Time to Modernize Our Approach to Data

We are living in a golden age of data; there's more information available than ever before. Consequently, the U.S. government increasingly depends on data to tackle societal challenges. The 2018 President's Management Agenda states that the "data is transforming society, business, and the economy." If the government fails to use information effectively, "it will no longer be able to fulfill the trust placed in it by the American people."

The federal government has historically struggled to collect and interpret data. Consider the boondoggle of USASpending.gov, a government website launched in 2007 to track spending across all federal agencies. The site failed to properly report $619 billion in federal spending, according to a 2014 report from the Government Accountability Office.  

In June, the Office of Management and Budget launched a website to solicit public input on a new Federal Data Strategy, which will, if properly deployed, begin the process of properly managing the fuel necessary to power IT Modernization.  

Many people want this strategy to prioritize open data transparency. It's a necessary goal—transparency for government unique data is essential to a healthy democracy. But well-meaning but overly broad transparency requirements will create unintended consequences by curtailing access to other essential types of data.

That's because much of the data essential to government operations and analysis arises in the private sector. Private data is a category that isn't collected by the government. It's leased from private companies who have gathered and analyzed that information, often at great expense. Releasing private data to the world under an “open data” mandate would deter these firms from partnering with the government.

If federal agencies lose access to the highest-quality data from whatever source, everyday citizens would suffer, since incomplete or inaccurate data leads to bad decision-making. Government officials must have access to information—some call it "actionable intelligence"—not the underlying raw data. That requires unfettered access to all forms of data, whether public, private, or a mashup of the two.

Many federal agencies have wisely turned to a mix of public and private data to complete their mission. Today, it's increasingly common for agencies to utilize a mix of public and privately collected data—what I call "hybrid data" in a new paper.

For instance, the Transportation Department recently launched a project that will integrate government data with data from Waze, a smartphone navigation app. The collaboration could help reduce fatalities from car accidents. Similarly, a new government effort to provide timelier flood warnings to residents of Virginia Beach combines data from software developer Esri with data from the National Institute of Science and Technology and the U.S. Geological Survey.

Private companies invest billions of dollars and decades of time into developing these high-quality, accurate datasets. To protect these investments, both government and its contractors naturally require some restrictions on public release or use.

Unfortunately, many legislators, regulators and other public officials believe that all forms of data in the government's possession should be "open;" they argue it's akin to a public right. But that could make public-private data collaboration far less common.

If data became publicly available because of a well-intentioned mandate, few data firms would be willing to do business with federal agencies. The quality of government services would decline, as policymakers would have to base critical decisions solely on the data gathered by the government itself.

Even worse, such changes could lead to inaccurate or outdated data. Private companies would no longer be able to control the quality of information; anyone could use or corrupt it.

Look at what happened in the early 2000s in the Defense Department. Officials sought to establish a mandate that all software adopted by the department should be freely distributed without limitation as so-called “open source” software. The policy caused a great deal of confusion—and required a detailed OMB memorandum to clarify restrictions.

Now, with a new Federal Data Strategy on the horizon, we must encourage open government without destroying the incentive for public-private data collaboration. To achieve this balance, Congress should initiate a Government Accountability Office study that evaluates public, private and hybrid datasets. This study can build off of existing open government efforts—for example, the Commission on Evidence-Based Policymaking's final report, which provides recommendations for how government can improve its use of data.

Policymakers can then craft carefully tailored restrictions that enable firms to share data while keeping their intellectual property.

More than ever before, the public can use government-created data to power private-sector innovation, increase civic activism, and fight corruption. That's as it should be. But when government tries to extend this principle to privately owned or hybrid data by default, the consequences can be devastating.

We must create a Federal Data Strategy that properly addresses all forms of data: public, private and hybrid. The best decision-making demands the best data, and we need government, and the private sector to work together to ensure it can still be produced.

Richard Beutel is a principal at Cyrrus Analytics LLC.