Our ability to use data as a tool for systems change is complicated by several factors that still need to be addressed.

Data can be a powerful tool for transforming the systems that perpetuate urban poverty. It allows us to understand current conditions, identify problems, set goals, and track progress for sustained impact. That said, our ability to use data as a tool for systems change is complicated by several factors that still need to be addressed.

The potential power of data is maximized when local decision-makers from multiple sectors form a collective impact partnership to tackle complex social problems. In these cases, data is used to continuously track indicators to identify which strategies are having an impact and which are not (with resources then redirected accordingly). Data is already being used in this way to improve educational outcomes in collective impact cradle-to-career partnerships across the country. While these efforts aim to prepare low income people to take advantage of future opportunities, we also need to consider how the cities and neighborhoods where low income people live in enable them to ‘connect’ to opportunities. But identifying, collecting, and tracking the right data can be trickier when you move away from individuals as units of analysis, and begin to instead look at places.

To learn about how communities around the country are using data to track neighborhood-level outcomes, I recently attended an annual meeting of the National Neighborhood Indicators Partnership (NNIP) - a network run by the Urban Institute of 37 data intermediary partners who are working to build local information systems, compile data publically for practical use, and support local capacity to improve low-income neighborhoods. By focusing on neighborhood-level data, local NNIP partners shed light on the importance of place in improving the lives of low-income people and the communities where they live.

As I thought about how cities can harness neighborhood-level data to lead collective impact for systems change, three themes emerged:

Neighborhood data is powerful, yet imperfect, messy, and often misused. Many organizations are doing amazing work with neighborhood data – such as the Institute for Urban Policy Research at UT Dallas’ work to help a local health foundation identify specific blocks for targeted interventions around preventable ER visits. Yet while the idea of data as a tool to inform decision- and policy-making is incredibly appealing, it’s often very hard to realize. The lack of standardized data across geographies and the lack of reliable data are just a few challenges to data quality. As a result, when a nuanced analysis of complex data is shared, it can easily be misused and misinterpreted. Local context that explains abnormalities is easily ignored for quick sound bites. To counter this, we need to invest in building robust local information systems, which will often require new levels of collaboration across government agencies, researchers, funders, service providers, and other groups that require, request or collect data.

Access to data is only the tip of the iceberg. Gaining access to data from government agencies can be a struggle in itself. But even after access is granted, lots of work is needed to clean, update, analyze, and properly share data. Local NNIP partners wear many hats in attempts to address these needs. Some have launched educational workshops around data-driven decision-making; others have spent years getting local agencies and funders on the same page around indicators to build robust information systems; and all spend significant time responding to requests for data. The biggest challenge appears to be the sheer capacity needed to analyze data while handling these other functions. A successful collective impact effort that uses data-driven decision-making will need to take into consideration the significant maintenance and analysis that appropriate and accurate data requires.

We don’t always know what the right data is. Depending on the big audacious goal of the collective impact effort, the right data to track may not even be known. One exception is with cradle-to-career efforts that measure student-level data to track progress; in this case, it makes perfect sense to look at student performance as the target outcome for the educational system. Yet the link between data and outcomes is not as clear within the economic or workforce development fields, especially on a neighborhood-level. In fact, many places have struggled with developing appropriate indicators for identifying and tracking progress in these other systems. The lack of clarity on appropriate data presents a challenge to any collective impact effort seeking to re-engineer the economic or workforce development systems. What does the “right” data look like for these systems? As we think about how to improve low-income neighborhoods, we’ll also need to improve the tracking and measurement of outcomes in fields other than education.

While data is essential in collective impact for systems change, the effective use of it is more complicated than most realize. We need to invest in the infrastructure for the collection, analysis, and effective use of data, especially if we want to harness its potential to improve the actual neighborhoods where low-income people live.