Potential presidential candidate Martin O’Malley speaks about his data-driven approach to governing at the Brookings Institution in Washington, D.C. in March. (Madeline Fox/MNS)
WASHINGTON — Martin O’Malley wouldn’t bite.
“Frankly, I’m a little sick of the email drama,” the potential 2016 presidential candidate said in response to a question about Hillary Clinton’s latest controversy in early March.
The “drama,” the former Maryland governor and Baltimore mayor might say, would be a waste of time. And a potential response to the flap, his calculations may indicate, would do more harm than good.
So, O’Malley abstained.
It was indicative of the data-driven governing style he used as a chief executive — and had spoken about a few minutes earlier at the Brookings Institution last week. Business-like and pragmatic, O’Malley detailed how he used statistical models to improve policy outcomes for the agencies he oversaw.
If the numbers added up and pre-determined outcomes were reached, a policy worked. If the data didn’t align, and performance fell short, the policy had to be tweaked.
“It’s not about excuses, deflecting blame, or ignoring problems; it’s about transparency; it’s about accountability; it’s about performance management. It’s not about left or right…” he said in a speech at Brookings. “It’s about setting clear goals, measuring progress, and, quite simply, getting things done.
“The old ways of governing — bureaucracy, hierarchy — are fading away,” he said. “A new way of governing is emerging.”
When O’Malley took the reins in Baltimore in 1999, he implemented CitiStat, a performance-based management system that used metrics and digital mapping to address policies in sectors such as education, justice and infrastructure. It was based off CompStat, a program installed in the New York City Police Department in 1995.
After the Democrat won the governorship in 2007, he scaled up the system to crunch data from across the state. After speaking with experts from each agency, O’Malley and his team would then use available metrics to map out performance goals and implementation plans. The program didn’t necessarily collect new data, but organized and aligned figures it already had to help meet policy goals.
“We used data as a backdrop for dialogue that’s ongoing, consistent, regular and very routine,” said Beth Blauer, who led O’Malley’s StateStat program in Annapolis. “It wasn’t that we were just taking the data and analyzing it in a vacuum, but we were having weekly or bi-monthly conversations with agencies around their data. We were using that data as a backdrop and we were using the outcomes as sort of the North Star.”
The idea, Blauer said, is to use accurate data, in the proper context, to drive policy discussions with the goal of improving the lives of citizens.
But a caution flag. According to Harry Hatry, the director of the public management program for the Urban Institute in Washington, the data-driven model uses sound logic. But he is concerned about providing relevant context in order for the numbers to be used responsibly.
“If the data-driven government — and there’s a danger of this — does not also consider the analysis element to try to explain what the data means, you’re going to have a problem and a misuse of data,” Hatry said.
In an effort to avoid this problem, StateStat publicly tracked easily understandable goals.
For example, the O’Malley administration aimed to “improve student achievement and school, college and career readiness” in Maryland by the end of 2015 — a goal not unique to the Maryland governor, but common among many chief executives across the nation.
But with StateStat and weekly meetings with education leaders, his administration determined that a 25 percent improvement in student achievement and career readiness by the end of 2015 was a logical and reasonable goal.
The metrics set smaller goals, such as increasing the number of AP exams taken with a score of three or higher. The state tracked progress toward its goal on a website, where the public could view how students were performing in different areas.
By the end of O’Malley’s second term in 2014, Maryland had reached its goal. Education Week magazine ranked the state’s public school system as the best in the nation for five consecutive years during his tenure.
But it didn’t come without a cost.
O’Malley raised taxes throughout his eight years in office and his handpicked successor, Anthony Brown, lost in the 2014 gubernatorial election.
Maryland’s system, as even Blauer would admit, was not perfect. For example, the administration’s goal to reduce carbon emissions was not met. In fact, she said, Maryland actually emitted more greenhouse gas during O’Malley’s two terms.
The reason for falling short, she said, is because it’s difficult to measure factors not easily picked up in data. In the case of emissions, Blauer said it was near impossible to stop people from using cars to drive to work, for example. The goal served more as a “call-to-action.”
After working for O’Malley, Blauer moved to the private sector to work for cloud software company Socrata. In her role, she works with software developers to create Open Performance, formerly named GovStat, a data software program adapted for the public sector and federal government.
The current state of the federal government’s use of data is inefficient and lacks transparency, according to Blauer. Not only is it possible to expand a StateState-type program to the federal level, but it’s necessary, she said.
“I was an inefficient bureaucrat myself,” she said in a phone interview. “I think that a lot of the norms in government were created before the Internet and so there are processes in place that have not evolved consistent with what technology would allow…There are a lot of norms involved that … that are completely irrelevant in government now, but, out of convenience, they’re still in place.”
Agencies such as the Department of Housing and Urban Development and the Federal Emergency Management Agency use effective data-driven policies, she said. In 2011, President Barack Obama signed an update to the Government Performance and Results Act, requiring federal agencies to publish strategies, goals, measurements and outcomes to improve transparency.
In his Brookings speech, O’Malley said that he increased transparency through the use of open data — figures the public can readily access and easily understand.
Hatry said that the open system “sounds excellent on paper,” but he doesn’t think that prevents the government from foul play.
“Certainly, government should be more transparent, but the question is: what does that mean?” Hatry said. “We’ve got all these databanks that, now, are becoming very, very available. New technology, now, opens up a tremendous potential source of data. How that data is [able to be manipulated] is going to be an issue. And how it can be used properly should be very important to everybody.”
Privacy is another concern with the data-driven model. It is something Hatry has heard from people with worries in both the public and private sector. Some laws, both at the state and federal level, bar agencies sharing personally identifiable information with one another.
But according to Hatry, there are ways to get around these laws, which could allow the government to share information once thought to be private.
“If you have the right security and you can get the agency personnel and the lawyers together, you can work out memos of understanding and agreements,” Hatry said about privacy concerns.
Blauer said privacy was a barrier in using certain datasets during her time working for O’Malley, but the administration made a conscious effort to find “the best way to view [data] without sacrificing [privacy].”
Although still close to the former governor, Blauer did not say if O’Malley would run for the presidency in 2016.
If following his governing style, he might still be making calculations.
