Sunday, September 25, 2011

Government IT investments

A couple of days ago, the New York Times published an article by David Halbfinger that highlights major problems with a City of New York information technology project. Shortly after Mayor Michael Bloomberg took office, the city opted to allocate $66 million to modernizing the system that manages information about the municipal workforce. At the time of this writing, the city has spent $363 million for a system that does far less than initially planned. What went wrong? According to Halbfinger, who relied heavily upon records requested in accordance with the New York State Freedom of Information Law, several things:
  • The Bloomberg administration was so intent on proving that outside consultants could streamline government that it failed to heed early signs that things were going wrong
  • The individuals given responsibility for administering the project did not have the power to make pivotal decisions and overlooked numerous cost-saving opportunities
  • The administration reacted to an early security failure by giving Accenture, the consulting firm responsible for developing the software, responsibility for defining precisely what the new system should do -- something that private corporations are loath to do for cost reasons
  • City agencies kept fighting over the system's features and functionality -- the city was slow to give a single person responsibility for setting policy -- and Accenture repeatedly made and undid changes as a result
  • Managers realized late in the game that they never wrote user instructions or provided training for staff responsible for using the new system and scrambled to cobble together documentation in the days prior to its rollout
At present, the new system is functional -- sort of. Although most City of New York employees can now access their personnel information online, retired employees and tens of thousands of Dept. of Education paraprofessonals and support staffers are currently not represented in the system. Moreover, at some point the city decided not to integrate management of its civil service system into its new personnel system. As a result, the civil service system that was developed in the 1980s is going to remain operational for the foreseeable future.

This is not the only government IT investment gone horribly wrong. The Bloomberg administration planned to spend $70 million developing CityTime, a modern municipal employee payroll system. To date, the city has spent $740 million, has a system that still isn't fully functional, and has suffered the humiliation of having the feds indict a bunch of CityTime contractors and consultants on charges of offering and accepting kickbacks. Several years ago, the Federal Bureau of Investigation junked its Virtual Case File system, which cost over $100 million, after it determined that the system was already outdated and simply couldn't replace its paper-based records management system. Similar horror stories abound at the federal, state, and local government level. Billions of taxpayer dollars that could have been put to better use have been squandered, and I don't even want to think about how difficult it's going to be to pull archival data out of some of these hideously expensive, deeply flawed systems.

Given the frequency with which government IT investments go bad, it's not surprising that some people -- including several of the readers who commented on Halbfinger's article -- conclude that governments are inherently bad at making IT purchases and that we simply shouldn't expect otherwise.

I don't think that this is the case, and I started writing this post with the intent of articulating a few strategies for avoiding the sorts of problems that the City of New York, the Federal Bureau of Investigation, and countless other government entities have encountered. However, upon second thought, I think I'm going to save these ideas for another post, simply because the problems that Halbfinger's article highlights are so fundamental that mere "strategies" can't solve them.

Personnel system project leaders and the Bloomberg administration itself knew that major problems existed and did nothing about them. The rollout of the first component of the personnel management system, which was completed in 2002, resulted in a massive security breach. Staff responsible for monitoring the system's development asserted in 2003 that "no sense of economy, efficiency or value is evident in any area of the project," but no one in a position of power paid any attention to their findings. Moreover, no government official has been fired or demoted as a result of the problems associated with this project.

In essence, what we have here is a woeful failure of leadership. No one in a position of power kept the scope and mission of the project from shifting and expanding, prevented city agencies from issuing multiple, competing change requests, or intervened as the contractor's bills skyrocketed. The City of New York did hire a private-sector project manager who was nominally responsible for keeping the project on track, but he quit after less than a year and has become an outspoken critic of the city's handling of the project. It also hired a human relations expert who was supposed to establish one city-wide personnel policy -- more than eight years after the personnel system project got underway. He returned to the private sector after less than six months in the city's employ.

At best, the Bloomberg administration failed to appreciate the differences between a multi-agency, public-sector operating environment and a single-entity, private-sector operating environment. At worst, it simply assumed that the public sector was inherently incapable of operating effectively and that the consultant would simply do an end run around the dysfunction -- real and perceived -- of the public sector. Instead of tackling the dysfunction head-on and ensuring that the personnel system project remained on track and under control, it sought to take the easy way out and gave its contractor free rein. City of New York taxpayers are now paying the price: the unanticipated $300 million that this project consumed would pay the salaries and benefits of more than a few teachers, nurses, firefighters, police officers, and sanitation workers.

No comments: