On June 18, 2013, Stephen Goldsmith testified at "Reinventing Government," a hearing of the House Committee on Oversight and Government Reform.
Testimony of Stephen Goldsmith,
Professor of Government at Harvard Kennedy School,
Before the U.S. House of Representatives
Committee on Oversight and Government Reform
June 18, 2013
Mr. Chairman and members of the Committee, thank you for the invitation to speak on government efficiency and performance in the context of the possible creation of a Government Transformation Commission. My name is Stephen Goldsmith and I am currently the Daniel Paul Professor of the Practice of Government at Harvard Kennedy School. In my time leading the Innovations in Government program at Harvard and my years of public service in Indianapolis and New York City, I have studied, evaluated and participated in numerous efficiency projects.
Across the country at every level, public employees each day produce effective services and even innovative breakthroughs, but they often do so despite the structures of government which restrict discretion, value activities over outcomes and fail to utilize the best of new technologies. I endorse a Commission which would broadly and fundamentally suggest new ways for taxpayers to receive the services they deserve at a price they can afford. Such a Commission should look at the following areas.
1. Government needs to produce solutions, not activities, through outcome-oriented governance and a better balance of accountability and discretion in civil service.
For 100 years now, we have developed civil service, oversight and procurement systems to limit corruption and abuses of discretion by limiting discretion itself. Narrow job descriptions, layers of outdated oversight, and hyper-technical and protracted procurement processes have produced a government that manufactures widgets designed for a single application, while entangling their production in red tape for reasons that virtually no one can remember. Today, as digital systems replace paper-based ones, analytics allow ongoing visibility into the actions of public employees—GPS chips in mobile tools tell us where employees perform their work, how long it takes to accomplish a task, and data helps supervisors better identify high performing employees and those who need help. All these technologies allow us to grant more discretion to workers, more flexibility to businesses with good records, and more trust to private companies partnering to deliver public services. Concurrently these approaches allow government to better concentrate its resources and attention—whether on regulating bad actors whose conduct requires thorough regulation and enforcement, helping individuals with the most intractable problems or producing public goods. For now though, the status quo encourages “safe” routine activity at the expense of innovation and solutions.
Almost inadvertently, bureaucrats remain preoccupied with tracking inputs and activities while spending too little time producing results. And if we pay for activities, as governments routinely do, we will always have more demand for those activities than we can afford and less demand for real results. For example, if we pay for visits to the doctor, that is what we will get. But if we pay for preventive services, we will be much more likely to get improved health outcomes.
Too little thought at the appropriation or agency level involves a true definition of public value—what outcomes are important to the public. Instead, we fund and track inputs—homeless shelter beds rather than the reduction of homelessness, or visits to the doctor instead of good health. Neither those receiving services nor the taxpayers who pay for them are well-served by spending money on activities that are not focused on solutions or delivering services that are too narrow, wasteful and impersonal.
Risk avoidance permeates the context in which most government employees work. Promotion depends on following rules and avoiding mistakes, not on reform. The problem here is one of incentives. We try to reduce risk by conditioning the behavior of public decision-makers to adhere to rules and procedures.
We have a government built on hierarchies that dramatically increase expenses while simultaneously suffocating continuous improvement. Reorganization needs to unlock ideas from those who do the work, rewarding their suggestions and implementing them. For example, IdeaHub, an online community for employees of the U.S. Department of Transportation, enables substantial virtual ideation and collaboration within the department. IdeaHub empowers employees to contribute innovative ideas and work together to develop ideas for improved agency performance and efficiency. In the past three years, the program has collected about 6,000 ideas serving as a model for how to digitally encourage innovation and ideation from within government.
Funding outcomes, whether internally in what an agency shows Congress it has accomplished or in a procurement will drive results. For example, the New York City Center for Economic Opportunity combines public and private venture funding, an avowed interest in innovation and willingness to tolerate failure to develop new approaches to reducing poverty. Its most recent initiative, "Social Impact Bonds" (SIBs), uses creative funding to purchase outcomes rather than the more traditional process of awarding funds based on a response to a highly prescriptive RFP. The first effort, an "Adolescent Behavioral Learning Experience" program, will pay for reductions in recidivism among black and Latino adolescents at the Rikers Island correctional facility. The vendor can use whichever programs it bundles together but it will be paid for results, not how much drug counseling or the amount of mental health services it provides.
2. Change the structures of government to incent results.
A government that often hires by list, not skill, that promotes irrespective of managerial ability, and that almost never penalizes or terminates bad performers and rarely rewards excellence cannot expect excellence. Similarly, legislative bodies need to align agency-level incentives—good performers need more autonomy. For example, allowing agencies and their subdivisions to retain part of any efficiency they generate changes conduct. During a meeting at which I solicited suggestions for improvements, New York City’s wastewater plant operators responded, almost in unison, that the procurement rules caused bad decisions. They wanted procurement flexibility, allowing them to purchase in real time rather than observing imposed calendar deadlines and use it or lose it rules which caused them to stockpile parts well before they might be needed. In most jurisdictions, procurement stifles innovation and effectiveness. Procurement that provides flexibility, speed, encourages innovation and good bureaucratic behavior, uses the web to purchase, places everything online for maximum transparency, and utilizes analytic tools to monitor the behavior of vendors and procurement officials alike will drive maximum benefit. For the most part, public employees do what they do to serve the public. They would rather take pride in their work than lapse into waste, sloth or malingering. But that motivation is dampened when systems and supervisors ignore excellence and treat the high and low performers equally in terms of pay and promotion. Though difficult, we need to begin to identify ways, free of favoritism, that reward employees who excel and quickly, fairly and firmly discipline those who take advantage of the public with wasteful ways.
Outmoded job classifications, slow hiring, stagnant promotion, too much overhead, too little recruitment of key new talent sets and a reluctance to hold employees or agencies accountable paralyzes progress. Congressional oversight and appropriation should spend more time monitoring whether agency outcomes will produce better value.
Structural changes specifically designed to promote innovation are also key. On the municipal level, Innovation Delivery Teams have been working in Atlanta, Chicago, Louisville, Memphis, and New Orleans to increase the cities’ innovation capacity and address city challenges. Another model is LAUNCH, a public-private partnership of NASA, USAID, the State Department, and NIKE working to develop innovative solutions to large-scale issues like clean drinking water. By creating innovation capacity in terms of time, information and funding, allowing new approaches and tolerating informed risk outside of the pressures of day-to-day operations, new structures will result in transformative progress.
3. Realign the public/private relationship.
Competition forces all of us to be better. This concept applies equally to bureaucracies unchallenged by competition and long-term incumbent vendors. Competition pushes those involved to improve. Allowing public-private partnerships and competition translates into better services and reinvigorates public workers. A recent example came when Chicago Mayor Rahm Emanuel had city employees compete against private companies to see who could pick up recycling more efficiently.
Yet performance based contracts involve complicated work on the part of the acquisition workforce and the federal government needs to enhance the skills of those who design RFPs, moving away from prescription to purchasing innovation and results in a manner that induces fair competition. Yet this does not need to be a zero sum game. The issue no longer is whether a public service should be privatized but how to integrate the public and private pieces to provide the best value. Almost every complicated activity requires public, private and often nonprofit partnerships, clear contract deliverables, good project management and high quality oversight.
4. Data analytics and open source transparency will reveal root causes and save money.
Data analytics centers are now evolving into one of the most powerful drivers of public innovation, allowing predictive problem solving to be delivered to the field through decision support tools. Several governors and mayors are now actively involved in determining how to set up special initiatives dedicated to driving reforms through digital analytics. These data-driven efforts help officials target services, allowing them to both anticipate problems such as weather emergencies, and improve services like child welfare outcomes.
Data analytics can save money through fraud detection in areas such as taxation or benefit eligibility. Data can also reveal duplicative government programs and those that do not target the root causes of social problems. Outcome-oriented government also intersects with data analytics. In my first week as deputy mayor of New York, top fire department officials explained to me the futility of trying to inspect every building in the city, which led them to rethink their approach and goals. The new approach included analyzing data from the departments of buildings, health, and the police to identify structures and operators with a history of serious infractions that needed more urgent attention and remediation. The mission of reducing fire, not increasing the number of inspections, would reset the allocation of resources at NYFD.
Yet we will not recognize these benefits at the federal level without materially changing the way we operate. Truly significant discoveries occur by integrating data systems across silos. Too many agencies have their own data standards and systems, preventing the kind of analysis that reduced fire in New York City.
5. Government needs to operate horizontally rather than vertically.
Big governments produce inefficiency and restrict problem solving because they are organized hierarchically, with too many layers as mentioned above, but also because the vertical nature of public agencies increasingly prevents us from designing effective solutions to public problems. Citizens live in neighborhoods, not in a transportation or workforce agency. Our government frequently operates as a set of independent and disconnected entities. Real efficiencies are possible by treating government as an integrated enterprise both around priority objectives (for example, food safety) as well as support functions (IT, HR, finance etc.). Both Congress and the Executive Branch should pay more attention to managing portfolios across agencies, aligning services, reducing duplications, discovering solutions and addressing citizens with more relevance.
We need horizontal government not just in how we react to citizens but also in how agencies relate to each other. Most government officials now talk about shared services—the consolidation of functions across governments or government agencies—but few have done it broadly and correctly. As deputy mayor working to implement shared services in fleet, accounts receivable, real estate, IT and human resources, I found that each agency had what at first appeared to be a compelling reason why it should be exempted from a shared service. Coupled with not wanting to lose control was a reasonable concern that commissioners did not want their service quality hindered by a shared service monopoly over which they had no control. Yet potential savings when completed would reach a compounding $500 million a year.
Capturing savings from shared services requires setting the bar quite high, strong executive leadership, clear customer service agreements so that participating agencies do not become prisoners of a larger bureaucracy, carefully aligned incentives including shared savings, moving quickly, and finally setting up a real market so constituent agencies can still access outside options on occasion.
The federal government obviously is a massive operation. Yet every single dollar spent by every single employee needs to drive maximum value. That requires structural and cultural changes. When I started our efficiency work some time ago in Indianapolis, I noticed how difficult it was both for my hardworking employees and the legislative branch to conceptualize a bold new way of doing business. A well-intentioned defense of the status quo crept in. Eighty managed competitions later, we had saved over $400 million, repeatedly reduced taxes, invested over $1B in infrastructure and produced a group of union workers who were proud of what they accomplished and pleased to be rewarded for it. It can be done in Washington, but not without new rules, new approaches and a new definition of oversight.