Innovation Files has moved! For ITIF's quick takes, quips, and commentary on the latest in tech policy, go to

Lessons from the ACA Health Insurance Marketplace Failure header

One can’t pass a single day it seems without seeing in the news coverage of the problems with the Affordable Care Act’s Health Insurance Marketplace (HIM). But what is perhaps most surprising is not that the web site had problems, but that people are surprised that it had problems. The current process of managing and acquiring federal IT is largely broken and the failure of the HIM is simply the newest reminder of that dysfunction. We can just go down the list of past high-profile failures, including the delayed launch last year of, the FBI’s Virtual Case Files program, the Census Bureau’s handheld PC debacle, and the FAA modernization.

There are several reasons for this dysfunction. First, the contracting process does not work as it should. Larded up with an accretion of rules and requirements from past scandals and failures, only the most intrepid firms are able to manage the labyrinth called federal contracting. Moreover, as Congress has tried to use federal contracting to fulfill social policy goals that should be addressed with other policy tools, agencies must give preferences to a wide variety of businesses—small businesses, women-owned businesses, minority-owned businesses, veteran-owned businesses, disabled-owned businesses, and the like. There are even firms that market their IT services to the federal government because they have managed to affiliate themselves with Native American tribes in Alaska who because of Congressional action, have a special exemption from some procurement rules, making it easier for government agencies to contract with them. Federal procurement decisions should be decided on the basis of the best combination of price and quality. Unfortunately, these set-asides are sacred because no elected official wants to be seen as against (fill in the blank for selected group).

In addition, agencies are largely not able to simply identify the best contractor and hire them. They must go through a complicated and byzantine bidding process designed to ensure fairness. But in reality some, if not many, bids are decided ahead of time and the bidding process is just a formality. In these cases, losing bidders waste valuable time and money submitting bids. In addition, given the complex system, government agencies cannot easily “think outside the box” to identify the best firm to do the work, and are often forced to choose from among those already in their existing contractor stable. This is the major reason why CMS chose the two firms that it issued the HIM prime contracts to.

The bidding process has also become so contentious that most major bids are challenged by the loser(s). This delays project implementation and raises costs. In fact, some agencies will pay outside organizations money to review their bidding process, simply so that when the contract award is made they can challenge more quickly.

Second, the process by which federal agencies manage contracting and IT systems development is broken. Federal agency executives are largely incapable of firing poor performers within their agencies. Moreover, they have a hard time hiring the best and the brightest, in part because of restrictive federal hiring practices and because the Feds can’t match the pay that applicants could get working for the private sector, including government IT contractors. In addition, failure is an option. In other words, while it is not uncommon in the private sector for individuals who lead projects that fail to lose their jobs or suffer other negative consequences (e.g. Apple fired the product manager responsible for its weak Maps offering and Microsoft fired the executive responsible for Windows 8) in the federal government this is the exception. Case in point, it is not likely that anyone in CMS will lose their job over the HIM problems. Likewise, individuals will not lose their jobs for not telling senior officials the whole truth about particular problems and challenges.

Now federal rules regarding hiring, acquisition and operations do have a purpose. They were designed to minimize fraud, increase competition, and ensure accountability of public funds. But perhaps there are more efficient ways to achieve some of the same outcomes. As my colleague Daniel Castro argued earlier this year, perhaps we can use technology to increase transparency and promote accountability by, for example, making information about all expenses (including salaries above a certain level) of federal agencies publicly available online.

The acquisition rules were also designed to be the same across all agencies to make it easier for businesses to work with the federal government and create more competition across the system. However, in practice, the rules have become a barrier to competition because they are so complicated. This same barrier does not exist in the private sector, where private companies can simply hire the contractor they want without going through a byzantine process. Good choices provide competitive advantage, and where there are bad choices, executives are held accountable.

And while it makes sense to retain some rules related to protecting individuals over efficiency, such as those to protect whistleblowers, many of the regulations seem unnecessary. As mentioned above, some rules put in place by congress are actually thinly-veiled attempts to set social policy. For example, FAR Subpart 23.5 outlines the requirements to create a drug-free awareness program that a government contractor must create for all its employees. Why should government have a different policy than the private sector? And notably, many of the rules are designed to benefit specific groups, such as small businesses, women, minorities or veterans, which increase costs. While there is a presumption that small businesses and diverse ownership offers the government with innovative solutions, these rules have instead created a cadre of small businesses whose only customer is the federal government.

These problems illustrate that given the huge costs involved policymakers should give more thought to creating a more flexible government bureaucracy that actually works.

So what’s the solution? Here’s one answer: the House and Senate Government Reform Committees should pass a law that would allow one federal agency to be completely exempt from all civil service and procurement rules, for a period of three years. The agency would be bound only by civil laws. At the end of the trial period, the Government Accountability Office would do a soup-to-nuts evaluation of how it worked. Ideally, Congress should pick a small agency, like the Federal Housing Finance Board or a federal lab. Or alternatively, agency heads could apply to be the test case; so that the flexibility of these rules is matched with an agency leader who wants the increased freedom. Given how onerous federal agency regulations are, there could be a long line of applicants for relief.

The idea would be to free the agency from all of the hiring practices, acquisition rules and other restrictions that make federal operations so inefficient. Under this system managers could more easily fire employees who deserve to be fired, hire the right employees faster, and manage operations in the most efficient and innovative ways.

Some experimentation in this area has already taken place. For example, in 1970, Congress passed the Postal Reorganization Act and exempted USPS from federal acquisition requirements to give it more flexibility in its procurement practices (although, as I’ve noted before, Congress has hamstrung the Postal Service in other ways). More recently, Congress initially exempted TSA from FAR requirements, in part so it could rapidly focus on its mission without getting bogged down by government contracting requirements. One of the major differences between FAR and the Acquisition Management System (AMS) used by TSA at the outset is that AMS allowed “managed competition.” As explained by Rick Gunderson, then Assistant Administrator for Acquisition at TSA, “Whereas the FAR requires full and open competition, AMS is based on managed competition. This is consistent with how industry conducts its own purchasing and supply chain management. As a result, government resources are not spent on firms that have no chance of receiving awards, and industry maximizes the impact of their bid and proposal costs.” Nevertheless, Congress took away this exemption in 2008, even as TSA agency leaders protested the change.

Here’s a second idea if the first one is too far a stretch. Congress should enable agencies and the federal government as a whole to better manage IT and IT acquisition. This is especially important as IT is now central to enabling agencies to achieve their mission. Congress took a step in that direction in 1996 when it passed the Clinger-Cohen Act that established Chief Information Officers in all federal agencies. But the CIO system has not lived up to its promise. There are a number of reasons for this. One is that in most agencies there are multiple CIOs for various divisions, with some 3 or even 4 levels down from the Secretary, giving these positions little real power or ability to shape strategic direction. For example, even though HHS has a CIO, many of the agencies within HHS, such as NIH, FDA, CMS, etc., each have their own CIOs. A related reason is that too many CIOs spend too much of their time managing IT operations (e.g. data centers, email, etc.). This responsibility should be tasked to the deputy CIO in each agency or even to the director of administration, leaving the CIO to focus on broader strategic planning and agency-wide IT oversight.

To address this problem, Congress should allow only one CIO per agency who reports directly to the agency head, and empower this new CIO with more authority for strategically managing IT resources, specifically by including the position in the budget planning process and giving it responsibility for funds allocated for IT resources by Congress. In addition, the CIO should have full authority to approve or deny the hiring of any federal workers who will be responsible for managing IT resources within their agency. These types of changes are included as part of H.R. 1232, the Federal Information Technology Acquisition Reform Act (FITARA) sponsored by Rep. Issa (R-CA). In addition, FITARA would make the CIO position a presidential appointee. So instead of just calling in the “best and brightest” to repair broken government systems, future presidents will be able to directly appoint seasoned executives to these positions. CIOs can then focus on formulating and implementing their agency’s digital strategy and managing the procurement process to advance government mission goals.

FITARA is a good step, but it could go further. To ensure that this is an improvement over the existing system, CIOs should be allowed to do “pay banding” (flexibility over pay, including performance based pay) and be exempted from personal rules regarding hiring and firing. They would be subject to the same laws against discrimination that government contractors must abide by, allowing government to compete on a level playing field for the best talent.

In addition, Congress should require government agencies to consider public-private partnership solutions, where instead of the government building and operating systems, they open up their back end interfaces and data and let certified private companies build and operate systems. Under this framework, governments can move beyond engaging with the private sector as IT vendors and instead empower third-party, for-profit and nonprofit organizations as partners in the provision of e-government services.

Government and the private sector have already engaged in successful partnerships in numerous areas. One of the most widely used is tax preparation and filing. A host of companies, including H&R Block, Intuit (maker of TurboTax software) and TaxAct, have used software to simplify the complicated task of filing taxes. Because these firms are competing intensely for market share, they have strong incentives to make their programs as user-friendly and comprehensive as possible.

It’s time to build on this model by empowering for-profit and nonprofit organizations to help citizens and businesses interact electronically with government, particularly in areas that are inherently complex or involve cross-agency and cross-government functions.

To do this, governments must think of themselves less as direct providers of e-government services and more as enablers of third-party integrators that tie together multiple agencies across multiple levels of government to package information, forms, regulations, and other government services and requirements in user-friendly ways.

Moving to this “turbo-government” model has the potential to dramatically boost the implementation of digital government services, cut costs for both government and users, and make the experience of dealing with government less frustrating. Intermediaries can play a key role in two kinds of tasks: building and operating function-based portals, and creating digital integration tools.

Going even further, Congress should learn from our friends in the UK who radically transformed their government IT procurement practices after wasting £12 billion over six years on an IT system for the National Health Service (perhaps the most costly government IT failure in the world). Following this debacle, the UK created Government Digital Services (GDS), which serves as a central IT operator, managing all substantive IT acquisitions for UK government agencies and serving as a technical support service for government operators. As one description of they agency notes, “Think agile development, user-centric design, human-readable copy on government websites. All the stuff that startups and software companies work with every day, but applied to the behemoth that is a government IT infrastructure.”

GDS works for several reasons. First, the head of GDS is a cabinet-level position, and can hire and fire who he wants, pay market wages, has real power over procurement decisions, and is exempt from many existing procurement rules. GDS has apparently worked so well that the president of CGI (yes, the same contractor involved in the dustup) has threatened to take its services elsewhere (nobody seems to mind).

In the United States, such an initiative would be like putting the Presidential Innovation Fellows program on steroids. Instead of having a few technical gurus floating around in government, we would have an entire office dedicated to this purpose. In addition, this office would be charged with making final decisions on IT procurement, not the procurement offices of the agencies which often are not fully qualified to judge complicated IT procurement decisions. Too often they make decisions solely on the basis of cost, not best value.

In closing, Congress can do two things in response the HIM failure. It can use it to score political points to attack the Affordable Care Act and the Obama Administration. Or it can utilize the opportunity to drive real reform in federal IT contracting and management. Let us hope that they don’t overlook the latter.

Print Friendly, PDF & Email