Stephen Goldsmith: The Basics of Big Data
This is part of a series of interviews with leading practitioners and thought leaders on new approaches and solutions that have been proven to work. Stephen Goldsmith is a Professor of Practice and Director of the Innovations in American Government Program at Harvard’s Kennedy School of Government. He currently directs Data-Smart City Solutions, a project focused on government efforts to use and blend new technologies, big data analytics and community input. He previously served as Deputy Mayor of New York and Mayor of Indianapolis. Goldsmith also writes a column for Governing magazine entitled Better, Faster, Cheaper.
What advice would you provide to a newly elected mayor and would you have different suggestions for small, mid-sized or large cities?
Common principles apply despite issues of size. Successful mayors are those that combine an aspirational vision for their city with management and operational excellence. You can’t aspire to be great and have basic services that don’t work and the converse is also true: effective service delivery without a guiding vision marginalizes results. Too often mayors are elected who are really good at one of those two and not so good at both. You don’t have to necessarily excel at both, but you do need a balance; so if you’re good on vision and not on operations, then you simply must bring someone in who is.
Many cities are focusing more on data. This can mean new performance measurements, ‘big data’ or data analytics. If a city wants to advance a strategy where should it begin?
Start with basics. Big data seems daunting as many cities don’t have the money or requisite skills to advance such a strategy. Cities do not need to go straight to big data—it can be overwhelming. Municipal leaders need to just pick a few discrete areas to begin. A place to start is to simply ask how do you measure performance? Do you have a stat-like program or another way of gathering basic data? The goal should be a continuous effort that provides value in specific areas and which produces useful insights into what is and is not working. Analytics can highlight recurring problems and their causes. A framework can then be established to resolve the underlying causes, which can then translate into continuous results. This whole process involves asking realistic questions that drive solutions.
That is a great explanation of getting started. But what if you do want to advance and take on higher-level analysis and data analytics?
I would use a very simple analogy to frame a more advanced approach. I tend to think about this as ‘verticals’ and ‘horizontals.’ People live horizontally, yet government and government agencies tend to operate vertically. This is a fundamental mismatch. So think about what it takes to unlock answers to big problems: if you’re trying to understand crumbling families it may be drug issues or housing or something else. The point is that all of our really important problems have multiple sources and we need to get a handle on all of them. So if we are talking about truly reengineering cities and tackling the big stuff, a horizontal approach is required. To do this, cities must go outside of just tracking data in a vertical (agency) silo. These challenges require multiple agencies interacting with each other through many different approaches, including the sharing of data.
So in sum, cities should start with a data driven performance program – in many cases this may be transactional data that is collected at the agency level – then cities should introduce basic data mining and analytics to identify the root cause of specific service challenges. From there, cities can advance to predictive data analytics to solve more complicated and systematic problems.
Tell me what you see on the horizon or new directions to pursue in the area of data.
The development I’m most interested in is cloud computing. I think this will quickly blossom as a major area for cities; one that will greatly advance their data utilization and the convergence of mobility and a data driven government approach. Previously only big cities could afford the best when it came to integrating data and analytics. Today, cloud solutions will bring the best even to small and mid-sized cities. Cloud computing coupled with ubiquitous hand held devices has the potential to revolutionize governance across all governments, regardless of size.
What are your recommendations about economic development and job creation?
A mayor needs to understand the local or regional economy, understand the limits and opportunities—a mayor can’t invent an economy but he or she might be able to bend the curve. Of late, I have seen a lot of interesting work (around) eds and meds—universities and hospitals. Cities are (increasingly) working with these anchor institutions and aligning them with their competitive economic clusters. You see this happening in Phoenix and San Francisco where local universities are helping to stimulate the IT industry. Now, that alone is good economic development, but where it gets really interesting is when cities combine this approach of job creation with geographic rejuvenation. Some cities are looking at how this university or hospital-fueled development and affiliated job creation can improve a specific neighborhood or part of town, which is particularly interesting
I do also want to point out the need to think about equity. A recent Brookings study shows that cities with high growth rates have the highest income disparity rates. So cities want to close disparity by creating jobs but by creating such (knowledge-based) jobs they are left with a growing opportunity gap. This is a huge conundrum for cities. So I’m really impressed when places take this on with a focus for creating wealth and opportunity. One good example is the health-based career ladders work being led in Boston by the Boston Foundation. The most creative cities taking on economic development are also looking at upward mobility; they’re targeting economic clusters, but are also committing to lifting up struggling residents working within that sector.
This interview was conducted by Neil Kleiman, National Resource Network Director of Policy, Research and Evaluation.