Written byAndrew McMahon
Two weeks ago, I highlighted a number of structural issues impinging on the federal government’s ability to procure and develop modern technology. Additionally, I did my best to offer a few important steps that government agencies and private sector companies could take to make dents in the technology procurement-industrial complex. These steps are not a panacea but are critical to shedding light on how good products could solve the government’s needs and how agencies could avoid overpaying for unnecessarily complex solutions. To recap from my previous post:
The structural problem is threefold, (1) agencies try and avoid risk outright rather than manage it and the punishment for failure far exceeds the reward for success; the government maintains (2) a procurement process that has become more important than the outcome; and (3) a hiring process that is torturous for the applicant and rarely relies on subject matter experts to identify talent.
In addition to government agencies and private sector, there are two other stakeholder groups that are responsible for the maintenance of the problem but could also be a driver of change: the Press and Congress.
In the press
One might wonder what the press has to do with fixing government’s technology, but the press plays an important role. From major publications (think New York Times) to sector specific ones (think Wired), and trade publications (think FedScoop), each can move the government forward or contribute to the pull of inertia. By creating or reducing a fear of failure and risk aversion through their reporting, the press has quite a bit of influence. The press could certainly help change behavior and the acceptance of risk by agencies in a few key ways.
First, trade publications need to write more about emerging technology. Without frequent articles about what the private sector is adopting or anticipating as the next disruptive technology, a large part of government will believe blockchain equals bitcoin and artificial intelligence equals Skynet. There needs to be a broader conversation about what is driving other industries forward and reinforcing that government isn’t “too unique” to be early adopters.
Just as critical is positive news around testing, use and outcomes from emerging technology that would encourage broader adoption and risk taking among agencies. If the average agency program sees the emerging technology market as the realm of DARPA alone, the government will never move forward from a technical perspective. Increased reporting on what other agencies are learning and scaling could encourage other agencies to follow suit.
The same can be said for sector specific publications. Here, you are more likely to read about the government’s maintenance of decades old mainframes than about the use of robotics to help amputees at the VA. But still, there is a growing interest among the technology community in working for or building companies/products for the public sector. There are real, complex, and meaningful challenges to solve in this space. Broad discussion of the opportunities and problem set in government by publications like Fast Company, Wired, Re:code would encourage more people with modern technology skill sets to enter this sector of the economy. And by showing technologists and entrepreneurs the impact they can have — by highlighting those who have successfully done it before — would only compound that interest.
Finally, because the relationship between agencies and oversight is critically important to our style of government, press coverage of these moments are a spotlight. When audits and oversight become embarrassing moments, it freezes progress. Sometimes this is warranted, but other times the amplification of normal oversight causes an agency to stop progress for fear of bad press.
Unfortunately, this is when major publications like the Washington Post, that don’t regularly write about federal technology, step into the fray. Again, in some cases the oversight is warranted and there is bad behavior occurring. In others, the oversight and conclusions are completely normal or, occasionally, incorrect. But rarely do agencies push back on poor oversight or reporting and would simply rather not take additional risk.
if I were writing a GAO newsletter of some kind I would title it “Opportunities Exist”
— Eric Mill (@konklone) April 9, 2017
Short of widespread drug use and illegal activity, an audit report isn’t normally sensational. Most frequently and correctly, audit reports highlight areas for improvement with the common phrase “Opportunities Exist”. The press should quit sensationalizing Government Accountability Office reports, Inspector General alerts, and hearings that aren’t sensational. Talk about the content of the oversight and allow for the possibility of multiple conclusions. A number of months back, Seamus Kraft wrote one of the best pieces of commentary about a Congressional hearing, in this case, related to 18F and USDS. Seamus isn’t a reporter, but this is the type of reporting that should have occurred. Many reporters did write about this event and most of it was closer in tone to the coverage of the Army-McCarthy hearings. If the press does a better job of picking and choosing when to yell ‘fire’, agencies will react more effectively to oversight and not clamp down on risk inappropriately.
In many ways, Congress is the gatekeeper that allows the rest of government to tolerate risk and, at times, failure. Congressional oversight plays a critical role, but in most cases it is reactive. By being proactive with agencies in the emerging technology space, House and Senate Committees (and individual members) could have broad ranging benefit to the adoption of emerging technology through their own tolerance for risk and failure.
First, Congress needs to allow for more pilots, emerging technology set asides, and unique contracting authorities across a wider set of agencies. This past year’s National Defense Authorization Act opened up a few authorities for the Defense Department, Department of Homeland Security, and General Services Administration to acquire “innovative commercial items” (see Sec. 878). Congress even did everyone a favor and defined innovation!
(f) DEFINITION. — In this section, the term ‘‘innovative’’ means —
(1) any technology, process, or method, including research and development, that is new as of the date of submission of a proposal; or
(2) any application that is new as of the date of submission of a proposal of a technology, process, or method existing as of such date.
These new authorities are welcome, but more can be done here. Congress should closely monitor how any of the emerging technology purchased through these pilots is adopted across government and not simply contained in the silo of the programs. If Congress can work to ensure these agencies report on progress and show how technology is scaling, a model can be created that may be repeatable. In my previous post, I highlighted the DHS Science and Technology Directorate. They are doing some excellent work now, but if the technologies they procure don’t see adoption across the rest of the agency, most of their procurement work will be essentially a novelty.
While allowing for greater latitude in procuring and adopting any relevant technology, legislating on specific technologies such as cloud computing, mobile, or artificial intelligence with the intent to increase adoption in government is a poor use of the legislative tool. Speed of adoption is an organizational and cultural issue, not a legal one. Congress would be much better served by holding hearings and writing letters to agencies to gather information on how they are integrating specific technologies into the agency, what is resulting from the use, and following up on progress. Agencies respond to this type of Congressional interaction much more rapidly than implementing a legislative requirement. Agencies should be held accountable to modernize and make use of all of the authorities provided to them, if Congress encourages use and adoption it could have a long ranging effect.
The important (boring) work of reforming government
Again, we should all be weary of large government reform efforts without much detail that promise grand improvements in what is certainly the largest enterprise in the U.S. and possibly the world. I’ve highlighted some practices that should end across a number of stakeholders and others that should be adopted. But honestly, if all of these things magically happened tomorrow, it would make a small dent in the overall effectiveness and efficiency of the federal government. The adoption of emerging technology would certainly get a boost but there is a much, much more to do.
Without significant budget, procurement, and hiring reform across the executive branch, the federal government will make only marginal progress. It will take years to make these improvements and longer to see the benefit. Companies a fraction of the size of government take years, even up to a decade, to complete a turnaround. The government should be afforded this amount of time but also be held accountable to keep focus and pivot when necessary. The Obama Administration saw a half dozen different reform agendas that changed tack across the Executive Branch each time a new one was rolled out.
Until a focused agenda around government reform occurs, anyone who is interested in this space should be working on these or other smaller, targeted steps. Dcode42 will continue to identify, teach, and push great companies and their products into the federal market. If and when the government wants to dig deep and reform the procurement system, we’ll be waiting with lots of good suggestions built from trying to knock down the door from the outside.
This week we launched our spring cohort focused on Artificial Intelligence and Machine Learning. The companies we’ve selected could improve the government in the areas of productivity, data science, legal work and security, but unless agencies are prepared to adopt new ways of working the opportunity will be lost. We hope that by preparing our companies for the federal market they’ll stand a better chance against the inertia of government.
Stayed tuned, as Dcode42 and its companies will be writing a recurring blog on FedScoop discussing emerging technology, the application in government, and how the government can continue to improve the procurement of technology products.
Andrew McMahon is the director of strategic partnerships for DCode42. A former Obama Administration appointee focused on technology policy and building sustainable innovation in government, McMahon worked to change how the federal government invests in people, technology and companies so that it can better serve the public and businesses. During his time in government he helped to build 18F, the Presidential Innovation Fellows Program, and the GSA’s Technology Transformation Service.