been a fundamental curiosity in how to cure the cobra problem, or a deeply held belief that the city should have fewer cobras, the motive would be aligned with the activities.

Companies spend a lot of fruitless time trying to align incentives to behaviors.  We’ve been asked by organizations in the past to create perfect scorecards that make sure that every possible cobra effect has been accounted for.  This pipe dream is simply impossible.

Instead of spending so much time aligning incentives, we should instead spend the time aligning motives.  For example, rather than publish-or-perish, could we not create a promotion model where researchers work more closely with the people and organizations that would benefit from their work?  Career decisions could be linked more closely to impact versus prestige or publications.  

In the call center, could we not structure roles so that representatives feel most beholden to their customer rather than their performance management system? 

We need to construct roles, organizations, career paths, and compensation systems that encourage people to work because they enjoy and believe in the work itself.  Then all the time and money we invest in judging performance could be better spent on improving performance.  

When we solve this problem, we unlock the whole of human potential. 


Pick up a copy of our bookPrimed to Perform, to learn more about The Cobra Effect and the science of total motivation. And join the movement to fundamentally transform workplace culture by subscribing to our newsletter