What does predicting the location of improvised explosive devices have to do with better decisions in government?
According to Shell Culp, data analytics — discovering meaningful patterns in data using statistics, computer programming and operations research — can reveal opportunities overlooked in the normal process of government decision-making. Culp, — chief deputy director of the Office of Systems Integration of California’s Health and Human Services Agency — went on to say that “Most departments are good at telling you what they did: counting things. What they can’t tell you is ‘how they are doing’ in terms of outcomes,” she said. “That’s where the opportunities are.” She added that analytics does not require massive warehouses of data — that small bits of data can have enormous benefit if analyzed and applied intelligently, and that such processes as permitting are fertile ground for improvement.
Tuesday morning, May 13, Culp will moderate Data Analytics 2014, a Techwire event featuring Ann Boynton, deputy executive officer, CalPERS; Chris Cruz, CIO, Department of Health Care Services, State of California; Tamara Dull, director of emerging technologies, SAS; Scott Gregory, geographic information officer, State of California; and others.
Why should you attend? The simple reason is that what you don’t know can hurt you, and what you do know can reap huge benefits to your agency, department and community.
For example, Michael P. Flowers, a number cruncher in the NYC Office of Policy and Strategic Planning, saw how the military used analytics in Afghanistan to predict where improvised explosive devices were most likely to be planted. The experience prompted Flowers to start using analytics to solve city problems — like tenement fires.The city has 2,000 serious fires each year in multi-family buildings. The city also gets 20,000 complaints each year about illegal conversions of buildings — jammed with partitions, hotplates, extension cords and extra families. Each illegal conversion is a potential catastrophe, but with few inspectors, how could the city prioritize the complaints so that inspectors would visit the worst buildings first?
Flowers and his team inspected the data and found four predictors of fires. The buildings most likely to burn were constructed before building codes were revised in 1938, had unpaid taxes, an owner undergoing foreclosure, and were located in a low socio-economic neighborhood. A 13 percent vacate rate — meaning that the buildings inspected were so dangerous that they were unfit for human habitation — shot up to 70 percent when these data were used to prioritize inspections.
Far from being an isolated success story, this “data analytics” methodology can be applied to many other scenarios. And that is the gold at the end of the numbers, the understanding that helps drive better decisions and derive better results. So join Culp and colleagues tomorrow for this exciting event.
