October 9, 2019

REI Systems is pleased to have joined Johns Hopkins University and ACT-IAC on October 2, 2019 for a panel discussion that addressed the question “Are Analytics Important to Government…?” Jeff Myers of REI Systems and Dr. Jennifer Bachner of Johns Hopkins presented results of a 2019 survey of government analytics professionals.  We conducted the survey to help members of the analytics community understand whether they are on par with or significantly different from their peers. The survey touched on use of resources, types of analytic methods and tools, results obtained, and how impactful those results have been for the agency or organization.

Key findings from the survey included:

  1. When making significant decisions, the majority of agencies use analytics.
  2. If your analytics focus on money, they will have less impact on strategy.
  3. Analytics professionals spend more time gathering data than they dedicate to analyzing it.
  4. Artificial Intelligence gets a lot of talk, but not much action.
  5. Staffing is the biggest hurdle to the government’s ability to gain more value from analytics.

The good news is that most agencies use analytics, and, analytics are used heavily when it counts the most.  In fact, more than 80 percent of respondents indicated for their agency’s three most significant decisions, analytics played a dominant role, important role, or a role at least equal to that of other factors. This is certainly a heartening result for those who advocate evidence-based decision-making!

Perhaps bad news is that simply gathering data takes up more time than any other activity reported by analytics professionals. Having been influenced by the results of this survey, REI Systems is working on potential solutions that will help automate data gathering in several different contexts.  On another perhaps surprising note, another significant use of time for analytics professionals is communication. In fact, survey respondents reported spending more time on communication than on designing and testing analytic models or operating analytic models. Perhaps this should lead to communication training for analysts, or lead agencies to hire communication specialists to work with their analytics teams.

One hypothesis that our survey confirmed was that artificial intelligence has garnered lots of interest, but little action. Respondents believe AI is one of the most promising tools for improving government, but less than 4% indicated that they currently work with AI. Thus, there is almost certainly a skill gap between the few people using AI now and the future need. But there are also are clearly hurdles to adoption, else more agencies and survey respondents would be using AI now.  REI has conducted multiple experiments on relevant customer-specific problems to demonstrate what works and what doesn’t, using AI/ML for Grants, NASA, HHS, State of Utah, FEMA, GSA, and USCIS. In our experience while prediction with AI/ML is getting easier, converting that prediction into a judgement that guides organizational decision-making is much more difficult.

If you would like to obtain the detailed survey results, or view a video of a panel discussion about the survey, please visit Government Analytics Survey – Results and Analysis.  The panel that discussed the survey included:

  • Michael Conlin, Chief Data Officer for the Department of Defense
  • Patricia Hu, Director of the Bureau of Transportation Statistics
  • Collin Paschall, PhD, Sr. Lecturer in Government Analytics at Johns Hopkins
  • William Wiatrowski, Deputy Commissioner of the Bureau of Labor Statistics

 

And finally, if you have questions, please contact AnalyticsOffering@reisystems.com.