Charles Chieppo is a research fellow at the Ash Center of the Harvard Kennedy School.E-mail: Charlie_Chieppo@hks.harvard.edu
As local government revenue has decreased in each of the last several years, many cities and counties have responded by slashing infrastructure investment. The result is an unsustainable model. But a new report from IBM proposes a way to reduce operating budgets and use the savings to make the infrastructure investments that are needed to attract investment and create jobs. Now all we have to do is keep politics from standing in the way of some very good ideas.
In 1992, local governments were dedicating 16.6 percent of their spending to capital projects. By 2008, when the financial crisis hit and local government revenue began drying up, infrastructure spending already had fallen by about one-quarter, to 12.7 percent. In 2009, the latest year for which data are available, it had fallen further, to 12.2 percent.
And local officials learned some hard lessons. Preserving public safety jobs doesn't mean much when first responders' vehicles fail; emergency response personnel are helpless when their communication systems don't work. They also learned that aging infrastructure costs more to maintain.
In 2009, the American Society of Civil Engineers estimated that the nation's total five-year infrastructure spending needs were a staggering $2.2 trillion, about half of which was unfunded. Local governments' share of the five-year needs was about $550 billion. But according to the IBM report, achieving efficiency-driven cost reductions of 9 percent and investing those savings in infrastructure would eliminate the local infrastructure shortfall.
Among IBM's proposals for reducing local government operating costs is the use of advanced data analytics. The power of this approach is illustrated by the experience of Richmond, Va. Richmond was the fifth most dangerous city in the country in 2005, before it began using analytics technology to sift historical crime information and predict when and where crimes were likely to be committed. In 2010, the city deployed police based on the data on New Year's Eve, when crime usually spiked. As a result, random-gunfire incidents dropped by nearly half and the number of weapons seized more than doubled. Crime has plummeted, and Richmond is now the 99th most dangerous city in the country.
But advanced analytics aren't just about helping officials make better public-safety decisions. For example, they can tell us what the likely impact of building a new road would be not just on mobility, but also on police response times and access to green space. Officials could then compare the results to the likely effects from building a new park or another potential investment. Armed with this data, leaders can compare the relative return on investment for various expenditures.
Of course, cutting local government spending on a scale that would bridge the infrastructure funding gap would require political will. Richmond's predictive analytics software dramatically reduced police overtime, but will city officials follow through by cutting the department's budget by that amount?
Nationally, crime has fallen dramatically over the last 20 years, to the point where crime rates in many cities are similar to what they were in the 1950s. But during that time, local government spending on police services rose. Police organizations would surely argue that it was the steady increases in law-enforcement spending that contributed to the drop in crime.
When it comes to local government revenue, most economists agree that the future is likely to look a lot more like the recent past than the preceding decades. If they're right, the existing financial model just isn't sustainable. But even if they're wrong, do we really want to spend more if we could get the same or better results for less?