What 85 Years of Research Says About What Makes Relationships Work
The Harvard Study of Adult Development has tracked hundreds of lives since 1938. Its central finding is not what most people expect.
Opening Hook
In 1938, researchers at Harvard University began tracking the lives of 268 Harvard sophomores. They measured everything: physical health, mental health, career success, income, personality traits, habits, relationships. The study, eventually expanded to include a second cohort of inner-city Boston youth, has continued for over 85 years, making it one of the longest longitudinal studies of human development ever conducted.
The question the study set out to answer was simple: What makes a good life?
After eight decades of data collection, physical exams, psychological assessments, and interviews, the answer was not career achievement. It was not wealth. It was not fame, IQ, or social class. The single strongest predictor of health and happiness across the entire lifespan was the quality of a person's close relationships.
The Research
The Harvard Study of Adult Development has been directed by a series of researchers, each contributing to its findings. George Vaillant, who directed the study for over three decades, published his analysis in "Triumphs of Experience" (2012), documenting how the men's lives unfolded from young adulthood through old age.
Vaillant's central observation was that the men who fared best, by nearly every metric the study tracked, were those who had invested in close relationships. This was true controlling for childhood advantages, intelligence, physical health, and socioeconomic status. Men who had warm relationships with their mothers earned significantly more than those who did not. Men who had warm relationships with their fathers were less likely to develop anxiety disorders. Men who had strong social connections at midlife were healthier at age 80 (Vaillant, 2012).
Robert Waldinger, who took over the study's direction in 2003 and expanded it to include the original participants' spouses and children, has further developed these findings. Waldinger and Schulz (2010) found that people who were most satisfied in their relationships at age 50 were the healthiest at age 80. The quality of relationships at midlife was a better predictor of late-life health than cholesterol levels, blood pressure, or any other biomedical measure the study collected.
Waldinger has been explicit about what "relationship quality" means in this context. It does not mean the absence of conflict. Many of the happiest couples in the study argued regularly. What distinguished satisfying relationships was not the absence of friction but the presence of trust, the felt sense that you could count on the other person when things got difficult.
The study also documented what did not predict a good life. Professional achievement, beyond a baseline level of occupational stability, did not predict happiness. Income, beyond a moderate threshold, did not predict happiness. IQ did not predict happiness. The men who were most successful professionally but lacked close relationships were, on average, less healthy and less happy than men of modest professional achievement who had strong connections.
One of the study's most striking findings concerned the trajectory of relationships over time. Waldinger (2015) noted that the men who were happiest at age 80 were not the ones who had always had easy relationships. Many had experienced periods of difficulty, distance, or conflict. What characterized the happiest men was not the absence of relational struggle but the persistence of relational effort, the willingness to continue investing in connection even when it was difficult.
The second cohort, composed of men from disadvantaged backgrounds in inner-city Boston, showed the same fundamental pattern. Despite dramatically different starting conditions, including differences in education, income, neighborhood quality, and childhood stability, the strongest predictor of late-life wellbeing was the same: the quality of close relationships (Waldinger & Schulz, 2010).
The Commentary
There is something both comforting and uncomfortable about this finding.
It is comforting because it suggests that the things that matter most are available to almost everyone. You do not need a Harvard degree, a high salary, or exceptional talent to build a good life. You need people you trust and who trust you.
It is uncomfortable because most people, if they are honest, are not treating relationships as their primary investment. They are treating them as something that will be attended to once the more pressing matters of career, finances, and personal achievement are handled.
The Harvard data suggest that this prioritization is backwards. The men who deferred relational investment in favor of professional achievement did not arrive at a later life stage where they could then enjoy the fruits of their labor in the company of people they loved. They arrived at old age with professional accomplishments and, in too many cases, relational poverty.
What makes this especially important is that the study controlled for the variables people typically assume matter most. It was not the case that healthy, wealthy, well-adjusted people simply happened to have good relationships. Relationship quality predicted outcomes independently of health, wealth, and baseline adjustment. It was an independent variable, not a confound.
Consider what this means for how you spend your time. If you are like most ambitious people, you allocate the majority of your waking hours to professional development, skill building, financial planning, and career advancement. The Harvard data do not suggest these are unimportant. They suggest that beyond a baseline of occupational stability and financial adequacy, additional investment in these domains produces diminishing returns, while investment in relationships continues to produce substantial returns throughout the lifespan.
What This Means
The Harvard Study did not discover that relationships are nice. It discovered that relationships are fundamental, that they are the single strongest predictor of whether a life goes well or poorly, and that this finding holds regardless of the advantages or disadvantages a person starts with.
This is not a sentimental claim. It is an empirical one, supported by 85 years of data collection across diverse populations.
The practical implication is straightforward, even if it is difficult to act on: the time you spend maintaining, deepening, and repairing your close relationships is not time away from building a good life. It is the primary activity of building a good life.
If 85 years of continuous longitudinal research cannot settle that question, it is hard to imagine what could.
References
Vaillant, G. E. (2012). Triumphs of experience: The men of the Harvard Grant Study. Harvard University Press.
Waldinger, R. J. (2015). What makes a good life? Lessons from the longest study on happiness [TED talk]. TEDxBeaconStreet.
Waldinger, R. J., & Schulz, M. S. (2010). What's love got to do with it? Social functioning, perceived health, and daily happiness in married octogenarians. Psychology and Aging, 25(2), 422-431.