Like most big employers across the country, Pennsylvania State University has spent several million dollars over the last decade on voluntary wellness programs, hoping to cut health costs and improve workers' health.
The strategy hasn't worked.
"Only a very small percentage of faculty and staff have participated," university president Rodney Erickson wrote this month to all 17,500 employees and their 22,700 dependents.
So Penn State, like a small but fast-growing number of employers, is switching from carrots to sticks. In January, employees will face a $100 monthly surcharge if they haven't filled out health-screening forms and gotten a physical exam. If they smoke, that's an additional $75 a month.
The health-coverage changes - announced last month to howls of outrage on the State College campus - put Penn State at the forefront of a movement aimed at promoting personal responsibility, especially through rewards or penalties.
The problem, researchers say, is that there is scant evidence that wellness programs, with or without financial incentives, work.
Harald Schmidt, a researcher with the University of Pennsylvania's Center for Health Incentives and Behavioral Economics, said workplace wellness initiatives were built on lots of assumptions that haven't been proved.
"And there is no federal requirement to evaluate all these programs," he said, "so a huge number of employers are doing these natural experiments, and we can't learn from them."
Last year in the journal BMJ, Penn and Carnegie Mellon University health economists concluded that while some components of health incentive programs were effective, "many programs are not particularly well designed, and most have been implemented without testing."
The decadelong rise of wellness programs is rooted in "behavioral economics," which recognizes that people do not necessarily do what they know is best for them. The idea is to nudge them by, say, reimbursing the cost of joining a health club, or paying $500 for quitting smoking.
The Affordable Care Act firmly endorses wellness incentives. It allows rewards or penalties based on health factors - such as body mass index or cholesterol levels - of up to 20 percent of the total cost of coverage this year, and up to 30 percent next year. Smokers could be on the hook - or rewarded - for as much as 50 percent, although they can avoid the surcharge if they are in a smoking-cessation program.
The average annual cost of coverage for an employee with a family is $8,000 a year, according to the Kaiser Family Foundation, so the incentive could be as much as $4,000.
But "very few are anywhere near 20 percent" for maximum penalties, said LuAnn Heinen, a vice president of the nonprofit National Business Group on Health, which represents large employers' perspective on health policy.
Employers are increasingly opting for penalties. The business group's latest survey of 583 organizations with $103 billion in annual health-care expenditures found that, this year, 18 percent were increasing premiums or deductibles for workers who didn't comply with wellness programs. But next year, 36 percent of employers plan to do so.
The shift reflects another premise of behavioral economics, noted in the BMJ article: "It is usually far more efficient to penalize than to reward, because most rewards will typically be wasted on people" who would behave prudently anyway.
However, this is counterbalanced by the fact that employers don't want disgruntled employees. It reduces productivity and increases turnover.
"No employer is trying to create disaffection," Heinen said, "so when that happens, it's not the goal."
Penn State officials have stirred considerable disaffection with the new wellness initiative, managed by WebMD.
They emphasize that the health questionnaire, physical exam, and "biometric screening" - checks of weight, blood pressure, cholesterol level - are not mandatory; that the information will "not be used for any type of punitive action under any circumstances now or in the future"; and that privacy and confidentiality laws will be upheld.
But art history professor Brian Curran is among those who don't buy the assurances. About 2,100 colleagues have signed his online petition demanding that the health-coverage changes be suspended pending consultation with the university community.
"It's been called Unhappy Valley for a while," said Curran, referring to fallout from the university's sexual-molestation scandal. "Serious damage has been done with this plan. It's a cost-cutting program, not just a wellness program."
Indeed, the president's letter said cutting costs was essential. Health expenditures are projected to be $217 million next year, up 13 percent from this year. "Passing large increases year after year onto employees or our students," Erickson wrote, "is not a sustainable strategy."
In addition to the wellness initiative, Penn State employees whose spouses could get coverage through their own employers will be dinged $100 a month. Last year, the university said it saved more than $5 million by rejecting dependents who could not prove eligibility.
Some research suggests wellness programs can generate savings. A Harvard analysis of studies of dozens of programs, including counseling and seminars, found that for every dollar spent on the programs, medical costs fell $3.27 and absentee costs fell $2.73. Fewer than a third of these programs used financial incentives to drive participation.
In contrast, a recent federally commissioned report by the Rand Corp. found the impact of workplace wellness initiatives on medical costs and employee health was negligible, even though the programs have become a $6 billion industry.
The report quoted a wellness-program representative who said, "When it comes to changing your health, the real motivation has to be internal, and you have to want it. If somebody forced me to participate, or if people participated to get an iPod or something like that, I don't know how genuine that participation is."