Internet Explorer 11 is not supported

For optimal browsing, we recommend Chrome, Firefox or Safari browsers.

The Governance Trinity, and How Information Can Power It

Governments are getting much quicker and clearer feedback than ever before. It’s data that could be better used to solve problems.

As social animals, we mostly work together. That opens up benefits of scale (armies are stronger when they are bigger) and specialization (workers are better when they do what comes naturally and learn through extensive repetition).

But when we work together, we must decide who does what and, more controversially, who gets what from the results. This is the governance problem, and it's always been tough.

However, I'd argue that governance has become even more difficult in recent times. More interactions now flow across borders of authority, so we don't have a well-tested and accepted process for resolving conflicts. And more of our interactions, even within borders of authority, now require innovative rather than routine behavior, so again we don't have an effective process for resolving conflicts.

Fortunately, we do have a powerful capacity that we didn't have before, one that helps solve important governance problems when used well: much clearer and quicker feedback made available through the productivity improvement of digital information, processing and communications. Systems with quick and clear feedback are famously easy to learn and control. A bicycle is the classic example.

Feedback in government, however, is often anything but clear, largely because those who receive the benefits are usually not those who bear the costs. That's different from commercial transactions, where sales can be used to track value received and profits are a reasonable measure of value added. Government has no comparable measures of value added. And government feedback is often greatly delayed; elections are critical, but neither clear nor timely.

But if we're smarter, we can make progress by taking straightforward advantage of the newly available information and analysis. What we need to do is to follow the basic trinity of governance:

Goals: Set motivating and verifiable goals for groups and the units and individuals within.

Gaps: Measure and analyze the differences between goals and results.

Guidance: Decide on the rules and roles needed to close the gaps.

A sad reality is that too many governments today fail to make effective use of the information that is now available. We don't set very good goals, don't analyze the gaps between goals and results, and don't decide on clear but minimalist rules and roles for corrective guidance and action.

To see an example of what we need to do, let's look at the recent history of policing, and particularly at the CompStat methods developed under William Bratton when he was New York City police commissioner. In taking over the NYPD, Bratton initiated CompStat as a key element of his aggressive and comprehensive effort to improve departmental performance and reduce crime. Here's how Bratton made effective use of the governance trinity:

Goals: Where history and the city's police culture suggested that a 1 percent reduction in crime might be all that was possible, Bratton's team set a goal of 10 percent, to be achieved year after year. The goal was translated downward into specific plans for objectives such as getting guns off the streets, reclaiming public spaces, curbing youth violence in the schools and streets, driving drug dealers out of the city, and breaking the cycle of domestic violence. Each goal contained measurable targets.

Gaps: The CompStat process was based on feedback on robberies, homicides, burglaries, gun-related deaths and other statistics that had in many cases been available but not aggressively used for timely problem-solving. The digital age makes much more data available for analysis including operational data (such as real-time car locations via GPS) and external data (such as from Facebook and other social-networking sources). CompStat used such feedback for a problem-solving and learning process. Using up-to-the-minute feedback, CompStat ranked precincts against themselves and against other precincts year over year, month over month and day over day.

Guidance: Guided by data analysis and by collaboration across levels of the police hierarchy and with other city institutions, CompStat created techniques for solving the problems that were identified. In general, commanders were freed up from many bureaucratic rules that had accreted over the years. In return, the commanders were held much more accountable for improving outputs (such as arrests) and outcomes (such as overall reductions in crime). As Bratton wrote later, "CompStat hinged on transparency, sharing and learning, accountability — and collaboration."

The results were truly impressive. Reported crime in New York was down 12 percent the first year, compared to 1 percent nationally, and it kept falling as Bratton continued with CompStat. CompStat took straightforward advantage of recent improvements in information availability and processing power. With many of the old rules eliminated, clarity of goals and gaps promoted better local problem-solving.

Many other governments and agencies have adopted the CompStat-style approach, and those that haven't should take advantage of similar strategies. What we need — and we need it desperately — is to use our new information availability to strengthen that basic and age-old governance trinity.

Special Projects