Clues to Longer Living

“How long will I live?”

It is a tantalizing question, and one for which there is rarely a satisfying answer until it is too late. Now a group of Bloomberg School faculty and students is teaming up with an insurance giant to try to get answers, by collaborating on a study of mortality of unprecedented scale and scope.

The academic researchers are interested in understanding past increases in life expectancy so they can anticipate future increases, says Gerard Anderson, PhD, a professor of Health Policy and Management who is leading the group.

For the insurer (which did not want to be identified) a data-driven understanding of mortality may improve its decision making in a number of important ways. The results could also have profound implications for the Social Security and Medicare trust funds. A one- or two-year difference in Americans’ life expectancy in 2025 and beyond will have major implications for how long the trust funds will be solvent, says Anderson.

The insurance giant has made a unique dataset available to the researchers: records of the 14 million current and former policyholders, with all personal identifiers removed to protect privacy. “Nobody has ever had this large a dataset looking at mortality from a research point of view,” says Anderson, “so we’re kids in a candy store.”

Even more important than the dataset’s breadth is its collective depth. It contains remarkably detailed information from driving records to minutia like whether or not policyholders are scuba divers.  “All the other datasets have limitations,” says Anderson. “This one has all the pieces.”

While the researchers are still a year from publishing any results from the partnership that began in March 2012, Anderson is confident the findings will be important. “We’re going to be able to predict in much better specificity and much better ways how long people are going to live,” he says.