IE 11 Not Supported

For optimal browsing, we recommend Chrome, Firefox or Safari browsers.

How to Make Analytics Work for Your Government

A California pension and health services executive highlights tips and progress in analytic initiatives.

Use of analytics in government is on the rise around the country, as proven by Chicago, the Sacramento Municipal Utility District in California, Indiana, Pittsburgh and a host of other jurisdictions.

And on May 13, a senior official from California Public Employees' Retirement System (CalPERS) shared her insights on government analytics -- the practice of using big data for predictive or statistical decision-making. For CalPERS, success in analytics is credited to the agency’s freedom to embrace risk, said agency Deputy Executive Officer Ann Boynton, who delivered this message at Data Analytics 2014, an event hosted in Sacramento, Calif., by Government Technology sister publication

CalPERS completed 22 analytics initiatives in the last three years, Boynton said, which meant there were many lessons to learn about the practice. One such initiative was the creation of MyCalPERS, a massive overhaul of 109 different data systems that manages the organization’s account portfolio of $290 billion — the largest in the nation. 

One piece of advice Boynton has relates to government’s attempt to be perfect -- which she says can unintentionally limit government to significantly imperfect solutions and restrain its ability to innovate next to the private sector.

“I think CalPERS as an organization tries to take the position of, ‘Go for it!’ because I can guarantee you we won’t get any better if we don’t do anything," she said. "And even if we fail, we will have learned what not to do the next time out."

Drawing upon examples of her work at CalPERS, which began in 2010, Boynton said the ability to take leaps and experiment has led the organization to make sizable efficiency improvements that employ analytics to gain business insights.

She identified four main areas where analytics has helped the agency: innovations in services, timely and enhanced insights of member needs, deeper understanding of market trends, and comprehensive impact analysis of policy and legislation.

“The beauty of business intelligence," Boynton said, "is that it creates insights that enable an organization to take a holistic view of information for instant decision-making.”

To set the stage for analytics, CalPERS completed a number of initiatives in 2011, which included conducting a study to assess analytics needs, drafting a five-year road map for its operational goals and strategies, and consolidating 109 databases into MyCalPERS, a transactional database system housing agency information.

“Conversion from legacy systems to MyCalPERS required a lot of patience and hard work,” she recalled, and then joked that some employees in the room may still have a few nightmares about the needful — yet massive — undertaking.

Notwithstanding Boynton’s levity, the labor was a crucial step for the organization as it seeks to improve services and innovate. As a use-case example, Boynton said CalPERS has harnessed analytics and its unified database to dramatically reduce the time spent on research for annual rate plan negotiations.

“Research for annual rate adjustments and negotiations took days to accomplish and multiple spreadsheets,” Boynton said. “Now [analytics] enables the center to perform the analysis in seconds or minutes, rather than weeks.”

Additionally, areas where analytics have improved efficiency include payroll audits, contracted service assessments and a project that uncovered a few instances of erroneous sick leave paid out to retirees.

Like some states, CalPERS has the practice of converting remaining sick leave time into retirement compensation. Typically, most state employees must work 21 years before they’re able to accrue one year of sick leave. However, an analysis found pockets of discrepancies where service didn’t match time awarded.

“What happened when we looked at the data is that some members had more leave than we thought was actually possible based on our information of their service records,” she said.

Boynton also advised governments to nurture a work culture for analytics use, focusing hiring practices toward business analytics skill sets and harboring creative solutions with an open mind.

Jason Shueh is a former staff writer for Government Technology magazine.