Last week, I participated in the kickoff session for the State of Texas Technology Strategic Plan. We’re pretty familiar with government planning, and this was a great way to learn about the future of technology in the state.
Spearheaded by the Texas Department of Information Resources, the technology strategic plan will be available later this year. The advisory board includes the CIOs of major universities, state departments, and private companies. On a sunny April afternoon in Austin, we workshopped our way into a decent outline for the future.
Strategic plans can be a helpful reference for government employees. They can justify difficult decisions when there are no other roadmaps. But many strategic plans have a reputation for outlining lofty goals with no practical way to achieve them. We wanted to avoid this trap.
The session began with brainstorming about technology trends. While it was a wide ranging discussion that included 3D printing and self-driving cars, this group honed in on two pressing topics: cybersecurity and data. I focused on data.
From there, the workshop facilitator pushed us to revise our ideas into actionable recommendations for the State. I offered two things:
- Publish more data
- Encourage open source development
I believe that these two action items could have a significant impact on the State’s technology decisions for years to come. Here’s why:
Publish more data
Releasing data to the public isn’t an easy process. The data usually lives inside an application, and needs to be “exported” or “reported” to create a raw dataset. In order to be truly open, this data should publish automatically to a data portal. The whole process can be cumbersome, and a lot of agencies just give up.
It doesn’t have to be this way.
The State of Texas already has a great infrastructure. We have an open data policy (check out Sec. 2054.1265), an open data portal, and years of experience with open data procedures.
Despite this groundwork, I don’t think the state is publishing and updating enough “high value” datasets.
Currently, the State of Texas open data portal has only 123 datasets, of which only 12 datasets have more than 1,000 views in 2015. And only 23 datasets have more than 300 views. Almost all the popular datasets have something to do with DIR contracting. I suspect the recent software procurement scandal has something to do with it.
It’s great that we have so much procurement data, but there is more to government than spending. We miss a lot of opportunities to publish useful data, and to promote the use of that data to improve operations and citizen service.
The state should double down on its commitment to open data. They should publish a lot more datasets, analyze usage, and engage the community in an ongoing dialogue about the value of the data.
Encourage open source development
Open source solutions exist across every government technology stack. For example, Drupal is the content management system for 150+ countries and thousands of cities. But it’s not just limited to city websites. Any server running Linux and any app built with PHP relies on open source technology. This doesn’t even account for the work of Data.gov or Code for America.
Open source is here to stay.
The state should embrace it. The total cost of ownership of many open source solutions is a fraction of their proprietary counterparts. But embracing open source requires competent people to support the various systems and applications. That doesn’t happen overnight. I think a good place to start is an audit and inventory of existing open source solutions used by the state. Then they can create a support plan to ensure the security and ongoing development.
Next, the state should build competency in these technologies by adopting modern recruitment methods. For example, every job application requires a resume, but resumes are a bad proxy for a software developer’s experience. A good recruiter will look at an applicant’s Github profile as well as their resume. To encourage the right culture, the recruiter could give preference to active open source contributors. The goal is to fill the building with engaged, curious, and productive software engineers.
The state could also do more to encourage software development in the community. They have experimented with community challenges in the past, but they need to commit to more than just a one-off PR stunt. Community engagement should be part of someone’s job description. Representatives should appear at local hackathons with sample projects and datasets in hand. Each agency should research their counterparts in other cities, and encourage the technology community to fork and adapt. Then each agency would do everything possible to remove barriers to implementation.
I’m hopeful that the State of Texas will incorporate these recommendations. To do it right requires a bit of operational change. But it ends up encouraging a more open, transparent, and accountable government.
