As advocates both for transparency and for open data, we spend quite a lot of time at Sunlight thinking about precisely how open data relates to government transparency. Though my colleagues and I have provided many examples of how open data can be used to improve transparency, we also know that open data by itself does not automatically produce the transparency that allows us to achieve accountability.

Open data is a tool, not an end in itself, much as a hammer is not an end in itself.

Open data has a wide variety of supporters because it is a principle that can be used to accomplish a number of disparate goals. Sunlight and other transparency organizations who want to increase the amount of open government data use it in order to hold government actors accountable. Private companies looking to increase profits want more “high value” data that can help them make more money. Civic hackers want to apply technology to open government data to solve governmental problems. Community organizations want to use open government data to support community health and growth. All of these organizations want more open government data in order to advance their work, but we’re using open data to pursue different goals.

The recent success of the DATA Act in the U.S. provides an example of the way that open government data represents a neutral tool that different advocates value for different reasons. Despite the challenge of passing almost anything through our famously gridlocked and partisan Congress, the DATA Act overwhelmingly passed both chambers. This is because the measure provided different parties with a tool to achieve different goals. Republican Rep. Darryl Issa, Calif., celebrated the passage of DATA Act as a way “to root out waste, fraud & abuse.” Democratic sponsor Sen. Mark Warner, Va., described the DATA Act as “but the first shot of a technology revolution that will transform the way we govern,” linking the effort to his committee’s work on improving governmental performance and efficiency.

Meanwhile, the White House’s Office of Management and Budget (OMB) managed to use the flexible tool of open data in its effort to inhibit transparency. When OMB wanted to limit the degree of spending transparency enabled by the DATA Act, it did this by substituting language advocating use of “open data structures (e.g. using a widely accepted, nonproprietary, searchable, platform-independent formats)” in place of existing language that required more rigorous disclosure standards. This cynical use of open data to avoid increased transparency evokes the argument made several years ago by Harlan Yu and David Robinson that open data efforts can be used to disparate ends – and even against actual transparency. It is not simply beneficial by default.

So if open data doesn’t produce benefits by itself, how does it work? Through the active and intentional involvement of human actors. Advocates need to ask for the specific data they need to achieve their goal. Advocates need to persist in seeking and obtaining the specific data they need, even if they are confronted with obstacles.

This is particularly true where the topic is controversial, as is often the case with transparency work. Sunlight’s mission of identifying improper influence over the policymaking process requires that we get data reflecting efforts to influence lawmakers, like campaign finance and lobbying disclosures. We need data about policymaking outcomes, like legislative data, expenditures, contracts and procurement processes, and we need information about the relationships that might enable or predict improper influence, such as asset disclosures, corporate ownership structures and nonprofit funding data.

Just because we now seek this information under the rubric of open data doesn’t necessarily make it easier to get. While the democratic system as a whole benefits from transparency work, individual government actors may not. At a minimum, these datasets mean greater work and scrutiny for individual political actors; at maximum, these datasets create a greater likelihood that they’ll be caught doing something embarrassing, unethical or illegal.

Once open data is in place as a tool, it is up to the organizations involved to wield it properly to achieve their desired ends. From our perspective, that means we need to ask the question: Will the open government data initiatives we witness result in improved transparency? Will they help us better achieve accountability?

Where “open data” does not help us achieve our ends as transparency advocates, is this a problem with open data itself? At what point do we call the hammer a bad hammer, and at what point do we reexamine our technique?

Using open data to achieve your goals first requires being clear about what your goals are and then advocating for the specific data that will help you meet them. Sunlight has a history of advocating for transparency data on the federal level, including our recent decision to file complaints against 11 television stations for failing to disclose mandated information about campaign advertisements. At the international level, building consensus around the need for strong political finance transparency norms is the main focus of our new Money, Politics and Transparency project.

We advocate for transparency data in different ways on the state and local level as well. Most recently, we have called for the opening of public municipal finance data centrally collected by the Municipal Securities Rulemaking Board. We recommended that D.C.’s recent campaign finance reform require e-filing as a way to improve the quality and speed with which campaign finance data could be made available. We continue to look for opportunities to be sure that transparency data is represented within collaborative efforts to describe specific data that municipalities should proactively make available, like the US City Open Data Census.

If we want open data to open government, we need to say so loudly, clearly and often. We must let governments know when the specific transparency data that we want is not there, is incomplete, or is insufficiently accessible. When the National Weather Service tried to change the data it released, the multi-billion dollar industries that rely on weather data effectively lobbied them to define their product in a way that worked best for those private industries. Organizations focused on transparency and accountability need to do the same thing. Using open data to achieve our goals means clearly demanding the specific data that we want to be opened.