What’s Left for Government to Do?
From the January/February 2008 Issue
Outsourcing champion STEPHEN GOLDSMITH examines the question of whether governments really have ‘inherent’ or ‘core’ functions.
It is difficult to get Democrats and Republicans to agree about much of anything these days, but politicians of every stripe are likely to concur on this proposition: government at all levels has promised much more than it can deliver, and every year the gap grows worse. One potential solution to this intractable problem is the outsourcing of government functions to the private sector, but it appears almost certain that any significant steps toward privatization are going to occur at the state level. Why? Because that’s where the most exciting work is going on right now, and because Washington appears allergic to public-private innovation. Congress not only clings to the view that government workers must produce all government services, but also even tries to impede progress when it surfaces outside the Beltway—witness the provision tucked inside the 2007 farm bill that would prohibit outsourcing by states and require government employees to process all applications for food stamps.
Congress’s resistance to innovation involving the private sector reminds me of my service a few years ago on a working group appointed by the head of the Government Accountability Office (GAO), Comptroller General David Walker, to evaluate the outsourcing of U.S. government activities. The previous decade had seen a dramatic rise in the number of private contractors performing the federal government’s work—with a corresponding drop in the total number of public employees—and Walker’s Commercial Activities Panel seemed to be bringing together the right people at the right time to grapple with the full complexity of outsourcing issues. Yet almost immediately the panel got bogged down, wrestling instead with the question of what constitutes an “inherently governmental” function—a technical term for work that is too important, or too public, to be done by contractors.
Functions such as law enforcement and public safety are typically considered inherently governmental. But as the mayor of Indianapolis for two terms (1992–1999), I outsourced the operation not only of a jail but also of an airport and a utility—with good results. It was hard not to be amused, as the GAO panel debated the definition of an inherently governmental function, by the fact that the federal building we were meeting in was under the protection of a private security firm. Even if we could agree on core government functions that had to be walled off from contractors, we would be left with a thorny question: what happens when government turns out not to be very good at inherently governmental work?
Even if we could agree on what constitutes an 'inherently governmental' function, what happens when government turns out not to be very good at inherently governmental work?
At the state level, several governors and legislatures are not getting hung up on philosophical questions and are instead rushing to try private-sector solutions to pernicious public-service problems. Officials do so at significant political risk, but the potential for success—and for breaking with age-old failures—seem worth the gamble. In 2004, Mayor Richard Daley secured a $1.8 billion deal to lease the operation of the 7.8-mile elevated toll road known as the Chicago Skyway to a private consortium for 99 years. Two years later, Indiana made a deal with the same consortium: a 75-year lease for $3.8 billion to operate the Indiana Toll Road. Other states have outsourced the operation of toll roads and bridges, and more are considering it. Privatizing state lotteries is also under discussion.
But leasing out toll roads and lotteries is relatively simple; some states are showing much bigger ambitions. Texas has attempted to outsource the way that welfare eligibility is determined, for instance, and Florida has privatized its human-resources operations. Neither venture has found much success, but they offer lessons that might help officials elsewhere as they attempt their own innovations. Certainly, the efforts in Texas and Florida are an indication of how today’s public-private partnerships differ significantly from those of recent decades. No longer limited to ownership of a public asset, privatization initiatives nowadays are increasingly adventurous. Businesses and government are clearly ready to consider tackling complex policy problems using modern technology and the creative spark of private enterprise.
Perhaps no state or local effort in the last several years better represents the future of privatization than the welfare-eligibility modernization project launched last year in Indiana. The complex undertaking points to the substantial potential and pitfalls of government outsourcing.
In Indiana, a million Hoosiers depend on the Family and Social Services Agency (FSSA) for disability payments, food assistance, medical care, and other services. In 2005, Governor Mitch Daniels assigned a daunting task to FSSA Secretary Mitch Roob: reduce the growth of Medicaid by half, fix the struggling welfare agency, and transform its philosophy to one that encouraged self-sufficiency. Roob took the charge seriously. “In way too many instances, we are turning away people who probably do qualify for some of our services, and we are accepting people who don’t qualify,” he said at the time. “That is just no way to run our business. Our customers, the citizens of Indiana, deserve a system that addresses their needs for public assistance.”
Roob and his project manager, Zach Main, visited county offices across the state. Their report to Governor Daniels decried “terrible customer service” in the offices they visited. “Indiana’s neediest citizens, the ones who have the least access to transportation, have to drag along children or leave a job in order to jump through a bunch of hoops to access the system.” Analysts found that citizens in need of the FSSA’s help were forced to make more than two million unnecessary trips a year. Roob and Main had encountered firsthand a broken system in the state’s largest agency—it has 9,700 employees and a budget of $6.5 billion—that had inadequately served its disadvantaged clients while simultaneously avoiding media or government scrutiny.
Analysts found that citizens in need of help from the state Family and Social Services Agency were forced to make more than two million unnecessary trips to county offices a year.
Roob then took a critical step—one that is too often overlooked during bureaucratic overhauls. He hired independent third parties to assess the FSSA’s service quality, establish benchmarks for expected performance, and identify organizational risks. A KPMG review exposed what Roob and Main already sensed: caseworkers did not apply eligibility rules in a uniform manner. As Main later pointed out, the FSSA “had 107 county offices and 107 different ways of doing business. More likely we had 2,200 caseworkers and probably 2,200 different ways of doing business.”
Simply put, Indiana’s government was not very good at providing this important governmental service. And, unfortunately, governments that are not very good at producing public goods are often not very good at the contracting process either. Indiana avoided this pitfall by securing substantial outside consulting advice and relying on Roob’s experience, which included serving as the president of the Marion County Health and Hospital Corporation in Indiana and as the chief operating officer of Indianapolis Water, a privately owned utility.
In their quest to repair the dilapidated FSSA, Governor Daniels and Secretary Roob refocused the agency’s goals. First, the state sought to establish a strategic direction for the agency that emphasized big-picture issues: supporting work rather than paying benefits, and enhancing health rather then simply funding Medicaid help. Roob insisted that these strategic goals be accomplished in a system that would be made to work better for clients, caseworkers, and taxpayers.
Over the years, thousands of FSSA employees had become adept at applying rules and pushing paper but lacked the critical skills necessary to handle the agency’s transformation—regardless of whether it involved insourcing or outsourcing. Roob attacked this problem by adding to his team, hiring consultants and lawyers proficient at managing Request for Proposals processes, and establishing appropriate contract-monitoring techniques. Further, he was determined to rally state employees around his aspiration that caseworkers spend more time supporting people in need instead of wasting countless hours on paperwork.
By establishing these specific priorities, Indiana was able to concentrate on what is inherently governmental: expressing democratic values through strategy and policy while ensuring the equitable implementation of that policy. Shuffling pieces of paper or creating digital images of application forms is not inherently governmental, even when these actions support critical governmental functions. Determining the range of financial benefits and services necessary to promote self-sufficiency among struggling citizens—the FSSA’s mission—is inherently governmental.
Having envisioned the transformation of the FSSA, Roob and Main had to decide whether to produce the change internally or look outside the government for help. They found ample reasons to seek an external solution, including a lack of capital—both financial and managerial—and an abundance of constraints on internal change.
A transformation of the size and scope envisioned by Roob would likely require years to plan, develop, and implement. Consistent commitment and direction from the highest levels of the agency would be a necessity. But a lack of commitment and direction in the upper levels of the FSSA over the years was what had helped produce its problems. Since the agency’s founding in 1991, at least ten different secretaries had commanded the organization. As Main explains it: “A new guy would start every year and say, ‘All right, I’ve got the answer. It came to me from whatever the last management book was that I read.’ So maybe it’s ‘Six Sigma’ or maybe it’s ‘ISO 9000,’ or maybe it’s this or that. He’d start to pilot some things. They might work or they might not. But by the time he figured that out, he was gone. A new person was in who said, ‘Oh, I don’t want to do it that way. I’m going to do it my way.” This discontinuity is a prescription for failure. Locking the agency into a long-term agreement with an outside partner would impose a certain consistency even as changes were underway.
‘A new guy would start every year and say, “All right, I’ve got the answer. It came to me from whatever the last management book was that I read.”’
Indiana mirrors most other governments in that it lacks a separate capital budget; this makes it difficult to commit to the investment necessary to produce dramatically better services. Indiana’s strategic goals required a sizable up-front capital investment, which private partners could spread over the life of the contract. Even if the state had the financial capital readily available, Roob knew that another resource—proven technical expertise—was only available in the private sector.
To implement deep changes in the way that the FSSA’s 2,200 caseworkers and 107 county office directors had been doing business for years would require flexibility in the state’s merit worker rules. As written, the rules made changing an employee’s responsibilities extremely difficult. Even if the “new” private hires came from among existing state caseworkers, Roob and Main believed that bringing in a private company would solve the challenge of working around the innovation-stifling rules that applied to state employees.
Undertaking a transformation of the agency internally rather than working on it from the outside would have been difficult. Only a strong and experienced management could attempt it, and the FSSA didn’t have one. Its managers consisted primarily of county-office directors who were qualified but had risen through the ranks of the broken system, and its central office was staffed by political appointees who came and went with administrations. The middle management required to implement Roob and Main’s vision simply did not exist.
On a broader level, the state’s leaders had ample experience in outsourcing. Governor Daniels was an adviser to the city of Indianapolis during its period of comprehensive privatization reform, and he had been the director of the Office of Management and Budget for a period under President George W. Bush. In addition, Roob had been involved in competitive sourcing in the past for the city of Indianapolis. Indiana also benefited from consultants and vendors—many of them former public officials—who could see the challenges through the eyes of Roob and Main.
Despite all these reasons to look to the private sector, Roob and Main had to consider three factors that favored an internal transformation. None was related to the inherently governmental nature of the FSSA’s services.
First, no state had ever successfully outsourced its eligibility-determination system for all major benefits programs. Florida and Texas were the only states that had attempted anything similar, and, at the time, their prospects didn’t look good. Florida was on the verge of deciding to return to a strictly in-house approach, while Texas was encountering difficulties in the earliest stages of its program. The Texas experience, in fact, continued to provide Roob and Main important lessons. In 2005, Texas signed a five-year, $900 million contract with Accenture to perform an ambitious modernization of its eligibility system. The effort was almost immediately plagued by cost overruns, customer service problems, and coverage issues—thousands of children lost healthcare insurance—and Texas canceled the contract in March 2007. What went wrong? Texas Health and Human Services Commissioner Albert Hawkins has said that administrators “didn’t draw the line between state and vendor in the right place.” Additionally, Texas coupled its outsourcing effort with major system and policy changes that resulted in too many large moving parts.
Second, an approach toward internal transformation would avoid the inevitable political considerations typically involved in a major privatization initiative. Governor Daniels was facing opposition to the privatization of the Indiana Toll Road, and there was a chance that similar opposition from state elected officials, interest groups, or the general public might create obstacles to the transformation of the FSSA.
Third, the KPMG audit identified the FSSA’s severe lack of contract-monitoring capacity as a major risk factor. Despite holding more than 10,000 contracts with service providers, the FSSA made few efforts to ensure vendor accountability. A contract of the size, scope, and complexity being contemplated would expose the FSSA to even greater liabilities, from abuse to poor performance.
As Indiana’s decision-making process illustrates, the line between public and private is permeable. A state could use some combination of private and public elements, such as management, employees, and technology. However, the critical question is less about which sector to employ than about the availability of the requisite skills and technologies.
The manager reported back that the agency ‘had 107 county offices and 107 different ways of doing business. More likely we had 2,200 caseworkers and probably 2,200 different ways of doing business.’
Having weighed the FSSA’s available capacity and resources, Roob and Main decided to take the work outside. Indiana awarded a $1 billion-plus, seven-year contract—with an option to extend to ten years—to a group of private and nonprofit agents led by IBM and its subcontractor ACS. Indiana charged the IBM coalition not only with taking over the client-intake and welfare-to-work functions that state caseworkers had been performing, but also with finding solutions to major problems within the state’s social service system: inefficient and outdated business processes, out-of-control Medicaid expenses, poor customer service, and one of the nation’s worst records for implementing federal welfare reforms. Any one of these challenges would be a major test for an agency or contractor; taken together, they presented a monumental task.
Indiana imposed one critical condition on its hiring of outside help: the vendor would be required initially to hire all caseworkers not retained by the state. This stipulation proved to be an important difference between the approach in Indiana and the failed privatization in Texas, where new hires were undertrained and underinformed. As Texas works to unwind its privatization, the state is having to cope with a staffing shortage—many caseworkers had quit when first told about layoffs caused by the privatization. Roob understood that his commitment to state employees would reduce taxpayer savings, but the FSSA’s caseworkers knew the system and its rules—their knowledge was a valuable asset. Also, Roob now explains that “we intentionally bought ourselves an insurance policy in case IBM failed. Nor do I think we would have ever gotten federal approval or state legislative acquiescence if we had done otherwise.”
Requiring its vendor to hire the state’s caseworkers was not the only step that reduced the potential near-term and long-term savings of the privatization move; Indiana also resolved to maintain an open and staffed office facility in each county. Despite these and other burdens the state assumed, the project still promises to result in substantial savings: the ten-year price tag for both the IBM contract and the state’s retained costs will total roughly $1.6 billion—more than $300 million less than the projected ten-year cost of the current system and $500 million less than the projected ten-year cost of internally modernizing the system.
Indiana expects that the modernization will radically improve its ability to reduce mistaken benefit expenditures for ineligible applicants. The resulting savings—plus the fact that the vendor will be responsible for errors in eligibility determinations—mean that the true reduction in costs will be considerably higher. Governor Daniels estimates total savings will equal $1 billion over ten years.
While it is too early to determine whether Indiana’s daring plan will succeed in bringing about such substantial cost reductions, saving taxpayers’ dollars cannot be regarded as the sole criterion of success. It will be just as essential for the program to show that a true, effective transformation of government services has been accomplished—an achievement that will provide a blueprint for other public officials as they try to meet escalating demands on resources.
Three emerging developments explain why now, more than at any other time, we are on the verge of a true transformation of the public sector: digital platforms consolidate important data sources and make them more accessible; private-sector partners, with their experience as network integrators, are more adept than a patchwork of state agencies at meeting the multiple needs of individual clients; and decision-support systems employ algorithms to analyze large amounts of data, helping public employees identify problems and structure individualized solutions.
The state will have only one hand to shake if things go well, or, if the modernization goes sour, one neck to wring.
Aided by a large corporation such as IBM, Indiana will replace its antiquated paper-based system with online capabilities, document processing centers, call centers, and new information-system interfaces. The state will have only one hand to shake if things go well, or, if the modernization goes sour, one neck to wring. Rather than hiring dozens of private and nonprofit agencies to perform the work of over 2,000 caseworkers, support staff, and contractors, the state is contracting with IBM to serve as the integrator of the entire network. But true transformation for Hoosiers will only come when Indiana capitalizes on knowledge-based technologies to better organize and manage information across the system. Accurate program assessments, early and effective interventions, and accessible and transparent information will result in more families finding their way to healthier and more meaningful lives.
Success will be far from automatic with such an enormous undertaking. The modernization of eligibility in Indiana places the state ahead of its peers. But the project, regardless of its ultimate success, can’t help but lay down guidelines that will improve the chances of success in future public-private partnerships of such size and complexity. Officials in other states are closely watching how Indiana deals with a major hurdle that still remains to be cleared by the state: resistance from Congress and from national labor unions.
But a few tenets of the modernization drive seem almost certain to be included in any future effort by other states. Key values of the project include never letting the quality or quantity of service decrease during the transition; staying open to dynamic change by concentrating on public value, not just cost; and increasing the government’s capacity, with the aid of experienced private-sector experts, to design and manage sophisticated networks.
Indiana’s progress represents an important early step in what may well become a dramatic revision of the way we think about public-sector services. Public-private collaboration in the 21st century will not only provide solutions to complex public-sector problems, but it will also harness the power of data and information to unlock value hidden even deeper inside government programs.
Stephen Goldsmith, formerly mayor of Indianapolis, is now the Daniel Paul Professor of Government and Director of the Innovations in American Government Program at Harvard’s Kennedy School of Government. This essay was prepared with the help of Tim Burke, policy research coordinator at the Kennedy School of Government.
Image credit: illustration by Melinda Beck.
Image credit: illustration by Melinda Beck.