In October 2015, two senior California officials: Marybel Batjer, Secretary for Government Operations, and Michael Wilkening, Undersecretary for the Health and Human Services Agency, seized on an idea that had the potential to turn the state’s long dysfunctional technology procurement process on its head.
After years of planning, California was about to request bids for a new child welfare management system to replace a twenty-year-old technology that, as part of the country’s largest child welfare program, was accessed by approximately 25,000 state and county employees, to serve millions of California’s children. The request for bids called for a “waterfall” approach to software development, where all aspects of the project would follow a pre-ordained sequence and likely cost the state nearly half a billion dollars.
Just weeks before the request was to be released Batjer and Wilkening met with former U.S. Chief Technology Officer Todd Park and members of a tech nonprofit, Code for America, who warned that both the waterfall approach and the large size of the project posed significant risks. First, they proposed an alternative “architecture” that would break up the “monolith” proposal into its component parts. Rather than build and replace one giant system with another, why not replace the system in parts—or in modules? Second, for each of these modules, teams of programmers and social workers could work together to build a prototype and iterate till the final product satisfied the needs of the state—in a process known as “agile” development.
But could California’s bureaucracy, with its stringent procurement rules, pivot to a modular approach? And could Batjer and Wilkening convince state staff and county partners to experiment with agile development, a methodology never attempted in California government before?
In taking an in-depth look at California’s decision to try both an agile and modularized approach to procurement, students learn about the risks involved in attempting digital transformation in government. Students identify and analyze the underlying incentives that drive individuals and institutions, and when incentives are misaligned making bringing about change even harder. Last, students learn that an agile approach in and of itself does not provide the magic bullet for governments trying to fix technology projects. Changes in both management of and execution of these projects will have to follow for success.