Data enthusiasts must do more to help public-sector leaders understand how data techniques and tools can help them change real outcomes.
I recently saw a T-shirt that said, “Data is the new bacon.” And it certainly seems that way — everyone is hungry to find, acquire and consume data, and the market is answering the call.
In the past few months, we have seen the White House launch a new Smart Cities Initiative and host a forum on citizen science and crowdsourcing. General Electric started rebranding itself as a digital company helping cities become more intelligent. My own organization, Johns Hopkins University’s Center for Government Excellence, through our partnership with Bloomberg Philanthropies’ What Works Cities program, is helping 100 mid-size cities accelerate their use of data and evidence to improve people’s lives.
As this data consumption becomes increasingly widespread, practitioners must avoid using technical jargon that not every public servant, let alone the public, can decipher. The esoteric language can be alienating and seem far removed from the problems governments need to solve.
How does using "a Hadoop cluster to mine unstructured data" help a city fight crime? How is using "R to build predictive models" going to increase the supply of affordable housing? How is an "open data platform" going to increase the quality of classroom instruction? As data enthusiasts, we must do more to help public-sector leaders understand how these techniques and tools can help them change real outcomes. Here are four simple steps to get government leaders — who don’t know the difference between a sequel and SQL — on the data diet.
All the data and technology in the world will not help governments solve problems if its leaders are not clear what they want to improve and commit to improving it. This requires making tough choices about what to improve (and what to temporarily ignore). It requires being public and transparent about those choices. And then, finally, it requires publishing data and demanding a steady operational focus on the problem until progress is made and sustained.
Public-sector managers are often searching for more public policy graduates to fill ubiquitous “program analyst” positions. We need to think differently. Computer science, not political science, is the lifeblood governments rely on to serve and communicate with constituents. Almost every problem in every jurisdiction requires layered data analysis, and that requires coding. So stop outsourcing 21st-century talent and bring modern skills into the halls of government.
Government leaders want to be good at everything, but, honestly, no one has that luxury. So get selective. Pick one outcome to dramatically improve. Study it top to bottom. Publish all the relevant data. Meet and talk about it relentlessly. Instead of trying to be good at everything, be great at a few things. It will have a cascading effect on everything else.
For example, in 2007, the Washington, D.C., mayor took control of the chronically underperforming public schools and focused relentlessly on improving education. Subsequent administrations have continued this laser-like focus, and it is paying off. Recently released testing data shows District of Columbia Public Schools as the fastest improving urban school district in the country.
Once, after a tornado, I spent hours in an assembly line of volunteers passing bricks one-by-one from house to curb. When I stepped out of line and walked up a small hill to get water, I could see a nearby bulldozer moving thousands of bricks to the curb in seconds. I was instantly reminded how our best intentions don’t justify the continuation of mediocre practices or the underuse of technology. Don’t get stuck passing bricks by hand because that’s how things have always been done or because that’s what people know how to do. Zoom out to find a new vantage point. Data is an enterprise asset — use it, study it and find a bulldozer strategy to clear the bricks faster than previously imagined.
After watching its planning department take action on less than five blighted properties one year, the mayor of Jackson, Miss., zoomed out and realized police officers were well positioned to enforce building codes at blighted properties. Within seven months, more than 200 lots were designated as “menaces to public health, safety and welfare,” and 107 vacant houses were demolished by September 2015.
Data and computer science are making game-changing contributions to public service. It’s time to embrace that wave of change and look to data as a solution. Start by making a strong commitment, rethinking the talent pipeline, focusing more intensely on fewer things and zooming out for a better perspective on the challenges we face. And remember to use language everyone can understand.
Carter Hewgley is the director of Analytics and Performance at Johns Hopkins University’s Center for Government Excellence. He is the former head of analytics for FEMA.