Principal investigator: Adriano Valladão Pires Ribeiro
Article title: WHAT DO WORKPLACE WELLNESS PROGRAMS DO? EVIDENCE FROM THE ILLINOIS WORKPLACE WELLNESS STUDY
Article authors: Damon Jones, David Molitor and Julian Reif
Location of the intervention: Illinois, USA
Sample size: 4.834 employees
Sector: Health Insurance
Type of intervention: Offer of a wellness program for employees
Primary variable of interest: Health and productivity metrics
Evaluation method: Experimental Evaluation (RCT)
Evaluation Context
In the United States, the federal government adopted an affordable healthcare measure in 2010 that encouraged companies to contribute up to 30% of the total cost of health plans, and some states also included wellness incentive programs. As a result, more than 50 million workers have some type of coverage for wellness programs, and the responsible industry has tripled its revenue since 2010. Among the factors that may explain the popularity of these programs is employers' belief in reduced medical expenses and increased worker productivity. Furthermore, workers who value health more may be attracted to or remain with companies that support wellness care.
In this context, the effects of wellness programs will depend on the type of worker participating, that is, whether or not they already had healthy habits before the program. For those who already had healthy habits, the cost of entering the program is small, but the potential returns are lower compared to those with poor habits. Furthermore, the effects can be far-reaching, impacting healthcare spending, worker productivity, absenteeism, and health and wellness habits, both in the short and long term.
Intervention Details
To study the issue, a wellness program, iThrive, was implemented for University of Illinois employees in the cities of Urbana and Champaign over a period of two years. Three types of interventions were employed: (i) an annual in-person biometric health exam; (ii) an online health risk assessment; and (iii) weekly wellness activities. Clinical health data, information on healthy habits, and the promotion of a healthier lifestyle were collected, respectively.
In the first year, 12.459 eligible beneficiaries were invited to participate in the program, of which 4.834 accepted. Of these, 3.300 formed the treatment group and were invited to attend wellness activities and receive remuneration for them, in addition to completing the two other interventions mentioned above. The remuneration for those who completed the two-year program ranged from $50 to $650, an amount decided randomly and announced at the beginning of each year. The remaining 1.534 people formed the control group and were not allowed to engage in the activities; they only underwent biometric testing and answered a questionnaire at the end of each year.
Along with biometric exams and health risk assessments, administrative data on participants such as age, gender, race, salary, occupation, and medical leave were also collected. Participants had to complete questionnaires about their health condition, use of health insurance, job satisfaction, and productivity. Data from health insurance plans were also collected to determine when and why they were used (clinic, hospital, pharmacy, etc.), as well as participation in marathons or 10km and 5km races in the city of Champaign and a technical index of worker productivity. All this combined data allows us to study, following the methodology described in the subsequent section, the impact of the wellness program on worker productivity, the formation of healthy habits, healthcare spending, among others.
First, it was possible to identify the characteristics of workers most likely to participate in wellness programs based on information from the treatment group. Second, since the participants in each group were selected randomly, the difference in the average results of the variables of interest can be explained by the employee's participation or non-participation in the wellness program. In other words, the results of the treatment group are compared with those of the control group. The variables of interest included medical expenses related to health, measures of work productivity, health habits, and health status.
Results
It was found that employees who participated in the wellness program spent, on average, $115,3 less per month on healthcare in the 13 months prior to the program compared to the control group. On the other hand, they were more likely to have a positive expenditure. In other words, the people in the treatment group who participated in the activities already had moderate healthcare spending, while those in the control group were at the extremes, spending either nothing or a considerable amount. Furthermore, the participants in the activities occupy an intermediate position in the income distribution, have slightly lower productivity, and were already more likely to engage in physical activity before the program.
First YearAfter the first year of the program, no change was observed in healthcare spending for the treatment group. The impact on productivity was also nil across all metrics, whether for administrative data (annual salary, probability of promotion and termination, and sick leave), questionnaires (such as job satisfaction or feeling more productive at work), or the productivity index. Measures of health habits, such as gym attendance, participation in marathons and 5km and 10km races, were also not impacted by the treatment.
Second YearThe program design was similar for the treatment group in the second year, and the results were reported after 30 months of the program. The long-term effects were similar to those obtained after the first year; that is, there was no impact on the variables that measure health expenditure, productivity, and physical activity. The main difference lies in the workers' perception of managers; after 12 months, managers are seen as prioritizing health and safety, but this effect disappears after 30 months.
Lessons in Public Policy
The main lesson from the study lies in the lack of impact of the wellness program on medical expenditure, productivity, and the promotion of a healthy routine for the worker. The program's greatest beneficiaries were not at the bottom of the income distribution, already had some medical expenses, and practiced physical activity. This suggests a transfer of costs from program participants to lower-paid employees with high healthcare expenses and poor health habits. The difficulty of stimulating healthy routines on a large scale is also highlighted, as financial incentives are not always sufficient.
Reference
JONES, Damon; MOLITOR, David; REIF, Julian. What do workplace wellness programs do? Evidence from the Illinois workplace wellness study. The Quarterly Journal of Economics, vol. 134, no. 4, p. 1747-1791, 2019.
